From Theory to Application: How the DMREF Program Helps Accelerate Advanced Materials Research
What began as a small program with roughly 14 projects has now expanded to more than 200 projects. DMREF teams from universities across the U.S. collaborate to discover, develop, and help manufacture new materials.
By Divya Abhat
The year 2020 was dark for much of the world engulfed in a pandemic. However, it was also the year that researchers developed the world’s brightest fluorescent material.
In the solid state, traditional fluorescent dyes are unable to glow to their full potential as their molecules stick to each other and, in the process, lose their brightness. To work around this issue, researchers at Indiana University, the University of Copenhagen, and the University of Southern Mississippi mixed positively charged fluorescent dyes with a solution containing donut-shaped molecules—a combination that produced crystals referred to as small-molecule ionic isolation lattices or SMILES. In this crystal structure, molecules in the dyes not only stay separate from each other but can also maintain their brilliance.
SMILES is believed to be 30 times brighter than the fluorescent materials that are currently used in medical imaging, and researchers hope the new material could be used in medical dye lasers to treat a range of medical conditions including blemishes and wounds.
“Light capture, light upconversion, and light emission are all areas where advanced materials based on the SMILES platform could have a substantial positive impact,” says Amar Flood, a chemistry professor at Indiana University and researcher who worked on developing SMILES. “The significance and value of advanced materials are to ensure the sustainable stewardship of our planet, such that we can make efficient use of materials and energy flow.”
A Materials Initiative is Born
Materials are used in every facet of our lives. Their advancements are key to developing new tools and technologies, which go far in improving our lives as individuals and as a society. With this overarching premise in mind, the Materials Genome Initiative (MGI) for Global Competitiveness in the United States was launched in 2011, stimulating a nascent archetype for integrating computation and experiment in materials research. The goal of the MGI was to create a highly collaborative environment that enabled the use and progress of computational and experimental tools as well as digital data—all of which could be applied to efficient, expedited, and lower-cost development of advanced materials.
Several federal agencies joined the effort including the National Science Foundation, which created the DMREF program, short for Designing Materials to Revolutionize and Engineer Our Future. Flood’s research on fluorescent dyes is a product of the DMREF program, which was launched in 2012. Today, nearly 20 federal agencies including the Department of Energy, the Department of Defense, and the National Institutes of Science and Technology are part of the effort and have invested more than $1 billion in resources and infrastructure. They each strive toward developing and strengthening the country’s materials innovation infrastructure, achieving national goals with the help of advanced materials, and laying the groundwork for continued efficient and effective use of these materials.
Building a Solid Foundation
MGI has three overarching objectives—building a strong materials innovation infrastructure, achieving national goals with advanced materials, and establishing a strong foundation for continued use of these materials. While all of these components are a critical part of the DMREF program, a lot of focus is placed on workforce and infrastructure development. This includes developing a feedback loop between experimental and computational research to improve software and computational tools for the integrated design of materials. “There are tools that need to be used,” says Michael Chabinyc, Professor and Department Chair of the Materials Department at the University of California, Santa Barbara and a member of the DMREF program. “For example, you can collect thousands and thousands of images of data but then how do you organize them? How do you extract and archive information, let alone interpret it? These are convergence issues that are common, and the program will help with that.”
Anything that is developed is required to be open to the community. As a result, people can not only share their databases, but also their code along with techniques and strategies. “There’s always a discussion about how to share approaches across sub-disciplines,” says Chabinyc. “Methods being used to develop metals could end up helping people studying soft materials. Having repositories for common codes ensures that there will be crossover.”
In order to build a strong infrastructure, researchers also rely on developing experimental tools that help create and develop models. In fact, this kind of testing complements computational tools by filling gaps that might arise in theoretical frameworks.
Both of these approaches rely on data—another key focus of the DMREF program. As noted in the Materials Genome Initiative for Global Competitiveness 2011 report, “Data inform and verify the computational models that will streamline the development process.”
“The data piece was envisioned as being a key piece of MGI, and that’s probably one of the fasted growing aspects of DMREF and MGI,” says John Schlueter, NSF Program Director in the Division of Materials Research. Schlueter manages the DMREF program and works with more than a dozen NSF program managers to coordinate efforts across multiple NSF divisions. “Our objective has always been to be able to make the digital output of the project—the code, the experimental data, the computational capabilities—available to the community to be able to use as they see fit.”
A team of researchers at the Massachusetts Institute of Technology, the University of California, Berkeley, and the University of Massachusetts Amherst are doing their part to not only help sift through swathes of data but also help make data easily accessible. Scientists traditionally have had to spend hours sifting through published journal papers, reports, and articles, and painstakingly highlighting relevant sections of literature. In the last decade, however, computational systems have evolved to scan thousands of journal papers. As part of the DMREF program, the team of researchers tapped into that capability and created a natural language processing system that can scan millions of papers on inorganic materials synthesis to gain insights into how to accelerate the discovery of new materials or optimize synthesis of known materials. Based on what the extracted data showed, they were then able to not only better understand the structure of inorganic materials but also predict how they could be made, helping create a recipe of sorts.
Within this effort, referred to as the Synthesis Project, the team has focused on a case involving tens of thousands of journal papers on zeolites—a kind of porous material primarily comprised of aluminum and silicon. As a catalysis material, zeolites have a lot of industrial relevance, however, very few zeolites are commercially available. In fact, there’s a sizable gap between the hundreds of thousands of zeolites that are theoretically predicted to be synthetical stable and the actual number that are in use. “There’s a playground that’s very ripe for new materials design,” says Elsa Olivetti, a materials science and engineering professor at MIT who is part of this project. She notes that more research would also bring down that large number of potentially stable zeolites on the theory side and subsequently help close the gap on both ends.
For Schlueter, the growth and expansion of tools like artificial intelligence, machine learning, natural language processing, and data mining, is especially exciting. “These are all focused on computational-led, data-driven type research that’s able to accelerate this process of taking fundamental research to deployment,” he says.
The Power of Collaboration
Perhaps the most crucial component of MGI is collaboration, not only across sectors such as universities, government agencies, and small and large companies but also among experimental and theoretical scientists as well as computer and industrial engineers—all with diverse backgrounds and skills.
The idea was to break apart the old approach, which was for researchers to work in silos and potentially swap notes only after the research was well underway, if at all. Instead, says Chabinyc, MGI sought to “come up with a scheme where the theorists and experimentalists would work together from the beginning. And they would iterate to not only improve the theory, but to also use the theory to [provide feedback on] the experiment at an early stage to be more predictive.”
"DMREF supports a holistic approach to materials research," says Bryan Boudouris, Program Director for NSF’s Division of Materials Research. "The philosophy of this program inherently fosters partnership among academic, federal laboratories, and industrial entities."
The benefit of these kinds of partnerships is evident in the recent and timely effort to address surface contamination. In 2019, researchers at Penn State University and the University of Minnesota joined forces with Japan’s Tohoku University and the University of Tokyo to determine whether a new class of materials known as correlated materials could be effectively incorporated into handheld UV light devices.
UV rays have long been used to kill bacteria and viruses that might be found on surfaces, offering a particularly promising solution that could be used to tackle the novel coronavirus. But current UV sources come in the form of mercury-containing lamps and, as a result, are bulky, expensive, and tend to use a lot of power.
While researchers saw UV LEDs as a potentially good fit, they needed to identify an electrode material that is not only transparent to visible light but also a strong conductor of electricity. Correlated materials—strontium niobate in particular—seemed to fit the bill. Using a thin-film coating technique known as sputtering, researchers were able to create an extremely thin film of strontium niobate—10,000 times thinner than a strand of hair—that had all the required physical properties to be transparent to UV light as well as conduct electricity. While still in the research phase, this early breakthrough paves the way towards UV-LED based handheld devices that could be used to sanitize public surfaces in stadiums, transport hubs such as train stations and airports, as well as the vehicles themselves.
What's more, according to Roman Engel-Herbert, a Penn State associate professor who is part of the research effort, "the value of this discovery to have a high-performance UV transparent conductor material like strontium niobate potentially allows us to replace the bulky, mercury-containing gas discharge lamps with UV LEDs, offering an environmentally benign and low power consumption solution with longer lifetimes, smaller form factor and instant on/off response."
“These examples highlight the potential of the DMREF to address some of the most pressing global challenges of the day through deliberate innovation with an eye towards translation through strategic and impactful partnerships,” says Boudouris.
DMREF Today and Beyond
What began as a small program with roughly 14 projects has now expanded to more than 200 projects. DMREF teams from universities across the U.S. collaborate to discover, develop, and help manufacture new materials. These teams often find collaborators at other federal institutions such as national laboratories associated with the Department of Energy, the National Institute of Standards and Technology, the Department of Defense laboratories—and most recently, the Air Force Research Laboratory. "The future of DMREF is one that continues to expand the vision of MGI such that its impact is felt on myriad fronts," says Boudouris.
When reviewing applications, the DMREF Management Team seeks to determine if there is a vision for the application of a particular advanced material. “We’re trying to understand if this were successful, how would it find its way into a system that would have a significant impact on society,” says Schlueter. As part of that process, it’s key for the computational data aspect of a project to make predictions based on which materials can be synthesized and validated. “Having an effective feedback loop is key,” he says.
Competitions are held every two years, with the next competition scheduled for 2023. Projects are four years in duration. Each project has a team of at least two Principal Investigators and is typically comprised of approximately 10-12 people, including graduate students and a small number of postdoctoral researchers. As part of the effort, DMREF also ensures that there is educational and workforce development and that these researchers receive training in a multidisciplinary fashion. As all data is mandated to be readily searchable, accessible, and useful to the community, reviewers also pay close attention to each project’s digital output plans.
A community that works together and learns from its successes and failures can achieve great results. And whether those results translate into high quality production software toolkits, better energy storage capabilities, more refined imaging, or sturdier electronics, it all begins with a multifaceted, interdisciplinary team of researchers who come at a single problem from different directions.
“It’s critical that materials find their way into real-world applications. That’s DMREF’s goal to be able to do that,” says Schlueter. “But we can’t do it all ourselves—collaboration and partnerships are key. It’s all part of the MGI vision.”