Speeding Toward Discovery

Chemical engineers employ new, efficient method to study catalysis

A catalyst is a substance that speeds up or steers a reaction without itself becoming consumed. Catalysts are the specialty of W. Nicholas Delgass, the Maxine Spencer Nichols Professor of Chemical Engineering, who leads the catalyst design group in the School of Chemical Engineering. You could call him a catalyst in the advancement of the science of catalysis, but the pun would be slightly off-base; this scientist, unlike a chemical catalyst, is indeed consumed by his work.

Delgass and his team have been successful in attracting major funding for the pursuit of their scientific passion, including two three-year grants from the Basic Energy Sciences Catalysis Sciences program for Catalyst Design by Discovery Informatics within the Department of Energy. The first was for $2.2 million and the second was for $2.4 million. The six-year program will be up for renewal in September 2009. They also have worked with ExxonMobil for the past four years, receiving $297,000 from the company in the most recent year.

In his work leading the catalyst design group, Delgass collaborates with fellow chemical engineers: Jim Caruthers, Fabio Ribiero, and Venkat Venkatasubramanian, professors; Kendall Thomson, associate professor; and Chelsey Baertsch, assistant professor. The group is investigating an exciting new methodology for studying catalytic reactions called "discovery informatics."

"Discovery informatics is a framework that enables management of complexity, accumulation of knowledge, systematic testing of hypotheses by interaction with experiments, and the efficient search for new materials with desired performance characteristics," Delgass says.

The ultimate goal is to make catalysis a predictive science. Currently, designing a new catalytic reaction or improving a known one is essentially accomplished by trial and error. The group aims to improve these "best-guess" odds. One step is to find higher level descriptors for the materials uses in a given reaction. "A descriptor must have a quantitative value, and it must carry enough information so that I can make the catalyst," Delgass says.

Systems involved in a catalytic reaction are very complicated, so the next huge piece of the catalysis puzzle will involve building the so-called "forward model" to act as a kind of bridge between descriptors and the performance outcome. "It's a huge job. No one person can do it," says Delgass, who explains that a team of chemical engineers in consultation with chemists, computer scientists, and others will develop a forward model. They then will run the reactions using catalysts consistent with the model.

If the performance is as hoped, all the necessary information is already in the model. If the model is wrong, they will use the new data to revise the model, which becomes a continuous "knowledge archive." Computers use the new forward model to predict the best catalysts, which will be run next. The model-building process itself contributes to the fundamental science of catalysis.

This scientific streamlined method of extracting knowledge from data will speed the discovery of new materials and will play a crucial role as engineers tackle some of the grand challenges facing our environment and our economy. Delgass and Ribeiro are using discovery informatics in collaboration with Rakesh Agrawal, the Winthrop E. Stone Distinguished Professor of Chemical Engineering, in his quest to turn nature's leftovers—grass clippings, farm waste, and forest debris—into fuel.

The process, called H2Bioil, will use solar or nuclear power to produce hydrogen that, in turn, will remove the oxygen from the biomass to make diesel fuel. There is little literature on the process, called hydrodeoxygenation (HDO), to date, Delgass says. He plans to contribute to the knowledge base by applying the new methods he and his team are perfecting at Purdue. "Discovery informatics will be used to discover new HDO catalysts," he says with excitement. "There's enough sustainable biomass, under proper conditions, to fill the entire transportation fuel need of the nation."

-Gina Vozenilek