ABSTRACT
We report on a numerical study of the thermal conductivity of core-shell particle packed bed columns. Covering a variety of packing structures and a broad range of mobile phase and porous zone conductivities, it was in all cases found that switching to particles with a highly conducting core (e.g., with a gold or copper core instead of a silica core) would produce a much smaller increase of the effective heat conductivity of the bed (keff) than previously expected in literature. We found maximal increases on the order of some 20-70%, which is much lower than the potential increases up to 2000% assumed in literature. The overestimation in literature could be attributed to the fact that this literature was based on an incorrect extrapolation of the Zarichnyak-model which was the heat conductivity model predominantly used up till now. On the other hand, the computed relationships between keff and the core conductivity obtained in the present study are in good agreement with an analytical solution derived from the effective medium theory, a theory which is physically much more relevant for the case at hand than the Zarichnyak-model. The results also show that the observed increase in effective bed conductivity between fully porous and core-shell particle beds frequently observed in literature is not only due to the presence of the core, but that differences in the shell layer conductivity can play an equally important role. In addition, it could also be demonstrated that, if ways could be found to increase the conductivity of the shell layer, this would produce a much stronger increase of the overall bed conductivity than will ever be possible by increasing the conductivity of the cores.
