Soft template inducted hydrothermal BiYO₃ catalysts for enhanced formic acid formation from the photocatalytic reduction of carbon dioxide

Abstract

Bismuth yttrium oxide (BiYO₃) photocatalysts were hydrothermally prepared using sodium dodecyl benzene sulfonate (SDBS), disodium ethylenediamine tetraacetate (EDTA), and citric acid (CA) as soft templates. The effects of templates on the formed BiYO₃ nanostructures were revealed. The BiYO₃ nanostructures prepared without the existing template and with the existence of SDBS, EDTA, and CA as templates had flaked structures, sliced, rectangular lath structures with fine particles, and scaly uniform particles, respectively. The photocatalytic reduction of CO₂ under visible-light irradiation with these synthesized BiYO₃ nanostructures was studied and compared. Due to the larger specific surface area, higher adsorbed edge, and lower band gap, BiYO₃ prepared using EDTA as a template produced fewer hydroxyl free radicals (·OH) to give a higher yield of formic acid (HCOOH) in the process of the photocatalytic reduction of CO₂. The HCOOH yield from the photocatalytic reduction of CO₂ by the EDTA synthesized BiYO₃ was 1.68 μmol L⁻¹, which was 1.9 times higher than that produced from BiYO₃ prepared without a template, and 2.6 times higher than that of BiYO₃ prepared by a solid-state reaction.

---

1. Introduction

With accelerated industrialization and urbanization processes and continuously upgraded consumption structures, CO₂ generated from the combustion of coal, oil, natural gas and other hydrocarbon fuels has increased unbelievably over the past decade, and the subsequent energy crisis and global warming have become increasingly serious.^1,2 Therefore, it is very urgent to transform and utilize CO₂. At present, the conversion technology of CO₂ includes biological,³ electrochemical,⁴ photochemical,⁵ and photocatalytic reduction⁶ methods. Due to the required clean and renewable solar energy, mild reaction conditions, and lack of pollution, the photocatalytic reduction is considered favorably for CO₂ conversion to valuable chemicals.

The photocatalytic conversion of CO₂ into hydrocarbons uses solar energy to excite semiconductor photocatalysts to produce photogenerated electron–hole pairs and to induce a redox reaction to synthesize hydrocarbons from CO₂.^7–8 For example, Halmann⁹ used GaP as a photoelectric pole to reduce CO₂ to CH₃OH in an aqueous solution in 1978. Inoue et al. used TiO₂, CdS, GaP, ZnO, and SiC photocatalysts to photocatalytically reduce CO₂ in 1979.¹⁰ TiO₂ is the most popular material used in the photocatalytic reduction of CO₂, however, the band gap of TiO₂ is 3.2 eV, which can only respond to ultraviolet light (UV) (about 3–5% in the solar light) for the photocatalytic reaction. In recent years, to utilize visible light, modification of TiO₂ with Ag,¹¹ Cu₂O,¹² and CeO₂ ¹³ has been reported to shift the absorption edge of TiO₂ from the UV range to the visible light range. On the other hand, developed new “second generation” photocatalysts have been used in the photocatalytic reduction of CO₂, such as Zn₂GeO₄,¹⁴ HNb₃O₈,¹⁵ ZnGa₂O₄,¹⁶ and BiVO₄.¹⁷ Although the photocatalytic reduction of CO₂ has made significant progress, there are still many problems, for example, the low utilization rate of solar energy, low quantum yield, low adsorption of CO₂ on the photocatalyst, and poor light stability. All these lead to a very low yield of hydrocarbons. Meanwhile, a template can effectively control crystal nucleation, growth and the assembly process of inorganic materials by simulating the biological mineralization process, and subsequently can influence the surface morphology, catalyst particle size and the interaction between the photocatalyst and the reactant.^20,25 Bi-based oxide semiconductors have unique electronic structures, forming a valence band between the Bi-6s and the O-2p,¹⁸ for example, BiVO₄,¹⁹ Bi₂MoO₆,²⁰ BiWO₆,²¹ and BiYO₃.^22–23 However, their application for the photocatalytic reduction of CO₂ has rarely been reported.

In this work, BiYO₃ photocatalysts were prepared using a hydrothermal method. The influence of a soft template sodium dodecyl benzene sulfonate (SDBS), disodium ethylenediamine tetraacetate (EDTA), and citric acid (CA) on the photocatalysts were studied. The photocatalytic reduction of CO₂ conversion was studied and compared.

2. Experimental

2.1. Preparation of photocatalysts

The BiYO₃ photocatalysts with different templates were prepared by a hydrothermal method. Briefly, 0.5 mol L⁻¹ Bi(NO₃)₃ aqueous solution and 0.5 mol L⁻¹ Y(NO₃)₃ aqueous solution were mixed evenly. Then, 4 mol L⁻¹ NH₄OH was added into the mixture with stirring at 100 rpm to adjust the pH = 8.0; after that 2.0 g SDBS, EDTA, or CA were added into the mixture as a template. After being stirred at room temperature for 0.5 h, the mixture was transferred into a Teflon-lined stainless sealed autoclave and heated at 180 °C for 12 h. After natural cooling, the reactant was vacuum filtrated and washed with deionized water, and then dried at 90 °C for 10 h and calcined at 700 °C for 2 h to obtain BiYO₃. The BiYO₃ prepared by SDBS, EDTA, and CA were marked as BiYO₃-SDBS, BiYO₃-EDTA, and BiYO₃-CA, respectively. For comparison, the BiYO₃ catalyst (marked as BiYO₃-blank) was prepared following the same method without adding a template. Meanwhile, the BiYO₃ catalyst was further prepared by a solid state reaction following the prior procedures for comparison.²²

2.2. Photocatalytic activity tests

The photocatalytic reduction of CO₂ was carried out in a SGY-I photochemical reactor (Nanjing Stonetech Electric Equipment Co., Ltd.). A 300 W Xe lamp was used as the light source and the UV light was removed by UV cut-off filter. 200 mL NaOH (0.25 mol L⁻¹) and Na₂SO₃ (0.1 mol L⁻¹) aqueous solution was used to maintain an alkaline environment in the reaction process, and 2.00 g L⁻¹ BiYO₃ catalysts were added into the reactor. Before illuminating, CO₂ was allowed to flow into the reactor at 100 mL min⁻¹ for 0.5 h with solution stirring to reach an absorption equilibrium. During the reaction process, about 5 mL of the reaction suspension was extracted every 1.0 h. The products, mainly formic acid (HCOOH), were detected by a TU-1901 dual beam UV-visible spectrophotometer (Beijing Purkinje General Instrument Co., Ltd) according to the method described in the literature.²⁴

2.3. Characterizations and measurements

XRD was conducted on a Bruker D8 Advance X-ray diffractometer with Cu Kα radiation at 36 kV, 55 mA using a graphite monochromator. The morphology of the photocatalysts was investigated using a Hitachi SU-8020 scanning electron microscope (SEM). The oxidation state and surface composition were analyzed using an X-ray photoelectron spectrometer (XPS) (Kratos Ultra Axis DLD), equipped with an Al Kα radiation source, and the XPS analysis was conducted at 150 W with a pass energy of 40 eV. The Fourier transform infrared (FTIR) spectrum of BiYO₃ samples was studied in a Nicolet NEXUS 470 infrared spectrometer (Thermo Nicolet, USA), and BiYO₃ was pressed into a tablet after being dispersed in a KBr matrix, and scanned 20 times at a resolution of 4 cm⁻¹ from 400 to 4000 cm⁻¹. The N₂ adsorption/desorption was conducted using a Micromeritics TriStar II 3020 surface area and porosity analyzer with N₂ (at 77.3 K) as sorbate, the total specific surface areas were calculated by the multi-point Brunauer–Emmett–Teller (BET) method and the pore size distributions were evaluated using the standard Barrett–Joyner–Halenda (BJH) method. The UV-vis diffuse reflection spectra of the catalysts were collected by a TU-1901 dual beam UV visible spectrophotometer (Beijing Purkinje General Instrument Co., Ltd.), and were converted from reflectance to absorbance by the Kubelka–Munk method. A photoluminescence (PL) technique with terephthalic acid as a probe molecule was used to detect the formation of ·OH free radicals on the surface of the UV-illuminated photocatalyst, the details of the PL test were documented elsewhere.²⁵ The zeta potentials of the BiYO₃ samples were obtained in a Zeta Potential Analyzer (BrookHaven, American) in 0.1 mM KCl aqueous solution and at 25 °C.

3. Results and discussion

3.1. XRD and zeta potential analysis

Fig. 1 shows the XRD patterns of the BiYO₃ photocatalysts prepared by SDBS, EDTA and CA and with no template using a hydrothermal method after being calcined at 700 °C for 2 h. The XRD patterns of the four photocatalysts can be indexed to BiYO₃ (JCPDF card no. 27-1047); the peaks at 2θ = 28.58°, 33.01°, 47.36°,56.14°, 58.85°, 69.06°, 76.44°, and 78.69° corresponded to the crystal planes of (111), (200), (220), (311), (222), (400), (331), and (420) of BiYO₃, respectively, indicating that BiYO₃ can be obtained by a hydrothermal method with SDBS, EDTA, and CA as a template or without a template.

Fig. 1 XRD patterns of the BiYO₃ photocatalysts prepared by SDBS, EDTA and CA and without template using a hydrothermal method.

A weak diffraction peak observed in all the four samples was found at 2θ = 30.40°, attributed to the diffraction peak of Bi₂O₃, which was the intermediate product during the hydrothermal reaction process. The formation of Bi₂O₃ would consume the Bi³⁺ ions, leading to a decreased Bi³⁺ concentration in the reaction solution. A lower concentration of the reactants (Bi³⁺) can promote the growth of the BiYO₃ crystal.²⁶ The diffraction peak of the BiYO₃ photocatalyst without a template was narrow, while the diffraction peaks of the BiYO₃ photocatalysts prepared with adding SDBS, EDTA, and CA were slightly wide. A decreasing grain size of crystallite was observed using Scherrer’s equation according to the BiYO₃ (111) peaks, the crystallite sizes of photocatalysts were 21.44, 18.42, 18.02, and 16.34 nm, respectively.

The interactions between the adsorbents and the BiYO₃ samples strongly depend on the surface charge of the templates. The zeta potentials of the surface charged particles were measured and are shown in Table 1. The zeta potential of the BiYO₃ prepared using SDBS, EDTA, and CA as templates was 2.98, 4.70 and 3.09 mV, respectively, lower than that of the BiYO₃ prepared by no template (16.47 mV). This indicates that the addition of anionic surfactant caused a decrease of the zeta potential.²⁷ SDBS and CA are anionic surfactants, and EDTA is a chelate with four carboxyl groups, which may easily dissociate negative charge. The decreased zeta potential for the BiYO₃ prepared using SDBS, EDTA, and CA as templates with adsorbed positive and negative charges had a significant effect on the crystallite size.²⁸ This indicates that the template has a structure directing effect on the formed BiYO₃ during its crystal nucleation and growth in the reaction process. According to the measured zeta potential value, the stronger the adsorption force between the template and BiYO₃, the more the inhibition of the crystal growth, and thus, the crystallite size of BiYO₃ after adding SDBS, EDTA, and CA was decreased.

Table 1 The effects of the type of template on the zeta potential

Sample name	Zeta potential (mV)
BiYO3-blank	16.47
BiYO3-SDBS	2.98
BiYO3-EDTA	4.70
BiYO3-CA	3.09

---

3.2. SEM analysis

Fig. 2 shows SEM images of the BiYO₃ photocatalysts prepared with no template, and with SDBS, EDTA, and CA as templates by a hydrothermal method. As shown in Fig. 2(a), the BiYO₃ photocatalyst prepared without template was flaky with a flake length of 100–600 nm. In Fig. 2(b) it can be seen that the BiYO₃ prepared using SDBS as a template formed slices with a length of 100–800 nm and a thickness of about 10 nm. SDBS is an anionic surface active agent and can produce negatively charged surface active agents in aqueous solution. These adsorbed negative charges cover the photocatalyst surface to promote the crystal growth and form a thin slice.²⁶ In Fig. 2(c) it can be seen that the BiYO₃ catalysts prepared using EDTA as a template showed a rectangular lath structures with fine particles with size less than 10 nm on the surface, different from the BiYO₃ catalysts formed with SDBS. When the chelating agent EDTA with six atoms was added into the aqueous solution, the Bi–EDTA chelate compound decreased the concentration of free Bi³⁺ in the solution, slowed down the growth rate of BiYO₃ crystal,²⁹ and thus generated tiny nanoparticles²⁶ with uniform dispersion. When the photocatalytic reaction was carried out, the edge of the rectangular lath structure serves as sites for electronic transition, and thus an accelerated photocatalytic reaction was observed on the BiYO₃ powders. In Fig. 2(d) it can be seen that the BiYO₃ photocatalyst prepared using CA as a template formed scaly uniform particles. When the complexing agent CA with three carboxyl groups was added into the solution, three H⁺ were dissociated and the complex compound of –COOBi would be formed to promote the growth rate of BiYO₃ crystal.³⁰

Fig. 2 SEM images of the BiYO₃ catalysts prepared (a) without any template, and with a template of (b) SDBS, (c) EDTA, and (d) CA using a hydrothermal method.

3.3. XPS analysis

The elementary oxidation states and surface composition of BiYO₃ prepared without template, and with SDBS, EDTA and CA as templates were investigated using XPS analysis, Fig. 3. The binding energies of the core electrons of the four BiYO₃ photocatalysts are shown in Table 2. In the Bi 4f spectrum, the binding energy of Bi 4f7/2 in the four samples was about 164 and 156 eV,³¹ and the binding energy of Bi 4f5/2, a weak peak, was 156 eV,³² suggesting that Bi occurred in the form of Bi³⁺ in BiYO₃. As shown in Table 2, with the addition the template of SDBS, EDTA, or CA, the binding energies of Bi 4f7/2 and Bi 4f5/2 of BiYO₃ had a blue shift of 0.1 eV. As mentioned previously, SDBS, an anionic surface active agent, can produce negative charges in aqueous solutions and absorb on the surface of the photocatalyst, increasing the extra nuclear electron density of the Bi atomic nucleus. The chelating agent EDTA with six atoms can form Bi–EDTA chelates with Bi³⁺ in alkaline conditions; the complexing agent CA with three carboxyl groups can dissociate and give three H⁺ in water, and form three –COOBi complexes with Bi³⁺ ions. When the chemical bonds were formed between the Bi atom and EDTA or CA, the nuclear electron density of Bi increases and the binding energy decreases due to the transfer of valence electron to the Bi.³³ For the Y 3p spectra in Fig. 3, the Y 3p3/2 (312 eV) and Y 3p1/2 (300 eV) of the catalysts after using three templates were similar to the catalysts without template and could be assigned to the surface of Y³⁺ species.³⁴ In the O 1s spectrum, three kinds of binding oxygen were observed, namely, (a) oxygen in the lattice (O²⁻) located at 529.3 eV,³⁵ (b) oxygen in the hydroxyl group on the photocatalysts surface (–OH) located at 531.4 eV,³⁶ and (c) oxygen absorbed on the surface (O⁻) located at 533.2 eV.^35–36 From the spectra and the data in Table 2, the content of surface adsorbed oxygen of the BiYO₃ prepared by EDTA (0.92%) was significantly lower than that of the other three kinds of catalysts.

Fig. 3 The Bi 4f, Y 3p, and O 1s XPS spectra of BiYO₃ prepared (a) without template, and with a template of (b) SDBS, (c) EDTA, and (d) CA by a hydrothermal method.

Table 2 Binding energy (BE, eV) of core electrons of BiYO₃ catalysts prepared by using SDBS, EDTA, and CA as template and no template by a hydrothermal method

XPS spectra	Element valence	BE (percent of valence state, %)
		BiYO3-blank	SDBS-BiYO3	EDTA-BiYO3	CA-BiYO3
Bi 4f	Bi^3+(Bi 4f7/2)	164.41(34.63)	164.32 (34.75)	164.34(31.50)	164.33(32.06)
	Bi^3+(Bi 4f5/2)	159.03(59.67)	158.92(62.07)	158.91(61.91)	158.90(62.94)
	Bi^3+(Bi 4f7/2)	156.61(5.71)	156.38(3.19)	156.54(6.59)	156.38(5.00)
Y 3p	Y^3+(Y 3p3/2)	311.89(38.28)	311.98(35.60)	311.91(36.57)	311.90(36.35)
	Y^3+(Y 3p1/2)	300.03(61.72)	300.12(64.40)	300.03(63.43)	300.02(63.65)
O 1s	OH^−	531.36(35.18)	531.39(49.53)	531.44(47.92)	531.41(48.51)
	O^−	533.16(3.72)	533.39(2.07)	533.20(0.92)	533.28(2.80)
	O^2−	529.36(61.10)	529.31(48.40)	529.26(51.16)	529.23(48.68)

---

In a photocatalytic reaction, the oxygen species played an important role in the reaction process.^37–38 Especially, in a photocatalytic reduction reaction system, the adsorbed oxygen species can be used as the electron acceptor to compete for the photoelectron, thus the presence of adsorbed oxygen species will prevent the photocatalytic reduction reaction.³⁹ The oxygen species of chemical adsorption, such as surface hydroxyl groups, are scavengers of photo-induced holes and can generate active species ·OH free radicals. The ·OH in the adsorbed phase and solution phase is a strong oxidant and can easily initiate an oxidation reaction. The physically adsorbed oxygen species, such as O₂, act as scavengers of the photo-induced electrons. Thus, an increased amount of the physically adsorbed oxygen would help to improve the capture of the photo-induced electrons and decrease the recombination probability of the photo-induced electrons and the photo-induced holes, which will inhibit the photocatalytic reduction reaction. The total adsorbed oxygen species (including the surface hydroxyl oxygen and surface adsorbed oxygen) content of BiYO₃ prepared with no template, and using SDBS, EDTA, and CA as templates were 38.9%, 51.6%, 48.84%, and 51.32%, respectively.

3.4. N₂ adsorption–desorption analysis

Fig. 4 shows N₂ adsorption–desorption isotherms and the pore size distribution curves of the different BiYO₃ catalysts. The specific surface areas of the different BiYO₃ photocatalysts obtained by a multi-point BET method and other parameters are shown in Table 3. It can be found that the N₂ adsorption isotherm of the four samples belonged to a III-type isotherm with an H3-type hysteresis loop. At low pressures (P/P₀ = 0–0.8), the adsorption capacity of N₂ increased slowly with increasing relative pressure, indicating that the adsorption of N₂ molecules on the surface of catalyst was converted to a multi-layer adsorption.⁴⁰ At high pressures (P/P₀ = 0.8–1.0), the isotherms obtained by desorption were not coincident with the adsorption isotherms, attributed to the occurrence of capillary condensation. From Fig. 4(A) it can be seen that the hysteresis loop of BiYO₃ prepared using EDTA as template began at P/P₀ = 0.8 and ended at P/P₀ of about 1.0, the beginning relative pressure was 0.05 lower that other samples, indicating that the amount of adsorption and desorption by the BiYO₃ prepared using EDTA as template was larger than that of other catalysts according to the capillary condensed theory and Kelvin equation.⁴¹

Fig. 4 (A) Nitrogen adsorption/desorption isotherms and (B) pore size distribution of BiYO₃ prepared by (a) no template and using (b) SDBS, (c) EDTA, and (d) CA as templates with a hydrothermal method.

Table 3 Textural properties of the BiYO₃ catalysts prepared using SDBS, EDTA, and CA as templates and no template with a hydrothermal method

Sample name	BET surface area (m^2 g^−1)	Pore diameter (nm)	Pore volume (mL g^−1)
BiYO3-blank	6.75	4.72	0.034
BiYO3-SDBS	15.62	2.14	0.080
BiYO3-EDTA	20.87	20.20	0.164
BiYO3-CA	12.20	30.17	0.109

---

As shown in Fig. 4(B) and Table 2, the four BiYO₃ photocatalysts were mesoporous. The pore size of BiYO₃ prepared without a template was 4.72 nm, while it was 2.14, 20.20, and 30.17 nm for the BiYO₃ prepared using SDBS, EDTA, and CA as templates, respectively, showing that the templates had a significant effect on the pore size of the photocatalyst.⁴²

The specific surface areas of BiYO₃ prepared using SDBS, EDTA, and CA are 15.62, 20.87, and 12.20 m² g⁻¹, respectively, which are higher than that of the BiYO₃ prepared without a template (6.75 m² g⁻¹). Obviously, after using templates, the surface areas and pore volume of BiYO₃ were increased. Among these samples, the BiYO₃ prepared using EDTA as template had a larger specific surface area and a bigger pore volume than the other three BiYO₃ samples.

3.5. UV-vis diffuse reflectance spectra and FTIR analysis

Fig. 5 shows the UV-vis diffuse reflectance spectra and the FTIR spectra of BiYO₃ prepared without a template, and using SDBS, EDTA, and CA as templates with a hydrothermal method. Obviously, these four BiYO₃ samples absorbed visible light, and the absorbed edges of the BiYO₃ prepared without a template, and using SDBS, EDTA, and CA as templates were 517, 533, 569, and 539 nm, indicating that the adsorbed edges had a red shift after the templates were added. For a semiconductor sample, it is possible to determine the optical absorption near the band edge using a Kubelka–Munk method,⁴³ as show in eqn (1).

αhν = A(hν − E_g)^n/2 (1)

where α is the absorption coefficient; ν is the radiation frequency; h is the Planck constant; A is a constant; and E_g is the band gap. According to eqn (1), the band gap values of the BiYO₃ prepared without a template, and using SDBS, EDTA, and CA as templates were 2.39, 2.32, 2.18, and 2.30 eV, respectively. After the introduction of a template, red shifts were observed on the three samples, because the Bi–EDTA or –COOBi covalent bonds were formed between Bi atom and three templates through the negative charges in the aqueous solution. Therefore, the transfer of valence electrons to the Bi atom and the hybrid orbital of Bi 6s and O 2p became narrow.³⁰ This increased the activity of photoelectrons and holes and electronic transition becomes easier after being photo-excited.

Fig. 5 UV-vis diffuse reflectance spectra (A) and FT-IR spectra (B) of the BiYO₃ catalysts prepared by (a) no template, and using (b) EDTA, (c) SDBS, and (d) CA as templates with a hydrothermal method.

The FT-IR spectra (Fig. 5(B)) of the four samples show main absorption bands at 3440 and 1600 cm⁻¹, which correspond to the O–H stretching and H–O–H bending vibrations of the adsorbed moisture in the samples,⁴⁴ and the peaks at about 848 cm⁻¹ were attributed to the stretching vibrations of the Bi–O in BiYO₃.⁴⁵ The weak bands at about 2990 and 2831 cm⁻¹ were ascribed to the stretching vibrations of the aliphatic CH₂ and CH₃ groups,²⁷ respectively, which were not found in the spectrum of the BiYO₃ prepared without a template, indicating very small amount of the templates were found on the BiYO₃ surface. This further confirmed the zeta potential results.

3.6. Fluorescence spectroscopy analysis

The charges (electron e⁻ and hole h⁺), photogenerated from the inside of the photocatalyst to the surface in a photocatalytic process, react with water to produce a very strong oxidation ability species, i.e., hydroxyl radicals. Hydroxyl radicals are an important active species in preventing the activity of the photocatalytic reduction. The higher the content of hydroxyl radicals, the easier it is to capture the light, and the more difficult it is to undergo the photocatalytic reduction reaction.^46–47 Terephthalic acid reacts with hydroxyl groups readily to produce a highly fluorescent product, 2-hydroxyterephthalic acid, whose PL peak intensity is in proportion to the amount of hydroxyl radicals produced in water.⁴⁸ A photoluminescence (PL) technique with terephthalic acid as a probe molecule was used to detect the formation of free hydroxyl radicals on the surface of the BiYO₃ photocatalyst with different template agents, Fig. 6.

Fig. 6 PL spectra observed after illumination of BiYO₃ prepared by no template and using SDBS, EDTA, and CA as templates with a hydrothermal method in 2 × 10⁻³ mol L⁻¹ NaOH solution of terephthalic acid (excitation at 315 nm), and the blank was the ​the fluorescence spectrum of pureterephthalic acid.

From Fig. 6, an obvious difference in PL intensity at about 425 nm can be observed when using BiYO₃ prepared by different templates as photocatalysts. No PL intensity at about 425 nm was observed without a photocatalyst. The PL intensity order of the as-prepared BiYO₃ photocatalysts was: EDTA < SDBS < CA < no template, indicating that the addition of a template can effectively inhibit the formation of hydroxyl radicals, which were the active substances in the photocatalytic oxidation. The PL intensity of the 2-hydroxyterephthalic acid with the BiYO₃-EDTA photocatalyst was much lower than that of other three BiYO₃ photocatalysts, indicating that the BiYO₃ prepared by EDTA could promote the electron transfer and prevent the recombination of photogenerated electrons and holes.⁴⁹ Furthermore, the relatively lower PL intensity of BiYO₃-EDTA indicates that the electron–hole recombination on the surface of catalysts is largely inhibited to generate more photoelectrons and holes to trigger the photocatalytic reaction.⁵⁰

3.7. Photocatalytic reduction of CO₂ under visible-light irradiation

Fig. 7(A) shows the HCOOH products of the BiYO₃ photocatalysts prepared by no template and using SDBS, EDTA, and CA as templates with a hydrothermal method. The yields of HCOOH using BiYO₃ prepared by no template and using SDBS, EDTA, and CA as templates were 0.90, 1.36, 1.68, and 1.06 μmol L⁻¹, respectively. For comparison, the yield of HCOOH using BiYO₃ prepared by a solid–state reaction was measured to be 0.64 μmol L⁻¹ and there was almost no yield of HCOOH in the dark condition. Obviously, there was an enhancement of photocatalytic activity after using SDBS, EDTA, and CA as template, the yield of HCOOH by photocatalytic reduction of CO₂ by BiYO₃ prepared by using EDTA as template was 1.9 times higher than BiYO₃ prepared without template, and 2.6 times higher than that of BiYO₃ prepared by a solid-state reaction, or other photocatalysts, such as RuReCl,⁵¹ and GN–TiO₂,⁵² for the photocatalytic reduction of CO₂. Combined with the results of characterization, when SDBS was used as the template, the surface of the photocatalyst was covered by the adsorbed negative charges to inhibit the crystal growth, while EDTA and CA mainly formed Bi–EDTA chelation and –COOBi complexing agent with Bi³⁺ in the reaction solution to inhibit the crystal growth, then slice-like BiYO₃, rectangular lath-like BiYO₃, and scale-like BiYO₃ were obtained. At the same time, the addition of SDBS, EDTA, and CA also increased the specific surface area of BiYO₃ from 6.75 m² g⁻¹ to 15.62, 20.87, and 12.20 m² g⁻¹, respectively, and decreased the band gap of the BiYO₃ photocatalysts from 2.39 eV to 2.32, 2.18, and 2.30 eV. Among all the BiYO₃ photocatalysts, the BiYO₃ prepared using EDTA as a template had a higher specific surface area (20.79 m² g⁻¹) and a larger visible light absorbed edge (569 nm) or lower band gap (2.18 eV), which greatly increased the adsorption capacity of CO₂, and provided more active sites on the catalyst surface, and accelerated the mass transfer rate of the photogenerated electron and the hole, and enhanced the separation efficiency of the photogenerated electron–hole pairs, which led to a higher photocatalytic reduction activity.⁷

Fig. 7 (A) Yield of HCOOH, and (B) the kinetics (ln(HCOOH concentration) vs. reaction time) of the photocatalytic reduction of CO₂, and (C) the cycle time effect on HCOOH yield with the BiYO₃ catalysts prepared using EDTA as template.

The photocatalytic reduction of CO₂ can be described by eqn (2).

CO₂ + H₂O ↔ 1/2O₂ + HCOOH (2)

For most of the photocatalytic processes, the kinetics can be described by a pseudo first-order reaction.⁵³ Therefore, using the pseudo first-order relationship between the reaction time, rate constant, and the concentration of HCOOH, the linear relationship was described by eqn (3), and Fig. 7(B) shows the plot of reaction time t and ln c.

ln(c/c₀) = −kt (3)

where c₀ is the initial concentration and c is the concentration of the HCOOH, t is the reaction time, and k is the first order rate constant. The logarithmic plot of the concentration data gives a straight line. From the data in Fig. 7(B), the reaction rate constants of the photocatalytic reduction of CO₂ on BiYO₃ prepared using EDTA, SDBS, CA as templates and with no template, and prepared by a solid state reaction, and the reduction of CO₂ in a dark reaction were calculated to be 0.0147, 0.0136, 0.0110, 0.0102, 0.0096, and 0.0056 min⁻¹, indicating that the BiYO₃ prepared using EDTA exhibited the fastest reaction rate among the studied samples.

Fig. 7(C) shows the reproducibility of the BiYO₃ catalyst on the photocatalytic reduction of CO₂. In the same conditions, the HCOOH yield was 1.68 μmol mL⁻¹ after the 1^st round of reaction. When the photocatalysts were used for their sixth round, the HCOOH yield was reduced by 6.5% to 1.57 μmol mL⁻¹. This shows that the catalyst maintains a good stability after a six-time reaction. The photo-generated electrons, exhibiting strong reducing capability in the presence of H₂O, and with their reducing ability dependent on the reduction potential and the number of the transferred electrons, caused the reduction of CO₂ to HCOOH. Based on our obtained experimental results and reports in the literature,^54–55 the possible reaction mechanism for the photocatalytic reduction of CO₂ in aqueous solution is suggested in Fig. 8. The CO₂ is first absorbed on the BiYO₃ surface and becomes a distorted molecule. The C=O bond of absorbed CO₂ becomes long and its π bond is activated, thus, the activated C=O bond is attacked by an electron from the conduction band of BiYO₃ to form ·CO₂⁻. The reaction of H⁺ with an electron generates ·H, and the ·CO₂⁻ by reaction with ·H is subsequently transformed to HCOOH.

Fig. 8 Possible mechanism for the reduction of CO₂ to formic acid.

4. Conclusions

Three templates, including SDBS, EDTA, and CA, were added to prepare BiYO₃ using a hydrothermal method. The effects of templates on the structure and the photocatalytic reduction properties of BiYO₃ were studied. The results show that, the surface area and band gap of BiYO₃ prepared with no template and using SDBS, EDTA, and CA as templates were 6.75, 15.62, 20.87, 12.20 m² g⁻¹ and 2.99, 2.84, 2.64, and 2.86 eV, respectively. The addition of templates affected the production of ·OH, and the ·OH concentration was the lowest when using the BiYO₃ catalysts prepared by EDTA. When using BiYO₃ prepared by EDTA as a photocatalyst for the photocatalytic reduction of CO₂, the HCOOH yield was 1.68 μmol L⁻¹, 1.9 times higher than that produced using the BiYO₃ prepared with no template, and 2.6 times higher than that of BiYO₃ catalysts prepared by a solid-state reaction, due to the higher surface area and lower band gap, and the accelerated mass transfer rate of the photogenerated electrons and the holes.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (21366004, 21425627), Guangxi Zhuang Autonomous Region special funding of distinguished experts and the Open Project of Guangxi Key Laboratory of Petrochemical Resource Processing and Process Intensification Technology (2015K004).

References

1. B. K. Sovacool and P. L. Blyth, Environ. Sci. Policy, 2015, 54, 304–315.
2. S. Zheng, H. Yi and H. Li, Renewable Sustainable Energy Rev., 2015, 49, 386–394.
3. D. Suryaman and K. Hasegawa, J. Hazard. Mater., 2010, 183, 490–496.
4. M. Murugananthan, M. Kumaravel, H. Katsumata, T. Suzuki and S. Kaneco, Int. J. Hydrogen Energy, 2015, 40, 6740–6744.
5. Y. Zhang, C. Xu, J. Chen, X. Zhang, Z. Wang, J. Zhou and K. Cen, Appl. Energy, 2015, 156, 223–229.
6. J. Mao, T. Peng, X. Zhang, K. Li and L. Zan, Catal. Commun., 2012, 28, 38–41.
7. J.-H. Kim, F. Nishimura, S. Yonezawa and M. Takashima, J. Fluorine Chem., 2012, 144, 165–170.
8. T.-m. Su, Z.-z. Qin, H.-b. Ji, Y.-x. Jiang and G. Huang, Environ. Chem. Lett., 2016, 14, 99–112.
9. M. Halmann, Nature, 1978, 275, 115–116.
10. T. Inoue, A. Fujishima, S. Konishi and K. Honda, Nature, 1979, 277, 637–638.
11. K. KoČÍ, K. ZatloukalovÁ, L. ObalovÁ, S. KrejČÍKovÁ, Z. LacnÝ, L. ČApek, A. HospodkovÁ and O. ŠOlcovÁ, Chin. J. Catal., 2011, 32, 812–815.
12. J. Wang, G. Ji, Y. Liu, M. A. Gondal and X. Chang, Catal. Commun., 2014, 46, 17–21.
13. H. Abdullah, M. R. Khan, M. Pudukudy, Z. Yaakob and N. A. Ismail, J. Rare Earths, 2015, 33, 1155–1161.
14. P.-W. Pan and Y.-W. Chen, Catal. Commun., 2007, 8, 1546–1549.
15. X. Li, H. Pan, W. Li and Z. Zhuang, Appl. Catal., A, 2012, 413–414, 103–108.
16. S. C. Yan, S. X. Ouyang, J. Gao, M. Yang, J. Y. Feng, X. X. Fan, L. J. Wan, Z. S. Li, J. H. Ye, Y. Zhou and Z. G. Zou, Angew. Chem., Int. Ed., 2010, 49, 6400–6404.
17. J. Mao, T. Peng, X. Zhang, K. Li and L. Zan, Catal. Commun., 2012, 28, 38–41.
18. R. a. He, S. Cao, P. Zhou and J. Yu, Chin. J. Catal., 2014, 35, 989–1007.
19. J. H. Sun and H. Yang, Ceram. Int., 2014, 40, 6399–6404.
20. M. Zhang, C. Shao, P. Zhang, C. Su, X. Zhang, P. Liang, Y. Sun and Y. Liu, J. Hazard. Mater., 2012, 225–226, 155–163.
21. W. Dai, H. Xu, J. Yu, X. Hu, X. Luo, X. Tu and L. Yang, Appl. Surf. Sci., 2015, 356, 173–180.
22. Z.-z. Qin, Z.-l. Liu, Y.-b. Liu and K.-d. Yang, Catal. Commun., 2009, 10, 1604–1608.
23. M.-g. Fan, Z.-z. Qin, Z.-l. Liu and T.-m. Su, Adv. Mater. Res., 2011, 287–290, 1640–1645.
24. L.-m. Xue, F.-h. Zhang, H.-j. Fan, B. Chen and X.-f. Bai, Min. Metall. Eng., 2011, 31, 84–87.
25. T.-m. Su, Z.-l. Liu, Y. Liang, Z.-z. Qin, J. Liu and Y.-q. Huang, Catal. Commun., 2012, 18, 93–97.
26. W.-t. Cheng, Z.-b. Li and J.-j. Ke, Chin. J. of Nonferrous Met., 2008, 18, 230–235.
27. S. W. Jiang, L. Yang, J. N. Pang, H. Lin and Z. Q. Wang, Surf. Coat. Technol., 2016, 286, 197–205.
28. R. Marsalek, APCBEE Proc., 2014, 9, 13–17.
29. Y. Chen, K.-c. Zhou, S.-p. Huang, Z.-y. Li and G.-c. Liu, Chin. J. of Nonferrous Met., 2011, 21, 1570–1579.
30. A. M. Dumitrescu, A. I. Borhan, A. R. Iordan, I. Dumitru and M. N. Palamaru, Powder Technol., 2014, 268, 95–101.
31. A. T. Kozakov, A. G. Kochur, A. V. Nikolsky, K. A. Googlev, V. G. Smotrakov and V. V. Eremkin, J. Electron Spectrosc. Relat. Phenom., 2011, 184, 508–516.
32. S. Poulston, N. J. Price, C. Weeks, M. D. Allen, P. Parlett, M. Steinberg and M. Bowker, J. Catal., 1998, 178, 658–667.
33. P. S. Bagus, E. S. Ilton and C. J. Nelin, Surf. Sci. Rep., 2013, 68, 273–304.
34. A. Fujimori and L. Schlapbach, J. Phys. C: Solid State Phys., 1984, 17, 341–351.
35. V. Abdelsayed, S. Moussa, H. M. Hassan, H. S. Aluri, M. M. Collinson and M. S. El-Shall, J. Phys. Chem. Lett., 2010, 1, 2804–2809.
36. L. Hu, Z. Du, X. Tai, Q. Li and Y. Zhao, Chin. J. Catal., 2008, 29, 571–576.
37. T.-m. Su, Z.-z. Qin, H.-b. Ji and Y.-x. Jiang, Int. J. Photoenergy, 2014, 2014, 794057.
38. M. R. Hoffmann, S. T. Martin, W. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69–96.
39. K. Sutthiumporn and S. Kawi, Int. J. Hydrogen Energy, 2011, 36, 14435–14446.
40. J. L. Mendoza de la Cruz, I. V. Castellanos-Ramírez, A. Ortiz-Tapia, E. Buenrostro-González, C. d. l. A. Durán-Valencia and S. López-Ramírez, Colloids Surf., A, 2009, 340, 149–154.
41. H. Qi, J. Ma and P. Z. Wong, Colloids Surf., A, 2002, 206, 401–407.
42. M. R. Housaindokht and A. Nakhaei Pour, Solid State Sci., 2012, 14, 622–625.
43. Z. C. Orel, M. K. Gunde and B. Orel, Prog. Org. Coat., 1997, 30, 59–66.
44. N. Guangdi, Z. Liang, L. Xiaofeng, B. Xiujie, S. Weining and W. Ce, Dalton Trans., 2013, 42, 14006–14013.
45. N. Tian, Y. Zhang, H. Huang, Y. Guo and Y. He, Appl. Surf. Sci., 2014, 322, 249–254.
46. W. Li, D. Du, T. Yan, D. Kong, J. You and D. Li, J. Colloid Interface Sci., 2015, 444, 42–48.
47. C.-Y. Chang, Y.-H. Hsieh, L.-L. Hsieh, K.-S. Yao and T.-C. Cheng, J. Hazard. Mater., 2009, 166, 897–903.
48. R. Lamba, A. Umar, S. K. Mehta, W. A. Anderson and S. K. Kansal, J. Mol. Catal. A: Chem., 2015, 408, 189–201.
49. R. J. Isaifan, S. Ntais and E. A. Baranova, Appl. Catal., A, 2013, 464–465, 87–94.
50. W. Dai, H. Xu, J. Yu, X. Hu, X. Luo, X. Tu and L. Yang, Appl. Surf. Sci., 2015, 356, 173–180.
51. A. Nakada, K. Koike, T. Nakashima, T. Morimoto and O. Ishitani, Inorg. Chem., 2015, 54, 1800–1807.
52. F. Yoshitomi, K. Sekizawa, K. Maeda and O. Ishitani, ACS Appl. Mater. Interfaces, 2015, 7, 13092–13097.
53. I. U. Gaya, in Heterogeneous Photocatalysis Using Inorganic Semiconductor Solids, Springer Netherlands, Dordrecht, 2014, pp. 43–71,  DOI:10.1007/978-94-007-7775-0_2.
54. D. Uner and M. M. Oymak, Catal. Today, 2012, 181, 82–88.
55. G. Qin, Y. Zhang, X. Ke, X. Tong, Z. Sun, M. Liang and X. Song, Appl. Catal., B, 2013, 129, 599–605.

---