Tailoring Electronic Structures via Ce/C Co-Doping and Oxygen Vacancy in TiO2 Aerogels for Enhanced Solar Fuel Production

TL;DR

Researchers improved TiO2 aerogels by adding cerium and carbon, creating oxygen vacancies that boost solar fuel production from CO2 under sunlight.

Contribution

A novel Ce/C co-doping strategy combined with oxygen vacancies enhances TiO2 aerogels for efficient CO2 reduction without co-catalysts.

Findings

Ce/C co-doped TiO2 aerogels achieved CH4 and CO production rates 82.0 and 5.7 times higher than pristine TiO2 under simulated sunlight.

Oxygen vacancies formed by C-doping increase light absorption and extend charge carrier lifetimes, improving photocatalytic performance.

DFT calculations and experiments confirm that defect states from oxygen vacancies enable efficient CO2 reduction.

Abstract

A targeted modification approach involving the synthesis of Ce/C co-doped TiO2 aerogels (CeCTi) via a sol–gel method combined with supercritical CO2 drying and subsequent heat treatment is employed to enhance the photocatalytic CO2 reduction performance of cost-effective and stable TiO2 aerogels. The results demonstrate that the CeCTi exhibits a pearl-like porous network structure, an optical band gap of 2.90 eV, and a maximum specific surface area of 188.81 m2/g. The black aerogel sample shows an enhanced light absorption capability resulting from the Ce/C co-doping, which is attributed to the formation of oxygen vacancies. Under simulated sunlight irradiation, the production rates of CH4 and CO reach 27.06 and 97.11 μmol g−1 h−1 without any co-catalysts or sacrificial agents, respectively, which are 82.0 and 5.7 times higher than those of the pristine TiO2 aerogel. DFT reveals that C-doping facilitates the formation of oxygen vacancies, which introduces defect states within the calculational band gap of TiO2. The proposed photocatalytic mechanism involves the light-induced excitation of electrons from the valence band to the conduction band, their trapping by oxygen vacancies to prolong the charge carrier lifetime, and their subsequent transfer to adsorbed CO2 molecules, thereby enabling efficient CO2 reduction, which is experimentally supported by photoluminescence measurements.

Figures

Full text

1. Introduction

In recent years, the expanding scope of industrial activities has precipitated a dual crisis of escalating carbon emissions and energy shortages, where the excessive release of CO_2_ into the atmosphere not only drives climate change but also signifies a substantial waste of carbon resources [1]. Further, designing photocatalysts with CO_2_ reduction performance is of great significance for promoting industrial progress and reducing carbon emissions. Researchers have developed a variety of photocatalysts with different energy band structures, such as TiO_2_ [2], WO_3_ [3], GaN [4], C_3_N_4_ [5], and BiVO_4_ [6], as well as ceramic catalysts fabricated via 3D printing [7]. For a long time, TiO_2_ has been very promising in the field of photocatalytic CO_2_ reduction [8,9] due to its advantages such as low cost, non-toxicity, stable physicochemical properties, and abundant reserves. Lundberg et al. [10] systematically investigated the robust nanoparticle catalysts, including TiO_2_, for CO_2_ photoreduction, and revealed a universal volcano-plot relationship between electron transfer kinetics and catalytic performance. Xu et al. [11] constructed a 2D porphyrin MOF–TiO_2_ hybrid photocatalyst through in situ growth, achieving 94.1% selective CO_2_-to-CO conversion via interfacial Ti-O-N bonds that enhance charge separation and CO_2_ activation. Despite its potential, TiO_2_ suffers from a wide bandgap, easy recombination of photogenerated electron–hole pairs, poor light capture ability, and insufficient active sites, necessitating modifications for improved efficiency [12,13].

An ideal photocatalyst should feature an appropriately tuned bandgap for visible-light absorption, high optical activity with efficient charge separation, and a three-dimensional porous structure with a large specific surface area to maximize active site exposure [14,15]. The aerogels, which are porous materials composed of low-density, high-specific-surface-area, and high-porosity nanoparticles, have been widely used in chemical sensors, catalytic barriers, thermal insulation, and adsorbents [16,17,18,19]. The high specific surface area of aerogels provides abundant reactive sites on the photocatalyst surface, while the continuous porous 3D nano-network skeleton enhances the light-harvesting efficiency through multiple reflections [20,21]. These advantages make aerogels one of the most promising photocatalysts for CO_2_ reduction. Li et al. [22] developed cyanamide-functionalized crystalline carbon nitride aerogels via a molten-salt self-assembly approach, achieving an excellent CO_2_-to-CO conversion rate of 25.7 μmol g^−1^ h^−1^ due to the cyanamide groups acting as active sites to lower the energy barrier for *COOH formation. Saure et al. [23] utilized the high porosity of hybrid aerogel materials fabricated from hollow silica microtubes and trace amounts of reduced graphene oxide to regulate the light scattering and absorption properties, achieving an enhanced volumetric photothermal response. These structural and functional advantages make aerogel-based photocatalysts a key factor in promoting the development of solar-driven CO_2_ conversion technologies. However, the fabrication process is complex, often requiring extended reaction times and post-treatment, which creates formidable barriers for large-scale manufacturing.

Concurrently, doping engineering with elements such as cerium or carbon has been widely employed to modify the wide band gap of TiO_2_, effectively enabling its visible-light photocatalytic activity [24,25]. The Ce-doping boosts photocatalytic performance through their unique redox properties, oxygen vacancy (O_V_) generation, and band structure modulation capabilities. Specifically, the incorporation of Ce element introduces an impurity energy level near the Fermi level, which effectively suppresses the electron–hole pair recombination and extends the optical absorption range from ultraviolet to visible light [26]. Li et al. [27] proposed S-scheme Ni_2_P@ Ce-BDC-CeO_2_ hollow octahedrons with hierarchical pores and O_V_, where Ce-doping enhanced the charge separation and CO_2_ adsorption, achieving a 61.6 μmol g^−1^ h^−1^ CO production rate. Zhang et al. [28] synthesized a Ce-doped TiO_2_ via in situ calcination, achieving a high Li^+^ adsorption capacity of 58.42 mg/g and enhanced selectivity/stability for efficient lithium recovery from brines. Although Ce-doping effectively enhances photocatalysis through O_V_ generation, the limited CO yield and energy-intensive lactic acid etching process necessitate optimized material design and synthesis for performance breakthroughs. C-doping also serves as a pivotal strategy to boost photocatalytic performance through its unique capability to modify the local electronic structure, induce O_V_, and regulate charge carrier dynamics [29]. Wang et al. [30] designed a dual S-scheme SrTiO_3_/SrCO_3_/C-doped TiO_2_ heterojunction via carbon doping, demonstrating enhanced photocatalytic hydrogen production under UV–Vis light compared to individual components. C-doping often limits its efficiency in photocatalysis due to drawbacks such as increased charge recombination centers, insufficient structural stability, and weakened oxidation capability.

However, the systematic understanding of how Ce and C dopants cooperatively function within an aerogel architecture and promote CO_2_ photoreduction without any co-catalysts or sacrificial agents remains limited.

Herein, this study successfully synthesized a pearl-like porous Ce/C co-doped TiO_2_ aerogel (CeCTi) through a sol–gel process combined with supercritical CO_2_ drying and subsequent thermal treatment. The introduction of Ce creates electron traps that prolongs the carrier lifetime and promotes the interface charge transfer. When carbon is co-doped into the lattice, it exhibits strong electronic coupling with Ce, which collectively regulates the spatial distribution and density of O_V_. At elevated temperatures, carbon further participates in the partial reduction of the lattice, reacting with lattice oxygen and evolving as CO_2_, thereby creating additional O_V_ and fine-tuning the local defect architecture. As a result, the CeCTi by heat treatment at 450 °C exhibits outstanding photocatalytic CO_2_ reduction performance under simulated solar irradiation, with CH_4_ and CO production rates reaching 27.06 and 97.11 μmol g^−1^ h^−1^, respectively. The DFT calculations further reveal that Ce/C co-doping progressively narrows the calculational band gap of TiO_2_ from 1.404 eV to 0.463 eV. This work provides a solid foundation for the rational design of Ce/C co-doped TiO_2_-based photocatalysts with enhanced CO_2_ reduction performance in future studies.

2. Results and Discussion

2.1. Chemical Composition and Structural Analysis

Figure 1a illustrates the preparation process of the pearl-like porous CeCTi, which is successfully synthesized through a combined sol–gel method involving supercritical CO_2_ drying and subsequent thermal treatment. During the sol–gel synthesis, TBOT and Ce(NO_3_)3_6H_2_O are used as precursors, and the introduction of R, followed by continuous stirring, promotes condensation reactions between the surface hydroxyl groups of the TiO_2 sol and the phenolic hydroxyl groups of R. After aging and solvent exchange, the wet gel is subjected to supercritical CO_2_ drying and the obtained aerogel is subsequently thermally treated at 450 °C for 2 h under an Ar atmosphere. During the thermal treatment R undergoes pyrolysis accompanied by C-H bond cleavage, generating carbon that is co-doped into the TiO_2_ lattice, and carbon escapes in the form of CO_2_ to create O_V_. Then, the samples combine synergistically with the O_V_ induced by Ce^3+^ substitution for Ti^4+^ to produce the O_V_-CeCTi-450 aerogel, referred to as CeCTi. The XRD patterns of the CeTi, CTi, CeC_0.5_Ti, CeC_0.75_Ti, and CeC_1.0_Ti samples after heat treatment at 450 °C in Figure 1b all show distinct diffraction peaks at 2θ of 25.3°, 37.8°, 47.9°, 54.2°, 55.4°, 62.6°, 70.2°, and 75.1°. The positions of these peaks are in complete agreement with the anatase TiO_2_ (PDF NO. 1-562), and can be indexed to the (101), (004), (200), (105), (211), (204), (220), and (215) crystal planes, respectively [31,32,33]. The pattern of CeTi-450 is nearly identical to TiO_2_, and the absence of characteristic diffraction peaks for metallic Ce or CeO_x_ indicates that Ce does not form other phases but is incorporated into the TiO_2_ lattice as a dopant [34]. The CTi-450 exhibits a reduction in diffraction peak intensity and a weakening in crystallinity, primarily attributed to the C-doping which alters the gelation kinetics of the precursor and inhibits the formation of an ordered titanium dioxide gel network. Notably, the CeCTi-450 shows a recovered diffraction intensity relative to CTi-450, suggesting that Ce incorporation mitigates the inhibitory effect of carbon on gelation, thereby enhancing the overall structural stability. Increasing C-doping leads to an initial enhancement, followed by a reduction in diffraction peak intensity, indicating that an optimal doping level promotes orderly grain growth and prevents full width at half maximum (FWHM) broadening and intensity attenuation. The thermal treatment of CeC_0.75_Ti at different temperatures revealed that the sample achieved an optimal crystalline state at 450 °C, evidenced by a maximum in diffraction intensity and a minimum in FWHM (Figure 1c). It is attributed to the moderate reducing effect of carbon on TiO_2_ at 450 °C that enhances structural ordering, whereas, at 550 °C, excessive carbonization leads to lattice distortions and degraded crystal integrity. The Rietveld refinement of the XRD pattern for CeC_0.75_Ti-450, showing a TiO_2_ phase, achieved a high-quality fit, as reflected by the reliability indices of Rwp = 7.42% and GOF = 0.91 (Figure 1d). The refined unit cell parameters, a = b = 3.791832 Å and c = 9.516982 Å, show a close agreement with those reported for the standard anatase TiO_2_ (Table S1). The absence of significant lattice expansion suggests that the actual incorporation of the Ce atom into the structure is limited, which aligns well with the Ce doping concentration. The FT-IR spectra show that the infrared bands of CeTi and TiO_2_ are highly overlapping, indicating single Ce-doping did not significantly alter the framework structure of TiO_2_ (Figure S1). Compared to CeTi and TiO_2_, the band of CTi at 3419 cm^−1^ is noticeably broadened with enhanced absorbance, which originates from the O-H groups enriched on the surface of carbon. The increased intensity of the band at 1630 cm^−1^ suggests contributions from both O-H bending vibrations and C=O functional groups, with the appearance of a new C-H stretching band at 1413 cm^−1^ [35]. The characteristic Ti-O-Ti band at 826 cm^−1^ remains present in all samples, confirming that the main framework structure of TiO_2_ is preserved, which is consistent with the XRD results. As shown in Figure S2a, the CeTi exhibit a typical porous network structure, with pores exceeding 100 nm attributed to the incomplete phase separation caused by rapid gelation. In the case of the CTi (Figure S2b), the addition of R influences the hydrolysis and condensation processes of TiO_2_, leading to reduced resistance to interfacial tension, pore collapse, and overall densification. Although all samples in the CeCTi system maintain a complete three-dimensional porous nanostructure without pore channel collapse, the variation in carbon content significantly modulates their microstructure. Moderate carbon doping in CeC_0.5_Ti leads to a microstructure dominated by mesoporous material, alongside a minor fraction of macropores (Figure 1e), as a result of the insufficient carbon amount to fully fill the interstices of the TiO_2_ framework. In contrast, in CeC_1.0_Ti, the excessively high carbon content disrupts the continuity of the TiO_2_ network, leading to obvious particle agglomeration and a non-uniform pore distribution (Figure 1f). Notably, as shown in Figure 1g, the uniform “pearl-like” porous structure of CeC_0.75_Ti indicates a significant effect between Ce and C elements that optimizes the mesostructured material. Ce modulates the hydrolysis-condensation kinetics of the precursor, slowing the reaction rate and suppressing excessive agglomeration, while carbon enhances the mechanical strength of the network through the formation of Ti-O-C bonds. The aerogel possesses tightly interconnected particles, an absence of agglomeration, and an ideal pore distribution for high-density active sites. Furthermore, the EDS mapping in Figure S4 shows a homogeneous distribution of Ce, Ti, O, and C elements throughout the selected area. However, the signal intensity and distribution density of Ce and C elements are lower than those of Ti and O elements, indicating that those are uniformly distributed within the TiO_2_ aerogel in a low-concentration doping state.

2.2. Multiscale Structure Analysis

The TEM images of the CeC_0.75_Ti-450 reveal a uniformly dispersed nanostructure composed of spherical particles (10–20 nm) that form pores of 50–70 nm, the morphological characteristics consistent with the SEM results (Figure 2a and Figure S5). As shown in Figure 2b, the distinct polycrystalline diffraction rings in the SAED pattern are assigned to the (101), (103), and (004) crystal planes of anatase TiO_2_. The HRTEM reveals distinct lattice fringes with measured spacings of 0.35 and 0.23 nm, corresponding to the (101) and (103) planes of anatase TiO_2_ [36], respectively, which further confirms the formation of the anatase phase (Figure 2c). As shown in Figure 2d, all adsorption isotherms exhibit mixed type II and IV characteristics, indicative of mesoporous materials, with a distinct H3-type hysteresis loop observed in the relative pressure (P/P_0_) range of 0.8–1.0 [37]. The pore size distribution curves in Figure 2e further reveal a bimodal mesoporous structure for all samples, with relatively consistent distributions centered primarily around 7 and 40 nm, aligning with the mesoporous characteristics in Figure 2d. Among all samples, the CTi shows the lowest peak intensity at 7 nm and the highest around 30 nm, indicating that its pore structure is dominated by mesopores of about 30 nm. In contrast, the pore size distributions of the CeCTi overall shift towards larger pore sizes, exhibiting a weak peak near 80 nm; yet, the pores in the 40–50 nm range remain predominant. Notably, the main peak near 30 nm for CeC_0.75_Ti is significantly narrower than that of other CeCTi samples, suggesting a more concentrated pore size distribution and a better uniformity of the pore structure within this size range, which is consistent with the SEM result. The increased BET surface area of CeCTi with carbon content is due to the carbon decomposition during high-temperature treatment, which releases small organic molecules that diffuse out, creating a more developed pore structure and enhanced porosity (Table 1). Notably, the BET surface area of CeC_0.75_Ti reaches 188.81 m^2^·g^−1^, which is significantly higher than those of similar materials reported in the literature [38,39]. The average pore size of CeC_0.75_Ti is approximately 26 nm, which indicates that the carbon decomposition process contributes to the formation of smaller pores, which is beneficial for improving the mass transfer efficiency and the accessibility of active sites, thereby promoting the efficient progression of catalytic reactions.

Figure 2f shows that all samples exhibit two distinct peaks at binding energies of approximately 464.5 and 458.8 eV, corresponding to the spin-orbit doublets of Ti 2p_1_/2 and Ti 2p_3_/2, respectively, which are in agreement with the characteristic signatures of Ti^4+^ [40] (Table S2). The Ti 2p peaks of the CeTi sample shift towards a lower binding energy compared to the others in the XPS spectrum, indicating an increased electron cloud density around Ti. This phenomenon can be attributed to an electronic interaction between Ce^3+^/Ce^4+^ and Ti^4+^, leading to a partial electron transfer from Ce to Ti, thereby reducing the apparent binding energy of Ti. In contrast, the binding energy of the Ti 2p core level in CeCTi samples shows no significant shift relative to CeTi, suggesting that the introduction of carbon does not substantially alter the local electron density around the Ti active centers. This observation implies that the electronic modulation induced by the Ce element incorporation likely dominates the local electronic structure of Ti. In the high-resolution C 1s XPS spectra (Figure 2g), both CeTi and TiO_2_ samples show peaks at approximately 286.3 eV and 288.6 eV, which are assigned to C-O and C=O bonds, respectively [41]. In the CeCTi samples, the intensity of the C 1s peak at 286.3 eV is significantly enhanced and exhibits a slight shift towards a higher binding energy. This indicates an increased content of C-O bonds, likely resulting from the decomposition of the carbon component during treatment. In the O 1s spectra of Figure 2h, a strong characteristic peak at 530.0 eV for all samples is assigned to the Ti-O bond in the TiO_2_ lattice [42], confirming that the Ce/C co-doping does not alter the fundamental crystal structure. Additionally, a shoulder peak typically appears at approximately 531.9 eV, attributed to the surface-OH resulting primarily from surface adsorption after air exposure. As depicted in Figure S6, the characteristic XPS peaks for Ce are notably weak, a phenomenon attributed principally to its low doping level relative to the major constituents of the system. The EPR spectra show an anisotropic signal at g = 2.003 for both CeTi and O_V_-CeCTi (Figure 2i), which is characteristic of O_V_ [42,43]. The signal intensity is higher in CeCTi, indicating that the co-doping of Ce and C promotes the formation of more O_V_ compared to Ce-doping alone.

2.3. Photoelectrochemical Properties and Reduction Performance

The arc radius in EIS reflects the electrical conductivity of the photocatalyst, with a smaller radius typically indicating superior conductivity, lower resistance, and a higher electron transfer rate [44]. As shown in Figure 3a, under identical testing conditions, a successive decrease in the EIS arc radius is observed from TiO_2_ to CeTi, CTi, and CeCTi, demonstrating a progressively lower charge transfer resistance. This trend clearly demonstrates that C-doping is more effective than Ce-doping in enhancing electrical conductivity, while their combination achieves the optimal synergistic effect, substantially improving interfacial charge migration. The photocurrent response measurements in Figure 3b show that the CeCTi sample exhibits the highest photocurrent density, significantly surpassing that of the CeTi sample. The latter, in turn, shows a markedly stronger response than both the CTi and pristine TiO_2_ aerogels, which display only marginal differences between them. This trend demonstrates that Ce-doping plays a dominant role in suppressing the charge recombination and extending the carrier lifetime, while the synergistic effect of Ce/C co-doping further enhances the charge migration and separation efficiency, leading to the optimal utilization of photogenerated carriers [45]. Combined with the smaller EIS arc radius of the CeCTi, these findings indicate that Ce/C co-doping boosts the charge carrier concentration while lowering the transport resistance, collectively enhancing photocatalytic activity. Figure 3c shows that all samples exhibit a distinct PL peak at around 460 nm, which originates from the recombination of photogenerated electron–hole pairs that release energy radiatively [46]. The pristine TiO_2_ aerogel shows the highest PL intensity, indicating the most severe charge recombination. Notably, the PL intensity of the CeTi sample is higher than that of CTi, suggesting that C-doping contributes more effectively to the initial charge separation. The CeCTi sample exhibits the lowest PL intensity, demonstrating the strongest suppression of recombination, which can be attributed to the effect of Ce/C co-doping. To further investigate this enhanced charge transfer dynamic, time-resolved PL decay measurements were performed on TiO_2_ and CeCTi. As shown in Figure 3d, the fluorescence lifetime of TiO_2_ is 0.63 ns, while that of CeCTi is prolonged to 0.83 ns. This extension directly reflects the slower carrier recombination in CeCTi, providing clear evidence of the more efficient charge separation. To further investigate the effect of Ce/C co-doping on the charge transfer efficiency, the PL decay curves of TiO_2_ and CeCTi were measured and fitted using an exponential decay function. As shown in Figure 3d, the fluorescence lifetime of TiO_2_ is 0.63 ns, while that of CeCTi is prolonged to 0.83 ns. The observed extension in the lifetime is a clear signature of the suppressed recombination of photogenerated charge carriers, which, in turn, leads to a significantly improved charge separation efficiency [47]. The UV–Vis spectra in Figure 3e show that CTi and CeCTi clearly exhibit characteristic absorption bands in the range of 300–400 nm, which are attributed to the intrinsic absorption of TiO_2_. With an increasing C-doping content, the absorption of the CeCTi composites demonstrates an initial red shift followed by a slight blue shift, a trend driven by the C-doping optimization of the defect distribution in CeTi. Compared to CTi, the CeCTi samples demonstrate significantly enhanced absorbance in the visible light region (400–700 nm), with CeC_0.75_Ti showing the highest absorption, indicating its superior visible-light-harvesting capability. The CTi aerogel retains the optical band gap of 3.20 eV in the Tauc plots, consistent with the pristine TiO_2_ [48], suggesting that singular C-doping only mildly tunes the electronic structure without significantly altering the optical band gap. In contrast, the optical band gap of the CeCTi series narrows from 3.06 eV to 2.90 eV with an increasing C-doping content, then slightly widens to 3.00 eV at higher C levels. This trend demonstrates that an optimal C-doping concentration, in synergy with Ce, most effectively stabilizes the defect states and extends the visible-light absorption. To further investigate the detailed electronic properties of the CeCTi samples, Mott–Schottky measurements were carried out on its representative CeC_0.75_Ti sample, as illustrated in Figure 3f. The Mott–Schottky test performed on the CeC_0.75_Ti sample to further investigate the electronic band structure of CeCTi shows that the conduction band (CB) position is −0.81 V vs. SCE. Using conversion Equation (1), the CB potential is calculated to be approximately −0.57 V vs. NHE, and the 2.90 eV band gap yields a valence band (VB) potential of +2.33 V vs. NHE via Equation (2). This band alignment confirms that CeC_0.75_Ti possesses sufficient thermodynamic driving force to participate in various photocatalytic redox reactions.

$$[eqn]$$ $$[eqn]$$

C-doping significantly enhances the conductivity of the material and helps stabilize the defect structure, while Ce-doping effectively prolongs the carrier lifetime, and promotes the interface charge transfer. Thus, Ce/C co-doping achieves a comprehensive improvement in the photocatalytic CO_2_ reduction performance through coordinated band engineering, defect engineering and structural optimization. Based on the Mott–Schottky and UV–Vis analyses, a plausible mechanism for the photocatalytic CO_2_ reduction over the CeCTi is proposed. As illustrated in Figure 3g, the CB potential of CeCTi is −0.57 V vs. NHE, which is more negative than the standard redox potentials for CO_2_/CO (−0.53 eV) and CO_2_/CH_4_ (−0.24 eV). Meanwhile, its VB potential (+2.33 V vs. NHE) remains sufficiently positive for water oxidation. This band alignment provides the necessary thermodynamic driving force for both CO_2_ reduction and H_2_O oxidation. However, the photocatalytic performance is governed not only by the thermodynamic feasibility but also by the charge separation efficiency and surface reaction kinetics. Ce/C co-doping effectively introduces abundant O_V_ and optimizes the electronic structure of the material. First, they act as electron traps, prolonging the lifetime of photogenerated charge carriers and suppressing the electron–hole recombination. More importantly, they drive a dynamic Ti^4+^/Ti^3+^ redox cycle at the catalyst surface. Specifically, the electron-donating property of Ce^3+^ reduces adjacent Ti^4+^ ions to Ti^3+^. CO_2_ molecules adsorbed on nearby O_V_ acquire electrons from these Ti^3+^ sites, leading to their activation and initial reduction. The resulting Ti^4+^ is then regenerated back to Ti^3+^ by photogenerated electrons captured by O_V_, thereby sustaining an efficient electron shuttle that continuously supplies reducing equivalents to the adsorbed CO_2_. Furthermore, the three-dimensional porous architecture of the CeCTi aerogel offers a high specific surface area, which exposes abundant active sites and facilitates reactant adsorption and electron transport. Figure 3h,i present the photocatalytic CO_2_ reduction performance of the synthesized photocatalysts under full-spectrum light irradiation, demonstrating their promising potential [49]. It should be emphasized that the detected CO and CH_4_ are exclusively derived from CO_2_, as no external carbon source was introduced in the reaction system. As seen in Figure 3h, after 2 h of illumination, the CH_4_ and CO production rates over the TiO_2_ aerogel were 0.33 and 17.01 μmol g^−1^ h^−1^, respectively. The CeTi shows enhanced rates for both products, while the CTi gave a lower CH_4_ yield, indicating that carbon incorporation alone does not facilitate effective light utilization. Overall, the CeC_0.75_Ti-450 demonstrated the better performance, achieving CH_4_ and CO production rates of 27.06 and 97.11 μmol g^−1^ h^−1^, respectively, which are approximately 82.0 and 5.7 times higher than those of the TiO_2_ aerogel. Table S3 presents a comparison of various CO_2_ photocatalytic materials (including the materials developed in this study) in terms of reaction conditions and yields. Unlike many systems that require precious metals or complex heterogeneous structures, our catalyst achieves a comparable or even superior performance through a simple Ce/C co-doping process. It demonstrates that the electronic structure optimized through Ce/C co-doping is favorable for product conversion. Furthermore, as shown in Figure 3i, the total yields of CH_4_ and CO over CeC_0.75_Ti-450 reach 69.81 and 376.62 μmol g^−1^, respectively, after 5 h of simulated sunlight irradiation. Figure S7a shows that, under visible light irradiation, the CeTi and CTi samples both exhibited dramatically higher CH_4_ and CO production rates than the TiO_2_ aerogel, which registered only 0.41 and 0.13 μmol g^−1^ h^−1^, respectively. The CeC_0.75_Ti-450 exhibited the best performance, with CH_4_ and CO production rates reaching 5.21 and 11.23 μmol g^−1^ h^−1^, respectively, representing increases by factors of approximately 12.7 and 86.4 compared to the TiO_2_ aerogel. This significant improvement is attributed to the effect of Ce/C co-doping, which optimizes O_V_ distribution, enhances visible-light absorption through defect-state introduction, and improves the charge carrier separation efficiency. Additionally, when the CeC_0.75_Ti-450 sample was irradiated under visible light for 5 h, the yields of CH_4_ and CO continued to increase, reaching 20.58 and 37.25 μmol g^−1^ in Figure S7b. As shown in Figure S7c, the stability of the CeC_0.75_Ti-450 catalyst is confirmed through three consecutive recycling tests (2 h per cycle), with no significant decrease in CH_4_ and CO production yields observed. The results indicate that, during this test period, the yields of CH_4_ and CO do not show a significant decline, suggesting that the catalyst exhibits good catalytic activity retention over this timespan. These current findings can serve as preliminary evidence of the potential stability of the catalyst and provide a basis for a further systematic investigation.

2.4. DFT Calculation

Figure 4a shows that, for TiO_2_, the spin-up and spin-down electrons are symmetrically distributed around the Fermi level in the total density of states (TDOS), indicating no net spin and confirming its non-magnetic nature. The density of states (DOS) of the TiO_2_ distribution reveals that the CB is primarily contributed by O 2p orbitals, while the VB originates mainly from Ti 3d orbitals. Compared to the pristine TiO_2_, O_V_-CeCTi introduces an impurity energy level near the Fermi level (Figure 4b). This level can act as a charge carrier trap, which is expected to facilitate more efficient charge separation and potentially enhance visible-light absorption [50]. To elucidate the individual contributions of Ce and C elements, it analyzes the enlarged view of the partial density of states (PDOS) for the Ce and C elements in CeCTi, as presented in Figure 4c. The hybrid peaks near the Fermi level mainly arise from the combined contributions of Ce 3d and C 2p orbitals, with the C 2p orbitals exhibiting a more pronounced density of states. This demonstrates the dominant role of carbon in regulating the computational band, while the Ce element enhances this effect through synergistic doping. Figure 4d shows that the DFT-calculated band gap of the pristine TiO_2_ is 1.404 eV, which is significantly lower than the experimental optical band gap of 3.2 eV. This systematic underestimation is a well-known limitation of standard DFT due to its inherent self-interaction error in the exchange–correlation treatment [51]. Therefore, the DFT results presented here are used primarily for analyzing the relative trends in electronic structure modulation, rather than for predicting absolute band gap values. In comparison to the pristine TiO_2_, CeTi leads to a slight increase in the calculated band gap to 1.445 eV, consistent with surface state passivation (Figure S8a). Notably, for the CeCTi, the DFT-derived electronic transition threshold drops drastically to 0.102 eV. This extremely low value corresponds not to the intrinsic optical band gap, but to defect-induced in-gap electronic excitations. The latter arise from strong C 2p–lattice hybridization, which generates near-Fermi-level impurity states and elevates the VB maximum (Figure S8). Thus, the 0.102 eV value characterizes the energy required for excitations involving these defect levels, not the fundamental band-to-band transition. Furthermore, introducing O_V_ into CeCTi causes the transition energy to rebound slightly to 0.463 eV, as localized electrons from the vacancies partially fill the C-doping impurity levels and shift their energy positions. This further confirms that the low-energy electronic transitions in these doped systems are governed by defect-level physics, rather than by the pristine band-to-band excitation. It is important to emphasize that the DFT-calculated transition energies for CeCTi and O_V_-CeCTi are much lower than the experimental optical band gap of 2.90 eV obtained from the UV-Vis/Tauc analysis. This discrepancy underscores that the DFT values represent defect-related electronic excitations, while the UV–Vis measurement probes the intrinsic optical band gap of the material. To obtain more quantitatively accurate band gap predictions, advanced computational methods such as DFT + U would be required in future studies. To further investigate the charge transfer behavior in the O_V_-CeCTi material, a differential charge density analysis was conducted. According to the results of the sectional plane shown in Figure S9b, distinct blue regions are observed around the Ce atoms, indicating that they lose electrons during the bonding process and act as electron donors (Figure 4g). Meanwhile, red regions directed towards Ti atoms are present between the Ce and Ti atoms, suggesting that electrons migrate from Ce to Ti, with Ti serving as an electron acceptor and undergoing a significant charge redistribution. This phenomenon indicates that the introduced O_V_ effectively modulates the charge distribution between Ce and Ti atoms, enhancing their interfacial electron interactions. The redistribution subsequently facilitates the separation and migration of photogenerated charge carriers, thereby improving the photocatalytic and electrocatalytic performance of the material. An electron localization analysis reveals that, in the pristine TiO_2_, VB electrons are localized on surface O atoms while CB electrons reside on Ti atoms (Figure S9). In contrast, for O_V_-CeCTi, the VB electron density concentrates around Ce/C dopants, indicating dominant contributions from C 2p and Ce 3d orbitals to the highest occupied state, whereas CB electrons accumulate around O_V_ (Figure 4h,i). Upon photoexcitation, electrons migrate from the VB to the CB and are effectively captured by O_V_ sites, prolonging the carrier lifetime. These trapped electrons are then transferred to the adsorbed CO_2_, driving its reduction. This spatial separation between VB and CB suppresses the electron–hole recombination and enhances visible-light absorption, providing a clear electronic-structure basis for understanding the improved photocatalytic activity of CeCTi [52].

3. Conclusions

This study successfully synthesized CeCTi via a sol–gel method combined with supercritical CO_2_ drying and subsequent heat treatment. The results show that the O_V_-CeCTi-450 exhibits a pearl-like porous network structure, an optical bandgap of 2.90 eV, and a high specific surface area of 188.81 m^2^/g. The Ce/C co-doping enhances light absorption, with the Ce element prolonging the carrier lifetime and promoting the interface charge transfer, while carbon further contributes to the generation of additional O_V_. Under simulated sunlight irradiation, the O_V_-CeCTi aerogel achieves CH_4_ and CO production rates of 27.06 and 97.11 μmol g^−1^ h^−1^ without any co-catalysts or sacrificial agents, representing 82.0- and 5.7-fold improvements over the pristine TiO_2_ aerogel. DFT calculations confirm that, under Ce/C co-doping, carbon promotes O_V_ formation and introduces defect states within the TiO_2_ band gap. Upon light irradiation, electrons are photoexcited from the VB to the CB, captured by O_V_ traps to prolong their lifetime, which prolongs the carrier lifetime as evidenced by the extended fluorescence lifetime from 0.63 ns (TiO_2_) to 0.83 ns (CeCTi). This enhanced charge separation effectively facilitates electron delivery to the adsorbed CO_2_, driving its photocatalytic reduction and offering insights for the rational design of high-performance photocatalysts.

4. Materials and Methods

4.1. Materials

Tetrabutyl titanate (TBOT, ≥99%) was purchased from Shanghai Alfa Aesar (China) Chemical Co., Ltd. (Shanghai, China). Nitric acid (HNO_3_, 69%) and resorcinol (R, 99%) were bought from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Cerium (III) nitrate hexahydrate (Ce(NO_3_)3·6H_2_O, 99%) was purchased from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China). DI water and anhydrous ethanol (ETOH) were purchased from Nanjing Wanqing Chemical Glass Ware & Instrument Co., Ltd. (Nanjing, China). Ar and CO_2_ were bought from Nanjing Changyuan Industrial Gases Co., Ltd. (Nanjing, China). All chemicals were used directly without further purification.

4.2. Method

4.2.1. Synthesis of the CeCTi Aerogels

First, TBOT (4.40 mL, 12.9 mmol) and HNO_3_ (0.24 mL, 3.79 mmol) were mixed in EtOH (10.42 mL, 178 mmol) under constant magnetic stirring at 500 rpm for 20 min to achieve complete homogenization. Next, Ce(NO_3_)3·6H_2_O (0.19 g, 0.437 mmol) was gradually introduced into a separate EtOH (5.21 mL, 89.2 mmol) and stirred at ambient temperature for an additional 20 min to guarantee thorough dissolution. At the same time, a mixture of R, EtOH, and DI water was continuously stirred for 20 min in another beaker. After mixing all prepared solutions and stirring for 15 min, the homogeneous mixture was transferred into a plastic mold and allowed to undergo wet gel by standing undisturbed at room temperature for 24 h. The gels were subsequently aged by soaking in an ethanol solution for 3 days, with the ethanol being replaced three times daily to ensure proper solvent exchange. Following aging, the wet gels were subjected to supercritical CO_2_ drying to obtain CeRTi aerogels. Finally, the aerogels were heat-treated in a tube furnace under an Ar atmosphere at temperatures of 350 °C, 450 °C, and 550 °C for 2 h, corresponding to the CeCTi samples designated as O_V_-CeCTi-350, O_V_-CeCTi-450, and O_V_-CeCTi-550, respectively. It was important to note that the molar ratios of R to TBOT were maintained at 0.5:1, 0.75:1, and 1:1, corresponding to the CeC_x_Ti samples designated as CeC_0.5_Ti, CeC_0.75_Ti, and CeC_1.0_Ti, respectively. The sample without Ce/C co-doping was named TiO_2_, the sample without Ce-doping was named CTi, and the sample without C-doping was named CeTi.

4.2.2. Characterizations

Scanning electron microscopy (SEM, Zeiss Gemini Sigma 300 VP SEM, Carl Zeiss AG, Oberkochen, Germany) and transmission electron microscopy (TEM, JEM 2100F, JEOL, Tokyo, Japan) were utilized to analyze the microstructure and morphology of the samples. Fourier-transform infrared spectroscopy (FT-IR, Thermo Scientific Nicolet iN10 MX, Thermo Fisher Scientific, Waltham, MA, USA) was employed to investigate the functional groups of the samples. The elemental composition of the samples was analyzed using X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, Thermo Fisher Scientific, Waltham, MA, USA) equipped with an Al Kα excitation source (1487.6 eV). The crystalline phase evolution of the samples was examined using X-ray diffraction (XRD, Mini Flex-600 X, Rigaku, Tokyo Akishima, Japan) with a Cu Kα source (λ = 0.154 nm) at a scanning rate of 10°/min. Information regarding the pores was obtained through a brunauer–emmett–teller (BET) analysis using a V-Sorb 2800P surface area and pore size analyzer (Jin Aipu Technology, Beijing, China). O_V_ was characterized using electron paramagnetic resonance (EPR, EMXplus-6/1, Bruker, Walzbachtal, Germany) with the following parameters: sweep time of 30.0 s, power of 2.0 mW, scan width of 100.0 G, modulation amplitude of 3.0 G, and modulation frequency of 100.0 kHz. The light response range and band structure of the samples were characterized using ultraviolet–visible absorption spectroscopy (UV–Vis, UV-2600, Shimadzu Corporation, Kyoto, Japan). Steady-state/transient fluorescence spectroscopy (FLS1000, Edinburgh, Livingston, UK) was employed to measure the photoluminescence (PL) and time-resolved photoluminescence (TRPL) spectra, which were used to study the charge transfer kinetics and fluorescence lifetime of the samples. Electrochemical impedance spectroscopy (EIS), Mott–Schottky analysis, and photocurrent tests were conducted using a CS2350 electrochemical workstation (Wuhan Kost Instrument, Wuhan, China) to evaluate the impedance frequency and potential. A three-electrode quartz cell was used, with 0.5 M Na_2_SO_4_ solution as the electrolyte. Platinum (Pt) electrode, saturated calomel electrode (SCE), and fluorine-doped tin oxide (FTO) glass were employed as the counter electrode, reference electrode, and working electrode. The photocurrent (i-t) was measured via constant-potential polarization testing using a 1 cm^2^ electrode under illumination from a 300 W xenon lamp positioned 15 cm away.

4.2.3. Photoactivity Test

Then, 5 mL H_2_O was added to the bottom of a 500 mL quartz reactor, and a glass fiber membrane loaded with 3 mg of catalyst is placed above the water droplet, avoiding direct liquid–solid contact. The reactor was purged by three cycles of CO_2_ evacuation and refilling to eliminate residual air, then maintained at 80 kPa to establish a reaction atmosphere comprising gaseous CO_2_ and H_2_O. A 300 W xenon lamp (Microsolar 300, Perfect Light, Beijing, China, 320–780 nm, 100 mW·cm^−2^) was used to simulate solar irradiation, positioned vertically 10 cm away from the catalyst during the photocatalytic reaction. Product analysis was performed using a GC-126N gas chromatograph (INESA, Shanghai, China), equipped with a flame ionization detector and a thermal conductivity detector for quantitative detection. Gas samples were collected hourly, and product concentrations were determined by converting chromatographic peak areas via external standard calibration. The final product yield was calculated based on the reaction time. Additionally, several control experiments with varying parameters were conducted to eliminate the influence of external environmental factors on the catalytic results.

4.2.4. Theoretical Calculations

DFT calculations were performed using the DMol3 module to obtain geometric structure optimization and electronic structure calculations. The generalized gradient approximation (GGA) with the Perdew–Burke–Ernzerhof (PBE) functional was employed to describe electron exchange and correlation effects. To match the experimental conditions, the (101) crystal plane of the TiO_2_ phase was selected to construct the CeCTi model. A 2 × 3 × 1 supercell is built, where a Ti atom was replaced with a Ce atom to form the CeTi model, and adjacent O atom was removed to create O_V_, yielding a CeTi model with O_V_. The C atom was introduced into the CeTi model with O_V_ to construct the O_V_-CeCTi catalyst model. During the geometric structure optimization, the bottom three layers of atoms were fixed, while the remaining atoms were allowed to relax freely to their ground states. A vacuum layer of 15 Å was applied along the Z-direction to eliminate interactions between periodic images. The Brillouin zone was sampled using a 2 × 3 × 1 Monkhorst–Pack k-point grid, with energy, maximum force, and maximum displacement thresholds set at 1.0 × 10^−5^ Hartree, 2.0 × 10^−3^ Hartree/Å, and 5.0 × 10^−3^ Å, respectively. Additionally, a 4 × 4 × 1 k-point grid was used for electronic structure calculations in this work.