Zetian Mi awarded $3M DARPA grant to scale next generation semiconductor materials on silicon
Bringing the power of today’s supercomputers to next-generation portable and wearable devices like cell phones, wristwatches, eyeglasses, and more, will require some changes to the fundamental components of computer chips. The performance of microelectronics, critical building blocks in every electronic device we use, relies on the material properties of their components. Current semiconductor manufacturing and foundry processes are designed for silicon, but researchers must take advantage of compounds with better semiconductor properties to continue improving electronic performance.
“We must find new materials that have fundamentally better properties than silicon, as far as transistor functionality, but they need to be integrated into the current manufacturing processes,”
explained Zetian Mi, professor of Electrical and Computer Engineering. “It’s not likely for the industry to give up many billions of dollars of infrastructure. As such, the new material should ideally be CMOS compatible.”
Mi’s research group has focused on developing these new materials for many years, growing pristine single-crystalline ferroelectric III-V compounds, such as gallium nitride, in the laboratory at nanoscales. Mi knows that these materials must be scalable and compatible with silicon to transition into devices, and his newest grant support from the Defense Advanced Research Projects Agency (DARPA) will bring his team closer to this goal.
Mi’s project, called “CMOS compatible, defect-free universal growth of III-N and III-V multilayer heterostructures on Si (001),” was awarded $3M as part of DARPA’s Material Synthesis Technologies for Universal and Diverse Integration Opportunities (M-STUDIO) initiative. The goal of M-STUDIO is to “realize a universal heterogeneous integration technology, compatible with leading edge and future advanced-node semiconductor manufacturing processes, via atomic-precision nano-scale multi-layer material synthesis.”
To achieve this goal, Mi and his collaborators, Kai Sun (U-M Materials Science and Engineering) and Patrick Fay (University of Notre Dame), will use a new method to grow ultra-thin layers of III-V compound crystals without a foreign metal catalyst, which often introduces impurities. These tens-of-atoms thick layers will allow Mi’s team to grow the semiconductor materials on silicon lattice without defects. The final goal of this project will be to partner with the manufacturing company Intelligent Epitaxy Technology, Inc. (IntelliEPI) to show viability of this material at scale for use in industry.
“We are looking into the fundamental, long-standing challenge of integrating compound semiconductors with silicon,” said Mi. “It’s a very difficult problem, but if we are successful, there is the possibility that this material will go into every computer, every cell phone, virtually every electronic device that we use.”