Development of efficient ab initio methods for the accurate prediction of hyperfine and quadrupole interaction parameters in large molecules

High-resolution electron paramagnetic resonance (EPR) spectroscopy has reached a level of sophistication that allows for a multitude of small hyperfine and quadrupole couplings to be experimentally determined. In principle, this data contains a wealth of selective geometric structure information about the paramagnetic center and its surrounding. Two severe problems arise: (a) proper data reduction through computer simulation of the experimental spectra and (b) interpretation of the results in terms of molecular geometric and electronic structure. Both problems need to be thoroughly addressed in order for EPR spectroscopy to develop into an even more powerful structure determination tool than it is today. Quantum chemistry is a powerful partner of experiment in approaching both fundamental problems. Quantum chemical calculations provide full sets of spin-Hamiltonian (SH) parameters for anygiven structure (or structural proposal) that can be used as starting points for the computer simulation of the actual spectra. One arrives in this way at a reliable assignment of individual hyperfine and quadrupole tensors to individual nuclei. This is of critical importance as the number of parameters required to completely fit high-resolution spectra quickly becomes unmanageable. Secondly, once it is established that the calculations provide a realistic picture of the spin distribution, the results can be interpreted in terms of geometric and electronic structure. Hence, molecular level insight can be obtained that serves as a basis for understanding molecular reactivity or molecular properties. Todays quantum chemical approaches to EPR spectroscopy are strongly dominated by density functional theory (DFT).However, there are still significant method inherent errors in these calculations that arise from unknown shortcomings of the density functionals used. Rather than trying to engage in the endeavor of empirically trying new functionals, we propose an ab initio wavefunction based route to EPR parameters that is based on the extremely powerful and accurate coupled cluster (CC) theory. It is well known that CC theory converges quickly to the exact solution of the molecular Schrödinger equation. Already at the level of only incorporating single- and double- excitation operators the results for molecular properties, especially hyperfine couplings are known to be excellent. Yet, the computational effort to obtain these results is unmanageable for larger molecules and grows as the sixth power of the system size. In this proposal we propose to combine coupled cluster linear response theory with the concept of local pair natural orbitals in order to arrive at a new and systematically accurate theory for the calculation of EPR properties. Collaborative applications inside the SPP are evident and will initially focus on the experimental group Prof. Marina Bennati.

F. Neese

Prof. Dr. Frank Neese

Max-Planck Institut für Bioanorganische Chemie, Abteilung „Molekulare Theorie und Spektroskopie“
Stiftstr. 34-36, 45470 Mülheim an der Ruhr
Tel.: +49 208 306-3656 Email:
Fax: +49 208 306-3955 Website: http://