Abstract
photocatalyst was prepared by spray pyrolysis process and tested as photocatalyst for water splitting under UV light. Precursor solution was prepared from and in nitric acid solution and spray-pyrolyzed in air at between 973 and 1273 K. Considerable enhancement of photocatalytic activity was achieved by loading wt% of NiO on the surface of . The NiO loading was more effective on the synthesized by spray pyrolysis in comparison with that synthesized by solid-state reaction. The quantum yield (QY) of photocatalyst was measured by chemical actinometry using potassium ferrioxalate and compared with the apparent photocatalytic activities (APA) which would be more useful for the purpose of photocatalytic reactor design than the quantum yield. The apparent photocatalytic activity (APA) was defined by the rate of hydrogen production divided by weight of catalyst, volume of reactant mixture, duration of irradiation, and power of UV lamp. The validity of the apparent photocatalytic activity (APA) was discussed based on our results and reported activities of photocatalyst loaded with or without NiO.
1. Introduction
Photocatalytic splitting of water by using oxide semiconductors is one of the promising schemes to produce hydrogen from water. Among various metal oxides, NaTaO3 was reported to be one of the most efficient photocatalysts for water decomposition without cocatalyst. The higher potential of the conduction band of NaTaO3 could lead to be more effective to reduce the water to produce H2 since the tantalates possess conduction bands of Ta5d orbital which are more negative position than titanate (Ti3d) or niobate (Nb4d) [1–3].
Kudo and Kato synthesized NaTaO3 by means of solid-state reaction at high temperature (1420 K) for a long reaction time (10 hours), which resulted in the localized segregation of component and loss of stoichiometry due to evaporation of the constituent components [2, 3].
In this report, a spray pyrolysis was proposed as a way of avoiding problems of solid-state reaction method. The spray pyrolysis process is an aerosol process in which one droplet of precursor solution converts into one particle in short residence time at lower processing temperature than the solid-state reaction process. Because of short residence time and low temperature, the spray pyrolysis method produces powders of high phase-purity and uniformity in composition. Particles produced by the spray pyrolysis are usually spherical and aggregation-free [4, 5].
The quantum yield (QY) has been used to compare the intrinsic activity of photocatalyst. The quantum yield includes the effect of loss of photons scattered or reflected by the photocatalyst itself and absorbed by some impurities or reactants. The number of photons absorbed by photocatalyst has been determined by two methods. One is using the light-flux meter such as a 1815-C (Newport) or a power meter (Advantest, Si diode, TQ8210 or Molectron Power Max 5200). The other is a chemical actinometry method. Potassium ferrioxalate method developed by Parker et al. [6, 7] has been a standard chemical actinometer and used over a broad band of wavelength. However, the exact evaluation of quantum yield for the heterogeneous reaction has not been easy because of the difficulty of measuring the intrinsic number of photons absorbed by photocatalyst. In addition, the quantum yield has been strongly dependent upon the determination methods as well as experimental conditions. It has been frequently encountered that the values of quantum yield have been directly related only to the rate of H2 evolution, although the experimental conditions such as the amount of catalyst, reactor type, and power of the lamp are quite different.
Therefore, in this report, a new definition of photocatalytic activity was proposed for the purpose of comparing photocatalysts from the reactor design point of view.
2. Experiment
2.1. Synthesis of photocatalyst
To synthesize the NaTaO3, a spray pyrolysis system which consists of an ultrasonic atomizer, vertical furnace reactor, and particle collector was employed (see Figure 1). A commercial ultrasonic atomizer has four oscillators (Dong-Lim Eng.) with atomizing frequency of 1.7 MHz. The droplets generated by atomizer were transported into vertical furnace reactor by carrier gas (air) at the flow rate of L/min and passed through a 0.84 m long and 30 mm i.d. quartz tube. The reactor temperature was controlled at between 973 K and 1273 K. The prepared particles were collected by a thimble filter (28 mm 100 mm, Toyo Roshi Kaisha Ltd.).
Ta(OC2H5)5 (Aldrich, purity 99.98%) and NaNO3 (Junsei, purity 99.5%) were used as precursors. For the preparation of 0.125 M colloid solution, Ta(OC2H5)5 was dissolved in distilled water containing excess nitric acid, and it was stirred vigorously by magnetic stirrer and heated to K or use without heating at 298 K. When the precursor solution became clear or colloid without any precipitates, NaNO3 was added and stirred for dissolution. NiO as cocatalyst was loaded by an impregnation method from aqueous solution of Ni(NO3)26H2O (Aldrich). NaTaO3 was added to the solution and stirred by magnetic stirrer. Then, water was removed by using a rotary vacuum pump. Finally, the NiO/NaTaO3 powder was calcined at 540 K for 3 hours under air.
2.2. Characterization of photocatalyst
Synthesized powders were identified by X-ray diffraction (Rigaku, D/MAX-RB), and the surface morphology and average particle size of the powders were observed by field-emission SEM (Philips, 533M). The dispersion of NiO on surface of powder was observed by field-emission TEM (FEI, Technai G2 F3 S-Twin). The surface area was determined by BET measurement (Micromeritics, ASAP 2000). A diffuse reflection spectrum was obtained by using a UV-vis-NIR spectrometer (Jasco, Ubest V-570) and was converted from reflection to absorbance by the Kubelka-Munk method.
2.3. Photocatalytic activity measurement
Hydrogen production reactions were carried out in a closed circulation system as shown in Figure 2. The system consisted of reactor, reservoir, and detector (gas chromatograph). The reactor was a donut shape but it had only one Pyrex outer tube without inner quartz tube so that the wall of UV lamp was in direct contact with reactant mixture. Removal of the quartz inner tube reduced the construction cost of this type of slurry reactor. Inner diameter and length of the outer tube were 34 mm and 320 mm, respectively. UV-C lamp of 15 W (SANKYO, Japan) was used as an illuminating source. The wave length of this lamp was predominantly 254 nm (4.9 eV). The photocatalyst powder of g was dispersed in mL of distilled pure water or with methanol in the reservoir and stirred with a magnetic stirrer. Before turning on the light, the system was deaerated with N2 gas for an hour and kept at atmospheric pressure. The amount of gas evolved was determined online with gas chromatograph equipped with TCD detector.
2.4. Measurement of quantum yield
The moles of photons absorbed by photocatalyst were measured by a chemical actinometry employing potassium ferrioxalate (K3Fe(C2O4)33H2O) synthesized from K2C2O4H2O and FeCl3 to calculate quantum yield. Solution of 0.006 M K3Fe(C2O4)33H2O in 0.06 M H2SO4 was prepared and poured into reaction system for irradiation. After 0.1 M K3Fe(CN)6 was added to sample irradiated, the absorbance was measured by UV-vis spectrophotometer (AIS, DT 1000CE). This work was carried out in a dark room. Finally, the light intensity of an examined light source was determined by[8] where A is absorbance, d is length of absorption path, is extinction coefficient, is QY at wavelength of light used, V1 is volume of sample, V2 is volume of actinometer irradiated, V3 is volume of diluted solution of V1, and t is irradiated time.
3. Results and Discussion
The XRD pattern of NaTaO3 powders synthesized by spray pyrolysis is shown in Figure 3. The main peak appeared at 2θ = 22.8°. The XRD pattern confirmed that the homogeneous NaTaO3 crystal of high crystallinity was synthesized. SEM images of NaTaO3 were shown in Figure 4. The prepared powders were generally spherical. However, some hollow and shell-like particles are present together. To optimize the preparation condition, the effect of the precursor preparation temperature on the crystal characteristics was tested. In Figure 5, the optimum precursor preparation temperature was 353 K, and most of samples were prepared at this temperature.
(a)
(b)
(c)
(d)
SEM images and XRD patterns of NaTaO3 powders with variation of spray pyrolysis temperature are shown in Figures 6 and 7, respectively. Based on these two results, the optimum spray pyrolysis temperature was found to be 1173 K. According to the diffuse reflectance spectra of NaTaO3 powder in Figure 8, the bandgap of NaTaO3 was estimated to be 4.0 eV. It was determined from the onset of the absorption using the following equation:
(a)
(b)
(c)
(d)
Physical properties such as BET surface area, pore volume, and pore size of NaTaO3 were summarized in Table 1. The BET surface area was larger by ten times than that of particles synthesized by conventional solid-state reaction.
NiO was loaded on NaTaO3 synthesized by spray pyrolysis at 1173 K to enhance the photocatalytic activity for water splitting. Figure 9 is TEM image of NiO-loaded NaTaO3 synthesized by spray pyrolysis. Figure 9(b) is a magnified image of the circle in Figure 9(a). In Figure 9(a), NiO particles (10 nm) were highly dispersed on the surface of NaTaO3 synthesized by spray pyrolysis.
(a)
(b)
Table 2 shows the photocatalytic activities for water splitting in pure water over NaTaO3 with and without NiO. The rate of hydrogen evolution of naked sodium tantalate photocatalysts prepared by spray pyrolysis was 1.76moL/hr while that of NaTaO3 loaded with NiO was dramatically increased to 222moL/hr. The activity enhancement factor by NiO/NaTaO3 synthesized in the present work was much larger than that of NiO/NaTaO3 synthesized by means of the conventional solid-state reaction [2, 3]. Figure 10 shows that the cumulative amount of gas evolution increases almost constantly with irradiation time and the ratio of hydrogen and oxygen is maintained at 2.
In order to determine the optimum loading of photocatalyst, effect of photocatalyst loading (NiO (0.2 wt%)/NaTaO3) on the rate of H2 evolution and quantum yield was tested, and results are shown in Figure 11. The optimum loading was found to be around 0.4 g/L.
The quantum yield (QY), which is defined by (3), has been used to evaluate the activity of photocatalyst over a broad band of wavelength:
In order to design an optimum photocatalytic reactor system, consumption of photocatalyst, electric power, volume of slurry reactor, and duration of reaction should be minimized, if catalysts with high-quantum yield were available.
To evaluate photocatalysts for the purpose of optimal design of photocatalytic reactor, the apparent photocatalyst activity (APA) is proposed aswhere and .
This definition is more useful for designing a photocatalytic reactor than the quantum yield. The calculated values of the apparent photocatalytic activity from literature were summarized in Table 3, along with the reported values of quantum yield. For comparison, the values of APA and quantum yield of the current work were summarized in Table 4. It should be noted that the reactors, types of lamps, and measurement methods of photon intensity were different in those tables.
Figure 12 is a plot of APA versus quantum yield for data summarized in Tables 3 and 4. It is clear that the high-quantum yield is not always preferable criterion from the reactor design point of view. For example, the rate of H2 evolution reported by Abe et al. [11] is 220moL/hr and that of present study by using NaTaO3 is 237.5moL/hr, and the quantum yields are also very similar to each other. However, the APA for the catalyst of Abe et al. [11] is 2.7moL/gLhrW whereas that of the present work is 439.8moL/gLhrW, although the rate of hydrogen evolution is very similar.
Figure 13 is a plot of apparent photocatalytic activity of current work and published results. The APA of current work is almost equal or slightly better than the highest APA reported by Kudo and Kato [16] although the quantum yield of the current work is lower than that of Kudo and Kato’s result.
However, the purpose of this paper is not to claim that we prepared superior photocatalyst, but to claim that both quantum yield and apparent photocatalytic activity are equally important when comparing the performance of photocatalyst for water splitting.
4. Conclusion
(1) NaTaO3 photocatalyst was synthesized by means of spray pyrolysis in relatively mild experimental conditions: K of reaction temperature and seconds of residence time. The reaction temperature and duration of reaction are 1420 K and 10 hours, respectively, when NaTaO3 was prepared by the conventional solid-state reaction
(2) To enhance the photocatalytic activity for water splitting, wt% of NiO was loaded on the surface of NaTaO3 powder synthesized by spray pyrolysis, which resulted in a considerable increase of photocatalytic activity in comparison with that of NaTaO3 powder synthesized by means of conventional solid-state reaction.
(3) Anew definition of photocatalytic activity was proposed from the reactor design point of view and named as apparent photocatalytic activity (APA), which is defined by the rate of hydrogen production divided by amount of catalyst, volume of reactant, duration of reaction, and power of lamp irradiated.
(4) Quantum yield is an important criterion to screen photocatalyst. However, the high-quantum yield is not always preferable for designing an economically feasible photocatalytic reactor system. Small size reactor, small amount of photocatalyst, short reaction time, and low consumption of power are also important for economically competitive production of hydrogen. Therefore, it is strongly recommended to report the apparent photocatalytic activity along with quantum yield when improvement of photocatalytic activity is to be reported.
(5) In this report, only the case of slurry reactor was discussed. If the type of photocatalytic reactor is a flat-surface reactor, the apparent photocatalytic activity would be reported in terms of unit surface area in place of the volume of reactor, because the goal of reactor design is to maximize the hydrogen evolution per unit surface area.