A computational mechanistic investigation of hydrogen production in water using the [RhIII(dmbpy)2Cl2]+/[RuII(bpy)3]2+/ascorbic acid photocatalytic system

We recently reported an efficient molecular homogeneous photocatalytic system for hydrogen (H-2) production in water combining [Rh-III(dmbpy)(2)Cl-2](+) (dmbpy = 4,4'-dimethyl-2,2'-bipyridine) as a H-2 evolving catalyst, [Ru-II(bpy)(3)](2+) (bpy = 2,2'-bipyridine) as a photosensitizer and ascorbic acid as a sacrificial electron donor (Chem. - Eur. J., 2013, 19, 781). Herein, the possible rhodium intermediates and mechanistic pathways for H-2 production with this system were investigated at DFT/B3LYP level of theory and the most probable reaction pathways were proposed. The calculations confirmed that the initial step of the mechanism is a reductive quenching of the excited state of the Ru photosensitizer by ascorbate, affording the reduced [Ru-II(bpy)(2)(bpy(center dot-))](+) form, which is capable, in turn, of reducing the Rh-III catalyst to the distorted square planar [Rh-I(dmbpy)(2)](+) species. This two-electron reduction by [Ru-II(bpy)(2)(bpy(center dot-))](+) is sequential and occurs according to an ECEC mechanism which involves the release of one chloride after each one-electron reduction step of the Rh catalyst. The mechanism of disproportionation of the intermediate Rh-II species, much less thermodynamically favoured, cannot be barely ruled out since it could also be favoured from a kinetic point of view. The Rh-I catalyst reacts with H3O+ to generate the hexa-coordinated hydride [Rh-III(H)(dmbpy)(2)(X)](n+) (X = Cl- or H2O), as the key intermediate for H-2 release. The DFT study also revealed that the real source of protons for the hydride formation as well as the subsequent step of H-2 evolution is H3O+ rather than ascorbic acid, even if the latter does govern the pH of the aqueous solution. Besides, the calculations have shown that H-2 is preferentially released through an heterolytic mechanism by reaction of the RhIII(H) hydride and H3O+; the homolytic pathway, involving the reaction of two Rh-III(H) hydrides, being clearly less favoured. In parallel to this mechanism, the reduction of the Rh-III(H) hydride into the penta-coordinated species [Rh-II(H)(dmbpy)(2)](+) by [Ru-II(bpy)(2)(bpy(center dot-))](+) is also possible, according to the potentials of the respective species determined experimentally and this is confirmed by the calculations. From this Rh-II(H) species, the heterolytic and homolytic pathways are both thermodynamically favourable to produce H-2 confirming that Rh-II(H) is as reactive as Rh-III(H) towards the production of H-2.