Researchers in Switzerland have made the first working transistors made from flakes of molybdenite just one molecule thick. A mixture of molybdenum and sulphur, the material is a semiconductor with direct bandgap, which means that could be better than silicon for making certain photonic devices. Indeed, the scientists believe molybdenite could even rival “wonder material” graphene in its potential for use in future electronic circuits.
In 2005 Nobel laureates Konstantin Novoselov and Andre Geim showed that extremely thin flakes of molybdenite (MoS2) can be produced using the same “sticky tape” method that they used to obtain graphene. However, this preliminary work did not show that molybdenite had any interesting properties or potential applications. Since then, most scientists have shown little interest in this and other 2D materials, but have largely focused on graphene instead.
However, Andras Kis and colleagues of the Ecole Polytechnique Federale de Lausanne (EPFL) decided to study molybdenite and have now revealed some remarkable results. “We have spent the last 18 months or so improving our molybdenite fabrication technique to make working transistors that can be turned on and off at reasonable single-digit voltages,” said Kis.
“We have also shown that we can incorporate some advanced features like high-k dielectrics (for example, HfO2, the material used in Intel processors today) and turn the transistor off using a local gate,” Kis added. Although these two features are not unique in themselves, they do show the potential of molybdenite, he explained.
Electron-hole pair excitations
So why could this material rival both conventional silicon, and newcomer graphene? Firstly, and most importantly, single-layer molybdenite is a direct-bandgap semiconductor, unlike silicon, which has an indirect gap. It is easier to make devices like LEDs, solar cells and photodetectors and any other photonic devices that exploit electron-hole pair excitation with direct rather than indirect gap semiconductors.
The material also has good charge mobility – greater than 200 cm2/Vs, according to Kis and colleagues, and sometimes up to 800 cm2/Vs. This compares well to state-of-the-art silicon but is not as high as gallium arsenide – making the semiconductor “suitable for everyday use,” according to Kis. “While the figure may seem insignificant compared to graphene, we have to remember that the oft-quoted 120,000 cm2/Vs refers to graphene in its bandgap-free form.”
Finally, single-layer molybdenite is only about 0.65 nm thick – which means that very short transistors can be made from it. Such devices would dissipate less waste heat than conventional transistors made from silicon. “Molybdenite is the thinnest semiconducting ‘sheet’,” said Kis. “Our back-of-the-envelope calculations imply that we should be able to make a molybdenite transistor that is 3–4 nm long and still be able to turn it off. To do the same with silicon, you would need a layer 0.3 nm thick, which is not possible.”
Mind the bandgap
Kis says that although graphene is a fascinating material (his team is also working on it) that may have “great future applications”, it may not be best suited for making digital electronics because its lack of a bandgap. “The entire semiconductor industry is based on the idea of having a bandgap,” he stressed. “Graphene in its basic form is simply not a semiconductor whereas molybdenite is. It is better to make a transistor from a semiconductor because this means the device can be turned off easily.”
Although a bandgap can be induced in graphene by cutting the material into thin ribbons or applying an electric field to bilayer graphene, this is complicated – according to Kis. “You need to first make ribbons that are around 5–10 nm wide and at the end of the day you have a mobility that is similar, if not lower than that of molybdenite. Bilayer graphene requires applied voltages of around 100 V to create a bandgap of about 0.25 eV and you would simply not have access to such high voltages in devices like cell phones, laptops or tablets.”
“In molybdenite, we have a 1.8 eV bandgap from the start and don’t need to make super-narrow ribbons or apply voltages that could zap you,” he said. The larger gap also means that a molybdenite transistor can turn off more completely than those with made with bilayer graphene.
“Why bother with all these difficult things – why spend so much effort in creating a bandgap when you could just start with a semiconducting material and go from there?”
All well and good, but the researchers do stress that much more work is still needed to really assess molybdenite. Questions that remain include is it economically feasible to produce the material in large quantities – and can it be easily doped?
Stability also important
Andre Geim told physicsworld.com, “The Swiss team has managed to get the quality of molybdenite 100 times better than what we achieved back in 2005. This makes a whole sea of difference. However, stability at ambient conditions will always be a big issue, reducing quality. But, if their trick with HfO2 is reproducible, this area will explode.”
Kis hopes that his group’s work will motivate other researchers to start working on alternative 2D materials, especially semiconductors and single-layer molybdenite itself. “It would also be nice if the semiconductor industry could devote molybdenite at least a fraction of resources and attention it does to graphene today,” he added. “Molybdenite can be bought online fairly cheaply, like graphite for graphene production, and a lot of the know-how from graphene could be directly transferred to molybdenite.”
The transistor is described in Nature Nanotechnology DOI: doi:10.1038/nnano.2010.279.
