Accelerating cellulose nanocomposite design through coarse-grained molecular dynamics simulation and rule of mixture analysis

Abstract The spatial arrangement of crystalline cellulose nanofibers and amorphous hemicellulose in the coconut endocarp remains unclear. Inspired by the hierarchical structures of nacre and bone, we proposed a brick-and-mortar model in which cellulose bundles are arranged in a staggered pattern and then fully embedded within the hemicellulose matrix. Such complex, high-dimensional biopolymer composite systems pose a significant challenge for all-atom molecular dynamics (AAMD) simulations to study microstructure-property relationships and reveal the underlying deformation mechanisms. Therefore, in this work, we developed a coarse-grained (CG) potential model that explicitly incorporates hydrogen bonding, the key interfacial interaction in cellulose-hemicellulose composites. The CG model parameters were carefully fitted and validated against AAMD simulations, ensuring accurate predictions of stiffness, strength, toughness, and failure mechanisms. It was proven that the CG model enables efficient simulations of large-scale 3D systems with millions of atoms, providing crucial insights into mechanical behavior while maintaining computational efficiency. The staggered 3D distribution of long cellulose bundles was found to optimize the reinforcement effect by maximizing strain energy absorption during deformation. Given the vital role of hydrogen bonding, we modified the rule of mixture (ROM) to quantify their contributions across different models by incorporating an interphase term to account for interfacial interactions, facilitated by comprehensive data from MD simulations. This integrated approach of CGMD and ROM analysis not only enhances our understanding of the mechanical behavior of cellulose-hemicellulose composites but also provides a generalizable framework for studying and optimizing natural and bioinspired materials at large scale.