Estimation of diffusion parameters in functionalized silicas with modulated porosity. Part I: chromatographic studies.

The diffusion parameters of binary gas mixture He (tracer gas)-N2 (carrier gas) in hybrid organic-inorganic SiO2-X porous solids which have suffered gradual functionalization with functional groups X of increasing length (X = psi, [triple bond]Si-H, [triple bond]Si-CH2OH, [triple bond]Si-(CH2)3OH, [triple bond]Si-(CH2)11CH3) are reported. The effective diffusivities Deff, the Henry law constants K as well as the tortuosity factors tau for the examined solids were estimated by a typical pulse gas chromatographic method. Analysis of the experimental results was carried out by the well-known method of linearization of moments. The moments s analysis provides a powerful means for extracting diffusion parameters from the experimental response curves The proposed methodology is simple compared to other similar studies and provides rapidly reliable data. The results of this work indicate that the effective diffusivity Deff in porous networks drops markedly as the initial porosity of the parent SiO2 sample is blocked by the functionalization of the pore surfaces with functional groups of increasing size, [triple bond]Si-H, [triple bond]Si-CH2OH, [triple bond]Si-(CH2)3OH and [triple bond]Si-(CH2)11CH3. The low values of the Henry law constants K found indicate that the adsorption of He on the porous surfaces for all the solids is weak. Also, the tortuosity factor r is proportionally correlated to the pore blocking effects and the percolation phenomena of gases flowing into the porous network.

Estimation of diffusion parameters in functionalized silicas with modulated porosity. Part II: pore network modeling.

In this work, the pore structure of those five (5) silicas SiO2-X (see Part I) which have suffered gradual functionalization with functional groups X of increasing length (X = psi, [triple bond]Si-H, [triple bond]Si-CH2OH, [triple bond]Si-(CH2)3OH, [triple bond]Si-(CH2)11CH3), is modeled as a three-dimensional cubic network of cylindrical pores. Those hybrids organic-inorganic SiO2-X samples are characterized by different extent of pore blocking effects. The pores of samples are represented in a 9 x 9 x 9 lattice by the nodes as well the bonds that are interconnected in a so-called dual site-bond model, DSBM, network. The pore network is developed using a Monte Carlo statistical method where the cylindrical pores (nodes and bonds) are randomly assigned into the lattice, until matching of the theoretical results to the experimental data of N2 adsorption-desorption measurements. Thus, a visual picture of the porous solid is possible. This realistic network is used next in order to study the steady-state gas transport (Knudsen gas-phase and viscous diffusion) properties for the examined materials and how flow processes depend on the morphology of the pore structure. The pore diffusivity Dp and total permeability P of each porous medium is determined based on theoretical calculations and the structural statistical parameters, such as porosity epsilonp, tortuosity factor tau and connectivity c of pores is discussed with the corresponding experimental data described in Part I of this work. The results indicate clearly that the diffusion model made it possible to predict pore effective diffusivity in these porous media in very good agreement with the corresponding experimental results for all the examined solids (Part I). The pore diffusivity increases significantly as the value of the pore connectivity increases but the transport properties of the network are influenced strongly at lowest connectivity. Also the predicted tortuosity factor is related inversely to the extent of interconnection of pores in these solids, which indicates that the influence of pore branching to the tortuosity factor of the pore network decreases, as connectivity increases.