Correlation equations for prediction of gas chromatographic separation of hydrocarbons on squalane

Summary The retention index of several hydrocarbon classes on squalane was found to be predictable (within two index units) with help of a physico-chemical index and a correction-structural number. A large number of structural elements were examined by means of correlation analysis and regression equations were obtained for calculation of the structural number. The multinomial regression equations were simplified by combining in one factor the important structural elements. A set of a limited number of selected representatives for every hydrocarbon class was found to give reliable estimations for the regression coefficient equations.