运用静电势图像分割方法预测超分子结合能  

Evaluation of binding strength for supramolecules through image threshold segmentation

作　　者：柳平英[1,2] 陈秋凤[3] 马晶[1] 

机构地区：[1]南京大学化学化工学院,南京210093 [2]景德镇陶瓷学院材料科学与工程学院,景德镇333403 [3]中南大学信息科学与工程学院,长沙410083

出　　处：《中国科学：化学》2016年第1期69-79,共11页SCIENTIA SINICA Chimica

基　　金：国家重点基础研究发展计划(2011CB808600);国家自然科学基金(21290192;21273102)资助

摘　　要：[2]-轮烷是由大环与分子链组成的两组分超分子体系,计算评估大环和链之间的结合强度,对于理解超分子识别和超分子体系可控设计尤为重要.本文借鉴计算机图像处理技术,通过全局阈值分割方法将大环和链的静电势图像分成两个片段,两者的静电势差值近似为[2]-轮烷体系的结合能.该计算方法可以直观简便地预测超分子体系的结合能.基于静电势的图像阈值分割方法计算的结合能定性符合实验测定的结合常数和密度泛函理论(DFT)计算的结合能.对于不同分子链和大环分子组成的多种[2]-轮烷体系的对比研究表明,大环与链的结合强度与结合位点的给质子能力、形成氢键的数目、大环的尺寸以及大环与链之间?-?堆积作用相关.[2]Rotaxane is composed of macrocyclic compound （macrocycle） and an axle-shaped molecule （thread）. The evaluation of macrocycle-thread binding strength is important for understanding the specific recognition and controllable design of molecular devices. With the aid of computer image processing technology, the electrostatic potential graphs of supramolecules were segmented to two fragments. And the binding strength of [2]rotaxane was approximated by the electrostatic potential difference of two fragments. The method provides an alternative way to predict binding strength for supramolecular systems. The predicted relative binding strengths of [2]rotaxanes are in good agreement with the experimental association constants and DFT-calculated binding energies. Further comparison on the binding manners between different threads and macrocycles indicates that the binding strength of [2]rotaxane is related to the strength of H donor, the number of H-bonds, and the size of macrocycle and the π-π stacking.

关 键 词：-轮烷 结合能 图像分割 静电势 DFT 

分 类 号：O641.3[理学—物理化学]