Atomically thin materials and nano-transistor fabrication are being used by the ARC Centre of Excellence in Future Low-Energy Electronics Technologies (FLEET), a new Australian research collaboration developing ultra-low energy electronics for an impending computing energy crunch.

FLEET’s research is at the very boundaries of what is possible in modern condensed matter physics and nanotechnology. At the nano scale and smaller, synthesis of atomically thin materials and nano-fabrication of functioning devices will be key to the centre’s mission success.

Atomically thin, or two-dimensional (2D), materials are the material of choice at FLEET, a collaboration of researchers from seven Australian universities developing novel electronics and electronic devices that will operate at ultra-low energies. These would replace the silicon of traditional CMOS (complementary metal-oxide-semiconductor) devices in a new generation of electronics in which electrical current can flow with zero, or near-zero, resistance.

The driving force behind FLEET is the significant and increasing amount of energy being used in information technology, computing and communication, which already represents 5% of global electricity use, and is doubling every decade. As electrical current traverses the material of normal, silicon computer chips, electrons scatter and consume energy. It’s only a tiny amount of energy per ‘switch’, but with billions of switches per chip, switching billions of times per second, in millions of processors in data centres around the world, it adds up.

To head off this looming power crunch, FLEET will use two systems in which electricity can be carried with very, very low dissipation of energy:

  • First is the relatively new science of topological materials, which are insulators in the centre of the material, but can carry current on their edges or surfaces, potentially with near-zero resistance. This is the science that was recognised by the 2016 Nobel Prize in Physics.
  • The second system is exciton superfluids, in which pairs of bound, oppositely-charged particles can be made to ‘flow’ in a superfluid with zero electrical resistance. Exciton transistors will switch off and on just like conventional transistors, but without dissipating energy.

FLEET also uses light-transformed materials, which can be temporarily forced into either a topological or superfluid state.

Each of these systems for ultra-low energy electrical conduction depend heavily on atomically thin, 2D materials. For example, in order for topological insulators to carry current with near-zero resistance, they must be extremely thin, such that their conducting edges are one-dimensional wires in which current can only go forwards or backwards. To achieve large bandgaps in topological insulators sufficient for operation at room temperature, the materials must be atomically thin. Similarly, exciton condensation, which requires a large binding energy between charged particles, is limited in any 3D material by the relatively large (tens of nanometres) distances between the oppositely-charged particles. In atomically thin semiconductors, this distance can be reduced to less than one nanometre.

Synthesis of atomically thin materials

FLEET’s work depends heavily on the recent revolution in atomically thin (2D) materials. To provide the required materials for future electronics, FLEET draws on extensive expertise in materials synthesis in Australia and internationally, from bulk crystals to thin films to atomically thin layers.

The most well-known atomically-thin material is graphene, which was first isolated in the lab in 2008 (winning its discoverers the 2010 Nobel Prize in Physics). Graphene is a single, 2D sheet of carbon atoms arranged in a hexagonal ‘honeycomb’ lattice. It is an extraordinarily good electrical conductor, with electron speeds more than a hundred times faster than in silicon.

FLEET’s main research uses other atomically thin materials, with researchers seeking materials possessing the necessary properties for topological and exciton superfluid states, and that can be temporarily forced into those states.

For example, for topological function, materials must form two-dimensional structures, be electrically insulating in the centre of the material, and have a large inverted bandgap (i.e. be topological). The inverted bandgap results from spin-orbit coupling, which is a relativistic quantum mechanical effect that is much stronger in heavy atoms. Hence large bandgap topological insulators will require heavy elements such as lead or bismuth.

To manufacture such materials, FLEET synthesises materials with robust two-dimensional quantum spin Hall effect systems, with large bandgaps. Magnetic topological insulators can offer even better protection from resistance, and FLEET is studying doping with magnetic impurities to introduce magnetic ordering into these systems.

FLEET researchers Lan Wang (RMIT University) and Xiaolin Wang (University of Wollongong’s Australian Institute for Innovative Materials) use semiconductor fabrication facilities at those institutions to synthesise several materials that are predicted to have electrically-tunable large-gap quantum spin Hall systems, using mechanical and chemical exfoliation techniques to obtain single-layer and multiple-layer 2D flakes.

At Monash University, Michael Fuhrer and Mark Edmonds are growing atomically thin layers of bismuth-containing materials by molecular beam epitaxy onto atomically flat substrates.

Meanwhile at UNSW, Nagarajan Valanoor and Jan Seidel are taking a different approach, assembling a 2D topological insulator at the interface of two different complex oxide materials, which can be deposited by pulsed laser deposition. Deposition at the junction eliminates the issue of the atomically-fine material dissipating on contact with air.

A functioning transistor must also be switchable, and so other, novel substrate materials with the correct electric and magnetic ordering are needed to provide control of the properties of atomically thin substrate.

Nanofabrication of electrical devices

For the new technology to compete with silicon, it will be necessary that these 2D and other new materials be incorporated into nano-scale devices. As well as developing theoretical electronic systems, FLEET is charged with developing functioning electronic devices in which that ultra-low resistance electrical current can be controlled – switched off and on, as in the switches and transistors that comprise traditional, silicon-based systems.

The Centre will develop practical, functioning electronic devices based on the new technology. FLEET will use a range of advanced fabrication techniques to incorporate the atomically thin materials developed into novel device structures with suitable performance. Atomically thin topological insulators will be integrated with electrical control gates to create topological transistors. And atomically thin semiconductors will be integrated with optical cavities to create exciton transistors in which a superfluid current can be switched on and off. Topological transistors and exciton transistors will form the backbone of future electronics, replacing the relatively high-energy silicon transistors of our current systems.

FLEET nanofabrication researchers use the expertise and facilities of the Australian National Fabrication Facility (ANFF) at RMIT University in Melbourne (the Micro Nano Research Facility), the Melbourne Centre for Nanofabrication (MCN), and at the Australian National Fabrication Facility –New South Wales (ANFF-NSW).

The MCN is an open-access research and development facility that allows users such as FLEET researcher Qiaoliang Bao to fabricate structures on the nanometre scale using techniques like electron-beam lithography.

Nearby at Monash University’s Faculty of Engineering labs, Bao’s team grows atomically thin materials on copper foil using chemical vapour deposition, and uses scanning tunneling microscopy and other imaging methods to confirm the resulting structures are atomically thin.

Elsewhere in FLEET, Oleh Klochan, Alex Hamilton, and Oleg Sushkov at ANFF-NSW are using patterning at the nano-scale to turn a conventional 2D semiconductor layer into a topological insulator.

Why do it? The energy costs of computing

FLEET addresses a grand challenge: reducing the energy used in information and communication technology (ICT), which already accounts for at least 5% of global electricity use, and is doubling every ten years.

And it’s going to get worse. The current silicon-based technology will stop becoming more efficient in the next decade as ‘Moore’s Law’ comes to an end. Moore’s law is an unofficial phenomenon, first cited by Intel co-founder Gordon Moore in the 1960s, whereby the number of chips per unit area doubled every 18 months. As transistors got smaller, they also generally got more efficient in terms of energy use. But Moore’s Law is slowing and expected to come to an end in the next decade, with transistors no longer getting smaller or more efficient.

The so-called ‘Internet of Things’ would drive this energy demand even higher. However, within a decade, the financial and environmental cost of electricity use will limit the growth of computing.

FLEET is a collaboration of researchers from seven Australian universities and 13 Australian and international partner organisations. The centre received funding in the September 2016 Australian Research Council funding round. FLEET’s participating organisations are: Monash University, the University of New South Wales, Australian National University, RMIT University, Swinburne University of Technology, University of Queensland and University of Wollongong.

FLEET’s national and international partners include Australian Nuclear Science and Technology Organisation (ANSTO); the Australian Synchrotron; California Institute of Technology (Caltech); Columbia University, New York; Johannes Gutenberg University, Mainz; University of Maryland Joint Quantum Institute & National Institute of Standards and Technology; Max Plank Institute of Quantum Optics; the National University of Singapore; the University of Colorado, Boulder; University of Maryland Center for Nanophysics and Advanced Materials; the University of Texas, Austin; Tsinghua University, Beijing; and the University of Wurzburg.