Environ. Sci. Technol. l991,25, 1565-1572
Harwell, J. H.; Liapis, A. I.; Litchfield, R.; Hanson, D. T.
Bennett, P. G. Presented at the Western Regional Symposium on Mining and Mineral Processing Waste, Berkeley, CA, 1990. Bradley, W. G.; Sweed, N. H. In Adsorption and Zon Exchange; Zweibel, I,, Sweed, N. W., Eds.; AIChE Symposium Series 152; American Institute of Chemical Engineers: New York, 1975; pp 59-65. Helfferich, F. G. Ind. Eng. Chem. Fundam. 1967, 6, 362. Helfferich, F. G.; Klein, G. Multicomponent Chromatography; Marcel Dekker, Inc.: New York, 1970. Klein, G.; Tondeur, D.; Vermeulen, T. Znd. Eng. Chem. Fundam. 1967,6,339. Lin, K. H. In Adsorption and Zon Exchange; Zweibel, I., Sweed, N. W., Eds.; AIChE Symposium Series 152; American Institute of Chemical Engineers: New York, 1975; pp 86-95. Tondeur, D.; Klein, G. Znd. Eng. Chem. Fundam. 1967,6, 351. Smith, R. P.; Woodburn, E. T. AIChE J . 1978, 24, 577. Saunders, M. S.; Vierow, J. B.; Carta, G. AIChE J . 1989, 35, 53. Yu, J.-W.; Neretnieks, I. Znd. Eng. Chem. Res. 1990, 29, 220. Suwanayuen, S.; Danner, R. P. AZChE J. 1980, 26, 68. Tsezos, M.; Noh, S. H.; Baird, M. H. I. Biotechnol. Bioeng. 1988, 32, 545. Santacesaria,E.; Morbidelli, M.; Danise, P.; Mercenari, M.; Carra, S. Znd. Eng. Chem. Process Res. Dev. 1982,21,440.
Santacesaria, E.; Morbidelli, M.; Servida,A.; Sorti, G.; Carra, S. Znd. Eng. Chem. Process Res. Dev. 1982, 21, 446. Steiner, E. C.; Blau, G. E.; Agin, G. L. Introductory Guide: SZMUSOLV Modeling and Simulation Software; Mitchell and Gauthier Assoc., Inc.: Concord, MA, 1986.
Chem. Eng. Sci. 1980,35, 2287. Wong, Y. W.; Niedzwiecki, J. L. In Adsorption and Zon Exchange; Ma, W. A,, Ed.; AIChE Symposium Series 219; American Institute of Chemical Engineers: New York, 1982; pp 120-127. Moon, H.; Tien, C. In Adsorption and Zon Exchange; LeVan, M. D., Ed,;AIChE Symposium Series 264; American Institute of Chemical Engineers: New York, 1988; pp 94-108. Coroyannakis,P.; Basmadjian, D. In Adsorption and Zon Exchange: Recent Developments; Ausikaitis, J. P., Myers, A. L., Eds.; AIChE Symposium Series 242; American Institute of Chemical Engineers: New York, 1985; pp 9-16. Rhee, H.-K.; Amundson, N. R. AIChE J . 1982, 28, 423. Carra, S.; Santacesaria, E.; Morbidelli, M.; Storti, G.; Gelosa, D. Znd. Eng. Chem. Process Res. Dev. 1982, 21, 451. Gu, T.; Tsai, G. J.; Tsao, G. T. AZChE J. 1990, 36, 784. Pirkle, J. C.; Schiesser, W. E. DSS/2: A Transportable Fortran 77 Code for Systems of Ordinary and One-,Two-,
and Three-Dimensional Partial Differential Equations. Presented at the 1987 Summer Computer Simulation Conference, Montreal, 1987. Fergusen, C. R. U.S.Bureau of Mines, Salt Lake City, UT, personal communication, 1989.
Received for review October 2,1990. Revised manuscript received March 20,1991. Accepted March 27, 1991. Some of this work was sponsored by a grant from the US.Environmental Protection Agency, R-817440-01-0. A grant of computer time from the Utah Supercomputing Institute, which isfunded by the State of U t a h and the ZBM Corp., is also gratefully acknowledged.
Oxidation-Reduction Capacities of Aquifer Solids Mlchael J. Barcelona+~tand Thomas R. Holm$ Department of Chemistry and Institute for Water Sciences, Western Michigan University, 1024 Trimpe, Kalamazoo, Michigan 49008, and Office of Environmental Chemistry, Water Survey Division, Illinois Department of Energy and Natural Resources, 2204 Griffith Drive, Champaign, Illinois 61820
rn
Measurements of the oxidation (i.e., of aqueous Cr2+) and reduction (i.e., of aqueous Cr2O7* and H202)capacities of aquifer solids and groundwater have been made on samples from a sand-and-gravel aquifer. The groundwater contributed less than 1%of the system oxidation or reduction poising capacity. Reduction capacities averaged 0.095, 0.111, and 0.136 mequiv/g of dry solids for oxic, transitional, and reducing E h conditions, respectively. Measured oxidation capacities averaged 0.4 mequiv/g of dry solids over the range of redox intensity conditions. These capacities represent considerable resistance to the adjustment of redox conditions even at uncontaminated sites. Hydrogen peroxide reduction by aquifer solid samples proceeds rapidly relative to microbially mediated decomposition. This study indicates the need for closer scrutiny of the predictability and cost effectiveness of attempts to manipulate redox conditions in poorly poised aquifer systems. 1. Introduction
Oxidation-reduction processes play a major role in the mobility, transport, and fate of inorganic and organic chemical constituents in natural waters. Therefore, the manipulation of redox conditions in natural and treated t Western Michigan University. 8 Illinois Department of Energy and Natural
0013-936X191/0925-1565$02.50/0
Resources.
water systems is assumed to be a common option for the control of contaminant concentrations. Both of these general themes are of great current concern with respect to the monitoring, prediction, and remediation of subsurface contaminant distributions in aquifers. There has been considerable research activity focused on the characterization of redox-potential or intensity (Eh) conditions in groundwater systems defined as the redox activity of dissolved chemical species. Early observations of significant Eh trends along groundwater flow paths led to hypotheses of successive redox zones characterized by the activity of specific thermodynamically favored electron acceptors (1-3). These redox zones may be classified as oxic (i.e., detectable dissolved 02),suboxic or postoxic (i.e., no detectable O2or sulfide, detectable Fe), and reducing (i-e.,detectable Fe and sulfide, no detectable 02).Further investigations correctly postulated that oxidation-reduction processes were mediated by natural microbial populations that catalyze electron-transfer reactions (4-7). More recent work noted considerable temporal and spatial variability in measured subsurface redox conditions (8)and that the succession of electron acceptors under oxic, suboxic, or reducing conditions was not strictly predictable by either chemical equilibrium calculations or platinum electrode measurements (9). The need exists to protect groundwater supplies from further contamination and, where feasible, to implement remedial actions to clean up existing contamination. Op-
0 1991 American Chemical Society
Environ. Sci. Technol., Vol. 25, No. 9, 1991 1585
tions under consideration to meet this need call for the disturbance of groundwater movement, chemistry, and microbiology via selective pumping and surface or in situ treatment strategies (10, 11). The effectiveness and the long-term consequences of such strategies on groundwater resources remain open questions. Of particular concern are the effects of subsurface injections of chemical oxidants/reductants and nutrient solutions to destroy or immobilize contaminant mixtures. In these situations, the reactivity of the total concentrations of redox-active species, i.e., system oxidation-reduction poising capacity (12), toward injected reagents will strongly influence both the dimensions of the "treated zone" and the potential immobilization or release of other chemical constituents (13, 14). This study was undertaken to determine the oxidation and reduction capacities of aquifer solids and groundwater and to identify the principal oxidized or reduced species contributing to the capacities over a range of redox conditions. 2. Experimental Section 2.1. Site Description and Sampling. Subsurface solid and groundwater samples were collected at two sites in Illinois. The first site, Sand Ridge State Forest, is a pristine environment with an unconfined, water table aquifer with a strong vertical redox gradient from the surface to a depth of over 30 m (8,16). The second site has similar regional hydrogeologic characteristics, but had been disturbed by the operation of a leaking anaerobic treatment impoundment for meat processing wastes. The groundwater redox intensities (Pt electrode potentials) ranged from +500 to +lo0 mV with depth at the pristine site. At the contaminated site, upgradient redox potentials ranged from +200 to +150 mV, while the downgradient groundwater exhibited potentials from +130 to -25 mV. The hydrology and chemical characteristics of these sites in the same aquifer have been described previously (8,15, 16). The subsurface solids were collected by either splitspoon or continuous coring techniques with the use of an auger drill rig in 1985 and 1987. The outer portions of the recovered cores were pared away with a stainless steel blade to minimize contamination from the corer and placed in clean, sterilized glass jars with tight lids. The solids were then stored in the dark at 4 "C until they were prepared for analytical measurements. The solids were prepared in several different ways and at different times to determine the effects of storage, air-drying, and grinding (to pass 100-mesh screens) on measurements of oxidizing and reducing capacities. Split samples were taken prior to analysis for determination of moisture content, organic and inorganic carbon (Coulometrics 120), total Fe and Mn (atomic absorption spectrophotometry), and extractable Fe(I1) and Fe(II1) content (17). Analytical details for the dissolved redox-active constituents (i.e., 02,H202,Mn, S2-, SO:-, NO3-, NO2-, NH,, and CHI) have been provided previously (8,16). All reagents were of ACS reagent grade and fresh standards or reagent solutions were made up on each day of analysis. 2.2. Oxidation Capacity. The oxidation capacities of the subsurface solids were measured by the oxidation of chromous ion (Cr(I1)). Stock standard solutions of acidic potassium dichromate (0.10 N) were first reduced by the addition of an excess of zinc amalgam. The resulting chromic (Cr(II1))solution was then further reduced to the chromous state with a slight excess of freshly diluted hydrogen peroxide, which had been standardized by iodometric titration. The final reduction step and all subsequent transfers or measurements were made in a glovebox purged 1588
Envlron. Sci. Technol., Vol. 25, No. 9, 1991
with 02-scrubbedN2 or in sealed glass cylindrical cuvettes. Duplicate wet solid samples (approximately 1 g) were placed in tared cuvettes and weighed in the glovebox prior to the addition of 5 mL of the Cr(I1) solution. The tubes were then sealed, agitated manually in the glovebox, and then left to react for 2 h. Standards were prepared by serial dilution of the Cr(I1) stock solution. They were also handled in the glovebox in sealed cuvettes. The sealed sample tubes containing the oxidized samples were then centrifuged (500g) for 1h to clear the supernatant solution for spectrophotometric measurement of unreacted Cr(I1). The oxidation capacity of the solids (in electron equivalents of Cr(I1) oxidized) was calculated relative to the series of Cr(I1) standards. The half-reactions below were used to assign respective oxidative equivalents to identifiable oxidized chemical species in the aquifer solids or groundwater. Mn02 + 4H+ + 2e- = Mn2+ + 2H20
(1)
+ 3H+ + e- = Fe2++ 3H20 C6H402+ 2H+ + 2e- = C6H602 SO12- + 9H+ + 8e- = HS-+ 4H20 NO3- + 10H+ + 8e- = NH4+ + 3H20 NOz- + 8H+ + 6e- = NH4++ 2H20 O2 + 4H+ + 4e- = 2H20 Q + 2H+ + 2e- = H2Q
(2)
Fe(OH),
(3) (4) (5) (6) (7) (8)
Where Q = C6H402and H2Q = C6H602representing quinone and hydroquinone. The organic carbon content of groundwater was assumed to be quinone, to provide an upper bound estimate of the oxidation capacity. The oxidation capacity of groundwater was calculated according to oxidation capacity = 4[O2] + 8[N03-] + 8[S0d2-] + 6[N02-] + TOC/72 (9) where brackets indicate molar concentrations. The oxidation capacities of the groundwater samples were calculated from the average analytical concentrations of oxidized forms of Fe (calculated by difference between total Fe and Fe(I1)) (181, Mn, and organic carbon (i.e., as quinone) measured over a period of 18 months of biweekly sampling and analysis (8). The oxidation capacity of aquifer solids was calculated according to oxidation capacity = (Fe(III)}+ 2(Mn(IV)}+ TOC/72 (10) where braces indicate solids concentrations in moles per gram. Manganese was assumed to be in the form of Mn02 2.3. Reduction Capacity. The reduction capacities of the subsurface solids were measured in duplicate by the reduction of acid dichromate in a modified closed-tube chemical oxygen demand (COD) procedure (19). The samples were prepared for the procedure by several methods. Replicate samples were either dried at 105 "C, dried and ground to e100 mesh with a stainless-steel mechanical grinder, or transferred wet to the reaction cuvettes directly from the storage jars. The weights of the wet samples were corrected for moisture content determined on split portions. Approximately 1.0 g dry weight of solids was placed in each of the reaction tubes. Then 5 mL of cold concentrated sulfuric acid/dichromate reagent was added and it was allowed to react at room temperature for 15 min or until CO, evolution ceased. The
tubes were then sealed and heated to 150 "C for 3 h. After being cooled, the tubes were centrifuged as above (2.2) and the unreacted Cr(V1) was determined colorimetrically (600 nm). Potassium hydrogen phthalate (KHP) standards were run simultaneously. Measured reduction capacities were corrected for chloride in the groundwater. The chloride corrections were found to be negligible. The reverse half-reactions for Fe, Mn, N, and S from section 2.2 were used to calculate reduction capacities from analytical data. For organic matter, the assumed halfreaction was C&@4 12H20 = 8C02 + 30H' + 30e- (11)
+
Phthalic acid (HP) was assumed to be a reasonable model compound for humic substances reduction capacity since its carbon, oxygen, and aromatic carbon contents closely match those for humics (20). The reduction capacity of groundwater was calculated according to reduction capacity = [Fe] + 2[Mn] S[S(-II)] + 8[N(-III)] + TOC/3 (12)
+
where brackets indicate molar concentrations, S(-11) is the sum of HS- and H2S, and N(-111) is the sum of NH4+and NH3(aq). The reduction capacity of aquifer solids was calculated according to reduction capacity = (Fe(I1))+ TOC/3 (13) where braces indicate solids concentration in moles per gram. 2.4. H202 Reduction Kinetics. Duplicate solid samples were reacted in slurries with 0.01 M aqueous hydrogen peroxide, a commonly proposed oxidant or oxygen source for in situ remediation (21,22). Approximately 10 g of fresh, wet solids was suspended in 100 mL of degassed M NaHC03solution and placed in Al foil covered beakers. The beakers were placed on a magnetic stirrer and a Pt electrode/calomel reference electrode pair was placed through an opening in the top of the vessel to record an initial E h reading. The peroxide solution was added at time zero. At selected time intervals, Eh readings were taken and aliquots of the slurry were removed for the determination of H202. H202determinations were done by the titanium(1V) chloride method (23). Controls were run in M NaHC03 solutions simultaneously with the solids samples in order to correct for trace reductants in the procedure and peroxide autodecomposition effects. Within experimental errors, the corrections were negligible. Some runs were made at 45 "C in thermostated vessels. 3. Results and Discussion 3.1. Oxidation and Reduction Capacities. The redox capacity measurements were accurate and reproducible with respect to KHP standards, Cr(I1) standards, and replicate samples. The mean accuracies of oxidation and reduction capacity determinations were 96% and 87%. The relative standard deviations (RSD) of oxidation capacity determinations were 0.3% for Cr(I1) standard run as samples ( n = 20) and 9.4% for samples (n = 25). The RSDs of reduction capacity determinations were 7 % for KHP standards ( n = 20) and 14% for samples ( n = 40). The results are shown in Tables I and 11. The oxidation capacities of aquifer solids averaged approximately 0.4 f 0.06 mequiv/g. There were no significant differences between those subsets of samples from oxic, suboxic, or reducing zones of the aquifer. In all cases, the oxidation of chromous ion was essentially complete within the 2-h reaction period with negligible change noted after 4-6 h. Replicate determinations on samples that were either stored at 4 "C or dried at 105 "C both showed slight
Table I. Average Oxidation Capacities and Oxidized Chemical Species'
parameter mV measd oxidn capac, mequiv/g N calcd oxidn capac, mequiv/g calcd/measd, % organic C, mg/g organic C oxidn capac, mequiv/g total extrctble Fe, mg/g extrctble Fe3+,mg/g Fe3+ oxidn capac, mequiv/g total Mn, mg/g Mn total oxidn capac, mequiv/g
study sites Beardstown (contam) Sand Ridge updown(uncontam) gradient gradient 470-300 0.374 (0.043)b 26 0.024 (0.024)b 6.4 0.422 0.008
220-170 225-50 0.376 0.410 (0.076)b (0.056)* 28 42 0.020 0.018 (0.024)b (O.O1O)b 4.4 5.3 0.610 0.648 0.012 0.012
0.184 0.154 0.003
0.052 0.030 0.001
0.115 0.075 0.001
0.394 0.014
0.218 0.008
0.111 0.004
'Oxidation capacities expressed on a dry weight basis. standard deviation.
One
Table 11. Average Reduction Capacities and Reduced Chemical Species'
parameter mV measd redctn capac, mequiv/g N calcd redctn capac, mequiv/g calcd/measd, % organic C, mg/g organic C redctn capac, mequiv/g total Fe, mg/g extrctble Fe, mg/g extrctble Fez+,mg/g Fez+ redctn capac, mequiv/g total Mn, mg/g Mn redctn capac, mequiv/g Eht
Sand Ridge
study sites Beardstown downU?gradient gradient 225-50 0.136 (0.012)b 42 0.106 (0.050)b 78 0.648 0.104
470-300 0.095 (0.066) 26 0.075 (0.050)b 79 0.422 0.068
220-170 0.111 (0.058) 28 0.102 (0.051)b 92 0.610 0.098
6.8 0.184 0.030