Diamond Is Almost the Perfect Power Semiconductor. Doping Is the Problem

Diamond is almost unfair on paper. It has a bandgap of about 5.47 electron volts, extremely high thermal conductivity, and a breakdown field far above silicon. Those properties make it an obvious candidate for power electronics that must handle high voltage, high temperature, and high switching density. Yet diamond has spent decades remaining more impressive as a material than as a semiconductor industry. The reason is not that researchers cannot make transistors. It is that diamond still struggles with something silicon solved long ago: reliable control of both carrier types. Making p-type diamond is relatively straightforward. Boron can create holes, and hydrogen-terminated diamond surfaces can form high-mobility p-type channels. That is why some of the strongest diamond transistor results are p-channel devices. In March 2026, Japanese startup Power Diamond Systems reported a diamond MOSFET that combined 550 volts of breakdown with 0.8 amps of drain current and demonstrated 200-volt, 1-amp switching. The company later used a diamond MOSFET in a working DC/DC converter. The harder half is n-type diamond. Conventional donors such as phosphorus tend to sit too deep in diamond's wide bandgap, so relatively few electrons become mobile at room temperature. A December 2025 study in Research offered a more complicated route: oxygen-assisted boron-nitrogen codoping. The researchers reported electron concentrations above 10^19 per cubic centimeter and a shallow activation energy of about 21 meV. That is meaningful progress because it attacks the donor problem instead of working around it. But the same result also shows why the problem is not solved. Electron mobility in the reported n-type material was only about 4 cm²/V·s, far below diamond's theoretical transport potential. The useful growth condition was also narrow, centered around roughly 1,020 K. A separate 2026 review still lists n-type doping, defect control, surface stability, and scalable single-crystal growth among the major barriers to commercialization. This creates an unusual split in diamond electronics. Engineers can already build useful p-channel devices and combine them with n-channel devices made from other wide-bandgap materials. That may become commercially useful before an all-diamond complementary circuit ever does. In other words, diamond does not need a perfect n-type transistor to enter power electronics. Still, solving n-type doping would change the material's role completely. It would allow true diamond p-n junctions, complementary logic, and more flexible power-device architectures without relying on another semiconductor for half of the circuit. Diamond's strongest properties were never the mystery. The difficult part has been convincing electrons to behave inside it. In 2026, that problem looks more solvable than it did a decade ago, but it remains the line between a remarkable material and a complete semiconductor platform.