Molecular assemblies for photodriven hydrogen formation: Empirical design principles and predictive computational methods

Artificial photosynthetic systems that efficiently drive water oxidation and hydrogen formation could provide sustainable, local sources of energy-dense fuel. In one design paradigm, linked photosensitizer-catalyst assemblies proximally position chromophore, electron relay, and catalyst components to promote ultrafast intramolecular charge separation and facilitate the use of robust organic photosensitizers with short excited state lifetimes. However, this proximity of system components also introduces competitive deactivation pathways including energy transfer, fast charge recombination, and spin or heavy-atom induced intersystem crossing. The objective of this work is to empirically derive a set of design principles and identify predictive computational methods to guide the development of assemblies that limit competitive processes and efficiently produce long-lived charge separated states for photocatalysis. To meet this goal, several photosensitizer-catalyst architectures are explored, each formed via the coordination of a pyridyl-substituted organic photosensitizer to the [Co((difluoroboryl)dimethylglyoximato)2(H2O) 2] proton reduction catalyst. Singular value decomposition and global fitting methods are developed to analyze transient absorption data and characterize photoinduced dynamics. For a simple dyad incorporating a perylene-3,4-dicarboximide chromophore, rapid excited state decay without the productive formation of a long-lived charge separated state is observed. Conversely, with a donor-bridge-acceptor-catalyst architecture comprising a perylene donor, xylene bridge, and 1,8-naphthalimide electron acceptor, decoupling of the chromophore and catalyst prevents similar deactivation and facilitates multi-step electron transfer to reduce the catalyst in > 0.99 quantum yield. To investigate the molecular design principles that would allow the accumulation of multiple charges at a single acceptor site, a key requirement for photocatalytic hydrogen formation, a series of model structures are investigated using pump-pump-probe transient absorption spectroscopy. Dianion formation via sequential photoinduced electron transfer to a central acceptor is shown and a competing energy transfer pathway identified as the main cause of low quantum yields. Finally, in a step toward computationally guided molecular design from first principles, variationally optimized configuration interaction singles theory is shown to accurately predict the energy difference and electronic coupling between charge transfer and non-charge transfer excited states for six small donor-acceptor dyads. As a result, it is expected this method may accurately predict the rates of key photoinduced electron transfer reactions.