Technical Insight
Graphene: opening the bandgap
A host of lesser-known compound semiconductors are being used to exploit graphene's electronic properties. Compound Semiconductor reports.
Graphene/Boron Nitride heterostructures are just one materials combination University of Manchester researchers are working on. [Manchester Graphene]
Some ten years ago, UK-based University of Manchester researchers, Kostya Novoselov and Andre Geim, published their seminal Science paper on graphene. Using scotch tape, the physicists peeled layers of graphene from a graphite block, dissolved the tape to leave ultra-thin flakes of graphene that they used to fabricate rudimentary silicon oxide-based transistors.
The rest is history. Just one atom thick, graphene forms a high quality crystal lattice with no vacancies and no dislocations, yielding remarkable properties.
Studies by Geim and Novoselov - since awarded the Nobel Prize for Physics - revealed the material's conducting electrons arrange into quasi-particles that behave more like neutrinos, moving close to the speed of light. With their first transistors displaying ballistic transport, linear current-voltage characteristics and huge sustainable currents, the researchers could only conclude: "graphene may be the best possible metal for metallic transistor applications".
However, a clear obstacle to the use of graphene as an alternative to silicon electronics is its lack of bandgap; with electrons flowing at any energy, low power dissipation in the OFF state is very difficult. Given this, researchers worldwide have been exploring ways to create an artificial bandgap - applying an electric field, modifying graphene's structure or doping with impurities - but device characteristics have fallen short of the wonder material's intrinsic properties.
On a different track, Novoselov and a handful of other researchers around the world, have been fabricating heterostructures based on single layers of graphene and novel compound semiconductors. These structures comprise layers of alternating 2D crystals - graphene and the insulating material - stacked up to form a device.
As Novoselov says: "Research on heterostructures is gaining momentum, and the possibilities for controlling properties of heterostructures might become very useful for future applications."
Novoselov and his team at the University of Manchester have largely focused on using 2D crystals of hexagonal boron nitride to fabricate graphene heterostructures. Just over two years ago the researchers unveiled a bipolar field-effect transistor based on layers of graphene and atomically thin boron nitride with ON-OFF ratios an order of magnitude greater than planar graphene FETs. At the same time, other researchers have fabricated devices with mobilities almost an order of magnitude better than devices on silicon dioxide.
As Novoselov's colleague, Colin Woods, explains, hBN has a very similar lattice to graphene, making it an attractive candidate for building graphene heterostructures.
"It's also a high bandgap semiconductor and has a high crystal purity so we've been getting very good results with these heterostructures," he adds.
Excitingly, recent research published in Nature Physics, shows that growing graphene on hBN can alter its crystal structure, opening a gap between its electron energy bands.
Novoselov, Woods and colleagues found that when the two lattices are oriented in a certain way, the distance between graphene's carbon atoms is stretched so its lattice matches that of hBN, and the bandgap is opened. If this alignment is off, the carbon atoms revert to original positions and the bandgap disappears.
“It was extremely exciting to see the properties of graphene change so dramatically by simply twisting the two crystals only a fraction of a degree," says Woods. "We're not yet sure whether the bandgap is a result of [lattice] strain, but obviously a bandgap in graphene is very desirable."
"It is very small but you never know where this might go in the future," he adds.
And crucially, as the researcher points out: "This discovery provides another way to control the properties of graphene heterostructures. Not all the questions are answered and this is certainly a phenomenon worth investigating."
But the research doesn't stop with hBN. Novoselov's group is looking at other 2D crystals - from molybdenum disulphide to niobium diselenide - all of which hold considerable promise for graphene.
"Each material has its own set of properties and its difficult to single out a single crystal and say 'yes its properties are going to be good' as they change as you alter the number of layers in the heterostructure," says Woods. "But this is a very exciting time to be looking at graphene; people honestly don't know what's going to happen until they do it."
