ReaxFF molecular dynamics simulation of the thermal decomposition reaction of bio-based polyester materials

Abstract

Bio-based polyester elastomers have been widely concerned by researchers in recent years because of their wide sources of monomers and environmentally friendly characteristics. However, compared with traditional petroleum-based elastomers, the...

Predicting the glass transition temperature and solubility parameter between rubber/silica and rubber/resins via all‐atom molecular dynamics simulation

“…Molecular dynamics simulation is a computational method used to simulate the temporal evolution of molecular systems, enabling researchers to gain insights into intermolecular interactions and the physical properties of molecules 26–33 . For this work, an all‐atom molecular dynamics simulation 34–38 is adopted, utilizing Materials Studio as the primary tool and employing the COMPASS all‐atom force field 39,40 . In the COMPASS all‐atom force field, the total energy () of the simulation system is calculated as the sum of bonding and non‐bonding interactions: …”

Predicting Natural Rubber Crystallinity by a Novel Machine Learning Algorithm Based on Molecular Dynamics Simulation Data

“…Thus, it is crucial to meticulously investigate the impacts caused by elements other than rubber, particularly phospholipids and proteins, on the behavior of NR’s SIC in order to attain a more comprehensive understanding. Impressively, existing studies have acknowledged the substantial potential of molecular dynamics (MD) simulation as a robust approach for unraveling the complexities of materials with intricate structures. − MD simulation has been widely acknowledged as an effective approach for studying the crystallization behavior of polymers, as demonstrated by numerous studies. It offers a powerful tool to investigate the molecular-level mechanisms and dynamics involved in the process of polymer crystallization. − However, it has been observed that in the fields of chemistry and materials science, the process of model training typically involves a limited subset of high-quality data points, which usually range from a few hundred to even fewer .…”