聚甘油脂肪酸酯QSPR的研究

Study on QSPR of the polyglycerol fatty acid esters

采用量子化学MOPAC-AM1和单点能方法计算聚甘油脂肪酸酯的分子结构参数,然后用逐步线性回归方法建立聚甘油脂肪酸酯HLB值的定量结构性质(QSPR)模型,所得的预测模型中包含4个参数[单位质量分子所含氧原子数Xo、生成热fHm、电子能Ee和水合能Eh,预测值及外部检验的复相关系数(R2)和标准偏差(SD)分别为0.9553、0.73722和0.9678、6.34426。结果表明,量子化学方法计算简单,对聚甘油脂肪酸酯结构的表征能力较强,所建定量结构性质模型具有能较好的预测能力和较强的稳健性,并在一定程度上阐明了聚甘油脂肪酸酯HLB值与其分子结构之间的关系。

Semi-empirical quantum chemical AM1 method and single point method implemented in MOPAC software were used to calculate the molecular structure parameters of 77 polyglycerol fatty acid esters. Based on stepwise multiple regession （SMR）, the obtained parameters were taken as theoretical descriptors to establish the quantitation structure-property relationship （QSPR） models for predicting HLB values of polyglycerol fatty acid esters. The predicted model included 4 parameters, namely the oxygen atom number contained in each unit mass members ）to, heat of formation A/Hm, electronic energy E~ and hydration energy Eh. The correlation coefficient （R2） was 0.9553 and standard deviation （SD） was 0.73722, the appearance test R2 was 0.9678 and SD was 6.34426, respectively, for the model. The results suggest that quantum chemistry methods have multiple advantages, such as easy to get and good structural charaeterizationability for study on polyglycerol fatty acid esters. The established QSPR model possess good predictive ability and stability. To some extent, it can tell the relations between HLB values and the molecular structure of polyglycerol fatty acid esters.