Kinetic Studies on Sb(III) - American Chemical Society

Although antimony, a pollutant of prime interest to the U.S.. EPA (1) and the E.U. (2), occurs widely ..... Council Directive 76/. 464/EEC of 4 May 19...
1 downloads 0 Views 116KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 2843-2848

Kinetic Studies on Sb(III) Oxidation by Hydrogen Peroxide in Aqueous Solution F R A N C¸ O I S Q U E N T E L , † M O N T S E R R A T F I L E L L A , * ,‡ CATHERINE ELLEOUET,† AND CHRISTIAN-LOUIS MADEC† Laboratoire de Chimie Analytique, UMR-CNRS 6521, Universite´ de Bretagne Occidentale, 6, Avenue V. Le Gorgeu, BP 809, F- 29285 Brest, France, and Department of Inorganic, Analytical and Applied Chemistry, University of Geneva, Quai Ernest-Ansermet 30, CH-1211 Geneva 4, Switzerland

Knowledge of antimony redox kinetics is crucial in understanding the impact and fate of Sb in the environment and optimizing Sb removal from drinking water. The rate of oxidation of Sb(III) with H2O2 was measured in 0.5 mol L-1 NaCl solutions as a function of [Sb(III)], [H2O2], pH, temperature, and ionic strength. The rate of oxidation of Sb(III) with H2O2 can be described by the general expression: -d[Sb(III)]/dt ) k[Sb(III)][H2O2][H+]-1 with log k ) -6.88 ((0.17) [k: min-1]. The undissociated Sb(OH)3 does not react with H2O2: the formation of Sb(OH)4- is needed for the reaction to take place. In a mildly acidic hydrochloric acid medium, the rate of oxidation of Sb(III) is zeroth order with respect to Sb(III) and can be described by the expression -d[Sb(III)]/dt ) k[H2O2][H+][Cl-] with log k ) 4.44 ((0.05) [k: L2 mol-2 min-1]. The application of the calculated rate laws to environmental conditions suggests that Sb(III) oxidation by H2O2 may be relevant either in surface waters with elevated H2O2 concentrations and alkaline pH values or in treatment systems for contaminated solutions with millimolar H2O2 concentrations.

trations of reduced species are found in oxic waters and oxidized species in anoxic waters (6). Several authors have invoked the kinetic stabilization of thermodynamically unstable species to explain these observations, without providing experimental evidence to support their claims. Very few systematic studies exist on Sb(III) oxidation kinetics in natural water conditions. Published observations seem to indicate that, at natural antimony levels, Sb(III) is oxidized to Sb(V) (7-9). Cutter (10) estimated the Sb(III) oxidation rate by using the depth profiles for antimony species in the upper 100 m of the Black Sea. He calculated a pseudo-firstorder rate constant of 0.008 day-1 (residence time ) 125 days). This rate includes all forms of removal (i.e., Sb(III) oxidation but also Sb(III) scavenging by sedimenting particles). Recently, the same authors concluded that Sb(III) probably has longer residence times in seawater because this value was calculated at the suboxic-oxic interface of the Black Sea where the presence of manganese and iron oxides is likely to increase the oxidation rate (11). Indeed, Belzile and co-workers (12), in a study on the oxidation of Sb(III) in the presence of natural and synthetic iron and manganese oxyhydroxides, showed that Sb(III) can rapidly be oxidized to Sb(V) by both compounds. For this study, Sb(III) oxidation by hydrogen peroxide was studied over a wide range of pH, ionic strength, and temperature conditions. Hydrogen peroxide was measured in a variety of natural surface waters (13-20) at concentrations exceeding 10-7 mol L-1 and is thought to play a key role in the redox chemistry of trace elements in aquatic environments. Hydrogen peroxide is produced primarily by photochemical processes through the interaction of UV radiation and organic matter (13). Rainwater is known to have high concentrations of hydrogen peroxide (19, 21-23), and its contribution to surface water hydrogen peroxide concentrations may be significant in some systems (24, 25). Studies have been published concerning the oxidation of iron (2640), copper (37, 41-44), chromium (45, 46), and arsenic (4751) by hydrogen peroxide. No study exists on Sb(III) oxidation by this natural oxidant.

Experimental Procedures Introduction Although antimony, a pollutant of prime interest to the U.S. EPA (1) and the E.U. (2), occurs widely in the environment as a result of natural processes and human activity, it has barely been touched upon by environmental studies. Antimony production and use have steadily increased, and more importantly, its uses have changed over the years (3). Traditionally, antimony was used in lead-antimony alloys. Bulk secondary antimony can be recovered as antimonial lead, most of which is regenerated and then consumed by the battery industry. Nowadays, the form of Sb most commonly used is Sb2O3 (flame retardants, catalyst, ceramics, glass, pigment) that cannot be recycled and is released into the environment. The behavior of antimony in the environment and in water and wastewater treatment systems depends to a great extent on its oxidation state. Antimony speciation in natural waters is more complicated than equilibrium thermodynamic calculations predict (4, 5). In particular, significant concen* Corresponding author phone: 41-22-3796046; fax: 41-223796069; e-mail: [email protected]. † Universite ´ de Bretagne Occidentale. ‡ University of Geneva. 10.1021/es035019r CCC: $27.50 Published on Web 04/07/2004

 2004 American Chemical Society

The reaction kinetics were monitored by following the decay of Sb(III) during the course of the oxidation. Sb(III) was measured by DPASV following the method described in ref 9. All voltammetric measurements were performed with a computer controlled electrochemical system Autolab PGSTAT12/GPES (General Purpose Electrochemical System Version 4.8, Eco Chemie). A static mercury drop electrode (SMDE) Metrohm model 663 VA with a mercury drop size of 0.52 mm2 was used. All potentials were referred to an Ag/AgCl, 3 mol L-1 KCl, reference electrode. The counter electrode was a platinum wire. For all measurements, the differential pulse mode was used. The quartz voltammetric cell was thermostated at 25 ((0.1) °C during the experiments to maintain reproducible conditions. The analytical method used involves Sb(III) measurements being made at pH 2. Prior to the kinetic runs, tests showed that the presence of H2O2 had no effect on the measured signal. Oxidation rates were studied in NaCl solutions over a wide range of pH, ionic strength, and temperature conditions. Reactions were carried out in a 500 mL glass reaction vessel thermostatically controlled to (0.1 °C and purged continuously with N2. Reactions were carried out at 25 °C unless stated otherwise. The pH was measured throughout the experiments with an Orion combination electrode. VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2843

FIGURE 1. Oxidation of Sb(III) by H2O2 in 0.5 mol L-1 NaCl, 4.2 × 10-4 mol L-1 H3BO3 at 25 °C, and different pH values (2, 8.10; 1, 8.62; b, 8.96; (triangle pointing right), 9.10; (triangle pointing left), 9.22; [, 9.55; 9, 9.69). [Sb(III)] ) 0.410 µmol L-1; [H2O2 ] ) 50 µmol L-1. All solutions were prepared with 18 MΩ Millipore water, and all chemicals were of analytical reagent grade, except hydrochloric acid, sodium hydroxide, and boric acid that were of Suprapur grade. Standard solutions of Sb(III) were prepared by diluting SbCl3 Fisher Scientific standard solution for atomic absorption spectrometry (1 g L-1 in 20% HCl). A 5 × 10-2 mol L-1 H2O2 solution was prepared daily from 30% H2O2 (Merck Suprapur). The concentration of the stock solution of hydrogen peroxide was periodically determined by KMnO4 titration. In the experiments in the nonacidic medium, the pH of the samples was adjusted with a borate buffer and small amounts of dilute HCl or NaOH solutions to fix the desired pH for the reaction. Speciation calculations were performed with the JESS (Joint Expert Speciation System) modeling package, Version 6.4 (52). This computer software is designed to process thermodynamic data for chemical reactions to automatically achieve thermodynamic consistency. Previously, the available thermodynamic data had been critically evaluated and entered into a thermodynamic database. See ref 5 for details of the method used.

FIGURE 2. Oxidation of Sb(III) by H2O2 for three different concentrations of Sb(III) (1, 0.205 µmol L-1; 2, 0.410 µmol L-1; b, 0.820 µmol L-1) in 0.5 mol L-1 NaCl, 4.2 × 10-4 mol L-1 H3BO3 at 25 °C, pH ) 8.5. [H2O2] ) 200 µmol L-1.

FIGURE 3. Values of log k′ for the oxidation of Sb(III) by H2O2 as a function of the H2O2 concentration (mol L-1) in 0.5 mol L-1 NaCl, 4.2 × 10-4 mol L-1 H3BO3 at 25 °C, and two pH values (1, 8.96; 2, 9.55). [Sb(III)] ) 0.410 µmol L-1. k′ is defined in eq 2.

Results and Discussion Oxidation in a Nonacidic Medium. The overall rate equation for the oxidation reaction of Sb(III) by H2O2 may be expressed as

-d[Sb(III)]/dt ) kexp[Sb(III)]m[H2O2]n

(1)

where m and n are the order of the reaction with respect to the concentrations of Sb(III) and H2O2, respectively, and kexp is the corresponding kinetic rate constant. Under pseudo-first-order conditions (excess of H2O2), k′ ) kexp[H2O2]n and eq 1 becomes

-d[Sb(III)]/dt ) k′[Sb(III)]m

(2)

where [Sb(III)] is the concentration of Sb(III) at any time. Figure 1 shows some typical oxidation runs for Sb(III) with H2O2 at different pH values. Under our experimental conditions, values of ln[Sb(III)] were linear as a function of time, and similar slopes, k′ (min-1), were obtained when different initial concentrations of Sb(III) were used at constant pH, temperature, ionic strength, and H2O2 concentrations (Figure 2). A value of log k′ ) -2.19 ((0.08) [k′: min-1] (mean value for three different Sb concentrations) was calculated in 0.5 mol L-1 NaCl at 25 °C and pH ) 8.5. Hydrogen peroxide concentration was 200 µmol L-1. A first-order dependence of the reaction was also observed with respect to H2O2. Kinetic runs were performed in 0.5 mol L-1 NaCl solutions at pH 8.96 and 9.55 containing 0.410 µmol L-1 Sb(III). As shown in Figure 3, values of log k′ can be 2844

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

FIGURE 4. Values of log kexp for the oxidation of Sb(III) by H2O2 as a function of pH in 0.5 mol L-1 NaCl and 4.2 × 10-4 mol L-1 H3BO3 at 25 °C for different initial concentrations of H2O2 (µmol L-1). [Sb(III)] ) 0.410 µmol L-1. kexp is defined in eq 1. described as a function of log[H2O2] by linear equations with slopes 1.26 ( 0.07 (r2 ) 0.986) and 0.99 ( 0.06 (r2 ) 0.992) at pH values 8.96 and 9.55, respectively. The effect of pH on the reaction rate was studied for the oxidation of 0.410 µmol L-1 Sb(III) for different initial concentrations of H2O2. Results are given in Figure 4. No oxidation was detected below pH 7. Kinetic results as a function of pH have been fitted to (Figure 4)

log kexp ) (-6.42 ( 0.45) + (0.94 ( 0.05)pH (r2 ) 0.985) (3) The lack of oxidation below pH 7 and the effect of pH on the overall constants can be analyzed as a function of Sb(III) speciation. In the neutral pH range, Sb(III) is present in solution as Sb(OH)3, which behaves as a weak acid with the

FIGURE 5. Values of log kexp for the oxidation of Sb(III) by H2O2 in 0.5 mol L-1 NaCl and 4.2 × 10-4 mol L-1 H3BO3 at 25 °C as a function of ionic strength at pH ) 8.20 (1) and 8.98 (2). [Sb(III)] ) 0.410 µmol L-1. The smooth curves were calculated by accounting only for the effect of the ionic strength on the changes of the speciation of Sb(OH)3 according to eq 4. kexp is defined in eq 1.

FIGURE 6. Values of log k′ for the oxidation of Sb(III) by H2O2 as a function of temperature in 0.5 mol L-1 NaCl and 4.2 × 10-4 mol L-1 H3BO3. [Sb(III)] ) 0.410 µmol L-1; [H2O2] ) 200 µmol L-1. k′ is defined in eq 2.

following dissociation equilibrium:

Sb(OH)3 + H2O S Sb(OH)4- + H+

(4)

with log K ) -11.447 at 0.5 mol L-1 and 25 °C. See ref 5 for a detailed description of the approach used in the calculation of this thermodynamically consistent value from published equilibrium constants and ref 53 for details on the ionic strength correction method applied. The overall rate constant can be expressed as a function of pH as

kexp ) koRSb(OH)3 + k1RSb(OH)4-

(5)

where Ri are the molar fractions of Sb(III) species. Eq 5 can be expressed as

kexp RSb(OH)3

K ) ko + k1 + [H ]

(6)

leading to ko ) (-0.07 ( 6.10) L mol-1 min-1 and k1 ) (2.66 ( 0.08) × 104 L mol-1 min-1. The value of ko is clearly close to zero. This result indicates that Sb(OH)3 does not react with H2O2 and that, for the oxidation to take place, the formation of Sb(OH)4- is needed. This is confirmed by the lack of oxidation observed at pH < 7, where speciation calculations show that Sb(OH)4- is not formed to any significant extent. However, the presence of minute amounts of Sb(OH)4- is enough for the oxidation reaction to occur. For instance, speciation calculations show that Sb(OH)4- only represents 2% of the total amount of Sb(III) at pH 9.7 and 4% at pH 10.0. Since kexp shows a dependence on pH, eq 1 can be generalized as

-d[Sb(III)]/dt ) k[Sb(III)][H2O2][H+]-1

(7)

where k is the corresponding kinetic rate constant. Combining the results from all experiments (29 independent kinetic runs with total antimony concentrations ranging from 0.205 to 0.820 µmol L-1, total hydrogen peroxide from 5 × 10-5 to 2 × 10-4 mol L-1 and pH from 8.0 to 9.5) gives a log k ) -6.88 ((0.17) [k: min-1] (Table 1 in Supporting Information). The effect of ionic strength (I ) 0.01-3.0 mol L-1) was studied in NaCl solutions with [Sb(III)] ) 0.410 µmol L-1 at 25 °C and pH values 8.20 and 8.98 (Figure 5). Measured oxidation rates significantly increased with the increase in the ionic strength and do not change as might be expected with the variation of the corresponding deprotonation equilibria with ionic strength (continuous lines in Figure 5).

FIGURE 7. Oxidation of Sb(III) by H2O2 for different concentrations of Sb(III) (2, 1.642 µmol L-1; 1, 0.821 µmol L-1; [, 0.410 µmol L-1; (triangle pointing right), 0.164 µmol L-1; b, 0.082 164 µmol L-1) in 0.5 mol L-1 NaCl and 0.25 mol L-1 HCl at 25 °C. [H2O2] ) 500 µmol L-1. To test whether the observed trend was the result of the formation of new antimony chloride-containing species in the presence of higher concentrations of chloride ions, measurements were repeated in solutions 0.5 mol L-1 NaCl and 1 mol L-1 NaNO3 instead of 1.5 mol L-1 NaCl. However, similar results were obtained (log k′ ) 2.29 and 2.35, respectively). The possible catalytic effect of any trace metal impurities present in the NaCl salt on the hydrogen peroxide action used was also checked and proved not to be the cause of the observed dependence. See values in Table 2 in the Supporting Information. The effect of the temperature on the oxidation of Sb(III) with H2O2 was examined in 0.5 mol L-1 NaCl solutions, pH ) 8.98, with [Sb(III)] ) 0.410 µmol L-1 and [H2O2] ) 200 µmol L-1 at temperatures ranging from 15 to 40 °C. Conventional plots of log k′ against 1/T gave a value of 50 ( 2.5 kJ mol-1 for the apparent activation energy (Figure 6). A general equation giving the variation of log k with ionic strength and temperature can be fitted to

log k ) 1.26 + 1.01xI - 0.22I - 2603/T

(8)

Oxidation in a Mild Acidic Medium. The oxidation of Sb(III) was studied in solutions containing low concentrations of hydrochloric acid. An approach based on experimentation and calculation, similar to the one used in nonacidic media, was adopted. As observed in Figure 7, the variation of the concentration of Sb(III) as a function of time behaved in a linear fashion, and parallel lines were obtained for different initial Sb(III) concentrations at the same temperature, ionic strength, and HCl and H2O2 concentrations. These results demonstrate that the reaction is zeroth order with respect to Sb(III), with k′ ) 2.12 ((0.29) mol L-1 min-1. VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2845

FIGURE 8. Values of log k′ for the oxidation of 0.410 µmol L-1 Sb(III) by H2O2 as a function of the H2O2 concentration (mol L-1) in 0.5 mol L-1 NaCl and 0.25 mol L-1 HCl at 25 °C. k′ defined in eq 2.

FIGURE 9. Values of log k′ for the oxidation of 0.410 µmol L-1 Sb(III) by H2O2 as a function of the chloride concentration (mol L-1) in 0.25 mol L-1 HCl at 25 °C. [H2O2] ) 500 µmol L-1. k′ defined in eq 2. A first-order dependence of the reaction rates was found with respect to H2O2. Kinetic runs were performed in 0.5 mol L-1 NaCl solutions with 0.410 µmol L-1 Sb(III) and 0.25 mol L-1 HCl. Values of log k′ can be described as a function of log[H2O2] by a linear equation with a slope of 1.16 ( 0.04 (r2 ) 0.998) (Figure 8). The oxidation rate constant was found to be first-order with respect to the chloride concentration (Figure 9), with a slope of 1.01 ( 0.04 (r2 ) 0.978). To check whether the presence of chloride was needed for Sb(III) oxidation to occur, experiments were performed with perchlorate instead of chloride as the inert electrolyte, in the same conditions of pH and ionic strength. No oxidation was observed even when the acidity was increased in perchlorate media. As shown in Figure 10a, the reaction rate increased when the amount of HCl present in the solution increased. However, it is not possible to isolate the effect that increasing proton and chloride concentrations has on the reaction rates of such experiments. The effect of acidity on the reaction rate could be better evaluated through a comparison of the results obtained from experiments in which the same total amount of chloride was present but the pH was different (i.e., solutions containing 0.5 mol L-1 HCl (JESS calculated pH ) 0.63) and 0.25 mol L-1 HCl and 0.25 mol L-1 NaCl (JESS calculated pH ) 0.33) (Figure 10b)). Such a comparison demonstrates the positive effect of the acidity on the reaction rate and infers a first-order dependence of the reaction rate on the proton concentration. The fact that the reaction rate is zeroth order with respect to Sb(III) implies a reaction rate-determining step that is independent of antimony. A possible oxidation mechanism would be the formation of chlorine by hydrogen peroxide (slow reaction under the conditions of our experiments) 2846

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

FIGURE 10. Oxidation of Sb(III) by H2O2 at 25 °C (a) for different concentrations of HCl ((triangle pointing right), 0.01 mol L-1; [, 0.10 mol L-1; 1, 0.25 mol L-1; 2, 0.50 mol L-1) in 0.5 mol L-1 NaCl; (b) in 0.5 mol L-1 HCl (2) and 0.25 mol L-1 HCl in 0.25 mol L-1 NaCl solutions (1), respectively. [Sb(III)] ) 0.410 µmol -1; [H2O2] ) 500 µmol L-1. In panel b, SbCl4- accounts for 83% of total initial Sb(III), SbCl63for 14%, and SbCl3 for 2% in both solutions. followed by the rapid oxidation of Sb(III) by the chlorine formed. The production of chlorine by hydrogen peroxide in acidic chloride-containing solutions can be expected given the free energies of the substances involved and has been known for a long time (54-57). This reaction mechanism would be consistent with (i) the observed first-order dependence with respect to chloride and H2O2 concentrations and (ii) the absence of oxidation when chloride ions are not present. However, formation of chlorine has only been described in experiments performed at high acidities (i.e., ref 58) and studies where hydrogen peroxide acts as an oxidant in mild hydrochloric acid solutions do not mention chlorine formation (59, 60). We have qualitatively checked that Sb(III) oxidation by chlorine is a fast reaction, but we have not tried to prove that chlorine is effectively formed in our kinetic solutions. The speciation of Sb(III) in initial solutions has been calculated, and the results are shown in Figures 9 and 10a and in the caption of Figure 10b. SbCl4- is the main species present in most experimental solutions. Higher chloride concentrations favor the formation of SbCl63-. Although calculated using the best existing equilibrium constant values, these speciation results must be viewed with some caution. It is well-known that any calculated species distribution depends on the species considered to be present in the system, as well as on the reliability of the corresponding equilibrium constant values. In the case of the Sb(III)chloride system, published equilibrium constants are not particularly reliable. Most of the constant values have been obtained at very high acidities, thus making extrapolation to low ionic strengths problematic. Moreover, the few log K

values determined in relatively more dilute conditions are subject to a degree of uncertainty in determining the species stoichiometry because of the experimental techniques used. In particular, the formation of the SbCl63- species is questionable. The possible presence of species such as Sb(OH)Cl3- in diluted HCl solutions cannot be excluded (61), but reasonable formation constants do not exist for such species and could not be included in the speciation calculations. See ref 5 for a detailed discussion of the reliability of the constant values used and the references of all original values. Although significant formation of Sb-chloride species is observed when oxidation occurs, it is impossible to ascertain whether Sb(III) speciation plays any role in the fast oxidation stage. When noncontaining chloride species predominate, as is the case in the experiment [HCl] ) 0.01 mol L-1 in Figure 10a (where 78% of the initial Sb(III) is present as hydroxilated species), reaction kinetics is very slow, but this may be due to the low production of chlorine at this pH value and not to the lack of formation of chloride-containing species. There is little evidence available concerning the nature of the species formed when Sb(V) is present in diluted HCl solutions. Antimony(V) is present as the anionic Sb(OH)6form over a wide pH range. The pK of Sb(OH)6- is 2.72 at infinite dilution (62); thus, formation of Sb(OH)5 would be expected at the pH range of our study in the absence of chloride. Formation of SbCl6- has been reported in concentrated HCl solutions (i.e., 9-12 mol L-1 HCl (63); 6 mol L-1 HCl (64)), but the Sb(V) species formed at lower acidities and chloride concentrations remain unidentified. Formation of SbCl6- would be extremely attractive because of the possible formation of transition species containing chloride bridges between SbCl6- and SbCl3 or SbCl4- where electron transfer would be facilitated. Chloride would act as a common ligand facilitating the transition state. However, Sb(V) is probably extensively hydrolyzed (65) in our solutions. Studies published on the Sb(III)-Sb(V) exchange reactions in HCl solutions (66-70) do not help to explain the oxidation mechanism under our conditions because they were all performed in extremely acidic media. The integration of results from all experiments (12 independent kinetic runs with total chloride concentrations ranging from 0.1 to 1.0 and total proton concentrations from 0.1 to 0.5 mol L-1) gives a log k ) 4.44 ((0.05) [k: L2 mol-2 min-1] (Table 3 in Supporting Information) for the oxidation rate

-d[Sb(III)]/dt ) k[H2O2][Cl-][H+]

(9)

Environmental Implications. The common presence of significant amounts of Sb(III) in oxic waters (6) requires some kind of kinetic stabilization of this species. No extensive studies have yet been carried out on oxidation rates for Sb(III) in natural aquatic systems (7, 8, 10, 71). Kinetic results obtained in this study show that the neutral Sb(OH)3 species is unreactive to H2O2 oxidation, the presence of Sb(OH)4being necessary for the formation of Sb(V) to occur. This is also the case for arsenic (48), but while Sb(OH)3 has a pK of 11.7, the corresponding arsenic species has a pK of 9.2 (72). Arsenic oxidation will thus start at a pH value closer to the range of natural pH values than antimony. Measured rate constants allow the estimation of the Sb(III) half-life under conditions close to natural ones. For example, in standard seawater conditions (temperature ) 25 °C, I ) 0.5 mol L-1, pH ) 8.1, [H2O2] ) 10-7 mol L-1), Sb(III) has a half-life of 284 days. In freshwater conditions (temperature ) 25 °C, I ) 0.02 mol L-1, pH ) 7.5, [H2O2] ) 10-7 mol L-1), the Sb(III) half-life would be 3303 days. Lowering the temperature decreases the reactivity of Sb(III) (a temperature shift from 25 to 15 °C doubles the residence

time). Decreasing the H2O2 concentration by a certain factor will increase the residence time by the same factor. Comparisons with arsenic show that, for an ionic strength of 0.02 mol L-1 at 25 °C and pH 8.5, a H2O2 concentration of 0.5 µmol L-1 gives a half-time of 65 days for Sb(III) as compared to 89 days for As(III) under the same conditions. All these results suggest that Sb(III) oxidation by hydrogen peroxide might be relevant either in surface waters with high H2O2 concentrations and alkaline pH values or in treatment systems for contaminated solutions with millimolar H2O2 concentrations. Otherwise, in the absence of the catalytic effect of trace metals present in the system, Sb(III) seems to be rather refractory vis-a`-vis H2O2 oxidation. Should this stability be confirmed vis-a`-vis other natural oxidants, it may help to explain the significant Sb(III) concentrations generally found in oxic surface waters.

Supporting Information Available Tables 1-3. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) United States Environmental Protection Agency. Water Related Fate of the 129 Priority Pollutants; Doc. 745-R-00-007; USEPA: Washington, DC, 1979; Vol. 1. (2) Council of the European Communities. Council Directive 76/ 464/EEC of 4 May 1976 on Pollution Caused by Certain Dangerous Substances Discharged into Aquatic Environment of the Community; Official Journal L 129, 1976; pp 23-29. (3) http://minerals.usgs.gov/minerals/. (4) Filella, M.; Belzile, N.; Chen, Y.-W. Earth-Sci. Rev. 2002, 59, 265-285. (5) Filella, M.; May, P. M. Geochim. Cosmochim. Acta 2003, 67, 4013-4031. (6) Filella, M.; Belzile, N.; Chen, Y.-W. Earth-Sci. Rev. 2002, 57, 125-176. (7) Sun, H.-W.; Shan, X.-Q.; Ni, Z.-M. Talanta 1982, 29, 589-593. (8) Bertine, K. K.; Lee, D. S. Antimony content and speciation in the water column and interstitial waters of Saanich Inlet. In Trace Metals in Seawater; Wong, C. S., Boyle, E., Bruland, K. W., Burton, J. D. Goldberg, E. D., Eds.; Plenum Press: New York, 1983; pp 21-38. (9) Quentel, F.; Filella, M. Anal. Chim. Acta 2002, 452, 237-244. (10) Cutter, G. A. Mar. Chem. 1992, 40, 65-80. (11) Cutter, G. A.; Cutter, L. S.; Featherstone, A. M.; Lohrenz, S. E. Deep Sea Res. II 2001, 48, 2895-2915. (12) Belzile, N.; Chen, Y.-W.; Wang, Z. Chem. Geol. 2001, 174, 379387. (13) Cooper, W. J.; Zika, R. G. Science 1983, 220, 711-712. (14) Draper, W. M.; Crosby, D. G. Arch. Environ. Contam. Toxicol. 1983, 12, 121-126. (15) Zika, R. G.; Moffett, J. W.; Petasne, R. G.; Cooper, W. J.; Saltzman, E. S. Geochim. Cosmochim. Acta 1985, 49, 1173-1184. (16) Zika, R. G.; Saltzman, E. S.; Cooper, W. J. Mar. Chem. 1985, 17, 265-275. (17) Cooper, W. J.; Shao, C.; Lean, D. R. S.; Gordon, A. S.; Scully, F. E. Factors affecting the distribution of H2O2 in surface waters. In Environmental Chemistry of Lakes and Reservoirs; Lawrence, A., Baker, A., Eds.; Advances in Chemistry Series 237, American Chemical Society: Washington, DC, 1994; pp 391-422. (18) Herrmann; R.; Herrmann, K. Aqua Fennica 1994, 24, 3-8. (19) Herrmann, R. Environ. Toxicol. Chem. 1996, 15, 652-662. (20) Price, D.; Fauzi, C.; Mantoura, R.; Worsfold, P. J. Anal. Chim. Acta 1998, 377, 145-155. (21) Cooper, W. J.; Lean, D. R. S. Environ. Sci. Technol. 1989, 23, 1425-1428. (22) Zuo, Y.; Hoigne´, J. Environ. Sci. Technol. 1992, 26, 1014-1022. (23) Yuan, J.; Shiller, A. M. Atmospheric Environ. 2000, 34, 39733980. (24) Cooper, W. J.; Saltzman, E. S.; Zika, R. G. J. Geophys. Res. 1987, 92, 2970-2980. (25) Willey, J. D.; Kieber; R. J.; Lancaster, R. D. J. Atmos. Chem. 1996, 25, 149-165. (26) Haber, F.; Weiss, J. Proc. R. Soc. (London) A 1934, 147, 332-351. (27) Baxendale, J. H.; Evans, M. G.; Park, G. S. Trans. Faraday Soc. 1946, 42, 155-169. (28) Barb, W. G.; Baxendale, J. H.; George, P.; Hargrave, K. R. Trans. Faraday Soc. 1951, 47, 462-500. VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2847

(29) Uri, N. Chem. Rev. 1952, 50, 375-454. (30) Dainton, F. S.; Sutton, H. C. Trans. Faraday Soc. 1953, 49, 10111025. (31) Rigg, T.; Taylor, W.; Weiss, J. J. Chem. Phys. 1954, 22, 575-577. (32) Hardwick, T. J. Can. J. Chem. 1957, 35, 428-436. (33) Hardwick, T. J. Can. J. Chem. 1957, 35, 437-443. (34) Wells, C. F.; Salam, M. A. Trans. Faraday Soc. 1967, 63, 620629. (35) Wells, C. F.; Salam, M. A. J. Chem. Soc. A 1968, 24-29. (36) Po, H. N.; Sutin, N. Inorg. Chem. 1968, 7, 621-624. (37) Waite, T. D.; Morel, F. M. M. Environ. Sci. Technol. 1984, 18, 860-868. (38) Moffett, J. W.; Zika, R. G. Environ. Sci. Technol. 1987, 21, 804810. (39) Millero, F. J.; Sotolongo, S. Geochim. Cosmochim. Acta 1989, 53, 1867-1873. (40) Millero, F. J.; Sotolongo, S.; Stade, D. J.; Vega, C. A. J. Solid Chem. 1991, 20, 1079-1092. (41) Moffett, J. W.; Zika, R. G. Mar. Chem. 1983, 13, 235-251. (42) Moffett, J. W.; Zika, R. G. In Photochemistry of Environmental Aquatic Systems; Zika, R. G., Cooper, W. J., Eds.; ACS Symposium Series 327; American Chemical Society: Washington, DC, 1987; pp 116-130. (43) Sharma, V. K.; Millero, F. J. Geochim. Cosmochim. Acta 1989, 53, 2269-2276. (44) Millero, F. J.; Sharma, V. K.; Karn, B. Mar. Chem. 1991, 36, 7183. (45) Pettine, M.; Millero, F. J. Limnol. Oceanogr. 1990, 35, 730-736. (46) Pettine, M.; Millero, F. J.; La Noce, T. Mar. Chem. 1991, 34, 29-46. (47) Woods, R.; Kolthoff, I. M.; Meehan, E. J. J. Am. Chem. Soc. 1964, 86, 1698-1700. (48) Pettine, M.; Campanella, L.; Millero, F. J. Geochim. Cosmochim. Acta 1999, 63, 2727-2735. (49) Pettine, M.; Millero, F. J. Mar. Chem. 2000, 70, 223-234. (50) Voegelin, A.; Hug, S. J. Environ. Sci. Technol. 2003, 37, 972-978. (51) Hug, S. J.; Leupin, O. Environ. Sci. Technol. 2003, 37, 27342742. (52) May, P. M. Chem. Commun. 2000, 1265-1266. (53) May, P. M.; Murray, K. J. Chem. Eng. Data 2000, 46, 1035-1040.

2848

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

(54) Livingstone, R. S.; Bray, W. C. J. Am. Chem. Soc. 1925, 47, 20692082. (55) Livingstone, R. S.; Bray, W. C. J. Am. Chem. Soc. 1926, 48, 405406. (56) Livingstone, R. S. J. Am. Chem. Soc. 1928, 50, 3204-3211. (57) Makower, B. J. Am. Chem. Soc. 1934, 56, 1315-1319. (58) Mishra, S. K.; Gupta, Y. K. J. Inorg. Nucl. Chem. 1968, 30, 29913008. (59) Nicol, M. J. South Afr. J. Chem. 1982, 35, 77-79. (60) Dimitrijevic, M.; Antonijevic, M. M.; Dimitrijevic, V. Minerals Eng. 1999, 12, 165-174. (61) Fridman, Ya. D.; Veresova, R. A.; Luk’yanets, A. N. Khim. i Tekhnol. Sur’my, Akad. Nauk. Kirg. SSR, Inst. Neorgan. i Fiz. Khim. 1965 13-24; CA 64 13453 g. (62) Baes, C. F.; Mesmer, R. E. The Hydrolysis of Cations; Wiley: New York, 1976. (63) Newmann, H. M. J. Am. Chem. Soc. 1954, 76, 2611-2615. (64) Lingane, J. J.; Nishida F. J. Am. Chem. Soc. 1947, 69, 530-533. (65) Neumann, H. M.; Ramette, R. W. J. Am. Chem. Soc. 1956, 78, 1848-1851. (66) Whitney J. E.; Davidson, N. J. Am. Chem. Soc. 1949, 71, 38093816. (67) Bonner, N. A. J. Am. Chem. Soc. 1949, 71, 3909-3914. (68) Newmann, H. M.; Brown, H. J. Am. Chem. Soc. 1956, 78, 18431847. (69) Chek, C. H.; Bonner, N. A.; Wahl, A. C. J. Am. Chem. Soc. 1961, 83, 80-84. (70) Bonner, N. A.; Goishi, W. J. Am. Chem. Soc. 1961, 83, 85-99. (71) Chen, Y.-W.; Deng, T.-L.; Filella, M.; Belzile, N. Environ. Sci. Technol. 2003, 37, 1163-1168. (72) Smith, R. M.; Martell, A. E.; Motekaitis, R. J. NIST Critically Selected Stability Constants of Metal Complexes Database; NIST Standard Reference Database 46.Version 6.0; NIST: Gaithersburg, MD, 2001.

Received for review September 16, 2003. Revised manuscript received March 1, 2004. Accepted March 3, 2004. ES035019R