(723f) Predicting Properties of Carbon Nanofibers from Polyurethane Based Precursors Using Experimental Characterization and Molecular Dynamics Simulation

Molecular dynamics (MD) simulation approach is taken to study polyurethane precursors and their impact on carbon nanofiber (CNFs) properties, which is validated experimentally. Polyurethane precursors are used in the MD model, which is carbonized using reactive force field (ReaxFF) scheme. Results are reported by analyzing pore size distribution and porosity for the resultant CNFs with different densities. To validate the model, CNFs are produced experimentally by electrospinning polyurethane precursors into fiber mats and carbonizing them subsequently. From the resulting carbon nanofiber mats, pore properties are determined using Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) adsorption data. Surface functional groups are quantified using XPS, while density is measured directly using X-ray scanning technique. Experimental data are compared to the MD simulation in terms of densities, surface composition, and pore size distribution. The discrepancy between ReaxFF modeling and experimental parameters will be fed into a machine learning tool to shine light on better predication power of the model and to provide guidance for future experimental design for desired CNF structures and properties in targeted applications, such as energy storage devices.