Gas Chromatography of Alkanes

The generation of a quantitative structure retention relationship (QSRR) model used to identifiy acyclic alkanes containing three to seven carbons is presented here as a complement to calibration-based experiments in instrumental analysis. Students are introduced to the concept of a topological index (TI) and to two TIs in particular, the Wiener index and the Randić molecular connectivity index, which relate the gas chromatographic (GC) retention of alkanes to their molecular structure in a QSRR model. A training set for each QSRR model is based on the retention of three normal alkanes (pentane, hexane, and heptane) on a thick film, non-polar, capillary GC column. Linear regression is applied to the logarithm of the adjusted retention time vs. calculated TI values for the training set in order to build each QSRR model. Of the two QSRR models examined, the Randić molecular connectivity index performed best by correctly identifying all normal alkanes and branched alkanes with either a single alkyl substituent, or with two methyl substituents located on the second carbon (i.e. 2,2-dimethylbutane and 2,2-dimethylpentane). Limitations in the Randić QSRR model in predicting the identity of highly branched alkanes is discussed with respect to the solvent accessible surface area-to-molecular volume (SASA/MV) ratio of branched vs. normal alkanes for each set of structural isomers. Branched alkanes with a SASA/MV ratio closest to that of the normal alkane isomer were successfully identified using the Randić QSRR model. Students use the Randić QSRR model in this experiment to identify normal and branched alkanes in refill canisters for butane torches.