SusChEM: Multiscale Interaction Potentials for Cellulose

University Of Arkansas, Fayetteville AR

Investigators

Abstract

NON-TECHNICAL SUMMARY The Division of Materials Research and the Chemistry Division contribute funds to this award. This SusChEM project involves computational research on the conversion of cellulosic biomass to biofuel as a sustainable energy source and a sustainable feedstock for new materials. Biofuel derived from perennial plants, such as grass, is most desirable since these plants grow on marginal land and can be harvested repeatedly. One major roadblock for economical utilization of plant biomass is the resistance of cellulosic fibrils to pretreatment to facilitate their conversion to usable fuels. Many questions regarding the detailed structures of fibrils and their interactions with water, other chemical solvents, and enzymes are poorly understood. The team will use high quality quantum mechanical computer simulations to develop accurate computational models to describe these interactions, with the ultimate goal of improving the efficiency of biomass conversion and for applications for the discovery and modeling of cellulose-based materials. The research team will engage undergraduate and graduate students in sustainability research which aims to find solutions to enable the balance of carbon emission with carbon sequestration. Students will have the opportunity to visit Oak Ridge National Laboratory and experience research in a government laboratory. The team will develop a self-contained computational chemistry USB memory stick with packages to perform electronic structure and other modeling. The computer programs contained will include many that can be used without a detailed knowledge of molecular quantum mechanics. The team will disseminate the USB memory stick to regional colleges and help the faculty to incorporate modeling in their classrooms. The PI will also develop modeling modules and tutorials to teach concepts in physical and organic chemistry curricula. TECHNICAL SUMMARY The Division of Materials Research and the Chemistry Division contribute funds to this award. Through this SusChEM project, the research team will develop a multiscale model for cellulose fibrils and investigate fundamental properties of fibrils and fiber bundles with application to sustainable energy and the discovery of sustainable cellulose-based materials. The research team will develop an accurate potential for cellulose by fitting to accurate electronic structure forces using the adaptive force matching method. Through an iterative procedure, adaptive force matching provides both high quality reference forces and representative training sets for fitting. This allows accurate force fields to be developed without using very complex energy expressions. As a consequence, larger structures can be modeled efficiently. Once the adaptive force matching cellulose force field is available, an accurate coarse-grained potential will be developed using the multiscale coarse-graining approach. The coarse-grained potential will allow long cellulose fibrils and fibril bundles to be modeled. The cellulose potential will be developed with only electronic structure information as input. The model will be validated to reproduce experimental properties, such as lattice constants and rotamer distributions. The validated all-atom and coarse-grained potentials will be used to address fundamental problems of cellulosic fibrils, including the number of chains in a fibril, the tendency for twisting, rotamer conformations at the interface, free energy of polymorph transformations, and the persistence length of the fibril. The accurate multiscale potentials to be developed will enable reliable modeling of hydrated cellulosic fibrils, and further research in modeling cellulosic biomass and cellulose-based materials. The cellulose force field will be a stepping stone for the developments of additional models that involve alternative solvents, such as ionic liquids, and enzymes. Successful development of an accurate cellulose potential will represent a major advance of adaptive force matching. The development of adaptive force matching into a reliable protocol for mapping an expensive electronic structure potential to a simple molecular mechanics force field will have broad impact for material research in general. This award also supports educational activities to integrate computational modeling into Chemistry education. A self-contained computational chemistry USB memory stick will be developed to facilitate computer modeling by undergraduate students and the use of computer modeling in classrooms. No installation or licensing is needed, allowing the user to focus on the problem instead of computational details. The PI will disseminate the USB memory stick to regional colleges and help the faculty incorporate modeling in their classrooms. The PI will also develop modeling modules and tutorials to teach important concepts in physical and organic chemistry. The PIs will engage undergraduate and graduate students in carbon neutral sustainability research and provide them with the opportunity to visit Oak Ridge National Laboratory and experience research in a national laboratory.