Facile Preparation of Highly Active CO 2 Reduction (001)TiO 2 /Ti 3 C 2 T x Photocatalyst from Ti 3 AlC 2 with Less Fluorine

: To date, (001)TiO 2 /Ti 3 C 2 T x hybridized photocatalyst is usually prepared through the complicated treatment of Ti 3 AlC 2 in the presence of corrosive ﬂuorine with a molar ratio of nF:nTi of more than 20. To reduce the use of corrosive ﬂuorine, herein, exploiting beyond the conventional method, we report a facile synthetic method for (001)TiO 2 /Ti 3 C 2 T x , elaborately using HF as both an etchant for Al elimination and a morphology control agent for the growth of (001)TiO 2 nanosheets, with a sharply diminished use of ﬂuorine (nF:nTi = 4:1) and simpliﬁed operation procedures. After optimization, the resulting (001)TiO 2 /Ti 3 C 2 T x heterojunction exhibited markedly high photocatalytic activity with the CO 2 reduction rate of 13.45 µ mol g − 1 h − 1 , which even surpasses that of P25 (10.95 µ mol g − 1 h − 1 ), while the photoelectron selectivity to CH 4 is approaching 92.84%. The superior photoactivity is interpreted as the fact that Ti 3 C 2 T x with a lower work function induces photoinduced hole transfer and suppresses the charge recombination, thus facilitating the CO 2 multi-electron reduction. This study provides a novel and simple synthesis for (001)TiO 2 /Ti 3 C 2 T x towards sustainable energy transformations.


Introduction
TiO 2 is an extremely attractive candidate for photocatalysis, given the multiple advantages of low price, chemical durability, and ecological friendliness [1,2]. However, realistic applications of TiO 2 are restricted due to the rapid recombination of photogenerated electron-hole pairs. Therefore, numerous methods have been developed with the focus on the promotion of photoinduced charge transfer, such as loading co-catalyst [3,4], morphology modulation [5], and the construction of heterojunction [6][7][8]. Thereinto, loading co-catalyst on TiO 2 has hitherto been a satisfactory choice for improving its photoactivity and stability by accelerating the separation efficiency of photogenerated carriers. Nevertheless, the widespread utilization of noble-metal cocatalysts, such as Au and Pt, remains challenging for both high cost and scarcity. Hence, there is a considerable incentive to exploit cost-effective noble-metal-free cocatalysts for the enhancement of the TiO 2 photocatalytic performance.
MXenes, a new family of two-dimensional layered transition metal carbides, nitrides, or carbonitrides, possess a structural formula of M n+1 X n T x , where M refers to the early transition metal, X stands for carbon and/or nitrogen, T represents the surface functional groups (-O, -OH, -F), and n ranges from 1 to 3. Along with a mushrooming in the study of MXenes, up to now, more than 40 MXenes have been synthesized by selectively etching the metal layer from the precursor MAX phases [9]. Moreover, various syntheses for

Results and Discussion
The phase structure of the sample was analyzed using X-ray diffraction (XRD) spectroscopy. As shown in Figure 2a, the diffraction peak of Ti3AlC2 MAX is consistent with that reported in the literature [12]. In comparison to Ti3AlC2, the HF etching process causes the (104) diffraction peak at 39° to disappear in Ti3C2Tx, as well as the (002) and (004) diffraction peaks to shift towards lower angles, which are indications of Al layer elimination and increased layer spacing in the Ti3C2Tx structure [28], as likewise evidenced by the observation of SEM images for both Ti3AlC2 and Ti3C2Tx ( Figure S2). For conciseness, the TiO2/Ti3C2Tx samples prepared by hydrothermal method are hereafter named TT-x, where x refers to the added volume (unit: mL) of the HF. All of the TiO2/Ti3C2Tx samples are well indexed to the peaks of both Ti3C2Tx and anatase phase TiO2 (JCPDS No. 21-1272), indicating that TiO2/Ti3C2Tx heterojunctions were successfully prepared ( Figure 2a). However, there remains a faint (104) diffraction peak of Al at 39° for the TT-2 sample, which is mainly since the amount of HF added during the preparation was too small to etch the Al layer of Ti3AlC2 completely. When the amount of HF added was greater than 3 mL, the peak of Al fully disappeared in the corresponding position of the sample. Notably, the intensity of the diffraction peaks of TiO2 decreased with the increasing HF addition higher than 5mL. This is mainly because the excessive amount of HF will further etch the TiO2 nanosheets and make them less crystalline. Among them, the crystallinity of TiO2 in the TiO2/Ti3C2Tx composites prepared by the HF addition in the range of 3 mL to 5 mL was better.

Results and Discussion
The phase structure of the sample was analyzed using X-ray diffraction (XRD) spectroscopy. As shown in Figure 2a, the diffraction peak of Ti 3 AlC 2 MAX is consistent with that reported in the literature [12]. In comparison to Ti 3 AlC 2 , the HF etching process causes the (104) diffraction peak at 39 • to disappear in Ti 3 C 2 T x , as well as the (002) and (004) diffraction peaks to shift towards lower angles, which are indications of Al layer elimination and increased layer spacing in the Ti 3 C 2 T x structure [28], as likewise evidenced by the observation of SEM images for both Ti 3 AlC 2 and Ti 3 C 2 T x ( Figure S2). For conciseness, the TiO 2 /Ti 3 C 2 T x samples prepared by hydrothermal method are hereafter named TT-x, where x refers to the added volume (unit: mL) of the HF. All of the TiO 2 /Ti 3 C 2 T x samples are well indexed to the peaks of both Ti 3 C 2 T x and anatase phase TiO 2 (JCPDS No. 21-1272), indicating that TiO 2 /Ti 3 C 2 T x heterojunctions were successfully prepared ( Figure 2a). However, there remains a faint (104) diffraction peak of Al at 39 • for the TT-2 sample, which is mainly since the amount of HF added during the preparation was too small to etch the Al layer of Ti 3 AlC 2 completely. When the amount of HF added was greater than 3 mL, the peak of Al fully disappeared in the corresponding position of the sample. Notably, the intensity of the diffraction peaks of TiO 2 decreased with the increasing HF addition higher than 5 mL. This is mainly because the excessive amount of HF will further etch the TiO 2 nanosheets and make them less crystalline. Among them, the crystallinity of TiO 2 in the TiO 2 /Ti 3 C 2 T x composites prepared by the HF addition in the range of 3 mL to 5 mL was better.
stretching vibrations of graphitic carbon [12,22]. For the other two samples, the Ti3C2Tx sample showed the expected Raman vibrational modes [15,29]. Compared to the Ti3C2Tx sample, the TT-5 sample displayed an ascending Raman signal at 150 cm −1 (Eg), together with three other peaks at 393 cm −1 (B1g), 507 cm −1 (A1g), 630 cm −1 (Eg), which are originated from anatase (001)TiO2 [12,30]. Besides, the D-and G-modes peaks of Ti3C2Tx were also observed clearly in TT-5. Reasonably, both XRD and Raman results confirm the transformation of Ti3C2Tx to TiO2 during hydrothermal treatment. Another direct piece of evidence for the successful preparation of TiO2/Ti3C2Tx is provided by the morphology and microstructure observation via field emission scanning electron microscopy (FESEM) and high-angle annular dark field scanning transmission electron microscope (HAADF-STEM). The most striking feature of the typical heterojunction is the existence of TiO2 nanosheets embedded in Ti3C2Tx that are deliberately created using a fluorine source (HF) under the hydrothermal process. This characteristic is directly observed with FESEM (Figure 2c), showing square TiO2 nanosheets with a thickness In addition to XRD, Raman spectra were used to reveal the constituents of the asprepared samples, yielding similar results. As shown in Figure 2b, Ti 3 AlC 2 contains several characteristic bands that appeared at ca. 169, 268, 411, and 598 cm −1 , which are associated with the ω1, ω2, ω3, and ω4 Raman-active phonon vibration modes, respectively, while the two peaks between 1000 and 1800 cm −1 (D and G band) are referred to the stretching vibrations of graphitic carbon [12,22]. For the other two samples, the Ti 3 C 2 T x sample showed the expected Raman vibrational modes [15,29]. Compared to the Ti 3 C 2 T x sample, the TT-5 sample displayed an ascending Raman signal at 150 cm −1 (E g ), together with three other peaks at 393 cm −1 (B 1g ), 507 cm −1 (A 1g ), 630 cm −1 (E g ), which are originated from anatase (001)TiO 2 [12,30]. Besides, the D-and G-modes peaks of Ti 3 C 2 T x were also observed clearly in TT-5. Reasonably, both XRD and Raman results confirm the transformation of Ti 3 C 2 T x to TiO 2 during hydrothermal treatment.
Another direct piece of evidence for the successful preparation of TiO 2 /Ti 3 C 2 T x is provided by the morphology and microstructure observation via field emission scanning electron microscopy (FESEM) and high-angle annular dark field scanning transmission electron microscope (HAADF-STEM). The most striking feature of the typical heterojunction is the existence of TiO 2 nanosheets embedded in Ti 3 C 2 T x that are deliberately created using a fluorine source (HF) under the hydrothermal process. This characteristic is directly observed with FESEM (Figure 2c), showing square TiO 2 nanosheets with a thickness of 130 nm embedded in the layered Ti 3 C 2 T x . The TEM image (Figure 2d) of TiO 2/ Ti 3 C 2 T x displays several square thin nanosheets on a substrate, and its partially enlarged image (Figure 2e) shows the obvious lattice spacing of 0.235 nm, which is identified for the TiO 2 (001) plane. Usually, the growth rate of crystals is exponentially related to the surface energy [31]. During the growth of TiO 2 nuclei, the (001) planes with high surface energy (0.90 J/m 2 ) decrease rapidly, and the TiO 2 nanocrystals evolve spontaneously into a truncated bipyramidal structure dominated by stable (101) facets because of their low surface energy (0.44 J/m 2 ) [32]. It is worth noting that modifying the surface adsorption species of the crystal can change the surface energy, exposing specific highly active facets of the crystal. F is a species with the affinity for Ti atoms so that it can bind Ti atoms and reduce the surface energy of the (001) facets, making them more stable than the (101) facets during the anatase TiO 2 crystal growth process [27,33]. Therefore, in our case, the absorbed F atom functioned as a capping agent to modify the surface energy of (001) facets of TiO 2 , resulting in exceptional stabilization. Furthermore, the elemental distribution in the composite was visualized by STEM-energy-dispersive X-ray spectroscopy (EDS) mapping analysis (Figure 2f). By focusing on an area containing square sheets with lying and standing positions, it is clearly revealed that the distributions of Ti atoms and O atoms are perfectly coincided with the square region, verifying the formation of TiO 2 nanosheet structure. Therefore, it can be concluded that the (001)TiO 2 /Ti 3 C 2 T x has been successfully prepared.
As evidenced above, after HF pretreatment and hydrothermal process for precursor Ti 3 AlC 2 , the Al layer has been removed and the TiO 2 /Ti 3 C 2 T x heterojunction was formed.
To further confirm the changes in the surface composition of the prepared samples Ti 3 AlC 2 , TT-5, and TT-5-u (the sample pretreated with 5 mL HF but without hydrothermal treatment) at different stages in synthesis, XPS measurements were carried out. As a consequence of the Al layer removal, Al 2s and 2p peaks in Ti 3 AlC 2 disappeared in TT-5 and TT-5-u, which can be observed via the XPS survey spectra ( Figure S3). The comparison of the high-resolution Ti 2p spectrum between TT-5 and TT-5-u is shown in Figure 3a [12,16,22]. In comparison to TT-5-u, the TT-5 sample shows two more peaks of TiO 2 (459.4 eV and 465.0 eV) and the fade of the other Ti species, demonstrating the gradual transformation from Ti 3 C 2 T x to TiO 2 . Furthermore, the C 1s spectrum (Figure 3b) of TT-5 exhibited several distinct peaks at 288.8, 286.4, 284.8, 282.5, and 281.9 eV, corresponding to C-F, C-O, C-C, C-Ti-O, and Ti-C, respectively [16,22,34]. Apparently, C-F is derived from HF; Ti-C and C-C originate from Ti 3 C 2 T x and adventitious carbon, respectively. It is noteworthy that the TT-5 showed a new C-Ti-O peak at 282.5 eV as compared with TT-5-u (Figure 3e), demonstrating that the TiO 2 /Ti 3 C 2 T x heterojunction is linked by a C-Ti-O bond. The O 1s spectrum of TT-5-u is fitted by two components located at 532.18 and 530.0 eV (Figure 3f), which are corresponding to Ti-OH, Ti-O [35]. After hydrothermal treatment, the Ti-O intensity of TT-5 significantly increased as compared to Ti-OH (Figure 3c), which suggests the existence of TiO 2 . The XPS results verify the successful preparation of the TiO 2 /Ti 3 C 2 T x heterojunction, which is well consistent with the results of XRD, Raman, and EDS analyses. Catalysts 2022, 12, x FOR PEER REVIEW 6 of 13 The photocatalytic activity of the samples was evaluated by photocatalytic CO2 reduction in a gas-solid reaction system under simulated solar light. Mild conditions were kept without using any photosensitizer or organic sacrificial reagents. Methane (CH4) and carbon monoxide (CO) are generated as the predominant products via the two-electron and eight-electron reduction processes, respectively ( Figure 4a). Ti3C2Tx with metallic properties is generally not photocatalytically active [28]. However, a weak CO2 photoreduction activity of Ti3C2Tx was detected here, resulting from the TiO2 (partial oxidation Ti3C2Tx) during storage or light exposure. As for the other samples, the photoactivity showed a trend to increase and then decline as the increasing addition of HF. From TT-2 to TT-5, the photoactivity gradually enhanced due to the increase in (001)TiO2 nanosheets on Ti3C2Tx, which is consistent with the trend seen in the XRD and SEM results ( Figures  2a and S4). Besides, for nitrogen adsorption-desorption isotherms and corresponding pore-size distribution analysis ( Figure S5 and Table S2), TT-5 shows larger surface area and more micropores (<2 nm) as compared to TT-2, which may favor the adsorption of CO2 gas molecules. However, following adding excessive HF, the TiO2 may further be etched (as shown in Figure S4), resulting in a decrease in the photoactivity for TT-6 and TT-8. The results suggest that there is an optimum amount (5 mL) of HF in the synthesis of TiO2/Ti3C2Tx for achieving high photocatalytic CO2 reduction rates. Among all the samples, TT-5 exhibited the best CO2 photoreduction rate of 13.45 μmol g −1 h −1 (10.28 μmol g −1 h −1 for CH4 and 3.17 μmol g −1 h −1 for CO), which was higher than that of P25 TiO2 (10.95 μmol g −1 h −1 ). Note that methane was not detected for commercial P25, while TT-5 was up to 92.84% methanogenic selectivity ( Figure 4b). As reported previously, the existence of Ti3C2 favors the selectivity for photocatalytic CO2-CH4 conversion [36,37]. Combining the reported studies and above experimental results, the reason for the extraordinary selectivity for CH4 production could be interpreted as Ti3C2 with lower work function induces the photogenerated hole transfer and suppresses the charge recombination, facilitating the CO2 multi-electron reduction. The photocatalytic activity of the samples was evaluated by photocatalytic CO 2 reduction in a gas-solid reaction system under simulated solar light. Mild conditions were kept without using any photosensitizer or organic sacrificial reagents. Methane (CH 4 ) and carbon monoxide (CO) are generated as the predominant products via the two-electron and eight-electron reduction processes, respectively (Figure 4a). Ti 3 C 2 T x with metallic properties is generally not photocatalytically active [28]. However, a weak CO 2 photoreduction activity of Ti 3 C 2 T x was detected here, resulting from the TiO 2 (partial oxidation Ti 3 C 2 T x ) during storage or light exposure. As for the other samples, the photoactivity showed a trend to increase and then decline as the increasing addition of HF. From TT-2 to TT-5, the photoactivity gradually enhanced due to the increase in (001)TiO 2 nanosheets on Ti 3 C 2 T x , which is consistent with the trend seen in the XRD and SEM results (Figures 2a and S4). Besides, for nitrogen adsorption-desorption isotherms and corresponding pore-size distribution analysis ( Figure S5 and Table S2), TT-5 shows larger surface area and more micropores (<2 nm) as compared to TT-2, which may favor the adsorption of CO 2 gas molecules. However, following adding excessive HF, the TiO 2 may further be etched (as shown in Figure S4), resulting in a decrease in the photoactivity for TT-6 and TT-8. The results suggest that there is an optimum amount (5 mL) of HF in the synthesis of TiO 2 /Ti 3 C 2 T x for achieving high photocatalytic CO 2 reduction rates. Among all the samples, TT-5 exhibited the best CO 2 photoreduction rate of 13.45 µmol g −1 h −1 (10.28 µmol g −1 h −1 for CH 4 and 3.17 µmol g −1 h −1 for CO), which was higher than that of P25 TiO 2 (10.95 µmol g −1 h −1 ). Note that methane was not detected for commercial P25, while TT-5 was up to 92.84% methanogenic selectivity (Figure 4b). As reported previously, the existence of Ti 3 C 2 favors the selectivity for photocatalytic CO 2 -CH 4 conversion [36,37]. Combining the reported studies and above experimental results, the reason for the extraordinary selectivity for CH 4 production could be interpreted as Ti 3 C 2 with lower work function induces the photogenerated hole transfer and suppresses the charge recombination, facilitating the CO 2 multi-electron reduction. A series of control experiments were conducted to reveal the source of reduct products; the results are shown in Figure S6. The product concentration is below the li of detection when the experiments were carried out in the absence of catalyst or irrad tion or in Ar, proving that the detectable CO and CH4 were derived from photocataly CO2 reduction over (001)TiO2/Ti3C2Tx. To further disclose the origin of the products, isotopically labeled 13 CO2 experiment was performed. As shown in Figure 4c,d, the stro signals were observed at m/z = 17 and = 29, which were attributed to 13 CH4 and 13 CO, spectively. This result further confirms that the reduction products indeed originate fr the CO2 photoreduction rather than from the impurity or the sample.
With efforts to acquire an in-depth understanding of the reaction intermediates CO2 photoreduction for TT-5, in situ diffuse reflectance infrared Fourier transform sp troscopy (DRIFTS) was performed ( Figure 5). Obviously, various peaks can be observ during the photocatalytic reaction that corresponds to the formation of carbonate spec in which CO3 2− and HCO3 -are generated from CO2 and -OH [38]. The peaks at 1868, 18 1828, 1792, 1698, 1684, 1576, and 1558 cm −1 were attributed to HCOO [12,28,39], while band for COOH* was located at 1569 cm −1 [40]; they might serve as possible intermedia for CO production. Furthermore, the peaks at 1772, 1750, 1716, 1457, and 1419 cm −1 belo to HCHO [12,28,41], and the peak at 1734 cm −1 originates from CH3O [41], both of wh are important intermediate species for the production of CH4. Notably, the signal of C was undetectable on DRIFT spectra, as it is a non-polar molecule. Therefore, the CO2 p toconversion process could be deduced as follows: CO2 molecules are firstly adsorbed the surface of the TT-5 sample and reacted with H + dissociated from H2O to form COO (CO2* + e − + H + → COOH*). Then, COOH* transformed to CO* by a protonation proc (COOH* + e − + H + → CO* + H2O) [40,42]. One part of the adsorbed CO* is released to fo gaseous CO (CO* → CO); more importantly, the other part of CO* continues to particip A series of control experiments were conducted to reveal the source of reduction products; the results are shown in Figure S6. The product concentration is below the limit of detection when the experiments were carried out in the absence of catalyst or irradiation or in Ar, proving that the detectable CO and CH 4 were derived from photocatalytic CO 2 reduction over (001)TiO 2 /Ti 3 C 2 T x . To further disclose the origin of the products, an isotopically labeled 13 CO 2 experiment was performed. As shown in Figure 4c,d, the strong signals were observed at m/z = 17 and = 29, which were attributed to 13 CH 4 and 13 CO, respectively. This result further confirms that the reduction products indeed originate from the CO 2 photoreduction rather than from the impurity or the sample.
With efforts to acquire an in-depth understanding of the reaction intermediates in CO 2 photoreduction for TT-5, in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was performed ( Figure 5). Obviously, various peaks can be observed during the photocatalytic reaction that corresponds to the formation of carbonate species, in which CO 3 2− and HCO 3 − are generated from CO 2 and -OH [38]. The peaks at 1868, 1844, 1828, 1792, 1698, 1684, 1576, and 1558 cm −1 were attributed to HCOO [12,28,39], while the band for COOH* was located at 1569 cm −1 [40]; they might serve as possible intermediates for CO production. Furthermore, the peaks at 1772, 1750, 1716, 1457, and 1419 cm −1 belong to HCHO [12,28,41], and the peak at 1734 cm −1 originates from CH 3 O [41], both of which are important intermediate species for the production of CH 4 . Notably, the signal of CH 4 was undetectable on DRIFT spectra, as it is a non-polar molecule. Therefore, the CO 2 photoconversion process could be deduced as follows: CO 2 molecules are firstly adsorbed on the surface of the TT-5 sample and reacted with H + dissociated from H 2 O to form COOH (CO 2 * + e − + H + → COOH*). Then, COOH* transformed to CO* by a protonation process (COOH* + e − + H + → CO* + H 2 O) [40,42]. One part of the adsorbed CO* is released to form gaseous CO (CO* → CO); more importantly, the other part of CO* continues to participate in the reduction reaction to form HCHO* and CH 3 O* (CO* + 2e − + 2H + → HCHO*; HCHO* + e − + H + → CH 3 O*). Eventually, the CH 4 production could be achieved by the further protonation of CH 3 O* (CH 3 O* + 3e − + 3H + → CH 4 ↑+ H 2 O) [43]. in the reduction reaction to form HCHO* and CH3O* (CO* + 2e − + 2H + → HCHO*; HCHO* + e − + H + → CH3O*). Eventually, the CH4 production could be achieved by the further protonation of CH3O* (CH3O* + 3e − + 3H + → CH4↑+ H2O) [43]. Furthermore, considering that the separation and migration of photoinduced electron-hole pairs play an important role in achieving the enhancement of photocatalytic efficiency, the transient photocurrent response test was performed. As shown in Figure 6a, the anode photocurrent increases during irradiation and declines without irradiation. It can clearly be seen that TT-5 exhibited improved photocurrent responses compared with other samples, suggesting the efficient separation of the charge carriers in the composites of TT-5. Notably, the rapid decay characteristic of P25 could be assigned to fast electronhole recombination kinetics due to the absence of Ti3C2Tx with high conductivity [12,44]. For the others, their photocurrent is relatively stable, because the presence of Ti3C2Tx slows down the recombination rate of photogenerated carriers, which is beneficial for the multielectron reduction of CO2 (CO2 + 8H + + 8e − → CH4↑ + 2H2O).
In an effort to better understand the pathway of interfacial charge transfer between (001)TiO2 nanosheets (synthesized according to our previous report [45]) and Ti3C2Tx, the work functions (Wsample) of them were studied via Kelvin probe using a gold mesh as the reference. Specifically, the Wsample is calculated by Equation (1) [46]: where CPDsample is the contact potential difference (CPD) of the given sample, which was measured by the Kelvin probe; Wprobe represents the work function of the gold mesh (≈4.25 eV); and e is the electronic charge. Based on the probed CPD results (Figure 6b), the W of (001)TiO2 nanosheets and Ti3C2Tx are calculated to be 4.66 and 4.42 eV, respectively (Figure 6c). Accordingly, the Fermi levels (Ef) of the samples are determined via the conversion formulas (Equations (2)-(4)) [28], as shown below: Ef(vs.SHE, pH7) = Ef(vs.SHE, pH0) − 0.059 pH (4) where the Ef,vac represents the Fermi levels of the samples at the vacuum level, and Evac is the energy of a stationary electron at the vacuum level (regarded as 0 eV). Thus, the Ef of Ti3C2Tx and TiO2 are calculated to be −0.49 and −0.25 eV (vs. SHE at pH = 7), respectively. Obviously, the Fermi energy level of Ti3C2Tx is more negative than that of (001)TiO2 nanosheets (i.e., the work function of Ti3C2Tx is lower than that of (001)TiO2), and thus the photogenerated holes rather than electrons of TiO2 nanosheets are trapped by Ti3C2Tx. Furthermore, considering that the separation and migration of photoinduced electronhole pairs play an important role in achieving the enhancement of photocatalytic efficiency, the transient photocurrent response test was performed. As shown in Figure 6a, the anode photocurrent increases during irradiation and declines without irradiation. It can clearly be seen that TT-5 exhibited improved photocurrent responses compared with other samples, suggesting the efficient separation of the charge carriers in the composites of TT-5. Notably, the rapid decay characteristic of P25 could be assigned to fast electron-hole recombination kinetics due to the absence of Ti 3 C 2 T x with high conductivity [12,44]. For the others, their photocurrent is relatively stable, because the presence of Ti 3 C 2 T x slows down the recombination rate of photogenerated carriers, which is beneficial for the multi-electron reduction of CO 2 (CO 2 + 8H + + 8e − → CH 4 ↑ + 2H 2 O).
In an effort to better understand the pathway of interfacial charge transfer between (001)TiO 2 nanosheets (synthesized according to our previous report [45]) and Ti 3 C 2 T x , the work functions (W sample ) of them were studied via Kelvin probe using a gold mesh as the reference. Specifically, the W sample is calculated by Equation (1) [46]: where CPD sample is the contact potential difference (CPD) of the given sample, which was measured by the Kelvin probe; W probe represents the work function of the gold mesh (≈4.25 eV); and e is the electronic charge. Based on the probed CPD results (Figure 6b), the W of (001)TiO 2 nanosheets and Ti 3 C 2 T x are calculated to be 4.66 and 4.42 eV, respectively ( Figure 6c). Accordingly, the Fermi levels (E f ) of the samples are determined via the conversion formulas (Equations (2)-(4)) [28], as shown below: E f (vs.SHE, pH7) = E f (vs.SHE, pH0) − 0.059 pH (4) where the E f,vac represents the Fermi levels of the samples at the vacuum level, and E vac is the energy of a stationary electron at the vacuum level (regarded as 0 eV). Thus, the E f of Ti 3 C 2 T x and TiO 2 are calculated to be −0.49 and −0.25 eV (vs. SHE at pH = 7), respectively. Obviously, the Fermi energy level of Ti 3 C 2 T x is more negative than that of (001)TiO 2 nanosheets (i.e., the work function of Ti 3 C 2 T x is lower than that of (001)TiO 2 ), and thus the photogenerated holes rather than electrons of TiO 2 nanosheets are trapped by Ti 3 C 2 T x . Based on the above experimental results, we inferred the mechanism of (001) TiO2/Ti3C2Tx for photocatalytic CO2 reduction (Figure 6d). Under illumination, the electrons on the valence band of (001)TiO2 nanosheets are excited to the conduction band, and these photoelectrons gathered on the conduction band react with the adsorbed CO2 to generate CH4 and CO via the photocatalytic process. Because of the differences in work function for (001)TiO2 nanosheets and Ti3C2Tx with metallic properties, a Schottky barrier could be formed at the interface of (001)TiO2 and Ti3C2Tx [16,22], which would promote the holes rather than electrons transfer to Ti3C2Tx, while enabling the accumulation of electrons on TiO2. Undoubtedly, this is favorable for the efficiently multi-electron CO2 photoreaction.

Chemicals
All the chemicals were of analytic grade and used as received. Deionized (DI) water used in this research was from local sources. Ti3AlC2 power was supplied by 11 Technology Co., Ltd. (Changchun, China). Hydrofluoric acid (HF), sulfuric acid (H2SO4), and sodium bicarbonate (NaHCO3) were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).

Sample Preparation
Preparation of Ti3C2Tx: Ti3C2Tx was prepared by an HF-etching method. In detail, 1 g Ti3AlC2 powder was slowly put into the aqueous HF solution (40 wt%) under magnetic stirring for 24 h to eliminate the Al layer. Then, the black suspension was filtered and washed with distilled water until the pH of the filtrate was about 7. After drying at 60 ℃ under vacuum for 12 h, the Ti3C2Tx powder was obtained. Based on the above experimental results, we inferred the mechanism of (001) TiO 2 /Ti 3 C 2 T x for photocatalytic CO 2 reduction (Figure 6d). Under illumination, the electrons on the valence band of (001)TiO 2 nanosheets are excited to the conduction band, and these photoelectrons gathered on the conduction band react with the adsorbed CO 2 to generate CH 4 and CO via the photocatalytic process. Because of the differences in work function for (001)TiO 2 nanosheets and Ti 3 C 2 T x with metallic properties, a Schottky barrier could be formed at the interface of (001)TiO 2 and Ti 3 C 2 T x [16,22], which would promote the holes rather than electrons transfer to Ti 3 C 2 T x , while enabling the accumulation of electrons on TiO 2 . Undoubtedly, this is favorable for the efficiently multi-electron CO 2 photoreaction.

Chemicals
All the chemicals were of analytic grade and used as received. Deionized (DI) water used in this research was from local sources. Ti 3 AlC 2 power was supplied by 11 Technology Co., Ltd. (Changchun, China). Hydrofluoric acid (HF), sulfuric acid (H 2 SO 4 ), and sodium bicarbonate (NaHCO 3 ) were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).

Sample Preparation
Preparation of Ti 3 C 2 T x : Ti 3 C 2 T x was prepared by an HF-etching method. In detail, 1 g Ti 3 AlC 2 powder was slowly put into the aqueous HF solution (40 wt%) under magnetic stirring for 24 h to eliminate the Al layer. Then, the black suspension was filtered and washed with distilled water until the pH of the filtrate was about 7. After drying at 60°C under vacuum for 12 h, the Ti 3 C 2 T x powder was obtained.
Preparation of (001) TiO 2 /Ti 3 C 2 T x : (001) TiO 2 /Ti 3 C 2 T x composite photocatalysts were synthesized by a one-step hydrothermal method. Typically, 2 g Ti 3 AlC 2 powder was slowly added to HF solution (5 mL, 40 wt%) and kept agitating at room temperature for 24 h in order to remove the Al layer. Then, distilled water was directly added to the above suspensions to a total volume of 120 mL. After sonicating for 10 min, the homogeneous mixture was sealed into a 200 mL Teflon-lined autoclave with a stainless steel jacket for a hydrothermal process at 180 • C for 12 h, in which HF was delicately used as both an etchant for Ti 3 AlC 2 and a morphology control agent for the growth of (001)TiO 2 nanosheets on Ti 3 C 2 T x . After natural cooling, the precipitate was collected by filtration, washed thoroughly with distilled water until the pH of the filtrate turned to 7, and dried at 60 • C. Subsequently, the (001)TiO 2 /Ti 3 C 2 T x was obtained and labeled as TT-5 (5 mL HF) for short.
Similarly, a serial of composites was synthesized by varying the usage of HF under other identical conditions, and the acquired samples were named TT-x (x = 2, 3, 4, 5, 6, and 8), where x represents the usage (unit: mL) of HF. To study the transformation process of Ti 3 AlC 2 to (001)TiO 2 /Ti 3 C 2 T x , samples pretreated with 5 mL HF but without hydrothermal process were collected and labeled TT-5-u.

Characterization
The X-ray diffraction (XRD) patterns of the samples were recorded on an X-ray diffractometer (D8 ADVANCE Bruker, Karlsruhe, Germany) with Cu Kα radiation at a scan rate of 0.02 • s −1 . Raman spectra were collected through a Raman spectroscopy (DXR2 XI, Thermo Fisher, Waltham, MA, USA) with an excitation wavelength of 633 nm. The morphology and microstructure of the samples were analyzed using a field emission scanning electron microscope (FESEM, SU8010 Hitachi, Tokyo, Japan) and high-angle annular dark-field scanning/transmission electron microscopy (HAADF-STEM, Talos F200S, Thermo Scientific, Waltham, MA, USA), respectively. X-ray photoelectron spectroscopy (XPS) measurements were performed using a VG Multilab 2000 XPS system (Thermo VG Scientific, London, UK) to obtain information related to the surface chemical state and composition of the sample. All the spectra were calibrated to the C 1 s peak at 284.8 eV. The UV-vis diffuse reflectance spectra (DRS) were obtained with a UV-vis spectrophotometer (UV2600, Shimadzu, Kyoto, Japan) using BaSO 4 as the reference standard. The Brunauer-Emmett-Teller (BET) surface area and pore size distribution data of the samples were analyzed using a nitrogen adsorption apparatus (ASAP-2020, Micromeritics, Norcross, GA, USA). The samples were degassed at 175 • C for 4 h before N 2 adsorption measurement. A Kelvin probe apparatus (Instytut Fotonowy, Cracow, Poland) with a sensitivity of 1 mV was utilized to determine the contact potential difference (CPD) of the samples with a gold mesh Kelvin probe as the reference. Surface work function (W) was monitored using an LED lamp as a light source in conjunction with a monochromator. In situ FTIR spectra were measured on a Bruker Tensor II FTIR spectrometer (Bruker, Karlsruhe, Germany).

Electrochemical Measurements
Electrochemical characterization for transient photocurrent responses (TPR) was conducted on an electrochemical analyzer (CHI 760e, CH Instruments, Shanghai, China). Pt wire, Ag/AgCl (saturated KCl), and 0.4 M Na 2 SO 4 solution were functioned as the counter electrode, reference electrode, and electrolyte, respectively. For the working electrode, 30 mg of catalyst was ground in 1 mL water/absolute ethanol mixed solvent (v/v = 1/1) and 30 µL of Nafion solution to make a slurry, which was coated onto indium tin oxide (ITO) conductive glass with an exposed area of 1 cm 2 after ultrasonic dispersion. A 365 nm LED (3 W, Shenzhen Lamplic, Shenzhen, China) was used as the light source.

Photocatalytic CO 2 Reduction and Isotope-Labelling Measurement
The photocatalytic CO 2 reduction of the samples was conducted in a glass automatic online gas analysis system (Labsolar-6A, Beijing, China) ( Figure S1a). The light source was a 300 W xenon lamp (PLS-SXE300+, Beijing, China). The photocatalytic activity test was carried out in three main steps: coating catalyst, in situ CO 2 production, and the occurrence of the CO 2 photocatalytic reduction. (1) Coating catalyst: 30 mg of the photocatalyst was dispersed in 15 mL distilled water to form a suspension, which was then transferred to a petri dish (φ = 60 mm). After drying, the catalyst film was deposited on the bottom of the dishes. (2) In situ CO 2 generation: 2.1 g of NaHCO 3 powder was placed on the bottom of a homemade double-neck reactor, and the coated dish in (1) was supported in the reactor using a quartz tripod ( Figure S1b). Before illumination, the photocatalytic system was vacuumed to exclude the air. Finally, 7 mL of H 2 SO 4 (2 M) was injected into the reactor to react with NaHCO 3 for generating CO 2 and H 2 O vapor within the reactor.
(3) CO 2 photoreduction ( Figure S1c): After adsorption equilibrium, a 300 W xenon lamp (PLS-SXE300+, Perfect Light) was used as the light source to initiate the photocatalytic reaction. The concentration of the CO 2 reduction products was analyzed by gas chromatography (GC-2014, Shimazu, Kyoto, Japan) with FID and TCD detectors. For determining the carbon source of the reduction products, 13 CO 2 isotope labeling experiment was carried out under the same conditions as the above-mentioned CO 2 reduction test, except for the substitute of NaH 12 CO 3 using NaH 13 CO 3 (SHANGHAI ZZBIO Co., Ltd., Shanghai, China). The gas products were verified by a mass spectrometer (MS, HPR-20, Hiden Analytical, Warrington, UK). The electron selectivity for CH 4 production during the photocatalytic CO 2 reduction process is evaluated using the following equation [47]:

Conclusions
In summary, a facile one-step hydrothermal approach of (001)TiO 2 /Ti 3 C 2 T x composite was developed, where HF was delicately used as both an etchant for Ti 3 AlC 2 and a morphology control agent for TiO 2 nanosheets. This method simultaneously addressed the issues of the large usage of HF and complex operation procedure. Furthermore, by tailoring the HF amount within a narrow range, the morphology of TiO 2 nanosheets could be purposively manipulated. The optimized sample (TT-5) shows a superior photoactivity with a CO 2 reduction rate of 13.45 µmol g −1 h −1 , as well as a CH 4 production selectivity of 92.84%. The excellent photocatalytic performance is mainly since purposeful spatial isolation of photogenerated charge carriers induced by Ti 3 C 2 T x dramatically improves the electron density of TiO 2 , thereby facilitating the CO 2 multi-electron reduction of (001)TiO 2 /Ti 3 C 2 T x heterojunction. This work may provide an easy and novel approach to fabricating highly efficient MXene-derived composites for sustainable energy conversion.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/catal12070785/s1, Figure S1: Photocatalytic CO 2 reduction device (a), in situ CO 2 production (b), CO 2 photoreduction reaction (c); Figure S2: SEM image of Ti 3 AlC 2 (a) and Ti 3 C 2 (b), showing that the morphology of bulk Ti 3 AlC 2 changed to accordion-like Ti 3 C 2 T x , as a consequence of the eliminated Al layer via HF etching; Figure S3: X-ray photoelectron spectroscopy (XPS) survey spectra (a) and high-resolution XPS spectra of Al 2p (b) of Ti 3 AlC 2 , TT-5-u and TT5; Figure S4: SEM images of TiO 2 /Ti 3 C 2 T x samples prepared with different usage of HF; Figure S5: Nitrogen adsorption/desorption isotherms and the corresponding pore-size distribution curves (inset) of the TT-2, TT-5 and TT-8 samples; Figure S6: The control experiments of photocatalytic CO 2 reduction performance over TT-5 under altered conditions; Table S1: Synthesis conditions of (001)TiO 2 /Ti 3 C 2 T x composites; Table S2: BET surface area and the corresponding pore volume and pore size of samples [48][49][50][51].