Studies of recombination in organic and hybrid solar cells using electroluminescence

The performance of solution processed solar cells such as organic bulk heterojunction (OPV) devices is limited by strong recombination. However, the mechanisms are still unclear. In this thesis, I develop a toolbox using steady-state spectroscopy measurements to explore the recombination mechanisms in a range of solution-processed solar cells. In the first results chapter, I use the reciprocity relation between light absorption and light emission to explore theoretical and practical performance limits for solar cells based on organic semiconductors and perovskites and compare the results with data for state-of-the-art photovoltaic cells made from GaAs, c-Si, and CIGS. In OPV systems, I show that the energetic losses due to the mismatch of the bandgap have been significantly reduced through optimisation of the donor polymer, but the non-radiative recombination losses remain the same and become the major barrier to higher performance. In the next two chapters, I use light intensity dependence of open-circuit voltage measurement (suns-VOC) and electroluminescence – injection current measurement (EL-J) to disentangle recombination mechanisms in OPV and perovskite cells, respectively. First, I identify the present of Shockley-Read-Hall recombination and surface recombination in OPV devices. I intentionally control the sample geometry to modulate the amount of surface recombination and demonstrate that surface recombination can significantly affect the device performance. In the following chapter, I analyse time dependent suns-VOC and EL-J measurements on perovskite cells with different architectures and pre-conditioning regimes used. I identify the changes in recombination mechanisms with delay time and pre-conditions. The effects of ion migration are used to interpret the results. In the final chapter, I apply luminescence spectroscopy techniques to investigate the degree of fullerene crystallinity in polymer:fullerene blends. Charge-transfer state emission is used to probe the onset of the crystallisation of fullerenes in an amorphous polymer. I relate the CT peak shift directly to the change in microstructure of a blend film.