Physicists unveil new technique to detect tiny defects in semiconductors

Michigan State University combined terahertz laser light, shown as a red wavy arrow, with a scanning tunneling microscope, or STM, tip — the dark pyramid shape exchanging a red electron with a sample shown with a blue surface. Credit: Eve Ammerman.

One of the challenges in creating smarter and more powerful electronics is analyzing the materials that make up these devices with extreme precision.

Physicists at Michigan State University (MSU) have made a significant step forward in this area by developing a method that combines high-resolution microscopy with ultrafast lasers.

This new technique allows researchers to spot single-atom defects in semiconductors with unmatched accuracy.

The technique, detailed in the journal Nature Photonics, enables scientists to detect misfit atoms, or “defects,” in semiconductors.

While the term “defects” might sound negative, these atoms are usually added on purpose to improve the performance of semiconductors in modern electronics.

“This is especially important for components with nanoscale structures,” said Tyler Cocker, the Jerry Cowen Endowed Chair in Experimental Physics at MSU and leader of the study.

These tiny structures are found in things like computer chips and other electronic devices. Researchers are also working on materials that are only one atom thick, which represent the future of semiconductor technology.

“When you have nanoscale electronics, it’s crucial to ensure that electrons can move as intended,” Cocker explained. Defects play a big role in how electrons move, so scientists like Cocker are eager to know exactly where these defects are and how they behave.

Cocker’s team has created a technique that lets them obtain this information with great ease. “One of my colleagues said, ‘I hope you went out and celebrated,'” Cocker recalled.

The project was led by Vedran Jelic, who was a postdoctoral researcher in Cocker’s group and is now with the National Research Council Canada. The research team also included doctoral students Stefanie Adams, Eve Ammerman, and Mohamed Hassan, as well as undergraduate researcher Kaedon Cleland-Host.

Cocker noted that the technique is straightforward to use with the right equipment, and his team is already applying it to various atomically thin materials like graphene nanoribbons. “We’ve got a number of projects where we’re using the technique with different materials,” Cocker said. “We’re basically incorporating it into everything we do.”

Existing tools, like scanning tunneling microscopes (STMs), can help scientists spot single-atom defects. Unlike traditional microscopes, STMs don’t use lenses and light to magnify objects. Instead, they scan a sample’s surface using an atomically sharp tip, similar to a record player’s stylus. This tip doesn’t touch the surface but gets close enough for electrons to jump between the tip and the sample.

STMs gather data about where and how many electrons jump, providing atomic-scale information about samples. However, STM data alone isn’t always enough to clearly identify defects, especially in materials like gallium arsenide, an important semiconductor used in radar systems, high-efficiency solar cells, and telecommunication devices.

In their latest work, Cocker and his team focused on gallium arsenide samples that were deliberately infused with silicon defect atoms to control how electrons move through the semiconductor. “The silicon atom basically looks like a deep pothole to the electrons,” Cocker said.

Although theorists have studied this type of defect for decades, experimentalists have not been able to detect these single atoms directly until now. Cocker’s new technique still uses an STM, but it also shines laser pulses right at the STM’s tip. These pulses consist of terahertz light waves, which oscillate a trillion times per second. Theorists had predicted that silicon defects would oscillate at the same frequency inside a gallium arsenide sample.

By combining STM and terahertz light, the MSU team created a probe with exceptional sensitivity to these defects. When the STM tip reached a silicon defect, a sudden, intense signal appeared in the data. Moving the tip just one atom away made the signal disappear.

“Here was this defect that people have been hunting for over forty years, and we could see it ringing like a bell,” Cocker said. “At first, it was hard to believe because it was so distinct. We had to measure it in every possible way to be sure it was real.”

Thanks to decades of theoretical research, explaining the signal was straightforward once they were convinced it was real. “When you discover something like this, it’s really helpful to have existing theoretical research thoroughly characterizing it,” said Jelic.

Cocker’s lab is leading the way in this field, but other groups worldwide are also combining STMs and terahertz light. Various materials could benefit from this technique for applications beyond detecting defects. Now that the MSU team has shared their method, Cocker is excited to see what other discoveries will follow.