by Brian Clegg
Concerned that a chemical reaction had taken place that could have rendered the gas explosive, Plunkett took the cylinder outside his lab and rigged up a blast shield before cutting through the casing. Inside was a slippery, white, waxy solid – PTFE. The high pressure the gas was at, combined with the iron interior of the cylinder acting as a catalyst, had enabled the polymer to form. PTFE was first used, as it still is, to coat joints to make a good seal, but we also come across it in non-stick pans, where it’s mostly known by the brand name Teflon.
Those non-stick pans were first made in France in 1954, when Colette Grégoire suggested to her husband Marc that he find a way to prevent food sticking to pans using the slippery PTFE he had put on his fishing tackle to make it run more smoothly. It proved a non-trivial task, as the PTFE didn’t easily stick to the metal, but Grégoire achieved it by first pitting the aluminium surface with acid, then heating PTFE powder on the surface so it gripped the uneven metal. The Grégoires would start a company, Tefal, selling pans with Teflon coating. PTFE naturally repels water and has very little for other molecules to cling on to – it even resists the van der Waals forces (see page 80 ) that enable a gecko to walk up a vertical sheet of glass.
Conventional PTFE molecules are effectively onedimensional structures in the form of long chains of linked carbon atoms, each attached to a pair of fluorine atoms. Fluorographene takes PTFE up to two dimensions (admittedly with only one fluorine atom per carbon atom), which may prove particularly beneficial in obtaining extremely even layers for high-tech equipment, or duplicating PTFE’s low adhesion abilities on much smaller objects.
The versatile film
Whether using graphene, which continues to have the most extraordinary capabilities of any substance known, or the various alternative two-dimensional materials, there has been an explosion of potential applications of these versatile substances. Many are still at the development phase – not surprising, given that graphene was only first produced in 2004. However, the range is remarkable.
* More accurately you could describe boron nitride as a colleague rather than a rival as it is often used alongside graphene.
† In quantum computing, rather than bits, the computer works with qubits, where each qubit is the combined states of a quantum particle. The capabilities of the computer increase exponentially as you add qubits, but as yet it has proved difficult to make a stable, large-scale quantum computer.
6
THE ULTRATHIN WORLD
With hundreds of scientists working on ultrathin materials around the world, exploring both the underlying physics and chemistry of their two-dimensional structures as well as developing ways to use them, we are already seeing some remarkable products at the prototype stage, even though at the time of writing there is little yet in full production.
Almost inevitably, one of the first areas to be given consideration is the future of ultrathin electronics.
Graphene transistors
Given the observation of the field effect that Geim and Novoselov made with devices constructed from sticky tape, silver paint and toothpicks, it is no great surprise that one application that sprang to mind early was finding a way to produce graphene transistors. With its incredible thinness, a layer of graphene would make for a tiny circuit that could even be flexible if required. Although, of course, it requires a substrate to support the graphene’s inherent floppiness, which limits the overall thinness possible, this substrate can itself be thin, and it’s possible to imagine thousands of graphene-based circuits, perhaps piled on top of each other with suitable insulating layers.
Field effect transistors (see page 65 ) made from graphene are of extremely high quality, allowing control over high levels of quasiparticle flows by an imposed electric field, beating the best silicon/metal oxide field effect transistors. They are particularly effective for the high frequency applications which are often required by modern electronics.
There is, however, what appears initially to be a serious problem when it comes to producing integrated circuits based on graphene rather than silicon. To work together to provide the logic gates of a computer chip, for example, the transistors in the tiny circuitry have to be able to turn flow on and off – to act as switches, rather than as amplifiers. As we have seen, graphene is a wonderfully good conductor, so much so that it is very difficult to avoid leakage of quasiparticles even when their transport is suppressed as much as possible.
This is not an impossible problem to overcome, though. One approach that has been tried, with success, is to use graphene-based structures, such as multi-layered ribbons or ‘quantum dots’ which aren’t made up of a pure sheet of graphene, but can exhibit more of the on/off control required for a logic gate and still have many of graphene’s benefits.
Another option would be to selectively react carbon atoms with fluorine. As we have seen (page 112 ), fluorographene, with a fluorine atom attached to each carbon atom, is another potentially useful ultrathin film. And it is possible in principle (though not yet in practice) to set up a hybrid of graphene and fluorographene by adding fluorine to selected carbon atoms in the lattice. Fluorographene is a very effective insulator, and adding its insulating properties to limit conductivity could make it possible to devise a fully operational circuit complete with logic gates all embedded within a single sheet of graphene.
Finally, there is the possibility of using dual-layer graphene. Putting together two layers of graphene does not return it to being boring old graphite, but modifies the properties of the graphene. Depending on the way the two layers line up, it can have interesting electronic properties of its own, which are different to pure graphene. In the so-called Bernal arrangement, where half the atoms in one layer are above other carbon atoms, but the other half are above the centres of the hexagonal spaces, dual-layer graphene or bigraphene has shown promise in having the right kind of band gap to facilitate the production of computer gates. Alternatively, we could still look to lithography, the technique used on conventional silicon chips, but with a new twist, as will be discussed in the next section.
Graphene transistors can not only make use of field effects for amplification, but also provide the possibility of producing ultrafast transistor-based photodetectors, which work across the visible and infrared light spectrum. Not only would such detectors be flexible in the ways they can be used, they would literally have the potential to be flexible, providing light detection facilities and potentially the mechanism for a miniature camera that can be adapted to any shape and size.
Working two-dimensional chips
Although there are undoubted problems with using graphene as a basis for integrated circuits, there are obvious benefits in using a material that is cheaper, more resilient and far thinner than silicon. By 2011, IBM had announced the first prototype chip that was based on graphene.
The demonstration was a radio frequency mixer chip. These devices take two different signals and produce an output based on some combination of the two – most commonly, a mixer outputs the sum of the two signals. The circuit was made using two to three layers of graphene. To form the circuit, the graphene was first spin-coated with a polymer. This is a process where a dot of the polymer is placed in the middle of a piece of graphene, which is then rotated at high speed – typically around 50 revolutions per second – so that the coating spreads out to form a thin layer over the surface.
The polymer was itself then given a coat of hydrogen silsesquioxane. This may sound like a toothpaste additive, but is in reality a compound of hydrogen, silicon and oxygen which forms a polymer that works well as a resisting material when producing the pattern that is required to form an integrated circuit chip. The structure of the circuit was then produced using electron beam lithography. This is a process using a tightly focused beam of electrons, which changes the structure of the hydrogen silsesquioxane, making it soluble. The treated parts are then removed, leaving behind channels which will act as a mask so the unwanted graphene can be removed with a laser.<
br />
After the lithography was complete, the graphene chip was cleaned using acetone, resulting in a final device less than 1 square millimetre in size. Although this approach isn’t suitable for mass manufacturing – electron beam lithography is slow and expensive – it did demonstrate that it is possible to make a functioning graphene-based chip. What’s more, the device worked with very high radio frequencies, up to 10 gigahertz. For comparison, FM radio is typically around 0.1 gigahertz, digital radio is around 0.22 gigahertz, and mobile phones operate at around 0.85 to 1.9 gigahertz. The ultra-high frequencies that the graphene chip can handle make it ideal for secure, close-range military communications.
Perhaps the biggest difficulty with pure graphene electronics to date has been that it is just too good a conductor. As we have seen, graphene transistors work fine as amplifiers, but without modifying the molecule, it has proved impossible to get them to switch off entirely – which has limited the application of electronics based on graphene alone. However, in late 2017, researchers at Rutgers University in New Jersey announced that they had discovered a way to make graphene pause its conductivity – which has the potential to provide an alternative route to making fully functional graphene transistors a reality. To do this, they made use of the probe of a scanning tunnelling microscope – a device we encountered briefly in Chapter 1, whose calibration method facilitated Geim and Novoselov’s early supply of Scotch tape-graphene samples (see page 14 ).
Scanning tunnelling microscopes are remarkable devices, which bring a tiny, electrically charged point extremely close to the surface of the material to be observed. When the point is almost touching the surface, electrons from the charged point of the device undergo quantum tunnelling across the barrier formed by the gap between the tip and the material beneath. Any tiny changes in separation between this point and the surface have a strong effect on the electrons’ ability to tunnel, so as the tip is moved back and forth over the surface it can plot out an object’s surface in great detail.
It might seem that the tip needs to be incredibly sharp to work on this scale – and it does – but achieving this is rather easier than might be expected. Ideally the point should have a single atom of the metal it is made from protruding from its end. The sophisticated technology used to hone the metal to this invisibly fine point is a pair of wire cutters. A thin piece of metal wire is simply snipped with the wire cutters – in practice, there is always a single atom sticking out somewhere on the top, and this works fine as the microscope’s tip.
By varying the electrical current applied, the tip of a scanning tunnelling microscope can also be used to move individual atoms – famously, IBM scientists spelled out the company’s name in 35 xenon atoms using one of these devices back in 1989. The Rutgers researchers used the same effect to create an electrical force field in the graphene beneath it, which stops the charge-carriers in their tracks or forces them to travel along certain paths while the probe is present, like a lens acting on light.
‘IBM’ in atoms.
(Image originally created by IBM Corporation)
Although a scanning tunnelling microscope is small compared to traditional electron microscopes, it is still far too bulky to be part of an electronic circuit. However, the vast majority of the microscope is not necessary for this use. All that should be required is a set of tiny wires on top of the graphene, which can set up the electrical field to control electron flow and turn the graphene into circuits containing fully functional transistors.
Spin doctors
As we have seen (page 111 ), as well as better conventional electronics, graphene and its fellow ultrathin materials are also likely to play a part in the still infant field of spintronics, where not only electrical charge but also the spin of the electrons is used in producing logic circuits.
Researchers in Spain, the Netherlands and Germany, as part of the Graphene Flagship initiative (see page 150 ), made significant steps forward in 2017 on the kind of development needed to produce a practical spintronic device. Just as it stands out in so many other ways, graphene has been discovered to have a uniquely high ‘spin lifetime anisotropy’ – in effect the ability to keep spins locked in a particular direction for far longer than normal. The graphene-based device acts as a filter which will only transmit certain directions of spin, allowing it to act as an ultra-sensitive detector of spin changes, necessary to make use of spin as the equivalent of the state of a traditional electronic bit in computing.
When graphene is combined with other ultrathin materials, such as molybdenum disulphide and tungsten disulphide, the interaction between the layers can provide a mechanism for controlling the lifetime of the spin directions before the electrons undergo ‘spin relaxation’, reverting to random directions – the equivalent of going from a 1 to a 0 in an electronic logic circuit. If reliable spintronic logic devices can be constructed – and the Graphene Flagship’s collaboration between researchers and development companies seems ideal to fast-track this – they could benefit both conventional devices and the new field of quantum computing.
However, not every electronic or even spintronic application of graphene requires as complex a structure as an integrated circuit. In some cases, it is enough that it is a superb conductor that just happens to be transparent too.
Light fantastic
Anyone involved in the production of solar cells is certainly keeping an eye on the development of graphene. To make an effective solar cell, producing electricity from sunlight, requires the use of a conducting layer of material that is transparent. To date, the need to combine transparency and conductivity has meant using very thin layers of metal, or metal oxides. However, these tend to be less transparent than graphene, * are expensive to make, are more variable in the frequencies of light that they absorb and tend to be more likely to undergo unwanted chemical reactions.
Prototype solar cells using graphene produced by chemical vapour deposition (see page 90 ) have been made now for some years and it should not be long before graphene is transforming this industry. What’s more, combined with the capabilities of a flexible ultrathin light-absorbing semiconductor such as molybdenum disulfide, graphene could help make far better flexible films that are still able to generate electricity from sunlight, wrapped around any shape of surface.
The same requirement for a transparent conductor is also true of LCD screens and touchscreen technology. These require significantly larger sheets than a solar cell, which is typically only a few centimetres across (solar panels are made up of arrays of multiple cells), which in the early days of ultrathin film production was a problem, but the newer production techniques are making it possible to produce two-dimensional materials on the kind of scale required for modern displays. Bear in mind how LCD displays, which were first limited to a few centimetres across, are now routinely produced for TVs with diagonals of 50 to 70 inches (1.2 to 1.8 metres).
We can expect larger scale graphene panels within a few years. But they are already produced on a scale that could make one of our most familiar domestic disasters less of a problem.
Smashing screens
Anyone who has a smartphone knows how easy it is to break the screen. Drop your phone on a hard surface, and the chances are you will end up with a crazed piece of glass and an expensive repair bill. But a team at the University of Sussex have come up with a flexible alternative. What’s more, their robust screen has the potential to use less energy and will be more responsive than a traditional glass touch screen. It could also be used in other devices where a flexible touch connection would be useful.
The Sussex screen uses a flexible acrylic plastic, coated with a grid of silver nanowires and pieces of graphene, which are floated on water, picked up with a rubber stamp and pressed onto the silver grid in whatever pattern is required – it’s a bit like potato printing with graphene. The benefits are remarkable. The resultant screen is flexible, and thanks to the remarkable conductivity of graphene is about 10,000 times as conductive as it would be wi
th silver alone, making it more sensitive and far less power-consuming.
The amount of metal in the nanowires used is also reduced hugely from a traditional screen, making the graphene and silver approach significantly cheaper than the current most frequently used alternative, indium tin oxide. And the graphene prevents the silver from tarnishing – always a problem when that metal is used in air.
While the flexible screen, like many of graphene’s uses, relies on the substance’s electronic special nature, other possible applications may depend on its other properties: for example, how it deals with magnetism.
Magnetic trick
Appropriately, given Andre Geim’s history of levitating frogs and other diamagnetic objects, it turns out that graphene too is a diamagnetic material. As we have discovered, this is a medium that itself becomes a (relatively) weak magnet in opposition to a magnetic field that it is exposed to. In effect, a diamagnetic object is repelled by a magnetic field. In the case of graphene, the special electronic structure of the lattice makes it a strong enough diamagnetic material to be levitated by conventional neodymium permanent magnets, not requiring the massively strong electromagnets needed to levitate frogs.
Although there isn’t an immediate application for this effect – we aren’t going to see graphene holding maglev trains above the track as a result of diamagnetism, as the repulsive force is too weak – it is likely that this is one more ability of the two-dimensional wonder material that may come in useful in the future alongside its other properties. It’s possible to imagine that a graphene film might be floated over an air gap, for example, using the variable magnetic field of an electromagnet to fine tune its interaction with another material. This would make it possible to vary the charge held in a capacitor at the touch of a button, making it easier to control the speed at which a supercapacitor replacement for a battery (see page 136 ) discharges.
