Peiquan
Ling‡
a,
Juncheng
Zhu‡
a,
Zhiqiang
Wang
a,
Jun
Hu
a,
Junfa
Zhu
a,
Wensheng
Yan
a,
Yongfu
Sun
*ab and
Yi
Xie
*ab
aHefei National Research Center for Physical Sciences at the Microscale, National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei 230026, China. E-mail: yfsun@ustc.edu.cn; yxie@ustc.edu.cn
bInstitute of Energy, Hefei Comprehensive National Science Center, Hefei 230031, China
First published on 10th June 2022
Arduous CO2 activation and sluggish charge transfer retard the photoreduction of CO2 to CH3OH with high efficiency and selectivity. Here, we fabricate ultrathin Ti-doped WO3 nanosheets possessing approving active sites and optimized carrier dynamics as a promising catalyst. Quasi in situ X-ray photoelectron spectroscopy and synchrotron–radiation X-ray absorption near-edge spectroscopy firmly confirm that the true active sites for CO2 reduction are the W sites rather the Ti sites, while the Ti dopants can facilitate charge transfer, which accelerates the generation of crucial COOH* intermediates as revealed by in situ Fourier-transform infrared spectroscopy and density functional theory calculations. Besides, the Gibbs free energy calculations also validate that Ti doping can lower the energy barrier of CO2 activation and CH3OH desorption by 0.22 eV and 0.42 eV, respectively, thus promoting the formation of CH3OH. In consequence, the Ti-doped WO3 ultrathin nanosheets show a superior CH3OH selectivity of 88.9% and reach a CH3OH evolution rate of 16.8 μmol g−1 h−1, about 3.3 times higher than that on WO3 nanosheets. This work sheds light on promoting CO2 photoreduction to CH3OH by rational elemental doping.
To meet the above requirements, constructing ultrathin two-dimensional (2D) nanosheets with appropriate elemental doping is a very promising solution. On the one hand, ultrathin 2D nanosheets can provide abundant active sites for CO2 adsorption and activation due to their high specific surface area and sufficient uncoordinated atoms.21–23 Their light absorption ability is also different from that of the traditional bulk catalysts, while the efficiencies of photogenerated carrier separation and charge transfer can be promoted thanks to their reduced migration paths.24–28 On the other hand, elemental doping can regulate the electron structure of active atoms, further expediting CO2 activation and reduction.29,30 It can also change the band structure of catalysts and extend their light absorption range, making the catalysts more suitable for CO2 photoreduction.31–33 Furthermore, suitable elemental doping can optimize the photogenerated carrier dynamics, alter the reaction energy barrier and control reaction intermediates, enhancing product selectivity towards the desired species.29,30,34 Therefore, it is prospective to boost CO2 photoreduction to CH3OH through synthesizing ultrathin 2D nanosheets with elemental doping. In this regard, inexpensive and eco-friendly tungsten oxide (WO3) is one of the promising candidates, thanks to its suitable band gap, good light-harvesting ability and easily tunable electron structure.35,36 Although WO3 nanosheets have been applied in many photochemical fields,36–38 their performance in CO2 photoreduction to CH3OH still lacks exploration.
In this work, Ti-doped WO3 ultrathin nanosheets were fabricated via an acid-assisted method. X-ray photoelectron spectroscopy (XPS), synchrotron–radiation X-ray absorption near-edge spectroscopy (XANES) and element mappings show the successful doping of Ti into WO3 nanosheets, while photoluminescence (PL) and time-resolved photoluminescence spectra (TRPL) reveal the enhanced separation of photogenerated carriers due to the Ti doping. Besides, quasi in situ XPS spectra and quasi in situ XANES spectra firmly validate that the true active sites are the W sites rather the Ti sites, where the Ti dopants facilitate the charge transfer, which is beneficial for the formation of COOH* species, one of the most important intermediates for carbon products during CO2 reduction. In situ Fourier-transform infrared (FTIR) spectroscopy demonstrates this conclusion by the stronger intensity of COOH* species on Ti-doped WO3 nanosheets. In addition, density functional theory (DFT) calculations confirm that Ti doping can strengthen the bonding between COOH* intermediates and catalysts, while the energy barrier of CO2 activation and CH3OH desorption is decreased by 0.22 eV and 0.42 eV, respectively, which dramatically facilitates the formation of CH3OH. As a result, the Ti-doped WO3 ultrathin nanosheets show a superior CH3OH selectivity of 88.9% and achieve a CH3OH evolution rate of 16.8 μmol g−1 h−1, about 3.3 times higher than that on the WO3 nanosheets. This work offers an effective approach to enhance the conversion of CO2 and H2O into CH3OH through elemental doping.
To obtain a non-toxic and efficient catalyst for CO2 photoreduction, the ultrathin Ti-doped WO3 nanosheets (called Ti–WO3 nanosheets for short) were fabricated via an acid-assisted method. The powder X-ray diffraction (XRD) patterns of the Ti–WO3 nanosheets could be indexed well to JCPDS no. 89-1287 and no additional peak was detected, indicating the successful synthesis of pure tetragonal WO3 (Fig. 1A). The transmission electron microscopy (TEM) image showed that Ti-doped WO3 had a sheet-like morphology (Fig. 1B). Meanwhile, the high-resolution TEM (HRTEM) image showed two interplanar spacings of 0.367 nm and 0.382 nm with a dihedral angle of 90°, corresponding to the (200) and (001) planes of tetragonal WO3, which demonstrated their [010] orientation (Fig. 1C). The atomic force microscopy (AFM) image demonstrated that the thickness of the Ti–WO3 nanosheets was about 2.20 nm (Fig. 1D), which was approximately the thickness of the triple-unit-cell slab along the [010] direction, illustrating the successful fabrication of ultrathin Ti–WO3 nanosheets. For comparison, the undoped WO3 nanosheets (called WO3 nanosheets for short) were obtained without the addition of Ti sources during the synthesis. The WO3 nanosheets had the same crystalline phase, growth orientation and thickness as the Ti–WO3 nanosheets (Fig. S1†). It is worth mentioning that the crystallinity of the Ti–WO3 nanosheets was lower than that of the WO3 nanosheets (Fig. 1A), which could be ascribed to the successful Ti doping into WO3.39,40 In addition, the Ti doping also slightly decreased the interplanar spacings in the Ti–WO3 nanosheets (Fig. 1C and S1†), which was in agreement with a previous study.40
To further confirm that the Ti dopants had been doped in WO3 nanosheets, X-ray photoelectron spectroscopy (XPS) of W 4f and Ti 2p was performed. As shown in Fig. 1E, there were two peaks at around 37.90 eV and 35.75 eV, corresponding to W 4f5/2 and W 4f7/2 states, respectively. Notably, both W 4f5/2 and W 4f7/2 in the Ti–WO3 nanosheets exhibited a slightly negative shift, implying that some electrons were transferred to W6+ after Ti doping, resulting in more W5+ species.35 This phenomenon was due to the partial replacement of W6+ by the Ti atoms.41 In addition, there was no Ti 2p signal detected in the WO3 nanosheets, while Ti 2p in the Ti–WO3 nanosheets showed two peaks at 458.7 eV and 464.4 eV, which could be ascribed to the Ti(IV) 2p3/2 peak and the Ti(IV) 2p1/2 peak (Fig. 1F). This result was in accordance with the conclusion drawn from synchrotron–radiation X-ray absorption near-edge spectroscopy (XANES), which showed that the valence state of Ti in the Ti–WO3 nanosheets was +4 (Fig. S2†). The content of Ti in the Ti–WO3 nanosheets determined using inductively coupled plasma atomic emission spectroscopy (ICP-OES) was 0.87 wt%, which was close to that determined by the XPS results (0.90 wt%). Furthermore, the annular dark-field TEM image and the corresponding element mappings revealed that there was no TiO2 particle observed and the W, O and Ti elements were distributed uniformly (Fig. S3†). The Raman spectra also demonstrated that there was no additional peak detected in the Ti–WO3 nanosheets (Fig. S4†), excluding the possibility of TiO2/WO3 heterojunction formation. Thus, these results showed that the Ti dopants had been doped in the WO3 nanosheets and the valence state of Ti in the Ti–WO3 nanosheets was +4.
To investigate the effect of Ti dopants on the electronic band structures, ultraviolet–visible (UV–vis) absorption spectra were recorded. As shown in Fig. 2A, the absorption spectrum of the Ti–WO3 nanosheets showed a red-shift compared with that of the WO3 nanosheets, implying the stronger visible light absorption on the Ti–WO3 nanosheets.21,31,35 In addition, the bandgap of the Ti–WO3 nanosheets determined using the absorption spectra was 2.59 eV (Fig. 2B), which was smaller than that of the WO3 nanosheets (2.66 eV). To obtain the band edge potentials of the samples, synchrotron–radiation photoemission spectroscopy (SRPES) was applied to measure the work functions and the valence band maxima (VBM).36 As shown in Fig. 2C and D, the work functions of the Ti–WO3 nanosheets and the WO3 nanosheets could be calculated to be 4.30 and 4.53 eV, whereas their valence-band edges were located at 2.27 eV and 2.48 eV. Combined with the bandgaps obtained above, the electronic band structures of the Ti–WO3 nanosheets and the WO3 nanosheets could be estimated (Table S1†). It was obvious that both samples had suitable electronic band structures for reducing CO2 to CH3OH, confirming their potential applications in CO2 photoreduction.5
To evaluate the effect of Ti dopants on the catalytic activity, photocatalytic CO2 reduction experiments were carried out. The liquid products were analyzed using nuclear magnetic resonance (NMR) spectroscopy, while the gas products were analyzed using a gas chromatograph (GC) equipped with a thermal conductivity detector (TCD) and a flame ionization detector (FID). As displayed in Fig. 3A, CH3OH was the major product of CO2 photoreduction on both the samples, since the CO evolution rate was much lower than the CH3OH evolution rate. The CH3OH evolution rate on the Ti–WO3 nanosheets could reach 16.8 μmol g−1 h−1, which was about 3.3 times higher than that on the WO3 nanosheets, suggesting that Ti doping could efficiently promote the catalytic activity of WO3 nanosheets. The selectivity towards CH3OH also increased from 80.0% to 88.9%. In addition, the CH3OH evolution rate was related to the amount of Ti dopants (Fig. S5†), indicating that Ti doping was a feasible strategy to adjust the performance of WO3 nanosheets during the CO2 photoreduction. What's more, the control experiments demonstrated that CO2, the Ti–WO3 nanosheets (catalyst) and illumination were prerequisites for the CH3OH formation, indicating that CH3OH was derived from photocatalytic CO2 reduction over the Ti–WO3 nanosheets (Fig. 3B). This conclusion could be further verified by the 13CO2 labelling experiments performed with synchrotron-based vacuum ultraviolet photoionization mass spectrometry (SVUV-PIM). As shown in Fig. S6,† N2, CO and CO2 could not be ionized when the photon energy was set at 11.50 eV, and thus one could eliminate the interference from their fragments.42 Hence, the Ti–WO3 nanosheets produced the product 13CH3OH (m/z = 33), confirming that evolved CH3OH indeed originated from the photoreduction of CO2 (Fig. 3C). Furthermore, the CH3OH evolution rate showed almost no decay after 4 cycles of the photoreduction test (Fig. 3D), indicating the good photocatalytic stability of the Ti–WO3 nanosheets. The corresponding XRD patterns and TEM images after catalysis could also affirm the good stability of the Ti–WO3 nanosheets (Fig. S7†).
To unravel the true active sites during CO2 photoreduction over the Ti–WO3 nanosheets, quasi in situ XPS spectra were used to investigate the change in electron density on the W sites. As shown in Fig. 4A and B, the content of W5+ increased distinctly upon light irradiation of the reaction system, indicating that the W sites accepted the photoexcited electrons and generated W5+ active species during the reaction. Once CO2 was introduced into the system, the ratio of W5+/W6+ decreased from 0.38 to 0.25, indicating that the W5+ ions were reoxidized into the original W6+ ions by donating electrons to the adsorbed CO2 molecules.43 In addition, the valence state of the W atoms recovered to the initial state after removing the light irradiation (Fig. 3C) and the same trend was also detected on the WO3 nanosheets (Fig. S8†). Notably, the W5+ content in the Ti–WO3 nanosheets was higher than that in the WO3 nanosheets in the dark and it increased after the introduction of CO2, implying that CO2 was adsorbed on the W sites (Fig. 4D). On the contrary, the quasi in situ X-ray absorption near-edge spectroscopy (XANES) spectra revealed that the L-edge of Ti did not have any distinct changes during the whole process (Fig. S9†), indicating that the doped Ti atoms did not participate in the CO2 photoreduction directly. This further certified that the true active sites over the Ti–WO3 nanosheets were the W sites rather the Ti sites. Interestingly, the ratio of W5+/W6+ on the Ti–WO3 nanosheets was lower than that on the WO3 nanosheets under light irradiation, suggesting that more photoexcited electrons were donated to the adsorbed CO2 molecules for photoreduction. In other words, the Ti dopants could facilitate the charge transfer on the Ti–WO3 nanosheets to enhance the CO2RR performance, where the separation of photogenerated carriers on the Ti–WO3 nanosheets was also promoted, as revealed by the impedance test, transient photocurrent response, photo-luminescence (PL) and time-resolved photoluminescence (TRPL) spectra (Fig. S10 and S11†).17,31
To find out the reaction intermediates of CH3OH formation on the Ti–WO3 nanosheets, in situ FTIR measurement was carried out. As shown in Fig. 5A, three peaks appeared at 1158 cm−1, 1365 cm−1 and 1530 cm−1. Similar peaks also appeared in the in situ FTIR spectra of the pristine WO3 nanosheets, indicating the same CO2 reduction process (Fig. S12†). The peak at 1530 cm−1 could be assigned to COOH*, which was one of the most important intermediates for carbon products during CO2 reduction.36,42,44 Meanwhile, it was found that the peak density of COOH* on the Ti–WO3 nanosheets was higher than that on the WO3 nanosheets (Fig. S12†), implying that the Ti dopants might benefit CH3OH formation through promoting the generation of COOH* intermediates. The peak at 1365 cm−1 was assigned to m-CO32−, which could be due to the dissolved CO2 in the water.45 In addition, the peak at 1158 cm−1 could be ascribed to CH3O*, which was one of the intermediates for CH3OH formation.42,44 Based on the results of in situ FTIR spectra, the possible reaction pathways could be summarized as follows:
* + CO2 + e− + H+ → COOH* | (1) |
COOH* + e− + H+ → CO* + H2O | (2) |
CO* + e− + H+ → CHO* | (3) |
CHO* + e− + H+ → CH2O* | (4) |
CH2O* + e− + H+ → CH3O* | (5) |
CH3O* + e− + H+ → CH3OH + * | (6) |
To further reveal the reason for the promotion of CH3OH formation on the Ti–WO3 nanosheets, the Gibbs free energy of these key reaction pathways was calculated using DFT calculation (Fig. 5B and C and Table S4†). The calculation results suggested that the rate-limiting step was the formation of COOH* intermediates for both the samples,46,47 while the reaction energy from CO2 to COOH* was decreased from 1.37 to 1.15 eV over the Ti–WO3 ultrathin layer slab, indicating a lower activation barrier of CO2 on the WO3 nanosheets after Ti doping. This could be attributed to the strengthened bonding between COOH* intermediates and the catalyst (Fig. S13†),44 which could stabilize and generate more COOH* intermediates during the CO2 photoreduction, well consistent with the in situ FTIR results. Furthermore, the energy barrier of CH3OH desorption over the Ti–WO3 ultrathin layer slab was −0.06 eV, obviously lower than that of 0.36 eV over the WO3 ultrathin layer slab, implying that the Ti dopants could also facilitate the CH3OH desorption, which was an important process for CH3OH formation. Thus, it was rational to conclude that doping Ti into WO3 nanosheets could lower the reaction energy barrier, stabilize the COOH* intermediates and make CH3OH easier to desorb, thus improving the CH3OH yield during the CO2 photoreduction.
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2nr02364d |
‡ These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2022 |