Contrasting performance of donor-acceptor copolymer pairs in ternary blend solar cells and two-acceptor copolymers in binary blend solar cells.

Here two contrasting approaches to polymer-fullerene solar cells are compared. In the first approach, two distinct semi-random donor-acceptor copolymers are blended with phenyl-C61-butyric acid methyl ester (PC61BM) to form ternary blend solar cells. The two poly(3-hexylthiophene)-based polymers contain either the acceptor thienopyrroledione (TPD) or diketopyrrolopyrrole (DPP). In the second approach, semi-random donor-acceptor copolymers containing both TPD and DPP acceptors in the same polymer backbone, termed two-acceptor polymers, are blended with PC61BM to give binary blend solar cells. The two approaches result in bulk heterojunction solar cells that have the same molecular active-layer components but differ in the manner in which these molecular components are mixed, either by physical mixing (ternary blend) or chemical "mixing" in the two-acceptor (binary blend) case. Optical properties and photon-to-electron conversion efficiencies of the binary and ternary blends were found to have similar features and were described as a linear combination of the individual components. At the same time, significant differences were observed in the open-circuit voltage (Voc) behaviors of binary and ternary blend solar cells. While in case of two-acceptor polymers, the Voc was found to be in the range of 0.495-0.552 V, ternary blend solar cells showed behavior inherent to organic alloy formation, displaying an intermediate, composition-dependent and tunable Voc in the range from 0.582 to 0.684 V, significantly exceeding the values achieved in the two-acceptor containing binary blend solar cells. Despite the differences between the physical and chemical mixing approaches, both pathways provided solar cells with similar power conversion efficiencies, highlighting the advantages of both pathways toward highly efficient organic solar cells.