1. Introduction
The production of alternative and renewable energy sources is a significant challenge for sustaining the development of human society. Photocatalytic water splitting for converting solar energy to hydrogen energy is one of the ideal solutions for our energy problems. To date, various metal oxide semiconductors, especially Ta-based oxides, have been reported as highly active photocatalysts for water splitting [
1,
2,
3,
4,
5,
6,
7,
8]. However, these compounds have wide band gaps, which is a disadvantage for solar energy harvesting. Domen
et al. have reported that oxynitride compounds have the potential to be efficient water splitting photocatalysts [
9,
10]. Oxynitride compounds have suitable band gaps for visible light absorption and the energy levels of the conduction and valence bands are more negative than the H
2 evolution potential and more positive than the O
2 evolution potential, respectively. Tantalum oxynitride (TaON) is a good candidate for a water splitting photocatalyst due to its stability, suitable conduction and valence band energies [
11,
12]. The photocatalytic activity of TaON for H
2 and O
2 formation from aqueous solution has been investigated using sacrificial reagents, which are electron donors (e.g., methanol) or acceptors (e.g., silver nitrate) under visible light illumination [
13,
14,
15]. Compared to the activity for O
2 formation, the activity of TaON for H
2 formation is an order of magnitude lower [
13,
14,
15]. Thus, improvement of H
2 formation activity is essential for applying TaON as a water splitting photocatalyst.
In our previous studies, we focused on the modification effect of organic dyes (porphyrins, phthalocyanines, and carotenoids) on the photocatalytic activity of inorganic compounds such as KTaO
3 [
16,
17,
18,
19,
20] or SnS [
21]. Specifically, the water splitting activity of the KTaO
3 photocatalyst was much improved by modification with chromium (III) tetraphenyl porphyrin chloride (Cr-TPPCl). The Cr-TPPCl physically adsorbed on the KTaO
3 surface and worked as a photo-excited electron acceptor for KTaO
3, thus prolonging the charge separation state of KTaO
3. Furthermore, from the results of the wavelength dependency of the photocatalytic activity, the water splitting reaction on the dye-modified KTaO
3 proceeded via a two-step excitation of both KTaO
3 and dye similar to photosynthesis in green plants. The reduction potential of the photo-excited electron was increased from the conduction band level of KTaO
3 to the LUMO level of the porphyrin dyes. Thus, H
2 formation activity was improved, showing that dye modification has the potential to improve the reduction ability of inorganic photocatalysts. In this study, we applied this dye modification method to improve the H
2 formation activity of tantalum (oxy)nitride. The tantalum (oxy)nitrides were modified with five different types of porphyrin dyes, and the influence of adsorption states (physical adsorption and covalent attachment) and the photocatalytic activities of each were investigated.
2. Results and Discussion
Crystal structures of (oxy)nitride samples were determined by X-ray diffraction (XRD).
Figure 1 shows the XRD patterns for Ta
2O
5 and (oxy)nitride samples prepared at 800 °C for various heat treatment times under NH
3 gas flow. In this paper, the prepared (oxy)nitride samples are named according to the heat treatment time, e.g., ON-12 is the (oxy)nitride sample treated for 12 h. The XRD patterns showed that the crystal structures of the (oxy)nitride samples were influenced by the treatment time. The ON-12 consisted of TaON (JCPDS No. 71-0178) and a small amount of orthorhombic Ta
2O
5 (JCPDS No. 25-0922). From energy dispersive X-ray spectrometry (EDX), it was confirmed that ON-12 contains 2.9% of Ta
2O
5. The XRD pattern of ON-14 was assigned to that of TaON. In these heat treatment processes, TaON was formed according to the following equation [
22]:
As shown in
Figure 1(d), excessive treatment time led to the formation of Ta
3N
5 (JCPDS No. 79-1533), which is the nitrided phase of TaON formed as in the following equation:
Because these nitridation processes are endothermic [
23], the content of nitrogen increased with an increasing heat treatment period under the preparation conditions in this study. In the case of ON-24, the content of TaON was calculated as 6.6%.
Figure 1.
XRD patterns of (a) Ta2O5 and tantalum (oxy)nitride samples: (b) ON-12; (c) ON-14, and (d) ON-24.
Figure 1.
XRD patterns of (a) Ta2O5 and tantalum (oxy)nitride samples: (b) ON-12; (c) ON-14, and (d) ON-24.
The colors of ON-12, ON-14, and ON-24 are yellow, dark yellow and orange, respectively.
Figure 2(a) shows UV-vis diffuse reflectance spectra of Ta
2O
5 and (oxy)nitride samples. The absorption spectrum of ON-12 is not shown in
Figure 2(a) because it is almost the same as that of ON-14. Compared with the spectrum of Ta
2O
5, both (oxy)nitride samples showed strong visible light absorptions in the wavelength range from 350 to 600 nm. The absorption edges of ON-14 and ON-24 were approximately 550 and 615 nm, respectively. The band gaps of TaON and Ta
3N
5 were reported as 2.4 and 2.1 eV from electrochemical analysis and ultraviolet photoelectron spectroscopy (UPS) [
24]. Strong visible light absorption of ON-24 was attributed to the photoabsorption of Ta
3N
5.
Figure 2(b) shows UV-vis diffuse reflectance spectra of ON-14 modified with various porphyrin dyes. The porphyrin dyes used in this study are as follows: chromium tetraphenyl porphyrin chloride (Cr-TPPCl), hemin, pentamethylene bis[4-(10,15,20-triphenylporphine-5-yl) benzoate]-dizinc (II) (Zn-TPPD), tetrakis(4-carboxyphenyl)porphyrin (TCPP), and 5,10,15,20-tetrakis(pentafluorophenyl)-21H,23H-porphyrin (TPFPP). Compared with the absorption spectrum of ON-14, dye-modified ON-14 showed photoabsorption peaks in the wavelength range from 500–700 nm. These absorption peaks are attributed to the Q-band absorption of porphyrin dyes adsorbed on the surface of ON-14. Therefore, it is expected that visible light from 500 to 700 nm could be used for the photocatalytic reaction.
Figure 2.
UV-vis diffuse reflectance spectra of (a) Ta2O5 and tantalum (oxy)nitride samples and (b) ON-14 modified with various porphyrin dyes.
Figure 2.
UV-vis diffuse reflectance spectra of (a) Ta2O5 and tantalum (oxy)nitride samples and (b) ON-14 modified with various porphyrin dyes.
X-ray photoelectron spectroscopy (XPS) was used to investigate the oxidation state of the (oxy)nitride samples and confirm the nitrogen doping.
Figure 3 shows the O 1s, N 1s, and Ta 4d orbitals of XPS spectra of Ta
2O
5, ON-14, and ON-24. The binding energy was normalized by adjusting the C 1s peak to 284.6 eV. The peak area of the O 1s spectra decreased with increasing heat treatment period, while that of the N 1s spectra increased. The nitrogen content of the (oxy)nitride samples was expected to increase based on the results of the XRD analysis, and the XPS spectra of O 1s and N 1s clearly indicated that N atoms replaced O atoms by the heat treatment under NH
3 gas flow. Compared with the Ta 4d peaks of Ta
2O
5, the Ta 4d peaks of the (oxy)nitride samples shifted to lower binding energy after heat treatment. Some Ta atoms were reduced during nitridation, and the reduced Ta might have produced defects that functioned as trap sites for photogenerated electrons and holes. Therefore, the heat treatment time for preparation of (oxy)nitride samples was optimized at 14 h.
Figure 4 shows Pt 4f XPS spectra of the Pt-loaded Zn-TPPD/ON-14 sample. The Pt 4f
5/2 and 4f
7/2 peaks of the loaded Pt are fitted to four Gaussian peaks. The XPS peaks at 74.4 eV (4f
7/2) and 78.3 eV (4f
5/2) were almost the same as those for PtO
2. The other peaks at 73.4 eV and 76.4 eV appeared between the Pt 4f peaks of PtO
2 and Pt foil. These peaks were attributed to PtO, which has been reported in several papers [
25,
26,
27]. Therefore, it was found that the oxidation states for the Pt particles were Pt
2+ (PtO) and Pt
4+ (PtO
2) on the Zn-TPPD/ON-14. It has been reported that oxidized Pt effectively improved the photocatalytic activity when used as a co-catalyst [
28]. Therefore, PtO and PtO
2 (PtO
x) as co-catalysts loaded on the dye-modified (oxy)nitride were expected to serve as reaction sites and decrease the overpotential of hydrogen formation.
Figure 3.
XPS spectra of O 1s, N 1s, and Ta 4d in (a) Ta2O5; (b) ON-14 and (c) ON-24.
Figure 3.
XPS spectra of O 1s, N 1s, and Ta 4d in (a) Ta2O5; (b) ON-14 and (c) ON-24.
Figure 4.
XPS spectra of Pt4f in (a) PtO2; (b) a Pt foil, and (c) Pt loaded on Zn-TPPD/ON-14.
Figure 4.
XPS spectra of Pt4f in (a) PtO2; (b) a Pt foil, and (c) Pt loaded on Zn-TPPD/ON-14.
The photocatalytic activities of dye-modified (oxy)nitrides were tested in sodium sulfide (Na
2S) aqueous solution. Since some porphyrin dyes dissolved in aqueous alcohol (ethanol or methanol) solutions, Na
2S was employed as a sacrificial reagent for hydrogen formation. The hydrogen formation process under this experimental condition is as follows:
Table 1 shows the effect of dye modification on the catalytic activity of Pt-loaded (oxy)nitride (ON-14) under visible light irradiation. The modification dyes were selected on the basis of the results from dye-modified KTaO
3 photocatalysts for water splitting [
17]. Except for hemin, the porphyrin dyes showed a positive effect on the photocatalytic H
2 formation. In these dyes, TCPP and hemin formed covalent attachments with the ON-14 surface with their carboxyl groups, and the other dyes attached by physical adsorption. The porphyrin dyes attached with physical adsorption showed more positive effects on the photocatalytic activity of ON-14. These dyes were adsorbed onto the ON-14 surface via the porphyrin ring, which may function as the electron acceptor for ON-14. The distance between the porphyrin ring and the TaON surface was shorter than that for TCPP or hemin. This is an advantage for electron transfer between porphyrin dyes and TaON, and thus, the photocatalytic activity of ON-14 was much improved by modification with physically adsorbed dyes. Among the porphyrin dyes examined, the H
2 formation rate for PtO
x/Zn-TPPD/ON-14 was the highest (95.6 μmol g
−1h
−1). In the case of PtO
x/Zn-TPPD/ON-14, the turnover number (TON) calculated from the amount of formed H
2 and porphyrin dye used was 173.6. From the UV-vis absorption spectral measurement, it was confirmed that Zn-TPPD remained after the photocatalytic reaction under visible light irradiation for 6 h. Thus, the porphyrin dyes improved the photocatalytic activity without decomposition.
Table 1.
Dye modification effect on H2 formation from Na2S aqueous solution over PtOx (0.2 wt%)/Dye (0.5 wt%)/(oxy)nitride (ON-14) photocatalyst.
Table 1.
Dye modification effect on H2 formation from Na2S aqueous solution over PtOx (0.2 wt%)/Dye (0.5 wt%)/(oxy)nitride (ON-14) photocatalyst.
Dye | H2 formation rate
/μmol g−1h−1 | TON
(formed H2/dye) |
---|
Zn-TPPD | 95.6 | 173.6 |
Cr-TPPCl | 75.1 | 59.9 |
TPFPP | 59.3 | 69.4 |
TCPP | 30.6 | 29.0 |
None | 7.7 | - |
Hemin | 1.2 | 1.0 |
In order to reveal the role of the modification dye in PtO
x/Zn-TPPD/ON-14, the photocatalytic activities of the catalyst components combined in various ways for H
2 or O
2 formation reactions were investigated. Silver nitrate (AgNO
3) was used as a sacrificial reagent for the O
2 formation reaction.
Figure 5 shows the photocatalytic activities of the catalyst components for H
2 and O
2 formation from sacrificial reagent aqueous solution. For the H
2 formation from Na
2S aqueous solution, PtO
x/Zn-TPPD/ON-14 showed the highest photocatalytic activity. Despite the existence of all components, the H
2 formation rate for Zn-TPPD/PtO
x/ON-14, which was prepared by coating Zn-TPPD on PtO
x/ON-14, was lower than that for PtO
x/Zn-TPPD/ON-14. This strongly suggests that PtO
x works as a co-catalyst for H
2 formation at the surface of Zn-TPPD coated ON-14. PtO
x/ON-14 and Zn-TPPD/ON-14 showed low photocatalytic activity, and no H
2 formation was observed over ON-14 and PtO
x/Zn-TPPD. Thus, the existence of all components (ON-14, dye, and co-catalyst) in conjunction with the preparation procedure was important to improve the photocatalytic activity of dye-modified (oxy)nitride photocatalysts for H
2 formation. On the other hand, in the case of the O
2 formation reaction from AgNO
3 aqueous solution, almost the same formation rates were observed with the exception of that for PtO
x/Zn-TPPD. These results suggest that the reduction and oxidation reaction sites of the PtO
x/Zn-TPPD/ON-14 photocatalyst are the Pt co-catalyst and ON-14 surface, respectively, unlike the case for general dye-sensitized photocatalysts [
29,
30]. Although the PtO
x/Zn-TPPD/ON-14 photocatalyst could form H
2 and O
2 from sacrificial reagent solutions, complete photocatalytic water splitting did not proceed on this photocatalyst. The oxygen or nitrogen defects in (oxy)nitride may trap the excited charge and complete photocatalytic water splitting reaction on the PtO
x/Zn-TPPD/ON-14 photocatalyst is prevented.
Figure 5.
Photocatalytic activities of the combination of PtOx, Zn-TPPD, and TaON (ON-14) for H2 and O2 formation reaction from sacrificial reagent aqueous solution.
Figure 5.
Photocatalytic activities of the combination of PtOx, Zn-TPPD, and TaON (ON-14) for H2 and O2 formation reaction from sacrificial reagent aqueous solution.
Figure 6 shows the expected charge transfer mechanism for PtO
x/Zn-TPPD/ON-14. The conduction and valence band energies for TaON were adapted from reference [
24]. In the case of Zn-TPPD, the HOMO and LUMO levels were determined from the redox potential and the band gap was estimated from the photoabsorption and photoluminescence spectra. The LUMO level for Zn-TPPD was estimated to be −1.25 V, which is more negative than the conduction band level for TaON. Different from the charge transfer mechanism in the dye sensitization system, photo-excited electrons in Zn-TPPD transfer to the PtO
x co-catalysts, which might be the reaction site for H
2 formation. The oxidized Zn-TPPD is reduced by photo-excited electrons in ON-14, and the photo-generated holes in ON-14 oxidize the hole scavenger (S
2−). Zn-TPPD plays a role in absorbing visible light and improving the reduction ability of TaON. Although complete photocatalytic water splitting was not achieved on the dye-modified TaON photocatalyst, this study reveals that modification with porphyrin dye is an effective method for improving the photocatalytic activity of the hydrogen formation reaction on TaON.
Figure 6.
Expected charge transfer mechanism in PtOx/Zn-TPPD/ON-14 photocatalyst.
Figure 6.
Expected charge transfer mechanism in PtOx/Zn-TPPD/ON-14 photocatalyst.
3. Experimental Section
For the preparation of the tantalum (oxy)nitride, Ta
2O
5 powder (5.0 g, Kojundo Chemical Laboratory Co., Ltd., Kariya, Japan) was used as the starting material and placed onto an alumina combustion boat (width: 15 mm; height: 13 mm; length: 90 mm; Nikkato Corp., Sakai, Japan) and heated in an alumina-tube furnace (φ: 30 mm) at 800 °C under a NH
3 gas flow (25 mL min
−1). Except for Cr-TPPCl, all porphyrin dyes, Zn-TPPD [Tokyo Chemical Industry Co., Ltd. (TCI), Tokyo, Japan], TPFPP (TCI), TCPP (TCI), and hemin (TCI) were commercially available and used without further purification. Cr(III)-TPPCl was prepared by a refluxing method published by Adler and coworkers [
31]. H
2TPP (1.0 g, >98.0%, chlorine free, TCI) was dissolved in 300 mL of
N,
N-dimethylformamide (DMF). Several minutes after the reflux started, CrCl
2 (1.0 g, 99.99%, Sigma-Aldrich Co., Ltd., St. Luis, MO, USA) was added to the DMF solution. The reaction mixture was cooled to room temperature, and then poured into a beaker containing 300 mL of chilled ice water. The obtained products were collected by filtration, and dried in a vacuum oven for 1 h at 100 °C. The yield of crude Cr-TPPCl was 64.1%. Purification of Cr-TPPCl was performed by column chromatography. Crude Cr-TPPCl (0.2 g) was dissolved in 50 mL of dichloromethane (CH
2Cl
2), and then applied to a silica gel column (Merck, silica gel 60, 40–63 μm) and eluted with CH
2Cl
2. Unreacted H
2TPP was promptly eluted as a red band with the solvent front, and Cr-TPPCl was eluted as a green band. After the Cr-TPPCl was eluted from the column, the CH
2Cl
2 was removed with a rotary evaporator. The complex was dried for 3 h in a vacuum oven at 100 °C. The yield of purified Cr-TPPCl was 41.2%. The dye modification of tantalum (oxy)nitrides was performed by an evaporation to dryness method with pyridine as the solvent. The (oxy)nitride powder was suspended in dye-dissolved pyridine (0.5 mM) and the solvent was evaporated to dryness by stirring with a glass rod at 130 °C. The amount of modification dye was calibrated as 0.5 wt% of the amount of tantalum (oxy)nitride sample in this study. The platinum co-catalyst was loaded on the dye-modified tantalum (oxy)nitride by an evaporation to dryness method with Pt(NH
3)
4(NO
3)
2 (Sigma-Aldrich Co., St. Luis, MO, USA) aqueous solution, followed by drying on a hot plate at 100 °C. The loading amount of co-catalyst on the dye-modified tantalum (oxy)nitride photocatalysts was 0.1 wt%.
The photocatalytic reaction was performed with a conventional closed circulating system with a dead volume of ca. 500 mL. The catalyst (100 mg) was suspended in 30 mL of reaction solution, which was pre-saturated with Ar. The concentrations of Na2S and AgNO3 aqueous solution were 0.24 M and 0.05 M, respectively. A quartz reaction cell was irradiated by an external light source consisting of a 500 W Xe lamp (Ushio Inc., SX-UI500XQ, Tokyo, Japan). During the photocatalytic reaction, the suspension was mixed using a magnetic stirring bar. Ar gas (10.67 kPa) was used as the circulating carrier gas. Before the reaction started, the absence of gas leakage was confirmed for 1 h by performing a reaction in the absence of light. The amount of H2 or O2 formed was measured by a gas chromatograph with a thermal conductivity detector (Shimadzu Corp., GC-8A, Kyoto, Japan), which was connected to a conventional volumetric circulating line with a vacuum pump.
The crystal structure was analyzed by X-ray powder diffraction (RINT-2500, Rigaku, Tokyo, Japan), in conjunction with Cu Kα radiation, operated at 40 kV, 80 mA, and a scanning rate of 4°/min within the range of 2θ from 10 to 80°. UV-vis absorption spectra were obtained by a reflection method using a spectrophotometer (U-3310, HITACHI, Tokyo, Japan) with an integrating sphere attachment and a KBr pellet was used as the reference. X-ray photoelectron spectroscopy (XPS) was performed with an X-ray photoelectron spectrometer (Shimadzu Co. Ltd., ESCA-3400, Kyoto, Japan) using a Mg Kα irradiation source. Each spectrum was normalized by adjusting the C 1s peak to 284.6 eV. The compositions of the tantalum (oxy)nitrides were determined using scanning electron microscopy-EDX (SEM-EDX, VE-7800, Keyence Corp., Osaka, Japan). The catalyst sample was mounted on a circular metallic sample holder. The EDX analysis was performed at an acceleration voltage of 20 kV without a coating conductive layer.