A thiomolybdate cluster for visible-light-driven hydrogen evolution: comparison of homogeneous and heterogeneous approaches

This study investigates the hydrogen evolution reaction (HER) efficiency of two photosystems incorporating an all-inorganic molecular thiomolybdate [Mo3S13]2− cluster as a HER catalyst. First, we delve into the performance of a homogeneous [Mo3S13]2−/[Ru(bpy)3]2+ (Mo3/Ru) dyad which demonstrates high turnover frequencies (TOFs) and apparent quantum yields (AQYs) at 445 nm approaching the level of 0.5%, yet its performance is marked by pronounced deactivation. In contrast, a heterogeneous approach involves anchoring [Mo3S13]2− onto graphitic carbon nitride (GCN) nanosheets through weak electrostatic association with its triazine/heptazine scaffold. [Mo3S13]2−/GCN (Mo3/GCN) displays effective H2 generation under visible light, with TOF metrics on par with those of its homogeneous analog. Although substantial leaching of [Mo3S13]2− species from the Mo3/GCN surface occurs, the remaining {Mo3}-based centers demonstrate impressive stability, leading to enduring HER performance, starkly distinguishing it from the homogeneous Mo3/Ru photosystem. Photoluminescence (PL) quenching experiments confirm that the performance of Mo3/GCN is not limited by the quality of the inorganic interface, but could be optimized by using higher surface area supports or a higher concentration of [Mo3S13]2− sites. Our findings showcase complexities underlying the evaluation and comparison of photosystems comprising well-defined catalytic centers and pave the way for developing analogous surface-supported (photo)catalysts with broad use in energy applications.


XRD and FTIR of the Mo-Clusters
Successful formation of [Mo3S13] 2-clusters was confirmed using FTIR spectroscopy.Figure S1a shows the FTIR spectra of Na2[Mo3S13] revealing three peaks at 542 cm -1 , 505 (510/501 doublet) cm -1 , and 458 cm -1 in the fingerprint region that correspond to bridging, terminal, and apical sulfur ligands in the cluster framework. [1,2]dditionally, only OH bending and stretching vibrations are observed at ~1600 cm -1 and ~3300 cm -1 , which correspond to residual and crystalline water.The purity of the product (e.g.complete cation exchange) is manifested by no extra peaks observed in the IR spectrum.]

DRS
Optoelectronic properties of as-prepared (GCN) and protonated (H-GCN) carbon nitride (details in Experimental section) have been investigated using diffusereflectance spectroscopy (DRS).Figure S2a shows as-obtained profiles that highlight strong absorption of both GCN materials in visible-light range (400-700 nm).Tauc analyses (Figure S2b) reveal the band gap of ca.2.75 eV, which corresponds well to the literature. [4]g. S2 | DRS plots of (a) GCN and H-GCN powders, (b) Tauc plots of bare GCN and H-GCN.Shaded area corresponds to the range of visible-light absorption (i.e.>400 nm).
Fig. S3 | DRS plots of [Mo3S13] 2-(Mo3) and 10Mo3/GCN composite, the latter is measured taking GCN as a baseline.The broad peak centered at ~456 nm is the qualitative indication for the presence of Mo3 on the surface of GCN.The quantitative assessment of the loading values is performed via TXRF (Table 1).

Comparison of Mo3/GCN and Mo3/Ru in terms of excitation source
First, we verified if the heterogenized [Mo3S13] 2-clusters of the Mo3/GCN composite are able to promote the desired HER under visible-light illumination.Figure 1b compares absorption spectra of the GCN (measured in reflectance mode) and [Ru(bpy)3] 2+ (measured in transmission mode) compared to the output spectrum of the visible-light emitter used in this work to evaluate light-driven HER performance of both Mo3/GCN and Mo3/Ru.The band gap value of GCN can be estimated to around 2.75 eV (see Figure S2), which allows 445 nm photons to trigger band-to-band excitation (C 2p to N 2p) and generate an electron-hole pair.We note, however, that -compared to the extent of the [Ru(bpy)3] 2+ absorption -only edgeto-edge transition in GCN is likely to be triggered by the light source leaving the generated electron-hole pairs with little-to-no overpotential for further reaction.

Optimization of HER conditions
Table S1.Optimization of parameters for hydrogen evolution reaction under homogeneous conditions.The concentration of [Ru(bpy)3]PF6 photosensitizer was kept 0.645 mM for all the experiments mentioned in the table and the reaction mixture was illuminated for 30 min with LED lamp 445 nm.
[7] However, ascorbic acid (H2A) produces an intermediate during the hole scavenging process which is reported to oxidize Rubased PS and therefore reduces the overall hydrogen evolution performance. [5,8]This makes the stability of the Mo3/Ru photosystem questionable as Ru-PS degrades with increase in illumination time as well as it recombines with oxidized intermediate of ascorbic acid promoting charge recombination.However, when assessing overall HER performance of the photosystem, one should also consider other less direct contributions.First of all, the choice of the solvent system and the associated dielectric constant of the reaction medium affect the final availability of sacrificial donor as well as the efficiency of the charge transfer process. [6]esides this, the mechanism of electron donor oxidation (via hole trapping) often involves intermediate species that may affect the reaction in a number of ways e.g.act as recombination centers. [5]

Role of sacrificial agent
To verify the role of H2A as an electron donor in our HER experiments performed with the Mo3/GCN photosystem, we conducted PL quenching experiments using [Mo3S13] 2--free and [Mo3S13] 2--loaded GCN suspensions in water in and without the presence of H2A.The reason for this is that the two-electron oxidized species of ascorbate (i.e., dehydroascorbate) as well as radicals generated from H2A are known to react with the reduced from of the ruthenium dye.Despite the first reaction is slowed by its bimolecular nature, we still consider both being relevant under our reaction conditions. [8,5]As summarized in Figure S7

Choice of H2A concentration
To further justify the choice of H2A concentration (0.1 M) used in our HER experiments performed with the Mo3/GCN photosystem, we conducted additional PL quenching experiments.Figure 8d shows that the addition of low concentrations (0-0.05M) of H2A to Mo3/GCN gradually reduces the PL emission intensity of GCN, which suggests that reductive quenching mechanism is in place and it leads to improved electron/hole separation (i.e.thus affording higher HER performances).It is noteworthy, however, that higher concentrations of H2A (˃ 0.05 M) do not lead to further quenching (i.e.saturation is reached).This result indicates that hole extraction by H2A is not a performance-limiting factor at H2A concentrations above ~0.1 M and that the use of this H2A concentration can be justified when conducting HER studies.

Reloading experiments
To investigate the stability of Na2[Mo3S13] in Mo3/Ru photosystem, a solution containing 2 mL (4.5:4.5:1) of [Ru(bpy)3] 2+ photosensitizer (PS, 0.645 mM in MeOH), the [Mo3S13] 2-catalyst (50 μM in MeOH), and H2A proton donor (0.1 M in H2O) solvent mixture was irradiated and the H2 was detected by gas chromatography until saturated, indicated by a plateau (red curve in Figure S9, the point of 120 min).After this point was reached, the reaction mixture was recharged with 100 μL of a freshly prepared solution of [Ru(bpy)3] 2+ photosensitizer (0.645 mM) and 100 μL H2A to yield 2.2 mL of a reloaded reaction mixture with 545 mM [Ru(bpy)3] 2+ PS, 0.9 mM H2A and 45 μM [Mo3S13] 2-, thereby mimicking the initial Mo3/Ru (1:13) molar ratios of the first HER cycle.Afterwards, sealing, de-gassing and irradiation of the reloaded reaction mixture initiated the second HER cycle (blue curve in Figure S9).In another experiment, after the first HER cycle, the solution was recharged with just the PS (grey curve in Figure S9).Both secondary datasets (grey and blue curves) show that the original activity of the freshly-made Mo3/Ru photosystem can not be reached when PS (or PS and H2A) are reloaded.This, in turn, indicates that at least partial degradation of [Mo3S13] 2-takes place along with PS and SA depletion.

Postcatalytic characterization
Table S2.TXRF of Mo3/GCN composites before and after hydrogen evolution reaction in ascorbic acid and TEOA sacrificial donors to quantify remaining cluster loadings in wt.%.
, the addition of 0.1M H2A to the GCN suspensions ([Mo3S13] 2--free in a, with 1 wt.% [Mo3S13] 2-in b and with 10 wt.% [Mo3S13] 2-in c) reduce the PL intensities in all cases compared to those obtained in pure water.This result confirms the role of H2A as an efficient hole acceptor, which leads to a better separation of charge carriers photoexcited in GCN under our reaction conditions.

Fig. S9 |
Fig. S9 | Reloading experiments showing the effect of catalytic components on the HER performance.The reaction solution (50 µM [Mo3S13] 2-, 0.645 mM [Ru(bpy)3] 2+ (PS), and 0.1 M H2A) after the first HER cycle (red) was recharged with [Ru(bpy)3] 2+ (PS) and H2A (blue) as described above.After purging, the second HER cycle didn't recover the original HER activity.Reloading only the PS after the first HER cycle (grey) accounted for the quarter of original HER activity.