Hydrogen sulphide needs to be removed from natural gas for economic and environmental reasons. Adsorption is a commonly used operation to remove H2S, and the development of novel adsorbents with larger capacity and selectivity is an active area of research. This thesis focuses on adsorption and separation of hydrogen sulphide using metal organic frameworks (MOFs). MOFs are constructed from a metal cluster and an organic linker. The organic ligands can be tailored by changing their chemical nature, tuning their length, or modify by adding functional groups. The pore shape can be tuned by changing or mixing the metal ion. The ability to tailor the MOFs makes them interesting materials for industrial applications because their properties can be optimised, but it also creates an extraordinarily large number of possibilities that need to be tested to find the optimum material for a given separation. Molecular simulations are an attractive tool to study the properties of MOFs, as it allows to predict their performance based on intermolecular interactions, and provides insight into the molecular level structure of the system. In this work, two different families of MOF materials were studied: FPYEu and UiO. The first one has applications for sensors due to its low capacity while the second can have applications for gas separations. The validity of the force fields to describe the gas molecules and the solid frameworks were assessed, as well as the influence of using different approximations in the description of the solid framework. In general, it was found that force field applicability needs to be treated carefully, as in some cases large deviation with experimental results are obtained. Nevertheless, for a single family of materials, like UiO-66, UiO-67, UiO-68, it is possible to use the same force field parameters to describe correctly the adsorption in all of them. More importantly, it was also shown that using simplified descriptions of the functionalised material structure gives results within 10% of those obtained with the most accurate descriptions. Overall, it was found that functionalised UiO-67 with four carboxylic acid groups has the best compromise in terms of selectivity, adsorbent selection parameter and adsorption capacity for H2S separations. The enhancement in selectivity due to the functional groups is approximately 80%. The selection of an adsorbent requires a lot of additional information beyond the selectivity and capacity, which are the focus of this work, but having fast and accurate approaches to calculate adsorption selectivity and capacity can provide an initial screening of potentially interesting adsorbents for a given application.