New Computational Method To Simulate Behavior of Cellulose Fibers
Cellulose fibers provide the structural framework for plant cell walls and are critical for plant growth, stability, and normal function. These same properties of cellulose fibers are also the main obstacle for efficient conversion of biomass to biofuels. Molecular dynamic simulations can aid in understanding cellulose fiber crystallinity and its resilience to deconstruction; however, since the fibers are very large, realistic molecular simulations require extensive run times on leadership-class supercomputers. Recently Scientific Discovery through Advanced Computing (SciDAC) supported researchers at Oak Ridge National Laboratory, in collaboration with RIKEN National Lab in Japan, developed a coarse-grained simulation method termed REACH (Realistic Extension Algorithm via Covariance Hessian) that will enable more efficient simulation of large cellulose fibers. The REACH method reduces the complexity of the simulation (coarse graining) and directly relates molecular force parameters from the more complex all-atom simulation to the faster REACH simulation. Using this method, the researchers simulated the behavior of a cellulose fiber of 36 chains and 40 to 160 degrees of polymerization with a speed of up to 24 nanoseconds per day of computation. The REACH simulations are in agreement with previous findings that the hydrophobic face of the cellulose fiber is more easily deconstructed than the hydrophilic face. An extension of REACH is now being developed that will account for larger amplitude strand separation motions of the fibers thought to precede subsequent deconstruction.
Glass, D. C., K. Moritsugu, X. Cheng, and J. C. Smith. 2012. “REACH Coarse-Grained Simulation of a Cellulose Fiber,” Biomacromolecules 13(9), 2634-44. DOI: 10.1021/bm300460f.