基于高阶基团贡献法与COSMO-SAC模型的溶剂设计方法

提出了一种基于高阶基团贡献法与类导体屏蔽片段活度系数模型（conductor like screening model-segment activity coefficient, COSMO-SAC）的计算机辅助溶剂设计方法（computer-aided molecular design, CAMD）。首先，基于高阶基团贡献法（higher-order group contribution, GC⁺）与COSMO-SAC模型构建GC⁺-COSMO方法，关联分子基团组合与表面屏蔽电荷密度分布σ-profiles, p(σ)]、分子空腔体积V_c，实现对二者的高通量预测；然后结合基于简化分子线性输入系统（simplified molecular input line entry system, SMILES）的异构体生成算法与GC⁺-COSMO方法实现CAMD技术对异构体的识别及性质区分；最后，通过目标函数与约束方程组成的混合整数非线性规划模型（mixed integer nonlinear programming, MINLP）来建立溶剂设计问题，进一步采用分解式算法优化求解，实现溶剂优化设计目标。基于以上模型和方法开展了狄尔斯-阿尔德（Diels-Alder, DA）竞争性反应溶剂设计，验证了提出的方法的可行性与有效性。

Solvent design method based on higher-order group contribution method and COSMO-SAC model

A computer-aided molecular design method (CAMD) based on the higher-order group contribution method (GC⁺) and the conductor like screening model-segment activity coefficient (COSMO-SAC) model is proposed. First, based on the GC⁺ method and the COSMO-SAC model, the GC⁺-COSMO method is constructed, which correlates the molecular group combination and the surface screening charge density distribution σ-profiles, p(σ)] as well as the volume of molecular cavity (V_c), so as to achieve high-throughput prediction of them. Then, the simplified molecular input line entry system (SMILES) based isomer generation algorithm and the GC⁺-COSMO method are combined to realize the recognition and property differentiation of solvent molecular isomers by CAMD technology. Finally, the solvent design problem is established by the mixed integer nonlinear programming model (MINLP) composed of objective function and constraint equations, and the decomposition algorithm is further used to optimize the solution to achieve the solvent optimization design goal. Based on the above models and methods, the Diels-Alder (DA) competitive reaction solvent design case is carried out, and the feasibility and effectiveness of the proposed method are verified.