AMHERST, Mass. – One would think that scientists had long ago cracked the secret of cellulose, the most abundant polymer on Earth, in order to break its chemical bonds and harness its wealth of energy. But in fact, only recently have theoretical chemist Scott Auerbach and colleagues at the University of Massachusetts Amherst discovered how cellulose chains break down with heat, which is critical information for efficiently converting cellulose to biofuels.
Reporting in the current issue of the Journal of the American Chemical Society, Auerbach and chemical engineer Paul Dauenhauer, with others, for the first time model at the molecular level the activation energies needed for the chemical reaction known as “fast pyrolysis” to proceed in cellulose. The model meets the tight strictures of chemical accuracy, within 5 kilojoules per mole of cellulose. “We’re quite sure that experiments testing our model will confirm it,” says Auerbach.
A basic building block of plants, cellulose is a naturally occurring crystalline polymer carbohydrate that can take many forms but is usually rigid, like uncooked spaghetti. Researchers have tried for years to convert the abundant, cheap material to biofuels and valuable chemicals, using heat to break or “depolymerize” the chemical bonds to yield a vapor, the necessary precursor to biofuels. But the process has been unpredictable, with different outcomes derived from different heating protocols.
“No one knew how cellulose thermal depolymerization works at an atomic scale,” Auerbach says. “There had been a great deal of uncertainty and controversy about the fundamental reactions and processes. But for the first time our theoretical calculations reveal the dynamics of these bond-breaking events. Given this new knowledge, we can begin to build a picture for how cellulose depolymerizes and how it can be done better and more efficiently.”
Results of the UMass Amherst team’s modeling are presented in a table with 18 reaction pathways at two temperatures, 327 and 600 degrees C.
Overall, Auerbach says, “The key chemical feature that holds cellulose together, hydrogen bonding, can also be the seed of its own destruction.” That’s because below about 260 degrees Celsius (500 degrees Fahrenheit) hydrogen bonds hold the cellulose chains together in much the same way as they hold water molecules together in ice. But the UMass Amherst team found that above this temperature, these same hydrogen bonds begin to insert themselves between other atoms, promoting, or catalyzing, the breakup of chemical bonds.
Auerbach points out, “We modeled not only these 18 pathways, but the activation energy for each. So now we know where the transition from the intermediate liquid to vapor takes place. The table shows the different kinds of processes and transition points for each in kilocalories per mole, at a particular temperature. The amazing new thing is that hydrogen bonding keeps cellulose intact in one situation and catalyzes its own destruction in another. Cellulose carries the key to its own destruction within its chemical makeup.”
From previous experimental work, it was known that rapid heating, or fast pyrolysis, turns cellulose into a mysterious but unknown new phase dubbed “active cellulose,” with properties different from cellulose, but still not the desired vapor. Auerbach says, “This active cellulose stuff is very difficult to characterize, with different heating methods yielding different results, in some cases solid and reversible, and in others liquid and irreversible. Nobody knew what this ‘active cellulose’ really is.”
As a result of the modeling, however, he and colleagues now know cellulose’s secret, the pathways and barriers by which cellulose chains are likely to break. To make these discoveries, the researchers applied special Car-Parrinello molecular dynamics (CPMD) software, named after its developers, in a new way. This accomplishes two tasks that are usually done separately: Making accurate quantum calculations of chemical energies and performing exhaustive sampling of atomic configurations.
Auerbach explains, “The cellulose pyrolysis mystery was a classic: to engage in accurate calculations we needed to know which atomic configurations to consider, but to know which atomic configurations were important we needed accurate calculations. CPMD allowed us to do them both in one shot.”
But still, Auerbach adds, one problem remained to prevent the cellulose mystery from being solved. Simulations can sample molecular motions only for short times, very much shorter than those required to observe bond breaking. To address this, Auerbach and colleagues applied a very recently developed method for accelerating the dynamics while still getting the right answer. “It worked,” he says. “Accelerating the dynamics was the key, the technical breakthrough making this discovery possible.”
At last confirming their simulations, the researchers found the statistically most likely product of cellulose pyrolysis to be the compound levoglucosan (LGA), which also is the major product in experiments. “We knew we were on the right track when we saw LGA,” says Auerbach. He and Dauenhauer are applying these new insights with their colleagues to find more efficient ways to make biofuels from cellulose.