The electronic properties of macromolecular semiconductor thin films depend profoundly on their solid-state microstructure, which in turn is governed, among other things, by the processing conditions selected and the polymer's chemical nature and molecular weight. Specifically, lowmolecular-weight materials form crystalline domains of cofacially π-stacked molecules, while the usually entangled nature of higher molecular-weight polymers leads to microstructures comprised of molecularly ordered crystallites interconnected by amorphous regions. Here, we examine the interplay between extended exciton states delocalized along the polymer backbones and across polymer chains within the π-stack, depending on the structural development with molecular weight.Such two-dimensional excitations can be considered as Frenkel excitons in the limit of weak intersite coupling. We combine optical spectroscopies, thermal probes, and theoretical modeling, focusing on neat poly(3-hexylthiophene) (P3HT) -one of the most extensively studied polymer semiconductors -of weight-average molecular weight (M w ) of 3-450 kg/mol. In thin-film structures of high-molecular-weight materials (M w > 50 kg/mol), a balance of intramolecular and intermolecular excitonic coupling results in high exciton coherence lengths along chains (~4 thiophene units), with interchain coherence limited to ~2.5 chains. In contrast, for structures of low-M w P3HT (<40 kg/mol), the interchain exciton coherence is dominant (~20% higher than in architectures formed by high-molecular-weight materials). In addition, the spatial coherence within the chain is significantly reduced (by nearly 30%). These observations give valuable structural information; they suggest that the macromolecules in aggregated regions of high-molecular-weight P3HT adopt a more planar conformation compared to low-molecular-weight materials. This results in the observed increase in intrachain exciton coherence. In contrast, shorter chains seem to lead to torsionally more disordered architectures. A rigorous, fundamental description of primary photoexcitations in π-conjugated polymers is hence developed: two-dimensional excitons are defined by the chain-length dependent molecular arrangement and interconnectivity of the conjugated macromolecules, leading to interplay between intramolecular and intermolecular spatial coherence.

The dissociation of photogenerated excitons and the subsequent spatial separation of the charges are of crucial importance to the design of efficient donor-acceptor heterojunction solar cells. While huge progress has been made in understanding charge generation at all-organic junctions, the process in hybrid organic:inorganic systems has barely been addressed. Here, we explore the influence of energetic driving force and local crystallinity on the efficiency of charge pair generation at hybrid organic:inorganic semiconductor heterojunctions. We use x-ray diffraction, photoluminescence quenching, transient absorption spectroscopy, photovoltaic device and electroluminescence measurements to demonstrate that the dissociation of photogenerated polaron pairs at hybrid heterojunctions is assisted by the presence of crystalline electron acceptor domains. We propose that such domains encourage delocalization of the geminate pair state. The present findings suggest that the requirement for a large driving energy for charge separation is relaxed when a more crystalline electron acceptor is used.

The bis and tris adducts of [6,6]phenyl-C(61)-butyric acid methyl ester (PCBM) offer lower reduction potentials than PCBM and are therefore expected to offer larger open-circuit voltages and more efficient energy conversion when blended with conjugated polymers in photovoltaic devices in place of PCBM. However, poor photovoltaic device performances are commonly observed when PCBM is replaced with higher-adduct fullerenes. In this work, we use transmission electron microscopy (TEM), steady-state and ultrafast time-resolved photoluminescence spectroscopy (PL), and differential scanning calorimetry (DSC) to probe the microstructural properties of blend films of poly(3-hexylthiophene-2,5-diyl) (P3HT) with the bis and tris adducts of PCBM. TEM and PL indicate that, in as-spun blend films, fullerenes become less soluble in P3HT as the number of adducts increases. PL indicates that upon annealing crystallization leads to phase separation in P3HT:PCBM samples only. DSC studies indicate that the interactions between P3HT and the fullerene become weaker with higher-adduct fullerenes and that all systems exhibit eutectic phase behavior with a eutectic composition being shifted to higher molar fullerene content for higher-adduct fullerenes. We propose two different mechanisms of microstructure development for PCBM and higher-adduct fullerenes. P3HT:PCBM blends, phase segregation is the result of crystallization of either one or both components and is facilitated by thermal treatments. In contrast, for blends containing higher adducts, the phase separation is due to a partial demixing of the amorphous phases. We rationalize the lower photocurrent generation by the higher-adduct fullerene blends in terms of film microstructure.

In this study, a generally applicable strategy is described to manipulate the optical properties of a wide range of polymer semiconductors in the solid state. Blending these materials with a non-conjugated, polar polymer matrix is found to be the processing key to a drastic change and red-shift of the absorption characteristics.

We demonstrate that blend films containing poly(3-hexylthiophene-2,5-diyl) and in situ grown CdS display a greater yield of photogenerated charges than a blend containing an equivalent amount of pre-synthesised CdS quantum dots. Moreover, we show that the greater charge yield in the in situ grown films leads to an improvement in device efficiency. The present findings also appear to suggest that charge photogeneration at the CdS/polymer heterojunction is facilitated by the formation of nanoparticle networks as a result of CdS aggregation.