Graphene has revealed itself from a direction that seems opposite to what one might have expected. First came the zero-dimensional form: C60 and the other fullerenes, nanoscopically finite in every direction. Then came the carbon nanotube, whose one-dimensional form set everyone thinking in terms of fibres and wires. It was just two years ago that the 2D form, graphene itself, appeared: flat sheets of carbon one atom thick K. S. Novoselov et al. Nature. 306, 666–669; 2004), which, when stacked in the third dimension, returns us to familiar, lustrous graphite.

It's tempting to wonder if the earlier focus on reduced dimensionality and curvature may have been misplaced. C60 is a fascinating molecule, but useful materials tend to be extended in at least one dimension. Carbon nanotubes can be matted into 'bucky paper', but without exceptional strength. Long, thin, single-molecule transistors are fine, but microelectronics is inherently 2D. Graphene is the master substance of all these, and perhaps, for materials and electronics, sheets were what we needed all along.

You can cut these sheets into device-styled patterns — but that's best done with chemistry (etching with an oxygen plasma, say), as attempts to tear single-layer graphene with a diamond tip are apt to make the tip blunt. (As carbon nanotubes have shown, graphite has a false reputation for weakness.) And graphene is a semi-metal with a tunable charge-carrier density that makes it suitable for the conducting channel of transistors.

But its conductivity is more extraordinary than that. For one thing, the electron transport is ballistic, free from scattering. That recommends graphene for ultrahigh-frequency electronics, as scattering processes limit the switching speeds. More remarkably, the mobile electrons behave as Dirac fermions K. S. Novoselov et al. Nature. 438, 197–200; 2005) which mimic the characteristics of electrons travelling close to the speed of light.

From the perspective of applications, however, one key question is how to make the stuff. Peeling away flakes of graphite with Scotch tape, or just rubbing a piece of graphite on a surface (popularly known as drawing) will produce single-layer films — but neither reliably nor abundantly. Walt de Heer and co-workers have recently flagged up the value of a method several years old, by which silicon carbide heated in a vacuum will decompose to form graphitic films one layer at a time (C. Berger et al. Science Express doi:10.1126/science.1125925; 2006).

But maybe wet chemistry will be better still. Graphite was separated into layers nearly 150 years ago by oxidation, producing platelets of water-soluble oxidized graphene, possibly including single sheets. But reducing them triggers aggregation via hydrophobic interactions. This can be prevented by the use of amphiphilic polymers (S. Stankovich et al. J. Mater. Chem. 16, 155–158; 2006). Anchoring bare, single graphene sheets to a surface remains a challenge — but one that may benefit from the wealth of experience of organic chemists.