Meteoritics amp Planetary Science 39 Nr 3 425ndash451 (2004)Abstract available online at httpmeteoriticsorg

425 copy Meteoritical Society 2004 Printed in USA

Geology petrography shock petrography and geochemistry of impactites and target rocks from the Kaumlrdla crater Estonia

V PUURA1 H HUBER2dagger J KIRS1 A KAumlRKI3 K SUUROJA4 K KIRSIMAumlE1 J KIVISILLA4

A KLEESMENT5 M KONSA5 U PREEDEN1 S SUUROJA5 and C KOEBERL2

1Institute of Geology University of Tartu Vanemuise strasse 46 51014 Tartu Estonia2Department of Geological Sciences University of Vienna Althanstrasse 14 A-1090 Vienna Austria

3Institute of Geosciences University of Oulu Box 3000 FIN-90401 Oulu Finland4Geological Survey of Estonia Kadaka tee 8082 12618 Tallinn Estonia

5Institute of Geology Tallinn Technical University Estonia pst 7 10143 Tallinn EstoniadaggerPresent address Institute of Geophysics and Planetary Physics University of California Los Angeles

595 Charles Young Drive Los Angeles California 90095 USACorresponding author E-mail puurautee

(Received 20 December 2002 revision accepted 7 January 2004)

AbstractndashThe Kaumlrdla crater is a 4 km-wide impact structure of Late Ordovician age located onHiiumaa Island Estonia The 455 Ma-old buried crater was formed in shallow seawater inPrecambrian crystalline target rocks that were covered with sedimentary rocks Basement and brecciasamples from 13 drill cores were studied mineralogically petrographically and geochemicallyGeochemical analyses of major and trace elements were performed on 90 samples fromallochthonous breccias sub-crater and surrounding basement rocks The breccia units do not includeany melt rocks or suevites The remarkably poorly mixed sedimentary and crystalline rocks weredeposited separately within the allochthonous breccia suites of the crater The most intensely shock-metamorphosed allochthonous granitoid crystalline-derived breccia layers contain planardeformation features (PDFs) in quartz indicating shock pressures of 20ndash35 GPa An apparent K-enrichment and Ca-Na-depletion of feldspar- and hornblende-bearing rocks in the allochthonousbreccia units and sub-crater basement is interpreted to be the result of early stage alteration in animpact-induced hydrothermal system The chemical composition of the breccias shows no definitesign of an extraterrestrial contamination By modeling of the different breccia units with HMX-mixing the indigenous component was determined From the abundances of the siderophile elements(Cr Co Ni Ir and Au) in the breccia samples no unambiguous evidence for the incorporation of ameteoritic component above about 01 wt chondrite-equivalent was found

INTRODUCTION

In the Baltic Sea region 15 meteorite impact structures ofPaleozoic age have been identified to date (Abels et al 2002)During the Paleozoic era the Baltic area developed as a largeand shallow epi-continental sea or flat sedimentary lowlandAt the end of the Paleozoic era a thin (mostly lt500 mmaximum ~3-km-thick) sequence of unconsolidatedsediments covered the Precambrian basement From the latePaleozoic era and during the Meso-Cenozoic era the presentFennoscandian Shield was re-exposed due to uplift anderosion Consequently in the Paleozoic era asteroids orcomets hit a two-layered target either in marine or continentalenvironments depending on high or low sea levels Among

these 15 impact craters (see Fig 1) six structures of Middle toLate Ordovician (450ndash470 Ma) age form a prominent groupie Granby Hummeln Karikkoselkauml Lockne Tvaumlren andKaumlrdla The majority of them formed in a marine environment(Puura et al 1994 Abels et al 2002) Kaumlrdla is the best-preserved crater among the Paleozoic marine impactstructures in this region The Kaumlrdla crater formed during theearly Late Ordovician (see IUGS 2000) and has been dated at455 Ma (Puura et al 1989 Puura and Suuroja 1992) based onthe stratigraphic position of the Kaumlrdla distal ejecta in thesequence of the Caradoc Series

Kaumlrdla is also a good target for the study of unmeltedsuites of impactites produced in marine environments andcombined sedimentary-crystalline targets Impactites from

426 V Puura et al

the Kaumlrdla crater both allochthonous crater-fill andautochthonous sub-crater rocks are of silicate-dominatedcomposition

During impact-induced processes these rocks wereenriched in potassium and depleted in sodium and calcium(Puura and Suuroja 1992 Puura et al 2000a b) Thedominant type of alteration is metasomatic replacement of therock-forming feldspars hornblende and biotite withsecondary minerals The occurrence and mechanisms oflarge-scale K-metasomatism in impact structures remainpoorly studied At many craters fissures vugs and pores inimpactites are filled with Ca- and Na-rich minerals asdescribed for the Manson crater in the USA (McCarville andCrossey 1996) or the Popigai Kara and Puchezh-Katunkicraters in Russia (Naumov 1999 2002 Masaitis et al 2004)These secondary minerals include zeolites clay minerals andcarbonates However at Kaumlrdla veins and vugs filled with theCa-Mg-rich minerals calcite dolomite and chlorite are only aconsiderably small admixture in K-enriched impactites

Drilling of the Kaumlrdla crater was performed by theGeological Survey of Estonia at 30 different sites withmaximum depths of 800 m but most cores penetrated to anaverage depth of 300 m (Suuroja 1996 Puura and Suuroja1992) The studied samples are from 13 of these drill cores(for locations see Figs 2 and 3) that penetrated between 797and 8152 m into the crater-fill deposits and sub-craterbasement in the crater moat central uplift crater rim and

ejecta layer outside the crater proper Core K1 was drilled in1989ndash1990 in impactites with diameters of 60 and 40 mmdown to 8152 m and core K18 was drilled in 1984 with adiameter of 60 mm to a depth of 4314 m Both cores arestored at the Arbavere field station of the Geological Surveyof Estonia Short descriptions of these two cores includingcore logs were given by Suuroja (1996) and in unpublishedreports of the Geological Survey of Estonia A detaileddescription of core K1 is given by Suuroja et al (2002b)Puura and Suuroja (1992) published preliminary data onchemical and mineralogical changes of Precambrianbasement silicate target rocks

The aim of the present paper is to present results ofsystematic studies of impact lithologies and target rocks of theKaumlrdla crater The research focused on the geochemical-mineralogical composition and metasomatic alteration on thespatial distribution of shocked minerals and on the mixing oftarget rocks and the admixture of a meteoritic component

GEOLOGICAL SETTING OF THE CRATER AND IMPACT LITHOLOGIES

The Multilayer Target

The buried crater structure is located in the northwesternmarginal part of the Russian platform within the exposure ofthe Upper Ordovician and Lower Silurian The crater site is

Fig 1 Geographic map of the 15 known Paleozoic impact structures in the Baltic Sea region Among others the Kaumlrdla crater in Estoniabelongs to the Middle to Late Ordovician age group the most common in this region The location of the Kaumlrdla crater (lat 58deg59acuteN long22deg46acuteE) on Hiiumaa Island is shown in the close-up The rectangle in the insert indicates the part shown in the geological map (Fig 2)

Impactites and target rocks from the Kaumlrdla crater 427

Fig 2 Subsurface geological map of the Kaumlrdla crater Post-impact formations are removed and contour lines show the subsurface topographyof target rocks and impact breccias that underlie the post-impact cover-rocks The Paleozoic rocks in the rim area are mainly Cambriansandstones and claystones Cross sections AB and CD are shown in Fig 3

Fig 3 Generalized cross sections of the Kaumlrdla crater Top SW-NE section across the uplifted crystalline basement in the rim (drill cores atPalukuumlla and Tubala) Bottom NW-SE section through the gullies in the rim The locations of the cross sections are shown in Fig 2 (AB CD)Post-impact cover sediments are shown in gray Logs of the allochthonous breccias of boreholes K1 and K18 are given in Figs 4andash4c For thesubsurface geology below the cover see Fig 2

428 V Puura et al

on Hiiumaa Island Estonia in a flat low near-shore areaOnly low hills with up to 23 m elevation mark the mostelevated buried fragments of the crystalline rim (eg thePalukuumlla and Tubala hills see Fig 2) A flat wetlandcorresponds to the crater proper

At present the composite lithologies of the Kaumlrdla areaconsist of 170 m-thick uppermost Middle Ordovician toCambrian and uppermost Vendian (Neoproterozoic III afterIUGS 2000) sedimentary rocks underlain by Paleoproterozoic(Svecofennian 19ndash18 Ga) crystalline basement (see Fig 3)The sea at the impact site was a shallow-water near-shorezone of the Paleo-Baltic basin which opened to the IapetusOcean far in the southwest The lower Paleozoic sedimentarycover and Paleo-Proterozoic crystalline basement were bothaffected by the impact At the time of the impact event thetarget consisted of three principal layers (from the top)

1 Seawater less than 100 m yet probably only 20ndash50 mdeep (Puura and Suuroja 1992)

2 Sediments estimated thickness ~200 m3 The crystalline basement

The sedimentary cover consisted of (from the top)1 Limestones and clayish limestones with an admixture of

goethite-rich oolites and glauconite grains in itslowermost layers 25 m

2 Organic-rich (metalliferous) mud (presently blackshale) weakly consolidated ~2 m

3 Quartz silts with clay inter-layers and sands weaklyconsolidated ~170 mThe crystalline basement consisted of the upper (5ndash25 m)

weathered (usually kaolinite-rich) and lower fresh parts witha gradual transition between the two The basement iscomposed of igneous mainly migmatitic granite andmetamorphic (amphibolites and granitoid gneisses) rocks(Koppelmaa et al 1996) Microcline-plagioclase granites formveins and more massive bodies Metamorphic rocks in thesurroundings of the Kaumlrdla crater contain migmatized biotiteor biotite-amphibole gneisses biotite-amphibole-plagioclaseamphibolites and quartzites For the characterization ofimpact-affected rocks and mixing calculations the chemicalcomposition of pre-impact rocks were used from Koppelmaaet al (1996) and Kivisilla et al (1999)

The Impact Structure and Its Post-Impact History

The sub-surface map of the crater (without post-impactdeposits see Fig 2) and the cross sections (Fig 3) show thestructure of the crater To restore the possible morphology ofthe transient cavity at Kaumlrdla we can compare it to the Brentcrater Canada (B A Ivanov 2000 personal communication)For the present study we calculated approximate volumes inkm3 as follows (Figs 2ndash4) transient cavity (8 km3) presentcrater below the target surface (23) uplifted (displaced) sub-crater basement (35) excavated rocks (35) (includingsedimentary cover [23] and crystalline basement [12]) and

allochthonous breccias within the crater (05) (includingsediment-dominated [03] and crystalline [02]) For thecalculations the vertical parabolic cross section of the craterwas divided into truncated cones considering the thickness oftarget layers in the volume of the transient cavity and impact-stratigraphic units in the present crater The morphology ofbreccia units and the proportions of sedimentary to crystallinerocks in the impact breccia units were estimated from logs ofcores K1 K18 and K12 The results of the calculations wererounded to the nearest 01 km3

The crater rim wall consists of four horseshoe-shapedelevated parts including Palukuumlla High in the east-northeastand Tubala in the west-southwest The crest of the wallsegments was eroded deeply enough to expose the crystallinecore before the final burial under post-impact sediments Inthe gullies (compare profiles in Fig 2) the post-impactbreccia layer overlies deformed Cambrian sedimentary rocksThe amount of stratigraphic uplift of the basement in the rimranged between 100 and 250 m

In Fig 2 the semi-annular area of the eroded breccialayer outlines the perimeter of the structure The top ofimpact-deformed target rocks and ejecta is truncated by adiscontinuity which formed due to post-impact (MiddleOrdovician) local marine shoreline erosion Within the crateras a result of post-sedimentary compaction the top of thebreccias is at depths of 295 to 100 m below present-day sealevel (bsl) which is probably deeper than during theaccumulation Within the semi-annular zone of the missingallochthonous breccia this discontinuity surface lies atvariable depths between 150 and 30 m relating to the localtopography Within the rim the surface of the crystallinebasement ranges from depths of 30 m to heights of 6 m abovethe sea level Further outside the rim the surface of the distalejecta is 50 to 100 m bsl In the further surroundings of thecrater the top of the pre-impact target is at depths of 60 m inthe north to 100 m in the south due to the post-impact obliqueregional sinking of the Baltic sedimentary basin The ejectalayer that is preserved in the surroundings of the crater isthickest up to 25 m at distances of 3ndash6 km from the cratercenter Erosion of the proximal part of the ejecta layer tookplace soon after the impact event because during 5ndash10 Mathe crater rim remained as an atoll-like feature where it wassubjected to erosion in the shallow sea The crater depressionrepresented a sedimentary trap with a higher accumulationrate especially during the early post-impact times

The crater was gradually buried under sediments over 5ndash10 Ma During the Silurian and Devonian eras the structurewas probably buried under a 500 m-thick blanket ofsediments The Oslo rifting and tectonic-thermal event inwestern part of the East European Craton (including theBaltic Sea region) happened in the Late Carboniferous-EarlyPermian era eg at about 300 Ma Regional dolomitizationand base metallic sulfide mineralization in Estonia is amanifestation of this activity (Puura et al 1996) The post-

Impactites and target rocks from the Kaumlrdla crater 429

Fig 4a Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 1) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

430 V Puura et al

Fig 4b Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 2) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

Impactites and target rocks from the Kaumlrdla crater 431

Fig 4c Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K18 Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second) Thelithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

432 V Puura et al

impact lead and zinc sulphide and related carbonatemineralization in the Kaumlrdla rim wall and coveringsedimentary rocks belong to this event and consequently arenot related to the impact-event During the Mesozoic andCenozoic eras it was gradually unroofed The Cenozoicerosion (see Puura and Flodeacuten 1997) essentially uncoveredthe Kaumlrdla impact site

Settings of Studied Rock Units

Mixing of the debris produced from the almost 200 m-thick sedimentary layers and 300-m-thick portion of thecrystalline basement appears to be incomplete Theapproximately 50-m-thick allochthonous breccia unit thatoverlies the sub-crater basement is derived mainly fromcrystalline granitoids and amphibolites of the basement (coreK1 471ndash523 m see Fig 4b) This layer contains the moststrongly shocked mineral fragments (shock pressures of 20ndash35 GPa) The other 30-m-thick layer of the same compositionand with similarly shocked mineral clasts was penetrated atdepths between 356 and 380 m being separated from the lowerlayer by the mainly sedimentary rock-derived allochthonousslumped breccia layer (380ndash471 m) without any admixture ofshock metamorphic minerals The second crystalline-derivedbreccia layer is covered with the uppermost dominantlysedimentary-derived slumped resurge and air-fall breccias(300ndash356 m) with occasional admixture of clasts of crystallinerocks and crystalline-derived breccias We assume that theupper layer of crystalline-derived breccias with shockedminerals (356ndash380 m) is derived due to slumping of theuppermost parts of the rim wall where the lower breccia unitis presently lacking into the annular moat of the crater Theimpact breccias of these two layers show the most intenseevidence for the potassium enrichment and the sodium andcalcium depletion Inter-layers and patches of this breccia withsignatures of K-metasomatism are met also within thesedimentary-derived layers of the slump breccia

The sediment-derived breccia layers consist mainly oflarge blocks cobbles and pebbles of very soft rocks subjectedto low shock pressure The material probably slid back intothe crater via gullies that existed in the crater rimObservations on drill cores indicate that the sediment-derivedallochthonous breccias lack any other major secondarymineralization

The sub-crater basement has also been subjected toimpact-induced fracturing and brecciation at low shockpressures reaching ~2 GPa in the par-autochthonous brecciaunit in drill core K1 The K-enrichment and Na-Ca-depletionis gradually decreasing downward However the total volumeof K-enriched rocks is much larger than the volume of rockswith obvious shock-metamorphic signatures

In the geochemical study we concentrate on 1) impact-metamorphosed rocks of the allochthonous crystalline-derived breccia units 2) rocks of the sub-crater basement as

the main source for the breccias and 3) the possibleadmixture of an extraterrestrial component

SAMPLES AND ANALYTICAL TECHNIQUES

Optical Microscopy

Optical microscopy was performed on 175 thin sectionsat the Institute of Geology University of Tartu and theInstitute of Earth Sciences Oulu University (see theAppendix for detailed descriptions) The orientations ofplanar deformation features (PDFs) in quartz grains weremeasured in 32 thin sections from drill cores K1 and K18using standard universal stage (U-stage) techniques (seeFrench 1998) Electron microscopic studies were performedusing a JEOL 6300 scanning electron microscope (SEM)equipped with an Oxford ISIS energy dispersive spectrometerand a JEOL JSM-6300F field emission SEM at the Institute ofElectron Optics University of Oulu Finland

Electron Microprobe Analyses

Electron microprobe analyses (EPMA) of 20 thinsections were done with an automatized JEOL Superprobe733 using 15 kV voltage a 15 nA sample current and a 5 or10 m electron beam diameter at the Institute of ElectronOptics Oulu University Finland The quantitative chemicalcompositions of silicate minerals were calculated using theZAF correction program

X-Ray Powder Diffractometry

The mineralogical composition of bulk-rock samples andK-feldspar fractions was determined using a Dron-3Mdiffractometer with Ni filtered CuK -radiation at the Instituteof Geology University of Tartu Estonia Powdered rocksamples were scanned at step widths of 002deg 2 in the rangeof 2 and 70deg 2 with time steps of 5ndash10 sec Quantitativemineral composition of bulk-rock samples was found by full-profile Rietveld fitting using the Siroquant 25trade (Taylor1991) code The XRD patterns of K-feldspar fractions weredeconvoluted in the 20deg to 40deg 2 region into elementarypeaks using the Axes-19a code (Maumlndar et al 1996) Themeasured diffraction peaks were fitted with symmetricalpseudo-Voigt function shape curve doublets Relativeabundances of the K-feldspar structural varieties(crosshatched twin structure microcline and orthoclase) wereestimated from integral intensities of triclinic microcline(131) and (131) and monoclinic orthoclase (131) peaks

Geochemical Analyses

Geochemical analyses were performed on 90 samplesfrom both drill cores K1 and K18 and target rock samples

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 427

Fig 2 Subsurface geological map of the Kaumlrdla crater Post-impact formations are removed and contour lines show the subsurface topographyof target rocks and impact breccias that underlie the post-impact cover-rocks The Paleozoic rocks in the rim area are mainly Cambriansandstones and claystones Cross sections AB and CD are shown in Fig 3

Fig 3 Generalized cross sections of the Kaumlrdla crater Top SW-NE section across the uplifted crystalline basement in the rim (drill cores atPalukuumlla and Tubala) Bottom NW-SE section through the gullies in the rim The locations of the cross sections are shown in Fig 2 (AB CD)Post-impact cover sediments are shown in gray Logs of the allochthonous breccias of boreholes K1 and K18 are given in Figs 4andash4c For thesubsurface geology below the cover see Fig 2

428 V Puura et al

on Hiiumaa Island Estonia in a flat low near-shore areaOnly low hills with up to 23 m elevation mark the mostelevated buried fragments of the crystalline rim (eg thePalukuumlla and Tubala hills see Fig 2) A flat wetlandcorresponds to the crater proper

At present the composite lithologies of the Kaumlrdla areaconsist of 170 m-thick uppermost Middle Ordovician toCambrian and uppermost Vendian (Neoproterozoic III afterIUGS 2000) sedimentary rocks underlain by Paleoproterozoic(Svecofennian 19ndash18 Ga) crystalline basement (see Fig 3)The sea at the impact site was a shallow-water near-shorezone of the Paleo-Baltic basin which opened to the IapetusOcean far in the southwest The lower Paleozoic sedimentarycover and Paleo-Proterozoic crystalline basement were bothaffected by the impact At the time of the impact event thetarget consisted of three principal layers (from the top)

1 Seawater less than 100 m yet probably only 20ndash50 mdeep (Puura and Suuroja 1992)

2 Sediments estimated thickness ~200 m3 The crystalline basement

The sedimentary cover consisted of (from the top)1 Limestones and clayish limestones with an admixture of

goethite-rich oolites and glauconite grains in itslowermost layers 25 m

2 Organic-rich (metalliferous) mud (presently blackshale) weakly consolidated ~2 m

3 Quartz silts with clay inter-layers and sands weaklyconsolidated ~170 mThe crystalline basement consisted of the upper (5ndash25 m)

weathered (usually kaolinite-rich) and lower fresh parts witha gradual transition between the two The basement iscomposed of igneous mainly migmatitic granite andmetamorphic (amphibolites and granitoid gneisses) rocks(Koppelmaa et al 1996) Microcline-plagioclase granites formveins and more massive bodies Metamorphic rocks in thesurroundings of the Kaumlrdla crater contain migmatized biotiteor biotite-amphibole gneisses biotite-amphibole-plagioclaseamphibolites and quartzites For the characterization ofimpact-affected rocks and mixing calculations the chemicalcomposition of pre-impact rocks were used from Koppelmaaet al (1996) and Kivisilla et al (1999)

The Impact Structure and Its Post-Impact History

The sub-surface map of the crater (without post-impactdeposits see Fig 2) and the cross sections (Fig 3) show thestructure of the crater To restore the possible morphology ofthe transient cavity at Kaumlrdla we can compare it to the Brentcrater Canada (B A Ivanov 2000 personal communication)For the present study we calculated approximate volumes inkm3 as follows (Figs 2ndash4) transient cavity (8 km3) presentcrater below the target surface (23) uplifted (displaced) sub-crater basement (35) excavated rocks (35) (includingsedimentary cover [23] and crystalline basement [12]) and

allochthonous breccias within the crater (05) (includingsediment-dominated [03] and crystalline [02]) For thecalculations the vertical parabolic cross section of the craterwas divided into truncated cones considering the thickness oftarget layers in the volume of the transient cavity and impact-stratigraphic units in the present crater The morphology ofbreccia units and the proportions of sedimentary to crystallinerocks in the impact breccia units were estimated from logs ofcores K1 K18 and K12 The results of the calculations wererounded to the nearest 01 km3

The crater rim wall consists of four horseshoe-shapedelevated parts including Palukuumlla High in the east-northeastand Tubala in the west-southwest The crest of the wallsegments was eroded deeply enough to expose the crystallinecore before the final burial under post-impact sediments Inthe gullies (compare profiles in Fig 2) the post-impactbreccia layer overlies deformed Cambrian sedimentary rocksThe amount of stratigraphic uplift of the basement in the rimranged between 100 and 250 m

In Fig 2 the semi-annular area of the eroded breccialayer outlines the perimeter of the structure The top ofimpact-deformed target rocks and ejecta is truncated by adiscontinuity which formed due to post-impact (MiddleOrdovician) local marine shoreline erosion Within the crateras a result of post-sedimentary compaction the top of thebreccias is at depths of 295 to 100 m below present-day sealevel (bsl) which is probably deeper than during theaccumulation Within the semi-annular zone of the missingallochthonous breccia this discontinuity surface lies atvariable depths between 150 and 30 m relating to the localtopography Within the rim the surface of the crystallinebasement ranges from depths of 30 m to heights of 6 m abovethe sea level Further outside the rim the surface of the distalejecta is 50 to 100 m bsl In the further surroundings of thecrater the top of the pre-impact target is at depths of 60 m inthe north to 100 m in the south due to the post-impact obliqueregional sinking of the Baltic sedimentary basin The ejectalayer that is preserved in the surroundings of the crater isthickest up to 25 m at distances of 3ndash6 km from the cratercenter Erosion of the proximal part of the ejecta layer tookplace soon after the impact event because during 5ndash10 Mathe crater rim remained as an atoll-like feature where it wassubjected to erosion in the shallow sea The crater depressionrepresented a sedimentary trap with a higher accumulationrate especially during the early post-impact times

The crater was gradually buried under sediments over 5ndash10 Ma During the Silurian and Devonian eras the structurewas probably buried under a 500 m-thick blanket ofsediments The Oslo rifting and tectonic-thermal event inwestern part of the East European Craton (including theBaltic Sea region) happened in the Late Carboniferous-EarlyPermian era eg at about 300 Ma Regional dolomitizationand base metallic sulfide mineralization in Estonia is amanifestation of this activity (Puura et al 1996) The post-

Impactites and target rocks from the Kaumlrdla crater 429

Fig 4a Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 1) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

430 V Puura et al

Fig 4b Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 2) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

Impactites and target rocks from the Kaumlrdla crater 431

Fig 4c Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K18 Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second) Thelithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

432 V Puura et al

impact lead and zinc sulphide and related carbonatemineralization in the Kaumlrdla rim wall and coveringsedimentary rocks belong to this event and consequently arenot related to the impact-event During the Mesozoic andCenozoic eras it was gradually unroofed The Cenozoicerosion (see Puura and Flodeacuten 1997) essentially uncoveredthe Kaumlrdla impact site

Settings of Studied Rock Units

Mixing of the debris produced from the almost 200 m-thick sedimentary layers and 300-m-thick portion of thecrystalline basement appears to be incomplete Theapproximately 50-m-thick allochthonous breccia unit thatoverlies the sub-crater basement is derived mainly fromcrystalline granitoids and amphibolites of the basement (coreK1 471ndash523 m see Fig 4b) This layer contains the moststrongly shocked mineral fragments (shock pressures of 20ndash35 GPa) The other 30-m-thick layer of the same compositionand with similarly shocked mineral clasts was penetrated atdepths between 356 and 380 m being separated from the lowerlayer by the mainly sedimentary rock-derived allochthonousslumped breccia layer (380ndash471 m) without any admixture ofshock metamorphic minerals The second crystalline-derivedbreccia layer is covered with the uppermost dominantlysedimentary-derived slumped resurge and air-fall breccias(300ndash356 m) with occasional admixture of clasts of crystallinerocks and crystalline-derived breccias We assume that theupper layer of crystalline-derived breccias with shockedminerals (356ndash380 m) is derived due to slumping of theuppermost parts of the rim wall where the lower breccia unitis presently lacking into the annular moat of the crater Theimpact breccias of these two layers show the most intenseevidence for the potassium enrichment and the sodium andcalcium depletion Inter-layers and patches of this breccia withsignatures of K-metasomatism are met also within thesedimentary-derived layers of the slump breccia

The sediment-derived breccia layers consist mainly oflarge blocks cobbles and pebbles of very soft rocks subjectedto low shock pressure The material probably slid back intothe crater via gullies that existed in the crater rimObservations on drill cores indicate that the sediment-derivedallochthonous breccias lack any other major secondarymineralization

The sub-crater basement has also been subjected toimpact-induced fracturing and brecciation at low shockpressures reaching ~2 GPa in the par-autochthonous brecciaunit in drill core K1 The K-enrichment and Na-Ca-depletionis gradually decreasing downward However the total volumeof K-enriched rocks is much larger than the volume of rockswith obvious shock-metamorphic signatures

In the geochemical study we concentrate on 1) impact-metamorphosed rocks of the allochthonous crystalline-derived breccia units 2) rocks of the sub-crater basement as

the main source for the breccias and 3) the possibleadmixture of an extraterrestrial component

SAMPLES AND ANALYTICAL TECHNIQUES

Optical Microscopy

Optical microscopy was performed on 175 thin sectionsat the Institute of Geology University of Tartu and theInstitute of Earth Sciences Oulu University (see theAppendix for detailed descriptions) The orientations ofplanar deformation features (PDFs) in quartz grains weremeasured in 32 thin sections from drill cores K1 and K18using standard universal stage (U-stage) techniques (seeFrench 1998) Electron microscopic studies were performedusing a JEOL 6300 scanning electron microscope (SEM)equipped with an Oxford ISIS energy dispersive spectrometerand a JEOL JSM-6300F field emission SEM at the Institute ofElectron Optics University of Oulu Finland

Electron Microprobe Analyses

Electron microprobe analyses (EPMA) of 20 thinsections were done with an automatized JEOL Superprobe733 using 15 kV voltage a 15 nA sample current and a 5 or10 m electron beam diameter at the Institute of ElectronOptics Oulu University Finland The quantitative chemicalcompositions of silicate minerals were calculated using theZAF correction program

X-Ray Powder Diffractometry

The mineralogical composition of bulk-rock samples andK-feldspar fractions was determined using a Dron-3Mdiffractometer with Ni filtered CuK -radiation at the Instituteof Geology University of Tartu Estonia Powdered rocksamples were scanned at step widths of 002deg 2 in the rangeof 2 and 70deg 2 with time steps of 5ndash10 sec Quantitativemineral composition of bulk-rock samples was found by full-profile Rietveld fitting using the Siroquant 25trade (Taylor1991) code The XRD patterns of K-feldspar fractions weredeconvoluted in the 20deg to 40deg 2 region into elementarypeaks using the Axes-19a code (Maumlndar et al 1996) Themeasured diffraction peaks were fitted with symmetricalpseudo-Voigt function shape curve doublets Relativeabundances of the K-feldspar structural varieties(crosshatched twin structure microcline and orthoclase) wereestimated from integral intensities of triclinic microcline(131) and (131) and monoclinic orthoclase (131) peaks

Geochemical Analyses

Geochemical analyses were performed on 90 samplesfrom both drill cores K1 and K18 and target rock samples

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

428 V Puura et al

on Hiiumaa Island Estonia in a flat low near-shore areaOnly low hills with up to 23 m elevation mark the mostelevated buried fragments of the crystalline rim (eg thePalukuumlla and Tubala hills see Fig 2) A flat wetlandcorresponds to the crater proper

At present the composite lithologies of the Kaumlrdla areaconsist of 170 m-thick uppermost Middle Ordovician toCambrian and uppermost Vendian (Neoproterozoic III afterIUGS 2000) sedimentary rocks underlain by Paleoproterozoic(Svecofennian 19ndash18 Ga) crystalline basement (see Fig 3)The sea at the impact site was a shallow-water near-shorezone of the Paleo-Baltic basin which opened to the IapetusOcean far in the southwest The lower Paleozoic sedimentarycover and Paleo-Proterozoic crystalline basement were bothaffected by the impact At the time of the impact event thetarget consisted of three principal layers (from the top)

1 Seawater less than 100 m yet probably only 20ndash50 mdeep (Puura and Suuroja 1992)

2 Sediments estimated thickness ~200 m3 The crystalline basement

The sedimentary cover consisted of (from the top)1 Limestones and clayish limestones with an admixture of

goethite-rich oolites and glauconite grains in itslowermost layers 25 m

2 Organic-rich (metalliferous) mud (presently blackshale) weakly consolidated ~2 m

3 Quartz silts with clay inter-layers and sands weaklyconsolidated ~170 mThe crystalline basement consisted of the upper (5ndash25 m)

weathered (usually kaolinite-rich) and lower fresh parts witha gradual transition between the two The basement iscomposed of igneous mainly migmatitic granite andmetamorphic (amphibolites and granitoid gneisses) rocks(Koppelmaa et al 1996) Microcline-plagioclase granites formveins and more massive bodies Metamorphic rocks in thesurroundings of the Kaumlrdla crater contain migmatized biotiteor biotite-amphibole gneisses biotite-amphibole-plagioclaseamphibolites and quartzites For the characterization ofimpact-affected rocks and mixing calculations the chemicalcomposition of pre-impact rocks were used from Koppelmaaet al (1996) and Kivisilla et al (1999)

The Impact Structure and Its Post-Impact History

The sub-surface map of the crater (without post-impactdeposits see Fig 2) and the cross sections (Fig 3) show thestructure of the crater To restore the possible morphology ofthe transient cavity at Kaumlrdla we can compare it to the Brentcrater Canada (B A Ivanov 2000 personal communication)For the present study we calculated approximate volumes inkm3 as follows (Figs 2ndash4) transient cavity (8 km3) presentcrater below the target surface (23) uplifted (displaced) sub-crater basement (35) excavated rocks (35) (includingsedimentary cover [23] and crystalline basement [12]) and

allochthonous breccias within the crater (05) (includingsediment-dominated [03] and crystalline [02]) For thecalculations the vertical parabolic cross section of the craterwas divided into truncated cones considering the thickness oftarget layers in the volume of the transient cavity and impact-stratigraphic units in the present crater The morphology ofbreccia units and the proportions of sedimentary to crystallinerocks in the impact breccia units were estimated from logs ofcores K1 K18 and K12 The results of the calculations wererounded to the nearest 01 km3

The crater rim wall consists of four horseshoe-shapedelevated parts including Palukuumlla High in the east-northeastand Tubala in the west-southwest The crest of the wallsegments was eroded deeply enough to expose the crystallinecore before the final burial under post-impact sediments Inthe gullies (compare profiles in Fig 2) the post-impactbreccia layer overlies deformed Cambrian sedimentary rocksThe amount of stratigraphic uplift of the basement in the rimranged between 100 and 250 m

In Fig 2 the semi-annular area of the eroded breccialayer outlines the perimeter of the structure The top ofimpact-deformed target rocks and ejecta is truncated by adiscontinuity which formed due to post-impact (MiddleOrdovician) local marine shoreline erosion Within the crateras a result of post-sedimentary compaction the top of thebreccias is at depths of 295 to 100 m below present-day sealevel (bsl) which is probably deeper than during theaccumulation Within the semi-annular zone of the missingallochthonous breccia this discontinuity surface lies atvariable depths between 150 and 30 m relating to the localtopography Within the rim the surface of the crystallinebasement ranges from depths of 30 m to heights of 6 m abovethe sea level Further outside the rim the surface of the distalejecta is 50 to 100 m bsl In the further surroundings of thecrater the top of the pre-impact target is at depths of 60 m inthe north to 100 m in the south due to the post-impact obliqueregional sinking of the Baltic sedimentary basin The ejectalayer that is preserved in the surroundings of the crater isthickest up to 25 m at distances of 3ndash6 km from the cratercenter Erosion of the proximal part of the ejecta layer tookplace soon after the impact event because during 5ndash10 Mathe crater rim remained as an atoll-like feature where it wassubjected to erosion in the shallow sea The crater depressionrepresented a sedimentary trap with a higher accumulationrate especially during the early post-impact times

The crater was gradually buried under sediments over 5ndash10 Ma During the Silurian and Devonian eras the structurewas probably buried under a 500 m-thick blanket ofsediments The Oslo rifting and tectonic-thermal event inwestern part of the East European Craton (including theBaltic Sea region) happened in the Late Carboniferous-EarlyPermian era eg at about 300 Ma Regional dolomitizationand base metallic sulfide mineralization in Estonia is amanifestation of this activity (Puura et al 1996) The post-

Impactites and target rocks from the Kaumlrdla crater 429

Fig 4a Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 1) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

430 V Puura et al

Fig 4b Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 2) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

Impactites and target rocks from the Kaumlrdla crater 431

Fig 4c Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K18 Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second) Thelithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

432 V Puura et al

impact lead and zinc sulphide and related carbonatemineralization in the Kaumlrdla rim wall and coveringsedimentary rocks belong to this event and consequently arenot related to the impact-event During the Mesozoic andCenozoic eras it was gradually unroofed The Cenozoicerosion (see Puura and Flodeacuten 1997) essentially uncoveredthe Kaumlrdla impact site

Settings of Studied Rock Units

Mixing of the debris produced from the almost 200 m-thick sedimentary layers and 300-m-thick portion of thecrystalline basement appears to be incomplete Theapproximately 50-m-thick allochthonous breccia unit thatoverlies the sub-crater basement is derived mainly fromcrystalline granitoids and amphibolites of the basement (coreK1 471ndash523 m see Fig 4b) This layer contains the moststrongly shocked mineral fragments (shock pressures of 20ndash35 GPa) The other 30-m-thick layer of the same compositionand with similarly shocked mineral clasts was penetrated atdepths between 356 and 380 m being separated from the lowerlayer by the mainly sedimentary rock-derived allochthonousslumped breccia layer (380ndash471 m) without any admixture ofshock metamorphic minerals The second crystalline-derivedbreccia layer is covered with the uppermost dominantlysedimentary-derived slumped resurge and air-fall breccias(300ndash356 m) with occasional admixture of clasts of crystallinerocks and crystalline-derived breccias We assume that theupper layer of crystalline-derived breccias with shockedminerals (356ndash380 m) is derived due to slumping of theuppermost parts of the rim wall where the lower breccia unitis presently lacking into the annular moat of the crater Theimpact breccias of these two layers show the most intenseevidence for the potassium enrichment and the sodium andcalcium depletion Inter-layers and patches of this breccia withsignatures of K-metasomatism are met also within thesedimentary-derived layers of the slump breccia

The sediment-derived breccia layers consist mainly oflarge blocks cobbles and pebbles of very soft rocks subjectedto low shock pressure The material probably slid back intothe crater via gullies that existed in the crater rimObservations on drill cores indicate that the sediment-derivedallochthonous breccias lack any other major secondarymineralization

The sub-crater basement has also been subjected toimpact-induced fracturing and brecciation at low shockpressures reaching ~2 GPa in the par-autochthonous brecciaunit in drill core K1 The K-enrichment and Na-Ca-depletionis gradually decreasing downward However the total volumeof K-enriched rocks is much larger than the volume of rockswith obvious shock-metamorphic signatures

In the geochemical study we concentrate on 1) impact-metamorphosed rocks of the allochthonous crystalline-derived breccia units 2) rocks of the sub-crater basement as

the main source for the breccias and 3) the possibleadmixture of an extraterrestrial component

SAMPLES AND ANALYTICAL TECHNIQUES

Optical Microscopy

Optical microscopy was performed on 175 thin sectionsat the Institute of Geology University of Tartu and theInstitute of Earth Sciences Oulu University (see theAppendix for detailed descriptions) The orientations ofplanar deformation features (PDFs) in quartz grains weremeasured in 32 thin sections from drill cores K1 and K18using standard universal stage (U-stage) techniques (seeFrench 1998) Electron microscopic studies were performedusing a JEOL 6300 scanning electron microscope (SEM)equipped with an Oxford ISIS energy dispersive spectrometerand a JEOL JSM-6300F field emission SEM at the Institute ofElectron Optics University of Oulu Finland

Electron Microprobe Analyses

Electron microprobe analyses (EPMA) of 20 thinsections were done with an automatized JEOL Superprobe733 using 15 kV voltage a 15 nA sample current and a 5 or10 m electron beam diameter at the Institute of ElectronOptics Oulu University Finland The quantitative chemicalcompositions of silicate minerals were calculated using theZAF correction program

X-Ray Powder Diffractometry

The mineralogical composition of bulk-rock samples andK-feldspar fractions was determined using a Dron-3Mdiffractometer with Ni filtered CuK -radiation at the Instituteof Geology University of Tartu Estonia Powdered rocksamples were scanned at step widths of 002deg 2 in the rangeof 2 and 70deg 2 with time steps of 5ndash10 sec Quantitativemineral composition of bulk-rock samples was found by full-profile Rietveld fitting using the Siroquant 25trade (Taylor1991) code The XRD patterns of K-feldspar fractions weredeconvoluted in the 20deg to 40deg 2 region into elementarypeaks using the Axes-19a code (Maumlndar et al 1996) Themeasured diffraction peaks were fitted with symmetricalpseudo-Voigt function shape curve doublets Relativeabundances of the K-feldspar structural varieties(crosshatched twin structure microcline and orthoclase) wereestimated from integral intensities of triclinic microcline(131) and (131) and monoclinic orthoclase (131) peaks

Geochemical Analyses

Geochemical analyses were performed on 90 samplesfrom both drill cores K1 and K18 and target rock samples

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 429

Fig 4a Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 1) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

430 V Puura et al

Fig 4b Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 2) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

Impactites and target rocks from the Kaumlrdla crater 431

Fig 4c Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K18 Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second) Thelithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

432 V Puura et al

impact lead and zinc sulphide and related carbonatemineralization in the Kaumlrdla rim wall and coveringsedimentary rocks belong to this event and consequently arenot related to the impact-event During the Mesozoic andCenozoic eras it was gradually unroofed The Cenozoicerosion (see Puura and Flodeacuten 1997) essentially uncoveredthe Kaumlrdla impact site

Settings of Studied Rock Units

Mixing of the debris produced from the almost 200 m-thick sedimentary layers and 300-m-thick portion of thecrystalline basement appears to be incomplete Theapproximately 50-m-thick allochthonous breccia unit thatoverlies the sub-crater basement is derived mainly fromcrystalline granitoids and amphibolites of the basement (coreK1 471ndash523 m see Fig 4b) This layer contains the moststrongly shocked mineral fragments (shock pressures of 20ndash35 GPa) The other 30-m-thick layer of the same compositionand with similarly shocked mineral clasts was penetrated atdepths between 356 and 380 m being separated from the lowerlayer by the mainly sedimentary rock-derived allochthonousslumped breccia layer (380ndash471 m) without any admixture ofshock metamorphic minerals The second crystalline-derivedbreccia layer is covered with the uppermost dominantlysedimentary-derived slumped resurge and air-fall breccias(300ndash356 m) with occasional admixture of clasts of crystallinerocks and crystalline-derived breccias We assume that theupper layer of crystalline-derived breccias with shockedminerals (356ndash380 m) is derived due to slumping of theuppermost parts of the rim wall where the lower breccia unitis presently lacking into the annular moat of the crater Theimpact breccias of these two layers show the most intenseevidence for the potassium enrichment and the sodium andcalcium depletion Inter-layers and patches of this breccia withsignatures of K-metasomatism are met also within thesedimentary-derived layers of the slump breccia

The sediment-derived breccia layers consist mainly oflarge blocks cobbles and pebbles of very soft rocks subjectedto low shock pressure The material probably slid back intothe crater via gullies that existed in the crater rimObservations on drill cores indicate that the sediment-derivedallochthonous breccias lack any other major secondarymineralization

The sub-crater basement has also been subjected toimpact-induced fracturing and brecciation at low shockpressures reaching ~2 GPa in the par-autochthonous brecciaunit in drill core K1 The K-enrichment and Na-Ca-depletionis gradually decreasing downward However the total volumeof K-enriched rocks is much larger than the volume of rockswith obvious shock-metamorphic signatures

In the geochemical study we concentrate on 1) impact-metamorphosed rocks of the allochthonous crystalline-derived breccia units 2) rocks of the sub-crater basement as

the main source for the breccias and 3) the possibleadmixture of an extraterrestrial component

SAMPLES AND ANALYTICAL TECHNIQUES

Optical Microscopy

Optical microscopy was performed on 175 thin sectionsat the Institute of Geology University of Tartu and theInstitute of Earth Sciences Oulu University (see theAppendix for detailed descriptions) The orientations ofplanar deformation features (PDFs) in quartz grains weremeasured in 32 thin sections from drill cores K1 and K18using standard universal stage (U-stage) techniques (seeFrench 1998) Electron microscopic studies were performedusing a JEOL 6300 scanning electron microscope (SEM)equipped with an Oxford ISIS energy dispersive spectrometerand a JEOL JSM-6300F field emission SEM at the Institute ofElectron Optics University of Oulu Finland

Electron Microprobe Analyses

Electron microprobe analyses (EPMA) of 20 thinsections were done with an automatized JEOL Superprobe733 using 15 kV voltage a 15 nA sample current and a 5 or10 m electron beam diameter at the Institute of ElectronOptics Oulu University Finland The quantitative chemicalcompositions of silicate minerals were calculated using theZAF correction program

X-Ray Powder Diffractometry

The mineralogical composition of bulk-rock samples andK-feldspar fractions was determined using a Dron-3Mdiffractometer with Ni filtered CuK -radiation at the Instituteof Geology University of Tartu Estonia Powdered rocksamples were scanned at step widths of 002deg 2 in the rangeof 2 and 70deg 2 with time steps of 5ndash10 sec Quantitativemineral composition of bulk-rock samples was found by full-profile Rietveld fitting using the Siroquant 25trade (Taylor1991) code The XRD patterns of K-feldspar fractions weredeconvoluted in the 20deg to 40deg 2 region into elementarypeaks using the Axes-19a code (Maumlndar et al 1996) Themeasured diffraction peaks were fitted with symmetricalpseudo-Voigt function shape curve doublets Relativeabundances of the K-feldspar structural varieties(crosshatched twin structure microcline and orthoclase) wereestimated from integral intensities of triclinic microcline(131) and (131) and monoclinic orthoclase (131) peaks

Geochemical Analyses

Geochemical analyses were performed on 90 samplesfrom both drill cores K1 and K18 and target rock samples

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

430 V Puura et al

Fig 4b Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K1 (part 2) Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second)The lithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

Impactites and target rocks from the Kaumlrdla crater 431

Fig 4c Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K18 Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second) Thelithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

432 V Puura et al

impact lead and zinc sulphide and related carbonatemineralization in the Kaumlrdla rim wall and coveringsedimentary rocks belong to this event and consequently arenot related to the impact-event During the Mesozoic andCenozoic eras it was gradually unroofed The Cenozoicerosion (see Puura and Flodeacuten 1997) essentially uncoveredthe Kaumlrdla impact site

Settings of Studied Rock Units

Mixing of the debris produced from the almost 200 m-thick sedimentary layers and 300-m-thick portion of thecrystalline basement appears to be incomplete Theapproximately 50-m-thick allochthonous breccia unit thatoverlies the sub-crater basement is derived mainly fromcrystalline granitoids and amphibolites of the basement (coreK1 471ndash523 m see Fig 4b) This layer contains the moststrongly shocked mineral fragments (shock pressures of 20ndash35 GPa) The other 30-m-thick layer of the same compositionand with similarly shocked mineral clasts was penetrated atdepths between 356 and 380 m being separated from the lowerlayer by the mainly sedimentary rock-derived allochthonousslumped breccia layer (380ndash471 m) without any admixture ofshock metamorphic minerals The second crystalline-derivedbreccia layer is covered with the uppermost dominantlysedimentary-derived slumped resurge and air-fall breccias(300ndash356 m) with occasional admixture of clasts of crystallinerocks and crystalline-derived breccias We assume that theupper layer of crystalline-derived breccias with shockedminerals (356ndash380 m) is derived due to slumping of theuppermost parts of the rim wall where the lower breccia unitis presently lacking into the annular moat of the crater Theimpact breccias of these two layers show the most intenseevidence for the potassium enrichment and the sodium andcalcium depletion Inter-layers and patches of this breccia withsignatures of K-metasomatism are met also within thesedimentary-derived layers of the slump breccia

The sediment-derived breccia layers consist mainly oflarge blocks cobbles and pebbles of very soft rocks subjectedto low shock pressure The material probably slid back intothe crater via gullies that existed in the crater rimObservations on drill cores indicate that the sediment-derivedallochthonous breccias lack any other major secondarymineralization

The sub-crater basement has also been subjected toimpact-induced fracturing and brecciation at low shockpressures reaching ~2 GPa in the par-autochthonous brecciaunit in drill core K1 The K-enrichment and Na-Ca-depletionis gradually decreasing downward However the total volumeof K-enriched rocks is much larger than the volume of rockswith obvious shock-metamorphic signatures

In the geochemical study we concentrate on 1) impact-metamorphosed rocks of the allochthonous crystalline-derived breccia units 2) rocks of the sub-crater basement as

the main source for the breccias and 3) the possibleadmixture of an extraterrestrial component

SAMPLES AND ANALYTICAL TECHNIQUES

Optical Microscopy

Optical microscopy was performed on 175 thin sectionsat the Institute of Geology University of Tartu and theInstitute of Earth Sciences Oulu University (see theAppendix for detailed descriptions) The orientations ofplanar deformation features (PDFs) in quartz grains weremeasured in 32 thin sections from drill cores K1 and K18using standard universal stage (U-stage) techniques (seeFrench 1998) Electron microscopic studies were performedusing a JEOL 6300 scanning electron microscope (SEM)equipped with an Oxford ISIS energy dispersive spectrometerand a JEOL JSM-6300F field emission SEM at the Institute ofElectron Optics University of Oulu Finland

Electron Microprobe Analyses

Electron microprobe analyses (EPMA) of 20 thinsections were done with an automatized JEOL Superprobe733 using 15 kV voltage a 15 nA sample current and a 5 or10 m electron beam diameter at the Institute of ElectronOptics Oulu University Finland The quantitative chemicalcompositions of silicate minerals were calculated using theZAF correction program

X-Ray Powder Diffractometry

The mineralogical composition of bulk-rock samples andK-feldspar fractions was determined using a Dron-3Mdiffractometer with Ni filtered CuK -radiation at the Instituteof Geology University of Tartu Estonia Powdered rocksamples were scanned at step widths of 002deg 2 in the rangeof 2 and 70deg 2 with time steps of 5ndash10 sec Quantitativemineral composition of bulk-rock samples was found by full-profile Rietveld fitting using the Siroquant 25trade (Taylor1991) code The XRD patterns of K-feldspar fractions weredeconvoluted in the 20deg to 40deg 2 region into elementarypeaks using the Axes-19a code (Maumlndar et al 1996) Themeasured diffraction peaks were fitted with symmetricalpseudo-Voigt function shape curve doublets Relativeabundances of the K-feldspar structural varieties(crosshatched twin structure microcline and orthoclase) wereestimated from integral intensities of triclinic microcline(131) and (131) and monoclinic orthoclase (131) peaks

Geochemical Analyses

Geochemical analyses were performed on 90 samplesfrom both drill cores K1 and K18 and target rock samples

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 431

Fig 4c Lithological log of the drill core grain size distribution and lithological composition of allochthonous breccias from the crater interiorin borehole K18 Visual logs with -ray graphs are given in American Petroleum Institute (API) units (normalized for counts per second) Thelithological composition is based on visual counting and the -ray log ldquoxrdquo marks the positions of samples taken for geochemical analyses

432 V Puura et al

impact lead and zinc sulphide and related carbonatemineralization in the Kaumlrdla rim wall and coveringsedimentary rocks belong to this event and consequently arenot related to the impact-event During the Mesozoic andCenozoic eras it was gradually unroofed The Cenozoicerosion (see Puura and Flodeacuten 1997) essentially uncoveredthe Kaumlrdla impact site

Settings of Studied Rock Units

Mixing of the debris produced from the almost 200 m-thick sedimentary layers and 300-m-thick portion of thecrystalline basement appears to be incomplete Theapproximately 50-m-thick allochthonous breccia unit thatoverlies the sub-crater basement is derived mainly fromcrystalline granitoids and amphibolites of the basement (coreK1 471ndash523 m see Fig 4b) This layer contains the moststrongly shocked mineral fragments (shock pressures of 20ndash35 GPa) The other 30-m-thick layer of the same compositionand with similarly shocked mineral clasts was penetrated atdepths between 356 and 380 m being separated from the lowerlayer by the mainly sedimentary rock-derived allochthonousslumped breccia layer (380ndash471 m) without any admixture ofshock metamorphic minerals The second crystalline-derivedbreccia layer is covered with the uppermost dominantlysedimentary-derived slumped resurge and air-fall breccias(300ndash356 m) with occasional admixture of clasts of crystallinerocks and crystalline-derived breccias We assume that theupper layer of crystalline-derived breccias with shockedminerals (356ndash380 m) is derived due to slumping of theuppermost parts of the rim wall where the lower breccia unitis presently lacking into the annular moat of the crater Theimpact breccias of these two layers show the most intenseevidence for the potassium enrichment and the sodium andcalcium depletion Inter-layers and patches of this breccia withsignatures of K-metasomatism are met also within thesedimentary-derived layers of the slump breccia

The sediment-derived breccia layers consist mainly oflarge blocks cobbles and pebbles of very soft rocks subjectedto low shock pressure The material probably slid back intothe crater via gullies that existed in the crater rimObservations on drill cores indicate that the sediment-derivedallochthonous breccias lack any other major secondarymineralization

The sub-crater basement has also been subjected toimpact-induced fracturing and brecciation at low shockpressures reaching ~2 GPa in the par-autochthonous brecciaunit in drill core K1 The K-enrichment and Na-Ca-depletionis gradually decreasing downward However the total volumeof K-enriched rocks is much larger than the volume of rockswith obvious shock-metamorphic signatures

In the geochemical study we concentrate on 1) impact-metamorphosed rocks of the allochthonous crystalline-derived breccia units 2) rocks of the sub-crater basement as

the main source for the breccias and 3) the possibleadmixture of an extraterrestrial component

SAMPLES AND ANALYTICAL TECHNIQUES

Optical Microscopy

Optical microscopy was performed on 175 thin sectionsat the Institute of Geology University of Tartu and theInstitute of Earth Sciences Oulu University (see theAppendix for detailed descriptions) The orientations ofplanar deformation features (PDFs) in quartz grains weremeasured in 32 thin sections from drill cores K1 and K18using standard universal stage (U-stage) techniques (seeFrench 1998) Electron microscopic studies were performedusing a JEOL 6300 scanning electron microscope (SEM)equipped with an Oxford ISIS energy dispersive spectrometerand a JEOL JSM-6300F field emission SEM at the Institute ofElectron Optics University of Oulu Finland

Electron Microprobe Analyses

Electron microprobe analyses (EPMA) of 20 thinsections were done with an automatized JEOL Superprobe733 using 15 kV voltage a 15 nA sample current and a 5 or10 m electron beam diameter at the Institute of ElectronOptics Oulu University Finland The quantitative chemicalcompositions of silicate minerals were calculated using theZAF correction program

X-Ray Powder Diffractometry

The mineralogical composition of bulk-rock samples andK-feldspar fractions was determined using a Dron-3Mdiffractometer with Ni filtered CuK -radiation at the Instituteof Geology University of Tartu Estonia Powdered rocksamples were scanned at step widths of 002deg 2 in the rangeof 2 and 70deg 2 with time steps of 5ndash10 sec Quantitativemineral composition of bulk-rock samples was found by full-profile Rietveld fitting using the Siroquant 25trade (Taylor1991) code The XRD patterns of K-feldspar fractions weredeconvoluted in the 20deg to 40deg 2 region into elementarypeaks using the Axes-19a code (Maumlndar et al 1996) Themeasured diffraction peaks were fitted with symmetricalpseudo-Voigt function shape curve doublets Relativeabundances of the K-feldspar structural varieties(crosshatched twin structure microcline and orthoclase) wereestimated from integral intensities of triclinic microcline(131) and (131) and monoclinic orthoclase (131) peaks

Geochemical Analyses

Geochemical analyses were performed on 90 samplesfrom both drill cores K1 and K18 and target rock samples

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

432 V Puura et al

impact lead and zinc sulphide and related carbonatemineralization in the Kaumlrdla rim wall and coveringsedimentary rocks belong to this event and consequently arenot related to the impact-event During the Mesozoic andCenozoic eras it was gradually unroofed The Cenozoicerosion (see Puura and Flodeacuten 1997) essentially uncoveredthe Kaumlrdla impact site

Settings of Studied Rock Units

Mixing of the debris produced from the almost 200 m-thick sedimentary layers and 300-m-thick portion of thecrystalline basement appears to be incomplete Theapproximately 50-m-thick allochthonous breccia unit thatoverlies the sub-crater basement is derived mainly fromcrystalline granitoids and amphibolites of the basement (coreK1 471ndash523 m see Fig 4b) This layer contains the moststrongly shocked mineral fragments (shock pressures of 20ndash35 GPa) The other 30-m-thick layer of the same compositionand with similarly shocked mineral clasts was penetrated atdepths between 356 and 380 m being separated from the lowerlayer by the mainly sedimentary rock-derived allochthonousslumped breccia layer (380ndash471 m) without any admixture ofshock metamorphic minerals The second crystalline-derivedbreccia layer is covered with the uppermost dominantlysedimentary-derived slumped resurge and air-fall breccias(300ndash356 m) with occasional admixture of clasts of crystallinerocks and crystalline-derived breccias We assume that theupper layer of crystalline-derived breccias with shockedminerals (356ndash380 m) is derived due to slumping of theuppermost parts of the rim wall where the lower breccia unitis presently lacking into the annular moat of the crater Theimpact breccias of these two layers show the most intenseevidence for the potassium enrichment and the sodium andcalcium depletion Inter-layers and patches of this breccia withsignatures of K-metasomatism are met also within thesedimentary-derived layers of the slump breccia

The sediment-derived breccia layers consist mainly oflarge blocks cobbles and pebbles of very soft rocks subjectedto low shock pressure The material probably slid back intothe crater via gullies that existed in the crater rimObservations on drill cores indicate that the sediment-derivedallochthonous breccias lack any other major secondarymineralization

The sub-crater basement has also been subjected toimpact-induced fracturing and brecciation at low shockpressures reaching ~2 GPa in the par-autochthonous brecciaunit in drill core K1 The K-enrichment and Na-Ca-depletionis gradually decreasing downward However the total volumeof K-enriched rocks is much larger than the volume of rockswith obvious shock-metamorphic signatures

In the geochemical study we concentrate on 1) impact-metamorphosed rocks of the allochthonous crystalline-derived breccia units 2) rocks of the sub-crater basement as

the main source for the breccias and 3) the possibleadmixture of an extraterrestrial component

SAMPLES AND ANALYTICAL TECHNIQUES

Optical Microscopy

Optical microscopy was performed on 175 thin sectionsat the Institute of Geology University of Tartu and theInstitute of Earth Sciences Oulu University (see theAppendix for detailed descriptions) The orientations ofplanar deformation features (PDFs) in quartz grains weremeasured in 32 thin sections from drill cores K1 and K18using standard universal stage (U-stage) techniques (seeFrench 1998) Electron microscopic studies were performedusing a JEOL 6300 scanning electron microscope (SEM)equipped with an Oxford ISIS energy dispersive spectrometerand a JEOL JSM-6300F field emission SEM at the Institute ofElectron Optics University of Oulu Finland

Electron Microprobe Analyses

Electron microprobe analyses (EPMA) of 20 thinsections were done with an automatized JEOL Superprobe733 using 15 kV voltage a 15 nA sample current and a 5 or10 m electron beam diameter at the Institute of ElectronOptics Oulu University Finland The quantitative chemicalcompositions of silicate minerals were calculated using theZAF correction program

X-Ray Powder Diffractometry

The mineralogical composition of bulk-rock samples andK-feldspar fractions was determined using a Dron-3Mdiffractometer with Ni filtered CuK -radiation at the Instituteof Geology University of Tartu Estonia Powdered rocksamples were scanned at step widths of 002deg 2 in the rangeof 2 and 70deg 2 with time steps of 5ndash10 sec Quantitativemineral composition of bulk-rock samples was found by full-profile Rietveld fitting using the Siroquant 25trade (Taylor1991) code The XRD patterns of K-feldspar fractions weredeconvoluted in the 20deg to 40deg 2 region into elementarypeaks using the Axes-19a code (Maumlndar et al 1996) Themeasured diffraction peaks were fitted with symmetricalpseudo-Voigt function shape curve doublets Relativeabundances of the K-feldspar structural varieties(crosshatched twin structure microcline and orthoclase) wereestimated from integral intensities of triclinic microcline(131) and (131) and monoclinic orthoclase (131) peaks

Geochemical Analyses

Geochemical analyses were performed on 90 samplesfrom both drill cores K1 and K18 and target rock samples

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 433

from outside the crater rim (cores F361 and F364 seeKoppelmaa et al 1996)

Major minor and trace element abundances of 51samples were determined using instrumental neutronactivation analysis (INAA) and X-ray fluorescencespectrometry (XRF) Rock samples were crushed manually inplastic wrap then mechanically in an alumina jaw crusherand were powdered using an automatic agate mill XRFanalyses were performed at the University of theWitwatersrand Johannesburg South Africa for thedetermination of major and some trace element abundances(Sr Y Zr Nb Ni Cu V and Ba) following proceduresdescribed by Reimold et al (1994) These authors also giveinformation on the precision and accuracy of this method

Another 15 samples were analyzed for major and traceelement composition using XRD (Toumlroumlk and van Grieken1992) in the central laboratory of the Federal Institute ofGeosciences and Natural Resources (BGR) in Hannover(Germany) For quantitative analysis a wavelength-dispersive X-ray spectrometer (WD-XRF) was used with achromium tube (PW 1480) and a rhodium tube (PW 2400)For details of measurement procedures precision andaccuracy see Toumlroumlk et al (1996)

In the laboratory of the Geological Survey of Estonia 58samples were crushed in a jaw crusher with carbon steel jawsand powdered in a ceramic disc-mill for wet chemical analysesof major elements For details see Kivisilla et al (1999)

Aliquots of about 150 mg were analyzed at theUniversity of Vienna Austria with INAA together with thewell-characterized standard reference materials the granitesAC-E and USGS G-2 (Govindaraju 1989) the Allendemeteorite standard reference powder (Jarosewich et al 1987)and (for Au and Ir) the mineralized gabbro PGE standardWMG-1 (Canmet 1994) Koeberl (1993) and Koeberl et al(1987) described the details of instrumentation methodologyprecision and accuracy

In addition the iridium content of seven samples wasdetermined with the -coincidence spectrometry system atthe Institute of Geochemistry (University of Vienna) Fordetails on the measurement procedure and instrumentationsee Koeberl and Huber (2000)

RESULTS AND DISCUSSION

Impact Stratigraphy and Lithology

The sequence of crater-related lithologies consists offrom the bottom to the top a) presumably intact crystallinerocks of the deep basement b) autochthonous and par-autochthonous fractured and brecciated rocks beneath thecrater floor (sub-crater basement) and in ring walls c)allochthonous breccias within the crater (lt250 m thick) andoutside the wall (ejecta layer ~25 m thick) and d) post-impact(cover) deposits

To compare the different impact stratigraphic units westudied a variety of samples from core sections especially inthe main core K1 (originally labeled as F373) and K18 asshown in Figs 4andash4c and described in the Appendix

Unaffected basement rocks were drilled outside the crater(F361 F364) Autochthonous fractured and brecciated rockscontinue below the bottom of 8152 m in drill hole K1 Asestimated from geophysical data and reconstruction of thetransient cavity these deformed rocks may extend to a depthof 1000 m (Plado et al 1996) Their total thicknessconsequently may exceed 500 m In the subcrater rocksdikes of fine-grained breccia are common but decrease innumber and size with depth (in K-1 to a depth of 800 m) Inbrecciated rocks the same matrix fills interstitial spaces Thecontact between the fractured and brecciated rocks is gradualThe thickness of the strongly brecciated zone at the craterfloor in core K1 is estimated at ~30 m in K18 at over 30 mand in the Palukuumlla rim wall at over 100 m This layer iscovered by par-autochthonous breccias forming theuppermost part of the sub-crater basement in core K1

In the studied cores in the crater (K1 K18) up to fivedifferent impact-derived allochthonous crater-fill layers canbe distinguished (Fig 4) Cores K12 412 and 383 (forlocations see Fig 2) were investigated only macroscopically

In the crater depression horizontal allochthonous crater-fill layers conceal the central uplift They perfectly leveled thecrater floor before the post-impact sedimentation started Onthe inner slopes of the crater breccia layers thin out at depthsthat are not much higher than the central uplift (see Fig 3)The youngest impact-related sediments (laminated carbonate-siliciclastic muds) cover a large area and were probablydeposited from post-impact suspension-rich water In gulliesthis sediment also covers the lowest portions of the rim

The regular concentric internal structure of theallochthonous impact breccias suggests that they formed earlyduring the active modification stage The lowermost unit ofcrystalline-derived breccias (K1 523ndash471 m) is of almostpar-autochthonous composition and probably formed alongthe contact of the impact plume and transient cavity floor Thesecond unit (K1 471ndash380 m) composed of sedimentary-derived breccia probably filled the moat around the risingcentral uplift during the early slumping process The thirdlayer of crystalline-derived breccias (K1 380ndash356 m) issimilar in texture and composition to the first unit and mayhave formed by slumping of parts of the first layer that weredeposited at or near the crater rim The crystalline-derivedbreccias of the first and third layers are dominantly monomictcontaining only small grains of sedimentary material Thefourth layer (K1 356ndash300 m) is composed mainly ofsedimentary rocks with a few clasts of crystalline rocks In theupper part of this unit (~315 m) the abundance of crystallineclasts increases In the northeast of the crater rim at PalukuumllaHigh (core P11) matrix-rich impact breccia was found inpockets within the fractured crystalline rock

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

434 V Puura et al

In the first and third crystalline-derived breccia units thematrix is also usually crystalline-derived Crystalline-derivedallochthonous breccia matrix is more or less strongly weldedsuggesting hot deposition Mixing of crystalline andsedimentary material in the plume is indicated by fine clasticsand to pelitic particles in the upper layers of the brecciamatrix Sedimentary-derived breccias of the third and fourthlayers consist of large blocks and smaller fragments ofsandstone claystone and limestone (Fig 4) They arecement-poor and mainly of slump origin Inter-clast cement ispresent as soft sedimentary rock powder with clasts ofsedimentary origin Larger fragments of crystalline materialare found mainly as lumps of larger clasts already envelopedin fine impact-welded crystalline-type cement

Slumping processes were essential during the formationof the within-crater stratigraphy Thinning of the lowerbreccia layers toward the central uplift and all breccia layerstoward the crystalline rim wall at Palukuumlla can be observedReplacement deformation doubling and even tripling ofearly breccia layers could have taken place

Late resurge processes accumulated coarse-grainedmaterial such as blocks cobbles pebbles and gravel andwere replaced by gradational sedimentation first of gravel andsand and then of carbonate mud Post-impact Late Ordoviciancarbonate sedimentation filled the crater depression and laterSilurian sediments covered the impact site

Spatial Distribution of Shock Metamorphic Features andImpact-Induced Alteration of Minerals

Shock metamorphic features were seen in quartzmicrocline and biotite Due to impact-induced alterationplagioclase and hornblende were replaced by submicroscopicorthoclase and chlorite aggregates (see Table 1) Distinctshock metamorphic features are found only in minerals of the

allochthonous breccia The volume of rocks subjected tochemicalmineralogical alteration is considerably largercomprising also parautochthonous breccias and the fracturedsub-crater basement

Deformation of QuartzQuartz with micro-scale fractures and voids is

characteristic for a large volume of both allochthonous andautochthonous impact-affected rocks (Puura et al 2000a)PDFs in quartz were found predominantly in crystalline rockclasts of the granitoid-dominated polymict breccia from coresK1 and K18 and in slump breccia from K1 In additionweakly developed PDFs were found in quartz grains fromproximal ejecta (cores K14 K15 K19 in Fig 3) which aremainly composed of rounded quartz from Cambrian sand-and siltstone with few mineral clasts derived from thecrystalline basement No PDFs were found in quartz from theautochthonous and par-autochthonous sub-crater basementfrom crystalline rocks of the rim wall or from the lowerpolymict breccia unit with sedimentary clasts and upper unitswith sedimentary and crystalline clasts Diaplectic glass wasnot found

The PDF orientation data obtained by U-stage studies ofthin sections from the granitoid polymict breccia samples aregiven in Fig 5 The number of PDF sets in quartz crystalsfrom different units is typically one or two with a maximumof three sets (Fig 6) The abundance of PDFs decorated withfluid inclusions is low amounting to only 13 of theundecorated PDF systems The crystallographic orientationof PDF systems against quartz optical axis gives very broadmaxima at 1013 ( ) (Figs 5a and 5b) The distributionpatterns of the upper parts of K1 and K18 display a secondmaximum at an angle slightly lower than that of the 1012( ) orientation This could be due to the binning (3deg inFig 5a) or could indicate a problem with the precision of the

Table 1 Summary of the state of rock-forming minerals in different units (rim K1 K18) of the Kaumlrdla crater that were subjected to impact-induced alteration and replacementa

Unit Depth (m) Microcline Plagioclase Hornblende Biotite

Subcrater basementK1 6200ndash8152 Fresh Or relict or fresh Pl Fresh relict Hbl Fresh rare Chl

5890ndash6200 Fresh Or relict Pl Chl Fresh rare ChlK18 4000ndash4314 ndash Or relict Pl Chl Fresh

Rim wall ndash Fresh Or fresh relict or weathered Pl Chl relict Hbl Fresh Chl

Par-autochthonous basementK1 5228ndash5890 Fresh Or relict Pl Chl Fresh rare Chl

Allochthonous crystalline-derived brecciaK1 4710ndash5228 Fresh Or Or Chl Fresh rusty patches

3560ndash3800 Fresh Or Or Chl Fresh partly ChlK18 3560ndash4000 Fresh Or Or Chl Fresh partly ChlK1 3000ndash3400 Fresh Or Or Chl FreshK18 2927ndash3400 Fresh Or Chl Fresh

aOr = orthoclase Pl = plagioclase Hbl = hornblende Chl = chlorite

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 435

measurements This could also be the reason for the highpercentage of unindexed planes (about 50 by the method ofGrieve et al [1996]) In the lower parts of cores K1 and K18the distribution of PDF orientations changes indicating adominance of orientations The PDF orientations indicate ashock pressure in the range of ~10 to 35 GPa with highervalues characteristic for core K1and lower ones for K18

Deformation and Alteration of MicroclinePrimary microcline (Mc) with well-developed

crosshatched twin structure has been mostly deformed andfractured by shock metamorphism Inside the cratermicrocline clasts with fractures and submicroscopic voidsrare deformation bands and kink bands are observedMicrocline with planar fractures (PFs) and a ladder structure

Fig 5 Distribution of PDF orientations in quartz of granite-derived polymict breccia samples from cores K1 and K18 Top histogramsshowing the number of PDF planes in 3deg binning versus the angle between the c-axis and PDFs Bottom histogram combining all sets fromthe top with 5deg binning

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

436 V Puura et al

Fig 6 Photomicrographs a and b) allochthonous breccia on top of the central uplift thin section K18-3830 (plane polarized light) a) a quartz(Qtz) clast with 3 sets of PDFs b) microcline (Mc) replaced by secondary orthoclase (Or) that is pigmented by hematite (Hem) This large Orclast is kink-banded and crosscut by 3 systems of PDFs The surrounding fine-grained matrix contains medium-sized quartz clasts c)allochthonous breccia thin section K18-3812 crossed polars The breccia contains quartz clasts with one or two PDF systems and orthoclasereplacing strongly deformed microcline which has lost its primary twin structure d) breccia dike injected into the uppermost part of the sub-crater basement thin section K1-6060 crossed polars Quartz clasts are irregularly fractured and blocky Microcline (Mc) either survived as afragment with slightly deformed (kinked) twin structure or is replaced by secondary orthoclase without a crosshatched twin structure e) shearedgranite from near the base of the parautochthonous breccia layer thin section of sample K1-5870 (plane polarized light) The large quartzfragment in the center displays crossing planar and sub-planar to conical fracture systems The black rectangle corresponds to the close-up of(f) f) near the center of the cones impact-related planer fractures divide the quartz grain into lamellae resembling tectonic Boumlhm lamellae

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 437

occasionally occurs in rocks which also contain PDF-bearingquartz (Fig 6a) In places in the allochthonous breccia themicrocline twins are intensely deformed along multipleparallel fractures In those samples many microcline grainshave partially or completely lost their optical evidences ofpolysynthetic twinning which is in accordance with theresults of XRD study According to XRD data the orthoclasemicrocline (OrMc) ratio of K-Fsp megacrysts and clasts(Fig 6b) varies from ~1 and less in the fractured sub-craterbasement to 6ndash7 in the allochthonous breccia (Table 2 Fig7) Within the K-feldspar-yielding fine-grained allochthonousbreccia matrix and also in mm-size K-feldspar-yielding lithicclasts the OrMc ratio increases up to 21 In crosshatchedmicrocline from outside the crater the OrMc ratio is around01ndash02

Large microcline clasts that derived from megacrysts ofgranitoids were found in the allochthonous breccia thatincludes abundant PDF-bearing quartz and were studied usingbackscattered electron images EPMA and XRD techniquesIt was observed that in the course of alteration K-Fspindistinguishable from the surrounding K-Fsp replacesperthitic lamellae During this process the Na and Cacontents of the primary perthitic microcline decreased (Puuraet al 2000b) In breccias devoid of shocked quartz and othershocked minerals microcline clasts are mostly intact

Alteration of Plagioclase and Hornblende Primary plagioclase in biotite gneisses granites and

amphibolites is in part replaced by submicroscopic orthoclaseaggregates (Puura et al 2000a) Partially altered plagioclasecan contain relict plagioclase with polysynthetic twinning aswell as cryptocrystalline orthoclase replacing the plagioclase(eg Fig 8) The composition of primary and secondaryfeldspars is given in Table 3 Replacement by orthoclase isprogressing along the plagioclase crystal rim and withincrystal fractures and related micro-channels The boundary

between primary plagioclase and secondary orthoclase issharp at the resolution of the microprobe (of about 5 m)Voids less than 10 m in diameter follow the impact-inducedmicro-fracture systems in replaced parts of the primaryplagioclase crystals This replacement is well-developed inboth felsic and mafic rocks In the allochthonous brecciacontaining quartz with PDF secondary K-feldspar stronglydominates over primary plagioclase In the sub-craterbasement partially altered plagioclase crystals and zonedrock clasts with both altered and unaltered plagioclases arefound Hornblende is the other mineral of mafic rockssensitive to impact-induced alteration In the allochthonousbreccia hornblende is totally replaced by an iron-rich chloriteand corrensite (Kirsimaumle et al 2002) In the sub-craterbasement some cm- to m-sized amphibolite bodies have azoned structure In their interior almost fresh hornblende andplagioclase are preserved along the perimeter and infractures hornblende is almost totally replaced by chloriteand plagioclase by orthoclase In the sub-crater basement(cores K1 and K18) the content of altered hornblendedecreases with depth

Geochemistry

Major and Trace Elemental CompositionThe results of the geochemical analysis are summarized

in Table 4 The results of the individual analysis are availableon request from the authors The target rocks comprise mainlygranites and amphibolites and accessory quartzites andlimestone The elemental abundances of the target rocks showconsiderable variations but the different rock types still havedistinct average compositions (Table 4)

Major element variations versus SiO2-contents are shownin Harker diagrams (Fig 9) for both target rocks (opensymbols) and impact breccias (filled symbols) Amphibolitesplot to the left (lower SiO2-contents) limestone to the lower

Fig 7 Variation of orthoclase-to-microcline (Or-Mc) ratios with depth in rocks from cores K1 and K18 and for comparison from core F364from outside the crater a = K-Fsp clasts (megacrysts) in the breccia matrix lithic clasts in the allochthonous impact breccia and in rocks ofthe sub-crater basement b = K-Fsp-rich allochthonous breccia matrix c = fine-grained K-Fsp in lithic clasts from the allochthonous brecciaand the sub-crater basement

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

438 V Puura et al

Table 2 Orthoclasemicrocline ratio (OrMc) determined by X-ray diffraction in various clasts megacrysts and bulk rock powder of impact breccias and sub-crater basement The sub-divisions are according to Fig 8 The numbers of the samples consist of the drill core designation (K1 K18 or F364) and the depth of sampling (in 01 m)

Sample OrMc Lithology

K-feldspar megacrysts and clasts

Uppermost allochthonous breccia unitsK18-3066b 104 K-Fsp-Qtz aggregate clast in breccia matrixK18-3112 034 K-Fsp-Qtz aggregate in granitoid clastK1-3420 132 K-Fsp clast in breccia

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3803 097 K-Fsp clast in breccia matrixK18-3830 336 K-Fsp clast in breccia matrixK18-3930 578 K-Fsp clast in breccia matrixK1-3570 097 K-Fsp clast in breccia matrixK1-3610 137 K-Fsp-Qtz aggregate clast in breccia matrixK1-4750a 710 K-Fsp inside of a 5 cm-size granite clastK1-4750b 162 K-Fsp of fine chips from breccia matrixK1-4836 559 K-Fsp megacryst in a migmatite clastK1-4891 172 K-Fsp megacryst in granite block

Par-autochthonous breccia unitK1-5299 264 K-Fsp clast in breccia matrixK1-5368 402 K-Fsp small clast (less than 1cm) in breccia matrix

Sub-crater basement unitsK18-4243 168 K-Fsp-Qtz aggregate in graniteK1-5895 112 K-Fsp-Qtz aggregate in graniteK1-6647 134 K-Fsp megacryst in graniteK1-7603 069 K-Fsp megacryst in graniteK1-8064 026 K-Fsp-Qtz aggregate in granite

Bulk rock powders from breccia matrix

Crystalline basement-derived allochtonous breccia units with PDFs in quartzK18-3830 206 Breccia matrixK18-3865 708 Breccia matrixK1-3620 673 Breccia matrixK1-3960 331 Breccia matrixK1-5130 256 Breccia matrix

Par-autochthonous breccia unitK1-5770 147 Breccia matrixK1-5870 758 Breccia matrix

Rock powders from basement breccia matrix and rock clasts

Uppermost allochthonous breccia unitsK1-3255 208 Biotite gneiss clastK1-3405b 135 Amphibolite clastK1-3405d 150 Amphibolite clastK1-3530 147 Breccia matrix

Sub-crater basement unitsK1-6030 271 AmphiboliteK1-6060 263 GraniteK1-7053 086 Granite

Basement rocks outside the crater rimF364-3425 006 GraniteF364-3110 019 Granite

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 439

left quartzites to the far right and granites at about 70 wtsilica While the contents of Fe2O3 and Al2O3 are linearlycorrelated in target rocks and breccias K2O is enriched andNa2O and CaO are generally depleted in the brecciascompared to the target rocks As mentioned the K-enrichment results from the replacement of plagioclase byorthoclase and is prominent at many impact craters eg at

Boltysh (Gurov et al 1986) Manson (Koeberl et al 1996)Ames (Koeberl et al 1997) Gardnos (French et al 1997) orIlyinets (Gurov et al 1998) Generally the targetamphibolites have distinctly higher Sc V Cr Co and Ni butlower Hf and Th contents than the other basement rocks Theelement ratios such as KU ThU LaTh and HfTa alsoindicate characteristic differences for each target lithology

Fig 8 Amphibolite K1-6030 from the sub-crater basement a) photomicrograph parallel polarizers Electron microprobe analysis was usedto determine the chemical composition of fresh and altered plagioclase (Pl) The numbers at the squares correspond to analyzes given inTable 3 The light colored area is a large crystal of plagioclase replaced by orthoclase (Or) at its margin and along micro-fractures Plagioclaseis surrounded by hornblende (Hbl) which is partly replaced by chlorite (Chl) The black frame corresponds to (b) b) BSE image Plagioclase(grey) is replaced by orthoclase (white) at the lower right corner In unaltered plagioclase orthoclase is formed along fractures and porousregions The inclusion of hornblende in altered pagioclase is partially replaced by chlorite

Table 3 Electron microprobe (EMP) data for the chemical composition of a plagioclase grain partially replaced by orthoclase in amphibolite K1-6030 from the sub-crater basement (for analyzed spots see Fig 8)a Sample 1 2 3 4 5 6 7 8 9 10 11wt Pl Or Pl Or Pl Or Or Pl Or Pl Pl

SiO2 569 618 557 614 572 602 625 572 617 578 574TiO2 ndash ndash 004 ndash 001 002 ndash 002 ndash 004 006Al2O3 236 167 244 170 236 166 173 235 168 236 235MgO ndash ndash ndash 001 ndash ndash 006 001 ndash ndash ndashFeO 006 ndash 003 005 015 023 030 013 015 009 006MnO ndash ndash 001 008 004 011 ndash ndash 001 002 ndashCaO 668 ndash 745 ndash 651 023 ndash 647 002 687 690Na2O 796 007 732 004 800 012 009 814 005 799 817K2O 013 141 029 148 022 143 145 018 145 025 027

Total 9533 9267 9524 9338 9573 9181 9475 9565 9323 9666 9636

Normative composition of the plagioclase grainb

Ab 678 07 630 04 681 13 09 688 05 669 672An 315 00 354 00 306 13 00 302 01 318 314Or 07 993 16 996 13 974 991 10 994 13 14

aPl = plagioclase Or = orthoclasebAll data in Ab = Albite An = Anorthite Or = Orthoclase Calibration of the EMP was done with the registered standard No MAC 3056 of Agar Scientific

Ltd UK The detection limits of the analyzed elements are about 001 wt

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

440 V Puura et al

Table 4 Average compositions of target rocks and impact breccias from Kaumlrdla cratera

Granite(25)

Quartzite(5)

Amphibolite(12)

Limestone(2)

Breccia matrix(13)

Breccia with granite clasts (17)

Breccia withamph clasts (4)

Avg Avg Avg Avg Avg Avg Avg

SiO2 724 30 830 27 478 32 492 121 679 62 773 84 515 11TiO2 040 017 021 007 097 036 017 002 071 047 037 022 084 023Al2O3 123 16 722 127 152 13 079 001 122 13 100 34 133 17Fe2O3 329 130 174 050 124 11 121 112 465 356 242 183 937 176MnO 002 002 001 000 012 004 009 001 004 002 003 002 013 008MgO 063 036 068 055 102 12 020 013 204 173 079 100 808 235CaO 064 057 023 017 540 329 265 61 047 085 024 012 339 373Na2O 165 125 047 029 160 090 011 003 018 010 043 035 010 003K2O 681 119 510 106 259 148 039 026 854 146 702 219 575 055P2O5 009 010 004 001 013 008 018 013 009 008 005 002 010 001LOI 102 063 084 064 338 247 213 51 307 236 087 041 743 230

Total 9912 9962 9973 10011 9972 9848 10006

Sc 523 274 209 138 397 54 137 063 128 49 461 365 371 179V 177 220 218 104 291 72 143 95 845 442 343 280 205 49Cr 37 27 191 158 263 130 104 05 469 208 166 163 130 36Co 43 40 25 07 428 102 21 10 115 80 42 43 237 49Ni 54 29 61 47 933 347 105 07 210 159 85 107 458 161Cu 503 124 62 116 358 389 10 00 153 293 30 22 78 90Zn 539 225 178 97 100 117 54 28 650 427 302 187 105 193As 028 018 023 017 069 038 334 278 029 030 028 018 050 033Rb 185 461 120 172 50 24 140 03 150 19 156 44 103 4 7Sr 664 451 632 141 143 671 585 205 579 66 773 21 455 49Y 471 202 134 59 286 176 80 14 423 450 282 207 385 58Zr 382 192 109 26 83 61 135 66 226 45 218 127 131 61Nb 150 44 72 29 118 68 70 00 155 53 122 52 200 83Sb 343 605 223 490 031 021 021 011 501 696 577 649 026 024Cs 134 093 110 081 067 042 036 017 238 113 182 113 142 112Ba 735 514 584 556 148 329 907 356 742 2781 869 288 365 229La 131 735 318 235 130 104 100 41 654 396 409 137 239 185Ce 226 127 503 240 284 197 176 72 105 452 786 308 497 349Nd 843 582 194 83 142 78 91 48 397 192 318 119 237 157Sm 138 98 338 154 336 191 194 115 762 412 585 293 529 325Eu 116 027 057 003 113 044 039 018 128 056 103 066 092 018Gd 114 66 319 156 420 223 200 107 717 513 502 249 651 324Tb 158 096 045 015 081 043 035 018 122 104 076 045 125 056Tm 061 029 024 010 055 038 016 002 062 052 029 026 071 018Yb 383 191 148 072 387 285 100 002 402 323 181 165 489 102Lu 050 024 020 011 059 042 015 001 055 039 025 024 072 013Hf 944 357 313 092 197 125 412 226 438 168 339 175 363 161Ta 127 100 077 100 074 063 020 001 156 081 145 117 130 062Au (ppb) 7 3 5 3 5 2 1 1 5 2 5 1 6 1Th 208 73 137 126 164 210 255 025 152 13 170 44 774 545U 146 147 133 055 165 135 181 091 348 141 240 169 325 072

KU 26520 14544 31918 16460 36144 47072 2370 2390 21760 8814 33527 16577 15084 2354ThU 915 408 103 897 162 201 157 064 490 161 1105 777 260 208LaTh 617 282 313 235 109 354 385 124 428 255 369 198 301 068ZrHf 461 290 353 423 423 219 334 233 593 254 739 557 363 357HfTa 164 168 131 128 502 401 213 123 410 294 625 585 328 248LaNYbN 186 862 1282 107 276 271 674 267 901 407 211 102 306 182EuEu 035 015 062 024 103 045 062 008 057 013 056 007 055 017

aMajor element data in wt trace elements in ppm except as noted all Fe as Fe2O3 The number of samples is given in parentheses Individual analyses areavailable from V Puura

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 441

Fig 9 Harker diagrams of target rocks and impact breccias from Kaumlrdla crater The contents of Fe2O3 Al2O3 and MgO are inversely correlatedwith that of SiO2 with correlation coefficients of 087 079 and 086 respectively On the contrary chemical alteration and K-metasomatismin the hydrothermal environment during and after the impact event led to highly variable patterns for CaO Na2O and K2O

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

442 V Puura et al

In Fig 10 the variation of compatible and incompatibletrace element compositions of allochthonous breccia samplesand target rocks between elements is shown and indicates aminor admixture of amphibolitic component to mostlygranitic breccias

The average trace element composition of the impactbreccias shows enrichments in the Ba Cs and U contentscompared to the target rocks (Table 4) These enrichments arein good agreement with trace element distributions found forexample at the Manson Ames or Gardnos craters (Koeberlet al 1996 1997 French et al 1997) Impact-inducedhydrothermal systems have been shown to be responsible forthe mobilization of these elements in earlier impact craterstudies (see eg Reimold et al 1994)

To further study the variations in the alkali contents weused the ternary alteration diagrams by Nesbitt and Young(1989) shown in Fig 11 In the A-CN-K (Al2O3-CaO + Na2O-K2O) diagram unaffected amphibolites and mildly tostrongly altered granites can be distinguished In this diagramK-metasomatism leads to a plot as shown in Fig 11a whileweathering would lead to compositions in the direction of thedashed arrow (see Fedo et al 1995) Two samples from thesedimentary allochthonous breccias plot outside of the rangeof these groups One sample of sandstone seems to beunaffected while a claystone-breccia is highly weatheredwhich is also visible in the A-CNK-FM diagram (Fig 11b)

Chondrite-normalized rare earth element (REE)abundance patterns of three sample classes (breccias withgranitic clasts breccias with amphibolite clasts and breccia

matrix) are shown in Fig 12 Breccia matrix includes sampleswithout clasts larger than 2 mm Averages and ranges incomposition for the different rock types are presented inTable 4 and Fig 12 Figure 12a shows the averagecomposition of the three groups of breccia samples comparedto the average compositions of the target rock types granitequartzite amphibolite and limestone Figure 12b illustratesthe ranges in abundances The breccias display largevariations in REE abundances reflecting the wide range oftarget rocks Patterns for granites quartzites and limestoneshave steep slopes and negative europium anomaliesAmphibolites show an almost flat pattern with no distinct Eu-anomaly and slight enrichments in the heavy REEs (HREE)The three types of breccia samples have REE patterns similarto the granites The HREE enrichment indicates a moderateadmixture of amphibolites

Mixing CalculationsAccording to major and trace element data (see Figs 9

10 and 12) most breccia samples result from the migmatizedgranitic basement with only minor contributions ofamphibolite or limestone For quantitative modeling we useda harmonic least-squares mixing (HMX) program(Stoumlckelmann and Reimold 1989) which allows inclusion ofmeasurement and averaging errors Separate treatment of themixing parameters (ie major and trace elements) and use ofa large number of input components (target rocks andbreccias) characterize this highly versatile program Anotheruseful feature is the so called Pivot component which allows

Fig 10 Ternary diagram of La-Th-Sc showing the poor mixing of polymict allochthonous breccias The two major types of target rocksgranites and amphibolites plot close to the La (granites) and Sc (amphibolites) corner

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 443

one to force the presence of a certain component known to bepresent from petrographic studies in the mixture

Calculations were performed on different breccia unitsfrom drill cores K1 and K18 These units comprise theallochthonous crystalline-derived polymict breccias at 471ndash523 m and 356ndash380 m depth in core K1 and 3685ndash3992 mand 3561ndash3685 m in core K18 (Figs 4 and 5) For the targetrocks (=input components) we used average compositions ofgranite quartzite amphibolite and limestone (see Table 4)Average breccia compositions were modeled for threedifferent types breccia-bearing granite clasts breccia withamphibolite clasts and fine-grained breccia matrix (Table 4)

The parameters given in Table 5a were chosen tooptimize the mixing conditions as indicated by the lowerdiscrepancy factors These factors are a measure for thequality of the mixing calculation zero means statisticallyoptimized results while results with factors of higher thanfive are of bad quality The first run included all majorelements except sodium and potassium using all componentswith totals set to 100 Run 2 was the same as run 1 butexcluded limestone as a target component according topetrographic observations (Figs 4andash4c) Runs 3 and 4 weredesigned similar to runs 1 and 2 but included Sc Co Y Yband Th These trace elements have the largest variationsbetween different breccia types but the smallest variations

within one breccia type All mixing runs were first appliedwith and without fixed totals of 100 wt to indicate anymissing components As differences between fixed and non-fixed totals were less than 1 only results of runs with fixedtotals are shown The inclusion of trace elements led to aslight increase in the discrepancy factors In spite of the highalteration of many samples discrepancy factors are fairlylow For the granitic clast bearing breccia and the averagebreccia matrix the use of trace elements enlarge the amountof incorporated amphibolite while for the breccia withamphibolite clasts a quartzite contribution could beexcluded

The calculations (Table 5a) yield the following resultsThe breccia matrix is mainly composed of graniteamphibolite and subordinate quartzite The breccia matrixcan be reproduced by an admixture of ~80 granite andquartzite with a small contribution of amphibolite materialThe amphibolite clast bearing breccia is composed of 70amphibolites mixed with granite The composition determinedin run 4 is given in Table 5b The average relative differencebetween observed and calculated abundances is 07 for allelements and lt02 for the major elements The highestdifferences between observed and calculated contents arefound for the granite clast bearing breccia and the amphiboliteclast bearing breccia shows the lowest delta values Generally

Fig 11 Ternary diagrams for target rocks and impact breccias A = Al2O3 C = CaO (in silicate fraction only corrected for phosphates andcarbonates) N = Na2O K = K2O F = FeOtotal and M = MgO (in molar proportions) CIA = chemical index of alteration (for details see Nesbittand Young [1989]) CB = claystone breccia SM = sandstone matrix a) In the A-CN-K diagram the general trends of alteration are drawn asdashed arrows The breccias are mixtures of the amhibolites and granites of the basement yet alteration causes a shift toward potassium Plagand K-Fsp are idealized plagioclase and K-feldspar compositions respectively (after Fedo et al 1995) b) in the A-CNK-FM diagram suchhigh alteration values would shift the samples toward the A corner but K-metasomatism neutralizes this effect The stippled line representsthe composition of unaltered granites Only one breccia and one target sample show the ldquonormalrdquo weathering trend along the dashed arrowsIn the A-CN-K diagram these samples plot close to the Al2O3 corner

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

444 V Puura et al

Fig 12 Rare earth element plots of basement and brecciated samples normalized to C1-chondritic composition (Taylor and McLennan 1985)comparing (a) average breccias with average target rocks and (b) the ranges of different breccia types

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 445

SiO2 Al2O3 FeO Th and Y show the smallest differenceswhile the values for MnO CaO and Sc differ highly

Extraterrestrial ComponentFor the detection of an extraterrestrial component in

impact breccias contents of the siderophile element (eg CoNi Cr and Ir) can be used which are highly enriched inchondritic and iron meteorites but have very low abundancesin the upper continental crust Due to the possible enrichmentof these elements in amphibolites the terrestrial input(indigenous component from the target rocks) has to bedetermined first This was done on the basis of the HMXmixing results The patterns of C1-normalized ranges for thesiderophile transition metals (Fig 13) show no or only minorenrichments indicative of an extraterrestrial component Inaddition the averages corrected for the indigenouscomponents are shown All averages plot below the

normalized average upper continental crust (UCC)abundances (Taylor and McLennan 1985) Therefore on thebasis of this pattern no accurate information on the amount ofany extraterrestrial material can be given

Iridium on the other hand is an element withabundances that can be measured with specific techniques(chemical separation or -coincidence spectrometry) in therange characteristic of upper crustal values (about 20 pptafter Taylor and McLennan 1985) The distribution of Irwithin the seven samples analyzed by -coincidencespectrometry points to only minor enrichment Sample K18-3791 a granite clast bearing breccia has a content of 128 pptIr that compared to a background between 28 and 62 ppt Irmay indicate a small extraterrestrial contribution Thissample was corrected for the indigenous component and thepattern is shown in Fig 13b Although iridium and goldcontents are slightly above average UCC abundances the

Table 5a Parameters and results of mixing calculations to reproduce Kaumlrdla impact breccias The originally DOS-based HMX program (Stoumlckelmann and Reimold 1989) was adapted for Windows-based multi-sample input from an Excel spreadsheet (Huber 2003) to enable extensive mixing calculations for a large number of breccia samplesa

Mixing componentsMix Run Granite Quartzite Amphibolite Limestone Discrep factor

Breccia matrix 1 866 plusmn 20 lt1 111 plusmn 17 lt07 272 866 plusmn 21 lt09 112 plusmn 18 ndash 093 740 plusmn 26 69 plusmn 20 175 plusmn 12 lt001 384 738 plusmn 30 84 plusmn 23 179 plusmn 14 ndash 15

Granite clast bearing breccia 1 684 plusmn 57 316 plusmn 55 lt07 lt08 332 684 plusmn 58 316 plusmn 57 lt07 ndash 113 574 plusmn 41 419 plusmn 40 08 plusmn 03 lt001 514 574 plusmn 45 419 plusmn 44 08 plusmn 02 ndash 18

Amphibolite clast bearing breccia 1 85 plusmn 34 114 plusmn 23 751 plusmn 09 lt13 092 85 plusmn 34 114 plusmn 16 751 plusmn 14 ndash 023 264 plusmn 14 lt001 685 plusmn 14 lt001 534 264 plusmn 15 lt001 685 plusmn 15 ndash 12

aRuns (1) total to 100 all comp all major elements without Na2O K2O LOI (2) 1 without limestone (3) 1 + trace elements (Sc Co Y Yb Th) and (4) 3without limestone

Table 5b Comparison of calculated and observed composition for the three breccia sample types based on run 4 of the mixing calculationsa

Breccia matrix Breccia with granite clasts Breccia with amphibolite clasts

Calc (Obs Calc) Calc (Obs Calc) Calc (Obs Calc)

SiO2 6874 082 7662 005 5158 004TiO2 050 021 033 005 081 003Al2O3 1231 008 1018 008 1356 023Fe2O3 478 013 271 001 937 000MnO 004 000 002 001 009 003MgO 233 028 073 004 722 085CaO 104 057 034 005 374 035Sc 114 13 432 063 289 81Co 111 04 408 137 258 22Y 411 11 317 35 374 11Yb 371 031 251 071 464 025Th 152 01 175 01 679 096

aData given in wt (major elements) and ppm (trace elements)

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

446 V Puura et al

Fig 13 C1-chondrite normalized siderophile element patterns of different breccia types In (a) the ranges for the different breccia types areshown while in (b) averages corrected for the indigenous component are presented For comparison the abundances of the upper continentalcrust (UCC) are shown as a dashed line The iridium contents are below the detection limit of INAA at 05 ppb (black arrow) All averagesiderophile element contents in the breccias corrected for the indigenous component are below the average upper continental crust abundancesSample K18-3791 showing elevated iridium abundances of 128 ppt was plotted after correction for an indigenous contribution leading to apattern below the 005 mixing line for CI-chondritic material in UCC (dotted line data from Taylor and McLennan [1985])

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 447

pattern suggests a maximum incorporation of less than 005of a C1-chondrite This trend is in good agreement withrecently presented mixing-parameters of projectile quantitiesfor impacts on wet targets (Ryder 2001) A previouslysuggested iron meteorite as projectile for the Kaumlrdla crater(Suuroja et al 2002a) is rather unlikely based on thedistribution patterns given in Fig 13 For an iron projectilevariations of the Ni Co and Ir contents of at least threeorders of magnitude would be necessary The observedpatterns are in better agreement with a very minor chondriticaddition but no clear assignment is possible due to the lowabundances observed for a confirmation of an extraterrestrialcomponent abundances and inter-elemental ratios of allplatinum group elements (see McDonald et al 2001) orisotopic studies would be necessary

K-Enrichment and Na-Ca-Depletion of Minerals and RocksThe Na-Ca-depletion and K-enrichment developed in

strongly crushed and porous allochthonous and par-allochthonous crystalline-derived breccias and to a lesserdegree in fractured sub-crater basement rocks The alterationcaused the replacement of plagioclase and minor microclineby orthoclase and of hornblende by chlorite-corrensite Inaddition some biotite was altered These processes yielded aNa-Ca-depletion and enrichment with K Besides theextensive formation of orthoclase pseudomorphshydrothermal chlorite and corrensite and carbonates(Kirsimaumle et al 2002) appear as fracture and vug fillings Insome places the latter are accompanied by minor veinlets ofK-feldspar yet formation of the orthoclase pseudomorphspredates the vein- and vug-filling minerals

In impactites the formation of K-feldspar is usuallyexplained by K-release during post-impact hydrothermalalteration of impact melt rocks (eg Grieve 1994 McCarvilleand Crossey 1996 Gurov et al 1998 Masaitis et al 1999Naumov 2002 Vishnevsky and Montanari 1999) In theKaumlrdla structure however melt rocks and even suevites werenot found so far Our preliminary results indicate aparagenetic succession of mineral associations starting withthe wide-spread formation of orthoclase pseudomorphs afterplagioclase from hot possibly vapor-dominated fluids tosubdued late microcrystalline K-feldspar dolomite andcalcite which crystallized from water-dominated fluidsystems at approximately ambient conditions Kirsimaumle et al(2002) give homogenization temperatures of 150ndash300 degC forfluid inclusions in quartz and 100ndash200 degC for the formation ofcorrensite-chlorite mineral pairs corresponding to conditionsduring the hydrothermal stage

CONCLUSIONS

The Kaumlrdla crater was formed in a layered sedimentaryand crystalline target in a shallow epicontinental sea Thecrater-fill consists of sedimentary- and crystalline-derivedallochthonous breccias with their upper parts in secondary

slump and resurge position Only the crystalline-derivedpolymict allochthonous breccias contain intensely shock-metamorphosed minerals including quartz containing PDFsindicating shock pressures of about 10ndash35 GPa Theallochthonous sedimentary-derived breccias par-autochthonous crystalline-derived breccias and sub-craterbasement rocks lack quartz with PDFs In the par-autochthonous crystalline-derived breccias and the uppermostpart of the sub-crater basement the presence of kinked biotitesuggests shock pressures between ~2 and 6ndash10 GPaCrystalline rocks are the major precursor rocks for thepolymict breccias Slight Ir enrichments indicate theincorporation of extraterrestrial material in very minoramounts but an iron meteorite is unlikely as a projectileaccording to the inter-elemental ratios of siderophileelements The K-enrichment and Na-Ca-depletion isubiquitous at Kaumlrdla in the crystalline-derived allochthonousand par-autochthonous breccias In the sub-crater basement itgradually decreases with increasing depth

AcknowledgmentsndashThis work was supported by the Ministryof Education of Estonia (Project 0551) and Estonian ScienceFoundation (Grant 4417 to V Puura) and the Austrian ldquoFondszur Foumlrderung der wissenschaftlichen Forschungrdquo projectSTART Y58-GEO (to C Koeberl) The authors are verygrateful to O Taikina-Aho and J Paaso operators at theInstitute of Electron Optics Oulu University for SEM andEMPA analyses We are grateful to the reviewers Erik Sturkelland Ray Anderson and to associate editor Alex Deutsch fordetailed and critical comments and helpful advice forimproving the manuscript

Editorial HandlingmdashDr Alex Deutsch

REFERENCES

Abels A Plado J Pesonen L J and Lehtinen M 2002 The impactcratering record of Fennoscandia A close look at the database InImpacts in Precambrian shields edited by Plado J and PesonenL J Berlin Springer Verlag pp 1ndash58

CANMET 1994 Catalogue of certified reference materialsCanadian Certified Reference Materials project 94-1E 82 p

Fedo C M Nesbitt H W and Young G M 1995 Unraveling theeffects of potassium metasomatism in sedimentary rocks andpaleosols with implications for paleoweathering conditions andprovenance Geology 23921ndash924

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

French B M Koeberl C Gilmour I and Shirey S B 1997 TheGardnos impact structure Norway Petrology and geochemistryof target rocks and impactites Geochimica et CosmochimicaActa 61873ndash904

Govindaraju K 1989 1989 compilation of working values andsample description for 272 geostandards GeostandardsNewsletters 131ndash113

Grieve R A F 1994 An impact model of the Sudbury structure In

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

448 V Puura et al

Proceedings of the Sudbury-Norilrsquosk symposium edited byLightfoot P C and Naldrett A J Special Vol 5 Toronto OntarioGeological Survey pp 119ndash132

Grieve R F Langenhorst F and Stoumlffler D 1996 Shockmetamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Gurov E P Gurova E P and Kolesov G M 1986 Impactitecomposition of the Boltysh astrobleme Meteoritika 45150ndash155In Russian

Gurov E P Koeberl C and Reimold W U 1998 Petrography andgeochemistry of target rocks and impactites from the IljinetsCrater Ukraine Meteoritics amp Planetary Science 331317ndash1333

Huber H 2003 Application of -coincidence spectrometry for thedetermination of iridium in impact-related glasses tektites andbreccias PhD thesis Institute of Geochemistry UniversityVienna Austria

IUGS 2000 International stratigraphic chart 2nd edition ParisUNESCO Division of Earth Science

Jarosewich E Clarke R S Jr and Barrows J N editors 1987 TheAllende meteorite reference sample Smithsonian Contributionsto Earth Science 271ndash49

Kirsimaumle K Suuroja S Kirs J Kaumlrki A Polikarpus M Puura Vand Suuroja K 2002 Hornblende alteration and fluid inclusionsin Kaumlrdla impact crater Estonia Evidence for impact-inducedhydrothermal activity Meteoritics amp Planetary Science 37449ndash457

Kivisilla J Niin M and Koppelmaa H 1999 Catalogue ofchemical analyses of major elements in the rocks of thecrystalline basement of Estonia Tallinn Geological Survey ofEstonia 94 p

Koeberl C 1993 Instrumental neutron activation analysis ofgeochemical and cosmochemical samples A fast and reliablemethod for small sample analysis Journal of Radioanalyticaland Nuclear Chemistry 16847ndash60

Koeberl C and Huber H 2000 Multiparameter -coincidencespectrometry for the determination of iridium in geologicalmaterial Journal of Radioanalytical and Nuclear Chemistry 244655ndash660

Koeberl C Kluger F and Kiesl W 1987 Rare earth elementdetermination at ultratrace abundance levels in geologicmaterials Journal of Radioanalytical and Nuclear Chemistry112481ndash487

Koeberl C Reimold W U Kracher A Traumlxler B Vormaier A andKoumlrner W 1996 Mineralogical petrological and geochemicalstudies of drill cores from the Manson impact structure Iowa InThe Manson impact structure Anatomy of an impact crateredited by Koeberl C and Anderson R R Special Paper 302Boulder Geological Society of America pp 145ndash219

Koeberl C Reimold W U Brandt D Dallmeyer R D and PowellR A 1997 Target rocks and breccias from the Ames impactstructure Oklahoma Petrology mineralogy geochemistry andage In Ames structure in northwest Oklahoma and similarfeatures Origin and petroleum production edited by Johnson KS and Campbell J A Circular 100 Norman OklahomaGeological Survey pp 169ndash198

Koppelmaa H Niin M and Kivisilla J 1996 About the petrographyand mineralogy of the crystalline basement rocks of the Kaumlrdlacrater area Hiiumaa Island Estonia Bulletin of the GeologicalSurvey Estonia 64ndash24

Maumlndar H Vajakas T Felsche J and Dinnebier R E 1996AXES 14 A program for the preparation of parameter inputfiles for FULLPROF Journal of Applied Crystallography 29304

Masaitis V L Naumov M V and Mashchak M S 1999 Anatomy

of the Popigai impact crater Russia In Large meteorite impactsand planetary evolution II edited by Dessler B O and SharptonV L Special Paper 339 Boulder Geological Society ofAmerica pp 1ndash17

Masaitis V L Pevzner L A Deutsch A and Ivanov B A editors2004 Deep drilling in the Puchezh-Katunki impact structureRussia Heidelberg Springer Verlag

McCarville P and Crossey L J 1996 Post-impact hydrothermalalteration of the Manson impact structure In The Manson impactstructure Anatomy of an impact crater edited by Koeberl C andAnderson R R Special Paper 302 Boulder Geological Societyof America pp 347ndash376

McDonald I Andreoli M A G Hart R J and Tredoux M 2001Platinum-group elements in the Morokweng impact structureSouth Africa Evidence for the impact of a large ordinarychondrite projectile at the Jurassic-Cretaceous boundaryGeochimica et Cosmochimica Acta 65299ndash309

Naumov M V 1999 Hydrothermal-metasomatic mineralization InDeep drilling at Puchezh-Katuniki impact crater edited byMasaitis V L and Pevzner L A St Petersburg All-RussiaGeological Research Institute pp 276ndash286 In Russian

Naumov M V 2002 Impact-generated hydrothermal systems Datafrom Popigai Kara and Puchezh-Katunki impact structures InImpacts in Precambrian shields edited by Plado J and PesonenL J Heidelberg Springer Verlag pp 117ndash172

Nesbitt H W and Young G M 1989 Formation and diagenesis ofweathering profiles Journal of Geology 97129ndash147

Plado J Pesonen L J Puura V and Suuroja K 1996 Geophysicalresearch on the Kaumlrdla impact structure Hiiumaa Island EstoniaMeteoritics amp Planetary Science 31289ndash298

Puura V and Flodeacuten T 1997 The Baltic Sea drainage basin A modelof Cenozoic morphostructure reflecting the early Precambriancrustal pattern In Proceedings of the fourth marine geologicalconference ldquoThe Balticrdquo Uppsala 1995 edited by Cato I andKlinberg F Research Papers Ca 86 Uppsala SverigesGeologiska Undersoumlkning pp 131ndash137

Puura V and Suuroja K 1992 Ordovician impact crater at KaumlrdlaHiiumaa Island Estonia Tectonophysics 216143ndash156

Puura V A Kala E A and Suuroja K A 1989 The structure of theKaumlrdla astrobleme (Hiiumaa Island NW Estonia) Meteoritika48150ndash161 In Russian

Puura V Lindstroumlm M Flodeacuten T Pipping F Motuza G LehtinenM Suuroja K and Murnieks A 1994 Structure andstratigraphy of meteorite craters in Fennoscandia and the Balticregion A first outlook Proceedings Estonian Academy ofSciences Geology 4393ndash108

Puura V Amantov A Tikhomirov S and Laitakari I 1996 Latestevents affecting the Precambrian basement Gulf of Finland andsurrounding areas In Explanation to the map of Precambrianbasement of the Gulf of Finland and surrounding area 11million edited by Koistinen T Special Paper 21 HelsinkiGeological Survey of Finland pp 115ndash215

Puura V Kaumlrki A Kirs J Kirsimaumle K Kleesment A Konsa MNiin M Plado J Suuroja K and Suuroja S 2000a Impact-induced replacement of plagioclase by K-feldspar in granitoidsand amphibolites at the Kaumlrdla crater Estonia In Impacts and theearly Earth edited by Gilmour I and Koeberl C HeidelbergSpringer Verlag pp 415ndash445

Puura V Kaumlrki A Kirs J Kirsimaumle K Konsa M and Suuroja K2000b Alteration of crater silicate rocks Examples of dissimilarenvironments from Kaumlrdla and Manson structures (abstract) InCatastrophic events and mass extinctions Impacts and beyondLPI Contribution 1053 Houston Lunar and Planetary Institutepp 172ndash173

Reimold W U Koeberl C and Bishop J 1994 Roter Kamm impact

Impactites and target rocks from the Kaumlrdla crater 449

crater Namibia Geochemistry of basement rocks and brecciasGeochimica et Cosmochimica Acta 582689ndash2710

Ryder G 2001 Mixing of target melt and projectile during impactsWet and dry target contrast (abstract) Meteoritics amp PlanetaryScience 36A180

Stoumlckelmann D and Reimold W U 1989 The HMX mixingcalculation program Mathematical Geology 21853ndash860

Suuroja K 1996 Kaumlrdla impaktkraatri teke ja areng MSc thesisInstitute of Geology University of Tartu Tartu Estonia

Suuroja K Suuroja S All T and Flodeacuten T 2002a Kaumlrdla (HiiumaaIsland Estonia) The buried and well-preserved Ordovicianmarine impact structure Deep Sea Research II Topical Studiesin Oceanography 491121ndash1144

Suuroja K Puura V Kirs J Niin M and Potildeldvere A 2002bGeological setting of the Kaumlrdla area In Soovaumllja (K-1) drillcore edited by Potildeldvere A Tallinn Geological Survey ofEstonia pp 6ndash10

Taylor J C 1991 Computer programs for standardless quantitative

analysis of minerals using the full powder diffraction profilePowder Diffraction 62ndash9

Taylor S R and McLennan S M 1985 The continental crust Itscomposition and evolution Oxford Blackwell ScientificPublications 312 p

Toumlroumlk S B and van Grieken R E 1992 X-ray spectrometryAnalytical Chemistry 64180ndash196

Toumlroumlk S B Labar J Injuk J and van Grieken R E 1996 X-rayspectrometry Analytical Chemistry 68467ndash485

Vishnevsky S and Montanari A 1999 Popigai impact structure(Arctic Siberia Russia) Geology petrology geochemistry andgeochronology of glass-bearing impactites In Large meteoriteimpacts and planetary evolution II edited by Dressler B O andSharpton V L Special Paper 339 Boulder Geological Society ofAmerica pp 19ndash60

Webby B D 1998 Steps toward a global standard for Ordovicianstratigraphy Newsletter on Stratigraphy 361ndash33

APPENDIX

LITHOLOGIES AND PETROGRAPHY OF THE IMPACT-STRATIGRAPHIC UNITS

Kaumlrdla Core Sequences K1 and K18 Description ofLithological Units

The lithologic units are classified and describedaccording to their primary mineralogical composition andstratigraphic position Thin section numbers refer to drill coresamples analyzed chemically Abbreviations of minerals areBt = biotite Cal = calcite Chl = chlorite Hbl = hornblendeHem = hematite K-Fsp = K-feldspar Mc = microcline Or =orthoclase Pl = plagioclase Qtz = quartz

Core K1 (3000ndash8152 m)

6200ndash8152 m Fractured Sub-Crater Basement Impact-affected crystalline rocks granites and migmatites stronglyprevailing over folded amphibolite bodies fractured locallycrushed andor penetrated with minor breccia dikes down tomillimeters in width The texture of the rock is massivesporadically vesicular with concentric shock cracks aroundvoids

Granitoids QtzmdashXenomorphic with strong undulatoryextinction sporadically cataclastic chipped in crushed zonesMcmdashfresh crosshatched and with fissures of brecciationcutting the twinning Plmdashmore or less replaced by earthymasses of a fine aggregate of Or In relict Pl polysynthetictwinning is partly preserved Btmdashslightly deformed andsporadically chloritized Thin sections K1-806431 K1-806432 K1-753033 K1-753034

Amphibolites HblmdashIn places deformed and partlyreplaced by Chl especially in brecciated zones Plmdashall grainsare more or less replaced by a fine aggregate of K-Fsp as inthe granites Thin sections K1-7810 K1-7734 K1-7660

Breccia dikes Veinlets of fine-grained aggregates ingranitoids Thin sections K1-806431 32 K1-753033 34

5890ndash6200 m Strongly Fractured Sub-Crater Basement Strongly fractured granites with folded minor

amphibolite bodies and widespread impact breccia dikes Granite Qtzmdashas in the lowermost unit sporadically

more severely crushed and with sharp blocky extinction (K1-5960) but without PDFs Mcmdashfresh crosshatched Mc Inbrecciated zones chipped and deformed (K1-6060) Plmdashpseudomorphs of Or with relic polysynthetic twinning (K1-59901) Btmdashdeformed locally chloritized Thin sections K1-606035 K1-606036 K1-59901 K1-59902 K1-5960 K1-5895

Amphibolite bodies are severely altered in zones ofbrecciation Hbl is mechanically fragmented and replaced byChl Plmdashlargely replaced by K-Fsp aggregates in places withhematitic earthy masses Thin sections K1-6030 K1-6018K1-5970

5228ndash5890 m Par-Autochthonous Breccia Brecciatedcrystalline rocks matrix-rich in lower part and clast-rich inupper part derived mostly from granitic and in fewer casesfrom amphibolitic rocks In places very intensively foldedBt-gneiss (K1-5887) exists Secondary calcite aggregatessporadically exist in matrix voids

Most mineral clasts are Qtz Besides sinuous inclusions-trailed micro-fissures (K1-5425A) one PDF system rarelywas established K-FspmdashClasts are fresh but deformedcrosshatched Mc without PDF Pl is almost totally replaced byearthy K-Fsp and rusty hematite aggregate in which rarerelict areas with polysynthetic twinning occur Bt shows kinkbands and is partly chloritized Few amphibolite clasts arerepresented as dark semi-opaque masses

Breccia matrix consists of the same mineral clastscrushed only to a finer grain size Thin sections K1-5887 K1-5777 K1-5535 K1-5425A K1-5425B K1-5368 K1-5231

450 V Puura et al

4710ndash5228 m Allochthonous Breccia Crystalline rock-derived matrix-supported breccia Clasts are predominantlygranite with minor amounts of amphibolite Sporadicallymatrix shows flow texture via kink-banded biotite flakes

Clasts (mainly from granitoids) Qtzmdash1ndash2 (seldom up to3 eg thin section K1-50732) systems of PDF In additionvariably oriented rows of inclusions are seen Inside Qtzthere are tiny patches with reduced refraction indexmdashapossible occurrence of high-pressure Qtz polymorph (French1998) In places clasts with 1ndash2 PDF systems have beencoalesced into one Mc clasts are formed by relatively freshbut deformed sporadically heavily kinked (K1-50752)crosshatched varieties Pl shows relict polysynthetic twinningbut is mainly replaced by submicroscopic Or aggregatecontaining voids with ripped edges Sporadically up to 2 PDFsystems is seen in Mc (K1-49471) Bt is kink-bandedsporadically containing patches of opaque or rusty masses(K1-50731) Calmdashdispersed aggregates and clasts in matrix(K1-5038A)

Amphibolite clasts are replaced by chloritic masses as inpar-autochthonous breccia

Matrix consists of the same clastic material crushed to afiner grain size Thin sections K1-51621 K1-51622 K1-51591 K1-51592 K1-50751 K1-50752 K1-50731 K1-50732 K1-5032B K1-49471 K1-49472 K1-49451 K1-49452

3800ndash4710 m Allochthonous Breccia Sedimentary rock-derived breccia clast-supported Blocks and minor clasts areCambrian sandstones and locally claystones crystallineclasts are rare (see -ray log Fig 4)

3560ndash3800 m Allochthonous Breccia Crystalline rock-derived breccia predominately matrix-supported with agranitic block in the middle Sporadically fragments ofclaystone are seen Microscopic appearance ie features andintensity of mineral deformation are exactly the same as in theunit 4710ndash5228 m

Qtzmdashin clasts 1 to 3 systems of PDFs are seen Sinuousrows and zones of fluid inclusions are distributed in variousdirections Often clasts with different directions of PDFs arewelded into one McmdashDeformed patchy crosshatchedtwinning is characteristic Earthy and opaque inclusions arespreading partly along cleavage fractures partly alongsinuous crevasse-like fissures partly in the form of irregularmasses In places 1 PDF system is seen diagonal in relationto crosshatching Pl is replaced by earthy masses of Or Btshows kink-bands is partly decomposed and chloritized Caloccurs as dispersed aggregate in breccia groundmass andfissure-fillings cutting clasts Thin sections K1-3750A K1-37371 K1-37372 K1-37351 K1-37352

3000ndash3560 m Allochthonous Breccia Predominatelysedimentary rock-derived clast-supported breccia composed

of blocks of Cambrian claystones with minor sandstones andblocks of Lower to Middle Ordovician limestone Crystallinerocks clasts are rare in lower but more abundant in the upperpart of the unit Thin sections K1-34051 K1-34052 K1-34053 K1-34054 K1-34055

3000ndash~340 m Allochthonous Breccia Locally abundantclasts of crystalline-derived breccias

Mineral associations and clast population together withdeformational features are exactly the same as those in theunit of 5228ndash5890 m A grain of Qtz aggregate with PDFsin a crystalline breccia clast was found at a depth of 324 mIn Qtz of larger lithic clasts no PDFs were found Calcite inthe form of veins and pore fillings is more abundant than inunits 5228-5890 and 3560-3800 m Thin sections K1-3255A K1-3255B K1-3240A K1-3240B K1-3240 K1-32401ndash5

Core K18

4000ndash4314 m Sub-Crater Basement (Central Uplift) Strongly fractured or brecciated crystalline rocks

(migmatized biotite gneisses microcline granites with foldedamphibolite bodies) with impact breccia dikes Mineralcomposition and deformational structures are the same as inthe lowermost units of core K1 Rare carbonate veins andaggregates are seen in amphibolites

In amphibolites Pl is replaced by rusty masses ofsubmicroscopic Or Voids with irregular edges are visibleHbl is replaced by pseudomorphs of Chl Thin section K18-4000

In granitoids the fragmented Qtz grains andcrosshatched Mc show mechanical deformation Despite thelarge number of thin sections PDFs of PFs in Qtz and Mcwere not found only submicroscopic inclusions fillingvarious micro-fractures Bt is slightly deformed partlydecomposed and replaced with Chl In places kink-bandsand alignment of Bt flakes around clasts is visible Thinsections K18-40207 K18-40208 K18-40209 K18-402010 K18-40101 K18-40102 K18-40103 K18-40104K18-40105 K18-40106 K18-400510 K18-400512-21K18-4003

3560ndash4000 m Allochthonous Breccia This is the mostthoroughly studied impact breccia unit in Kaumlrdla Themineral-textural fabric of this crystalline rock-derivedmatrix-supported fragmental breccia (K18-3830) follows

Minerals and rock clasts predominantly derived fromgranites subordinately from amphibolites composed of Chland Or with minor carbonate aggregates Matrixmdashconsists ofwelded microscopic-sub-microscopic clasts with massive inplaces fluidal structure partly with voids

Qtz clasts mostly 01ndash1 mm in size very often showheavily patchy extinction with 2ndash3 systems of PDFs and

Impactites and target rocks from the Kaumlrdla crater 451

sinuous lines of fluid and opaque inclusions K-Fsp existseither as fresh crosshatched Mc (from basement granitoids)with refined banded twin lamellae (with rare laddertextures) or as turbid sub-microscopically aggregates withrusty-pigmented Or patches replacing Mc In largeaggregates clasts of Mc showing mosaicism 1ndash2 systems ofPDFs and partly decorated with iron oxide earthy massesand beads were observed Pl is replaced by sub-microscopicOr aggregates with some relict polysynthetic twinning Btflakes show kink- bands partly chloritized partly filled withhematitic ribbons in several directions Thin sections K18-3997 K18-3984 K18-39751 K18-39752 K18-39641K18-39642 K18-3960 K18-3960A K18-3960B K18-396028 - 30 K18-393025A K18-393026 K18-393027K18-393029 K18-39041 K18-39042 K18-3900A K18-3900B K18-3888 K18-3830 K18-38131 K18-38132K18-38121 K18-38122 K18-38051 K18-38052 K18-3802A K18-3802B K18-38022 K18-3650 K18-3620K18-3610

2927ndash3560 m Allochthonous Breccia Predominantlysedimentary rock-derived breccia composed of Cambriansandstone and claystone blocks and clasts with minoradmixture of crystalline-derived clasts

Crystalline-derived breccia mainly composed of graniticrocks dominates in the interval of 300ndash314 m or is presentedas an admixture in the sedimentary breccia in the lower (314ndash340 m) and upper part (2927ndash300 m) of the core Thesedimentary-rock-derived breccias are composed ofCambrian sandstones claystones and Ordovician limestonesCarbonate mineralization is abundant in the uppermost part ofthis unit Thin sections K18-3010A K18-3010B

Crater Rim

Core Sections S8 F175 F241 F374 F376 F379 F380 F382 413 415

The crystalline core of the rim wall at Palukuumlla andTubala is made of fractured and locally brecciated granitesgranite-migmatites and amphibolites with impact brecciadikes and pockets Locally in these cores rich veins and porefillings of calcite occur In the groundmass of the rock thereplacement of primary Mc Pl and Hbl with above-describedsecondary aggregatic pseudomorphs is seen No PDFs wereobserved in Qtz or other minerals Thin sections F241-1673F241-1330 F241-1063 F241-1053 F241-1004 F241-0967F241-0795 F241-0574 F241-0546

Core P11mdashdepth interval 17ndash28 m allochthonous breccialens (pocket) on top of the crystalline core of the Palukuumlla rimCrystalline-derived breccia is composed of granitic andamphibolitic rocks with fine clastic matrix and also withcarbonate cement Calcite veins cement brecciated granitoidsin the uppermost part of crystalline rock sequence In theclastic breccia matrix partial relicts of Mc Pl Bt and Hbl arevisible No PDFs were found in quartz or other minerals Thinsections P11-0178AndashC P11-0264AndashE P11-0280AndashD

Crystalline Basement Rocks from Outside the Crater

Samples were collected from drill cores located within amaximum radius of 25 km around the crater Koppelmaa et al(1996) made the petrographic characterization Thin sectionsF361-3408 F361-2947 F361-2915 F361-2885 F364-3310F364-3231 F364-3170 F364-3032 F364-3004 F364-2985F364-2915

Impactites and target rocks from the Kaumlrdla crater 451

sinuous lines of fluid and opaque inclusions K-Fsp existseither as fresh crosshatched Mc (from basement granitoids)with refined banded twin lamellae (with rare laddertextures) or as turbid sub-microscopically aggregates withrusty-pigmented Or patches replacing Mc In largeaggregates clasts of Mc showing mosaicism 1ndash2 systems ofPDFs and partly decorated with iron oxide earthy massesand beads were observed Pl is replaced by sub-microscopicOr aggregates with some relict polysynthetic twinning Btflakes show kink- bands partly chloritized partly filled withhematitic ribbons in several directions Thin sections K18-3997 K18-3984 K18-39751 K18-39752 K18-39641K18-39642 K18-3960 K18-3960A K18-3960B K18-396028 - 30 K18-393025A K18-393026 K18-393027K18-393029 K18-39041 K18-39042 K18-3900A K18-3900B K18-3888 K18-3830 K18-38131 K18-38132K18-38121 K18-38122 K18-38051 K18-38052 K18-3802A K18-3802B K18-38022 K18-3650 K18-3620K18-3610

2927ndash3560 m Allochthonous Breccia Predominantlysedimentary rock-derived breccia composed of Cambriansandstone and claystone blocks and clasts with minoradmixture of crystalline-derived clasts

Crystalline-derived breccia mainly composed of graniticrocks dominates in the interval of 300ndash314 m or is presentedas an admixture in the sedimentary breccia in the lower (314ndash340 m) and upper part (2927ndash300 m) of the core Thesedimentary-rock-derived breccias are composed ofCambrian sandstones claystones and Ordovician limestonesCarbonate mineralization is abundant in the uppermost part ofthis unit Thin sections K18-3010A K18-3010B

Crater Rim

Core Sections S8 F175 F241 F374 F376 F379 F380 F382 413 415

The crystalline core of the rim wall at Palukuumlla andTubala is made of fractured and locally brecciated granitesgranite-migmatites and amphibolites with impact brecciadikes and pockets Locally in these cores rich veins and porefillings of calcite occur In the groundmass of the rock thereplacement of primary Mc Pl and Hbl with above-describedsecondary aggregatic pseudomorphs is seen No PDFs wereobserved in Qtz or other minerals Thin sections F241-1673F241-1330 F241-1063 F241-1053 F241-1004 F241-0967F241-0795 F241-0574 F241-0546

Core P11mdashdepth interval 17ndash28 m allochthonous breccialens (pocket) on top of the crystalline core of the Palukuumlla rimCrystalline-derived breccia is composed of granitic andamphibolitic rocks with fine clastic matrix and also withcarbonate cement Calcite veins cement brecciated granitoidsin the uppermost part of crystalline rock sequence In theclastic breccia matrix partial relicts of Mc Pl Bt and Hbl arevisible No PDFs were found in quartz or other minerals Thinsections P11-0178AndashC P11-0264AndashE P11-0280AndashD

Crystalline Basement Rocks from Outside the Crater

Samples were collected from drill cores located within amaximum radius of 25 km around the crater Koppelmaa et al(1996) made the petrographic characterization Thin sectionsF361-3408 F361-2947 F361-2915 F361-2885 F364-3310F364-3231 F364-3170 F364-3032 F364-3004 F364-2985F364-2915