The project “Accurate and efficient calculation methods based on the formalism of the renormalization group of DMRG density matrix for large molecules” received more than PLN 800 thousand funding for the research, which will be led by Professor Katarzyna Pernal from the Faculty of Technical Physics, Information Technology and Applied Mathematics. It will be implemented in cooperation with Professor Libor Veis from the Czech Academy of Sciences.
As Prof. Pernal explains:
Quantum-chemical computing now plays a full-fledged role in chemical research. Computational chemistry helps to analyse and explain experimental observations, assists the design of new materials, and allows predicting the properties of new molecules. Despite the huge progress in the development of computational methods, there is still a category of chemical problems, waiting for a reliable theoretical description. A long-known problem and challenge to existing methods are so-called quasidegenerative chemical systems, also known as systems with highly correlated electrons. Contrary to appearances, strongly correlated electrons are not exotic creations and occur in complex molecular systems of biological importance as well as in many modern materials of technological importance.
The aim of the proposed project is to develop and implement new DMRG-based computational tools that are suitable for extending strongly correlated molecules and their application to difficult problems such as the electron structure of biologically significant metal complexes transient. How the research team plans to solve the target is interpreted by Professor Katarzyna Pernal.
To achieve this goal, we plan to develop theoretical methods based on DMRG, including quantum embedding methods, which will significantly expand the possibilities of applying DMRG to large systems. The second goal of the project is fast computing algorithms for new methods and their efficient parallel computer implementation in a generally available computing package. The new computing tools will allow the limitations in accuracy and size of currently available DMRG-based methods to be broken. One of the world's fastest DMRG calculation programs has been created in Prof. Veis’ group, which will be used as the main solver for the project's tasks. The project will use L. Veis' unique experience in parallel programming of algorithms to solve quantumochemical problems.
Projects involving international cooperation were evaluated based on the Lead Agency Procedure (LAP), in which National Science Centre acted as lead agency. The Czech Academy of Sciences will be funded by the Czech Science Foundation (GACR) agency.