Abstract
The photocatalytic production of hydrogen from water using solar energy is potentially a clean and renewable source for hydrogen fuel. This study examines the production of hydrogen over In, P-TiO2s photocatalysts. 1 mol% In-TiO2 and P-TiO2 were produced using the solvothermal method and were treated at 500 and 800∘C to obtain anatase and rutile structure, respectively. The photocatalysts were characterized by X-ray diffraction, photoluminescence spectra, X-ray spectroscopy, UV-visible spectroscopy, and scanning electron microscopy. The production of H2 from methanol photodecomposition was greater over the rutile structure than over the anatase structure of TiO2. Moreover, the amount of hydrogen was enhanced over In-TiO2 and P-TiO2 compared to that over pure TiO2; the production increased by about 30%. The structural effect and the addition of In, P have significant influence on the H2 production from methanol/water decomposition.
1. Introduction
The importance of hydrogen as a source of fuel in the near future is due to its application as a compact energy supplier in fuel cells and batteries. Hydrogen fuel can be produced from clean and renewable energy resources, and thus, its life cycle is clean and renewable. Compared to the conventional and expensive methods, the photocatalytic splitting of water using TiO2 offers a promising way for clean, low-cost, and environmentally friendly production of hydrogen using solar energy. Early work of TiO2-assisted photo-electrochemical hydrogen production was reported by Fujishima and Honda [1]. Since then, the technology for generating hydrogen by the splitting of water using a photocatalyst has attracted much attention. Recently, hydrogen production has been extended to the photocomposition of methanol which has a lower splitting energy than water. Kawai and Sakata suggested the following overall methanol decomposition reaction [2, 3]: Overall,
Consequently, the decomposition energy for methanol is 0.7 eV. Most investigations of hydrogen production via methanol photodecomposition have focused on modifying TiO2 in many reports, and the photocatalytic method is modified either by anion doping or metal ion doping or by loading with noble metals. Anpo and Takeuchi suggested Pt/TiO2 [4], Bamwenda et al. compared Au-loaded TiO2 and Pt-loaded TiO2 [5], Sakthivel et al. used Pt, Au, and Pd-loaded TiO2 [6], Wu et al. studied CuOx/TiO2 [7], and Park and Kang used Ag-incorporated TiO2 [8]. In the case of metal ion doping, transitional metal ion doping and rare earth metal ion doping have been extensively developed for enhancing the TiO2 photocatalytic activities. Choi et al. studied the photocatalytic activity of 21 metal ions doped into TiO2, and among of them, Fe, Mo, Ru, Os, Re, and V ions exhibited good activity [9]. Some of the recent reports have studied the photo-electrochemical characteristics of anion (N, F, C, S, etc.) doped TiO2 powder and colloidal materials under UV-visible and visible illumination [10–13]. However, rapid electron-hole recombination and thermal instability are thought to be the major impediments. However, the number of known photocatalysts is limited, and their activity is still low. There is an urgent need to develop new photocatalysts that have greater hydrogen-producing activity under visible light irradiation.
In this study, we evaluated two new materials, In-TiO2 and P-TiO2, to reduce the large band gap of pure TiO2. To investigate the structural effect of the photocatalysts, we prepared the photocatalysts with anatase and rutile structures using thermal treatment at 500 and 800°C, respectively which were then used to produce hydrogen gas via methanol photodecomposition. To determine the relationship between In, P species and catalytic performance for H2 production, the In-TiO2 and P-TiO2 photocatalysts were examined using X-ray diffraction, photoluminescence, X-ray photon spectroscopy, and UV-Visible spectroscopy.
2. Experimental
2.1. Catalysts Preparation
In this study, a commonly used solvothermal method was employed to synthesize In and P-TiO2 catalysts. The reagents used for the preparation of sol mixture were titanium tetraisopropoxide (99.95%, Junsei Chemical, Japan), indium trichloride (98%, Aldrich, USA) and phosphoric acid (Wako Pure Chem. Ltd., Japan) as Ti, In, and P precursors, respectively. The metal precursors were mixed with ethyl alcohol (99%, Wako Pure Chem. Ltd., Japan) in an autoclave (model R-211, Reaction Engineering Inc., Korea) heated at 200°C for 10 h at a rate of 10°C min−1. During the thermal treatment, Ti, In, and P were hydrolyzed by the hydroxyl group in solvent leading to the formation of nanosized In and P-TiO2 crystals. The resulting powder was washed with distilled water until pH-7 was reached and then dried. Finally, the samples were thermally treated at two different temperatures, 500 and 800°C, for 3 h to remove impurities like the residual carbon or chloride ions on the powder surface and also to study the effect of calcination on the phase formation in the prepared catalysts.
2.2. Characterizations
The prepared catalysts were identified through powder X-ray diffraction analysis (XRD, model PW 1830, Philips, The Netherlands) with nickel-filtered Cu Kα radiation (40 kV, 100 mA) at 2θ angles of 5 to 90°. The scan speed was 10° min−1, and the time constants were 1 sec. The scanning electron microscopy (SEM) images were obtained using Leica 440, Beam source is tungsten filament, resolution is 4.5 nm, and probe current was in between 1 pA to 1 μA. For analysis of the binding energy among In3d, P2p, Ti2p, and O 1s, the X-ray photoelectron spectroscopy (XPS, PHI 5700, PHI com) was employed. The 1.0 mol% In and P incorporated TiO2 powder were pelletized at 2.0 × 104 kPa for 10 min; the 1.0 mm sized pellets were then maintained overnight in a vacuum oven (1.0 × 10−7 Pa) to remove the surface water molecules prior to measurement. The case pressure of the ESCA system was below 1 × 10−9 Pa. The experiments were recorded using a 200 W power source and angular acceptance of ±5°, with the analyzer axis set at an angle of 90° to the specimen surface. Wide scan spectra were measured over the binding energy range 0 to 1200 eV, with pass energy of 100.0 eV. The Ar+ bombardments of the 1.0 mol% In, P incorporated TiO2s were performed with ion currents between 70 and 100 nA, over a 10.0 × 10.0 mm area, with a total sputtering time of 2400 s, divided into 60 s intervals. The UV-visible spectrum was obtained using a JASCO V-570 spectrometer equipped with a reflectance sphere. The spectral range varied from 200 to 800 nm. Photoluminescence spectroscopy was measured at room temperature using a Phillips H-9000 using HeCd as the beam source at 325 nm.
2.3. H2 Production from Methanol/Water Decomposition over In, P-TiO2
In general, water decomposition systems are based on the evaluation of accumulated H2 from water, and most of the systems employ semi-batch type reactors like plug flow reactor. In our study, we used a custom-made liquid photoreactor setup for the photodecomposition of methanol/water. For hydrogen generation, 1.0 g of the powdered catalysts was added to 500 ml each of distilled water and methanol in a 1300 mL Pyrex reactor. UV-lamps (3 × 6 Wcm−2, 15 cm length × 2.0 cm diameter) emitting radiation at 365 nm were used. A flask for hydrogen storage collects the product from reactor using a vacuum pump. The produced hydrogen was analyzed by GC/TCD (Shimadzu 17A, Japan) using molecular seive-5A column.
3. Results and Discussion
3.1. H2 Production from Hydrolysis of Water with Methanol on In, P TiO2s Photocatalysts
Figure 1 shows the catalytic activity of In and P-TiO2 with anatase and rutile phase for the production of H2 from methanol/water photodecomposition. Unfortunately, the GC detector could not sense the H2 production until 2 h had elapsed because too little product had decomposed. Then rate of H2 evolution becomes very stable after 2 h, and activities reach to steady-state at 8 h. However In, P incorporated TiO2 increase rate of H2 evolution until 12 h. In general, previous researchers plotted continuously increasing graph because H2 production accumulated by using batch system [4–13]. While, in this study, used system for H2 evolution is semi-batch system, it can be calculated H2 production per hour. So, the time to reach steady state can be obtained, and the graph could not increase continuously. The catalysts with rutile phase exhibit relatively higher activity than those with anatase phase in a semi-batch type liquid photosystem. The maximum H2 production of 3,300 μmol at about 12 h was achieved for In-TiO2 calcined at 800°C (rutile structure), and with further increase in time, a small decrease in its activity was observed. The increasing activity of the added In and P catalysts may be caused by increased reduction activity, substantial increase in the rate of photo absorption, and change of atomic bonding and composition of catalyst surface.
3.2. Characterization of In, P-TiO2 Catalysts
Figure 2 shows the XRD patterns of P-TiO2, In-TiO2, and TiO2 powders treated at 500 and 800°C. Depending on the calcination treatment temperature, well-developed anatase (500°C) and rutile (800°C) phases were formed, and their diffraction pattern are labeled as “A” and “R”, respectively. The peaks of the In, P are very well inserted into the TiO2 structure and that explains the absence of their peaks in Figure 2. The peaks corresponding to anatase were assigned as 2θ = 25.2° (101), 37.5°(004), 47.5° (220), 53.8° (105), 54.9° (211), 63.0° (204); for rutile as 2θ = 27.5° (110), 36.1°(101), 41.3° (111), 54.4° (211), 56.7° (220). Especially, rutile phase of In incorporated TiO2 appears peak of In2O3 at 2θ = 30.6°, 31.4°. It is considered that indium is formed crystalline easily than phosphorus by effect of sintering [14].
Figure 3 shows SEM micrographs of TiO2, In-TiO2, and P-TiO2 particles. The photocatalysts consisted of relatively irregular and spherical particles of varied size between 50 to 500 nm. Among them, the particles with anatase structure prepared at 500°C were smaller than those with the rutile structure prepared at 800°C. This result is related to a sintering effect in which the particle size increases with the calcination temperature due to the increased agglomeration of the particles. Also, TiO2 without additive material is composed of granules with a smooth surface. However, after incorporating In and P on TiO2, the granules became irregular and angular with the grain size of about 0.3–50 μm.
(a) TiO2
(b) P-TiO2
(c) In-TiO2
XPS analysis of O 1s and Ti 2p orbital of TiO2, In-TiO2, and P-TiO2 are shown in Figure 4. The Ti 2p1/2 and Ti 2p3/2 spin-orbital splitting photoelectrons for anatase TiO2 were located at biding energies of 464.34 and 458.57 eV, which were assigned to the presence of general Ti4+ [15, 16] as shown in Figure 4(a). From Figure 4(b), the binding energy of Ti 2p orbital was observed to decrease due to the added In and P which infers the transformation from higher Ti4+ to lower Ti3+ valence state. Generally, a high binding energy means that the metal has a high valence [17]. At this situation, an increase in catalytic activity is expected to change valence as Ti3+. In the case of O 1s (Figure 4(c)), the two overlapping peaks were isolated using the Gaussian method. The O 1s peaks at 529.8 and 531.20 eV were assigned to metal oxidation and metal hydroxyl group, respectively. With increasing area of the hydroxyl group, the hydrophilic and catalytic activities are expected to increase.
(a)
(b) Ti 2p3/2
(c) O 1s
From Figure 5, the In 3d5/2 and P 2p3/2 spin-orbital splitting photoelectrons were located at the biding energies of 444.03 eV and 133.02 eV which is identified with the oxidation state of In3+ and P5+, respectively in TiO2. Especially, the existence of phosphorous in the pentavalent oxidation state has a decreasing effect on valence-hole recombination because of the unbalanced electrical charge [18].
(a) In 3d5/2
(b) P 2p3/2
UV-Vis spectroscopy was used to examine the light absorption ability of the investigated photocatalysts as shown in Figure 6. The onset absorption wavelength and corresponding band gap energy of the photocatalysts used in this study as obtained from the UV-Vis spectra are summarized in Table 1. It is clearly seen that the absorption band of the synthesized In and P-TiO2 is in the UV light range of 200 to 400 nm, and the shift of the onset absorption edges toward longer wavelength with rutile and anatase form [19].
Figure 7 shows photoluminescence emission on In-TiO2, P-TiO2, and pure TiO2. The objective of this experiment is to study the addition of In and P on the electro-hole recombination rate. According to the test results, the shape of the peak discloses no change in the spectrum of the used catalysts. However, the peak intensity decreases depending on the doping level of In and P in TiO2. Generally, when photoluminescence emission value is large, the numbe of recombination electrons should also large number of electrons are photoexcited. Therefore, the larger the emission value is, the photocatalytic reaction might be. However, in this study the higher the photoluminescence emission values were, the lower the photocatalytic activities that could be indicated. The result could be thought that the metal incorporated TiO2, the amount of recombination of electron-hole pairs is decreased, which could be due to photoexcited electrons joining in photocatalytic reaction rather than recombining with holes emitting the absorbed energy [20–23].
4. Conclusions
This study focused on using In, P incorporated TiO2 catalysts synthesized by solvothermal method and used for the production of H2 from methanol/water photodecomposition. Compared to anatase structure, the rutile form exhibited better H2 production from methanol/water photodecomposition. Compared to P-TiO2, the production of H2 was a little better in the case of In-TiO2 and also showed a 30% increase in the activity than pure TiO2. The well-defined dispersion of the metal in the TiO2 was confirmed from XRD result. Increase of the OH group on the catalytic surface, change of Ti structure, increase in the absorption of visible light, and decrease of valence-hole recombination were confirmed using XPS, UV-Vis, and PL data. The prepared catalysts were found to exhibit good activity than pure TiO2, and our future studies will focus on further improving the catalytic activity.
Acknowledgment
The authors are grateful to the Financial Grant (KRF20090088941) provided by the Korea Research Foundation.