by Brian Clegg
Because, as we have already seen, graphene is so conductive, it is only necessary to include a small percentage of graphene in a plastic – around 1 per cent of the whole material – to make the plastic conducting, which then gives it a whole range of extra potential applications. This can be done using cheaply produced flakes of graphene just millionths of a metre across. Similarly, powdered graphene could be used to replace the graphite or carbon fibres in batteries, significantly increasing their efficiency.
A sensitive surface
One surprising possible use for graphene is in providing ultra-sensitive gas detectors that can pick up the presence of a single atom of a gas. Such a detector would involve an exposed graphene surface, onto which gas molecules would adsorb – effectively sticking to the surface. Because graphene is such a superb conductor, and its conduction is due to interaction between its crystal lattice and conduction band electrons, even a single atom sticking to it will make a small change to its conductivity, which can be picked up and analysed.
In tests, graphene detectors were inserted into a glass tube which contained either helium or nitrogen, plus a range of contaminants. In the first trials, for common air pollutants from nitrogen dioxide to carbon monoxide, concentrations of one part per million were easily detected from a change in the current flow, with nitrogen dioxide detected almost immediately the pollutant was added. By the end of the tests it was possible to detect the impact of individual gas molecules, making detection in fractions of parts per billion possible.
This kind of application is just the start, though, of graphene’s capabilities, particularly once the existence of other two-dimensional materials is added into the mix.
* Dutch physicist Heike Kamerlingh Onnes again, who discovered some of the remarkable physics of ultra-low temperatures, such as superconductivity.
† Far less well known than Kamerlingh Onnes, American physicist Percy Bridgman’s work on high pressures led to many discoveries about physical materials, though these arguably have fewer practical applications than the extremes of low temperature.
‡ Scientists are rarely happy for long with a simple-sounding name. The method of using sticky tape to remove a graphene layer is now referred to in academic circles as the ‘micromechanical cleavage technique’.
§ There’s an echo here of the development of the laser at the end of the 1950s, when the big players assumed it was impossible to use a ruby to make a laser due to an incorrect set of data they didn’t check properly – but a lone player, Theodore Maiman, used rubies anyway and created the first laser in May 1960.
¶ Though there is something rather appealing about the thought of a whole factory filled with row after row of robots, each applying pieces of sticky tape to graphite and then to a silicon oxide wafer.
|| ‘Epitaxial’ refers to materials grown by epitaxy, which is growing a crystal on a substrate that determines the orientation of the new crystal. The structure of the substrate acts as a kind of template for the crystalline structure.
** The massless quasiparticles are formally known as massless Dirac fermions and have the rare honour of appearing in a TV comedy programme – specifically the episode of The Big Bang Theory titled ‘The Einstein Approximation’, in which the character Sheldon is seen puzzling over their behaviour.
†† Although many requirements for strength aren’t just about pulling the substance, tensile strength often reflects ability to stand up to other stresses and strains. So, for example, a bulletproof vest’s ability to stand impact depends on how easy it is to stretch the material in the vest at right angles to the bullet’s direct of travel, as the bullet will try to stretch the material apart to get through it.
‡‡ This is not some strange material, but the amount of pull required to dislodge a limpet when it is stuck to a surface.
§§ It’s hard to visualise 1020 – it is 100,000,000,000,000,000,000 atoms.
5
OTHER FLATTIES
There’s more to two dimensions than carbon
Graphene sprung to the fore as a result of Geim and Novoselov’s work, but once they had proved that it was possible to make stable two-dimensional materials, graphene was never going to be the only kid on the block. Carbon is not the only atom capable of forming two-dimensional sheets, particularly once compounds are considered. From boron nitride to molybdenum disulfide and the mysterious sounding dichalcogenides, the field of the ultrathin is proving unstoppable.
Graphene goes white
Perhaps the best-known of graphene’s rivals * is boron nitride, which we’ve already met as one of the few substances to come close to graphene in tensile strength. In its two-dimensional form, boron nitride is sometimes known as ‘white graphene’; despite having a totally different chemical structure and very different properties from the carbon-based equivalent, it is another winner of the ultrathin world. Boron nitride is a simple inorganic compound made of pairs of boron and nitrogen atoms. This combination has a similar versatility to carbon in its ability to make bonds. The result is an equivalent set of allotropes with structures similar to diamond and fullerenes. Most importantly here, though, boron nitride can form single-layer hexagonal lattice sheets like graphene, which will initially form a block equivalent to graphite.
In its two-dimensional sheet form, like graphene, the atoms of boron nitride are arranged in a hexagonal lattice, but with the significant difference that atoms around each hexagon alternate between being boron and nitrogen. Each atom is connected to one of its neighbours by a double covalent bond and to its other two neighbours by single covalent bonds. With four bonds per atom, there are no freely available electrons, meaning that hexagonal boron nitride has a wide band gap and is a good insulator.
The hexagonal lattice of boron nitride.
The combination of a similar physical structure but different electron availability means that when boron nitride takes this hexagonal form resembling graphite, it shares some properties with graphite, while being very different in others. For example, like graphite it is a good lubricant. Because it also doesn’t react well with other chemicals, it has gained an unlikely role as a lubricating agent in cosmetics, as well as having more traditional applications such as being embedded in ceramics designed to withstand high temperatures and in providing the slipperiness for self-lubricating bearings. It has even joined graphite as a component of some pencil leads – it doesn’t make a good writing material alone, but it renders the reconstituted graphite more stable. But of particular interest here, as with graphite, is the weakness of the bonds between layers. It can shed single atom-deep layers: the hexagonal boron nitride equivalent of graphene.
One possible use for nano-scale sheets of boron nitride is to reduce water pollution. A sheet made up from the boron nitride layers can absorb up to 33 times its own weight in potential pollutants such as oil and organic solvents. However, the boron nitride repels water molecules, so it proves very effective at removing these kinds of pollutants from water. Once the sheet has become saturated, it can be cleaned by heating up to a high temperature, staying in one piece as the pollutants burn off. The ability to repel water also means that a layer of boron nitride could be used in self-cleaning electronic display screens, which won’t get misty with condensation when the air is humid.
As with all the two-dimensional compounds, one of the most far-reaching potential applications of boron nitride sheets is in electronics. Combining layers of the insulating boron nitride with layers of superbly conducting graphene is a recipe for all kinds of electronic devices. Because we are operating at the scale of atoms, quantum effects can be very strong between the different layers. This means, for instance, that with single boron nitride sheets quantum tunnelling (see page 50 ) can take place, giving the potential to make tiny components making use of this effect. With a number of boron nitride sheets between a pair of graphene sheets it becomes possible to construct miniature high-capacity electrical storage devices, as we will discover in the next chapter.
/> As with all these kinds of multi-layer concepts, we are currently in the early days of development, but the combinations are exciting indeed to those whose job it is to produce smaller and smaller electronic components.
Slippery moly
The compound molybdenum disulfide is probably most familiar to engineers as ‘Moly’ (pronounced molly rather than moley), an additive for grease lubricants that reduces wear on the equipment it lubricates. This naturally occurring compound isn’t quite as thin as graphene in its thinnest hexagonal layer form – rather than a single atom layer, it’s three atoms thick, as a central layer of molybdenum atoms link to sulfur atoms that stick out either side of the sheet. However, it’s still thin enough to count practically speaking as a two-dimensional substance.
The ultrathin version of molybdenum disulfide was not discovered until 2011, produced by the same Scotch tape peeling technique as graphene, though it is now also produced chemically, grown on a silicon wafer. Like the thin layers of carbon in graphene, the molybdenum disulfide layers move easily over each other, giving it its lubricating properties.
Also like graphene and boron nitride, molybdenum disulfide is interesting electronically, but in this case, it completes the set of useful ultrathin materials by being a semiconductor, sitting between graphene’s impressive conductivity and boron nitride’s insulating capabilities. The band gap of molybdenum disulfide is valuable as it’s just right for the change in energy between the conduction and valence bands to match the energy in a light photon, so molybdenum disulfide has plenty of potential both as a light detector, where a photon is absorbed as an electron jumps up in energy from valence to conduction band, and as a light source when an electron jumps down into the valence band.
Perhaps the biggest potential, though, is as an alternative to silicon in developing a much wider range of solid state electronics. Production of traditional silicon-based electronics is getting ever closer to its physical limits of miniaturisation, but in October 2016 what is currently the smallest ever transistor was made with molybdenum disulfide and carbon nanotubes. The transistor’s active part, the gate, was just 1 nanometre across – contrasting with the 20 nanometre gates in the smallest commercially available silicon chip transistors.
By December 2016, researchers from Stanford had moved on from single transistors to show how sheets of molybdenum disulfide can be used to make practical electronic circuits on a scale smaller than their silicon equivalents. It’s still early days for this material, but it is rapidly moving from an experimental material to a mass-market contender.
As with the other two-dimensional films, the physical properties of the transparent sheets of molybdenum disulfide make for extremely thin transistors, which could be built into any sheet of glass. This would make it possible to turn windows, car windscreens or spectacles into information displays. And, once again, the sheet is flexible, which makes molybdenum disulfide one of the options for circuitry that can be built into paper-like screens, go-anywhere solar panels, clothing and more. And the structures aren’t limited to simple sheets – like carbon, molybdenum disulfide can form fullerene structures, including nanotubes. These are showing promise as electrodes in experimental high-performance lithium ion batteries.
In parallel, like its fellow thinnies, the structure of molybdenum disulfide is proving interesting as a filter – in this case, for producing drinkable water from seawater. For some time, experiments have been undertaken using graphene as a porous membrane to allow water through but to block the flow of salt ions in a mechanism called reverse osmosis. After a number of thin film materials were computer modelled for this role in 2015, molybdenum disulfide came out ahead, coping with more than half as much water again as a graphene filter. This benefit seems to be because the pores allowing water through tend to be surrounded by molybdenum, which pulls the water towards the pore, while the adjacent sulfur atoms push the water away, encouraging it to clear the pore and move beyond.
These three leaders of the early ultrathin revolution – graphene, boron nitride and molybdenum disulfide – are likely to provide the backbone of the first generation of ultrathin applications, but they are by no means the only contenders.
From silicene to dichalcogenides
Study the papers being written on the ultrathin and you will find a number of other names cropping up regularly. One is silicene. The element silicon is probably the closest in behaviour to carbon – some scientists have even suggested that there could be silicon-based life somewhere in the universe to complement our familiar carbon-based lifeforms. So, it seems reasonable that silicon could form a structure something like graphene, and it does – given the name of silicene.
Unlike graphene, silicene does not occur naturally and was not discovered until 2010, when the substance was produced in small quantities by depositing silicon onto a silver substrate. And there are significant differences in the way the structure forms. While silicene has the familiar hexagonal lattice, it isn’t absolutely flat like carbon’s, giving a buckled structure that makes it, for instance, less useful as a lubricant, but that can have actual benefits for electronics applications.
The crumpled surface results in a band gap that can be modified easily with an external electrical field, and its atomic structure makes it much easier to react with doping agents. This means that silicene may be better than graphene for producing field effect transistors in an integrated circuit – so graphene isn’t necessarily the end of the road for the silicon chip. One essential that hasn’t been fully explored is what will be the best substrate for silicene to operate on – as yet, these have been expensive materials compared with the substrates used with the core ultrathin materials.
Other ultrathin possibilities are represented by the mysterious-sounding compounds dichalcogenides. In practice, these aren’t as exotic as one might think. A chalcogen is a fancy name for an element in group 16 of the periodic table, which includes oxygen, sulfur, selenium, tellurium and polonium. A chalcogenide is a compound of one of these with another element, typically a metal (by convention, oxygen tends not to be considered as forming a chalcogenide).
The most likely forms to be used are transition metal dichalcogenides, where the other element is from a particular block of the periodic table, such as molybdenum or tungsten. Molybdenum disulfide, described in the previous section, is in fact a transition metal dichalcogenide, and a number of other compounds in this category are also showing promise for electronics, with the right-sized band gap for emission of photons as a light source or absorption of photons as a detector. These include tungsten disulfide, molybdenum diselenide and molybdenum ditelluride. Similarly, this kind of application can be used in field effect transistors (see page 65 ) – increasing still further the range of ultrathin materials available to produce extremely compact and flexible electronics.
Some of these materials, for example tungsten ditelluride, have properties that make them suitable materials for experimental ‘spintronic’ devices. Normal electronics deals only with one property of the electron – its electric charge. However, the electron has other properties too, notably its spin (see page 59 ), which is a quantum property with a value of either up or down when measured in any particular direction. Although it’s early days, a lot of effort is being put into spintronics as it would make it possible to pack more information into a single bit, without all the complexity required to make quantum computing † work.
As semiconductors, the transition metal dichalcogenides are very much complementary to graphene – but the final option to go further with ultrathin materials returns to graphene itself, with an added twist.
Compound interest
In effect, a sheet of graphene is a single, enormous carbon molecule, which can undergo chemical reactions to produce a new two-dimensional substance like any other molecule. To date this has been done with hydrogen to produce graphane, with one hydrogen atom per carbon atom, and with fluorine to produce fluorographene, with a fluorine atom per carbon ato
m. Both compounds are stable substances, though fluorographene seems the more robust of the two and has been explored more to date.
Fluorographene retains the familiar hexagonal lattice of graphene, but with a fluorine atom attached to each carbon. As yet, the best way to produce this seems to be exposing a sheet of graphene to XeF2 , one of the rare compounds of the noble gas xenon. Fluorographene has only been made in small quantities so far, but the evidence is that it is an excellent insulator – so may be useful in multi-layered construction with highly conducting conventional graphene. The degree to which a compound is an insulator depends on the size of the band gap (see page 56 ). Graphene has no band gap, while fluorographene has a very wide one. It is speculated that other graphene compounds could sit between the two, giving them properties more like a semiconductor. This would provide a wider range of substances that could be used in ultrathin electronics.
Fluorographene is also able to resist temperatures up to around 200°C and is chemically inert. Apart from acting as a two-dimensional insulator, fluorographene could provide a thin film equivalent of polytetrafluoroethylene (PTFE), which is a long chain carbon molecule with fluorine atoms attached. It was originally made by accident in 1938 when an American chemist, Roy Plunkett, was trying out different compounds as possible new refrigerant gases. He was using a cylinder of tetrafluoroethylene gas – a simple compound comprising a pair of carbon atoms and four fluorine atoms. The gas in the cylinder seemed to have run out, yet it felt far too heavy to have nothing remaining in it.