烷基咪唑双-三氟甲烷磺酰亚胺室温熔盐物理化学性质的人工神经网络模型训练

ANN Model Training of Alkylimidazolium-bis (trifluoro-methylsulfonyl)imide RTMS' Physico-chemical Properties

利用分子路径指数矢量表示出室温熔盐阳离子的结构,采用误差反向传播人工神经网络模型对烷基咪唑双-三氟甲烷磺酰亚胺室温熔盐的物理化学性质进行了数值处理,给出了室温熔盐阳离子结构和它们物理化学性质之间的关系.室温熔盐结构与熔点的关系:熔盐熔点随着取代基团链的增长而下降;熔盐分子对称性越好,熔点越高;R2位上的取代基团使得熔盐熔点上升;R4上的取代基团使得熔盐熔点下降.室温熔盐结构与密度的关系:熔盐的密度随着N 烷基链长的增加而呈线性减小(R>0.99);取代基在R4位上的分子比在R2位上的熔盐密度低;如果R2上没有取代基团,R4上有取代基会使熔盐密度降低.室温熔盐结构与粘度的关系:R2位上的取代基团使得熔盐粘度上升,R4位上的取代基团会使熔盐粘度下降.室温熔盐结构与电导率的关系:R2位取代基团使得熔盐电导率下降;R2上无取代基团时,R4的取代基团使得熔盐电导率上升.这些规律对设计不同物理化学性质的室温熔盐分子具有很好的指导作用.

By using the expression of the molecular path vector(MPV)for the structure of the room temperature molten salts'(RTMS)cation and the model training of the error back propagation artifical neural net(ANN),the relation of RTMS cation's structure and their physico-chemical properties can be given:The melting points fall steadily with the length of alkyl chains. The more symmetrical RTMS cations,the higher melting points. The substituting group R_2 makes the RTMS' melting points fall and R_4 makes melting points increase. RTMS' densities fall in linear with the length of the alkyl substituents of imidazolyl nitrogen increase,(the linear correlation｜R｜>0.99). The density of RTMS with a R_4 substituent is lower than that of RTMS with an R_2 substituent. If there is no substituting group at the R_2 site,the R_4 will cause the RTMS' densities to fall. The substituting group R_2 decreases the conductivity and increases viscosity,while the substituting group R_4 brings on opposite results. These rules between RTMS structures and their physico-chemical properties are significatively applied to choose and synthesize RTMS for some special aims.