New study gives insight into graphene grain boundaries

Jan 15, 2013 by Steve Mcgaughey
This image shows a graphene grain boundary. Credit: Courtesy Justin Koepke, Joe Lyding

(Phys.org)—Using graphene – either as an alternative to, or most likely as a complementary material with – silicon, offers the promise of much faster future electronics, along with several other advantages over the commonly used semiconductor. However, creating the one-atom thick sheets of carbon known as graphene in a way that could be easily integrated into mass production methods has proven difficult.

When graphene is grown, lattices of the carbon grains are formed randomly, linked together at different angles of orientation in a hexagonal network. However, when those orientations become misaligned during the growth process, defects called (GBs) form. These boundaries scatter the flow of electrons in graphene, a fact that is detrimental to its successful electronic performance.

Researchers Joe Lyding and Eric Pop from the University of Illinois' Beckman Institute and their research groups have now given new insight into the electronics behavior of graphene with grain boundaries that could guide toward lessening their effect. The researchers grew polycrystalline graphene by (CVD), using and spectroscopy for analysis, to examine at the grain boundaries on a silicon wafer. They reported their results in the journal ACS Nano.

"We obtained information about electron scattering at the boundaries that shows it significantly limits the electronic performance compared to grain boundary free graphene," Lyding said. "Grain boundaries form during graphene growth by CVD, and, while there is much worldwide effort to minimize the occurrence of grain boundaries, they are a fact of life for now.

"For electronics you would want to be able to make it on a wafer scale. Boundary free graphene is a key goal. In the interim we have to live with the grain boundaries, so understanding them is what we're trying to do."

Lyding compared graphene lattices made with the CVD method to pieces of a cyclone fence.

"If you had two pieces of fence, and you laid them on the ground next to each other but they weren't perfectly aligned, then they wouldn't match," he said. "That's a grain boundary, where the lattice doesn't match."

The research involved Pop's group, led by Beckman Fellow Josh Wood, growing the graphene at the Micro and Nanotechnology Lab, and transferring the thin films to a silicon (Si02) wafer. They then used the STM at Beckman developed by Lyding for analysis, led by first author Justin Koepke of Lyding's group.

Their analysis showed that when the electrons' itinerary takes them to a grain boundary, it is like, Lyding said, hitting a hill.

"The electrons hit this hill, they bounce off, they interfere with themselves and you actually see a standing wave pattern," he said. "It's a barrier so they have to go up and over that hill. Like anything else, that is going to slow them down. That's what Justin was able to measure with these spectroscopy measurements.

"Basically a grain boundary is a resistor in series with a conductor. That's always bad. It means it's going to take longer for an electron to get from point A to point B with some voltage applied."

Images from the STM reveal grain boundaries that suggest two pieces of cloth sewn together, Lyding said, by "a really bad tailor."

In the paper, the researchers were able to report on their analysis of the orientation angles between pieces of graphene as they grew together, and found "no preferential orientation angle between grains, and the GBs are continuous across graphene wrinkles and Si02 topography." They reported that analysis of those patterns "indicates that backscattering and intervalley scattering are the dominant mechanisms responsible for the mobility reduction in the presence of GBs in CVD-grown graphene."

Lyding said that the relationship between the orientation angle of the pieces of graphene and the wavelength of an electron impinges on the electron's movement at the grain boundary, leading to variations in their scattering.

"More scattering means that it is making it more difficult for an electron to move from one grain to the next," he said. "The more difficult you make that, the lower the quality of the electronic performance of any device made from that graphene."

The researchers work is aimed not just at understanding, but also at controlling grain boundaries. One of their findings – that GBs are aperiodic – replicated other work and could have implications for controlling them, as they wrote in the paper: "Combining the spectroscopic and scattering results suggest that GBs that are more periodic and well-ordered lead to reduced scattering from the GBs."

"I think if you have to live with grain boundaries you would like to be able to control exactly what their orientation is and choose an angle that minimizes the scattering," Lyding said.

Explore further: Engineered proteins stick like glue—even in water

More information: pubs.acs.org/doi/full/10.1021/nn302064p

Related Stories

Graphene's strength lies in its defects

Nov 11, 2010

The website of the Nobel Prize shows a cat resting in a graphene hammock. Although fictitious, the image captures the excitement around graphene, which, at one atom thick, is the among the thinnest and strongest ...

Graphene grains make atom-thick patchwork 'quilts'

Jan 05, 2011

(PhysOrg.com) -- A quick look at new Cornell research hints at colorful patchwork quilts, but they are actually pictures of graphene -- one atom-thick sheets of carbon stitched together at tilted interfaces. ...

Hydrogen may be key to growth of high-quality graphene

Jul 18, 2011

A new approach to growing graphene greatly reduces problems that have plagued researchers in the past and clears a path to the crystalline form of graphite's use in sophisticated electronic devices of tomorrow.

STM of individual grains in CVD-grown graphene

Jun 24, 2011

Users from Purdue University, working collaboratively with staff in the CNM Electronic & Magnetic Materials & Devices Group, studied CVD-grown graphene on polycrystalline copper foil for the first time at ...

Recommended for you

Engineered proteins stick like glue—even in water

5 hours ago

Shellfish such as mussels and barnacles secrete very sticky proteins that help them cling to rocks or ship hulls, even underwater. Inspired by these natural adhesives, a team of MIT engineers has designed new materials that ...

Smallest possible diamonds form ultra-thin nanothreads

5 hours ago

For the first time, scientists have discovered how to produce ultra-thin "diamond nanothreads" that promise extraordinary properties, including strength and stiffness greater than that of today's strongest ...

User comments : 9

Adjust slider to filter visible comments by rank

Display comments: newest first

that_guy
not rated yet Jan 15, 2013
Interesting.

Is it possible to dope the graphene in a way to mend the boundaries by a better tailor?
antialias_physorg
1 / 5 (1) Jan 16, 2013
Basically a grain boundary is a resistor in series with a conductor

Don't they mean: "a resistor in series with a capacitor" ?

Is it possible to dope the graphene in a way to mend the boundaries

Why would you think doping would have any effect?
that_guy
not rated yet Jan 16, 2013
Is it possible to dope the graphene in a way to mend the boundaries

Why would you think doping would have any effect?

Because doping has an affect on growth patterns, grain boundaries, and properties of other materials...

While there may be a better solution or method, it is still perfectly possible that a dopant may help reduce the resistance of the boundary without destroying the positive properties of the material.

Course, I don't know how hard it is to dope this type of carbon.
antialias_physorg
1 / 5 (1) Jan 16, 2013
Doping is a stochastic process. To get an effect that would get you a straight edge you need a regular process. Doping isn't going to solve this.
that_guy
not rated yet Jan 16, 2013
Doping is a stochastic process. To get an effect that would get you a straight edge you need a regular process. Doping isn't going to solve this.

"Straight edge" is not what I meant, but I see how you pulled my metaphor too far. My comment was pretty general, about giving the grain boundaries better chemical or physical properties to help bridge the gap. Having a straight edge boundary may or may not fix the issue anyways.

I am suggesting that they dope the graphene with similar goals as they are trying to do with this ceramic to reduce grain boundary impedance, which happens to be a nearly identical issue:
http://onlinelibr...abstract

Clearly, empirically, doping can have an effect on grain boundary conductivity.

There are multiple ways and properties that doping can effect. No need to limit the possibilities to one.

A better growth or deposition process could be other routes to attempt to solve this issue.
antialias_physorg
not rated yet Jan 17, 2013
I am suggesting that they dope the graphene with similar goals as they are trying to do with this ceramic to reduce grain boundary impedance, which happens to be a nearly identical issue:

As I said: Doping is a stochastic process. What you linked to does not 'mend' the boundary and does not lead to better growth. It simply changes the resistance characteristic ON AVERAGE.

For ultrathin/ultrasmall electronics you need something that can be PATTERNED in a very deterministic way (via (ion) etching). For this you need a REGULAR structure. For those kinds of applications we're in the range of a few atoms where 'averaging' effects via doping aren't enough to mkae sure that LOCAL, GEOMETRIC defects don't spoil the element characteristics.
that_guy
not rated yet Jan 17, 2013
As I said: Doping is a stochastic process. What you linked to does not 'mend' the boundary and does not lead to better growth. It simply changes the resistance characteristic ON AVERAGE.


1. In silicon doping, for example - the dopants often integrate with silicon in a regular, PERIODIC way during the growth process, and can alter the growth process itself.
Example: http://www.jim.or.../237.pdf
2. Doping can and does affect internal and boundary chemical/electonic properties differently - Such as the example provided in the previous comment.

that_guy
not rated yet Jan 17, 2013
As I said: Doping is a stochastic process. What you linked to does not 'mend'

Doping during the production of the material can affect the growth process to help it grow without the same defects. I was extending the metaphor, but I didn't mean that it had to be after the fact.

For ultrathin/ultrasmall electronics you need something that can be PATTERNED in a very deterministic way (via (ion) etching). For this you need a REGULAR structure. For those kinds of applications we're in the range of a few atoms where 'averaging' effects via doping aren't enough to mkae sure that LOCAL, GEOMETRIC defects don't spoil the element characteristics.

Assuming that the defects remain the same despite the doping, I would probably agree that there would be a limit to the feature size - Which would be an identical scenario to polysilicon electronics.

http://en.wikiped..._silicon
that_guy
not rated yet Jan 17, 2013
I'm failing to see any way that ion etching would be constructive to fix this issue. Please explain your line of thinking. Are you saying that you would use ion etching to help smooth the edges? As graphene occurs in single layer sheets, anything that takes material away (as in ion etching) can have a drastic affect on the properties of the material. And that affect would not bring those edges closer together...

Did you mean ion [vapor?] deposition? (...a way to dope the material)