Abstract
The various approaches developed for prediction and optimization in linear gradient elution in reversed-phase liquid chromatography are critically reviewed. These approaches concern both single-mode and multi-mode gradient elution, the latter involving either gradients related exclusively to the mobile phase composition or combined gradients of the mobile phase composition with flow rate and/or column temperature or combined gradients of flow rate and column temperature. The advantages and disadvantages of each method are discussed and special attention is devoted to the factors that affect the quality of the prediction and their impact to the optimisation. Finally, the fitting techniques and optimisation methods adopted in linear gradient elution are presented and the most effective algorithms used for this purpose are indicated and discussed.