Summary
Computational materials science, using ab initio simulations and high-performance computing, is expected to play a key role in realizing the vision of ‘materials by design’. However, the goal to discover game-changing materials with scientific and industrial relevance requires highly accurate ab initio methods for excited state as well as ground state properties of atoms, molecules and solids. So far, due to the computational complexities involved, methods with systematically improvable accuracy for condensed matter systems, such as coupled-cluster theories, are mostly limited to the study of ground state properties in the clamped-nuclei approximation. This ambitious proposal aims at inducing a computational paradigm shift in the study of vibrational and optical properties of real materials by implementing a multitude of novel methods. On the one hand, we propose to reduce the computational cost of time-dependent equation-of-motion coupled-cluster theory by several orders of magnitude compared to existing approaches. On the other hand, coupled-cluster atomic forces will be implemented for machine-learning force fields in the Gaussian approximation potentials framework. Together, the proposed methods have the potential to achieve an unprecedented level of accuracy and system size for the prediction of a wide range of material properties including optical spectra and phonon frequencies. We seek to employ the newly developed approaches to resolve a number of long-standing discrepancies between theoretical predictions and experimental findings for dynamic properties of defects, molecular crystals and layered materials. These carefully selected systems highlight key problems of currently available ab initio methods and novel approaches that go beyond the state of the art will have an enormous impact in all areas of physics, chemistry and computational materials science.
Unfold all
/
Fold all
More information & hyperlinks
| Web resources: | https://cordis.europa.eu/project/id/101087184 |
| Start date: | 01-08-2023 |
| End date: | 31-07-2028 |
| Total budget - Public funding: | 1 999 288,00 Euro - 1 999 288,00 Euro |
Cordis data
Original description
Computational materials science, using ab initio simulations and high-performance computing, is expected to play a key role in realizing the vision of ‘materials by design’. However, the goal to discover game-changing materials with scientific and industrial relevance requires highly accurate ab initio methods for excited state as well as ground state properties of atoms, molecules and solids. So far, due to the computational complexities involved, methods with systematically improvable accuracy for condensed matter systems, such as coupled-cluster theories, are mostly limited to the study of ground state properties in the clamped-nuclei approximation. This ambitious proposal aims at inducing a computational paradigm shift in the study of vibrational and optical properties of real materials by implementing a multitude of novel methods. On the one hand, we propose to reduce the computational cost of time-dependent equation-of-motion coupled-cluster theory by several orders of magnitude compared to existing approaches. On the other hand, coupled-cluster atomic forces will be implemented for machine-learning force fields in the Gaussian approximation potentials framework. Together, the proposed methods have the potential to achieve an unprecedented level of accuracy and system size for the prediction of a wide range of material properties including optical spectra and phonon frequencies. We seek to employ the newly developed approaches to resolve a number of long-standing discrepancies between theoretical predictions and experimental findings for dynamic properties of defects, molecular crystals and layered materials. These carefully selected systems highlight key problems of currently available ab initio methods and novel approaches that go beyond the state of the art will have an enormous impact in all areas of physics, chemistry and computational materials science.Status
SIGNEDCall topic
ERC-2022-COGUpdate Date
31-07-2023
Images
No images available.
Geographical location(s)
