CAREER: Many-body expansions for strongly correlated systems

点击翻译按钮获取中文摘要

Nicholas Mayhall, of Virginia Tech, is supported by an award from the Chemical Theory, Models, and Computational Methods program in the Division of Chemistry to develop new theoretical models and computer software capable of simulating, so-called "strongly correlated' molecules, which have been difficult to study with conventional quantum mechanical simulation approaches. These types of molecules constitute a highly important class of systems due to their role in a number of important processes, such as photosynthesis and homogeneous transition-metal catalysis. Strongly correlated systems are difficult to model, in part, due to the inability to divide up the molecule into small parts ("inseparability") to focus on only one part at a time. This requires simulations to be performed on an entire system, which is usually far too complicated for even the world's fastest supercomputers. Prof. Mayhall's research is focused on studying the extent of this inseparability in different types of molecular systems. This knowledge will then be used to design better theories and software which start from a "separated" initial description, reintroducing the effects of inseparability in a slow and controlled manner, to avoid the large growth in computational complexity while maintaining accurate results. This work has the potential to contribute to technologies involving energy production and chemical catalysis. Prof. Mayhall's outreach program involves developing two sets of educational modules targeting distinct demographics. The first is focused on inmates in regional detention facilities, and the second is directed toward female high-school students interested in STEM fields. The latter work is done in collaboration with established activities at Virginia Tech.The main goals of this project are two-fold, 1) obtain better understanding of inseparable systems, such as strongly-correlated molecules and collective excited states in molecular aggregates and materials, and 2) develop new computational methods capable of modeling these systems in a way which exploits approximate separability to obtain computational algorithms which can be effectively parallelized to run on large scale computing resources. To achieve these goals, Prof. Mayhall's research group is pursuing two distinct approaches. The first approach builds on recent work from the PI in which an approximate tensor decomposition is used to model ground and excited states in strongly correlated systems. The second approach draws on recent work in the field based on quantum embedding theories, with direct applicability to large scale parallelization.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.