Solving nearly century-old problem: Using graphene, professor finds out what causes low-frequency electronic 1/f noise

Mar 07, 2013 by Sean Nealon
Alexander Balandin, a professor electrical engineering at UC Riverside

(Phys.org) —A University of California, Riverside Bourns College of Engineering professor and a team of researchers published a paper today that show how they solved an almost century-old problem that could further help downscale the size of electronic devices.

The work, led by Alexander A. Balandin, a professor of electrical engineering at UC Riverside, focused on the low-frequency electronic 1/f , also known as pink noise and flicker noise. It is a signal or process with a power spectral density inversely proportional to the frequency. It was first discovered in in 1925 and since then it has been found everywhere from fluctuations of the intensity in music recordings to human heart rates and electrical currents in materials and devices.

The importance of this noise for electronics motivated numerous studies of its physical origin and methods for its control. For example, the signal's phase noise in a radar or communication gadget such as smart phone is determined, to a large degree, by the 1/f noise level in the transistors used inside the radar or smart phone.

However, after almost a century of investigations, the origin of 1/f noise in most of material systems remained a mystery. A question of particular importance for electronics was whether 1/f noise was generated on the surface of or inside their volumes.

A team of researchers from the UC Riverside, Rensselaer Polytechnic Institute (RPI) and Ioffe Physical-Technical Institute of The were able to shed light on 1/f noise origin using a set of multi-layered graphene samples with the thickness continuously varied from around 15 atomic planes to a single layer of graphene. Graphene is a single-atom thick carbon crystal with unique properties, including superior electrical and heat conductivity, and unique .

In addition to Balandin, who is also the founding chair of the program at UC Riverside, the team of researchers included: The team included: Guanxiong Liu, a research associate in Balandin's Nano-Device Laboratory (NDL); Michael S. Shur, Patricia W. and C. Sheldon Roberts Professor of Solid State Electronics at RPI; and Sergey Rumyantsev, research professor at RPI and Ioffe Institute.

"The key to this interesting result was that unlike in metal or semiconductor films, the thickness of graphene multilayers can be continuously and uniformly varied all the way down to a single atomic layer of graphene – the ultimate "surface" of the film," Balandin said. "Thus, we were able to accomplish with multilayer graphene films something that researchers could not do with metal films in the last century. We probed the origin of 1/f noise directly."

He added that previous studies could not test metal films to the thicknesses below about eight nanometers. The thickness of graphene is 0.35 nanometers and can be increased gradually, one atomic plane at a time.

"Apart from the fundamental science, the reported results are important for continuing the downscaling of conventional electronic devices," Balandin said. "Current technology is already at the level when many devices become essentially the surfaces. In this sense, the finding goes beyond graphene field."

He also noted that the study was essential for the proposed applications of graphene in analog circuits, communications and sensors. This is because all these applications require acceptably low levels of 1/f noise, which contributes to the of communication systems and limits sensor sensitivity and selectivity.

The results of the research have been published in the journal Applied Physics Letters.

Explore further: All directions are not created equal for nanoscale heat sources

More information: The paper, "Origin of 1/f Noise in Graphene Multilayers: Surface vs. Volume" is available at: apl.aip.org/resource/1/applab/v102/i9/p093111_s1

Related Stories

Turning down the noise in graphene

Aug 06, 2010

(PhysOrg.com) -- Graphene is a two-dimensional crystalline sheet of carbon atoms - meaning it is only one atom thick - through which electrons can race at nearly the speed of light - 100 times faster than ...

Recommended for you

New absorber will lead to better biosensors

5 hours ago

Biological sensors, or biosensors, are like technological canaries in the coalmine. By converting a biological response into an optical or electrical signal, they can alert us to dangers in our external and internal environments. ...

Ultrafast remote switching of light emission

Sep 30, 2014

Researchers from Eindhoven University of Technology can now for the first time remotely control a miniature light source at timescales of 200 trillionth of a second. They published the results on Sept. 2014 ...

Nanotube cathode beats large, pricey laser

Sep 30, 2014

Scientists are a step closer to building an intense electron beam source without a laser. Using the High-Brightness Electron Source Lab at DOE's Fermi National Accelerator Laboratory, a team led by scientist ...

User comments : 10

Adjust slider to filter visible comments by rank

Display comments: newest first

ian_j_allen
5 / 5 (14) Mar 07, 2013
Phys.org, can you PLEASE include the actual results in your articles? I dont know how many times I am reading an item here, and it just happens to skip right over the single most important point of the entire discovery.

In this case you have a paragraph outlining the main question, then an out of place paragraph introducing the rest of the research team (should have been in the beginning or put at the end), and then a paragraph discussing why graphene was key in producing the result.

THERE IS NO PARAGRAPH DESCRIBING THE ACTUAL RESULT!

At least you hotlinked the paper, this time. Thank you for that.
ian_j_allen
5 / 5 (6) Mar 07, 2013
1/f noise is dominated by volume noise when graphene is more than 7 layers thick (2.5 nm). 1/f noise, therefore, is a surface phenomenon that shows up in situations that are thinner than 2.5 nm (at least for graphene).
Lurker2358
2.3 / 5 (3) Mar 07, 2013
1/f noise is dominated by volume noise when graphene is more than 7 layers thick (2.5 nm). 1/f noise, therefore, is a surface phenomenon that shows up in situations that are thinner than 2.5 nm (at least for graphene).


Isn't that a bad thing for computers, especially any potential 3D architecture, since computer circuits will ultimately have a very high surface area, and a high surface to volume ratio?!
Ophelia
not rated yet Mar 07, 2013
fyi, it appears an early version of the paper was published on the arxiv entitled: "Direct Probing of 1/f Noise Origin with Graphene Multilayers: Surface vs. Volume", arXiv:1211.5155, at http://arxiv.org/...11.5155.
Tausch
1 / 5 (3) Mar 07, 2013
"When the thickness reduces down to one atomic layer, the total noise consists of the surface noise only."

http://arxiv.org/...5155.pdf

@Lurker
In superconducting substrates the question vanishes.
Tausch
1 / 5 (3) Mar 07, 2013
Maybe tamed noise is superconducting.
swordsman
1 / 5 (3) Mar 08, 2013
I think I agree with Ian. What is it the fundamental physics that produces 1/f noise? So it is layer-dependent? what does that tell us? Since it varies inversely with frequency, it most likely is an electromagnetic inductive effect, which also varies with path length.
Tausch
1 / 5 (3) Mar 08, 2013
Lan quoted the research paper. You are agreeing with the researchers.
KBK
1 / 5 (3) Mar 10, 2013
Maybe tamed noise is superconducting.


NOW YOU'RE GETTING IT. (everyone in general)

FINALLY.
william_fairholm
5 / 5 (1) Mar 11, 2013
This article was not about what causes 1/f noise. It was about what is the primary source of 1/f noise, the surface or the bulk of the material.

Not surprisingly it has found to be both, with the surface being the predominant source for very thin layers (the only source for a one atom thick layer)and the bulk taking over for thicker layers. And they modeled this combined behaviour. Interesting, but not earth shattering.

In the pdf linked by @Ophelia they made an algebraic error in their equation for the bulk resistance. When rearranging the equation for parallel circuits, they reversed the order of two of the resistances and so would produce a negative resistance for the bulk.