A unique photo-activation mechanism by “in situ doping” for photo-assisted selective NO reduction with ammonia over TiO₂ and photooxidation of alcohols over Nb₂O₅

Abstract

This paper reviews the recent development of the photocatalytic emission control reaction and photooxidation with molecular oxygen, specifically focusing on efforts based on revealing the reaction mechanism by the authors' group. TiO₂ acts as an effective catalyst for the photo-assisted selective catalytic reduction of NO with NH₃ in the presence of O₂ (photo-SCR). Photooxidation of alcohols to carbonyl compounds proceeds selectively over Nb₂O₅ without organic solvents. Usually, both TiO₂ and Nb₂O₅ work only in the ultraviolet (UV) region because of the limit of their bandgap energies. However, both photo-SCR over TiO₂ and photooxidation of alcohols over Nb₂O₅ proceed even under visible light irradiation up to ca. 450 nm. This indicates that these two reactions take place by the different photo-activation mechanism from the classical electron transfer mechanism in semiconductor photocatalysis, that is, the formation of an excited electron in the conduction band and the positive hole in the valence band. A mechanistic study using UV-Vis, ESR, FT/IR, kinetic study, and DFT calculations revealed the reaction mechanisms of photo-SCR and photooxidation of alcohols, and that the surface complex consisting of the adsorbed molecule and catalyst plays an important role in the photo-activation step. The surface complex is converted to the photo-activated species even under visible light irradiation, because the direct electron transition from a donor level derived from the adsorbed molecule to the conduction band of a photocatalyst takes place and a photo-generated hole is trapped on the adsorbed molecule to form the photo-activated radical species. The effective wavelength is shifted to a longer wavelength by the formation of the donor level derived from the adsorbed molecule during a chemical reaction (called here “in situ doping”). This unique photo-activation mechanism by “in situ doping” gives us attractive ways for removing the limit of bandgap energy, and the utilization of visible light.

---

1. Introduction

As shown in Scheme 1, photocatalytic reactions on a semiconductor powder involve several steps. Photocatalysis is generally explained in terms of band theory (the classical electron transfer mechanism) accompanied by the interaction of reactants with the photo-generated electrons and holes, and is potentially available to make the catalytic reactions proceeding at low temperature. The band structure of the photocatalysts determines the utilizable light energy, oxidizability, and reducing ability. It has been considered that suppressing recombination between photo-generated electrons and holes in a photocatalyst is important to achieving the reaction, since the lifetime of the charge separation contributes to the photocatalytic activity. Therefore, a number of studies are related to the control of band structure and charge separation. On the other hand, little information about the adsorbed species and the intermediates in photocatalytic reactions is available. Photocatalytic reactions take place on the surface of the photocatalysts as well as the ordinary catalysts. The difference between the photocatalysts and the ordinary catalysts is just the driving force to activate the adsorbed reactants; the photocatalysts use the photo-energy and the ordinary catalysts use the thermal energy. Therefore, the kinetic interpretation and the knowledge of surface structure, surface property, and surface species during the photo-reaction are required to understand the photocatalysis; generally a catalytic reaction consists of several elementary steps and one or two of the elementary steps involves absorption of light in the case of a photocatalytic reaction. Hence, there is the same thermodynamic restriction in the photocatalysis with the ordinary catalysis. Evidently, it is necessary to consider the photo-activation mechanism in detail. The clarification of the reaction mechanism provides the beneficial information on the further improvement of the photocatalysis and a new insight of photocatalytic chemistry.

Scheme 1 Model of reaction, charge separation, and recombination over photocatalyst.

This article summarizes our recent work on the photo-activation mechanisms of NH₃ over TiO₂ in photo-SCR, and alcohols over Nb₂O₅ in the photooxidation with molecular oxygen. We show a detailed investigation of reaction mechanisms of photo-SCR over TiO₂ and selective photooxidation of alcohols over Nb₂O₅ using in situ characterization, kinetic study and DFT calculations. A redshift of effective wavelength due to the direct electron transition from the donor level derived from the adsorbed molecule to the conduction band (“in situ doping”) is demonstrated.

2. Photo-assisted selective reduction of NO with NH₃ over TiO₂ based catalysts

2.1 Low-temperature NH₃-SCR systems

The growth of the global economy involved the environmental problems such as solid, air and water pollutions. These pollutions cause serious damage to human and nature. In order to solve these problems, many efforts are undergoing for the development of the environmental technology.

NO_x is one of the environmental pollutants and causes acid rain and photochemical smog. Therefore, it is desirable to remove NO_x (de-NO_x) from the stationary emission source and the mobile emission source. In the stationary emission source such as a thermal power station, an industrial boiler and a waste incinerator, NO_x is conventionally removed from the exhaust gas by the selective catalytic reduction system with NH₃ as a reductant (NH₃-SCR) in the presence of the excess O₂ over V₂O₅–WO₃ (or V₂O₅–MoO₃)/TiO₂ catalyst.^1–4 This technology was invented by three Japanese corporations (Hitachi Ltd., Babcock-Hitachi K. K., Mitsubishi Petrochemical Corp.).⁵ The reaction stoichiometry in the typical NH₃-SCR is shown as follows;

4NO + 4NH₃ + O₂ → 4N₂ + 6H₂O (1)

N₂ is formed by the reaction of NO with NH₃.^6–11N₂O is a by-product. This system shows high NO conversion (99%), high N₂ selectivity (>90%) and resistance for H₂O and SO_x, although the catalyst requires high operating temperature (573–673 K).¹² Since the exhaust gas contains various pollutants and materials such as SO_x, halogen compounds, particulate matter (PM) and fly ash in addition to NO_x, the NH₃-SCR system is used together with de-SO_x, de-halogen and dust collection systems. Thus, the NH₃-SCR system is often located downstream of the de-SO_x, de-halogen and dust collection systems in order to inhibit deactivation of V₂O₅–WO₃ (or V₂O₅–MoO₃)/TiO₂ catalyst. In this case, the inlet temperature of the exhaust gas in the NH₃-SCR system falls below 453 K. Consequently, it is necessary to re-heat the catalysis bed and the gas up to the operating temperature of the catalyst. Therefore, it is desired to develop a new de-NO_x system working at low temperature (<453 K). Since the 1990s, the low-temperature NH₃-SCR has been investigated to develop the new NH₃-SCR system capable of operating under 453 K. The following are requested features of the low-temperature NH₃-SCR system: (1) NO_x conversion must be high (more than 90%), (2) N₂ selectivity must be high (more than 90%), (3) catalysts must have high durability to H₂O, and (4) the NH₃-SCR system must be operated in the presence of an excess O₂. Since H₂O vapor gets mixed with the exhaust gas in the de-SO_x and the de-halogen processes, the catalysts must have high durability to H₂O. The low-temperature NH₃ SCR systems reported are listed in Table 1.^13–19 The first report of the NH₃-SCR is the MnO₂–Al₂O₃ catalyst reported by Singoredjo et al.¹⁴ The MnO₂–Al₂O₃ catalyst had poor durability to H₂O and the activity decreased with the course of the reaction time.

Table 1 Low-temperature SCR reaction systems

Catalyst	Reaction gas composition				Reaction conditions		Activity		Ref.
	NO (ppm)	NH3 (ppm)	O2 (ppm)	H2O (%)	T/K	SV/h^−1	X NO (%)	S N2 (%)
T: reaction temperature; SV: space velocity; XNO: NO conversion; SN2: N2 selectivity.
MnO2–Al2O3	550	550	2	0	435	31 000	98	92	11
MnO2–carbonized silica-alumina	800	800	3	0	413	12 000	94	91	13
MnO2–NaY	1000	1000	10	7	443	48 000	88	93
MnO2–TiO2	2000	2000	2	0	393	8000	100	100	17
MnO2–TiO2	400	400	2	11	448	50 000	98	97
Fe–Mn oxide	1000	1000	2	0	393	15 000	100	100	19
Fe–Mn oxide	1000	1000	2	2.5	413	15 000	98	100
MnO2–CeO2	1000	1000	2	0	393	42 000	99	100	16
MnO2–CeO2	1000	1000	2	19	393	42 000	95	100
V2O5-sulfated carbon	500	600	3	0	453	34 000	92	100	15

---

Subsequently, several Mn-containing catalysts, which show the high activity with the durability to H₂O at a low-temperature, are reported.^{13–15,17,18} Among the Mn-containing catalysts reported, MnO₂–CeO₂ shows the highest activity (95% NO conversion and 100% N₂ selectivity in the presence of 19% H₂O).¹⁷

2.2 Photo-SCR with NH₃ over TiO₂ photocatalysts

Over the past decades, the most attentions of the photocatalysis have been gathered to the water photolysis and the removal of harmful compounds over photocatalysts responding to visible light. In contrast, although photocatalysts have the advantage that the re-heating of the catalyst bed is unnecessary because of their possibility of application at low temperatures, the reports of the NH₃-SCR reaction over photocatalysts (photo-SCR) were limited. Cant and Cole reported that NO reduction by NH₃ to form N₂ proceeds over TiO₂ under photo-irradiation.²⁰ The reaction was carried out in a closed system and the activity was very low. In contrast, we found that TiO₂ and metal oxide promoted TiO₂ are effective for the photo-SCR with NH₃ in the presence of O₂ proceeding at room temperature.^21–29

Table 2 shows the catalytic activity and physicochemical property of photo-SCR with NH₃ over various TiO₂ photocatalysts (JRC-TIO-1-13 supplied from Catalysis Society of Japan). JRC-TIO-11, a mixture of rutile and anatase phases, exhibited the highest activity among all TiO₂ tested. JRC-TIO-8 and JRC-TIO-3 were the most active catalysts consisting of anatase or rutile single phases. There is poor correlation between the activity and crystal phase, crystallite size, and specific surface area. This indicates that these properties are not the important factors to determine the activity of photo-SCR.

Table 2 Activities of photo-SCR reaction with NH₃ over various TiO₂ photocatalysts

Catalysts	Sa/m^2 g^−1	Phase	D c/Å	NO conv. (%)	N2 sel. (%)
A: anatase, R: rutile. Reaction conditions; NO: 1000 ppm, NH3: 1000 ppm, O2: 2%, Ar balance, GHSV = 32 000 h^−1.
JRC-TIO-1	71.1	A	187	36	100
JRC-TIO-2	15.6	A	535	14.5	100
JRC-TIO-3	45.6	R	219	53	100
JRC-TIO-4	47.8	R 29.4%	382	35.5	100
		A 70.6%	259
JRC-TIO-5	3–4	R 92.4%	2200	31	100
		A 7.6%	1000
JRC-TIO-6	58.0	R	240	20	100
JRC-TIO-7	108	A	197	35	100
JRC-TIO-8	93.2	A	155	51.7	98.6
JRC-TIO-9	95.2	A	197	31	100
JRC-TIO-10	100	A	169	35.5	100
JRC-TIO-11	76.6	R 8.7%	200	63	100
		A 91.3%	153
JRC-TIO-12	98.7	A	159	41	100
JRC-TIO-13	71.1	A	237	33	100

---

2.3 Mechanism of photo-SCR with NH₃ over TiO₂ photocatalysts

Fig. 1 shows the EPR spectra of TiO₂. After evacuation at 673 K, the signals are derived from the Ti³⁺ species (Fig. 1(a)).^30–33 There is little change in the EPR signal by the exposure of NH₃ to TiO₂ in the dark. On the other hand, the EPR signal changed drastically after photo-irradiation. New signals assignable to the NH₂ radical^34–38 were detected together with signals assigned to Ti³⁺. These new signals were quite stable even after more than 1 hour at 123 K without photo-irradiation. However, these signals immediately vanished by the exposure to NO in the dark whereas the intensity of signals due to Ti³⁺ species increased. This suggests that (1) the photo-generated electron is trapped on Ti⁴⁺ to form Ti³⁺ and the positive hole is captured by adsorbed NH₃ species and is converted to an active NH₂ radical, and that (2) NO in the gas phase attacks the NH₂ radical on TiO₂ rapidly. As both the NH₂ radical and NO are doublet state species, it follows that the NH₂ radical reacts with NO easily without irradiation. Moreover, the formation of a NH₂NO intermediate was confirmed by FTIR spectroscopy after admittance of NO to TiO₂ adsorbing NH₃ under photo-irradiation (vide infra). As described above, the signals due to Ti³⁺ species increased in intensity after the introduction of NO. It seems that the electron transfer took place from the N atom of adsorbed NH₃ to the Ti atom of TiO₂ bulk. In other words, the photo-generated electron was trapped on the Ti atom and the photo-generated hole was captured by the NH₂⁻ species derived from adsorbed NH₃. As a result, the NH₂⁻ species is converted to the active NH₂ radical. On the other hand, the electron may move into the inside of TiO₂ bulk as a stable free electron. Before the exposure to NO, recombination took place between a part of Ti³⁺ species and the NH₂ radical. On the other hand, after the exposure to NO, the electron could not recombine because of losing an opponent (NH₂ radical). The electron was localized and stabilized in the inside of TiO₂, and the signals assigned to the Ti³⁺ species increased in intensity.

Fig. 1 EPR spectra of TiO₂ (a) after pretreatment, (b) after introduction of NH₃ in the dark, (c) under photo irradiation and (d) after introduction of NO in the dark.

Fig. 2 shows the time course of the N₂ evolution rate of photo-SCR. NO conversion and N₂ selectivity attained 100% and 96%, respectively, in the conventional fixed bed flow system (GHSV = 8000 h⁻¹). The N₂ evolution rate gradually increased at the initial stage and reached a steady rate at 1.5 h. However, when the reaction gas (a mixture of NO/NH₃/O₂) was passed in the dark for 0.5 h, and then photo-irradiation was started, the N₂ evolution rate immediately jumped to the level of the steady rate.^23,25 This clearly indicates that the induction period shown in Fig. 2 is the time for saturation of the adsorption equilibrium of the reactant molecule. When NH₃ were passed for 1.5 h in the dark, then the gas was switched to a mixture of NO/O₂ and the photo-irradiation was started, N₂ was evolved. The N₂ evolution rate gradually decreased and the total amount of evolved N₂ was consistent with that of equilibrium adsorption of NH₃ on TiO₂. In contrast, when a mixture of NO/O₂ was firstly passed and then switched to NH₃, neither N₂ nor N₂O was formed. These results suggest that NH₃ species adsorbed on the Lewis acid site is excited by photo-irradiation and reacts with NO in the gas phase to produce N₂. Furthermore, this is supported by the fact that only ¹⁵N¹⁴N was evolved in the photo-SCR of ¹⁵NO with ¹⁴NH₃ in the presence of O₂.²²

Fig. 2 Time course of N₂ (circle) and N₂O (square) in the photo-SCR with NH₃ over TiO₂ JRC-TIO-11, reaction conditions; GHSV = 8000 h⁻¹, NO: 1000 ppm, NH₃: 1000 ppm, O₂: 2%, Ar balance.

The adsorbed species and intermediates of photo-SCR were identified by in situ FT/IR spectra (Fig. 3). After NH₃ adsorbed on TiO₂, the bands (1136, 1215, and 1599 cm⁻¹) due to adsorbed NH₃ species on the Lewis acid site of TiO₂ appeared.^39–41 The bands at 1599 and 1215 cm⁻¹ retained their intensity after evacuation (Fig. 3(b)) and exposure to NO in the dark (Fig. 3(c)). The bands due to adsorbed NH₃ decreased gradually in intensity with irradiation time. On the other hand, the band at 1624 cm⁻¹, which is assignable to the deformation vibration of H₂O,⁴² grew. Furthermore, new bands between 1400 and 1600 cm⁻¹ were observed and then disappeared. These new bands are assigned to the nitrosamide species (NH₂NO) by comparing the FT/IR spectrum of TiO₂ exposed to ¹⁴NO and NH₃ to that exposed to ¹⁵NO and NH₃.²⁴

Fig. 3 FT-IR spectra of adsorbed species on TiO₂ in the photo-SCR with NH₃. (a) After introduction of NH₃, (b) after evacuation, (c) after introduction of NO in the dark, (d) under photo irradiation for 10 min, (e) for 30 min, (f) for 60 min, and (g) for 120 min.

These results indicate that the intermediate of photo-SCR is the nitrosamide species (NH₂NO) and the nitrosamide species is decomposed to N₂ and H₂O. Moreover, it was confirmed that the Ti³⁺ species of TiO₂ reduced by H₂ was re-oxidized to the Ti⁴⁺ species by exposure to O₂ easily even at room temperature using UV-Vis spectroscopy.²⁷ On the basis of these results, we proposed the Eley–Rideal type mechanism as follows (Scheme 2);^24–29 (1) the NH₃ adsorbs on the Lewis acid site of TiO₂, (2) the adsorbed NH₃ species is excited by photo-irradiation, (3) the excited species (NH₂ radical) reacts with NO in the gas phase to form the nitrosamide species (NH₂NO), (4) the nitrosamide species is decomposed to N₂ and H₂O, and (5) the Ti³⁺ site is re-oxidized by molecular oxygen.

Scheme 2 Reaction mechanism of photo-SCR with NH₃ over TiO₂.

According to the dependencies of the partial pressure of NO, NH₃, and O₂, the reaction rate (r) of photo-SCR is expressed as [eqn (2)].

Here, the rate constant, the partial pressure of NH₃, NO, and O₂ are abbreviated to k, P_NH3, P_NO, and P_O2, respectively. This equation indicates that NH₃ adsorbs strongly on TiO₂ (Step 1) and that re-oxidation of Ti³⁺ to Ti⁴⁺ (Step 5) proceeds rapidly. By comparing the obtained rate equation [eqn (2)] with the rate equation derived from the proposed reaction mechanism by steady-state approximation, we concluded that Step 4, the decomposition of the nitrosamide species, is the rate-determining step of photo-SCR.

2.4 Enhancement of activity by improving the lifetime of photo-activated species

An efficient charge separation promotes the chemical reactions competing with a recombination of the photo-generated electrons and holes. Einaga et al. reported that benzene as a model VOC (volatile organic compound) can be abated by total oxidation over TiO₂ photocatalysts effectively.^43–45 However, the specific activity of TiO₂ is absolutely low. They reported that 120 ppm of benzene cannot be removed over TiO₂, but 80 ppm of benzene is efficiently decomposed to CO₂ and CO in the presence of H₂O vapor. This would be caused by the insufficient lifetime of the charge-separated state over TiO₂. It is widely thought that the photo-generated electrons and holes are consumed by recombination much more rapidly than by the photocatalytic reaction and the recombination is the main reason of too short lifetime of the charge-separated state and resulting in low activity of TiO₂. Indeed, the half-life of the charge-separated state of TiO₂ was estimated to be below 100 ps.^46,47 Up to now, the limit of the extended lifetime of the charge-separated state is only several tens of ns despite the careful effort. Therefore, in the case of photo-SCR over TiO₂, it seems that the half-life of NH₂ radicals, which are formed by capturing photo-generated positive holes, is below 100 ps and that the rate-determining step of photo-SCR is the reaction of the NH₂ radical with NO to form the nitrosamide species (NH₂NO). However, as described above, the kinetic study indicated that the decomposition of the nitrosamide species (Step 4) is the rate-determining step of photo-SCR. This strongly suggests that the strongly adsorbed NH₃ on the Lewis acid site of TiO₂ lengthened the lifetime of the charge-separated state by trapping holes and consequently the recombination of the photo-generated electrons and holes is inhibited. Fig. 4 shows the decay curve of the NH₂ radical signal recorded by EPR after irradiation stopped. All decay curves can be approximated to a hyperbolic curve, indicating that the NH₂ radical is quenched by the secondary reaction between the NH₂ radical and an electron. The half-life of the NH₂ radical at reaction temperature of photo-SCR (323 K) is calculated to be 1.4 min by using the Arrhenius equation (Fig. 5). Since this half-life of the NH₂ radical at 323 K is much longer than that of the charge-separated state of TiO₂ (<100 ps), the concentration of the NH₂ radical on the surface of TiO₂ increased and the activity of photo-SCR is enhanced. To the best of our knowledge, the intermediate and active species having a longer lifetime such as the NH₂ radical has not been reported.

Fig. 4 Decay curve of NH₂ radical and the approximated curve (lines) at (a) 113 K, (b) 123 K, (c) 133 K and (d) 143 K.

Fig. 5 Arrhenius plots of the half-lives of NH₂ radical at each temperature (dot) and the approximated line (line).

2.5 Mechanism of formation of the NH₂ radical over TiO₂

Fig. 6 shows the apparent quantum efficiency of the photo-SCR over TiO₂ as a function of the incident light (action spectrum) and a UV-Vis spectrum of TiO₂. The band gap of this TiO₂ is estimated to be 3.28 eV (photo-excitation energy is 385 nm). The action spectrum is in good agreement with the UV-Vis spectrum of TiO₂ in the region of wavelength <385 nm. Although TiO₂ is unable to absorb light at wavelengths >385 nm, photo-SCR proceeded under irradiation up to ca. 450 nm. The feature of this action spectrum is similar to the UV-Vis spectrum of N-doped TiO₂.⁴⁸ In order to reveal whether a new energy level derived from the adsorbed molecule is located between the HOMO and LUMO levels or not, density functional theory (DFT) calculations were employed. DFT calculations revealed that the N 2p electron donor level is located between O 2p and Ti 3d when NH₂ species are formed on a TiO₂ surface by the dissociative adsorption of NH₃.²⁷

Fig. 6 Action spectrum of photo-SCR (dot) and UV-Vis spectrum of JRC-TIO-11 (line); reaction conditions of action spectrum: NH₃: 1000 ppm, NO: 1000 ppm, O₂: 2%, flow rate: 100 ml min⁻¹.

On the basis of these results, we conclude that the photo-activation of NH₃ adsorbed on TiO₂ to an NH₂ radical occurs through two paths as shown in Fig. 7. One is the electron transition from the valence band consisting of O 2p orbitals to the conduction band consisting of Ti 3d orbitals of TiO₂ under UV irradiation. The other is the direct electron transfer from N 2p of adsorbed NH₃ to Ti 3d. This N 2p electron donor level formed between O 2p and Ti 3d enables the photo-SCR to proceed even under visible light irradiation (400–450 nm). It can be thought that the expansion of the effective wavelength of TiO₂ by adsorption of NH₃, called here, “in situ doping”, is one of the factors for high activity of TiO₂ in the photo-SCR.

Fig. 7 Formation mechanism of NH₂ radical over TiO₂.

Many researchers reported that the “pre-doped” or “pre-modified” photocatalysts with metals or ions such as N-doped TiO₂ adsorb the visible light.^48–51 However, there has been no report that the effective wavelength of photo-reaction is shifted to a longer wavelength by the formation of the donor level derived from the adsorbed molecule on the catalyst during a chemical reaction (“in situ doping”). Moreover, “in situ doping” was found not only in photo-SCR with NH₃ over TiO₂ but also in the photooxidation of alcohols over Nb₂O₅. The details of the photooxidation of alcohols over Nb₂O₅ are described in the next section.

3. Photooxidation of alcohols with molecular oxygen over Nb₂O₅ catalysts

3.1 Oxidation of alcohol with molecular oxygen

Catalytic alcohol oxidation to carbonyl compounds is one of the most important chemical transformations used in the industrial chemistry and in organic syntheses.^52–54 Non-catalytic methods with stoichiometric, toxic, corrosive and expensive oxidants such as ClO⁻, dichromate, permanganate, and peroxy acids under stringent conditions of high pressure and/or temperature have been widely used for alcohol oxidations.^52–55 In addition, these reactions are often carried out with high concentration of bases and environmentally unfriendly organic solvents. Therefore, much attention has been paid to the development of heterogeneous catalytic systems that use clean and atom efficient oxidants like molecular oxygen or H₂O₂ without organic solvents.^55–66

Recently, the aerobic oxidation of alcohols was successfully carried out by using heterogeneous catalysts such as tetrapropylammonium perruthenate (TPAP)/MCM-41,⁵⁹Ru/CeO₂,⁶⁰Ru–hydrotalcite,⁶¹Ru/hydroxyapatite (Ru–HAP),⁶²[RuCl₂(p-cymene)]₂/activated carbon,⁶³Ru/Al₂O₃,⁶⁴Pd–hydrotalcite which requires the addition of pyridine,⁶⁵ and Pd or Pt on activated carbon.^65,66 These systems require the use of organic solvents. Wu et al. reported on solvent-free aerobic oxidation of alcohols by Pd/Al₂O_3.⁶⁷ However, the use of the noble metal, Pd, is an essential requirement. Despite the advantage of using heterogeneous catalysts without organic solvents or additives for oxidation of alcohols, few reports have appeared on the use of highly active solvent-free heterogeneous catalysts with only molecular oxygen as an oxidant.

In this respect, photoreactions are promising processes and the development of photocatalysts is a subject that is now receiving noticeable attention. TiO₂ has been identified as one example of a practical and useful photocatalysts,^68–71 and widely used in degradation of organic pollutants in air and water. However, in the most part of these reports, TiO₂ is used in vapor phase oxidations at high temperature,⁶⁷oxidation of only lower alcohols,^69,70oxidation using solvents such as benzene⁷¹ and a low selectivity to partial oxidized products due to excess photo-activation of target products which leads to deep oxidation. Zhao et al.^72,73 reported that the photooxidation of alcohols on TiO₂ could be dramatically accelerated without any loss of selectivity by adsorption of Brønsted acid and this effect by Brønsted acid results from the decomposition of the relatively stable side-on peroxide promoted by the protons, which effectively clean the catalytic Ti–OH₂ sites. However, this system requires the use of benzotrifluoride as a solvent.

Recently, we found that the photooxidation of alcohols to carbonyl compounds proceeded selectively at low temperature over Nb₂O₅ without organic solvents or any additives (Table 3).^74–76 Various metal oxides (SiO₂, MgO, Al₂O₃, ZrO₂, V₂O₅, Ta₂O₅, MoO₃, and WO₃) showed no activity and the activity of ZnO was very low. TiO₂ showed higher activity than Nb₂O₅, however, the Nb₂O₅ catalyst showed higher selectivity than TiO₂ at the same conversion level.⁷⁴Nb₂O₅ is suitable for selective oxidation. The photooxidation did not take place in the dark. Autooxidation proceeded when 1-phenylethanol, cyclohexanol and benzylalcohol were irradiated without the catalyst. This was due to the formation of radical species by the photo-decomposition of carbonyl compounds (Norrish Type I reaction) which were present as impurities in the alcohols (entries 1 to 3). The Nb₂O₅ catalyst improved the conversions and/or selectivities to carbonyl compounds greatly. The less reactive primary alcohol, 1-pentanol was also photooxidized over the Nb₂O₅ catalyst. The Nb₂O₅ catalyst was reusable and showed the same conversion and selectivity without any pretreatment as the catalyst as prepared.

Table 3 Photooxidation of various alcohols over Nb₂O₅ with molecular oxygen^a,b

Entry	Substrate	Product	T/h	Conv. (%)	Sel. (%)
a Reaction conditions were as follows: alcohol (10 mL), Nb2O5 (100 mg), 323 K, under 0.1 MPa of O2, O2 flow rate (2 cm^3 min^−1): conversion and selectivity were determined by gas chromatography with an internal standard. b Figures in parentheses are the results of photochemical reaction without catalysts.
1			240 (72)	99 (14)	96 (69)
2			168 (96)	76 (46)	64 (36)
3			72 (72)	67 (79)	90 (43)
4			168 (119)	23 (2)	85 (82)
5			192 (121)	18 (5)	83 (81)
6			84 (24)	14 (0)	92 (−)

---

3.2 Mechanism of photooxidation of alcohol over Nb₂O₅

The adsorbed species and intermediates of photooxidation were identified by in situ FT/IR spectra of adsorbed cyclohexanol on Nb₂O₅. Fig. 8 shows the FT/IR spectra of adsorbed cyclohexanol on Nb₂O₅. The bands at 1467 and 1452 cm⁻¹ were assigned to δ_s(CH₂) and the bands at 1363 and 1347 cm⁻¹ were assigned to ω(CH₂), respectively. The new bands at 1091 and 1126 cm⁻¹ appeared after the adsorption of cyclohexanol on Nb₂O₅. Therefore, these bands are assigned to the stretching mode of a C–O bond in the alcoholate species on the Nb₂O₅, because the formation of the alcoholate species by the adsorption of alcohol is usually accompanied by a shift of the stretching mode of a C–O bond to a higher wavenumber.^77–79 The change in FT/IR spectra by UV irradiation (<390 nm) was shown in Fig. 9. The intensity of the band assigned to ν(C–O) (around 1090 cm⁻¹) decreased as the irradiation time increased, whereas the bands assigned to ν(C = O) (1676 cm⁻¹) and the symmetric-stretching of the carboxylic acid anion (1554 cm⁻¹) gradually grew. This result indicates that the alcoholate species on Nb₂O₅ was excited by photons and oxidized to carbonyl compounds. Interestingly, the carbonyl compounds were formed even under visible light irradiation (>390 nm).

Fig. 8 FT-IR spectra of adsorbed cyclohexanol on Nb₂O₅. (a) Cyclohexanol was exposed to Nb₂O₅ for 1 h (physisorption + chemisorption), (b) evacuated for 2 h (chemisorption), (c) difference spectrum ((a)–(b): physisorption). Nb₂O₅ was evacuated at 773 K for 1 h and oxidized at 773 K with 10.7 kPa of O₂ and then evacuated at 773 K for 1 h before FT-IR measurements.

Fig. 9 FT-IR spectra of adsorbed species on Nb₂O₅ in the photo-reaction of adsorbed cyclohexanol with O₂. (a) Cyclohexanol was exposed to Nb₂O₅ for 1 h and evacuated for 2 h, (b) under UV irradiation for 1, (c) 5, (d) 7, (e) 10, (f) 15 and (g) 30 min. Nb₂O₅ was evacuated at 773 K for 1 h and oxidized at 773 K with 10.7 kPa of O₂ and then evacuated at 773 K for 1 h before FT-IR measurements.

Fig. 10 shows the EPR spectra of Nb₂O₅. A broad EPR signal around g = 1.9 was observed at 123 K (Fig. 9(c)), when 1-pentanol was adsorbed on Nb₂O₅ under UV-irradiation. This broad signal at g = 1.9 was assignable to Nb^4+80,81 and immediately disappeared by the exposure to O₂ in the dark, indicating that Nb⁴⁺ was oxidized to Nb⁵⁺ rapidly even at 123 K. On the other hand, when 1-pentanol was adsorbed on Nb₂O₅ under UV-irradiation at 77 K, the EPR signal (g = 2.006, A_H1 = 2.0 mT, A_H2 = 4.4 mT) assigned to alkenyl radical species was observed (Fig. 11). These new signals were stable at 77 K without photo-irradiation, but disappeared at room temperature. The signal was restored by UV-irradiation at 77 K. Moreover, the signal did not change in the presence of O₂ even under UV-irradiation (Fig. 11(f)).

Fig. 10 ESR spectra of Nb₂O₅ recorded at 123 K. (a) After pretreatment, (b) in the dark in the presence of 1-pentanol, (c) under irradiation for 5 h in the presence of an excess of 1-pentanol, (d) after introduction of O₂. Nb₂O₅ was evacuated at 773 K for 1 h and oxidized at 773 K with 10.7 kPa of O₂ and then evacuated at 773 K for 1 h before ESR measurements.

Fig. 11 ESR spectra of Nb₂O₅ recorded at 77 K. (a) After pretreatment, (b) under irradiation, (c) in the dark in the presence of 1-pentanol, (d) under irradiation in the presence of 1-pentanol, (e) in the dark after the sample was heated up to RT and then cooled to 77 K, (f) under re-irradiation. Nb₂O₅ was evacuated at 773 K for 1 h and oxidized at 773 K with 10.7 kPa of O₂ and then evacuated at 773 K for 1 h before ESR measurements, (g) alkenyl radical.

This indicates that the alkenyl radical species does not react with O₂. Therefore, it suggests that (1) the photo-formed electron is trapped on Nb⁵⁺ to form Nb⁴⁺ and the positive hole is captured by alcoholate species and is converted to an active alkenyl radical, (2) the active alkenyl radical is dehydrogenated to a carbonyl compound (the reduction of Nb⁵⁺ to Nb⁴⁺ takes place simultaneously), and (3) O₂ in the gas phase re-oxidizes Nb⁴⁺ to Nb⁵⁺.

On the basis of these results, we proposed the reaction mechanism as shown in Scheme 3;^75–77 (1) alcohol is adsorbed on Nb₂O₅ as alcoholate species, (2) alcoholate adsorbed on Nb₂O₅ is activated by transferring an electron to the conduction band reducing Nb⁵⁺ to Nb⁴⁺ and leaving a hole on alcoholate, (3) the formed alkenyl radical is converted to a carbonyl compound, (4) the product desorbs, and (5) the reduced Nb⁴⁺ sites are re-oxidized by the reaction with O₂.

Scheme 3 Reaction mechanism of alcohol photooxidation with molecular oxygen over Nb₂O₅.

In this mechanism, oxygen anion radical species (O₂⁻ and O₃⁻), which are formed by irradiation over TiO₂^80–87 and often responsible for total oxidation, do not contribute to the photooxidation over Nb₂O₅. For instance, when Nb₂O₅ was irradiated in the presence of O₂, no EPR signal due to oxygen anion radical species was observed. This presumably explains why the photooxidation of alcohol to a carbonyl compound proceeds selectively over the Nb₂O₅ catalyst.

We carried out the photooxidation of 1-pentanol under the various concentrations of 1-pentanol, O₂ and the different light intensity to determine each reaction order. On the basis of these results, the reaction rate (r) of photooxidation of alcohol is expressed as [eqn (3)].

The rate constant, the substrate concentration, the light intensity and the pressure of the oxygen are abbreviated to k, [S], I and P_O2, respectively. By comparing the obtained rate equation [eqn (3)] with the rate equation derived from the proposed reaction mechanism by steady-state approximation, it is suggested that Step 3 or Step 4 is the rate-determining step of the photooxidation of alcohol over Nb₂O₅. The alkenyl radical species was not obtained at 123 K, whereas Nb⁴⁺ was observed. This suggests that Step 3, the conversion of the formed alkenyl radical to a carbonyl compound took place even at 123 K because of a high reactivity of the alkenyl radical species. FT/IR spectra showed that the carbonyl compounds remained on the surface of Nb₂O₅ at room temperature, indicating that the desorption of the carbonyl compound was slow. Therefore, we concluded that the rate-determining step of the photooxidation of alcohol over Nb₂O₅ is Step 4, the desorption process of the formed carbonyl compound.

3.3 Mechanism of formation of the alkenyl radical over Nb₂O₅

Fig. 12 shows the apparent quantum efficiency of photooxidation of 1-pentanol as a function of the wavelength of the incident light (action spectrum) and a UV-Vis spectrum of Nb₂O₅. Although the Nb₂O₅ catalyst is not able to absorb visible light (>390 nm), the photooxidation of 1-pentanol took place under irradiation up to ca. 480 nm. This result is consistent with the change in the FT/IR spectra under visible light irradiation and a red shift of the effective wavelength of photo-reaction is similar to that of photo-SCR over TiO₂. In the case of photo-SCR over TiO₂, we found that the adsorbed NH₃ was photo-activated by the direct electron transfer from the N 2p electron donor level formed between O 2p and Ti 3d of TiO₂ to the conduction band.²⁷

Fig. 12 Action spectrum of photooxidation of 1-pentanol (dot) and UV-Vis spectrum of Nb₂O₅ (line). Reaction conditions of the action spectrum were as follows: 1-pentanol (10 ml), Nb₂O₅ (100 mg), 323 K, under 0.1 MPa of O₂, O₂ flow rate (2 cm³ min⁻¹).

In order to investigate the formation of a new energy level derived from the adsorbed molecule, DFT calculations were employed and showed that donor levels were generated between the HOMO and LUMO levels of Nb₂O₅ by adsorbed alcohol on Nb₂O₅ and that the electron transitions from the O 2p donor level derived from the adsorbed alcoholate species to the conduction band of Nb₂O₅ (Nb 4d orbitals) had lower energy than those from O 2p of Nb₂O₅ (the conduction band) to Nb 4d.⁷⁵ On the basis of these results, we concluded that the photooxidation of alcohol over Nb₂O₅ takes place through the direct electron transfer from the O 2p orbital of adsorbed alcoholate species to the conduction band consisting of Nb 4d orbitals as shown in Fig. 13 (“in situ doping”). As a result of “in situ doping”, the photooxidation of alcohol proceeded even under visible light irradiation.

Fig. 13 Formation mechanism of alkenyl radical over Nb₂O₅.

4. Conclusions and outlook

By means of UV-Vis, ESR, FT/IR with the aid of kinetic study and DFT calculations, the detailed reaction mechanisms of photo-assisted selective reduction of NO with NH₃ (photo-SCR) over TiO₂ and photooxidation of alcohol with O₂ over Nb₂O₅ were revealed and a unique photo-activation mechanism by “in situ doping” in both photo-SCR and photooxidation of alcohol was demonstrated.

TiO₂ acts as an effective catalyst for photo-SCR even at room temperature. In this photo-SCR system, the re-heating of the catalyst bed is unnecessary because of their possibility of application at low temperatures. Thus, the photo-SCR system can miniaturize the reactor. Moreover, in this photo-SCR system, the TiO₂ photocatalyst can activate NH₃ effectively even in the presence of excess O₂. Indeed, we found that TiO₂ acts as an effective catalyst for photo-assisted selective catalytic oxidation of NH₃ (photo-SCO: 4NH₃ + 3O₂ → 2N₂ + 6H₂O).^85–87 Although further improving the activity may be needed, it seems that this photo-SCO system can be used for removing unreacted NH₃ in the SCR process and for removing NH₃ from a small and isolated source such as daily firm.

Photooxidation of alcohols to carbonyl compounds proceeds selectively over Nb₂O₅ without organic solvents. Nb₂O₅ shows higher selectivity to a partial oxidation product than that of a commonly used TiO₂ photocatalyst, and efficient conversion under a solvent-free condition.

In the case of photo-SCR over TiO₂, a new electron donor (N 2p) was located between O 2p and Ti 3d by the adsorption of NH₃ on TiO₂. The direct electron transition from N 2p to Ti 3d took place to form NH₂ radical species by visible light irradiation. As a result, the photo-SCR proceeded even under visible light irradiation. The high activity of TiO₂ was caused by the expansion of the effective wavelength of TiO₂ by adsorption of NH₃ and the long lifetime of the NH₂ radical. In the case of photooxidation of alcohol over Nb₂O₅, as well as the adsorbed NH₃ on TiO₂, the new electron donor level was generated between O 2p and Ti 3d by the adsorption of alcohol on Nb₂O₅. The direct electron transition from O 2p derived from alcohol to Ti 3d took place to form alkenyl radical species by photo-irradiation. Then, this alkenyl radical was immediately dehydrogenated to a carbonyl compound. As shown in the present review, the unique photo-activation mechanism by “in situ doping” gives us attractive ways for removing the limit of bandgap energy, and the utilization of visible light.

Acknowledgements

A part of this work was partially supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

Notes and references

1. S. Cho, Chem. Eng. Prog., 1994, 90, 39–45.
2. P. Forzatti and L. Lietti, Heterog. Chem. Rev., 1996, 3, 33–51.
3. F. Nakajima, Catalysis, 1990, 32, 236–239.
4. S. Wood, Chem. Eng. Prog., 1994, 90, 32–38.
5. A. Kato, S. Matsuda, F. Nakajima, M. Imanari and Y. Watanabe, J. Phys. Chem., 1981, 85, 1710–1713.
6. K. Otto, M. Shelef and J. T. Kummer, J. Phys. Chem., 1970, 74, 2690–2696.
7. F. Janssen, F. Van den Kerkhof, H. Bosch and J. Ross, J. Phys. Chem., 1987, 91, 5931–5934.
8. F. Janssen, F. Van den Kerkhof, H. Bosch and J. Ross, J. Phys. Chem., 1987, 91, 6633–6637.
9. B. Duffy, H. Curryhyde and N. Cant, J. Phys. Chem., 1994, 98, 7153–7161.
10. U. Ozkan, Y. Cai and M. W. Kumthekar, J. Catal., 1994, 149, 375–389.
11. U. Ozkan, Y. Cai and M. W. Kumthekar, J. Phys. Chem., 1995, 99, 2363–2371.
12. V. Parvulescu, P. Grange and B. Delmon, Catal. Today, 1998, 46, 233–316.
13. R. Long and R. Yang, J. Catal., 2002, 207, 158–165.
14. L. Singoredjo, R. Korver, F. Kapteijn and J. Moulijn, Appl. Catal., B, 1992, 1, 297–316.
15. M. Richter, A. Trunschke, U. Bentrup, K. Brzezinka, E. Schreier, M. Schneider, M. Pohl and R. Fricke, J. Catal., 2002, 206, 98–113.
16. E. Garcia-Bordeje, J. Pinilla, M. Lazaro, R. Moliner and J. Fierro, J. Catal., 2005, 233, 166–175.
17. G. Qi and R. Yang, Chem. Commun., 2003, 848–849.
18. D. Pena, B. Uphade and P. Smirniotis, J. Catal., 2004, 221, 421–431.
19. L. Gang, B. Anderson, J. van Grondelle and R. van Santen, Catal. Today, 2000, 61, 179–185.
20. N. Cant and J. R. Cole, J. Catal., 1992, 134, 317–330.
21. T. Tanaka, K. Teramura and T. Funabiki, Phys. Chem. Chem. Phys., 2000, 2, 2681–2682.
22. T. Tanaka, K. Teramura, T. Yamamoto, S. Takenaka, S. Yoshida and T. Funabiki, J. Photochem. Photobiol., A, 2002, 148, 277–281.
23. T. Tanaka, K. Teramura, K. Arakaki and T. Funabiki, Chem. Commun., 2002, 2742–2743.
24. K. Teramura, T. Tanaka and T. Funabiki, Langmuir, 2003, 19, 1209–1214.
25. K. Teramura, T. Tanaka, S. Yamazoe, K. Arakaki and T. Funabiki, Appl. Catal., B, 2004, 53, 29–36.
26. K. Teramura, T. Tanaka and T. Funabiki, Chem. Lett., 2003, 32, 1184–1185.
27. S. Yamazoe, K. Teramura, Y. Hitomi, T. Shishido and T. Tanaka, J. Phys. Chem. C, 2007, 111, 14189–14197.
28. S. Yamazoe, T. Okumura, K. Teramura and T. Tanaka, Catal. Today, 2006, 111, 266–270.
29. S. Yamazoe, Y. Masutani, Y. Hitomi, T. Shishido and T. Tanaka, Appl. Catal., B, 2008, 83, 123–130.
30. P. Meriaudreau, M. Che and C. K. Jorgensen, Chem. Phys. Lett., 1970, 5, 226–228.
31. C. Hauser and P. Naccache, Chem. Phys. Lett., 1971, 8, 45–48.
32. R. F. Howe and M. Gratzel, J. Phys. Chem., 1985, 89, 4495–4499.
33. D. Hurum, A. G. Agrios, K. A. Gray, T. Rajh and M. C. Thurnauer, J. Phys. Chem. B, 2003, 107, 4545–4549.
34. S. N. Forer, E. L. Cochran, V. A. Bowers and C. K. Jen, Phys. Rev. Lett., 1958, 1, 91–94.
35. E. F. Vansant and J. H. Lunsford, J. Phys. Chem., 1972, 76, 2716–2718.
36. O. I. Brotikovskii, G. M. Zhidomirov, V. B. Kazanskii, A. I. Mashchenko and B. N. Shelmimov, Kinet. Katal., 1971, 12, 616–619.
37. S. Nagai, Bull. Chem. Soc. Jpn., 1973, 46, 1144–1148.
38. N. Shimamoto, K. Hatano, T. Katsu and Y. Fujita, Bull. Chem. Soc. Jpn., 1975, 48, 18–21.
39. G. Ramis, G. Busca, F. Bregani and P. Forzatti, Appl. Catal., 1990, 64, 259–278.
40. M. C. Kung and H. H. Kung, Catal. Rev., 1985, 27, 425–460.
41. C. C. Chuang, J. S. Shiu and L. J. Lin, Phys. Chem. Chem. Phys., 2000, 2, 2629–2633.
42. E. Ito, Y. J. Mergler, B. E. Nieuwenhuys, H. P. A. Calis, H. vanBekkum and C. M. vandenBleck, J. Chem. Soc., Faraday Trans., 1996, 92, 1799–1806.
43. H. Einaga, S. Futamura and T. Ibusuki, J. Jpn. Pet. Inst., 1999, 42, 363–364.
44. H. Einaga, S. Futamura and T. Ibusuki, Phys. Chem. Chem. Phys., 1999, 1, 4903–4908.
45. H. Einaga, S. Futamura and T. Ibusuki, Appl. Catal., B, 2002, 38, 215–225.
46. S. Ikeda, N. Sugiyama, B. Pal, G. Macri, L. Palmisano, H. Noguchi, K. Uosaki and B. Ohtani, Phys. Chem. Chem. Phys., 2001, 3, 267–273.
47. S. Ikeda, N. Sugiyama, S. Murakai, H. Kominami, Y. Kera, H. Noguchi, K. Uosaki, K. Torimoto and B. Ohtani, Phys. Chem. Chem. Phys., 2003, 5, 778–783.
48. R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga, Science, 2001, 293, 269–271.
49. S. Sato, Chem. Phys. Lett., 1986, 123, 126–128.
50. H. Luo, T. Takata, Y. Lee, J. Zhao, K. Domen and Y. Yan, Chem. Mater., 2004, 16, 846–849.
51. M. Miyauchi, A. Nakajima, T. Watanabe and K. Hashimoto, Chem. Mater., 2002, 14, 4714–4720.
52. R. A. Sheldon and J. K. Kochi, Metal-Catalyzed Oxidations of Organic Compounds, Academic Press, New York, 1981.
53. M. Hudlicky, Oxidations in Organic Chemistry, ACS Monograph Series, American Chemical Society, Washington, DC, 1990.
54. C. L. Hill, Advances in Oxygenated Process, ed. A. L. Baumstark, JAI, London, 1998, vol. 1, p.1.
55. R. C. Larock, Comprehensive Organic Transformations, VCH, New York, 1989.
56. R. A. Sheldon, I. W. C. E. Arends and A. Dijksman, Catal. Today, 2000, 57, 157.
57. T. Mallat and A. Baiker, Chem. Rev., 2004, 104, 3037–3058.
58. N. Srinivas, V. R. Rani, M. R. Kishan, S. J. Kulkarni and K. V. Raghavan, J. Mol. Catal. A: Chem., 2001, 172, 187–191.
59. A. Bleloch, B. F. G. Johnson, S. V. Ley, A. Price, D. S. Shephard and A. W. Thomas, Chem. Commun., 1999, 1907–1908.
60. F. Vocanson, Y. P. Guo, J. L. Namy and H. B. Kagan, Synth. Commun., 1998, 398, 1907–1908.
61. T. Matsushita, K. Ebitani and K. Kaneda, Chem. Commun., 1999, 265–266.
62. (a) K. Yamaguchi, K. Mori, T. Mizugaki, K. Ebitani and K. Kaneda, J. Am. Chem. Soc., 2000, 122, 7144–7145; (b) K. Mori, S. Kanai, T. Hara, T. Mizugaki, K. Ebatani, K. Jitsukawa and K. Kaneda, Chem. Mater., 2007, 19, 1249–1256.
63. E. Choi, C. Lee, Y. Na and S. Chang, Org. Lett., 2002, 4, 2369–2371.
64. K. Yamaguchi and N. Mizuno, Angew. Chem., Int. Ed., 2002, 41, 4538–4540.
65. N. Kakiuchi, Y. Maeda, T. Nishimura and S. Uemura, J. Org. Chem., 2001, 66, 6620–6623.
66. T. Mallat and A. Baiker, Catal. Today, 1994, 19, 247–283.
67. H. Wu, Q. Zhang and Y. Wang, Adv. Synth. Catal., 2005, 347, 1356–1357.
68. U. R. Pillai and E. S. Demessie, J. Catal., 2002, 211, 434–444.
69. J. Chen, D. F. Ollis, W. H. Rulkens and H. Bruning, Water Res., 1999, 33, 661–668.
70. (a) D. S. Muggli, J. T. Mccue and J. L. Falconer, J. Catal., 1998, 173, 470–483; (b) J. L. Falconer and K. A. M. Bair, J. Catal., 1998, 179, 171–178.
71. F. H. Hussein, G. Pattenden, R. Rudham and J. J. Russell, Tetrahedron Lett., 1984, 25, 3363–3364.
72. M. Zhang, Q. Wang, C. Chen, L. Zang, W. Ma and J. Zhao, Angew. Chem., Int. Ed., 2009, 48, 6081–6084.
73. Q. Wang, M. Zhang, C. Chen, W. Ma and J. Zhao, Angew. Chem., Int. Ed., 2010, 49, 7976–7979.
74. T. Ohuchi, T. Miyatake, Y. Hitomi and T. Tanaka, Catal. Today, 2007, 120, 233–239.
75. T. Shishido, T. Miyatake, K. Teramura, Y. Hitomi and T. Tanaka, J. Phys. Chem. C, 2009, 113, 18713–18718.
76. S. Furukawa, Y. Ohno, T. Shishido, K. Teramrau and T. Tanaka, Catalysts and Catalysis, 2011, 53, 132–134.
77. L. H. Little, A. V. Kiselev and V. I. Lygin, Infrared Spectra of Adsorbed Species, Academic Press Inc., London, 1966.
78. G. Socrates, Infrared Characteristic Group Frequencies: Table and Charts, Wiley, New York, 1994.
79. V. Z. Fridman, A. A. Davydov and K. Titievsky, J. Catal., 2004, 222, 545–557.
80. M. Sugantha and U. V. Varadaraju, J. Solid State Chem., 1994, 111, 33–40.
81. C. Veríssimo, F. M. S. Garriodo, O. L. Alves, P. Calle, A. Martínez-Juárez, J. E. Iglesias and J. M. Rojo, Solid State Ionics, 1997, 100, 127–134.
82. P. Meriaudeau and J. C. Vedrine, J. Chem. Soc., Faraday Trans. 2, 1976, 72, 472–480.
83. A. R. Gonzalez-Elipe, G. Munuera and J. Soria, J. Chem. Soc., Faraday Trans. 1, 1979, 75, 748–761.
84. Y. Nakaoka and Y. Nosaka, J. Photochem. Photobiol., A, 1997, 110, 299–305.
85. S. Yamazoe, T. Okumura and T. Tanaka, Catal. Today, 2007, 120, 220–225.
86. S. Yamazoe, T. Okumura, Y. Hitomi, T. Shishido and T. Tanaka, J. Phys. Chem. C, 2007, 111, 11077–11085.
87. S. Yamazoe, Y. Hitomi, T. Shishido and T. Tanaka, Appl. Catal., B, 2008, 82, 67–76.

---