Abstract
The intermixing of donor and acceptor materials in the active layer of bulk-heterojunction organic solar cells leads to a complex morphology. Many processes such as the splitting of excitons, charge dynamics, and charge recombination are based on the morphology, in particular on the interface between the two materials. To enhance the solar cell performance that is mainly based on these processes, two different modeling techniques are established today. In drift-diffusion (DD) simulations, the common assumption is to neglect the morphology including all interfaces and treat the active layer by one single effective material. On the contrary, kinetic Monte Carlo simulations offer a method to investigate organic solar cells considering a realistic blend morphology. We establish a common basis between the real blend Monte Carlo simulations and DD simulations with an effective medium approximation and compare the two models. Our results show that in the effective medium, the absence of local interface effects such as an explicit treatment of the short-ranged Coulomb interaction of charges across an interface lead to a difference in the charge density profile across the active layer. We provide hints that the effective medium approach might not be sufficient to treat all internal effects induced by the blend.