Suresh K. Bhatia, Division of Chemical Engineering, The University of Queensland, Brisbane, Australia and David Nicholson, Chemical Engineering, University of Queensland, Department of Chemical Engineering, Brisbane, 4072, Australia.
We develop a novel theory to predict the density dependence of the diffusivity of simple fluids in a molecularly sized nanopore with diffusely reflecting walls, incorporating nearest neighbor intermolecular interactions within the framework of the recent oscillator model of low density transport arising from this laboratory. It is shown that when the pore width is about two molecular diameters, at sufficiently high densities these interactions lead to a repulsive inner core, as a result of which the diffusing molecules undergo more frequent reflections at the wall. This leads to a reduction in diffusivity with increase in density, which is consistent with molecular dynamics simulation results, and contrasts with the behavior in larger pores where the transport coefficient has previously been shown to increase with increase in density due to viscous-like intermolecular interactions. At low densities the behavior is opposite, with the inner core becoming more attractive with increase in density, which can lead to an increase in diffusivity. The theory consistently explains molecular dynamics simulation results when the inhomogeneous pair distribution function of moving particles in the pore is axially periodic, suggesting concerted motion of neighboring molecules. It is also shown that a potential of mean force concept is inadequate for describing the influence of intermolecular interactions on transport. In the range of pore sizes at which the above anomalous diffusion is noted, there is also a levitation effect, whereby a maximum in the diffusivity with respect to pore size occurs at the size at which there is a transition from a single to a double potential well. The oscillator theory of the author describes this effect theoretically, explaining earlier molecular dynamics simulation results that have previously not been well understood. Our work in the above areas will be reviewed.