Overloaded gradient elution chromatography on heterogeneous adsorbents in reversed-phase liquid chromatography

Overloaded band profiles of phenol were measured on a C18-Kromasil column in gradient elution conditions. The mobile phase used was a mixture of methanol and water. The volume fraction of methanol was allowed to vary between 0 and 0.5. A general adsorption model, which expresses the amount of phenol adsorbed q∗ as a function of both its concentration C and the composition ϕ of the organic modifier (methanol) in the mobile phase, was empirically derived from previous independent adsorption experiments based on frontal analysis (FA) and frontal analysis by the characteristic point (FACP). Accordingly, the general model was an extension of the simplest heterogeneous model, the Bilangmuir model, to non-isocratic conditions. The low-energy sites followed the classical linear solvent strength model (LSSM), but not the high-energy sites whose saturation capacity linearly decreased with ϕ. The general model was validated by comparing the experimental and simulated band profiles in gradient elution conditions, in linear and non-linear conditions, as well. The band profiles were calculated by means of the equilibrium-dispersive model of chromatography with a finite difference algorithm. A very good agreement was observed using steps gradient (Δϕ) from 0 to 50% methanol and gradient times tg of 20, 25, 30, 40, 60, 80 and 100 min. The agreement was still excellent for steps gradient from 5 to 45% (tg=25 min), 5 to 35% (tg=50 min), 5 to 25% (tg=50 min) and 5 to 15% (tg=50 min). Significative differences appeared between experience and simulation when the slope of the gradient (Δϕ/tg) became too strong beyond 3.3% methanol per minute. This threshold value probably mirrored the kinetic of arrangement of the G18-bonded chains when the methanol content increased in the mobile phase. It suggested that the chromatographic system was not in a full thermodynamic equilibrium state when very steep mobile phase gradients were applied.