The Ultimate Silicon Integrated Nanodevice: Toshiro Hiramoto

The Ultimate Transistor Structure: Silicon Nanowire Transistors

The device architecture of VLSI has been changed from bulk planar to FinFET. It is anticipated that the device structure will be further changed to the gate-all-around nanowire transistor. We initiated the study on silicon nanowire transistors in late 1990s and reported the increase in threshold voltage due to the quantum confinement effects and higher mobility than the universal mobility for the first time. Recently, the low temperature characteristics have been studied for the application for quantum computing. We can fabricate silicon nanowire transistors with approximately 5nm nanowire width in our clean room.

Silicon single-electron transistors and quantum bits

We have studied the silicon single electron transistor, which consists of a silicon quantum dot, for a long time. A single electron transistor acts as a very sensitive charge sensor. On the other hand, silicon quantum dot is promising for a quantum bit (qubit) which is key to the quantum computing. We have proposed the integration of silicon qubits using stacked silicon layers.

New Developments in Silicon Power Devices

Although power transistors using wide bandgap such as SiC and GaN have been studied, silicon power transistors are still the mainstream technology for power electronics. Especially, IGBT (Integrated Gate Bipolar Transistors) are widely utilized in various applications of rate voltage from 650V to 6500V. Based on the new concepts of “IGBT Scaling” and “Double-gate IGBT”, we have demonstrated the performance improvements of silicon IGBT by fabrication and measurements. In the IGBT scaling, the gate drive voltage is scaled from 15V to 5V. Then, digital gate drivers can be used, leading to intelligent power electronics. We can fabricate 3300V class silicon IGBTs in our clean room.

The Extended CMOS concept

New devices operating based on new physics or new phenomena are generally called “Beyond CMOS”, which is expected for future high speed and low power devices with higher functionalities. Based on the idea that it is impossible for Beyond CMOS to replace conventional CMOS, we proposed a concept of “Extended CMOS” in 2000s, where Beyond CMOS should be merged to conventional CMOS in order to complement the weakness of CMOS. This concept has been widely accepted and the figure below has been published in the famous semiconductor roadmap (ITRS, International Technology Roadmap for Semiconductor). The research in our lab is conducted based on this concept.