(411f) Understanding Hierarchical Structures within Aqueous Methylcellulose Solutions & Gels Using Multiscale Molecular Dynamics Simulations

Methylcellulose (MC) is a cellulose derivative where some or all the hydroxyl groups (-OH) are replaced by methoxy (-OCH₃) groups. At low temperatures, commercial MC chains (with extent of methylation or degree of substitution or DS = 1.8) are soluble in water, while at temperatures above 323K, these MC chains undergo thermoreversible gelation, forming a gel network of semi-flexible fibrils. The fibrillar network makes MC chains ideal for applications as thickeners or stabilizers in food and drug formulations. Understanding the hierarchical structure and thermo-responsive assembly of aqueous MC that gives rise to the desirable properties has been the subject of many experiments and simulation studies in the past two decades. Small-angle scattering data interpreted using analytical model and machine learning-enhanced Computational Reverse Engineering Analysis for Scattering Experiments (ML-CREASE) consistently show that the diameters of the assembled MC fibrils do not change with varying MC solution concentration (<3 wt%) and MC chain length. To understand why and how MC chains pack into fibrillar networks with consistent fibril diameters, we have conducted multi-scale molecular dynamics (MD) simulations. First, we used coarse-grained (CG) MD simulations in implicit water to reveal the self-assembly of MC chains. We found that the MC chains assemble into fibrillar network with consistent fibril diameters regardless of the concentration or chain length; further, in all the fibrils the MC chains align parallel to the fibril axis, with a few chains adopting twisted conformations. To understand the role of explicit water on these conformations, we then conducted atomistic (AA) MD simulations with explicit water. We observed that increasing temperature and methylation increases the twisting of MC chains suggesting the role of hydrophobicity and bulkiness of methyl groups in stabilizing the twists. We conjecture that the MC chain twisting entropically limits the number of chains that assemble into fibrils, and as a result, the fibril diameter is limited to a certain value regardless of chain length or concentration. In ongoing work, we are using these AA MD simulation results to reparametrize CG models of MC chains for varying extent and pattern of methylation along the MC chain to predict assembly and phase behavior for new designs of MC chains not yet studied experimentally.