Limits and optimization of a solar energy conversion system consisting of a photochemical charge separating unit coupled to an energy storage state are explored by multi-objective genetic algorithms. Pareto fronts were evaluated to obtain information about the ideal parameter combinations, guaranteeing highest efficiency. The light absorbing and charge separating unit is described by a chain of chromophores and electron carriers, connected by Marcus type electron transfer processes. It is coupled to the thermal equilibrium of charge conduction and transport in an energy storage system according to the principle of detailed balance. In addition to our previous findings for an optimal charge separation unit, consisting of a minimum number of charge carriers with adapted recombination and reaction rates, the complete photochemical unit must fulfil further requirements. Low reorganization energies are found to be essential for the initial charge separation steps and can be realized by a low dielectric constant in the local environment. The identified optimal operation rates can be realized by antenna systems adapted to the illumination conditions. For standard solar illumination and a realistic parameter setting energy conversion efficiencies up to 26.8% are predicted, comparable to the limit (31.8%) of ideal single junction semiconductor solar cells.