| The retention time is one of most important index. With the purpose of optimization of GC separation and identification of eluted components, it is popular to predict retention time in temperature-programmed gas chromatography (TPGC). With aid of computer software and mathematical method, the retention time can be accurately calculated.In this paper, different kinds of retention models were used to predict retention time, based on non-linear plate theory, under three heating rates including 25?C/min, 15?C/min, 5?C/min.The crystallinity formula of polymer was carefully researched and innovatively used in predicting retention time in TPGC. Based on 3 isothermal data, the crystallinity model was established. There were two hypothesises: firstly, the relation of crystallinityχc and logarithm of capacity factor lnk'can be described by adjusted crystallinity formula; secondly, crystallinityχc has a linear relationship with temperature T in a certain temperature range. The monotonic lnk'-T curves derived from crystallinity formula can only reflect chromatographic retention behavior in certain range. Predicting the retention time of 30 samples in three heating rates, 95% of the relative errors between prediction value and experimental data are below±3%. The average relative errors of 30 samples are: 25?C/min, -1.73%; 15?C/min, -1.11%; 5?C/min, -0.25%. All errors are under control.As the number of experimental data used in establishing models decreases, the workload of isothermal pre-experiment is simplified. Investigate the prediction results of the cubic retention model based on 6 and 4 isothermal data, the square retention model based on 3 isothermal data, the linear retention model based on 2 isothermal data respectively and compare with the results of crystallinity model. It is proved that the letter is a excellent prediction method which gives consideration to the accuracy and experimental cost, meanwhile has outstanding advantages in application value and practical guiding significance. |
PDF Full Text Request