Revealing Spatially Resolved Exciton Dynamics of Dual Active Centers for Dehydrogenation Oxidation of Benzyl Alcohol Coupled by Hydrogen Evolution

ABSTRACT

Photooxidation of biomass alcohols coupling H₂ evolution has received increasing attention due to its higher economy. However, the insufficient photoexciton utilization has become a bottleneck restricting its practical application, thereby precisely regulating the microstructure of photocatalysts is crucial for achieving efficient photoactivity. Strategically introducing functional active centers has been proven to be a valid strategy for optimizing the exciton dynamics of photocatalysts. Herein, the cobalt (Co) clusters and single‐atom Co─O sites were co‐decorated on the heterostructured MoS₂/CdS composite to integrate the merits of heterojunction and non‐precious metal cocatalysis, which was demonstrated by the x‐ray absorption fine structure spectroscopy and spherical aberration‐corrected transmission electron microscope techniques. Remarkable collaborative photoactivity for benzyl alcohol (BA) oxidation (181.8 mmol_BAD·g⁻¹·h⁻¹) coupling H₂ evolution (64.6 mmol_H2·g⁻¹·h⁻¹) was achieved under artificial sunlight. Femtosecond transient absorption spectroscopy and Kelvin probe force microscopy techniques proved the improved spatial separation/transfer kinetics and hindered recombination kinetics of photoexcitons. Based on density functional theory, the Co clusters act as the oxidation sites for BA, and the Co─O sites serve as the H₂ evolution reduction centers, enabling the synergistically enhanced REDOX half‐reactions due to more effective spatial separation of photoexcitons. This study proposes a strategic catalyst construction insight for highly efficient collaborative REDOX photocatalysis.

The Co clusters and single‐atom Co─O sites were co‐decorated on the heterostructured MoS₂/CdS composite to integrate the merits of heterojunction and non‐precious metal cocatalysis. The former and the latter respectively act as the oxidation and reduction centers to achieve the remarkable collaborative photoactivity for benzyl alcohol oxidation coupling H₂ evolution under artificial sunlight due to effective spatial separation of photoexcitons.

1. Introduction

Since Fujishima and Honda jointly discovered the photoactivity of superhydrophilic TiO₂ in 1972 [1], the photocatalysis technology based on semiconductor micro‐nano structures has become the research focus in the field of heterogeneous catalysis. It is well known that the essence of a chemical reaction is a process of gaining and losing electrons. Semiconductor‐based micro‐nano photocatalysts can undergo electron transitions under photoirradiation and simultaneously generate photogenerated h⁺, obtaining a higher surface chemical potential, thereby producing chemical activity to drive the progress of related chemical reactions. Such as organic matter degradation and transformation [2], CO₂ reduction or polycarbonization (C₂₊) [3], water splitting [4], hydrogen peroxide (H₂O₂) synthesis [5], nitrogen fixation [6], transformation and recovery of metal ions [7], and so forth. Recently, light‐driven collaborative REDOX process has been highly recognized due to its higher photoexciton utilization efficiency and economic value. Among them, the dehydrogenation oxidation of biomass alcohols coupled by H₂ evolution has received increasing attention, which can achieve higher H₂ evolution efficiency without the aid of sacrificial agents and realize the value‐added conversion of biomass resources, showing immeasurable practical research value [8, 9]. However, photocatalytic reactions are typical dynamic processes involving exciton separation, transfer, recombination and REDOX reactions at surface active sites under photoactivation, among which the separation and recombination processes of photoexcitons are the most discussed classical competitive relationship in photocatalytic research. Generally, higher photoexciton utilization efficiency and photoactivity can be enabled by strengthening the transfer kinetics and weakening the recombination kinetics [10, 11].

In fact, the photoexciton recombination behavior is impossible to be completely eliminated. We can only try to improve the utilization efficiency of photoexcitons by intervening in the dynamic balance of their transfer and recombination processes. However, the related photocatalytic mechanisms are extremely complex, and the difficulty of mechanism investigation becomes even greater as the catalyst components increase. Therefore, exploring the photocatalytic mechanism has always been a highly concerning and perplexing scientific challenge in this field [12, 13]. For instance, Grätzel et al. pioneered the research field of energy and charge transfer in mesosightseeing photocatalytic systems [14], and Li et al. demonstrated the spatial separation and transfer phenomenon of photocarriers on micro‐nano photocatalysts [15]. Furthermore, Ye et al. innovatively proposed a high‐entropy photocatalytic mechanism [16], and Yu's team successively proposed the Z‐scheme and S‐scheme heterostructured mechanisms [17]. Subsequently, numerous scholars, represented by the above scientists, have continuously expanded the cognitive boundaries of photocatalytic mechanisms through interdisciplinary collaboration, providing the theoretical basis for highly efficient photocatalysis. However, there are still many deep‐seated mechanismatics that are difficult to explain clearly in photocatalysis research, so more pioneering studies are needed to support the development of this technology.

Strategically introducing functional active centers has been proven to be a valid strategy for optimizing the exciton dynamics and improving the catalytic activities of semiconductor photocatalysts [18, 19]. Especially, cobalt (Co)‐based cocatalyst has been proven to be a potential candidate for improving the activity of semiconductor photocatalysts due to its non‐precious metal character and highly efficient active site capacity [20, 21]. Herein, this study strategically co‐decorated Co clusters and single‐atom Co─O sites on the heterostructured MoS₂/CdS (MCS) photocatalyst to integrate the merits of heterojunction and non‐precious metal cocatalysis, and a remarkable collaborative photoactivity for benzyl alcohol (BA) oxidation coupled by H₂ evolution was achieved under artificial sunlight. X‐ray absorption fine structure (XAFS) spectroscopy and spherical aberration transmission electron microscope (STEM) techniques were adopted to demonstrate the successful co‐decoration of Co clusters and single‐atom Co─O sites, and the exciton dynamics were investigated to confirm the photoelectric enhancement mechanism by the femtosecond transient absorption (fs‐TA) spectroscopy and Kelvin probe force microscopy (KPFM) techniques. Finally, the enhanced REDOX mechanism based on dual active centers was revealed through density functional theory (DFT) calculations. This study proposes a strategic catalyst construction insight for collaborative REDOX photocatalysis with high‐efficiency utilization of photoexcitons.

2. Results and Discussion

2.1. Preparation and Topological Characterizations of MCS‐Co Composite Photocatalyst

As displayed in Figure 1a, the pre‐synthesized CdS NPs was added to the precursor system of MoS₂ to hydrothermally prepare the heterostructured MCS composite, and the interfacial interaction between MoS₂ and CdS was precisely regulated by adjusting the addition amount of CdS NPs. Subsequently, the Co species was decorated on the optimized MCS_40% composite to integrate the merits of heterojunction and non‐precious metal cocatalysis, and the MCS‐Co composite photocatalyst was optimized through the fine modulation of surface adsorption and wet chemical reduction processes. The crystal structure of the above‐prepared samples were determined by powder x‐ray diffraction (XRD) characterization. The XRD patterns in Figure 1b show that the pure MoS₂ presents a coexisting crystal feature of 1T phase and 2H phase, while the pure CdS shows a single hexagonal crystal phase, which are consistent with our previous researches [22, 23]. The MCS_40% composite simultaneously exhibits all the diffraction peaks belonging to MoS₂ and CdS, and the diffraction intensity of CdS also enhances with the increase of its content (10%–50%, Figure S1a), indicating that the proportion of CdS in the MCS composite can be effectively adjusted by changing its addition amount during the preparation process. Moreover, the partial magnification of Figure 1b reveals that the diffraction peak of the 1T‐phase MoS₂(004) plane in MCS composite shows a slight blue shift compared to pure MoS₂, which can be attributed to the crystal plane coupling between MoS₂ and CdS [24]. After growing the Co species on the surface, the crystal feature belonging to metallic Co (PDF#15‐0806) cannot be observed even after increasing the growth amount (Figures 1b and S1b). In fact, the actual capacity of Co element in the optimized MCS‐Co_11% composite photocatalyst was determined to be 1.04 wt% by the inductively coupled plasma mass spectrometry (ICP‐MS) technique.

FIGURE 1.

Preparation and topological characterizations: (a) Schematic diagram of catalyst preparation. (b) XRD patterns of MoS₂, CdS, MCS_40%, and MCS‐Co_11%. (c) SEM image, (d) TEM image, and (e) HAADF image and EDX mappings of MCS_40%. (f) SEM, (g) TEM, (h) HAADF‐ACTEM, and (i) HAADF‐Co‐EDX images of MCS‐Co_11%. N₂ adsorption‐desorption isotherms of MoS₂ (j), MCS_40% (k), and MCS‐Co_11% (l). The inset image is the SEM image of the corresponding sample.

The morphology and crystal structure of the samples were observed by electron microscope techniques. As shown in the SEM and TEM images of Figure S2a,b, the pure MoS₂ is a flower‐spherical morphology with a size of approximately 1.5–2.5 µm. In contrast, the SEM and TEM images of the MCS_40% composite clearly show that a large number of CdS NPs with a size of 20∼30 nm are tightly bound to the surface of MoS₂ nanosheets (NSs) (Figure 1c,d), and the MoS₂/CdS heterointerface is clearly visible. Moreover, the energy dispersive x‐ray (EDX) maps in Figure 1e show uniformly distributed Cd element, which is obviously conducive to the formation of the tightly heterostructured MoS₂/CdS interface. The N₂ adsorption‐desorption isotherms in Figure 1j,k show that the MCS_40% composite has a BET specific surface area (S _BET = 14.72 m²/g) 14 times that of MoS₂ (1.08 m²/g), which indicates that the tightly accumulated CdS NPs on MoS₂ microspheres creates a typical mesoporous structure [25, 26]. Furthermore, it can be determined from the SEM images of Figure S3a–h that the depositing amount of CdS NPs on MoS₂ microspheres gradually increases with the increase of its addition amount from 30% to 50%, which supports the conclusion of XRD analysis. Subsequently, it can be determined from the SEM and TEM images of the MCS‐Co_11% composite photocatalyst in Figure 1f,g that the Co decoration does not change the morphology of MCS_40% composite, and only the lattice fringes of d = 0.331 nm belonging to the hexagonal CdS(002) plane can be observed from the high‐angle annular dark field‐spherical aberration‐corrected TEM (HAADF‐ACTEM) image in Figure 1h [27]. No morphological feature and lattice fringes belonging to the Co species can be observed, and its S _BET is very close to that of the MCS_40% composite (14.44 m²/g, Figure 1l). However, its EDX map captured in the HAADF‐ACTEM pattern reveals the uniformly distributed Co element in the nano‐size region of MCS‐Co_11% surface (Figure 1i), which proves the successful Co decoration.

The chemical states of the surface constituent elements and the interfacial interactions of the optimized composites were analyzed by x‐ray photoelectron spectroscopy (XPS) technique. The XPS survey spectra in Figure S4 show that obvious Mo 3d, Cd 3d, and S 2p characteristic peaks were detected on the surfaces of MCS_40% and MCS‐Co_11% composites. However, no obvious characteristic peak corresponding to Co 2p was detected on MCS‐Co_11% surface, and this is similar to the result of XRD characterization, indicating the low content of Co species. Subsequently, high‐resolution fittings were performed on the above features, and calibration was carried out according to the C 1s standard peak at 284.8 eV. Among them, the Mo 3d (Figure 2a) and S 2p fine spectra (Figure 2b) show the similar results to our previous study, and the chemical characteristics of both 1T‐phase and 2H‐phase MoS₂ can be observed simultaneously [22]. Moreover, the Cd 3d fine spectra (Figure 2c) also show similar features to previous study, and the typical 5/2 and 3/2 spin states of Cd²⁺ in CdS can be observed [23]. Noticeably, compared with the pure samples, the characteristic peaks of Mo 3d, S 2p, and Cd 3d chemical states of MCS_40% composite show slight shifts to varying degrees towards lower and higher binding energy directions, accompanied by an increase and decrease in peak area, which is attributed to the strong electronic interaction at MoS₂/CdS heterointerface [28]. After surface decoration of Co species, the above‐mentioned characteristic peaks once again show slight shifts to varying degrees towards higher binding energy directions, which is a result of the strong electronic interaction between Co species and the MCS_40% composite. XAFS spectroscopy was employed to accurately determine the local coordination state of the Co species on MCS‐Co_11% surface. The standardized Co K‐edge x‐ray absorption near‐edge structure (XANES) spectra in Figure 2d shows that the absorption edge of the MCS‐Co_11% composite lies between the Co‐foil and CoO, determining the average chemical state between 0 and +2 valence, and revealing the coexistence of elemental and oxidized Co species [20]. The Fourier transform extended x‐ray absorption fine structure (FT‐EXAFS) spectra in Figure 2e shows that the MCS‐Co_11% composite photocatalyst presents a distinct scattering signal peak near 1.5 Å, and its peak position and shape are highly consistent with that of the Co─O bond of Co‐based oxides. Meanwhile, the peak position of the scattering signal near 2.2 Å is similar to the scattering path of the Co─Co bond in Co‐foil, but shows a much lower and wider absorption peak. The above results prove that the Co species on MCS‐Co_11% surface coexists in the forms of monodispersed Co─O coordination structure and extremely small clusters containing 2∼3 Co atoms, without long‐range Co─Co aggregation state [21]. In the wavelet transform EXAFS (WT‐EXAFS) map of Figure 2f, the Co─O signal in the low region and the Co─Co signal in the middle and high regions can be simultaneously observed from the MCS‐Co_11% composite photocatalyst, which are highly consistent with the characteristics of the Co‐based standard references (Figure 2g–i). Moreover, its Co─Co signal is significantly weaker than that of the Co‐foil reference and presents a more dispersed state, which indicates that the atomic number weight of Co─Co scattering in the MCS‐Co_11% composite photocatalyst is weaker, and proves the co‐decoration of single‐atom Co─O and Co cluster.

FIGURE 2.

Characterizations of electronic structures and local coordination environment: High‐resolution XPS spectra of MoS₂, CdS, MCS_40%, and MCS‐Co_11%: (a) Mo 3d, (b) S 2p, and (c) Cd 3d. (d) Standardized Co K‐edge XANES spectra, (e) FT‐EXAFS spectra, and (f–i) WT‐EXAFS maps of MCS‐Co_11% and Co‐based standard references (Co‐foil, CoO, and Co₂O₃).

2.2. Collaborative Photoactivity for Artificial Sunlight‐Driven BA Oxidation Coupled by H₂ Evolution

To fully utilize oxidizing and reducing photoexcitons and achieve more sustainable hydrogen production, this study collaboratively conducted BA dehydrogenation oxidation and H₂ evolution under artificial sunlight‐irradiation. As shown in Figure 3a, the MCS composites exhibit significantly enhanced collaborative photoactivity compared to pure MoS₂ and CdS. With the increase of CdS content (10%–40%), both the generation rates (R) of BAD (29.1 → 87.6 mmol_BAD·g⁻¹·h⁻¹) and H₂ (10.3 → 27.3 mmol_H2·g⁻¹·h⁻¹) show an increasing trend. Subsequently, further increasing the CdS content leads to the decreased R _BAD (87.6 → 59.7 mmol_BAD·g⁻¹·h⁻¹) and R _H2 (27.3 → 20.1 mmol_H2·g⁻¹·h⁻¹). Therefore, the MCS_40% composite can be determined to possess the highest collaborative photoactivity, and its R_BAD and R_H2 are 87.6 mmol_BAD·g⁻¹·h⁻¹ and 27.3 mmol_H2·g⁻¹·h⁻¹, respectively, reaching 146 (BAD)/54.6 (H₂) times that of MoS₂ and 19.5 (BAD)/12.4 (H₂) times that of CdS. Specifically, in the MCS composite, the number of heterostructured MoS₂/CdS interface increases with the increase of CdS content (e.g., 10%–40%), which is conducive to the charge conduction at the heterointerface, resulting in the improved photoexciton utilization efficiency and enhanced collaborative photoactivity. Excessive CdS (e.g., 50%) will cause severe agglomeration in the composite, which is not conducive to the synergistic electronic interaction of the heterostructured MoS₂/CdS interface, resulting in a decreased collaborative photoactivity. Although the heterostructured MoS₂/CdS interface can result in a collaborative photoactivity far superior to that of pure samples, further improvement is still desired due to the absence of effective active sites. To this end, near‐atomic‐level Co species was decorated on the MCS_40% surface. As shown in Figure 3b, with the amount increase of Co species (8%–11%), the R_BAD (96.6 → 181.8 mmol_BAD·g⁻¹·h⁻¹) and R _H2 (36.2 → 64.6 mmol_H2·g⁻¹·h⁻¹) are increased accordingly. However, further amount increase of Co species leads to the decreased R _BAD (181.8 → 89.1 mmol_BAD·g⁻¹·h⁻¹) and R _H2 (64.6 → 30.0 mmol_H2·g⁻¹·h⁻¹). Accordingly, the MCS‐Co_11% composite photocatalyst can be determined to possess the highest collaborative photoactivity, and its R _BAD (181.8 mmol_BAD·g⁻¹·h⁻¹) and R _H2 (64.6 mmol_H2·g⁻¹·h⁻¹) even reach 303_(BAD)/129.2_(H2) times that of MoS₂ and 40.4_(BAD)/29.4_(H2) times that of CdS. The above material structure characterizations have demonstrated the co‐decoration of single‐atom Co─O and Co cluster on MCS‐Co_11% surface. When its amount increases in a lower range (e.g., 8%–11%), the number of effective active centers increases accordingly, resulting in the gradually enhanced collaborative photoactivity. While excessive decoration will lead to the serious aggregation of Co species, destroying the advantage of near‐atomic‐level active centers and reducing the collaborative photoactivity.

FIGURE 3.

Collaborative photoactivity for BA oxidation coupled by H₂ evolution: (a) Photoactivities of MoS₂, CdS, and MCS composites with different CdS contents (10%–50%). (b) Photoactivities of MCS‐Co composites with different amounts of Co species (8%–12%). (c) Photoactivities and (d) AQE of MCS‐Co_11% composite photocatalyst under different wavelengths of irradiated light (380, 420, 500, 550 nm). (e) Photoactivity comparation with the catalysts reported in literature. (f) Cyclic photoactivities for 10 cycles. The inset images show the digital photos of the flexible P/MCS‐Co composite membrane before and after the cycle.

The photoactivity is highly dependent on the wavelength and intensity of the irradiated light, thereby the collaborative photoactivities of the MCS‐Co_11% composite photocatalyst under different wavelengths of irradiated light were evaluated. As shown in Figure 3c, under artificial sunlight‐illumination, this composite photocatalyst exhibits much higher photoactivity than that under monochromatic illumination, which is attributed to the full harvesting to different wavelengths of light in artificial sunlight. Further processing the irradiated light with bandpass filters (T380, T420, T500, T550), the significantly reduced collaborative photoactivity can be attributed to the decreased intensity of irradiated light due to the filtering out of the majority of photon energy, resulting in insufficient activation of the MCS‐Co_11% composite photocatalyst following a generation of low‐density photoexcitons. Taking into account the factors of wavelength and light intensity comprehensively, it can be seen from Figure 3d that the apparent quantum efficiencies (AQE) of the MCS‐Co_11% composite photocatalyst under different wavelengths of irradiated light (10.4% (380 nm), 10.3% (420 nm), 4.3% (500 nm), 0.01% (550 nm)) basically conforms to the intensity distribution trend of its light absorption, demonstrating the light‐driven collaborative catalytic ability over a wide wavelength range. The photoactivity is significantly superior to that of the recently reported related catalytic systems (Figure 3e) [29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40]. Moreover, under the stabilization of PVDF‐networked membrane, the MCS‐Co_11% composite photocatalyst maintains highly stable collaborative photoactivity during 10 cycles (Figure 3f), and the increased photoactivity is attributed to the enlarged interstitial space of the membrane fibers, which enables the active catalyst to be exposed more fully [41].

2.3. Optical and Photoelectric Properties

Endowing photocatalytic materials with excellent broadband light‐harvesting capability is crucial for enhancing their light energy utilization efficiency. As shown in the UV‐Vis absorption spectra in Figure S5a, due to the interfacial coupling between MoS₂ and CdS, the MCS_40% composite possesses significantly enhanced light absorption capacity compared to the pure samples and simultaneously exhibits the typical semiconductor absorption characteristics similar to that of CdS NPs (Figure S5b), suggesting the significantly improved broadband‐light harvesting capacity. After co‐decorating the single‐atom Co─O and Co cluster, its light harvesting capacity was further enhanced so much, and its maximum absorption edge was redshifted to 774 from 638 nm, revealing more outstanding broadband light response capability. Noticeably, although the Co species decoration did not significantly change the morphology and structure of the MCS_40% composite, a remarkable improvement was presented on the light‐harvesting capacity. The photoelectric performance of photocatalytic materials is highly determined by the exciton dynamics under photoactivation, so it can be used to indirectly insight the exciton separation and transfer characters of photocatalytic materials. Under photoactivation, an extremely small arc radius of pure MoS₂ can be observed from the Nyquist plots in Figure S6 due to its metal‐like 1T phase, which reveals a very low charge transfer resistance (R _c) at the electrode–electrolyte interface [22]. However, the significantly increased R _c of the MCS_40% composite is attributed to the surface deposition of semiconductor‐like CdS NPs. Encouragingly, the MCS‐Co_11% composite photocatalyst shows a significantly weakened R _c than that of the MCS_40% composite, which is attributed to the synergistic electronic interaction of surface Co species [42]. Furthermore, the exciton separation ability under photoirradiation was evaluated according to the transient photocurrent (TRPC) curves in Figure S7. The good broadband light absorption capacity of pure MoS₂ does not lead to a considerable TRPC intensity (0.12 µA·cm⁻²) under artificial sunlight‐irradiation, but the MCS_40% composite exhibits a significantly enhanced TRPC intensity (0.53 µA·cm⁻²), suggesting higher exciton separation efficiency under photoirradiation due to the interfacial coupling of the MoS₂/CdS heterointerface. Comparatively, the MCS‐Co_11% composite photocatalyst exhibits a further enhanced TRPC intensity (2.21 µA·cm⁻²), reaching 6.3 times that of MCS_40% composite and 18.4 times that of pure MoS₂, respectively. This proves the effectiveness of the co‐decoration of single‐atom Co─O sites and Co clusters.

2.4. Spatially Resolved Exciton Dynamics Under Photoactivation

Investigating the spatially resolved exciton dynamics under photoactivation is crucial for establishing a more accurate structure‐activity relationship [43]. To deeply explore the exciton separation kinetics of the MCS_40% and MCS‐Co_11% composites under photoactivation, the surface photovoltage microscopic (SPVM) images shown in Figure S8 were captured by light‐assisted KPFM technique, and the SPV maps shown in Figure 4a,b,d,e were fitted [4]. The contact potential difference (ΔCPD) can be determined, as shown in Figure 4c,f, by subtracting the dark potential value with that measured under photoirradiation. As shown in Figure 4a,d, the SPVM maps of MCS_40% and MCS‐Co_11% composites show inhomogeneous spatial charge distributions in dark condition, while the SPV signals are significantly enhanced under photoirradiation (Figure 4b,e), suggesting obvious exciton spatial separation under photoirradiation. Furthermore, the MCS‐Co_11% composite shows a greater ΔCPD (57 mV) than that of the MCS_40% composite (35 mV). Moreover, based on their surface morphology and surface charge distribution, the normalized potential intensity spectra, as shown in Figure S9, can be obtained. Similarly, the normalized potential intensities of the MCS_40% (0.135 mV/nm) and MCS‐Co_11% (0.446 mV/nm) composites under photoirradiation are higher than those (0.0252 and 0.154 mV/nm) in dark condition, and the MCS‐Co_11% composite shows a higher normalized surface potential intensity in both dark and light conditions than that of the MCS_40% composite, demonstrating that the co‐decoration of single‐atom Co─O sites and Co clusters can significantly promote the exciton spatial separation kinetics and slow down its recombination kinetics, thereby obtaining a higher photoexciton utilization efficiency.

FIGURE 4.

Photoelectric properties and exciton dynamics: (a, b, d, e) SPV maps, (c, f) SPV spectra, (g, j) 2D mapped TA spectra, (h, k) attenuated TA spectra, and (i, l) decay kinetic curves of MCS_40% and MCS‐Co_11%. The SPV maps in panels (b) and (f) were obtained in dark, and those in panels (c) and (g) were obtained under photoirradiation.

fs‐TA spectroscopy is an effective technique for revealing the exciton dynamics of photocatalysts, thereby the fs‐TA spectra of the MCS_40% and MCS‐Co_11% were tested on a 380 nm pump pulse excitation and a 450–750 nm white light pulse probe. As shown in Figure 4g,j, both the two‐dimensional (2D)‐mapped TA spectra of the MCS_40% and MCS‐Co_11% composites show a distinct excited‐state absorption (ESA) signal at 475 nm, which corresponds to the transition of the ground‐state electrons (GS‐e⁻) of MoS₂ and CdS to higher energy levels [44]. Meanwhile, both samples emit a distinct ground‐state bleaching (GSB) signal at 510 nm, which can be attributed to the transition of GS‐e⁻ to its excited state (ES) after absorbing the pumped pulsed laser, resulting in a reduced number of GS‐e⁻ [45]. The TA spectra of the MCS_40% (Figures S10a and 4h) and MCS‐Co_11% (Figures S10b and 4k) composites at different relaxation times indicate that both their TA intensities of GSB peaks show a trend of first rapidly increasing in the fs range and then gradually decreasing in the ps and ns ranges. The former and the latter correspond to the generation of instantaneous photoexcitons and their recombination kinetics, respectively [46]. In addition, the maximum TA intensity of the GSB signal of the MCS‐Co_11% composite is significantly lower than that of the MCS_40% composite. Combined with its significantly enhanced TRPC and SPV signals (Figures S7 and 4k) under photoactivation, it also proves the effectiveness of co‐decoration of single‐atom Co─O sites and Co clusters for promoting the rapid transfer of photoexcitons and suppressing their recombination kinetics, thereby an increased surface reactivity is presented.

Based on the attenuated TA spectra in Figure 4h,k, the decay kinetic curves shown in Figure 4i,l were fitted through the triple exponential function [47]. Both the MCS_40% and MCS‐Co_11% composites exhibit two rapid attenuated processes (τ ₁ and τ ₂) and one slow attenuated process (τ ₃). The τ ₁ of the MCS‐Co_11% composite (τ ₁ = 3.71 ps) is shorter than that of the MCS_40% composite (τ ₁ = 3.75 ps), indicating the easier capture of its photoexcited e⁻ [48], which is obviously beneficial for reducing the recombination frequency of photoexcitons. Moreover, the MCS‐Co_11% composite (τ ₂ = 40.9 ps, τ ₃ = 663 ps) has a longer TA signal decay lifetime than the MCS_40% composite (τ ₂ = 38.9 ps, τ ₃ = 562 ps), indicating that further co‐decorating the single‐atom Co─O sites and Co clusters on the surface of MCS_40% composite can significantly promote the capture behavior of photoexcitons, resulting in a hindered recombination behavior and higher utilization efficiency of photoexcitons, thereby facilitating the efficient BA oxidation coupled by H₂ evolution.

2.5. Catalytic Mechanism of Collaborative Photocatalytic Process Based on DFT Calculations

Based on the relevant photoelectric data (Figures S5, S11 and S12a,b) and the calculation methods (shown in Supporting Information), the S‐scheme heterostructured mechanism can be confirmed, as shown in Figure 5a, due to the MoS₂‐to‐CdS electron transfer, which is conducive to the spatial separation and rapid transfer of photoexcitons. Based on the microstructure analysis of the catalyst, Co clusters containing 2–3 Co atoms and single‐atom Co─O coordination structures coexist on surface, so its local conformation diagram, as shown in Figure 5b, was optimized based on DFT calculation to analyze the transfer directions and enrichment sites of positive and negative charges (e⁻ and h⁺) and precisely locate the main active centers of oxidation half‐reaction and reduction half‐reaction from the perspective of adsorption and transformation energy barriers of reactant molecules, main intermediates, and products. Figure 5c shows the differential charge density diagram of the local conformation of the MCS‐Co_11% composite. The blue cloud is mainly concentrated on the Co cluster (Q _Bader = 0.76 e), while the yellow cloud is mainly concentrated on the Co atom of Co─O site (Q _Bader = 0.14 e), indicating that, in a static state, the former has a higher electron density, while the latter is in an electron dissipation state. Therefore, when the catalyst is under photoactivation, due to the spatial separation of photoexcitons on the semiconductor side (MCS_40%), the density of photoexcited e⁻ significantly increases, thereby they tend to transfer to the Co─O site with e⁻ dissipation after migrating to the MCS_40% surface, while the photogenerated h⁺ tend to transfer to the Co cluster site. This transfer mechanism is conducive to achieving the spatial separation of photoexcitons and suppressing their recombination. Based on the EPR analysis (Figure S13), it has been proven that the main intermediates of BA dehydrogenation oxidation are the ·CH(OH)Ph radicals, and this transformation process is the rate‐controlling step of the entire oxidation half‐reaction, thereby the adsorption and transformation energy barriers of BA molecules and their dehydrogenation oxidation intermediates and product molecules at the Co─O and Co cluster sites were calculated. As shown in the adsorption energy level diagram of Figure 5d, the adsorption energy barrier of BA molecules at the Co cluster site (−0.74 eV) is much lower than that at the Co─O site (−0.25 eV), and the energy barrier that needs to be overcome for Cα‐H bond cleavage to transform into ·CH(OH)Ph intermediate at the Co cluster site (0.32 eV) is also much lower than that at the Co─O site (0.36 eV), indicating that BA molecules are more inclined to adsorb at the Co cluster site and undergo dehydrogenation oxidation to transform into the ·CH(OH)Ph intermediates. Therefore, the relatively e⁻‐concentrated Co cluster sites on the MCS‐Co_11% surface act as h⁺‐acceptors under photoactivation to receive photogenerated h⁺ from the semiconductor side, achieving the BA dehydrogenation oxidation. Figure 5e shows the adsorption energy barrier diagram of H protons at the Co─O and Co cluster sites. The Gibbs free energy change (∆G _H*) of H proton adsorption is a vital indicator for theoretically evaluating hydrogen evolution capacity. Obviously, the ∆G _H* of H proton adsorption at the Co─O site (−2.28 eV) is closer to 0 eV than that at the Co cluster site (−2.46 eV), indicating that the Co─O site is more friendly to the H proton adsorption and the subsequent H₂ evolution half‐reaction as the main active center.

FIGURE 5.

Catalytic mechanism of collaborative photocatalysis: (a) Band structure of MCS_40% composite. (b, c) Local conformation diagram of the MCS‐Co_11% composite and its corresponding differential charge density diagram. (d) Adsorption energy level diagram for BA adsorption/oxidation and (e) adsorption energy level diagram of H protons at the Co─O and Co cluster sites. (f) Schematic diagram of the dehydrogenation oxidation of BA coupled by H₂ evolution using MCS‐Co_11% composite catalyst.

In summary, due to its typical metallic property, the Co cluster concentrates a relatively high density of e⁻, thereby exhibiting a strong ability to receive photogenerated h⁺ under photoactivation and becoming the main active center for BA adsorption and dehydrogenation oxidation. However, the strong electronegativity of the O atom significantly reduces the electron density of the Co atom in the Co─O site, thereby gaining a strong ability to receive photoexcited e⁻ under photoactivation and becoming the main active center for adsorbing H⁺ and H₂ evolution half‐reaction. The significant electronegativity difference of the two sites promotes the spatial separation of photoexcitons, enabling the two half‐reactions to work in synergy with each other, and achieving more efficient BA dehydrogenation oxidation and H₂ evolution (Figure 5f).

3. Conclusion

In this study, the Co clusters and single‐atom Co─O sites were co‐decorated on the heterostructured MoS₂/CdS composite to integrate the merits of heterojunction and non‐precious metal cocatalysis, which was demonstrated by adopting the XAFS spectroscopy and AC‐TEM characterizations. Under artificial sunlight‐irradiation, a remarkable collaborative photoactivity for BA oxidation (181.8 mmol_BAD·g⁻¹·h⁻¹) coupled by H₂ evolution (64.6 mmol_H2·g⁻¹·h⁻¹) was achieved on this innovation system. fs‐TA spectroscopy and KPFM techniques proved the improved spatial separation/transfer kinetics and hindered recombination kinetics of photoexcitons. Based on the DFT calculations, the Co clusters act as the oxidation sites for BA, and the Co─O sites serve as the H₂ evolution reduction centers, enabling the synergistically enhanced REDOX half‐reactions due to more effective spatial separation of photoexcitons. This study proposes a strategic catalyst construction insight for high‐efficient collaborative REDOX photocatalysis.

4. Experimental Section

4.1. Preparation of Heterostructured MCS composites

The synthesized CdS NPs was added to the synthetic precursor system of MoS₂ to hydrothermally prepare the heterostructured MCS composite. In 80 mL of mixed solution containing 4.84 g Na₂MoO₄·2H₂O (A.R.) and 4.56 g NH₂CSNH₂ (A.R.), 0.4 g CdS NPs was ultrasonically dispersed to form a uniform mixed system, and it was transferred to a 100 mL sealed high‐pressure reactor and heated to 200°C for 8 h. After the reaction, the product was collected and rinsed several times with deionized water, and a black powdered MCS composite was obtained after drying in a vacuum oven for 12 h at 60°C. The interfacial interaction between MoS₂ and CdS was precisely regulated by adjusting the addition amount of CdS NPs, and the prepared composites were represented as MCS_x, where x represents the percentage content of CdS NPs in the composite (x = 10%, 20%, 30%, 40%, 50%). The synthetic methods of CdS NPs and MoS₂ microspheres are included in Supporting Information.

4.2. Preparation of MCS‐Co Composite Photocatalysts

Through fine regulation of surface adsorption and wet chemical reduction processes, the Co species was successfully decorated on the surface of the optimized MCS composite to prepare the MCS‐Co composite photocatalyst. Optimized MCS composite weighing 0.2 g was ultrasonically dispersed in 20 mL of deionized water. After adding 0.02 g Co(NO₃)₂·6H₂O (A.R.), it was magnetically stirred for 3 h, and 10 mL of deionized water dissolving excess of 10% chemical equivalent NaBH₄ (A.R.) was gradually added under continuous stirring. After 2 h of reaction, it was rinsed several times with deionized water and dried in a vacuum oven for 12 h at 60°C to obtain the MCS‐Co composite photocatalyst. In this process, the loading capacity of Co species on the surface of the optimized MCS composite can be precisely regulated by adjusting the dosage of Co(NO₃)₂·6H₂O. The prepared composite catalysts were represented as MCS‐Co_y, where y represents the percentage content of Co(NO₃)₂·6H₂O adding to the preparation system (y = 8%, 9%, 10%, 11%, 12%).

4.3. Photoactivity Tests for Artificial Sunlight‐Driven BA Oxidation Coupled by H₂ Evolution

All photoactivity tests were performed on a commercial closed glass reactor and gas circulation test system (Labsolar‐III, AG, Beijing Perfect Light Technology Co., CN) equipped with a 300 W xenon lamp (MC‐PF300C, 212.9 mW·cm⁻², CN), which was connected to a gas chromatograph (Fuli, GC 7900, CN). Ten milligrams of catalyst were used in each test, and the liquid phase was composed of 47 mL of deionized water and 3 mL of BA (A.R.). The detailed operation can be referred to in our previous research [22] and is provided in Supporting Information. The stability evaluation of the catalyst also adopted the membrane‐assisted method, and the methods for membrane preparation and stability evaluation can be referred to in our previous study [22].

The supplemented descriptions of chemical reagents, preparation of CdS NPs, MoS₂ microspheres, and P/MCS‐Co composite membrane, parameters for photoelectrical tests, materials characterizations, and methods for DFT calculations are provided in Supporting Information.

Author Contributions

Wen Duan: data curation, investigation, validation, formal analysis, writing–original draft, supervision, visualization, methodology, software. Wei Li: conceptualization, investigation, funding acquisition, writing–original draft, writing – review and editing, methodology, visualization, formal analysis, project administration, resources, supervision, data curation. Rongxia Cui: software, data curation, investigation, validation, formal analysis, methodology, writing – original draft. Tenghao Ma: investigation, validation, formal analysis, software, visualization, data curation. Ying Su: funding acquisition, supervision, resources, project administration, writing – original draft, writing – review and editing. Chunping Hou: writing – review and editing, investigation. Chuanyi Wang: writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: anie72734‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.