Recent technical improvements in the molecular dynamics (MD) simulation tec
hnique have led to re-evaluation of the transport properties of fluids conf
ined in narrow capillary pores of several molecular diameters in width (or
nanofluids). Coincident with these developments, it has also become clear t
hat unambiguous predictions of the transport properties of nanofluids may o
nly be made when a rigorous analysis based on statistical mechanical theory
is considered in conjunction with molecular simulation studies. In this pa
per, the theoretical analysis embodied in the Pozhar-Gubbins [L.A. Pozhar a
nd K.E. Gubbins, J. Chem. Phys., 99 (1993) 8970; L.A. Pozhar and K.E. Gubbi
ns, Phys. Rev., E56 (1997) 5367] statistical mechanical theory of transport
in strongly inhomogeneous fluid mixtures is combined with nonequilibrium a
nd equilibrium molecular dynamics techniques to investigate self-diffusion
in a dense fluid confined within a model crystalline nanopore. The results
obtained demonstrate that the spatial dependence of the transport parameter
s should be taken into consideration to reliably predict the diffusion flux
es within zeolitic systems. For the comparatively simple pore structure exa
mined in this work, the local self-diffusivity varies significantly in magn
itude over nanometer length scales with corresponding implications for the
interpretation of the rate processes taking place within crystalline nanopo
rous media. (C) 2001 Elsevier Science B.V. All rights reserved.