Environ. Sci. Technol. 1996, 30, 371-376
Chemical Interactions between Cr(VI) and Hydrous Concrete Particles C. H. WENG Department of Civil Engineering, Kaohsiung Polytechnic Institute, Ta-Hsu Hsiang, Kaohsiung County, Taiwan 84008, Republic of China
C. P. HUANG* AND HERBERT E. ALLEN Department of Civil & Environmental Engineering, University of Delaware, Newark, Delaware 19716
PETER B. LEAVENS Department of Geology, University of Delaware, Newark, Delaware 19716
PAUL F. SANDERS Division of Science & Research, New Jersey Department of Environmental Protection and Energy, Trenton, New Jersey 08625
Chemical interactions between Cr(VI) and concrete particles in synthetic groundwater solutions were studied. Results indicate that redox, adsorption, and precipitation are the three major reactions occurring at the concrete-water interface. The solution pH plays a significant role in determining the characteristics of chromium reaction behavior with concrete materials. In acidic systems, Cr(VI) species are removed from the solution by reduction reactions. At pH < 4.0, Cr(VI) reacts with aqueous Fe(II) ion, derived from the Fe-containing minerals of concrete materials, and is subsequently reduced to Cr(III). A simple stoichiometric relationship was observed for Cr(VI) reduction by concrete particles. As Cr(III) is not adsorbable in this pH region, it becomes released into the solution phase. The Cr(III) concentration is then governed by the solubility of Cr(OH)3(s) and/ or FexCr1-x(OH)3(s) precipitates as solution pH increases. In the pH range from 4.0 to 9.0, an adsorption reaction is mainly responsible for Cr(VI) removal.
Introduction In Hudson County, New Jersey, approximately 2.75 million tons of chromite ore processing residue (COPR) waste containing 2-5% chromium was generated from three chromate-processing facilities operating between 1905 and 1976. This residue has been used widely as fill material and has been deposited in wetlands, industrial sites, factories, and residential areas. Over 400 chromium* To whom correspondence should be addressed: telephone: 302831-8428; fax: 302-831-3640; e-mail address:
[email protected].
0013-936X/96/0930-0371$12.00/0
1996 American Chemical Society
contaminated sites were identified in this region (2). Weng et al. (3) have reported that at pH > 4.5, a great amount of Cr(VI) (=4.5 µmol/g) leaches from the chromium-contaminated soil (Cr-soil) derived from these residues, posing a serious environmental problem. Chromium salts can deteriorate concrete structures in chromium-contaminated areas (2, 4). Yellow crystals of chromate salts have been found on the surface of concrete blocks of many buildings (5). Remediation of these sites is necessary since they are hazardous to public health. Knowledge of the chemical interactions between chromium and concrete materials is needed to establish a cleanup strategy for the chromiumcontaminated concrete sites. To what extent the infected soil must be cleaned so as to prevent the chromium from being transported through the concrete block wall is one of the most urgent questions needing to be answered. However, little is known about the chemical interactions between chromium and concrete in the aqueous system. In the aquatic environment, chromium (Cr) primarily exists in two oxidation states: Cr(VI) and Cr(III). Public concerns with chromium are mostly related to hexavalent compounds, because these compounds are toxic to microorganisms, plants, animals, and humans (6-9). Unlike Cr(III), which behaves as a “hard” Lewis acid and can form insoluble Cr(OH)3(s) and CrxFe1-x(OH)3(s) precipitates, Cr(VI) is a Lewis base and is present in aqueous solution mainly as the anion. The current limit for soluble Cr in drinking water is 0.05 mg/L (10-6 mol/L) (10). While little or no information is available on chromium behavior with concrete, there has been extensive study on chromium behavior in soils. It is generally agreed that adsorption and desorption processes play a significant role in controlling the concentration of Cr species, and therefore their mobility, in soils. Much has been reported on chromate adsorption onto solids such as soil particles, ferric oxide, aluminum oxide, kaolinite, montmorillonite, coconut husk, and palm pressed fibers (4, 11-16). Generally, the adsorption of chromate on these solids occurs markedly under low-pH conditions. The extent of anionic chromate adsorption increases as pH decreases, reaches a peak value, and then decreases upon further increases in pH. Apparently, in the acidic pH region, competition between solution and surface protons disfavors the adsorption of anionic chromium, i.e., HCrO4-. As pH increases, the solid surface becomes progressively more deprotonated, which in turn discourages the formation of surface complexes. Surface complexation models have been successfully used to describe Cr(VI) ion adsorption onto some solid surfaces. Zachara et al. (14) and Davis and Leckie (15) have suggested that CrO42- ion adsorption onto amorphous iron oxy hydroxide and kaolinite can be described by outer-sphere complexation, while Hsia et al. (11) have suggested that Cr(VI) adsorption onto amorphous iron oxide was via innersphere complexation. Generally, outer-sphere complexes only adsorb weakly onto the soil particle surface. Thus, according to Davis and Leckie (15), the CrO42- ion is not strongly held onto soil particles and can be readily leached at high pH. Leaching of chromate into groundwater occurs readily under these conditions. The study of Cr(III) adsorption onto soil and soil components receives little attention because it is not perceived as an environmental
VOL. 30, NO. 2, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
371
FIGURE 1. XRD patterns of concrete particles. The patterns were analyzed on the basis of the search manual published by the Joint Committee on Powder Diffraction Standard (JCPDS), Swarthmore, PA, 1980.
hazard. In general, the adsorption characteristics of Cr(III) onto hydrous solids are similar to those of divalent metal ions (16-18). The extent of cation adsorption increases abruptly at a specific pH value, reaches a plateau, and then remains constant upon further increase in pH. Crawford et al. (19) and Charlet and Manceau (20) have shown that the sorption of Cr(III) by hydrous Fe oxides involve adsorption, surface precipitation, and coprecipitation processes. They have also reported that Cr(III) adsorption onto goethite or hydrous ferric oxide was brought by the formation of strong inner-sphere surface complexes (20). Redox reactions also affect aqueous concentrations of chromium. The oxidation of Cr(III) to Cr(VI) by Mn oxides is thermodynamically possible in soil and aquatic systems (25, 30-32). Under favorable conditions, the reduction of Cr(VI) to Cr(III) by organic matter and Fe(II) ion derived from iron-containing minerals will occur, which drastically decreases chromium mobility and toxicity through the formation of Cr(OH)3(s) and/or (Crx, Fe1-x)Cr(OH)3(s) precipitates. Under acidic conditions, Cr(VI) can be reduced to Cr(III) by activated carbon, Fe(II), and organic matter (21-29). Chromate reduction by ferrous ions is nearly spontaneous. Early and Lai have reported that Cr(VI) can be readily reduced to Cr(III) by Fe(II) ions dissolved from Fe(II)-containing minerals such as hematite and biotite in acidic media (24). They have also demonstrated that the rate of Cr(VI) reduction by hematite and biotite increases with decreasing pH values. Eary and Rai (24) have proposed a two-step process for chromate reduction by these minerals. First, the Fe(II) is released to solution due to dissolution or surface redox reaction. Second, the released Fe(II) is rapidly oxidized to Fe(III) by reaction with Cr(VI) present in the aqueous phase. The resulting Fe(III) was spontaneously reduced to Fe(II) by a coupled electron-cation transfer (CECT) reaction at the biotite-water interface. The Fe(II) ions so generated were then ready to react with Cr(VI), yielding further chromate reduction. The objective of this study is to investigate specific chemical interactions between Cr(VI) and concrete particles in aqueous systems. Parameters, such as pH and chromium concentrations, that may affect the chemical reactions were studied. Adsorption/desorption, redox, and precipitation processes, which affect the concentration of chromium in its interaction with concrete, were also investigated.
372
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 30, NO. 2, 1996
Materials and Methods Concrete Particles. Concrete blocks were purchased from Yardville Supply Co. (Yardville, NJ). The concrete blocks are prepared from ordinary Portland cement. After airdrying, the concrete block was crushed by a compressive testing machine (Tinius Olsen Testing Machine Co., Willow Grove, PA). The particles were then passed through ASTM standard sieves and used as adsorbent material without treatment unless otherwise noted. A size fraction between 0.18-0.5 mm was used for all experiments. The major chemical components of the concrete material, such as Si, Al, Fe, Ca, and Mg, are determined with a scanning electron microscope equipped with an energy-dispersive spectrometer (SEM-EDS, (Phillips 501 scanning electron microscope and Phillips EDAX 9100). The major crystal phases of the concrete particles are quartz (SiO2), calcite (CaCO3), gehlenite (Ca2Al2SiO7), orthopyroxene ((Fe, Mg)SiO3), and olivine ((Fe, Mg)2SiO4), which were identified by X-ray diffraction analysis (XRD) (Phillips semiautomatic X-ray diffractometer) (Figure 1). Other crystal phases such as CSH (calcium silicate hydrates), a hydration product of calcium silicate in cement, may also be present. The specific surface area of the concrete particles was 2.5 ( 0.1 m2/g, determined by the BET N2 gas adsorption method using a Model QS-7 Quantasorb surface area analyzer (Quantachrom Co., Greenvale, NY). The pHzpc of concrete particles was 5.9, as determined by a zeta-meter (Lazer Zee Meter, Model 500 Pen Kem Inc., Bedford Hills, NY) (Figure 2). Analytical Methods. Metals in the aqueous leachate, except for Fe and Cr, were determined with Perkin-Elmer Zeeman 5000 atomic absorption spectrophotometer (AA). Fe(II) and total Fe were analyzed by the 1,10-phenanthroline method at a wavelength of 510 nm (33). Fe(III) was determined from the difference between total Fe and Fe(II). Cr(VI) was analyzed by the 1,5-diphenylcarbohydrazide-chromate complex at 540 nm (34). Total Cr was determined by oxidizing the Cr(III) to Cr(VI) with potassium permanganate (27). Cr(III) was determined from the difference between total Cr and Cr(VI). Concentrations of NH4+, NO3-, and Cl- were determined by using an ionselective electrode. The concentration of SO42- was determined by the turbidimetric method at a wavelength of 420 nm (33). Preparation of Synthetic Groundwater. It was hypothesized that, during wet periods, Cr-contaminated
FIGURE 2. Zeta potential of hydrous concrete as a function of pH in different concentrations of electrolyte.
FIGURE 3. Major ions leached from Cr-soil used as the recipe for the synthetic groundwater preparation.
groundwater transports Cr(VI) ions into concrete block. Therefore, synthetic groundwater was prepared and used as the solution matrix throughout this study. Chromium-contaminated soil (Cr-soil) samples were collected from Liberty State Park, New Jersey. Physicochemical properties of Cr-soil have been reported (35). To determine the appropriate groundwater constituent that would be derived from this material, 50 g of Cr-soil was added to 50 mL of simulated rainwater (initial pH 4.3) in a plastic bottle. Preparation of simulated rainwater followed the recipe described by Weng et al. (35). The samples were placed on an orbital shaker (Model G33, New Brunswick Scientific, Edison, NJ) at a slow speed (50 rpm) at room temperature (25 °C) for 15 days. The steady-state pH was then recorded, and aliquots of the suspension were filtered through 0.45 µm Gelman filters. The concentrations of the major metal ions and other constituents in the supernatant were analyzed as described earlier. Results of major ion species leached from Cr-soil are shown in Figure 3. The high amount of leachable NO3may pose a severe groundwater contamination problem. The major source of nitrate is soil organic matter, as Weng et al. have reported an organic content of up to 9% in the Cr-soil (35). The concentration of Cr(III) in the leached solution, which was under the detection limit (0.01 mg/L), is controlled by the Cr(OH)3 precipitate and/or the adsorbed Cr(III) species onto soil. The high concentration of Ca(II) is attributed to the moderately soluble CaCO3 in the Crsoil (35). The amount of Cr(VI) leached, 4.4 µmol/g, is close to that reported by Weng et al. (3). They reported a concentration of 4.5 µmol/g and that chromate was largely leached from the Cr-soil. Synthetic groundwater was prepared on the basis of the amount of major ions, i.e., Ca(II), Mg(II), Na+, K+, NO3-, and Cl-, leached from the Cr-soil supplemented with an amount of HCO3- appropriate to maintain electroneutrality (Figure 3). Other ions such as Cr(VI) (0.13 mmol/L) and SO42- (0.01 mmol/L) also found in the leachate are not
FIGURE 4. Remaining concentrations of Cr(VI) and Cr(III) as a function of pH in the synthetic groundwater and 0.01 M NaNO3 solutions.
included in the recipe. The synthetic groundwater has an ionic strength of ca. 0.01 M and a pH of 7.8. Batch Experiments. The study of Cr(VI) reactions with concrete materials was conducted with batch equilibrium experiments in synthetic groundwater (except as mentioned otherwise) at room temperature (25 °C). In a series of plastic bottles containing 100 mL of synthetic groundwater, the initial pH was adjusted with 1 N NaOH and/or HNO3, and then a given amount of the sorted concrete particles was added. The samples were shaken on a shaker (Eberbach Co., Ann Arbor, MI) at 150 excursions per minute. After agitation for 24 h (the 24 h reaction time was found to be adequate to reach equilibrium), the final pH was recorded and an aliquot of the suspensions was taken and filtered (Gelman, 0.45 µm). The concentration of chromium in the supernatant was determined. The concentrations of other metals in the supernatant were analyzed with AA, except for Fe(II) and Fe(III) ions, which were analyzed by the colorimetric method, as described earlier.
Results and Discussion Effect of pH and Cr(VI) Concentration on Cr(VI) Removal. Figure 4 shows the residual Cr(VI) in the solution as a function of pH after the 24 h reaction time. Cr(III) was found below pH 4.5 and above pH 10.0, indicating that Cr(VI) was reduced to Cr(III) in this pH range. There was no apparent difference in the chromium concentration between synthetic groundwater and 0.01 M NaNO3 solution. This clearly indicates that the composition of synthetic groundwater does not affect the percentage of Cr(VI) removal. On the basis of a simple mass balance relationship, the total amount of Cr(VI) removed by concrete materials can be expressed as
Cr(VI)rem ) Cr(VI)added - Cr(VI)sol
(1)
where Cr(VI)added is the amount of Cr(VI) present in the solution at the beginning of the experiment, Cr(VI)sol is the Cr(VI) remaining in the solution after the experiment, and Cr(VI)rem is the Cr(VI) that was removed from the solution after the experiment. Figure 5 shows the percent of Cr(VI) removed at different initial Cr(VI) concentrations as a function of pH. The results indicate that the removal of Cr(VI) by concrete materials was greatly affected by pH and the Cr(VI) concentration. In general, the percent Cr(VI) removed increases with
VOL. 30, NO. 2, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
373
FIGURE 5. Cr(VI) removed determined as the difference between initial and final Cr(VI) solution concentrations.
FIGURE 6. (a) Cr(III) in the solution as a function of pH and different Cr(VI) concentrations added. (b) The ratio of Cr(III) produced to Cr(VI) reduced as a function of pH.
FIGURE 7. (a) Fe(II) leached as a function of pH and different Cr(VI) concentrations added. (b) The ratio of Fe(II) oxidized to Cr(VI) reduced as a function of pH.
decreasing pH and Cr(VI) concentration. Nearly 100% Cr(VI) removal was observed over the pH < 4.0 region. Redox Reaction. The observed concentrations of Cr(III), Fe(II), and Fe(III) are shown in Figures 6a, 7a, and 8a, respectively. At low pH values, a significant amount of Cr(III) was found in the solution, indicating that reduction of Cr(VI) had occurred. The behavior of Cr(III) produced as a function of pH is similar to that of Al(III), Fe(III), and Cr(III) leached from the Cr-soil reported by Weng et al. (3), in that the amount of metals leached increases sharply with decreasing pH. It appears that pH has a significant
374
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 30, NO. 2, 1996
FIGURE 8. (a) Fe(III) leached as a function of pH and different Cr(VI) concentrations added. (b) The ratio of Fe(III) produced to Cr(III) produced as a function of pH.
effect on the formation of Cr(III) from Cr(VI) reaction with the concrete material. At pH < 4.0, the amounts of Cr(III) and Fe(III) produced increase sharply with decreasing pH. The concentration of Cr(III) is governed by the solubility of (CrxFe1-x)(OH)3(s) precipitates. No soluble Cr(III) and Fe(III) were detected at neutral pH, which may be attributed to the formation of hydroxide precipitates Cr(OH)3(s) and Fe(OH)3(s) or other precipitates such as (CrxFe1-x)(OH)3(s). Cr(III) is produced primarily from the reduction of Cr(VI) by Fe(II) in acidic solution. Figure 7a shows the presence of Fe(II) in the concrete suspension. Up to 4 mg of Fe(II) per gram of concrete material was leached at pH 3.0 in the absence of Cr(VI). Under acidic conditions with increasing amounts of Cr(VI) added, from 0.5 to 10 mg/L in a 1 g/L concrete particle solution, the amount of Fe(II) leached decreases (Figure 7a) while Fe(III) increases (Figure 8a). Apparently, Cr(VI) is reduced to Cr(III) by Fe(II) derived from the concrete material. As shown in Figure 5, there is 100% Cr(VI) removal at ca. pH 6.0 when 0.5 mg/L concrete is added. It is interesting to note that the amount of Cr(VI) removed is equal to the amount of Cr(III) produced (Figure 6b). Stoichiometrically, in homogeneous solutions, the reduction of 1 mol of Cr(VI) requires 3 mol of Fe(II), yielding 1 mol of Cr(III) and 3 mol of Fe(III). In this study, the stoichiometric relationships among Cr(VI), Cr(III), Fe(II), and Fe(III) are presented in Figures 6b, 7b, and 8b. From Figure 6b, the molar ratio of Cr(III)/Cr(VI) is consistently equal to 1 below pH 3.5, indicating that all Cr(VI) added was reduced to Cr(III). Apparently, at pH less than 3.5, oxidation/reduction is the dominant reaction between Cr(VI) and the concrete interface and is responsible for Cr(VI) removal from the solution in this pH region. Aqueous Fe(II) is the major agent for Cr(VI) reduction. As depicted in Figure 7a, the Fe(II) concentration dropped markedly upon the addition of Cr(VI). It is interesting to note that no Fe(II) ions were found in the solution when Cr(VI) was added at amounts of 5 and 10 mg/L, individually. Apparently, all Fe(II) present in the solution was consumed for Cr(VI) reduction. An insufficient amount of Fe(II) for Cr(VI) reduction led to the remainder of a portion of Cr(VI) still in the solution after equilibration. The ratio of Fe(II) oxidized to Cr(VI) reduced, Fe(II)/ Cr(VI), and the ratio of Fe(III) produced to Cr(III) produced, Fe(III)/Cr(III), are approximately equal to 1, which is less than the theoretical value of 3 (Figures 7b and 8b). Two possible reactions may be responsible for the reduction of Cr(VI) under acidic conditions: coupled electron-cation
transfer (CECT) reactions and/or direct surface redox reaction. According to White and Yee (36), the CECT reactions occur on the surface of Fe(II)-containing silicates as in the following expression:
[Fe2+, (1/z)Mz+]silicate + Fe3+ w [Fe3+]silicate + Fe2+ + (1/z)Mz+ (2) where the brackets represent the solid surface and M denotes a nonreacting cation of charge z+. The sources of Fe(III) are either from the dissolution of the Fe(III) component of concrete or from the oxidation of aqueous Fe(II) in the presence of Cr(VI). Dissolution of Fe(II) oxide, Fe(II) silicates, and Fe(III) oxide minerals has been documented (24, 36-40). The CECT reaction can produce Fe(II) ions (eq 2), which reduces aqueous Fe(III) to Fe(II). This Fe(II) ion serves as a reducing agent for Cr(VI) reduction. Consequently, additional Cr(VI) was reduced, which decreases the Fe(II)/Cr(VI) ratio. As a result, the Fe(II)/Cr(VI) ratio is rendered less than the theoretical value of 3. Figure 1 clearly shows the presence of Fe(II)-containing silicates in the concrete. Direct surface redox reaction can also contribute Cr(VI) reduction. It is possible that, under acidic conditions, the adsorbed chromate reacts with a nearby surface Fe(II) site and becomes directly reduced on the concrete, without structural Fe(II) ever being released to solution. Because of the unfavorable adsorption of Cr(III) under acidic conditions, the direct surface redox reaction product, Cr(III), is released to the solution. Both reaction mechanisms, i.e., CECT and direct surface redox reactions, can alter the stoichiometry of Cr(VI) reduction by concrete particles under acidic conditions. Brown precipitates observed under acidic conditions are believed to be CrxFe1-x(OH)3(s) precipitate. Similar observations have also been reported in the Fe-Cr reaction system (24, 25, 35, 41). Under acidic conditions, the Cr(III) concentration in the concrete particle suspension is less than that from the solubility of pure Cr(OH)3(s), but follows the solubility of (CrxFe1-x)(OH)3(s). Sass and Rai (41) have reported a composition-dependent solubility equation to predict the aqueous Cr concentration in equilibrium with (CrxFe1-x)(OH)3(s). At neutral pH, the ratio of Fe(II) reduced to Cr(VI) oxidized cannot be determined because of the formation of both Fe(OH)3(s) and Cr(OH)3(s) precipitates. On the basis of the experimental results presented earlier, under acidic conditions, one can conclude that the removal of Cr(VI) by concrete particles is mainly due to the reduction reaction. However, further study on the kinetics of the reduction reaction and surface spectroscopic observations are needed in order to establish the reaction mechanism. Adsorption Reaction. In addition to the removal of Cr(VI) due to reduction, adsorption also plays an important role in the removal of Cr(VI) from solution by concrete particles. The total amount of Cr(VI) adsorbed can be determined from the following expression:
[Cr(VI)]ads ) [Cr(VI)]add - [Cr(VI)]sol - [Cr(III)]sol [Cr(III)]ads - [Cr(III)]precip (3) where [Cr(III)]sol is the Cr(III) remaining in the solution, [Cr(III)]ads is the Cr(III) adsorbed on the concrete particles, and [Cr(III)]precip is the primary hydroxide precipitates, i.e., Cr(OH)3(s) or CrxFe1-x(OH)3(s). Quantitatively, it is not possible to differentiate the amount of Cr(VI) adsorbed, Cr(III) adsorbed, Cr(VI) reduced,
FIGURE 9. Comparison of Cr(OH)3(s) and CrxFe1-x(OH)3(s) precipitates. Synthetic groundwater and a particle size of 0.1-0.5 mm were used in the experiments, except as mentioned in the figure.
FIGURE 10. Cr(VI) adsorbed as a function of pH and different Cr(VI) concentrations added. At low pH (
