Surface Anchoring and Active Sites of [Mo3S13]2- Clusters as Co-Catalysts for Photocatalytic Hydrogen Evolution

Abstract

Achieving light-driven splitting of water with high efficiency remains a challenging task on the way to solar fuel exploration. In this work, to combine the advantages of heterogeneous and homogeneous photosystems, we covalently anchor noble-metal- and carbon-free thiomolybdate [Mo₃S₁₃]^2- clusters onto photoactive metal oxide supports to act as molecular co-catalysts for photocatalytic water splitting. We demonstrate that strong and surface-limited binding of the [Mo₃S₁₃]^2- to the oxide surfaces takes place. The attachment involves the loss of the majority of the terminal S₂ ^2- groups, upon which Mo-O-Ti bonds with the hydroxylated TiO₂ surface are established. The heterogenized [Mo₃S₁₃]^2- clusters are active and stable co-catalysts for the light-driven hydrogen evolution reaction (HER) with performance close to the level of the benchmark Pt. Optimal HER rates are achieved for 2 wt % cluster loadings, which we relate to the accessibility of the TiO₂ surface required for efficient hole scavenging. We further elucidate the active HER sites by applying thermal post-treatments in air and N₂. Our data demonstrate the importance of the trinuclear core of the [Mo₃S₁₃]^2- cluster and suggest bridging S₂ ^2- and vacant coordination sites at the Mo centers as likely HER active sites. This work provides a prime example for the successful heterogenization of an inorganic molecular cluster as a co-catalyst for light-driven HER and gives the incentive to explore other thio(oxo)metalates.

Figure 1

Cluster structure. (a) Powder XRD pattern of the Na₂[Mo₃S₁₃] and (NH₄)₂[Mo₃S₁₃] along with the ICDD 04-021-7028 reference pattern of the ammonium salt, (b) overlayed ATR-FTIR and Raman spectra of the Na₂[Mo₃S₁₃] powder featuring characteristic molecular vibrations and the corresponding ranges, (c) molecular model of the [Mo₃S₁₃]^2– compared to the edge structure of the MoS₂ sheet, similar bonding motifs are highlighted; and (d) UV–Vis absorption spectrum of the 0.025 mM Na₂[Mo₃S₁₃] solution in water and water/methanol featuring a characteristic absorption band.

Figure 2

Cluster immobilization. (a) Digital photographs of the xMo₃/TiO₂ samples with 0.1, 2, 5, and 7 wt % cluster loadings, (b) difference DRS spectra of the Mo₃/TiO₂ composites with different cluster loadings (0.1–20 wt %) subtracted from TiO₂ absorption (see details in Supporting Information, methods) along with the DRS spectrum of Na₂[Mo₃S₁₃] (dashed line), (c) real vs intended loading plot depicting the range of theoretical monolayer capacity (see details in Supporting Information, Section 6), a linear increase between the real and intended loadings for loadings <9 wt % and its saturation at higher loadings. (d) XRD pattern of Mo₃/TiO₂ composites (10 wt % loading) and of a physical mixture of Na₂[Mo₃S₁₃] and TiO₂ (1:9 wt %), (e) EDS-derived elemental mappings of Ti (ii), Mo (iii), and S (iv) in an exemplary 3Mo₃/TiO₂ composite.

Figure 3

Cluster attachment. (a,b) HAADF-STEM images of 3Mo₃/TiO₂ composites; Fourier filtered images are shown in the bottom panels; examples of detected {Mo₃} cores are circled and magnified in the insets in (a), where they are compared with the model of tilted [Mo₃S₁₃]^2– cluster cores, (c–e) XPS spectra of the Na₂[Mo₃S₁₃] and the clusters after attachment to the TiO₂ surface (c) Mo 3d region, (d) S 2p, and (e) O 1s regions with corresponding fits; real [Mo₃S₁₃]^2– loading is derived via TXRF to be 2 wt %.

Figure 4

Photocatalytic performance and active sites. (a) Hydrogen evolution rates plotted against the [Mo₃S₁₃]^2– cluster content and the degree of surface coverage in %; green area shows the standard deviation of hydrogen evolution rate values obtained from multiple measurements for each loading value, (b) PL spectra obtained from the catalyst solutions containing TA as OH trap after UV pre-illumination; inset shows intensities of the peak maximum (ca. 425 nm corresponding to hydroxyterephthalic acid emission) plotted against the real [Mo₃S₁₃]^2– loadings, (c) long-term photocatalytic hydrogen evolution experiments of Mo₃/TiO₂ composites with 2 and 10 wt % loadings and their comparison with Pt/TiO₂ in terms of HER stability, (d) comparison of the hydrogen generation rate of Mo₃/TiO₂ composites heat-treated (25–400 °C) under air and the N₂ atmosphere; full HER profiles are shown in Figures S12 and S19.