Basic Research is Now Applied
04/01/2009
 |
Pete Singer
Editor-in-Chief
The American Recovery and Reinvestment Act (HR 1), which was signed into law on February 17th by President Obama, includes more than $15 billion for vital scientific research programs. It includes significant increases in funding for the National Science Foundation (NSF), the National Institute for Standards and Technology (NIST), the Department of Energy (DoE) Office of Science, and the Advanced Research Project Agency-Energy. This is great news for the semiconductor industry. “The inclusion of funds for renewable energy, energy efficiency, broadband deployment, and accelerated adoption of information technology by doctors and hospitals will enable many new services to be readily available at lower cost,” noted George Scalise, president of the Semiconductor Industry Association (SIA).
Those $15 billion will hopefully be used in a responsible way. There is obviously a need to immediately stimulate the economy, but there’s also a need to continue with basic research that will make continued advances possible. As we all know, the limits of CMOS scaling are upon us and a “switch” other than the transistor will eventually be needed.
In the past, I have wondered if the government was funding too much basic research, or what many consider “blue sky” research — research done for theoretical reasons without immediate commercial value. Way back in 2000 (is it already 2009?) I said that it’s debatable whether the semiconductor industry’s long-term success will depend that much on what researchers are learning by “fiddling around with quantum dots and the like.”
With the perspective of time, I now realize that it was rather silly to think that there was inherently more value in short-term research. The very research that I found too blue sky in 2000 is now entering the penultimate stage of development, close to entering production/commercialization.
A great example is some recent work from Rensselaer Polytechnic Institute (RPI) where researchers have developed a method for producing slim and sticky nanorods that appear to be a perfect solution for “gluing” together stacked components in 3D integration.
“When fabricating and assembling 3D chips, and when bonding the silicon wafers together, you want as low a temperature as possible,” said Pei-I Wang, research associate at Rensselaer’s Center for Integrated Electronics. “Slimmer nanorods, by virtue of their smaller diameters, require less heat to anneal. These lower temperatures won’t damage or degrade the delicate semiconductors. The end result is a less expensive, more reliable device.”
The slimmer copper nanorods were formed by periodically interrupting the growth process. The vapor-deposition process was occasionally halted, and the fledgling nanorods were exposed to oxygen. This resulted in a forest of nanorods with diameters between 10-50nm — far smaller than the typical 100nm diameter copper nanorods grown conventionally without interruption.
Vast forests, or arrays, of copper nanorods are produced by vapor deposition at an oblique angle. In a conventional setting, with an uninterrupted stream of copper atoms deposited in a vacuum onto a substrate, the deposition angle naturally results in taller, thicker nanorods. Periodically interrupting the deposition, and exposing the copper nanorods to ambient air, however, leads to oxygen being absorbed into the surface of the nanorods. During subsequent depositions, this oxidized copper helps to prevent the vaporized copper atoms from migrating away from the very tips of the nanorods. This ensures the nanorods grow taller, without necessarily increasing in diameter. The more growth interruptions, the thinner the resulting nanorods, Wang said.
I would imagine that back in 2000, such research would have definitely been considered “blue sky” (although RPI even then had a strong focus on real-world applications). At the time, we needed a better understanding of the physics of atomic-scale materials and how to synthesize novel materials. That’s still true today. The International Technology Roadmap for Semiconductors (ITRS) still contains a “sea of red” indicating technical challenges for which there are no known solutions.
The $15 billion provided in the recovery and reinvestment act will go a long way in helping researchers find those solutions.