Environ. Sci. Technol. 2004, 38, 1589-1594
Heterogeneous Photocatalytic Reduction of Chromium(VI) over TiO2 Particles in the Presence of Oxalate: Involvement of Cr(V) Species JUAN J. TESTA Facultad de Ingenierı´a, Universidad de Buenos Aires, Paseo Colo´n 850, 1063 Buenos Aires, Argentina M A R IÄ A A . G R E L A Departamento de Quı´mica, Universidad Nacional de Mar del Plata, Funes 3350, B7602AYL Mar del Plata, Argentina MARTA I. LITTER* Unidad de Actividad Quı´mica, Centro Ato´mico Constituyentes, Comisio´n Nacional de Energı´a Ato´mica, Av. Gral. Paz 1499, (1650) San Martı´n, Prov. de Buenos Aires, Argentina
Cr(VI) photocatalytic reduction experiments over TiO2 particles under near UV irradiation in the presence of excess oxalate were performed at acid pH (2 and 3) and under air and N2 bubbling. Initial photonic efficiencies for Cr(VI) reduction are nearly the same under aerobic and anaerobic conditions, but show a significant increase at the lowest pH. At pH 2, the addition of oxalate facilitates Cr(VI) reduction, hindering the electron-shuttle mechanism taking place in pure water. The oxalate synergistic effect at pH 2 is lower than that previously found for EDTA and negligible at pH 3. Chromium(V) oxalate concentration profiles were obtained by EPR spectroscopy in the presence of excess oxalate at pH 1.5. Coordinated Cr(V) complexes [CrV(O)(Ox)2]-, [CrV(OH2)(Ox)2]-, and [CrV(O)(OH)2(Ox)]- were identified, on the basis of the comparison of their corresponding g values with recent literature data. The kinetic analysis of the temporal evolution of the paramagnetic Cr(V) species indicates also an effective photocatalytic degradation of chromium(V) oxalate complexes. This new evidence reinforces previous findings regarding sequential one-electron-transfer processes in Cr(VI) photocatalytic reduction, suggesting that this route may represent a general behavior for the Cr(VI) reduction over UV-irradiated TiO2 particles.
Introduction Chromium exists in natural and anthropogenically modified waters in two common oxidation states, Cr(VI) and Cr(III). [Caution! Cr(VI) is mutagenic and carcinogenic. Inhalation and skin contact of chromium solutions should be avoided.] The first is very toxic to most organisms, and has been classified as carcinogenic and mutagenic. Due to its acute toxicity and high mobility in water, Cr(VI) has been included in the list of priority pollutants by the U.S. Environmental * Corresponding author phone: +54-11-6772-7016; fax: +54-116772-7886; e-mail:
[email protected]. 10.1021/es0346532 CCC: $27.50 Published on Web 01/28/2004
2004 American Chemical Society
Protection Agency, and its concentration in drinking water has been regulated in many countries (1, 2). For example, the allowable limit in drinking water in Sweden and Germany is 0.05 ppm, and the same value has been recommended by the Argentine legislation (3). Chromium(VI) is a frequent contaminant in wastewaters arising from industrial processes such as leather tanning, paint making, and others. As an example, the manufacturing process of leather tanning requires considerable quantities of water, and it discharges nearly 30-35 L of water for every kilogram of leather processed. Cleaner technologies proposes among other actionssthe treatment of effluents for reusing and saving water in industrial processes. One of the most preferred methods to treat Cr(VI) in waters is the transformation to the less noxious Cr(III), which is considered nontoxic and an essential trace metal in human nutrition. Furthermore, Cr(III) can be precipitated and removed as a solid waste. Thus, the complete procedure involves reduction of hexavalent chromium to the trivalent state, followed by precipitation of the insoluble chromium(III) hydroxide at neutral pH. Common reducing agents are sodium thiosulfate, ferrous sulfate, sulfur dioxide gas, and sodium bisulfite/ metabisulfite, but these methods require expensive use of chemicals. The heterogeneous photocatalytic reduction of chromium(VI) using semiconductorssan advanced oxidation technology or processshas been proposed as an economical and simple method of treatment (4, 5 and references therein). It is worthwhile to remark that surface-catalyzed Cr(VI) reductionsalthough very slowsis described as a feasible process in the presence of oxide surfaces such as TiO2 (6). Furthermore, it has been largely demonstrated that the addition of organic donors able to chelate the TiO2 surface accelerates the reduction of Cr(VI) in photocatalytic systems, the synergy being very dependent on the nature of the reducing agent (5-10). In several cases, these organic compounds occur simultaneously with Cr(VI) in wastewaters as a result of different industrial processes. In the last several years, we became interested in mechanistic and kinetic studies on the photocatalytic Cr(VI) reduction over naked and surface-modified TiO2 particles. We have thoroughly studied the influence of different conditions (pH, absence or presence of oxygen, surface morphology). The effect of the presence of EDTA was also analyzed (11-13). Regarding the mechanism of reduction, we proposed that the actual process occurs via sequential one-electron-transfer steps, and we demonstrated this hypothesis for the first time, obtaining direct experimental evidence of the involvement of chromium(V) in the form of aquo and EDTA complexes through EPR spectroscopic experiments (14). Since a Cr(V) species, [Cr(O)(H2O)5], was also detected in the absence of the donor, we believe that this seems to be a general behavior for the Cr(VI) photocatalytic reduction. In another context, Cr(V) intermediates are commonly involved in the mechanisms of Cr(VI) reduction, and Cr(V) chemistry has recently received a lot of attention since Cr(V) complexes with biomolecules are implicated in the mechanism of Cr(VI) carcinogenesis (15, 16). Lately, Millero et al. postulated that the reduction of Cr(VI) to Cr(III) with H2O2s another possible engineered system for the removal of Cr(VI) from wastewaterssoccurs in sequential one-electron steps. On the basis of the experimental evidence, the authors postulate that the reduction of Cr(VI) is controlled by the formation of Cr(V) intermediates (17, 18). In this paper, we analyze in detail the effect of oxalic acid on Cr(VI) reduction. To our knowledge, the Cr(VI)/oxalate/ VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1589
TiO2 system has never been investigated, despite that oxalate is a very simple species and a good reducing agent that has previously been used in other photocatalytic systems (19). On the other hand, oxalic acid is a model for organic compounds present in natural waters, and its effect on Cr(VI) reduction by photocatalysis can afford explanations for solarlight-promoted processes in aquatic environments.
Experimental Section Materials and Methods. Degussa P-25 was a commercial sample gently supplied by the manufacturers and used as provided. K2Cr2O7 (Carlo Erba), oxalic acid (Riedel de Ha¨en), 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO; Aldrich), and all other chemicals were of analytical reagent grade. Doubly distilled water was used for the preparation of all solutions. Cr(VI) Decay. Photocatalytic irradiations were carried out in a 125 mL cylindrical cell, irradiated from the top, using a high-pressure xenon arc lamp (Osram XBO, 450 W). The setup was thermostated at 298 K. The cell was provided with a Teflon cup with a gas inlet, a sampler, and a holder for a band-pass filter (Schott no. BG25, 3 mm thickness, 270 nm < λ < 510 nm, 87% maximum transmission at 360 nm). The IR fraction of the incident light was removed by a Schott no. KG5 filter. Actinometric measurements were performed by the ferrioxalate method (20) using 20 mL of actinometric solutions to keep the same conditions as in the photocatalytic experiments. A photon flow per unit volume of (1.0 ( 0.1) × 10-5 einstein s-1 L-1 was calculated. For all photocatalytic runs, a fresh solution (20 mL) containing 0.4 mM K2Cr2O7 and 4.0 mM C2O4H2 was adjusted to the desired pH with diluted perchloric acid or NaOH, and the catalyst was suspended at 1 g L-1 concentration. Prior to irradiation, suspensions were ultrasonicated for 20 s and magnetically stirred in the dark at 298 K for 20 min to ensure substrate-surface equilibration. The concentration of Cr(VI) after equilibration was taken as the initial concentration (C0), to discount the adsorption in the dark (which varied between 4% and 7% in the conditions explored in this work) and to evaluate only changes due to irradiation. Irradiations for a fixed period were performed under magnetic stirring and under air or nitrogen bubbling (0.2 L min-1). Samples of 500 µL were periodically withdrawn for quantitative analysis and filtered through a 0.22 µm Millipore filter. At least duplicated runs were carried out for each condition, averaging the results. Changes in Cr(VI) concentrations were followed by UV spectroscopy at 349 nm (21). UV-vis absorption measurements were performed employing a Shimadzu 210A spectrophotometer. Detection of Intermediates by EPR Spectrometry. EPR spectra were obtained at 298 K using a Bruker ER 200 X-band spectrometer (Bruker Analystische Messtechnik GMBH, Germany). TEMPO (g ) 2.0051) was used as a concentration standard and as a standard for determination of g factors, as recommended elsewhere (22, 23). Typical instrumental conditions were as follows: central field, 3498 G; sweep width, 50 G; 1-20 scans; microwave power, 43 mW; modulation frequency, 100 kHz; time constant, 1-50 ms; sweep time, 1-5 s; modulation amplitude, 1.25 Gpp; receiver gain, 5 × 105. Appropriate volumes of stock solutions of 1.6 mM K2Cr2O7 and 16 mM oxalic acid were added to previously sonicated 0.1 g L-1 aqueous TiO2 suspensions, and the pH was adjusted to 1.5 with diluted HClO4. The samples were stirred in the dark for 10 min and then irradiated in a quartz EPR flat cell, inside the EPR cavity, with a 400 W, medium-pressure metal halide lamp (Phillips, HPA 400), emitting predominantly light between 300 and 450 nm. The output passed successively through a 10 cm water infrared filter and a long-pass glass filter, to isolate λ g 340 nm. 1590
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
FIGURE 1. Time profile of Cr(VI) reduction during the irradiation of 1 g L-1 aqueous TiO2 suspensions in the presence of oxalate at two different pH values under continuous air or N2 bubbling. Conditions: [C2O4H2] ) 4.0 mM, [K2Cr2O7] ) 0.4 mM, T ) 298 K, photon flow per unit volume (1.0 ( 0.1) × 10-5 einstein s-1 L-1, (a) pH 3, (b) pH 2. Comparative curves in similar conditions in the absence of donor and in the presence of EDTA, extracted from refs 12 and 13, are also included. EPR amplitude signals were transformed in radical concentrations by comparing the area under the EPR firstderivative spectrum of the sample with that of a standard aqueous solution of TEMPO, recorded at the same microwave power, modulation amplitude, and amplification gain.
Results and Discussion Chromium(VI) Decay. Figure 1 shows the time profiles of Cr(VI) reduction during the irradiation of 1 g L-1 aqueous TiO2 suspensions in the presence of excess oxalate, at two different pH values (2 and 3) and under continuous air or N2 bubbling. The results obtained in pure water and in the presence of EDTA under similar conditions were extracted from refs 12 and 13 and added to Figure 1 for the sake of comparison. At pH 3, after 60 min of irradiation, a low extent of Cr(VI) reduction took place in the absence of donor and in the presence of oxalate, while a much higher conversion was attained in the presence of EDTA; the Cr(VI) reduction extents attained in pure water, in oxalate, and in EDTA were 35%, 40%, and 86%, respectively (Figure 1a). Conversely, at pH 2, an almost complete transformation of Cr(VI) was achieved after 60 min in the presence of oxalic acid, either in air or under nitrogen (97% and 98%, respectively), while, in the absence of donor, only 52% Cr(VI) reduction was attained after 60 min (Figure 1b). Again EDTA proved to be a much more efficient donor, with a complete transformation after only 15 min (12). An electron-shuttle mechanism, in which Cr(VI) is continuously reduced and reoxidized by valence band holes, was previously invoked to
TABLE 1. Initial Photonic Efficiencies for Cr(VI) Reduction under Different Experimental Conditionsa conditions
pH
ζ0 (%)
conditions
pH
ζ0 (%)
no donor no donor EDTA
2 3 2
1.2 0.8 9.8
EDTA oxalate oxalate
3 2 3
4.7 2.5 (air)/2.0 (N2) 0.8 (air)/1.2 (N2)
a
Conditions are those of Figure 1 and refs 12 and 13.
explain the results in pure water. This mechanism is nonoperative in the presence of a reducing (sacrificial) agent that can strongly compete for the holes with reduced Cr species, producing a synergistic cooperative effect (13). The fact that EDTA enhances Cr(VI) reduction more than oxalate can be explained because the last species readily oxidizes to CO2 during the photocatalytic process, while EDTA produces several intermediates in route to its mineralization, which may also contribute to the synergistic effect. However, we were rather surprised to see that oxalate has no influence on Cr(VI) reduction at pH 3. A plausible explanation is that different surface titanium(V) oxalate complexes of dissimilar reactivity, whose stability is strongly dependent on protolytic changes occurring in the interface (24, 25), prevail at the two pH values. Our results indicate that, at pH 3, surfacecoordinated oxalate cannot prevent Cr(V) reoxidation. From the plots, initial photonic efficiencies ζ0 were calculated according to
ζ0 )
(-dC/dt)t)0 P0
where (-dC/dt)t)0 is the initial reaction rate and P0 the incident photonic flow per unit volume. The results are indicated in Table 1. At pH 3, initial efficiencies are similar in oxalate and in the absence of donor (see Figure 1a), but at pH 2 they are higher in the presence of oxalate (see Figure 1b). At both pH values, EDTA promotes a faster reaction from the beginning of the irradiation. The photocatalytic reduction of Cr(VI) becomes faster as pH is diminished, as previously reported in pure water or in the presence of donors (5, 8, 9, 12 and references therein). This can be explained due to the fact that the global Cr(VI) reduction consumes protons. Concerning specifically the Cr(VI)/Ox profiles, a fairly good first-order kinetics (correlation coefficients around 0.960.99) is followed under all conditions; taking into account the low initial Cr(VI) concentration (0.4 mM), this agrees with a general Langmuir-Hinshelwood kinetic behavior, as proposed elsewhere (9, 26-28). It can be observed that oxygen has a minimum effect on Cr(VI) decay, according to previous results under different conditions (9, 10, 13) (see also Table 1). This behavior was ascribed to the fact that, at these working pH values, Cr(VI) reduction by conduction band electrons is much more favorable than O2 reduction. This result is a characteristic of the Cr(VI)/TiO2 system, which has been controversial in the first papers on the system (see ref 5). Photocatalytic reduction of Cr(VI) seems to be an exception compared to that of the majority of metallic ions. For example, the Hg(II) reduction rate as well as the nature of the products deposited onto the catalyst is very sensitive to the presence of oxygen (29). Additionally, the fact that Cr(VI) decays in the presence of oxalate are almost the same under N2 or air indicates that the intermediates formed during the oxidation of oxalic acid are not involved in Cr(VI) reduction, as discussed below. Oxalate degradation is supposed to occur through the cleavage of the C-C bond to form CO2 plus a strong reducing agent, •CO2- (19). The reduction potential for the (CO2/•CO2-)
couple, E0(CO2/•CO2-) ) -1.8 V (30), makes this intermediate capable of reducing Cr(VI) (E0(Cr(VI)/Cr(V) ) 0.55 V (31)), according to reaction 1.
Cr(VI) + •CO2- f Cr(V) + CO2 (k1)
(1)
However, the reaction of •CO2- with dissolved molecular oxygen (E0(CO2/•CO2-) ) -0.16 V (31)) (eq 2) or its oxidation to CO2 via electron injection into the conduction band (VFB ) -0.3 V at pH 0 for TiO2 Degussa P-25 (32)) (eq 3) must also be considered when the possible fate of the carboxyl anion radical is discussed.
CO2- + O2 f CO2 + O2•- (k2)
(2)
CO2- f CO2 + ecb- (k3)
(3)
•
•
A lack of oxygen effect could only be explained if the fraction of •CO2- radicals that reacts with Cr(VI)
f)
k1[Cr(VI)] k1[Cr(VI)] + k2[O2] + k3
remains unaltered when N2 is replaced by O2. The last condition requires k1[Cr(VI)] + k3 . k2[O2]. Using the reported k2 value, 4 × 109 M-1 s-1 (33), and considering [O2] ) 0.3 mM in air-saturated atmosphere, we calculate a pseudo-firstorder constant k2[O2] ) 1.2 × 106 s-1. On the other hand, Hug and co-workers derived a tentative value of k1 ) (6-12) × 107 M-1 s-1 from a kinetic modeling of the iron(III)-catalyzed reduction of Cr(VI) by oxalate in aqueous solutions, which implies k1[Cr(VI)] ) 7.2 × 104 s-1 for the highest Cr(VI) concentration explored in this work (0.8 mM) (34). Thus, we are led to conclude that, if formed, carboxyl anion radicals should rapidly (k3 . 1.2 × 106 s-1) be removed by electron injection into the conduction band. Although it has previously been recognized that intermediate radicals formed by reaction of suitable donors with HO• or h+ will often inject an electron into the conduction band of the semiconductor, there are no available experimental data to compare with this estimation (35, 36). Nevertheless, current-doubling effects have been observed in the photooxidation of formiate on a semiconductor electrode, a phenomenon that was attributed to the fact that the •CO2- intermediate is simultaneously oxidized with the release of electrons into the conduction band (36). On the other hand, it has been determined by transient absorption spectroscopy that injection of redox dye photosensitizers chemically bound to TiO2 is a very fast process ranging from femtoseconds to nanoseconds (37). Detection of Intermediates by EPR Spectrometry. In homogeneous systems, it has been demonstrated that the Cr(VI) oxidation of oxalic acid involves Cr(V)-Ox complexes as intermediates. These complexes have been thoroughly characterized by Lay and co-workers (38-41), on the basis of previous studies of Rocˇek (42, 43). In agreement with these results, we were able to detect Cr(V) species in the heterogeneous TiO2/Cr(VI)/Ox system by EPR spectrometry, as follows. The experiments were performed at pH 1.5 because it is reported that the Cr(V) signal can be only detected at values lower than 1.6, due to the stability of Cr(V)-Ox complexes under this condition (38). Figure 2 shows a typical EPR spectrum obtained at 298 K, under continuous UV irradiation of a sample containing 0.1 g L-1 TiO2, 4.0 mM oxalic acid, and 0.8 mM K2Cr2O7, pH 1.5. From this figure, we determined g values of 1.976 and 1.971, using TEMPO as internal reference, which agrees, within experimental error, with the freely tumbling structures a and b in solution, shown in Scheme 1 (38, 43). VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1591
SCHEME 1
FIGURE 2. EPR spectra of the Cr(V) intermediates formed under continuous irradiation of a TiO2 suspension (0.1 g L-1) at pH 1.5. Conditions: [C2O4H2] ) 4.0 mM, [K2Cr2O7] ) 0.8 mM, central field 3498 G, modulation amplitude 1.25 Gpp, gain 5 × 105. No evidence of [Cr(O)(H2O)5] (g ) 1.9707) could be obtained; thus, if formed, Cr(V)-aquo species should attain a considerably low concentration, making these species undetectable. In fact, the same appearance of the spectra shown in Figure 2 was invariably recorded in the whole range of conditions explored in this work (1 e [oxalic acid]/mM e 8.0, 0.2 e [K2Cr2O7]/mM e 0.8). Moreover, the ratio of the areas of the two signals (g ) 1.971 and 1.976), i.e., 3.5 ( 0.1, remains unaltered during the whole irradiation experiment. Kinetic runs were thus performed by recording the signal amplitude of the most intense line of the first-derivative EPR spectra shown in Figure 2sat a fixed magnetic fieldsas a function of time. At least three runs were acquired and averaged for each set of experimental conditions. Figure 3 shows Cr(V) concentration as a function of irradiation time in the heterogeneous and homogeneous systems (1 mM oxalic acid and 0.2 mM K2Cr2O7). As observed, a small paramagnetic signal of Cr(V)-Ox complexes (corresponding to 0.15 µM in this case) appears immediately after the solutions are mixed, either in the presence or in the absence of TiO2, and remains constant in the absence of irradiation in the time range of our EPR experiments. However, the temporal evolution of the paramagnetic species obtained under irradiation is remarkably different. In the presence of TiO2, the initial rate of formation of Cr(V) intermediates is more than 1 order of magnitude greater than that obtained in the homogeneous system under steady irradiation and, which is more remarkable, Cr(V) complexes readily disappear in the heterogeneous photocatalytic system at variance with the behavior in the homogeneous medium (38, 43). Both facts indicate that we are observing a distinct 1592
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
FIGURE 3. Time dependence of the [Cr(V)-Ox] intermediates obtained under continuous irradiation in the EPR experiments (see the text). Conditions: [C2O4H2] ) 1.0 mM, [K2Cr2O7] ) 0.2 mM, pH 1.5, T ) 298 K, (open circles) [TiO2] ) 0, (solid circles) [TiO2] ) 0.1 g L-1. The solid line is the result of the kinetic model with fitted rate constants. and heterogeneous process. Nearly the same pattern, characteristic of a consecutive mechanism, is obtained for all concentrations explored (cf. Figures 3 and 4 in which different concentrations of both reagents were used). Instead of a full description of the Cr(VI) photocatalytic degradation, we look forward to a simplified mechanism to rationalize the experimental findings. We found that the set of reactions comprised by eqs 4-7:
Cr(V) + eCB- f Cr(IV) (k5)
(4)
Cr(V) + ecb- f Cr(IV) (k5)
(5)
Cr(V) + Ox S Cr(V)-Ox (k6, k-6)
(6)
Cr(V)-Ox + ecb- f Cr(IV)-Ox (k7)
(7)
is able to account for the generation and decay of Cr(V)-Ox complexes during the photocatalytic degradation of Cr(VI). Here, Ox stands for oxalic acid, independently of its speciation at the corresponding pH (pK1 ) 1.25), Cr(V) and Cr(V)-Ox represent, respectively, the aquo and oxalate complexes (the last in the form of structure a or b), and Cr(IV) and Cr(IV)Ox are the corresponding Cr(IV) complexes, which are silent to EPR detection (43). An iterative computer program (FACSIMILE) (44) was used to determine optimal values for the set of constants of reactions 4-7. The program minimizes the difference between the experimental and the simulated Cr(V) profiles derived from the numerical integration of the set of differential equations corresponding to the simplified mech-
FIGURE 4. Time dependence of the [Cr(V)-Ox] intermediates obtained under continuous irradiation in the EPR experiments (see the text). Conditions: [C2O4H2] ) 4.0 mM, [K2Cr2O7] ) 1.0 mM, pH 1.5, T ) 298 K, (open circles) [TiO2] ) 0, (solid circles) [TiO2] ) 0.1 g dm-3. The solid line is the result of the kinetic model with fitted rate constants.
to a first-order exponential decay, from which we estimated a half-life value of 23 s for the Cr(V)-Ox complexes under this condition. This value is larger than the one obtained from k7 (8 s), thus verifying that conduction band electrons are involved in the decay of Cr(V) species. In summary, we showed that at pH 2 oxalic acid is a relatively good synergistic agent for Cr(VI) reduction. Although oxalate is not as efficient as EDTA, the presence of oxalate reduces the half-life of Cr(VI) nearly 4-fold. This cooperative action is postulated to occur by the suppression of the electron-shuttle mechanism of continuous Cr(VI) reduction and reoxidation by holes, since the experimental evidence presented here demonstrates that oxalate intermediates are not involved in Cr(VI) depletion. On the other hand, we proved that Cr(VI) is reduced via a sequential one-electron-transfer step, as previously found with EDTA and also in the absence of added donors. Additionally, in the case of oxalate, Cr(V)-organic complexes are also efficiently degraded by conduction band electrons. Environmental and Technical Significance. Oxalic acid has a synergic action on the photocatalytic reduction of Cr(VI) over TiO2 particles. This can be a useful method to replace more conventional treatment techniques that involve expensive chemical reagents, just using fluorescent UV lamps with a low consumption of electrical energy or solar illumination, without any electrical waste energy. On the other hand, oxalic acid or other similar organic donors are usually present in industrial wastes together with chromium, making unnecessary the external addition of the reagent. The catalyst can be recovered and reused.
Acknowledgments This work was performed as part of Comisio´n Nacional de Energı´a Ato´mica CNEA-CAC-UAQ Project No. 00-Q-03-08. M.I.L. thanks CONICET for financial support (Grant PIP 662/ 98) and ANPCyT (Grant PICT98-13-03672). M.A.G. also thanks ANPCyT (Grant PICT98-06-0431). M.I.L. and M.A.G. are members of CONICET.
Literature Cited FIGURE 5. [Cr(V)-Ox] generation under continuous irradiation of a 0.1 g L-1 TiO2 suspension (open circles) and time evolution of the [Cr(V)-Ox] intermediates after interruption of the irradiation (solid circles). Conditions: [C2O4H2] ) 1.0 mM, [K2Cr2O7] ) 0.2 mM, pH 1.5, T ) 298 K. The solid line is a fit to first-order decay. See the text. anism. Solid lines in Figures 3 and 4 show the results of the simulations with the derived set of constants k7 ) 8.5 × 10-2 s-1, k6 ) 1.4 × 105 M-1 s-1, k-6 ) 2.2 × 101 s-1, k5 ) 1.0 × 101 s-1, and k4[Cr(VI)]0 ) 9.6 × 10-5 M s-1. However, several combinations of the individual values of k5, k6, and k-6, with slight changes, may also reproduce quite reasonably the profiles, provided that the group satisfies the condition k5 × k6/k-6 ) 6.4 × 104 M-1 s-1. These figures are also consistent with the fact that we did not observe Cr(V)-aquo complexes, since the simulated concentration of this species remains below our detection limit. As said before, Cr(V)-Ox complexes are readily decomposed under irradiation in the photocatalytic system, at variance with the behavior observed in the homogeneous medium. We suggest that this is due to the fact that Cr(V)Ox complexes are additionally reduced by conduction band electrons, i.e., in a heterogeneous reaction, as stated in step 7. In a separate experiment, we produced the Cr(V)-Ox species heterogeneously, under the same conditions as in Figure 3, and when the EPR signal amplitude achieved its maximum value (t ≈ 10 s), we turned off the light and followed the decay of the paramagnetic signal (see Figure 5). The experimental profile produced in the dark could be adjusted
(1) Heavy Metals, Problems and Solutions; Salomons, W., Fo¨rstner, U., Mader, P., Eds.; Springer-Verlag: Berlin, Heidelberg, 1995; (a) p 386, (b) p 36. (2) Sreeram, K. J.; Rao, J. R.; Sundaram, R.; Nair, B. U.; Ramasami, T. Green Chem. 2000, 37. (3) Secretarı´a de recursos hı´dricos de la Repu ´ blica Argentina, http://www.medioambiente.gov.ar/mlegal/residuos/dec831/ dec831_93.htm, 1991. (4) Chenthamarakshan, C. R.; Rajeshwar, K.; Wolfrum, E. J. Langmuir 2000, 16, 2715. (5) Litter, M. I. Appl. Catal., B 1999, 23, 89. (6) Deng, B.; Stone, A. T. Environ. Sci. Technol. 1996, 30, 2484. (7) Colo´n, G.; Hidalgo, M. C.; Navı´o, J. A. J. Photochem. Photobiol., A 2001, 138, 79. (8) Prairie, M. R.; Evans, L. R.; Stange, B. M.; Martı´nez, S. L. Environ. Sci. Technol. 1993, 27, 1776. (9) Ku, Y.; Jung, I.-L. Water Res. 2001, 35, 135. (10) Colo´n, G.; Hidalgo, M. C.; Navı´o, J. A. Langmuir 2001, 17, 7174. (11) Navı´o, J. A.; Colo´n, G.; Trillas, M.; Peral, J., Dome`nech, X.; Testa, J. J.; Padro´n, J.; Rodrı´guez, D.; Litter, M. I. Appl. Catal., B 1998, 16, 187. (12) Navı´o, J. A.; Testa, J. J.; Djedjeian, P.; Padro´n, J. R.; Rodrı´guez, D.; Litter, M. I. Appl. Catal., A 1999, 178, 191. (13) Siemon, U.; Bahnemann, D.; Testa, J. J.; Rodrı´guez, D.; Litter, M. I.; Bruno, N. J. Photochem. Photobiol., A 2002, 148, 247. (14) Testa, J. J.; Grela, M. A.; Litter, M. I. Langmuir 2001, 17, 3515. (15) Pattison, D. L.; Lay, P. A.; Davies, M. J. Inorg. Chem. 2000, 39, 2729. (16) Lay P. A.; Levina, A. J. Am. Chem. Soc. 1998, 120, 6704. (17) Pettine, M.; Campanella, L.; Millero, F. J. Environ. Sci. Technol. 1995, 29, 222. (18) Pettine, M.; Campanella, L.; Millero, F. J. Environ. Sci. Technol. 2002, 36, 901. (19) Forouzan, F.; Richards, T. C.; Bard, A. J. J. Phys. Chem. 1996, 100, 18123. VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1593
(20) Hatchard, C. G.; Parker, C. A. Proc. R. Soc. London, A 1956, 235, 518. (21) Wei, C.; German, S.; Basak, S. R.; Rajeshwar, K. J. Electrochem. Soc. 1993, 140, L60. (22) Weil, J. A.; Bolton, J. R.; Wertz, J. E. Electron Paramagnetic Resonance: Elementary Theory and Practical Applications; Wiley: New York, 1994; p 498. (23) Chappell, J.; Chiswell, B.; Canning, A. Talanta 1998, 38, 23. (24) Hug, S. J.; Sulzberger, B. Langmuir 1994, 10, 3587. (25) Weisz, A. D.; Regazzoni, A. E.; Blesa, M. A. Solid State Ionics 2001, 143, 125. (26) Lin, W.-Y.; Wei, C.; Rajeshwar, K. J. Electrochem. Soc. 1993, 140, 2477. (27) Fu, H.; Lu, G.; Li, S. J. Photochem. Photobiol., A 1998, 114, 81. (28) Schrank, S. G.; Jose´, H. J.; Moreira, R. F. P. M. J. Photochem. Photobiol., A 2001, 147, 71. (29) Botta, S. G.; Rodrı´guez, D. J.; Leyva, A. G.; Litter, M. I. Catal. Today 2002, 76 247. (30) Surdhar, P. S.; Mezyk, S. P.; Armstrong, D. A. J. Phys. Chem. 1989, 93, 3360. (31) Standard Potentials in aqueous solutions; Bard, A. J., Parsons, R., Jordan, J., Eds.; Marcel Dekker Inc.: New York, 1985. (32) Martin, S. T.; Herrmann, H.; Hoffmann, M. R. J. Chem. Soc., Faraday Trans. 1994, 90, 3323. (33) Perissinotti, L. L.; Brusa, M. A.; Grela, M. A. Langmuir 2001, 17, 8422. (34) Hug, S. J.; Laubscher, H. U.; James, B. R. Environ. Sci. Technol. 1997, 31, 160.
1594
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004
(35) Schwitzgebel, J.; Ekerdt, J. G.; Sunada, F.; Lindquist, S.-E.; Heller, E. J. Phys. Chem. B 1997, 101, 2621. (36) Hykaway, N.; Sears, W. M., Morisaki, H., Morrison, R. J. Phys. Chem. 1986, 90, 4429. (37) Huber, R.; Moser, J. E.; Gratzel, M. Wachtveitl, J. J. Phys. Chem. B 2002, 106, 6494. (38) Farrell, R. P.; Lay P. A.; Levina, A.; Maxwell, I. A.; Bramley R.; Brumby, S.; Ji-Ying, J. Inorg. Chem. 1998, 37, 3159. (39) Barr-David, G.; Charara, M.; Codd, R.; Farrell, R. P.; Irwin, J. A.; Lay P. A.; Bramley R.; Brumby, S.; Ji-Ying, J.; Hanson, G. R. J. Chem. Soc., Faraday Trans. 1995, 91, 1207. (40) Farrell, R. P.; Hudd, R. J.; Lay, P. A.; Bramley, R.; Brumby, S.; Ji, J.-Y. Inorg. Chem. 1989, 28, 3401. (41) Bramley R.; Farrell, R. P.; Ji, J.-Y.; Lay, P. A. Aust. J. Chem. 1990, 43, 263. (42) Hasan, F.; Rocˇek, J. J. Am. Chem. Soc. 1972, 94, 9073. (43) Srinivasan, J.; Rocˇek, J. J. Am. Chem. Soc. 1974, 96, 127. (44) Curtis, A. R.; Sweetenham, W. P. FACSIMILE: a computer program for flow and chemistry simulation and general initial value problems; Computer Science and System Division, Harwell Laboratory: Oxfordshire, Harwell, 1987.
Received for review June 25, 2003. Revised manuscript received October 28, 2003. Accepted December 12, 2003. ES0346532
