In Massachusetts, University of Massachusetts Amherst computational chemist Scott Auerbach has been awarded a three-year, $330,000 grant from the National Science Foundation to improve basic understanding and optimize the process of producing fuels such as gasoline from plant biomass instead of from petroleum.
For more than 50 years, Auerbach explains, chemists and chemical engineers have refined crude oil using synthetic catalysts inspired by minerals known as zeolites. Zeolites feature uniform arrays of tiny, molecule-sized pores with the right size and shape to convert the hydrocarbon molecules that make up petroleum into high-octane gasoline, in a process Auerbach calls “shape-selective catalysis.”
“Shape-selective catalysis is a relatively old technology, from the 1970s; this is still how our gasoline is refined today,” he says. “Zeolite can handle the long carbon chains of crude oil, cutting them into shorter pieces, re-forming them into high-octane gasoline.” Auerbach and others who study zeolites across the globe are asking whether zeolites may also be used to convert plant biomass such as cellulose into gasoline.
The challenge in applying zeolites to make biofuels, he says, is that biomass is generally made of carbohydrates, sugars such as glucose and starches such as cellulose, and these react very differently within the pores of zeolites. Auerbach plans to apply computational chemistry, using computers to reveal the microscopic structures and motions of molecules, to understand how carbohydrates react in zeolite pores.
Another problem that must be overcome, in both petroleum and biomass refining, he adds, is that zeolites promote side reactions that create sheets of carbon also known as “coke.” It can deactivate zeolite catalysts by clogging their microscopic pores, like enveloping a sponge in plastic wrap, shutting down the refining process.
The problem of coke formation in zeolite-based refining has been around for a long time. Auerbach is taking a fresh look at understanding this process with a battery of new computational techniques. “The process of forming coke is rather complex, but we have a good handle on the kinds of coke molecules that eventually form, so we’ll apply the trick of beginning with the end product and working backwards to reveal how coke forms,” he explains.