Afterglow Effects as a Tool to Screen Emissive Nongeminate Charge Recombination Processes in Organic Photovoltaic Composites

Disentangling temporally overlapping charge carrier recombination events in organic bulk heterojunctions by optical spectroscopy is challenging. Here, a new methodology for employing delayed luminescence spectroscopy is presented. The proposed method is capable of distinguishing between recombination of spatially separated charge carriers and trap-assisted charge recombination simply by monitoring the delayed luminescence (afterglow) of bulk heterojunctions with a quasi time-integrated detection scheme. Applied on the model composite of the donor poly(6,12-dihydro-6,6,12,12-tetraoctyl-indeno[1,2-b]fluorene-alt-benzothiadiazole) (PIF8BT) polymer and the acceptor ethyl-propyl perylene diimide (PDI) derivative, that is, PIF8BT:PDI, the luminescence of charge-transfer (CT) states created by nongeminate charge recombination on the ns to μs timescale is observed. Fluence-dependent, quasi time-integrated detection of the CT luminescence monitors exclusively emissive charge recombination events, while rejecting the contribution of other early-time emissive processes. Trap-assisted and bimolecular charge recombination channels are identified based on their distinct dependence on fluence. The importance of the two recombination channels is correlated with the layer's order and electrical properties of the corresponding devices. Four different microstructures of the PIF8BT:PDI composite obtained by thermal annealing are investigated. Thermal annealing of PIF8BT:PDI shrinks the PDI domains in parallel with the growth of the PIF8BT domains in the blend. Common to all states studied, the delayed CT luminescence signal is dominated by trap-assisted recombination. Yet, the minor fraction of fully separated charge recombination in the overall CT emission increases as the difference in the size of the donor and acceptor domains in the PIF8BT:PDI blend becomes larger. Electric field-induced quenching measurements on complete PIF8BT:PDI devices confirm quantitatively the dominance of emissive trap-limited charge recombination and demonstrates that only 40% of the PIF8BT/PDI CT luminescence comes from the recombination of fully-separated charges, taking place within 200 ns after photoexcitation. The method is applicable to other nonfullerene acceptor blends beyond the system discussed here, if their CT state luminescence can be monitored