English

• With the continuous increase in global energy demand and the intensification of environmental crises, the search for renewable and green fuels becomes one of the most significative and prospective missions of human society. Artificial photocatalysis using solar energy has attracted extensive attention on account of its clean and sustainable advantages, and has been considered as one of the most hopeful approaches to eliminate both the world energy dilemma and pollution of the environment [1-4].

  Among various photoactive materials, cadmium sulfide (CdS) is regarded as a promising semiconductor material for establishing highly efficient artificial photosystems [5-7], because of its relatively narrow bandgap for visible light response and more suitable band edge positions that meet the standards related to thermodynamics for lots of mainstream photocatalytic reactions [8-11]. However, the insufficient photogenerated charge carriers separation and poor photostability inhibit blank CdS from practical application [5].

  To resolve these handicaps, many tactics have been employed to construct CdS-based composite photocatalyst, aimed to harvest more efficient separation of photoinduced charge carriers and obtain higher photoredox-catalyzed stability than blank CdS, such as hybridizing with earth-abundant carbon materials and decorating with appropriate cocatalyst [12-14]. Among various carbon materials, reduced graphene oxide (GR) has unique physicochemical characteristics, such as two-dimensional (2D) planar structure, ultrahigh electron conductivity and mobility, good light transmittance, and has been proven to be an ideal 2D building block to construct semiconductor hybrid materials and separate/transfer photoexcited holes and electrons [15-19], thereby promoting the overall catalytic performance [20-22]. Furthermore, decorating semiconductor components with noble metal cocatalysts is capable of further accelerating the separation of photoexcited charge carriers and serve as active sites for a customized photoredox-catalyzed reaction [23-26].

  Herein, we have reported a ternary GR-CdS-Pd composite photocatalyst by in-situ growing CdS nanoparticles (NPs) on GR nanosheets (NSs) via a one-step solvothermal method and further decorating with Pd NPs via a facile photodeposition method. The synergistic effect of GR NSs and Pd NPs on fostering the separation of charge carriers derived from the bandgap photoexcitation of CdS endows the composite catalyst with distinctly increased photocatalytic activity and good recyclability toward selective conversion of benzyl alcohol (BA) into benzaldehyde (BAD) with concomitant formation of hydrogen (H₂).

  Fig. S1 (Supporting information) schematically illustrates the preparation procedure of reduced graphene oxide (GR)-CdS-Pd composites (unless otherwise noted, the following characterizations are mainly focused on 5% GR-CdS-1% Pd, abbreviated as GR-CdS-Pd, since it shows the optimal catalytic performance according to the subsequent activity test). Firstly, with natural graphite powder as raw material, and adopt an improved Hummers method to synthesize graphene oxide (GO) [13]. Then, the GR-CdS hybrids are obtained through a facile dimethyl sulfoxide (DMSO) solvothermal method, during this process, GO is turned to be GR at the same time. Blank CdS is synthesized through the same process as that for GR-CdS composites without the need for precursor of the GR.

  The morphologies of prepared GO NSs and CdS NPs are investigated by the scanning electron microscope (SEM). The GO NSs depicted in Fig. S2a (Supporting information) are smooth and thin, and Fig. S2b (Supporting information) displays the blank CdS NPs with an average diameter of 200 nm. The SEM characterization of GR-CdS (Fig. 1a) suggests that the 2D GR NSs have adequate interfacial contact with the CdS NPs. Distinctively, the introduction of GR NSs provides a huge platform for in-situ growth of CdS NPs, leading to the average diameter of spherical CdS NPs decreases from 200 nm to about 10 nm. A smaller particle size will result in more exposed atoms on the surface as potential catalytic sites [27]. Furthermore, Pd NPs are crafted onto the GR-CdS-Pd ternary composite through an in-situ photodeposition strategy, and the morphology of GR-CdS-Pd composite is basically unchanged (Fig. 1b). Transmission electron microscope (TEM) characterizations and elemental mapping have been carried out to obtain the microscopic structure information and the elemental distribution. The TEM image (Fig. 1c) of GR-CdS-Pd composite presents the uniform distribution of CdS NPs and Pd NPs on the GR NSs and the intimate interface contact between GR and CdS. Besides, the coexistence of CdS and Pd can be recognized from the high-resolution TEM (HRTEM), the lattice spacing of 0.198 nm is observed in Fig. 1d, which is indexed to the (200) crystalline plane of Pd [28]. And the identified lattice spacing of 0.336 and 0.205 nm can be attributed to the (111) and (220) crystal planes of cubic CdS [29]. Moreover, the elements of C, O, Cd, S and Pd are presented in Fig. S3 (Supporting information) with even distribution, indicating that CdS and Pd NPs carpet the GR NSs uniformly.

  Figure 1

  

  Figure 1.  The SEM images of (a) GR-CdS and (b) GR-CdS-Pd. (c) The TEM image of GR-CdS-Pd composite. (d) The HRTEM image of GR-CdS-Pd composite.

  DownLoad: Full-Size Img PowerPoint

  The chemical element compositions of the as-prepared GR-CdS-Pd composite are analyzed by X-ray photoelectron spectroscopy (XPS). As displayed in the survey XPS spectrum of GR-CdS-Pd composite (Fig. S5a in Supporting information), all elements (Cd, S, O, C and Pd) associated with CdS, GR and Pd are determined, which is consistent with the elemental mapping results. The C 1s spectrum of GR composite is showed in Fig. S5c (Supporting information), which displays three fitted peaks including the C−C/C−H at 284.80 eV, the C−OH bond at 285.40 eV and the HO−C=O bond at 287.90 eV. The C 1s spectrum of GR-CdS composite describes the C−C/C−H at 284.80 eV, the C−OH bond at 285.55 eV and the HO−C=O bond at 288.42 eV (Fig. S5d in Supporting information). The C 1s and O 1s spectra of GR-CdS-Pd are exhibited in Figs. S5e and f (Supporting information). In the C 1s spectrum of GR (Fig. S5e), three fitted peaks related to C atoms represents different functional groups, such as the C−C/C−H at 284.80 eV, the C−OH bond at 285.85 eV and HO−C=O bond at 288.50 eV [29]. In the XPS results of the GR-CdS-Pd hybrids, the two peaks of 405.40 and 412.15 eV depicted in the Cd 3d spectrum (Fig. 2a) are respectively ascribed to Cd 3d_5/2 and Cd 3d_3/2, corresponding to Cd²⁺ [30]. For the S 2p XPS spectra, the binding energies of S 2p_3/2 at 161.85 eV and S 2p_1/2 at 162.90 eV are ascribed to S²⁻ in CdS (Fig. 2b) [31]. The Pd 3d spectrum (Fig. 2c) displays that the binding energies fixed at 341.00 and 335.70 eV are ascribed to Pd 3d_3/2 and Pd 3d_5/2, respectively, indicating the valence state of Pd is 0 [32]. Noticeably, as depicted in Figs. S5c-e and Figs. 2a and b, with the intervention of GR, the characteristic peaks of C 1s and CdS shift toward higher binding energy, which indicate that more intimate interface contact and stronger electron interaction between CdS and GR have been generated [29].

  Figure 2

  

  Figure 2.  XPS spectra of (a) Cd 3d, (b) S 2p and (c) Pd 3d of blank CdS, GR-CdS and GR-CdS-Pd composites, among which GR-CdS-Pd is etched in Ar gas for 30 s. (d) XRD patterns of blank CdS, GR-CdS, CdS-1% Pd and GR-CdS-x% Pd (x = 0.5, 1, 1.5) composites. (e) Raman spectra and (f) FT-IR spectra of blank CdS, blank GO, GR-CdS and GR-CdS-Pd composites.

  DownLoad: Full-Size Img PowerPoint

  The crystal structure properties of the different composites have been investigated by X-ray diffraction (XRD) analysis. The crystalline structure of blank CdS and GR-CdS samples with different weight ratios of GO (denoted as x% GR-CdS, x = 1, 3, 5, or 7, respectively) are displayed in Fig. S6a (Supporting information). As displayed in Fig. 2d, the peaks of blank CdS situate in 2θ values of ca. 26.50°, 43.96°, and 52.13° can be assorted to (111), (220) and (311) facets of cubic phase CdS (JCPDS No.10–0454) [29]. However, GR-CdS-Pd composites with different weight ratios of Pd (the GR-CdS here means 5% GR-CdS, denoted as GR-CdS-x% Pd, x = 0.5, 1, or 1.5, respectively) show nearly three identical diffraction peaks to those of blank CdS NPs, which can be ascribed to the relatively weak diffraction intensity and well dispersion of GR and Pd [33].

  Ultraviolet–visible (UV–vis) diffuse reflectance spectra (DRS) have been used to analyze the optical properties of the GR-CdS-Pd composites. As displayed in Fig. S7a (Supporting information), the well-defined absorption edge of CdS NPs is located at 530 nm, correlating to intrinsic absorption of the CdS NPs [34]. Besides, the optical band gap of the CdS NPs is calculated to be 2.42 eV, which is obtained from the Tauc plot of the transformed Kubelka-Munk function (Fig. S7b in Supporting information) [35]. In addition, the optical absorption performance of GR-CdS-Pd composites are obviously affected by the different content of GR and Pd. With the increase of GR and Pd proportion, the absorption in the visible region (500–800 nm) gradually increases. This can be attributed to the inherent absorption of dark-colored Pd and the electron transfer between GR, Pd and CdS [36,37]. On the grounds of the Mott-Schottky curves (Fig. S8 in Supporting information), the CdS NPs exhibit the trend of n-type semiconductor, of which the flat band potential is situated in −0.68 V vs. Ag/AgCl. For the conduction band (CB) position of n-type semiconductor is closer to its flat band potential, so the CB position of CdS is estimated at −0.48 V vs. NHE. Considering the band gap obtained before, the valance band (VB) position of CdS is converted to be 1.94 V vs. NHE based on the formula E_VB = E_CB + E_g (E_VB, E_CB, and E_g are the energy values of VB, CB, and band gap, respectively). It has been reported that the calculated Fermi level of GR is ca. −0.08 V vs. NHE [38]. Besides, to calculate the Fermi level of Pd NPs, Kelvin probe technology has been employed to determine the surface electronic potential of Pd NPs. As shown in Fig. S9 (Supporting information), the Fermi level of Pd NPs is calculated to be 0.84 V vs. NHE, which is in concert with the reported study [39].

  The Raman spectra and Fourier transformed infrared (FT-IR) spectra over blank GR, blank CdS, GR-CdS and GR-CdS-Pd are shown in Figs. 2e and f. From the Raman spectra, two conspicuous peaks situated at 301.07 and 601.20 cm⁻¹ are discovered, which can be assigned to the E₁ and 2E₁ vibration modes of CdS, respectively [40]. Two adjacent peaks located at 1349.53 and 1607.26 cm⁻¹ are ascribed to the D band and G band of GR, respectively [41]. As depicted in the FT-IR spectra, the peak at 1610 cm⁻¹ is indexed to the C=C stretching vibration, the peak located at 1725 cm⁻¹ is attributed to the C=O bond from COOH group [10]. Compared with the spectrum of blank GO, the significant decrease of graphene oxide functional groups in the composites confirms that the GO is successfully reduced to GR. The FT-IR spectrum of blank CdS sketches a sharp peak at 1266 cm⁻¹ corresponding to the Cd−S bond of CdS. The wide peak in 3464 cm⁻¹ is indexed to the stretching vibration modes of O−H bond of water molecules adsorbed by CdS [27]. And the peak observed in all samples at 1630 cm⁻¹ is also assigned to O−H bond [14]. The aforementioned results clearly confirm the successful synthesis of GR-CdS-Pd ternary composite.

  The photocatalytic activities of blank CdS NPs, GR-CdS and GR-CdS-Pd hybrids have been evaluated via a dual-function photocatalytic redox system gathered the conversion of BA to BAD and H₂ production (Fig. 3a). As shown in Fig. S10b (Supporting information), after UV–vis light irradiation for 2 h, the photocatalytic activity of blank CdS is very low, but when the weight ratio of GR is added to 1%, the conversion of BA is promoted and the H₂ generation is also improved synchronously. The best photocatalytic performance is achieved at 5% GR-CdS; however further increasing the GR weight ratio in GR-CdS brings about a decrease in the photocatalytic performance. The relatively high GR weight ratio in the GR-CdS binary composites would mask the contact area of the semiconductor CdS NPs with light, and the strong adsorption capacity of GR is not conducive to the desorption of the product BAD from the photocatalyst surface [42]. After photodepositing Pd NPs on the optimal binary composite GR-CdS, the photocatalytic activity is further enhanced. Under UV–vis light illumination for 2 h, the photoactivities of blank CdS, GR-CdS and GR-CdS-Pd hybrids with different weight ratio of Pd are sketched in Fig. 3b. There is a volcanic-type relationship between the weight ratio of Pd and the conversion of BA. Specifically, the best photocatalytic activity is obtained over the GR-CdS-1% Pd composite for the conversion reaches the highest values (80.6 µmol for H₂ production and 75.2 µmol for BAD yield), about 16.4-fold and 31.3-fold than those of the blank CdS, respectively. The GR-CdS-1% Pd composite performs an improved BA conversion of 82.4% and a good BAD selectivity of 91.2%, which are comparatively high when compared with other CdS or GR-based materials used in this field (Table S3 in Supporting information). And the photocatalytic performance over 5% GR+CdS and 5% GR+CdS+1% Pd composites made by mechanical mixing is listed in Fig. S11 (Supporting information), which is lower than the ones prepared by solvothermal and photodeposition methods. Besides, the structural information of the liquid products is verified by gas chromatography-mass spectrometry (Fig. S12 in Supporting information).

  Figure 3

  

  Figure 3.  (a) The formula of photocatalytic performance tests of selective BA conversion into BAD paired with H₂ evolution. (b) Photocatalytic performance over blank CdS, GR-CdS, GR-CdS-x% Pd (x = 0.5, 1, 1.5) and CdS-1% Pd composites. (c) Photocatalytic cyclic tests over GR-CdS-Pd composites.

  DownLoad: Full-Size Img PowerPoint

  Recycle tests have been implemented for disclosing the recyclability and sustainability of GR-CdS-Pd composite. As depicted in Fig. 3c, no significant inactivation is observed on GR-CdS-Pd after 7 cycles, demonstrating the good recyclability and sustainability of the ternary composite. The morphology of GR-CdS-Pd composite shows no obvious change after photocatalytic reaction (Fig. S4 in Supporting information). In addition, the XPS and XRD spectra of used GR-CdS-Pd composite show inappreciable change compared with the fresh ones as displayed in Figs. S2a and b and S4b, which prove the structural stability of the materials [35].

  For the purpose of inspecting the universality in conversion of aromatic alcohols over GR-CdS-Pd composite, we have extended the alcohol substrates to 4-methoxybenzyl alcohol, 4-methylbenzyl alcohol, 4-chlorobenzyl alcohol, 4-fluorobenzyl alcohol and phenethyl alcohol under the same reaction conditions. As revealed by the results in Fig. S13 (Supporting information), the GR-CdS-Pd photocatalyst can efficaciously promote the conversion from these different aromatic alcohols to corresponding carbonyl compounds accompany with almost equal yield of H₂. Above results prove the GR-CdS-Pd photocatalyst can simultaneously achieve the photoredox reaction of aromatic alcohols oxidation paired with H₂ production.

  To further confirm the origin of the enhanced photocatalytic activity of GR-CdS-Pd composite compared with that of blank CdS NPs, a series of photoelectrochemical tests have been carried out [9,35]. The electrochemical impedance spectroscopy (EIS) Nyquist curves exhibited in Fig. S14 (Supporting information) reveal that the three samples all show semicircle at high frequencies. The GR-CdS-Pd composite owns the smallest arc radius, implying that it has the fastest interfacial charge carriers transfer when compared to that of GR-CdS binary composite and blank CdS [34,36]. As depicted in Fig. 4a, the enhanced photocurrent density on GR-CdS composite indicates that the introduction of GR can efficiently inhibit the recombination of photoexcited charge carries, while GR-CdS-Pd composite shows the strongest photocurrent response indicating that it can more effectively separate the photoexcited electron-hole pairs [42], which are attributed to the effective synergistic acceleration effect of GR NSs and Pd NPs on boosting the separation of charge carriers originated from the bandgap photoexcitation of CdS [43].

  Figure 4

  

  Figure 4.  (a) Transient photocurrent spectra. (b) EPR spectra of blank CdS, GR-CdS and GR-CdS-Pd suspensions in CH₃CN solution containing 0.1 mmol BA with or without light irradiation. (c) Tentative reaction mechanism of photoredox-catalyzed coupling reaction for BA conversion and H₂ production over GR-CdS-Pd under light irradiation.

  DownLoad: Full-Size Img PowerPoint

  The potential mechanism of photocatalytic reaction is discussed through a sequence of controlled experiments. As shown in Fig. S10c (Supporting information), the addition of hole scavenger triethanolamine (TEOA) in the photoredox-catalyzed reaction system can significantly inhibit the formation of BAD, indicating that the oxidation half reaction of BA is driven by holes [33]. When carbon tetrachloride (CCl₄) is added as an electron scavenger, the yield of BAD increases, but the yield of H₂ decreases significantly, indicating that the reduction of H⁺ is driven by electrons [44]. In addition, the controlled experiments reveal that the production of BAD and H₂ is almost inactivated without light irradiation or photocatalyst, which strongly proves that the reaction is driven by light [45]. Moreover, after adding 5,5-dimethyl-1-pyrrolin-N-oxide (DMPO), the conversion rate of BA decreases sharply, indicating that the active intermediates in the reaction are radicals [46].

  To get more detailed information about the reactive intermediates in the photoredox-catalyzed system, the in-situ electron paramagnetic resonance spectrometer (EPR) trapping measurement has been performed with DMPO as the spin-trapping agent to detect the α-hydroxybenzyl radicals (C_α radicals). As shown in Fig. 4b, no any signal peak is detected in the dark state, indicating that light is a prerequisite for free radical formation [35]. When the reaction system is illuminated, all samples show similar six characteristic signal peaks at α_H = 22 and α_N = 16, which are classified as the DMPO-C_α radical adduct, indicating the presence of radical intermediates in the photoredox-catalyzed system [47,48]. It is worth noting that the GR-CdS-Pd composite exhibits the strongest signal intensity of DMPO-C_α radicals, expressing that a more number of such C_α radicals can be photoexcited in the ternary GR-CdS-Pd photoredox-catalyzed system [44], which is attributed to the prominently improved light collection capability of GR-CdS-Pd compared with that of blank CdS.

  In light of the above discussion, a plausible mechanism of photocatalytic BA conversion to BAD coupled with H₂ production over GR-CdS-Pd is put forward as the following. As depicted in Fig. 4c, under the illumination of UV–vis light, the electrons in CdS NPs are excited and then transferred to the CB, leaving holes in the VB of CdS. Subsequently, owing to the Fermi level of GR is lower than the CB edge of CdS, the photoinduced electrons can flow from the CB of the CdS NPs to GR NSs based on the suitable band alignment between them [49]. Meanwhile, the Pd NPs capture these photoexcited electrons directly from the CdS NPs or indirectly from the GR layers [50]. Furthermore, the accumulated photoinduced holes in the VB of CdS NPs first oxidize the C–H bond of the BA, producing ^•CH(OH)pH species and protons. Then, the holes further oxidize the ^•CH(OH)pH species to generate BAD or C–C coupled products. Consequently, the protons originated from BA are reduced by electrons to yield H₂.

  In summary, a ternary GR-CdS-Pd composite has been constructed for efficient coupling photoredox reaction of selective dehydrogenation of BA into BAD and H₂ evolution, which presents much improved photoactivity than blank CdS. Complementary characterizations suggest that the utilization of GR is not only conducive to the separation and migration of photoinduced charge carriers, but also beneficial to improve the resistance to photocorrosion of CdS. The deposited Pd NPs co-catalyst enable more efficient charge separation of ternary GR-CdS-Pd composite and facilitate the formation of intermediate products, therefore significantly enhancing the photocatalytic activity. It is anticipated that this work could offer new avenues to the judicious design of high-performance CdS-based composite photocatalysts for photoredox-catalyzed production of clean solar fuels and selective organic synthesis of biomass-derived intermediates.

  Declaration of competing interest

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

  Acknowledgments

  This work was supported by the Natural Science Foundation of China (Nos. 22172030, 22072023, 21872029, U1463204 and 21173045), the Program for National Science and Technology Innovation Leading Talents (No. 00387072), the Program for Leading Talents of Fujian Universities, the 1st Program of Fujian Province for Top Creative Young Talents, and the Natural Science Foundation of Fujian Province (Nos. 2017J07002 and 2019J01631).

  
  
• 1. [1]

     P.V. Kamat, S. Jin, ACS Energy Lett. 3 (2018) 622–623. doi: 10.1021/acsenergylett.8b00196

  2. [2]

     H. Kisch, Acc. Chem. Res. 50 (2017) 1002–1010. doi: 10.1021/acs.accounts.7b00023

  3. [3]

     M.Y. Qi, M. Conte, M. Anpo, Z.R. Tang, Y.J. Xu, Chem. Rev. 121 (2021) 13051–13085. doi: 10.1021/acs.chemrev.1c00197

  4. [4]

     X. Liu, C. Huang, B. Ouyang, et al., Chem.: Eur. J. 28 (2022) e202201034.

  5. [5]

     Q. Li, B. Guo, J. Yu, et al., J. Am. Chem. Soc. 133 (2011) 10878–10884. doi: 10.1021/ja2025454

  6. [6]

     K. Zhang, G. Lu, Z. Xi, et al., Chin. Chem. Lett. 32 (2021) 2207–2211. doi: 10.1016/j.cclet.2020.12.021

  7. [7]

     H.Q. Yang, Q.Q. Chen, F. Liu, R. Shi, Y. Chen, Chin. Chem. Lett. 32 (2021) 676–680. doi: 10.1016/j.cclet.2020.06.022

  8. [8]

     L. Cheng, Q.J. Xiang, Y.L. Liao, H.W. Zhang, Energy Environ. Sci. 11 (2018) 1362–1391. doi: 10.1039/C7EE03640J

  9. [9]

     Y.H. Li, F. Zhang, Y. Chen, J.Y. Li, Y.J. Xu, Green Chem. 22 (2020) 163–169. doi: 10.1039/c9gc03332g

  10. [10]

      M. Darwish, A. Mohammadi, N. Assi, J. Hazard. Mater. 320 (2016) 304–314. doi: 10.1016/j.jhazmat.2016.08.043

  11. [11]

      P. Zhang, D.Y. Luan, X.W. Lou, Adv. Mater. 32 (2020) 2004561. doi: 10.1002/adma.202004561

  12. [12]

      M.Q. Yang, Y. Zhang, N. Zhang, Z.R. Tang, Y.J. Xu, Sci Rep. 3 (2013) 3314. doi: 10.1038/srep03314

  13. [13]

      N. Zhang, M.Q. Yang, Z.R. Tang, Y.J. Xu, J. Catal. 303 (2013) 60–69. doi: 10.1016/j.jcat.2013.02.026

  14. [14]

      C. Zhang, Y. Lu, Q. Jiang, J. Hu, Nanotechnology 27 (2016) 355402. doi: 10.1088/0957-4484/27/35/355402

  15. [15]

      Q. Li, X. Li, S. Wageh, A.A. Al-Ghamdi, J. Yu, Adv. Energy Mater. 5 (2015) 1500010. doi: 10.1002/aenm.201500010

  16. [16]

      X.Z. Li, B. Weng, N. Zhang, Y.J. Xu, RSC Adv. 4 (2014) 64484–64493. doi: 10.1039/C4RA13764G

  17. [17]

      S. Liu, B. Weng, Z.R. Tang, Y.J. Xu, Nanoscale 7 (2015) 861–866. doi: 10.1039/C4NR04229H

  18. [18]

      Q. Quan, X. Lin, N. Zhang, Y.J. Xu, Nanoscale 9 (2017) 2398–2416. doi: 10.1039/C6NR09439B

  19. [19]

      B. Weng, J. Wu, N. Zhang, Y.J. Xu, Langmuir 30 (2014) 5574–5584. doi: 10.1021/la4048566

  20. [20]

      X. Xie, N. Zhang, Z.R. Tang, Y.J. Xu, Chem. Sci. 9 (2018) 8876–8882. doi: 10.1039/c8sc03679a

  21. [21]

      M.Q. Yang, C. Han, N. Zhang, Y.J. Xu, Nanoscale 7 (2015) 18062–18070. doi: 10.1039/C5NR05143F

  22. [22]

      M.Q. Yang, N. Zhang, M. Pagliaro, Y.J. Xu, Chem. Soc. Rev. 43 (2014) 8240–8254. doi: 10.1039/C4CS00213J

  23. [23]

      A. Currao, V.R. Reddy, G. Calzaferri, Chemphyschem 5 (2004) 720–724. doi: 10.1002/cphc.200301211

  24. [24]

      T. Yan, H. Zhang, Y. Liu, et al., RSC Adv. 4 (2014) 37220–37230. doi: 10.1039/C4RA06254J

  25. [25]

      T. Jiang, C. Jia, L. Zhang, et al., Nanoscale 7 (2015) 209–217. doi: 10.1039/C4NR05905K

  26. [26]

      A. Li, E. Kan, S.M. Chen, et al., Small 18 (2022) 2200073. doi: 10.1002/smll.202200073

  27. [27]

      S.C. Tudu, M. Zubko, J. Kusz, A. Bhattacharjee, Appl. Phys. A: Mater. Sci. Process. 127 (2021) 85. doi: 10.1007/s00339-020-04245-3

  28. [28]

      W. Liu, P. Rodriguez, L. Borchardt, et al., Angew. Chem., Int. Ed. 52 (2013) 9849–9852. doi: 10.1002/anie.201303109

  29. [29]

      M.Q. Yang, C. Han, Y.J. Xu, J. Phys. Chem. C 119 (2015) 27234–27246. doi: 10.1021/acs.jpcc.5b08016

  30. [30]

      Z.B. Yu, Y.P. Xie, G. Liu, et al., J. Mater. Chem. A 1 (2013) 2773–2776. doi: 10.1039/c3ta01476b

  31. [31]

      L.J. Zhang, S. Li, B.K. Liu, D.J. Wang, T.F. Xie, ACS Catal. 4 (2014) 3724–3729. doi: 10.1021/cs500794j

  32. [32]

      K. Ament, N. Kowitsch, D. Hou, et al., Angew. Chem. Int. Ed. Engl. 60 (2021) 5890–5897. doi: 10.1002/anie.202015138

  33. [33]

      J.Y. Li, Y.H. Li, F. Zhang, Z.R. Tang, Y.J. Xu, Appl. Catal. B: Environ 269 (2020) 118783. doi: 10.1016/j.apcatb.2020.118783

  34. [34]

      C. Han, Z. Chen, N. Zhang, J.C. Colmenares, Y.J. Xu, Adv. Funct. Mater. 25 (2015) 221–229. doi: 10.1002/adfm.201402443

  35. [35]

      C.L. Tan, M.Y. Qi, Z.R. Tang, Y.J. Xu, Appl. Catal. B 298 (2021) 120541. doi: 10.1016/j.apcatb.2021.120541

  36. [36]

      N. Zhang, M.Q. Yang, Z.R. Tang, Y.J. Xu, ACS Nano 8 (2014) 623–633. doi: 10.1021/nn405242t

  37. [37]

      B. Krishnakumar, S. Kumar, J.M. Gil, et al., J. Mol. Struct. 1153 (2018) 346–352. doi: 10.1016/j.molstruc.2017.09.120

  38. [38]

      Y.H. Zhang, Z. Chen, S.Q. Liu, Y.J. Xu, Appl. Catal. B: Environ 140-141 (2013) 598–607. doi: 10.1016/j.apcatb.2013.04.059

  39. [39]

      L. Chen, S. Mao, P. Wang, et al., Adv. Opt. Mater. 9 (2021) 2001505. doi: 10.1002/adom.202001505

  40. [40]

      B. Cao, Y. Jiang, C. Wang, et al., Adv. Funct. Mater. 17 (2007) 1501–1506. doi: 10.1002/adfm.200601179

  41. [41]

      V. Nagyte, D.J. Kelly, A. Felten, et al., Nano Lett. 20 (2020) 3411–3419. doi: 10.1021/acs.nanolett.0c00332

  42. [42]

      N. Zhang, Y.H. Zhang, X.Y. Pan, et al., J. Phys. Chem. C 115 (2011) 23501–23511. doi: 10.1021/jp208661n

  43. [43]

      N. Zhang, S.Q. Liu, X.Q. Fu, Y.J. Xu, J. Mater. Chem. 22 (2012) 5042–5052. doi: 10.1039/c2jm15009c

  44. [44]

      M.Y. Qi, Y.H. Li, F. Zhang, et al., ACS Catal. 10 (2020) 3194–3202. doi: 10.1021/acscatal.9b05420

  45. [45]

      J.Y. Li, L. Yuan, S.H. Li, Z.R. Tang, Y.J. Xu, J. Mater. Chem. A 7 (2019) 8676–8689. doi: 10.1039/c8ta12427b

  46. [46]

      Q. Lin, Y.H. Li, M.Y. Qi, et al., Appl. Catal. B: Environ 271 (2020) 118946. doi: 10.1016/j.apcatb.2020.118946

  47. [47]

      M.Y. Qi, Y.H. Li, M. Anpo, Z.R. Tang, Y.J. Xu, ACS Catal. 10 (2020) 14327–14335. doi: 10.1021/acscatal.0c04237

  48. [48]

      Q. Guo, F. Liang, X.B. Li, et al., Chem 5 (2019) 2605–2616. doi: 10.1016/j.chempr.2019.06.019

  49. [49]

      N. Zhang, Y.H. Zhang, M.Q. Yang, Z.R. Tang, Y.J. Xu, J. Catal. 299 (2013) 210–221. doi: 10.1016/j.jcat.2012.11.021

  50. [50]

      B. Weng, Q. Quan, Y.J. Xu, J. Mater. Chem. A 4 (2016) 18366–18377. doi: 10.1039/C6TA07853B

• 

  Figure 1  The SEM images of (a) GR-CdS and (b) GR-CdS-Pd. (c) The TEM image of GR-CdS-Pd composite. (d) The HRTEM image of GR-CdS-Pd composite.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  XPS spectra of (a) Cd 3d, (b) S 2p and (c) Pd 3d of blank CdS, GR-CdS and GR-CdS-Pd composites, among which GR-CdS-Pd is etched in Ar gas for 30 s. (d) XRD patterns of blank CdS, GR-CdS, CdS-1% Pd and GR-CdS-x% Pd (x = 0.5, 1, 1.5) composites. (e) Raman spectra and (f) FT-IR spectra of blank CdS, blank GO, GR-CdS and GR-CdS-Pd composites.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) The formula of photocatalytic performance tests of selective BA conversion into BAD paired with H₂ evolution. (b) Photocatalytic performance over blank CdS, GR-CdS, GR-CdS-x% Pd (x = 0.5, 1, 1.5) and CdS-1% Pd composites. (c) Photocatalytic cyclic tests over GR-CdS-Pd composites.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Transient photocurrent spectra. (b) EPR spectra of blank CdS, GR-CdS and GR-CdS-Pd suspensions in CH₃CN solution containing 0.1 mmol BA with or without light irradiation. (c) Tentative reaction mechanism of photoredox-catalyzed coupling reaction for BA conversion and H₂ production over GR-CdS-Pd under light irradiation.

  下载: 全尺寸图片 幻灯片
•