Beyond the Orbital Picture: New Computational Approaches to Chemical Phenomena

化学家通过制造具有有用性质的分子和材料来解决问题。有时，一个化学问题可以通过找到一种有效和环保的方法来解决，这种方法可以从现成的原料中制造出已知的物质。计算化学软件可以让化学家在开始耗时的物质制造任务之前探索可能的合成途径。有时解决一个化学问题需要创造一种全新的物质。系统地测试所有可能的物质是不可能的，因此化学家使用定性的理论规则和定量的计算，在天文数字的可能性中导航，直到他们找到一种具有理想性质的物质。

Chemists solve problems by making molecules and materials with useful properties. Sometimes a chemical problem can be solved by finding an efficient and environmentally friendly way to make a known substance from readily available starting materials. Computational chemistry software lets chemists explore possible synthetic pathways before they embark on the time-consuming quest to make a substance. Sometimes solving a chemical problem requires creating an entirely new substance. It is impossible to systematically test all possible substances, so chemists use qualitative theoretical rules and quantitative computations to navigate through the astronomical number of possibilities until they find a substance with desirable properties. I develop computational methods and theoretical models for predicting chemical reaction pathways, describing chemical phenomena, and designing molecules with desirable properties. At the most fundamental level, I find new computational approaches for describing how electrons bind atoms into molecules. Conventional approaches to this problem are based on the orbital picture. In the orbital picture, electrons are assumed to move quasi-independently. Conventional approaches therefore fail to describe technologically important substances, like unconventional superconductors and spintronic materials, in which the electrons are strongly correlated. My research group is developing alternatives to the orbital picture. First, we are developing alternative computational methods for computing molecular properties; these methods can help chemists decide whether a proposed reaction pathway is plausible. Second, we are developing orbital-free tools for chemical insight. These tools provide intuitive rules that chemists can use to understand and explore new chemical phenomena. In addition, these tools can be used by machine-learning methods, where a computer (instead of a human chemist) gathers data about the features and properties of known molecules, discovers rules for organizing this data, and uses these rules to predict new molecules with desirable properties.