论文部分内容阅读
用Chemoffice建构和优化了90个饱和酯的分子结构,运用Dragon6.0计算了2337个结构参数,采用Minitab15进行逐步回归筛选出2个结构参数SpPosA_Dz(Z)和RDF020p,通过多元线性回归方法构建了饱和酯化合物在7种不同极性色谱固定相上的定量结构-色谱保留相关关系(QSRR)模型,模型方程的相关系数在0.99以上。在此基础上建立交互柱QSRR模型,考察色谱固定相极性参数麦克雷诺常数(Mp)对保留指数的影响。研究结果表明:包含固定相极性参数Mp的交互柱QSRR模型方程对饱和酯化合物在不同极性色谱柱上的色谱保留指数都能得到较准确的预测结果。
The molecular structure of 90 saturated esters was constructed and optimized by Chemoffice. The structure parameters of 2337 were calculated by using Dragon6.0. Two structural parameters, SpPosA_Dz (Z) and RDF020p, were screened by stepwise regression using Minitab15. Multiple linear regression methods were used to construct Quantitative structure-chromatographic retention correlation (QSRR) models of saturated ester compounds on seven different polar stationary phases were obtained. The correlation coefficients of the model equations were above 0.99. On this basis, an interactive column QSRR model was established to investigate the influence of the Michaelis-Menten constant (Mp), a polar parameter of chromatographic stationary phase, on the retention index. The results show that the chromatographic retention index of saturated ester compounds on different polar columns can be more accurately predicted by QSRR equation of reciprocal column containing Mp as stationary phase.