Earlier this year I was writing a review about transparent conducting materials for organic electronic applications. As part of it, I wrote a fair bit about graphene. One of the key problems is that it’s difficult to scale the desirable properties of small pieces of graphene to large areas without resorting to quite challenging techniques.
In 2004, when Geim and Novoselov managed to isolate single-layer and few-layer graphene, they did it using Scotch tape to “mechanically exfoliate” graphite—basically using Scotch tape to peel off a layer of graphene from the graphite.1
This technique gives very high quality crystals of up to 10 micrometres in size, but it’s limited to laboratory scale production—you’re not going to use this in a factory to make components for electronic devices. When I was writing the review, I wanted to write something along the lines of “…mechanical exfoliation, whilst producing very high quality graphene, is impossible to scale up…” but my supervisor told me to change it to something less definite, hinting these things have a habit of a coming back to prove you wrong.
To my surprise, someone has managed to scale it up. Chen and co-workers recently published a paper describing mechanical exfoliation of graphite using a three–roll mill.2 Here’s a pen and paper sketch of their set up.
First they prepared an adhesive (polyvinylchloride (PVC) in dioctyl phthalate) which was poured between the feed and centre rolls. They then started rolls rotating then spread the graphite powder onto the adhesive. The adhesive runs in an S shape around the rolls. The graphene is continuously exfoliated to give graphene, unlike the scotch tape method where you exfoliate once with each application and removal of the scotch tape. After 12 hours of operation, you collect the material, wash it and then burn off the PVC to get the graphene. There’s a video (direct link) in the supplementary information showing the mill in operation.
I’m not entirely sure why anyone would want that much graphene in this form. For transparent conducting applications (where you want to maximise transparency and minimise electrical resistance) it’s likely you would still have high electrical resistances between flakes of graphene and poor performance. But it’s still a clever way of scaling up what, to me, looked like a completely unscalable process. I won’t be so certain when writing in the future.