Newsroom
We are delighted to announce that our member Fred Hamprecht, head of the Scientific AI lab at the Interdisciplinary Center for Scientific Computing (IWR) and principal investigator of STRUCTURES, has been awarded one of the European Research Council's prestigious ERC Advanced Grants.
The grant, endowed with € 2.5 million in funding over five years, will support his project Learning Orbital-Free Density Functional Theory (LearningOFDFT). The aim of this project is to develop new methods for the precise and stable calculation of molecular energies based on electron densities, using what is called an orbital-free approach.
Orbital-free approaches provide a major simplification of the quantum-mechanical calculations required to study molecular systems. Understanding molecular structure and reactivity requires precise calculations of molecular energies. Traditionally, this relied on a wave-function description, which depends on the position of every electron in space. While highly accurate, this becomes computationally demanding as molecular systems grow in size and complexity. In 1964, Walter Kohn created a formalism that promised to eliminate the need to account for every single electron's position, instead requiring only the average density of electrons in the space. A formal proof of this claim earned Walter Kohn the Nobel Prize in 1998. Yet, a central challenge remains unsolved: an explicit formula for the kinetic energy functional, the central object of the theory, has remained unknown to this day.
Originally started as an Exploratory Project of STRUCTURES, Fred Hamprecht and his collaborators, notably Roman Remme, pioneered a machine learning method to solve this long-standing problem by learning the mapping from electron density to molecular energy – using physics-inspired AI methods. This work has since developed into a new research line of the cluster. Through close interdisciplinary cooperation with chemists and physicists, it was possible to create the STRUCTURES-25 method, which obtained both chemically precise energies relative to the reference and stable, practical optimization of electron density.
Building on this success, Prof Hamprecht and his team will use the ERC funding to develop fast and precise predictions for increasingly complex systems – thereby contributing to solving the decades-old challenge of finding an orbital-free density functional. If successful, this research promises to open new opportunities for applied research, from the analysis of biomolecular reactions to the development of environmentally friendly materials.
About the ERC Advanced Grant
The highly competitive ERC Advanced Grant is awarded by the European Research Council to support outstanding established researchers in the pursuit of groundbreaking, ambitious research projects that could lead to major scientific breakthroughs. The grants are part of the European Union’s Horizon Europe programme, with a maximum funding period of five years.
Further information: