Cornell Team Sets Semiconductor Bandgap Record Above 7 eV Using Silicon-Doped Oxide on Sapphire

A team at Cornell University has built a semiconductor film with a wider bandgap than any before it that can still carry current, a combination that could lead to power electronics that waste far less energy as heat. The work, led by Darrell Schlom, was published in Nature.

A bandgap is the energy an electron needs to jump from the valence band, where it is bound tightly to its atom, into the conduction band, where it moves freely and carries current. A wider gap lets a material withstand stronger electric fields, which is why wide-bandgap semiconductors such as gallium nitride and silicon carbide already appear in fast phone chargers and electric vehicles.

Aluminum gallium oxide is an especially promising candidate. Changing its ratio of aluminum to gallium tunes the bandgap anywhere from 5.4 to 8.6 electronvolts, the widest range of any known semiconductor. The problem has been that the more aluminum the film contains, the harder it is to supply free electrons. Until now, aluminum-rich films conducted electricity several orders of magnitude more poorly than gallium-rich ones, which made them impractical for devices.

Schlom's team grew the films on a sapphire substrate using molecular-beam epitaxy, a technique that builds a material one atomic layer at a time. As each film grew, they added small amounts of silicon, whose atoms donate spare electrons. They then measured the films' bandgaps and electrical properties and tested them in two key components, a diode and a transistor.

The silicon-doped films conducted electricity at bandgaps above 7 eV, beating the previous record held by cubic boron nitride. Above 6 eV, a range the team calls "colossal bandgap," the films conducted at room temperature more than 100 million times better than any previous report. Both the diode and the transistor worked with these materials. The paper is titled "A 7-eV bandgap semiconductor based on silicon-doped α-(AlxGa1−x)2O3" and carries the DOI 10.1038/s41586-026-11085-3.

The substrate matters for cost. Other ultra-wide-bandgap materials, such as diamond and cubic boron nitride, are hard to produce at scale. Sapphire, by contrast, is abundant, cheap and already manufactured at high quality in very large volumes. That could make the approach far easier to bring to production than the alternatives.

Schlom's team now hopes the films can open the way to a new generation of power electronics, the devices that convert and control electricity in everything from chargers to the electric grid. Because wider-gap devices can handle far higher voltages in smaller packages, the payoff would be circuits that run cooler and lose less energy in conversion. The work shows the main obstacle, getting electrons into an aluminum-rich oxide, can be overcome at least in thin films, though moving from lab films to commercial parts is the next test.