Liquid chromatography looks deceptively simple. A constant flow-rate stream of a solution percolates through an isothermal column, carrying the components of a mixture through a bed of an adsorbent. The components equilibrate between the solid and liquid phase, migrating at a velocity depending on the constant of this equilibrium. They elute from the column, separated provided a suitable system was chosen.

Yet, all physicochemical properties involved (densities, viscosities, diffusivities, equilibrium constants) depend on the pressure (that decreases almost but not quite linearly along the column); they depend on the temperature (that is rarely constant along the column, due to radial exchanges and to viscous friction); they depend on the concentration of the mixture components (that affect viscosities and equilibrium constants). Column beds are not homogeneous and their properties (permeability, density) vary along and across columns. Flow instabilities may take place.

Many of these effects are nearly negligible under the conventional conditions used in most industrial applications of preparative chromatography. Yet, they may suddenly have drastic consequences for those who attempt to increase the performance of such separations. Engineers have made major contributions to the solution of most fundamental and practical problems of HPLC in the last half-century. Their role is not yet finished. An important series of new problems may be lurking.