Modelling of functional hybrid organic-inorganic materials: from structure to properties

Computational modelling of porous hybrid organic-inorganic materials is a challenging task. The dual nature and flexibility of frameworks require the delicate description of systems. In addition, the absence of reliable force-fields and the size of systems makes the density functional theory (DFT) the best choice. The poor description of long-range dispersion interactions, a crucial deficiency of DFT, has been circumvented by introducing the dispersion correction schemes. Zeolitic imidazolate frameworks (ZIF), extended analogues of inorganic zeolites by topological similarity, are the focal point of this thesis. Their prospective applications in gas adsorption/separation and catalysis are driving the search of new not-yet-syntesised structures and methods to tune their properties. A set of new hypothetical ZIF has been explored, including lightweight LiB-based frameworks. To establish structure-property relationship, the impact of topologies and linker has been analysed. It has been revealed the crucial role of linker-linker interaction and packing for the thermodynamics stability of frameworks. Furthermore, the mechanical properties of ZIFs have been explored. Close collaboration with experimentalist groups has allowed the cross-linked analysis and rationalization of the observed of phenomena. Finally, an empirical force-field fitting protocol from ab initio data has been improved. The idea is to exploit the redundant internal coordinates in fitting as it makes the problem more linear, but an existing code was able to handle non-periodic models only. The periodicity has been implemented to allow fitting force-field for material, which could not be partitioned into non-periodic models, including ZIFs. The new protocol has been tested with several systems