Understanding Reaction Kinetics by Tailoring Metal Co-catalysts of the BiVO4 Photocatalyst.

In this study, the reaction mechanisms of metal-semiconductor composites used as photocatalysts were demonstrated by first preparing bismuth vanadate (BiVO4) and then performing photodeposition of metal nanoparticles. The photocatalytic activity of metal-BiVO4 (M-BiVO4, where M = Pt, Au, Ag) composites were evaluated through dye decomposition under UV-vis irradiation. The photocatalytic efficiency was significantly enhanced after Pt deposition as compared to other M-BiVO4 composites. The size or shape of BiVO4 was not the main factor for the efficiency of Pt-BiVO4. However, a deposited Pt co-catalyst was essential for the photocatalytic decomposition of dye on the BiVO4 surface. Radical scavengers were employed to elucidate the reaction mechanism during the photocatalytic reaction with the Pt-BiVO4 composite. This study provides details on the reaction mechanism of the photocatalytic reaction on Pt at the BiVO4 surface under solar irradiation.

Introduction

Efficient purification of industrial wastewaters is highly desirable to address the diminishing availability of water resources.[1−4] Recently, semiconductor photocatalytic processes have shown great potential as an environmentally benign and sustainable treatment technology in the wastewater industry.[5−7] Although a multitude of metal oxide photocatalysts have been investigated for potential use in photocatalytic degradation of organic pollutants, most photocatalysts developed thus far are wide band gap semiconductors. These include TiO2, WO3, SrTiO3, and ZnO, which are active only under UV light.[8−11] In this context, developing a photocatalyst with a high photon-to-charge carrier efficiency under visible light is required.[12]

Bismuth vanadate (BiVO4) has tremendous potential for use as a heterogeneous photocatalyst that can be easily activated under the visible-light wavelength range.[13] However, the slow carrier mobility in the semiconductor contributes to the poor catalytic efficiency of BiVO4.[14] Previous studies have mainly focused on the size control to improve catalytic properties when attempting to enhance the activity of BiVO4. However, fast recombination at the semiconductor–solution interface remains a performance bottleneck. Although efficient electrocatalysts have been intensively studied for application in photocatalytic reactions,[15,16] there is no guarantee that the best electrocatalysts will perform equally well when they have been integrated into a photocatalyst. Although noble metals have been intensively studied as efficient electrocatalysts,[17−19] the origin of enhanced catalytic activity remains unclear. Further studies must be conducted to elucidate the detailed mechanism.

Herein, we report the synthesis of metal–BiVO4 (M–BiVO4, where M = Pt, Au, Ag) composites to improve the photocatalytic efficiency toward organic dye decomposition under solar light irradiation. Our results showed that the Pt–BiVO4 catalyst considerably improved the decomposition of methyl orange (MO) and mixed dye as compared to other BiVO4 catalysts. The decomposition rates were studied using radical scavengers to clarify the degradation pathways.

Results and Discussion

Preparation of BiVO4 and M–BiVO4 Catalysts

A mixture of 20 mM Bi(NO3)3·5H2O, 20 mM NH4VO3, and 1.0 g poly(vinylpyrrolidone) (PVP) precursor solution was prepared and then transferred into a Teflon-lined stainless steel autoclave. The sealed autoclave was heated in an electric oven at 180 °C for 6 h in air. The as-prepared BiVO4 powder was annealed at 500 °C for 3 h in air to yield crystallized BiVO4.[20] An X-ray diffraction (XRD) analysis was performed to produce reliable structural information about this material (Figure a). The diffraction peaks of the BiVO4 crystals were indexed to the monoclinic scheelite structure (JCPDS No. 83-1699), indicating a well-crystallized BiVO4.[13]

Figure 1

(a) XRD patterns; (b) Raman spectrum of monoclinic phase BiVO4 crystals; and X-ray photoelectron spectroscopy (XPS) spectra of (c) Bi 4f and (d) V 2p for BiVO4 crystals.

Raman spectroscopy is a nondestructive experimental technique used to analyze the vibrational and structural properties of materials. Figure b shows the Raman spectrum of BiVO4 in the range of 100 to 1000 cm–1. The Raman peaks of BiVO4 at 127, 213, 326, 368, and 827 cm–1 were in agreement with those of the monoclinic BiVO4.[21] Peaks of tetragonal phase BiVO4 were not observed in the spectrum, indicating that the monoclinic BiVO4 grew uniformly on the entire particle. The surface composition and chemical states of BiVO4 crystals were further characterized by X-ray photoelectron spectroscopy (XPS) (Figure c,d). The two dominant peaks at 155.2 and 160.5 eV were assigned to Bi 4f, which is the characteristic of Bi3+ species (Figure c).[18] The observed V 2p peaks at 512.3 and 520.1 eV were attributed to the oxide form of V5+ in BiVO4 (Figure d).[22] The band gap was estimated from the onset of UV–visible absorbance (Figure S1). From the absorbance data, the monoclinic BiVO4 crystals showed a direct transition with a band gap of approximately 2.3 eV, which is also consistent with the value in the literature.[13]

Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) analyses were performed to identify the structural composition of the BiVO4 crystals (Figure ). The SEM image shows that the BiVO4 crystals had an average size of 716 ± 87 nm (Figure a). A high-resolution TEM (HRTEM) image of the edge of a BiVO4 shows a lattice spacing of 0.309 nm, corresponding to the (112) planes (Figure b). The energy-dispersive X-ray spectrometry of a BiVO4 crystal yielded an average atomic ratio of 30:28 (Bi/V), indicative of the 1:1 atomic composition. A spatial elemental mapping was also performed on a BiVO4 crystal to evaluate the distribution of each element (Figure c–f). The elemental mapping confirmed that Bi, V, and O atoms were uniformly distributed throughout the particle. Therefore, the XRD, Raman, XPS, and TEM results verified the preparation of the monoclinic BiVO4 crystals.

Figure 2

(a) SEM and (b) HRTEM images of a BiVO4 crystal; (c) spatial elemental map of BiVO4 crystals; and the corresponding elemental distribution maps of (d) Bi (green), (e) V (blue), and (f) O (yellow).

Light irradiation applied to semiconductor particles in an aqueous solution of metal salts results in the deposition of metal nanoparticles on the surface of the semiconductor, characterizing a behavior known as photodeposition.[10] Various M–BiVO4 composites were prepared by photodeposition using three types of noble-metal precursors (H2PtCl6, HAuCl4, and AgNO3), and methanol was used as a hole scavenger. The three composites were denoted as Pt–BiVO4, Au–BiVO4, and Ag–BiVO4 (Figure a). Figures and S2 show the TEM images and XRD patterns of the M–BiVO4 crystals. The HRTEM images of the edge of Pt–BiVO4, Au–BiVO4, and Ag–BiVO4 show the typical lattice spacing of each metal (Figure S2), confirming the metal deposition on BiVO4. The HRTEM image of Pt–BiVO4 shows a lattice spacing of 0.228 nm, which corresponds to the (111) planes of Pt metal (Figure S2d). The XRD peaks of all samples were identical to those of the monoclinic scheelite structure of BiVO4 (Figure S2). The XRD patterns of deposited metals could not be observed because of the small grain size of metal nanoparticles. The Raman peaks were also in agreement with those of the monoclinic BiVO4 (Figure b), and the metal deposition did not influence the crystal structures of BiVO4. The XPS at the Pt 4f level showed splitting peaks at 71.4 and 74.4 eV, which corresponded to the metallic state of Pt species (Figure S3).[23]

Figure 3

(a) Schematic of the photodeposition of metals on BiVO4 crystals; (b) Raman spectra of BiVO4 and Pt–BiVO4 crystals; (c) spatial elemental map of Pt–BiVO4 crystals; and (d) corresponding elemental distribution maps of Pt (red).

Spatial elemental mapping was also performed for M–BiVO4 crystals to evaluate the distribution of each element (Figures c,d and S4). The composition and elemental distribution of the Pt–BiVO4 sample is shown in Figure c,d. The elemental mapping shows that Pt was homogeneously dispersed on the BiVO4 surface with 1 wt % Pt loading. Au and Ag were also homogeneously deposited on BiVO4 (Figure S4).

Photocatalytic Activity of BiVO4 and M–BiVO4 Catalysts

The photocatalytic performance of BiVO4 and M–BiVO4 samples were evaluated through the degradation of methyl orange (MO) under UV–visible light (one-sun, 100 mW cm–2). Figures and S5 show the absorption spectra of the MO solution as a function of illumination time in the presence of BiVO4 and M–BiVO4 catalysts. The blank test (MO without catalyst) under UV–visible light showed negligible self-photolysis of MO. For pure BiVO4, the absorption peak at 465 nm gradually decreased, indicating a partial degradation of the dye (approximately 20% after 240 min) under UV–visible light (Figure a, green squares). Pt–BiVO4, Au–BiVO4, and Ag–BiVO4 were also tested for photocatalytic MO degradation under the same conditions. After irradiation for 20 min, the degradation percentages were 2, 13, 22, and 99% for BiVO4, Au–BiVO4, Ag–BiVO4, and Pt–BiVO4, respectively (Figure a). The Pt–BiVO4 composite exhibited a much higher photocatalytic activity for MO degradation as compared to the other samples. Furthermore, the photocatalytic activity of the Pt–BiVO4 samples improved with increasing Pt content, and the 1.0 wt % of the Pt deposited sample (15 min photodeposition) exhibited the best activity (Figure S6). No obvious enhancement of the photocatalytic activity was observed when more Pt particles were introduced.

Figure 4

(a) Photocatalytic degradation rates of MO when using BiVO4 (green squares), Au–BiVO4 (blue triangles), Ag–BiVO4 (pink triangles), and Pt–BiVO4 (red circles) crystals and without BiVO4 (black squares); and (b) Stability experiments of Pt–BiVO4 crystals.

The recyclability of Pt–BiVO4 was evaluated through cycling experiments, which resulted in practically unchanged photocatalytic activity (Figure b). The stability of the Pt–BiVO4 sample was further investigated by XRD and XPS analyses following the stability experiments (Figure S7). The XRD and XPS results showed nearly identical peaks before and after the stability test (Figure S7), indicating that Pt–BiVO4 retained its chemical structure. In other words, the Pt–BiVO4 composite was photocatalytically stable.

To elucidate photocatalytic activities, small BiVO4 crystals were prepared using ultrasonic methods (BiVO4 sono-crystals).[20] The XRD peaks of the BiVO4 sono-crystals were indexed to the monoclinic scheelite structure (JCPDS No. 83-1699) (Figure S8a), indicating a well-crystallized BiVO4. The SEM image shows that the BiVO4 sono-crystals had an average size of 495 ± 54 nm (Figure S8b). The BiVO4 sono-crystals showed a higher photocatalytic activity for MO degradation than did the previous BiVO4 crystals. However, their photocatalytic efficiency was nearly identical to that of BiVO4 sono-crystals after Pt deposition (Figure S9). The catalytic activity of BiVO4 was not the main factor for efficiency. However, the deposited Pt co-catalyst was essential for the photocatalytic decomposition of dye on the BiVO4 surface.

The photocatalytic performance of the Pt–BiVO4 composite was also evaluated by the degradation of cationic dye and anionic dye mixtures (rhodamine B (RhB), and MO) under the same conditions. The dye mixture was used because real-world wastewater from industries commonly contain more than one type of organic dye.[19] The results clearly confirm that both RhB and MO dyes were rapidly degraded over the Pt–BiVO4 composite (Figure ).

Figure 5

Photocatalytic degradation rates of a mixed solution of MO and RhB using Pt–BiVO4 crystals and time-dependent UV–vis absorption spectra (inset).

Understanding the mechanism of photocatalytic reactions is critical to improve the catalytic efficiency for practical applications. The deposition of noble metals on the BiVO4 surface has been reported to improve the photocatalytic efficiency by acting as an electron trap due to the formation of the Schottky barrier, thus reducing the electron–hole recombination process.[8,24] The highly improved photocatalytic activity of Pt–BiVO4 is also attributed to the Pt co-catalyst, which enhances the production rate of oxidizing species. To evaluate the mechanism involved in the high photocatalytic activity of the Pt–BiVO4 catalyst, experiments on trapping reactive oxygen species (hydroxyl radicals: •OH; and superoxide radical anions: •O2–) were performed and analyzed (Figure a). The •OH radical formed on the semiconductor surface is an extremely powerful oxidizing agent. It attacks the adsorbed organic molecules nonselectively and participates in the further oxidation process. To investigate the role of •OH in the photocatalytic degradation of MO by Pt–BiVO4, a control experiment was conducted in the presence of methanol (Figures b and S10). Methanol has been used as a scavenger to determine the role of •OH in photocatalysis because it reacts rapidly with •OH (inhibiting pathway 1).[25] The addition of methanol reduces the degradation rate of the targeted contaminants by Pt–BiVO4, indicating that the •OH-mediated oxidation process is the predominant decomposition pathway. Although the photogenerated hole reacts with the surface-bound water or OH– to produce the hydroxyl radical (pathway 1), an electron in the conduction band is removed by the oxygen (O2) or MO (pathways 2 or 3 in Figure a), thus maintaining electron neutrality within BiVO4.[8] A photocatalytic experiment was also performed under inert conditions (argon) to identify the formation and role of •O2– during the photocatalytic degradation of MO by Pt–BiVO4 (inhibition of pathway 2).[26] The photocatalytic activity remained nearly unchanged (Figure b), thus suggesting the absence of inhibition under inert conditions (without O2) and the minor role of •O2– in the MO degradation. Therefore, pathway 1 is the predominant oxidative decomposition pathway, and pathway 3 prevents electron–hole recombination.

Figure 6

(a) Schematic representation of the dye degradation processes. (b) Photocatalytic degradation rates of MO with Pt–BiVO4 in the presence of methanol and O2 as scavengers under UV–vis light.

To further investigate the interaction of metal catalysts on the semiconductor, BiVO4 and Pt–BiVO4 electrodes were prepared on a fluorine-doped tin oxide (FTO) substrate (Figure S11). The photoelectrochemical (PEC) performance of BiVO4 and Pt–BiVO4 electrodes was studied by linear sweep voltammetry (LSV) for water oxidation (0.1 M phosphate buffered, pH 7). The LSV was conducted from +0.3 to +1.0 V vs Ag/AgCl at a scan rate of 20 mV s–1 with chopped light under UV–vis irradiation (light intensity: 100 mW cm–2). The Pt–BiVO4/FTO electrode showed an enhanced photocurrent as compared to BiVO4/FTO (Figure S11), and this result agreed well with the photocatalytic MO degradation outcome. This result indicates that the photogenerated hole transfer could be enhanced when the Pt catalyst is deposited on the BiVO4 surface.

Based on the PEC and photocatalytic results, the enhanced photocatalytic efficiencies were attributed to the rapid charge transfer from BiVO4 to the attached Pt nanoparticles, which enhanced the production rate of hydroxyl radicals. Although the Pt–BiVO4 composite exhibits distinct advantages because of the rapid charge transfer from BiVO4 to the attached Pt nanoparticles, the oxidation and reduction products remain unclear. Further studies must be conducted to elucidate the detailed organic products and reaction mechanism.

Conclusions

Metal–semiconductor composites were synthesized by preparing BiVO4 and then conducting photodeposition of metal nanoparticles in BiVO4. The photocatalytic activities of M–BiVO4 composites were evaluated based on the decomposition of MO and mixed dye (MO and RhB) under UV–vis irradiation. The photocatalytic efficiency was significantly enhanced when Pt deposition was used as compared to the other M–BiVO4 samples. The catalytic activity of BiVO4 semiconductors was not the main factor in the efficiency of Pt–BiVO4. However, the deposited Pt co-catalyst was essential for the photocatalytic dye decomposition on the Pt–BiVO4 surface. A variety of radical scavengers were employed to elucidate the reaction mechanisms of the Pt–BiVO4 composite during the photocatalytic reaction. The results indicate that the enhanced activity was mainly attributed to the •OH radicals that mediated oxidation during the dye degradation on Pt at the BiVO4 surface.

Methods

Materials

Bi(NO3)3·5H2O (99.999%, Sigma-Aldrich), NH4VO3 (99%, Daejung Chemicals) poly(vinylpyrrolidone) (K 30, Daejung Chemicals) were used as metal precursor salts and utilized as received. H2PtCl6·6H2O (≥37.50%, Sigma-Aldrich), HAuCl4·3H2O (≥99.9%, Sigma-Aldrich), AgNO3 (>99%, Sigma-Aldrich), methyl orange (SHOWA), RhB (Daejung Chemicals), FTO (TEC 15, WY-GMS) coated glass, and methyl alcohol (99.8%, Daejung Chemicals) were also used as received. Deionized (DI) water was used as a solvent in all experiments.

Preparation of BiVO4 Crystals

BiVO4 crystals were prepared using a hydrothermal synthetic method. A mixture of 20 mM Bi(NO3)3·5H2O, NH4VO3, and 1.0 g PVP precursor in solution (a mixture of 0.6 mL of 5 M NaOH and 35 mL of DI water) was prepared and then transferred to a Teflon-lined stainless steel autoclave. The sealed autoclave was heated in an electric oven at 180 °C for 6 h and then cooled to room temperature. The prepared powders were annealed at 500 °C for 3 h (with a 2 h ramp time) in air to form the BiVO4 crystals.

Photodeposition of Metals (Pt, Au, Ag) on BiVO4 Crystals

The as-prepared BiVO4 crystals were added to a 2 mM metal precursor solution (H2PtCl6, HAuCl4, AgNO3) with a methanol hole scavenger. The reaction mixture was stirred for 15 min under UV–vis illumination (light intensity: 100 mW cm–2) to form the M–BiVO4 crystals. Ethanol (10 mL) was added to the M–BiVO4 suspension, the suspension was centrifuged, and the supernatant was removed. The residue was washed three times with ethanol (10 mL) to provide yellow M–BiVO4 crystals.

Photocatalytic Activity

In all tests, the intensity of the lamp on the sample was measured as 100 mW cm–2 using a Si solar cell (AIST). Forty-eight milligrams of photocatalyst and 80 mL of an aqueous solution of MO or a mixture solution (3:1 ratio of MO and RhB) were placed in a quartz photoreactor. Prior to irradiation, the suspensions were magnetically stirred in the dark for 20 min to establish an absorption and desorption equilibrium between the dyes and the surface of the catalyst under normal atmospheric conditions. At given time intervals, aliquots of the mixed solution were collected and centrifuged to remove the catalyst for analysis. The residual concentration of dye was measured using a UV–vis spectrometer.

Material Characterization

The BiVO4 samples were characterized by SEM (Magellan 400, operated at 10 kV). TEM and HRTEM were performed using a Talos F200X instrument at 200 kV. The XRD pattern was measured using Cu Kα radiation at 40 kV and 300 mA (Rigaku, D/MAX-2500). The UV–vis–NIR absorption spectra were acquired by a UV-3600 UV–vis–NIR spectrophotometer using a solid sample holder for wavelengths from 300 to 1500 nm with the FTO substrate as the reference. No unexpected or unusually high safety hazards were encountered.