Rene Janssen1 2 Mengmeng Li1 Dario Di Carlo Rasi1 Pieter van Thiel1 Martijn Wienk1 2

1, Eindhoven University of Technology, Eindhoven, , Netherlands
2, Dutch Institute for Fundamental Energy Research, Eindhoven, , Netherlands

Using diketopyrrolopyrrole (DPP) polymers with different chemical structure and molecular weights, the device performance of polymer:fullerene solar cells was systematically optimized by considering device polarity, morphology, and light absorption. More soluble derivatives show a 10-25% enhanced PCE in inverted device configurations, while the device performance is independent of device polarity for less soluble DPP polymers. Optimization of the nature of the cosolvent to narrow the fibril width of polymers in the blends towards the exciton diffusion length enhanced charge generation. Additionally, the use of a retroreflective foil increased absorption of light. Combined, the effects afforded a PCE of 9.6% for small band gap polymers.

Using a combination of poly(3,4-ethylenedioxythiophene):polystyrene sulfonate, diluted in near azeotropic water/n-propanol dispersions as hole transport layer, and metal (Zn, Sn) oxide nanoparticles, dispersed in alcohols as electron transport layer, novel, versatile charge recombination layers have been developed for solution-processed multi-junction solar cells in n-i-p and p-i-n architectures. These have been incorporated in multi-junction solar cells, employing a range of different polymer-fullerene photoactive layers, without the need of adjusting the formulations or deposition conditions. The approach permitted realizing complex devices in good yields, providing PCE up to 10%.

Further a new protocol, involving optical modeling, has been developed to correctly measure the EQE of triple-junction organic solar cells. Apart from correcting for the build-up electric field, the effect of light intensity is considered with the help of representative single-junction cells. The short-circuit current density determined from integration of the EQE with the AM1.5G solar spectrum differs by up to 10% between corrected and un-corrected protocols. The results were validated by comparing the EQE experimentally measured to the EQE calculated via optical-electronic modeling, obtaining an excellent agreement.