Article pubs.acs.org/est
Biogeochemical Controls on Hexavalent Chromium Formation in Estuarine Sediments Amar R. Wadhawan,#,¶ Alan T. Stone,# and Edward J. Bouwer*,# #
Johns Hopkins University, Department of Geography and Environmental Engineering, 3400 N. Charles Street, 313 Ames Hall, Baltimore, Maryland 21218, United States ¶ Geosyntec Consultants, 10220 Old Columbia Road, Suite A, Columbia, Maryland 21046, United States S Supporting Information *
ABSTRACT: Predicting the aquatic and human health impacts of chromium (Cr) necessitates one to determine its speciation as either relatively nontoxic CrIII or toxic CrVI and elucidate the influence of biogeochemical changes on its behavior and fate. In the Baltimore Harbor, Cr predominantly exists as CrIII associated with sediments. While reduction of CrVI to CrIII is dominant in these anoxic sediments, the potential of CrIII oxidation and CrVI reoccurrence during sediment resuspension and oxygenation resulting from dredging, bioturbation, and flood events poses a serious concern. In batch experiments, aqueous CrVI spiked into continuously mixed anoxic suspensions was reduced to product CrIII under anaerobic conditions. No CrVI reoccurrence was observed when conditions remained anaerobic. Aeration caused CrVI reoccurrence from the abiotic oxidation of product CrIII. Rates of aeration-driven CrVI reoccurrence increased with pH, and CrVI reoccurrence positively correlated with dissolved manganese (Mn) decline at pH ≥ 7. Aeration-driven oxidation of MnII to MnIII,IV(hydr)oxides was the underlying mechanism causing product CrIII oxidation. CrVI reoccurrence decreased with sediment loading and negatively correlated with the acid volatile sulfide (AVS) concentration. Although sediment resuspension and oxygenation may create temporary conditions conducive to CrVI formation, long-term CrVI persistence is unlikely in the presence of sediment reductants. While such natural attenuation in reducing environments mitigates the risk associated with Cr toxicity, this risk may still persist in Mn-rich and reductant-deficient environments.
■
INTRODUCTION
reduction−oxidation, adsorption−desorption, precipitation− dissolution, and complexation.11−14 The species-dependent mobility, bioavailability, and toxicity of Cr have made it attractive for implementing in situ remediation where CrVI is reduced to sparingly soluble CrIII.15,16 Several reductants in the natural environment serve as electron donors and reduce CrVI to CrIII. In oxic environments, organics 17−21 and Fe II -bearing mineral phases22−24 act as primary reductants of CrVI, whereas in anoxic environments, CrVI attenuation is achieved through its reduction by dissolved FeII,25−28 FeS(s)/FeS2(s) phases,29,30 and other reduced sulfur species.31 The stability of the product Cr III at circum-neutral pH conditions, Cr(OH) 3 (s)/ CrxFe1‑x(OH)3(s), is assured under reducing conditions with excess reductant supply, which prevents CrIII oxidation and creates an environmental sink for CrVI. However, biogeochemical conditions may change over a prolonged period of time favoring CrIII oxidation and CrVI reoccurrence. CrIII oxidation
The widespread problem of sediment contamination originating from years of navigational, maritime, and industrial activities in the coastal bays and harbors across the U.S. has raised concerns regarding aquatic biota and human health. The complex mixture of metals and organic contaminants found in sediments has proven toxic to benthic organisms, aquatic life, and subsequently to humans by bioaccumulation in the food web.1,2 The Baltimore Harbor-Patapsco River system in the Chesapeake Bay is a Region of Concern for contaminated sediments, and chromium (Cr) is one of the specific contaminants of concern.3,4 The behavior, transport, and fate of Cr in such estuarine environments depend heavily on its presence as trivalent Cr (CrIII) or as hexavalent Cr (CrVI).5 Under acidic pH, CrIII predominantly exists as dissolved cationic species while at circum-neutral pH it precipitates as Cr(OH)3(s) and as CrxFe1‑x(OH)3(s) in the presence of iron (hydr)oxides, limiting CrIII solubility and bioavailability.6 CrIII is considerably less toxic than CrVI, which is a carcinogen and occurs mainly as a soluble oxyanionic species, making it readily bioavailable.7−10 This speciation of Cr is governed by the biogeochemical conditions that influence processes such as © 2013 American Chemical Society
Received: Revised: Accepted: Published: 8220
July 19, 2012 June 26, 2013 June 26, 2013 June 26, 2013 dx.doi.org/10.1021/es401159b | Environ. Sci. Technol. 2013, 47, 8220−8228
Environmental Science & Technology
Article
Table 1. Analyses of Baltimore Harbor Sediments and Porewater (PW)a
a
sample
PWpH
BC-1007 IH-1007 DMT-109 54-109 45-109 38-109 DMT-909 68-909 45-909 33-909
7.71 7.83 7.70 7.66 7.73 7.68 7.99 7.92 7.58 7.52
% solids 20.0 22.8 66.3 43.3 28.4 31.0 62.4 28.6 28.0 18.1
(0.13) (0.85) (0.50) (0.87) (0.16) (0.17) (0.10) (0.25) (0.34) (0.34)
% TOC 5.89 3.61 0.41 1.58 4.37 3.38 NM 3.87 4.23 5.27
(0.13) (0.21) (0.13) (0.10) (0.05) (0.04) (0.06) (0.01)
AVS (μmol/g dry wt) 307 (9.08) 147 (18.1) 0.34 (0.03) 12.8 (3.52) 142 (1.41) 132 (2.63) 25.0 (0.12) 75.4 (7.65) 110 (33.5) 175 (6.11)
SEM (μmol/g dry wt) 13.8 NM 0.75 1.20 5.65 6.67 3.50 6.30 5.75 13.2
(0.57) (0.16) (0.01) (0.03) (0.04) (0.62) (0.11) (0.11) (0.54)
CrT (mg/kg dry wt)
FeT (g/kg dry wt)
600 (41.8) 294 (48.3) 214 (13.5) 83.5 (7.1) 276 (10.7) 239 (18.2) 1274 (143) 355 (21.3) 245 (4.2) 525 (55.1)
87.0 40.8 10.5 34.6 78.3 42.9 27.0 44.0 80.2 73.3
(8.9) (1.7) (0.1) (1.9) (2.1) (0.5) (1.4) (0.8) (2.9) (5.8)
MnT (mg/kg dry wt) 612 483 348 445 618 559 803 533 700 487
(89.5) (20.8) (12.9) (24.3) (17.9) (8.2) (29.3) (3.1) (28.2) (53.4)
PW-Mn (μM) 9.11 26.9 17.3 47.4 12.9 28.0 25.3 5.66 16.1 7.61
(1.08) (1.51) (13.6) (1.45) (6.44) (1.27) (1.25) (0.21) (0.92) (3.62)
PW-FeT (μM) 20.8 10.4 4.69 21.8 15.2 16.4 6.25 5.11 41.2 29.5
(6.11) (0.72) (0.27) (2.17) (2.17) (9.62) (1.27) (0.21) (26.2) (32.8)
Note: The values in parentheses are the difference between duplicate samples. NM stands for not measured.
to CrVI occurs predominantly in the presence of manganese oxyhydroxides (MnIII,IV(hydr)oxides), the only known naturally occurring minerals capable of oxidizing CrIII.32−39 The present study is the first of its kind aimed at elucidating the biogeochemical conditions and mechanisms that influence CrIII stability with respect to its oxidation in estuarine sediments. Cr accumulation in the Baltimore Harbor sediments stems from the leaching of CrVI from chromium ore processing residue (COPR) that was used for years as a fill material throughout the harbor and surrounding areas.40,41 Under the anoxic sulfide-rich reducing conditions in the harbor sediments, CrVI is rapidly reduced to particulate CrIII and the rate of this reduction positively correlates with the acid volatile sulfide (AVS) concentrations in the sediments.42 The anoxic sediments in quiescent conditions are not expected to be a source of toxic CrVI to the water column.43 Sediment oxygenation, however, has been demonstrated to result in the oxidation of metal sulfides and other reductants.44 This lowering of reductant capacity may create an oxidizing environment conducive to CrIII oxidation and CrVI persistence. Hence, the compelling need arises to determine the long-term stability of CrIII under the varying biogeochemical conditions of the Baltimore Harbor sediments. In this study, we monitored the change in Cr speciation and redox-cycling in sediment suspensions when the suspension environment was altered from its primary anoxic and reduced state to an oxic and oxidized state by exposure to air. The main objectives were to (i) determine the likelihood of CrVI reoccurrence in these sediments after its reduction to product CrIII and (ii) identify the corresponding biogeochemical factors and underlying mechanism(s) controlling CrVI reoccurrence if CrIII oxidation was observed. The resulting information is expected to provide predictive insight into Cr speciation, fate, and transport not only in Baltimore Harbor sediments but also in other similarly Cr-contaminated environmental settings.
name followed by the month and year of sampling; for example DMT-207 refers to the sample collected from the DMT site in February 2007. The sediments were characterized for solids content, total organic carbon (TOC), total metals (Cr, Mn, and Fe), CrVI, AVS, and simultaneously extracted metals (SEM). Porewater was extracted by centrifugation and analyzed for pH and dissolved metals (Cr, Mn, and Fe). Reagents and Solutions. All reagents used in this study were of trace metal grade or better, and all solutions were prepared using distilled, deionized water (DDIW) (Milli-Q water, 18 MΩ-cm resistivity, Millipore Corp., Milford, MA). The experiments were buffered for pH using noncomplexing buffers (Supporting Information) that have been used previously to study Cr redox reactions.42,45 A 0.0019 M (100 mg/L) K2CrO4 (J.T. Baker, Phillipsburg, NJ) solution was the aqueous CrVI (CrVIaq) source. Batch Reaction Experiments. Batch reaction experiments were performed in 125 mL polypropylene wide-mouth bottles (Fisher Scientific, Fair Lawn, NJ) and were maintained at a constant temperature (22.5 ± 0.5 °C) in an environmental chamber. The experimental procedure comprised three sequential steps. In the first step, the reactors, each containing a 100 mL suspension, were prepared in an anaerobic glovebag (Coy Laboratory Products, Grass Lakes, MI) by mixing homogenized wet sediments with deaerated (N2-sparged) 10 mM NaCl and 10 mM pH buffer solutions. After stirring the suspensions overnight, in the second step, they were spiked with CrVIaq to give the desired final CrVI concentration. Reduction of the added CrVIaq to product CrIII was monitored under these anaerobic conditions. In the third and final step, the reactors were removed from the anaerobic glovebag and aerated. Aeration was performed by uncapping the reactors and exposing the suspensions to room air for 1 h while they were stirred (600 rpm) on a multiposition magnetic stir plate (Variomag Poly 15, Thermo Scientific). The suspensions reached dissolved oxygen saturation within 10 min of aeration. The O2 in the headspace of the reactors was sufficient to maintain aerobic conditions after the reactors were capped. The reactors were aerated once the added CrVIaq reduced to product CrIII, which was confirmed when dissolved Cr reached negligible levels and additionally ascertained by Cr speciation analysis using HPLC-ICP-MS (PerkinElmer, model: Elan DRC II, quadrupole ICP-MS). Thereafter, the reaction was monitored for CrVI reoccurrence during the aeration period. For both anaerobic and aerobic sampling, 5 mL samples were withdrawn at regular time intervals and amended with 0.025 mL of 1 M phosphate buffer (pH 7). The samples were then
■
MATERIALS AND METHODS Sediment Collection and Characterization. Sediment and porewater collection and characterization methods are described in the Supporting Information. In summary, grab samples of surficial sediments were collected from multiple locations in the Baltimore Harbor (Supporting Information Figure S1) to account for spatial heterogeneity and site-to-site differences in biogeochemical conditions. DMT, 68, 54, 45, and 33 sites were sampled more than once over different months and years to examine long-term temporal variations in sediment characteristics. Samples were designated by their respective site 8221
dx.doi.org/10.1021/es401159b | Environ. Sci. Technol. 2013, 47, 8220−8228
Environmental Science & Technology
Article
suspension (5 g/L loading, pH 7) was spiked with 100 μM CrIIIK(SO4)2·12H2O (Sigma Aldrich, St. Louis, MO) and monitored for CrVI formation. While the suspension remained in the anaerobic glovebag, no CrVI formation was observed for 360 h. However, within 24 h of removing the CrIII amended suspension from the anaerobic glovebag and exposing it to air, CrVI formation was observed and CrVI concentration increased with time during the aeration period (Figure 1a). Oxidation of
shaken on a rotary shaker for 1 h to extract any adsorbed CrVI. Following the 1 h extraction, the samples were filtered through a 0.2 μm (pore size) nylon syringe filter (model 431224, Corning Inc., Corning, NY). The first 1 to 2 mL of the filtrate was discarded, and the rest was collected for total dissolved Cr and Mn, and Cr speciation analyses. Experiments were performed under conditions of varying pH, sediment loading, and Cr concentration. Analytical Techniques. Cr speciation was achieved through a reverse-phase ion-pairing HPLC-ICP-MS method described in our previous work.43 The filtrate was diluted with HPLC mobile phase (3 mM tetrabutylammonium hydroxide (TBAH), 0.3 mM EDTA, 1 mM MOPS, adjusted to pH 7.0) for the formation of ion pairs prior to Cr speciation analysis. Total dissolved Cr and Mn concentrations were determined using ICP-MS after diluting the filtrate with 2% HNO3 prior to analysis. The HPLC-ICP-MS analyses were performed in dynamic reaction cell (DRC) mode using NH3 as the reaction gas to remove polyatomic interferences by 40Ar12C+, 40Ar12CH+, and 37Cl16O+, which have the same m/z as either 52Cr+ or 53 Cr+. The pH in the reactors was measured using a combination electrode (Thermo Scientific) and stayed within ±0.2 units of the desired value throughout each experiment.
■
Figure 1. Aeration-driven conversion of CrIII to CrVI in suspended sediments. N2-sparged DMT-207 suspension (5 g/L loading at pH 7) was spiked with 100 μM CrIIIaq under anaerobic conditions. 360 h later, the CrIII-spiked suspension was aerated. CrVI formation was monitored using HPLC-ICP-MS.
RESULTS AND DISCUSSION Sediment Characterization and Effect of Oxygenation. Characterization results of the sediments and porewater collected from the Baltimore Harbor are presented in Table 1. The porewater pH stayed fairly consistent between 7.5 and 8 across the sampling locations, and the total organic carbon (TOC) ranged between 0.4 and 6% w/w. The sediments were typically fine grained with low (20 to 30%) solids content except for the DMT site, which was sandy and high (about 60%) in solids. The AVS concentrations in the sediments from Bear Creek, Colgate Creek, and Curtis Creek sites were much greater than those in sediments from DMT and 54 sites. SEM concentration was at least seven times lower than the corresponding AVS concentration in all but the DMT-109 sample. The AVS/SEM ratio is considered a reliable predictor of metal toxicity in sulfide-rich sediments. Evidence suggests that when AVS > SEM, FeS(s) is present and the toxic metals likely exist as metal sulfides.46 The AVS also represents a surrogate measure of the reducing capacity in sulfide-rich sediments. Even though Cr species do not form Cr-sulfide phases, CrVI is readily reduced to sparingly soluble CrIII by FeS(s) and dissolved sulfide under environmental pH conditions. Therefore, AVS > SEM suggests that excess sulfides or FeS(s) are available for CrVI reduction. This process would describe the observed Cr speciation results. While total Cr concentrations varied from site to site, Cr was predominantly present as CrIII; no significant CrVI was detected in the sediments or porewaters. Low (
