MANHASSET, NY -- National Institute of Standards and Technology researchers have found in recent reliability tests that carbon nanotubes device reliability is a major issue.
NIST test results from numerous nanotube interconnects between metal electrodes show that nanotubes can sustain extremely high current densities -- tens to hundreds of times larger than that in a typical semiconductor circuit -- for several hours but slowly degrade under constant current.
And in about 40 hours the researchers found that the metal electrodes fail when currents rise above a certain threshold.
NIST is developing measurement and test techniques and studying a variety of nanotube structures, zeroing in on what happens at the intersections of nanotubes and metals and between different nanotubes.
"The common link is that we really need to study the interfaces," said Mark Strus, a NIST postdoctoral researcher, in a statement.
In a related study NIST researchers identified failures in carbon nanotube networks, where electrons physically hop from tube to tube. Failures in this case seemed to occur between nanotubes, the point of highest resistance, Strus said.
By monitoring the starting resistance and initial stages of material degradation, researchers could predict whether resistance would degrade gradually—allowing operational limits to be set—or in a sporadic, unpredictable way that would undermine device performance.
NIST developed electrical stress tests that link initial resistance to degradation rate, predictability of failure and total device lifetime. The test can be used to screen for proper fabrication and reliability of nanotube networks.
Strus speculates that carbon nanotube networks may ultimately be very useful for some electronic applications such as interconnects for flexible electronic displays or photovoltaics.
To @GroovyGeek: I am not sure why you would blame EE TImes for this? Clearly this is a controversial issue that requires further science, experiments or measurements. Any new material, like carbon nanotube, will take years of development (I work on a compound material that was discovered 50 years ago, we are still perfecting it). EE Times is not Bell Labs ;-)...Kris
And a bad data point at that. I attended the talk and it was devoid of any serious analysis. The problem was mostly with electromigration of their Au electrodes (duh!). When they did observe "something" with the CNTs they made no effort to do a physical analysis to determine what is going on. It is really disappointing to see the EE Times not do their homework and highlight such low quality research.
Agree, this is hardly news since the interface is always a challenge whenever you try to bridge two materials together. Adding a graphite or graphene layer between the CNT and the metal may help to spread out the current.
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.