Graphene - from science fundamentals to low-cost production

Keynote presentations on the third day of Graphene Week 2015 offered an eclectic mix of fundamental science and practical chemical engineering. Here we report briefly on each of the talks, beginning with an introduction to optoelectronics in 2d semiconductors and heterostructures, and concluding with an outline of a highly promising ‘kitchen sink’ approach to graphene production. Xiaodong Xu of the University of Washington in Seattle set the ball rolling with a look at the optoelectronics and valleytronics of 2d semiconductors and heterostructures, such as those based on the sulphides and selenides of molybdenum and tungsten. Xu began with a recap of the electronic properties of MoS2, including the change in band gap with number of layers. Of particular interest to Xu are valley-specific interlayer excitons in monolayer WSe2-MoSe2 vertical heterostructures. Optical pumping leads to coupled spin-valley polarisation of interlayer excitons, with measured lifetimes of more than 30 nanoseconds. Such long-lived polarisation allows for the visualisation of both spin and valley diffusion over length scales of several microns. In practical terms, the effect is important in laser technologies with a tuneable carrier density. Marco Polini of the Scuola Normale Superiore in Pisa discussed plasmon damping mechanisms, with a focus on graphene sheets encapsulated in boron nitride. With 2d materials, plasmon lifetimes are two orders of magnitude longer than with bulk materials such as silicon and silver. With graphene and boron nitride, lifetimes as long as a picosecond have been observed. Damping mechanisms outlined by Polini included electron-electron, electron-impurity, and electron-phonon collisions. He went on to discuss experimental and theoretical work on hybrid plasmon-phonon polaritons. In the second part of his talk, Polini looked at direct current transport in graphene, with hydrodynamic flow and current whirlpools observed at length scales of half a micron. Amir Yacoby from Harvard University in Cambridge, Massachusetts discussed observations of edge currents using Josephson interferometry. The idea here is to use superconductivity to study the intrinsic physical properties of graphene. As for the origin of the observed edge currents, Yacoby suggested that certain edge shapes may guide the current. An alternative explanation is guided electron fibre-optic states at the Dirac point. This guided mode theory can explain edge current observations in bilayer as well as in single layer graphene.
Relevant applications include imaging of topological currents at domain boundaries in bilayer graphene, and induced superconductivity in the quantum spin Hall regime.

Graphene is a two-dimensional polymer, noted Klaus Müllen of the Max-Planck Institute in Mainz, and this makes it something of a challenge for materials synthesis. Müllen looked at both bottom-up and top-down production protocols, including the flattening of 3d, propeller-like molecules. The most promising approach to graphene synthesis is electrochemical exfoliation.

Applications of electrochemically exfoliated graphene identified by Müllen include organic photodetectors and transparent conductive electrodes, with the ability to produce ultrathin and flexible devices. Energy storage is another possibility, using exfoliated graphene and colloidal nanoparticles. Such nanoparticles, wrapped in graphene, offer high reversible charge capacity, retention and Coulomb efficiency.

Müllen concluded his talk with some 3d simulations of carbon networks, and noted, with the illustration of a beehive, that nature sometimes makes mistakes.

Manish Chhowalla of Rutgers University in New Jersey began his talk with an overview of molybdenum and tungsten disulphides. These layered semiconductor materials have a number of interesting properties, but the key problem in using them for electronics applications has been high contact resistance with metals deposited on the semiconducting 2H phase.

Contact resistance in MoS2 can be reduced by inducing a metallic (1T) phase on 2H phase nanosheets. Hybrid field-effect transistors with 2H monolayer MoS2 as the channel, and 1T source and drain contacts, display high electron mobilities, low subthreshold swing values, high on/off ratios and drive currents, and excellent current saturation. Deposition of different metals has a limited influence on transistor performance, suggesting that the 1T-2H interface controls carrier injection into the channel. In practical terms, the MoS2 channel must be locally patterned in order to make such structures. This can be done with a PMMA mask to partially cover certain areas. The result is a contact resistance of 0.2 kiloohms per micrometre. In comparison, 2H phase MoS2 has a contact resistance of 1.12 kiloohms per micron. Jonathan Coleman from Trinity College Dublin spoke of his research group's much-lauded graphene production process known as liquid phase exfoliation, aka kitchen-blender graphene. And not only graphene, as the technique can be used to produce nanoscale flakes of a range of 2d materials. Coleman discussed the fundamentals and practicalities of liquid-phase exfoliation, focusing on such matters as control of flake size. The bulk of Coleman's presentation was given to applications, and here he identified a number of areas. These include the mechanical improvement of composite materials, strain and other motion sensors based on electrical conductivity changes, electrical energy storage and printed electronics. The next challenge for liquid exfoliation is to achieve industrial-scale production of graphene and related 2d materials. To this end, Coleman highlighted a collaboration between his research group and chemical manufacturer Thomas Swan.