First Principles Simulations of Hydrogen Storage via Spillover in MOF-5

Metal organic frameworks (MOF) have attracted considerable attention as hydrogen storage materials due to their high surface areas and ability to adsorb large quantities of hydrogen by weight ($\sim $12 wt. {\%} at 100bar). However, as a consequence of the weak H$_{2}$-MOF bonding interaction ($\sim $5 kJ/mol H$_{2})$, this uptake occurs only at cryogenic temperatures; at room temperature, gravimetric capacities do not exceed $\sim $0.5 wt. {\%}. As an ideal storage system would operate under ambient conditions, renewed interest in these materials has been sparked by recent experiments demonstrating RT uptake of $\sim $4 wt. {\%} via the so-called ``spillover'' mechanism [JACS \textbf{128}, 8136 (2006)]. In contrast to the conventional mechanism of MOF-based storage, where \textit{molecular} H$_{2}$ bonds directly to the MOF, spillover employs a hydrogen dissociation catalyst to generate \textit{atomic} hydrogen (H), presumably resulting in stronger H-MOF bonding. Recent computational studies of spillover have reported conflicting results regarding the nature of this interaction. In an effort to resolve these ambiguities and clarify the thermodynamics of MOF-based spillover, DFT calculations are used to evaluate binding energies for several H adsorption configurations on MOF-5. Importantly, our calculations avoid the cluster approximation to the MOF geometry--a source of significant uncertainty in previous studies--and account for finite-temperature contributions to the free energy of adsorption.