Nanoscientists suggest use of vacuums to overcome limits of conventional silicon-based semiconductor electronics

July 1, 2012

With the advent of semiconductor transistors—invented in 1947 as a replacement for bulky and inefficient vacuum tubes—has come the consistent demand for faster, more energy-efficient technologies. To fill this need, researchers at the University of Pittsburgh are proposing a new spin on an old method: a switch from the use of silicon electronics back to vacuums as a medium for electron transport—exhibiting a significant paradigm shift in electronics. Their findings were published online in Nature Nanotechnology July 1.

For the past 40 years, the number of transistors placed on integrated circuit boards in devices like computers and smartphones has doubled every two years, producing faster and more efficient machines. This doubling effect, commonly known as "Moore's Law," occurred by scientists' ability to continually shrink the transistor size, thus producing computer chips with all-around better performance. However, as transistor sizes have approached lower nanometer scales, it's become increasingly difficult and expensive to extend Moore's Law further.

"Physical barriers are blocking scientists from achieving more efficient electronics," said Hong Koo Kim, principal investigator on the project and Bell of Pennsylvania/Bell Atlantic Professor in the University of Pittsburgh's Swanson School of Engineering. "We worked toward solving that road block by investigating transistors and its predecessor—the ."

The ultimate limit of transistor speed, says Kim, is determined by the "electron transit time," or the time it takes an electron to travel from one device to the other. traveling inside a semiconductor device frequently experience collisions or scattering in the solid-state medium. Kim likens this to driving a vehicle on a bumpy road—cars cannot speed up very much. Likewise, the electron energy needed to produce faster electronics is hindered.

"The best way to avoid this scattering—or traffic jam—would be to use no medium at all, like vacuum or the air in a nanometer scale space," said Kim. "Think of it as an airplane in the sky creating an unobstructed journey to its destination."

However, says Kim, conventional vacuum electronic devices require high voltage, and they aren't compatible with many applications. Therefore, his team decided to redesign the structure of the vacuum electronic device altogether. With the assistance of Siwapon Srisonphan, a Pitt PhD candidate, and Yun Suk Jung, a Pitt postdoctoral fellow in electrical and computer engineering, Kim and his team discovered that electrons trapped inside a semiconductor at the interface with an oxide or metal layer can be easily extracted out into the air. The electrons harbored at the interface form a sheet of charges, called two-dimensional electron gas. Kim found that the Coulombic repulsion—the interaction between electrically charged particles—in the electron layer enables the easy emission of electrons out of silicon. The team extracted electrons from the silicon structure efficiently by applying a negligible amount of voltage and then placed them in the air, allowing them to travel ballistically in a nanometer-scale channel without any collisions or scattering.

"The emission of this electron system into vacuum channels could enable a new class of low-power, high-speed transistors, and it's also compatible with current silicon electronics, complementing those electronics by adding new functions that are faster and more energy efficient due to the low voltage," said Kim.

With this finding, he says, there is the potential for the vacuum transistor concept to come back, but in a fundamentally different and improved way.

Explore further: Return of the vacuum tube

Related Stories

Return of the vacuum tube

May 18, 2012

Vacuum tubes have been retro for decades. They almost completely disappeared from the electronics scene when consumers exchanged their old cathode ray tube monitors for flat screen TVs. Their replacement – the semiconductor ...

Can graphene nanoribbons replace silicon?

February 18, 2010

( -- "Graphene has been the subject of intense focus and research for a few years now," Philip Kim tells "There are researchers that feel that it is possible that graphene could replace silicon as ...

Recommended for you

Four elements make 2-D optical platform

September 21, 2017

Rice University scientists have discovered a two-dimensional alloy with an optical bandgap that can be tuned by the temperature used to grow it.


Adjust slider to filter visible comments by rank

Display comments: newest first

Jul 01, 2012
This comment has been removed by a moderator.
not rated yet Jul 02, 2012
Nice work. It sounds like there is an unmentioned drawback. Size? Cost? Something. Honestly though. This would be useful for handling a mobile devices functions while in an idle state. Quick and low power to draw out battery life. Then you wake up the power hogs only when needed. (Once again, size permitting)
5 / 5 (2) Jul 02, 2012
..the article is a bit unclear if a real vacuum is needed, or else just air/gas space...
It could actually work even at the presence of air at the moment, when the size of elements would drop bellow length of mean free path of air molecules, i.e. about 65 nm.
5 / 5 (1) Jul 02, 2012
I'm a little confused. Are they addressing switches (transistors), or connections between the switches? If the latter, then there is still the switch's speed ("ultimate limit of transistor speed ... time it takes an electron to travel from one device to the other").

Do longer interconnects require higher voltages?

The lack of a link to the article is annoying. The press release lacks substance.
5 / 5 (1) Jul 02, 2012
I would like to see how this stacks up against photonics.

It still seems to me that 'extracting' photons (e.g. at quantum dots) - and letting them do the transmission - should be orders of magnitude faster than usiong electrons.
not rated yet Jul 02, 2012
A method of using planar ceramic substrates (w/vacuum)was
developed and reported in the trade publication "Electronics"
some time ~(20 years) ago. I.E. Micro Tube arrays.

Does anyone recall reading that related article or link(s)to
that original work ?

I remember reading it, but haven't been able to reference it
since that publication


3.7 / 5 (3) Jul 02, 2012
I've never heard of a nanoscientist. How big are they, and where do they come from?
not rated yet Jul 02, 2012
This seems very similar to the method the SED display used to emit electrons.
5 / 5 (1) Jul 03, 2012
Gah, another "Moore's Law is almost dead" pronouncement.

It isn't. There are many, many avenues for continuing Moore's Law's progress. Chips are two-dimensional; they can be expanded to three. Advances in cooling physics have been experimentally developed, and advances in understanding quantum materials have been and are being made. Then there are memristors, which can both calculate and store information, and optical circuits, and lots more.

Just because Moore's Law surfed on two-dimensional photo-etched silicon transistors for four decades does not mean they're the only game in town.

The irony is that the cited research is yet another path to possibly exploit for better circuits... yet the article begins by lamenting how Moore's Law is nearly defunct. Cognitivie dissonance, anyone?
1.5 / 5 (4) Jul 04, 2012
WoW verkle is back
Great research. The article is a bit unclear if a real vacuum is needed, or else just air/gas space. I took it to mean the latter.
Very surprised you haven't quoted a bible reference as I hope you realise greater understanding of computing allows us to manage much better DNA mutations causing disease, which goes against of many sorts of deities, thus we can alleviate the bulk of human suffering caused by Eve - the poor 12yr old child...

So I'm keen to see your point if your particular uncommunicative deity has a thought on this, will he bring back Noah perhaps ?


not rated yet Jul 24, 2012
Just because Moore's Law surfed on two-dimensional photo-etched silicon transistors for four decades does not mean they're the only game in town.

The Moore's Law isn't even defined in any other terms than the number of transistors in an affordable silicon microchip. All the other technologies are irrelevant in that respect, and transitioning to something like optical chips would in and of itself mean that the Moore's Law is dead because the silicon transistor couldn't keep up with the curve.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.