At high concentrations, one boundary (the front if the isotherm is con
vex upward) of the elution band of a single component becomes very ste
ep. Similarly, in frontal analysis, the same boundary is self-sharpeni
ng. These steep boundaries are called shock layers. They result from t
he steady-state equilibrium between a nonlinear thermodynamics of phas
e equilibrium on the one hand, which tends to create a concentration d
iscontinuity, and axial dispersion and a finite rate of mass transfer
on the other hand, which tend to relax the concentration gradient. We
show that the width of experimental breakthrough curves recorded in fr
ontal analysis agrees well with the prediction of the shock layer theo
ry of Rhee and Amundson in the high concentration range. It is a funct
ion of the coefficients of the column HETP equation and the height of
the concentration step and is independent of the column length, after
the steady state is achieved. In the low concentration range, a discre
pancy is observed when the column length is short and the migration di
stance is insufficient to achieve constant pattern behavior.