Capillary condensation in a disordered mesoporous medium: a grand canonical Monte Carlo study

Abstract

The adsorption of rare gases in a disordered mesoporous silica glass has been studied by means of grand canonical Monte Carlo (GCMC) simulation. A series of porous samples has been obtained by using an off-lattice 3D reconstruction method recently introduced to reproduce topological and morphological properties of correlated disordered porous materials such as Vycor. The off-lattice functional of 115m2g−1 Vycor is applied to a simulation box containing silicon and oxygen atoms of cubic cristoballite with a homothetic reduction in order to obtain porous samples with mean pore size around 35 Å and specific surface around 220 m2 g−1. A realistic surface chemistry is then obtained by saturating all the dangling oxygen bonds with hydrogen. Some topological properties of the different 3D reconstructions of the Vycor-like material are analysed using chord length distributions and small angle scattering data. The GCMC Ar, Kr and Xe adsorption/desorption isotherms are calculated at different temperatures. At sufficiently low temperatures, they exhibit a capillary condensation transition with an adsorption branch having a finite slope accompanied by a hysteresis loop upon desorption. It has been shown on a set of simulated argon isotherms, that evolution with temperature of the GCMC results is similar to experiment. At the temperature at which the hysteresis loop disappears, it was found that the compressibility of the dense liquid-like phase at the maximum of the so-called hysteretic coexistence curves increases significantly. In the low pressure and temperature domain, different adsorption scenarios can be interpreted on the basis of a Zisman-type of criterion for wetting. The BET surface area is shown to be strongly related to this criterion. At higher pressure, it was found that the pore size distribution obtained by using the standard BJH analysis applied to both simulated adsorption and desorption data qualitatively reproduces the main features of the chord length distribution.