Errors in the measurement of retention indices on packed columns

Summary

Variations in retention index have been determined for several probes on both polar and non-polar phases as a function of sample size and also of the ratios of sample to bracketing hydrocarbons in the sample mixture. Significant effects were found in both cases, these effects being much greater on Carbowax 20 M than on the nonpolar phases used. It is suggested that sample-bracketing hydrocarbon mixtures should always be 1∶1∶1 by volume, or that a series of samples would provide retention data which could be extrapolated to zero sample concentration. The sample size effect on retention is such that improved methods of measurement are necessary, reliance upon the nominal volume of 1μl can lead to variations unacceptable in published data. It is further concluded that the only measurement of column dead time of sufficient accuracy is the actual measurement of the peak arising from air introduced along with the sample.