Effects of pH on Metals Precipitation and Sorption: Field Bioremediation and Geochemical Modeling Approaches

We tracked the changes in water chemistry (major and trace elements, S isotope compositions, pH, Eh) and At the Sanders Lead car-battery recycling plant, near Troy, AL, identified solid “mineral” phases produced by biochemigroundwater is highly acidic (pH varies from 3 to 3.5) and carries cal processes. Theoretical geochemical modeling was high concentrations of Pb, Cd, Zn, Cu, and Fe. Pilot field experiments conducted at the site show that in situ metabolism of sulfate reducing conducted to predict how the water chemistry evolves, bacteria (SRB) can produce desired geochemical effects to remove which minerals precipitate, and how ferric hydroxides heavy metal Pb, Cd, Zn, and Cu from groundwater. A reaction path adsorb metals during the bioremediation experiment. model of sulfate reduction shows the redox potential (Eh) effects on Themodeling results were compared with field observamineral precipitation and pH controls on the sorption of different tions of changes in water chemistry and precipitations metals. Lead strongly adsorbs to hydrous ferric oxide (HFO) present of solid materials after the injection. in the aquifer over a wide pH range. Both sorption (due to pH inIn the past few decades hydrogeologists have develcrease) and solid sulfide formation are important for removing Pb. oped geochemical models, with varying degrees of soAlthough theoretical modeling shows that the sorption of most cations phistication, to study water–rock interaction in natural is promoted as pH increases, HFO can only scavenge Zn, Cd, Co, systems (e.g., Helgeson et al., 1970; Wolery, 1979; Reed, and Ni at relatively neutral pH conditions. Thus concentrations of our primary contaminants Zn and Cd attenuate in acidic conditions 1982; Parkhurst et al., 1990; Bethke, 1996). The purpose primarily via precipitation or coprecipitation of solid sulfide phase as of constructing geochemical models is to quantify the Eh drops. The modeling result explains why the Pb plume is retarded chemical and biologic processes that could significantly inmigrationwith respect to theCd plume under the low-pH conditions impact water chemistry in hydrologic systems. Hydrogeat the site. For As, arsenate sorbs strongly onto the protonated weak ologists have spent less effort in predicting the fate and sites of ferric hydroxide for the pH range of calculation. Arsenite transformation of heavy metals during artificially insorption is also favored by increasing pH, however, arsenite desorbs duced (or engineered) bioremediation, even though the and becomes mobilized at very low oxidation state as it reacts with chemical evolution of water provides perhaps the best dissolved sulfide to form AsS complexes. In addition, intermittent opportunity to monitor the subsurface remediation prorainfall events could cause short-termEh increases, potentially leading cesses. We present how a reaction path model could be to oxidation of sulfide solids and subsequent pH decrease, and the remobilization of metals. This study argues for the importance of constructed to trace geochemical evolution (e.g., speciaaccounting for pH changes when evaluating the fate, transport, and tion, mineral precipitation, sorption) of groundwater long-term stability of metals at shallow contaminated sites. and sediments during in situ bioremediation experiments conducted at the Troy site. Geochemist’s Workbench (Rockware, Inc., Golden, CO) (Bethke, 1996; Lee A unconfined aquifer below a car battery and Bethke, 1996) was used to investigate how bacteria recycling plant in Troy, AL is heavily contamisulfate reduction induces the precipitation of metal sulnated with heavy metals and sulfate from sulfuric acid fides from the contaminated groundwater. It is well spills (Saunders et al., 2001). To assess the feasibility of known that bacteria sulfate reduction would produce bioremediation at this site, a field pilot study began in desired geochemical effects including an increase in pH December 1999 as an alternative to an on-going pump(e.g., Berner, 1970; Chapelle, 1993; Bottrell et al., 1995). and-treat remediation that proved to be ineffective. The Since the primary control on the process ofmetal attenuobjective of the field experiment was to inject required ation is acid neutralization, our calculations were aimed nutrients to stimulate the growth and metabolic activity at predicting the pH range at which adsorption of metals of SRB that we hypothesized existed in groundwater onto ferric hydroxides minerals becomes significant. In even under aerobic and low-pH conditions (see Postthis geologic setting, ferric hydroxidesmay adsorb heavy gate, 1979; Elliott et al., 1998; Edwards et al., 2000). We metals and help limit or diminish their mobility downfurther hypothesized that sulfate reduction could be stream. A geochemical model thus provides a powerful stimulated to the point where amorphous solid sulfide assessment of the effects of bacteria sulfate reduction on phases would precipitate, leading to metal removals by mineral precipitation and surface adsorption reaction. coprecipitation in solids. This study documents the principal biogeochemical processes that occur at the test MATERIALS AND METHODS site after the injection of bacteria-stimulating solutions. Injection Experiments At the field site, we hypothesized that the SRB were dorM.-K. Lee and J.A. Saunders, Department of Geology and Geogramant and lacked some metabolism-limiting nutrients in the phy, 210 PetrieHall, AuburnUniversity,Auburn,AL 36849. Received 8 Aug. 2002. Special Submissions—Contaminant Characterization, Abbreviations: DIRB, dissimilatory iron-reducing bacteria; EDAX, Transport, andRemediation in ComplexMultiphase Systems. *Correenergy-dispersive x-ray analysis; Eh, redox potential; HFO, hydrous sponding author ([email protected]). ferric oxide; ICP-MS, inductively coupled plasma mass spectrometry; SEM, scanning electron microscope; SRB, sulfate reducing bacteria. Published in Vadose Zone Journal 2:177–185 (2003).