The terpenes released from Mosla chinensis Maxim and fresh flowers of Chimonanthus praecox were investigated by means of solid-phase microextraction (SPME) coupled to gas chromatography-mass spectrometry (GC-MS). The structure of terpenes was identified by MS and retention index, by which 17 monoterpenes and 30 sesquiterpenes were identified in this study. The quantitative structure and retention relationships of these monoterpenes and sesquiterpenes were investigated by GFA models within Cerius2 molecular modeling program package and then validated by F-test and predictive-ability test. The retention indices of several terpenes randomly selected were then predicted by QSRR models. The results demonstrated that the models had perfect predictive ability. This investigation provided an effective method for predicting the retention indices of monoterpenes and sesquiterpenes. It is worthwhile in future study for GC-MS qualitative analysis.

Key words: solid-phase microextraction, gas chromatography-mass spectrometry, retention index, quantitative structure-retention relationship