Group Name:
Project B05: Multi-resolution methods including quantum chemistry, force fields, and hybrid particle-field schemes
URL Slug:
project-b05-multi-resolution-methods-including-quantum-chemistry-force-fields
Description:
Multiscaling techniques that involve a quantum-chemical treatment of the electronic structure for the part with the highest resolution are promising computational tools. They are particularly useful for dealing with problems involving large systems like enzymes, membranes, polymers, etc., where, for example, chemical reactions take place. Having completed in the previous funding period of the TRR (i.e., the first funding period of this project) a corresponding QM/MM implementation that allows to include high-accuracy quantum-chemical methods from either coupled-cluster (CC) theory (i.e., CCSD, CCSD(T), etc.) or of multiconfigurational nature (i.e., CASSCF), we intend to complete the envisioned QM/MM/CG/hPF implementation that extends the QM/MM approach to coarse-grained (CG) treatments. In particular, we plan on using hybrid particle-field (hPF) theory based on its Hamiltonian reformulation, where the latter has been accomplished in the first funding period of this project. This reformulation facilitates the coupling of the regimes of different resolution as well as the use of the intended QM/MM/CG/hPF scheme in molecular dynamics (MD) simulations. In order to reduce the computational cost for the QM part which is the time-determining step of QM/MM and QM/MM/CG/hPF treatments and limits their applicability when using high-accuracy quantum-chemical methods, we propose to use Cholesky decomposition (CD) of the two-electron integrals to speed up the treatment of the QM region. We intend to implement CD based QM/MM and QM/MM/CG/hPF schemes using CD based nuclear forces to perform corresponding MD simulations. We also propose to implement CD based QM/MM and QM/MM/CG/hPF schemes for the computation of spectroscopic properties (NMR, EPR, vibrational) such that an accurate joint experimental and theoretical spectroscopic characterization of soft-matter systems (e.g., biological systems, molecules in solution or non-crystalline solids) becomes possible. In addition, we intend to extend the CD based schemes to include also those from CC and equation-of-motion CC (for the treatment of excited states) theory, and plan to develop schemes, again using CD techniques in a QM/MM or QM/MM/CG/hPF framework that allow the investigation of large systems under the influence of magnetic fields in a non-perturbative manner. Finally, the project intends to use the developed schemes in applications, in close collaboration with projects B1, B3, and B4 to simulate force-probe experiments and to investigate force fields for hydrated ions.
Parent Group: