Multiscale Modeling of Precursor Molecule Alignment for Improving Pitch-Based Carbon Fiber Production

Carbon fiber (CF) is a versatile material widely applied in many fields, and one main method to produce CF is from pitch, a viscoelastic material composed of aromatic hydrocarbons. Though it is well known that the early stage alignment of precursor molecules has significant effects on the pitch-based fibers, a deeper understanding of its underlying microscopic mechanisms is lacking. Here we perform fully atomistic (FA) and coarse-grained (CG) simulations to study the alignment of pitch molecules. While the FA simulations generates accurate atomistic descriptions, the CG simulations provide a means to efficiently explore larger time and length scales. The ellipsoid algorithm is adopted to extract geometric features of these molecules from the FA trajectories, enabling a quantitative analysis of the intermolecular alignment as well as the construction of anisotropic CG particles. Our multiscale simulations identify the dependences of alignment on the molecule size, temperature and loading conditions. Further, the model suggests specific additives and operation conditions that can improve the molecule alignment. These calculations provide guidelines for optimizing the synthesis pathways for low cost and high-quality CF.