The Impact of Amount of Cu on CO 2 Reduction Performance of Cu/TiO 2 with NH 3 and H 2 O

: This study has investigated the impact of molar ratio of CO 2 to reductants NH 3 and H 2 O as well as that of Cu loading on CO 2 reduction characteristics over Cu/TiO 2 . No study to optimize the reductants’ combination and Cu loading weight in order to enhance CO 2 reduction performance of TiO 2 has been investigated yet. This study prepared Cu/TiO 2 film by loading Cu particles during the pulse arc plasma gun process after coating TiO 2 film by the sol-gel and dip-coating process. As to loading weight of Cu, it was regulated by change in the pulse number. This study characterized the prepared Cu/TiO 2 film by SEM and EPMA. Additionally, the performance of CO 2 reduction has been investigated under the illumination condition of Xe lamp with or without ultraviolet (UV) light. It is revealed that the molar ratio of CO 2 /NH 3 /H 2 O is optimized according to the pulse number. the amount of H + is the same as of electron produce CO CO with H 2 O or NH 3 , larger H + is needed with the increase in the pulse number. It is revealed that Cu of 4.57 wt% for the pulse number of 200 is the optimum condition, whereas the molar quantity of CO per unit weight of Cu/TiO 2 with and without UV light illumination is 34.1 mol/g and 12.0 mol/g, respectively. OH (purity of 99.5 wt%, produced by Nacalai Tesque Co., Kyoto, Japan) of 2.4 mol, distilled water of 0.3 mol, and HCl (purity of 35 wt%, produced by Nacalai Tesque Co., Kyoto, Japan) of 0.07 mol were mixed for preparing the TiO 2 sol solution. This study coats the TiO 2 film on a netlike glass fiber (SILIGLASS U, produced by Nihonmuki Co., Tokyo, Japan) by sol-gel and dip-coating processes. The glass fiber, having a diameter of about 10 m, weaved as a net, is collected to be the diameter of approximately 1 mm. The porous diameter of glass fiber and the specific surface area are about 1 nm and 400 m 2 /g, respectively from the specifications of netlike glass fiber. The netlike glass fiber is composed of SiO 2 of 96 wt%. The opening space of the netlike glass fiber is approximately 2 mm × 2 mm. Since the netlike glass fiber has porous characteristics, the netlike glass fiber can capture the TiO 2 film easily during sol-gel and dip-coating pro-cesses. Additionally, we can expect that CO 2 is more easily absorbed by the prepared photocatalyst due to the porous characteristics of the netlike glass fiber. This study cut the netlike glass fiber to be disc form having diameter of 50 mm and thickness of 1 mm. This


Introduction
Since the concentration of CO2 in the atmosphere has increased notably since the Industrial Revolution, each country has declared its goal in order to reduce the amount of CO2 emission. In Japan, the prime minister has declared to reduce the effective CO2 emission to zero by 2050. However, the global average concentration of CO2 in the atmosphere increased up to 410 ppmV in September 2020, which is an increase of 25 ppmV since 2009 [1]. Consequently, the break-through technology is required to reduce the amount of CO2 in the world.
It is known that there are some approaches to reduce CO2 emission, e.g., reduction of CO2 emission, CO2 capture and storage (CCS), and CO2 capture and utilization (CCU). This study focuses on photocatalytic CO2 reduction using photocatalyst. CO2 could be converted into fuel species such as CO, CH4, CH3OH, etc., by photocatalyst [2][3][4]. TiO2 is commonly used as the photocatalyst to reform CO2 into fuel species, such as CO, CH4, CH3OH, and H2, etc., with ultraviolet (UV) light [4][5][6], since it is useful, economical, and exhibits strong endurance for chemicals and corrosion [7]. However, pure TiO2 can only work under UV light illumination which accounts for only 4% in sunlight [3]. Therefore, it is not effective to utilize sunlight, which is a renewable energy source. If the visible light, of which 44% of solar energy reaching the earth is [3], it could be utilized for photocatalytic reduction by TiO2, and the CCU system utilizing renewable energy would be constructed.
Many approaches have attempted to enhance the CO2 reduction performance of TiO2 as a photocatalyst by expanding the wavelength of light absorbed by TiO2. One of the popular methods is to dope precious metals, such as Pd [8], Pt [9], Ru, Pt-Ru alloy [10], and Au [11], for preparing the visible light-driven TiO2. A hierarchical pore network and morphology to prepare the bio-templated TiO2 catalyst [12], heteroleptic iridium complex supported on graphite carbon nitride [13], TiO2 synthesis using superficial fluid technology [14], and N-doped reduced graphene oxide-promoted nano TiO2 [15] were attempted to prepare the visible light driven TiO2. Doped various metal ions have been applied, but among them, and this study considers that Cu is a promising dopant. Cu has an ability to absorb a wider wavelength, from 400 to 800 nm [16,17], which can cover most of visible light range. It was reported that Cu/TiO2 had the superiority to pure TiO2. Cu/Cu + fabricated Ti 3+ /TiO2 performed 8 mol/g of CH4 production, which was 2.6 times as large as Ti 3+ /TiO2 [18]. Cu/TiO2 prepared by a facile solvothermal method performed CO and CH4 production up to 4.48 mol/g and 5.34 mol/g, which is 10 times larger compared to that of TiO2 [19]. Cu/TiO2 which was prepared by a sonothermal-hydrothermal route that performed 6.6 mol/g of CH4 and 472.5 mol/g of CH3OH in KOH/H2O medium [20]. It was reported that the synthesized Cu2O/TiO2 showed the performance of 3.5 mol/g of CO production while that of TiO2 was 0.1 mol/g [16]. These results [16,[18][19][20] were achieved under the visible light illumination condition. The other well-known doped metals, e.g., Pd, Ag, and Au, are too precious to be applied in industrial usage. To spread a breakthrough technology to reduce the amount of CO2 in the world, this study thinks that we had better select an economical and abundant material. Due to the reported performances, as well as low cost and large reserves, Cu is thought to be a preferable candidate compared to precious metals.
For the CO2 reduction, a reductant is important as a partner for reaction. According to review papers [6,21], H2O and H2 are generally used as reductant. It is necessary to decide the optimum reductant which provides the proton (H + ) for the reduction reaction to enhance the CO2 reduction performance. From the past studies [22][23][24], the reaction scheme of CO2 reduction with H2O can be shown as below: <Photocatalytic reaction> CO2 + 8H + + 8e − → CH4 + 2H2O (11) Though the previous studies investigated CO2 reduction reacting with H2O or H2 [6,21], the effect of NH3 having 3H + , which is superior to H2O and H2, on photocatalytic CO2 reduction characteristics is not examined yet other than the previous studies conducted by Nishimura et al. using Fe [26] or Cu [27]. The previous study [26] investigated only one combination ratio of CO2, NH3 and H2O for Fe/TiO2 photocatalyst. The other previous study reported the effect of ratio of CO2, NH3 and H2O on CO2 reduction characteristics over Cu/TiO2 [27]. However, the effect of loading weight of Cu on the CO2 reduction characteristics with NH3 and H2O was not reported though the amount of loaded metal is important to improve the CO2 reduction performance using Cu/TiO2 [28]. Therefore, it is necessary to optimize the loading weight of Cu with a different combination condition of NH3 and H2O in order to enhance the CO2 reduction characteristics over Cu/TiO2. As to the reaction scheme to reduce CO2 with NH3, it is as follows [25,29]: <Photocatalytic reaction> <Oxidization reaction> H2 → 2H + + 2e − (14) <Reduction reaction> CO2 + e − → CO2 (16) HCOO -+ H + → CO + H2O The aims of this study are as follows: (1) To reveal the impact of loading weight of Cu on CO2 reduction characteristics using Cu/TiO2. (2) To reveal the effect of molar ratio of CO2 to reductants NH3 and H2O on CO2 reduction characteristics over Cu/TiO2.
After these aims are completed, it is expected that a high-performance photocatalyst, which can absorb the wide-ranging wavelength of light effectively, as well as provide sufficient H + matched with the number of electrons as shown in the reaction scheme, is developed. The key point to complete the aims is to optimize the amount of Cu as well as the combination of CO2, NH3, and H2O. This study focuses on the effective utilization of light energy with the aid of a doped metal and reductants combination. It matches with the scope of this journal which is the photocatalytic CO2 reduction utilizing light energy effectively.
This study investigates the CO2 reduction characteristics reacting with NH3 and H2O over Cu/TiO2 photocatalyst film under the illumination condition of Xe lamp with or without UV light. The combination of CO2/NH3/H2O is changed for 1:1:1, 1:0.5:1, 1:1:0.5, 1:0.5:0.5, 3:2:3, 3:8:12, 3:12:18 to decide the optimum molar ratio for CO2/NH3/H2O. The reaction scheme of CO2 reduction reacting with H2O or NH3 reveals that the theoretical molar ratio of CO2/H2O to produce CO or CH4 is 1:1 or 1:4, respectively, and that of CO2/NH3 to produce CO or CH4 is 3:2 or 3:8, respectively. Consequently, this study assumes that the molar ratios to produce CO and CH4 are CO2/NH3/H2O = 3:2:3 and 3:8:12, respectively, based on the theory. In addition, this study controlled the amount of Cu loaded on TiO2 film during pulse arc plasma gun process. This study changed the pulse number by 100, 200 and 500 to control the loading weight of Cu. Figure 1 represents SEM and EPMA (electron probe microanalyzer) images of Cu/TiO2 film which is coated on netlike glass disc. The SEM and EPMA data of Cu/TiO2 for the pulse number of 100, 200, 500 are shown. This study obtained the black and white SEM images at 1500 times magnification, which were available for EPMA analysis. As to the EPMA image, this study indicates the concentrations of each element in observation area by the diverse colors. When the amount of element is large, light colors, e.g., white, pink, and red are used. On the other hand, dark colors, e.g., black and blue, are used to display small amounts of elements. It is seen from Figure 1 that we can observe the TiO2 film having a teeth-like shape coated on the netlike glass fiber irrespective of pulse number. It is believed that the temperature distribution of TiO2 solution adhered on the netlike glass disc was not even during firing process since the thermal conductivity of Ti and SiO2 at 600 K which are 19.4 W/(m·K) and 1.82 W/(m·K), respectively [30]. Since the thermal expansion and shrinkage around netlike glass fiber occurred, a thermal crack formed within the TiO2 film. Accordingly, the TiO2 film on the netlike glass fiber was teeth-like. As to Cu, we can observe that nanosized Cu particles are loaded on TiO2 uniformly since nanosized Cu particles are emitted by the pulse arc plasma gun process. In addition, the amount of Cu particles increases with the increase in pulse number as expected. The observation area, i.e., the center part of netlike glass disc having a diameter of 300 μm, is analyzed by EPMA in order to measure the amount of loaded Cu within the TiO2 film. The ratio of Cu to Ti is calculated by averaging the data detected in the observation area. The amount of element Cu within Cu/TiO2 film for the pulse number of 100, 200 and 500 are counted by 1.62 wt%, 4.57 wt%, and 7.95 wt%, respectively, indicating that the weight of loaded Cu increases with the increase in the pulse number quantitatively. On the other hand, total weights of Cu/TiO2 for the pulse number of 100, 200, and 500, which were measured by an electron balance, are 0.05 g, 0.06 g, and 0.08 g, respectively.

The CO2 Reduction Characteristics Over Cu/TiO2 for a Pulse Number of 100
Figures 2 and 3 present the concentration change of formed CO with the time under the Xe lamp with or without UV light, respectively. In these figures, the produced CO is evaluated quantitatively by the molar quantity of CO per unit weight of photocatalyst having a unit of mol/g. The other fuels were not detected. As to a blank test, this study conducted the same experiment under no Xe lamp illumination condition as a reference case before the experiment. We detected no fuel during the blank test as we hoped. As to the stability, e.g., repeating use, this has not been tested in this study. As to the reproducibility of experiments, this study shows the data averaging three times experiments. The experimental error through the experiments investigated in this study is distributed approximately from 0% to 20%.
It is revealed from Figure 2 that the CO2 reduction performance for the molar ratio of CO2/NH3/H2O = 1:1:1 is the highest where the molar quantity of CO per unit weight of photocatalyst is 10.2 mol/g. According to the reaction scheme of CO2 reduction reacting with H2O or NH3, the molar ratio of CO2/H2O to produce CO or CH4 is 1:1 or 1:4, respectively, based on the theory. In addition, the theoretical molar of CO2/NH3 to produce CO or CH4 is 3:2, 3:8, respectively. Consequently, this study assumes that the theoretical molar ratios of CO2/NH3/H2O are 3:2:3 and 3:8:12 to produce CO and CH4, respectively. However, it is revealed that the molar ratio of CO2/NH3/H2O = 1:1:1 does not match with them. The ionized Cu which is doped on TiO2 can provide free electrons to be used during the reduction reaction [31]. Therefore, the reductants NH3 and H2O are adequate to produce CO in this study though they are smaller than the values according to the reaction scheme. It is also seen from Figures 2 and 3 that the produced CO decreases after attaining the maximum value. The decrease in the produced CO is thought to be caused due to the oxidization of CO with O2 [32] which is by-product as explained by Equation (2) in the reaction scheme of CO2 reduction reacting with H2O. Therefore, this study does not establish that it is caused by the deactivation of photocatalyst.
As to the impact of NH3 on CO2 reduction characteristics using Cu/TiO2, the authors' previous study [27] had revealed the following conclusions: The highest produced CO for the case of molar ratio of CO2/NH3/H2O = 1:0.5:0.5 exhibits approximately four times compared to that for the case of molar ratio of CO2/H2O = 1:0.5. In addition, the produced CO keeps some value approximately without rapid decrease before 24 h for CO2/NH3/H2O conditions compared to the molar ratio of CO2/H2O = 1:0.5. It is known from the theoretical reaction scheme to reduce CO2 with NH3 that the more reaction step is needed to produce CO since NH3 should be converted into H2. Therefore, it is believed that the time to produce CO is longer compared to the molar ratio of CO2/H2 = 1:0.5. When comparing the highest produced CO under the condition of CO2/NH3/H2O = 3:8:12 to that under the condition of CO2/H2O = 3:12, it is confirmed that the highest produced CO under the condition of molar ratio of CO2/NH3/H2O = 3:8:12 is approximately three times compared to that under the condition of molar ratio of CO2/H2O = 3:12. Therefore, it is clear that NH3 has the potential to enhance CO2 reduction characteristics of Cu/TiO2 investigated in this study.
On the other hand, it can be seen from Figure 3 that the CO2 reduction performance for the molar ratio of CO2/NH3/H2O = 1:0.5:0.5 is the highest where the molar quantity of CO per unit weight of the photocatalyst is 2.52 mol/g. Additionally, it is obvious that the amount of the total reductants is smaller than that in the case with UV light. When the Xe lamp is illuminated without UV light, the light intensity and wavelength range of light are smaller and narrower respectively, compared to the illumination condition with UV light as described before. It is known from the theoretical reaction scheme of CO2 reduction reacting with H2O or NH3 that an electron is produced by the photochemical reaction which is influenced by the light illumination condition. In addition, H + whose amount is the same as that of electron is needed to produce CO. Since the produced electron might be smaller due to the smaller light input, it is believed that the amount of needed H + is smaller. Therefore, it is revealed that the highest CO2 reduction characteristics is obtained under the condition of the molar ratio of CO2/NH3/H2O = 1:0.5:0.5, while total reductants are smaller compared to the case with UV light.

The CO2 Reduction Characteristics Over Cu/TiO2 for a Pulse Number of 200
Figures 4 and 5 present the change of formed CO with the time under the illumination condition of Xe lamp with or without UV light, respectively. As to these figures, the produced CO is evaluated quantitatively by the molar quantity of CO per unit weight of photocatalyst having a unit of mol/g. The other fuels were not detected in this study. The data obtained under the condition of molar ratio of CO2/NH3/H2O = 3:12:18 are shown in Figure 4 to investigate the CO2 reduction performance under the larger H + supply condition. As to a blank test, this study conducted the same experiment under the condition of no illumination of Xe lamp as a reference case before the experiment. We detected no fuel during the blank test, as we hoped.
It is revealed from Figures 4 and 5 that the CO2 reduction characteristics under the condition of the molar ratio of CO2/NH3/H2O = 3:8:12 is the highest where the molar quantity of CO per unit weight of photocatalyst are 34.1 mol/g and 12.0 mol/g, respectively. The highest CO2 reduction performance was obtained under the condition of the molar ratio of CO2/NH3/H2O = 3:8:12. According to the quantitative analysis by EPMA as shown above, the weight percentages of Cu within the Cu/TiO2 film for the pulse number of 100 and 200 are 1.62 wt% and 4.57 wt%, respectively. When the amount of Cu increases, the total amount of free electron emitted from Cu during the photochemical reaction increases. Since the amount of H + which is the same as that of electron is needed to produce CO referring to the reaction scheme of CO2 reduction reacting with H2O or NH3, larger H + is needed in the case of pulse number of 200. It is assumed that the theoretical molar ratios of CO2/NH3/H2O are 3:2:3 and 3:8:12 to produce CO and CH4 respectively. However, it is revealed that the molar ratio of CO2/NH3/H2O = 3:8:12 is the most suitable for CO production in this study. The larger H + condition such as the molar ratio of CO2/NH3/H2O = 3:12:18 was also examined to confirm whether larger H + is needed to produce more CO and CH4 or not, resulting that the CO2 reduction performance is lower compared to the molar ratio of CO2/NH3/H2O = 3:8:12. From these results, the optimum molar ratio of CO2/NH3/H2O is found to be 3:8:12.

The CO2 Reduction Characteristics Over Cu/TiO2 for a Pulse Number of 500
Figures 6 and 7 present the change of formed CO with the time under the Xe lamp with or without UV light, respectively. In these figures, the produced CO is evaluated quantitatively by the molar quantity of CO per unit weight of photocatalyst having a unit of mol/g. The other fuels were not detected in this study. The data obtained under the condition of molar ratio of CO2/NH3/H2O = 3:12:18 are shown in Figure 6 to investigate the larger H + supply condition. As to a blank test, this study conducted the same experiment under the condition of no illumination of Xe lamp as a reference case before the experiment. We detected no fuel during the blank test, as we hoped.
It is revealed from Figures 6 and 7 that the CO2 reduction characteristic under the condition of the molar ratio of CO2/NH3/H2O = 3:12:18 is the highest where the molar quantity of CO per unit weight of photocatalyst are 4.4 mol/g and 2.5 mol/g, respectively. It is revealed from Figures 6 and 7 that the highest CO2 reduction characteristic is obtained under the condition of the molar ratio of CO2/NH3/H2O = 3:12:18. According to the quantitative analysis by EPMA, the weight percentages of element Cu in the Cu/TiO2 film for the pulse number of 100, 200, and 500 are 1.62 wt%, 4.57 wt%, and 7.95 wt%, respectively. As mentioned above, the total amount of free electron emitted from Cu during the photochemical reaction increases when the amount of Cu increases. Since the amount of H + , which is the same as that of electron is needed to produce CO referring to the theoretical reaction scheme of CO2 reduction reacting with H2O or NH3, a larger amount of H + is needed in the case of a pulse number of 500 compared to the case of a pulse number of 200. Consequently, the optimum molar ratio of CO2/NH3/H2O is found to be 3:12:18. However, the molar quantity of CO per unit weight of photocatalyst is 4.4 mol/g under the condition of the molar ratio of CO2/NH3/H2O = 3:12:18, which is smaller than that under the condition of the molar ratio of CO2/NH3/H2O = 3:8:12 for a pulse number of 200. It might be thought that the CO2 reduction characteristic is promoted with the increase in Cu loading weight. However, it is thought that too much Cu loading brings to cover the surface of the TiO2 film [33,34]. As a result, CO2 and reductants cannot reach the surface of TiO2 film sufficiently. From these discussions, it is obvious that there is an optimum loading amount of Cu in order to promote CO2 reduction characteristics with NH3 and H2O.
The considered conditions shown in the case of pulse number of 100, 200, and 500 are different in this study. The aim of this study is to clarify the optimum molar ratio of CO2/NH3/H2O for the improvement of CO2 reduction performance of Cu/TiO2. In the case of pulse number of 100, the optimum molar ratio of CO2/NH3/H2O is 1:1:1 with UV light while that is 1:0.5:0.5 without UV light. Therefore, it is not necessary to investigate the larger reductant ratio condition such as CO2/NH3/H2O = 3:12:18. On the other hand, in the case of pulse number of 200, the optimum molar ratio of CO2/NH3/H2O is 3:8:12 under the condition both with and without UV light. To optimize the molar ratio, the molar ratio of CO2/NH3/H2O = 3:12:18 which was larger reductant ratio condition was investigated under the condition of Xe lamp with UV light in this study. However, in the case of pulse number of 500, the optimum molar ratio of CO2/NH3/H2O is found to be 3:12:18 under the condition both with and without UV light, resulting that it was necessary to investigate the larger reductant ratio condition. This is the reason why the reported experimental conditions are different among the cases of pulse number of 100, 200, and 500 in this study.
The highest CO2 reduction performance of Cu/TiO2 is the molar quantity of CO per unit weight of photocatalyst of 34.1 mol/g which is obtained in the case of pulse number of 200. According to previous reports on the CO2 reduction characteristic of the other metals doped TiO2, Pt/TiO2, and Ru/TiO2 performed the molar quantity of CO per unit weight of photocatalyst of 12 mol/g and 13 mol/g, respectively, after the Xe lamp illumination time of 20 h in the case of CO2/H2O [10]. Pd/TiO2 performed the molar quantity of CO per unit weight of photocatalyst of 10 mol/g after the Xe ark lamp illumination time of 3 h in the case of CO2/H2O [35]. On the other hand, Ag/TiO2 with activated carbon performed the molar quantity of CO per unit weight of photocatalyst of 1.6 mol/g after a Xe lamp illumination time of 4 h in the case of CO2/liquid H2O [36]. According to the comparison of the CO2 reduction performance of Cu/TiO2 prepared in this study with that of other metal-doped TiO2, the superiority of Cu/TiO2 is confirmed.

The Quantum Efficiency Evaluation
Quantum efficiency is a well-known factor used to indicate the photocatalytic activity and efficiency [37]. The quantum efficiency is calculated by the following equations [6,38]: where η is the quantum efficiency (%), Ninput is the photon number absorbed by photocatalyst (-), Noutput is the photon number used in photocatalytic reaction (-), I is the light intensity of UV light (W/cm 2 ), t is the time illuminating UV light (t), λ is the wavelength limit of light that the photocatalyst can absorb for the photocatalytic reaction (m), Are is the reaction surface area of photocatalyst assumed to be equal to the surface area of netlike glass disc (cm 2 Figures 8 and 9 show the quantum efficiencies among different molar ratios of CO2/NH3/H2O with and without UV light, respectively. In Figures 8 and 9, the results in the case of pulse number of 200 which provided the highest CO2 reduction characteristic in this study are shown. It is revealed from Figures 8 and 9 that the highest quantum efficiency is obtained under the condition of the molar ratio of CO2/NH3/H2O = 3:8:12 irrespective of illumination condition of Xe lamp, which follows the results shown in Figures 4 and 5. It is obvious that the largest molar quantity of CO per unit weight of photocatalyst is 4.4 mol/g in this study. Compared to the previous studies on CO2 reduction using Cu/TiO2 with H2O, as mentioned above [16,[18][19][20], the amount of molar quantity of CO per unit weight of photocatalyst is approximately the same level. The highest quantum efficiency is 4.69 × 10 -4 when the Xe lamp with UV light is illuminated. On the other hand, it is 2.47 × 10 -4 under the illumination condition of the Xe lamp without UV light in this study. According to the previous study [31], Cu/TiO2 (2 wt% of Cu) performed the quantum efficiency of 1.56 × 10 -2 in the case of CO2/H2O with UV light illumination. The other study reported that Cu/TiO2 (1 wt% of Cu) showed the quantum efficiency of 1.41 × 10 -2 in the case of CO2/H2O with UV light illumination [39]. The quantum efficiency obtained in this study is lower than that obtained in previous studies. The reason is thought to be that the total amount of electron needed in this study for photochemical reaction is too large due to the combination of two H + suppliers, i.e., NH3 and H2O. It is thought that (i) capturing the maximum visible light region, and (ii) draining the photogenerated charges on light irradiation towards Cu/TiO2 surface [40] may be possible ways to improve the quantum efficiency.  This study has revealed the relationship between the optimum loading weight of Cu and the combination of reductants, such as NH3 and H2O, in order to improve the CO2 reduction performance of TiO2. As the next step, it can be considered to improve the CO2 reduction performance of Cu/TiO2 with NH3 and H2O further. The combination of different doped metals may be tried to promote the CO2 reduction performance further in the near future. According to the previous studies [6,21], the co-doped TiO2 such as PbS-Cu/TiO2, Cu-Fe/TiO2, Cu-Ce/TiO2, Cu-Mn/TiO2, and Cu-CdS/TiO2 were conducted to promote the CO2 reduction performance of TiO2 with H2O. For the combination of CO2/NH3/H2O, the electron number emitted from the dopant had better fit with H + number according to the theoretical reaction scheme. The electron number should match with H + number to produce fuel. Although this study selects Cu + ion in order to enhance the CO2 reduction characteristic over TiO2, the co-doping Cu with the other metal which has a larger positive ion would provide the positive impact to enhance the CO2 reduction characteristic with reductants NH3 and H2O. Therefore, it is expected that the CO2 reduction performance is promoted by the combination of different metal doping in the case of CO2/NH3/H2O.

The Preparation Proceure of Cu/TiO2 Film
This study prepared TiO2 film by sol-gel and dip-coating process [27]. [(CH3)2CHO]4Ti (Purity of 95 wt%, produced by Nacalai Tesque Co., Kyoto, Japan) of 0.3 mol, anhydrous C2H5OH (purity of 99.5 wt%, produced by Nacalai Tesque Co., Kyoto, Japan) of 2.4 mol, distilled water of 0.3 mol, and HCl (purity of 35 wt%, produced by Nacalai Tesque Co., Kyoto, Japan) of 0.07 mol were mixed for preparing the TiO2 sol solution. This study coats the TiO2 film on a netlike glass fiber (SILIGLASS U, produced by Nihonmuki Co., Tokyo, Japan) by sol-gel and dip-coating processes. The glass fiber, having a diameter of about 10 m, weaved as a net, is collected to be the diameter of approximately 1 mm. The porous diameter of glass fiber and the specific surface area are about 1 nm and 400 m 2 /g, respectively from the specifications of netlike glass fiber. The netlike glass fiber is composed of SiO2 of 96 wt%. The opening space of the netlike glass fiber is approximately 2 mm × 2 mm. Since the netlike glass fiber has porous characteristics, the netlike glass fiber can capture the TiO2 film easily during sol-gel and dip-coating processes. Additionally, we can expect that CO2 is more easily absorbed by the prepared photocatalyst due to the porous characteristics of the netlike glass fiber. This study cut the netlike glass fiber to be disc form having diameter of 50 mm and thickness of 1 mm. This study immersed the netlike glass disc into TiO2 sol solution controlling the speed at 1.5 mm/s and drew it up controlling the fixed speed at 0.22 mm/s. After that, this study dried it out and fired under a controlled firing temperature (FT) and firing duration time (FD) to fasten the TiO2 film to the base material. This study set FT and FD at 623 K and 180 s, respectively.
After the coating of TiO2, this study loads Cu on the TiO2-coated netlike glass disc by the pulse arc plasma gun process [27] emitting Cu nano-particles uniformly via an applied high voltage. The pulse number can control the quantity of Cu loaded on TiO2. This study set the pulse number at 100, 200, and 500. This study applied the pulse arc plasma gun device (ARL-300, produced by ULVAC, Inc., Chigasaki, Japan) which has a Cu electrode. The diameter of the Cu electrode for Cu loading was 10 mm. The Cu nano-particles were emitted from Cu electrode with applying the voltage of 200 V after the TiO2 coated on netlike glass disc was set in the vacuumed vessel. The pulse arc plasma gun evaporates Cu electrode into fine particulate form over the TiO2 in the concentric area whose diameter is 100 mm when the distance between Cu electrode and the TiO2 is set to be 160 mm. Due to the distance between Cu electrode and TiO2 film of 150 mm, this study can spread Cu particles over TiO2 film uniformly.

The Characterization Procedure of Cu/TiO2 Film
This study evaluated the characteristics of external and crystal structure of Cu/TiO2 film prepared above by SEM (JXA-8530F, produced by JEOL Ltd., Tokyo, Japan) and EPMA (JXA-8530F, produced by JEOL Ltd., Tokyo, Japan) [27]. These procedures use electron to analyze characterization, resulting that the sample should conduct electricity. Since the netlike glass disc used for base material to coat Cu/TiO2 film cannot conduct electricity, the vaporized carbon was deposited by the carbon deposition device (JEE-420, produced by JEOL Ltd., Tokyo, Japan) on the surface of the Cu/TiO2 film before analyzing its characterization. The thickness of the deposited carbon was approximately 2030 nm. The electrons are emitted from the electrode to the sample setting the acceleration voltage and current of 15 kV and 3.0 × 10 -8 A respectively, in order to analyze the external structure of Cu/TiO2 film using SEM. After the character X-ray is analyzed using EPMA at the same time, the amount of chemical element is clarified referring to the relation between character X-ray energy and atomic number. SEM and EPMA have the space resolution of 10 mm. The EPMA analysis can support to clarify the structure of prepared photocatalyst as well as to measure the quantity of loaded metal within TiO2 film on the netlike glass disc as base material.  [27]. The reactor size for charging CO2 is 1.25 × 10 -4 m 3 . The light of Xe lamp positioned on the stainless tube is illuminated toward Cu/TiO2 film passing the sharp cut filter and the quartz glass disc located on the top of the stainless tube. The wavelength of light illuminated from Xe lamp is ranged from 185 nm to 2000 nm. The sharp cut filter can eliminate the UV from the Xe lamp, resulting in the wavelength of light illuminating the Cu/TiO2 film ranged from 401 nm to 2000 nm. Figure 11 shows the light transmittance characteristics of the sharp cut filter to remove the wavelength of the light, indicating that it can guarantee the removal of the wavelength of light under 400 nm [27]. The mean light intensity of light illuminated from Xe lamp without the sharp cut filter is 58.2 mW/cm 2 , while, with sharp cut filter, it is 33.5 mW/cm 2 .  After filling CO2 gas whose purity of 99.995 vol% in the reactor vacuumed by a vacuum pump for 15 min, the valves installed at the inlet and the outlet of reactor were closed during CO2 reduction experiment with NH3 + H2O. After that, this study confirmed the pressure of 0.1 MPa and gas temperature at 298 K in the reactor. Then, we injected the NH3 aqueous solution (NH3 purity of 50 vol%) into the reactor via the gas sampling tap, and turned on the Xe lamp at that time. The amount of injected NH3 aqueous solution was changed depending on the considered molar ratio. Due to the heat of the infrared light components illuminated by the Xe lamp, the injected NH3 aqueous solution was vaporized. The temperature in the reactor attained at 343 K within an hour, and it was maintained at 343 K during the CO2 reduction experiment. The molar ratio of CO2/NH3/H2O was changed by 1:1:1, 1:0.5:1, 1:1:0.5, 1:0.5:0.5, 3:2:3, 3:8:12, 3:12:18. The reacted gas filled in the reactor was extracted by gas syringe via gas sampling tap and it was analyzed by an FID gas chromatograph (GC353B, produced by GL Science) and a methanizer (MT221, produced by GL Science). The FID gas chromatograph and methanizer have minimum resolutions of 1 ppmV.

Conclusions
From the investigation in this study, the following conclusions could be drawn: (i) Cu particles whose size is nano-scale could be loaded on TiO2 uniformly by the pulse arc plasma gun process. The weight percentages of element Cu within the Cu/TiO2 film for pulse numbers of 100, 200, and 500 increase with the increase in the pulse number, which are 1.62 wt%, 4.57 wt%, and 7.95 wt%, respectively. (ii) As to the pulse number of 100, the CO2 reduction characteristic under the condition of the molar ratio of CO2/NH3/H2O = 1:1:1 is the highest where the molar quantity of CO produced for per unit weight of photocatalyst is 10.2 mol/g with UV light. However, the CO2 reduction characteristic under the condition of the molar ratio of CO2/NH3/H2O = 1:0.5:0.5 is the highest where the molar quantity of CO produced for per unit weight of photocatalyst is 2.52 mol/g without UV light. (iii) As to the pulse number of 200, the CO2 reduction characteristics under the condition of the molar ratio of CO2/NH3/H2O = 3:8:12 is the highest under the illumination conditions both with and without UV light, where the molar quantity of CO produced per unit weight of photocatalyst are 34.1 mol/g and 12.0 mol/g, respectively. (iv) In the case of pulse number of 500, the CO2 reduction characteristic under the condition of the molar ratio of CO2/NH3/H2O = 3:12:18 under the illumination conditions both with and without UV light is the highest where the molar quantity of CO produced per unit weight of photocatalyst are 4.4 mol/g and 2.5 mol/g, respectively. It is revealed that the optimum loading weight of Cu is 4.57 wt% in order to promote CO2 reduction characteristic with NH3 and H2O. Funding: This research received no external funding.

Data Availability Statement:
The authors agree with the data availability presented in this study.