Obtaining graphene with adhesive tape; graphene/the rest ratio

[Timothy S.]

Why can't the Novosiolov and Geim's method of getting graphe (with an adhesive tape) be upscaled to manufacturing?
It is said "because this method is based on the handwork", but aren't we really able to do this, for example, with robotics as an iterative process?
I guess, there is a problem with how many graphene emerges on the tape after a process.
Can you, please, name, what is an approximate ratio of tape's graphene square to the square of the rest of the carbon on the tape?

 

[berkeman]

Can you please post some links to what you are talking about? Thanks.

 

[Timothy S.]

Can you please post some links to what you are talking about? Thanks.

"They used a simple but effective mechanical exfoliation
method for extracting thin layers of graphite from a graphite crystal with Scotch tape and
then transferred these layers to a silicon substrate." [ https://www.nobelprize.org/prizes/physics/2010/advanced-information/ ]

The original Novosiolov and Geim's paper: https://www.science.org/doi/10.1126...=ori:rid:crossref.org&rfr_dat=cr_pub 0pubmed

 

[Tom.G]

There are two articles in the 11 October 2024 issue of Science magazine
( VOL 386 ISSUE 6718, pgs. 138-143 and pgs. 144-145)
that cover the current State-of-the-Art.

The first is COMING OF AGE by Mark Peplow
Then Graphene, beyond lab benches by Yixuan Zhao and Li Lin
(the second article has DOI 10.1126/science.ads6781)

They at least partially explain that the Scotch tape method is not being used because the Graphene particles it yields are of microscopic size; fine for lab investigation but not big enough to be useful out in the 'real world.'

Cheers,
Tom

 

[Timothy S.]

There are two articles in the 11 October 2024 issue of Science magazine
( VOL 386 ISSUE 6718, pgs. 138-143 and pgs. 144-145)
that cover the current State-of-the-Art.

The first is COMING OF AGE by Mark Peplow
Then Graphene, beyond lab benches by Yixuan Zhao and Li Lin
(the second article has DOI 10.1126/science.ads6781)

They at least partially explain that the Scotch tape method is not being used because the Graphene particles it yields are of microscopic size; fine for lab investigation but not big enough to be useful out in the 'real world.'

Cheers,
Tom

Thank you!

 

[bigfooted]

A visual inspection of the device also shows that the capacity for upscaling is very limited.

 

[crashcat]

My info is 6 years out of date, but without checking the lit first:

I have spent probably a hundred hours preparing graphene samples using the scotch tape method (or we use this low-residue blue tape from Nitto Denki). The problem is that your source HOPG or natural graphite has an uneven surface and miscellaneous grain size. After the transfer you have to hunt around the substrate for the actual monolayer pieces amongst the thicker deposit (1-100 layers thick pieces all around).

A high quality source material and good technique saves you time but doesn't eliminate the need to hunt around with a microscope.

You can actually then pick up the monolayer and re-place it on a new substrate. That has been done for a decade now in order to make the stacks with, e.g. hBN, to electrically isolate it from the substrate. Theoretically you could automate it with stamping, imaging, pick up and place onto your device wafer, then do further processing to make your device, but it would be crazy cost. I thought about this as a method to manufacture those 100 GHz devices at scale, but it would be for a niche market and devices, not for consumer CPUs, because of the crystal size limitation.

When people say "graphene" they often mean completely different things - like the "graphene" golf clubs where it's a slurry of exfoliated carbon that provides an improvement over other carbon additives. This tape method is about getting a single crystal monolayer which is what you need for the cool electronic properties. So chemical exfoliation, which you can do at scale, is right out.

The best is method to scale might be the CVD growth on copper which can grow large (100 um) single crystals which then could theoretically be automatically transferred to devices using camera based selection and processing. Still not cheap.

The easiest thing to do at scale is the lower quality, but still monolayer, <1 um grain size CVD growth. That completely covers the growth medium and can be directly transferred. You can't make 100 GHz transistors out of it, but you can still use it for optical applications and chemical sensing applications. Or do CVD growth for thicker 10-30 layer thick pieces and use it as a flexible conductor.