We show that quantum coherence of biomolecular excitons is maintained over exceedingly long times due to the constructive role of their non-Markovian protein-solvent environment. Using a numerically exact approach, we demonstrate that a slow quantum bath helps to sustain quantum entanglement of two pairs of Förster coupled excitons, in contrast to a Markovian environment. We consider the crossover from a fast to a slow bath and from weak to strong dissipation and show that a slow bath can generate robust entanglement. This persists to surprisingly high temperatures, even higher than the excitonic gap and is absent for a Markovian bath.

No abstract

We have developed a numerical approach to compute real-time path integral expressions for quantum transport problems out of equilibrium. The scheme is based on a deterministic iterative summation of the path integral (ISPI) for the generating function of the nonequilibrium current. Self-energies due to the leads, being non-local in time, are fully taken into account within a finite memory time, thereby including non-Markovian effects, and numerical results are extrapolated both to vanishing (Trotter) time discretization and to infinite memory time. This extrapolation scheme converges except at very low temperatures, and the results are then numerically exact. The method is applied to nonequilibrium transport through an Anderson dot.

We formulate and apply a nonperturbative numerical approach to the nonequilibrium current I (V ) through a voltage-biased molecular conductor. We focus on a single electronic level coupled to an unequilibrated vibration mode (Anderson-Holstein model), which can be mapped to an effective three-state problem. Performing an iterative summation of real-time path integral (ISPI) expressions, we accurately reproduce known analytical results in three different limits. We then study the crossover regime between those limits and show that the Franck-Condon blockade persists in the quantum-coherent low-temperature limit, with a nonequilibrium smearing of step features in the I V curve.