选择20种脂肪族及杂环化合物作为模型污染物,分别测定了三种不同纳滤膜对脂肪族及杂环化合物的截留率.结果表明,脂肪族及杂环化合物的截留率受到分子的分枝结构、环状结构、酸性和膜特性的影响:对于同分异构体,分枝结构愈多,截留率愈高;环状有机物与分子量相近的直链有机物相比,截留率明显偏高;脂肪酸的截留率高于绝大多数醇类、醚类和酮类等不离解的化合物;孔径小、荷电量大的纳滤膜截留率更高.通过基于遗传算法结合偏最小二乘回归法(GA-PLS)建立了纳滤膜对脂肪族及杂环化合物截留率的定量构效关系模型,通过分析回归方程,可以看出膜与脂肪族及杂环化合物之间的电性作用并不是影响截留率的主要因素,而分子形状和大小对截留率的影响很显著.
The rejections of 20 types of aliphatic and heterocyclic compounds by three nanofiltration membranes were tested by experiments, and the results showed that the rejections were affected by the molecular branch structure, ring structure, acidity and properties of nanofiltration membranes ; the isomeric compounds which had more branch structures had higher rejections; while the molecular weights were similar, the rejections of compounds which had ring structures were higher than those having chain structures ; the rejections of fatty acids were higher than those of most of alcohol, aether and ketone ; the membrane which had smaller pore radius and more charges a higher rejection .Quantitative structure-property relationships (QSPR) were developed by PLS combined with genetic algorithm. Based on the regression equations, the most important factors were molecularspace structure parameters, however, the electrical property parameters were not selected in the models.