Ultraviolet (UV) light, a powerful type of radiation invisible to the naked eye, has been used for decades to detect counterfeit money; investigate crime scenes; cure coatings, inks and other industrial materials; study DNA; and light up blacklight posters. For years, toxic, bulky, energy-inefficient mercury-based lamps were used to produce UV light, but in the last few decades, small, bio-safe, solid-state LEDs have replaced that technology, getting better and more efficient over time.
But when it comes to one stretch of the UV spectrum—the region between 280 and 365 nanometers known as UVB—technical hurdles mean the LEDs aren’t quite as good. That’s why Shubhra Pasayat, an assistant professor of electrical and computer engineering at the University of Wisconsin-Madison, will use a National Science Foundation CAREER award to design and fabricate new types of UVB LEDs for use in water and medical facility sterilization, vertical agriculture, food safety and other potential applications.
LEDs are actually semiconductors that release light when voltage passes through them. The wavelength of light produced depends on the semiconductor material the LED is made from and how it is oriented. When it comes to UV LEDs, versions of gallium nitride materials produce LEDs in the UVA range, while aluminum nitride makes UVC LEDs possible. These materials are grown in extremely thin layers on top of a substrate, like sapphire, which has a similar crystalline lattice structure at the atomic level. The more the lattices between the semiconductor and substrate match, the more efficient the LED.
Currently, however, there is no known substrate with a lattice matching the crystal semiconductors needed to produce efficient UVB LEDs. But Pasayat believes she has found a workaround. “What I’m proposing is to use a new class of material called porous nitrides which can reduce some of this lattice mismatch,” she says. “Because the material is flexible, I can deposit aluminum or gallium nitride, but stretch the spongy material underneath them and get rid of this problem.”
Pasayat says that the best commercially available UVB LEDs currently have an average efficiency of around 5%. In this project, she says she plans to improve that efficiency tenfold.
Pasayat is working with an industry partner to make sure her work is addressing real needs. “We always try to make sure our devices can readily get into the technology commercialization cycle,” she says. “I’m working with a partner to first assess the societal impact of the solution I’m proposing. And based on those assessments we’ll try to gauge interest in wavelengths that are more important for certain applications, whether it is medical or food industries.”
Key to the project is a new state-of-the-art aluminum nitride reactor the UW-Madison College of Engineering will bring online later in spring 2024.