Environ. Sci. Technol. 2002, 36, 982-988
Predicting Arsenic Solubility in Contaminated Soils Using Isotopic Dilution Techniques A . M . T Y E , † S . D . Y O U N G , * ,† N. M. J. CROUT,† H. ZHANG,‡ S. PRESTON,§ E. H. BAILEY,† W. DAVISON,‡ S. P. MCGRATH,| G. I. PATON,§ AND K. KILHAM§ School of Life and Environmental Sciences, University of Nottingham, University Park, Nottingham NG7 2RD, U.K., Agriculture and Environment Division, IACR Rothamsted, Harpenden, Hertfordshire AL5 2JQ, U.K., Institute of Environmental and Biological Sciences, University of Lancaster, Lancaster LA1 4YQ, U.K., and Department of Plant and Soil Science, University of Aberdeen, Aberdeen AB24 3UU, U.K.
An isotopic dilution assay was developed to measure radiolabile As concentration in a diverse range of soils (pH 3.30-7.62; % C ) 1.00-6.55). Soils amended with 50 mg of As kg-1 (as Na2HAsO4‚7H2O) were incubated for over 800 d in an aerated “microcosm” experiment. After 818 d, radiolabile As ranged from 27 to 57% of total applied As and showed a pH-dependent increase above pH 6. The radiolabile assay was also applied to three sets of soils historically contaminated with sewage sludge or mine-spoil. Results reflected the various geochemical forms in which the arsenic was present. On soils from a sewage disposal facility, radiolabile arsenate ranged from 3 to 60% of total As; mean lability was lower than in the equivalent pH range of the microcosm soils, suggesting occlusion of As into calcium phosphate compounds in the sludgeamended soils. In soils from mining areas in the U.K. and Malaysia, radiolabile As accounted for 0.44-19% of total As. The lowest levels of lability were associated with extremely large As concentrations, up to 17 000 mg kg-1, from arsenopyrite. Soil pore water was extracted from the microcosm experiment and speciated using “GEOCHEM”. The solidSsolution equilibria of As in the microcosm soils was described by a simple model based on competition between HAsO42- and HPO42- for “labile” adsorption sites.
Introduction Elevated concentrations of As in soils are associated with a number of contaminants, including residues from pesticides and preservatives (1), fly ash and colliery wastes (2), munitions manufacture (3), sewage sludge (4), and metalliferous mine wastes (5). The principal exposure pathways to humans resulting from As contamination of soils include direct ingestion of * Corresponding author phone: +44 0 115 951 6256; fax: +44 0 115 951 3251; e-mail:
[email protected]. † University of Nottingham. ‡ University of Lancaster. § University of Aberdeen. | IACR Rothamsted. 982
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 5, 2002
soil, water, and contaminated food; inhalation of soil and household dusts; and percutaneous exposure (3, 6-8). The effects of chronic As toxicity in humans have been welldocumented, largely through epidemiological studies in countries affected by groundwater contamination, such as Taiwan or Bangladesh and West Bengal (9). Guidelines defining threshold concentrations of As in soils vary between countries. In the United States, the “remedial action guideline” for residential soils is 10 mg kg-1 (10). However, factors that are expected to control As solubility and affect bioavailability in soil (such as pH, mineralogy, and phosphate status) are not generally included within the framework of statutory or advisory standards. Both the United States and Europe are moving toward more explicitly formulated risk-based assessments. Consequently, there is interest in developing improved assessments of As mobility and bioavailability in soils. Currently, methods used to estimate the “relative availability” of As in soils are based on single or sequential extraction schemes. Such methods generally reflect the importance of competition between arsenate and phosphate and adsorption/occlusion/coprecipitation of AsV and AsIII with Fe/Al/Ca compounds (11). However, sequential extraction schemes are often subject to operational problems such as poor selectivity and nonspecificity of the reagents, readsorption of metal during extraction, and dispute regarding interpretation (12). A recent variation intended to simulate direct ingestion of soil by humans is the physiologically based extraction technique (PBET) (13). The first part of this paper reports the development of an isotopic dilution assay to measure the “labile pool” of arsenate in soils. Isotopic dilution techniques may be the most appropriate method to assess the reactive pool of chemical species in soil and have been used previously to determine the size of radiolabile pools of metals such as Cd (14) and Zn (15). Radiolabile As was measured in 23 diverse soils amended with sodium arsenate and five metallic contaminants after incubation in “microcosms” for 818 d at constant temperature and moisture content. Radiolabile As was also measured on three sets of soils known to be (historically) contaminated with As including mining sites in England and Malaysia and a dedicated sewage sludge disposal site in England. In addition, we measured the concentration of As in soil pore water over a period of 818 d in the microcosm soils. The speciated solution data were used to parametrize a simple As solubility model based on competition between (HAsO42-) and (HPO42-) for labile adsorption sites on iron oxide.
Experimental Section Microcosm Experiment. A microcosm experiment was established in July 1997 to examine time-dependent changes in the solubility and bioavailability of six inorganic contaminants added to soils. Twenty-three topsoils (0-15 cm) were selected to represent a wide range of soil characteristics including land use (grassland, arable, deciduous woodland, and heathland), pH, %C, and texture (Table 1). The soil was sieved (6. Speciation of arsenate and phosphate in the soil pore water was carried out using GEOCHEM. Input files included [As], [Cl], [NO3], [PO4], [SO4], [NH4], [Ca], [K], [Mg], [Na], pH, and PCO2. Output from GEOCHEM included activities for the arsenate species AsO43-, HAsO42-, and H2AsO4- and for the phosphate species PO43-, HPO42-, and H2PO4-.
Single Extraction Method for Arsenic. A single extractant assay was investigated using several concentrations of (NH4)H2PO4. Originally, a small experiment involving three of the treated microcosm soils was used so that a suitable strength of extractant could be determined. The Arrow, Iveshead, and Evesham soils were chosen as they represent a range of pH values and textures (Table 1). Four strengths of (NH4)H2PO4 were used (0.005, 0.05, 0.15, and 0.5 M), which represented phosphate:arsenate mole ratios of 37, 374, 1022, and 3743, respectively, at the soil:solution ratio used (1:5). Triplicate samples of soil (5 g) were shaken with 25 mL of extracting solution. After being shaken for 24 h at room temperature, the tubes were centrifuged at 2200g for 25 min. A 10-mL aliquot was acidified with 1 mL of concentrated HCl prior to determination of As by hydride generation atomic absorption spectrophotometry (HGAAS). Measurement of Radiolabile As in Soils. A total of 5 g of soil was equilibrated with 25 mL of 0.005 M (NH4)H2PO4 for 5 d to partially solubilize labile As. Samples were then centrifuged at 2200g for 25 min, and a 10-mL aliquot was acidified with 1 mL of concentrated HCl prior to determination of As concentration [AsSol] by HGAAS. The remaining electrolyte and soil were resuspended, and a spike (0.2 mL) of 73As as sodium arsenate was added (approximately 0.006 MBq) and equilibrated by shaking for a further 2 d. After the sample was centrifuged at 2200g for 25 min, a 4-mL aliquot was removed and assayed for 73As by γ-spectrometry. The concentration of radiolabile As was determined as
(
[AsE] ) [AsSol] kd* +
V W
)
(1)
where [AsE] is the concentration of radiolabile soil As or E value (mg kg-1), [AsSol] is the concentration of As in solution (mg L-1), kd* is the isotopic distribution coefficient (L kg-1), V is the solution volume (L), and W is the mass of soil (kg). This method therefore includes the following assumptions: (i) there was no change in AsSol between the fifth and seventh day of equilibration of the soil suspension; (ii) the 73As spike fully equilibrated with the labile arsenate pool in 48 h; (iii) there was no transfer of 73As to the nonlabile pool during 2 d of contact with the soil suspension; (iv) both AsSol and 73As remained in the oxidized form (AsV) throughout the meaVOL. 36, NO. 5, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
983
surement process; (v) the low concentration of (NH4)H2PO4 used to suspend the soil did not release nonlabile As. We decided to include phosphate in the suspending electrolyte because measurements of As in the pore water of the microcosm soils suggested that some soils would produce extremely low 73As activities in solution if equilibration in an indifferent electrolyte was used. Other Data Sets. The radiolabile and single extraction assays were applied to three further sets of soils known to be historically contaminated with As. Soils were air-dried and sieved to
