258720 Nascent Decomposition Pathways of Cellulose From First Principles

Monday, October 29, 2012: 3:15 PM
315 (Convention Center )
Vishal Agarwal, Chemical Engineering, University of Massachusetts, Amherst, MA, Paul J. Dauenhauer, Department of Chemical Engineering, University of Massachusetts, Amherst, Amherst, MA, George W. Huber, Department of Chemical Engineering, University of Massachusetts Amherst, Amherst, MA and Scott M. Auerbach, Chemistry and Chemical Engineering, Univ. of Massachusetts, Amherst, MA

Fast pyrolysis is a burgeoning technology that converts lignocellulosic biomass to processable bio-oil.[1] Commercializing pyrolysis would require efficient process design, especially reactors as they are one of the most energy intensive units in the whole process. This would in turn require detailed understanding of pyrolysis chemistries. Biomass is mainly composed of the biopolymer cellulose; therefore, understanding complex cellulose pyrolysis chemistries is important for efficiently modeling and optimizing fast pyrolysis reactors. We have modeled nascent decomposition pathways of cellulose at 600 OC using Car-Parrinello molecular dynamics (CPMD)[2] simulations. We used a simulation cell of 4 cellobiose residues periodically repeated in all directions to model decomposition of cellulose matrix. We generated initial configurations by performing classical NPT molecular dynamics simulations on large periodic cells of cellulose[3] and then clipping out the required simulation cell of 4 cellobiose residues. By applying the metadynamics method[4] using multiple sets of collective variables, we have found various possible nascent processes that may occur during pyrolysis such as depolymerization, fragmentation, ring opening, and ring contraction. The nascent processes observed can explain the formation of precursors to major products observed during cellulose pyrolysis such as levoglucosan (LGA), hydroxy-methylfurfurral (HMF) and fragmentation products such as formic acid. We found that a low barrier process to a precursor of LGA is a concertered process with a likely intermediate/transition-state stabilized by resonance and nearby hydrogen bonding. We computed a barrier of 48.6 kcal/mol for this process, which is in excellent agreement with experimentally estimated activation energies, suggesting that depolymerization of cellulose to this precursor of LGA (pre-LGA) is an important, rate-limiting step. We found that free-energy barriers are in the order of pre-LGA < pre-HMF < formic acid. We also found LGA to be thermodynamically more stable than HMF. Kinetic and thermodynamic favorability taken together explain why LGA is the major product observed during pyrolysis.

[1] Y.-C. Lin, J. Cho, G. a. Tompsett, P. R. Westmoreland, G. W. Huber, The Journal of Physical Chemistry C 2009, 113, 20097.

[2] R. Car, M. Parrinello, Physical Review Letters 1985, 55, 2471.

[3] V. Agarwal, G. W. Huber, W. C. Conner, S. M. Auerbach, The Journal of chemical physics 2011, 135, 134506.

[4] A. Laio, M. Parrinello, Proceedings of the National Academy of Sciences of the United States of America 2002, 99, 12562.

Extended Abstract: File Not Uploaded
See more of this Session: Pyrolysis of Biomass
See more of this Group/Topical: Catalysis and Reaction Engineering Division