Lipophilicity data for some preservatives estimated by reversed-phase liquid chromatography and different computation methods

The chromatographic behavior of some preservatives was performed by reversed-phase high-performance liquid chromatography on C18 (LiChroCART, Purosphere RP-18e), C8 (Zorbax, Eclipse XDB-C8), CN100 (Säulentechnik, Lichrosphere) and NH2 (Supelcosil LC-NH2) columns. The lipophilicity estimated for the first time on the first three columns are comparable and very well correlated. The mobile phase was a mixture of methanol–water (0.1% formic acid) in different volume proportions from 40% to 60% (v/v) for RP-C18, RP-C8 and RP-CN100 column (exception for parabens on RP-C8 column—the methanol concentrations being from 55% to 65%) and from 30% to 50% (v/v) for RP-NH2. Highly significant correlations were obtained between different experimental indices of lipophilicity (log kw, S, φ0, mean of k and log k, and scores of k and log k corresponding to the first principal component) and computed log P values, and C8 column seems to be more suited for estimating the lipophilicity of the investigated compounds. These direct correlations offer a very good opportunity to derive powerful predictive models via Collander-type equations. The reliability of scores values as lipophilic indices is shown by their high correlation with the log Kow obtained using classical “shake-flask” technique, log kw and also some of the computed log P values. In addition, the results obtained applying PCA to the retention data may be used in interpreting the molecular mechanism of interactions between eluents and stationary phases with different polarity and to explain the chromatographic behavior of compounds. Finally, the “congeneric lipophilicity chart” described by the scores corresponding to the first principal component has the effect of separating compounds from each other more effectively from congeneric ((dis)similarity) point of view. The parabens and tert-butylhydroquinone appeared to be the most lipophilic preservatives.