Kinetics of Dimethylated Thioarsenicals and the ... - ACS Publications

Oct 4, 2016 - Kinetics of Dimethylated Thioarsenicals and the Formation of Highly. Toxic Dimethylmonothioarsinic Acid in Environment. Youn-Tae Kim,. â...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/est

Kinetics of Dimethylated Thioarsenicals and the Formation of Highly Toxic Dimethylmonothioarsinic Acid in Environment Youn-Tae Kim,†,‡ Hosub Lee,§ Hye-On Yoon,*,§ and Nam C. Woo*,† †

Department of Earth System Sciences, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea Natural Science Research Institute, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea § Seoul Center, Korea Basic Science Institute, 6-7 Inchon-ro 22-gil, Seongbuk-gu, Seoul 02855, Republic of Korea ‡

S Supporting Information *

ABSTRACT: Dimethylmonothioarsinic acid (DMMTAV) is a highly toxic, thiolated analogue of dimethylarsinic acid (DMAV). In comparison, a further thiolated analogue, dimethyldithioarsinic acid (DMDTAV), and DMAV both exhibit lower toxicity. To understand the environmental conditions responsible for forming DMMTAV, the kinetics of DMAV thiolation are examined. The thiolation of DMAV is pHdependent and consists of two consecutive first-order reactions under excess sulfide conditions. The first thiolation of DMAV to form DMMTAV is faster than the second one to DMDTAV. DMMTAV is therefore an intermediate. The first reaction is first-order in H2S at pH 6.0 and 20 °C; therefore, the overall reaction is second-order and the rate coefficient in this condition is 0.0780 M−1 s−1. The rate coefficient significantly decreases at pH 8.0, indicating that H2S(aq) triggers the thiolation of DMAV. The second reaction rate is significantly decreased at pH 2.5; therefore, reaction under strongly acidic conditions leads to accumulation of highly toxic DMMTAV in the early stages of thiolation. The transformation of DMDTAV to DMMTAV is catalyzed in the presence of ferric iron. Formation of DMMTAV should be considered when assessing risk posed by arsenic under sulfidic or sulfate reducing conditions.

■

INTRODUCTION Methylated arsenicals are well-known metabolites found in both humans and animals that have been exposed to inorganic arsenic (As). Dimethylarsinic acid (DMAV) has low toxicity (LC50 843 μM in cultured A431 human epidermoid carcinoma cells) compared with arsenite (iAsIII) or arsenate (iAsV) (LC50 5.49 or 571 μM, respectively),1 but the highly toxic dimethylarsinous acid (DMAIII) and dimethylmonothioarsinic acid (DMMTAV) (LC50 2.2 and 10.7 μM, respectively)1 are formed during metabolism of DMAV.2−4 Therefore, DMAV has been classified by the International Agency for Research on Cancer (IARC) into group 2B, a possible carcinogen for humans.2 As shown in the Table of Contents graphic, DMMTAV is formed from DMAV by substitution of oxygen by sulfur (S). If both oxygen atoms of DMAV are instead substituted by S, dimethyldithioarsinic acid (DMDTAV), which exhibits low toxicity (LC50 3660 μM in HepG2 human hepatocarcinoma cells), is formed.5 When As species are injected into femoral vein of hamsters, the distribution of DMMTAV is similar to that of DMAIII, in that over 60% of injected As is distributed in muscles and target organs like liver, kidney, and lung; whereas DMDTAV behaves similar to DMAV, which is excreted rapidly into urine without transformation.6 In urine collected from women exposed to As, dimethylated thioarsenicals accounted for up to 44% of the total As excreted, and a maximum concentration of 24 μg/L was observed.7 Recent research has focused on the metabolic pathways of © XXXX American Chemical Society

dimethylated arsenicals both in vivo and in vitro, with special interest paid to highly toxic species such as glutathione conjugates.3,8 In terms of environmental implications, inorganic As species are of great concern due to their abundance.9 The presence of thioarsenates, thiolated analogues of inorganic As species, has been reported in groundwater,10 geothermal water,11−13 and leachate,14 and the mechanism of the formation or transformation of thioarsenates in sulfidic or iron-rich sulfidic conditions has been studied by several research groups.15−17 Methylated As species are usually considered minor components in aqueous environments, despite the fact that they are still used as biocides and may be formed by microbial methylation. Indeed, Maguffin and colleagues18 estimate that ∼100 tons of inorganic As per year is transformed to methylated As species by biomethylation in aquifers. Therefore, the presence of their thiolated analogues, methylated thioarsenicals, cannot be neglected in natural aqueous conditions. The possible presence of methylated thioarsenicals and their detection have in fact been reported by several researchers. Wallschläger and London9 reported that methylated thioarsenReceived: May 27, 2016 Revised: August 26, 2016 Accepted: October 4, 2016

A

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

reactions, and one of them is determining the overall reaction rate. However, we are focused on the initial chemicals that can occur and be determined in environmental conditions. At first, the kinetics of DMAV thiolation were evaluated by conducting laboratory-scale tests under various pH conditions for time periods of up to 3 months. Several factors that can affect the formation and transformation of dimethylated thioarsenicals in environmental conditions were examined, but only chemically induced reactions in aqueous solution were considered. Finally, we suggest the conditions under which the transformation of As species could lead to the presence of highly toxic DMMTAV.

icals were detected in groundwater from reducing aquifers, which were directly impacted by methylated As biocides. Li and colleagues19 have detected various As species, including methylated and sulfur-containing organo-As species, in municipal landfill leachates; DMDTAV, up to a concentration of 27 μg/L, and DMMTAV were quantified in some samples. In a coastal acid sulfate soil-dominated catchment, unknown As chemicals that have been attributed to organo- or thiolated As species were observed to be important constituents of sediment pore water, accounting for up to half of the total As content, especially in the downstream, estuarine locations.20 Only a few studies have reported the formation and transformation of methylated thioarsenicals in natural media other than biological systems. Even though thioarsenicals are important in evaluating As toxicity and mobility in nature, there are difficulties in handling the necessary chemicals. For instance, DMMTAV and DMDTAV must be synthesized in a laboratory and identified before use, as they are not commercially available. These dimethylated thioarsenicals are the products of thiolation of DMAV by H2S, with short reaction time (1 h) for DMMTAV but longer (up to 24 h) for DMDTAV.21,22 The source of H2S is different according to the research groups: some use a gas cylinder of H2S, while others generate H2S by adding hydrochloric acid to iron sulfide21 or sulfuric acid to sodium sulfide solutions.22 There are several in vitro studies reporting DMMTAV formation from DMAIII, not directly from DMAV.3,8,22,23 Recently, both DMAIII and DMMTAV were also found to be possibly reconverted into DMAV during metabolic processes.3 Because DMAIII reduced from DMAV is very unstable and rapidly converted to fivevalent forms such as DMMTAV, DMAV is believed to be responsible for the formation of highly toxic DMMTAV in physiological condition.4,8 Conklin and colleagues24 have reported that the conversion of monomethylarsonic acid (MMAV), DMAV, and trimethylarsine oxide (TMAO) to their thiolated forms is pHdependent and has been observed in laboratory tests in the presence of sulfide at pH ≤ 7. Monothiolated forms are the primary species at pH 3, whereas >92% of initial DMAV is converted to DMDTAV and to a lesser extent DMMTAV at pH 5−7 after 46 h.24 The microbial formation of DMDTAV, DMMTAV, and monomethyldithioarsonic acid (MMDTAV) has been observed over a period of 43 days in municipal landfill leachate using an As spiking test.25 The bacterial generation of hydrogen sulfide was found to play a critical role in the transformation of As species within the landfill body. When DMAV was spiked into simulated landfill leachates containing sulfide (pH 7.5), DMDTAV was formed after 30 days.26 In groundwater samples, the transformation of As species via the slow conversion of mono- or dimethylated thioarsenicals into their oxy-analogues (such as DMDTAV → DMMTAV → DMAV) was observed during storage without preservation, with a half-life of ∼1 month at 4 °C.9 In addition, the loss of dithiospecies, which is accompanied by an increase in the corresponding monothio-species, was observed after acidification with hydrochloric acid.9 The purpose of this study was to understand the conditions leading to the transformation of dimethylated As species. Because highly toxic DMMTAV can be formed by DMAV thiolation or DMDTAV oxidation in environmental conditions, the study will subsequently suggest a reconsideration of the risk posed by changes in As species under these conditions. It is possible that the observed reaction involves several sub-

■

MATERIALS AND METHODS Preparation of Dimethylated Thioarsenicals. DMAV (C2H7AsO2, Sigma-Aldrich) was used to both synthesize thioarsenicals and make calibration standards. DMMTAV and DMDTAV were synthesized at Korea Basic Science Institute according to the protocol of Suzuki and colleagues.6,22,27 Briefly, DMA V was reacted with sodium sulfide and concentrated sulfuric acid at specific ratios (1:1.6:1.6 for DMMTAV and 1:7.5:7.5 for DMDTAV). The synthesized products after separation were analyzed using liquid chromatograph−electrospray ionization−mass spectrometer (LC-ESIMS, Thermo Scientific LCQ Deca XP plus) and identified as DMMTAV ([M+H]+ m/z 155) and DMDTAV ([M−H]− m/z 169). The As concentration of synthesized products was determined using inductively coupled plasma mass spectrometry (ICP/MS, Agilent 7700 series) after acid digestion with concentrated nitric acid at 70 °C for 1 h. Stock solutions of each synthesized thioarsenical were stored in a refrigerator to prevent further conversion and then used within 1 month. The distribution of As species in stock solutions was measured at the beginning of every analysis, and then the peak area was compared to DMAV with deviations below 7%. Thiolation Experiment. The kinetics of DMAV thiolation were examined by laboratory tests. The distribution of As species under sulfidic conditions was measured over a time period of up to 3 months under various pH conditions (2.5, 5.0, 6.0, 7.0, and 8.0). The pH conditions tested were selected according to the research of Conklin et al.24 Thiolation experiments were conducted in a 10 mM phosphate buffer solution that was adjusted to specific pH values with NaOH or concentrated phosphoric acid. DMAV solutions were prepared with each buffer solution, then spiked with a sulfide solution, prepared by dissolving sodium sulfide nonahydrate (Na2S· 9H2O, Sigma-Aldrich) in argon-purged distilled deionized water. To prevent alteration by oxidation, newly purchased sodium sulfide nonahydrate was used and placed under an airtight seal during storage. The final concentrations of DMAV and sulfide were 0.53 μM and 0.18 mM, respectively. The solutions were filled into airtight glass vials with a minimized head space and then stored at room temperature (∼20 °C) with a paper cover to prevent photochemical oxidation. Further tests were conducted at pH 6.0, because all thioarsenicals from DMAV were formed under this condition, and the autoxidation of H2S is slow enough to be negligible at pH < 6.28 To evaluate the reaction order in sulfide, reactions were examined for various DMAV (0.13−1.07 μM) and sulfide (0.009−0.36 mM) concentrations, with the S-to-As ratio ranging from 17 to 680 (Table S1). Samples were analyzed after 30 h of reaction. Because As mobility is significantly affected by iron (Fe),29 the effect of changing redox conditions was tested by spiking samples in which >90% of As exists as B

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology DMDTAV with ferric (FeIII) solution (FeCl3·6H2O, Kento Chemical). The samples were analyzed after 5 days, and the results compared with control reactions, which were not spiked with FeIII solution. In addition to 10 mM phosphate buffer solution, 10 mM ammonium acetate and 5 mM formic acid solutions that had been adjusted to pH 6.0 were also tested for 12 h to evaluate the matrix effect. The samples were opened just before analysis and then discarded. All tests were repeated in triplicate. The reaction products were confirmed by comparing the retention times (RTs) with those of known standards. In addition, the reaction product under strongly acidic condition was analyzed with LCESI-MS and confirmed to be DMMTAV by the mass-to-charge ratio (m/z 155 in positive mode) (Figure S1). The nature of the S species may change in each solution, but we did not control the S species during tests. Therefore, “sulfide” is used to refer to all S species, but the exact species are identified if required. Because our test set was not prepared in anaerobic condition, oxygen could be present in our system even though we tried to minimize air exposure. To examine the effect of oxygen, one test sample set, named “open-to-air set”, was prepared, then transferred into a normal glass vial without a septum, and occasionally aerated afterward. These samples were analyzed after 30 days and compared with reactions conducted under an airtight seal. Analytical Methods. Samples were analyzed by highperformance liquid chromatography (HPLC)−ICP/MS (Agilent 7700 series) using a reversed-phase C18 column (Thermo, HyPURITY C-18, 5 μm, 250 × 4.6 mm) with an eluent of 5 mM formic acid at a flow rate of 1.0 mL/min. The sample injection volume was 50 μL, and samples were diluted 4-fold with eluent to minimize the sample matrix effects and further reaction. The formate solution is considered to be sufficient to preserve lab samples during analysis, as our results showed that it slowed down the thiolation rate. As (m/z 75) and S (m/z 34, the second most abundant isotope) were measured. Calibration was performed for three As species: DMAV, DMMTAV, and DMDTAV. For details of instrumentation and analytical conditions, please refer to the Supporting Information (for Table S2 for instrumentation of ICP/MS and Figure S2 and Table S3 for the calibration of As species). DMDTAV showed a higher detection limit compared to other species, as the method has lower sensitivity toward this compound due to its lower efficiency in ionization by ICP (Table S4). Therefore, the loss of As in each sample was checked by comparing the sum of measured concentrations of all As species to the initial As concentration. The aqueous hydrogen sulfide, H2S, was detected at 5.6 min, and its peak area showed a linear correlation with the amount of spiked sulfide solution at pH 6.0 (Figure S3). The peak area was always checked to confirm the excess H2S condition because the sulfur species was not controlled.

The tests were conducted under excess sulfide conditions (S/ As ≥ 17) with a pseudo-first order approximation.30 The rate of reaction 1 (r) is expressed in eq 3 by the observed rate coefficient (kobs1) defined in eq 4, because the sulfide concentration is considered a constant. Reaction 2 is considered in the same way with kobs2. r = k1[DMAV ]m ·[H 2S]n = kobs1[DMAV ]m

kobs1 =

H 2S

DMMTAV ⎯⎯⎯→ DMDTAV

−1

where k is the reaction rate coefficient (M s ), m is the reaction order of DMAV, n is the reaction order of sulfide, and kobs is the observed reaction rate coefficient (s−1). The thiolation of DMAV under various pH conditions was investigated by measuring the distribution of As species with time. As shown in Figure 1, the amount of DMAV decreased

Figure 1. Changes in arsenic species by time under sulfidic and strongly acidic to near-neutral conditions ((a) pH 2.5; (b) pH 5.0; (c) pH 6.0; (d) pH 7.0; (e) pH 8.0). The initial DMAV and sulfide concentrations in 10 mM phosphate buffer were 0.53 μM and 0.18 mM, respectively. Error bars indicate the standard deviation calculated from triplicate experiments.

RESULTS DMAV Thiolation Process. Under sulfidic conditions, S substitutes oxygen in DMAV (thiolation). In the closed system, thiolation of DMAV is a process consisting of two consecutive reactions: H 2S

(4) −1

■

DMAV ⎯⎯⎯→ DMMTAV

(3)

k1[H 2S]n0

with time and its thiolated species, DMMTAV and DMDTAV, appeared. The reaction rate depends significantly on the pH. In acidic to neutral conditions (Figure 1a−d), DMAV shows an exponential decrease, and the rate of DMMTAV thiolation is significantly decreased at pH 2.5 (Figure 1a). At pH 5.0−7.0, DMMTAV appears as an intermediate, because reaction 1 is

(1) (2) C

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology Table 1. Distribution of Arsenic Species After Reaction for 3 Monthsa DMAV DMMTAV DMDTAV a

pH 2.5

pH 5.0

pH 6.0

pH 7.0

pH 8.0

ndb 0.303 ± 0.002 0.203 ± 0.008

nd 0.015 ± 0.001 0.513 ± 0.002

nd 0.013 ± 0.001 0.523 ± 0.004

nd 0.018 ± 0.001 0.513 ± 0.002

0.263 ± 0.001 0.155 ± 0.001 0.125 ± 0.001

Units = μM; [DMAV]0 = 0.53 μM. bnd = not detected.

Figure 2. Best fittings for evaluating the reaction rate coefficients of DMAV thiolation. Under excess sulfide condition, (a) DMAV concentrations with log scale by time show linear fittings, indicating the pseudo-first-order reaction and pH-dependent characteristics. The reaction order of sulfide is determined by (b) the fitting of observed rate coefficient at pH 6.0 with 30 h reaction and the initial sulfide concentration. The linear function indicates that the reaction is first-order in sulfide. Data is shown in Tables 2 and S1.

Table 2. Observed Reaction Rate Coefficients of DMAV and DMMTAV and Half-Life of DMAV under pH 2.5−8.0 in Excess Sulfide Condition pH 2.5 a

−1

observed reaction rate coefficient (kobs1) (s ) (r2) half-life (t1/2) from kobs1 observed reaction rate coefficientb (kobs2) (s−1) a

−5

1.90 × 10 (0.9974) 10.1 h

pH 5.0 −5

1.05 × 10 (0.9979) 18.3 h 2.65 × 10−6

pH 6.0 −5

1.53 × 10 (0.9999) 12.6 h 4.49 × 10−6

pH 7.0 −6

8.41 × 10 (0.9977) 22.9 h 4.73 × 10−6

pH 8.0 1.38 × 10−7 (0.9420)

Drawn by fitting over 4 points; r2 in bracket. bDrawn by modeling to minimize the normalized root-mean-square error.

slightly faster than reaction 2 (kobs1 > kobs2). At pH 8.0, a decrease in the rate of reaction 1 was clearly observed (Figure 1e). When the reactions were carried out for up to 3 months, the amount of DMDTAV as the end product was increased (Table 1). In the pH range of 5.0−7.0, DMDTAV was the dominant species with >80% yield after 10 days, and it maintained its prevailing status subsequently. Even at pH 2.5 and 8.0, the DMDTAV concentration increased to 40% and 23% of total As after 3 months, respectively. Reaction Rate of DMAV Thiolation. The reaction order of DMAV and the kobs1 value for each pH condition were deduced from the best fit of the natural log of DMAV concentration against time (Figure 2a). Because ln([DMAV]t) showed a linear correlation with time, [DMAV]t can be expressed according to eq 5. Here, the reaction order of DMAV, m, is 1, and kobs1 is the slope of the fitting line. The half-life time (t1/2) was calculated with eq 6 based on eq 5, and the results are shown in Table 2.

The reaction order for sulfide was evaluated with various concentrations of DMAV and sulfide at pH 6.0 during 30 h of reaction. Because ln([DMAV]t) was decreased linearly up to 1.7 × 105 s at pH 6.0 under excess sulfide condition (Figure 2a), the values of kobs1 were calculated by eq 7 based on eq 5 and are shown in Table S1. kobs1 = ln([DMAV ]t /[DMAV ]0 )/Δt

(7)

here t = 30 h = 0.108 × 106 s. The values of kobs1 showed a linear correlation with the initial sulfide concentration (r2 = 0.9920, Figure 2b), indicating firstorder reaction for sulfide (n = 1) in eq 4. Finally, the overall reaction of DMAV thiolation (eq 1) is a second-order reaction, and the rate coefficient (k1) is 0.0780 M−1 s−1 at pH 6.0 and 20 °C, from the slope of the fitting line in Figure 2b. Because DMMTAV appears as an intermediate at pH 5.0− 7.0, its concentration can be written according to eq 8: d[DMMTAV ]/dt = kobs1[DMAV ] − kobs2[DMMTAV ]

[DMAV ]t = [DMAV ]0 e−kobs1t ln([DMAV ]t ) = ln([DMAV ]0 ) − kobs1 × t

(5)

t1/2 = −ln(1/2)/kobs

(6)

[DMMTAV ]t = [DMAV ]0

kobs1 (e−kobs1t − e−kobs2t ) (kobs2 − kobs1) (8)

D

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology The values of kobs2 were determined by the best fit of [DMMTAV]t versus time by minimizing the normalized rootmean-square error (Figure S4). Only data up to 0.8× 106 s were used, and the deduced kobs2 values are shown in Table 2. Effect of Oxidation. The effect of exposure to oxygen (or aeration) was tested by comparing the open-to-air set with the airtight-closed set after 30 days of reaction, because the latter may contain oxygen. The distributions of As species in both sets were similar (Figure S5). The largest difference (3% decrease in DMAV concentration) was observed at pH 8.0, but it proved to be statistically insignificant (p = 0.12) in our results. The test set is a pure solution in a vial that is closed airtight, containing only DMAV and its thiolated analogues, sodium sulfide and associated sulfur species, and the matrix of ammonium phosphate. Therefore, only sulfide and As species are redox-sensitive under the test condition, and the amount of O2 is limited. The oxidation rate of aqueous H2S by O2 is very slow;31 however, when pH > 6.0 (where HS− exists), the oxidation rate increases with pH before being maximized at pH 8.0.28 Therefore, it is inferred that the effect of sulfide autoxidation is considerable at pH > 6.0. The other possible effects of O2 are direct oxidation of thioarsenicals and/or inhibition of thiolation by causing an oxidative atmosphere. Because the reactions proceed in the direction of thiolation even when open to the air, the direct oxidation of thioarsenicals is also believed to be insignificant within the period of 30 days. On the basis of our results, we cannot determine whether O2 inhibits the thiolation or accelerates the oxidation of thioarsenicals. Fe is one of the most abundant elements in soil and sedimentary environments. It is well-known to affect As mobility32 and moreover undergoes a rapid redox reaction when the redox condition changes.29 Therefore, ferric iron (FeIII) was investigated as the most likely oxidant after oxidized S species. Ten days after spiking DMAV solutions with sulfide at pH 6.0, DMDTAV accounted for 93% of As species, DMMTAV was 7%, and DMAV was below the reporting level (Figure 1c). These samples were then spiked with FeIII solutions at two concentrations: 0.018 and 0.107 mM. The reaction mixtures were then stored under airtight conditions. Another 5 days later, samples were analyzed and the results are shown in Figure 3. The DMMTAV concentration increased in the solution spiked with FeIII. The total As concentration was lower in

samples that had been spiked with FeIII than in control reactions without FeIII. Effect of Matrix. All of the above tests were conducted in 10 mM phosphate buffer solution, as phosphate is widely used as a preservative to prevent Fe precipitation and As coprecipitation,33 and it is also one of the common eluents used for As speciation with anion-exchange columns.34 To test the effect of the matrix on DMAV thiolation, two solutions of formate and acetate were separately prepared by adjusting the pH to 6.0 and tested. The two alternative matrices were chosen as formate, which is the eluent employed in our analysis and dilution procedures, and acetate, which is one of the major components in landfill leachate where methylated thioarsenicals have been detected.19 After 12 h of reaction time, which is close to the t1/2 of DMAV in 10 mM phosphate buffer solution, the matrix was found to have significant effects upon the reaction rates (Figure 4). For the reaction in 10 mM phosphate buffer at

Figure 4. Formation of dimethylated thioarsenicals in sulfidic condition under a different matrix at pH 6.0. The initial concentration of DMAV (C0) is 0.53 μM, and the reaction time is 12 h, close to the half-life time of DMAV in phosphate buffer.

pH 6.0, nearly half of the DMAV was transformed to DMMTAV, with only trace amounts of DMDTAV detected. However, very little transformation was observed in both 10 mM acetate and 5 mM formate solutions, with only trace amounts of DMMTAV detected. The concentration of unreacted DMAV was a little higher in 5 mM formate (0.497 ± 0.002 μM) than in 10 mM acetate (0.474 ± 0.003 μM).

■

DISCUSSION Kinetics for the Thiolation of Dimethylated Arsenics. Highly toxic DMMTAV occurs as an intermediate, because the reaction of DMMTAV to DMDTAV is slower than that of DMAV to DMMTAV (Figure 1 and Table 2). From our results, it is known that the reactions are first-order for DMAV and DMMTAV under excess sulfide condition, respectively, and the DMAV thiolation reaction is first-order for sulfide at pH 6.0 (Figures 2 and S3). The formation of dimethylated thioarsenicals is dependent on the pH. The reaction rate of DMAV thiolation followed the order pH 2.5 > pH 6.0 > pH 5.0 > pH 7.0 ≫ pH 8.0 (Figure 2a). However, for DMMTAV thiolation, the reaction rate at pH 2.5 was significantly slower than at other pH values (Figure 1). Consequently, the dominant species after 4 days of reaction are DMMTAV at pH 2.5, DMDTAV in pH 5.0−7.0, and DMAV at pH 8.0. This result is consistent with reports from Conklin and colleagues,24 who showed that, after 46 h, DMMTAV is the major product below pH 5 whereas DMDTAV is the major product in the pH range 5−7, although the test conditions differed from those

Figure 3. Changes in distribution of As species by spiking FeIII at two concentrations (0.018 and 0.107 mM) into DMDTAV dominant solution, 93% of total As, and that 10-day solution after spiking H2S at pH 6.0. When FeIII is spiked, DMMTAV concentration increases after 5 days. E

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology used here. If the reactions are allowed to continue, the final product under sulfidic conditions is DMDTAV. DMDTAV was the single dominant species after 10 days in the pH range of 5.0−7.0 (Figure 1) and continued to increase for three months even at pH 2.5 and pH 8.0 (Table 1). Because the airtight closed set maintained excess H2S condition during the observation period (except pH 8.0), the reactions proceed in the direction of thiolation. The reaction rates for the conversions of both DMAV to DMMTAV and DMMTAV to DMDTAV were significantly depressed at pH 8.0 compared with other pH conditions (Figure 1). This result is consistent with the findings reported by Conklin and colleagues,24,35 who noted that the conversion rates of methylated arsenicals (MMAV, DMAV, and TMAO) to their thiolated analogues were significantly lowered at pH ≥ 8.24 Similar results, namely, lowered rate at pH ≈ 7, were also reported for other As species, the thiolation of arsenosugars,35 and the reduction of arsenate.36 The species of sulfide present was considered as a possible cause of this observation. The pKa1 value of H2S in aqueous solution is 7.02, and the pKa2 value ranges from 12.35 to 15 at 25 °C.28,31,37 The aqueous hydrogen sulfide, H2S(aq), is a predominant species below pH 7.0, but, at pH 8.0, the HS− becomes the dominant species. Therefore, it was hypothesized that H2S(aq) is involved in the thiolation of methylated As species. This hypothesis corresponds to the synthesis method of DMMTAV and DMDTAV by bubbling H2S in DMAV solution.21 From our experimental data at pH 6.0 (Figure 2c), the thiolation of DMAV was revealed to be firstorder in sulfide concentration. Because most of the sulfide is present as H2S(aq) under these conditions, the decrease in H 2 S(aq) causes a slowing in the rate of thiolation. Consequently, it is supposed that HS− displays different kinetic behavior to H2S(aq) in the thiolation of methylated As species. There is a discrepancy between the formation of DMMTAV from DMAV observed here, and recent studies that the thioarsenicals can be formed from DMAIII, not DMAV.3,8,22,23 These studies were conducted in physiological conditions, where the reactions are controlled by enzymes, the pH is neutral (∼7.4), and the reaction time is within a few days. Therefore, their findings may not be applicable to the environmental conditions. However, there is a similar report for the formation of thioarsenates: in a batch test at pH 7.2, thioarsenates were formed only from iAsIII, and iAsV did not react with HS−.17 Hansen and colleagues38 have reported that DMAIII is formed as a byproduct at the initial stage of synthesizing DMMTAV from DMAV but disappears within 70 min. When summarizing these findings, DMAIII is considered as not a byproduct but an intermediate. As a result, the reaction pathway in environmental condition is inferred as follows. At first DMAV is reduced to DMAIII, and then DMAIII is oxidatively thiolated to DMMTAV by H2S, which is dominant at pH ≤ 7, similar to Suzuki’s suggestion for metabolic process.22 If this is the case, the reduction to DMAIII might be the rate-determining step; however, DMAIII was not detected in our experiments. As an aspect of chemistry, the pKa values of methylated As species also affect the reaction. DMAV and DMMTAV have similar pKa values of 6.2 and within the range of 6−7, respectively.39 However, MMAV is distinguishable, with a pKa1 of 4.1 and a pKa2 of 8.7.40 As we tested only for dimethylated arsenicals, we could not observe this effect. However, in Conklin’s result,24 the conversion rate of MMAV showed a difference depending on the pH; at pH 8, ∼40% conversion

was observed, whereas very little conversion was observed at pH 10, which could not be explained only by the pKa of the sulfide. The significant decrease of DMMTAV thiolation rate in pH 2.5 could not be explained here and warranted further research. Stability of Dimethylated Thioarsenicals. Because of the reductive character of sulfide, the thiolation of DMAV proceeds via the formation of thioarsenicals under sulfidic conditions. DMDTAV with low toxicity is rapidly excreted into urine as the intact form in hamsters6 and is believed to be a stable species under physiological conditions for at least several days.21 However, the stability of DMDTAV becomes a different concept in geological timeframes because the movement in groundwater and pore water is very slow. Wallschläger and London9 have reported the DMDTAV transformation to DMMTAV in an unpreserved groundwater sample. In their research, DMDTAV was found to have disappeared within 10 weeks, but highly toxic DMMTAV showed an increase in concentration, before finally decreasing to lower than the initial concentrations. Only DMAV was present after ∼20 weeks. The observation was made in terms of sample preservation but also gave an insight into changes in the risk posed by As species over time. The samples were taken from an aquifer under reducing conditions; however, they might have a certain constituent that catalyzes the oxidation of thioarsenicals, because the reaction was faster than when exposed to air. In our study, the test set was exposed to air but still with excess H2S condition. It showed the ongoing thiolation of DMAV for at least 30 days (Figure S5). However, the oxidation of DMDTAV started only after 5 days reaction, when FeIII was spiked (Figure 3). Fe is very likely to coexist with As under environmental conditions, because Fe-minerals and Fe-rich organic flocs are considered to be sources of As in groundwater, surface water, and sediment due to their known coprecipitation with As or sorption of it.16,29,32,41,42 Because the redox potentials of As, Fe, and S transformations are close to each other, these three elements are involved in the reaction orders in an intricate way.29 Fe in Fe-rich water is easily and rapidly oxidized and forms precipitates likes Fe-(hydr)oxides when exposed to air.33 In Fe-rich sulfidic conditions, the rapid reaction between sulfide and FeIII produces ferrous iron (FeII) and S0(aq).32 The FeII then precipitates as FeS(s).32 Arsenic can coprecipitate with these particles or be transformed to other species. Moreover, both FeII can form chelation with phosphate,33 which is the matrix in our test. Modeling of the reactions of inorganic As species in Fe-rich sulfidic conditions suggests that the oxidation of arsenite by S0 can produce thioarsenates,43 and this reaction was shown to occur in a lab condition.17 Similarly, it is a possible explanation that FeIII reduction forms S0, which then oxidizes methylated thioarsenicals (DMDTAV to DMMTAV). Although the detailed mechanism could not be clarified here, it is obvious that oxidized iron can catalyze the oxidation of DMDTAV even under sulfidic condition. The exact reactions in the solution will warrant further research. Factors that Enhance or Inhibit Thiolation. Landfill leachates are one of the possible environments in which methylated thioarsenicals can form. Li and colleagues19,25 reported the presence of methylated thioarsenicals in municipal landfill leachates and postulated a critical role played by sulfatereducing bacteria in the formation of methylated thioarsenicals. Thus, microbial action can increase both the concentration of methylated As species by methylation from inorganic species F

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology and the sulfide concentration by sulfate reduction,25 causing a rise in reaction rate. An et al.26 confirmed the chemical formation of DMDTAV in simulated landfill leachate and the relationship with sulfide. Landfill leachates under anaerobic conditions show high organic material concentrations containing carboxylic acids as major constituents.44 Because the rate of DMAV thiolation is significantly decreased in formate and acetate solutions compared with phosphate (Figure 4), the actual thiolation rate in leachate may be slower than that in Table 2. On the opposite side, the carboxylic acids can increase the thiolation rate according to the condition. When acetate is introduced into groundwater for the purpose of biostimulation, the result is increased As release under sulfate reducing conditions, leading to the formation of thioarsenic species with different sorption characteristics.17 In ref 17, among three wells with similar concentrations of sulfide, the site with the greatest acetate concentration showed the highest As release. A possible explanation is that the introduction of acetate increased the amount of arsenite that reacted to form thioarsenates. This example requires the thiolation of inorganic As under microbial-driven conditions. In such a scenario, acetate is a carbon source for the sulfate-reducing bacteria. This role of acetate is different, therefore, from that proposed in the current study, which only considers chemical-driven reactions. In the current study, acetate and formate are believed to act as electron donors, inhibiting the thiolation of DMAV by providing oxidative conditions. Exposure to air or the existence of oxidizers (e.g., oxidized S and Fe species) can cause the reverse reaction, namely, the oxidation of thioarsenicals, and further catalyze the reaction. Although we could not determine whether O2 affects DMAV thiolation or not, thiolation was observed to occur under sulfidic condition even when exposed to air, and oxidation of DMDTAV to highly toxic DMMTAV occurred when spiked with FeIII. Even though the autoxidation of sulfide at pH ≤ 6.0 and the direct oxidation of thioarsenicals within 30 days are negligible in our system; these reactions need to be considered in environmental conditions that have various constituents including minerals and microbes, especially when the condition is changing.

except in a strongly acidic environment where the conversion to less-toxic DMDTAV is very slow. Therefore, the strongly acidic and sulfidic condition at pH ∼2.5 is considered to have higher As toxicity than moderately acidic to neutral conditions. If the sulfidic condition is maintained in this condition, the As toxicity might decrease with time because of further reaction to form DMDTAV. Under the sulfidic and alkaline conditions, the existence of oxidative or reductive chemicals can be important for As transformation. DMDTAV is stable under sulfidic conditions; however, the existence of FeIII can trigger its oxidation to highly toxic DMMTAV and can catalyze this transformation more rapidly than is possible via autoxidation of sulfide in the presence of O2. Depending on the stability of Fe-oxides, sulfate reduction can be accompanied by FeIII reduction.45 Subsequently, in conditions in which the redox potential of the environment changes frequently (for example, seasonally saturated sediments), the transformation of dimethylated arsenicals within an Fe−S system must be carefully considered. The presence of formate or acetate inhibit the thiolation of DMAV, although under microbial-driven sulfate reducing conditions, the input of acetate as a carbon source can promote the formation of thioarsenates from inorganic As species.17 Thiolation of dimethylated As species and the oxidation of dimethylated thioarsenicals under field conditions requires further investigation, including consideration of the effect of sorption and microbial-driven redox change.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.6b02656. Test conditions and results to evaluate the reaction order in sulfide, instrumental conditions of ICP/MS, HPLCICP/MS results, total arsenic concentration, correlation between H2S peak area and the amount of spiked sulfide, and best-fit line of DMMTAV concentration (PDF)

■

■

AUTHOR INFORMATION

Corresponding Authors

*Phone: +82-2-6943-4192; fax: +82-2-6943-4149; e-mail: [email protected]. *Phone: +82-2-2123-2674; fax: +82-2-2123-8169; e-mail: [email protected].

ENVIRONMENTAL IMPLICATIONS The highly toxic DMMTAV can be formed in environmental conditions as well as via metabolism. Methylated As species may themselves be a potential source, as they are still used as biocides9 and formed in aquifers by microbial methylation.18 Further living organisms that are exposed to As, including algae20 and humans,7 produce methylated As species during metabolism and can release these to the environment upon excretion or decomposition. When the methylated As species are present in sulfidic or sulfate-reducing conditions, the thiolation of arsenicals occurs. However, the reaction is highly dependent on pH and is affected by the presence of oxidizers such as FeIII. The distribution of As species is therefore complicated under geological conditions and timeframes. Although our study has considered only chemical-driven reactions in aqueous solution under controlled laboratory conditions, we have proposed significant insights as to the conditions leading to the formation of highly-toxic DMMTAV from the less-toxic DMAV or DMDTAV. Under the sulfidic condition with pH ≤ 7, highly toxic DMMTAV is rapidly formed but exists as an intermediate,

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by a grant from the Korea Basic Science Institute (project no. T35760). Youn-Tae Kim was supported by the National Research Foundation of Korea (NRF) and the Center for Women In Science, Engineering and Technology (WISET) Grant under the Program for Returners into R&D (project no. KW-2016-0046). We thank Giehyeon Lee for providing valuable advice and Joo-Hee Chung for her support during ESI-MS measurements.

■

REFERENCES

(1) Naranmandura, H.; Ibata, K.; Suzuki, K. T. Toxicity of dimethylmonothioarsinic acid toward human epidermoid carcinoma A431 cells. Chem. Res. Toxicol. 2007, 20, 1120−1125.

G

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology (2) IARC. Arsenic and Arsenic Compounds. In IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 100C; IARC Publications: Lyon, France, 2012; pp 41−94. (3) Shimoda, Y.; Kurosawa, H.; Kato, K.; Endo, Y.; Yamanaka, K.; Endo, G. Proposal for novel metabolic pathway of highly toxic dimethylated arsenics accompanied by enzymatic sulfuration, desulfuration and oxidation. J. Trace Elem. Med. Biol. 2015, 30, 129−136. (4) Wang, Q. Q.; Thomas, D. J.; Naranmandura, H. Importance of being thiomethylated: Formation, fate, and effects of methylated thioarsenicals. Chem. Res. Toxicol. 2015, 28, 281−289. (5) Ochi, T.; Kita, K.; Suzuki, T.; Rumpler, A.; Goessler, W.; Francesconi, K. A. Cytotoxic, genotoxic and cell-cycle disruptive effects of thio-dimethylarsinate in cultured human cells and the role of glutathione. Toxicol. Appl. Pharmacol. 2008, 228, 59−67. (6) Naranmandura, H.; Iwata, K.; Suzuki, K. T.; Ogra, Y. Distribution and metabolism of four different dimethylated arsenicals in hamsters. Toxicol. Appl. Pharmacol. 2010, 245, 67−75. (7) Raml, R.; Rumpler, A.; Goessler, W.; Vahter, M.; Li, L.; Ochi, T.; Francesconi, K. A. Thio-dimethylarsinate is a common metabolite in urine samples from arsenic-exposed women in Bangladesh. Toxicol. Appl. Pharmacol. 2007, 222, 374−380. (8) Kurosawa, H.; Shimoda, Y.; Miura, M.; Kato, K.; Yamanaka, K.; Hata, A.; Yamano, Y.; Endo, Y.; Endo, G. A novel metabolic activation associated with glutathione in dimethylmonoarsinic acid (DMMTAV)induced toxicity obtained from in vitro reaction of DMMTAV with glutathione. J. Trace Elem. Med. Biol. 2016, 33, 87−94. (9) Wallschläger, D.; London, J. Determination of methylated arsenic-sulfur compounds in groundwater. Environ. Sci. Technol. 2008, 42, 228−234. (10) Stauder, S.; Raue, B.; Sacher, F. Thioarsenates in sulfidic waters. Environ. Sci. Technol. 2005, 39, 5933−5939. (11) Planer-Friedrich, B.; London, J.; McCleskey, R. B.; Nordstrom, D. K.; Wallschläger, D. Thioarsenates in geothermal waters of yellowstone national park: Determination, preservation, and geochemical importance. Environ. Sci. Technol. 2007, 41, 5245−5251. (12) Keller, N. S.; Stefánsson, A.; Sigfússon, B. Arsenic speciation in natural sulfidic geothermal waters. Geochim. Cosmochim. Acta 2014, 142, 15−26. (13) Maher, W. A.; Foster, S.; Krikowa, F.; Duncan, E.; St John, A.; Hug, K.; Moreau, J. W. Thio arsenic species measurements in marine organisms and geothermal waters. Microchem. J. 2013, 111, 82−90. (14) Zhang, J.; Kim, H.; Townsend, T. Methodology for assessing thioarsenic formation potential in sulfidic landfill environments. Chemosphere 2014, 107, 311−318. (15) Couture, R.-M.; Rose, J.; Kumar, N.; Mitchell, K.; Wallschläger, D.; Van Cappellen, P. Sorption of arsenite, arsenate, and thioarsenate to iron oxides and iron sulfides: A kinetic and spectroscopic investigation. Environ. Sci. Technol. 2013, 47, 5652−5659. (16) Suess, E.; Wallschlager, D.; Planer-Friedrich, B. Stabilization of thioarsenates in iron-rich waters. Chemosphere 2011, 83, 1524−1531. (17) Stucker, V. K.; Silverman, D. R.; Williams, K. H.; Sharp, J. O.; Ranville, J. F. Thioarsenic species associated with increased arsenic release during biostimulated subsurface sulfate reduction. Environ. Sci. Technol. 2014, 48, 13367−13375. (18) Maguffin, S. C.; Kirk, M. F.; Daigle, A. R.; Hinkle, S. R.; Jin, Q. Substantial contribution of biomethylation to aquifer arsenic cycling. Nat. Geosci. 2015, 8, 290−293. (19) Li, Y.; Low, G. K.-C.; Scott, J. A.; Amal, R. Arsenic speciation in municipal landfill leachate. Chemosphere 2010, 79, 794−801. (20) Kinsela, A. S.; Collins, R. N.; Waite, T. D. Speciation and transport of arsenic in an acid sulfate soil-dominated catchment, eastern Australia. Chemosphere 2011, 82, 879−887. (21) Fricke, M. W.; Zeller, M.; Sun, H.; Lai, V. W.-M.; Cullen, W. R.; Shoemaker, J. A.; Witkowski, M. R.; Creed, J. T. Chromatographic separation and identification of products from the reaction of dimethylarsinic acid with hydrogen sulfide. Chem. Res. Toxicol. 2005, 18, 1821−1829. (22) Suzuki, K. T.; Mandal, B. K.; Katagiri, A.; Sakuma, Y.; Kawakami, A.; Ogra, Y.; Yamaguchi, K.; Sei, Y.; Yamanaka, K.; Anzai,

K.; Ohmichi, M.; Takayama, H.; Aimi, N. Dimethylthioarsenicals as arsenic metabolites and their chemical preparations. Chem. Res. Toxicol. 2004, 17, 914−921. (23) Naranmandura, H.; Suzuki, K. T. Formation of dimethylthioarsenicals in red blood cells. Toxicol. Appl. Pharmacol. 2008, 227, 390−399. (24) Conklin, S. D.; Fricke, M. W.; Creed, P. A.; Creed, J. T. Investigation of the pH effects on the formation of methylated thioarsenicals, and the effects of pH and temperature on their stability. J. Anal. At. Spectrom. 2008, 23, 711−716. (25) Li, Y.; Low, G. K.-C.; Scott, J. A.; Amal, R. Microbial transformation of arsenic species in municipal landfill leachate. J. Hazard. Mater. 2011, 188, 140−147. (26) An, J.; Kim, K.-H.; Kong, M.; Kim, J.-A.; Shin, J. H.; Ahn, Y. G.; Yoon, H.-O. Formation of dimethyldithioarsinic acid in a simulated landfill leachate in relation to hydrosulfide concentration. Environ. Geochem. Health 2016, 38, 255−263. (27) Naranmandura, H.; Suzuki, N.; Suzuki, K. T. Trivalent arsenicals are bound to proteins during reductive methylation. Chem. Res. Toxicol. 2006, 19, 1010−1018. (28) Chen, K. Y.; Morris, J. C. Kinetics of oxidation of aqueous sulfide by O2. Environ. Sci. Technol. 1972, 6, 529−537. (29) Gorny, J.; Billon, G.; Lesven, L.; Dumoulin, D.; Madé, B.; Noiriel, C. Arsenic behavior in river sediments under redox gradient: A review. Sci. Total Environ. 2015, 505, 423−434. (30) Corbett, J. F. Pseudo first-order kinetics. J. Chem. Educ. 1972, 49, 663. (31) Hughes, M. N.; Centelles, M. N.; Moore, K. P. Making and working with hydrogen sulfide The chemistry and generation of hydrogen sulfide in vitro and its measurement in vivo: A review. Free Radical Biol. Med. 2009, 47, 1346−1353. (32) Kocar, B. D.; Fendorf, S. Thermodynamic constraints on reductive reactions influencing the biogeochemistry of arsenic in soils and sediments. Environ. Sci. Technol. 2009, 43, 4871−4877. (33) Kim, Y.-T.; Yoon, H.; Yoon, C.; Woo, N.-C. An assessment of sampling, preservation, and analytical procedures for arsenic speciation in potentially contaminated waters. Environ. Geochem. Health 2007, 29, 337−346. (34) Komorowicz; Barałkiewicz. Arsenic and its speciation in water samples by high performance liquid chromatography inductively coupled plasma mass spectrometry-Last decade review. Talanta 2011, 84, 247−261. (35) Conklin, S. D.; Creed, P. A.; Creed, J. T. Detection and quantification of a thio-arsenosugar in marine molluscs by IC-ICP-MS with an emphasis on the interaction of arsenosugars with sulfide as a function of pH. J. Anal. At. Spectrom. 2006, 21, 869−875. (36) Sun, F.; Dempsey, B. A.; Osseo-Asare, K. A. As(V) and As(III) reactions on pristine pyrite and on surface-oxidized pyrite. J. Colloid Interface Sci. 2012, 388, 170−175. (37) Li, Q.; Lancaster, J. R., Jr. Chemical foundations of hydrogen sulfide biology. Nitric Oxide 2013, 35, 21−34. (38) Hansen, H. R.; Raab, A.; Jaspars, M.; Milne, B. F.; Feldmann, J. Sulfur-containing arsenical mistaken for dimethylarsinous acid [DMA(III)] and identified as a natural metabolite in urine: Major implications for studies on arsenic metabolism and toxicity. Chem. Res. Toxicol. 2004, 17, 1086−1091. (39) Raml, R.; Goessler, W.; Francesconi, K. A. Improved chromatographic separation of thio-arsenic compounds by reversedphase high performance liquid chromatography-inductively coupled plasma mass spectrometry. J. Chromatogr. A 2006, 1128, 164−170. (40) IARC. Arsenic in drinking-water. In IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 84; IARC Publications: Lyon, France, 2004; pp 41−267. (41) Schröder, I.; Johnson, E.; de Vries, S. Microbial ferric iron reductase. FEMS Microbiology Rev. 2003, 27, 427−447. (42) ThomasArrigo, L. K.; Mikutta, C.; Lohmayer, R.; PlanerFriedrich, B.; Kretzschmar, R. Sulfidization of organic freshwater flocs from a minerotrophic peatland: Specification changes of iron, sulfur, and arsenic. Environ. Sci. Technol. 2016, 50, 3607−3616. H

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology (43) Couture, R.-M.; Van Cappellen, P. Reassessing the role of sulfur geochemistry on arsenic speciation in reducing environments. J. Hazard. Mater. 2011, 189, 647−652. (44) Kjeldsen, P.; Barlaz, M. A.; Rooker, A. P.; Baun, A.; Ledin, A.; Christensen, T. H. Present and long-term composition of MSW landfill leachates: A review. Crit. Rev. Environ. Sci. Technol. 2002, 32, 297−336. (45) Postma, D.; Jakobsen, R. Redox zonation: Equilibrium constraints on the Fe(III)/SO4−reduction interface. Geochim. Cosmochim. Acta 1996, 60, 3169−3175.

I

DOI: 10.1021/acs.est.6b02656 Environ. Sci. Technol. XXXX, XXX, XXX−XXX