Characterization of Multi-Scale Polymer Structure Evolution through Multivariate Decomposition of Process-Dependent Scattering Data

In this work, we have developed a novel generalized multivariate data decomposition approach and applied it to the study of multi-scale polymer structural evolution using scattering data. The method, based on Network Component Analysis (NCA), decomposes each experimental signature included in a large process-dependent dataset into the weighted sum of a reduced set of component signatures. These components represents scattering signatures from multiple independent structural units, while the weights can be correlated to the volume fraction of each unit in a single sample. This coarse-graining approach has been applied to the decomposition of small and wide-angle X-ray scattering data to study the time and temperature-controlled structural variation of branched polyethlyene copolymer samples at multiple characteristic length scales.