基于QSAR方法的离子液体毒性预测  

QSAR for predicting toxicity of imidazolium-based ionic liquids

作　　者：赵永升[1] 黄莺[1] 赵继红[2] 张香平[1] 

机构地区：[1]中国科学院过程工程研究所,北京100190 [2]郑州轻工业学院,河南郑州450002

出　　处：《化工学报》2014年第5期1616-1621,共6页CIESC Journal

基　　金：国家重点基础研究发展计划项目(2013CB733506);国家自然科学基金重点项目(21036007)~~

摘　　要：离子液体因其极低的蒸气压、独特的溶解催化性能而备受关注,在许多领域展现出良好的应用潜力。但实现离子液体的工业化,离子液体的潜在毒性及环境影响也是重要的指标。针对离子液体的毒性性质,采用启发式算法筛选了5个影响离子液体毒性的关键参数,以离子液体对大肠杆菌(E.coil)毒性数据MIC值作为研究对象,建立了40种咪唑类离子液体的定量结构-活性相关(QSAR)预测模型,使用外部测试集对模型进行了外部验证。结果表明,该模型具有良好的可靠性,可用于咪唑类离子液体的毒性预测。Due to the unique physical and chemical properties, ionic liquids （ILs） have shown great potential for industrial applications. However, their toxicity and potential environmental impact should be clearly understood. In this paper, five parameters selected by the heuristic method were used to study the minimum inhibitory concentration values of toxicity againstE.coli based on the quantitative relationship between toxicity and structure of 40 kinds of imidazolium ionic liquids by using the quantitative structure-activity relationships （QSAR） method. Test set was used to conduct the external validation. This model with good reliability could be used to predict the toxicity of imidazolium-based ionic liquids.

关 键 词：离子液体 毒性 启发式算法 预测 

分 类 号：X174[环境科学与工程—环境科学]