Identification and Quantification of Arsenic Species in Gold Mine Wastes Using Synchrotron-Based X-ray Techniques

Andrea L. Foster, PhDU.S. Geological Survey GMEG

Menlo Park, CA

Arsenic is an element of concern in mined gold deposits around the world

Don Pedro Harvard/Jamestown

Kelly/Rand (Au/Ag)

Ruth Mine (Cu)

Spenceville (Cu-Au-Ag)Lava Cap (Nevada)Empire Mine (Nevada) low-sulfide, qtz Au

Argonaut Mine (Au)Ketza River (Au)sulfide and oxide ore bodies

Goldenville, Caribou, and Montague (Au)

The common arsenic-rich particles in hard-rock gold mines have long been known

Arsenopyrite FeAsS

Primary Secondary

Scorodite FeAsO4 2H2OKankite : FeAsO4•3.5H2O

Jarosite KFe3(SO4)2(OH)6Tooleite [Fe6(AsO3)4(SO4)(OH)4•4H2OPharmacosiderite KFe4(AsO4)3(OH)4•6–7H2O

Secondary/Tertiary

Iron oxyhydroxide (“rust”) containing arsenic up to 20 wt%

arseniden = 1-3

n-

Arseniosiderite Ca2Fe3(AsO4)3O2·3H2OYukonite Ca7Fe12(AsO4)10(OH)20•15H2O

“arsenian” pyriteAs-1 pyrite Fe(As,S)2

Reich and Becker (2006): maximum of 6% As-1

But it is still difficult to predict with an acceptable degree of uncertainty which forms will be present

• thermodynamic data lacking or unreliable for many important phases

• kinetic barriers to equilibrium

• changing geochemical conditions (tailings management)

Langmuir et al. (2006) GCA v70

Lava Cap Mine Superfund Site, Nevada Cty, CA

Typical exposure pathways at arsenic-contaminated sites are linked to particles and their dissolution in aqueous fluids

ingestion of arsenic-bearing water

solid-phase arsenic near-neutral, low dissolved organic carbon, low salinity waters

inhalation or ingestion of arsenic-rich particles

solid-phase arseniclung and gastic/intestinal tract fluids (saline, aqueous solutions-high dissolved organic carbon, enzymes, bile, some low pH, most near-neutral)

• critical to know the form(s) of arsenic in the solid phase • only a subset of those forms may be reactive

dissolution

dissolution

This talk will review the results of synchrotron X-ray studies of arsenic speciation, with focus on gold mine wastes

Brilliant, high flux radiation produced at points tangent toa circular orbit of electrons moving at near-relativistic speeds

75 synchrotrons around the world10 in US (4 suitable for environmental work)

Stanford

Berkeley

Large ( 1-10 mm) or small (2 -150 μm) X-ray beams are available for most techniques

Bulk

Microbeam Precision stage(micron)

CCD detector(diffraction)

Solid-state detector

Gas ionization detector

• X-ray Fluorescence• X-ray Absorption Spectroscopy• X-ray DiffractionX-ray Photoelectron Spectroscopy• Vibrational Spectroscopies• Magnetic spectroscopy• Small Angle ScatteringCCD Detector

Synchrotron X-ray Fluorescence (XRF) Spectrometry

Bulk Microbeam

One EDS spectrumPer point (> 1000)

Elemental IDElement CorrelationSpatial distribution

As

Ca

Fe

1800 microns

Elemental ID

One average Spectrum

CaFe As

Voltage (energy)

Synchrotron-based X-ray Diffraction (SXRD)2-D pattern

goethite/lepidocrocitemixture

Synchrotron XRD20 min

1-D patterns

Conventional XRD8 hours

• Mineral ID• Mineral Mapping• Amorphous materials (scattering)

X-ray Absorption Fine Structure Spectroscopy

X-ray Absorption Near Edge Structure (XANES)Oxidation state, species “fingerprinting”

Abso

rban

ce

Energy (eV)

> 1 mg/kg

EXAF

S fu

nctio

n χ(

k)

Photoelectron Wave vector k (Å-1)

Extended X-ray Absorption Fine Structure (EXAFS)Speciation = identity, #, distance of neighboring atoms

> 75 mg/kg

Spectral Deconvolution by Least-Squares Fits

Orpiment (As3+-S)

As5+ on rust (iron oxyhydroxide)

As3+ in water (As3+-O)

13.5 % As5+

86.5% As3+

Bangladesh Aquifer Sediment10-10.4 meters depth

As3+

As5+

Unique spectral shape

Energy increases with oxidation state

Arsenopyrite (As1-)

Energy, (electron volts)11860 11880 11900 11920

XANES spectra

Principal Component Analysis of XANES or EXAFS Spectra

Energy (eV)

Com

pone

nt In

tens

ity

1

234567n

Energy

Set of Unknown XANES Spectra

PCA

Principal Components(= number of species)

Subset “best” model compounds For LSA

PC

2 (v

2*w

2,i)

22%

of v

aria

nce

PC 1 (v1*w1,i)35% of variance

Variance Plot

N = 19

Energy (eV)

Set of Model Compound XANES Spectra

Target TransformationUsing principal components N = 30

Synchrotron studies of As in Gold Mine WastesEarly Days: 1994-2002

individual field and lab projects at “targets of opportunity, typically with limited connection to regulators’ needs

Bulk XANES + EXAFS

2003- present

collaborative projects at high-profile sites; research focused on addressing needs

Microbeam studies: 2005 and laterCoupled XAFS, XRD XRFCoupled bulk and microbeamMulti-metal XAFS

Complimentary Lab-based techniquesMicro Raman, XRD

As Lα50 μm

Fe Kα

3 4 5 6 7 8 9 10 11 12

k (Å-1)

0 1 2 3 4 5 6

Radial Distance (Å)

datafit

As(V)-Gibbsite

As(V)-FeOOH

RM

a b As-O

As-Al/Fe

As-Fe

As-Al

Energy (eV)11850 1195011900

As5+

As-Fe = 3.25 Å

FeFe

As

As

AlAl

As-Al 3.16 Å

Ruth Mine: Ballarat District (Trona, CA)Tailings (ca 1000 mg/kg As) used for residential landscaping

Foster et al., (1998) American Mineralogist 83, 553-568

D. L

awle

r, BL

M

Ruth MineTrona, CA

Mesa De Oro: Should be “Mesa de Arsenico”

http://www.pbs.org/newshour/bb/environment/superfund_4-16.html (transcript of 1996 show called “Paying for the Past”)

• Gold tailings with 115 – 1320 mg/kg arsenic• 40 homes developed on Mesa between 1975-1985• EPA emergency response

• halted new home construction• removed and replaced about 1 foot of soil• shored up sides of Mesa

George Wheeldon, “geologist”: arsenic from the mine is in a form that is not dangerous

Time Magazine Sept 25, 2000

San Jose Mercury News

Residents won 2,000,000 for loss of property valuehttp://consumerlawpage.com/article/environmental_pollution_1.shtml

XANES spectra of Mesa de Oro soil samples demonstrate that arsenic is not in arsenopyrite form

Arsenopyrite FeAsS As5+ on rust (iron oxyhydroxide)

Arsenopyrite (As1-)

Energy (KeV)

11.86 11.88 11.90 11.92

Arsenic (V) on Rust

Range of Mesa de OroSamples (n =4)

Mr Wheeldon’s Error: assuming that arsenic stays in original form

A. Foster, R. Ashley, and J. Rytuba, USGS: unpublished data

Lava Cap Mine Superfund Site, Nevada Cty, CA

Arsenic species in mine-impacted sediment from the Lava Cap Mine

Foster et al., (2010) Geochemical Transactions

pyrite arsenopyrite

pronounced oxidation

minimal oxidation

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0

PC 2

(v2*

w2,

i)22

% o

f var

ianc

e

PC 1 (v1*w1,i)35% of variance

subareal tailings≈ 30% FeAsS

Pyrite-rich ore69% Fe(As,S)2

≈ 65% FeAsS

Arsenopyrite-rich ore(FeAsS)

submergedtailings

typical ore

2 4 6 8 10 12 14

subareal tailings

photoelectron wave vector k (Å-1)

k3 -w

eigh

ted

EXAF

S

Our first studies at Lava Cap showed that submerged tailings from the private lake contain As in its original forms. Tailings exposed to air after the burst of a log dam in 1998 have oxidized considerably.

Kelly/Rand Mines: Ultra-high arsenic in mine tailings• gold-silver mines operated until 1947• approximate volume: 100,000 tons • breach of tailings levee; migration of tailings into residences in Randsburg• 3000->10,000 mg/kg As • remote, Federal Land (BLM), popular with OHVers

Kim, C.S., Wilson, K.M., and Rytuba, J.J. (2011) Particle-size dependence on metal distributions in mine wastes: implications for water contamination and human exposure. Applied Geochemistry 26, 484-495.

Secondary arsenates and As5+-rich sulfate phases predominate in Kelly/Rand tailings

0

20

40

60

80

100

120

“background”soil

FeAsO4 2H2O

am-FeAsO4

Ca2Fe3(AsO4)3O2 3H2O

KFe3[(S,As)O4)](OH)6

As-FeOOH

Secondary crustsTailingsLow-gradeOre

Line

ar C

ombi

natio

n fit

s to

EXAF

S

• no evidence for primary sulfide phases (below detection? Oxidized?)

• solubility and kinetics of dissolution of precipitates is expected to be very different than that of arsenic on ferric oxyhydroxide

Kim, C.S., Wilson, K.M., and Rytuba, J.J. (2011) Particle-size dependence on metal distributions in mine wastes: implications for water contamination and human exposure. Applied Geochemistry 26, 484-495.

Lungs of tortoises collected near mines contain particles similar to those found in mine tailings

687 lung tissue

11860 11880 11900 11920

scorodite

jarosite

Spot 2

Spot 1

4.6 mm

3.5

mmAs

1 2

A. Foster, unpublished data

Ultra-high As gold mines in Nova Scotia, Canada

Walker et al (2009) Canadian Mineralogist v 47

Current and Future directions in synchrotron-based arsenic research

• Validated method for As bioaccessibility test– Coupled to geochemistry, As speciation

• Bioreactors (anaerobic and aerobic)– Aerobic: naturally-occurring microbial consortia?– Anaerobic: relative bioaccessibility of As in neo-

sulfides vs. organic compounds (biomass)• Plants

– Finding more accumulators, maximizing uptake– Coupling genetics, protein expression, and location of

metal (As) sequestration

Arsenic Relative Bioavailability Project (Empire Mine State Historic Park)

N =25

25

ChemistryX-ray diffractionElectron ProbeQEMSCANParticle size analysis“Bulk” Synchrotron studies “Microfocused”

Synchrotron studies

In vitro

In vivo

Model of mineralogical control on As bioaccessibility

Correlation with easily-measured parameter

Helping to find a cost-effective means of evaluating the potential for re-development of mined lands contaminated with arsenic

% Bioaccessible

% B

ioav

aila

ble

This study is conducted through a proposal by CA Department of Toxic Substances Control and USGS that was funded by US-EPA (Brownfields Program)

Principal Component Analysis predicts 4-5 unique arsenic species

-0.2

0.3

0.8

1.3

-0.2 0.3 0.8

-0.8

-0.4

0

0.4

0.8

-0.8 -0.4 0 0.4

Component 1

Com

pone

nt 2

Component 1

Com

pone

nt 2

N =19

N =18

Outlier removed

Variance can be visualized on additional Component axes (3, 4, 5)

Bioaccessibility and Bioavailability have similar trends with key arsenic species

% re

lativ

e bi

oava

ilabi

lity

R² = 0.3028

0

5

10

15

20

25

30

35

0.0% 10.0% 20.0% 30.0% 40.0%

Ca-Fe-Arsenate

bioday 6/7

bioday 9/10

bioday 12/13

bioall days

R² = 0.9484

0

5

10

15

20

25

30

35

0 0.1 0.2 0.3 0.4 0.5 0.6

Pyrite/Arsenopy

R² = 0.28

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0%Ca-Fe arsenate

R² = 0.385

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 0.1 0.2 0.3 0.4 0.5 0.6As-Sulfides

Ca-Fe-Arsenate

Pyrite/Arsenopyrite

SEE MORE AT ALPERS TALK TOMORROW SESSION 10

Arsenic in roasted ore treated by an anerobicbiochemical reactor

Maghemite-richMore As3+

Hematite-richLess As3+

Paktunc et al. (2008) Proceedings of the 9th Intl Conference for Appl. Mineralogy

Anerobic Reactor SamplesMostly As3+, but organic NOT SULFIDE (??) As3+

SCH3 CH3

New and improved pyrite: now with As3+ Deditius et al (2008)

Yanacocha , Peru

Should be present in other low-sulfidation epithermal deposits: Nevada Au?

(Fe,Au)(As,S)2

Arsenic attenuation by naturally-occurring microbial consortia: Lava Cap Mine (NPL), CA

0

5000

10000

15000

20000

25000

30000

Arse

nic

(mg/

kg)

LL1LL10LL1adjLL2sedLL2 Fe flocalgae

LL2

LL1LL1adjLL10

Foster et al, USGS Open File Report 2009-1268

LL1LL10

LL2 algae

LL2 fe floc

LL1 adj

LL2 sediment

Biogenic Fe-(hydroxide) accumulates arsenic to levels several times to orders of magnitude greater than the original mine tailngs (yellow horizontal line)

Monitoring the performance of a natural passive bioreactor

LL1 LL10

0

10

20

30

40

50

60

70

80

90

100

Jun 06 Oct 06 Nov 06 Feb 07 Mar 07 Aug 07 Oct 07 Mar 08

AsFeMn

As: 12-30 μg/l 10 μg/lMn: 235-2000 μg/l 300 μg/l

EPA cleanup goal:

% re

mov

ed b

etw

een

LL1

and

LL10

Concentration range at LL10

Arsenic is associated with Fe Oxyhydroxides rather than with biological materials (contrast the anerobic treatment of Paktunc)

-0.10

0.16

0.11 0.46

99AF-wLL5

00AF-LCD1apost-rinse

00AF-LCD1a

v 2*w

2,i

v1*w1,i

As(V)-FeOOHdominant

1000x

EPA region 9 Superfund has a pilot aerobic /anerobic treatment in place at the adit, but is not supplanting it with the native microorganisms

As Speciation in Pteris vittata HyperaccumulatingFern

Webb et al (2003) ES&TPickering et al. Environ. Sci. Technol., 40 (2006) 5010-5014.

Synchrotron techniques have had great utility in the study of arsenic speciation in gold mines..there is more to come!

Arsenopyrite

Primary (ore, concentrates)

Pyrite

n = -1, +3

n

3+

5+

O

S

Fe

As-poor

acidic

Near-neutralCa minerals

FeOOH

Ca-Fe arsenates

As-rich

Fe arsenates Fe sulfates

The End

Leptothrix ochracea from the Lava Cap Mine