Charge transfer in peptides and proteins can occur on different pathways, depending on the energetic landscape as well as the coupling between the involved orbitals. Since details of the mechanism and pathways are difficult to access experimentally, different modeling strategies have been successfully applied to study these processes in the past. These can be based on a simple empirical pathway model, efficient tight binding type atomic orbital Hamiltonians or ab initio and density functional calculations. An interesting strategy, which allows an efficient calculations of charge transfer parameters, is based on a fragmentation of the system into functional units. While this works well for systems like DNA, where the charge transfer pathway is naturally divided into distinct molecular fragments, this is less obvious for charge transfer along peptide and protein backbones. In this work, we develop and access a strategy for an effective fragmentation approach, which allows one to compute electronic couplings for large systems along nanosecond time scale molecular dynamics trajectories. The new methodology is applied to a solvated peptide, for which charge transfer properties have been studied recently using an empirical pathway model. As could be expected, dynamical effects turn out to be important, which emphasizes the importance of using effective quantum approaches which allow for sufficient sampling. However, the computed rates are orders of magnitude smaller than experimentally determined, which indicates the shortcomings of present modeling approaches.
■ INTRODUCTIONCharge transfer (CT) in biomolecules is a fundamental process that has been subject of intense scientific study for more than 50 years.1−4 It is highly relevant in the understanding of functional as well as pathological cellular processes, ranging from respiration and photosynthesis to DNA damage and repair.5−7 Our theoretical understanding of molecular CT has increased tremendously since the pioneering studies by Marcus and others, 8−16 but a complete detailed description on the molecular level remains elusive.Several approaches exist to compute the electronic couplings, ranging from empirical pathway models, 17−19 over tight-binding based 20,21 to ab initio or density functional calculations.
22−25Most of these quantum approaches are based on an atomic orbital description to compute charge transfer parameters, which becomes very costly for large systems, even when using extended Huckel or tight-binding Hamiltonians, because the computer time scales cubic with the number of atomic orbitals involved. An additional computational challange is the need for sampling along MD trajectories. The biochemical transfer of electrons over many nanometers through large protein assemblies involves dynamical changes that occur on time scales ranging from subpicosecond changes in electronic structure to microsecond or even slower conformational changes. Especially the interplay between the atomic structural fluctuations and the electron dynamics has bec...