A novel physisorption model based on mathematical morphology operators preserving exact pore morphology and connectivity

Abstract

Complex large models of multi-scale microstructures of porous media do not exist in a manner adequately efficient to render their comprehensive analysis of textural and structural properties possible. The simulation of experimental processes, such as gas adsorption, raises two issues. First, the modeling of these complex materials must be sufficiently realistic. This implies that the numerical twin of the porous material must consider three essential aspects: a spatial structure defining different scales of porosity, comparable textural and structural properties of the real material and irregular pore morphology. Secondly, efficient algorithms must be developed to mimic the quasi-static phase transition behavior of fluids in a realistic manner. The proposed simulation approach allows to handle the physico-chemical phenomena inside complex materials by means of well-established mathematical operators. Morphological operators are used to efficiently mimic processes such as surface adsorption and pore filling. Fluid percolation that provokes phase transition is simulated by labeled connected components. This method relies entirely on morphological and structural operators, which has the advantage of substantially reducing the calculation time compared to that of density functional theory and molecular simulations-based approaches. In contrast to oversimplified models characterized by ideal pore shape and unconnected pores, our approach enables us to calculate the adsorption isotherm of realistic random models where pore morphology and porous network topology are unknown beforehand. We demonstrate that our model succeed in reproducing the adsorption isotherm of two well-known model materials (SBA-15 and KIT-5) and mesoporous alumina, represented by Cox Boolean models.