AB INITIO POTENTIAL ENERGY SURFACES FOR SMALL H-C-N-O COMPLEXES

小 H-C-N-O 配合物的 AB 从头计算势能面

基本信息

批准号：
7601482
负责人：
BASTIAAN BRAAMS
金额：
$ 0.03万
依托单位：
CARNEGIE-MELLON UNIVERSITY
依托单位国家：
美国
项目类别：
财政年份：
2007
资助国家：
美国
起止时间：
2007-08-01 至 2008-07-31
项目状态：
已结题

项目摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. In recent years, working with the Bowman group in the Emory University chemistry department, the proposer has constructed potential energy surfaces for a variety of small molecules and reaction complexes, for example CH5, CH5+, H3O2-, H4O2, H5O2+, C3H2, C3H3O, with other molecules in the pipeline [1]. (A potential energy surface is a representation of the electronic energy of a molecule as a function of the nuclear configuration. It is a key tool in the computational study of chemical reaction dynamics and molecular spectroscopy.) Our potential energy surfaces are fully invariant under permutation of like nuclei and they describe important reaction channels as well as the reaction complex or entire molecule. The ability to construct such potential energy (and dipole moment surfaces) in an almost routine manner by fitting to ab initio data represents a major advance over the state of the art of only two or three years ago. Now the proposer is working to develop a comprehensive library of fitted potential energy surfaces for all H-C-N-O molecules and reaction complexes that have at most 7 atoms total of which at most 3 are C or N or O; neutral or once charged; singlet, doublet, or triplet spin state. Using a multi-method approach, with many ab initio calculations (several * 10^4 for each of the larger complexes) done at not so high level of theory and fewer calculations done at high level of theory (e.g., rccsd(t) on an aug-cc-pvtz basis, or also mrci calculations), such a library appears well within reach. The core computational effort are the millions of ab initio calculations; about 10^7 total is estimated at a total cost of about 10^6 CPU hours on current hardware, and most of it embarrassingly parallel. The project is well underway using local resources at Emory, but in the end these do not suffice. Hence the request for a development allocation of 30,000 SUs for initial exploration of the TeraGrid environment. I expect to follow this up with a more substantial allocation request later. At present the Molpro quantum chemistry code is the key tool. In view of this and in view of the coarse parallellism of the project the most suitable TeraGrid resource appears to be the p690 system at NCSA. [1] http://www.chemistry.emory.edu/faculty/bowman/ab_initio.htm

该副本是利用众多研究子项目之一由NIH/NCRR资助的中心赠款提供的资源。子弹和调查员（PI）可能已经从其他NIH来源获得了主要资金，因此可以在其他清晰的条目中代表。列出的机构是对于中心，这不一定是调查员的机构。近年来，该提议者与埃默里大学化学系的鲍曼集团（Bowman Group）合作，为各种小分子和反应配合物（例如CH5，CH5+，H3O2-，H4O2，H4O2，H5O2+，C3H2，C3H2，C3H3O）构建了势能表面，并在管道中使用其他分子[1]。（势能表面是分子的电子能与核构型的函数的表示。它是化学反应动力学和分子光谱的计算研究中的关键工具。）我们的势能表面在核心置换率下完全不变，并且它们描述了重要的反应通道以及反应复合物或整个分子。通过拟合从头算数据，几乎常规的方式构建这种势能（和偶极矩表面）的能力代表了仅两三年前的艺术状态的重大进步。现在，提议者正在努力为所有H-C-N-O分子和反应复合物开发一个综合的势能表面库，最多具有7个原子，最多3个原子是C或N或O；中性或一次充电；单线，双线或三重旋转状态。使用多方法方法，在不那么高的理论水平上进行了许多Ab的计算（对于每个较大的复合物的几个 * 10^4），并且在高度理论上进行的计算较少（例如，在Aug-CC-PVTZ基础上进行RCCSD（t），或MRCI计算），这样的库就在范围内。核心的计算工作是数百万的从头计算。当前硬件的总成本约为10^6 CPU小时，总成本约为10^7，并且大多数是令人尴尬的。该项目使用Emory的本地资源正在进行中，但最终还不够。因此，要求开发30,000 SU的开发要求，以初步探索Teragrid环境。我希望以后再通过更实质性的分配请求进行跟进。目前，Molpro量子化学法规是关键工具。鉴于该项目的粗略相似之处，最合适的Teragrid资源似乎是NCSA的P690系统。 [1] http://www.chemistry.emory.edu/faculty/bowman/ab_initio.htm