A. Felinger et G. Guiochon, MULTICOMPONENT INTERFERENCES IN OVERLOADED GRADIENT ELUTION CHROMATOGRAPHY, Journal of chromatography, 724(1-2), 1996, pp. 27-37
Citations number
18
Categorie Soggetti
Chemistry Analytical","Biochemical Research Methods
A theoretical study of multicomponent interferences has been performed
by calculating band profiles in nonlinear overloaded gradient elution
chromatography. The separation of binary and ternary mixtures has bee
n modeled by means of a finite difference algorithm similar to the Cra
ig mechanism, in the case when the adsorption behavior of the mixture
components is accounted for by Langmuir competitive isotherms. The inf
luence of the loading factor, the composition of the sample, and the g
radient steepness have been investigated. In most cases, the profiles
and the band interactions are qualitatively similar to those obtained
under isocratic conditions. Although the profile fronts of the individ
ual bands are very sharp, nearly vertical, the separation of closely e
luted bands of 'parallel' or 'divergent' solutes cannot be improved si
gnificantly by gradient elution. Obviously, their retention time can b
e reduced considerably, which may improve the production rate although
column regeneration must be carried out after each run, which increas
es the cycle-time. The elution profiles of impurities and trace compon
ents is usually very similar in gradient elution and in isocratic elut
ion. This means that the recovery yield is not changed significantly,
regardless of the gradient steepness. The most interesting results and
the only one which is not observed under isocratic conditions, are ob
tained for the separation of 'convergent' solutes. In this case, the e
lution order of the components may change. This results in a kind of p
eak splitting, a fraction of the impurity eluting before the major com
ponent, the rest forming a flat profile spread dong the profile of the
major component. This phenomenon can be controlled or eliminated by a
djusting the loading factor of the sample and the gradient steepness.