The Influence of the Ammonolysis Temperature on the Photocatalytic Activity of β-TaON
Phase-pure tantalum oxynitride (β-TaON) powders were synthesized by thermal ammonolysis of Ta2O5 powders. X-ray diffraction revealed an enlargement of the unit cell and an increase of the crystallite size with increasing ammonolysis temperature. Scanning electron microscopy showed reduced particle sizes for β-TaON synthesized at 800 and compared to the precursor oxide. With increasing nitridation temperature the Brunauer-Emmett-Teller surface area was reduced and the nitrogen content increased. UV-Vis spectroscopy showed a bandgap energy of 2.6–2.4 eV. The highest oxygen evolution rate of 220 μmol·g−1·h−1 was achieved for β-TaON synthesized at . The factors determining the photocatalytic activity of β-TaON powders were found to be the specific surface area and defects in the β-TaON.
Solar radiation is one of the most promising alternative energy resources to fossil fuels and nuclear power. The conversion of sunlight into electricity or other forms of energy can help to meet future energy needs . However, the production of alternative fuels through the direct utilization of sunlight by photocatalysis or photoelectrochemical (PEC) cells is still far from being marketable as neither acceptable efficiency nor sufficient stability have been demonstrated [2–4]. To achieve efficient production of hydrogen by solar water splitting several challenges have to be overcome [5, 6], in particular, the inefficient absorption of the solar spectrum and the high concentration of defects in photocatalytic material which act as recombination centers for the charge carriers. Since Fujishima and Honda  first reported water splitting on a TiO2 electrode in the early 1970s, various concepts of PEC cells have been developed [8–10]. The advantage of TiO2 is its stability and low cost, and the disadvantage is that the low intensity of the solar spectrum in the ultraviolet (UV) can be utilized due to its large bandgap (~3 eV). Materials with a bandgap of around 2.0–2.4 eV are best suited to collect the most intense part of the solar spectrum in the visible light region. This is why oxynitrides are considered as promising semiconductor photoanodes, since the nitridation shifts the bandgap of the oxide to longer wavelengths into the visible light region [11–19]. Especially tantalum oxynitride (TaON) attracted much attention due to its suitable bandgap energy (2.4–2.5 eV) [20, 21]. Synthesis and crystal structure of β-TaON were first reported by Brauer and Weidlein  followed by additional experimental and theoretical studies on its physical, chemical, and structural properties [23–26]. Hitoki et al.  have proven that β-TaON acts as a stable and efficient photocatalyst for the oxidation of water into O2 under visible light irradiation. Banerjee et al.  showed that TaON nanotube arrays can absorb a broad band of the visible light which in turn leads to an enhanced photocurrent. Abe et al. successfully demonstrated efficient overall water splitting with nanogranular TaON thin films fabricated by electrophoretic deposition of β-TaON powder and subsequent loading of IrO2 nanoparticles . Recently, TaON photocatalysts modified with ZrO2  or polyaniline  exhibited a significant improvement of the photocatalytic activity. Repeatedly, critical factors for the enhancement of photocatalytic H2 and/or O2 evolution reaction have been investigated [21, 30, 31]. The density and distribution of defects, crystal size, surface area, surface morphology, and cocatalyst loading have been found to be decisive factors. Since the efficiency of the catalyst is governed by a complicated interplay between bulk and surface properties, the control of morphology, surface area, crystallinity, and defect concentration via a tailored synthesis procedure is essential for the improvement of the photocatalytic activities in β-TaON.
TaON can be obtained by annealing Ta2O5 in an ammonia [22, 23]. However, the synthesis of phase-pure β-TaON is a particularly challenging task. Small variations in the synthesis procedure often lead to undesirable nitrogen-rich phases or polymorphs of TaON. For instance, Ta3N5 can be easily formed when Ta2O5 reacts with gaseous ammonia in the temperature range of 700°C–900°C  suggesting that the temperature is one of the most important experimental parameters that determines the resulting composition of the compounds. Apart from the chemical composition, the particle size, crystallinity, and surface area are affected by the ammonolysis conditions. Therefore, the objective of this study is to synthesize phase-pure β-TaON by thermal ammonolysis and to tune the surface area, the amount of nitrogen-incorporation, and the crystal size in order to improve the photocatalytic activity without using precious metals as a cocatalyst. To this end, the influence of the ammonolysis temperature on the crystal structure, microstructure, nitrogen content, and bandgap energy of β-TaON was studied to identify the best synthesis conditions. Systematic comparison of the experimental results obtained by various techniques is expected to provide insight into the particularly crucial factors for the photocatalytic activity of β-TaON.
2. Experimental Procedures
A series of tantalum oxynitride (β-TaON) powders was synthesized by thermal ammonolysis of crystalline Ta2O5 powder (Fluka, 99.9%) in a tubular bed reactor using flowing NH3 (Messer, ≥99.98%) at different temperatures (800, 850, 900, 950, and 1000°C). Around 3 g of the white Ta2O5 powder were heated (20°C/min) to the desired temperature in an NH3 stream (15–20 mL/min) and the ammonolysis was carried out for 16.5 h. Subsequently, the oxynitride samples were cooled down to room temperature in the same NH3 stream. At 800°C a prolonged ammonolysis reaction time of 68 h was required to achieve the complete transformation into the phase-pure tantalum oxynitride.
X-ray diffraction (XRD) patterns were obtained using a PANalytical X’Pert PRO θ-2θ scan system equipped with a Johansson monochromator and an X’Celerator linear detector. The incident X-rays had a wavelength of 1.5406 Å (Cu-Kα1). The diffraction patterns were recorded from 20° to 100° (2θ) with an angular step of 0.0167°. XRD patterns were analyzed by the Le Bail method  integrated in the program Fullprof  to determine the lattice parameters. The crystal sizes are calculated by Scherrer equation . The Thompson-Cox-Hastings pseudo-Voigt function  was chosen as profile function and the background was modeled with a sixth-order polynomial function. The instrumental contribution to peak broadening was estimated by measuring the standard reference sample CeO2 (NIST SRM674b).
A scanning electron microscope (SEM, LEO JSM-6300F) was used for microstructure analysis. The Brunauer-Emmett-Teller (BET) surface area of the powders was determined by N2 physisorption at the temperature of liquid nitrogen (–196°C). Thermal reoxidation studies with thermogravimetric analysis (TGA) were carried out with a NETZSCH STA 409 CD thermobalance. Around 50 mg of the β-TaON powders were heated in an alumina crucible at a heating rate of 10°C/min up to 1500°C in synthetic air (50 mL/min). UV-visible diffuse reflectance spectra were collected using a UV-Vis spectrophotometer (UV-3600 Shimadzu UV-VIS NIR) equipped with an integrating sphere. Spectra were recorded in the range of 200 to 1200 nm. The Kubelka-Munk conversion was applied to the obtained reflectance spectra and the bandgap wavelength was estimated by extrapolating the absorption onset wavelength to the wavelength axis.
Photocatalytic reactions were performed in a closed constant-volume system connected to a gas chromatograph. A 150 W halogen lamp (Osram, HLX 64640) was used as a light source. 0.1 g β-TaON and 0.2 g La2O3 powders were dispersed in an aqueous AgNO3 solution (0.01 M, 250 mL). La2O3 was used as a buffer to maintain the pH of the solution at around 8 during the reaction  and the Ag+ ions were employed as sacrificial electron acceptors for O2 evolution. The reaction system was made of gas-tight glass and evacuated sufficiently to remove air. Before light illumination, 100 mbar of He was introduced into the reaction system to suppress vaporization of water. Accordingly, the leakage of atmospheric oxygen was prevented during the O2 evolution reaction.
3. Results and Discussion
X-ray diffraction (XRD) patterns of β-TaON powders synthesized at different temperatures (800°C, 850°C, 900°C, 950°C, and 1000°C) are shown in Figure 1(a). All observed reflections were assigned to pure β-TaON (JCPDS-PDF 71-0178) and no diffraction peaks arising from any secondary phases or impurities were detected. It is well known that Ta3N5 can be easily formed as a secondary phase, when dry NH3 gas is used in the ammonolysis. In this study, it was essential to apply a low flow rate of dry NH3 (15–20 mL/min) in order to obtain phase-pure β-TaON phases. Whenever the flow rate exceeds 20 mL/min, Ta3N5 and/or polymorphs of TaON were often found as secondary phases. The full width at half maximum (FWHM) of the reflections decreased with increasing ammonolysis temperature mainly due to the crystal growth. An exception to this is β-TaON synthesized at 800°C (Figure 1(b)) where the ammonolysis was conducted for 68 h. The prolonged reaction time leads to a reduction of the FWHM thus an increase of the crystal size.
The XRD pattern and Le Bail fitting profile of the exemplarily selected β-TaON powder synthesized at 1000°C are presented in Figure 2. The difference plot of observed and calculated diffraction profiles is shown together with the Bragg positions. The Le Bail fitting results for the cell parameters of β-TaON as a function of temperature are listed in Table 1. A monoclinic crystal structure with space group P21/c was identified. With increasing ammonolysis temperature, the lattice parameters b, c, and β gradually increase resulting in an expanded unit-cell volume. Since the anionic radius of nitrogen is larger than that of oxygen, the unit-cell expansion gives indirect evidence that more nitrogen is incorporated into the crystal lattice of β-TaON with increasing ammonolysis temperature. Interestingly, the lattice parameter decreased with increasing ammonolysis temperature. Our fitting results show that with increasing ammonolysis temperature the unit-cell expands along the y- and -axis. Yashima et al.  explained similar findings with complete occupational ordering of oxygen and nitrogen in alternating layers normal to the direction in β-TaON. Using Rietveld refinement of the neutron and synchrotron X-ray diffraction patterns, they found that oxygen and nitrogen atoms in β-TaON are coordinated with three and four tantalum atoms (OTa3 and NTa4), respectively. The structure exhibits a stacking of alternative NTa4 and OTa3 layers. With the increasing degree of substitution, crystallographically favored positions in the NTa4 layers are filled with more nitrogen resulting in an anisotropic expansion of the unit cell. The mechanism, however, leading to the observed crystallographic anisotropy is not yet fully understood and remains to be elucidated. The crystallite sizes were calculated based on the Scherrer equation with the Rietveld refinement and they are summarized in Table 2. The average size increases with increasing ammonolysis temperature. β-TaON synthesized at 800°C shows a slightly larger crystal size compared to those obtained at 850 and 900°C, since the ammonolysis time was extended.
Figure 3 shows SEM images of β-TaON. The size of individual particles of the precursor oxide Ta2O5 was less than 1 μm and the surface of the particles was rather smooth. After ammonolysis at 800°C and 900°C, pores were observed on the surface of β-TaON particles and the particle size was reduced compared to the precursor oxide. Increasing the ammonolysis temperature to 1000°C led again to a smoothing of the surface due to sintering. These observations are consistent with the BET surface area of the powders (Table 2). The BET surface area of the starting substance Ta2O5 was 3.5 m2/g. With increasing ammonolysis temperature the surface area first increased going through a maximum at 850°C and eventually decreased again. The powder synthesized at 800°C shows a slightly lower specific surface area (SSA) compared to that at 850°C. The lower SSA of samples prepared at 1000°C compared to the precursor oxide is attributed to the sintering of β-TaON particles.
Thermal stability and nitrogen content of the oxynitride powders were determined by TGA during a reoxidation process. Figure 4 illustrates the weight gain and loss upon heating of the oxynitride powders in synthetic air. All samples were thermally stable up to around 400°C at least. Above this temperature, the uptake of oxygen leads to gradual mass increase up to 750°C followed by a sharp increase peaking at 850°C. The subsequent mass loss above 850°C is attributed to the release of nitrogen which is commonly observed in oxynitrides during the reoxidation process [37, 38]. It should be mentioned that the nitrogen is already released gradually as N2 during the uptake of oxygen which otherwise could not be incorporated into the lattice . In many reoxidation reactions, a dinitrogen adduct intermediate phase is formed between the original oxynitride and the final oxide . It explains a higher weight than that of the final oxide due to nitrogen retention [39–41]. The intermediate phase is stable in a certain temperature range during reoxidation before it converts into the oxide. Interestingly, in this study no stable intermediate phase was found. Instead, a sharp mass increase peaking at 850°C was observed. XRD proved the reoxidized powder to be identical to the precursor oxide Ta2O5. The amount of nitrogen in β-TaON was estimated from the weight gain during the reoxidation. The calculated nitrogen content representing the ratio of nitrogen to the sum of oxygen and nitrogen is given in Table 2. It increased with increasing ammonolysis temperature indicating an enhanced nitridation at elevated temperatures excluding the powder synthesized at 800°C.
Figure 5 shows UV-visible diffuse reflectance spectra of the oxynitrides and the precursor oxide. The absorption edge wavelength of the precursor oxide Ta2O5 is in the UV region at 310 nm (bandgap energy eV). Thermal ammonolysis causes a redshift of the absorption edge into the visible light region. With increasing temperature the absorption edge shifted to longer wavelengths resulting in bandgap energies of 2.6 to 2.4 eV which differs only marginally from previous reports [15, 20]. This finding is also in good agreement with the reoxidation study which showed that the nitrogen content increases with increasing ammonolysis temperature. Consequently the color of the β-TaON powders gradually changes from green yellowish to dark green. Furthermore, the background absorption at wavelengths above the absorption edge increases with increasing temperature, with the exception of β-TaON synthesized at 850°C. This indicates temperature-dependent generation of defects in β-TaON due to reduced tantalum and/or anionic vacancies [18, 37, 42]. However, it is not yet clear why β-TaON synthesized at 850°C has a higher background level than the samples synthesized at 900 and 950°C.
The oxygen evolution rate as a function of the ammonolysis temperature is shown in Figure 6 and Table 2. The highest rate of 220 μmol·g−1·h−1 was obtained for β-TaON synthesized at 800°C. The reference substance TiO2 (Degussa, P-25) showed a rate of 70 μmol·g−1·h−1 under the same experimental conditions. With increasing ammonolysis temperature the oxygen evolution rate of β-TaON drastically decreased to negligible values. Unfortunately the photocatalytic activity of β-TaON presented in this study cannot be compared directly to previous findings [15, 31], because they have not been performed exactly in same conditions.
Important factors influencing the photo- and catalytic activity of β-TaON are given in Table 2: crystallite size, surface area, nitrogen content, bandgap, bandgap position, and oxidizing strength of the valence band are the determining factors for oxygen evolution rate. The data in Table 2 exhibit general trends: with increasing ammonolysis temperature crystallite and particle sizes increase, while the surface area of the β-TaON powders decreases due to sintering processes. Furthermore, the increasing nitrogen content is correlated with a lowering of the bandgap energy, as expected. Chun et al.  have reported that the top of valence band in TaON is far lower than the water oxidation potential, indicating that the valence band position does not largely affect the photocatalytic water oxidation activity. In general, the water oxidation activity of β-TaON is governed by all these factors. The drastic loss of the catalytic activity at higher temperatures shows that factors like the increasing crystallite size and the bandgap shift into the red by increased N incorporation do not have a major influence in our case, since these effects are believed to increase the activity rather than decreasing it. Obviously, factors known for leading to a reduction of the catalytic activity dominate in our case. These variables are the reduced surface area and the increased defect concentration with increasing temperature. The highest oxygen evolution rate with 220 μmol·g−1·h−1, however, was obtained with β-TaON synthesized at 800°C followed closely by β-TaON synthesized at 850°C (200 μmol·g−1·h−1) featuring the largest surface area among all samples with 7.5 m2/g. Large surface areas are generally known to have a positive influence on the photocatalytic activity, since the number of active sites scale in first approximation with the surface area. Hence we explain the excellent activity observed at 800°C with the combination of high surface area and low defect concentration. At 850°C, however, the beneficial effect of a large surface area is counterbalanced by a high defect concentration leading to comparable, but slightly reduced, activity. This indicates that surface area and defect concentration are of equal importance for the catalytic activity in our system.
In conclusion, the small differences in the average crystal size and/or the bandgap do not influence significantly the photocatalytic activity of β-TaON. A possible reason could be that these quantities were not changed by the temperature variation in the synthesis condition on such a large scale that their influence on the catalytic activity would be measurable. Our experiments show that a high surface area and a minimal defect concentration improve the catalytic activity of β-TaON and can be optimized by adapting the synthesis conditions.
β-TaON powders were successfully synthesized at different temperatures by thermal ammonolysis of the precursor oxide Ta2O5. All samples were characterized structurally as well as chemically, and the associated photocatalytic activities were measured. Defect concentration and surface area were found to be the main factors to have an essential influence on the photocatalytic activities of β-TaON. An improvement by a factor of three in the oxygen evolution rate was obtained versus a reference measurement on TiO2 by tuning the synthesis condition of β-TaON.
Conflict of Interests
The authors declare no conflict of interests.
This research has been funded by the European Commission (Project NanoPEC, contract number 227179).
A. J. Bard and M. A. Fox, “Artificial photosynthesis: solar splitting of water to hydrogen and oxygen,” Accounts of Chemical Research, vol. 28, no. 3, pp. 141–145, 1995.View at: Google Scholar
A. J. Bard, G. M. Whitesides, R. N. Zare, and F. W. Mclafferty, “Holy grails in chemistry,” Accounts of Chemical Research, vol. 28, no. 3, pp. 91–91, 1995.View at: Google Scholar
B. O'Regan and M. Grätzel, “A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO2 films,” Nature, vol. 353, no. 6346, pp. 737–740, 1991.View at: Google Scholar
G. Hitoki, T. Takata, J. N. Kondo, M. Hara, H. Kobayashi, and K. Domen, “An oxynitride, TaON, as an efficient water oxidation photocatalyst under visible light irradiation (λ ≤ 500 nm),” Chemical Communications, no. 16, pp. 1698–1699, 2002.View at: Google Scholar
M. Hara, G. Hitoki, T. Takata, J. N. Kondo, H. Kobayashi, and K. Domen, “Ta3N5 and TaON as novel photocatalysts responding to visible light,” Studies in Surface Science and Catalysis, vol. 145, pp. 169–172, 2002.View at: Google Scholar
G. Brauer and J. R. Weidlein, “Synthesis and properties of tantalum oxide nitride TaON,” Angewandte Chemie International Edition, vol. 4, no. 10, p. 875, 1965.View at: Google Scholar
J. H. Swisher and M. H. Read, “Thermodynamic properties and electrical conductivity of Ta3N5 and TaON,” Metallurgical Transactions, vol. 3, no. 2, pp. 489–494, 1972.View at: Google Scholar
D. Armytage and B. E. F. Fender, “Anion ordering in TaON-powder neutron-diffraction investigation,” Acta Crystallographica Section B, vol. 30, no. 3, pp. 809–812, 1974.View at: Google Scholar
M. Weishauptand and J. Strahle, “Preparation of oxide nitrides VON, NbON, and TaON-crystal-structure of NbON and TaON,” Zeitschrift für Anorganische und Allgemeine Chemie, vol. 429, no. 1, pp. 261–269, 1977.View at: Google Scholar
S. Banerjee, S. K. Mohapatra, and M. Misra, “Synthesis of TaON nanotube arrays by sonoelectrochemical anodization followed by nitridation: a novel catalyst for photoelectrochemical hydrogen generation from water,” Chemical Communications, no. 46, pp. 7137–7139, 2009.View at: Publisher Site | Google Scholar
K. Maeda, H. Terashima, K. Kase, M. Higashi, M. Tabata, and K. Domen, “Surface modification of TaON with monoclinic ZrO2 to produce a composite photocatalyst with enhanced hydrogen evolution activity under visible light,” Bulletin of the Chemical Society of Japan, vol. 81, no. 8, pp. 927–937, 2008.View at: Publisher Site | Google Scholar
Le Bail, Duroy, and Fourquet, “Ab-initio structure determination of LiSbWO6 by X-ray powder diffraction,” Materials Research Bulletin, vol. 23, no. 3, pp. 447–452, 1988.View at: Google Scholar
J. Rodríguez-Carvajal, “Recent advances in magnetic structure determination by neutron powder diffraction,” Physica B, vol. 192, no. 1-2, pp. 55–69, 1993.View at: Google Scholar
P. Scherrer, “Inteferenzen an regellos orientierten Teilchen im Rontgenlicht,” Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen, Mathematisch-Physikalische Klasse, vol. 2, pp. 98–100, 1918.View at: Google Scholar
P. Thompson, D. E. Cox, and J. B. Hastings, “Rietveld refinement of Debye-Scherrer synchrotron X-ray data from Al2O3,” Journal of Applied Crystallography, vol. 20, pp. 79–83, 1987.View at: Google Scholar
Le Gendre, Marchand, and Laurent, “A new class of inorganic compounds containing dinitrogen-metal bonds,” Journal of the European Ceramic Society, vol. 17, no. 15-16, pp. 1813–1818, 1997.View at: Google Scholar