Chemometrics have found widespread use to develop selectivity optimization procedures in liquid chromatography. This approach allows the systematic development of optimal chromatographic methods. An essential step of any optimization strategy is a clear definition of the goal(s) of the process. The mathematical description of such a goal is called an optimization criterion. A number of such criteria have been evaluated in this work. We have demonstrated that different requirements ofchromatographers lead to different optima, necessitating the selection and adaptation of optimization criteria for specific problem. <BR> Many different criteria have been suggested in the literature. However, the selection of an adequate criterion and its correct application can be a problem for potential users, because they do not have sufficient information. In practice, the various criteria must be used with care. For example, some criteria may yield arbitrary results. Other ones can only be applied in ideal cases (Gaussian peaks of similar areas). <BR> In the present work, we have shown that the used of flexible criteria adapted to practical situations is a considerable improvement for optimization procedures. Indeed, most of the chromatographic separation problems encountered in practice are complex, due to the presence of non ideal peaks (asymmetrical peaks or peaks of vastly different areas) or to the existence of irrelevant peaks, which may interfere with the separation of the relevant ones. More complex resolution expressions have to be applied to describe the quality of the separation correctly. We have also defined criteria dealing with limited optimization problems (i.e. situations in which a limited number of solutes in a complex matrix (biological samples, environmental samples, etc.). Therefore, in our opinion, an important requirement for optimization criteria is their ability to deal with limited optimization. <BR> some criteria defined in the context of limited optimization (normalized resolution products) have the additional advantage of promoting the coelution of irrelevant peaks and of favouring the separation of the solutes into groups (ideally, the relevant peaks are equally distributed over the chromatogram and the irrelevant peaks are confounded). The behaviour opens the way towards the optimization of multi-dimentional separations (LC-LC, LC-GC …) and, possibly, group-type separations. To achieve these objectives, the subject needs to be investigated more profoundly. <BR> Another fundamental aspect of this work has been to consider the robustness of a method with respect to the experimental conditions as an explicit objective of the optimization procedure. We have defined robustness criteria, which can be included in optimization strategies. By considering robustness at an early stage of method development, the chance of failure during the validation process is reduced. This new concept may be systematically integrated in any optimization procedure. <BR> With this work, we hope to have contributed to a better use of optimization criteria and to a more successful use of optimization procedures. <BR> the examples od selectivity optimization discussed here concern the reversed-phase mode, which is the most common from of liquid chromatography (ion-pair chromatography, ion-exchange chromatography, etc.) and to numerous other analytical techniques. For example, the continuing development of capillary electrophoresis leads to interesting perspectives. <BR> From a more general point of view, the intensive research on chemometrics is resulting in the continuous development of new concepts and increasingly flexible strategies (experimental designs, mathematical models better adapted to practical situations, etc.). The availability of efficient and user-friendly software will contribute to a more systematic use of this kind of approach. In conclusions, we think that chemometrics should be progressively integrated in laboratories where the concept of quality is an important concern. The introduction of the optimum robustness as an objective of optimization strategy has been a priority for us. A more general study of this subject is a logical follow-up of the present work.
Affiliations
UCLouvainMD/FARM/CHAM - Unité d'analyse chimique et physico-chimique des médicaments
Citations
APA
Chicago
FWB
Vanbel, P. (1997). Les critères d’optimatisation en chromatographie liquide-liquide à haute performance. https://hdl.handle.net/2078.5/111297