Accelerating Bioinspired Cellulose Nanocomposite Design Through a Coarse-Grained Molecular Dynamics Approach

Nature creates intricate materials with exceptional properties and versatile functions. Typically, the compositions of these materials are simple components, such as minerals and proteins, and their performance is dictated by their nanoscale features including structural arrangement, size, and interactions between different phases. Among these materials, coconut shells are distinguished not only by their exceptional performance but also due to their abundance as natural resources, thanks to the fast growth rate.
The coconut shell comprises three distinct layers: the outermost exocarp, the middle fibrous mesocarp, and the inner endocarp. Notably, the endocarp layer, which is primarily composed of cellulose, hemicellulose, and lignin, demonstrates an exceptional combination of stiffness, hardness, strength, and fracture toughness despite this simple composition. While the hierarchical structure of the endocarp is widely believed to be key to its remarkable performance, a clear relationship correlating this hierarchy with its properties remains elusive. Given the role of nanoscale structure in enhancing the mechanical properties of other biomaterials, we infer that the nanoscale organization of the coconut endocarp is crucial. However, comprehensive morphology detailing cellulose arrangement at the nanometer scale, as well as precise polymer composition, are still lacking.
In this study, we developed a coarse-grained (CG) model of the coconut endocarp that accurately replicates its mechanical and interfacial properties as observed in experiments and atomistic molecular dynamics (AA MD) simulations. The model effectively captures the erratic slip-stick phenomena characteristic of atomistic simulations, while also predicting important mechanical properties of the endocarp such as strength, stiffness, and toughness. An additional key feature of this model is its ability to simulate the interfacial characteristics of the endocarp.
We refined the CG model by systematically varying parameters like cellulose bundle length, degree of polymerization, bundle arrangement, and distribution dimensionality. This approach allowed us to explore the impact of interfacial properties in various configurations. When comparing fiber-matrix polymer pairs, the simulations showed that cellulose-hemicellulose combinations provided better mechanical performance than cellulose-lignin pairs.
In terms of the brick-and-mortar structure, the model demonstrated ductile behavior and a two-step failure mechanism. Initially, the soft amorphous matrix failed at the short interfaces, followed by crack propagation along the edges of the longer microfibrils. The models with different cellulose bundle lengths (10 nm and 20 nm) and degrees of polymerization (DP = 40 and 60) exhibited similar trends in strength, stiffness, and toughness. However, composites with longer cellulose chains and higher degrees of polymerization did not experience soft matrix failure at the short interfaces.
The arrangement of cellulose bundles within the hemicellulose matrix also played a significant role in the failure mechanisms. In structures with evenly distributed cellulose bundles, localized cracking between the bundles led to catastrophic failure. On the other hand, the staggered brick-and-mortar arrangement exhibited superior mechanical performance, with better stiffness, toughness, and strength. At the nanoscale, failure involved the degradation or fracture of the interfibrillar matrix, which affected the overall structural integrity.
Although this study focused on specific cases, the CG models offer broad potential for more complex engineering applications. They can be extended to model defects such as porosity, variations in matrix density, and moisture content. Additionally, these models could be applied to investigate anisotropic properties and a wider range of structural configurations.