Photocatalytic Water Splitting for Hydrogen Production with Gd₂MSbO₇ (M = Fe, In, Y) Photocatalysts under Visible Light Irradiation

Abstract

Novel photocatalysts Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇ were synthesized by the solid state reaction method for the first time. A comparative study about the structural and photocatalytic properties of Gd₂MSbO₇ (M = Fe, In, Y) was reported. The results showed that Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇ crystallized with the pyrochlore-type structure, cubic crystal system and space group Fd3m. The lattice parameter a for Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was 10.276026 Å, 10.449546 Å or 10.653651 Å. The band gap of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was estimated to be 2.151 eV, 2.897 eV or 2.396 eV. For the photocatalytic water-splitting reaction, H₂ or O₂ evolution was observed from pure water with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (wavelength > 420 nm). Moreover, H₂ or O₂ also spilt by using Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst from CH₃OH/H₂O or AgNO₃/H₂O solutions under visible light irradiation (λ > 420 nm). Gd₂FeSbO₇ showed the highest activity compared with Gd₂InSbO₇ or Gd₂YSbO₇. At the same time, Gd₂InSbO₇ showed higher activity compared with Gd₂YSbO₇. The photocatalytic activities were further improved under visible light irradiation with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ being loaded by Pt, NiO or RuO₂. The effect of Pt was better than that of NiO or RuO₂ for improving the photocatalytic activity of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇.

1. Introduction

Since water splitting reaction with TiO₂ as catalyst was discovered in 1972 [1], photocatalysis, as one energy-conserving and green emerging technology, has been studied massively and in detail for producing renewable hydrogen. However, early photocatalysts were applied conditionally for their ultraviolet light response. In order to make good use of exhaustless solar energy, plenty of endeavors were made to develop visible light responsive photocatalysts. By summarizing recent research achievements, there were mainly two approaches to obtain visible light responsive photocatalysts. The first approach was to modify TiO₂, such as by ion doping [2,3,4] or forming heterojunction [5,6,7], which was studied most repeatedly due to its non-toxic performance, excellent stability and low cost. The second approach was to develop new photocatalysts, which had also achieved great progress [8,9,10,11,12,13,14]. Especially, the reported CdS photocatalyst [15], with Pt and PdS together as cocatalysts, realized a high quantum efficiency up to 93% in photocatalytic H₂ production, which had aroused wide attention across the world. It was known that TiO₂ could not be used in the visible light region and could only split water under ultraviolet light irradiation. Moreover, ultraviolet light only occupied 4% of sunlight, which was a limitative factor for photocatalytic technology with TiO₂ as the catalyst. Thus, the efficient photocatalysts which could produce electron–hole pairs under visible light irradiation should be developed because visible light occupied 43% of sunlight.

Bi₂MNbO₇ (M = Fe³⁺, In³⁺) [16] photocatalysts, as one remarkable representation of A₂B₂O₇ compound family, were first found to realize H₂ evolution which was obtained from pure H₂O during ultraviolet light irradiation. Thereafter, our research group found that Y₂GaSbO₇ or Y₂GdSbO₇ [17] compound also successfully realized photocatalytic water splitting for hydrogen production under visible light irradiation. Therefore, we could reasonably deduce that the substitution of metallic element in A₂B₂O₇ compounds by another suitable metallic element might provide a way to find new photocatalysts which could efficiently catalyze water splitting reaction even under visible light irradiation. In addition, the substitution might result in the lattice O²⁻ and O⁻ ionosorbed on the surface, which created another way to generate hydroxyl radicals and the highly reactive atomic oxygen species and thus could enhance the photocatalytic activity of solid-solution photocatalysts [18]. Moreover, the substitution of metal ion in A₂B₂O₇ compounds would probably affect the energy level and narrow the band gap of the A₂B₂O₇ compound. As a result, some impurity energy levels or narrow band gaps which had low band gap energy would promote carrier concentration, and thus led to huge change in photocatalytic properties of photocatalysts [19]. Therefore, we speculated Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ catalyst might have phocatalytic potential for water splitting reaction, of which we were more convinced when we found that above novel photocatalysts successfully realized degrading rhodamine B under visible light irradiation [20,21].

In this experiment, we synthesized aforementioned photocatalysts Gd₂MSbO₇ (M = Fe, In, Y) again by a solid-state reaction method. Different from previous studies, Gd₂MSbO₇ (M = Fe, In, Y), herein, served as photocatalysts for splitting water into hydrogen under visible light irradiation. Meanwhile, the structural, photophysical and photocatalytic properties of photocatalysts Gd₂MSbO₇ (M = Fe, In, Y) were also analyzed more comprehensively.

2. Experimental Section

The novel photocatalysts were synthesized by a solid-state reaction method. Gd₂O₃, In₂O₃, Y₂O₃, Fe₂O₃ and Sb₂O₅ with purity of 99.99% (Sinopharm Group Chemical Reagent Co., Ltd., Shanghai, China) were used as starting materials. All powders were dried at 200 °C for 4 h before synthesis. In order to synthesize Gd₂YSbO₇, the precursors were stoichiometrically mixed, then pressed into small columns and put into an alumina crucible (Shenyang Crucible Co., Ltd., Shenyang, China). Finally, calcination was carried out at 1320 °C for 50 h in an electric furnace (KSL 1700X, Hefei Kejing Materials Technology Co., Ltd., Hefei, China). For the sake of preparing Gd₂FeSbO₇, the precursors were stoichiometrically churned up, subsequently pressed into small columns and put into an alumina crucible. Eventually, calcination was performed at 1250 °C for 65 h in an electric furnace. Similarly, Gd₂InSbO₇ was prepared by calcination at 1320 °C for 65 h. The heating rate of calcination was 0.24 °C/s. The crystal structure of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was analyzed by the powder X-ray diffraction method (D/MAX-RB, Rigaku Corporation, Tokyo, Japan) with CuKα radiation (λ = 1.54056 Å). The voltage was 40.0 kV and current was 30.0 mA. The data were collected at 295 K with a step-scan procedure in the range of 2θ = 10°–100°. The step interval was 0.02° for Gd₂FeSbO₇ or Gd₂YSbO₇ and the time per step was 1.2 s. The step interval was 0.01° for Gd₂InSbO₇ and the time per step was 1.0 s. The chemical composition of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was determined by scanning electron microscope-X-ray energy dispersion spectrum (SEM-EDS, LEO 1530VP, LEO Corporation, Krefeld, Germany). The scanning accelerating voltage was 20 kV and linked with an Oxford Instruments X-ray analysis system (Oxford, UK) and X-ray fluorescence spectrometer (XFS, ARL-9800, ARL Corporation, Geneva, Switzerland). The diffuse reflectance spectra of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was analyzed with an UV-visible spectrophotometer (Lambda 40, Perkin-Elmer Corporation, Waltham, MA, USA) in a UV-Vis diffuse reflectance experiment by the dry-pressed disk samples and BaSO₄ was used as the reference material. The surface area of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was measured by the Brunauer-Emmett-Teller (BET) method (MS-21, Quantachrome Instruments Corporation, Boynton Beach, FL, USA) with N₂ adsorption at liquid nitrogen temperature. All the samples were degassed at 180 °C for 8 h prior to nitrogen adsorption measurements. The BET surface area was determined by a multipoint BET method using the adsorption data in the relative pressure (P/P₀) range of 0.05–0.3. A desorption isotherm was used to determine the pore size distribution by the Barret-Joyner-Halender (BJH) method, assuming a cylindrical pore model. The nitrogen adsorption volume at the relative pressure (P/P₀) of 0.994 was used to determine the pore volume and average pore size.

The photocatalytic water splitting was carried out under visible light irradiation in a gas closed circulation system with an inner-irradiation type reactor (quartz cell). A light source (300 W Xe arc lamp, Beijing Dongsheng Glass Light Source Factory, Beijing, China) with the incident photon flux I₀ of 0.056176 µmol cm⁻² s⁻¹ or 0.078245 µmol cm⁻² s⁻¹ was focused through a shutter window and a 420 nm or 390 nm cut-off filter onto the window face of the cell. The incident photon flux I₀ was determined by a Ray virtual radiation actinometer (FU 100, silicon ray detector, Thorlabs Corporation, Newton, NJ, USA). According to the measured I₀ values with 420 nm and 390 nm cut-off filter, the percentage of UV light passing 390 nm cut-off filter was estimated to be 28.2%. The gas which evolved was determined with a thermal conductivity detector TCD-based gas chromatograph (6890 N, Agilent Technologies, Tempe, AZ, USA), which was connected to the gas closed circulation system. One gram of catalyst was suspended in 300 mL H₂O under stirrer. Before reaction, the closed gas circulation system and the reaction cell were degassed until O₂ and N₂ could not be detected. Then, about 35 Torr of argon was charged into the system. H₂ evolution reaction was carried out in CH₃OH/H₂O solution (50 mL CH₃OH, 300 mL H₂O) with Pt, NiO or RuO₂-loaded powder as the catalyst.

For H₂ evolution reaction, Pt, NiO or RuO₂ which was loaded on the surface of the catalysts, was prepared. Pt was loaded on the catalyst surface by an in situ photodeposition method by using aqueous H₂PtCl₆ solution (Shanghai Chemical Reagent Research Institute, Shanghai, China) as the Pt source. A typical synthesis procedure was as follows: 1 g catalyst powder and a calculated amount of (0.2 wt%) H₂PtCl₆ solution were mixed in 150 mL deionized water, and the suspension was then irradiated by a 300 W Xe lamp (λ > 420 nm) under continuous stirring. After 5 h photo-deposition, the suspension was filtered, washed with de-ionized water for 4 times, and finally dried in vacuum at 60 °C for 12 h. NiO or RuO₂ which was loaded on the surface of the catalysts, was prepared by the impregnation method by using Ni(NO₃)₂ or RuCl₃ solution (Sinopharm Group Chemical Reagent Co., Ltd., Shanghai, China), separately. Normally, 1 g catalyst powder was ultrasonically dispersed in 20 mL deionized water, and then a calculated amount of metal precursors (1.0 wt%) Ni(NO₃)₂ or RuCl₃ solution was added to the catalyst powder dispersed water. After magnetically stirring for 1 h, the mixed solution was boiled at 100 °C, and next dried at 60 °C for 12 h. Lastly, the as obtained sample was put into an electric furnace to calcine for 4 h at 450 °C in moving air.

3. Results and Discussion

3.1. Characterization

Figure 1 shows the X-ray powder diffraction patterns of Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇. It could be seen from Figure 1 that Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was a single phase. The calculations of lattice parameters were performed with the program of Cambridge serial total energy package (CASTEP) and first-principles simulation. The CASTEP package was provided by Materials Studio software and the CASTEP calculation was composed of the plane-wave pseudopotential total energy method according to the density functional theory. Thus, our calculations were based on the plane-wave-based density functional theory (DFT) in generalized gradient approximations (GGA) with Perdew–Burke–Ernzerh of (PBE) exchange-correlation potential.

In order to obtain the crystal lattice parameters, Rietveld refinement from X-Ray Diffraction (XRD) data was performed with DBWS software, experimental XRD data and simulation XRD data. The uncertainty of the refined lattice parameters lay in the estimated standard deviation (e.s.d.), calculated by the full pattern fitting program. However, e.s.d. was a measure of precision rather than of accuracy, and these two terms must not be confused. For a sound estimation of the measurement uncertainty of lattice parameters that were refined from XRD data, more information was needed than just the e.s.d. that was provided by the Rietveld refinement of the diffraction pattern of the sample. The outcome of refinements for Gd₂InSbO₇ generated the unweighted R factors, R_p = 12.13% with space group Fd3m. As for Gd₂YSbO₇, R_p was 12.16% with space group Fd3m. As for Gd₂FeSbO₇, R_p was 16.20% with space group Fd3m. According to the Rietveld analysis, Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ had the pyrochlore-type structure and a cubic crystal system which owned a space group Fd3m. The atomic coordinates and structural parameters of Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇ are listed in Table 1, Table 2 and Table 3, respectively. The lattice parameter a for Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was 10.276026 Å, 10.449546 Å or 10.653651 Å. Moreover, the XRD results showed that two theta angles of each reflection of Gd₂FeSbO₇ changed with Fe³⁺ being substituted by In³⁺ or Y³⁺. The lattice parameter α increased from α = 10.276026 Å for Gd₂FeSbO₇ to α = 10.449546 Å for Gd₂InSbO₇, which indicated a decrease in the lattice parameter of the photocatalyst with a decrease of the M ionic radii, Fe³⁺ (0.78 Å) < In³⁺ (0.92 Å). The lattice parameter α also increased from α = 10.276026 Å for Gd₂FeSbO₇ to α = 10.653651 Å for Gd₂YSbO₇, which indicated a decrease in lattice parameter of the photocatalyst with decrease of the M ionic radii, Fe³⁺ (0.78 Å) < Y³⁺ (1.019 Å). Meanwhile, The lattice parameter α also increased from α = 10.449546 Å for Gd₂InSbO₇ to α = 10.653651 Å for Gd₂YSbO₇, which indicated a decrease in the lattice parameter of the photocatalyst with a decrease of the M ionic radii, In³⁺ (0.92 Å) < Y³⁺ (1.019 Å).

Figure 1. X-ray powder diffraction pattern of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ prepared by a solid-state reaction method at 1250 °C, 1320 °C or 1320 °C

Figure 1. X-ray powder diffraction pattern of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ prepared by a solid-state reaction method at 1250 °C, 1320 °C or 1320 °C

Table 1. Structural parameters of Gd₂YSbO₇ prepared by solid-state reaction method. x, y and z refer to the position coordinates of different atoms in one unit cell, respectively.

Atom	x	y	z	Occupation Factor
Gd	0.00000	0.00000	0.00000	1.0
Y	0.50000	0.50000	0.50000	0.5
Sb	0.50000	0.50000	0.50000	0.5
O(1)	−0.16519	0.12500	0.12500	1.0
O(2)	0.12500	0.12500	0.12500	1.0

Table 1. Structural parameters of Gd₂YSbO₇ prepared by solid-state reaction method. x, y and z refer to the position coordinates of different atoms in one unit cell, respectively.

Atom	x	y	z	Occupation Factor
Gd	0.00000	0.00000	0.00000	1.0
Y	0.50000	0.50000	0.50000	0.5
Sb	0.50000	0.50000	0.50000	0.5
O(1)	−0.16519	0.12500	0.12500	1.0
O(2)	0.12500	0.12500	0.12500	1.0

Table 2. Structural parameters of Gd₂BiSbO₇ prepared by solid-state reaction method. x, y and z refer to the position coordinates of different atoms in one unit cell, respectively.

Atom	x	y	z	Occupation Factor
Gd	0.00000	0.00000	0.00000	1.0
Bi	0.50000	0.50000	0.50000	0.5
Sb	0.50000	0.50000	0.50000	0.5
O(1)	−0.14538	0.12500	0.12500	1.0
O(2)	0.12500	0.12500	0.12500	1.0

Table 2. Structural parameters of Gd₂BiSbO₇ prepared by solid-state reaction method. x, y and z refer to the position coordinates of different atoms in one unit cell, respectively.

Atom	x	y	z	Occupation Factor
Gd	0.00000	0.00000	0.00000	1.0
Bi	0.50000	0.50000	0.50000	0.5
Sb	0.50000	0.50000	0.50000	0.5
O(1)	−0.14538	0.12500	0.12500	1.0
O(2)	0.12500	0.12500	0.12500	1.0

Table 3. Structural parameters of Gd₂FeSbO₇ prepared by solid-state reaction method. x, y and z refer to the position coordinates of different atoms in one unit cell, respectively.

Atom	x	y	z	Occupation Factor
Gd	0.00000	0.00000	0.00000	1.0
Fe	0.50000	0.50000	0.50000	0.5
Sb	0.50000	0.50000	0.50000	0.5
O(1)	−0.20249	0.12500	0.12500	1.0
O(2)	0.12500	0.12500	0.12500	1.0

Table 3. Structural parameters of Gd₂FeSbO₇ prepared by solid-state reaction method. x, y and z refer to the position coordinates of different atoms in one unit cell, respectively.

Atom	x	y	z	Occupation Factor
Gd	0.00000	0.00000	0.00000	1.0
Fe	0.50000	0.50000	0.50000	0.5
Sb	0.50000	0.50000	0.50000	0.5
O(1)	−0.20249	0.12500	0.12500	1.0
O(2)	0.12500	0.12500	0.12500	1.0

Figure 2 represents the diffuse reflection spectra of Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇. Compared with well-known photocatalyst TiO₂ whose absorption edge was only 380 nm, the absorption band edge of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was found to be 586 nm, 428 nm or 479 nm. Clearly, the obvious absorption did not result from reflection and scattering. Consequently, the apparent absorbance at sub-band gap wavelengths (600–800 nm for Gd₂FeSbO₇, and 425–800 nm for Gd₂InSbO₇, and 490–700 nm for Gd₂YSbO₇) was higher than zero.

Figure 2. The diffuse reflection spectrum of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇.

Figure 2. The diffuse reflection spectrum of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇.

For a crystalline semiconductor, the optical absorption near the band edge followed the equation: αhν = A (hν − E_g)ⁿ [22,23]. Here, A, α, E_g and ν were proportional constant, absorption coefficient, band gap and light frequency, respectively. E_g and n could be calculated by the following steps: (i) plotting ln(αhν) vs. ln(hν − E_g) by assuming an approximate value of E_g; (ii) deducing the value of n according to the slope in this graph; (iii) refining the value of E_g by plotting (αhν)^1/n vs. hν and extrapolating the plot to (αhν)^1/n = 0. According to the above method, the band gap of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was estimated to be 2.151 eV, 2.897 eV or 2.396 eV.

3.2. Photocatalytic Activity of Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇

Generally speaking, the semiconductor photocatalysis started from the direct absorption of supra-band gap photons and the generation of electron–hole pairs in the semiconductor particles. Subsequently, the diffusion of the charge carriers to the surface of the semiconductor particle was followed. Under visible light irradiation, we measured H₂ or O₂ evolution rate by using Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as photocatalyst from CH₃OH/H₂O or AgNO₃/H₂O solution, respectively. The wavelength (λ) dependence on the photocatalytic activity under light irradiation from full arc up to λ = 420 nm was measured by using different cut-off filters.

Figure 3a shows the photocatalytic H₂ evolution from pure water with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g powder sample, 250 mL pure water). It could be found from Figure 3a that under visible light irradiation, the rate of H₂ evolution in the first 28 h with Gd₂FeSbO₇ as catalyst was 6.329 μmol h⁻¹ g⁻¹, and that with Gd₂InSbO₇ as catalyst was 5.157 μmol h⁻¹ g⁻¹, and that with Gd₂YSbO₇ as catalyst was 4.314 μmol h⁻¹ g⁻¹. Besides, under dark condition, no H₂ evolution was detected from pure water with above three catalysts, which reflected the phocatalytic H₂ evolution activities from pure water of three synthesized catalysts. The reason that water could be split for H₂ evolution from pure water with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm) was as following: Water could be split at a wavelength higher than 420 nm. However, the wavelength was not cut in exactly at 420 nm, in fact, the wavelength was cut by +50 or −50 nm, which meant that the wavelength up to 370 nm was probably absorbed by Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇, which could split water to provide tiny amounts of hydrogen generation in our experiment. The recycling experiments were performed three times with the same experimental conditions of Figure 3a, and the results were almost the same as the above results in Figure 3a. It could be seen that the photocatalysts we had produced had good stability and were thus desirably recyclable.

Figure 3b shows the photocatalytic O₂ evolution from pure water with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g powder sample, 250 mL pure water). It could be found from Figure 3b that under visible light irradiation, the rate of O₂ evolution in the first 28 h with Gd₂FeSbO₇ as catalyst was 3.158 μmol h⁻¹ g⁻¹, and that with Gd₂InSbO₇ as catalyst was 2.574 μmol h⁻¹ g⁻¹, and that with Gd₂YSbO₇ as catalyst was 2.149 μmol h⁻¹ g⁻¹.

Figure 3c shows the photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL methanol solution, 200 mL pure water). It could be found from Figure 3c that under visible light irradiation, the rate of H₂ evolution in the first 28 h with Gd₂FeSbO₇ as catalyst was 18.271 μmol h⁻¹ g⁻¹, and that with Gd₂InSbO₇ as catalyst was 12.479 μmol h⁻¹ g⁻¹, and that with Gd₂YSbO₇ as catalyst was 10.986 μmol h⁻¹ g⁻¹, indicating that the photocatalytic activity of Gd₂FeSbO₇ was much higher than that of Gd₂InSbO₇ or Gd₂YSbO₇.

Figure 3. (a) Photocatalytic H₂ evolution and photocatalytic O₂ evolution from pure water with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g powder sample, 250 mL pure water). Light source: 300 W Xe lamp; (b) Photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL methanol solution, 200 mL pure water). Light source: 300 W Xe lamp.

Figure 3. (a) Photocatalytic H₂ evolution and photocatalytic O₂ evolution from pure water with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g powder sample, 250 mL pure water). Light source: 300 W Xe lamp; (b) Photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL methanol solution, 200 mL pure water). Light source: 300 W Xe lamp.

We would estimate apparent quantum yield in this paper because scattering effects were assumed to be the same for all the photocatalysts and our system was a suspension rather than a homogeneous solution. The apparent quantum yield for hydrogen evolution at 420 nm with Gd₂FeSbO₇ as catalyst was 0.446%, and that with Gd₂InSbO₇ as catalyst was 0.305% and that with Gd₂YSbO₇ as catalyst was 0.268% under visible light irradiation. Moreover, Gd₂InSbO₇ showed higher photocatalytic activity than Gd₂YSbO₇. This also proved that the conduction band level of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was more negative than the reduction potential of H₂O for forming H₂. The formation rate of H₂ increased with decreasing the M ionic radii within Gd₂MSbO₇ (M = Fe, In, Y), Fe³⁺ (0.78 Å) < In³⁺ (0.92 Å) < Y³⁺ (1.019 Å). The reason was that the surface area of the photocatalyst increased with decreasing the M ionic radii, and the creation of more active sites was realized. As a result, the hydrogen generation rate increased. Moreover, the decrease of the M ionic radii would result in a decrease for the migration distance of photogenerated electrons and holes to reach the reaction site on the photocatalyst surface. Thus, the photogenerated electrons and holes could get to the photocatalyst surface more quickly. Above factors would suppress the electron–hole recombination, therefore, the photocatalytic activity would be enhanced. Such results were in good agreement with the optical absorption property of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ (see Figure 2). The rate of H₂ evolution also increased with increasing illumination time. The photocatalytic activity of Gd₂FeSbO₇ increased by about 166% than that of Gd₂YSbO₇.

Figure 4 shows the photocatalytic O₂ evolution from AgNO₃ solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g photocatalyst, 1 mmol AgNO₃, 270 mL pure water). It could be seen from Figure 4 that under visible light irradiation, the rate of O₂ evolution in the first 28 h with Gd₂FeSbO₇ as catalyst was 34.329 μmol h⁻¹ g⁻¹, and that with Gd₂InSbO₇ as catalyst was 23.264 μmol h⁻¹ g⁻¹, and that with Gd₂YSbO₇ as catalyst was 17.200 μmol h⁻¹g⁻¹, indicating that the valence band level of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was more positive than the oxidation potential of H₂O for forming O₂. The formation rate of O₂ increased with decreasing the M ionic radii within Gd₂MSbO₇ (M = Fe, In, Y), Fe³⁺ (0.78 Å) < In³⁺ (0.92 Å) < Y³⁺ (1.019 Å). The apparent quantum yield for the oxygen evolution at 420 nm with Gd₂FeSbO₇ as catalyst was 1.677%, and that with Gd₂InSbO₇ as catalyst was 1.136%, and that with Gd₂YSbO₇ as catalyst is 0.840% under visible light irradiation.

Figure 4. Photocatalytic O₂ evolution from AgNO₃ solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g photocatalyst, 1 mmol AgNO₃, 270 mL pure water). Light source: 300 W Xe lamp.

Figure 4. Photocatalytic O₂ evolution from AgNO₃ solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm, 0.5 g photocatalyst, 1 mmol AgNO₃, 270 mL pure water). Light source: 300 W Xe lamp.

Figure 5 shows the photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under light irradiation (390 nm cut-off filter, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL CH₃OH, 200 mL pure water). It was depicted in Figure 5 that under light irradiation (390 nm cut-off filter), the rate of H₂ evolution in the first 28 h with Gd₂FeSbO₇ as catalyst was 49.707 μmol h⁻¹ g⁻¹, and that with Gd₂InSbO₇ as catalyst was 35.900 μmol h⁻¹ g⁻¹, and that with Gd₂YSbO₇ as catalyst was 29.457 μmol h⁻¹ g⁻¹, indicating that the effect of wavelength (λ) dependence on the photocatalytic activity was very important. The apparent quantum yield for hydrogen evolution at 390 nm with Gd₂FeSbO₇ as catalyst was 0.871%, and that with Gd₂InSbO₇ as catalyst was 0.629% and that with Gd₂YSbO₇ as catalyst was 0.516% under light irradiation (390 nm cut-off filter).

The photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under light irradiation (No cut-off filter, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL CH₃OH, 200 mL pure water) are shown in Figure 6. It could be found from Figure 6 that under light irradiation without using any filters, the rate of H₂ evolution in the first 28 h with Gd₂FeSbO₇ as catalyst was 94.614 μmol h⁻¹ g⁻¹, and that with Gd₂InSbO₇ as catalyst was 70.893 μmol h⁻¹ g⁻¹, and that with Gd₂YSbO₇ as catalyst was 57.100 μmol h⁻¹ g⁻¹, indicating that Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ shows high photocatalytic activity under full arc irradiation. The apparent quantum yield for hydrogen evolution at 420 nm with Gd₂FeSbO₇ as catalyst was 2.311%, and that with Gd₂InSbO₇ as catalyst was 1.731%, and that with Gd₂YSbO₇ as catalyst was 1.394% under light irradiation without using any filters. The photocatalytic activity decreased with increasing incident wavelength λ. As to Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇, the turnover number—the ratio of total amount of gas evolves to catalyst—exceeded 1 for Gd₂FeSbO₇ after 46 h reaction time, exceeded 1 for Gd₂InSbO₇ after 57 h reaction time, and exceeded 1 for Gd₂YSbO₇ after 66 h reaction time under visible light irradiation (λ > 420 nm). Under the condition of full arc irradiation, after 28 h of reaction time, the turnover number exceeded 1.60 for Gd₂FeSbO₇, and the turnover number exceeded 1.32 for Gd₂InSbO₇, and the turnover number exceeded 1.02 as to Gd₂YSbO₇. Above results were enough to prove that the reaction occurred catalytically. The reaction stopped when the light was turned off in this experiment, showing the obvious light response.

Figure 5. Photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under light irradiation (390 nm cut-off filter, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL CH₃OH, 200 mL pure water). Light source: 300 W Xe lamp.

Figure 5. Photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under light irradiation (390 nm cut-off filter, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL CH₃OH, 200 mL pure water). Light source: 300 W Xe lamp.

Figure 6. Photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under light irradiation (No cut-off filter, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL CH₃OH, 200 mL pure water). Light source: 300 W Xe lamp.

Figure 6. Photocatalytic H₂ evolution from aqueous methanol solution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under light irradiation (No cut-off filter, 0.5 g 0.1 wt% Pt-loaded powder sample, 50 mL CH₃OH, 200 mL pure water). Light source: 300 W Xe lamp.

It was known that TiO₂ has very high photocatalytic activity under ultraviolet light irradiation. By contrast, the photocatalytic activity was not obtained with Pt/TiO₂ as catalyst under visible light irradiation (λ > 420 nm), while an obvious photocatalytic activity was observed with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst, showing that Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ could respond to visible light irradiation. The formation rate of H₂ evolution with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst was much larger than that with TiO₂ as catalyst under visible light irradiation. This indicated that the photocatalytic activity of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ for decomposing CH₃OH/H₂O solution was higher than that of TiO₂. The structure of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ after photocatalytic reaction was also checked by using X-ray diffraction method, and no change in their structures was observed during this reaction, which indicated that the H₂ evolution was induced from the photocatalytic reaction of H₂O.

Figure 7 shows the effect of Pt, NiO and RuO₂ co-catalysts on the photoactivity of Gd₂FeSbO₇ under visible light irradiation (λ > 420 nm, 0.5 g powder sample, 50 mL methanol solution, 200 mL pure water). In principle, the photoinduced electrons preferentially enriched on the surface of co-catalyst particles and the recombination of the photoinduced electrons with the photoinduced holes was therefore markedly suppressed. It could be found from Figure 7 that in the first 28 h under visible light irradiation, the rate of H₂ evolution was estimated to be 41.471 μmol h⁻¹ g⁻¹ with 0.2 wt%-Pt/Gd₂FeSbO₇ as catalyst, and that was estimated to be 32.064 μmol h⁻¹ g⁻¹ with 1.0 wt%-NiO/Gd₂FeSbO₇ as catalyst, and that was estimated to be 24.114 μmol h⁻¹ g⁻¹ with 1.0 wt%-RuO₂/Gd₂FeSbO₇ as catalyst, indicating that the photocatalytic activities could be further improved under visible light irradiation with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ being loaded by Pt, NiO or RuO₂. The apparent quantum yield for hydrogen evolution at 420 nm with 0.2 wt%-Pt/Gd₂FeSbO₇ as catalyst was 1.013%, and that with 1.0 wt%-NiO/Gd₂FeSbO₇ as catalyst was 0.783%, and that with 1.0 wt%-RuO₂/Gd₂FeSbO₇ as catalyst was 0.589% under visible light irradiation (λ > 420 nm). The effect of Pt was better than that of NiO or RuO₂ for improving the photocatalytic activity of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇.

Figure 7. Effect of Pt, NiO and RuO₂ co-catalysts on the photoactivity of Gd₂FeSbO₇ under visible light irradiation (λ > 420 nm, 0.5 g powder sample, 50 mL methanol solution, 200 mL pure water). Light source: 300 W Xe lamp.

Figure 7. Effect of Pt, NiO and RuO₂ co-catalysts on the photoactivity of Gd₂FeSbO₇ under visible light irradiation (λ > 420 nm, 0.5 g powder sample, 50 mL methanol solution, 200 mL pure water). Light source: 300 W Xe lamp.

It was known that the process for photocatalysis of semiconductors was the direct absorption of photon by band gap of the materials and generated electron–hole pairs in the semiconductor particles, and the excitation of an electron from the valence band to the conduction band was initiated by light absorption with energy equal to or greater than the band gap of the semiconductor. Upon excitation of photon, the separated electron and hole could follow the solid surface. This suggested that the narrow band gap could more easily excite an electron from the valence band to the conduction band. If the conduction band potential level of the semiconductor was more negative than that of H₂ evolution, and the valence band potential level was more positive than that of O₂ evolution, decomposition of water could occur even without applying electric power [1]. According to the above analysis, the photon absorption of Gd₂FeSbO₇ was much easier than that of the Gd₂InSbO₇ or Gd₂YSbO₇, which resulted in higher photocatalytic activity of Gd₂FeSbO₇.

The information about the specific surface area, pore volume, average pore size of Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇ are presented in Table 4, from which we could see that the specific surface area of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was measured to be 4.12 m² g⁻¹, 3.26 m² g⁻¹ or 1.28 m² g⁻¹, which was significantly smaller than that of TiO₂ photocatalyst (53.8 m² g⁻¹). Above results indicated much higher potential efficiency of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇. Although the surface area of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was smaller than that of TiO₂, but Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ showed higher photocatalytic activity for H₂ evolution under visible light irradiation, which indicated that the high photocatalytic activity of the Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was not owing to a big surface area, but rather due to the narrow band gap. It was obvious that further increase in photocatalytic activity might be prospected from increasing the surface area of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇. Since an efficient photocatalytic reaction process occurred on the photocatalyst surface, the increase of the surface area for the photocatalysts might lead to the increase of their photocatalytic activity.

Table 4. The detailed information about the specific surface area, pore volume, average pore size of Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇ catalysts.

Catalyst	Specific Surface Area (m2 g−1)	Pore Volume (cm3 g−1)	Average Pore Size (nm)
Gd2FeSbO7	4.12	0.034	33
Gd2InSbO7	3.26	0.033	41
Gd2YSbO7	1.28	0.019	59

Table 4. The detailed information about the specific surface area, pore volume, average pore size of Gd₂FeSbO₇, Gd₂InSbO₇ and Gd₂YSbO₇ catalysts.

Catalyst	Specific Surface Area (m2 g−1)	Pore Volume (cm3 g−1)	Average Pore Size (nm)
Gd2FeSbO7	4.12	0.034	33
Gd2InSbO7	3.26	0.033	41
Gd2YSbO7	1.28	0.019	59

4. Conclusions

In the present work, we prepared single phase of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ by solid-state reaction method and studied the structural, optical and photocatalytic properties of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇. Rietveld structure refinement results revealed that Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ crystallized with the pyrochlore-type structure, cubic crystal system and space group Fd3m. The lattice parameter a for Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was 10.276026 Å, 10.449546 Å or 10.653651 Å. The band gap of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ was estimated to be 2.151 eV, 2.897 eV or 2.396 eV. Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ showed optical absorption in the visible light region, indicating that above photocatalysts had the ability to respond to the wavelength of visible light region. For the photocatalytic water-splitting reaction, H₂ or O₂ evolution was observed from pure water with Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst under visible light irradiation (λ > 420 nm). In addition, under visible light irradiation (λ > 420 nm), H₂ or O₂ was also produced by using Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ as catalyst from CH₃OH/H₂O or AgNO₃/H₂O solutions. Gd₂FeSbO₇ showed the highest activity compared with Gd₂InSbO₇ or Gd₂YSbO₇. At the same time, Gd₂InSbO₇ showed higher activity compared with Gd₂YSbO₇. The photocatalytic activities were further improved under visible light irradiation with Gd₂FeSbO₇ being loaded by Pt, NiO or RuO₂. The effect of Pt was better than that of NiO or RuO₂ for improving the photocatalytic activity of Gd₂FeSbO₇. Moreover, the synthesis of Gd₂FeSbO₇, Gd₂InSbO₇ or Gd₂YSbO₇ offered some useful insights for the design of new photocatalysts for the photocatalytic evolution of H₂ or O₂.