Environ. Sci. Technol. 2005, 39, 2376-2382
Simultaneous and Synergistic Conversion of Dyes and Heavy Metal Ions in Aqueous TiO2 Suspensions under Visible-Light Illumination
SCHEME 1. Visible-Light-Induced Simultaneous Oxidation of Dyes and Reduction of Metal Ions on TiO2 Particlesa
HYUNSOOK KYUNG, JAESANG LEE, AND WONYONG CHOI* School of Environmental Science and Engineering, Pohang University of Science and Technology, Pohang 790-784, Korea
This study reports synergistic effects in the simultaneous conversion of dyes and heavy metal ions in aqueous TiO2/ dye/metal ion systems (ternary components) under visible light (λ > 420 nm). TiO2/Cr(VI)/Acid Orange 7 (AO7), TiO2/ Cr(VI)/Rhodamine B (RhB), TiO2/Ag+/AO7, and TiO2/Ag+/RhB were chosen as test systems. Although dyes can be degraded in TiO2 suspensions under visible light, their removal rates were markedly enhanced in the presence of metal ions. Similarly, the reduction rates of metal ions in visible-light-illuminated TiO2 suspensions were negligible, but they were highly accelerated with dyes present. In particular, the synergistic effect in the ternary system of TiO2/ Cr(VI)/AO7 was outstanding. The presence of dissolved oxygen increased the photoreduction rate of Cr(VI) despite the fact that Cr(VI) and O2 are competing electron acceptors. This is ascribed to in-situ photogenerated H2O2 from O2, which acts as a reductant of Cr(VI). RhB and Ag+ ions could be also converted simultaneously under visible light both in the presence and absence of TiO2. The visible-light-induced reduction of Ag+ did not occur at all in TiO2/Ag+ system, but it was enabled in both TiO2/Ag+/ RhB and TiO2/Ag+/AO7 to generate Ag particles. On the other hand, the binary systems of Cr(VI)/AO7, Ag+/AO7, and Ag+/RhB show significant visible-light activities for the conversion of both dye and metal ion. In this case, metal ions and dyes seem to form complexes that induce intracomplex electron transfers upon visible-light absorption. The Cr(VI)/RhB system, however, exhibited insignificant visible-light reactivity.
Introduction Dyes are one of the most notorious contaminants in aquatic environments because of their huge volume of production from industries, slow biodegradation and decoloration, and toxicity. About 700 000 tons of dyes are annually produced in the world, and approximately 15% of the synthetic textile dyes used are lost in waste streams during manufacturing or processing operations (1). Biological oxidation and many physical-chemical treatments are usually inefficient in removing dye colors (2). TiO2 photocatalysis has been extensively studied as a mean of controlling recalcitrant pollutants such as dyes (1, 3-7) and demonstrated successful performances in many cases (8-11). Although TiO2 is not activated by visible light, the * Corresponding author phone: +82-54-279-2283; fax: +82-54279-8299; e-mail:
[email protected]. 2376
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005
a The numbers represent main elementary paths: 1, dye excitation by visible light; 2, electron injection from the excited dye into the CB; 3, back-transfer of electrons to the oxidized dye; 4, reduction of metal ions; 5, degradation of dye.
degradation of dyes on TiO2 under visible light is enabled through a dye-sensitization process in which dyes are excited by absorbing visible-light photons and immediately inject electrons into the TiO2 conduction band (CB) with initiating the degradation of dyes. This visible-light-induced degradation of dyes on the TiO2 surface has been intensively investigated (1, 4-7). Such a dye-sensitized process can be also applied to reductive conversion of substrates (e.g., perchloro compounds) on TiO2 under visible light provided that the reduction potentials are positive with respect to the TiO2 CB edge (12, 13). On the other hand, TiO2 can reductively convert aqueous heavy metal ions under UV illumination (14-20). Photocatalytic reduction of metal ions leads to conversion to lower oxidation states [e.g., Cr(VI) f Cr(III)] (14-20) or deposition onto TiO2 surface as a zerovalent metal (e.g., Ag+ f Ag0) (20). Aqueous chromium species has two stable oxidation states, Cr(VI) and Cr(III). Cr(VI) is toxic, carcinogenic, and mobile in aquatic environments, whereas Cr(III) is much less toxic and less mobile and precipitates at neutral or basic pH (16). Therefore, the reduction of Cr(VI) to Cr(III) is desirable in water treatment. Many research groups have studied the photocatalytic conversion of dyes or heavy metal ions in TiO2 suspensions. Since dyes and heavy metal ions often coexist in real wastewaters, the effect of mutual interactions on the photocatalytic conversion process needs to be investigated. For example, Zhao and co-workers investigated the effect of transition metal ions on dye degradation on TiO2 under visible irradiation (21). They reported that Cu2+ and Fe3+ markedly depress the photodegradation of dyes, whereas other metal ions, such as Zn2+, Cd2+, and Al3+, have only a slight influence. Colo´n et al. studied Cr(VI) reduction in the presence of carboxylic acid in UV-illuminated TiO2 suspension and reported that the conversion rate of Cr(VI) was higher with mixed substrates (18, 19). Fu et al. investigated the photoinduced reduction of Cr(VI) in the TiO2/Cr(VI)/4-cholorophenol system under UV or sunlight irradiation to find that the presence of 4-chlorophenol in TiO2 suspension enhanced the conversion efficiency of Cr(VI) (22). Schrank et al. reported simultaneous Cr(VI) reduction and dye oxidation in a TiO2 slurry under UV illumination (23). However, there has been no report that demonstrates the simultaneous conversion of dyes and heavy metal ions in TiO2 suspension under visiblelight illumination. In this paper, we report synergistic effects in the simultaneous conversion of dyes and heavy metal ions in visiblelight-illuminated TiO2 suspensions. As illustrated in Scheme 1, excited dyes transfer electrons to metal ions through the TiO2 CB. As a result, dyes are oxidized and metal ions are 10.1021/es0492788 CCC: $30.25
2005 American Chemical Society Published on Web 02/18/2005
reduced simultaneously under visible light. Model dyes and metal ions tested in this study include Acid Orange 7 (AO7),
Rhodamine B (RhB), Cr(VI), and Ag+, respectively. The ternary systems (TiO2/dye/metal ion) show highly enhanced conversion efficiencies for both dyes and heavy metal ions under visible light, compared with the binary systems (TiO2/dyes or TiO2/metal ions). The synergistic effect found in this study might be utilized for the solar light treatment of wastewaters containing dyes and heavy metal ions.
Experimental Section Reagents and Materials. TiO2 (Degussa P25), a mixture of 80% anatase and 20% rutile, was used as a photocatalyst. AO7 (87% dye content), RhB (80% dye content), Na2Cr2O7, AgNO3, which represent an anionic dye, cationic dye, oxometal anion, and metal cation, respectively, were purchased from Aldrich and used as received. N2 gas (99.999% purity) that was used for deaerating suspensions was purchased from BOC-Gases. Deionized water used was ultrapure (18 MΩ‚cm) and prepared by a Barnstead purification system. Photolysis. TiO2 suspensions were prepared in water at 0.5 g/L and were dispersed by sonication and shaking for 30 s. An aliquot of AO7 (1 mM) or RhB (500 µM) stock solution was added to the suspension and then the heavy metal ion reagent of Na2Cr2O7 or AgNO3 was added to get desired initial concentrations. The initial concentrations used in all experiments of this study were 50 µM for RhB, 100 µM for AO7 and Na2Cr2O7, and 1 mM for AgNO3. The total suspension volume in the photoreactor was 30 mL for all cases. The pH of suspensions was adjusted to 3.0 with HCl or HClO4 standard solutions. The suspensions were equilibrated in the dark for 30 min prior to illumination and were stirred magnetically throughout the photolysis. Deaerated suspensions were prepared by purging nitrogen gas continuously before and during the illumination. The light source was a 300-W Xe Arc lamp (Oriel). The light was passed through an IR water filter and a UV cutoff filter (λ > 300 nm or λ > 420 nm for visible-light illumination). The sample aliquots were taken from the reactor in regular time intervals and filtered through a 0.45 µm PTFE filter (Millipore) to remove TiO2 particles. More than duplicate experiments were carried out for a given condition. The incident light intensity was measured using a power meter (Newport 1830-C). The light intensities of the 300 nm cutoff (I>300), 420 nm cutoff (I>420), and 550 nm cutoff filtered irradiation (I>550) were measured. The active portion of the UV intensity (Iuv, 300 < λ < 420 nm) and visible-light intensity (Ivis, 420 < λ < 550 nm) was roughly estimated from I>300 I>420 and I>420 - I>550, respectively. The average light intensity of the Xe Arc lamp was 139 mW/cm2 (Iuv) and 154 mW/cm2 (Ivis). Solar light experiments were carried out on the roof of the Environmental Engineering Building at POSTECH (Pohang, Korea: 36° N latitude) under clear sky conditions between June and August. The average solar light intensity was 7.7 mW/cm2 (Iuv) and 10.5 mW/cm2 (Ivis). Analysis. The removal of AO7 and RhB color was monitored using a UV/vis spectrophotometer (Shimadzu UV2401PC). The sampled dye solutions were diluted by 5-fold
FIGURE 1. Conversion of (a) AO7 (C0 ) 100 µM) and (b) Cr(VI) (C0 ) 100 µM as Cr2O72-) in the binary or ternary systems under visiblelight illumination. ([Cr(VI)] ) 2[Cr2O72-] + [HCrO4-]) The experimental conditions were air-equilibrated; pH ) 3.0; [TiO2] ) 0.5 g/L. prior to the absorbance measurements. The monitored absorption peaks were λ ) 485 nm for AO7 (24) and λ ) 554 nm for RhB (5). The conversion of Cr(VI) was monitored using an ion chromatograph (IC, Dionex DX-120), equipped with a Dionex IonPac AS14 column and a conductivity detector. The eluent solution was 3.5 mM Na2CO3/1 mM NaHCO3. The concentrations of in-situ photogenerated H2O2 were determined by a colorimetric method using iodide reagent (25). The sample aliquot was mixed with the solutions of potassium biphthalate, iodide reagent, and ammonium molybdate, and then the absorbance was measured at 350 nm (A350) to determine [H2O2]. Since dyes and Cr(VI) have absorbance at 350 nm, H2O2 concentrations were corrected by subtracting A350 of the control sample without reagents from A350 of the sample with reagents. The concentrations of Ag+ ion in solution were measured by inductively coupled plasma-atomic emission spectroscopy (ICP-AES, Spectro Flame Modula). Transmission electron microscopic images (TEM, Philips CM-200) were taken for Ag particles formed in the visible-light illuminated ternary (dye/TiO2/Ag+) or binary (dye/Ag+) systems. TEM sample was prepared on a copper mesh substrate by drying for 24 h at the room temperature. Dissolved organic carbon contents were quantified using a total organic carbon analyzer (TOC, Shimadzu TOC-VCSH).
Results and Discussion Conversion of AO7 and Cr(VI) under Visible Light. Figure 1 shows that the visible-light-induced conversion rates of VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2377
FIGURE 2. UV/vis absorption spectra of AO7 (C0 ) 100 µM), Cr(VI) (C0 ) 100 µM as Cr2O72-), and the mixture of AO7 and Cr(VI) in water at pH 3.0. The arithmetic sum of spectra of AO7 and Cr(VI) is compared with the spectrum of the mixture (AO7 + Cr(VI)). The solutions were diluted by 2-fold prior to the absorbance measurements.
FIGURE 4. (a) Simultaneous conversion of AO7 and Cr(VI) in the binary and ternary systems under UV-lamp (λ > 300 nm) or solarlight illumination. (b) TOC removal in the binary and ternary systems of AO7 and RhB under UV illumination. The initial TOC values (at time zero) were measured in the absence of TiO2 to avoid dye adsorption on TiO2. (air-equilibrated; [AO7]0 ) 100 µM, [RhB]0 ) 50 µM, [Cr2O72-] ) 100 µM; pH ) 3.0; [TiO2] ) 0.5 g/L). (4). The injected CB electrons subsequently transfer to O2 or Cr(VI) on TiO2 (reactions 2-4) (14, 15).
Dye*-TiO2 f Dye+•-TiO2 + ecb-
(1)
ecb- + O2 f O2-•
(2)
3ecb- + 7H+ + HCrO4- f Cr3+ + 4H2O
(3)
6ecb- + 14H+ + Cr2O72- f 2Cr3+ + 7H2O
(4)
+•
dye
FIGURE 3. Conversion of (a) RhB (C0 ) 50 µM) and (b) Cr(VI) (C0 ) 100 µM as Cr2O72-) in the binary or ternary systems under visiblelight illumination. (air-equilibrated; pH ) 3.0; [TiO2] ) 0.5 g/L) AO7 and Cr(VI) in the ternary mixture (TiO2/Cr(VI)/AO7) are markedly enhanced from those in the binary systems. AO7 alone and Cr(VI) alone were not converted at all under visible light. Neither AO7 nor Cr(VI) in the ternary mixture was transformed in the dark. Under visible-light illumination, excited dyes transfer electrons to the TiO2 CB (reaction 1) and then undergo an immediate degradation (reaction 5) 2378
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005
+ O2/O2
-•
f f degradation
(5)
The initial AO7 removal rate in TiO2/AO7 was 0.3 µM min-1, whereas that in TiO2/Cr(VI)/AO7 was about 1.7 µM min-1. The fact that the dye removal rate is synergistically enhanced in the presence of Cr(VI) indicates that the electron transfer to Cr(VI) (path 4 in Scheme 1) competes with the fast back-transfer of electrons (path 3 in Scheme 1). The difference in the initial AO7 concentrations in Figure 1a is ascribed to the surface adsorption on TiO2 or the complexation with Cr(VI). At pH 3, both AO7 and Cr(VI) take the anionic form, but the TiO2 surface is positively charged (pHzpc ∼ 6.5 (8)). Under this condition, AO7 and Cr(VI) adsorb on TiO2 surface competitively, and the AO7 adsorption on TiO2 is slightly lower in TiO2/Cr(VI)/AO7 than in TiO2/AO7, as reflected in the different initial AO7 concentrations (Figure 1a).
FIGURE 5. Effect of dissolved O2 in TiO2 suspensions (air-equilibrated vs N2-purged) on the conversion of (a) dyes (AO7 or RhB) and (b) Cr(VI) under visible-light illumination. N2 was continuously purged into the suspensions throughout the illumination. The experimental conditions were [AO7]0 ) 100 µM; [RhB]0 ) 50 µM; [Cr2O72-]0 ) 100 µM; pH ) 3.0; [TiO2] ) 0.5 g/L. Although the removal rate of AO7 was significantly enhanced in the presence of Cr(VI), the full mineralization of AO7 cannot be attained under the present experimental condition. In the binary system (AO7/TiO2) containing the initial AO7 concentration of 100 µM (initial TOC 17.3 ppm), the dissolved (nonadsorbed) TOC content increased from 8.5 to 13.7 ppm after 4-h illumination; in the ternary system, the dissolved TOC content increased from 10.7 to 13.3 ppm after 4-h illumination. The fact that the dissolved TOC content increased after the illumination indicates that AO7 adsorbed on TiO2 was partially oxidized and then desorbed into the solution phase. However, this does not necessarily mean that AO7 mineralization was completely absent in the present case, since the measured value is the dissolved TOC, not the total TOC. When the fraction of TOC that was desorbed from the TiO2 surface as a result of the sensitized degradation of AO7 is larger than the mineralized TOC, the net effect would be an increase in the dissolved TOC. A recent study (25) reported that the chemical oxygen demand (COD) of 285 µM AO7 in TiO2 suspension was reduced to about 45% of its initial value after 80 h visible-light illumination beyond which no further decrease in COD was observed because the dye was completely discolored by this time (hence no sensitization). Although their results cannot be directly compared with ours because of many differences in the experimental conditions and monitoring parameters (TOC vs COD), both studies show that AO7 mineralization in visible-light-illuminated TiO2 suspension is very slow and only partial. No other previous studies have demonstrated the full mineralization of dyes in TiO2/visible light systems to our knowl-
FIGURE 6. Conversion of (a) AO7 (C0 ) 100 µM) and (b) Ag+ (C0 ) 1 mM) in the binary or ternary systems under visible-light illumination. (air-equilibrated; pH ) 3.0; [TiO2] ) 0.5 g/L). edge. Although the removal of TOC cannot be achieved in this ternary system either, the oxidized intermediates from dye degradation are expected to be more biodegradable in general and more suitable to biological post-treatment processes. The accelerated discoloration alone is beneficial, since the color removal of the dye wastewaters is often the primary concern. As for the Cr(VI) part, the Cr(VI) conversion rate was almost negligible in TiO2/Cr(VI), but it drastically increased to 2.4 µM min-1 in TiO2/Cr(VI)/AO7 (Figure 1b). It is interesting to note that both Cr(VI) and AO7 were slowly converted under visible light even in the absence of TiO2 in Cr(VI)/AO7. This might be caused by the complex formation between Cr(VI) and AO7, as suggested in Figure 2. The UV/vis absorption spectrum of the mixed solution of Cr(VI) and AO7 was different from the arithmetic sum of the spectra of Cr(VI) and AO7 alone: the visible band intensity at 485 nm was reduced upon mixing. This is very similar to the case of the Fe3+ + AO7 system, in which the complex formation between Fe3+ and AO7 reduced the visible absorption at 485 nm and the visible-light-induced degradation of AO7 occurred (24). It has been also suggested previously that Cr(VI) and salicylic acid (or its degradation intermediates) may form a complex (19). While the standard reduction potential of AO7+•/AO7* is known to be -1.24 V (vs NHE) (4), that of Cr(VI)/Cr(III) is 1.35 and 1.23 V (vs NHE) at pH 0 for HCrO4- and Cr2O72species, respectively (26). Therefore, excited AO7 has an enough driving force to directly reduce Cr(VI). Neither AO7 VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2379
FIGURE 7. TEM images of (a) Ag deposited on TiO2 particles in the ternary TiO2/Ag+/AO7 suspension; (b) Ag particle formed in the binary Ag+/AO7 solution; (c) Ag deposited on TiO2 particles in the ternary TiO2/Ag+/RhB suspension; (d) Ag particles formed in the binary Ag+/RhB solution. All images were taken after 4-h visible-light illumination. The experimental conditions were [AO7]0 ) 100 µM; [RhB]0 ) 50 µM; [Ag+]0 ) 1 mM; air-equilibrated; pH ) 3.0; [TiO2] ) 0.5 g/L. nor Cr(VI) was converted at all in TiO2 suspension at pH 11, because both Cr(VI) and AO7 hardly adsorb on TiO2 surface at basic pH and the reduction of Cr(VI) should be favored at acidic pH according to reactions 3 and 4 (14, 16). Conversion of RhB and Cr(VI) under Visible Light. The simultaneous conversion of RhB and Cr(VI) is shown in Figure 3. RhB alone was not removed at all under visible light. The synergistic effect in the ternary system was not obvious for the RhB removal but was outstanding for Cr(VI) reduction, as in the case of TiO2/Cr(VI)/AO7. Unlike the case of Cr(VI)/AO7, the binary system of Cr(VI)/RhB had little visiblelight reactivity for both RhB and Cr(VI) conversion. The adsorption of RhB on TiO2 is insignificant unlike AO7 because RhB is a cationic dye and the TiO2 surface is positively charged at pH 3. Less than 10% of RhB adsorbed on the TiO2 surface at pH 3, whereas about 50% of AO7 adsorbed. The dye adsorption is a prerequisite for the visible-light sensitization and the important factor in determining the dye conversion rate. Much lower conversion rates of both RhB and Cr(VI) in TiO2/Cr(VI)/RhB than in TiO2/Cr(VI)/AO7 might be ascribed to the hindered adsorption of RhB. However, the fact that conversion rates of AO7 and RhB in the binary systems (TiO2/dye) were comparable (Figure 1a vs Figure 3a) implies that the overall mechanism in the ternary system should be more complex. Not only the dye adsorption but also the dye-metal interaction on TiO2 seems to influence the conversion process somehow. The TOC content was reduced by 25% (from 16.2 to 12.2 ppm) after 4-h illumination in both TiO2/RhB and TiO2/Cr(VI)/RhB systems. Conversion of Dyes and Cr(VI) under UV Illumination. The photoconversion of dyes and Cr(VI) under UV illumina2380
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005
tion was also carried out for comparison. Figure 4a compares the simultaneous conversion of AO7 and Cr(VI) in the ternary system with that in the binary system under UV or solarlight illumination. Similar to the case of visible light, the UV-induced conversion rates of AO7 and Cr(VI) in ternary systems were faster than those in binary systems. The initial conversion rates of AO7 and Cr(VI) were about 5.2 and 7.3 µM min-1, respectively, in the ternary system under UV or solar light illumination, which are significantly higher than the visible-light-induced conversion rates. The direct UVinduced conversion of AO7 alone or Cr(VI) alone (without TiO2) was much slower. Since TiO2 as well as AO7 is excited in the presence of UV light and generates not only additional CB electrons but also VB holes that can directly oxidize the dye, the conversion rates of both AO7 and Cr(VI) should be accelerated under UV light. However, it is unexpected to find that the UV-induced TOC removal kinetics (shown in Figure 4b) is very different from the dye-removal kinetics: the TOC reduction in the ternary system was much less than that in the binary system (TiO2/dye). In particular, the TOC in the ternary system remained constant after 1 h illumination, which implies that the catalyst was deactivated in some way. Colo´n et al. (18) also observed a similar phenomenon during the photocatalytic conversion of Cr(VI) and salicylic acid (SA) mixture: the conversion of both species stopped after 2 h UV illumination, whereas near complete conversions could be achieved in the single-component systems (TiO2/ Cr(VI) and TiO2/SA). The main difference between the TiO2/ Cr(VI)/dye system (this work) from TiO2/Cr(VI)/SA system (18) is that the inhibiting effect was observed only for the dye mineralization, not for the Cr(VI) conversion. It seems that
accumulate in the solution. Since H2O2 is a reductant of Cr(VI) at acidic pH (reaction 7) (27),
O2 + 2ecb- + 2H+ f H2O2
(6)
2HCrO4- + 3H2O2 + 8H+ f 2Cr3+ + 3O2 + 8H2O (7)
FIGURE 8. Conversion of (a) RhB (C0 ) 50 µM) and (b) Ag+ (C0 ) 1 mM) in the binary or ternary systems under visible-light illumination. (air-equilibrated; pH ) 3.0; [TiO2] ) 0.5 g/L).
dye degradation intermediates and chromic species form some complexes that subsequently deactivate the surface of TiO2. Not only the ternary/visible light system but also the ternary/UV system is not very efficient in removing TOC of dyes. Dissolved O2 Effect on the Conversion of Dyes and Cr(VI). Metal ions and dissolved oxygen are competing electron acceptors, which suggests that the metal ion reduction could be enhanced in deaerated suspensions. Effects of dissolved O2 on the conversion of dyes and Cr(VI) are compared in Figure 5. Contrary to the expectation, the absence of O2 (under N2 saturation) little changed the removal rate of dyes and decreased the conversion rate of Cr(VI). This indicates that the role of O2 is not simply to scavenge CB electrons to compete with metal ions. The fact that the presence of O2 markedly enhanced the conversion rate of Cr(VI) implies that O2 might serve as an electron-transfer mediator between CB and Cr(VI) or that additional reductants for Cr(VI) can be produced from the photoreaction of O2. For example, it is known that H2O2 is produced through the reduction of O2 by CB electrons (reaction 6). To investigate the role of O2, the visible-light-induced production of H2O2 was monitored in TiO2/AO7 and TiO2/Cr(VI)/AO7 systems. About 30 µM of H2O2 was detected in TiO2/AO7 after 4-h visible-light illumination, whereas the generation of H2O2 was insignificant in TiO2/Cr(VI)/AO7 under the same illumination condition (see Figure S1 in the Supporting Information). This implies that H2O2 should be consumed as soon as it is generated in the ternary system and does not
Cr(VI) can be reduced with a faster rate in the presence of the precursor of H2O2 (i.e., O2). An enhanced Cr(VI) photoreduction in the presence of O2 was also observed in the studies of the ZnO/Cr(VI)/UV system and was ascribed to the role of reaction 7 (17, 28). On the other hand, the reduction of Cr(VI) was reportedly retarded in the presence of O2 in the TiO2/Cr(VI)/UV system at neutral and basic pH (15, 16). The different O2 effects between ZnO and TiO2 systems seem to be due to the different H2O2 formation yield: the UV-induced generation of H2O2 in ZnO suspension is higher than in TiO2 suspension by more than 100 times (29). H2O2 formed on TiO2 is not readily desorbed and degraded rapidly via reacting with VB holes under UV irradiation. Therefore, the role of reaction 7 in reducing Cr(VI) is not important in the TiO2/UV system and O2 competes with Cr(VI) for CB electrons to decrease the photoreduction efficiency. However, the present system uses visible light only, and hence, VB holes are not generated in TiO2. As a result, H2O2 generated on illuminated TiO2 is not oxidized by VB holes, and a significant amount of H2O2 is released into the solution phase, which is subsequently followed by reaction 7. Conversion of AO7 and Ag+. The visible-light-induced conversion of AO7 and Ag+ is shown in Figure 6. The simultaneous and synergistic conversion of AO7 and Ag+ in this ternary system is apparent, as in TiO2/Cr(VI)/AO7. The removal rate of AO7 in TiO2/Ag+/AO7 was, however, slower than that in TiO2/Cr(VI)/AO7. Unlike the case of TiO2/Cr(VI)/AO7, the presence of Ag+ did not change the initial adsorption of AO7 on TiO2 surface, because both Ag+ and the TiO2 surface are positively charged at pH 3. Both AO7 and Ag+ ions in the binary Ag+/AO7 system were also successfully converted under visible light, and this indicates complex formation between AO7 and Ag+ ion. The UV/vis absorption spectrum of the mixed solution of Ag+ and AO7 was different from the arithmetic sum of the spectra of Ag+ and AO7 alone (see Figure S2 in the Supporting Information). Ag+ was not reduced at all in TiO2/Ag+ under visible light, whereas about 50% of Ag+ was reduced in both Ag+/AO7 and TiO2/Ag+/AO7 systems after 4-h illumination. The photoreduced Ag+ ions should be transformed into Ag0 particles (20), whose TEM images are shown in Figure 7a,b. The standard reduction potential of Ag+/Ag0 is 0.80 V (vs NHE), and hence, Ag+ can be directly reduced by either excited AO7* or the CB electron of TiO2. Figure 7a shows the TEM image of Ag deposits (
