Calculation of Excited-State Dynamics in Light-Harvesting Complexes

Theoretical description of excitation energy transfer in light-harvesting complexes, for example, in natural photosystems, is a complicated computational problem, though, in principle, it boils down merely to solution of time-dependent Schrodinger equation. To solve the equation one needs the information about the eigenstates and the initial state of the given system. While the second requirement is not very hard to fulfill, the first one poses serious difficulties. This is because of the large size of such systems: they are usually rather big aggregates of chromophore molecules, surrounded by some media protein or solvent. For example, chromophores (chlorophylls and carotenoids) of cyanobacterial photosystem I contain several tens of thousands of electrons. The presence of protein molecules makes the problem even more difficult. One of the ways to overcome these computational difficulties is to use system plus reservoir models, where system refers to chromophores, and reservoir, a much bigger part, to surrounding media. The treatment of reservoir as a thermally equilibrium media and averaging of coupling between system and reservoir over degrees of freedom of the media allows one to reduce the number of states involved in the computation. A variation of such an approach is the Redfield theory [1, 2], which is widely used for modelling the excitation energy and charge transfer in light-harvesting complexes, including natural photosystems. Its practical application depends heavily on the availability of the data on the so-called spectral density the generalized properties of the reservoir. Being an elusive property, the spectral density requires nontrivial experimental techniques for its determination [3]. We propose a Redfield theory-based model, which uses several assumptions about the properties of the media in order to obtain spectral density by theoretical means. We begin with microscopic description of the reservoir,