The new method makes large InSe wafers that work better than silicon. This can help build faster, smaller, low-power chips for future devices.
Researchers at the International Center for Quantum Materials, Peking University, in collaboration with Renmin University of China, have developed a method to produce large-area, high-quality wafers of two-dimensional indium selenide (InSe). This advancement marks a significant leap toward building faster, smaller, and more energy-efficient chips—surpassing silicon in critical performance metrics.
Using this new process, the team fabricated InSe transistors that operate efficiently at room temperature, achieving record electronic performance. These devices showed electron mobility up to 287 cm²/V·s and an average subthreshold swing of 67 mV/dec. Even at sub-10 nm gate lengths, the transistors demonstrated reduced drain-induced barrier lowering (DIBL), low operating voltages, high on/off current ratios, and efficient ballistic transport. Notably, the results exceeded the 2037 IRDS targets for delay and energy-delay product (EDP), positioning InSe as a strong alternative to silicon for future electronics.
The breakthrough is based on a new “solid–liquid–solid” (SLS) conversion method. Researchers began by depositing an amorphous InSe film on sapphire using magnetron sputtering. A layer of low-melting-point indium was added, and the entire structure was sealed inside a quartz enclosure. When heated to about 550°C, the indium created a controlled local environment that enabled uniform dissolution and recrystallization, resulting in single-phase crystalline InSe films.
This process successfully produced 2-inch wafers with excellent crystallinity, phase purity, and thickness uniformity—overcoming longstanding challenges related to the unstable vapor pressure and multiple phases in the indium–selenium system. Traditional methods, by contrast, have only been able to create small flakes unsuitable for large-scale applications.
Indium selenide, often called a “golden semiconductor,” offers a valuable combination of a suitable bandgap, low effective mass, and high thermal velocity. But until now, the difficulty of growing large-area crystals has held back its practical use.
With silicon nearing its physical limits and Moore’s Law slowing, the success at Peking University represents a key step in the search for new materials to power next-generation devices. The reviewers described this work as “an advancement in crystal growth,” underscoring its potential to impact fields ranging from AI and autonomous vehicles to smart devices and beyond.
