化工学报 ›› 2014, Vol. 65 ›› Issue (1): 131-142.DOI: 10.3969/j.issn.0438-1157.2014.01.017

• 催化、动力学与反应器 • 上一篇    下一篇

乙烯选择性齐聚取代PNP-Cr催化体系的二维定量构效关系

唐思扬, 刘振, 占兴稳, 程瑞华, 何雪莲, 刘柏平   

  1. 华东理工大学化学工程联合国家重点实验室, 上海 200237
  • 收稿日期:2013-07-01 修回日期:2013-10-29 出版日期:2014-01-05 发布日期:2014-01-05
  • 通讯作者: 刘振,刘柏平
  • 作者简介:唐思扬(1987-),女,博士研究生。
  • 基金资助:

    国家自然科学基金项目(21174037,21004020);引智计划项目(B08021);中央高校基本科研业务费专项资金项目。

2D-QSPR studies on substituting PNP-Cr catalyst systems for highly selective ethylene oligomerization

TANG Siyang, LIU Zhen, ZHAN Xingwen, CHENG Ruihua, HE Xuelian, LIU Boping   

  1. State Key Laboratory of Chemical Engineering, East China University of Science and Technology, Shanghai 200237, China
  • Received:2013-07-01 Revised:2013-10-29 Online:2014-01-05 Published:2014-01-05
  • Supported by:

    supported by the National Natural Science Foundation of China (21174037, 21004020), the Program of Introducing Talents of Discipline to Universities (B08021) and the Fundamental Research Funds for the Central Universities.

摘要: 1-己烯和1-辛烯是制备高性能聚烯烃产品的重要共聚单体,近年来,PNP型铬系乙烯选择性三聚/四聚催化体系成为本领域研究热点。针对实验研究报道的系列氮上取代PNP型铬系催化剂,采用二维定量构效关系(QSPR)方法结合密度泛函理论(DFT)建立了该催化体系1-己烯、1-辛烯选择性的线性回归模型。研究中同时采用了启发式方法和最佳多元线性回归方法,考察了自定义描述符和铬金属活性中心价态(包括一价和二价)对线性回归模型的影响。通过分析发现:基于DFT的自定义描述符的引入能够明显提高模型的相关性和稳定性;基于一价Cr的活性中心模型更适合关联乙烯三聚选择性,二价Cr的活性中心模型更适合关联乙烯四聚选择性;PNP-Cr骨架几何结构尤其是较小的PNP角度是获得较高1-己烯和1-辛烯选择性的关键。最后,根据最佳线性回归模型对新型PNP配体进行了预测,发现了9种新的PNP配体结构可能具有更高的1-辛烯或1-己烯/1-辛烯共选择性,为进一步的实验开发奠定了良好的理论基础。

关键词: 二维定量构效关系, 计算化学, 乙烯三聚/四聚催化剂, 分子模拟, 启发式方法, 最佳多元线性回归

Abstract: 1-Hexene and 1-octene are important comonomers for the synthesis of high performance polyolefins. Recently, Cr-bis(diphenylphosphino)amine (PNP-Cr) catalysts show the potential as excellent candidates for highly selective ethylene trimerization/tetramerization. In this work, a series of substituting PNP-Cr catalysts were studied by two-dimensional quantitative structure-property relationship (QSPR) method based on density functional theory (DFT) calculations. The heuristic method (HM) and best multi-linear regression (BMLR) were used for establishing the best linear regression models to describe the relationship between catalyst selectivity and its structure. Both Cr(Ⅰ) and Cr(Ⅱ) active site models for ethylene trimerization/tetramerization were considered. It was found that using self-defined descriptors from DFT calculations could increase the relativity and stability of the models. Monovalent Cr(Ⅰ) center was the most plausible active site for ethylene trimerization, while ethylene tetramerization was most possibly proceeded over divalent Cr(Ⅱ) active site. The skeleton structures of the PNP-Cr system especially a small PNP angle were crucial for achieving excellent catalytic selectivity. Nine new PNP ligands with high selectivity towards ethylene trimerization/tetramerization were predicted based on the best linear regression models providing a good basis for further development of novel catalyst systems with better performance.

Key words: 2D-QSPR, computational chemistry, ethylene trimerization/tetramerization catalysts, molecular simulation, heuristic method, best multi-linear regression

中图分类号: