the whole media. The main limitation of this method is not to consider the changes of pore connectivity and pore volume because of blocking the pores. A Probability Density Function (PDF) was used to estimate the normal Pore Size Distribution (PSD) with mean pore radius, r a . Transient Fickian mass transfer equations with Langmuir isotherm for adsorption phenomena followed by Higashi, Ito and Oishi (HIO) surface diffusion model were solved in each pore to obtain concentration profiles in the network. The effective diffusivity was predicted using trial and error in such a way that the usual continuum model accurately correlates the network simulation results. This effective diffusivity was then used to predict pore or surface tortousity factors. It was also shown that surface tortuosity factor decreases exponentially with increase in monolayer concentration but increases as the adsorption constant increases. In this kind of simulations unlike the continuous methods, the exact information of effective diffusivity and tortuosity factor is achieved. In pore network model, the porous media structure is considered as a three dimensional regular network which is composed of nodes and these nodes connect to each other by some pores. Helium as a non-adsorbing gas is used to determine the characteristics of the porous medium. The permeability of Helium versus presser was obtained linear, because of the effect of two diffusion mechanism, the Knudsen diffusion and the viscous flow diffusion. In this thesis by investigating the adsorption of light mercaptans on Activated Carbon in a three dimensional network, diffusivity and tortuosity factors are achieved according to Knudsen diffusion, viscous flow and surface diffusion conditions. In adsorption volume of pores of sorbent are reduced and this reduction affects on the permeability of adsorbed phase. Therefore in this modeling, the effect of the adsorbed molecules on the cross section area of pores has been investigated. So it was observed that the predicted results show better agreement to the experimental data. The result shows that if the amount of adsorption be small, we can neglect the variation of effective diffusivity (till saturation) with time and place. The equations were solved by orthogonal collocation method to obtain the concentration distribution and the effective diffusivity of methyl mercaptan (CH 3 OH) in to activated carbon. The effective diffusivity for two concentration of methyl mercaptan was achieved by solving the model and it was seen about 4 percent difference between the achieved effective diffusivity and experimental data. This difference seems to be the effect of pore volume reduction by adsorption of methyl mercaptan on small pores. So this pore volume reduction affects on diffusivity of adsorbed phase.