A wide range of conjugated polymers form thin films with oriented morphologies, typically characterized with an imaging technique such as AFM. To provide quantitative metrics to describe and compare these morphologies, we developed an automated image analysis software package called GTFiber. GTFiber was originally designed for the high-throughput extraction and measurement of fibers from images, and has enabled the determination of quantitative process-structure-property relationships for poly(3-hexylthiophene)-based field effect transistors. These included both the effect of the spin- and blade-coating flow fields on fiber alignment during thin film deposition as well as the correlation between fiber alignment and field effect mobility. Data visualizations produced using GTFiber revealed striking nanofiber bundling behavior during solution processing. More recently, we have demonstrated GTFiber's generalizability to fibrillar systems in other materials, as well as the oriented morphologies observed in thin films of PNDI or strain-stretched thin films of conjugated polymer blends.
The extraction and vectorization of fibers from images also enabled us to simulate tie chain behavior on a length scale unattainable by molecular dynamics. By performing a Monte Carlo simulation of P3HT chain packing overlaid on the real, extracted fiber backbones, we were able to define a quantitative measure of grain interconnectivity: the tie-chain length density. Tie chain length density quantifies the expected number of tie chain connections to a given grain per unit lenth, and is sensitive to fiber packing, orientational order, and the assumed molecular weight distribution. While experimental observation of tie chains remains difficult, our simulation provides an upper bound on grain interconnectivity, while laying the quantitative groundwork necessary to describe this important feature of conjugated polymer thin films.