To alleviate the climate change caused by carbon cycle imbalance and reduce the dependence on fossil fuels, conversion of CO₂ into value added chemicals and fuels is one promising choice [1-4]. Considering that solar energy is clean and inexhaustible to a certain extent, photocatalysis is thought to play a crucial role in the future industry system and has been widely investigated [5,6]. Thus, photocatalytic CO₂ conversion is a desirable approach for CO₂ utilization. However, insufficient efficiency limits the application of photocatalytic CO₂ conversion in large-scale industrial production. Although various strategies have been applied to improve the photocatalytic performance, the efficient CO₂ conversion is still a vital challenge.

Photothermal catalysis, which has the potential to utilize the whole solar spectrum, is one desirable choice for this issue [7,8]. During the photothermal catalysis, while the short-wavelength light could be converted into photogenerated electron-hole pairs to drive the photochemical reactions, the long-wavelength light could be converted into heat, accelerating the reaction, especially for the surface endothermic reaction [9,10]. Considering the high efficiency of thermochemical method, the combination of photocatalysis and thermocatalysis could strike a balance between energy consumption and conversion efficiency, which has a broad application prospect in CO₂ methanation [11-15].

To achieve desirable synergy between photochemical and thermochemical, the construction design of catalyst is crucial. TiO₂ has been widely applied in photocatalysis [16,17]. However, the photocatalytic activity of TiO₂ is greatly limited due to its poor response ability to visible light [18,19]. To solve this issue, Chen et al. prepared TiO_2-x (black TiO₂) from TiO₂ by strong reduction treatment, and black TiO₂ exhibits excellent spectral response range and photocatalytic activity [20,21]. Thus, black TiO₂ has attracted increasing attention and been widely applied in photocatalysis such as photocatalytic water splitting. Additionally, because of its inherent black color, black TiO₂ could perform high photothermal conversion ability. Considering the broad spectral response range and excellent photothermal conversion efficiency of black TiO₂ [22-25], it is reasonable to speculate that black TiO₂ could balance the photoelectric conversion and photothermal conversion during the photothermal catalytic CO₂ conversion, giving promising catalytic activity. Therefore, in the present work, supported black TiO₂ was synthesized for light-driven catalytic hydrogenation of CO₂ to CH₄ using ruthenium as the active metal, and the synergic mechanism between photocatalysis and thermocatalysis in the CO₂ conversion was investigated.

Firstly, as shown in Table 1, no CH₄ was detected in all entries without irradiation. Considering the high thermodynamic stability of CO₂, it is difficult to drive the hydrogenation of CO₂ without additional energy input at low temperature. Subsequently, the introduction of irradiation brought significant increase in CH₄ yield (Table 1, entries 3 and 4). Most importantly, the CH₄ yield achieved over 2 wt% Ru/black TiO₂ under irradiation was 40.1%, which was 2 times higher than that achieved over 2 wt% Ru/TiO₂. Additionally, small amount of CO was also produced in entries 3 and 4. CO species are commonly considered as the important intermediate during the CO₂ hydrogenation. Additionally, we performed the experiments without CO₂ and no carbon containing products were detected, indicating that the injected CO₂ is the sole carbon source. There are two possible reasons for the remarkable improvement of the catalytic activity of Ru/black TiO₂: (1) The enhancement of CO₂ adsorption capacity; (2) the improvement of spectral response capacity and photothermal conversion efficiency. In order to verify the above conjectures, the CO₂ adsorption capacity and spectral response ability of the catalysts were investigated, and the photothermal conversion performances of the catalysts were evaluated.

Firstly, the light response characteristics of the prepared materials were investigated and Fig. 1a shows the UV-vis diffuse reflectance spectra (DRS) of TiO₂ and black TiO₂. The spectral response ability of black TiO₂ in long wavelength range is much higher than that of TiO₂. It is possible that, after the hydrogenation at high temperature, the intrinsic symmetry of TiO₂ crystal was broken, forming an amorphous shell on the TiO₂ surface [26,27]. Thus, a secondary narrow bandgap would be introduced into TiO₂, enhancing the light absorption. Additionally, considering the more significant thermal effect of long wavelength light, the enhanced long wavelength light response ability of black TiO₂ can make black TiO₂ perform better photothermal conversion efficiency under the broad wavelength irradiation. Subsequently, the CO₂ adsorption observation was conducted. As shown in Fig. S1 (Supporting information), the CO₂ adsorption amount of both TiO₂ and black TiO₂ increased with the increase of pressure and the CO₂ adsorption ability of black TiO₂ was better than that of TiO₂. The enhanced CO₂ adsorption ability of black TiO₂ gives the prepared catalysts better catalytic activity. Furthermore, nitrogen adsorption/desorption isotherms of TiO₂ and black TiO₂ were compared. The Brunauer-Emmett-Teller (BET) specific surface areas of TiO₂ and black TiO₂ are 99.8 m²/g and 100.1 m²/g, respectively (Figs. S2a and b in Supporting information). The hydrogenation treatment has not brought significant change of specific surface area to black TiO₂. Thus, it is reasonable to speculate that the enhanced CO₂ adsorption capacity can be attributed to the formation of oxygen vacancy in black TiO₂ [28,29].

Then, the textural properties were characterized by X-ray diffraction (XRD). As shown in Fig. 1b, the XRD patterns of the catalysts displayed no obvious differences in crystal form compared to the original TiO₂. However, it is worth noting that the peak intensity of black TiO₂ is significantly lower than that of TiO₂ and the (101) plane characteristic peak of black TiO₂ shifts to larger diffraction angle, which can be owing to the decrease of the intrinsic material crystallinity and formation of lattice defects [30]. Subsequently, X-ray photoelectron spectroscopy analysis was applied to investigate the 2 wt% Ru/black TiO₂ (Figs. S3 and S4 in Supporting information). In the XPS diagram of O 1s (Fig. S4), it can be seen that the O 1s can be divided into two peaks, among which the peak at 529.52 eV is attributed to the lattice oxygen (O_L), while the peak at 530.73 eV can be attributed to the oxygen vacancy (O_V) in black TiO₂ [31,32]. During the high temperature reduction of TiO₂, part of the oxygen atoms were deprived, causing the decrease of the electron cloud density and thereby increasing the binding energy of O 1s. Additionally, the existence of oxygen vacancies was proved by electron paramagnetic resonance (EPR) measurements. As shown in Fig. 1c, a strong EPR signal (centering at g = 2.003) can be observed in black TiO₂ EPR spectra while TiO₂ only exhibits an extremely weak EPR signal [33]. It indicates that lots of oxygen vacancies were produced in black TiO₂, which accords with the XPS analysis results.

To further investigate the morphology and structure of the catalysts, high-resolution transmission electron microscopy (HRTEM) were conducted. The measured average particle sizes of TiO₂ and black TiO₂ are 22.5 ± 3.0 nm and 28.1 ± 4.7 nm, respectively (Fig. S5 in Supporting information). The agglomeration of the catalysts after calcination at high temperature is one possible reason for the increase of average particle size. Additionally, as shown in Figs. 1d and e, Ru nanoparticles in the 2 wt% Ru/TiO₂ and 2 wt% Ru/black TiO₂ catalysts are loaded on the support surface in the form of small particle and the contact interface between Ru and support can be clearly observed. The lattice fringes belonged to the Ru (002) crystal facet are clearly observed [34]. Owing to the well-contacted interface, efficient photogenerated carriers transferring can be constructed during the reaction. In addition, the lattice fringes of TiO₂ in 2 wt% Ru/TiO₂ catalyst are orderly arranged, and the lattice spacing is 0.355 nm, belonging to (101) crystal facet of anatase TiO₂ [35]. However, it is difficult to observe continuous lattice fringes in 2 wt% Ru/black TiO₂. The removal of O during reduction treatment broke the stoichiometric balance of TiO₂, forming amorphous phase [20,36]. This result also agrees with the XRD analysis result that the crystallinity of black TiO₂ has been weakened structurally.

As aforementioned that the increase of catalytic activity of Ru/black TiO₂ can also be attributed to the enhanced photothermal conversion ability, therefore, the photothermal conversion performances of Ru/TiO₂ and Ru/black TiO₂ were compared. Firstly, surface temperatures of the prepared catalysts under irradiation were directly observed based on IR temperature measuring device. As shown in Fig. S6 (Supporting information), the rising rate of the surface temperature of 2 wt% Ru/black TiO₂ was significantly higher than that of 2 wt% Ru/TiO₂ under irradiation. After 300 s, the surface temperature of 2 wt% Ru/black TiO₂ reached 283.7 ℃, while the temperature of 2 wt% Ru/TiO₂ was only 223.1 ℃. It can be seen from Fig. S7 (Supporting information) that the overall temperature of reaction system raised gradually during the reaction due to the photothermal conversion effect of the catalyst.

Furthermore, the photothermal conversion efficiency of 2 wt% Ru/black TiO₂ was evaluated. The materials were dispersed in aqueous solution with water as the negative control. Initial temperature of the solution and room temperature were controlled at 20 ℃. The solutions were irradiated under the Xe lamp and the lamp was turned off after 1200 s, and the temperature change was observed for 3000 s (Figs. 2a and b). The highest temperature of 2 wt% Ru/black TiO₂ aqueous solution is 79.6 ℃ while the temperature of TiO₂ aqueous solution is only 53.8 ℃ (Fig. 2a). It is worth noting that the highest temperature of 2 wt% Ru/black TiO₂ solution is slightly lower than that of black TiO₂. Considering that near-infrared (NIR) absorption ability has marked impact on photothermal conversion [37], UV-vis-NIR DRS of the catalysts were conducted. As shown in Fig. S8, TiO₂ displays poor NIR response and the loading of Ru improve its NIR absorption. However, there is no significant difference between the spectra of black TiO₂ and Ru/black TiO₂, which means that the intrinsic NIR absorption of black TiO₂ is higher than that of Ru NPs. Therefore, the photothermal conversion capacity of black TiO₂ is slightly higher than that of Ru/black TiO₂. The photothermal conversion efficiency (η) can be calculated based on Eq. 1 [37-39]:

(1)

where h, A, I, ΔT_{max, mix} and ΔT_{max, H2O} represent the heat transfer coefficient, the container surface area, the total energy Ru/black TiO₂ nanoparticles absorbed, the temperature change of the 2 wt% Ru/black TiO₂ suspension and water at the maximum stable temperature, respectively. Based on the fitting parameters of the cooling period in Fig. 2c, the η value of 2 wt% Ru/black TiO₂ is calculated to be 37.9%. The calculation details can be seen in the Supporting Information. Owing to the efficient photothermal conversion based on black TiO₂, strong heat center can be produced on the catalyst surface under irradiation, improving the substances (CO₂ and H₂) activation and accelerating the reaction [40].

Subsequently, to verify the participation of photogenerated carriers during the CO₂ conversion, the irradiation intensity was decreased to 1 sun (100 mW/cm²) and a series of experiments at different temperatures were conducted. The experiments were carried out at 180, 200, 220 and 240 ℃, respectively. As shown in Fig. 3, the methane yield increased with the increase of temperature, which indicates that the increase of reaction temperature is conducive to the methanation of CO₂. Most importantly, the CH₄ yields increased at all temperatures after the introduction of light and highest synergy was obtained at 220 ℃, which means that intrinsic excitation process of the catalysts may played an important role during the reaction.

Considering the large band gap of Al₂O₃, Ru/Al₂O₃ was prepared to avoid the excitation of catalyst and its catalytic performance was observed to further confirm the effect of photogenerated carriers. As displayed in the Fig. 3, the CH₄ yield showed no significant increase over Ru/Al₂O₃ after the introduction of light. Thus, it is reasonable to speculate that, during the hydrogenation of CO₂ over Ru/black TiO₂, not only the heat input generated from photothermal conversion but also the photogenerated carriers promote the conversion of CO₂. The role photogenerated carriers played during CO₂ conversion were investigated in mechanism discussion section.

According to the above results and the reported CO₂ hydrogenation mechanism, the functionary mechanism of photogenerated carrier was proposed (Fig. S9 in Supporting information). Due to the low Fermi level of Ru and the well-contacted interface between Ru nanoparticle and black TiO₂ nanoparticle (Fig. 1f), it is reasonable to speculate that the photogenerated electrons would be gathered in Ru nanoparticles, forming electron-rich Ru surface. Furthermore, to monitor the electron density variation of Ru nanoparticles, in situ high-resolution XPS was carried out. As displayed in Fig. 4, the binding energy of Ru 3d_5/2 shows a significant decrease after the introduction of light. In addition, with the gradual increase of irradiation intensity and irradiation time, the binding energy of Ru 3d_5/2 displayed a further shifting tendency.

Considering that the increase of electron density can enhance the shielding effect of electron clouds, which would decrease the binding energy of Ru 3d core level, it is reasonable to speculate that an electron injection process is achieved during the reaction. In contrast, the introduction of light caused the increase of O 1s binding energy (Fig. S10 in Supporting information), indicating that oxygen is in electron deficient state, which could be owing to the photogenerated holes. It is worth noting that the activation of H₂ is essential for the CO₂ hydrogenation. The formation of electron-rich Ru surface can provide more negative hydride, thus enhancing the nucleophilic attack reactivity of the hydride to the carbon center of CO₂ [41,42]. Besides, the light with larger wavelength could be converted into heat due to the photothermal conversion, further improving the CO₂ conversion. It is one promising way to achieve efficient hydrogenation of CO₂ using light as the only driving force based on the synergy between photocatalysis and thermocatalysis.

Additionally, a series of Ru/black TiO₂ catalysts with different Ru loading amount were prepared and the circulation stability test was carried out. The CH₄ yield increased with the increase of Ru loading amount and 93.8% of CH₄ yield was achieved within 2 h over 5 wt% Ru/black TiO₂ (Fig. S11 in Supporting information). The H₂-temperature programmed desorption (TPD) was carried out to investigate the adsorption strength between the adsorbed hydrogen and catalysts. As displayed in Fig. S12 (Supporting information), all tested catalysts displayed two significant hydrogen desorption peaks. The desorption peaks around 270 ℃ could be ascribed to the weakly adsorbed hydrogen on Ru surface and the peaks at high temperature (around 340 ℃) could be ascribed to the strongly adsorbed hydrogen on the surface of black TiO₂ caused by hydrogen spillover [43-45]. Such spillover could do contribution to the hydrogenation of CO₂. Additionally, no significant decrease in CH₄ yield was observed during the stability test of 5 wt% Ru/black TiO₂ (Fig. S13 in Supporting information). The XRD patterns of the collected catalysts after stability test display similar XRD patterns (Fig. S14 in Supporting information) and no high temperature phases such as rutile TiO₂ were produced after reaction, indicating that the catalyst can maintain desirable stability under the current reaction conditions.

In summary, a black TiO₂ supported catalyst was constructed for the photothermal catalytic CO₂ hydrogenation. Owing to the introduction of oxygen vacancies, the CO₂ adsorption ability of the black TiO₂ supported catalyst was significantly enhanced, improving the CO₂ conversion. The wide spectral response range and high photothermal conversion efficiency of black TiO₂ brought desirable catalytic performance for photocatalytic hydrogenation of CO₂ to the prepared Ru/black TiO₂. Additionally, based on the well-contacted interface between Ru nanoparticle and black TiO₂ nanoparticle, an electron injection process was achieved to form electron-rich Ru metal nanoparticles, further improving the hydrogenation of CO₂. This study opens the step for full wavelength range utilization of solar energy in light-driven conversion of CO₂.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors gratefully acknowledge the support from the National Natural Science Foundation of China (No. 21978170), the National Key R & D Program of China (No. 2017YFC0506004), the Natural Science Foundation of Shanghai (No. 19ZR1424800), and the Center of Hydrogen Science, Shanghai Jiao Tong University, China.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2021.07.046.