The present study intends to develop the quantitative structure–retention relationship (QSRR) models to predict the retention factor of some alkyl-benzenes on micro-dispersed sintered nano-diamond (MSND) and silica-gel stationary phases as new and common stationary phase, respectively, in normal-phase HPLC. Genetic algorithm–multiple linear regression (GA–MLR) method employed for implementation of QSRR models. The square of correlation coefficient (R2) was 0.997 and 0.961, and the SE was 1.06 and 1.93, respectively, for the training and test sets of GA–MLR model on MSND stationary phase. In addition, the R2 was 0.994 and 0.984, and the SE was 0.01 for both the training and test sets of GA–MLR model on silica-gel stationary phase. The statistical parameters and the result of validation tests of these models confirm the fitness, robustness, and predictability of developed QSRR models. The mean effect analysis on the best models introduced the polarizability as the most significant factor that effects on the retention factor of solutes on both studied stationary phases. This similarity confirms the suggestion of MSND as a common stationary phase for normal-phase HPLC. The developed models enable to predict the retention factor of other alkyl-benzenes on both MSND and silica-gel stationary phases on their applicability domain.
