Directional Entropy Bands for Surface Characterization of Polymer Crystallization

Area of Science:

• Materials Science
• Computational Chemistry
• Polymer Science

Background:

• Molecular dynamics (MD) simulations offer atomistic detail for polymer nucleation and crystallization.
• Interpreting complex spatiotemporal data from MD simulations is challenging.
• Existing order parameters struggle with directional alignment and spatial resolution, limiting analysis of anisotropic and heterogeneous crystallization.

Purpose of the Study:

• To introduce novel local order parameters, directional entropy bands, for enhanced analysis of polymer crystallization.
• To overcome limitations of conventional metrics in capturing directional alignment and surface phenomena.
• To provide a more accurate and interpretable framework for studying polymer crystallization kinetics.

Main Methods:

• Development and application of directional entropy bands, extending scalar entropy descriptors with angular moments.
• Comparison of directional entropy bands against conventional metrics like entropy, crystallinity index, and SOAP descriptors.
• Utilizing MD simulations of polymer crystallization and UMAP embeddings for data visualization and classification.

Main Results:

• Scalar entropy bands outperform SOAP descriptors in polymer phase separation analysis at single-snapshot resolution.
• Directional extensions effectively identify the evolving crystal-melt interface, enabling earlier nucleation detection.
• Quantitative surface profiling and identification of a continuous melt-surface-core manifold were achieved.

Conclusions:

• Directional entropy bands provide a robust and efficient method for analyzing polymer crystallization.
• The novel parameters offer superior spatial resolution and directional sensitivity compared to existing metrics.
• This framework facilitates deeper understanding of polymer crystallization kinetics and surface growth phenomena.