Dynamics Of Growing Carbon Nanotube Interfaces Probed By Machine Learning-enabled Molecular Simulations

DeepCNT-22 is a machine learning force field that enables near-microsecond timescale simulations of single-walled carbon nanotube (SWCNT) growth on iron catalysts. This work reveals the timescales and mechanisms of SWCNT growth, including the role of the tube-catalyst interface and the interplay between growth rate and temperature in defect formation and healing. MLFFs, trained on a diverse dataset of atomic configurations, offer a powerful tool for modeling materials at length and timescales approaching experiment.