Accurate Many-Body Electronic-Structure Methods for Extended Molecular Assemblies

弗吉尼亚理工大学的Edward Valeev获得了化学部化学理论，模型和计算方法计划，材料研究部凝聚态物质和材料理论计划以及计算和数据支持科学与工程计划（CDS E）的支持，以从根本上提高对分子和材料中电子的理解。 电子是将分子和材料结合在一起的通用“胶水”，因此决定了它们最相关的性质。预测电子的行为对化学和材料科学具有重要意义。在这个项目中开发的新的计算方法将使科学家能够以前所未有的精度看到分子中电子舞蹈的微妙细节。新模型将有助于解释一些新的迷人材料如何能够将二氧化碳（一种主要的温室气体）化学转化为有价值的产品。该项目的长期目标是在计算机上设计-而不是通过试错来发现-新的化学过程和新材料，以帮助改善我们的日常生活。 Valeev和他的研究小组开发的软件免费提供给研究界。该项目的技术重点是新的，有效的计算方法，用于描述大分子组件（100-500个原子）中基态和激发态的电子，其置信度仅适用于小分子（5-10个原子）。这些目标是通过使用更有效的电子状态的数值表示来实现的，例如显式相关的数值表示，它比标准方法更有效地模拟电子对的短程相关性，以及利用电子相关性的短视性的缩小尺度表示。这些想法也将被用来将“弱”（短程）电子相关耦合到描述“强”（长程）相关的模型中，例如组态展开或重整化群方法。这项工作有机会通过使其成为目前占主导地位的密度泛函方法的实用替代品，从而大大改变精确电子结构方法的最新技术水平。 在这项研究中开发的算法和方法将导致开源的公开可用的计算机软件，用于大型分子系统的量子力学建模。

Edward Valeev of Virginia Tech is supported by an award from the Chemical Theory, Models and Computational Methods program in the Chemistry Division, the Condensed Matter and Materials Theory program in the Division of Materials Research and the Computational and Data-Enabled Science and Engineering Program (CDS&E) to develop radically improved understanding of electrons in molecules and materials. Electrons are the universal "glue" that holds together molecules and materials and thus determine their most relevant properties. Predicting the behavior of electrons is of fundamental importance to chemistry and materials science. The new computational approaches being developed in this project will allow scientists to see with an unprecedented degree of precision the delicate details of the dance of electrons in molecules. The new models will then help to explain how some new fascinating materials are able to chemically transform carbon dioxide, a major greenhouse gas, into valuable products. The long-term objective of this project is to design on the computer -- rather than discover by trial-and-error -- new chemical processes and new materials that help improve our everyday lives. Valeev and his research group develop software that is freely available to the research community.The technical focus of this project is on new, efficient computational methods for describing the electrons in their ground and excited states in large molecular assemblies (100-500 atoms) with the level of confidence only accessible for small molecules (5-10 atoms). These objectives are achieved by using more efficient numerical representations for electronic states, such as explicitly correlated numerical representations that model short-range correlations of pairs of electrons much more efficiently than standard approaches, and reduced-scaling representations that take advantage of the short-sightedness of the electron correlations. These ideas will be also leveraged to couple "weak" (short-range) electron correlations into models describing "strong" (long-range) correlations, such as configuration expansion or renormalization group approaches. This work has a chance to substantially change the state of the art of accurate electronic structure methods by making them a practical alternative to the currently dominant density functional methods. The algorithms and methods developed in this research will result in open-source publicly-available computer software for doing quantum mechanical modeling of large molecular systems.