CAS OpenIR  > 研究所(批量导入)
基团贡献法物性估算程序的开发
Alternative TitleThe Development of Physical Property Estimation Program by Group Contribution
冷建飞
Subtype硕士
Thesis Advisor赵月红 ; 温浩
2012-05-01
Degree Grantor中国科学院研究生院
Degree Discipline应用化学
Keyword基团贡献法   物性数据   分子结构拆解   基团优先级   物性估算
Abstract化合物的物性数据是化学化工学科重要的基础数据,化学数据库中收集整理了大量的实验数据,这为物性数据的获取提供了极大的便利,但因实验测量难度较大、实验成本较高以及新物质的数量快速增长等问题,物性数据不足仍是普遍存在的情况。目前,基团贡献法物性估算是解决这一问题的有效手段。基团贡献法是一类将分子结构片段(基团)对物性的影响(贡献值)加和来预测化合物的性质的方法,在化工计算中应用广泛。为了简便、快速的利用基团贡献法进行物性估算,现已开发了许多基团贡献法物性估算程序,物性估算程序的核心是分子结构的自动拆解,然而在分子结构自动拆解中,不同的基团匹配顺序有可能造成拆解的结果出现多解或歧义,有时甚至会导致分子结构拆解失败,导致物性估算错误或无法进行。 针对这一问题,本文分析了基团贡献法定义基团间的相互关系,研究了不同重叠情况下,各类基团在分子结构拆解过程中的先后顺序及其对拆解结果的影响,并以此为基础设计了分子结构拆解的基团优先级方法。以Joback和C-G等典型基团贡献法为例对多种分子结构进行了拆解测试,结果表明基团优先级方法能够有效地解决分子结构自动拆解的多解问题。 在以上研究的基础上,本文采用SMARTS编码建立了几种基团贡献法的基团组按优先级排序的编码表,采用CACTVS化学工具库和Tcl语言开发了物性估算程序,并以此为基础实现了Joback和C-G两种物性估算方法。与文献报道的同类方法估算结果相比,本文开发物性估算程序获得相同精度的结果。测试结果进一步表明了基团优先级方法的正确性及开发物性估算程序的可靠性。此外,程序还为新增基团贡献法的实现留下了接口。 考虑到物性估算程序与化学主题数据库的集成及其对外服务的需要,本文开发了物性估算Web Service,设计了基于XML的数据交换格式,在此基础上实现了化学主题数据库的物性估算功能,进一步提高了化学主题数据库服务的能力和水平。 关键词:基团贡献法;物性数据;分子结构拆解;基团优先级;物性估算
Other AbstractPhysical properties are important to scientific research and industrial applications. Various chemical databases make it easy to obtain the physical properties data. However, because of the high cost of the experimental measurement and fast increasing number of substances, the phenomena of no or incomplete properties data of some substances is very common. To overcome this obstacle, group contribution method is widely used to estimate the properties. The principle of a group contribution method is the determination of a component property by summing up the group contribution. To making better use of Group contribution method, many property estimation programs using group contribution method have been developed. The key step for such program is molecular disassembly according to the predefined groups for a method. But for a molecule, molecular disassembly may provide different solutions using different group match sequences, which may lead to unreliable estimation result. In this paper, multi-solution of molecular structure automatic disassembly in group contribution method was discussed. The effects of different group matching sequences on disassembly results were studied, and a method was developed to make sure accurate molecular disassembly. In this method, group priority was defined according to the overlap relationship between the group structures and was used to sequence the molecular groups. The tests by Joback and C-G methods using diverse molecular structures show that the group priority can help to obtain correct disassembly solution. Based on above result, group tables for some typical methods, i.e. Joback, C-G, were built using SMARTS code, and an estimation program was developed using Tcl and CACTVS. And it was tested by comparing the obtained result with literature data of same method. The results show that the program could provide reliable result as expected. Furthermore, a web service of the developed Group Contribution Method was built and was integrated in Chemistry database system. The test shows that the group contribution method will help the Chemistry Database provide better user experience in the case of queried data is unavailable. Key words:Group contribution; Physical properties; Molecular structure disassembly; Group priority; Property estimation
Pages79
Language中文
Document Type学位论文
Identifierhttp://ir.ipe.ac.cn/handle/122111/8308
Collection研究所(批量导入)
Recommended Citation
GB/T 7714
冷建飞. 基团贡献法物性估算程序的开发[D]. 中国科学院研究生院,2012.
Files in This Item:
File Name/Size DocType Version Access License
基团贡献法物性估算程序的开发.pdf(1320KB) 限制开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[冷建飞]'s Articles
Baidu academic
Similar articles in Baidu academic
[冷建飞]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[冷建飞]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.