Comments on the paper "Characterization of stationary phases by a linear solvation energy relationship utilizing supercritical fluid chromatography" by C. R. Mitchell, N. J. Benz, S. Zhang.

In a recent paper published by Mitchell et al. in this journal, some results obtained in supercritical fluid chromatography and interpreted with the solvation parameter model to characterize interactions for "novel stationary phases" were surprising to us. Indeed, we had already published results for most of the stationary phases reported, but, except for polar phases, our results were not in agreement with those, despite the use of identical mobile phases in both studies. These data were disturbing because they suggest that supercritical fluid chromatography is always a normal-phase mode, while we have shown that it is a reversed-phase mode when working with non-polar stationary phases. In the process of establishing the reason for the differences between our works, we examined several different factors. This paper deals with practice of linear solvation energy relationships: choice of dead-volume marker, choice of test-solutes to adequately probe the possible interactions and appropriate column length for characterization of chromatographic systems with highly eluting mobile phases are discussed. The importance of control experiments to validate retention models and confirm their accordance with the chromatographer's experience is evidenced. Recommendations for good linear solvation energy relationship practice are suggested in order to avoid the publication of results leading to erroneous conclusions.