Steven Lopez is looking for a molecule that can create a new material to take carbon dioxide out of the atmosphere. The task poses many challenges, but foremost among them is: Where to begin? 

“Imagine all the grains of sand on earth and under the oceans. Now multiply that by a million. That’s how many possible molecules there are,” says Lopez, assistant professor of chemistry and chemical biology at Northeastern. Put another way, there are about 10,000,000,000,000,000,000,000,000 potential molecular combinations to choose from, far more than any one computer can analyze. 

“Right now, we can have about one billion molecules in the database,” he continues. “That’s enough to at least scratch the surface.”

Lopez will conduct this research as one of four principal investigators in the newly launched Institute for Data-Driven Dynamical Design. The goal of the institute, which is funded by the National Science Foundation, is to use machine learning to sift through information and discover new sustainable materials.

The idea for the group started two years ago at what Lopez calls “summer camp for professors.” At an event hosted by the NSF, groups of researchers from different scientific backgrounds were randomly selected to work together to identify a problem and propose a potential solution using data analysis.  

After five days of collaboration, Lopez’s group pitched their proposal to the NSF, sowing the seed of what is today the Institute for Data-Driven Dynamical Design. 

One of the main strengths of the institute, Lopez says, is its multidisciplinary approach to materials science. Among the founding members, the fields of chemistry, data visualization, physics, ontology, and machine learning are all represented. 

Lopez’s role will primarily focus on molecular chemistry, but the overall scope of the institute is much larger than that. “We’re studying materials for things ranging from the impossibly large to the impossibly small,” he says. “From bridges to molecules, essentially.” 

The institute is also interested in how those materials change depending on circumstance. For a bridge, that could mean studying how it fairs in different seasons or during earthquakes. For a single molecule, that might involve looking at how it functions under certain levels of light or heat. 

Lopez is equipped to help the group discover the building blocks for more sustainable materials because of his background in photochemistry, a field concerned with how molecules react to light. Molecules that use light rather than heat as a catalyst for chemical reactions can lead to more efficient forms of energy or materials. 

Take a typical chemistry experiment, for example. Liquid A is poured into Liquid B then heated over a Bunsen burner to cause a reaction. “On the classroom level, it’s not a big deal to heat up one liter of liquid. But on an industrial scale, it takes a lot of energy to heat up, say, a thousand liters of a solution,” he says. Sunlight can serve as a faster and more efficient alternative to heat. 

Aided by the data analysis tools of the institute, Lopez hopes to combine his expertise on sunlight-powered chemistry with the computing power of machine learning to cover more ground and work toward creating more sustainable materials overall. 

For media inquiries, please contact media@northeastern.edu.