Density-functional theory is the most widely applied quantum-chemical method. The main idea is that any quantity, including the energy due to all complicated pairwise electron-electron interactions, can be estimated from the total electron density alone.
In recent years, a very promising extension makes use of the kinetic energy density too. This project is based on a further extension utilizing a tensorial quantity akin to the stress-eenergy tensor [J. Chem. Phys. 149, 144109 (2018); https://doi.org/10.1063/1.5041931], encoding much richer chemical information. Instead of model input data being a scalar density assigning a single number to each location in space, the input is extended to a tensor density assigning a 3x3 matrix to each point.
The goal of the project is to build models of how interaction energies depend on this tensor density. This will involve a mix of physical modelling and machine learning as well as development and implementation of supporting functionality within quantum-chemical software packages.
Requirements
- MSc in physics or chemistry, preferably in computational physics/chemistry.
- Candidates with documented experience in scientific programming, quantum chemistry or quantum mechanical modelling, and experience from machine learning will be prioritized.
Supervisors
Call 2: Project start autumn 2022
This project is in call 2, starting autumn 2022.