Environ. Sci. Technol. 2001, 35, 2381-2387
Study of Au/Au3+-TiO2 Photocatalysts toward Visible Photooxidation for Water and Wastewater Treatment X. Z. LI* AND F. B. LI Department of Civil and Structural Engineering, The Hong Kong Polytechnic University, Hunghom, Kowloon, Hong Kong
With an attempt to extend light absorption of TiO2-based photocatalyst toward the visible light range and eliminate the rapid recombination of excited electrons/holes during photoreaction, a new type of photocatalysts (Au/Au3+TiO2) powder was prepared by a photoreduction/sol-gel process. The crystal phase composition, surface structure, and light absorption of the new photocatalysts were comprehensively examined by X-ray differential detection (XRD), UV-visible absorption spectra, X-ray photoelectron emission spectroscopy (XPS), and photoluminescence (PL) spectra. The photooxidation efficiencies of the photocatalysts were also evaluated in the photodegradation of methylene blue (MB) in aqueous solutions under visible light irradiation from a high-pressure sodium lamp (λ > 400 nm). The results of PL analyses in this study indicated that the gold/ gold ion-doping on the surface of TiO2 could eliminate the electron/holes recombination and also increase the light absorption in the visible range. The analytical results of UVvisible diffuse reflection spectra (DRS) and optical absorption spectra indicated that a new energy level below 3.2 eV generated in the Au/Au3+-TiO2 promoted the optical absorption in the visible region and made it possible to be excited by visible light (E < 3.2 eV). The experiment demonstrated that the photooxidation efficiency of MB using the Au/Au3+-TiO2 powder were significantly higher than that using conventional TiO2 powder and an optimum molar content of gold doping/deposition in the TiO2 was 0.5%. The development of such photocatalysts may be considered a breakthrough in large-scale utilization of solar energy to address environmental needs.
Introduction TiO2 photocatalysis has been the focus of numerous investigations in recent years, particularly owing to its application for the mineralization of undesirable organic contaminants to CO2, H2O, and inorganic constituent (1). Among various metal oxide semiconductor materials, TiO2 is probably the most widely studied material in the field of photocatalysis because of its favorable physical/chemical properties, low cost, ease of availability, and high stability. Although the technique is gradually nearing the stage of preindustrial application, there are still fundamental problems concerning the efficiency of photocatalysis that need to be solved. First, * Corresponding author phone: (852) 2766-6016; fax: (852) 2334 6389; e-mail:
[email protected]. 10.1021/es001752w CCC: $20.00 Published on Web 05/03/2001
2001 American Chemical Society
TiO2 is a high band gap (Eg ≈ 3.2 eV) material that can only be excited by high energy UV radiation with a wavelength of no longer than 387.5 nm. This practically rules out the use of sunlight as an energy source (2). Second, a low rate of electron transfer to oxygen and a high rate of recombination between electron/hole pairs result in a low quantum yield rate and also a limited photooxidation rate (3). Up to now, photocatalysis studies have focused on seeking an effective way to eliminate the recombination of electrons and holes by increasing the charge separation, and also a way to extend the wavelength range response by photosensitization. Numerous investigations have reported that the addition of group VIII metals and transition metal ions to TiO2-based photocatalytic systems are two effective ways to enhance the photocatalytic reaction rate (3). The noble metals including gold (4, 5) and platinum (6, 7) are capable of producing the highest Schottky barrier among the metals facilitating the electron capture. Choi and co-workers in 1994 (8) presented the results of a systematic study of the effects of 21 different metal ion dopants on the photochemical reactivity of quantum-sized TiO2 with respect to both chloroform oxidation and carbon tetrachloride reduction. On the other hand, surface sensitization of TiO2 via chemisorbed or physisorbed dyes including porphyrin, phthalocyanine, and polypyrrole can increase the efficiency of the excitation process and expand the wavelength range in the excitation of the photocatalysts through excitation of the sensitizer followed by charge transfer to the TiO2 (9, 10). However, the photosensitizers may be gradually photodegraded and become less effective. Recently, the photodegradation of several dyes through exposure to visible light in the presence of TiO2 was reported by Zhang et al. (11), Zhao et al. (12), Wu et al. (13), Liu et al. (14) and Liu et al. (15). They thought the dye, not the TiO2, was excited by visible light and that the excited dye injected an electron into the conduction band of TiO2, from where it was scavenged by preadsorbed oxygen to form active radicals. Moreover, several studies focused on the optical properties of TiO2 film (16, 17) and Pb-doped TiO2 film (18, 19). The Pd impurity makes it possible for TiO2 to absorb visible light. The impurities energy level such as V, Cr, Mn and Fe doped in single crystalline rutile TiO2 was investigated by photocurrent measurement and theoretical calculations, and the broad absorption band in visible range was ascribable to metal ion doping (20, 21). However, up to now, the relationship between optical properties and photooxidation efficiency has been rarely studied. To further improve the TiO2-based photooxidation for water and wastewater treatment, this study was aimed to modify the pure TiO2 powder by doping either gold or gold ion (Au or Au3+) to (i) extend the light absorption spectrum into the visible region; (ii) hinder the recombination of electron/hole pairs; and (iii) modify the surface properties by enlarging the specific surface of photocatalyst and adjusting the zero charge point.
Experimental Section Preparation of Photocatalysts. The gold ion-doped TiO2 samples (Au3+-TiO2) were prepared by the sol-gel method. In which 17 mL of titanium(IV) butoxide (Ti(O-Bu)4) dissolved in 80 mL of absolute ethanol was added dropwise under vigorous stirring to 100 mL of the mixture solution containing 80 mL of 95% ethanol, 5 mL of 0.1 M tetrachloroauric acid (HAuCl4‚4H2O) and 15 mL of acetic acid (>99.8%). The resulting transparent colloidal suspension was stirred for 2 h and aged for 2 days until the formation of gel. VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2381
The gel was dried at 353 K under vacuum and then ground. The powder was calcined at 923 K for 2 h, and then golddoped TiO2 was obtained in a nominal atomic doping level of 1.0% labeled as 1% Au3+-TiO2. Other gold ion-doped TiO2 samples were also prepared by the above procedure labeled as 0.25% Au3+-TiO2, 0.5% Au3+-TiO2, 3% Au3+-TiO2, and 5% Au3+-TiO2, respectively. The 0.5% Au3+-TiO2 sample was treated by H2 gas in a tube-furnace for 2 h labeled as 0.5% Au3+(H2)-TiO2. Pure quantum-sized TiO2 was prepared without addition of tetrachloroauric acid. The gold-doped TiO2 samples (Au-TiO2) were prepared by the photoreduction method (4). In which the weighted amount of TiO2 was suspended in a mixture solution containing the required concentration of tetrachloroauric acid and 0.1 M methanol solution, as a hole scavenger. The suspensions were irradiated with a 125 W high-pressure mercury lamp for 60 min. The gold concentration in the mixture solution was detected by PS-1000AT ICP. The doped amount of gold on TiO2 was determined by the loss of gold concentration in the solution. The Au-TiO2 samples were separated by filtration and washed repeatedly with distilled water and dried at 403 K for 24 h. The gold concentration in the resulting samples were 0.25, 0.5, 1.0, 2.0, 4.0, 5.0%, respectively. All the dopant contents mentioned in this work were the nominal molar content. Characterization of Photocatalysts. To determine the crystal phase composition of the prepared photocatalysts (Au/Au3+-TiO2), X-ray diffraction (XRD) was carried out at room temperature using a Rigaku D/MAX-IIIA diffractometer with Cu KR radiation (λ ) 0.15418 nm). The accelerating voltage of 35 kV and emission current of 30 mA were used. To determine the iso-electric points of the photocatalysts, the zeta(σ)-potential of photocatalyst samples were measured using a Brook-haven Zeta Plus analyzer, in which 1 mM aqueous solution of potassium nitrate was used as suspension, and the photocatalysts were added into the suspension to make the solution to 1 g L-1. During the zeta-potential analysis, the pH value of the suspension was adjusted by adding either 0.1 M nitric acid or 0.1 M potassium hydroxide solution. To study the light absorption of the photocatalysts, the diffuse reflectance spectra (DRS) of the samples in the wavelength range of 260-460 nm was obtained using a spectrophotometer (Perkin-Elmer UVW-340), while MgO was used as reference. To determine the light absorption band after the Au/Au3+ doping in the structure of TiO2, a thin film of photocatalysts was prepared, and the UV-visible absorption spectra of the photocatalysts were measured in the range of 250-700 nm using a UV-visible scanning spectrophotometer (Shimadzu UV-2101 PC). To study the recombination of electrons/holes in the photocatalysts, the photoluminescence (PL) emission spectra of the samples were measured in the following procedure: at a temperature of 10 K, a 325 nm He-Cd laser was used as an excitation light source. The light from the sample was focused into a spectrometer (Spex500) and detected by a photomultiplier tube (PMT). The signal from the PMT was input into a photon counter (SR400) before being recorded by a computer. To study the valance state of the photocatalysts, X-ray photoelectron spectroscopy (XPS) was recorded with the PHI Quantum ESCA Microprobe system, using the MgKR line of a 250W Mg X-ray tube as a radiation source with the energy of 1253.6 eV, 16 mA × 12.5 kV and a working pressure lower than 1 × 10-8 N m-2. As an internal reference for the absolute binding energies, the C1s peak of hydrocarbon contamination was used. The fitting of XPS curves was analyzed with a software (Multipak 6.0A). Photoreactor and Light Source. A cylindrical Pyrex photoreactor was used in the experiments as shown in Figure 1, in which a 110 W high-pressure sodium lamp (Institute of Electrical Light Source, Beijing) was positioned inside the 2382
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
FIGURE 1. A sketch of photoreactor.
FIGURE 2. The molecular structure of methylene blue. cylindrical Pyrex vessel surrounded by a circulating water jacket (Pyrex) to cool the reaction solution. The lamp mainly provides visible light in the range of 400-800 nm. Experimental Procedures. The reaction suspensions were prepared by adding 0.2 g of photocatalyst powder into a 165 mL of aqueous methylene blue (MB) solution (12 mg L-1). The MB chemical was provided by BDH and used without further purification. The molecular structure of MB is shown in Figure 2. Prior to photooxidation, the suspension was magnetically stirred in a dark condition for 15 min to establish an adsorption/desorption equilibrium condition. The aqueous suspension containing MB and photocatalyst were irradiated under visible light with constant aeration. At the given time intervals, the analytical samples were taken from the suspension and immediately centrifuged at 70 rps for 20 min, then filtered through a 0.45 µm Millipore filter to remove the particles. The filtrate was analyzed as required. Analytical Methods. MB concentration was analyzed by the Spectronic Genesys-2 UV-visible spectroscopy at 664 nm. TOC concentration was determined by the Shimadzu TOC-analyzer 5000A equipped with an autosampler ASI-5000. The concentrations of photoreaction products including ammonium ion, nitrate ion, and sulfate ion were determined by ion chromatography with a conductivity detector (Shimadzu HIC-6A). A Shim-Pack IC-A1 anion column and mobile phase containing 2.5 mM phthalic acid and 2.4 mM tris(hydroxymethyl)aminomethane with a flow rate of 1.5 mL min-1 was used for determination of nitrite, nitrate, and sulfate, while a Shim-Pack IC-C1 cationic column and mobile phase containing 5.0 mM nitrate acid with a flow rate of 1.0 mL/min was used for determination of ammonium ion.
FIGURE 4. The iso-electric point of the photocatalysts.
D)
Kλ B cos θ
FIGURE 3. The XRD photograph of the photocatalysts.
TABLE 1. The Crystal Size (D), Particle Diameter (Dp), Iso-Electric Point, Weight Percentage of the Anatase Phase (WA), and the Distance between the Crystal Planes of A101 Peak of TiO2 Doped by Gold catalysts
D (nm)
iso-electric point
WA (%)
d(hkl)a
TiO2 0.5% Au3+(H2)-TiO2 0.5% Au3+-TiO2 0.5% Au-TiO2
18.3 14.2 11.8 19.4
6.4 6.02 5.15 5.98
85.3 86.5 86.5 84.7
3.507 3.517 3.515 3.504
a
hkl is the crystal plane index of A101 peak.
Results and Discussion Crystal Phase Composition of Photocatalyst. Four resulting powder samples, 0.5% Au-TiO2, 0.5% Au3+(H2)-TiO2, 0.5% Au3+-TiO2, and pure TiO2, after calcinations at 923 K in which the samples had undergone a phase transformation to form a mixture of anatase and rutile phases, were examined with the XRD method. The main components of anatase (A), rutile (R), and gold (Au) were identified in the XRD analysis and are labeled as shown in Figure 3. To estimate the fraction of anatase and rutile in the resulting Au-TiO2 powder, the weight percentage of the anatase phase, WA, was determined using the following equation (17):
WA )
1 1 + 1.265
IR IA
where IA denotes the intensity of the strongest anatase reflection, and IR is the intensity of the strongest rutile reflection. For a given sample, the ratio IA/IR is independent of fluctuations in diffractometer characteristics. In this investigation, the intensity of anatase peak at 2θ ) 25.4 ( 0.1° was considered as IA (A101), and the intensity of rutile peak at 2θ ) 62.7 ( 0.1° was considered as IR (R002). The calculated results of WA for all samples are shown in Table 1, which indicate that the 0.5% Au-TiO2 has a WA value (84.7%) slightly lower than that (85.3%) of the pure TiO2. While the 0.5% Au3+-TiO2 and 0.5% Au3+(H2)-TiO2 have the WA values of 86.5 and 86.5% slightly higher than that of the pure TiO2. Bokhimi (22) reported similar results relating to phase transformation of TiO2. It might be concluded that the content of rutile slightly increased owing to gold deposition, and slightly decreased owing to gold ion doping. Size and Surface Structure of Photocatalyst. On the basis of the XRD results, the crystal sizes (D) of Au-TiO2 powder can also be estimated using the Scherrer equation (23):
where B is the peak width, K ) 0.89 is a coefficient, θ is the diffraction angle, and λ is the X-ray wavelength corresponding to Cu KR radiation. The crystal sizes (D) of all samples as shown in Table 1 indicate that the crystal size of 0.5% AuTiO2 slightly increased from 18.3 to 19.4 nm, and the crystal sizes of 0.5% Au3+-TiO2 and 0.5% Au3+(H2)-TiO2 decreased to 11.8 and 14.2 nm, respectively. On the other hand, the distance (dhkl) between crystal planes of A101 peak in the anatase of 0.5% Au3+-TiO2 and 0.5% Au3+(H2)-TiO2 slightly increased. This means that gold ion might be modified into the lattice of TiO2. However, the TiO2 powder with gold deposition had only a slight change in its crystal phase, quantum size, and crystal parameter, which should not affect their photoactivity very much. Iso-Electric Point and Substrate Adsorption. To understand the adsorption behavior of the modified TiO2 catalyst in its aqueous solution, the zeta (σ) potentials of TiO2, 0.5% Au3+(H2)-TiO2, 0.5% Au3+-TiO2, and 0.5% Au-TiO2 samples were measured in the pH 2.4-8.4, and the iso-electric points of samples were determined as pH 6.40, pH 6.02, pH 5.15, and pH 5.98, respectively, as shown in Figure 4. The results demonstrated that the iso-electric points of modified TiO2 powder have been reduced from pH 6.4 to below pH 6. A lower iso-electric point means a higher concentration of hydroxide ions on the surface of the photocatalyst (24). More hydroxide ions may be expected to yield a higher photocatalytic activity. On the other hand, it is believed that the iso-electric point would greatly influence the adsorption of organic substrates and its intermediates on the surface of photocatalyst during photoreaction. In fact, the photocatalytic process mainly occurs on the photocatalyst surface but not in bulk solution. The adsorption of substrates plays a very important role in photocatalytic degradation. As a cationic dye, MB was more effectively absorbed on the Au3+(H2)-TiO2, Au3+-TiO2, and Au-TiO2 surface than on the pure TiO2. This is one of reasons why the photoactivity of goldmodified TiO2 was higher than that of pure TiO2. However, some intermediates with negative charge would be produced in the process of MB photodegradation. This will remain to be further investigated. Optical Absorption Properties of Photocatalyst. The UVVisible-diffuse reflectance spectra of samples were measured in the wavelength range of 260-460 nm and are shown in Figure 5. It was found that there was a marked decrease of reflectance in the region 380-460 nm when TiO2 was modified by either gold ion or gold. This implied that the absorption spectrum of gold/gold ion-doped TiO2 was shifted toward the visible region and might be meaningful for the enhancement of photocatalytic behavior under solar energy application. According to Mie theory (25), semiconductors absorb light below a threshold wavelength λg (the fundamental absorption VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2383
FIGURE 5. The DRS of the photocatalysts.
FIGURE 7. The proposed AuxTi1-xO2 energy level and photoinduced electron excitation.
TABLE 2. The Position of PL Emission Peaks
FIGURE 6. (a) The UV-visible absorption spectra of the TiO2 modified with gold; (b) the UV-visible absorption spectra of the TiO2 modified with gold ion. edge). The band-gap energy of anatase (TiO2) is about 3.2 eV with the threshold wavelength λg ) 387.5 nm. In this study, the UV-visible absorption spectra of all samples were measured and are shown in Figure 6a,b. The results indicate that all TiO2 modified with Au/Au3+ had better optical absorption in general from 300 to 700 nm than the pure TiO2. A special interest can be addressed on the longer wavelength of 480-600 nm in which all Au-TiO2 samples demonstrated a significant enhancement of light absorption at a wavelength of 530-540 nm due to plasmon resonance. It was also found that a wide absorption peak in the spectrum of 0.5% Au3+(H2)-TiO2 appeared at the wavelength of around 535 nm, which was caused by the fraction of gold dopant and had a good agreement with that in Yonezawa’s study in 1999 (26). To interpret the change of light absorption band from the near UV to the visible light range found in this study, we assume that a complex of AuxTi1-xO2 would be formed in the TiO2 doped with gold ion, which should have a lower energy level than that of TiO2 as illustrated in Figure 7. When hν g (Ec - Ev), electrons can be excited in the valence band of TiO2; when (Ec - Ev) > hν g (Ec - EAu ion), electrons can only be excited from the AuxTi1-xO2 energy level. However, this 2384
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
0.5% Au3+(H2)- 0.5% Au3+TiO2 TiO2
0.5% AuTiO2
photocatalysts
TiO2
peak 1 position (eV/nm) peak 2 position (eV/nm) peak 3 position (eV/nm) peak 4 position (eV/nm) peak 5 position (eV/nm)
3.32/373.5
3.31/374.5
3.32/374
3.27/379.5
3.26/380.5
3.27/379.5 3.26/380.5
3.21/385.5
3.22/385
3.21/386
3.11/398.5
3.10/399.5
3.11/398.5 3.11/398.5
2.36/526
2.29/542
2.34/530
3.32/374
3.22/385
2.34/530
assumption needs to be confirmed by defining the position of gold impurity (AuxTi1-xO2) energy level. The Valence State of Gold. To study the valence state of gold, the modified TiO2 samples were examined by the XPS method. On the basis of the XPS data, the Au 4f peak of gold or gold ion-doped samples was fitted using Multipak 6.0A software, and the results are listed in Table 3. The Au 4f peak of 0.5% Au3+-TiO2 and 0.5% Au-TiO2 consisted of three peaks, corresponding Au (0), Au (I), Au (III), respectively. The results of XPS showed the valence state of Au had a slight difference among them and there are only Au (0) and Au (I), but not Au (III) in the 0.5% Au3+(H2)-TiO2, lattice. Since gold ion was doped in TiO2 by the sol-gel process, a fraction of the gold impurity in 0.5% Au3+-TiO2 and 0.5% Au3+(H2)-TiO2 was implanted into the TiO2 lattice, and the gold impurity in 0.5% Au-TiO2 was only deposited on the surface of TiO2 by the photoreduction process. Au (0), Au (I), and Au (III) play important roles in affecting charge trapping, release and migration, recombination, and interfacial charge transfer, and then the photocatalytic behavior. PL Emission and Electron/Hole Recombination. The photoluminescence (PL) emission spectra are useful to disclose the efficiency of charge carrier trapping, immigration and transfer and to understand the fate of electron-hole pairs in semiconductor particles because PL emission results from the recombination of free carriers. In this study, the PL emission spectra of all samples were examined in the range of 3-3.8 and 1.8-3 eV, and the analytical results are shown in Figures 8 and 9, respectively. To illustrate the results in these figures, the positions of individual PL emission peaks are listed in Table 2. The PL emission spectra of pure TiO2 sample in Figure 8 showed that several peaks appeared at 3.32, 3.27, 3.21, and 3.12 eV, equivalent to the wavelength of 373.5, 379.5, 385.5, and 397 nm, respectively. The PL emission spectra of Au/Au3+-doped samples and pure TiO2 sample in Figure 8 showed the similar positions of most peaks, but different PL intensities. The PL intensity of pure TiO2 sample was significantly higher than that of Au/Au3+-modified TiO2 samples. If we agree that the PL emission mainly results from the recombination of excited electron and holes, a lower PL intensity may indicate a lower recombination rate of electron/ holes under light irradiation.
TABLE 3. Au 4f XPS Data and the Valence State Obtained by Fitting Curves Au (III) catalysts 0.5% Au-TiO2 0.5% Au3+-TiO2 0.5% Au3+(H2)-TiO2 a
Au (I)
Au (0)
binding energy (eV)
fwhma (eV)
area (%)
binding energy (eV)
fwhm (eV)
area (%)
binding energy (eV)
fwhm (eV)
area (%)
87.86 87.92
1.37 1.91
37.91 40.52
86.12 86.72 85.99
1.91 1.91 1.91
17.25 22.72 68.80
84.03 83.93 83.83
1.26 1.34 1.91
44.84 36.76 31.20
fwhm: Full width at a half of the maximum height of peaks.
TABLE 4. The Pseudo-First-Order Kinetic Constant of MB Photodegradation Using Different Photocatalysts
FIGURE 8. The PL emission spectra of photocatalysts in the range of 350-420 nm at 10 K.
photocatalysts
kinetics constants, k, min-1
correlation coefficient, R2
pure TiO2 0.25% Au-TiO2 0.5% Au-TiO2 1.0% Au-TiO2 2.0% Au-TiO2 4.0% Au-TiO2 5.0% Au-TiO2 0.25% Au3+-TiO2 0.5% Au3+-TiO2 1.0% Au3+-TiO2 3.0% Au3+-TiO2 5.0% Au3+-TiO2 0.5% Au3+(H2)-TiO2
0.0144 0.0258 0.052 0.0407 0.0321 0.0224 0.0093 0.0458 0.0927 0.0591 0.369 0.0286 0.0567
0.9964 0.9983 0.999 0.997 0.9952 0.998 0.9988 0.9964 0.9865 0.9949 0.9539 0.9985 0.9978
FIGURE 9. The PL emission spectra of photocatalysts in the range of 420-700 nm at 10 K. In the study, all samples exhibited a mixture of anatase and rutile phases. We may consider the main peaks at 3.32 (or 3.31 eV) and 3.27 (or 3.26 eV) as the band gaps of anatase phase and the peaks at 3.22 (or 3.21 eV) and 3.12 (or 3.11 eV) as the band gaps of rutile phase, since these band gaps are very similar to the reports by Suisalu (16) Tang (27, 28) and Amtour (29). Some other small peaks also appeared in the PL emission spectra, which were believed to be the result of quantum size effects. Comparing our experimental results with other researchers’ work, it may be concluded that the gold impurity doped/deposited in TiO2 does not change the band edge of anatase and rutile very much, but can reduce the PL emission significantly. The PL emission spectra of all samples in the range of 1.8-3 eV (420-700 nm) are shown in Figure 9. It is observed that an intensive yellowish green PL spectrum (broad band) of the pure TiO2 is much higher than any other spectra of gold/gold ion-doped TiO2 samples. These results indicate that the position of PL peaks apparently shifted to red direction owing to the presence of gold impurity.
FIGURE 10. (a) The kinetics of MB photodegradation using Au-TiO2 with initial MB concentration of 12 mg l-1 and pH 5.98; (b) The kinetics of MB photodegradation using Au3+-TiO2 with initial MB concentration of 12 mg L-1 and pH 5.98. In this photoreaction, the Au (I) and Au (III) doped inside TiO2 act as electron acceptors and Au deposited on TiO2 surface produces the highest Schottky barrier among the metals facilitating the electron capture. It indicates that either gold or gold ion dopant in TiO2 hindered the recombination of electron/hole pairs that resulted in a declined intensity of PL emission. The Role of Gold Impurity. The enhancement of photoactivity depends on the absorption of substrate, optical absorption, and the efficiency of charge carrier. In this study, the noble metal gold was used to produce the highest Schottky VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2385
FIGURE 11. The MB reduction in dark condition as compared to MB photooxidation. barrier among the metals facilitating the electron capture (4, 5). The presence of Au (0) on TiO2 surface favors the migration of photoproduced electron to gold, thus improving the electron/hole separation. Subsequently, electrons migrate from gold to O2 molecules (3). Moreover, the density of free electrons in the TiO2 with gold deposit is less than that in the TiO2 without gold deposit and will promote the photoadsorption of O2 molecules on the surface of photocatalysts. Therefore, the rate of electron transfer to oxygen increases and the recombination rate of electron/hole pairs decreases. These improvements would be beneficial to the photooxidation of organic compounds using 0.5% Au-TiO2. On the other hand, we propose that the gold ion (Au+) doped in the lattice of TiO2 acts as electron trap and the gold ion (Au3+) acts as electron and hole traps as described in eqs 1-3, respectively. The trapped electron is transferred to oxygen by eq 4. Therefore, the gold ion in the lattice of TiO2 promotes the charge trapping that favors the charge migration
FIGURE 12. The TOC removal and MB decolorization during MB photodegradation with initial MB concentration of 12 mg L-1 and pH 5.98. to O2 and increases the photoactivity.
Au3+ +2 ecb- f Au+
(1)
Au3+ + hvb+ f Au4+
(2)
Au+ + ecb- f Au
(3)
O2(ads) + ecb- (or Au+, Au) f O2(ads)-
(4)
The Photodegradation of MB. To evaluate the photoactivity of the Au/Au3+-TiO2 photocatalysts with an optimum dosage of gold or gold ion doped in TiO2, the photodegra-
FIGURE 13. The products of NO3-, NO2-, NH4+, SO42-, and T-N during MB photodegradation. 2386
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 11, 2001
dation of MB dye with the initial concentration of 12 mg L-1 was carried out. The experimental data were converted to a linear pattern using pseudo-first-order kinetics model, and the results are shown in Figure 10. The observed rate constants were also calculated as listed in Table 4. The experimental results demonstrated that the TiO2 doped with gold impurity had a significant increase of photodegradation rate as compared to the pure TiO2. It was found that the molar content of 0.5% impurity of either gold (Au) or gold ion (Au3+) doped in TiO2 was the best dosage to achieve the highest photodegradation efficiency under this experimental condition. Any higher dosage of gold doped in TiO2 would be detrimental to the efficiency of MB photodegradation. It may be explained that when the content of gold/gold ion was less than its optimal ratio, the gold impurity energy level would be a separation center. On the contrary, when the content of gold/gold ion was more than its optimal ratio, the gold impurity energy level would be a recombination center. Although the absorption of visible light was promoted with the increase of gold/gold ion content, the photoactivity may not increase owing to the enhancement of electron/hole recombination rate as well. Of all photocatalysts tested, 0.5% Au3+-TiO2 demonstrated to be the most photoactive to MB decolorization and photodegradation. The observed rate constants of the pure TiO2 was only one-sixth of that of 0.5% Au3+-TiO2, one-fourth of that of 0.5% Au-TiO2 and 0.5% Au3+(H2)-TiO2. The results also indicated that hydrogen gas treatment did not help the enhancement of its photooxidation rate. Three control tests without light irradiation were also carried using pure TiO2, 5% Au3+-TiO2, and 5% Au-TiO2, respectively, to determine the reduction of MB in a dark condition, and the results with comparison to photooxidation are shown in Figure 11. It was found that the decrease of MB concentration in the dark condition was less than 5% in the pure TiO2 suspension and less than 10% in the Au3+/Au-TiO2 suspensions. This insignificant loss of MB during the reaction could result by both adsorption and dark reaction. The fraction of MB loss due to dark reaction was not determined in this study, although the gold-doped TiO2 could play a role of catalyst to conduct a dark reaction. It is believed that the MB ion with positive charge in aqueous suspension might be more easily adsorbed by gold-doped TiO2 than by pure TiO2 catalyst, since the iso-electric point of gold-doped TiO2 is lower than that of pure TiO2. MB Mineralization and Oxidation Products. The experiment of MB photodegradation was carried out for 120 min and six samples were collected at time intervals of 0, 15, 30, 60, 90, and 120 min. The concentrations of total organic carbon (TOC) and the percent of MB color remaining were analyzed, and the results are shown in Figure 12. It was found that the TOC removals of 43.7, 63.3, 73.6, and 67.9% for TiO2, 0.5% AuTiO2, 0.5% Au3+-TiO2, and 0.5% Au3+(H2)-TiO2 respectively were achieved after a 120 min reaction, while decolorizaton of MB conducted much faster and was completed after a 30-60 min reaction. To study the reaction products of MB photodegradation, five samples collected at reaction times of 0, 30, 60, 90, 120 min were analyzed for determinations of ammonium ion (NH4+), nitrite ion (NO2-), nitrate ion (NO3-), and sulfate ion (SO42-). The analytical results are shown in Figure 13. The experiment demonstrated that the MB was partially mineralized into several products including ammonium ion (NH4+), nitrite ion (NO2-), nitrate ion (NO3-), and sulfate ion (SO42-). Of which, ammonium ion might be an intermediate product and finally oxidized to nitrate ion (30). However, it was a heterogeneous reaction and the adsorption of those inorganic ions on the catalyst surface would be an important factor to affect the dissolved amount. In general, we believe that ammonium ion with positive charge would be less easily adsorbed by the catalyst than other ions with negative charge,
since TiO2 power is positively charged at pH < 6.4. The experimental results indicated that the degree of MB mineralization using the TiO2 powder doped with Au/Au3+ was significantly higher than that of using the pure TiO2 powder, since the concentrations of sulfate and inorganic nitrogen in the MB solution after reaction using the 0.5% Au-TiO2, 0.5% Au3+-TiO2, and 0.5% Au3+(H2)-TiO2 was much higher than that of using TiO2. This also confirmed the improvement of photoactivity of gold or gold ion-doped TiO2.
Acknowledgments The authors thank the Hong Kong Government Research Grant Committee for financial support to this work under the RGC Grant (RGC No: PolyU 5030/98E). The authors would also like to thank Mr. Simon Y. Li for his assistance of drawing figures in this paper.
Literature Cited (1) Hoffmann, M. R.; Martin, S. T.; Choi, W. Y.; Bahnemann, D. W. Chem. Rev. 1995, 95, 69. (2) Goswami, D. Y. J. Sol. Energ. 1997, 119, 101. (3) Linsebigler, A.; Lu, G.; Yates, J. T. Chem. Rev. 1995, 95, 735. (4) Wang, C. Y.; Liu, C. Y.; Chen, J.; Shen, T. J. Colloid Interface Sci. 1997, 191, 464. (5) Wang, C. Y.; Liu, C. Y.; Zheng, X.; Chen, J.; Shen T. Colloid. Surf. A 1998, 131, 271. (6) Sclafani, A.; Palmisana, L.; Marcı´, G.; Venezia, A. Sol. Energy Mater. Sol. C 1998, 51, 203. (7) Yang, J. C.; Kim, Y. C.; Shul, Y. G.; Shin, C. H.; Lee, T. K. Appl. Surf. Sci. 1997, 121/122, 525. (8) Choi, W.; Termin, A.; Hoffmann, M. R. J. Phys. Chem. 1994, 98, 13669. (9) Chen, C. P.; Qi, X. Y.; Zhou, B. M. J. Photochem. Photobiol. A 1997, 109, 155. (10) Deng, H. H.; Lu, Z. H.; Chen, Y. C.; Mao, H. F.; Xu, H. J. Phys Chem. 1998, 231, 95. (11) Zhang, F. L.; Zhao, J. C.; Shen, T.; Hidaka, H.; Pelizzetti, E.; Serpone, N. Appl. Catal. B 1998, 15, 147. (12) Zhao, J. C.; Wu, K. Q.; Wu, T. X.; Hidaka, H.; Serpone, N. J. Chem. Soc. Faraday Trans. 1998, 94, 673. (13) Wu, T. X.; Lin, T.; Zhao, J. C.; Hidaka, H.; Serpone, N. J. Environ. Sci. Technol. 1999, 33, 1379. (14) Liu, G. M.; Wu, T. X.; Lin, T.; Zhao, J. C. Environ. Sci. Technol. 1999, 33, 2081. (15) Liu, G. M.; Li, X. Z.; Zhao, J. C.; Horikoshi S.; Hidaka H. J. Mol. Catal. A 2000, 153, 221. (16) Aguado, M. A.; Anderson, M. A.; Hill, C. G. J. Mol. Catal. 1994, 89 (1-2), 165-178. (17) Mardare, D.; Tasca, M.; Delibas, M.; Rusu, G. I. Appl. Surf. Sci. 2000, 156, 200. (18) Rahman, M. M.; Krishna, K. M.; Soga, T.; Jimbo, T.; Umeno, M. J. Phys. Chem. Solids 1999, 60, 201. (19) Krishna, K. M.; Mosaddeq-ur-Rahman, M.; Miki, T.; Krishna, K. M.; Soga, T.; Igarashi, K.; Tanemura, S.; Umeno, M. Appl. Surf. Sci. 1997, 113/114, 149. (20) Mizushima, K.; Tanaka, M.; Asai, A.; Iida, S. J. Phys. Chem. Solids 1979, 40, 1129. (21) Mo, S. D.; Lin, L. B.; Lin, D. L. J. Phys. Chem. Solids 1994, 55, 1309. (22) Bokhimi, X.; Morales, A.; Novaro, O. J. Mater. Res. 1995, 10, 2788. (23) Liu, J.; Yu, J. C.; Lo, D.; Lam, S. K. J. Catal. 1999, 183, 368. (24) Yu, J. C.; Lin, J.; Kwok R, W. M. J. Photochem. Photobiol. A 1997, 111, 199. (25) Serpone, N.; Pelizzetti, E. In Photocatalysis: Fundamentals and Application, Edit I; Wiley-Interscience: New York, 1989; p 123. (26) Yonezawa, T.; Matsune, H.; Kunitake, T. Chem. Mater. 1999, 11, 33. (27) Tang, H.; Berger, H.; Schmid, P. Solid State Commun. 1993, 87, 847. (28) Tang, H.; Berger, H.; Schmid, P. Solid State Commun. 1994, 92, 267. (29) Amtout, A.; Leonelli, R. Phys. Rev. 1995, B 51, 6848. (30) Low, G. K. C.; McEvoy, S. R.; Matthews, R. W. Environ Sci Technol. 1991, 25 (3), 460.
Received for review October 11, 2000. Revised manuscript received February 13, 2001. Accepted March 2, 2001. ES001752W VOL. 35, NO. 11, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2387
