Influence of Copper Valence in CuO_x/TiO₂ Catalysts on the Selectivity of Carbon Dioxide Photocatalytic Reduction Products

Abstract

The Cu cocatalyst supported on the surface of TiO₂ photocatalysts has demonstrated unique activity and selectivity in photocatalytic CO₂ reduction. The valence state of copper significantly influences the catalytic process; however, due to the inherent instability of copper’s valence states, the precise role of different valence states in CO₂ reduction remains inadequately understood. In this study, CuO_x/TiO₂ catalysts were synthesized using an in situ growth reduction method, and we investigated the impact of various valence copper species on CO₂ photocatalytic reduction. Our results indicate that Cu⁺ and Cu⁰ serve as primary active sites, with the selectivity for CH₄ and CO products during CO₂ photoreduction being closely related to their respective ratios on the catalyst surface. The adsorption and activation mechanisms of CO on both Cu⁺ and Cu⁰ surfaces are identified as critical factors determining product selectivity in photocatalytic processes. Furthermore, it is confirmed that Cu⁺ primarily facilitates CH₄ production while Cu⁰ is responsible for generating CO. This study provides valuable insights into developing highly selective photocatalysts.

1. Introduction

With the rapid advancement of global industry and the growth of population, human consumption of fossil energy is escalating on a daily basis. On the one hand, this leads to an energy shortage [1,2]. On the other hand, the substantial emissions of CO₂ give rise to global warming, glacier melting, the extinction of plankton, and other issues, posing a serious threat to human health and safety [3,4,5]. In the treatment of CO₂, converting CO₂ directly into energy substances and chemicals is of the utmost scientific significance. Common strategies for CO₂ resource utilization include CO₂ capture and storage [6,7], electrocatalytic CO₂ reduction [8,9], thermal-assisted catalytic CO₂ reduction [10,11], and photocatalytic CO₂ reduction [12,13]. Photocatalytic CO₂ reduction harnesses the abundant and renewable energy of sunlight to transform CO₂ into valuable chemicals, all while avoiding the consumption of precious electrical or thermal energy. This process is regarded as an optimal solution to the challenge posed by excessive CO₂ emissions [14]. Since 1979, it has been reported that TiO₂ can catalyze the conversion of CO₂ into CH₃OH, HCHO, and other chemical compounds, which has sparked a global research surge in photocatalytic CO₂ conversion to energy-rich substances [14]. Currently, a diverse array of semiconductor photocatalytic materials has been developed, with TiO₂-based photocatalysts being the most extensively studied and applied in the field of photocatalysis due to their superior efficiency and stability [15,16,17,18,19,20], suitable bandgap structure, low toxicity, and other characteristics [21,22,23]. Nevertheless, the photocatalytic reduction of CO₂ using TiO₂ still faces challenges related to limited photocatalytic efficiency and low product selectivity [24]. In order to improve the photocatalytic performance of TiO₂, surface modification can be carried out by the following common methods: doping [25,26], defect construction [27,28,29], morphology regulation [30], heterojunction construction [31,32], co-catalyst support [33,34], and surface sensitization [35,36]. Among them, supported cocatalyst is an effective means to improve the photocatalytic performance of TiO₂ [33,34,37,38].

Transition metals are often used as cocatalysts to improve the photocatalytic activity of semiconductors, among which Cu is widely used in the field of photocatalysis, especially in the study of photocatalytic CO₂ reduction because of its abundant reserves, low price, easy obtainability, and efficient charge separation ability [39,40,41,42,43]. In addition, the use of p-type semiconductors (i.e., copper oxide) in designing catalyst strategies not only enhances charge redistribution due to their narrower band gap but also enhances the selectivity of the reaction to methanol [44]. The remarkable characteristic of Cu is the diversity of valence states (Cu²⁺, Cu⁺ and Cu⁰). Through a series of experiments and characterization methods, it is revealed that in the presence of oxygen, the oxidation of cuprous copper and zero-valent copper affects its performance and stability in the photocatalytic reaction [45,46]. The role of different valence states of Cu in photocatalytic CO₂ reduction has attracted more and more attention. Many studies have reported the effect of copper on the selectivity of CO₂ products in photocatalytic reduction [47,48,49,50,51]. Kreft et al. [48] reported a study on the control of different valence Cu components by introducing O₂ and found that significantly increased product yield and complete selectivity to CO products could be observed in the presence of O₂, and Cu₂O was the most active species in the photocatalytic CO₂ reduction process. When the proportion of Cu₂O increases, the output of the corresponding product CO will also increase. Zhang et al. [52] reported a catalyst of oxygen-containing copper (Cu₄O), which showed that CO on the surface of Cu₄O with oxygen vacancy often continued to hydrogenate to produce high-value products, rather than desorption to produce CO. Different valence Cu components have different adsorption and activation capacities for CO and CO₂, so the valence of Cu components on the catalyst surface is an important factor affecting CO₂ reduction. However, due to the instability of the valence state of Cu, there is still a lack of research on the practical role of each valence Cu component in the process of CO₂ reduction.

The crystalline phases of TiO₂ are anatase, rutile, plate titanite, and a mixture of anatase and rutile (e.g., P25). Anatase-phase titanium dioxide has a good basis for photocatalytic activity and is easy to be treated and modified. Surface modification and modification can be performed by doping, loading, and other methods to improve its photocatalytic activity and stability [53,54]. More importantly, the crystal structure of anatase-phase titanium dioxide is relatively simple, the property is more stable, and it is easy to characterize and analyze. In this study, the surface of anatase TiO₂ was modified with different valence Cu species. The CuO_x/TiO₂ catalyst was prepared with an in situ growth reduction method to explore the effects of different valence Cu components on the photocatalytic reduction of CO₂. Combined with a catalyst characterization test and photocatalytic CO₂ reduction performance test, the results showed that the introduction of CuO_x did not change the structure, morphology, or redox potential of TiO₂. Cu⁺ and Cu⁰ are the main active sites on the catalysts. The selectivity of CH₄ and CO in the photocatalytic reduction of CO₂ products by CuO_x/TiO₂ is related to the ratio of Cu⁺ and Cu⁰ content on the catalyst surface. The mechanism of the influence of different valence copper components on product selectivity was analyzed from the perspective of thermodynamics and kinetics. *CO is more likely to experience desorption from the Cu⁰ surface to produce CO while continuing to adsorb on the Cu⁺ surface to produce CH₄. In this study, the adsorption of CO on different valence Cu components and the relationship between Cu valence and corresponding products in the process of CO₂ photocatalytic reduction were investigated, which provided a guide for the development of highly selective photocatalysts.

2. Experimental

2.1. Chemical Materials

The chemicals used in the experiment were purchased from commercial suppliers without further treatment. They are titanium dioxide (anatase, 99.8%, Aladdin, Shanghai, China), copper(II) acetate monohydrate (C₄H₆CuO₄·H₂O, ≥98.0%, Kermel, Colmar, France), ethylene glycol (EG, (CH₂OH)₂, AR), and ethanol (CH₃CH₂OH, AR).

2.2. Synthesis of the CuO_x/TiO₂

An amount of 500 mg TiO₂ and 32.14 mg Cu(CH₃COO)₂·H₂O were dispersed into a mixture of 30 mL H₂O and ethylene glycol, stirred at room temperature for 1 h, and then hydrothermal reaction was carried out in an oven at 200 °C for 2 h, cooled to room temperature, washed with H₂O and ethanol three times, respectively, and dried at 60 °C for 12 h. Solid powder catalyst was obtained by grinding. According to the addition of y mL ethylene glycol (reductant), the catalyst sample was denoted as CuO_x/TiO₂-y (y = 0 ~ 5).

2.3. Characterization

The crystal structures of CuO_x/TiO₂ were characterized using X-ray diffraction (XRD, D8-ADVANCE, Cu Kα radiation, 2θ = 20 ~ 80°, China), whose operation voltage and current were set at 40 kV and 30 mA, respectively; the morphologies were analyzed by transmission electron microscopy and high-resolution transmission electron microscopy equipped with FFT; (TEM, HRTEM, JEOL JEM-F200, Tokyo, Japan). The UV–Vis diffuse reflectance spectra (DRS) absorbance spectra were obtained with a Scan UV–Vis diffuse reflectance spectrophotometer (Shimadzu, Kyoto, Japan, UV-2600), using BaSO₄ as the reflectance sample. Specific surface area, pore size distribution, and CO₂ physical adsorption spectra were measured by BET surface area measurements (Quadrascorb SI-4), which were carried out by N₂ adsorption–desorption isotherms. CO₂/TPD was carried out on the ChemBET PULSARTMTPR/TPD chemisorption analyzer with argon (Ar) as the carrier gas. The surface chemical states of elements on different samples were characterized by X-ray photoelectron spectroscopy and Auger electron spectroscopy (XPS and AES, Thermo ESCALAB 250 Xi, Waltham, MA, USA), and the shift of the spectra is due to the relative surface charging corrected according to the standard binding energy of C 1 s at 284.6 eV. The oxygen vacancy was measured by electron paramagnetic resonance (EPR, Bruker A300-10/12, Billerica, MA, USA). The real loading amount of Cu in samples was measured by an inductively coupled plasma optical emission spectrometer (ICP-OES, Agilent ICPOES730, Santa Clara, CA, USA). The electrochemical measurement was performed on an electrochemical analyzer (CHI600E) with three electrode cells at room temperature. The working electrodes were made of ITO glass and the corresponding prepared samples. The Na₂SO₄ aqueous solution (0.1 M) was used as an electrolyte and a 300 W Xe lamp (PLS-SXE300/300UV) was used as the light source. Photoluminescence spectra (PL) were recorded on a confocal laser Raman microscope (HORIBA FLuoroMax+, Kyoto, Japan) using a 310 nm excitation light source; the time-resolved transient PL decay of samples was measured by transient fluorescence spectrometer (Edinburgh FLS 980, Edinburgh, UK).

The in situ FT-IR was tested on a Bruker Tensor II spectrometer (Billerica, MA, USA). The samples were loaded into the in situ reaction tank of the infrared spectrometer, and the sample was pretreated for 1 h under vacuum at 80 °C. The photocatalytic reaction process was simulated: CO₂ and 0.5 mL H₂O were slowly injected, the adsorption–desorption equilibrium was reached after adsorption for 1 h, and the reaction system was illuminated. The infrared spectra under different illumination times were collected and the process of CO₂ reduction catalyzed by catalysts was analyzed. FT-IR spectra of adsorbed CO were measured using the same pretreatment method. After pretreatment, CO was slowly injected, adsorbed for 1 h, and then excess CO was removed. The temperature of all samples gradually increased from 30 °C to 50 °C and was measured after being stabilized at each temperature for 5 min.

2.4. Photocatalytic Reduction Reaction of CO₂

The performance of photocatalytic CO₂ reduction was tested in a custom-made quartz glass reactor with a volume of 0.3 L. The 20 mg catalyst powder sample was laid on the surface of the small circular table at the bottom, and 1 mL deionized water was uniformly added around the circular table. After sealing with a quartz lid, CO₂ gas was poured into the reactor for 1 h to empty it, and the reactor was filled with CO₂ at the same time. After ventilation, both ends of the reactor were sealed. A 300 W xenon lamp (PLS-SXE300/300UV, China) was used as the lamp source for analysis by gas chromatograph (Shimadzu, GC-2018). Xenon lamps were illuminated from top to bottom through the quartz glass cover of the reactor, and 1 mL of the gas in the reactor was collected every 1 h and injected into the gas chromatograph for quantitative detection of CO and CH₄. The retention time and standard curve of each component gas were obtained by detecting the standard gas.

3. Results and Discussion

3.1. The Structure and Morphology

The crystal structure of the catalyst was analyzed using X-ray diffraction (XRD) and Raman spectroscopy. As shown in Figure 1a, XRD patterns of CuOx/TiO₂ show typical anatase-phase TiO₂ (JCPDS No. 84-1286) crystal structure, indicating that the introduction of copper does not change the crystal structure of TiO₂. The diffraction peaks observed at 2θ = 36.1° and 61.4° correspond to the (111) and (220) crystal faces of Cu₂O (JCPDS No. 78-2076), respectively. The diffraction peaks observed at 2θ = 43.3° and 50.4° correspond to the (111) and (200) crystal faces of Cu (JCPDS No. 85-1326), respectively, indicating that copper species are successfully supported on the TiO₂ surface in the form of Cu₂O or Cu. As shown in Figure 1b, Raman diagrams of different catalyst samples all show the Raman characteristic peaks of anatase-phase TiO₂, and the peak position of anatase-phase TiO₂ does not shift after the introduction of copper, further indicating that the crystal phase of anatase-phase TiO₂ remains unchanged before and after the reaction.

Figure 1.

(a) XRD patterns and (b) Raman of CuO_x/TiO₂-y (y = 0.5, 1, 2, 3, 5).

The surface morphology of the catalyst was analyzed by transmission electron microscopy (TEM). The results reveal that, as shown in Figure 2a,b, there is no significant difference in the morphology of CuO_x/TiO₂-2 and TiO₂, and the size is about 50 nm nano-sheets, indicating that the introduction of copper has no effect on the morphology of TiO₂. This is further confirmed by BET surface area measurements. As shown in the figure, TiO₂, CuO_x/TiO₂-2, and CuO_x/TiO₂-5 have no obvious differences in specific surface area. The results show that the addition of Cu cannot change the surface structure of TiO₂, which is consistent with the results of TEM. The HRTEM image of CuO_x/TiO₂-2 is shown in Figure 2c. The lattice fringes with a lattice spacing of 0.246 nm, 0.208 nm, and 0.351 nm correspond to the Cu₂O (111) crystal face, Cu (111) crystal face, and anatase-phase TiO₂ (101) crystal face, respectively. This is consistent with the XRD results. In Figure 2d–f, element distribution mapping shows the uniform distribution of Ti, O, and Cu elements. The loading capacity of Cu species on CuO_x/TiO₂-2 was 1.16 wt.% by ICP-OES. The above results show that Cu species are uniformly supported on the surface of TiO₂ in the form of Cu0 and Cu₂O, and are closely bound to TiO₂.

Figure 2.

(a) TEM diagram of TiO₂, (b) TEM diagram, (c) HRTEM diagram, and (d–f) mapping diagram of CuOx/TiO₂-2.

3.2. Copper Valence State on the Surface of CuO_x/TiO₂

In the process of the photocatalytic reduction of CO₂ by CuO_x/TiO₂, Cu species are the active center of the catalytic reduction reaction, and the valence of Cu plays an important role in the selectivity of CO₂ products by photocatalytic reduction. According to the XRD pattern in Figure 1a, it can be preliminarily concluded that the present state of Cu changes regularly with the additional amount of glycol as a reducing agent. The valence state of Cu in the catalyst was further analyzed qualitatively and quantitatively with XPS and Auger electron spectroscopy (AES). Firstly, XPS was used to characterize the peak position of each valence Cu component on the catalyst surface for qualitative analysis, and the proportion of the peak area of each valence Cu component to the total peak area of all Cu components was calculated for quantitative analysis. As shown in Figure S2a–e, in the XPS spectrum of Cu 2p, the characteristic peaks at the binding energies of 932.08 eV and 952.08 eV correspond to the Cu 2p_3/2 and Cu 2p_1/2 orbitals of Cu⁺/Cu⁰, respectively, occupying the majority of the peak area of copper components. Its proportion increased slightly with the increase in the amount of ethylene glycol. The characteristic peaks at 934 eV and 935 eV correspond to the Cu 2p_3/2 and Cu 2p_1/2 orbitals of Cu²⁺, respectively, and occupy a small part of the peak area of the copper component, which decreases slightly with the increase in the amount of ethylene glycol added. Figure S2f shows that the Cu component mainly exists in the form of Cu⁺/Cu⁰, but the two components cannot be distinguished by XPS and need to be further analyzed with Auger electron spectroscopy (AES).

The AES energy spectrum of Cu LMM is shown in Figure 3a–e. The characteristic peak at 569.6 eV binding energy corresponds to Cu⁺, the characteristic peak at 567.6 eV binding energy corresponds to Cu⁰, and the characteristic peak at 564.5 eV is related to the Ti 2s orbital. Figure 3f shows that the proportion of Cu⁰ on the catalyst surface increases with the addition of ethylene glycol, while the proportion of Cu⁺ decreases. Due to the fact that the catalyst cannot avoid contact with oxygen in the air during the test, and the copper valence state is unstable, the test results show that the copper component cannot exist completely in the form of Cu⁰.

Figure 3.

Cu LMM Auger spectra of (a–e) CuO_x/TiO₂-y (y = 0.5, 1, 2, 3, 5); (f) proportion of different states of Cu⁺ and Cu⁰ components in CuO_x/TiO₂.

3.3. Performance of Photocatalytic CO₂ Reduction

The performance of the catalyst for photocatalytic CO₂ reduction was tested. The yields and selectivity of CO₂ products for different samples are shown in Figure 4a. The photocatalytic CO₂ reduction performance of pure TiO₂ was poor, with a CO yield of 1.41 μmol·g⁻¹·h⁻¹ and CH₄ yield of 0.28 μmol·g⁻¹·h⁻¹. The photocatalytic CO₂ reduction performance of TiO₂ was improved after loading CuO_x. Comparing the photocatalytic activities of different samples, it can be found that with the increase of reduction degree, the Cu valence state decreases, and the CO yield gradually increases. The CO yield of CuO_x/TiO₂-5 is up to 10.68 μmol·g⁻¹·h⁻¹, and the CO selectivity is up to 80.12%. The CH₄ yield of CuO_x/TiO₂-2 was up to 10.8 μmol·g⁻¹·h⁻¹, and the selectivity of CH₄ was up to 71.9%. Comparing the selectivity of CO₂ products by photocatalytic reduction of different samples, it can be found that Cu²⁺ has little effect on the selectivity of products. With the gradual reduction of Cu²⁺, the selectivity of CO first decreases and then gradually increases, and the selectivity of CH₄ first increases and then decreases. According to the analysis of the proportion of copper content in different valence states in Figure 4b, it can be seen that copper species Cu⁺ and Cu⁰ help to improve the reduction ability of the catalyst. With the increase of Cu⁰ content, the selectivity of CO gradually increases, indicating that Cu⁰ is conducive to the selective conversion of CO, while Cu⁺ is conducive to the selective conversion of CH₄. These results indicate that the valence state of Cu may be the key to the selectivity of CO₂ products in photocatalytic reduction. As shown in Table 1, CuO_x/TiO₂ has higher photocatalytic activity than other photocatalysts.

Figure 4.

(a) Photocatalytic CO₂ reduction activity of CuO_x/TiO₂ and (b) proportion of different states of Cu components in CuOx/TiO₂.

Table 1.

Performance comparison between CuO_x/TiO₂ and other photocatalysts.

Catalyst Name	Intended Product	Productivity (µmol g−1 h−1)	Ref.
CuOx/TiO2	COCH4	10.6810.8	This work
TiO2/Cu2O	CO	10.22	[55]
TiO2/Cu2O	CH4	1.35	[56]
Cu:TiO2-CS	COCH4	4.485.34	[57]
Cu2O/S-TiO2/CuO	CH4	2.31	[58]
NH2–B–TiO2–CuO	COCH4	3.836.6	[59]

In the process of photocatalytic CO₂ reduction, considering that many factors have certain effects on the activity and selectivity of the reaction, we tested the photocatalytic CO₂ reduction activity of CuO_x/TiO₂-2 under different reaction conditions. As shown in Figure S3, a small amount of CO and CH₄ were produced in the Ar atmosphere, which was generated by a small amount of carbon-containing reagents remaining in the catalyst preparation process. Compared with the results of the reaction activity and product selectivity in the CO₂ atmosphere, the results were negligible, indicating that the carbon source of the photocatalytic reduction reaction products mainly came from CO₂ gas. In the dark state, no CO₂ reduction products were detected in the reaction system of the catalyst, indicating that light is the necessary condition for the CuO_x/TiO₂ photocatalytic reduction of CO₂ reaction.

3.4. The Selectivity of Products in Photocatalytic Reduction of CO₂

The adsorption of CO₂ by a catalyst is the first step of the photocatalytic CO₂ reduction reaction. The physical and chemical adsorption capacity of different samples for CO₂ is analyzed by using a specific surface area and an aperture analyzer and chemical absorption desorption instrument (CO₂/TPD). As shown in Figure 5a, the physical adsorption capacity of the catalyst remained basically unchanged after the introduction of copper, which was consistent with the test results of the N₂ resorption–desorption curve. In Figure 5b CO₂/TPD, the test results show that the CO₂/TPD curve of TiO₂ shows the desorption of CO₂ at low temperatures, and the corresponding temperatures of desorption peaks are 101 °C and 322 °C, respectively, indicating that the interaction between CO₂ and TiO₂ is relatively low. When the desorption temperatures of CuO_x/TiO₂-2 and CuO_x/TiO₂-5 are higher, the desorption temperatures are 356 °C, 500 °C, 345 °C, and 475 °C respectively, indicating that the interaction between CO₂ and CuO_x/TiO₂-2 and CuO_x/TiO₂-5 is stronger. The results show that CuO_x cocatalyst can enhance the interaction between CO₂ and photocatalyst. CuO_x/TiO₂-2 showed a higher desorption peak than CuO_x/TiO₂-5, indicating that Cu⁺ is more conducive to the chemical adsorption of CO₂ and the subsequent photocatalytic reduction of the CO₂ reaction process.

Figure 5.

(a) CO₂ physical adsorption image, (b) CO₂ chemisorption (CO₂/TPD) image, (c) UV–Vis DRS image, (d) I–t curve, (e) steady-state fluorescence spectrum image, and (f) transient fluorescence lifetime image of CuO_x/TiO₂.

The UV–Vis DRS spectra are shown in Figure 5c. Compared with pure TiO₂, CuO_x/TiO₂ had visible light absorption, indicating that the introduction of copper increased the light absorption range of the catalyst. In addition, the absorption band edge of the catalyst remains basically unchanged (387 nm), indicating that the supported Cu does not change the band gap of TiO₂, which is consistent with the results of XRD and Raman. As shown in Figure 5d, CuO_x/TiO₂ has a higher photocurrent response than pure TiO₂, among which CuO_x/TiO₂-2 has the strongest photocurrent response, indicating that the introduction of copper improves the transport capacity of photogenerated electrons and increases the mobility of electrons. However, the selectivity of the photocatalytic reduction of CO₂ products cannot be directly determined by electron migration. Therefore, steady-state fluorescence spectra and fluorescence lifetime test were used to further analyze the lifetime of photogenerated electrons. As shown in Figure 5e,f, the catalyst has strong characteristic peaks at wavelengths of 394 nm and 466 nm, and the fluorescence intensity decreases after the introduction of copper, which indicates that the presence of Cu can promote the migration of photogenerated electrons to the surface of the catalyst, inhibiting the recombination of photogenerated electrons and holes. The fluorescence lifetime of TiO₂ is 6.00 ns, the fluorescence lifetime of CuO_x/TiO₂-2 is 6.40 ns, and the fluorescence lifetime of CuO_x/TiO₂-5 is 6.21ns. The introduction of copper can migrate electrons from TiO₂ to CuO_x, extending the electron lifetime. From the perspective of reaction kinetics, since the generation of CH₄ is an eight-electron reaction, the catalyst with a long photo-generated electron lifetime is more likely to generate CH₄ products in the photocatalytic reduction of CO₂, which also explains the higher selectivity of CuO_x/TiO₂-2 photocatalytic reduction of CO₂ to CH₄.

According to the semiconductor band gap (E_g) formula, (αhν)ⁿ = k(hν − Eg), the Tauc of different samples is calculated, as shown in Figure 6a. The band gap of TiO₂ is about 3.2 eV, and the band gap remains basically unchanged after loading CuO_x. VB-XPS was used to directly test the valence band position of the catalyst, and the results are shown in Figure 6b. The valence band values of different samples under standard hydrogen electrodes were calculated according to the following formula: E_VB,NHE = φ + E_VB,XPS − 4.44, where φ is the work function of the instrument (4.5 eV). Therefore, the E_{VB, NHE} of CuO_x/TiO₂-2 and CuO_x/TiO₂-5 are calculated to be 2.42 eV and 3.32 eV, respectively. According to the formula E_VB = E_CB + E_g, the conduction bands (E_CB) of CuO_x/TiO₂-2 and CuO_x/TiO₂-5 are −0.75 eV and −0.83 eV, respectively. As shown in Figure 6c, the band conduction position of TiO₂ was calculated to be −0.41 eV. Based on the above results, the band gap relationship of different samples is shown in Figure 6d. The conduction position of CuO_x/TiO₂ is more negative than that of single TiO₂, indicating that loaded CuO_x can improve the photocatalytic reduction ability of TiO₂ and enhance the photocatalytic reduction activity of CO₂. The conduction positions of CuO_x/TiO₂-2 and CuO_x/TiO₂-5 are more negative than the reaction potentials of CH₄ (CH₄/CO₂, −0.24 V vs. NHE) and CO (CO/CO₂, −0.52 V vs. NHE). The results show that the photocatalytic reduction of CO₂ to produce CH₄ and CO is thermodynamically feasible. Combined with the UV–Vis DRS results, the effect of different valence states of Cu on the redox potential of CuO_x/TiO₂ catalyst is not significant. In other words, under these conditions, the changes in product yield and the selectivity of CuO_x/TiO₂ photocatalytic reduction of CO₂ are not determined by the redox potential of the catalyst.

Figure 6.

(a) Tauc, (b) VB-XPS spectrum, (c) Mott–Schottky curve of TiO₂ and (d) band structure of catalyst.

Carbon monoxide (CO) is not only a significant product of photocatalytic CO₂ reduction but also serves as an important reaction intermediate. The key to producing high-value products lies in the continued adsorption of CO on the catalyst surface, preventing its desorption during the photocatalytic CO₂ reduction process. Fourier Transform Infrared Spectroscopy (FT-IR) was employed to assess the adsorption capacity of various samples for CO. A stronger interaction between the catalyst and CO correlates with increased difficulty in desorption from the catalyst surface. As temperature rises, the rate of decrease in the *CO signal intensity diminishes. The reduction rates of the *CO characteristic peak intensities at 2171 cm⁻¹ and 2100 cm⁻¹ can be utilized to characterize the CO adsorption capacity of the photocatalyst [60,61,62].

As shown in Figure 7a, when the temperature rises to 30 °C, the *CO signal on the pure TiO₂ sample begins to decline rapidly, and when the temperature rises to 45 °C, the *CO adsorption peak completely disappears, indicating that the interaction force between CO and TiO₂ is weak. Compared with a pure TiO₂ sample, the *CO absorption peak of the catalyst CuO_x/TiO₂ decreased at a slower rate, indicating that the interaction force between CO and CuO_x/TiO₂ was strong, that is, the main adsorption site of CO was CuO_x. The FT-IR spectra of CO adsorption on CuO_x/TiO₂-2 and CuO_x/TiO₂-5 are shown in Figure 7b,c, respectively. The effects of different valence states of Cu on CO adsorption are compared and analyzed. The results show that the decline rate of *CO signal intensity is as follows: R_TiO2 > R_CuOx/TiO2-5 > R_CuOx/TiO2-2, so the adsorption capacity of CO is CuO_x/TiO₂-2 > CuO_x/TiO₂-5 > TiO₂. This is consistent with the change in the ratio of copper components in the AES spectra. With the decrease of the ratio of Cu⁺ and the increase of the ratio of Cu⁰, the adsorption capacity of CO decreases, and CO is more easily resolved to form CO products. This shows that the adsorption capacity of Cu⁺ for CO is stronger than that of Cu⁰, which is conducive to further hydrogenation of CO, and the final product CH₄ is formed through a carbene pathway.

Figure 7.

FT-IR spectra of CO adsorbed on (a) TiO₂, (b) CuO_x/TiO₂-2, and (c) CuO_x/TiO₂-5.

As shown in Figure 8a, the adsorption peaks of HCO₃⁻ (1415 cm⁻¹), m-CO₃²⁻ (1506 and 1447 cm⁻¹), and b-CO₃²⁻ (1576 and 1522 cm⁻¹) can be observed in the in situ FT-IR spectra of CuO_x/TiO₂-2. The intensity of these characteristic adsorption peaks increased gradually with the extension of adsorption time, but the location did not change. COOH* (1558 cm⁻¹), *CHO (1102 cm⁻¹), and CH₃O* (1041 cm⁻¹) peaks appeared and increased with the increase of light time. They are all important intermediate species in the process of photocatalytic CO₂ reduction. In addition, CH₃O*, *CHO, and CH₂· (1373 cm⁻¹) [41] participate in the reaction as important intermediates of CH₄, which explains the high selectivity of the CuO_x/TiO₂-2 catalyst for CH₄ generation. It is speculated that the main conversion pathway of CH₄ is the carbene pathway: CO₂ → *COOH → *CO → *CHO → C· → ·CH₂ → ·CH₃ → CH₄. As shown in Figure 8b, in the photocatalytic CO₂ reduction process of CuO_x/TiO₂-5, the intermediates are mainly CO₃²⁻, HCO₃⁻, and *COOH, and no absorption peak of methane intermediates is observed. This indicates that *CO does not accumulate on the catalyst surface for further conversion but is quickly released into the air and converted into the final product CO. This is consistent with the results of the CO adsorption FT-IR, which explains the high selectivity of the CuO_x/TiO₂-5 photocatalytic reduction of CO₂ to produce CO. It is speculated that the main conversion pathway of CO is as follows: CO₂ → HCO₃⁻/CO₃²⁻ → *COOH → CO. Different catalysts produce different reaction intermediates in the light process, which directly affect the yield and selectivity of the final photocatalytic reduction of CO₂ products.

Figure 8.

In situ FT-IR spectra of (a) CuO_x/TiO₂-2 and (b) CuO_x/TiO₂-5.

Building on the aforementioned results, a reaction mechanism for the CuO_x/TiO₂ photocatalytic reduction of CO₂ has been proposed. The CuO_x species supported on the TiO₂ surface serve as a reactive site that effectively harnesses photoelectron generation, enhances charge separation efficiency, and boosts the photocatalytic activity for CO₂ reduction. The results from Auger electron spectroscopy (AES) and evaluations of CO₂ photoreduction performance indicate that the selectivity towards methane (CH₄) and carbon monoxide (CO) in the products is closely linked to the relative content ratio of Cu⁺ and Cu⁰ present on the catalyst surface. Furthermore, it was observed that the adsorption and activation of *CO intermediates on this surface significantly influence final product formation. CO readily desorbs from Cu0 to yield CO, while CH₄ undergoes further adsorption and hydrogenation at Cu⁺, thereby confirming that Cu⁺ acts as an active site for CH₄ production, whereas Cu⁰ functions as an active site for CO generation (Figure 9).

Figure 9.

Schematic image of CuO_x/TiO₂ photocatalytic reduction of CO₂.

4. Conclusions

The CuO_x/TiO₂ photocatalyst was synthesized via an in situ growth reduction method, enabling selective regulation of CO₂ photoreduction products by modulating the valence state of copper. The valence state of copper is a critical determinant influencing the selectivity of CO₂ reduction products. Cu⁺ serves as the active site for methane (CH₄) formation, while Cu⁰ acts as the active site for carbon monoxide (CO) production. Notably, CO is not only a significant product of CO₂ photoreduction but also functions as an essential reaction intermediate. Cu⁺ exhibits strong adsorption and activation capabilities towards CO, thereby facilitating the conversion of *CO intermediates into high-value CH₄ products. This study offers valuable insights for advancing research on copper-based photocatalysts and identifying highly efficient and selective catalysts for CO₂ reduction.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/nano14231930/s1, Figure S1: N₂ adsorption and desorption isotherms of TiO₂, CuO_x/TiO₂-2 and CuO_x/TiO₂-5; Figure S2: XPS of Cu2p spectra (a–e) CuO_x/TiO₂-y (y = 0.5, 1, 2, 3, 5), (f) Proportion of different state of Cu²⁺ and Cu⁺/Cu⁰ components in CuO_x/TiO₂; Figure S3: Photocatalytic CO₂ reduction activity of CuO_x/TiO₂-2 under different reaction conditions.

Author Contributions

Methodology, S.N.; Investigation, W.W.; Data curation, Z.Y.; Writing—review & editing, M.Z.; Project administration, J.Y. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Data are contained within the article or Supplementary Materials.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Funding Statement

This work was financially supported by the Natural Science Foundation of Henan Province (No. 222300420406, 242300421346).