Abstract

Fe-doped TiO2 (Fe/TiO2) film photocatalyst was prepared by sol-gel and dip-coating process and pulse arc plasma method. The effect of pulse number on the CO2 reduction performance with the Fe/TiO2 was investigated in this study. In addition, the effect of reductants such as H2O, H2, and NH3/H2O on the CO2 reduction performance with the Fe/TiO2 photocatalyst was also investigated. The characteristics of the prepared Fe/TiO2 film coated on a netlike glass fiber which is a base material were analyzed by SEM, EPMA, EDX, and EPMA. Furthermore, the CO2 reduction performance of the Fe/TiO2 film was tested under a Xe lamp with or without ultraviolet (UV) light. The results show that the CO2 reduction performance with the pulse number of 100 is the best with H2O and/or H2 as reductant under UV light illumination, while that with the pulse number of 500 is the best when NH3/H2O is used as reductant. On the other hand, the CO2 reduction performance with the pulse number of 500 is the best under every reductant condition without UV light illumination. The highest CO2 reduction performance with the Fe/TiO2 is obtained under H2 + H2O/CO2 condition, and the best moral ratio of total reductants to CO2 is 1.5 : 1.

1. Introduction

Due to mass consumption of fossil fuels, global warming and fossil fuel depletion have become serious global environmental problems in the world. After the industrial revolution, the averaged concentration of CO2 in the world has been increased from 278 ppmV to 400 ppmV by 2015 [1]. Therefore, it is necessary to develop a new CO2 reduction or utilization technology in order to recycle CO2.

There are six vital CO2 conversions: chemical conversions, electrochemical reductions, biological conversions, reforming, inorganic conversions, and photochemical reductions [2, 3]. Recently, artificial photosynthesis or the photochemical reduction of CO2 to fuel has become an attractive route due to its economically and environmentally friendly behavior [2].

TiO2 is the principle catalyst for almost all types of photocatalysis reaction. It is well known that CO2 can be reduced into fuels, for example, CO, CH4, CH3OH, and H2 by using TiO2 as the photocatalyst under ultraviolet (UV) light illumination [49]. If this technique could be applied practically, a carbon circulation system would then be established: CO2 from the combustion of fuel is reformed to fuels again using solar energy, and true zero emission can be achieved. However, the CO2 reduction rate using pure TiO2 is very low, that is, the fuel concentration in the products is very low. In addition, pure TiO2 is only photoactive at a wavelength below 400 nm due to its relatively large bandgap energy (~3.2 eV) [10].

Recently, studies on CO2 photochemical reduction by TiO2 have been carried out from the viewpoint of performance promotion by extending absorption range towards visible region [1115]. Noble metal doping such as Pt, Pd, Au, and Ag [11], nanocomposite CdS/TiO2 combining two different band gap photocatalysts [12], N2-modified TiO2 [13], light harvesting complex of green plants assisted Rh-doped TiO2 [14], and dye-sensitized TiO2 [15] have been attempted to overcome the shortcomings of pure TiO2. They did improve the CO2 reduction performance; however, the concentrations in the products achieved in all the attempts so far were still low, ranging from 10 ppmV to 1000 ppmV [46, 9] or from 1 μmol/g-cat to 100 μmol/g-cat [7, 8, 1115]. Therefore, a big breakthrough in increasing the concentration level of products is necessary to advance the CO2 reduction technology in order to make the technology practically useful.

It was reported that doping transition metal was a useful technique for extending the absorbance of TiO2 into the visible region [16]. For doping, various metal ions have been used, but among them, Fe3+ is considered as a strong candidate as it has a similar radius to Ti4+ (Fe3+ = 78.5 pm, Ti4+ = 74.5 pm) [17] and can easily fit into the crystal lattice of TiO2 [16, 18, 19]. Moreover, the redox potential (energy differential) of Fe2+/Fe3+ is close to that of Ti3+/Ti4+, resulting in shifting its optical absorption into the visible region [16, 18, 19]. Due to easy availability as well as the above described characteristics, Fe is selected as the dopant in the present study.

In the present paper, TiO2 film is coated by sol-gel and dip-coating process on net glass fiber (SILIGLASS U, Nihonmuki Co.). Netlike glass fiber is a net composed of glass fiber whose diameter is about 10 μm. The fine glass fibers are knitted, resulting in a diameter of the aggregate fiber of about 1 mm. According to manufacturer specifications of the netlike glass fiber, the porous diameter of glass fiber is about 1 nm and the specific surface area is about 400 m2/g. The netlike glass fiber consists of SiO2 whose purity is over 96 wt%. The aperture of the net is about 2 mm × 2 mm. Since the netlike glass fiber has a porous characteristic, it is believed that the TiO2 film is captured by the netlike glass fiber easily during a sol-gel and dip-coating process. In addition, it can be expected that CO2 absorption performance of prepared photocatalyst is promoted due to the porous structure of the netlike glass fiber.

Then, Fe is loaded on the TiO2-coated netlike glass fiber by pulse arc plasma method which can emit nanosized Fe particles by applying high electrical potential difference. Since the amount of loaded Fe can be controlled by pulse number, the present paper investigates the impact of the pulse number on characterization and CO2 reduction performance of the prepared Fe/TiO2.

To promote the CO2 reduction performance of photocatalyst, it is important to select the optimum reductant which provides the proton for the reduction reaction. Generally speaking, H2O is used as a reductant for CO2 reduction by photocatalyst according to some review papers [2023]. The reaction scheme of CO2 reduction with H2O is as follows [4, 2426].

Photocatalytic reaction:

Oxidization:

Reduction:

H2 is also a good candidate for the reductant since H2 can be converted into a proton by photocatalyst [2736]. The reaction scheme of CO2 reduction with H2 is as follows [30].

Photocatalytic reaction:

Oxidization:

Reduction:

Though there are some reports on CO2 reduction with H2 [2736], there are few reports investigating the reaction with TiO2 as the photocatalyst is a few [2731]. Particularly, there is no study on CO2 reduction with Fe/TiO2. This paper investigates the reduction performance of Fe/TiO2 for CO2 with H2 as well as CO2 with H2O. Furthermore, the present paper also investigates the effect of NH3 + H2O on CO2 reduction performance with the Fe/TiO2 as the photocatalyst since NH3 is thought to be a good H2 carrier. According to the authors’ review, there is no report on CO2 reduction over Fe/TiO2 with NH3. NH3 is thought to be used as reductant directly since NH3 aqueous solution can decompose into H2 and N2 by photocatalyst [37]. After decomposition of NH3, it is believed that the reaction scheme of CO2 reduction with H2 can be applied.

In this paper, TiO2 film doped with Fe (Fe/TiO2) was prepared and characterized by scanning electron microscope (SEM), electron probe microanalyzer (EPMA), transmission electron microscope (TEM), and energy dispersive X-ray spectrometry (EDX) analysis to clarify the optimum amount of loaded Fe. The CO2 reduction characteristics of Fe/TiO2 coated on net glass fiber with H2O and/or H2 and NH3 under the condition of illuminating Xe lamp with or without UV light were investigated.

2. Experiment

2.1. Preparation of Fe/TiO2 Film

Sol-gel and dip-coating process was used for preparing TiO2 film. TiO2 sol solution was made by mixing [(CH3)2CHO]4Ti (purity of 95 wt%, Nacalai Tesque Co.) of 0.3 mol, anhydrous C2H5OH (purity of 99.5 wt%, Nacalai Tesque Co.) of 2.4 mol, distilled water of 0.3 mol, and HCl (purity of 35 wt%, Nacalai Tesque Co.) of 0.07 mol. The netlike glass fiber was cut to disc, and its diameter and thickness were 50 mm and 1 mm, respectively. The netlike glass fiber was dipped into TiO2 sol solution at the speed of 1.5 mm/s and pulled up at a fixed speed of 0.22 mm/s. Then, it was dried out and fired under the controlled firing temperature (FT) and firing duration time (FD), resulting in the TiO2 film fastened on the base material. FT and FD were set at 623 K and 180 s, respectively. Fe was loaded on the TiO2 film by pulse arc plasma method. The pulse arc plasma gun device (ULVAC Inc., ARL-300) having an Fe electrode whose diameter was 10 mm was applied for Fe loading. After the netlike glass fiber coated with TiO2 was set in the chamber of the pulse arc plasma gun device, which was vacuumed, the nanosized Fe particles were emitted from the Fe electrode applying an electrical potential difference of 200 V. The pulse arc plasma gun can evaporate an Fe particle over the target in the circle area whose diameter is 100 mm when the distance between the Fe electrode and the target is 160 mm. Since the difference between Fe electrode and TiO2 film was 150 mm in the present study, Fe particle can be evaporated over the TiO2 film uniformly. The amount of loaded Fe was controlled by pulse number. In the present paper, the pulse number was set at 100, 500, and 1000.

2.2. Characterization of Fe/TiO2 Film

The structure and crystallization characteristics of Fe/TiO2 film were evaluated by SEM (JXA-8530F, JEOL Ltd.), EPMA (JXA-8530F, JEOL Ltd.), TEM (JEM-2100F/HK, JEOL Ltd.), and EDX (JEM-2100F/HK, JEOL Ltd.). Since these measuring instruments use electron for analysis, the sample should be an electron conductor. Since the netlike glass disc is not an electron conductor, the carbon vapor deposition was conducted by the dedicated device (JEE-420, JEOL Ltd.) for Fe/TiO2 coated on a netlike glass disc before analysis. The thickness of carbon deposited on samples was approximately 20–30 nm.

The electron probe emits the electrons to the sample under the acceleration voltage of 15 kV and the current of 3.0 × 10−8 A, when the surface structure of sample is analyzed by SEM. The characteristic X-ray is detected by EPMA at the same time, resulting in the concentration of chemical element analyzed according to the relationship between the characteristic X-ray energy and the atomic number. The spatial resolution of SEM and EPMA is 10 μm. The EPMA analysis helps not only to understand the coating state of prepared photocatalyst but also to measure the amount of doped metal within TiO2 film on the base material.

The electron probe emits the electron to the sample under the acceleration voltage of 200 kV, when the inner structure of the sample is analyzed by TEM. The size, thickness, and structure of loaded Fe were evaluated. The characteristic X-ray is detected by EDX at the same time, resulting in the concentration distribution of chemical element toward thickness direction of the sample being analyzed. In the present paper, the concentration distribution of Ti and Fe were analyzed.

2.3. CO2 Reduction Experiment

Figure 1 shows the experimental setup of the reactor consisting of a stainless pipe (100 mm (H.) × 50 mm (I.D.)), a netlike glass disc coated with Fe/TiO2 film which is located on the Teflon cylinder (50 mm (H.) × 50 mm (D.)), a quartz glass disc (84 mm (D.) × 10 mm (t.)), a sharp cut filter which cuts off the light of wavelength below 400 nm (SCF-49.5C-42L, SIGMA KOKI CO. LTD.), a 150 W Xe lamp (L2175, Hamamatsu Photonics K. K.), a mass flow controller, a CO2 gas cylinder, and a H2 gas cylinder.

The reactor volume available for CO2 charge is 1.25 × 10−4 m3. The light of the Xe lamp, through the sharp cut filter and the quartz glass disc that are at the top of the stainless pipe, illuminates the netlike glass disc coated with Fe/TiO2 film, which is located inside the stainless pipe. The wavelengths of light from the Xe lamp are ranged from 185 nm to 2000 nm. The Xe lamp can be fitted with a sharp cut filter to remove UV components of the light. With the filter, the wavelengths of light from the Xe lamp are ranged from 401 nm to 2000 nm [38]. The average light intensity of the Xe lamp on the photocatalyst without and with setting the sharp cut filter is 57.5 mW/cm2 and 43.7 mW/cm2, respectively.

In the CO2 reduction experiment with H2O or NH3 + H2O, after purging the reactor chamber with CO2 gas of the purity of 99.995 vol% flowed through the reactor for 15 minutes, the valves located at the inlet and the outlet of reactor were closed. After confirming the gas pressure and gas temperature in the reactor at 0.1 MPa and 298 K, respectively, the distilled water of 100 μL or NH3 aqueous solution (NH3: 50 vol%) of 200 μL was injected into the reactor through gas sampling tap, and the Xe lamp illumination was turned on at the same time. The water and NH3 aqueous injected vaporized completely in the reactor. Due to the heat of the Xe lamp, the temperature in the reactor was attained at 343 K within an hour and kept at approximately 343 K during the experiment.

In the CO2 reduction experiment with H2, CO2 gas with the purity of 99.995 vol% and H2 gas with the purity of 99.99999 vol% which were controlled by mass flow controller were mixed in the buffer chamber and introduced in the reactor which was prevacuumed by a vacuum pump. After that, Xe lamp illumination was turned on. The mixing ratio of CO2 and H2 was confirmed by TCD gas chromatograph (Micro GC CP4900, GL Science) before introducing into the reactor. In the CO2 reduction experiment with H2 + H2O, the distilled water was injected into the reactor after charging CO2 and H2.

In the CO2 reduction experiment with H2O or H2, the molar ratio of H2O or H2 to CO2 was 1 : 1. In the CO2 reduction experiment with NH3 + H2O, the molar ratio of NH3 and H2O to CO2 was 0.7 : 1 : 1. In the CO2 reduction experiment with H2 + H2O, the molar ratio of H2 and H2O to CO2 was set at 0.5 : 0.5 : 1, 1 : 1 : 1, 2 : 2 : 1, 1 : 0.5 : 1 and 1 : 1 : 0.5.

The gas in the reactor was sampled every 24 hours during the experiment. The gas samples were analyzed by FID gas chromatograph (GC353B, GL Science) and methanizer (MT221, GD Science). Minimum resolution of FID gas chromatograph and methanizer is 1 ppmV.

3. Results and Discussion

3.1. Characterization of Fe/TiO2 Film

Figures 25 show EPMA images of TiO2, Fe/TiO2(100), Fe/TiO2(500), and Fe/TiO2(1000) film, respectively, where Fe/TiO2(100) means TiO2 with Fe loaded by the pulse number of 100. EPMA analysis was carried out for 1500 times magnification SEM images. In EPMA image, the concentrations of each element in observation area are indicated by the different colors. Light colors, for example, white, pink, and red, indicate that the amount of element is large, while dark colors like black and blue indicate that the amount of element is small.

From these figures, it can be observed that the TiO2 film with teeth-like shape was coated on a netlike glass fiber. It is also seen that TiO2 film builds a bridge among several glass fibers like reported also in [38]. During firing process, the temperature profile of TiO2 solution adhered on the netlike glass disc was not even due to the different thermal conductivities of Ti and SiO2. The thermal conductivities of Ti and SiO2 at 600 K are 19.4 W/(m·K) and 1.82 W/(m·K), respectively [39]. Due to the thermal expansion and shrinkage around the netlike glass fiber, a thermal crack formed on the TiO2 film. Therefore, the TiO2 film on the netlike glass fiber was teeth like.

It is observed from Figure 3 that nanosized Fe particles loaded on TiO2 uniformly. On the other hand, it is revealed that the amount of loaded Fe increases with increasing the pulse number from Figures 4 and 5. In other words, there is an Fe layer covering TiO2 for Fe/TiO2(500) and Fe/TiO2(1000).

To evaluate the amount of loaded Fe within the TiO2 film quantitatively, the observation area, which is the center of netlike glass disc, of diameter of 300 μm, is analyzed by EPMA. The ratio of Fe to Ti in this observation area is counted by averaging the data obtained in this area. Table 1 gives the weight percentages of elements Fe and Ti in the Fe/TiO2 film. From this table, it can be seen that more Fe is contained in the Fe/TiO2 film with increasing pulse number. From these results, it is proved that the amount of dopants on a photocatalyst can be controlled by pulse arc plasm method quantitatively.

Figures 68 show EDX images of Fe/TiO2(100), Fe/TiO2(500), and Fe/TiO2(1000) film, respectively. EDX analysis was carried out using 150,000 times magnification TEM images. In these figures, the yellow circle indicates the existence of Fe.

According to Figure 6, it is observed that Fe particle is loaded on TiO2 film. The average size of Fe particle for Fe/TiO2(100) in the present study is approximately 28 nm where the longest length of Fe particle was used as a representative length. As to Fe/TiO2(500) shown in Figure 7, there was an Fe layer covering the TiO2 film. The average thickness of Fe layer for Fe/TiO2(500) in the present study was found to be approximately 36 nm. For Fe/TiO2(1000), the thickness of Fe layer was thicker than that on the Fe/TiO2(500). The average thickness of Fe layer for Fe/TiO2(1000) in the present study is approximately 230 nm. Since the pulse arc plasm was repeated continuously with such a large pulse number of 1000, the tip of Fe electrode was melted and a lump of Fe was emitted. Therefore, the thickness of Fe layer for Fe/TiO2(1000) was remarkably longer than that for Fe/TiO2(500).

3.2. Effect of Amount of Fe on CO2 Reduction Performance of Fe/TiO2

Figures 911 show the comparison of molar quantity of CO per weight of photocatalyst along the time under the Xe lamp with the UV light on, for TiO2 or Fe/TiO2 film under H2O/CO2, H2/CO2, and NH3 + H2O/CO2 condition, respectively. In this experiment, CO is the only fuel produced from the reactions. Since the concentrations of CO in most experiments started to decrease after illumination of 48–72 hours for illumination conditions with UV light due to the reverse reaction by CO and O2, that is, the by-product, Figures 911 only show the concentration up to 72 hours. Before the experiments, a blank test, that was running the same experiment without illumination of Xe lamp, had been carried out to set up a reference case. No fuel was produced in the blank test as expected.

According to Figures 9 and 10, the molar quantity of CO per weight of photocatalyst for Fe/TiO2(100) is the highest among the prepared photocatalysts under H2O/CO2 and H2/CO2 conditions. Although the amount of Fe increases with increasing pulse number as shown in Table 1, Fe layer covers TiO2 film according to EPMA and EDX analysis. When the metal covers TiO2 surface, the light absorption ability is impaired and the contact between TiO2 and CO2 + reductant is blocked [40]. Consequently, the Fe/TiO2(100) which has enough area absorbing light and contacting reactants has the best CO2 reduction performance. On the other hand, the molar quantity of CO per weight of photocatalyst for Fe/TiO2(500) is the highest among the prepared photocatalysts under NH3 + H2O/CO2 condition. Since six electrons are necessary to decompose NH3 into H2 and N2 by photocatalyst [37] and CO2 reduction with H2 also need two electrons after NH3 decomposition as described in Section 1, more electrons are used under NH3 + H2O/CO2 condition compared to H2O/CO2 and H2/CO2 conditions. Therefore, the Fe/TiO2(500) which can provide more electrons is the best photocatalyst even though Fe layer covers TiO2 film. However, the CO2 reduction performance of Fe/TiO2 under NH3 + H2O/CO2 condition is lower compared to that under H2O/CO2 and H2/CO2 conditions due to complicity of the reactions. In addition, a very thick Fe layer covers TiO2 film for the Fe/TiO2(1000), resulting in a very low CO2 reduction performance. The improvement of photocatalytic performance by Fe doping under the illumination condition with UV light is thought to be caused by the generation of shallow charge traps in the crystal structure which decreases the recombination rate of electron-hole pairs [18] except for Fe/TiO2(1000).

Figures 1214 show the comparison of molar quantity of CO per weight of photocatalyst along the time under the Xe lamp illumination without UV light, for TiO2 or Fe/TiO2 film under H2O/CO2, H2/CO2, and NH3 + H2O/CO2 condition, respectively. In this experiment, CO is the only fuel produced from the reactions. Since the concentration of CO almost started to decrease after illumination of 72–96 hours for most cases due to the reverse reaction by CO and O2 which is by-product, Figures 1214 only show the concentration up to 96 hours.

From Figures 1214, it can be seen that the CO2 reduction performance of TiO2 is promoted by Fe doping due to extension of the photoresponsivity of TiO2 [41] to the visible spectrum as well as decrease in the recombination rate of electron-hole pairs by the generation of shallow charge traps in the crystal structure. According to Figures 1214, the molar quantity of CO per weight of photocatalyst for Fe/TiO2(500) is the highest among the prepared photocatalysts, which indicates that the larger pulse number is suitable for the illumination condition without UV light. Under the illumination condition without UV light, the amount of doped Fe is important to absorb the visible light in order to perform the photocatalytic reaction [41]. As to the Fe/TiO2(1000), since the too-high loading Fe covered the TiO2 surface, the light absorption ability was impaired and the contact between TiO2 and CO2 + reductant was blocked [40], resulting in lower CO2 reduction performance compared to the other pulse numbers.

3.3. Optimization of Reductant on CO2 Reduction Performance of Fe/TiO2

To investigate the combination effect of reductant, Figure 15 shows the comparison of molar quantity of CO per weight of photocatalyst along the time under the Xe lamp with UV light on, for Fe/TiO2(100) film under H2 + H2O/CO2 condition. Due to the UV light illumination condition, Fe/TiO2(100) is selected as the best photocatalyst as described in Section 3.2. For comparison, the results under H2O/CO2 and H2/CO2 conditions are also shown in this figure. Since the present study wants to know the CO2 reduction performance obtaining a larger amount of product, the results under the illumination condition with UV light are shown.

From Figure 15, it reveals that the molar quantity of CO per weight for Fe/TiO2 under H2 + H2O/CO2 condition is larger than the sum of that under H2O/CO2 and H2/CO2 conditions, which agrees with the report carried out by TiO2 [30]. Simultaneous presence of H2O and H2 in the system have not changed the reaction pathway, but have accelerated the photoreduction of CO2 with H2, since H2O donated electron to inhibit the recombination of electron and hole [30]. In addition, the molar ratio of H2 and H2O to CO2 of 1 : 0.5 : 1 gives the best CO2 reduction performance. However, the molar ratio of H2 and H2O to CO2 of 0.5 : 1 : 1 shows the second best performance, which means the moral ratio of total reductants to CO2 of 1.5 : 1 is optimal. According to reaction scheme of H2O/CO2 and H2/CO2 shown in Section 1, the theoretical molar ratio of CO2 to H2O or H2 is 1 : 1. It is therefore believed that the excess amount of reductant is necessary for CO2 reduction in practice.

Table 2 lists the maximum molar quantity of CO per weight of photocatalyst for each condition to summarize the CO2 reduction performance of prepared photocatalysts in this study. According to Table 2, the molar ratio of H2 and H2O to CO2 of 1 : 0.5 : 1 gives the best CO2 reduction performance in this study as described above. Table 3 lists the initial rate of CO production for each condition to investigate the CO2 reduction rate of prepared photocatalysts in this study. The initial rate of CO production is estimated by dividing the molar quantity of CO per weight of photocatalyst at 24 h by illumination time of 24 h. According to Table 3, the initial rate of CO production under H2 + H2O/CO2 condition is larger than that under the other conditions. This indicates that the combination of H2 and H2O leads the fast reaction of prepared photocatalyst since H2 and H2O promote the reduction and oxidation reaction during photocatalysis reaction, respectively.

Compared to the previous research on CO2 reduction with H2 + H2O over pure TiO2, the CO2 reduction performance of photocatalysts prepared in this study is almost 100 times as large as that reported in [30], which is owing to Fe doping. Though reference [30] reported the CO production performance of ZrO2 under H2 + H2O/CO2 condition, the CO production performance of Fe/TiO2 prepared by this study is approximately 30 times as large as that of ZrO2. In addition, the other research [42] also reported the CO2 reduction performance of TiO2 under H2 + H2O/CO2 condition. The CO production performance of Fe/TiO2 prepared by this study is approximately 100 times as large as that reported in reference [42]. Since Fe doping provides the free electron preventing recombination of electron and hole produced as well as improving the light absorption effect, the big improvement of CO2 reduction performance is obtained in the present study, which reaches almost the same level compared to the other studies [7, 8, 1115], in terms of CO2 reduction performance. One way to further promote the CO2 reduction performance is double overlapping arrangement of Fe/TiO2 coated on netlike glass disc. The authors have already reported that the double overlapping arrangement of Fe/TiO2 coated on netlike glass disc was effective to improve CO2 reduction performance of Fe/TiO2 under H2O/CO2 condition since the electron transfer between two overlapped photocatalysts was promoted by overlapping [38]. In addition, the other dopants like Cu, which can absorb the longer wavelength light than Fe [19], should be used at lower positioned layers since the wavelength of penetrating light becomes long through higher positioned photocatalyst by losing energy [38]. This idea is similar to the concept of the hybrid photocatalyst using two photocatalysts having different band gaps [4345].

In the rear future, this study would like to carry out the reusability evaluation on the structure characterization and photocatalytic activity, which are important to investigate the possibility of applying a photocatalyst in a practical way.

4. Conclusions

Based on the investigation in this study, the following conclusions can be drawn. (1)TiO2 could be coated on netlike glass fiber where it appears in teeth-like shape. Fe fine particles could be loaded onto the TiO2 using the pulse arc plasma method with various pulse numbers. When the pulse number was over 500, there would be a Fe layer formed on the TiO2.(2)With the UV illumination condition, the molar quantity of CO per weight of photocatalyst for Fe/TiO2(100) is the highest among the prepared photocatalysts under H2O/CO2 and H2/CO2 conditions, while that for Fe/TiO2(500) is the highest among the prepared photocatalysts under NH3 + H2O/CO2 condition.(3)Under the illumination condition without UV light, the molar quantity of CO per weight of photocatalyst for Fe/TiO2(500) is the highest among the prepared photocatalysts irrespective of reductant used, indicating that the larger pulse number is suitable for the illumination condition without UV light.(4)The CO2 reduction performance of Fe/TiO2 under NH3 + H2O/CO2 condition is lower compared to that under H2O/CO2 and H2/CO2 conditions due to complicity of the reaction.(5)Since a very thick Fe layer covers TiO2 film for the Fe/TiO2(1000) and blocks light absorption and reactant contact with TiO2, the CO2 reduction performance with Fe/TiO2(1000) is very poor.(6)In this study, the highest CO2 reduction performance of Fe/TiO2, that is, the molar quantity of CO per weight of photocatalyst for Fe/TiO2(100), is obtained under H2 + H2O/CO2 condition, and the best moral ratio of total reductants to CO2 is found to be 1.5 : 1.

Conflicts of Interest

The authors declare that there is no conflict of interest regarding the publication of this paper.

Acknowledgments

The authors would like to gratefully thank JSPS KAKENHI (Grant no. 16K06970) and joint research program of the Institute of Materials and Systems for Sustainability, Nagoya University, for the financial support of this work.