American Mineralogist, Volume 70, pages 838-844, 1985

The solubility and stability of scorodite,FeAsOa'2}I2O

PArnrct.l M.mrnr Dow lNo J. Douero Rnrsrtor
Departmentof GeologicalSciences
Virginia Polytechniclnstitute and State Uniuersity,Blacksburg,Virginia 24061

Ab6trrct
Scorodite,FeAsOo.2H2O, is an orthorhombic iron arsenatecommonly associatedwith
arsenic-bearingore deposits as a weathering product. This study investigatedthe mode of
formation and persistenceof scoroditein nature, and suggeststhat its presen@controls the
arsenatelevelsin watersdraining from oxidizing arsenic-richores.Solubility experimentsgive
K.oscor- 19-2r'7+o'sand AGi: -1263.52+0.28kJ mol-1. Scoroditeis metastableunder
most conditions and teods to dissolveincongruently,forming iron hydroxidesand releasing
arsenateto solution.
First-hand observationsat the Brinton Arsenic Mine, Floyd County, VA, and observations
by othersat similar arsenopyritedepositsshow a closespatial relationshipbetweenarsenopy-
rite and scorodite,suggestingthat the reaction which forms scoroditeis:
FeAsS(aspy) + 16H+
+ 14Fe3** 10H2O- l4Fe2* + SO?- * FeAsOa.2H2O(scor) (l)
This processis catalyzedby Fenooxitlazs bacteria which oxidize Fe2* to Fe3*. The forma-
tion of scorodite involves the oxidation of arsenopyrite at low pH to give locally high
arsenateand iron activities which lead to scorodite precipitation. After the arsenopyriteis
consumed,scoroditedissolvesincongruentlyto give goethiteand aqueousarsenate.Scorodite
has a stability field at pH valuesnear 2 and arsenateactivity greaterthan 0.01molal. Analysis
of the surfaceand ground waters at the Brinton Arsenic Mine site showed that scorodite
seemsto control maximum dissolvedarsenatelevels.

Introduction under oxidizing conditions found in the weatheringenvi-

Scorodite,orthorhombic FeAsOa' 2HrO, is isostructural
with mansfieldite (AlAsOn.2H2O), variscite (AIPO4.
2HrO), and strengite(Fe3+PO4'2H2O).Publishedchemi-
cal analysesof scoroditeshow little deviation from FeAsOn
.2HzO (Foshaget al., 1930;Foshag,1937; anrdMartens,
1924);however, in unusual geologic environments,inter-
mediate compositionsmay form (Allen and Fahey, 1948;
Yarosh, 1955; Shadlum and Nesterova, 1947).The forma-
tion of scorodite solid solutions is possible but their oc-
curren@ is limited by the unlikely coincidenceof arsenic
and phosphorus. Mansfieldite occurrenceis unlikely be-
causeof the low activity of Al3 +insolution.
The low valuesof arsenicin waters draining from wea-
thering sulfide/sulfarsenidedepositssuggestthat the buffer-
ing minerals have very low solubilities. Scorodite is a
common weathering product of arsenic-bearingore de-
posits (Boyle and Jonasson,1973),and its pcrsistencein
nature suggeststhat its solubility may control the con-
centration of arsenatein natural waters.Scoroditeis by far
the most cornmon natural arsenate; the only other
common arsenatesare annabergite,Ni3(AsOn)2.8HrO and
erythrite, Co.(AsOn), . 8H2O.
Figure 1 shows the Eh-pH relations of the As-HrO
systemconstructedfrom the thermodynamicdata listed in
the Appendix. The arsenate(Ass*) minerals predominate
0003--004x/8
5/07c8-o8
38$02.00
ronment. Arsenite (As3+) minerals are much lesscommon
becauseof the narrower stability range of this oxidation
state.Native arsenicand arsenideminerals form under the
most reducing conditions in the HrO stability field.
The solubility product of scorodite determined by
Chukhlantsev(1956)indicatesthat scoroditeis stable only
at geologically unreasonableiron and arsenateactivities.
The paragenesisof scorodite suggeststhat the solubility
product determinedby Chukhlantsev(1956)is too high.
Experimental design
Syntheticscoroditewaspreparedby mixing ferric chlorideand
sodiumarsenate solutions(Dove,1984).The resultingprecipitate
wasaged14 daysin the mothcrliquor at 9!-105'C and rinsed
with distilled water to remove electrolytcs and excess arsenate.
The product was a light tan powder (l-7 pm diameter), showing a
distinct X-ray diffraction pattern of scorodite with sharp peaks
indicating good crystallinity.
Each solubility experiment consisted of three replicates. Thirty
ml of distilled/deionized water was adjusted to initial pH values of
3.07.5.57or 10.38with HCI or NaOH and combinedwith 1.0g of
the powdered synthetic scorodite. This mixture was placed in a 35
ml polyethylenebottle and incubatedat 2510.5'C for times up to
8 weeks before sampling. Sampled fluids were centrifuged at 8fi)g
for 20 minutes to remove any suspendedsolids prior to analysis
for arsenic and iron. The pH of each sample was measured using
an Orion model 811 pH meter.
838
 DOVE AND RIMSTIDT: SOLUBILITY AND S,TABILITY OF SCORODITE 839

iron hydroxides.Figure 2 illustratesthe differencebetween

arsenicand iron concentrationversustime for two groups
of samples.Four samplesthat dissolvedcongruently had
+o.8 very small (As,o,",- Fe.,.) differences that were random
-bK and on the order of 10-a's m. The remainingsampleshad
much larger (As..', - Fe,o,.) values,resulting from a rap-
idly decreasingiron concentrationaccompaniedby an in-
creasingarsenicconcentration.The samplesthat dissolved
incongruently also showed a systematicpH change with
HAsOi- time (Fig. 3). The pH shift over time for experimentswith
an initial pH of 10.4can be best describedby reactions(3)
and (4). While the pH was greater than pK2 of arsenate
E (6.86),reaction (3) was dominant and two hydrogen ions
E were releasedfor each formula unit of scoroditedissolved.
l! Asoi- When the pH fell below 6.86,reaction(4) was predominant
and one hydrogen ion was releasedper formula unit of
scoroditedissolved.
-o.4 FeAsOo.2H2O + H2O=HASOi- + ne(OH)r(s)+ 2H+(3)
FeAsOn.2H2O+ HrO+HrAsO; + F{OH)3(s) + H+ (4)
Incongruent reactionswith an initial pH of 5.6 showeda
H 3 A s (A Q )
-o.8 small pH decreaseover time (Fig. 3). This decreasecan be
explained by reaction (4) in which one hydrogen ion is
releasedfor each formula unit of scoroditedissolved.The
pH pH shifts (Fig. 3) for the group of experimentswith an
Fig. L Eh-pH diagram of the arsenic-HrO system at dA!: initial pH of 3.1are bestdescribedby reactions(5) and (4):
l0-3. Derived from the programsot,upr,or(Bethke,1978). FeAsOn.2H2O+H2AsO* + Fe(OH)2++ OH- (5)
FeAsOn.2H2O+ H2O+H2AsOf * F{OH)r(s) + H+ 6)
The arsenicanalysiswas modified from colorimetric procedures
of Boltz and Mellon (1947),Kingsley and Schaffert (1951) and Initially, congruent dissolution by reaction (5) was pre-
Haywood and Riley (1976).This procedureis basedon the reac- dominant causinga small pH increasewith the releaseof
tion of the arsenateion with molybdateto form an arsenomolyb- one hydroxyl ion per formula
unit of scorodite dissolved.
date blue complex. Details are given by Dove (1984).A linear
relationship betweenabsorbanceat 840 nm and arsenatecon-
centration was obtained when arseniclevelsranged between0.10
pg and 2.0pg per 1.0ml of solution.
Iron analyseswere done using both colorimetric (modified from
Harvey et al. 1955;Wiersma and Rimstidt, 1984)and atomic ab-
-z--.-
a
sorption procedures.Details of this method which is basedon the I
a 'a
reaction of iron with 1,10 phenanthroline,are given in Dove
(1984).Iron concentrationsless than the colorimetric minimum E. aa
F rn-3
sensitivity of 2.Opg of iron per 1.0 ml were determinedusing a F
q)
Perkin-Elmer703atomic absorptionspectrophotometer. LL
I
./'
Results F /'
o
Only four of twenty experimentsshowedcongruent dis- { to-a
solution of scorodite. In 16 of 20 experiments,scorodite
dissolvedincongruently (detailed data are given in Dove,
198a)by reaction(2):
FeAsOa.2H2O + HrO+HrAsO; + F{OH)3(s) + H+ to-5

(2',)
These experimentsshowed one or more of the following
effects:(1) Concentrationsof total iron declinedrelative to
total arsenicovpr the duration of the experiment.(2) Con-
centration changeswere accompaniedby systematicpH
shifts.(3) Iron concentrationsapproachedequilibrium with
ro 30 50
T I M E ,d o y s
Fig. 2. Total arsenic concentration minus total iron con-
centration versus time for all experiments.Triangles denote expr-
imcnts in which scorodite dissolvcd congruently. Dots represcnt
experimentsin which incongruentdissolutionwaspredominant.
8,10 DOVE AND RIMSTIDT: SOLUBILI.TY AND STABILITY OF SCORODITE

Table 1. Hydrolysis constantsof arsenate(Wagman et al., 1982)

and ferric iron (Baesand Mesmer, 1976,p.489) used to calculate
the K"o of scorodite.

PK

Arsenate

H3ASO4=H2AsOO-*11' 224

t-.*-----... H2ASO4-=HAsOO2-*11' 6.86

1_ HAsoo2-=Asoo3-*1''1* t1.49
a Ferric ircn

pH A F"3* r H2o = Fe(oH)22. . H' 2.19

l- ---\^. Fe(oH)22. . Hrg = P.16H1r. . u* 3.48

.-A=--_ A
Fe(oH)2.. H2o = Fe(oH)3' H 633

F e ( O H ) 3* H 2 O = F e ( O H ) 4 -. H 9.6

where 4g* : 10-eH of the solution; As,o,"1 : arsenic con-

-r--------a-- centration in solution (molal); and Kl, K2, ata.: hydroly-
sis constantsof arsenatespeciesgiven in Table 1. Activity
coeffrcientswere rounded to 1.0 becausecalculated ionic
30 strengthswerealwayslessthan 10-4.
T I M E ,d o y s The solubility product calculated for scorodite
Fig. 3. pH versustime,showingthe systematic pH changefor (19-zr't to'5 at 25"C)is approximately1.5log units lower
exprimentswith threedifferentinitial pH conditions.
Dotsrepre- than the K"o reported by Chukhlantsev(1956).Based on
sentthehighpH group,initial pH: 10.38;trianglesrepresent the this K,o and the free energiesgiven in the Appendix, the
intermediatepH group,initial pH: 5.57;squaresrepresnt the AGi of icorodite is - 1263.52+0.28 kJ mol- 1.
low pH group,initial pH:3.08. The reactionscausingthe sys- The higher solubility product reported by Chukhlantsev
tematicvariationsof pH with timeareexplained in the text.Data (1956)is probably due to grain size effects.The effect of
usedin thisdiagramarefromDove(1984). grain size on solubility measurementsis given by Enustun
and Turkevich (1960).
At approximatelyday 12,incongruentdissolutionoccurred ln (C,/C.n) : l(213)ovull[2.303vRT(2r)] (10)
by reaction (a). The releaseof one hydrogen ion per formu-
la unit of scorodite dissolvedagain causeda decreasein where C,: apparent solubility at a given radius r (m);
pH. C"q : bulk solubility; o : surface free energy (mJ m-2)
Four experimentsshowed congruent dissolution. Their from equation (12); V : molar volume of scorodite (6.99
results were u$ed to calculate the solubility product of x 10-5 m3 mol-l calculatedfrom cell dimensionsgiven in
scorodite.The solubility product definedby the congruent Kitahama et al.t 1975); d : geometric factor (6.0 for
dissolutionof scoroditeis sphereswas usedhere);v : 2 (ions contributed to solution
r
per formula unit of scoroditedissolved);R : 8.314J mol-
FeAsOa'2HrO(scor)+Fe3*+ AsO!- +2H2O (6) K-l and T: temperature(K). The surfacefree energy of
K, : (ap.r*[c6s.r-) (7) scoroditewas estimatedusing a relationship developedby
Sohnel(1982).
The activity ofFe3+ was calculatedusing the equation
o : - l7 .8 log C.n + 34.8 ( 11 )
Fe,o,.,
4pss+ : (8)
. Kr . KrKz . KrK2Kt KrK2K3K4
, -i*-F
'* - gr; - -
o*, {or; Table 2. Results of experiments showing congruent dissolution
usedto calculatesolubility product of scorodite.From thesedata,
where cs* : 10-eHof the solution; Fe.,"' : concentration AGi scorodite: -1263.52+0.28 kJ mol-r.
of iron in solution (molal); and K1, K2, etc.: hydrolysis
coustantsof ferric iron speciesgiven in Table 1. Similarly, TIME
the activity of AsO!- was calculatedfor total arsenicdata E L A P S E D I N I T I A L F I N A L T O T A L T O T A L scoR
pH pH pA$,m pFe,m PKsP
DAYS
as
5 55? 5.97 4.ls 4 84 21.88
8 s.57 4.27 4.32 2't 71
AS.r"r
4r"orr-:ffi (9) 25
J'
557
J fI 5.53
4 61
3 86
4.50
4.69
22.14
21.32
'*x.--r'',r=-EF Average 21.h0.5
 DOVE AND RIMSTIDT: SOLUBILITY AND STABILITY OF SCOR2DITE 841

where o : surfacefree energy (mJ m-2) and C"o : bulk Table 3. Field data and calculatedsolubility quotient, Q, from
solubility (molal). The scorodite solubility determined in waterscollectedat Brinton ArsenicMine, Flovd Countv. VA.
this study (2 x l0-s m) was usedfor C.o in equation(12)
to give o:118 mJ m-2. Figure4 showsthe solubilityof TOTAL TOTAL
5t tE pH pAs,m pFe,m pQ
scorodite as a function of grain size at 25"C. Grain size I About l00m upstream 5 80 6 57 605 2545
afects solubility only when the grain diameter is lessthan from mine site in small
perennial stream
about 2.0 trrm.The material used in this study had diam-
2 Shallow spring near 5.39 23 21
etersof 1.0-7.0pm, and thus the measuredsolubility prod- mine entaance
5 50 5 06

uct is very closeto that of the bulk solid. Syntheticscoro- 3 Seep at base of waste 4 20 3 95 s90 2261
dite prepared by the procedure of Chukhlantsev (1956) pile.

contained much smaller grains with many below the 4, About5mdownstream 412 483 s.84 23 39
from mine site in
measurablelimit of 0.5 p. same perennial stream
as site I

Scorodite occurrence
The Brinton Arsenic Mine site (Dietrich, 1959)in Floyd sampleswere filtered and acidified with 2% nitric acid to
County, Virginia, provided a site to observethe formation give a pH of less than 2.0. Total arsenic and iron were
and breakdownof natural scorodite.The mine site consists determinedby atomic absorption.The resultsare given in
mainly of two wastepiles,one of roastedore and the other Table3.
of arsenopyrite-bearing,quartz sericite schist. The coarse From the measuredpH, total arsenicand total iron data,
sandy soils of the mine site contain up to 42,000ppm total a solubility quotient, Q, was calculatedfor eachsamplesite
arsenic(Rocovich and West, 1975);microscopicexamina- usrng
tion of dump samplesshowedarsenopyriteto be the most
abundant ore mineral with chalcopyrite present in small Q: (ae"r*\ae"of;-) (12)
amounts(Craig,pers.comm.). The value of Q for eachsamplelocation is given in Table 3.
The presenceof scorodite in the mine waste was con- The calculated Q value from site I showed the stream
firmed by X-ray diffraction. Scorodite occurs as thin water entering the area to be undersaturatedwith respect
yellow-green, botryoidal layers on the surface of to scorodite by approximately four orders of magnitude.
arsenopyrite-bearing rocks. This is reasonableas this water is upstreamfrom the high
Severalwater sampleswere collectedfrom the mine area arsenicenvironment.Sites2, 3 and 4 wereunsaturatedwith
on September19, 1983, near the end of a dry summer. respectto scorodite by approximately I to 1.5 orders of
During collection, pH was measuredin the field using a magnitude.The observedundersaturationis probably the
Beckman portable meter (model H5). The sampleswere result of: (1) dilution by streamwater,surfaceand ground-
collectedin polyethylenebottles that had been leachedin water; (2) slow equilibration ratesof the waterswith scoro-
10% nitric acid. Sampleswere protected from direct sun- dite. It is important to note that none of the sample lo-
light during transport. Within a few hours of collection the cations have Q values exceedingthe scorodite solubility
product. This suggeststhat scorodite controls the maxi-
mum arsenatelevelsin solution.
The formation of scoroditeat the Brinton Arsenic Mine
involves the oxidation of arsenopyriteto give locally low
pH and high arsenateand iron activities.After the arse-
nopyrite has been depleted, iron and arsenate con-
centrations decline andlor the pH increases.This allows
E scorodite to dissolveincongruently,precipitating goethite
o and releasingaqueousarsenate.
ASpy- SCOR- cOE + Arsenate(aq) (13)
J It is likely that arsenopyriteis oxidized by ferric iron (in a
6 manner analogous to pyrite oxidation in mine wastes
:
J
o (Wiersma and Rimstidt, 1984) to give ferrous iron, arse-
at
nate,sulfateand a large quantity ofhydrogen ion.
FeAsS(aspy)
* 13Fe3+* 8H2O+
t4Fe2++ SO?- * H3AsOo+ t3H+ (14)
The ferric iron is regeneratedby Feruooxiilansbacteria
D | A M E T E Rg, . m which oxidize ferrousiron using Or.
Fig. 4. Scoroditesolubility as a function of grain diameter,as
calculatedfrom equation(10). Fe2*+ U4oz+H- l":Fe3* + rl2H2o (15)
 DOVE AND RIMSTIDT: SOLUBILITY AND STABILITY OF SCORODITE

Ehrlich (1964)showed that these bacteria persist in high scorodite stability as predicted from the experimentsre-
arsenicenvironmentsand catalyzearsenopyriteoxidation. ported here.
The iron is thus recycledto react with more arsenopyrite
Conclusions
according to equation (15). Scorodite precipitatesquickly
from the acidic,highly concentratedsolutionsof Fe3+ and Scoroditeis metastableunder most conditions and tends
AsOl-: to break down to form iron hydroxidesand aqueousarse-
nate. In many of our experiments,nucleation of iron hy-
Fe3* + H.AsOn * 2HrO+FeAsOn' 2H2O(scor)+ 3H+ droxide causediron concentrationsto declineto levelsthat
(16) approachequilibrium with amorphousFe(OH)3,while the
arsenateconcentrationsincreased.This produced arsenic
Figure 5 shows that scorodite has a stable field at pH concentrationsthat wereas much as 103times greaterthan
valuesnear 2.0 when dp.t+:10-3 and a^.ol- is greater iron concentrations.In only four casesdid iron hydroxides
than 0.1. When the rate of arsenopyriteoxidation declines, fail to nucleate and scorodite dissolvedcongruently.The
the pH increasesand the arsenateactivity may increaseas results of these experimentswere used to calculate Krn
the scorodite breaks down to goethite and aqueousarse- scorodite and AGi $able 2). Figure 5 shows the Eh-pH
nate: stability field of scorodite when iron activity is 10-3 and
FeAsOn'2H2O(scor) - FeOOH(goe) arsenate activity is 1.0 and 0.1. Scoroditeis not stable for
arsenateactivities less than 0.01; thus it persistsonly me-
+ H2AsOn(aq) + H* (17) tastably at lower arsenateconcentrations.Figure 6 illus-
scorodite and goethite solubilities as a function of
Thus, the final products of arsenopyrite weathering are trates
pH; scorodite is stable only when pH is less than 1.5 or
goethiteand arsenate(aq).
arsenateconcentrationsare groaterthan 0.01molal'
A listing of selectedscorodite o@urrencesis given by
Although scoroditeis metastableunder most conditions,
Dove (1984).At most localities,scoroditeis associatedwith
it does form in areaswhere arsenopyriteis oxidizing rap-
the oxidation of arsenopyrite-bearingsulfide deposits.
idly. The close spatial relationshipbetweentheseminerals
Scoroditehas also been reported associatedwith geother-
natural occurrencessuggeststhat the reaction forming
mal systemsand in a pegmatite.In most of these cases, in
scorodite is:
scoroditehas formed under conditions of low pH and high
activities of iron and arsenic.Emmon's (1917)report that FeAsS(aspy) * 14Fe3** l0H2O+
scorodite eventually alters to goethite is consistent with
l4Fe2+ + SOi- + FeAsOn'2HrO(scor)+ l6H* (18)

The process is catalyzed by Fenooxiilans bacteria which

oxidize Fe2* to Fe3*. Scoroditeis stableduring this oxida-
tion processbecausearsenateactivities are high and the
pH is low. After arsenopyriteoxidation ceases'arsenate
+o.8 activities decline and pH increases,causing scorodite to
\o-
o.o*
-..
+o.4

GOE
Fe2*
E F

d]
lrJ )J
a
o
-o4 o

Fig. 6. Logarithm of solubility versuspH for AsrO5, scorodite,

pH
Fig. 5. Eh-pH relations for the Fe-HrO systemat or.: l0-3
(solid lines). The arsenate-HrO systemis superimposedon this
diagram. Scorodite is stable when arsenateactivities are greater
than or equalto 10-1'2e.
amorphous Fe(OH)3 and goethite. The goethite and F(OH).
curveswrederivedfrom data given in Stumm and Morgan (1981,
p. 780).The AsrO, curve was derivedfrom data givenby Naumov
(1974).The scoroditecurve was derived from thc K"o determined
in this study.
 DOVE AND RIMSTIDT: SOLUBILITY AND STABILITY oF scoRoDITE
dissolve incongruently, to iron hydroxide and arsenate ion
solutions. Such a pro@ss was also observed in this labora-
tory study.
Measurements of the pH, hr. and m^" of surface and
ground waters at the Brinton Arsenic Mine site showed
that scorodite solubility probably sets an upper limit for
dissolved arsenate levels in the ground and surface waters.
Acknowledgments
Financial assistance for this research was provided through
Mining and Mineral Resources Resarch Institute Grant No.
Gl134l5l and Virginia PolytechnicInstitute and State Universitv
ResearchDivision.
We appreciate the assistanceof the Department of Civil Engi-
neering VPI and SU where the atomic absorption analyseswere
carried out. This work was also supported by NSF Grant No.
EAR-8318998.We also appreciatedthe critical reviewsof an early
versionof this paper by J. R. Craig and }V. R. Knocke. D. Lang-
muir, R. E. Meintzer and D. K. Nordstrom reviewed the final
version of the manuscript.
References
Allen, V. T. and Fahey, J. J. (1948)Mansfieldite,a new arsenate,
the aluminum analogue of scorodite and the mansfieldite-
scoroditeseries.AmericanMineralogist,33, 122-113.
Baes,C. F. and Mcsmer, R. E. (1976)The Hydrolysis of Cations.
John Wiley and Sons,New York.
Barton, P. B., Jr. (1969)Thermochemicalstudy of the systemF+-
As-S. Geochimicaet CosmochimicaActa, 33, 841-857.
Barton, P. B., Jr. and Skinner,B. J. (1979)Sulfide Mineral Stabil-
ities. In H. Barnes, Ed., Geochemistryof Hydrothermal Ore
Deposits,p.278-4O3.John Wiley and Sons,New York.
Bethke,C. M. (197E)Program soLupr.or.PennsylvaniaState Uni-
versityComputing Center,University Park, Pennsylvania.
Bolt4 D. F. and Mellon, M. G. (1947)Determination of phos-
phorous,germanium,silicon,and arsenicby the heteropolyblue
method.Analytical Chemistry,19,873-E79.
Boyle, R. W. and Jonasson,I. R. (1973)The geochemistryof arse-
nic and its use as an indicator element in geochemicalprospect-
ing. Journal of GeochemicalExploration, 2, 25l-296,
Chukhlantsev,V. G. (1956)Solubility products of a seriesof arse-
nates.Zhurnal AnalitichesikoiKhimii, 11,565-521.
Dietrich, R. V. (1959)Geology and mineral rcsour@sof Floyd
County of the Blue Ridge Upland, SouthwesternVirginia. V.p.l.
Engineering Experimental Station Bulletin, SeriesNo. 134, v.
Lil,t22-124.
Dove, P. M. (1984) The solubility and srability of scoroditg
FeAsOn'2H2O. M.S. Thesis,Virginia PolytechnicInstirute and
State University, Blacksburg Virginia.
Emmons,W. H. (1917)The enrichmentof ore deposits.U.S. Geo-
logical Survcy Bulletin 625, U.S. Government printing Office,
Washington,D.C.,,105.
Enustun,B. V. and Turkevich,J. (1960)Solubility offine parricles
of strontium sulfate.Journal of American ChemicalSociety,32,
450245@.
843
Ehrlich, H. L. (1964)Bacterialoxidation of arsenopyriteand enar-
gite. EconomicGcology, 59, l30Fl3l2.
Foshag W. F. (1937) Carminite and assosiatedminerals from
Mapimi, Mexico. AmericanMineralogis! 22, 47HE4.
Foshag W. H., Berman, H,, and Allen, R. (1930)Scoroditefrom
Gold Hill, Toole Co., Utah. American Mineralogist, 15, 390-
391.
Harvey, A. E., Smart, J. A., and Amis, E. S. (1955)Simulraneous
spectrophotometricdetermination of iron (II) and total iron
with I, l0 phenanthroline.Analytical Chemistry,27, 2G29.
Haywood, M. G. and Riley, J. P. (1976)The spectrophotometric
determination of arsenicin sea water, potable water and elllu-
ents.Analytica Chimica Acta,85, 219-230.
Kingsley, G. R. and Shaffert,R. R. (1951)Microdetennination of
arsenic and its application to biological material. Analytical
Chemistry, 23, 9 l4-9 19.
Kitahama, K., Kiriyama, R., and Baba, y. (1975)Refinementof
the crystal structure of scorodite. Acta Crystallography,B3l,
322-324.
Martens, J. H. C. (1924) Scorodite from Putnam County, New
York. American Mineralogist,9, 27-28.
Naumov, G. 8., Ryzhenko,B. N. and Khodakovosky,I. L. (1974)
Hahdbook of ThennodynamicData. National TechnicalInfor-
mation ServiceReport No. USGS-WRD-7+001.
Robie, R. A., Hemingway,B. S., and Fisher, J. R. (1978)Thermo-
dynamic Propertics of Minerals and Related Substancesat
298.15"Kand I Bar Pressureand at Higher Ternperatures.U.S.
GeologicalSurveyBulletin 1452,456.
Rocovich, S. E. and West, D. A. (1975)Arsenic tolerance in a
population of the grass Anihopogon scoparius Miclu. Scienae,
r88.263-264.
Shadlum, T. N. and Nesterova,Y. S. (1947) phosphoscorodlte
from Blyav (Southern Ural). Zapiski VserossiiMineral. Obsh-
chestva,76,2l2-215,in ChemicalAbstracts(1949)43,2898.
Sohnel, O. (1982)Electrolyte crystal-aqueoussolution interfacial
tensionsfrom crystallizationdata. Journal of Crystal Growth,
57. 101-108.
Stumm, W. and J. J. Morgan (1981)Aquatic Chemistry,2nd Edi
tion. John Wiley and Sons,New York.
Wagrnan,D. D., Evans, W. H., Parker, V. 8,, Schumm, R. H.,
Halow, I., Bailey, S. M., Churney, K. L., and Nuttall, R. L.
(1982) The NBS tables of chemical thermodynamic propcrties.
Journal of Physical and Chernical ReferenceData, 11, Sup-
plement2.
Wiersma, C. L. and Rimstidt, J. D. (1984) Rates of reaction of
pyrite and marcasite with ferric iron at pH 2. Geochimica et
CosmochimicaActa"48, Et-92.
Wilhehl E., Battino, R., and Wilcock, R. J. (1977)Low pressure
solubility of gasesin liquid water. Chemical Revicws,77,219-
262.
Yarosh, N. A. (1955) Mineralogy of the oxidation zone of the
Blagodat deposits (Middle Ural). Trudy Gorno-Geol. Inst.,
Akad. Nauk S.S.S.R.,26, 50-67, in Chemical Abstracts (1957)
5t.647.
Manuscriptreceioed, May 14, 1984;
accepteilforpublication,Much 14, 1985.
E44 DOVE AND RIMSTIDT: SOLUBILITY AND STABILITY OF SCORODITE

Appendix
Critically EvaluatedArsenic Theimodynsmic Drtr

aH{" AGf" Ref

kJmol-l kJmol-1 J.ol-1 K-1

00 00 35 69 (1)
!0 84

AsO (aq) -r63 80 (2)

AsrOa(arsn) -656 97 -s7s 96 107 41 (l)

!t 67 :1 88 :0 13

AsrOr(cld) -654 79 -575 55 ll3 33 (l)

!t 7l !1 05 10 13

lsor3-(ae) -663 71 -447 69 -186 6l (3)

lsoo3-(ac) -888 14 -648 52 -162 76 (2',)

HAsoa2-(ac) -6a9 10 -524 2s -r5 06 (3)

qc
rlt !2O 92

l-1Asoo2-(aq) -90634 -714 11 -t 67 (2)

H2AsO3-(ac) - 7 1 47 9 -5A1 22 ll0 26 (2)

HzAsO4-(aq) - 9 0 95 6 -753 29 1r7 ls (2)

H3AsO3(aq) -742 24 -639 90 (2)

H3AsO4(aq) -902 49 -766 09 184 10 (2)

H3As(aq) 49 37 90 55 92 80 ( 4 )I

HaAs(v) 66 44 68 9r 222 ?6 (2)

AsS(real) -37 94 -35 76 s9 88 (s)2

AsrSa(orc) -96 08 -9s 29 161 49 ( 6) "

Pb,(AsOO)r(s) -3482 34 (3)

A l A s O O . 2 H t (Om a n s) -r 708.75 (3)

Z n a( A s O o) t ' 5 H r O ( l e sr ) - 2 6 1 12 3 (3)

c u 3 ( A s o 4) . 6 H 2 o( s ) -2131 73 (3)

Ni , ( A s o o ) ( a nn a) -3482 34 (3)
r.8HrO
(erv)
C o a (A s O O ) . 8 H r O -3530 46 (3)

F e A s O O . 2 H t(O
scor) -1263 52 (7)

FeAss(aspy) - 1 11 7 5 -r25 70 il3 96 (6)-

-46 44 -66 68 ',16 (6)

FeAsr(loel) 138

M n 3 (A s O 4 ) 2 . 8 H 2 os() -4055 13

C a ,( A s O O ) , . 4 H r Os() -4018.73 (3)

References: (l) Robie et al (1978); (2) wagman et al (1982); (3)

Naumov et al. (1974); (4) Barton (1969); (5) Barton and Skinner
(1979); ( 6 ) Barton ( 1 9 7 9 ) ; ( 7 ) Dove (1984) This 3tudy

I
Footnote3: T"bul"t"d values obtained from Barton (1969) and Wilhelm
et al. (197?) for the reaction H3As(v) = H3As(aq) Consistent with

4
wers determined calorimetrically Consistent with Barton (1969)