M. Qamar*a and
A. Khanb
aCenter of Excellence in Nanotechnology (CENT), King Fahd University of Petroleum and Minerals, KFUPM Box 498, Dhahran 31261, Saudi Arabia. E-mail: qamar@kfupm.edu.sa; Fax: +966 3860 7264; Tel: +966 3860 7775
bEnvironmental Geochemistry Laboratory, Department of Earth System Sciences, New Science Research Center, Yonsei University, Shinchon-dong Seodaemun-gu, Seoul 120749, Republic of Korea
First published on 11th December 2013
The synthesis of flower-like hierarchical bismuth tungstate (Bi2WO6) consisting of a mesoporous surface was carried out by a hydrothermal method using the non-ionic surfactant Pluronic F127. The mesoporous and hierarchical surface of the bismuth tungstate was further modified with platinum nanoparticles and the photocatalytic activity was evaluated by studying the removal of rhodamine B under visible light (>420 nm). The effect of synthesis temperature and platinum amount on the photocatalytic activity was investigated and resulting photocatalytic activity of Pt/Bi2WO6 was compared with other visible-light-responsive photocatalysts, namely Pt/WO3, N-doped TiO2 and Pt/N-doped TiO2. Photoelectrochemical studies were performed to shed light on the involvement of excited charged carriers in the photooxidation of rhodamine B and a plausible mechanism was proposed based on the photocatalytic and photoelectrochemical behaviour of the catalysts.
Recently, owing to the layered structure and unique properties, semiconductor photocatalysts of the Aurivillius oxides Bi2An−1BnO3n+3 (A = Ca, Sr, Ba, Pb, Na, K, and B = Ti, Nb, Ta, Mo, W, Fe) have drawn particular attention.19 Among these Aurivillius oxides, Bi2WO6 has been the subject of many photocatalytic studies including water splitting and photodegradation of organic pollutants under visible light irradiation.20–26 Moreover, the stability of Bi2WO6 has been demonstrated to be excellent under photocatalytic conditions.27–29
Since the catalytic reactions take place on the surface of the catalysts, surface engineering is critical in designing active materials. Among the surface engineering strategies, creating roughness or porosity on the surface seems promising, in particular because it is likely to provide sufficient surface area for the adsorption of reactants and faster migration or diffusion of the parent as well as intermediate products, thereby enhancing the overall efficiency of the process. Moreover, materials having a hierarchical structure are of particular interest.30–32 The aim of the study presented here was to synthesize hierarchical Bi2WO6 in mesoporous form, modify the mesoporous surface with Pt, and then test & compare the photocatalytic activity with other active visible-light-driven semiconductor photocatalysts. Moreover, the study of the photoelectrochemical behavior and its correlation with photocatalytic activity was also included in the scope of this study.
Synthesis of N-doped TiO2 was carried out by annealing a TiO2 sample under NH3 flow following the procedure reported elsewhere.33
Scheme 1 Schematic of nucleation, growth, crystallization and self-assembling process of mesoporous Bi2WO6. |
The mesoporous structure of bismuth tungstate was confirmed by BET analysis and an exemplary nitrogen adsorption–desorption isotherm of this mesoporous sample is depicted in Fig. 3(A). Isotherms were found to be classical type IV, which is characteristic of mesoporous materials. The pore size distribution, shown as the inset, ranged between 2.5 and 5 nm and the average pore size distribution was around 3.5 nm. The average pore size distribution and BET surface area of mesoporous Bi2WO6 was investigated. The pore sizes were found to be 3.2, 3.3, 3.5 and 3.9 nm, whereas the BET surface areas were measured as 41.3, 38.1, 34.7 and 25.8 m2 g−1 for Bi2WO6 synthesized at 130, 150, 170 and 190 °C, respectively. The surface area of the non-mesoporous sample synthesized at 170 °C was found to be 24.6 m2 g−1. A decrease in the surface area with increasing synthesis temperature may be rationalized in terms of the greater extent of condensation of Bi2WO6 layers at higher temperature, which in turn leads to shrinkage or even destruction of the hierarchical structure, consequently resulting in a lower surface area. This assumption was substantiated by microscopic images which showed a partially destroyed hierarchy for the sample synthesized at 190 °C (Fig. S2†).
Fig. 3 (A) N2 adsorption–desorption isotherms together with pore size distribution (inset figure) and (B) DRS of Pt/mesoporous Bi2WO6. |
Optical properties of the bismuth tungstates were investigated by diffuse reflectance spectroscopy. No obvious difference in the absorption spectra of the various samples was found, and the absorption spectrum of mesoporous bismuth tungstate (prepared at 170 °C) is presented in Fig. 3(B). For crystalline and direct bandgap semiconductors, the absorption coefficient satisfies the equation (ahν)2 = A(hν − Eg) for a, where a, ν, A and Eg are the absorption coefficient, light frequency, proportionality coefficient and bandgap energy, respectively.36 The bandgap obtained by extrapolation of the plots (inset figure) of (ahν)2 vs. hν was 2.8 eV, which corresponds well into the visible region.
All the photocatalytic experiments were carried out under visible light irradiation (>420 nm) using a longpass filter. Apart from the fact that RhB is a widely used dye in various industries, it has also been chosen as the model pollutant to study the photocatalytic performance of bismuth tungstate in many previous investigations.20,21,23,29 Hence, RhB was selected as the model pollutant to study the photocatalytic activity so that the results of this study could be correlated to that of previous studies documented in the literature, as needed. Furthermore, noting the fact that dyes also absorb light, particularly visible light, and may lead to an elusive dye-sensitized photochemical process rather than a photocatalytic process, a test experiment was performed in which excitation of a bismuth tungstate aqueous suspension containing RhB was carried out at a wavelength >500 nm. The light intensity of >500 nm radiation was adjusted close to that of >420 nm radiation using neutral density filters. Analysis of the irradiated dye samples under >500 nm radiation did not show any noticeable change in RhB concentration, indicating that the photocatalytic performance evaluated in this study was truly photocatalytic.
Fig. 4(A) shows the change in relative concentration of RhB as a function of the irradiation time in the presence of bare mesoporous Bi2WO6, Pt/non-mesoporous Bi2WO6 and Pt/mesoporous Bi2WO6.
Fig. 4 (A) Comparative photocatalytic activity of Bi2WO6 samples and (B) temporal evolution of RhB absorption spectra in the presence of Pt/mesoporous Bi2WO6; (a) RhB, (b) N,N-diethyl-N′-ethyl rhodamine, (c) N,N-diethyl- or N-ethyl-N′-ethyl rhodamine, (d) N-ethyl rhodamine, and (e) rhodamine (Rh).37 |
The degradation curves can be fitted reasonably well by an exponential decay curve suggesting first order kinetics. As shown, a remarkable increase in the photocatalytic efficiency of the mesoporous catalyst was obtained after Pt modification, and the Pt-modified mesoporous sample exhibited better photocatalytic efficiency compared with Pt/non-mesoporous Bi2WO6; more than 95% of RhB removal was achieved within 20 min in the presence of Pt/mesoporous Bi2WO6, while only ∼75% of dye was removed with the Pt/non-mesoporous sample. The better efficiency of the mesoporous photocatalyst can be explained in terms of increased surface area. As stated above, the surface area of the mesoporous and non-mesoporous samples were found to be 34.7 and 24.6 m2 g−1, respectively. The higher surface area and porous structure seem to facilitate the adsorption–desorption kinetics of RhB, as well as by-products formed during the photocatalytic process making the photocatalytic removal of dye more efficient. A blank experiment (without photocatalyst) was also carried out under visible light and analysis of the irradiated dye solution did not show any noticeable change in concentration, as delineated in Fig. 4(A). Interestingly, in the presence of Pt/Bi2WO6, a blue shift in the dye absorption spectra, together with a change in concentration, was noticeable with respect to the irradiation time. A typical evolution of RhB absorption spectra in the presence of Pt/mesoporous Bi2WO6 with respect to the irradiation time (30 min) is depicted in Fig. 4(B). Apparently, the absorption intensity of RhB decreased at 553 nm upon irradiation and a gradual wavelength shift occurred from higher to lower wavelength (from 553 to ∼500 nm). Under continued irradiation, the peak around 500 nm also disappeared, implying complete removal of dye. Such a characteristic disappearance of RhB was also observed with pure Bi2WO6 in our study, as well as in a study reported by Fu et al.,29 which indicated that neither mesoporosity nor Pt seem to interfere in the degradation route of RhB but, nevertheless, enhance the overall photocatalytic removal of RhB. Since RhB is a tetra-ethylated organic dye, the hypsochromic transition could be correlated with a de-ethylation process followed by destruction of the chromophoric structure.37
As stated above, the synthesis temperature may have a significant impact on various properties of Bi2WO6 which determine the photocatalytic property. The effect of various temperatures, such as 130, 150, 170, and 190 °C, on the photocatalytic activity was investigated and the obtained results are delineated in Fig. 5(A). It can be readily seen that the photocatalytic activity increases with the increase in synthesis temperature from 130 to 170 °C, presumably owing to the formation of a hierarchical structure and improvement in crystallization as indicated by microscopic and XRD analysis. The highest photocatalytic activity was obtained with the sample synthesized at 170 °C followed by a decrease at 190 °C. The higher photocatalytic activity seems to be due to an optimum compromise between the surface area, degree of crystallinity and hierarchical structure. On the other hand, the lower activity shown by the samples prepared at 130 °C and 150 °C could be due to a poor degree of crystallinity and a smaller fraction of hierarchy attained, which was not compensated by the positive effect of high surface area. Furthermore, the decrease in photocatalytic activity of the catalyst synthesized at 190 °C could be explained by the large decrease in surface area, together with the loss of hierarchical structure, as observed by BET and SEM analysis.
Fig. 5 (A) Effect of synthesis temperature and (B) Pt amount on the photocatalytic efficiency of Pt/mesoporous Bi2WO6. |
Since Pt significantly triggered the photocatalytic efficiency of Bi2WO6, the amount of deposited Pt may play an important role in the optimization of the photocatalytic efficiency of the resulting bimetallic photocatalyst. To investigate the effect of the amount of Pt on the photocatalytic activity, Bi2WO6 synthesized at 170 °C was chosen as it showed the highest activity. Varying amounts of platinum (0.25, 0.5, 1.0, 1.5 and 2.0 wt%) were photodeposited onto the surface of mesoporous Bi2WO6 and the resulting photocatalytic activity was investigated. The maximum decomposition of RhB was obtained with 0.5 wt% Pt followed by a decrease at higher metal loadings, as illustrated in Fig. 5(B). The improvement in decomposition of RhB in the presence of platinized bismuth tungstate may be attributed to the possible formation of a Schottky barrier between Pt and Bi2WO6 and an enhanced surface area of Pt/Bi2WO6. In general, the formation of a Schottky barrier between noble metals and semiconductor photocatalysts has been discussed previously by other authors.3,38,39
Briefly, the enhancement in photocatalytic activity may be ascribed to the imbalance between electron and hole densities, which is caused by the quicker oxidation of reductants by holes than oxidant reduction by electrons.3 This gives the potential gradient according to Poisson's equations as well as the carrier concentration gradient.40 These two gradients, according to the hole and electron transport equations,40 are the driving factor for charge carriers to drift and diffuse. When the platinum nanoparticles, having an electron withdrawing capability, are deposited on the Bi2WO6 surface, a Schottky barrier is likely to be formed at the interface of Bi2WO6 and Pt, which facilitates the channelling of electrons from the bulk of Bi2WO6 to the newly formed interface. As a result, the number of electrons in Bi2WO6 decreases, which in turn prevents the electron–hole pair recombination (or enhances the availability of holes for oxidation) and hence a higher photocatalytic activity was observed. A decrease in photocatalytic efficiency at higher Pt loadings may be rationalized by the fact that when the electron density gradient in the bulk of Bi2WO6 decreases substantially owing to the transportation of electrons, the electrical potential gradient decreases and finally the rate of electron diffusion decreases. When the two gradients are too small to further increase the electron flux through the Pt–Bi2WO6 Schottky barrier, a new equilibrium is attained and the additional deposition of platinum particles is unable to create further separation of electrons and holes.41 A fast electron relay, due to the higher amount of Pt nanoparticles (>0.5 wt% in our study), from Bi2WO6 to Pt can deform the potential field in Bi2WO6 particles and draw a part of the holes near the Pt–Bi2WO6 junction, which can facilitate the electron–hole pair recombination. The increase in the Pt–Bi2WO6 contact area can enhance the probability for the recombination of charge carriers and reduce the overall photocatalytic activity. Moreover, a higher amount of Pt can also work as a shield and prevent the incident photons from impinging on the Bi2WO6 surface, thereby decreasing the photocatalytic performance.41 Interestingly, in this study, it appears that the role of Pt is not limited to only being an electron sink, as discussed above, but also to render the interaction between RhB and the photocatalyst surface. This observation was substantiated by performing controlled experiments in the absence of light, and the change in RhB absorption spectra, or concentration with respect to time, is presented in Fig. S5.† Controlled experiments that were carried out in the dark indicated that ∼82% RhB was adsorbed on Pt/mesoporous Bi2WO6 as compared to ∼60% adsorption on pure mesoporous Bi2WO6 after 60 min. The concentration of RhB decreased monotonically with time and no shift in the absorption spectrum, unlike with the photocatalytic process, was observed indicating strong adsorption of RhB on the Pt/Bi2WO6 surface. The greater adsorption of RhB on Pt/Bi2WO6 could be correlated to the enhanced surface area; interestingly, the surface area of platinized bismuth tungstate was found to be 50.35 m2 g−1, which was ∼45% higher than pure mesoporous Bi2WO6. The enhancement in surface area may be attributed to the deposition of small (∼2 nm) Pt nanoparticles. The higher surface area rendered greater adsorption of RhB and better interfacial charge transfer between the catalyst's surface and RhB molecules, which ameliorated the removal of dye.
The photocatalytic activity of Pt/mesoporous Bi2WO6 was compared with other photocatalysts which are widely studied and considered to be highly active visible-light-driven photocatalysts, namely N-doped TiO2, Pt/N-doped TiO2 and Pt/WO3 (ref. 6 and 42) and the obtained results are presented in Fig. 6. As shown, Pt/mesoporous Bi2WO6 showed much better activity than the other photocatalysts studied. For the first 20 min of irradiation, the photocatalytic removal values of dye in the presence of N-doped TiO2, Pt/WO3, Pt/N-doped TiO2 and Pt/Bi2WO6 were found to be ∼34%, ∼58%, ∼72% and ∼96%, respectively. The superior photocatalytic performance of Pt/Bi2WO6 may be rationalized in terms of the following: (a) strong interaction between Pt/Bi2WO6 and RhB, as discussed above, and (b) the standard redox potential of BiV/BiIII. The standard redox potential of BiV/BiIII is more negative (+1.59 V) than the required potential (OH/OH−, +1.99 V) for water oxidation to generate OH radicals.43 The theoretical assumption that Bi2WO6 does not have a sufficient redox potential to produce OH radicals was corroborated in earlier studies.44,45 With this noted fact that Bi2WO6 is incapable of producing OH˙ through water oxidation, it may be strongly anticipated that the separated holes, in the presence of Pt, will be directly involved in the photooxidation of dye which will cause faster removal of RhB in the presence of Bi2WO6 compared with the other studied photocatalysts.
Fig. 6 Comparison of the photocatalytic activity of Pt/mesoporous Bi2WO6, N-doped TiO2, Pt/WO3 and Pt/N-doped TiO2. |
In addition to the study of the photocatalytic properties, the photoelectrochemical behavior of mesoporous Bi2WO6with and without Pt was also investigated. Noting the fact that oxygen is an efficient electron acceptor, experiments were also performed in the presence and absence of O2. The potentiodynamic behavior of both the electrodes in the presence and absence of O2 monitored under visible light is shown in Fig. 7(A). Since the electrodes may show a certain degree of current drift over time scales of 5–10 min and hence create ambiguity between the photocurrents under illumination and the dark current, photocurrents (under illumination and dark current) were measured in a single experiment by alternately turning the light on and off every 20 seconds for 15 min at a constant applied voltage of 0.95 V and the obtained results are presented in Fig. 7(B) and (C). The photocurrent was generated instantaneously upon illumination and reached almost steady state, while a negligible current was observed in the dark, even at a high applied potential (0.95 V). As illustrated in Fig. 7(B) and (C), both of the electrodes possessed similar and reasonably good stability under the studied experimental conditions. Interestingly, the photocurrent density of Bi2WO6 was found to be highly dependent on Pt and O2. In the presence of O2, the photocurrent density of both the electrodes, pure Bi2WO6 as well as Pt/Bi2WO6, remained almost intact and comparable. Interestingly, in the absence of oxygen, the photocurrent generated by the Pt/Bi2WO6 electrode was again comparable but in the case of pure Bi2WO6, a dramatic decline in the photocurrent density was observed. It seems reasonable to rationalize this phenomenon in terms of the fact that the generated electrons, in the case of Pt/Bi2WO6, were scavenged by Pt nanoparticles and O2 appears to have a less obvious role to play as an electron acceptor on the photocatalyst’s surface. Hence, in the presence of Pt/Bi2WO6, the overall flow of photocurrent remained almost intact and irrespective of the presence or absence of oxygen. Moreover, in the absence of Pt or in the case of bare Bi2WO6, excited electrons seem to be taken up by oxygen, generating the analogous flow of photocurrent. However, when the analysis was performed in the absence of both oxygen and Pt, a remarkable decrease in the photocurrent density was observed presumably due to electron–hole pair recombination, which is extremely efficient in the absence of electron acceptors or donors. These findings advocated that both of the entities, Pt as well as O2, serve as an effective electron acceptor.
Fig. 7 (A) Potentiodynamic behavior and (B & C) voltammograms under intermittent illumination/darkness of Bi2WO6 and Pt/mesoporous Bi2WO6 in the presence and absence of O2. |
According to the photoelectrochemical observations, which showed analogous generation of charge carriers in pure and Pt/Bi2WO6 in the presence of O2, a comparable photocatalytic activity of bare and Pt/Bi2WO6 in the presence of O2 may be anticipated. However, in the presence of O2, the photocatalytic activity of Pt/Bi2WO6 was much better compared with pure Bi2WO6. This anomaly may possibly be ascribed to effective interaction of RhB dye with the surface of the platinized bismuth tungstate, as discussed above, and to valence band holes (as effective oxidants). Furthermore, in order to elucidate the role of O2 in the photocatalytic decomposition of RhB, photocatalytic experiments were carried out in the presence or absence of oxygen and the results are illustrated in Fig. 8. The removal of RhB, in the case of both pure as well as Pt/mesoporous Bi2WO6, was faster in the presence of O2 revealing the involvement of O2 in the photooxidation of RhB as an oxidant. In the case of bare Bi2WO6, O2 may take excited electrons from the photocatalyst's surface to generate reactive O2−˙ and more importantly make the valence band holes available for the photooxidation of dye. As a result, significant improvement in the photocatalytic efficiency of Bi2WO6 was observed. Interestingly, in the case of Pt/Bi2WO6, although the photocatalytic response was improved in the presence of O2, it was not as notable as in the case of bare Bi2WO6. So, despite the fact that O2 played a less significant role in the case of Pt/Bi2WO6, removal of RhB was significantly enhanced. This observation indicated the crucial involvement of valence band holes in the photocatalytic oxidation of RhB. As discussed above, Pt-modified Bi2WO6 has a higher surface area, which rendered greater adsorption of RhB and better interfacial electron transfer from RhB to the valence band holes.
Fig. 8 Effect of O2 on the photocatalytic performance of mesoporous Bi2WO6 samples; (a) Pt/Bi2WO6 (with O2), (b) Pt/Bi2WO6, (c) Bi2WO6 (with O2), and (d) Bi2WO6. |
So, based on the collective photocatalytic and photoelectrochemical observations, a plausible mechanism involving electrons, holes, Pt and RhB dye has been proposed in Fig. 9. Briefly, the excited conduction band electrons can be picked up by Pt, or by O2 in absence of Pt, and O2 can be reduced to O2−˙. These O2−˙ radicals can readily attack the adsorbed RhB and degrade it. On the other hand, valence band holes can directly oxidize the RhB, which was observed to be the dominant route, and destroy the chromophoric structure of RhB via the formation of different intermediate products as discussed above.
Fig. 9 Schematic showing the involvement of Pt, O2, electrons and holes in the degradation of RhB dye. |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra45948a |
This journal is © The Royal Society of Chemistry 2014 |