New AI method predicts properties of fluoride plastic without costly experiments
Researchers used computational modeling to predict how polyvinyl fluoride behaves at the atomic level, filling gaps where physical testing falls short. The approach could accelerate development of advanced materials for electronics and energy storage by reducing expensive lab work.
Originaltitel: Polyvinyl fluoride: Predicting polarization in a complex soft matter system
<p>We use first-principle density functional theory (DFT) to predict properties for semicrystalline polyvinyl fluoride (PVF) and compare with polyvinylidiene fluoride. We note that the crystalline regions of PVF are complex in the sense that we lack a complete experimental characterization of the detailed atomic organization. We therefore turn to DFT to predict both the structure and associated materials properties, illustrating a possible work flow for complex soft-matter modeling. We rely on the nonempirical consistent-exchange van der Waals density functional version [K. Berland and P. Hyldgaard, Phys. Rev. B 89, 035412 (2014)] and identify plausible ground-state and excited-state motifs. From there we predict the elastic response of the crystalline motifs, and an upper limit estimate of the PVF polarization at room temperature.</p>