LONDON – A combination of a radio frequency capable metallic MEMS cantilever beam and a carbon nanotube transistor shows promise as a non-volatile memory, according to researchers from the University of Edinburgh and Konkuk University and Seoul National University, Korea.
The research, published in Nature Communications, shows a device with power consumption performance superior to that of conventional flash memories, the authors claim.
Previous attempts to use carbon nanotube transistors for memory storage hit a stumbling block because they had low operational speed and short memory retention times. A team working under Professor Eleanor Campbell of the School of Chemistry at Edinburgh University has improved on that by using a mechanical arm to charge the floating gate electrode.
"This is a novel approach to designing memory storage devices. Using a mechanical method combined with the benefits of nanotechnology enables a system with superior energy efficiency compared with existing devices," said Professor Campbell, in a statement.
The cantilever is anchored at one end and suspended above the actuating electrode and the floating gate. The beam appears to measure about 1-micron across and be more than 10 microns long. When a bias voltage is applied to the actuating electrode, the resultant electrostatic force pulls down the cantilever until it contacts the floating gate. The cantilever is composed of Cr/Al/Cr triple layer and the floating gate is made of Au above an 80-nm aluminum-oxide insulating layer
The charge on the floating gate controls the source-drain current in the p-type CNT semiconducting channel. The on-off switching ratio was 10^4 to 10^5. Data retention is good to 4,000 seconds and cycling endurance of 500 cycles was demonstrated, but the known switching endurance of MEMS switches is of the order of 100 million cycles. This is greatly superior to flash memory.
The memory has the capability for multiple bit storage and the operational speed of the memory device is only limited by the speed of the cantiliver switch which is also much faster than flash memory, the authors claimed.
The one aspect of the design that is not discussed extensively is size, density and scalability. Although the CNT transistor could in theory scale well the cantilever MEMS may not.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.