Sometimes a breakthrough in fundamental research is placed on a shelf or filed away in a journal article until someone finds a practical use for it years or decades later. But when Jiamian Hu, an assistant professor of materials science and engineering at the University of Wisconsin-Madison, and graduate researcher Shihao Zhuang modeled a new thin-film heterostructure material that can emit signals in the terahertz spectrum, the light bulb came on almost immediately.
“We started with a fundamental study, but we saw this feature that could be very useful in terahertz imaging,” says Hu. “So I tried to think of ways this could be helpful and realized it could be used in public spaces like subway stations, concert halls and sports stadiums to detect explosives.”
Terahertz imaging is an emerging nondestructive method for analyzing materials. With wavelengths between 1 millimeter and 100 microns, it’s also effective for security applications, since terahertz waves can peer inside bags or packaging material without producing ionizing radiation, like x-rays. Terahertz imaging can also be used for spectroscopic identification of chemical species, reading the various wavelengths emitted by a material’s atoms and molecules to sniff out its chemical fingerprint nondestructively.
Jiamian Hu and Shihao Zhuang believe their terahertz emitter has security applications for large public spaces like subways and stadiums.
The problem is that current terahertz emitters only produce broadband terahertz waves, which are not exact enough to chemically identify compounds. By contrast, Hu and Zhuang’s design produces a narrowband terahertz wave that permits accurate spectral identification of a wide variety of chemical species. For example, it allows for differentiating a bag of sugar from a satchel of ammonium nitrate.
Using computational analysis, the researchers developed a method for producing the narrowband waves using a laser to zap a unique material called a multilayer thin-film heterostructure. The laser causes strain in its metal first layer, which also produces an acoustic wave. A thicker dielectric layer then dissipates the laser’s heat and allows the wave to continue on to the third magnetic layer. There, the wave interacts with the magnetic layer’s spin state to produce a narrowband terahertz signal through spin-charge conversion in a heavy-metal overlayer.
“Such a heterostructure can be fabricated by thin-film heteroepitaxy,” says Hu. “One of the challenges for fabrication is choosing the right materials and the right thickness for each layer. If we just tried it out experimentally, it would take a lot of time and effort. We created a very good computational model to design the entire heterostructure so we know what metal, what dielectric material, which magnetic material, and what heavy-metal overlayer to choose and what thickness to make them.”
Such heterostructure design requires a computational model that is both fast and accurate, something that is quite challenging for a multiphysics, multiphase material system. “To accurately simulate the multiple physical processes involved for converting an optical laser to a narrowband terahertz wave, we need to have a versatile computational model that considers interactions among a lot of different physical systems, including thermal, mechanical, magnetic and electrodynamic systems,” says Zhuang.
“The model also needs to be fast enough to perform high-throughput screening of materials, layer thickness and lateral size for heterostructure design. Equipped with state-of-the-art numerical and parallelization algorithms, we were able to develop a fully benchmarked, graphics processing unit (GPU)-accelerated multiphysics model that can reach up to 100 times speed gain in one single central processing unit (CPU).”
Their model showed that a four-layer aluminum, gadolinium gallium garnet, yttrium iron garnet, and platinum heterostructure is efficient at producing the narrowband waves. Future research is likely to reveal even more efficient combinations.
Hu and Zhuang are now working with Chang-Beom Eom, Raymond R. Holton Chair for Engineering and Theodore H. Geballe Professor in materials science and engineering, to fabricate the heterostructure thin-film emitter. Hu says they are getting close to producing a high-quality prototype device. Then they will collaborate with device engineers to test and measure the emitter before making more refinements.
Hu says he imagines that a security device based on the emitter would act something like a large smoke detector, searching an airport or subway station for the chemical signatures of dangerous materials in crowds without anyone noticing.
“Poor accuracy and potential privacy violations have been hurdles for terahertz technology,” says Hu. “We believe that the narrowband terahertz pulse like our device produces solves these two problems at once, because it’s accurate and only senses threats. It doesn’t image everything else.”
The research was published in 2021 in the journal ACS Applied Materials & Interfaces. In collaboration with the Wisconsin Alumni Research Foundation (WARF), Hu and Zhuang have filed and received two US Patents on this technology, US11112355B2 and US11199447B1. WARF also awarded Hu and Zhuang the 2021 WARF Innovation Award for the project.
Other authors include Peter B. Meisenheimer and John Heron from the University of Michigan. The UW-Madison authors acknowledge support from NSF award CBET-2006028, the Accelerator program from the Wisconsin Alumni Research Foundation, as well as the Pittsburgh Supercomputing Center for supercomputing resources through allocation TG-DMR180076, which is part of the Extreme Science and Engineering Discovery Environment (XSEDE) and supported by NSF grant ACI-1548562.