New materials are critical components of emerging technologies that promise to be major growth areas for the economy, such as less expensive solar power, electric-car batteries that can go longer between charges, lightweight portable electronic devices, and implantable medical devices for personalized medicine. But the journey from new material to product typically takes one to two decades. That’s in large part because new materials require advanced manufacturing technologies that can take many years to develop.
The White House hopes to cut that time in half by investing $100 million in a Materials Genome Initiative aimed at encouraging more efficient use of the computational modeling tools that researchers use to predict the properties of new materials. The initiative, which is part of the White House’s Advanced Manufacturing Partnership, will support open access to these models and databases across the materials science community in hopes of connecting academics with industry earlier in the development process.
As it stands now, scientists working with new materials don’t take manufacturing issues into account early enough, says Cyrus Wadia, assistant director for clean energy and materials R&D at the White House Office of Science and Technology Policy. As a result, their research can lead them into dead ends. The way to change that, he believes, is to encourage the whole materials science community, from academics to manufacturers, to share data and computational tools—the “materials genome.” Wadia says he wants researchers to ask themselves, “Who’s done it before, what did they learn, and what can the market bear?”
Materials scientists have been using predictive models with varying degrees of success over the past 20 years, manipulating data about properties such as melting point, conductivity, or the way a compound reacts with others to predict whether a material is suitable for a particular application such as a battery electrode. The computations involved are very complicated. But once the code to predict promising candidates for a particular application is written, it can be applied to test the potential of any material, says Gerbrand Ceder, a professor of materials science at MIT who specializes in computational modeling of new battery-electrode materials. Unfortunately, there’s been no infrastructure to help researchers share their data and the code used to crunch it, and few of the models have taken manufacturing issues into account.
“The problem with scaling and manufacturing is that you don’t understand everything,” says Ceder. “If we could make things exactly how we made them in the lab, there’d be no problem.” But it doesn’t work that way. Minor differences in manufacturing conditions are inevitable when scaling up from making grams of a material to making it by the ton. And the materials coming out of academic labs today are harder to make than the materials of the past. Many advanced materials gain their extraordinary properties through molecular or even atomic-scale structural precision, and making them is not like making, for example, steel. “You make steel by melting metals together in a huge vat,” says Alexander King, director of the Ames National Laboratory in Iowa. In manufacturing advanced materials, says King, “you have to use more controlled methods, or the atoms won’t do what you want.” Inconsistencies in temperature control, mixing, or other factors can lead to failure. And techniques used to achieve atomic-scale precision in the lab can be difficult to translate to large-scale manufacturing.