Abstract: The nanoscale morphology of polymer blends is a key parameter to reach high efficiency in bulk heterojunction solar cells. Thereby, research typically focusing on optimal blend morphologies while studying nonoptimized blends may give insight into blend designs that can prove more robust against morphology defects. Here, we focus on the direct correlation of morphology and device performance of thieno[3,4-b]-thiophene-alt-benzodithiophene (PTB7):[6,6]phenyl C-71 butyric acid methyl ester (PC71BM) bulk heterojunction (BHJ) blends processed without additives in different donor/acceptor weight ratios. We show that while blends of a 1:1.5 ratio are composed of large donor-enriched and fullerene domains beyond the exciton diffusion length, reducing the ratio below 1:0.5 leads to blends composed purely of polymer-enriched domains. Importantly, the photocurrent density in such blends can reach values between 45 and 60% of those reached for fully optimized blends using additives. We provide here direct visual evidence that fullerenes in the donor-enriched domains are not distributed homogeneously but fluctuate locally. To this end, we performed compositional nanoscale morphology analysis of the blend using spectroscopic imaging of low-energy-loss electrons using a transmission electron microscope. Charge transport measurement in combination with molecular dynamics simulations shows that the fullerene substructures inside the polymer phase generate efficient electron transport in the polymer-enriched phase. Furthermore, we show that the formation of densely packed regions of fullerene inside the polymer phase is driven by the PTB7:PC71BM enthalpy of mixing. The occurrence of such a nanoscale network of fullerene clusters leads to a reduction of electron trap states and thus efficient extraction of photocurrent inside the polymer domain. Suitable tuning of the polymer-acceptor interaction can thus introduce acceptor subnetworks in polymer-enriched phases, improving the tolerance for high-efficiency BHJ toward morphological defects such as donor-enriched domains exceeding the exciton diffusion length.

Abstract: Self-assembly (SA) of nanostructures has recently gained increasing interest. A clear understanding of the process is not straightforward since SA of nanoparticles is a complex multiscale phenomenon including different driving forces. Here, we study the SA between aluminum doped ZnO nanopyramids into couples by combining inorganic chemistry and advanced electron microscopy techniques with atomistic simulations. Our results show that the SA of the coupled nanopyramids is controlled first by morphology, as coupling only occurs in the case of pyramids with well-developed facets of the basal planes. The combination of electron microscopy and atomistic modeling reveals that the coupling is further driven by strong ligandligand interaction between the bases of the pyramids as dominant force, while screening effects due to Al doping or solvent as well as corecore interaction are only minor contributions. Our combined approach provides a deeper understanding of the complex interplay between the interactions at work in the coupled SA of ZnO nanopyramids.