JOURNAL OF RESEARCH of the National Bureau of Standards —A. Physics and Chemistry

Vol. 70A. No. 2, March-April 1966

Infrared Spectra of the Hydrated Borates

C. E. Weir

Institute for Materials Research, National Bureau of Standards, Washington, D.C.

(November 4, 1965)

Infrared absorption spectra of 42 different hydrated borates were obtained in the 2000-300 cm"1

range. A few spectra were obtained between 4000 and 2000 cm"1. Most spectra are complex andcannot be interpreted satisfactorily except in the case of the simplest anions. Many correlationsbetween spectra are possible, however, and possible anion types have been deduced. Differentiationbetween triangular and tetrahedral boron is possible on the basis of the spectra but is less certain thanin the case of the anhydrous borates.

Key Words: Hydrated borates, infrared spectra, absorption spectra, borates, boron coordination,borate anions

1. Introduction

A comprehensive survey of the infrared spectra ofthe crystalline anhydrous inorganic borates was re-ported recently [I].1 The spectra were correlatedwith known structures and predictions of unknownstructures were made on the basis of the observedspectra. Two of the predicted anion structures havebeen verified by subsequent x-ray crystallographicstudies [2, 3]. The obvious extension of the studiesto include the hydrated borates is the subject of thisreport.

The hydrated borates have been subjected to a moresystematic study than the anhydrous materials, largelyas a consequence of their mineralogical occurrence.Miller and Wilkins [4] reported data on eight hydratedborates obtained by the mull technique. Severalyears later, Takeuchi [5] studied the spectra of 17borates of which 12 were hydrated. Takeuchi alsoused the mull technique of sample preparation.Moenke and his coworkers, in a series of papers, [6-13]have presented spectra obtained on approximately17 hydrated borates using the KBr pellet technique.Akhamanova [14] has reported on seven hydratedborates and Plyusnina and Kharitonov [15] havestudied six, with spectra in both cases being obtainedusing films deposited from isopropyl alcohol on alkalihalide plates. The early work by Miller and Wilkins[4] and that of Takeuchi [5] had a lower frequencylimit of about 600 cm"1 which is too high to obtain someof the known fundamental vibrations of the borateanion. All the more recent studies, however, extendas far as 400 cm"1.

1 Figures in brackets indicate the literature references at the end of this paper.

In the present study, infrared spectra were obtainedfor 42 different hydrated borates. Most specimenswere natural minerals but in many instances syntheticmaterials were also available. Approximately 63different specimens were studied so that it was pos-sible to determine the reproducibility of the spectrafor materials of different origin —a question of impor-tance for natural minerals which may contain infraredactive impurities.

2. Experimental Method and Apparatus

Experimental techniques and equipment were sim-ilar to those described previously for the anhydrousborates [1]. A double-beam grating spectrometerwas used to cover the range 4000-200 cm"1. Thisrange was covered in two steps, 4000-400 cm"1 and2000—200 cm"1. In each case the lower frequencylimit imposed by the available energy was higher thanthe limits given. Most spectra were obtained onlyin the range 2000-300 cm"1 because the fundamentalsof the borate anions were known to occur in this range.For these studies powdered films deposited on CsBrplates from a volatile liquid were used. A few spec-tra in the higher frequency range were obtained tolocate the OH fundamental stretching modes. Forthese studies the spectra were obtained using a mullwith perfluorokerosene as the suspending medium.In both cases a blank cell was placed in the referencebeam of the spectrometer. The possibility of dehy-drating the specimens precluded using dry air in theinstrument and considerable care was exercised inthe 500-300 cm"1, region. In this range only thosebands were considered to originate in the specimenwhich were not obtained on a comparison spectrumrun with no specimen in the beam.

153

TABLE 1. Description of samples studied 3. Samples

Material

HambergiteBandyliteTeepleitePinnoite

Inyoite

Meyerhofferite

Colemanite

Inderite

Kurnakovite

InderboriteHydroboracite

Synthetic

Tunellite

Nobleite

Gowerite

Veatchite

Synthetic

Synthetic

Synthetic

Synthetic

Ginorite

Strontian GinoriteSynthetic

Priceite

Pandermite

Tertschite

Preobrazhenskite

Ulexite

Probertite

KaliboriteParahilgarditeKerniteTincalconiteSynthetic

SyntheticSynthetic

LarderelliteAmmonioboriteEzcurriteSyntheticSyntheticSyntheticSzaibelyiteSussexiteSynthetic

Empirical formula

4BeO • B2O3 • H2OCuO • CuCl2 • B2O3 • 4H2ONa2O • 2NaCl • B2O3 • 4H2OMgO • B2O3 • 3H2O

2CaO • 3B2O3 • 13H2O

2CaO • 3B2O3 • 7H2O

2CaO • 3B2O.} • 5H2O

2MgO • 3B2O3 • 15H2O

2MgO • 3B2O3 • 15H2O

M g O C a O - 3 B 2 O 3 l l H 2 OMgO • CaO • 3B2O3 • 6H2O

2SrO • 3B2O3 • 5H2O

SrO • 3B2O3 • 4H2O

CaO • 3B2O3 • 4H2O

CaO • 3B2O3 • 5H2O

SrO • 3B2O3 • 2H2O

CaO • 3B2O3 • 2H2O

MgO-3B2O3-7iH2O

MgO • 3B2O3 • 5H2O

SrO • 3B2O3 • 4H2O?

2CaO • 7B2O.{ • 8H2O

2SrO-7B2O3-8H2O2SrO • 7B2O3 • 8H2O

4CaO • 5B2O:! • 7H2O

4CaO • 5B2O3 • 7H2O

4CaO • 5B2O3 • 20H2O[?]

3MgO-5B2O3-4iH2O[?]

Na2O • 2CaO • 5B2O3 • 16H2O

Na2O • 2CaO • 5B2O3 • 10H2O

K2O-4MgOllB2O3-9H2O6CaO • 2CaCl2 • 9B2O3 • 4H2ONa2O • 2B2O3 • 4H2ONa2O • 2B2O3 • 5H2ONa2O • 2B2O3 • 10H2O

K2O • 2B2O3 • 4H2ONa2O • 5B2O3 • 10H2O

NH4 • 5B2O3 • 4H2ONH4 • 5B2O3 • 5VaH2O2Na2O • 5B2O3 • 7H2O2Na2O • 5B2O3 • 7H2O2Na2O • 5B2O3 • 5H2O2Na2O • 5B2O, • 4H2O2MgO • B2O3 • H2O2MnO • B2O3 • H2ONaBO3 • 4H2O

Origin of specimen a

C 4744103459.102798.C 4488; synthetic

Mgo • B2O3 • 3H2O (R. C. Erd).114260; Pacific Coast Borax

open pit, Boron, Calif.96072; Gower Gulch, Death

Valley Nat. Monument.96445; Pacific Coast Borax

open pit, Boron, Calif.11462; Pacific Coast Borax

open pit, Boron, Calif.R 11023; Pacific Coast

Borax open pit, Boron, Calif.Inder region, USSR.Inder region, USSR (R. C.

Erd via H. B. Mason).Synthetic strontium cole-

manite (R. C. Erd via C. R.Parkerson).

Pacific Coast Borax openpit, Boron, Calif.; synthetictunellite (R. C. Erd).

115278; N. of DeBely mine,Death Valley Nat. Monument;synthetic nobleite (R. C. Erd).

115277; Hard Scramble Claim,Death Valley Nat. Monument;synthetic gowerite (R. C. Erd).

117214; synthetic veatchite(C. R. Parkerson via R. C. Erd).

Synthetic calcium veatchite(C. R. Parkerson via R. C. Erd).

Synthetic macallisterite(R. C. Erd).

Synthetic aksaite (H. A.Lehmann via R. C. Erd).

Synthetic (C. R. Parkersonvia R. C. Erd), possibly a di-morph of tunellite but degree ofhydration uncertain.

112966; synthetic ginorite(R. C. Erd).

Natural mineral via R. C. Erd.Synthetic volkovite [?]

(R. C. Erd).Death Valley Nat. Monu-

ment (J. F. McAllister viaR. C. Erd).

Panderma, Asia Minor (W. T.Schaller via R. C. Erd).

Kurtpinari mine, Bigadicdistrict, Turkey (H. Meixner viaR. C. Erd).

Inder region, USSR (M. E.Mrose via R. C. Erd).

96994; Pacific Coast Boraxopen pit, Boron, Calif.; Mudd-mine, Boron, Calif.

96077; Mudd mine, Boron,Calif.; Harmony Borax Works,Death Valley Nat. Monument.

R5844.105822.95737; Mudd mine, Boron, Cal.107397.Synthetic borax, reagent

grade.Reagent grade.Synthetic (V. Morgan via

R. C. Erd).C 4481.R 6167.Tincalayu, Argentina.N. P. Nies via R. C. Erd.N. P. Nies via R. C. Erd.C. Cipriani via R. C. Erd.95111.C 4456.Sodium perborate.

a Number of specimens studied is equal to the number of listings for each materialseparated by semicolons. Numerals represent catalog numbers in the mineral collectionof the Smithsonian Institution.

The samples studied are listed in table 1. Column 1of this table lists the mineral or chemical name,column 2, the empirical formula, and column 3, thesource of the material. Throughout this paper, sam-ples will be referred to by the mineral names and syn-thetic samples will be designated by their compositionunless it is known that the synthetic and naturalminerals are equivalent. In the latter case the mineralname will be used. Samples were obtained generallyfrom two sources, the mineralogical collection of theSmithsonian Institution and the U.S. GeologicalSurvey. In column 3, designations denoted by num-bers correspond to catalog numbers in the Smith-sonian collection. Origins of specimens obtainedfrom the U.S. Geological Survey are listed in column 3.The total number of samples of each mineral typestudied is indicated by the number of listings incolumn 3 separated by semicolons. The order inwhich the samples are listed in table 1 corresponds tothe order followed in the discussion.

4. Results and Discussion

4.1. Presentation of Data

Data will be presented here as in the previousreport [1]. Compounds are grouped according tothe anion type expected based on the ideas presentedby Christ [16]. For each compound, observed bandsare tabulated with the type of band designated by theusual abbreviations.2 Typical spectra are shown onlyfor a few of the materials and only for the range 1800cm"1 to approximately 300 cm"1.

4.2. Borates With Simple Anions

In order to assess the extent to which analyses ofthe spectra of the hydrated borates can be pursued,it is instructive to consider first the spectra of ham-bergite, bandylite, teepleite, and pinnoite. Thesematerials have known structures with relatively simpleanions. Hambergite contains unhydrated planar BO73

groups [17, 18] and the formula Be2(0H)B0.3 repre-sents the structure. Both bandylite and teepleitecontain isolated B(OH)^1 groups and their structuresmay be represented by the formulas CuB(OH)4Cl andNa2B(OH)4Cl respectively [19, 20, 21]. Pinnoite con-tains anions of the form [B(OH)3OB(OH)3]"

2 which canbe considered as resulting from condensing twoB(OH)^1 groups with the elimination of H2O [22]. It isof interest to compare the spectra of these substanceswith those obtained in previous studies of the anhy-drous borates [1] where it was found that boron in

2 s —strong, v —very, b — broad, m —medium, w —weak, sh —shoulder.

154

three-fold coordination could be differentiated by thestrong absorption derived from v3 of the isolated ionswhich generally persisted in the complex anion. Itwas found that triangular boron even in complexanions exhibited broad strong absorption in the region1100-1300 cm"1 and tetrahedral boron in the region800-1100 cm"1. The v2 and v4 region (600-800 cm"1)of the isolated ions also was found useful for differ-entiating these anion types provided isotope substi-tution was possible (see ref. 1; figs. 1, 2, 19, and p.486).

The validity of these conclusions with respect tothe hydrated borates can be evaluated by consideringthe spectra of the hydrates with simple anions shownin figure 1 and the absorption bands listed in table 2.In hambergite the strong band at 1300 cm"1 can beascribed to v3 and the strong broad band centered at750 cm"1 probably arises from Be —O vibrations.The 600 and 650 cm"1 bands superimposed on theBe —O band are probably the two components of V4,and vz is probably obscured by the Be — O band. The1155 cm"1 band is unexpected. It does not seemstrong enough for a v3 band and it is tentatively con-sidered to be a bending mode involving BeOH.

In teepleite and bandylite the v3 bands are ascribedto the strong bands between 900 and 950 cm"1. Thesomewhat weaker band at 845 cm"1 in teepleite mayarise from a second component of v3 or from V\. Inboth materials the strong bands at lower frequenciesmay be due to v± or to an OH torsional mode. Inteepleite the bands at 1170 and 1302 cm"1 are quitestrong and might be erroneously considered to be v3fundamentals of triangular borate ions. In thisspectrum they are weaker than the v3 for the tetra-hedral borate group and there is no danger of misin-terpretation. These bands may represent eithercombinations or B — OH bending modes. In bandylitethe, 1255 cm"1 band is no weaker than the v3 bandand misinterpretation is possible. From spectralstudies on orthoboric acid, however, it seems that thisband most probably arises from an OH distortionvibration [23].

In the pinnoite spectrum the strong broad bandsin the 800-1000 cm"1 range are the most prominentfeatures of the spectrum and presumably are derivedfrom the antisymmetrical vibration, v3, of the tetra-hedral units. The complex series of bands below800 cm"1 are probably derived from the original v2and V4 vibrations. The region above 1100 cm"1

contains three broad strong bands which are some-what weaker than the antisymmetrical bands andare probably the result of B —OH distortion modes.These bands might easily be assumed to arise fromtriangular boron in a material with an unknown anionand show that interpretations must be made with greatcaution.

On the basis of these spectra it must be concludedthat hydration of the borate anions produces spectrawhich cannot be interpreted readily from a knowledgeof the spectra of the corresponding anhydrous anion.Hydration appears to produce absorption bands, someof which are very strong, throughout the spectra range

1800 1600 1400 1200 1000 800 600 400

FIGURE 1. Infrared absorption spectra of hydrated borates contain-ing simple anions.

TABLE 2. Infrared absorption spectra of borates with simple anions

flambergite

1450 ssh1300 vsvb

1155 s1025 w1012 w

825 sb792 s770 s740 s715 s645 m600 m448 m440 m

Pinnoite

3560 s3400 sb3140 svb2930 vb1295 sb1215 mb1150 sb1050 svb945 svb895 msh860 sb795 sb

610 wsh585 mb505 w485 m435 mb400 m335 m310 m

Teepleite

3530 s3400-2500 sb1450 mvb1302 sb1170 s935 vsb845 vs670 sh595 svb

522 w430 w

Bandylite

3480 s3440 sbsh2400 sb1255 svb1080 mvb1025 mvb905 svb640 w587 w

500 svb460 svb370 m300 w

studied. Bands apparently derived from v3 modes ofthe anhydrous BO ~̂3 and BO^5 groups still appear inthe spectra of the hydrated anions. These bands arevery strong and occur in the expected positions, i.e.,800-1100 cm"1 for tetrahedral boron and above 1100cm"1 for triangular boron. However, strong bandsapparently originating from B — OH distortion modescomplicate the spectra, particularly in the region1100-1400 cm"1.

Data on teepleite have been obtained by Plyusninaand Kharitonov [15]. In gross features the spectraagree reasonably well but there are significant dif-ferences. For example, they do not report the 1450or 430 cm"1 bands at all and band positions vary upto 21 cm"1 for the 845 cm"1 band which they found

155

at 866 cm"1. The spectrum of pinnoite has beenreported by Takeuchi [5] with considerable disagree-ment in the data. Comparison of the various spectrashows that there are considerable discrepancies innumbers of bands, band locations, and band contours.

5. Divalent Metal Borates

5.1. The MeO'3B2O3'jrH2O Compounds

The family of divalent metal borates of generalformula 2MeO-3B2O3*#H2O comprise the mineralsinyoite, meyerhofferite, colemanite, fabianite, inderite,kurnakovite, inderborite, and hydroboracite. Ex-cept for fabianite (2-3-1) all of these minerals, as wellas a synthetic strontium (2-3-5) analog of colemanitewere examined during the present study. The infra-red spectrum for fabianite has been reported by Kuhnand Moenke [13]. Among the other 2 • 3 -x compoundscrystal structures have been determined for the fol-lowing: Inyoite [25], meyerhofferite [26], colemanite[27], inderite [28, 29], and hydroboracite [30]. Theseborates all contain similar anions built up from six-membered rings formed from alternate boron and oxy-gen atoms by corner sharing among two BO4 tetra-hedra and one BO3 triangle. The oxygen atoms notshared by two boron atoms attach a hydrogen to formhydroxyl groups [16]. Isolated anions, [B3O3 • (OH)5]~

2

are found in inyoite, meyerhofferite and inderite. Incolemanite and hydroboracite, similar anions poly-merize into chains having the formula [B3O4(OH)3]^

2n.Typical spectra for some of these materials are

shown in figure 2 and the observed absorption bandsare listed in table 3. As seen in figure 2, the spectraare very complex and more than a cursory interpreta-tion is not possible. The spectra of inyoite and meyer-

INDERITE

hofferite are very similar with the predominantfeatures being a strong complex absorption band near1400 cm"1 and a similar band below 1200 cm"1. Theformer band probably is the counterpart of the anti-symmetric mode for the triangular boron atoms andthe latter band for the tetrahedral boron groups.

TABLE 3. Infrared absorption spectra of 2MeO • 3B2O3 • XH2Ocompounds

Inyoite

3530 s3440 s3200 vsvb

1640 mb1470 msh1420 s1385 s1330 s

1210 msh1165 s1105 sb1060 s1005 s955 sb885 sb865 sb800 s

705 sb675 msh610 w575 msh550 w540 m480 msh468s365 m330 w

Kurnakovite

3550 m3450 s3200 vsvb

1650 mvb1450 msh1400 sb1390 s1347 sb1257 s1215 s1160 m1080 msh1045 msh1010 sb987 sb860 svb

807 s735 m710 m650 mb

520 mb

460s

Inderite

1680 mb1640 mb1442 msh

1390 sb1340 sb1275 s1250 msh1170 mb1135 mb1040 s

980 svb885 msh850 sb790 m737 m

650 mvb

560 m

485 m450 m427 s325 mb

Inderborite

1660 mvb1460 msh1410 sb1390 msh1340 sb1290 msh1200 sb1150 wsh1065 s1020 msh

960 sb

860 sb810 m755 m715 mvb650 mbsh

530 mvb

435 m

TABLE 3. Infrared absorption spectra of 2MeO • 3B2O3 • XH2Ocompounds — Continued

1800 1600 1400 1200 1000 800 600 400

FIGURE 2. Infrared absorption spectra of 2ivfeO • 3B2O3 • XH2Ocompounds.

Colemanite

3590 s3510 b

3130 svb1455 msh1360 sb1320 sb1275 sb1225 s1160 mb1125 mvb1042 s

930 svb875 svb810 msh790 msh755 s730 s670 w580 m550 m515 m490 w

415 sb

Sr Colemanite

1450 msh1350 sb1300 sb1260 sb1190 mb1150 mb1060 msh1030 mb

925 b865 sb

790 msh750 s730 m710 msh575 m545 w510 w490 w

435 w

330 sb

Hydroboracite

1365 sb1300 s1280 s1188 s1130 mb1080 sb1050 mb985 mbsh950 sb882 s835 s800 msh760 s

580 m550 w530 w505 w470 m435 w425 w410 w

310 sb

Meyerhofferite

3600 s3470 s3420 sb3000 sb3020 svb1445 msh1400 sb1360 sb

1155 sb1085 sb1020 sb960 sb940 sb890 mb840 mb790 sb720 s675 m615 w580 m

527 m494 w468s427 m

390 w362 m330 s

156

Aside from the gross similarity it is not possible tocorrelate the spectra band for band because thereappear to be noticeable shifts of bands between thetwo spectra. These shifts are probably to be at-tributed to differences in hydrogen bonding largelyproduced by water molecules in the two structures,the formula of inyoite having 4H2O whereas that ofmeyerhofferite has only one H2O (table 1). Thenoticeable effects throughout the spectral range pro-duced by this water are the main deterrent to a moredetailed analysis of these spectra. The effect ofpolymerization on the isolated anions to form the chainsin colemanite and hydroboracite can be determinedby comparing their spectra with those of inyoite andmeyerhofferite. The bands in colemanite and hydro-boracite are broader and apparently more complex.In addition, there is a considerable apparent shiftingof bands as might be expected from loss of hydrogenbonding water and coupling of the anions. The spec-trum of colemanite gives a clear indication that thestretching vibrations involving the triangular boronatoms are represented by the band complex near 1300cm"1. In the spectra of inyoite and meyerhofferitethe strong bands, just below 1200 cm"1 might have beenconsidered part of the antisymmetric stretch of thetriangular borons. The fact that these bands becomemuch weaker in colemanite which contains fewerB —OH bonds implies that these bands may arisefrom B —OH bending modes which might be expectedin this frequency range [23]. It should be noted,however, that the spectrum of fabianite reported byKiihn and Moenke [13] appears to contradict thepresent conclusions, because it exhibits a strong bandnear 1142 cm"1.

Inderite and kurnakovite are polymorphs and theirspectra correspond very closely as shown by thetabulated bands. Inderborite, a mixed calcium-magnesium compound, also shows a spectrum similarto both the magnesium and the calcium compounds.The gross features of the spectra of the variousmagnesium and calcium compounds are also similar,but there are many differences in band positions andintensities.

Data on all of these materials, except the syntheticcompound, have been reported by one or more of theearlier workers [5, 7, 8, 9, 14, 15]. In place of com-paring the previous data for each material, it is usefulto consider reported results on colemanite which hasbeen studied by all these investigators. The fre-quency range considered is from 650 to 1350 cm"1

since this was the range used by Takeuchi [5]. Ingeneral appearance, of all five spectra for colemaniteare very similar, but the similarity holds only for thegross features. Between 650 and 1350 cm"1 Take-uchi [5] and Moenke [7] report twelve absorptionbands; Plyusnina and Kharitonov [15], fourteen;Akhamanova [14], nine; and the present work, thir-teen. Thus the most probable number of bands wouldseem to be between twelve and fourteen. However,the totals are obtained quite differently by the severalworkers. Consider, for example, the 1042 cm"1 band

of figure 2 which appears to be distinct and well sep-arated from neighboring bands. Akhamanova [14]reports a band at 1050 cm"1, Plyusnina and Kharitonov[15] a strong band at 1048 cm"1 plus two additionalbands at 1002 and 1062 cm"1, Moenke [7] a strongband at 1046 cm"1 and Takeuchi's [5] spectra indicatea band at about 1030 cm"1. In addition to theseuncertainties in band numbers and locations thereare considerable differences in relative intensitiesand band contours for many of the bands throughoutthe spectrum. The same considerations apply tocomparisons between the present and previous dataon the other members of the 2:3:x borates. Possibleexplanations and implications of these facts will bediscussed later.

5.2. The MeO 3B2O3vXH2O Compounds

Compounds studied with the empirical formulaMeO • 3B2O3 *^H2O are tunellite, nobleite, gowerite,synthetic aksaite, veatchite, synthetic calcium-veatchite [32], synthetic macallisterite [33], and asynthetic 1:3:4? strontium borate. Of these materials,the structure is known only for nobleite and tunellite[34] which are isomorphous. The anion consists ofthree six-membered rings forming a [B6O9(OH)2]~

2

unit. The six-membered rings are the same as thosefound in the 2MeO *3B2O3-^H2O compounds and eachcontains two tetrahedral and one triangular boron.The [B6O9(OH)2]"

2 ion contains three triangular andthree tetrahedral boron groups and is unique in thatone oxygen atom is common to all three rings. Theions are linked in sheets through certain of the off-ring oxygen atoms. The structures of the anions inthe other members of this series are not yet known,but it might be expected to be similar, probably linkedwith differing degrees of hydration and/or polymeriza-tion [16].

Data obtained on these materials are given in table 4and typical spectra are shown in figure 3. It wouldbe expected that these spectra should show somesimilarities with those of figure 2, since the anionsare built of similar units and indeed there are manysimilarities in band positions in the two sets of spec-tra. However, it was surmised that the coupling be-tween rings in tunellite would be strong enough toproduce considerable broadening and band overlap.The reverse situation actually occurs, many bandsbeing clearly resolved and relatively sharp. Fromthe similarities of the spectra (see data of table 4),it is clear that nobleite, tunellite, and the synthetic1:3:4? synthetic strontium borate have similar anions.Examination of figure 3, shows that it is probable thatgowerite has the same anion on the basis of the sim-ilarity of the spectra. In gowerite the bands aresharper and appear to be resolved better, a result tobe expected if depolymerization of the anion occurswith increased water content. It also appears thatveatchite possesses the same anion as tunellite withband broadening and loss of resolution accompanying

157

GOWERITE

TABLE 4. Infrared absorption spectra of MeO • 3B2O3 • XH2Ocompounds — Continued

1800 1600 1400 1200 1000 800 600 400

FIGURE 3. Infrared absorption spectra of MeO • 3B2O3 * XH2Ocompounds.

the dehydration and polymerization. From the tabu-lar data (table 4), it is clear that the spectra of veatchiteand synthetic calcium-veatchite are very similar andthe materials probably contain the same anion. Theconclusions with respect to the two magnesium com-pounds, synthetic aksaite, and synthetic macallisterite,are made with much less certainty; but again itseems probable that these materials also may containthe same anion with differing degrees of hydrationand polymeriztion. On the basis of the infraredspectra alone, therefore, it is concluded that all theMeO -3B2O3 *^H2O compounds probably containsimilar anions. These spectra do not appear to havebeen reported previously.

TABLE 4. Infrared absorption spectra of MeO • 3B2O3 * XH2Ocompounds

Veatchite

3420-2900 b1655 w1510 ssh1360 sb1270 sb1235 msh1210 wsh1160 m1095 mb1065 mb975 sb930 sb880 m830 w815 m765 m760 m700 m675 msh640 w625 m605 m570 w520 w510 w490 w470 m455 m430 msh395 mb

Synthetic Ca-Veatchite

1665 w

1390 sb1265 wsh1240 sb

1180 msh1130 m1075 mb975 sb910 sb870 m832 m810 m755 m745 wsh700 w

650 w610 m600 wsh572 wsh535 wb

495 w

TABLE 4. Infrared absorption spectra of MeO • 3B2O3 • XH2Ocompounds — Continued

Nobleite

3540 mb3380 svb3200 svb2400 svb1675 mvb

1385 sbsh1322 sb1300 msh1180 mb1160 mb1125 mb1107 m1033 msh965 sb930 msh880 sb827 m805 sb737 m705 m670 mb602 m570 w

460 wb425 w

Tunellite

1675 m b1440 msh1355 sb1315 sb

1177 mb1155 mb

1100 w1035 msh1000 sb965 msh880 sb827 msh810 s732 m700 w677 m598 m577 m528 w

420 m

SyntheticSrO • 3B2O3 • 4H2O?

1460 mb1375 svb1325 svb

1180 sb

1125 sb1100 w1010 sb962 svb927 mb875 sb

810 sb735 m

670 mb

Gowerite

1635 wb1480 msh1405 s1350 s1330 msh1250 s

1112 s

1048 ssh1010 s

967 msh951 s905 mb860 mb830 m759 m700 w675 w656 m630 w602 w567 w495 w475 w452 w410 w

Synthetic Aksaite

3600 s3450-2400 sb1652 mb1485 msh1425 sb1385 sb1325 sb1245 sb1210 sb1155 mb1095 sb

1020 sb

950 sb898 mb850 mb

800 sb725 w575 mb655 w

525 w455 w

410 w

SyntheticMacAUisterite

1470 msh1410 sb1358 sb

1240 s

1158 mb1122 sb1060 sb1005 msh980 m965 m880 msh

855 s812 s

698 wb670 m

580 wb

440 w

5.3. The 2MeO • 7B2O3 * 8H2O Compounds

Three members of this series were studied, thedata are given in table 5 and spectra are shown infigure 4. The marked similarity of the spectra andthe compounds makes it almost certain that the anionsare identical. The structures of the anions are notknown and the spectra are too complex to attemptto predict the anion on this basis alone. There is nodoubt that the anion contains both three-fold and four-fold coordinated boron as shown by the strong bands

158

1800 1600 1400 1200 1000 800 600 400

FIGURE 4. Infrared absorption spectra of 2MeO • 7B2O3 • 8H2Ocompounds.

TABLE 5. Infrared absorption spectracompounds

of 2MeO-7B2O3-8H2O

Ginorite

3250 vsvb1650 wvb1485 msh1435 msh1370 s1325 s1245 s

1180 s1160 s1035 m985 sb945 m880 s855 msh815 msh805 s

740 s675 sb605 m

Strontian Ginome

1660 mvb1480 mbsh1440 mbsh1360 svb1300 sb1250 m1230 mb1175 svb1125 mvb1045 m1010 sb975 svb940 sb870 s

800 sb

735 s655 sb605 m

435 m375 mvb

2SrO-7B2O3;8H2O- Synthetic

1650 wvb

1445 mbsh1365 s1320 s

1220 mb1185 s1125 m1040 s1005 sb965 m945 msh885 s835 m810 s780 msh735 s690 m675 m660 m615 w535 w

1800 1600 1400 1200 1000 800 600 400

FIGURE 5. Infrared absorption spectra of miscellaneous hydrateddivalent metal borates.

TABLE 6. Infrared absorption spectra of miscellaneous divalentmetal borate compounds

Priceite

1650 mvb1420 m1370 s1350 s1305 s1285 msh1190 msh1135 msh1060 sb1010 sb945 ssh885 svb

780 sb710 m655 mb

570 m515 m470 m

435 w

385 m355 w325 m

Tertsrhite

1650 mb1445 m1365 s1335 ssh

1265 w1225 m

1040 sb1015 msh965 sb915 sb875 msh825 m715 m

520 m

Preobrazhenskite

1540 mvb

1370 sb1320 s

1215 sb1125 msh1075 sb1000 svb940 mb900 m

860 mvb810 m790 m735 w700 sb665 mb615 m600 w575 w550 mb435 mvb375 m340 m325 m

near 1400 and 1000 cm"1 respectively. In additionthe spectra show a strong resemblance to the previousspectra for compounds containing six-memberedrings of alternate boron and oxygen atoms, with onetriangular and two tetrahedral boron atoms. Conse-quently, it appears likely that these rings may existin the anions. The relative weakness of lower fre-quency vibrations (i.e., below 600 cm"1) in thesespectra, however, is an indication that the anions arenot coupled six-membered rings and it is very unlikelythat the anions are so small that no low frequencyvibrations exist. It is probable that the symmetry ofthe anions is sufficiently high that lower frequencymodes are much weaker than in the previous com-pounds. Again there do not appear to be any previousreports of spectra of these materials.

5.4. Miscellaneous Divalent Metal Borates

Data on the 4CaO-5B2O3-ZH2O compounds aregiven in table 6 together with those for preobrazhens-

kite -3MgO-5B2(V 4iH2O[?] [35]. The spectra areshown in figure 5. The three materials show evidenceof both triangular and tetrahedral boron in the strongbands in the 1300-1400 cm"1 and 800-1000 cm"1 re-gion respectively. The spectra of priceite and tert-schite are very similar and these materials may containsimilar anions. The spectrum of preobrazhenskiteis indicative of a large anion with considerable inter-action between the anion structural units.

From the data obtained here the spectra of pander-mite and priceite are considered to be identical. Thesample of priceite yielded much the better spectrumof the two specimens, that for pandermite being lesswell resolved. However, for each band found in thespectrum of pandermite a corresponding band isobserved in the priceite with frequencies differing atmost by ± 1 cm"1. In addition, the same relativeintensity relationships are shown by the bands in thetwo materials. It seems possible to conclude thatthese materials are most probably identical [36].

159

The spectrum of priceite (pandermite) has beenreported by other workers [5, 7, 14, 15] and that oftertschite by Meixner and Moenke [12]. Comparisonof the present and previous results leads to the sameconclusions mentioned earlier. The spectra for agiven material are in superficial agreement. Thereare many discrepancies when details are examined.

Data obtained on other divalent metal borates aregiven in table 7 with spectra for ulexite and kaliboriteshown in figure 6. The anions of probertite and ulex-ite are known to be a ring structure containing threetetrahedra and two triangles with one tetrahedronbeing common to both rings [37, 38]. From the tabu-lated data it is clear that the infrared spectra cor-roborate the similarity of the anions in these materialsprobertite and ulexite. The spectrum of kaliboriteis very poorly resolved in the v3 region for tetrahedralborate groups, but is otherwise similar to the spectraof ulexite. The parahilgardite yielded a very poorspectrum and it is suspected that this material wascontaminated with CaCO3 as the 875 and 725 cm"1

bands strongly resemble the Vi and v± bands of calcite.

6. Monovalent Metal Borates

6.1. The Me>O 2B2O3 XH2O Compounds

Several borates of the type Me2O • 2B2O3 • ZH2Owere studied and the data are tabulated in table 8.Typical spectra for tincalconite and potassium tetra-borate tetrahydrate are shown in figure 7. Morimoto[39] has determined the structure of borax and foundthat the anion consists of isolated [B4O5(OH)4]"

2 rings.Marezio et al., [40] have found the same anion in potas-sium tetraborate tetrahydrate. The rings containtwo triangular and two tetrahedral borate groups joinedat common oxygen atoms. The two tetrahedral groupsare further linked by means of an oxygen bridgeacross the ring. All off-ring oxygen atoms are hy-drated in borax.

As noted in the previous study the alkali boratesyield very poorly resolved spectra under the conditionsused here. From the tabulated absorption bands itis noted that all the materials show similar spectraand probably all contain the basic [B4O5(OH)4]~

2 anion

TABLE 7. Infrared absorption spectra of miscellaneous boratecompounds

Ulexite

3562 s3400 sb3000 vsvb1650 mb1620 mb1470 m1415 sb1390 s1355 s1315 s1205 s1090 ssh1050 sb995 s975 s950 ssh925 msh855 s825 ssh740 m710 m640 m530 m

430 mb350 m

Probertite

3620 s3550 s3350 vsvb1650 mvb

1470 m

1425 msh1375 s1325 sb1210 m1130 ssh1080 sb1040 m985 msh955 s930 s900 s830 s750 m675 m555 w515 m

440m380 m

Kaliborite

1410 msh1350 sb

1185 svb

1080 sb1000-750 svb

705 s655 s625570 w445 vsb

Parahilgardite

1425 vsvb1350-1000 vsvb

875 s (CcCO3?)790 s775725 s (CaCO3?)

355 vbvs

TABLE 8.

Kernite

1700 mvb1460 m1425 msh1340 vsvb

1160 m

1075 mb1010 svb

950 vsvb865 svb820 svb

685 mvb

550 s495 mb455 mb425 mvb

Infrared absorption spectra ofcompounds

Tincalconite

3500 sb3340 vsb3000 svb2400 svb1630 mvb1455 msh1405 svb1340 vsvb1190 msh1155 sb1125 svb1075 svb1025 svb985 sb940 sb

805 sb750 mvb

665 mvb

600 mvb520 mvb425 mvb

Borax

3500-3000 vsvb

2000 svb1620 mvb1450 msh1390 svb1330 vsvb1275 mvb1155 mb1123 mvb1070 mb

980 sb935 sb

800 svb750 svb

665 mvb

595 wb505 mshvb420 sb395 s

Me2O • 2B2O3 • XH2O

K2O • 2B2O3 • 4H 2Osynthetic

1625 mvb

1395 svb1320 vsvb

1220 s1160 m1080 sb1060 msh970 sb905 sb

805 svb745 w710 m665 mb600560 mvb490 m450 mb422 mvb390 m

1800 1600 1400 1200 1000 800 600 400 1800 1600 1400 1200 1000 800 600 400

FIGURE 6. Infrared absorption spectra ofhydrated borates contain- FIGURE 7. Infrared absorption spectra of Me2O • 2B2O3 • XH2Oing both mono- and divalent metal ions. compounds.

160

in some state of hydration or polymerization. Thespecimen of K2B4O7 • 4H2O used was obtained from acommercially available reagent labeled K2B4O7 * 5H2O.The x-ray diffraction pattern of this compound can beindexed readily from the unit cell data of Marezio etal., J40] and it has been tabulated as the tetrahydrate.

The previous data obtained on this group of materialstogether with the present data exhibit a confusingpattern. The spectra for borax of Miller and Wilkins[4] and of Moenke [7] agree with each other but differconsiderably from that of Akhamanova [14] and thepresent results. Miller and Wilkins [4] reported thespectrum of K2B4O7*5H2O which closely resemblestheir spectrum of borax but differs considerably fromthe spectrum of K2B4C>7 • 4H2O obtained here. Take-uchi's [5] results for kernite agree well with the presentdata but both disagree in many respects with theresults of Moenke [7]. The overall lack of agreementbetween the spectra obtained by different workersappears to be greatest for these materials. Thisquestion will be discussed later.

6.2. The Me2O*5B2O3'XH2O Compounds

Data for the minerals ammonium larderellite andammonioborite [41] and the synthetic 1:5:10 sodiumborate are given in table 9. Representative spectraare not shown because there is strong absorption andpoor resolution throughout the entire spectral range.The strong absorption is to be expected in the twoammonium compounds because they contain theinfrared active NHJ ion in addition to water of hydra-tion and a complex borate anion. Christ [16] haspredicted that sborgite, larderellite, and ammonio-borite have the same anion and the data of table 9confirm this prediction for the ammonium compoundswith little question. The spectrum of the 1:5:10 syn-thetic sodium borate differs from those of the am-monium borates in many details but not in grossfeatures. Because of the unknown effect of theNHJ ion on the spectra, it is possible that all threecompounds contains similar anions.

TABLE 9. Infrared absorption spectra of Me2O • 5B2O3 • XH2Ocompounds

Na2O-5B2O3: 10H2Osynthetic

1670 mb

1375 vsvb1300 vsvb1195 m1145 svb1075 sb1045 m1010 sb950 m925 s910 s780 s765 s675 vsvb470 mb440 mb375 mvb

Ammonioborite

1650 mvb1425 vsvb1350 vsvb1235 vsvb1190 sb1160 mb1085 sb1055 sb1015 svb940 svb915 sb880 w820 mvb775 sb740 m675 sb545 s510 mb460 mb

Larderellite

1425 vsvb1350 vsvb1225 mvb1185 s1165 m1085 mvb1055 mb1015 mvb940 mb920 mvb880 w810 mb775 s

675 mvb545 s455 m425 mvb

2Na2O-5B2O3

7H20, ISYNTHETIC ^

1800 1600 1400 1200 . 1000 800 600 400

FIGURE 8. Infrared absorption spectra of 2Na2O • 5B2O3 • 7H2Ocompounds.

6.3. The 2Me2O • 5B2O3 • XH >O Compounds

Ezcurrite and the 2 : 5 : X synthetic sodium boratesshow rather poor spectra characterized by strongabsorption throughout the spectral range studied.The absorption bands observed are listed in table 10and figure 8 shows the spectra for a synthetic and anatural ezcurrite. It is obvious from figure 8 thatdespite the rather poor resolution that these materialsare almost certainly the same insofar as the anionsare concerned. In addition, the fact that, band forband the frequencies are the same within experimentalerror may be interpreted to mean that the materialsare chemically and structurally identical. This con-clusion is based on previous studies which showedthat change of anion or crystal structure generallyproduces shifting of some bands in the infrared spec-trum [42]. The close similarity in the spectra of allthree minerals listed in table 10 indicates that theanions are similar.

TABLE 10. Infrared absorption spectra of 2Me2O 5B2O3 XH2Ocompounds

Ezcurrite

1675 mvb1470 mb1425 m1400 mbsh1370 vsb1240 sb1195 msh1150 m1075 mb1020 mb950 sb915 m900 w865 m825 m760 m695 m670 s

550 w455 w410 mb

2Na2O-5B2O3-5H2OSynthetic

1625 mvb1465 mb

1400 mb1350 vsvb1245 svb

1045 mvb1015 svb945 svb

870 m825 mvb740 mvb695 mb670 mvb630 w550 mvb455 mvb410 mvb

2Na2O-5B2O3-4H2OSynthetic

1650 mvb1440 mvb

1360 svb1325 svb1250 mvb1230 svb1150 m1070 sb1030 svb925 sb

880 svb830 mvb740 mvb695 mvb650 mb600 mvb550 m440 mb

6.4. Miscellaneous Borates

Data for three borates not readily classified with theprevious groups are given in table 11 and the spectraare shown in figure 9. The data for szaibelyite and

161

1800 1600 1400 1200 1000 800 600 400

FIGURE 9. Infrared absorption spectra of miscellaneous hydratedborates.

TABLE 11. Infrared absorption spectra of miscellaneous boratecompounds

NaBO:! • 4H2O sodiumperborate

3590 s3500-3000 svb1640 wvb1225 msh1165 mvb1025 mb975 sb915 svb845 sb805 m775 mb550 svsb495 sb415 sb

Szaibelyite

3555 s sharp1425 vs1270 s1245 s1230 s1085 msh1010 s977 s922 m835 m

785 w700 vs625 s570 m538 m486 m432 m395 s330 m310 m

Sussexite

1450 sb1385 s1270 s1250 s1225 m995 m962 m907 m882 mb825 w696 s670 msh623 s540 m518 m450 m410 m370 sb

sussexite agree reasonably well with the results ob-tained by Takeuchi [5] for camsellite (szaibelyite) andsussexite and with the results of Plyusnina and Khari-tonov [15] for ascharite (szaibelyite). From hisspectra, Takeuchi [5] concluded that the anion in thesecompounds was the pyroborate ion while Plyusnina andKharitonov, apparently being unaware of Takeuchi'searlier work, noted that the spectrum of ascharitewas similar to that of LaBC>3 which contains a tri-angular borate anion. However, the latter workersconcluded tentatively that the complexity of thespectrum indicated a more complex ion than thesimple triangular ion [15]. Akhamanova independ-ently arrived at the same conclusion [14].

The structure of this anion has been reported to bea singly hydrated pyroborate group, with some un-certainty as to the exact structure [43] caused by trans-lation. It is not clear that both borons are coordinated

with three oxygen atoms. However, the spectrumclearly indicates that there is no tetrahedral boronpresent in the structure. From the earlier work onthe pyroborates [1] it appears that the spectrum ofanhydrous magnesium pyroborate becomes consid-erably less complex in the i>2 region on hydration butmore complicated in the vx region. Tentative assign-ments are possible for many of the bands in the spec-trum of szaibelyite on the basis of the spectrum ofanhydrous magnesium pyroborate and the knownspectrum of triangular borate groups. These areobvious and need not be specified here.

In sodium perborate the anion has been reported[44] to consist of two partly hydrated tetrahedral borategroups connected through peroxide linkages and to

have the structureOH X . 0 — O v .OH

R RRX O—O

RN OH

The

structural formula of sodium perborate may be writtenNa2[BO2(OH)2]2 • 6H2O. This anion is unlike anypreviously encountered here and its spectrum is like-wise unique as shown in figure 9. In this spectrumthe 3590 cm"1 band indicates some nonhydrogenbonded OH and the broad strong band between 3500and 3000 cm"1 shows that much of the OH is hydrogenbonded. The weak 1640 cm"1 band is attributed tothe bending mode of H —O —H, and the medium broadbands at 1225 and 1165 cm"1 to the B —OH bendingmodes plus overtones and combinations. The strongabsorptions at 915 and 845 cm"1 are derived from thevs tetrahedral modes and the doublet is indicativeof tetrahedral coupling as noted earlier. The strongbroad band at low frequencies probably contains modesderived from v2 and v4 of tetrahedral units plus OHtorsional modes. The complexity of the band indi-cates the presence of several overlapping bands.The spectrum for this material has been reported byMiller and Wilkins [4] with reasonably good agreementbetween the data.

7. Reliability of Infrared Spectra of HydratedBorates

In comparing the spectra of different workers it hasbeen noted that the agreement was not satisfactory.It is to be expected that on a given material, the sameinfrared bands having similar relative intensities shouldbe obtained. Some latitude in the exact frequenciesreported (i.e., a few cm"1) and perhaps in the detectionof weak bands might be expected. For the hydratedborates it is clear that data do not agree to this extentand it is desirable to examine possible reasons for thediscrepancies.

The discordance in the data must be traced to one ormore of three sources; first, the materials themselves;second, the preparation techniques used to preparethe materials for study; and third, the actual measure-ments. Of these three sources the third can probablybe dismissed as a real source of error. There is no

162

evidence that any appreciable errors in band positionsexist between the results obtained on different spec-trometers. In fact the evidence is that the spectrome-ters agree within experimental error because manystrong bands can be identified in spectra of variousworkers which agree within the frequency limits ex-pected. It is believed, therefore, that the real causesfor the discrepancies are to be attributed to variationin the materials or differences arising from techniquesof preparation.

Insofar as the identity of the materials investigatedis concerned, the previous reports give little informa-tion. Miller and Wilkins used chemical reagents intheir studies [4]. Takeuchi [5] reports that his sub-stances were carefully selected and were well-devel-oped crystals of highest purity. Plyusnina andKharitonov [15] used single crystals, but Akhamanova[14] and Moenke [7] give no details of the method usedfor selection of their specimens. In the present studymost of the materials were of the highest purity pres-ently available as judged by microscopic examinationand x-ray powder diffraction. It must be noted, how-ever, that neither these methods nor that of selectingnatural single crystals ensures that the specimens arefree of contaminants. If the contaminants should con-sist of other borates (all of which are strongly active inthe infrared) even in small quantity, it is clear that somedifferences in the spectra may result. In addition tothe question of purity, there remains a problem associ-ated with the degree of crystallinity of the specimen.Most of the borates yield complex spectra containingmany broad, obviously complex bands, containingunresolved structural details. Experience in thislaboratory has indicated that two specimens of identi-cal composition but differing either in degree ofcrystallinity or perhaps in perfection of crystallites canyield spectra which are very similar in gross appear-ance but remarkably different in detail. Broad feature-less bands frequently show many resolvable detailswhen the degree of crystallinity of the specimen isapparently improved. It appears most probable thatdiscrepancies in the spectra of various workers consist-ing of differences in numbers of bands reported is to beattributed either to actual contamination of the speci-mens by other borates (or other spectrally active chemi-cal species) or to differences in the crystallinity of thespecimens.

Insofar as the specimen preparation techniques areconcerned, the problem of purity is also involved butin a different manner. Here the concern is with theactual identity of the materials studied because ofchanges in degree of hydration or of interaction withthe alkali halide windows used. The various workersused the following techniques of sample study: Millerand Wilkins and Takeuchi — mulls with hydrocarbon orfluorocarbon liquids using salt (NaCl?) plates [4, 5];Moenke —pressed pellet technique using vacuumformation of KBr pellets [7]; Akhamanova —filmdeposited on KBr plates by evaporation of isobutylalcohol dispersing medium [14]; Plyusnina andKharitonov — films deposited on KC1 or NaCl platesby evaporation of isopropyl alcohol [15]; present

work —films deposited on CsBr plates by evaporationof CCI4. It is well known that solids (and in particularhydrates) interact with the alkali halides and that theinteraction produces changes in the spectra. Of themethods used, it would appear that the mull techniqueis the most preferable from the viewpoint of lesseningsuch interactions, the film method next, and the KBrpellet the most objectionable. In addition to the in-creased probability of interaction in KBr pellets mustbe added the undesirable procedure of subjecting thehydrated borates to vacuum in the pellet formingprocess. Reduction in pressure might affect the stateof hydration of the small quantity of hydrate present.It seems very probable that differences in positions ofbands observed by different workers may be causedby shifts due to interaction with the alkali halidewindows used.

In addition to the question of interaction with thealkali halide there is some uncertainty introducedby the mechanical and thermal effects of the finegrinding required to produce particles sufficientlysmall for infrared study. As noted earlier, the spectrafor borax, a chemically pure reagent, obtained bydifferent workers differ radically. C. J. Bowser [45]has obtained evidence showing that in KBr pellets,borax produces the spectrum of tincalconite, a lowerhydrate. Thus the spectra for borax of Miller andWilkins [4] and Moenke [7] which agree very wellactually appear to represent tincalconite. Dehydra-tion of borax can be understood in the KBr pelletingprocess but is hard to visualize in the mull technique,unless it occurs during grinding. Therefore, it isnecessary to include the fine grinding technique amongthe details which might produce differences in thespectra.

It is concluded that the infrared spectra of thehydrated borates, at present, are not of acceptablereliability. The lack of reproducibility is most prob-ably to be attributed to differences in chemicalcomposition and degree of crystallinity and to varyinginteractions between the hydrated borates and thealkali halide windows. Production of definitivespectra will require systematic study of the effect ofthese disturbances.

8. Identification of Borate Minerals byInfrared Spectroscopy

H. Moenke has studied the infrared spectra of sev-eral hydrated borate minerals in a series of papersgenerally oriented toward differentiating among variousborates [6--13]. As a result of his study he concludedthat the spectral method could be used to identifyhydrated borate minerals and has pointed out identify-ing features in the spectra of various minerals.

It is clear from the immediately preceding discussionand the comparisons between spectra made throughoutthis report that the present conclusions must disagreewith the idea of identification by infrared methods.The fact that there is a disturbing amount of disagree-

163799-883 O-66—4

merit between the spectra of various workers at thepresent time would appear to invalidate the use ofspectroscopy as a method of identification. Moreover,the spectra of most borates are so complex that itseems imperative to obtain at least a degree of under-standing of the origin of the spectral bands beforeidentification of minerals of this type is attempted onthe basis of their spectra. In the case of the simplerspectra, i.e., sussexite, hambergite, etc., identificationby means of infrared spectroscopy is believed to bepossible. However, if such minerals were contami-nated with other borates (as for example sussexite con-taminated with szaibelyite) identification would bealmost impossible.

9. Conclusions

The infrared spectra of the hydrated borates areconsiderably more complex than the spectra of theanhydrous analogues. The increased complexityarises from the effects of the hydration of the borateanions and hydrogen bonding between anions. Theeffect of the hydrogen bonding appears to producenoticeable changes in most of the borate anion funda-mentals and results in complicating the spectrumthroughout the frequency range 400-4000 cm"1 studiedhere. Thus the effects are by no means restricted tothe region commonly considered to be diagnostic forOH.

As a result of the increased number of bands thespectra are frequently poorly resolved and detailedinterpretation appears to be remote. It is unlikely thatany hydrated borate anion can be identified solelyfrom its infrared spectrum. However, the previousconclusions concerning identification of three-fold andfour-fold coordinated boron appear to be equally validin the hydrated and anhydrous borates. For the hy-flrated compounds the conclusions are more uncertainbecause strong bands near 1200 cm"1 apparently aris-ing from B —OH bending modes fall into the regionpreviously considered as indicative of trigonal borategroups. Any serious attempt to understand the spec-tra of the hydrated borates will probably requiredeuteration as well as boron isotope substitution.

At the present time agreement between spectraobtained by different investigators is not considered tobe satisfactory. The probable causes of the disagree-ment are believed to originate in differences in thematerials themselves or from varying degrees of inter-action with the alkali halide cells used. As a resultof the lack of agreement between various workers aswell as the inherent complexity of the spectra it is con-sidered that borate minerals cannot be identified at thistime on the basis of their infrared spectra alone.

furnished the samples which made this study possible.The very helpful comments of J. R. Clark and R. C. Erdof the U.S. Geological Survey on the structure andproperties of the borate minerals were invaluable.

10. References

[1] C. E. Weir and R. A. Schroeder, J. Res. NBS 68A (Phys. andChem.) No. 465 (1964).

[2] A. Perloff and S. Block, submitted to Acta Cryst.[3] A. Mighell, A. Perloff, and S. Block, submitted to Acta Cryst.[4] F. A. Miller and C. H. Wilkins, Anal. Chem. 2 4 , 1253 (1952).[5] Y. Takeuchi, Min. J. (Japan) 2 , 245 (1958).[6] H. Moenke, Jenaer Jahrbuch 1959 (II) 361.[7] H. Moenke, Jenaer Jahrbuch 1960 (I) 191.[8] H. Moenke, Jenaer Jahrbuch 1961 (I) 239.[9] H. Moenke, Naturwiss. 4 9 , 7 (1962).

[10] H. Moenke, Spectralanalyse von Mineralien und Gesteinen(Geest and Portig, Leipzig, 1962).

[11[12;[13;[14

H. Moenke, Fortschr. Min. 4 0 , 76 (1962).H. Meixner and H. Moenke, Kali u. Steinsalz 3 , 228 (1961).R. Kiihn and H. Moenke, Kali u. Steinsalz 12 , 399 (1963).M. V. Akhamanova, Zh. Strukt. Khim. 3,24 (1962).

The author is indebted to E. R. Lippincott and R. A.Schroeder for many discussions concerning the infra-red data and to G. Switzer of the Smithsonian Institu-tion and R. C. Erd of the U.S. Geological Survey who

[15] I. I. Plyusnina and Yu A. Kharitonov, Zh. Strukt. Khim. 4 , 506(1963).

[16] C. L. Christ, Am. Min. 4 5 , 334 (1960).[17] W. H. Zachariasen, Z. Krist. 7 6 , 289 (1931).[18] W. H. Zachariasen, H. A. Plettinger, and M. Marezio, Acta

Cryst. 16,1144(1963).[19] M. Fornaseri, Period. Min. (Rome) 18 , 103 (1949); 19 , 159

(1950).[20] R. L. Collin, Acta Cryst. 4 , 204 (1951).[21] V. Ross and J. O. Edwards, Am. Min. 44 ,875 (1959).[22] F. Paton and S. G. McDonald, Acta Cryst. 10 , 653 (1957).[23] D. F. Hornig and R. C. Plumb, J. Chem. Phys. 2 6 , 637 (1957).[24] C. R. Parkerson, Naval Ordnance Laboratory, Corona, Rept.

585 (1963b).[25] J. R. Clark, Acta Cryst. 12,162 (1959).[26] C. L. Christ and J. R. Clark, Z. Krist. 114 , 21 (1960).[27] C. L. Christ, J. R. Clark, and H. T. Evans, Acta Cryst. 1 1 , 761

(1958).[28] A. Ashirov, I. M. Rumanova, and N. V. Belov, Dokl. Akad. Nauk

SSR 143,331 (1962).[29] I. M. Rumanova and A. Ashirov, Kristallografiya, 8 , 517 (1963).[30] I. M. Rumanova and A. Ashirov, Kristallografiya, 8 , 828 (1963).[31] J. R. Clark, D. E. Appleman, and C. L. Christ, J. Inorg. Nucl.

Chem. 2 6 , 73 (1964).[32] C. R. Parkerson, Naval Ordnance Laboratory, Corona, Rept.

583 (1963a).[33] W. T. Schaller, A. C. Vlisidis, and M. E. Mrose, Am. Min. 5 0 ,

629 (1965).[34] J. R. Clark, Am. Min. 49,1549 (1964).[35] V. V. Kondrat'eva, Zap. Vses. Min. Obshch. 8 8 , 330 (1959).[36] See Dana's System of Mineralogy, C. Palache, H. Berman, and

C. Frondel, Vol. II, p. 343 (John Wiley & Sons, Inc., New York,N.Y. 1951).

[37] H. M. Kurbanov, I. M. Rumanova, and N. V. Belov, Dokl. Akad.Nauk. SSR 152,1100 (1963).

[38] J. R. Clark and D. E. Appleman, Sci. 145,1295 (1964).[39] N. Morimoto, Min. J. (Japan) 2 , 1 (1956).[40] M. Marezio, A. A. Plettinger, and W. H. Zachariasen, Acta

Cryst. 16,975(1963).[41] J. R. Clark, Am. Min. 4 5 , 1087 (1960); J. R. Clark and C. L.

Christ, Am. Min. 44,1150 (1959).[42] C. E. Weir and E. R. Lippincott, J. Res. NBS 65A (Phys. and

Chem.) No. 3,173 (1961).[43] C. P'eng, C. Y. Su, and P. H. Chang, Sci. Sinica (Peking) 12 ,

1761 (1963).[44] A. Hansson, Acta. Chem. Scan. 15 , 934 (1961).[45] C. J. Bowser, submitted to Am. Min.

(Paper 70A2-391)

164