Photocatalytic Reduction and Recovery of Copper by - ACS Publications

A series of polyoxometalates PW12O403-, SiW12O404-, and P2Mo18O626- have been used as photocatalysts for recovery of copper and production of fine ...
0 downloads 0 Views 173KB Size
Download PDF
Environ. Sci. Technol. 2002, 36, 5355-5362

Photocatalytic Reduction and Recovery of Copper by Polyoxometalates A . T R O U P I S , †,‡ A . H I S K I A , † A N D E . P A P A C O N S T A N T I N O U * ,† Institute of Physical Chemistry, NCSR Demokritos, 153 10 Athens, Greece, and Chemical Engineering Department, NTU, 157 80 Athens, Greece

A series of polyoxometalates PW12O403-, SiW12O404-, and P2Mo18O626- have been used as photocatalysts for recovery of copper and production of fine metal particles. The process involves absorption of light by polyoxometalates, oxidation of an organic substrate, for instance, propan-2ol as sacrificial reducing reagent, and reoxidation of the reduced polyoxometalates by Cu2+ ions, closing the photocatalytic cycle. Copper(II) ions are reduced to copper(I) and finally to zero-state particles in a 2-electron process, as also suggested by the half-order dependence. Increase of catalyst or propan-2-ol concentration, or both. accelerates the photodeposition of copper until a saturation value is reached. The method is operational at a wide range of copper concentrations varying from 3 to 1300 ppm, leading to very low final concentrations (99.95%) were used for deaeration or oxygenation of solutions. A typical experiment was as follows: 4 mL of aqueous CuSO4 solution containing propan-2-ol and POM catalyst was added to a spectrophotometer cell (1-cm path length), deaerated, and covered with a serum cap. The cell was mounted on a thermostated cell holder, under constant stirring, and the lamp focused with a lens on the contents. The temperature was 18 ( 1 °C. The pH was 5, and the ionic strength was adjusted at 0.1 M with NaClO4. Acidification of the solutions was done, whenever necessary, with HClO4. Analysis was performed using all 4 mL of the photolyzed solution every time. Photolysis was performed with an Oriel 1000-W Xe arc lamp equipped with a cool water circulating filter to absorb the near-IR radiation and a 320-nm cutoff filter in order to avoid direct photolysis of substrates. The incident radiation was reduced to ∼40% with a slit diaphragm in order to obtain reasonable photolysis times. The total photonic flux (320345 nm) determined by ferrioxalate actinometry was 7.9 × 10-6 einstein min-1. As a matter of comparison, for typical experiments (POM 0.7 mM, propan-2-ol 0.2 M, Cu2+ 2 mM), the initial rate of copper recovery was 47, 26, and 0.0 µM min-1 for PW12O403(pH 1), SiW12O404- (pH 0.5-5), and P2Mo18O626- (pH 1), respectively; see below. Under the above experimental conditions, the quantum yields (qy) of formation of the 1-equiv-reduced POM at 254 nm is ∼12% for PW12O404- and 6% for SiW12O405- as has been reported previously (13). For reduction precipitation of Cu2+ to Cu0, it is half of it, i.e., about 6 and 3%, respectively. The qy is independent of wavelength below ∼350 nm. The independence of qy from wavelength has been also verified recently (26). The degree of reduction of POM in photolyzed deaerated solutions was calculated from the known extinction coefficients of the 1-equiv-reduced 12-tungstophosphate, PW12O404- (752 ) 1600 Μ-1 cm-1), 1-equiv-reduced 12-tungstosilicate, SiW12O405- (730 ) 2100 Μ-1 cm-1), and 2-equivreduced 18-molybdodiphosphate, P2Mo18O628- (758 ) 11 000 Μ-1 cm-1), using a Hitachi U-2000 spectrophotometer. The absorption spectra of copper particles or catalyst were taken with a Perkin-Elmer Lambda 19 spectrometer, while the concentration of copper ions in the supernatant solution was measured, after filtration of the entire photolyzed solutions (4 mL), with a GBC flame atomic absorption spectrometer monitored at 324.7 nm. Characterization of the photodeposited copper particles was done with a D-500 Siemens powder X-rays diffractometer using Cu KR radiation, as well as a Philips 200-kV transmition electron microscope. The initial rate of copper recovery was measured from the slope of the curve obtained by monitoring the concentration of Cu2+ in the photolyzed filtered solutions versus time. This was done after the induction period of copper deposition and for the first ∼30% of copper removal using a least-squares treatment (see, for instance, Figure 1). Experimental details are described in the corresponding figures.

Results Blank experiments in the absence of POM were done to examine whether a direct photoreaction between propan2-ol and Cu2+ ions was involved. Photolysis of a deaerated solution containing 1 M propan-2-ol and 3.3 mM Cu2+ (pH 5, NaClO4 0.1 M) for up to 2.5 h did not give any precipitation, 5356

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002

FIGURE 1. Variation of Cu2+ concentration upon irradiation of deaerated aqueous solutions containing 0.5 M propan-2-ol, 0.07 mM K4SiW12O40, 2 mM Cu2+ at pH 5, 0.1 M NaClO4 (λ > 320 nm, T 18 °C), with illumination time. whereas no change in copper concentration was observed as measured by AAS. Prolonged irradiation, for 3 h, of a deaerated solution of 0.7 mM POM and 2 mM Cu2+ (pH 5, NaClO4 0.1 M) had no effect on the concentration of copper ions. Moreover, no change in the absorption spectra of POM was observed when mixed with Cu2+ ions. These last data exclude the possibility of a POM-Cu2+ complex interference in the photocatalytic process. Thermal Redox Reactions between Reduced POM and Cu2+. Deaerated solutions of blue, 1-equiv-reduced tungstates, 0.5 mM PW12O404- or SiW12O405-, produced after photolysis of aqueous solutions of 1.0 mM propan-2-ol, 0.7 mM POM, and 0.1 M HClO4, were mixed with a deaerated solution of 1 mM Cu2+. The process was monitored at ∼750 nm, characteristic absorbance of the 1-equiv-reduced tungstates. Upon mixing, a sudden decrease of the absorbance was observed within a second time frame, indicating that the 1-electron-reduced POM are rapidly reoxidized by copper ions in the absence of light. On the contrary, nothing happened in a similar experiment with the 2-equiv-reduced molybdate (P2Mo18O628-). In a series of thermal experiments between SiW12O405- (0.25 mM) and O2 (0.6 mM), reoxidation of the 1-electron-reduced POM took place within a few minutes (half-life 1.4 min), (0.1 M NaClO4 , pH 5) (20), whereas when Cu2+ (0.6 mM) was used instead of O2, the reaction was complete in less than 2 s. After total removal of copper, the photolyzed solution was filtered and the POM concentration was measured spectrometrically at 265 nm to verify that the catalyst remained intact through the whole process. Photocatalytic Reduction and Recovery of Copper. A deaerated solution containing 2 mM copper sulfate (127 ppm), 0.5 M propan-2-ol, and 0.07 mM SiW12O404- was irradiated with λ > 320 nm to examine the role of POM in photocatalytic reduction and recovery of copper. Small copper particles were formed after the first 12 min of photolysis, with an initial rate of ∼61 µM min-1. Less than 0.2 ppm copper was left in the solution after 150 min of illumination (Figure 1). Action spectra during photolysis are exhibited in Figure 2. The formation of copper particles is indicated by the gradual elevation of the baseline (scattering), whereas at 115 min of photolysis, the peak absorbance at 595 nm is characteristic of metal copper aggregates. Only after almost 120 min of photolysis, when the solution has been practically depleted of Cu2+, it gradually turns blue (peak absorption 752 nm) due to the formation of the 1-electronreduced catalyst SiW12O40.5-

FIGURE 2. UV-visible spectra of irradiated solutions for various irradiation times, showing the formation of metal copper particles and the 1-equiv-reduced tungstate, SiW12O405-, at 140 min. Experimental conditions as in Figure 1. Inset: detail of spectrum after 115 min of photolysis, indicating the formation of metal copper particles.

FIGURE 4. Influence of SiW12O404- concentration on the initial rates, R0, of (a) copper recovery or (b) formation of 1-equiv-reduced tungstosilicate (SiW12O405-) in the absence of Cu2+. Deaerated solutions 0.5 M propan-2-ol, 2 mM Cu2+, pH 5, and 0.1 M NaClO4 (λ > 320 nm, T 18 °C).

FIGURE 3. X-rays pattern of copper particles, photodeposited in a deaerated aqueous solution containing 0.5 M propan-2-ol, 0.7 mM K4SiW12O40, 7 mM Cu2+, and 0.1 M NaClO4 at pH 5 (λ > 320 nm, T 18 °C).

FIGURE 5. Influence of propan-2-ol concentration on the initial rate, R0, of copper recovery in deaerated or nondeaerated photolyzed solutions: 0. 7 mM K4SiW12O40, 2 mM Cu2+ at pH 5, and 0.1 M NaClO4 (λ > 320 nm, T 18 °C).

Characterization of Copper Particles. The precipitate produced after irradiation of PW12O403- or SiW12O404- copper solutions in the presence of propan-2-ol was of reddish color, characteristic for copper particles. Copper was characterized by a plasmon resonance absorption peak around 580 nm, attributed to interband transitions (27). A representative absorption spectrum of photolyzed PW12O403- or SiW12O404-/ Cu2+/propan-2-ol solutions is exhibited in the inset of Figure 2, with a peak at 595 nm, indicating the photoproduction of copper particles in the system. Powder XRD measurements revealed that copper ions were thoroughly reduced to zero-state copper with an fcc structure, whereas no formation of Cu2O was observed (2θ ) 25.9 not shown in the graph) (Figure 3). Transmission electron microscopy was used to obtain metal particle images and size distribution (400 nm-2 µm). Effect of POM Concentration. Figure 4 exhibits the influence of catalyst concentration on the initial rate of copper recovery (lower curve) and the initial rate of formation of the 1-equiv-reduced tungstosilicate (SiW12O405-) in the absence of Cu2+ (upper curve). For concentrations of SiW12O404- below ∼0.2 mM, a linear dependence is observed for both rates. For concentrations of SiW12O404- greater than 0.8 mM, the system reaches, practically, saturation in photon absorption and both rates become zero order with respect to catalyst.

Catalyst concentration affects the induction period before copper starts precipitating. For instance, increase of SiW12O404concentration from 0.1 to 6.5 mM lowers the induction time of copper formation from 13 to 1 min. Effect of Substrate Concentration. Figure 5 illustrates the variation of the initial rate of photocatalytic copper recovery with the concentration of organic substrate, in the presence or absence of oxygen. In both cases, there is a linear dependence on [propan-2-ol] for concentrations below ∼1 M and zero-order dependence for concentrations over ∼2 M. Moreover, this behavior is independent of dioxygen. It should be noted that copper recovery starts after an induction period, as mentioned earlier. The maximum rate of copper recovery is ∼1.75 × 10-4 M min-1 (Figure 5). The induction period, prior to copper formation, is a function of the concentration of propan-2-ol. For instance, increase of propan-2-ol concentration from 0.1 to 2.0 M lowers the induction time of copper formation from ∼17 min to less than 2 min. Effect of O2. Dioxygen influences the induction period observed before the commencement of copper formation. The induction period is prolonged by 25 min in airequilibrated solutions, whereas, the initial rate of copper precipitation is, practically, not altered as shown in Figure 5. VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5357

FIGURE 6. Effect of copper ions concentration on the initial rate of copper recovery: 0.7 mM K4SiW12O40, 0.5 M propan-2-ol at pH 5, and 0.1 M NaClO4 (λ > 320 nm, T 18 °C).

TABLE 1. Influence of Several Anions, ClO4-, Cl-, NO3-, or CH3COO- on the Time Elapsed before Copper Starts To Precipitate (Appearance of Red Copper Particles), and the Time Signaling the End of Copper Removal (i.e., Beginning of Development of the 1-Equiv-Reduced Blue Tungstate, SiW12O405-) anion, concn (M) ClO4- 0.1 ClO4- 0.7 NO3- 0.1 Cl- 0.01 Cl- 0. 1 M CH3COO0.1 M (pH 6-6.5) CH3COO0.1 M (pH 3)

start of Cu precipitation, min

end of Cu precipitation (formation of blue), min

3.5 3.5 3 4 10 18.5 yellow soln

35 35 39.5 39 35 17.5 (formation of blue)

3.5

35

FIGURE 7. Production of 1-equiv-reduced SiW12O404- after illumination of a deaerated aqueous solution containing 0.5 M propan-2-ol, 1.2 mM K4SiW12O40, and 0.1 M NaClO4 at pH 5 (λ > 320 nm, T 18 °C), in the absence or in the presence of 2 mM Cu2+. (right curve) is identical with the preceding experiment without Cu2+.

Discussion The following reactions summarize the mechanistic scheme suggested to take place during photocatalytic recovery of copper in the presence of POM (SiW12O404-):

POM + hν f POM*

I (absorbed intensity)

POM * f POM

k2

Effect of Copper Ion Concentration. The initial rate of copper recovery is also a function of [Cu2+] (Figure 6), whereas the induction time prior to copper photodeposition increases gradually with copper ion concentration from less than 1 min for 1 mM Cu2+ to ∼8 min for 20 mM Cu2+. Anion Effect on Copper Recovery. Deaerated solutions of 0.5 M propan-2-ol, 0.7 mM SiW12 O404-, and 2 mM copper sulfate, containing the sodium salt of perchlorate, nitrate, chloride, or acetate (at pH 5), were photolyzed in order to examine the role of these anions in the photocatalytic process. Table 1 illustrates the effect of these anions in (a) the time elapsed before copper starts to precipitate (appearance of reddish particles) and (b) the time signaling the end of copper removal (beginning of development of the 1-electron-reduced blue tungstate, SiW12O405-. Stability Tests of the Catalyst. Reproduction of the Process. Figure 7 presents another indication that the catalyst is unaffected by the recovery of copper. In this figure, the percent 1-equiv-reduced tungstosilicate (SiW12O405-) is monitored as a function of photolysis time. In one experiment, a deaerated propan-2-ol/SiW12O404is photolyzed, and the percent formation of SiW12O405- is observed (left curve). In the other experiment, Cu2+ has been added to the deaerated solution (propan-2-ol/SiW12O404-/Cu2+) and the percent formation of SiW12O405- is also observed as a function of photolysis time. After a period in which all Cu2+ has been removed from the solution, the rate of SiW12O405- formation 5358

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002

(2)

POM * + propan-2-ol f POM- + oxidized products POM- + Cu2+ f POM + Cu+ 2Cu+ a Cu2+ + Cu0 V

(1)

k4 K5

k3 (3) (4) (5)

and/or

POM- + Cu+ f POM + Cu0 V

k6

(6)

2POM- + O2 f 2POM + O22-

k7

(7)

Cu++ O2 f Cu2+ + O2-

k8

(8)

Reactions 1-3 provide an overall account of the photocatalytic oxidation of propan-2-ol. Details, i.e., formation of radicals involved in the process, etc., have been reported elsewhere (14, 28). When copper ions are in excess, a fast reoxidation of the reduced catalyst occurs (reactions 4 and 6) (as has also been shown in the blank experiments in the absence of light) and no blue color of the 1-equiv-reduced tungstate is observed. When the solution is depleted of Cu2+, the blue SiW12O405- is gradually developed as has been mentioned earlier. Deposition of copper as zero-state metal particles can be achieved either through disproportionation of Cu+ (eq 5) or via further reduction by the reduced catalyst (eq 6). Applying the steady-state approximation to the concentrations of the excited POM and reduced POM, gives

[POM*] ) I/(k2 + k3[S]) and

1/2

FIGURE 8. Linearity of (ppm Cu) as a function of illumination time, for various concentrations of SiW12O404-. Experimental conditions as in Figure 4. -

2+

+

[POM ] ) k3[POM*][S]/(k4[Cu ] + k6[Cu ])

(9)

FIGURE 9. Linear plot of R0-1 ([propan-2-ol]0-1) in deaerated or nondeaerated illuminated solutions. Experimental conditions as in Figure 5. Equation 12, for initial conditions and standard initial concentration of copper ions, becomes

d[Cu2+]/dt ) R0 ) k3I[S]0B/(k2 + k3[S]0)

Substituting [POM*] and taking into consideration the equilibrium constant of eq 5, where

K5 ) [Cu2+]/[Cu+]2

B ) k4K51/2[Cu2+]01/2/k6 ) constant

the steady-state concentration of reduced POM becomes Hence

[POM-] ) k3I[S]/(k2 + k3[S])(k4[Cu2+] +

k6K5-1/2[Cu2+]1/2) (10)

If we consider

k6K5-1/2[Cu2+]1/2 >> k4[Cu2+] eq 9 gives

[POM-] ) k3I[S]/(k2 + k3[S])k6K5-1/2[Cu2+]1/2 (11) From eq 4 the rate of Cu2+ reduction is

d[Cu2+]/dt ) k4[POM-][Cu2+] and substituting [POM-] from eq 11 we obtain

(14)

A plot of 1/R0 versus 1/[propan-2-ol]0 gives a straight line, in accordance with the proposed mechanistic scheme (Figure 9). It should be noted that during photooxidation of organic substrates by POM it is not necessary for the mechanism to involve a Langmuirian adsorption of the organic on POM in order to fit the linearity of 1/R0 as a function of 1/[S]0. This is rather a kinetic phenomenon, as has also been reported in TiO2 photocatalysis (29). It is interesting to note that the ratio of rates of SiW12O405formation and copper precipitation is 2:1, reflecting the overall stoichiometry of the reaction (Figure 4)

2SiW12O405- + Cu2+ f 2SiW12O404- + Cu0

d[Cu2+]/dt ) k3I[S]/(k2 + k3[S])k4K51/2[Cu2+]1/2/k6 (12) Integration of the last equation gives

[Cu2+]1/2 ) A/2t

1/R0 ) k2/k3I[S]0B + 1/IB

(13)

where

A ) k3I[S]/(k2 + k3[S])k4K51/2/k6 Equation 12 suggests a linear dependence of the square root of copper concentration with photolysis time, which is in accordance with the experimental data; Figure 8. Straight lines with good linearity, r > 0.990, for various concentrations of the catalyst and standard concentration of S are obtained. This pseudo-half-order model suggests a two 1-electron process for copper reduction, in accordance with similar results with TiO2 (4). The rate of copper removal with increasing propan-2-ol concentrations follows a Langmuirian behavior (Figure 5).

(15)

Induction Period in Copper Precipitation. Copper does not precipitate immediately when the solution is irradiated with UV-near-visible light. There is an induction period in copper photodeposition, which is a function of copper, catalyst, and propan-2-ol concentration. Increase in the concentration of the organic substrate or the catalyst accelerates the formation of copper particles, whereas, increase in copper ion concentration retards the induction period, as mentioned earlier. The results may be explained qualitatively as follows: The existence of an induction period in copper removal denotes that Cu2+ reduction is not a one-step process, suggesting that some intermediate, i.e., Cu+ ion, is involved, as the preceding mathematical treatment has also shown. Higher Cu2+ ion concentration favors eq 4, lowers the concentration of reduced catalyst, and shifts eq 5 to the left. In addition, (eq 6) slows down because of the competitive eq 4. On the contrary, higher concentrations of the catalyst, propan-2-ol, or both accelerate the formation of POM- (eq VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5359

3), as well as the recovery of copper, because larger quantities of Cu+ are produced through eq 4, favoring eq 5. A process with two different rates in copper recovery has also been observed in the TiO2 photocatalytic recovery of copper (30, 31). Effect of Oxygen. Oxygen is known to react fast with many reduced POM (eq 7) (20). However, as thermal experiments indicated, Cu2+ ions reoxidize SiW12O405- with a rate of more than 2 orders of magnitude greater than O2. Therefore, dissolved oxygen is not expected to compete efficiently with copper ions in abstracting electron from the reduced POM, as shown in Figure 5. Alternatively, O2 molecules have the ability to reoxidize Cu+ ions to Cu2+ (eq 8), preventing their further reduction to copper particles, thus prolonging the induction period until dioxygen has been consumed. The higher redox potential, E° (O2/OH-) ) 0.41 V versus NHE, compared with E° (Cu2+/Cu+) ) 0.153 V versus NHE, favors reaction 8, reoxidizing Cu+ to Cu2+ (32). Simple calculations indicate that copper starts to precipitate out after all dioxygen of the solution (∼1.2 mM for dioxygen saturated or ∼0.2 mM for air equilibrated aqueous solutions) has been consumed by the photogenerated 1-equiv-reduced tungstosilicate (SiW12O405-) (eq 7). Thus, dioxygen retards the start of photodeposition of copper until is totally consumed. Then, the rate of copper recovery is the same as in the absence of dioxygen (Figure 5), as mentioned earlier. Influence of Copper Concentration on Copper Recovery. In the absence of Cu2+, the initial rate of formation of the 1-equiv-reduced tungstosilicate, SiW12O405-, is 2.1 × 10-6 M s-1 (other conditions as in Figure 6). The reoxidation of reduced POM by Cu2+ ions occurs in less than 1 s, as mentioned earlier. Since the concentration of copper ions in these experiments is in excess (>5.0 × 10-5 M, Figure 6), early saturation is observed in the rate of copper recovery. As can also be seen in Figure 6, the process is effective in a wide range of copper concentrations, varying from 3 to 1320 ppm, whereas prolonged irradiation leads to complete decontamination from Cu2+ ions up to nondetected traces (detection limit 0.2 ppm). The wide range of copper concentrations at which the POM method is applied, as well the low final concentration of Cu2+ ions left in solution, can account for the environmental assessment of the process. Effect of Catalyst Redox Properties on Copper Recovery. Three POM catalysts, PW12O403-, SiW12O404-, and P2Mo18O626-, with different redox potentials (0.221, 0.057, and 0.664 V vs NHE for the initial 1-, 1-, and 2-electron reduction, respectively), as mentioned earlier, were used to examine the role of redox properties of POM anions in reduction and recovery of copper. The development of photoreduced blue POM versus illumination time, in the absence of dioxygen and Cu2+, is shown in Figure 10 (curves A-C). The reduced species, PW12O404-, SiW12O405-, and P2Mo18O628-, are followed spectrometrically until a steady state is reached. In these experiments, the relative rate of formation of the three reduced POM (i.e., eqs 1-3) has been adjusted to be the same in order to compare the rates of reaction 4. Thus, when the catalyst has lower relative rate of formation (13, 33), we add more concentration of propan-2-ol to equalize the velocity of reaction 3 and obtain equal rates of formation of reduced POM. Experiments have shown that the appropriate concentrations of propan-2-ol are 10-3, 0.2, and 5 M for PW12O403-, SiW12O404-, and P2Mo18O626-, respectively. Introduction of Cu 2+ in the system and photolysis of a propan-2-ol/POM/Cu 2+ solution led to the following results: No change in the rate of P2Mo18O628- reduction was observed (Figure 10, curve A), indicating that no reaction took place between the P2Mo18O628- and Cu 2+. This result 5360

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002

FIGURE 10. Production of 1-, 1-, or 2-equiv-reduced catalyst (i.e., PW12O40,4- SiW12O405- and P2Mo18O628-, respectively) after photolysis of a deaerated POM/propan-2-ol solution in the absence (curves A-C) or in the presence of 2 mM Cu2+ (curves A, D, E). pH 1, λ > 320 nm, T 18 °C. Initial concentrations: POM, 0.7 mM; propan-2-ol 1 × 10-3, 0.2, and 5 M for PW12O403-, SiW12O404- and P2Mo18O626respectively; see text.

FIGURE 11. Decontamination from Cu2+ ions in deaerated POM/ propan-2-ol/solutions, using three different POM catalysts H3PW12O40, K4SiW12O40, and Na6P2Mo18O62. Experimental conditions as in Figure 10. is in parallel with the thermal experiment between reduced P2Mo18O628- and copper ions, which did not give any reaction. On the contrary, in the case of PW12O403-, a lowering of the PW12O404- concentration took place (Figure 10, curve D), in comparison with the corresponding experiment in the absence of Cu2+ ions (Figure 10, curve B). In the case of 12-tungstosilicate, the concentration of SiW12O405- was reduced to zero (Figure 10, curve E). Only after all copper precipitated out, in ∼80 min, was formation of the blue 1-equiv-reduced tungstosilicate, SiW12O405-, observed. In addition, photolysis of PW12O403- or SiW12O404- solutions led to formation of copper particles after an induction period of 32 or 10 min, respectively, whereas no copper particles were observed during photolysis of a propan-2-ol/P2Mo18O626-/Cu 2+ solution (Figure 11). This figure also shows that when the concentrations of reduced catalysts in the solution (PW12O404-, SiW12O405-, P2Mo18O628-) are adjusted to be the same, the relative rates of copper recovery follow the order SiW12O405- > PW12O404- > P2Mo18O628-. This reflects the thermal rates of reaction 4 and are in accordance with the redox potentials of POM: E° (SiW12O404-/5-) ) 0.057 V versus NHE, E° (PW12O403-/4-) ) 0.221 V versus NHE and E° (P2Mo18O626-/8-) ) 0.664 V versus NHE.

solution of Cu2+, propan-2-ol, and poly(vinyl alcohol) (as colloidal stabilizer), adjusted at pH 6.2 (35). In support of this, when the acetate solution was acidified with HClO4 at pH 3, no yellow color was formed during photolysis and the same results were observed as in the absence of acetate ions.

Acknowledgments We thank Ministry of Development, General Secretariat of Research and Technology of Greece for supporting part of this work. We thank A. Bourlinos for his help with the XRDpattern.

Literature Cited

FIGURE 12. Lowering of copper concentration in deaerated photolyzed solution of 0.5 M propan-2-ol, 0.7 mM K4SiW12O40, and 2 mM Cu2+, ionic strength adjusted at 0.5 M (λ > 320 nm, T 18 °C), with various pH values. (Only the values at pH 5 are shown since the values at pH 3 and 1 are practically identical). Thermal experiments have also shown that, according to the equilibrium

PW12O404- + Cu2+ a Cu+ + PW12O403∼50% of PW12O404- is finally reoxidized by copper ions (23). In contrast, there is no reaction between P2Mo18O628- and Cu2+, as mentioned earlier. This reaction is thermodynamically forbidden as the reduction potential of the P2Mo18O626-/ P2Mo18O628- pair, 0.664 V, is much higher than those of Cu2+/ Cu° or Cu2+/Cu+ (0.337 and 0.153 V vs NHE). pH Effect on Copper Recovery. In Figure 12, the profile of copper ion depletion with photolysis time is shown as a function of several pH values. As can be seen, variation of pH from 5 to 0.3 did not alter the rate of copper recovery during photolysis of a SiW12O404-/propan-2-ol/Cu2+ solution in which the ionic strength was adjusted at 0.5 M. This is expected since the redox potentials for both copper reduction (4) and SiW12O404- reduction either electrolytically (24) or photochemically (34) are independent of pH. Anion Effect on Copper Recovery. Table 1 shows that introduction of perchlorate or nitrate sodium salts at 0.1 M concentration, did not influence the photodeposition of copper. On the contrary, an inhibition in copper precipitation was observed when 0.01 M sodium chloride was added. Formation of copper red particles started after 10 min of illumination, instead of 4 min in the absence of chloride ions. In addition, no blue color of the 1-equiv-reduced 12tungstosilicate, SiW12O405-, was formed until 35 min of photolysis. Addition of more chloride ions (0.1 M) caused further delay (17.5 min) in the photodeposition of copper, while the solution turned blue at the same time. It is known that chloride ions stabilize Cu+ and lower the reduction potential Cu+ /Cu0 from 0.521 to 0.137 V versus NHE, slowing the reduction to Cu0. Increase in the concentration of chloride ion further stabilizes Cu+ and the disproportionation (eq 5) shifts more to the left. When copper starts to precipitate through the slow eq 6, Cu0 is in coexistence with the 1-electron-reduced catalyst (SiW12O405-) and the blue color appears in the solution. When copper acetate (0.1 M) was added in the copper sulfate/propan-2-ol/SiW12 O404- solution and photolyzed, it turned yellow. The pH of the solution was measured to be 6-6.5. The yellow color of the photolyzed solution could be assigned to colloidal cuprous hydroxide, as has also been reported in the case of 60Co γ-irradiation of an aqueous

(1) Patterson, J. W. Director of the Industrial Waste Elimination Research Centre, Illinois Institute of Technology, as quoted in: Haggin, J. Chem. Eng. News 1986, (Sept 8), 37. (2) Salomons, W., Fo¨rstner, U., Mader, P., Eds. Heavy Metals, Problems and Solutions; Springer: Berlin, 1995, 386. (3) EC (European Communities). Council directive 98/83 on the quality of water intended for human consumption, L330/98, November 3, 1998. (4) Litter, M. I. Appl. Catal. B 1999, 23, 89. (5) Serpone, N.; Lawless, D.; Terzian, R.; Minero, C.; Pelizzetti, E. In Photochemical Conversion and Storage of Solar Energy; Pelizzetti, E., Sciavello, M., Eds.; Kluwer Academic Publ.: Dordrecht, The Netherlands, 1991; p 451. (6) Fujita, H.; Izawa, M.; Yamazaki, H. Nature 1962, 196, 666. (7) Marignier, J. L.; Belloni, J.; Delcourt, M. O.; Chevalier, J. P. Nature 1985, 317, 344. (8) Nagata, Y.; Watananabe, Y.; Fujita, S-i.; Dohmaru, T.; Taniguchi, S. J. Chem. Soc., Chem. Commun. 1992, 1620. (9) Yeung, S. A.; Hobson, R.; Biggs, S.; Grieser, F. J. Chem. Soc., Chem. Commun. 1993, 378. (10) Prairie, M. R.; Evans, L. R.; Stange, B. M.; Martinez, S. L. Environ. Sci. Technol. 1993, 27, 1776. (11) Byrne, J. A.; Eggins, B. R.; Byers, W.; Brown, N. M. D. Appl. Catal. B 1999, 20, L85. (12) Huang, M. H.; Tso, E.; Datye, A. K.; Prairie, M. R.; Stange, B. M. Environ. Sci. Technol. 1996, 30, 3084. (13) Papaconstantinou, E. Chem. Soc. Rev. 1989, 16, 1 and references therein. (14) Hiskia, A.; Mylonas, A.; Papaconstantinou, E. Chem. Soc. Rev. 2001, 30, 62. (15) Mylonas, A.; Roussis, V.; Papaconstantinou, E. Polyhedron 1996, 15, 3211. (16) Mylonas, A.; Papaconstantinou, E. J. Mol. Catal. 1994, 92, 261. (17) Mylonas, A.; Papaconstantinou, E. J. Photochem. Photobiol., A 1996, 94, 77. (18) Androulaki, E.; Hiskia, A.; Dimotikali, D.; Minero, C.; Calza, P.; Pelizzetti, E.; Papaconstantinou, E. Environ. Sci. Technol. 2000, 34, 2024. (19) Hiskia, A.; Androulaki, E.; Mylonas, A.; Boyatzis, S.; Dimotikali, D.; Minero, C.; Pelizzetti, E.; Papaconstantinou, E. Res. Chem. Intermed. 2000, 26, 235. (20) Hiskia, A.; Papaconstantinou, E. Inorg. Chem. 1992, 31, 163 and references therein. (21) Ioannidis, A.; Papaconstantinou, E. Inorg. Chem. 1985, 24, 439. (22) Amadelli, R.; Varani, G.; Maldotti, A.; Carassiti, V. J. Mol. Catal. 1990, 59, L9. (23) Troupis, A.; Hiskia, A.; Papaconstantinou, E. New J. Chem. 2001, 25, 361. (24) Pope, M. T.; Varga, G. M. Inorg. Chem. 1966, 5, 1249. (25) Pope, M. T. Heteropoly and Isopoly Oxometalates, Inorganic Chemistry Concepts 8; Springer-Verlag: West Berlin, 1983. (26) Ozer, R. R.; Ferry, J. L. J. Phys. Chem. 2002, 106, 4336. (27) Creighton, J. A.; Eadon, D. G. J. Chem. Soc., Faraday Trans. 1991, 87, 3881. (28) Mylonas, A.; Hiskia, A.; Androulaki, E.; Dimotikali, D.; Papaconstantinou, E. Phys. Chem. Chem. Phys. 1999, 1, 437. (29) Bahnemann, D.; Cunningham, J.; Fox, M. A.; Pelizzetti, E.; Pichat, P.; Serpone, N. In Aquatic and Surface Photochemistry; Helz, G. R., Zepp, R. G., Grosby, D. G., Eds.; Lewis Publ.: Boca Raton, FL, 1994; p 261. (30) Foster, N. S.; Lancaster, A. N.; Noble, R. D.; Koval, C. A. Ind. Eng. Chem. Res. 1995, 34, 3865. VOL. 36, NO. 24, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5361

(31) Foster, N. S.; Noble, R. D.; Koval, C. A. Environ. Sci. Technol. 1993, 27, 350. (32) Zang, L.; Liu, C.-Y.; Ren, X.-M. J. Chem. Soc., Faraday Trans. 1995, 91. 917. (33) Dimoticali, D.; Papaconstantinou, E. Inorg. Chim. Acta 1984, 87, 177. (34) Hiskia, A. Doctoral Dissertation 1989, University of Athens, p 128.

5362

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 24, 2002

(35) Khatouri, J.; Mostafavi, M.; Amblard, J.; Belloni, J. Chem. Phys. Lett. 1992, 191, 351.

Received for review September 10, 2002. Revised manuscript received October 8, 2002. Accepted October 8, 2002. ES020933Q