Transition metal complexes and chromophores with conjugated ? -bonds, both widely presented in biomolecules exhibit strong electron correlation. Their electronic structure features multiple near-degenerate electronic states, and therefore their computational treatment requires multiconfigurational and/or multireference methods. However, multiconfigurational wavefunction methods do not operate as “black-box” routines — they require selecting the so-called “active space”, a set of frontier orbitals for constructing important electronic configurations. Traditionally active space is selected manually based on chemical intuition and detailed inspection of orbitals. This way of selecting an active space appears to be impractical or impossible in the case of large and complex systems, such as occurring in biomolecules, or in the study of potential energy surfaces. Thus, automated and more systematic approaches are desirable. I will present two automatic algorithms for the active space selection, one targeted at transition metals and one at large ? -conjugated systems, which can be combined in practical applications. I then will demonstrate their performance for variety of different problems. These range from excitation energies of transition metal complexes and polyacenes, over the homolytic dissociation of hydrogen peroxide catalyzed by aqueous ferrous ion in Fenton chemistry, to the scanning potential energy surface along double proton transfer pathways in photoreceptor BLUF proteins.

Automatic algorithms for the active space selection

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.