Development of pore network method in simulation of non-catalytic gas–solid reactions – Study of sulfur dioxide chemisorption on copper oxide sorbents

•A model for non-catalytic gas solid reaction in a non-overlapped pore of known thickness.•Simulation of NCGSR in a pore network (NCGSR-PNS).•Effects of several parameters on NCGSR-PNS.•A model for pore thickness using sorbent characterization.•Validation of NCGSR-PNS using experimental data from literature. In this paper non-catalytic gas–solid reaction (NCGSR) was studied in three-dimensional pore network, as a media for discrete methods. Constructed networks consist of regular and nonoverlapping holes in three dimensions with the maximum number of neighbors equal to 6. The pore network model was then used to simulate the NCGSR. To obtain a model for NCGSR in the pore network, we first proposed a model for pore conversion. As the reaction progresses, the pore radius decreases due to formation of the product layer. The model was used to predict the conversion of the network. The pores were blocked sequentially until the reaction ceases due to pore blockage, solid reactant consumption or product layer resistance. The effects of several parameters on the network characteristic were investigated. It was shown that the product layer diffusivity has not much effect on the results, while initial connectivity, initial solid reactant concentration, pore thickness, pore diameter and expanding factor have significant impresses on the conversion-time diagrams. NCGSR accompanied by PNS was then used to reproduce the experimental data of SO2–CuO/Alumina reaction in thermal gravimetric analyzer (TGA) extracted form literature without using any adjustable parameter such as tortuosity factor. To guess the pore thickness for the network, a simple model was submitted based on structure of the actual sorbents. It was revealed that the model can provide a suitable estimate for conversion-time diagram of the sorbents if the active sites were dispersed mono layer over the catalyst surface. However, some inherent incapabilities of the model result in the deviation of the model from experimental.