General rules for the optimal external porosity of LC supports.

We present a series of numerically calculated plate height and flow resistance data obtained for an idealized chromatographic support mimic with variable bed porosity (0.3 </= epsilon </= 0.9), yielding a unique insight into how the main chromatographic performance parameters can be expected to vary with the external bed porosity, unbiased by any differences in molecular diffusivity or retention factor. The influence of pore heterogeneity effects is considered as well. It is found that the product h(min)nu(opt) depends only very weakly on epsilon and on the degree of pore heterogeneity. It is also found that the minimal separation impedance E(min) decreases monotonically with epsilon. This shows that the minimal plate height increase that can be expected for large porosity systems is always more than compensated by their reduced flow resistance, in agreement with the current observations in real silica monolith columns. Using the computed plate height data in an optimization analysis, it is found that large porosity supports can always potentially yield shorter analysis times or larger plate numbers than small porosity supports but need submicrometer feature sizes to actually achieve this. Assuming a lower limit on the producible or useable structural feature sizes, it is found that small N separations can best be performed with a small porosity packing, whereas large N separations require a large porosity packing if the column length (L) is left free. A plot yielding epsilon(opt) as a function of the required plate number has been established, showing that roughly epsilon(opt) approximately log(N) in both the ordered and the disordered support cases. It is also shown that the maximal increase in peak capacity ever to be expected from the use of high porosity supports is a factor of 2 (if the mobile-phase viscosity can be kept constant), potentially to be increased by a factor of 1.5 by increasing the homogeneity of the packing.