论文部分内容阅读
采用分子电性距离矢量(MEDV)表征酯类化合物的分子结构,同时运用多元线性回归技术建立了81个酯类化合物,在2种固定相(Carbowax1540和Cqualane)上,气相色谱保留指数(RI)与MEDV的定量结构-色谱保留关系(QSRR)模型。在2种固定相上的QSRR模型的建模计算值复相关系数(Rcum)、留一法(leave-one-out)交互校验复相关系数(QCV)分别为0.9968和0.9958(Carbowax1540);0.9935和0.9908(Cqualane),结果表明MEDV能理想地表征酯类化合物的分子结构,所建模型具有良好稳定性和预测能力。
The molecular structure of ester compounds was characterized by MEDV. At the same time, 81 ester compounds were established by using the multiple linear regression technique. The gas chromatography retention index (RI) of the two stationary phases (Carbowax 1540 and Cqualane) Quantitative Structure-Chromatographic Retention Relationships (QSRR) with MEDV. The calculated Rcum and leave-one-out cross-validation complex correlation coefficients (QCV) for the QSRR model on the two stationary phases were 0.9968 and 0.9958 (Carbowax 1540), respectively; 0.9935 And 0.9908 (Cqualane). The results show that MEDV can ideally characterize the molecular structure of esters and the model has good stability and predictability.