Chromium(III) Oxidation by Three Poorly-Crystalline Manganese(IV

Oct 11, 2012 - The Cr(III)-oxidizing capacity of three layered poorly crystalline Mn(IV)O2 phases, i.e. δ-MnO2, Random Stacked Birnessite (RSB), and ...
0 downloads 4 Views 2MB Size
Download PDF
Article pubs.acs.org/est

Chromium(III) Oxidation by Three Poorly-Crystalline Manganese(IV) Oxides. 1. Chromium(III)-Oxidizing Capacity Gautier Landrot,*,†,∥ Matthew Ginder-Vogel,† Kenneth Livi,‡ Jeffrey P. Fitts,§ and Donald L. Sparks† †

Plant and Soil Sciences Department, Delaware Environmental Institute, University of Delaware, 152 Townsend Hall, Newark, Delaware 19716, United States ‡ HRAEM Facility, Department of Earth and Planetary Sciences, Johns Hopkins University, Baltimore, Maryland 21218, United States § Department of Civil and Environmental Engineering, Princeton University, Engineering Quad, E430, Princeton, New Jersey 08544, United States S Supporting Information *

ABSTRACT: The Cr(III)-oxidizing capacity of three layered poorly crystalline Mn(IV)O2 phases, i.e. δ-MnO2, Random Stacked Birnessite (RSB), and Acid Birnessite (AB), was determined in real-time and in situ, using Quick X-ray Absorption Fine Structure Spectroscopy (Q-XAFS). The results obtained with this technique, which allows the measurement of the total amount of Cr(VI) produced in the system, indicated that the Cr(III) oxidation reaction had ceased between 30 min and 1 h under most experimental conditions. However, this cessation was not observed with a traditional batch technique, which only allows the measurement of Cr(VI) present in solution and thus neglects the amount of Cr(VI) that may be sorbed to Mn(IV)O2. This study also demonstrated that the Mn(IV)O2 phase oxidizing the highest amount of Cr(III), which is positively charged in solution, was the mineral featuring the most negatively charged surface. Also, the results indicated that the presence of Mn(II) and/or Mn(III) impurities inside the Mn(IV)O2 structure could enhance the mineral’s capacity to oxidize Cr(III). The information provided in this study will be useful in predicting the capabilities of naturally occurring Mn oxide minerals, which are similar to the three synthetic Mn(IV)O2 investigated, to oxidize Cr(III) to toxic and mobile Cr(VI) in the soil of contaminated sites. first few seconds of the reaction.1 In contrast to the traditional batch methods, this technique measures in real-time and in situ the total amount of Cr(VI) in the system, i.e. the amount of Cr(VI) present in solution and sorbed to the MnO2 surface. Therefore, this method is more suitable to measure the rates of Cr(III) oxidation by manganese oxide than the traditional batch techniques. The goal of this study is to determine the Cr(III)-oxidizing capacity of three synthetic manganese oxides, δ-MnO2, Random Stacked Birnessite (RSB), and Acid Birnessite (AB), using Q-XAFS. The three Mn(IV)O2 phases studied are known to be layered poorly crystalline minerals,10 similar to the naturally occurring biogenic manganese oxides.11,12 Some studies have compared the Cr(III)-oxidizing capacity of various MnO2 phases, whose degree of crystallinity and/or spatial structure (e.g., 3D tunnel vs 2D layer MnO2) were different from each other.6−8,13However, to our knowledge, no study has compared the Cr(III) oxidizing capacity of various poorly crystalline MnO2 phases with layered structures, even though a large portion of the natural Mn oxide minerals are poorly

1. INTRODUCTION Chromium can be found at very high concentrations in the soil of contaminated sites.1 For example, this heavy metal was measured at about 500 mM, which is about 2.7 million times higher than the chromium mean concentration in natural waters,2 in a soil near a leaking tank at a chrome-plating industrial site in Oregon, USA.3 Even if Cr(III), considered to be less toxic and mobile than Cr(VI),4 is the main form of Cr in a contaminated soil, its presence poses a potential threat to the environment−especially at high Cr(III) concentrations. Indeed, manganese oxides, which are ubiquitously found in soils,5 can rapidly oxidize Cr(III) to Cr(VI).4 Several studies have investigated the kinetics of Cr(III) oxidation by various manganese oxide phases.6−8 The rates measured in these studies were quantified from the amount of Cr(VI) in solution, using batch methods. These methods limited the accurate measurements of the Cr(III) oxidation kinetics, since Manceau and Charlet6 showed, after reacting Cr(III) with various MnO2 phases, that Cr(VI) produced in the system can be both present in solution and sorbed to the MnO2 surfaces. Alternatively, a phosphate buffer can be employed in traditional batch experiments to desorb the amount of Cr(VI) sorbed to Mn(IV)O2.9 Recently, Quick X-ray Absorption Fine Structure Spectroscopy (Q-XAFS) was employed to measure the “chemical” rates of Cr(III) oxidation by δ-MnO2 during the © 2012 American Chemical Society

Received: Revised: Accepted: Published: 11594

June 13, 2012 October 4, 2012 October 11, 2012 October 11, 2012 dx.doi.org/10.1021/es302383y | Environ. Sci. Technol. 2012, 46, 11594−11600

Environmental Science & Technology

Article

crystalline and/or nanocrystalline.12 Additionally, layered structure Mn-oxides are more abundant in the environment than tunnel structured phases.5 In this study, δ-MnO2, RSB, and AB are reacted with a high Cr(III) concentration (50 mM). Although this concentration is much higher than the Cr(III) concentrations employed in previous studies that investigated the kinetics of Cr(III) oxidation by MnO2 phase,6−8,14 it is comparable to concentrations found at Cr contaminated sites.3 Similarly, the pH values employed in our experiments (pH 2.5, 3, and 3.5) could be compared to those of some acidic chromium plumes found at contaminated sites, such as the one studied in Eary and Davis.15 The dense aqueous phase liquid (DAPL) plume investigated in this study, which was present in the soil of a former chemical manufacturing site in Massachusetts, USA, had a pH of 3 and a Cr concentration of 90 mM. Therefore, accurately measuring the kinetics of Cr(III) oxidation by δ-MnO2, RSB, and AB with Q-XAFS will be useful in predicting the Cr(III)-oxidizing capacity of naturally occurring Mn(IV)O2 phases present in the acidic soils of contaminated sites featuring high Cr concentrations.

MnO2), 280 mL of DI water, and 3.7 g of KCl were introduced in a 400 mL vessel. Therefore, after adding 20 mL of 1 M Cr(III) at t = 0 of the batch experiment, the vessel contained a 20 g/L manganese oxide suspension with 50 mM KCl and 50 mM Cr(III). The experiments were conducted at pH 2.5, pH 3, and pH 3.5. These experimental conditions were inserted in the chemical equilibrium program and associated thermodynamic database MINEQL+ (version 4.6), to predict the concentrations of Cr3+, Cr(OH)2+, and bulk Cr(OH)3 precipitate in solution as a function of pH (Figure 1). Figure 1 indicates that

2. MATERIALS AND METHODS Reactant and Standard Syntheses. All solutions were prepared with distilled dionized water, with a resistivity of 18.2 MΩ, produced from a Barnstead system. The Cr(III) stock solution was made from chromium nitrate Cr(NO3)3 (ACS grade), a few hours before conducting our experiments to minimize the effect of polymerization.16 Random Stacked Birnessite (RSB) was synthesized by slowly mixing two reactants, a 250 mL KOH solution at a concentration of 8 M, and a 250 mL Mn(II)Cl2 solution at 0.4 M.17 After bubbling both solutions with N2 for several hours, the Mn(II) solution was slowly titrated with KOH solution. The mixture was cooled down to approximately 5 °C for about one hour. Then, O2 was vigorously bubbled in the vessel for three hours. Acid birnessite (AB) was prepared by boiling a solution of 0.4 M KMnO4, which was titrated with a solution of 12 M HCl, with a 1/10 volumetric ratio of the acid over the permanganate solution.18 The δ-MnO2 phase was synthesized by slowly adding a 250 mL manganese(II) nitrate (Mn(NO3)2) solution at a concentration of 0.15 M to a 250 mL alkaline permanganate solution of 0.1 M KMnO4 and 0.2 M NaOH.19 The manganese oxides were synthesized no longer than a week before starting the experiments. All synthesized minerals were stored as a liquid at a suspension density of 80 g/L, at room temperature, and no longer than a week before conducting the experiments, since drying and storage may cause MnO2 phase transformations.20 The surface areas of the three minerals were measured by BET, using nitrogen as the adsorbate at 77.4 K.2 The X-ray powder diffraction (XRD) patterns and Scanning Electron Microscopy (SEM) images of the three minerals were compared to those reported in Villalobos et al.11 to verify that the Mn(IV)O2 phases were successfully synthesized. The XRD analyses were performed at the Institute of Energy Conversion, University of Delaware, using a Phillips/Norelco Diffractometer featuring a Cu−K(alpha) source, from 5 to 65 degree two-theta, with 0.05 degree steps. The SEM analyses were carried out at the Delaware Biotechnology Institute, University of Delaware, using a Hitachi 4700 FE-SEM, with a 3 kV accelerating voltage, a 4.5−9.3 mm free working distance, and a Secondary Electron Detector. Traditional Batch Experiments. One hundred milliliters of 80 g/L manganese oxide stock suspension (AB, RSB, or δ-

Figure 1. Concentrations of Cr3+, Cr(OH)2+, and bulk Cr(OH)3 precipitate as a function of pH, in a solution containing 50 mM Cr(NO3)3 and 50 mM KCl, calculated by MINEQL+ (version 4.6).

at pH 2.5, pH 3, and pH 3.5, the hydrated form of Cr(III) in solution is mostly Cr3+, and it is not expected to precipitate in bulk solution, since the pH limit at which Cr(III) starts to precipitate is pH 3.84 at this concentration. Since the addition of the concentrated chromium stock solution to the system resulted in an instantaneous variation in pH, a quantity of 4 M KOH or 2N HCl (volumes in S.I.) was manually added to the system at the beginning of the reaction, along with Cr(III), to reach the desired pH value of the experiment (pH 2.5, 3, or 3.5) within the first 10 s after t = 0 (t = 0 being the time when chromium is injected into the vessel). The H+ production, due to Cr(III) oxidation by manganese oxides, resulting in a constant decrease in pH throughout the reaction, was controlled by a pH Stat (Metrohm 751 Titrino) supplying 1 M KOH. After chromium was added at t = 0, 5 mL aliquots were taken from the vessel at different time intervals: 2 min, 5 min, 10 min, 20 min, 30 min, 60 min (1 h), and 360 min (6 h). Each aliquot was filtered through a 0.22 μm Millipore filter. The concentrations of Cr(III), Cr(VI), and Mnaqueous in the filtrate solution were measured with Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS) after chromatographic separation, using a G3268-80001 Agilent Column, in a 15 mM EDTA mobile phase at pH = 7. The batch experiments were performed in duplicate. Q-XAFS Batch Experiments. Q-XAFS experiments were conducted at beamline X18 B, at the National Synchrotron Light Source, Brookhaven National Laboratory, Upton, New York. The batch experimental setup was identical to the one used in a previous study where the chemical rates of chromium(III) oxidation were measured.1 In this present study, the batch vessel contained a suspension of Mn(IV)O2 at 20 g/L and 50 mM KCl mixed by magnetic stirring. A Cr(III) stock solution was injected at the beginning of the reaction to obtain a concentration of 50 mM in the vessel. The reaction 11595

dx.doi.org/10.1021/es302383y | Environ. Sci. Technol. 2012, 46, 11594−11600

Environmental Science & Technology

Article

Figure 2. (a) Percentage of Cr(VI) produced from normalized Q-XAFS XANES spectra at the Cr K-edge, which comes with a ±1−5% statistical error margin,1,22 using reaction conditions of 50 mM Cr(III), 20 g/L δ-MnO2 and AB at pH 2.5, 3, and 3.5. (b) Percentages converted to mM and normalized to the initial surface area of δ-MnO2 and AB.

was carried out at pH 2.5, 3, and 3.5. The H+ production, due to Cr(III) oxidation by manganese oxides, resulting in a constant decrease in pH throughout the reaction, was controlled inside the hutch by a pH Stat (Metrohm 751 Titrino) supplying 1 M KOH. The experiments were conducted at the Cr K edge (5989 eV). The gas mixture in I0 was 70% He and 30% N2, and the detuning was 30%. The data acquisition method collected 600,000 points/5 min using a Keithley current amplifier, a sixteen-channel VME analogdigital-converter (ADC), and custom programmed Linux based software. A five-minute Q-XAFS scan collected about 226 single XAFS spectra (1.3 s per spectrum), which were subsequently cropped into individual files. The Q-XAFS data acquisition started one minute before Cr(III) injection in the batch vessel via a syringe outside the beamline hutch1 and continued for four more minutes. After the first 5 min scan was finished, another scan automatically started; this cycle was repeated throughout the kinetic experiment, which lasted 72 min. Finally, a total of nine 5 min Q-XAFS scans were collected per kinetic experiment, since the Q-XAFS scanning system reset for a dead time of 179 s (3 min) between each 5 min scan. After cropping each scan to 226 single XAFS spectra, 116 single spectra collected around t = 0 and t = 4 min were averaged together to create one XAFS spectrum, representing t = 2 min of the reaction, since it is an average of spectra collected during the first four minutes. The same number of single XAFS spectra

was averaged to get merged spectra representing t = 10 min, t = 30 min, and t = 60 min. The data were processed similarly for the duplicate experiments conducted at the Mn K edge. Data were analyzed with the SIXPACK/IFEFFIT program.21 Collection of Q-XAFS spectra in the XANES region enables one to measure the total amount of Cr(VI) in the system, i.e. both in solution and that sorbed to the manganese oxide, by measuring the height of the Cr(VI) pre-edge feature in the XANES.1 Comparing this height to those of XANES standards of known Cr(VI) concentrations enables one to obtain the percentage of the total amount of Cr(VI) present in the system over the amount of Cr initially introduced in the batch vessel (i.e., the total amount of Cr present in the system). This approach is believed to be the most accurate method to quantify Cr(VI) in the system and comes with a 1−5% margin of error in the Cr(VI) percentage measured.22

3. RESULTS AND DISCUSSION Real-Time, in Situ Kinetics Experiments. The amount of Cr(VI) present in the batch reactor increases with time in all experiments (Figure 2 a), and the extent of Cr(III) oxidation is higher at low pH. Figure 2 (a) indicates that Cr(III) was rapidly oxidized during the first 30 min under all experimental conditions (e.g., 72% of 50 mM Cr(III) was oxidized by δMnO2 within 30 min at pH 2.5). Between 30 min and 1 h, 11596

dx.doi.org/10.1021/es302383y | Environ. Sci. Technol. 2012, 46, 11594−11600

Environmental Science & Technology

Article

Figure 3. a) Secondary electron images collected with SEM of unreacted δ-MnO2, AB, and RSB and b) XRD patterns of δ-MnO2, AB, and RSB.

however, Cr(III) was not oxidized, since the percentages of Cr(VI) measured during this period did not vary, except possibly for δ-MnO2 at pH 2.5. At these conditions, the Cr(VI) percentages measured at 30 min and 1 h were 72% and 77%, respectively (Figure 2 a). However, the difference between these two percentages is within the error margin of the Cr(VI) quantification method, i.e. ±1−5%.22 These results thus indicate that the Cr(III) oxidation is ceased between 30 min and 1 h under most experimental conditions. For the experiments at pH 3 and 50 mM of Cr(III) initially introduced in the system, 12 mM and 21 mM of Cr(VI) were

present in suspension when Cr(III) reacted for 30 min with AB and δ-MnO2, respectively (Figure 2 a). In comparison, 2 mM of Cr(VI) is present in the system after 30 min of Cr(III) reacting with RSB at pH 3 (not shown). The surface areas of the three unreacted minerals measured by BET were 78 m2/g, 84 m2/g, and 272 m2/g for RSB, AB, and δ-MnO2, respectively. The amount of Cr(III) oxidized in 30 min by each mineral phase divided by the initial mineral surface area is then 14.3 × 10−2 mM g/m2, 7.7 × 10−2 mM g/m2, and 2.6 × 10−2 mM g/m2 for AB, δ-MnO2, and RSB, respectively. Hence, if one assumes that the reactive surface site density is the same for the three poorly 11597

dx.doi.org/10.1021/es302383y | Environ. Sci. Technol. 2012, 46, 11594−11600

Environmental Science & Technology

Article

Figure 4. Kinetics of Cr(III) oxidation on δ-MnO2 and acid birnessite (AB) at pH 2.5, 3, and 3.5, using a batch technique and reaction conditions of 50 mM Cr(III) and 20 g/L δ-MnO2 or AB. The gray area represents the amount of Cr sorbed on Mn(IV)O2, which was measured by subtracting the amount of Cr(III) and Cr(VI) measured in solution at 2, 5, 10, 30, 60, and 360 min from the amount of Cr(III) introduced at t = 0 in the batch vessel (50 mM).

within the two minerals. Also, a difference in degree of crystallinity can be ruled out as well to explain the difference in Cr(III)-oxidizing capacity since AB is not more poorly crystalline than δ-MnO2. In fact, although both AB and δMnO2 are considered to be poorly crystalline mineral phases with layered structures,10,12 δ-MnO2 seems to be more poorly crystalline than AB based on observations from TEM,12 XRD, and SEM data (Figure 3 a). High-resolution transmission electron microscopy (HRTEM) analyses of unreacted AB and δ-MnO2 were reported in Livi et al.10 and Zhu et al.12 and also in part 2 of this study.23 Amorphous and crystalline sheet regions can be observed in the HRTEM images of δ-MnO2,12 while the HRTEM images of AB is characterized by sheets forming rosette shapes.10 These latter shapes are also observed in SEM images of AB and RSB, while at the same magnification, the sheets in δ-MnO2 seem to be too small and too disordered to be distinctly observed (Figure 3 a). Finally, the peaks featured in the XRD patterns of AB and RSB are sharper and better resolved than those in the XRD pattern of δ-MnO2

crystalline Mn(IV)O2 studied, RSB does not oxidize as much Cr(III) as δ-MnO2 and AB. Figure 2 (a) shows that δ-MnO2 oxidized more Cr(III) than AB. However, when the percentages reported in Figure 2 (a) are normalized to mM g/m2, by dividing the total Cr concentration (50 mM) by the initial mineral surface areas, and assuming that all poorly crystalline Mn(IV)O2 phases studied have the same reactive surface site density, AB oxidized more Cr(III) than δ-MnO2 under almost all experimental conditions (Figure 2 b). Therefore, although AB has a higher capacity to oxidize Cr(III) than δ-MnO2, the latter mineral phase produces the highest Cr(VI) amounts due to its high surface area. The crystal structures of these two layered minerals have been recently determined by Atomic Pair Distribution Function (PDF) analyses, which showed that both phases belong to the monoclinic C12/m1 space group with a disk-like shape.12 Therefore, since the two minerals share the same crystal structure, the higher capacity of AB than δ-MnO2 to oxidize Cr(III) is not due to differences in atomic spatial arrangements 11598

dx.doi.org/10.1021/es302383y | Environ. Sci. Technol. 2012, 46, 11594−11600

Environmental Science & Technology

Article

than δ-MnO2 when the amounts of Cr(VI) produced are normalized to the mineral surface area and assuming that all poorly crystalline Mn(IV)O2 phases studied have the same reactive surface site density, one can thus classify the factors controlling the ability of δ-MnO2, AB, and RSB to oxidize Cr(III) with the following order of importance: PZC of Mn(IV)O2 > oxidation states of Mn in MnO2> degree of crystallinity of MnO2. Batch Kinetics Experiments. Batch experiments were performed to complement the Q-XAFS data (Figure 4). The small differences in concentrations between each duplicate (e.g., a maximum difference of ∼5 mM in Cr(VI) concentrations between each duplicate) suggest that these measurements are quite reproducible. The concentrations of Cr(VI) in solution increased from 30 min to 6 h under almost all experimental conditions. These increases in Cr(VI) concentrations were the highest (∼10 mM) at low pH values for both the experiments conducted with AB and δ-MnO2. The results from Q-XAFS analyses indicated that Cr(III) oxidation ceased between 30 and 60 min of reaction under most experimental conditions (Figure 2). Therefore, the increase in Cr(VI) concentrations in solution observed in Figure 4 after 30 min under most employed experimental conditions could suggest that some amounts of chromate produced before 30 min were weakly bound to the MnO2 surface and progressively desorbed during the six-hour batch experiments. Without the results obtained from the Q-XAFS analyses, one could have concluded from the batch experiment results reported in Figure 4 that the Cr(III) kept on being oxidized by MnO2 for 6 h under almost all experimental conditions, since the concentration of Cr(VI) seemed to have progressively increased throughout most experiments, especially at low pH. This highlights the advantage of studying the kinetics of Cr(III) oxidation by MnO2 with the Q-XAFS technique over the traditional batch experiment method. The Cr(III) oxidation by Mn(IV)O2 ceased between 30 and 60 min under most employed experimental conditions (Figure 2). However, before that, the amounts of Cr(III) oxidized by AB and δ-MnO2 were high. Up to 28% and 77% of the initial 50 mM Cr(III) was converted to Cr(VI) by AB and δ-MnO2, respectively (Figure 2). This implies that the poorly crystalline Mn oxide phases commonly found in the environment, which are similar in nature to the three synthetic Mn(IV)O2 phases investigated in this study,10,12 may be capable of oxidizing large amounts of Cr(III), when the latter are present in soils at high concentrations, such as those at some Cr contaminated sites. Additionally, the results reported in this study and the literature8 indicate that the amount of Cr(III) oxidized by poorly crystalline Mn(IV)O2 phases and also other types of Mn oxide phases may depend on the mineral surface charge, which is constrained by the mineral PZC and the solution pH, the surface area and reactive surface site density of the Mn(IV)O2, and the number of Mn(II) and Mn(III) impurities in the Mn(IV)O2 phase. A crystalline Cr(OH)3 surface precipitate is believed to be responsible for the cessation of the Cr(III) oxidation by Mn(IV)O2 previously observed in former studies,26,27 which occurred in our experiments between 30 min and 1 h under all experimental conditions. The hypothesis of a Cr surface precipitate forming on AB and δ-MnO2 and being responsible for the cessation of Cr(III) oxidation observed in Figure 2 was tested by analyzing the Cr(III)-reacted MnO2 surfaces at the microscopic and molecular levels. The results of these analyses

(Figure 3 b), indicating a higher degree of crystallinity and/or particle sizes. Livi et al.10 measured by chemical titration and Electron Energy-Loss Spectroscopy (EELS) the average oxidation state of Mn in δ-MnO2, RSB, and AB. The oxidation states of Mn measured by chemical titration were 3.60, 3.89, and 3.96 for RSB, AB, and δ-MnO2, respectively. The oxidation states of Mn measured by EELS were 3.61, 3.89, and 4.04 for RSB, AB, and δ-MnO2, respectively. Since the three phases studied are Mn(IV) oxides with more or less Mn(II) and Mn(III) impurities,10 the average oxidation states of Mn reported in Livi et al.10 indicate that AB has more Mn(II) and Mn(III) impurities than δ-MnO2. A study that compared the Cr(III) oxidizing capacities of seven different manganese oxides found that the phases exhibiting the greatest Cr(III)-oxidizing ability were the minerals containing Mn(III) and Mn(II).8 Our results indeed indicate that AB has a higher Cr(III)-oxidizing capacity than δ-MnO2. However, the lowest amount of chromate was produced with RSB, which has the lowest average oxidation number according to Livi et al.10 The Point of Zero Charge (PZC) of this phase has been measured by electrophoretic mobility analyses in our laboratory and was equal to 3.8. The PZC of AB and δ-MnO2 are 1.824 and 2.4,25 respectively. The surface of RSB is thus mainly positive at the three pH values studied, which is not the case with AB and δMnO2. Therefore, Cr(III), which is positively charged, is more likely to sorb on the mineral surfaces of AB and δ-MnO2 than the one of RSB at pH 2.5, 3, and 3.5 from an electrostatic standpoint. This could explain the difference in the amounts of Cr(III) oxidized between RSB and AB. Weaver and Hochella8 showed that birnessite and hausmannite were the most powerful Cr(III) oxidizers among seven manganese oxides studied. The PZC of these two phases were below the pH at which the batch experiments were conducted,8 implying that the mineral surfaces were negatively charged. Additionally, lithiophorite, whose PZC is 6.9,7 was the least powerful Cr(III) oxidizer among the seven manganese oxides studied in Weaver and Hochella.8 Since this MnO2 phase was reacted with Cr(III) at pH 4.4, its mineral surface was mainly positively charged. Therefore, although not discussed by the authors, the differences in PZC between the seven mineral phases studied in Weaver and Hochella8 and the solution pH values employed in their experiments could thus partly explain the measured amounts of Cr(III) oxidized by each MnO2 investigated in this study. To summarize, our results and those reported in Weaver and Hochella8 indicate that several parameters simultaneously control the Cr(III) oxidizing capacity of a given manganese oxide phase. Weaver and Hochella8 found that the MnO2 phases exhibiting the greatest Cr(III) oxidizing ability are those containing Mn(II) and Mn(III). Based on their conclusions, RSB should have oxidized more Cr(III) than AB since the latter phase has a higher average oxidation number than the former phase.10 However, the opposite trend was observed during our experiments. Both AB and RSB have similar surface areas, XRD patterns (Figure 3 b), and exhibit similar morphologies, based on HRTEM images.10 Nevertheless, the PZC of RSB is higher than AB and above the highest pH used in our experiments (pH 3.5). Therefore, the mineral PZC value relative to the pH of the experiment seems to be the most important parameter to influence the Cr(III) oxidizing ability of δ-MnO2, AB, and RSB. As discussed previously, analyses by TEM, SEM, and XRD indicate that δ-MnO2 is more poorly crystalline than AB. The average oxidation number of Mn in AB is lower than the one in δ-MnO2. Since one can conclude that AB oxidizes more Cr(III) 11599

dx.doi.org/10.1021/es302383y | Environ. Sci. Technol. 2012, 46, 11594−11600

Environmental Science & Technology

Article

(11) Villalobos, M.; Toner, B.; Bargar, J.; Sposito, G. Characterization of the manganese oxide produced by pseudonomas putida strain MnB1. Geochim. Cosmochim. Acta 2003, 67 (14), 2649−2662. (12) Zhu, M.; Farrow, C. L.; Post, J. E.; Livi, K. J. T.; Billinge, S. J. L.; Ginder-Vogel, M.; Sparks, D. Structural study of biotic and abiotic poorly-crystalline manganese oxides using atomic pair distribution function analysis. Geochim. Cosmochim. Acta 2012, 81, 39−55. (13) Charlet, L.; Manceau, A. X-ray adsorption spectroscopy study of the sorption of Cr(III) at the oxide-water Interface. II. Adsorption, coprecipitation, and surface precipitation on hydrous ferric interface. J. Colloid Interface Sci. 1992, 148 (2), 443−468. (14) Fendorf, S. E.; Zasoski, R. J. Chromium(III) oxidation by δMnO2 1. Characterization. Environ. Sci. Technol. 1992, 26, 79−85. (15) Eary, E. L.; Davis, A. Geochemistry of an acidic chromium sulfate plume. Appl. Geochem. 2007, 22, 357−369. (16) Rotzinger, F.; Stünzi, H.; Marty, W. Early stages of the hydrolysis of chromium(III) in aqueous solution. 3. Kinetics of dimerization of the deprotonated aqua ion. Inorg. Chem. 1986, 25, 489−495. (17) Yang, D. S.; Wang, M. K. Syntheses and characterization of wellcrystallized birnessite. Chem. Mater. 2001, 13, 2589−2596. (18) McKenzie, R. M. The synthesis of birnessite, cryptomelane, and some other oxides and hydroxides of manganese. Mineral. Mag. 1971, 38, 493−502. (19) Gadde, R. R.; Laitinen, H. A. Studies of heavy metal adsorption by hydrous iron and manganese oxides. Anal. Chem. 1974, 46 (13), 2022−2026. (20) Ross, D. S.; Hales, H. C.; Shea-McCarthy, G. C.; Lanzirotti, A. Sensitivity of soil manganese oxides: Drying and storage cause reduction. Soil Sci. Soc. Am. J. 2001, 65, 736−743. (21) Ginder-Vogel, M.; Landrot, G.; Fischel, J.; Sparks, D. L. Quantification of rapid environmental redox processes using quick scanning x-ray absorption spectroscopy (Q-XAS). Proc. Natl. Acad. Sci. U.S.A. 2009, 106 (38), 16124−16128. (22) Peterson, M. L.; Brown, G. E.; Parks, G. A.; Stein, C. L. Differential redox and sorption of Cr(III/VI) on natural silicate and oxide minerals: EXAFS and XANES results. Geochim. Cosmochim. Acta 1997, 61, 3399−4413. (23) Landrot, G.; Ginder-Vogel, M.; Livi, K. J. T.; Fitts, J. P.; Sparks, D. L. Chromium(III) oxidation by three poorly-crystalline manganese(IV) oxides 2. Solid phase analyses. Environ. Sci. Technol. 2012, DOI: 10.1021/es302384q. (24) Matocha, C. J.; Elzinga, E. J.; Sparks, D. Reactivity of Pb(II) at the Mn(III, IV) (hydr)oxide-water interface. Environ. Sci. Technol. 2001, 35, 2967−2972. (25) Parikh, S. J.; Lafferty, B. J.; Sparks, D. L. An ATR-FTIR spectroscopic approach for measuring rapid kinetics at the mineral/ water interface. J. Colloid Interface Sci. 2008, 320, 177−185. (26) Fendorf, S. E. Oxidation and sorption mechanisms of hydrolysable metal ions on oxides surfaces. PhD Dissertation, University of Delaware, Newark, DE, 1992. (27) Fendorf, S. E.; Fendorf, M.; Sparks, D. Inhibitory mechanisms of Cr(III) oxidation by δ-MnO2. J. Colloid Interface Sci. 1992, 153 (1), 37−54.

are reported in a separate and complementary manuscript (part 2 of this study),23 which also aims to determine the binding mechanisms of Cr(III) and Cr(VI) on δ-MnO2, AB, and RSB.

■

ASSOCIATED CONTENT

S Supporting Information *

Quantity of HCl or KOH added at the beginning of the batch experiments, in mL. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +66 90 79 77 922. Fax: +66 29 42 81 06. E-mail: [email protected]. Present Address ∥

Department of Environmental Engineering, Kasetsart University, 10th floor, Building 14, 50 Ngamwongwan Road, Jatujak, Bangkok, 10900, Thailand. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank Gerald Hendricks and Caroline Golt, University of Delaware, for laboratory assistance and Brian McCandless, Institute of Energy Conversion, University of Delaware, for assistance in bulk XRD analyses. The National Synchrotron Light Source is supported by the US Department of Energy, Division of Material Sciences and Division of Chemical Sciences, under contract number DE-AC0298CH10886.

■

REFERENCES

(1) Landrot, G.; Ginder-Vogel, M.; Sparks, D. Kinetics of chromium(III) oxidation by manganese(IV) oxides using quick scanning x-ray absorption fine structure spectroscopy (Q-XAFS). Environ. Sci. Technol. 2010, 44, 143−149. (2) Sparks, D. L. Environmental Soil Chemistry; Academic Press: San Diego, CA, 2003. (3) Palmer, C. D.; Wittbrodt, P. R. Processes affecting the remediation of chromium-contaminated sites. Environ. Health Perspect. 1991, 92, 25−40. (4) Fendorf, S. E. Surface reactions of chromium in soils and waters. Geoderma 1995, 67, 55−71. (5) Post, J. E. Manganese oxide minerals: crytal stuctures and economic and environmental significance. Proc. Natl. Acad. Sci. U.S.A. 1999, 3447−3454. (6) Manceau, A.; Charlet, L. X-ray adsorption spectroscopy study of the sorption of Cr(III) at the oxide-water interface. I. Molecular mechanism of Cr(III) oxidation on Mn Oxides. J. Colloid Interface Sci. 1992, 148, 425−442. (7) Kim, J. G.; Dixon, J.; Chusui, C. C.; Deng, Y. Oxidation of chromium(III) to (VI) by manganese oxides. Soil Sci. Soc. Am. J. 2002, 66, 306−316. (8) Weaver, R. M.; Hochella, M. F. The reactivity of seven MnOxides with Cr3+aq: A comparative analysis of a omplex, environmentally important redox reaction. Am. Mineral. 2003, 88, 2016−2028. (9) James, B. R.; Bartlett, R. J. Behavior of chromiumin soils: VII. Adsorption and reduction of hexavalent forms. J. Environ. Qual. 1983, 12 (2), 177−181. (10) Livi, K. J. T.; Lafferty, B.; Zhu, M.; Shouliang, Z.; Gaillot, A.-C.; Sparks, D. Electron energy-loss safe-dose limits for manganese valence measurements in environmentally relevant manganese oxides. Environ. Sci. Technol. 2012, 46, 970−976. 11600

dx.doi.org/10.1021/es302383y | Environ. Sci. Technol. 2012, 46, 11594−11600