CO₂ Reduction Performance with Double-Layered Cu/TiO₂ and P₄O₁₀/TiO₂ as Photocatalysts under Different Light Illumination Conditions

Abstract

This paper presents an experimental study of using a double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ as photocatalysts for CO₂ reduction with an extended wavelength of range light from ultraviolet light (UV) to infrared light (IR). The lights studied were UV + visible light (VIS) + IR, VIS + IR and IR only. This study also investigated the impact of the molar ratio of CO₂:H₂O on the CO₂ reduction performance. This study revealed that the optimum molar ratio of CO₂:H₂O to produce CO was 1:1, irrespective of light illumination condition, which matched the theoretical molar ratio to produce CO according to the reaction scheme of CO₂ reduction with H₂O. Comparing the results of double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ with those of double-layered TiO₂ obtained under the UV + VIS + IR light illumination condition, the highest concentration of formed CO and the molar quantity of formed CO per unit weight of the photocatalyst increased by 281 ppmV and 0.8 μmol/g, in the case of the molar ratio of CO₂:H₂O = 1:1. With IR-only illumination, the highest concentration of formed CO and the molar quantity of CO formed per unit weight of the photocatalyst was 251 ppmV and 4.7 μmol/g, respectively.

1. Introduction

Global warming is a present problem–challenge currently facing the whole world. Photocatalytic CO₂ reduction, if developed as a potential technology, would assist in confronting this challenge. CO₂ can be converted, with the help of photocatalysts, from CO₂ into fuel, e.g., CO, CH₄, CH₃OH, etc. [1,2,3]. This study adopted TiO₂ as the photocatalyst. TiO₂ is the most widely investigated photocatalyst due to its chemical stability, abundance, low cost, and low toxic characteristics [4]. However, TiO₂ has some weak points. TiO₂ absorbs ultraviolet light (UV) only, which accounts for only 5% in sunlight [5]. The sunlight reaching the earth consists of UV, visible light (VIS), and infrared light (IR) which account for 5%, 45% and 50%, respectively [5]. This study aimed to develop new modified TiO₂ photocatalysts which could absorb the light from UV to IR.

According to previous studies [6,7,8,9,10,11,12,13,14,15,16,17,18,19,20], many approaches to extend the light absorption of photocatalysts from UV to VIS have been conducted. Metal doping is one of the popular approaches to extend the light absorption of photocatalysts from UV to VIS [6,7]. Though some metals such as Pt [8], Au [9] and Pd [10,11,12] have been investigated, this study focuses on Cu, which is abundant and has a low cost, so it might be easy to apply to industrial needs. According to the review paper [13], Cu/TiO₂ displayed good improvements compared to TiO₂. This phenomenon occurs since its estimated band gap is relatively the same compared to TiO₂ (3.1 eV), and Cu does not shift the absorption spectrum of TiO₂. Other previous studies have reported the impact of the amount of Cu loaded on TiO₂ on the performance of CO₂ reduction [14]. The highest production rate of CH₄ of 1.284 μmol/(g·h) and that of CO of 9.913 μmol/(g·h) were obtained for the weight amount of Cu of 1 wt% under the condition of CO₂/H₂O. When increasing the weight amount of Cu, the performance of photocatalytic CO₂ reduction decreased. Cu₂O/TiO₂ exhibited a production rate of CO, which is 20 times as large as that of TiO₂ under the condition of CO₂/H₂O [15]. Cu₂O also showed the photocatalytic performance by itself [16]. The C₂H₆ formation of 10 μmol/g was performed by Cu₂O, which worked under the light illumination condition of VIS in the case of CO_2/H₂O. The authors’ previous studies have reported the development of Cu/TiO₂ with a pulse arc plasma method to load a fine Cu particle after preparing TiO₂ film coated on a netlike glass disc [17,18]. According to these reports [17], Cu/TiO₂ exhibited the CO₂ reduction performance under the light illumination condition with VIS. Changing the pulse numbers of the fine Cu particles by 100, 200, and 500, the highest molar quantity of CO per weight of the photocatalyst obtained was 12.0 μmol/g for the pulse number of 200 under the light illumination condition without UV. Though the CO₂ reduction performance is promoted with the increase in the loading weight of Cu, it is thought that too much Cu loading covers the surface of the TiO₂ film [18,19]. It indicates that CO₂ and reductants cannot reach the surface of TiO₂ film sufficiently due to the optimum loading of Cu. According to previous studies [20], a Z-scheme heterojunction Cu₂O/PCN-250 photocatalyst has been utilized for photocatalytic CO₂ reduction with H₂O vapor. The significant reduction in the Cu²⁺ content means that the Cu²⁺ species accept photogenerated electrons for the production of Cu⁺ during the light irradiation. The Fermi energy level of Cu₂O nanoparticles is higher than that of PCN-250, which can lead to interlaced energy levels between two components, providing the possibility of constructing a Z-type heterojunction. From other previous studies [21], Co-MOF/Cu₂O heterojunction has conducted for photocatalytic CO₂ reduction with H₂O. The Schottky barrier can prevent the back transfer of injected electrons. Hence, under visible-light irradiation, the excited electrons in the CB of Cu₂O can quickly move to the CB of C_O-MOF which is above the Ev (CO₂/CO, —0.53 V vs. NHE) by the p-n junction induced electric field. Then, the photoexcited electrons remaining in the CB of Co-oxo clusters in Co-MOF react with the absorbed CO₂ molecules to produce CO₂, and then the photoinduced holes from the VB of Co-MOF move to Cu₂O. According to previous research, to extend the light absorption of the photocatalyst from UV to IR or near IR [22,23,24,25], many trials have been reported. W₁₈O₄₉/g-C₃N₄ composite exhibited the production performance of CO of 45 μmol/g and CH₄ of 28 μmol/g under the light illumination condition when the wavelength ranged from 200 nm to 2400 nm [22]. WS₂/Bi₂S₃ nanotube exhibited the production performance of CH₃OH of 28 μmol/g and C₂H₅OH of 25 μmol/g under the light illumination condition when the wavelength ranged from 420 nm to 1100 nm [23]. CuInZnS decorated g-C₃N₄ exhibited the production performance of CO of 38 μmol/g under the light illumination condition when the wavelength ranged from 200 nm to 1000 nm [24]. Hierarchical ZnIn₂S₄ nanorods, prepared using the solvothermal method, exhibited the production performance of CO of 54 μmol/g and CH₄ of 9 μmol/g [25]. According to the authors’ previous studies [26], P₄O₁₀/TiO₂ has been investigated. The largest molar quantity of the CO per unit weight of the photocatalyst for P₄O₁₀/TiO₂ film in the case of CO₂/H₂O was 2.36 μmol/g under the light illumination condition with IR, while that in case of CO₂/NH₃ was 33.4 μmol/g [26].

This study introduced metal doping to explain the basis for adopting Cu and P₄O₁₀ as a dopant for TiO₂ in this study. Cu is a promising dopant for expanding the light absorption to VIS since many previous studies have reported its good performance [13,14,15,16,17,18]. In addition, P₄O₁₀ is also a promising dopant to expand the light absorption to IR since previous studies carried out by authors have reported its good performance [26]. However, there is no study to clarify the impact of the combination of the photocatalyst absorbing the light with VIS as well as that with IR. If we could develop a photocatalyst absorbing the light ranging from UV to IR, it could then be able to utilize the sunlight for the photocatalytic CO₂ reduction in near future. The aim of this study is to clarify the impact of double-layered photocatalysts consisting of Cu/TiO₂ and P₄O₁₀/TiO₂ on the CO₂ reduction performance under the different light illumination conditions. This study adopts H₂O as a reductant for CO₂ reduction. The molar ratio of CO₂:H₂O is also changed and the impact on the CO₂ reduction performance is investigated. According to previous review papers [6,27], H₂O is adopted generally as a reductant. It is necessary to clarify the optimum molar ratio of CO₂:H₂O since H⁺ is needed for the reduction reaction to enhance the CO₂ reduction performance. From past studies [28,29,30], the reaction scheme of CO₂ reduction with H₂O can be explained as follows:

• Electron-hole pair generation process

TiO₂ + hν → h⁺ + e⁻

(1)

• Oxidization reaction process

H₂O + h⁺ → OH + H⁺

(2)

·OH + H₂O + h⁺ → O₂ + 3H⁺

(3)

• Reduction reaction process

CO₂ + 2H⁺ + 2e⁻ → CO + H₂O

(4)

CO₂ + 8H⁺ + 8e⁻ → CH₄ + 2H₂O

(5)

This study also compared the CO₂ reduction performance of double-layered photocatalysts consisting of Cu/TiO₂ and P₄O₁₀/TiO₂ with that of double-layered TiO₂. The light illumination, with a wavelength ranging from 185 nm to 2000 nm, 401 nm to 2000 nm, and 801 nm to 2000 nm, respectively, is changed. In addition, the molar ratio of CO₂:H₂O is changed by 1:0.5, 1:1, 1:2, and 1:4. The pulse number of Cu for preparing Cu/TiO₂ was 200, which followed the optimum pulse number decided by the authors’ previous study [17]. P₄O₁₀/TiO₂ has been prepared by sol-gel and the dip-coating processes following the authors’ previous studies [26].

2. Results and Discussion

2.1. Characterization of Cu/TiO₂ and P₄O₁₀/TiO₂

Figure 1 shows SEM and EPMA images of Cu/TiO₂, which was coated on a netlike glass disc. The black and white SEM images at 1500 times magnification were used for EPMA analysis. As for the EPMA images, the concentrations of each element in the observation area are indicated by the diverse colors. If the amount of element was large, light colors such as white, pink, and red were shown. On the other hand, if the amount of element was small, dark colors such as black and blue were shown. According to Figure 1, it is seen that TiO₂ film has a tooth-like shape, coated on the netlike glass fiber. The reason this was caused was thought to be that the temperature distribution of the TiO₂ solution adhered on the netlike glass disc was not even during the firing process due to the different thermal conductivities of Ti and SiO₂ at 600 K, which were 19.4 W/(m·K) and 1.82 W/(m·K), respectively [31]. Due to the thermal expansion and shrinkage around the netlike glass fiber, a thermal crack formed within the TiO₂ film. As a result, the TiO₂ film coated on the netlike glass fiber was tooth-like. As for Cu, it is seen that the nano-sized Cu particles are loaded on TiO₂ uniformly, since nano-sized Cu particles are emitted by the pulse arc plasma gun process. The center part of the netlike glass disc, with a diameter of 300 μm, was analyzed by EPMA for the measurement of the amount of loaded Cu within TiO₂ film. The ratio of Cu to Ti within Cu/TiO₂ was counted as 8.3 wt%.

Figure 2 shows SEM and EPMA images of P₄O₁₀/TiO₂, which was coated on the netlike glass disc. The black and white SEM images at 1500 times magnification were used for EPMA analysis. According to Figure 2, it can be seen that the TiO₂ film had a tooth-like shape coated on the netlike glass fiber as well, which was caused by the same reason explained before for Cu/TiO₂. In addition, it can be seen that the nano-sized P₄O₁₀ particles were loaded on the TiO₂ film. The center part of the netlike glass disc, which had a diameter of 300 μm, was analyzed by EPMA for the amount of loaded Cu within the TiO₂ film. The ratio of Cu to Ti within Cu/TiO₂ was counted by 6.4 wt%. On the other hand, the total weight of Cu/TiO₂ and P₄O₁₀/TiO₂ was measured by an electron balance which was 0.255 g.

This study has not tested PXRD patterns of prepared photocatalysts, but the crystal type of the TiO₂ photocatalyst made was thought to be anatase, since the photocatalyst was prepared at the controlled firing temperature of 623 K. According to the reference presented in [32], the crystal type of the TiO₂ photocatalyst would be anatase when prepared below 973 K.

2.2. CO₂ Reduction Characteristics of Double-Layered TiO₂ under the Light Illumination Condition of Xe Lamp with UV + VIS + IR

Figure 3 and Figure 4 exhibit the change of concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst with time, respectively. For the expression in the unit of μmol/g in Figure 4, the total weight of TiO₂ in double layers was measured by an electron balance, which was 0.136 g. In this study, the CO₂ reduction experiment was conducted without the introduction of H₂O and the CO₂ reduction experiment under no Xe lamp illumination conditions and the CO₂ reduction experiment with H₂O without a photocatalyst were conducted. As a result, no fuel was detected.

Figure 3 and Figure 4 show the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst of 262 ppmV and 9.5 μmol/g, respectively, which were obtained with the molar ratio of CO₂:H₂O = 1:1. This optimal molar ratio agrees with the theoretical molar ratio to produce CO as shown in Equations (1)–(5). The production rate of CO was saturated after 8 h of the illumination from the Xe lamp. The peak of the production rate of CO was observed at 2 h.

2.3. CO₂ Reduction Characteristics of Double-Layered Cu/TiO₂ and P₄O₁₀/TiO₂ with the Illumination of UV + VIS + IR

Figure 5 illustrates the double-layered Cu/TiO₂ and P₄O₁₀/TiO₂, which also shows the image of double-layered TiO₂ for readers to better understand the experimental condition. Figure 6 and Figure 7 show the concentration change of formed CO and the molar quantity of CO per unit weight of the photocatalyst with time, respectively. The data obtained under the light illumination condition with UV + VIS + IR are shown in these figures. In Figure 7, the total weight of double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ was measured by an electron balance, which was 0.255 g. In this study, the CO₂ reduction experiment without the introduction of H₂O, the CO₂ reduction experiment under no Xe lamp illumination condition, and the CO₂ reduction experiment with H₂O without a photocatalyst were conducted. As a result, no fuel was detected.

According to Figure 6 and Figure 7, the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst were obtained in the case of the molar ratio of CO₂:H₂O = 1:1. The highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst is 543 ppmV and 10.3 μmol/g, respectively. Compared to the results of double-layered TiO₂ shown in Figure 3 and Figure 4, the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst increased by 281 ppmV and 0.8 μmol/g, respectively, with the optimal molar ratio of CO₂:H₂O = 1:1. Due to the improvement of light absorption performance with the aid of loading Cu and P₄O₁₀, it is thought that the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst increased, respectively. Since Cu/TiO₂ was coated on both surfaces of the netlike glass disc, it is thought that the electron emitted from Cu/TiO₂ could reach P₄O₁₀/TiO₂, which was located under Cu/TiO₂. Additionally, it is thought that the balance of electrons and H⁺, which clarifies the optimum molar ratio according to the reaction scheme as shown in Equations (1)–(5), was not influenced by loading Cu and P₄O₁₀ on TiO₂ in this study. Therefore, the optimum molar ratio using double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ with UV + VIS + IR illumination in this study follows the theoretical molar ratio to produce CO as shown in Equations (1)–(5).

2.4. CO₂ Reduction Characteristics of Double-Layered Cu/TiO₂ and P₄O₁₀/TiO₂ with VIS + IR Illumination

Figure 8 and Figure 9 show the concentration change of formed CO and the molar quantity of CO per unit weight of the photocatalyst with VIS + IR illumination along time, respectively. The total weight of double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ was measured by an electron balance before the experiment, which was 0.255 g.

According to Figure 8 and Figure 9, the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst are obtained with the molar ratio of CO₂:H₂O = 1:1. The highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst was 373 ppmV and 7.1 μmol/g, respectively, as shown in Figure 8 and Figure 9. There was no CO found in the reference experiment conducted, which was with the double-layered TiO₂ and VIS + IR. Therefore, loading Cu and P₄O₁₀ on TiO₂ indeed provided the photocatalyst the absorption capability of VIS + IR. In addition, it also revealed that the optimum molar ratio with double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ and VIS + IR followed the theoretical molar ratio to produce CO as shown in Equations (1)–(5).

In the reference presented in [33], Cu/TiO₂ was prepared by the impregnation with aqueous solution of copper acetate, followed by reduction with sodium borohydride. The preparation procedure might influence the chemical/electrochemical state of Cu, which decided the absorption spectrum of TiO₂ and the spectrum of TiO₂ action in the oxidization reaction. Since this preparation procedure was based on a chemical reaction, it is different from the pulse arc plasma gun method adopted in this study. The pulse arc plasma gun method loads Cu fine particles on the surface of TiO₂ physically. Therefore, the role/state played by Cu to TiO₂ was different from that in the reference presented in [33].

2.5. CO₂ Reduction Characteristics of Double-Layered Cu/TiO₂ and P₄O₁₀/TiO₂ with IR Only Illumination

Figure 10 and Figure 11 show the concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst with IR only illumination with time, respectively. The total weight of double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ was measured by an electron balance, which was 0.255 g.

According to Figure 10 and Figure 11, the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst were obtained when CO₂:H₂O = 1:1. The highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst was 251 ppmV and 4.7 μmol/g, respectively. There was no CO found in the reference experiment conducted, which was with the double-layered TiO₂ and IR illumination. Therefore, the CO₂ reduction reaction could occur with IR only illumination by loading Cu and P₄O₁₀ into the TiO₂. In addition, the optimal molar ratio between CO₂ and H₂O was also 1:1 for the reaction with IR only, which is the same as with UV + VIS + IR and that with VIS + IR.

When compared to previous studies [22,23,24,25,26], the CO₂ reduction performance with IR only illumination obtained in this study was not high. However, the concept was thought-provoking, i.e., by doping Cu and P₄O₁₀ the wavelength of illumination required by CO₂ reduction reaction could be extended to IR range.

3. Experimental Procedure

3.1. Preparation Procedure of Cu/TiO₂ and P₄O₁₀/TiO₂

The TiO₂ film was prepared by sol-gel and dip-coating processes [34]. [(CH₃)₂CHO]₄Ti (purity of 95 wt%, Nacalai Tesque Co., Ltd., Kyoto, Japan) of 0.3 mol, anhydrous C₂H₅OH (purity of 99.5 wt%, Nacalai Tesque Co., Ltd.) of 2.4 mol, distilled water of 0.3 mol, and HCl (purity of 35 wt%, Nacalai Tesque Co., Ltd.) of 0.07 mol were mixed to prepare TiO₂ sol solution. The TiO₂ film was coated on netlike glass fiber (SILIGLASS U, Nihonmuki Co., Ltd., Yuki City, Japan) by means of sol-gel and dip-coating processes. The glass fiber, with a diameter of approximately 10 μm, weaved as a net, was collected to be at a diameter of about 1 mm. The porous diameter of the glass fiber and the specific surface area is approximately 1 nm and 400 m²/g, respectively, from the specification of the netlike glass fiber. The netlike glass fiber was composed of SiO₂, whose purity was 96 wt%. The opening space of the net glass is approximately 2 mm × 2 mm. The netlike glass fiber with the porous characteristics could capture TiO₂ film easily during the sol-gel and dip-coating processes. In addition, CO₂ was also more easily absorbed by the prepared photocatalyst due to the porous characteristics. In this study, the netlike glass fiber was cut to be a disc shape with a diameter and the thickness of 50 mm and 1 mm, respectively. The netlike glass disc was dipped into the TiO₂ sol solution at the speed of 1.5 mm/s and pulled up it at the fixed speed of 0.22 mm/s. After that, the disc was dried out and fired under the controlled firing temperature (FT) and firing duration time (FD) to coat TiO₂ film on the base material. The FT and FD was 623 K and 180 s, respectively, in this study.

After the coating of TiO₂, Cu was loaded on TiO₂ using a pulse arc plasma gun [34], which emitted nano-sized Cu fine particles uniformly with an applied high voltage potential difference of 200 V. Cu fine particles were loaded on both surfaces of TiO₂. The pulse number applied could control the quantity of metal loaded on TiO₂. This study set the pulse number at 200. This study used the pulse arc plasma gun device (ULVAC, Inc., Chigasaki City, Japan, ARL-300), which had a Cu electrode with a diameter of 10 mm. By setting the distance between the Cu electrode and TiO₂ film at 150 mm, Cu particles could be uniformly speared over the TiO₂ film. Figure 12 shows the photo of the final coated Cu/TiO₂ sample.

As to P₄O₁₀/TiO₂, P₄O₁₀ was made from red P by means of mechanical synthesis [34]. The red P, which had an average diameter of 75 μm (Nacalai Tesque) was filled in a ball mill crusher (AV-1, Asahi Rika Factory, Suginami City, Japan) with Al₂O₃ balls. The diameter of the Al₂O₃ balls was 3/8 inches (HD-10, NIKKATO COPORATION, Tokyo, Japan). The weight ratio of Al₂O₃ balls to red P particles in the ball mill crusher was set at 20 [34]. The rotation speed of 600 rpm was kept for 12 h, and P₄O₁₀ was identified to be one type of oxidized P by XPS (PHI Quantera SXM, ULVAC PHI Inc., Chigasaki City, Japan) [26]. The prepared P₄O₁₀ particles were put into the TiO₂ sol solution and were mixed with the TiO₂ sol solution by a magnetic stirrer for 60 min. After that, the netlike glass disc was immersed into this mixed solution. The following dipping and pulling process was the same as explained above. Figure 13 shows the photo of P₄O₁₀/TiO₂ coated on the netlike glass disc.

3.2. The Characterization Procedure of Cu/TiO₂ and P₄O₁₀/TiO₂

The characteristics of the external and crystal structure of Cu/TiO₂ and P₄O₁₀/TiO₂ film prepared above were evaluated by SEM (JXA-8530F, JEOL Ltd., Akishima City, Japan) and EPMA (JXA-8530F, JEOL Ltd.) [35]. Since the netlike glass disc adopted for base material to coat Cu/TiO₂ and P₄O₁₀/TiO₂ cannot conduct electricity, the vaporized carbon was deposited by the carbon deposition device (JEE-420, JEOL Ltd.) on the surface of Cu/TiO₂ and P₄O₁₀/TiO₂ before analyzing the characterization. The thickness of the deposited carbon was about 2030 nm. The acceleration voltage and the current were set at 15 kV and 3.0 × 10⁻⁸ A, respectively, in order to analyze the external structure.

3.3. The Experimental Procedure of CO₂ Reduction

Figure 14 illustrates the experimental apparatus, which is as follows: the reactor, composed of a stainless tube of 100 mm (H.) × 50 mm (I.D.); Cu/TiO₂, which is located over the P₄O₁₀/TiO_2, coated on a netlike glass disc of 50 mm (D.) × 1mm (t.) positioned on the Teflon cylinder of 84 mm (D.) × 10 mm (t.); a sharp cut filter removing the wavelength of light which is below 400 nm (SCF-49.5C-42L, SIGMA KOKI Co., Ltd., Tokyo, Japan); a 150W Xe lamp (L2175, Hamamatsu Photonics K. K., Hamamatsu City, Japan); mass flow controller; and CO₂ gas cylinder (purity of 99.995 vol%) [34]. The size of the reactor chamber is 1.25 × 10⁻⁴ m³. The light of Xe lamp positioned on the stainless tube illuminates toward Cu/TiO₂ and P₄O₁₀/TiO₂ passing the sharp cut filter and the quartz glass disc located on the top of the stainless tube. The wavelength of the Xe lamp light ranges from 185 nm to 2000 nm. The sharp cut filter can eliminate UV from the Xe lamp light, resulting in the wavelength of light after the filter ranging from 401 nm to 2000 nm or from 801 nm to 2000 nm. Figure 15 exhibits the light transmittance characteristics after the sharp cut filter to demonstrate, as an example, the removal of the wavelength of light under 400 nm [34]. Though this study has not taken the data on the absorption of light by synthesized catalyst samples, this study exhibits the light transmittance data of the sharp cut filter cutting the wavelength below 400 nm as shown in Figure 15, supporting the UV light illumination condition without UV, i.e., VIS + IR. The authors think this could be evidence for the light absorption range of prepared photocatalysts.

The mean light intensity of light from the Xe lamp without the cut filter was 72.0 mW/cm² in the wavelength ranging from 185 nm to 2000 nm, while that with the filter was 60.4 mW/cm² for the wavelength between 401 nm and 2000 nm. Additionally, the mean light intensity of light was 51.2 mW/cm² in the wavelength ranging from 801 nm to 2000 nm with the sharp cut filter. The light intensities above were measured by the light intensity meter located 55 mm away from the lamp, which was the distance between the Xe lamp and the prepared photocatalyst during the CO₂ reduction experiment. Unfortunately, it was impossible to measure the reflect radiation from the walls of reactor by this light intensity meter during the experiment. The light intensity spectrum data of the Xe lamp used in this study, according to the brochure of Hamamatsu Photonics Corp. [36], is shown in Figure 16.

CO₂ gas with the purity of 99.995 vol% was filled in the vacuumed reactor chamber. After that, the valves installed at the inlet and the outlet of reactor were closed during the experiment. The pressure was 0.1 MPa and the gas temperature was set at 298 K in the reactor. Liquid H₂O was injected into the reactor via the gas sampling tap and the Xe lamp was tuned on at the same time. The amount of injected H₂O was changed following the set molar ratio of CO₂:H₂O. The injected H₂O was vaporized because of the heat of the infrared light components illuminated by the Xe lamp. The temperature in the reactor attained 343 K within an hour, and it was kept at 343 K during the experiment. The molar ratio of CO₂:H₂O was changed by 1:0.5, 1:1, 1:2 and 1:4. The reacted gas filled in the reactor was extracted busing a gas syringe via gas sampling tap and it was analyzed by an FID gas chromatograph (GC353B, GL Science, Tokyo, Japan) and a methanizer (MT221, GL Science). The minimum resolution of the FID gas chromatograph and methanizer is 1 ppmV. Regarding the recyclability of photocatalysts, the experiment was conducted three times for the double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ in this study.

4. Conclusions

The impact of double-layered photocatalysts consisting of Cu/TiO₂ and P₄O₁₀/TiO₂ on the CO₂ reduction performance under different light illumination conditions was studied in this paper. The impact of the molar ratio of CO₂:H₂O on the CO₂ reduction performance was also investigated. As a result, the following conclusions were drawn:

(i)

The highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst were achieved when the molar ratio of CO₂:H₂O = 1:1, which matched the theoretical molar ratio to produce CO according to the reaction scheme of CO₂ reduction with H₂O, irrespective of the light illumination conditions.

(ii)

Comparing the results of double-layered Cu/TiO₂ and P₄O₁₀/TiO₂ with the results of double-layered TiO₂ obtained under the UV + VIS + IR illumination, the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst increased by 281 ppmV and 0.8 μmol/g, respectively, when CO₂:H₂O = 1:1.

(iii)

With VIS + IR illumination, the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst reached 373 ppmV and 7.1 μmol/g, respectively. This proves that CO₂ reduction could be activated with VIS + IR illumination if TiO₂ loaded with Cu and P₄O₁₀ was used as the photocatalyst.

(iv)

With only IR illumination, the highest concentration of formed CO and the molar quantity of CO per unit weight of the photocatalyst was 251 ppmV and 4.7 μmol/g, respectively. This proves that CO₂ reduction reaction could be activated with IR only illumination if TiO₂ loaded with Cu and P₄O₁₀ was used as the photocatalyst.