On your rapidly diminishing list of things graphene cannot improve, go ahead and cross off "light bulbs." An international team of researchers drawn from Columbia University, Seoul National University, and the Korea Research Institute of Standards and Science have for the first time ever accomplished just that: a graphene light bulb.
Emitting visible light is a behavior that goes against the very nature of graphene as a highly-efficient conductor of thermal energy—and was discovered quite by accident—but it may have technological applications in the future as "the world's thinnest light bulb," in the words of James Hone, a Columbia engineering professor and co-author of a new study describing the behavior.
The principle is pretty much the same as that which applies to the filament in a conventional light bulb, in which fine coils of tungsten—tungsten having a very high melting point—soak up electrical current and heat until they begin to emit thermal radiation. Light. The same basic technology has been going strong for over a century now, excepting some late-breaking energy-efficient modifications. The same idea applies to graphene: apply enough current and, eventually, there will be light.
The catch with graphene, however, is that because it's such a good conductor of thermal (and electrical) energy, that energy isn't especially inclined to stick around. It takes a whole lot of current before a filament of any material actually gets hot enough to see, around 2500 degrees Celsius. The only thing that allows graphene to get this hot in the first place is the fact that its thermal conductivity starts to plummet as the material gets hotter and hotter. This is due to what's known as Umklapp scattering, which is a strange behavior characteristic of heat in crystalline structures that finds excitations within those structures suddenly flipping direction as more energy/heat is applied.
So, as the thermal conductivity of graphene starts to drop precipitously, it becomes possible to heat the material beyond those previous limits and to the point where it's manifested as visible light.
There's a bit more to it, however. Heating graphene becomes really inefficient given the usual silicon substrate, which has a tendency to dissipate heat away. "An extremely small fraction of the applied energy is converted into light radiation," the researchers explain.
The answer, according to the current paper, is to freely suspend the graphene in a way somewhat analogous to the coiled light bulb filaments we're used to. Suspending the graphene above a separate substrate allows not only a reasonably efficient heating of the material, but also a fine tuning of the relationship between the graphene-emitted light and the reflectivity of the substrate below. This means that the visible light wavelengths released by the graphene strips are tunable.
How practical is this? For one thing, it should allow for much smaller scales of visible light emitters in electronics. As conventional light bulb-style filaments scale further and further down, they hit a fundamental limit—with that much thermal energy and that small of a conduit, the filament will just burn up. Graphene, not so much. "This new type of 'broadband' light emitter can be integrated into chips and will pave the way towards the realization of atomically thin, flexible, and transparent displays, and graphene-based on-chip optical communications," Hone notes in a statement.
The catch is that suspended graphene is rather hard to produce. This isn't quite a ready-made technology. As Andrea Ferrari, the engineer who first developed the suspension method, cautions in Physics World, "The technique [to produce suspended graphene] is very laborious. So the question is: can this be easily mass-produced? At the moment, the answer is probably no, but in the future, one can find a way of making useful devices."