Polarized resonant soft X-ray scattering for organic photovoltaics

The materials science of polymeric semiconductors is complex because their intrinsic semicrystallinity always results in structural heterogeneity. This is especially true in bulk heterojunction (BHJ) organic photovoltaics (OPVs), in which two or more organic semiconductors are solidified from the same solution to create a bicontinuous percolative network with length scales of (10 to 100) nm. Heterogeneities in composition, order, and molecular orientation strongly influence the performance of polymeric semiconductors in OPVs, diodes, transistors, and sensors. Our ability to measure and correlate these heterogeneities to performance has grown enormously over the past two decades. For example, X-ray diffraction and transmission electron microscopy provide composition maps and orientation distributions of well-ordered regions.

A powerful technique for measuring the structure of polymer OPVs is Polarized Resonant Soft X-ray Scattering (P-RSoXS). In fact, RSoXS of BHJs is by far the most published application of RSoXS despite its potential to measure almost any soft matter. Typical aspects of BHJ structure that are said to be measured by RSoXS include relative phase purity and the sizes or length scales of domains or phases.

 

Using a GPU-accelerated computational framework for the forward simulation of P-RSoXS from real-space representations, I will consider a realistic BHJ “toy” model built from a blend of polymer and small molecule that routinely delivers photovoltaic power conversion efficiencies of >16%. By considering reasonable hypothetical compositional heterogeneity, sample roughness, and orientational behaviors, I will demonstrate that the most common RSoXS data analysis workflow is very likely to misrepresent – both qualitatively and quantitatively – many aspects of structure that it is relied upon to measure. Happily, the rich chemical contrast afforded by the technique provides an accesible improved analysis workflow. I will describe model-free analysis protocols to verify that the technique is measuring the aspects of structure that it is expected to measure. I will then highlight opportunities for the extraction of unique structural information from P-RSoXS data via model-based analysis to accelerate the further development of OPVs and other functional soft matter technologies.