ABSTRACT CTS-9508680 Keith E. Gubbins Cornell University Micro-and meso-porous solids find widespread use in the chemical, oil, pharmaceutical and food industries as adsorbents and membranes for separation processes, and as catalysts and catalyst supports. At present the design of such processes and materials is largely empirical. The aim of this project is to provide more powerful methods, based in molecular theory and simulation, for prediction of absorbed fluid behavior and for the design of improved processes. Modern statistical mechanical methods, including density functional theory, simulation, and kinetic theory will be employed to carry our studies in four areas: 1. Development of an improved method for pore size distribution (PSD) analysis. Density functional theory proves much more powerful than classical methods for determining PSD's from absorption isotherms. This method, which is already findine commercial application for carbons, will be extended to gases other than nitrogen, and to materials such as aluminosilicate and oxide molecular sieves, which have cylindrical pore geometries. 2. Molecular simulations will be used to develop and understanding of the importance of connectivity effects in networks for both pure fluids and mixtures. The effects of connectivity on both diffusion rates and absorption isotherms will be investigated. Goals include incorporation of network effects into PSD analyses, and understanding how networks can hinder and promote separations. 3. Selective absorption of trace components in gas streams. Calculations will be made to elucidate the role of pore and system variables on the selective absorption of trace contaminants in nitrogen and air streams for materials with pores of cylindrical geometry. Specific case studies will be made for several trace halocarbons in a range of molecular sieves, carbons and oxides. The goal of this work is to develop methods for designing improved absorbents for removal of trace contaminants. 4. Viscou s flow and diffusion in pores. Nonequilibrium molecular dynamics methods will be used to study viscous flow and mutual diffusion in mixtures in pores. The goal of this work will be to develop an understanding of the role of fluid-wall interactions and pore shape and size on these transport processes.
