Thermodynamics of metal hydrides for hydrogen storage applications using first principles calculations.

摘要

Metal hydrides are promising candidates for H2 storage, but high stability and poor kinetics are the important challenges which have to be solved for vehicular applications. Most of recent experimental reports for improving thermodynamics of metal hydrides have been focused on lowering reaction enthalpies of a metal hydride by mixing other compounds. However, finding out metal hydride mixtures satisfying favorable thermodynamics among a large number of possible metal hydride mixtures is inefficient and thus a systematic approach is required for an efficient and rigorous solution. Our approaches introduced in this thesis allow a systematic screening of promising metal hydrides or their mixtures from all possible metal hydrides and their mixtures. Our approaches basically suggest two directions for improving metal hydride thermodynamics. First, our calculations for examining the relation between the particle size of simple metal hydrides and thermodynamics of their decomposition reactions provide that the relation would depend on the total surface energy difference between a metal and its hydride form. It ultimately suggests that we will be able to screen metal hydride nanoparticles having favorable thermodynamics from all possible metal hydrides by examining the total surface differences. Second, more importantly, we suggest that our thermodynamic calculations combined with the grand canonical linear programming method and updated database efficiently and rigorously screen potential promising bulk metal hydrides and their mixtures from a large collection of possible combinations. The screened promising metal hydrides and their mixtures can release H 2 via single step or multi step. Our additional free energy calculations for a few selected promising single step reactions and their metastable paths show that we can identify the most stable free energy paths for any selected reactant mixtures. In this thesis, we also demonstrate that a total free energy minimization method can predict the possible evolution of impurity other than H2 for several specified mixtures. However, it is not ready to predict reaction thermodynamics from a large number of compounds.