English

• As an important pharmaceutical and fine chemical intermediate, imine has high added value and wide applications [1-3]. Developing green and efficient imine preparation pathways has been a research emphasis in organic synthesis. Photocatalytic synthesis of organics has been paid increasing attention on account of the benign conditions, atomic economy and sustainability, which provides an effective approach for the synthesis of imines [4,5]. Now, photocatalytic processes of N-benzylidene benzylamine include the coupling of benzylamine via dehydrogenation [6], conversion of amines and alcohols [7,8] and self-coupling of benzylamine by oxidation [9-11]. Photocatalytic oxidation coupling of alcohols and amines provides a sustainable pathway for green synthesis of imines. Recently, a wide range of photocatalysts have been explored to synthesize imines, such as BiOCl [12], TiO₂ [13], sulfides (CdS [14], SnS₂ [15]) and C₃N₄ [16]. However, achieving precise production of imine still necessitates the fabrication of new photocatalysts.

  Metal-organic frameworks (MOFs) are a large type of crystalline materials composed of metal ions or metal clusters with organic ligands, have widespread applications owing to their large specific surface area, highly ordered pore structures and tunable functionalities. Moreover, MOFs can be considered as photocatalysts because of their semiconductor-like properties. These unique characteristics endow MOFs with a wealth of potential applications in H₂ production [17-22], CO₂ reduction [23-27], dye degradation [28-30] and organic transformation [31]. To further promote catalytic efficiency, defects are introduced into MOFs via multiple strategies, such as eliminating some organic ligands, regulating their morphology and doping other metals [32-35]. The abundant defects can chemisorb and activate the reactants molecules while promote charge separation for improving the catalytic performance. However, the defects can also serve as recombination centers for photogenerated charge carriers, reducing the catalytic performance [36]. Among above strategies of defects creation, they are all unable to effectively prevent this disadvantageous factor of defects. Therefore, developing a strategy to construct defects on catalysts that can both effectively activate reactant molecules and prevent the recombination of photogenerated charge carriers is key to enhancing catalytic performance, which may be a challenge topic.

  It is well known that constructing heterojunction catalysts is an extremely effective method for optimizing the dynamics of photogenerated charge carriers [37]. Based on this advantage of heterojunction, it may be a very effective approach for overcoming the shortcomings of defects via rationally constructing abundant defects on heterojunction catalysts. It is reported that there is interfacial interaction between different components of heterojunction, which would induce the defects formation. Two-dimensional (2D) MOFs as the focus photocatalysts have attracted extensive attention, such as Cu-TCPP [38], NiFc-MOF/NF [39], Pd/NH₂−MIL-125(Ti) [40] and FeNi-MOFs [41]. Compared with 3D MOFs, 2D MOFs exhibit special advantages for construct heterojunction due to their open 2D structure. Among these 2D MOFs, NH₂−MIL-125(Ti) nanosheet has attracted widespread attention owing to its low toxicity, easy synthesis, excellent visible photo-response and stability [42], leading to the NH₂−MIL-125(Ti) nanosheet potentially as a desirable photocatalyst in photocatalytic formation of imines [43-45]. Moreover, CdS has been commonly applied in photocatalysis owing to its low-cost, abundant source of raw materials, simple synthesis and sufficiently negative conduction band (CB) position [46-48]. Importantly, CdS has adjustable electronic structure, which is more conducive to the generation of strong interfacial interactions on heterojunction. Therefore, introducing CdS on NH₂−MIL-125(Ti) nanosheets would generate the strong interfacial interaction, which may change the coordination environment of Ti and S sites, thus generating abundant defects on CdS/NH₂−MIL-125(Ti) and further facilitating the coordination activation of reactant molecules. Meanwhile, the heterojunction can avoid the reconfiguration for photogenerated electrons-holes. Hence, CdS/NH₂−MIL-125(Ti) photocatalyst has a potential for production of imines.

  Herein, 2D NH₂−MIL-125(Ti) nanosheets (NMT-NS) loaded with different mass of CdS clusters were developed as photocatalysts for converting benzylamine (BZA) and benzyl alcohol (BZO) to N-benzylidene benzylamine (N-BZA). The structures and morphologies of S_1.8/NMT-NS were measured by X-ray diffraction (XRD), scanning electron microscope (SEM), transmission electron microscope (TEM) and atomic force microscopy (AFM). X-ray photoelectron spectroscopy (XPS) was applied to reveal the chemical state of Ti metal sites and electronic structures of S_1.8/NMT-NS. Moreover, response surface methodology (RSM) was employed to exploit the optimized combination of reaction conditions for BZA conversion. The adsorption behavior for BZO on photocatalyst surface and intermediate active species were further elucidated by in situ infrared spectroscopy (in situ FT-IR) and electron paramagnetic resonance spectroscopy (EPR). Finally, a possible mechanism was proposed to better understand the photocatalytic pathway at molecular level.

  The crystalline phases of NMT-NS with varying CdS loading masses were detected using powder XRD (Fig. 1a). The XRD profiles of the catalysts show principal peaks corresponding to the (101), (002), (211), (222), (312), and (004) crystalline planes, which are coherent with those of simulated NH₂−MIL-125(Ti) [38]. Compared with NMT-B, NMT-NS exhibits broader diffraction peaks, indicative of the presence of a 2D structure. Furthermore, no significant peak associated with CdS is observed in the spectra of S_x/NMT-NS, likely due to the low loading mass and small diameter of CdS. The morphologies of S_1.8/NMT-NS and NMT-NS were examined by scanning electron microscopy (SEM), revealing a thin sheet-like structure (Fig. 1b and Fig. S2 Supporting information). The lengths and widths of NMT-NS and S_1.8/NMT-NS are approximately 1.3 µm and 0.8 µm, respectively. Highly dispersed CdS clusters with a diameter of approximately 3.0 nm can be observed in the TEM image of S_1.8/NMT-NS (Fig. 1c). The HR-TEM image (Fig. 1d) clearly displays crystal lattice spacings of 0.341 nm and 1.36 nm, which are assigned to the (002) crystal facet of CdS clusters and the (110) crystal plane of NMT-NS. Importantly, some lattice distortion is observed in S_1.8/NMT-NS, as the in situ growth of CdS clusters generates lattice stress for NMT-NS. Lattice distortion can lead to the formation of defects and uncoordinated metal atoms, potentially improving catalytic performance. Elemental mapping images demonstrate that C, N, O, Ti, Cd, and S elements are dispersed throughout S_1.8/NMT-NS (Fig. 2 and Fig. S3 in Supporting information). AFM images in Fig. S4 (Supporting information) indicate that the average thickness of NMT-NS is approximately 10 nm. Porosity and specific surface area (S_BET) of S_1.8/NMT-NS and NMT-NS were investigated using N₂ adsorption-desorption isotherms. As shown in Fig. S5 and Table S1 (Supporting information), the S_BET and average pore size of NMT-NS are 1380.6 m²/g and 2.8 nm, respectively. However, the S_BET and pore size of S_1.8/NMT-NS are reduced to 1023.8 m²/g and 2.2 nm, respectively, indicating that some CdS clusters occupy the surface and pore structure of NMT-NS. These results confirm that NH₂−MIL-125(Ti) nanosheets decorated with CdS clusters have been successfully prepared.

  Figure 1

  

  Figure 1.  (a) Powder XRD patterns of the samples. (b) SEM image of S_1.8/NMT-NS. (c) TEM image of S_1.8/NMT-NS. (d) HRTEM image of S_1.8/NMT-NS.

  DownLoad: Full-Size Img PowerPoint

  Figure 2

  

  Figure 2.  (a, b) HAADF imaging. (c) EDX elemental mapping of C, N, Ti, O, Cd, S.

  DownLoad: Full-Size Img PowerPoint

  The surface chemical state of NMT-B, NMT-NS and S_1.8/NMT-NS was detected by XPS. Fig. 3a presents XPS survey of the samples. Distinctive peaks associated with C, N, O, Ti, Cd and S elements are observed. Fig. 3b exhibits that Cd 3d has a pair of spin-orbit bimodal peaks at 404.48 eV and 411.18 eV, attributing to Cd²⁺ 3d_5/2 and Cd²⁺ 3d_3/2 [49]. Compared with the pristine CdS, there is no shift for the Cd²⁺ 3d peak of S_1.8/NMT-NS. In the XPS profile of S 2p, the peaks at 160.98 eV and 162.31 eV can be assigned to S^2- 2p_3/2 and S^2- 2p_1/2 (Fig. 3c). The binding energy of S^2- 2p decreases about 0.2 eV in S_1.8/NMT-NS, compared with that in CdS, which suggests that the charge of NMT-NS can be transferred into CdS. The spectra of Ti 2p exhibit two peaks at 458.49 and 464.14 eV, ascribing to Ti⁴⁺ 2p_1/2 and Ti⁴⁺ 2p_3/2 of NMT-B and NMT-NS (Fig. 3d). Specially, Ti 2p spectrum of S_1.8/NMT-NS will be further deconvoluted into additional two peaks at 457.51 eV and 462.42 eV, assigning to Ti³⁺ 2p_1/2 and Ti³⁺ 2p_3/2, respectively. Fig. 3e shows the O 1s spectra of the samples. Three peaks decomposed from unsymmetrical O 1s peak can be allocated to lattice oxygen (O1), coordination of -OH (O2) and absorbed oxygen (O3). Table S2 (Supporting information) shows that the species of O3 in S_1.8/NMT-NS (24.19%) is higher than that in NMT-NS (19.17%) and NMT-B (15.36%), suggesting that more defects will be formed on the surface of S_1.8/NMT-NS. Electron paramagnetic resonance (EPR) was utilized to confirm the electron structure of the catalysts. Fig. 3f displays the obvious signal at g = 2.002, belonging to the unpaired electrons. The single intensity of S_1.8/NMT-NS is stronger than that of NMT-NS, NMT-B, and CdS, indicating a higher number of defects on S_1.8/NMT-NS. These results reveal that there are abundant Ti³⁺ and defect sites on S_1.8/NMT-NS, which may enhance the catalytic performance. Defects may serve as electron or holes trap to facilitate the separation of photogenerated carriers. Furthermore, defects can absorb and activate more O₂ from air. Ti³⁺ sites serve as Lewis acid sites to coordinate phenylcarbinol molecules.

  Figure 3

  

  Figure 3.  High-resolution XPS profiles of (a) survey scan. (b, c) Cd 3d and S 2p in S_1.8/NMT-NS and CdS. (d, e) Ti 2p and O 1s in NMT-B, NMT-NS, and S_1.8/NMT-NS. (f) EPR spectra of CdS, NMT-B, NMT-NS, and S_1.8/NMT-NS.

  DownLoad: Full-Size Img PowerPoint

  UV–vis diffuse reflection spectrum (UV–vis DRS) spectra reveal that the samples have strong absorption of visible light (Fig. S6a in Supporting information). The absorption edge of S_x/NMT-NS shows a blue shift due to the inter-band transition between CdS clusters and NMT-NS. Fig. S6b (Supporting information) indicates that the bandgaps of NMT-NS and CdS are 2.58 eV and 2.31 eV, respectively. Mott-Schottky plots show that the flat-band voltages of NMT-NS and CdS are calculated as −1.06 V and −1.59 V (vs. Ag/AgCl) (Fig. S7 in Supporting information). Therefore, the conduction bands (CB) of NMT-NS and CdS are −0.86 V and −1.39 V (vs. NHE), respectively. The valence bands (VB) of NMT-NS and CdS are located at 1.72 and 0.92 V, respectively. The EPR spectra of DMPO-^•OH for CdS, NMT-B, NMT-NS, and S_1.8/NMT-NS are shown in Fig. S8 (Supporting information). These quadruple peaks correspond to DMPO-^•OH. There is no obvious signal in the spectrum of CdS, suggesting that CdS is unable to produce ^•OH. Obviously, NMT-NS can produce ^•OH. Compared with CdS and NMT-NS, S_1.8/NMT-NS exhibits an enhanced peak, indicating that more holes are gathered on S_1.8/NMT-NS to produce ^•OH. Therefore, a Z-scheme heterojunction is likely formed between NMT-NS and CdS (Fig. 4a) [50-52]. The separation and transfer of electron-hole pairs were characterized by photocurrent test, photoluminescence (PL) spectroscopy, and electrochemical impedance spectroscopy (EIS) [53-55]. S_1.8/NMT-NS exhibits a stronger photocurrent response than NMT-NS and NMT-B (Fig. 4b), indicating rapid separation of photogenerated electron-hole pairs. S_1.8/NMT-NS has the smallest EIS arc radius compared to those of NMT-NS and NMT-B (Fig. 4c), suggesting the minimum impedance to surface charge migration. In Fig. S9 (Supporting information), compared with NMT-NS and NMT-B, S_1.8/NMT-NS shows lower signal intensity, indicating higher separation efficiency of photogenerated electron-hole pairs. This suggests that CdS clusters are key to improving the surface charge transfer efficiency of NMT-NS, facilitating the conversion of reactants.

  Figure 4

  

  Figure 4.  (a) Schematic description of the electronic band structures of NMT-NS and CdS. (b) Photocurrent response. (c) EIS Nyquist plots of S_1.8/NMT-NS, NMT-NS, and NMT-B.

  DownLoad: Full-Size Img PowerPoint

  In order to achieve the optimal conversion for BZA, RSM was employed to explore optimal combination of factors corresponding to conversion for BZA. Based on Box-Behnken analysis, the depletion of S_1.8/NMT-NS (X₁), loading mass of CdS clusters (X₂) and reaction time (X₃) were selected as independent variables with conversion for BZA as response value (Y). The factor codes and experiment designs are shown in Tables S3 and S4 (Supporting information). A quadratic mathematical model for BZA conversion (Y) with respect to depletion of S_1.8/NMT-NS (X₁), loading mass of CdS clusters (X₂) and reaction time (X₃) is as follows: Y = 96.82 + 1.10X₁ - 3.12X₂ + 0.45X₃ + 0.82X₁X₂ + 0.53X₁X₃ + 0.93X₂X₃ - 0.57X₁²–5.32X₂² - 0.57X₃².

  The relevance coefficient for the formulas is R² = 0.9912, demonstrating a well-fitting of the equation (Table S5 in Supporting information). The F-value (21.16) and P-value (< 0.0001) indicate that the mathematical model has a significant statistical significance. Comparing the F values of three factors, the order of influence for three factors on the conversion for BZA are ranked as follows: Loading mass of CdS clusters (X₂) > depletion of S_1.8/NMT-NS (X₁) > reaction time (X₃). The response surface graph clearly exhibits that the loading mass of CdS clusters, the consumption of photocatalysts and the reaction time have significant impact on the conversion for BZA (Figs. S10a-c in Supporting information).

  Based on mathematical model optimization of reaction conditions, photocatalytic oxidation coupling of BZA and BZO was implemented to explore performance of photocatalysts under 400 nm visible light, benzyl alcohol as the solvent (Table S6 in Supporting information). NMT-NS displays a high conversion for BZA (53.1%) and selectivity of N-BZA (96.7%). NMT-B shows a low conversion for BZA (29.8%) and selectivity of N-BZA (78.5%). CdS merely displays 21.8% conversion for BZA. NMT-NS has more defects and faster carrier separation, and thus exhibits an increasing conversion when loaded with CdS (Fig. 5a). BZA can be nearly fully converted to N-BZA over the S_1.8/NMT-NS (> 99%). The conversion of BZA decreases slightly over the S_2.0/NMT-NS, which can be attributed to the excessive CdS clusters that cover the active sites on NMT-NS. The turnover frequency (TOF) for photocatalysis over S_1.8/NMT-NS is measured as 32.7 h^-1. The optimal loading mass of CdS is 1.8 wt%, indicating that the results of RSM are reliable. Additionally, negligible products are obtained due to lack of light and photocatalyst, suggesting that this reaction is driven by photocatalytic process over S_1.8/NMT-NS. Under O₂ atmosphere, S_1.8/NMT-NS still shows excellent conversion (98.2%) for BZA and selectivity (99.1%) of N-BZA, demonstrating that O₂ molecules act as a key factor in BZA transformation. Under N₂ atmosphere, the conversion of BZA and selectivity of BZO are 14.8% and 74.6%, indicating that oxidation coupling of BZA and BZO will be driven by photogenerated holes. In Fig. S10d (Supporting information), photocatalytic performance over the S_1.8/NMT-NS was explored under different wavelength light irradiation. With the wavelength increasing from 400 nm to 450 nm, the conversion of BZA decreases from to 99.2% to 62.3% because of the decrease for light quantum. BZA can fully transform to N-BZA after 7 h (Fig. S11 in Supporting information). As the temperature rises, the conversion of BZA increases from 5.6% (30 ℃) to 74.5% (70 ℃). Heat also can drive this catalytic reaction (Table S7 in Supporting information).

  Figure 5

  

  Figure 5.  (a) Conversion for BZA and selectivity of N-BZA over NMT-NS with different loading mass of CdS clusters. (b) Cyclic test of the photocatalytic oxidation of benzylamine over S_1.8/NMT-NS. In situ FT-IR spectra of (c) S_1.8/NMT-NS, (d) NMT-NS. (1) After degassing at 150 ℃ for 3 h. (2) Adsorption of benzylamine for 30 min at room temperature (physisorption + chemisorption). (3) Further evacuation of excess benzylamine at 150 ℃ for 15 min under 5 Pa (chemisorption). (4) FT-IR spectra of BZO in KBr.

  DownLoad: Full-Size Img PowerPoint

  The universality of the prepared S_1.8/NMT-NS was explored via oxidation coupling of aryl-amines and benzyl alcohol to produce relevant imines (Table S8 in Supporting information). S_1.8/NMT-NS shows the high transformation for converting amine derivatives to the corresponding imines. The conversions for 4-fluorobenzylamine and 4-chlorobenzylamine are 86.7% and 84.3%, and selectivity of imines are 92.3% and 91.4%, respectively. The inductive effect of electron-withdrawing group improves the electrophilicity of α-carbon, bring about the strengthen of C—N bond and the weakening of N—H bond. Hence, it is more difficult to couple with aldehyde intermediate. Compared to the -F group, the -Cl group exhibits a weaker electron-inducing effect. Hence, the conversion for benzylamine with electron-withdrawing groups to their corresponding imines are more difficult, compared with benzylamine. In contrast, substrates with electron-donating groups are more efficient for converting amine to their corresponding imines (≥96%) owing to conjugation effect under identical reaction conditions. The stability of S_1.8/NMT-NS was investigated for photocatalytic coupling of BZA and BZO to N-BZA via cyclic experiments at air atmosphere. Moreover, compared with other catalysts employed in this reaction, S_1.8/NMT-NS exhibits an excellent photocatalytic performance (Table S9 in Supporting information). As shown in Fig. 5b, after five repeated photocatalytic reactions, the conversion of BZA and selectivity of N-BZA show no reduction. XRD pattern of S_1.8/NMT-NS after five cycles shows no significant change (Fig. S12 in Supporting information). The concentration of Cd²⁺ is only 0.03 mg/L in the solution after reaction (Fig. S13 in Supporting information). These results show that S_1.8/NMT-NS still well stabilized.

  In situ FT-IR was employed to explore the selective chemisorption for BZO on the surface of prepared catalysts. The catalyst is degassed at 150 ℃ for 3 h at 5 Pa to clean up surface impurities. BZO is adsorbed on the catalysts at room temperature for 30 min. Two absorption peaks at 1490 and 1448 cm^-1 are observed in Fig. 5c, which correspond for ν_C=C for aromatic ring of BZO molecule, respectively. The obvious peak at 1208 cm^-1 is attributed to ν_O-H, which confirms that BZO molecules are adsorbed on the surface of S_1.8/NMT-NS. Compared with S_1.8/NMT-NS, NMT-NS and NMT-B show weaker peaks of absorbed BZO (Figs. 5c and d and Fig. S15b in Supporting information), indicating that BZO is more strongly adsorbed on S_1.8/NMT-NS. However, no BZO adsorption is observed on the CdS surface (Fig. S14a in Supporting information). Thus, BZO molecules may adsorb on Ti³⁺ via C—O···Ti coordination bond. Figs. S14b-d and S15a (Supporting information) show that the C—N absorption peak of N-BZA cannot be detected on the S_1.8/NMT-NS, NMT-NS, CdS, and NMT-B, suggesting that N-BZA molecules can conveniently detach from the photocatalyst to inhibit over oxidation. XPS profiles of S_1.8/NMT-NS, NMT-NS, NMT-B and CdS after adsorption of benzyl alcohol (BZO) are displayed in Fig. S16 (Supporting information). There is no change for the binding energy of S 2p and Cd 3d in catalysts after adsorption of BZO. The binding energy of O 1s in NMT-NS and NMT-B has no change after adsorption of BZO. However, the binding energy of O 1s in S_1.8/NMT-NS decrease about 0.6 eV after adsorption of BZO, compared with that of S_1.8/NMT-NS, exhibiting that surface charges of S_1.8/NMT-NS are transferred from O to BZO. There is a blue shift about 0.1 eV of Ti 2p in the catalysts, suggesting that the electrons are transferred from Ti to BZO. Therefore, BZO molecules are absorbed on the Ti sites and defects.

  The role of active oxygen species was investigated using EPR with DMPO. Fig. 6a demonstrates that no evident peak is observed for NMT-B, NMT-NS and S_1.8/NMT-NS under dark. However, under visible light irradiation, a quartet of peaks can be obviously observed in the spectra, corresponding to ^•O₂^- [56]. The signals of NMT-B, CdS and NMT-NS are weaker than those of S_1.8/NMT-NS, exhibiting that S_1.8/NMT-NS can produce more ^•O₂^-. This may be attributed to the presence of more defects in S_1.8/NMT-NS and its higher efficiency in separating photogenerated charge carriers. Thus, S_1.8/NMT-NS exhibit better catalytic performance, compared with other catalysts. Additionally, to further explore the active species in this reaction, free radical capture experiments were conducted using 1 mmol/L methanol, benzoquinone and silver nitrate as scavengers to capture h⁺, ^•O₂^- and e^- [57]. The transformation of BZA without scavengers is about 99.2%. After the addition of, benzoquinone and silver nitrate, the conversion of BZA decreased to 50.6%, 37.9% and 72.3%, respectively (Fig. S17 in Supporting information). Therefore, h⁺ and ^•O₂^- are key active species in this photocatalytic reaction.

  Figure 6

  

  Figure 6.  (a) EPR spectra of ^•O₂^- generated by CdS, NMT-B, NMT-NS, and S_1.8/NMT-NS with DMPO. (b) Proposed mechanism of photocatalytic selective oxidation coupling of BZA and BZO over S_1.8/NMT-NS under visible light.

  DownLoad: Full-Size Img PowerPoint

  Based on above results, a possible photocatalytic mechanism for the formation of N-BZA by photocatalytic oxidation coupling of BZO and BZA at the S_1.8/NMT-NS interface is proposed (Fig. 6b). BZO molecules are chemisorbed on NMT-NS via a C—O···Ti coordination bond, resulting in a charge deviation from the O atoms of BZO to the Ti³⁺ sites of NMT-NS, thereby activating -C-O-bond. The surface defects can adsorb and activate O₂ molecules from air [58,59]. Visible light irradiation excites NMT-NS to produce photogenerated electrons and holes. In Z-scheme heterojunction, electrons of CB in CdS and holes of LUMO in NMT-NS would be left. Holes of VB in CdS would be combined with the electrons of LUMO in NMT-NS. Thus, Z-scheme heterojunction significantly decrease the impedance of catalyst and improve the transport efficiency of photogenerated carriers. The enriched photogenerated electrons quickly reduce the adsorbed O₂ molecule to ^•O₂^- radicals. Because benzyl alcohol acts as solvent, the ^•O₂^- radicals and holes would oxidize BZA to form Ph—C=O. Ph—C=O rapidly reacts with BZA molecules to synthesize N-BZA. Therefore, the collaboration of Ti³⁺ sites, defects and CdS clusters collectively achieves the accurately selective oxidation coupling of BZA and BZO to the N-BZA.

  In conclusion, the heterojunction catalysts, NMT-NS nanosheets decorated by CdS clusters, are developed successfully. The optimal catalyst (S_1.8/NMT-NS) shows a high conversion (> 99%) for BZA and selectivity (> 99%) of N-BZA under visible light. The interfacial interaction between CdS clusters and NMT-NS cause the lattice distortion of NMT-NS and the transfer of interfacial charges, thus generating abundant defect and Ti³⁺ sites. BZO and O₂ molecules can be activated through chemical adsorption on defects and Ti³⁺ sites. Z-scheme heterojunction between CdS clusters and NMT-NS optimizes the transfer and separation of photogenerated electrons-holes, thus accelerating the production of ^•O₂^-. Finally, a possible mechanism is proposed to elucidate photocatalytic coupling of BZA and BZO to N-BZA on S_1.8/NMT-NS. This work provides a potential strategy to induce defect on catalysts via utilizing the interfacial interactions between different components of heterojunction for improving the catalytic performance.

  Declaration of competing interest

  The authors declare that they have no conflict of interest.

  CRediT authorship contribution statement

  Hongtao Wang: Writing – original draft. Yingzhang Shi: Writing – review & editing. Jiayi Guo: Data curation. Shuzhao Sun: Methodology. Wenda Zhang: Investigation. Zhiwen Wang: Writing – review & editing. Yujie Song: Writing – review & editing, Conceptualization. Dongpeng Yan: Writing – review & editing, Conceptualization.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 22002030), Innovational Found for Scientific and Technological Personnel of Hainan Province (No. NHXXRCXM202301), Collaborative Innovation Center of Ecological Civilization, Hainan University (No. XTCX2022STC11), and the Shandong Laboratory of Advanced Materials and Green Manufacturing at Yantai (No. AMGM2024F23). We acknowledge the Analytical & Testing Center of Hainan University for the characterization of SEM and TEM.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110779.

  
  
• 1. [1]

     M.Y. Qi, M. Conte, Z.R. Tang, et al., ACS Nano 16 (2022) 17444–17453. doi: 10.1021/acsnano.2c08652

  2. [2]

     M.Y. Qi, M. Conte, M. Anpo, et al., Chem. Rev. 121 (2021) 13051–13085. doi: 10.1021/acs.chemrev.1c00197

  3. [3]

     J.Y. Li, Y.H. Li, M.Y. Qi, et al., ACS Catal. 10 (2020) 6262–6280. doi: 10.1021/acscatal.0c01567

  4. [4]

     P. Wang, X. Li, S. Fan, et al., Nano Energy 83 (2021) 105831. doi: 10.1016/j.nanoen.2021.105831

  5. [5]

     Z. Chen, C. Wu, W. Li, et al., Appl. Catal. B: Environ. 340 (2024) 123203. doi: 10.1016/j.apcatb.2023.123203

  6. [6]

     J. Zou, W. Zhou, L. Huang, et al., J. Catal. 400 (2021) 347–354. doi: 10.1016/j.jcat.2021.07.003

  7. [7]

     J. Masdemont, J.A. Luque-Urrutia, M. Gimferrer, et al., ACS Catal. 9 (2019) 1662–1669. doi: 10.1021/acscatal.8b04175

  8. [8]

     E. Ortiz, J. Shezaf, Y.H. Chang, et al., ACS Catal. 12 (2022) 8164–8174. doi: 10.1021/acscatal.2c02444

  9. [9]

     B. He, S. Zhang, Y. Zhang, et al., J. Am. Chem. Soc. 144 (2022) 4422–4430. doi: 10.1021/jacs.1c11577

  10. [10]

      P. Li, X. Dong, Y. Zhang, et al., Mater. Today Chem. 25 (2022) 100953. doi: 10.1016/j.mtchem.2022.100953

  11. [11]

      W. Sheng, F. Huang, X. Lang, Mater. Today Chem. 30 (2023) 101505. doi: 10.1016/j.mtchem.2023.101505

  12. [12]

      B. Zhang, J. Zhang, R. Duan, et al., Nano Energy 78 (2020) 105340. doi: 10.1016/j.nanoen.2020.105340

  13. [13]

      A.M. Nauth, E. Schechtel, R. Doren, et al., J. Am. Chem. Soc. 140 (2018) 14169–14177. doi: 10.1021/jacs.8b07539

  14. [14]

      Y. Huang, C. Liu, M. Li, et al., ACS Catal. 10 (2020) 3904–3910. doi: 10.1021/acscatal.0c00282

  15. [15]

      S. Mondal, L. Sahoo, C.P. Vinod, et al., Appl. Catal. B: Environ. 286 (2021) 119927. doi: 10.1016/j.apcatb.2021.119927

  16. [16]

      C. Xiao, G. Dong, T. Yao, et al., Appl. Surf. Sci. 654 (2024) 159453. doi: 10.1016/j.apsusc.2024.159453

  17. [17]

      X. Zhang, Z. Chen, Y. Luo, et al., J. Hazard Mater. 405 (2021) 124128. doi: 10.1016/j.jhazmat.2020.124128

  18. [18]

      S. Wang, Z. Ai, X. Niu, et al., Adv. Mater. 35 (2023) e2302512. doi: 10.1002/adma.202302512

  19. [19]

      J.B. Soberanas, N. Martín, M. Bacic, et al., Nat. Catal. 7 (2024) 452–463. doi: 10.1038/s41929-024-01130-7

  20. [20]

      P. Qiu, Y. An, X. Wang, et al., Chin. Chem. Lett. 34 (2023) 108246. doi: 10.1016/j.cclet.2023.108246

  21. [21]

      Z. Pan, W. Ding, H. Chen, et al., Chin. Chem. Lett. 35 (2024) 108567. doi: 10.1016/j.cclet.2023.108567

  22. [22]

      Z. Liang, J. Guo, P. Wang, et al., Chin. Chem. Lett. 34 (2023) 108001. doi: 10.1016/j.cclet.2022.108001

  23. [23]

      J. Liang, H. Yu, J. Shi, et al., Adv. Mater. 35 (2023) 2209814. doi: 10.1002/adma.202209814

  24. [24]

      T. Luo, L. Gilmanova, S. Kaskel, Chem. Soc. Rev. 490 (2023) 215210.

  25. [25]

      W. Ni, X. He, H. Chen, et al., Ener. Environ. Sci. 17 (2024) 3191–3201. doi: 10.1039/d4ee00720d

  26. [26]

      A.A. Zhang, R. Wang, H.B. Huang, et al., Chin. Chem. Lett. 34 (2023) 107311. doi: 10.1016/j.cclet.2022.03.034

  27. [27]

      L. Chen, Q. Tang, S. Wu, et al., Chin. Chem. Lett. 34 (2023) 107903. doi: 10.1016/j.cclet.2022.107903

  28. [28]

      X. Zheng, Y. Li, J. Yang, et al., Chem. Eng. J. 422 (2021) 130105. doi: 10.1016/j.cej.2021.130105

  29. [29]

      N.C. Chiu, J.M. Lessard, E.N. Musa, et al., Nat. Commun. 15 (2024) 1459. doi: 10.1038/s41467-024-45546-y

  30. [30]

      X. Li, K. Chen, R. Guo, Z. Wei, Chem. Rev. 123 (2023) 10432–10467. doi: 10.1021/acs.chemrev.3c00248

  31. [31]

      Y. Gong, Y. Ding, Q. Tang, et al., Chin. Chem. Lett. 35 (2024) 108475. doi: 10.1016/j.cclet.2023.108475

  32. [32]

      T. Jia, W. Wang, L. Zhang, et al., Appl. Catal. B Environ. 340 (2024) 123168. doi: 10.1016/j.apcatb.2023.123168

  33. [33]

      B. Yeh, S. Chheda, S.D. Prinslow, et al., J. Am. Chem. Soc. 145 (2023) 3408–3418. doi: 10.1021/jacs.2c10551

  34. [34]

      L. Jiao, H.L. Jiang, Chin. J. Catal. 45 (2023) 1–5. doi: 10.1016/S1872-2067(22)64193-7

  35. [35]

      Z. Wang, Y. Kang, Y. Shi, et al., Appl. Catal. B Environ. 340 (2024) 123162. doi: 10.1016/j.apcatb.2023.123162

  36. [36]

      Z.X. Sun, K. Sun, M.L. Gao, et al., Angew. Chem. Int. Ed. 61 (2022) e202206108. doi: 10.1002/anie.202206108

  37. [37]

      S.M. Sajjadinezhad, L. Boivin, K. Bouarab, et al., Coordin. Chem. Rev. 510 (2024) 215794. doi: 10.1016/j.ccr.2024.215794

  38. [38]

      J. Wang, Z. Song, M. He, et al., Nat. Commun. 15 (2024) 2125. doi: 10.1038/s41467-024-46439-w

  39. [39]

      J. Liang, X. Gao, B. Guo, et al., Angew. Chem. Int. Ed. 60 (2021) 12770–12774. doi: 10.1002/anie.202101878

  40. [40]

      H. Wang, J. Yu, S. Wei, et al., Chem. Eng. J. 441 (2022) 136020. doi: 10.1016/j.cej.2022.136020

  41. [41]

      C.P. Wang, Y. Feng, H. Sun, et al., ACS Catal. 11 (2021) 7132–7143. doi: 10.1021/acscatal.1c01447

  42. [42]

      X.L. Yang, Z.Y. Yang, R. Shao, et al., Adv. Mater. 35 (2023) 2304046. doi: 10.1002/adma.202304046

  43. [43]

      S. Jiang, T. Lv, Y. Peng, et al., Adv. Sci. 10 (2023) e2206887. doi: 10.1002/advs.202206887

  44. [44]

      X. Qin, M. Xu, J. Guan, et al., Nat. Energy 9 (2024) 154–162. doi: 10.1038/s41560-023-01418-1

  45. [45]

      W. Sheng, H. Hao, F. Huang, et al., Chem. Eng. J. 430 (2021) 133071.

  46. [46]

      Y. Zhang, X. Ran, H. Fu, et al., Adv. Funct. Mater. 34 (2024) 2404585. doi: 10.1002/adfm.202404585

  47. [47]

      Y. Zhang, H. Liang, Adv. Compos. Hybrid Mater. 7 (2024) 21. doi: 10.1007/s42114-024-00837-2

  48. [48]

      T. Zhang, T. Li, M. Gao, et al., Adv. Energy Mater. 14 (2024) 2400388. doi: 10.1002/aenm.202400388

  49. [49]

      H. Wang, P.H. Cui, J.-X. Shi, et al., Mat. Sci. Semicon. Pro. 97 (2019) 91–100. doi: 10.1016/j.mssp.2019.03.016

  50. [50]

      Q. Wang, W. Ma, J. Qian, et al., Environ. Pollut. 347 (2024) 123707. doi: 10.1016/j.envpol.2024.123707

  51. [51]

      Q. Wang, N. Li, M. Tan, et al., Sep. Purif. Technol. 307 (2023) 143587.

  52. [52]

      L. Wang, K. Zhang, J. Qian, et al., Chemosphere 344 (2023) 140277. doi: 10.1016/j.chemosphere.2023.140277

  53. [53]

      X. Ma, H. Du, M. Tan, et al., Sep. Purif. Technol. 339 (2024) 126644. doi: 10.1016/j.seppur.2024.126644

  54. [54]

      Q. Li, B. Xue, Z. Wang, et al., Sep. Purif. Technol. 338 (2024) 126450. doi: 10.1016/j.seppur.2024.126450

  55. [55]

      M. Deng, J. Guo, X. Ma, et al., Sep. Purif. Technol. 326 (2023) 124786. doi: 10.1016/j.seppur.2023.124786

  56. [56]

      T. Zhou, G. Dong, J. Geng, et al., Appl. Surf. Sci. 655 (2024) 159592. doi: 10.1016/j.apsusc.2024.159592

  57. [57]

      T. Dong, G. Dong, Y. Li, et al., Appl. Catal. B Environ. 349 (2024) 123889. doi: 10.1016/j.apcatb.2024.123889

  58. [58]

      X. Han, W. Chen, M. Xu, et al., Sep. Purif. Technol. 345 (2024) 127397. doi: 10.1016/j.seppur.2024.127397

  59. [59]

      Z. Zheng, F. Han, B. Xing, et al., J. Colloid Interface Sci. 624 (2022) 460–470. doi: 10.1016/j.jcis.2022.05.161

• 

  Figure 1  (a) Powder XRD patterns of the samples. (b) SEM image of S_1.8/NMT-NS. (c) TEM image of S_1.8/NMT-NS. (d) HRTEM image of S_1.8/NMT-NS.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a, b) HAADF imaging. (c) EDX elemental mapping of C, N, Ti, O, Cd, S.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  High-resolution XPS profiles of (a) survey scan. (b, c) Cd 3d and S 2p in S_1.8/NMT-NS and CdS. (d, e) Ti 2p and O 1s in NMT-B, NMT-NS, and S_1.8/NMT-NS. (f) EPR spectra of CdS, NMT-B, NMT-NS, and S_1.8/NMT-NS.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Schematic description of the electronic band structures of NMT-NS and CdS. (b) Photocurrent response. (c) EIS Nyquist plots of S_1.8/NMT-NS, NMT-NS, and NMT-B.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Conversion for BZA and selectivity of N-BZA over NMT-NS with different loading mass of CdS clusters. (b) Cyclic test of the photocatalytic oxidation of benzylamine over S_1.8/NMT-NS. In situ FT-IR spectra of (c) S_1.8/NMT-NS, (d) NMT-NS. (1) After degassing at 150 ℃ for 3 h. (2) Adsorption of benzylamine for 30 min at room temperature (physisorption + chemisorption). (3) Further evacuation of excess benzylamine at 150 ℃ for 15 min under 5 Pa (chemisorption). (4) FT-IR spectra of BZO in KBr.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) EPR spectra of ^•O₂^- generated by CdS, NMT-B, NMT-NS, and S_1.8/NMT-NS with DMPO. (b) Proposed mechanism of photocatalytic selective oxidation coupling of BZA and BZO over S_1.8/NMT-NS under visible light.

  下载: 全尺寸图片 幻灯片
•