新型拓扑指数和神经网络研究咪唑啉衍生物的缓蚀性能

Research on Corrosion Inhibition Efficiency of Imidazoline Derivatives With Novel Topological Indices and Artificial Neural Network

基于分子图论提出了一种用于表征咪唑啉衍生物分子局部化学微环境及原子杂化状态的新颖结构描述子,即电性连接性指数0Kv、1 Kv和咪唑啉环非氢原子平衡总电荷分数MCI,用于研究15种咪唑啉类缓蚀剂抗CO2、H2S腐蚀性能的定量构效关系。结果表明,模型计算值、留一法交互检验预测值的复相关系数分别为0.9764、0.9546,所建模型具有良好的稳定性和外部预测能力;同多元回归方法比较,运用人工神经网络法的复相关系数为0.9848,结果更精确;增加咪唑啉环上取代基长度、减小分子的支化度和降低咪唑环非氢原子平衡总电荷分数能显著提高咪唑啉缓蚀剂的缓蚀效率。

Based on the molecular graphic theory,novel molecular structure descriptors of electrical connectivity index 0^K^v,1^K^v and the imidazoline ring of non-hydrogen atoms balance total charge fraction（MCI） were proposed for expression of local chemical microenvironment and atomic hybridation state and used to quantitatively estimate the structure-activity relationship（QSAR） of corrosion behavior towards H2S and CO2 for fifteen imidazolines.The results showed that correlation coefficients of modeling calculated and leave-one-out cross-validation（LOO-CV） predicted values were 0.9764 and 0.9546,respectively.The QSAR model was of good stability and external predictive capability.For the same purpose,artificial neural networks were applied to get more improved result with the correlation coefficient of 0.9848.It is proposed that increasing substitution length of the imidazoline ring,reducing the molecular branching and lowering the imidazoline ring of non-hydrogen atoms balance total charge fraction had a significant effect for enhancing the corrosion inhibition efficiency of imidazolines.