ARTICLE IN PRESS

Water Research 38 (2004) 1155–1158

Arsenic-sulfides confound anion exchange resin speciation of aqueous arsenic Jenny Ayla Jaya,*, Nicole Keon Bluteb, Harold F. Hemondc, John L. Durantd a

Civil and Environmental Engineering Department, University of California Los Angeles, 5732 H Boelter Hall, Los Angeles, CA 90095-1593, USA b McGuire Environmental Consultants, Inc., 1919 Santa Monica Blvd., Suite 200, Santa Monica, CA 90404, USA c Parsons Laboratory, Civil and Environmental Engineering Department, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139, USA d Civil and Environmental Department, Tufts University, 113 Anderson Hall, Medford, MA 02155, USA Received 16 October 2002; received in revised form 27 October 2003; accepted 21 November 2003

Abstract A field-portable anion exchange resin method (often cited as the Ficklin method (1983)) has been extensively used to distinguish between dissolved arsenite (As(III)) and arsenate (As(V)) species in natural waters. As(III) occurs largely as As(OH)3, which is uncharged at ca. pH 7, while As(V) is negatively charged and will sorb to the resin. However, we show that negatively charged As(III)-sulfide (thioarsenite) species, important at sulfide concentrations >10 mM, also bind to the anion exchange resins, and therefore might be interpreted incorrectly as As(V). Furthermore, we show that nitrogen-purging, which results in a conversion of As(III)-sulfides to arsenite, can be used to obtain accurate arsenic speciation when resins are used on sulfidic water samples. r 2003 Elsevier Ltd. All rights reserved. Keywords: Thioarsenites; Adsorption; Speciation; Interference; Ion exchange; Water analysis

1. Introduction Arsenic’s toxicity and mobility in the environment, as well as its ease of removal during drinking water treatment, are largely determined by speciation. Aqueous arsenic exists most commonly as arsenite (As(III)) and arsenate (As(V)), depending on the redox status of the water. A common form of arsenite, AsIII(OH)3, has a pKa of 9.2 and is uncharged in most natural waters. Arsenate is negatively charged at neutral pH, either as V 2
H2AsVO
(pKa of 6.9). This charge 4 or HAs O4 difference results in disparate sorptive affinities between the two oxidation states, and is the basis for anion exchange methods that separate As(III) and As(V) by adsorbing the negatively charged species.

*Corresponding author. Tel.: +1-310-267-5365; fax: +1310-206-2222. E-mail address: [email protected] (J.A. Jay).

Column-based anion exchange resin methods have been developed for rapid, low-cost separation of As(III) and As(V) [1–4]. These methods have been used extensively in field studies [5–11] and at water utilities [4,12–15]. The field portability of the resin eliminates the need for arsenic oxidation-state preservation. Commonly, As(V) has been calculated by subtracting the uncharged arsenic, which passes through the column, from the total arsenic in the influent. The assumption that As(III) is uncharged and that As(V) is charged is not always correct, however. For example, Miller et al. [16] show that at a pH o2 the organoarsenicals, monomethylarsonate (MMA) and dimethylarsinate (DMA), elute with arsenite, although they are in the +V oxidation state. Under reducing, sulfidic conditions, soluble As(III)sulfide species form [17,18]. The exact formulae of the species are unknown, but it is generally thought that the 0
x dominant species, Asx S2x ; is negatively charged at ca. pH 7 [19–21]. In this work we investigated whether

0043-1354/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2003.11.014

ARTICLE IN PRESS J.A. Jay et al. / Water Research 38 (2004) 1155–1158

2. Methods Serum bottles were purged with high-purity nitrogen using a 30 cm-long Teflon needle, during which 50 mL of deoxygenated pH 6 buffer (Hydrion, containing sodium phosphate and potassium phosphate) was transferred to each bottle. Ionic strength was approximately 0.3 M. Each bottle was capped with a butyl rubber septum, crimped, and purged with high-purity nitrogen gas. A stock solution of hydrogen sulfide was prepared using washed crystals of Na2S  9H2O, standardized by iodometry [7]. Aliquots of stock solution were added via syringe to yield final concentrations of 0, 10, 100, and 1000 mM S(-II)T in duplicate bottles. S(-II)T is the sum of all sulfur in the (-II) oxidation state. In this system, S(II)T=(H2S)+(HS
)+(S2
). As(III), prepared from a stock solution of NaAsO2, was added to a final concentration of 3500 nM in all bottles except the blanks. Bottles were equilibrated for a week in the dark before analysis. Resin extractions were performed at pH 6, without pH adjustment. In a nitrogen-filled glove box, aliquots were withdrawn through the septum of each bottle via syringe and were added to purged 15 ml Falcon tubes containing 0.2 g BIO-RAD AG-1  8 anion exchange resin (acetate functional groups). Capped tubes were tumbled for five minutes and allowed to settle for 5 min. The supernatant was decanted by pipet, after which HCl (1 or 6 N) was added to the resin to elute sorbed arsenic. Arsenic was measured by hydride-generation atomic fluorescence spectrometry (HGAF, PSA Excalibur, detection limit of 1.3 nM). Samples were pre-reduced for at least 12 h using 13.9 g/L potassium iodide and 2.8 g/mL ascorbic acid in 1 N HCl, then reduced to arsine with 15 g/L NaBH4 and 100 mM sodium hydroxide immediately before measurement by HGAF. During 24-h purging of As(III)-sulfide solutions, it was necessary to hydrate the gas stream to prevent significant sample loss. High-purity nitrogen was passed through a gas dispersion tube into a tower of deoxygenated water, which was heated to 30 C in a water bath on a hot plate. The hydrated gas stream was then passed through Teflon tubing into a Teflon bottle containing the sample.

Equilibrium modeling of As(III) speciation at varying levels of S(-II)T (pH=6, pe=4, T=22 C) was done using PHREEQCI [20], adding the following reactions and constants for thioarsenite formation [9,21] to the Minteq database for dissolved and solid-phase arsenic compounds: H3 AsO3 þ2HS
þHþ ¼ AsS
2 þ3H2 O; log K ¼ 17:49; H3 AsO3 þ2HS
þ2Hþ ¼ HAsS2 þ3H2 O; log K ¼ 21:29:

3. Results To test whether soluble As(III)-sulfide species bind to the anion exchange resin, arsenic from bottles containing different sulfide concentrations was separated using the anion exchange resins. The amount of arsenic retained by the resin, referred to as ‘‘charged arsenic’’ increased as the sulfide concentration increased (Fig. 1a). This change was mirrored by a corresponding

4000 3500 Charged As (nM)

soluble As(III)-sulfide species bind to anion exchange resins, and compared their elution behavior with that of uncharged arsenite. We then tested the hypothesis that soluble trivalent thioarsenites are converted to uncharged arsenite when sulfide is purged from the system, thus providing a means of overcoming interference when using the resin technique for separating As(III) and As(V) species. Resin binding was compared before and after the sulfide purging step to distinguish between arsenate, uncharged arsenite, and As(III)-sulfide species.

3000 2500 2000 1500 1000 500 0 0

(a)

10

100

1000

S(-II)T (µM) 4000 3500

Uncharged As (nM)

1156

3000 2500 2000 1500 1000 500 0 0

(b)

10

100

1000

S(-II) T (µM)

Fig. 1. (a) Charged arsenic retained on the resin (not purged). Replicate samples are shown as individual bars. (b) Uncharged arsenic passed through the resin (not purged). Replicate samples are shown as individual bars.

ARTICLE IN PRESS J.A. Jay et al. / Water Research 38 (2004) 1155–1158

sulfide would normally be predicted to be removed during this time). However, almost full conversion of charged thioarsenites to uncharged arsenite was observed after 24 h of purging (Figs. 2a and b). At 1000 mM S(-II)T, somewhat less conversion was observed, possibly due to slow kinetics in the dissociation of thioarsenites.

4000 3500 Charged As (nM)

1157

3000 2500 2000 1500 1000 500 0

(a)

10

100

1000

S(-II)T (µM) 4000

Uncharged As (nM)

3500 3000 2500 2000 1500 1000 500 0 0

(b)

10

100

1000

S(-II)T (µM)

Fig. 2. (a) Charged arsenic retained on the resin (purged for 24 h). Replicate samples are shown as individual bars. (b) Uncharged arsenic passed through the resin (purged for 24 h). Replicate samples are shown as individual bars.

decrease in the amount of arsenic in the supernatant, or ‘‘uncharged arsenic’’ (Fig. 1b). Full recovery of arsenic from the resins was not accomplished using 1 N HCl for 10 min, although this elution regime was typically sufficient for arsenate. More concentrated acid and a longer elution time (6 N acid for 7 h) resulted in more complete arsenic recovery. At the highest level of S(-II)T, we did not recover 100% of the added arsenic. This is possibly due to the formation of arsenosulfide solids, e.g. orpiment (As2S3(s)), that are slow to form abiotically but are increasingly favorable at higher S(-II)T [22]. We expected that purging hydrogen sulfide from solution would result in the dissociation of thioarsenites, leaving arsenite in solution. Arsenic speciation by anion exchange was performed for nitrogen-purged thioarsenite solutions. Aliquots from all bottles were resinseparated after no purging (control), 20 min, and 24 h. We found that while purging for 20 min caused some changes in speciation in the expected direction, this amount of purging was insufficient to eliminate the thioarsenites from solution (even though almost all

4. Discussion and conclusions These results clearly show that soluble thioarsenites bind to anion exchange resins along with arsenate at environmentally relevant concentrations of sulfide. The data fit exceptionally well with expected results (considering dissolved phase speciation) based on PHREEQC modeling (Fig. 3). Sorption of thioarsenites has likely resulted in misinterpretation of speciation results in the literature, as the presence of charged arsenic has been interpreted as evidence for the presence of arsenate, even under sulfidic conditions. Caution must be observed in applying anion exchange methods to speciate arsenic in waters that may contain sulfide. Reducing waters should be tested for sulfide, and purged if necessary. It should be noted that the longevity of the thioarsenites in the absence of sulfide (demonstrated by the long required purge time for conversion to arsenite) could also indicate persistence in natural systems. This work may also have implications for anion exchange resin systems used in water treatment for arsenic. Charged As(III)-sulfide species would be effectively removed by ion exchange. However, the stronger elution conditions required for thioarsenite elution in this study may indicate a potential difficulty in resin regeneration.

4000 Concentration of arsenic species (nM)

0

3500 H3 AsO3

3000

AsS2

-

2500 2000 1500 1000 500 0 1

10

100

1000

S(-II)T (µM)

Fig. 3. Modeling of As(III) speciation with increasing amounts of sulfide.

ARTICLE IN PRESS 1158

J.A. Jay et al. / Water Research 38 (2004) 1155–1158

Acknowledgements We thank Eszter Gulacsy, Katherine Lin, and Lisa Walters for meticulous work in the laboratory. Funding for this study was provided by the Melvina Foundation.

[11]

[12]

References [1] Ficklin WH. Separation of arsenic (III) and arsenic (V) in ground waters by ion-exchange. Talanta 1983;30(5):371–3. [2] Henry FT, Thorpe TM. Determination of arsenic (III), arsenic (V), monomethylarsonate, and demthylarsinate by differential pulse polarography after separation by ion exchange chromatography. Anal Chem 1979;52(1):80–3. [3] Pacey GE, Ford JA. Arsenic speciation by ion-exchange separation and graphite-furnace atomic-absorption spectrophotometry. Talanta 1981;28:935–8. [4] Edwards M, Patel S, McNeill L, Chen H-W, Frey M, Eaton AD, Antweiler RC, Tayloe HE. Considerations in arsenic analysis and speciation. J Am Water Works Assoc 1998;90(3):103–13. [5] Wilkie JA, Hering JG. Rapid oxidation of geothermal arsenic (III) in streamwaters of the eastern Sierra Nevada. Environ Sci Technol 1998;32(5):657–62. [6] McCreadie H, Blowes DW, Ptacek CJ, Lambor JL. Influence of reduction reactions and solid-phase composition on porewater concentrations of arsenic. Environ Sci Technol 2000;34(15):3159–66. [7] Ficklin W. Extraction and speciation of arsenic in lacustrine sediments. Talanta 1990;37(8):831–4. [8] Kim M-J. Separation of inorganic arsenic species in groundwater. Bull Environ Contam Toxicol 2001;67:46–51. [9] Grimes DJ, Ficklin WH, Meier AL, McHugh JB. Anomalous gold, antimony, arsenic, and tungsten in ground water and alluvium around disseminated gold deposits along the Getchell Trend, Humboldt County, Nevada. J Geochem Explor 1995;52:351–71. [10] Villa-Lojo MC, Beceiro-Gonzalez E, Alonso-Rodriguez E, Prada-Rodriguez D. Arsenic speciation in marine

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20] [21]

[22]

sediments: effects of redox potential and reducing conditions. Intern J Environ Anal Chem 1997;68(3): 377–89. Azcue JM, Nriagu JO. Impact of abandoned mine tailings on the arsenic concentrations in Moira Lake, Ontario. J Geochem Explor 1995;52:81–9. McNeill LS, Edwards M. Soluble arsenic removal at water treatment plants. J Am Water Works Assoc 1995; 87(4):105–13. McNeill LS, Edwards M. Predicting As removal during metal hydroxide precipitation. J Am Water Works Assoc 1997;89(1):75–86. Hering JG, Chiu VQ. Arsenic occurrence and speciation in municipal ground-water-based supply system. J Environ Eng 2000;126(5):471–4. Fields KA, Chen A, Wang L. Arsenic removal from drinking water by coagulation/filtration and lime softening plants. Columbus, OH: Battelle; 2000. Miller GP, Norman DI, Frisch PL. A comment on arsenic species separation using ion exchange. Water Res 2000;34(4):1397–400. Wood SA, Tait CD, Janecky DR. A raman spectroscopic study of arsenite and thioarsenite species in aqueous solution at 25 C. Geochem Trans 2002;3(4):31–9. Helz G, Tossell J, Charnock J, Pattrick R, Vaughan D, Garner C. Oligomerization in As(III) sulfide solutions: theoretical constraints and spectroscopic evidence. Geochim Cosmochim Acta 1995;59(22):4591–606. Spycher N, Reed M. As(III) and Sb(III) sulfide complexes: an evaluation of stoichiometry and stability from existing experimental data. Geochim Cosmochim Acta 1989;53: 2185–94. Eary LE. The solubility of amorphous As2S3 from 25 C to 90 C. Geochim Cosmochim Acta 1992;56:2267–80. Mironova GD, Zotov AV. Solubility studies of the stability of As(III) sulfide complexes. Geochem Int 1980;17:46–54. Newman D, Beveridge T, Morel F. Precipitation of arsenic trisulfide by Desulfotomaculum auripigmentum. Appl Environ Microbiol 1997;63(5):2022–8.