Page 538 – Physics World

Home » Page 538

Topological insulators enter the fourth dimension

Materials known as topological insulators have the unusual property of being able to conduct electricity on their surface even though they are insulators inside. If that weren’t strange enough, physicists have now shown experimentally that such materials can exist in four spatial dimensions. The researchers, based in Singapore and the UK, achieved this feat by wiring up a stack of electric circuit boards to represent a highly interconnected lattice, and say that their scheme could be extended to five or even six dimensions.

The importance of topology can be seen by comparing the spatial characteristics of a sphere and a doughnut. It’s not possible to transform one into the other by gradually varying their shapes – no amount of stretching, squeezing or other distortion will turn a doughnut into a sphere, and vice-versa. Such a change can only come about through an abrupt, discontinuous action – the insertion or removal of a hole.

A similar idea underpins topological insulators, but the transformation in this case doesn’t involve a material’s spatial properties. Instead, it concerns the wavefunctions of the electrons flowing in the material. These contain what is in effect a knot, making it impossible to smoothly convert a topological insulator into a normal insulator. Instead, the material must first become a metal. This means that the material’s physical boundary with the outside world conducts electricity, even though its bulk is insulating.

This phenomenon has previously been seen in two and three dimensions. In the former case, first observed in 2007 and similar to the quantum Hall effect, the material in question consists of an insulating plane with one-dimensional electric currents flowing around the edge. The latter case, detected from 2008 onwards, sees currents flowing in a range of directions across the surfaces of a bulk insulator. In both cases, the currents are generated when an electron’s spin interacts with its orbital motion as it travels through space. These currents persist even when the edges or surfaces involved are deformed.

Beyond three dimensions

Theorists tell us that topological insulators should also exist in higher dimensions. Indeed, researchers have previously observed them in four dimensions, but they did so by mapping a 4D system to a 2D one. The latest work instead involves the direct creation of a 4D topological insulator that acts as a standard insulator when confined to three or fewer dimensions, and only exhibits its exotic properties once it is extended into the fourth.

The new 4D insulator is based on a model drawn up by Hannah Price of the University of Birmingham, and has been experimentally realized by a team led by Yidong Chong at Nanyang Technological University. To create their insulator, the team relied on an artificial construct known as a metamaterial. Building this metamaterial involved making dozens of electrical oscillators from capacitors and inductors, and then arranging them in grids on several circuit boards stacked on top of one another. With each oscillator behaving like an artificial atom, says Chong, the system as a whole acts like a crystal through which electric current can flow.

Two sets of four square grids of dots. The left set has dark red dots on the edge of the grid, representing high current flow. The right set has milder yellow-to-orange dots throughout, representing low current flow (insulator state) — **Surface effect** Experimental data showing waves flowing within a four-dimensional metamaterial. (Courtesy: Wang *et al.*)

The idea was to hook up the oscillators so that they behave electrically as they would in four dimensions, even though they are actually sitting in our 3D world. As Chong explains, this wouldn’t be possible if each oscillator were only linked to its nearest neighbours. But an extra dimension can be added to an otherwise one-dimensional row of oscillators by linking each of those objects to two others further up and down the line. This idea can then be extrapolated so that in 3D space connections beyond those to the (six) nearest neighbours yield a circuit that behaves as if it were sitting in 4D-space.

Testing predictions

To test their system, Chong and colleagues sent voltage pulses through the circuits. They found that when the circuits were wired up according to Price’s model those pulses propagated in waves across the connections made to create the material’s 3D “surface” but not within the 4D bulk. They also observed that by tweaking the circuit components they could make the waves disappear from the surface – so turning the material into a straightforward insulator.

According to Chong, scientists had previously thought it impossible to realize such a 4D system experimentally. “It was believed to be just the province of theoretical study,” he says. “The trick here is that we are not using an actual material but a synthetic one.”

Krzysztof Sacha of Jagiellonian University in Poland, who was not involved in the research, reckons that the work bodes well for possible applications of topological insulators, given that the surface states remained intact even though the components had imprecise values of capacitance and inductance. “Such robustness of topological systems to perturbation is very promising from the point of view of future applications where quantum computing is one of the most important candidates,” he says.

What is a topological insulator?

Price says that this particular research won’t be applied to quantum computers because the metamaterial used is a purely classical system that can’t support entanglement. But she argues that the work paves the way for other higher-dimensional systems. “There are suggestions that some really cool things could happen in 5D and 6D,” she says.

Chris Hooley of the University of St Andrews in the UK also thinks it unlikely that this particular system will have immediate applications. But he agrees that in the long term, technology could benefit. “The general idea that protected surface states might be of use in quantum computation has been around for a while,” he notes. “And this sort of system gives us some new kinds of protected surface state, so one never knows!”

The research is published in Nature Communications.

High-quality graphene foams are made from organic waste

High-quality graphene foams have been made from the waste gases produced during the high temperature treatment of organic waste. Researchers in China say their pyrolysis-based technique is cheaper and more environmentally friendly than conventional production methods.

The materials are structurally similar to conventionally manufactured graphene foams, say Hong Jiang at the University of Science and Technology of China in Hefei and colleagues. The foams also display similar electrical-transport performance and were able to filter and absorb liquids such as benzene and paraffin, just like other graphene products.

Graphene foams are porous, 3D versions of the 2D sheet of carbon that is conventional graphene. Like other forms of graphene, they are super strong, have a high electron mobility and are great conductors of heat. The foams have a wide range of potential uses, including in energy storage, environmental purification, chemistry and biosensing.

Chemical vapour deposition

Graphene foams are normally produced by chemical vapour deposition. Carbon containing gases, such as methane, and carrier gases are introduced to a heated metal foam substrate – usually nickel or copper. As the gases contact the substrate, they deposit a film of carbon atoms. Once the reaction is complete, the metal is etched away leaving behind a 3D lattice of graphene.

Chemical vapour deposition is expensive and uses large amounts of hydrogen, methane and other gases – which is why Jiang and colleagues set out to produce graphene using the carbon-rich waste gases from bio-refineries that run off organic waste. These facilities heat biomass to 500 °C or more in the absence of oxygen to produce alternatives to crude oil-based products, such as biofuels. This process, known as pyrolysis, has been used for thousands of years to produce charcoal.

The team set up a high-temperature pyrolysis system and produced graphene foams using two different feedstocks: powdered cellulose and powdered lignin – both derived from plants. The powders were pyrolysed at 800 °C and the waste gases filtered to obtain small-molecule gases. These purified gases were then fed into a chemical vapour deposition chamber and deposited on nickel foam.

Various techniques, including Raman spectroscopy and scanning electron microscopy, were used to probe the structure of the graphene foams. “Both products are good quality with no obvious defects,” Jiang told Physics World.

Straw and sawdust

But powdered cellulose and lignin are not the same as biomass waste. To simulate more conventional organic waste, the researchers next fed their system straw and sawdust. The graphene foams produced showed minor drops in quality compared with those from cellulose and lignin. The researchers say, however, that all the foams they made exhibit uniform structures and excellent performance in environmental or energy-storage applications.

Materials rich in lignin, cellulose and hemicellulose would be best for producing 3D graphene foams, Jiang says. But other organic waste could be used. “The quality of 3D graphene foam may be influenced by other elements of biomass waste,” he explains. “For example, nitrogen or sulphur could be doped into the obtained products if feedstocks contain high amounts [of these elements].”

Edward Randviir, a chemist at Manchester Metropolitan University, who was not involved with the research, told Physics World that as graphene production usually relies on fossil fuels or pure graphite, a critical rare material, looking into alternative production methods is a worthwhile pursuit. He adds that the work shows that producing graphene from renewable biomass is “quite clearly feasible”.

Life-cycle analysis

A life-cycle analysis found that producing graphene foams from the pyrolysis of saw dust uses less energy and has a smaller environmental footprint than a conventional chemical vapour deposition process using methane. Biomass-derived foams had a lower impact in 15 out of 18 categories, including global warming and land use.

As well as graphene foams, Jiang’s process also produces bio-oil. The team say that making a high-value product like graphene alongside such biofuels could make improve the economic viability of biomass pyrolysis and support its wider commercialization.

Randviir is not convinced by this, however. While he thinks there is potential in producing graphene from organic waste, he questions whether it can be done on a large scale. And, if it is possible, the economics may change. Graphene is currently expensive because no one has figured out how to produce it on an industrial scale, he explains. If they did, the price would drop, adding less value to the pyrolysis process.

Jiang and his colleagues are now looking to explore the relationship between the yield of graphene foams and bio-oil, and further evaluate the cost and value of their process. They also wish to enlarge the size of the graphene foams and better control the number of graphene layers, Jiang says.

The research is described in Nature Sustainability.

Sourdough tips from Fermilab, anti-5G USB stick does nothing, tracing a message in a bottle

One curious result of the COVID-19 lockdown is an explosion of interest in making sourdough bread. My wife and my eldest daughter are using their skills as a materials scientist and a biochemist respectively to see who can make the best sourdough (top tip from the biochemist, try baking yours in a covered Dutch oven).

So, it is not surprising that physicists are also getting into the act. There is a fantastic article in Symmetry about a sourdough starter at Fermilab that was brought to the accelerator facility in 2017 – when Ryan Mueller arrived from Texas A&M University with a jar of microbes.

As well as giving a bit of sourdough lore, the article has several recipes and some fantastic photos of bread baked by Fermilab staff. It is by Jerald Pinson and called The sourdough starter physics family.

And if that doesn’t sate your appetite, check out The physics of bread by Bob Crease.

Expensive nonsense

The BBC is reporting that trading standards officers in the UK have come to the not-so-surprising conclusion that a £339 USB stick claimed to offer protection from 5G cellular networks is expensive nonsense. They are seeking a court order to take down the website offering the 5GBioShield, which they describe as being no more than a basic USB memory. Never mind that there is no evidence that 5G is harmful in the first place.

The device claims to use a “holographic nano-layer catalyser” to protect against 5G and apparently has been endorsed by a member of the town of Glastonbury’s 5G Advisory Committee. Perhaps the nano-layer catalyser was too small to be spotted by the officers, who probably don’t have access to an electron microscope.

What puzzles me is that surely anyone who believes that 5G signals are harmful would also be wary of USB technology.

Finally, some serious science. Researchers at ETH Zurich and the Massachusetts Institute of Technology have devised a new algorithm that predicts how objects drift on the sea as a result of currents. This could be useful for speeding up search and rescue operations at sea and cleaning up spilled materials. And of course, you could use it to send a message in a bottle to a friend. You can read more in Search and rescue at sea aided by hidden flow structures.

Randomized measurements reveal topological quantum states

Topological materials – materials that have surface properties very different to those found in their bulk – are currently revolutionizing condensed-matter physics thanks to their unique characteristics. Researchers in Austria, France, the US and Germany, have now put forward a new technique to identify and characterize the global invariants that mathematically describe these materials in various experimental platforms in the laboratory. The work could advance our understanding of these structures, which might be used in next-generation energy-efficient electronics and quantum-computing applications.

Topology – sometimes called “rubber sheet geometry” – is a branch of mathematics in which two objects are assumed to be equivalent if they can be continuously deformed into one another by bending, twisting, stretching or shrinking (but not tearing or cutting). In this framework, a circle is topologically equivalent to an ellipse, for example, and a doughnut to a coffee mug. In both cases, the objects can be deformed into the other by stretching.

Topological materials show similar geometries on the molecular scale, which gives rise to several unusual mechanical and electrical properties. Topological insulators, for example, do not carry electrical currents in their bulk, but current does flow along their surfaces through special “edge” states. Crucially, the electrons in these states can only travel in one direction, and they also steer around imperfections or defects on the surface without backscattering. Since backscattering is the main energy-dissipating process in electronic devices, these “topologically protected states”, as they are known, might be useful ingredients in next-generation energy-efficient devices.

Another benefit is that in topological materials, a surface electron with a certain momentum cannot scatter into a state with opposite momentum because to do so it would have to flip its spin. Topologically protected states might thus also be ideal for quantum-computing applications, in which defects usually destroy quantum information (the spin state) carried by electrons.

Topological invariants

While characterizing these topologically protected states is important for classifying the different topological phases that can be realized for potential applications, doing so in laboratory experiments is difficult. This is because the mathematical invariants that describe them are considered as global, overall, quantities and thus cannot be probed on small, local scales.

A team of physicists led by Peter Zoller from the Centre for Quantum Physics at the University of Innsbruck and the Institute for Quantum Optics and Quantum Information at the Austrian Academy of Sciences and Benoit Vermersch from the University of Grenoble-Alpes in France have now put forward a new measurement technique that overcomes this problem.

“The topological invariants in these states of matter are very complex functions of their quantum states, which makes a direct measurement in an experiment an impossible task,” explains Vermersch. “What we propose instead is to extract ‘many-body’ topological invariants (MBTIs) from a data set, which we obtain from several measurements.”

Random operations

The experimental recipe consists of subjecting a quantum state to a number of different random operations and studying how it reacts, Vermersch says. “By then using random matrix theory, we have proven that we can estimate MBTIs from this data set obtained from such randomized measurements.”

The specific feature of this technique is that although the topological invariants are highly complex, non-local correlation functions, they can still be extracted from statistical correlations of randomized measurements, he adds. Such random measurements are possible in synthetic quantum matter (or “quantum simulators”) made from tried and trusted experimental platforms such as cold atoms, trapped ions and superconducting quantum bits, to name but three. “Our protocol for measuring the topological invariants can therefore be directly studied in these existing systems in the laboratory,” says Vermersch.

Spurred on by their preliminary results, which they report in Science Advances, the researchers say that they would like to generalize their toolbox to be able to classify different types of topological phases beyond the ones studied in the present work. “Such phases include exotic ones, like those with ‘intrinsic topological order,’” Vermersch tells Physics World. “Doing this will of course be a great technological challenge.”

Vermersch and Zoller worked with colleagues Andreas Elben and Jinlong Yu to develop their measurement technique. They also collaborated closely with Guanyu Zhu and Mohammad Hafezi from the Joint Quantum Institute in Maryland and Frank Pollmann from the Technical University of Munich.

Solitons from new fibre laser could improve eye surgery

A fibre laser that emits high-energy, dispersion-free light pulses has been produced by researchers in Australia and the US. These soliton pulses are held together by a high-order term in the optical dispersion equation that had previously been a nuisance to scientists producing solitons. The researchers hope that their work will encourage the study of higher-order terms in the dispersion equation. Practical applications of high-power solitons include laser eye surgery.

Optical dispersion occurs when light at different frequencies (colours) travel at different speeds in a medium – causing a light pulse containing different frequencies to spread out in space and time. Familiar examples of dispersion are light being broken into different colours by a prism and the formation of rainbows. The equations governing dispersion, however, are complex. Adding second-order effects leads to non-dispersive solutions called solitons. In this case first and second order effects balance each other out and the pulse does not spread out.

Solitons in optical fibres were first observed in 1973 and have since have found applications in areas such as laser spectroscopy. However, their power is limited. In the early 1990s, researchers attributed this to higher-order terms in the optical dispersion equations, which become increasingly important as pulse power increases. This makes the pulses increasingly unstable and eventually causing them to break apart. Today, lasers that are used to produce high-power ultrashort pulses are complex and expensive. They produce pulses that chirp (change in frequency as the pulse progresses) and use additional equipment to reshape these chirped pulses into stable pulses.

All the heavy lifting

In 2018, Martijn de Sterke of the University of Sydney and colleagues found that fourth-order (quartic) dispersion is not necessarily destructive when they discovered a new type of optical soliton in a silicon waveguide. By “serendipity” de Sterke and colleagues found that, rather than being a perturbation “[the quartic dispersion] really did all the heavy lifting”: “The waveguide was not designed that way,” he explains, “it just happened to be that way, and our colleague Andrea Blanco-Redondo was clever enough to recognize that.”

In this latest work, de Sterke, Blanco-Redondo (now at Nokia Bell Labs in the US), Antoine Runge and Kevin Tam demonstrate the value of that serendipitous 2018 observation using an ultrafast fibre-laser that emits “pure quartic solitons”. The laser cavity incorporates a spectral pulse shaper that engineers the second- and third-order dispersion to be zero, while imparting a strong quartic dispersion. The resulting pulses agree well with the researchers’ theoretical predictions.

Crucially, Runge notes that much higher-power soliton lasers than previously possible may be achievable. “In a normal, quadratic soliton, if you halve the pulse duration, the energy goes up by a factor of two,” explains Runge, “In these new pure quartic solitons, if you halve the pulse duration, the energy goes up by a factor of eight.” This could allow them to be used in new applications: “A 100 fs quartic soliton could contain nanojoule energies with peak powers in the tens or hundreds of kilowatts,” he says. “For laser surgery and non-linear imaging that’s probably enough.”

Simplification needed

He says, however, that their device needs simplification to eliminate the pulse shaper before it could be commercially viable as a soliton laser: “At present, we don’t have fibres with the right dispersion,” he says, “but in future, if you managed to fabricate fibre like that – and we already have designs – then you wouldn’t need to do phase modulation or anything like that, and you would have a laser that was much simpler. What we did here was a proof of principle that these pulses exist and have these properties, but we’re still far away from a commercial pure quartic soliton laser.”

Runge hopes the work will encourage researchers to explore the potential of higher-order optical dispersion terms: “Now we have potentially an infinite family of pulses we can generate because we can tailor the dispersion the way we want and see if there’s something interesting for such or such application.”

Ultrafast laser spectroscopist Georg Herink of the University of Bayreuth in Germany is cautiously optimistic: “You definitely would get more energy into a soliton, and it definitely makes it interesting to look at even higher order terms,” he says. He cautions, however, that current de-chirping schemes produce orders of magnitude more energetic and shorter pulses than does the Sydney technique.

Herink suspects, however, that tuning the dispersion inside an optical cavity and producing multiple stable solutions of the optical dispersion equation has significant research potential: “Can you do something that hasn’t been observed before? Now you can explore a lot of academic questions that are maybe uniquely accessible with this platform experimentally.”

The research is described in Nature Photonics.

So you think you know your physics and music trivia?

1 At which institution in London did Queen guitarist Brian May study physics? A. King’s College B. University College C. Imperial College D. Queen Mary

2 What instrument is Albert Einstein best known for playing? A. Piano B. Trumpet C. Cello D. Violin

3 Which of these was not a song by German electronic band Kraftwerk? A. Electricity B. Radioactivity C. Uranium D. Transistor

4 US physicist and polymath Douglas Hofstadter once wrote a book that in part examined the work of which composer? A. Bach B. Beethoven C. Berlioz D. Brahms

5 German pop star Lena Meyer-Landrut won the 2010 Eurovision Song Contest with a song that likens her relationship to her lover to which piece of hardware? A. Satellite B. Laser C. Fusion reactor D. Magnet

6 In 2017 hip-hop star, philanthropist and entrepreneur will.i.am met physicists on a tour of which UK university? A. Aberdeen B. Birmingham C. Cardiff D. Durham

7 Which 1970s rock band sang the lyric “I’m a Spaceship Superstar… got a solar-powered laser beam guitar”? A. Slade B. Prism C. Sweet D. Wizzard

8 Which of these was not a song by British rock band Muse? A. Supermassive Black Holes B. Starlight C. Neutron Star Collision D. Quasar

9 Which of these terms is not used to describe the shape of a concert hall? A. Shoebox B. Vineyard. C. Shell D. Horseshoe

10 Who sang this appallingly stereotypical description of a physicist? “The lasers are in the lab/The old man is dressed in white clothes/Everybody says he’s mad/No one knows the things that he knows”? A. Joni Mitchell B. James Taylor C. Neil Young D.Linda Ronstadt

Stuck for the answers? The solutions are given below.

Offering readers a basic understanding of sound, musical instruments and music equipment, Musical Sound, Instruments and Equipment by Panos Photinos from Southern Oregon University is an ideal book for anyone wanting an introduction to the fundamental properties of sound waves and acoustics. Perfect reading for those budding musicians and sound engineers, it’s free to download until 6 June 2020.

Solutions 1 C 2 D 3 A 4 A 5 A 6 D 7 B 8 D 9 C 10 C

Fluids flow faster in liquid-walled channels

Blood may be thicker than water, but in a narrow enough tube, both liquids flow like treacle. This sluggish behaviour arises because, as you reduce the size of the channel, friction between the liquid and the channel wall comes to dominate the flow dynamics. Researchers have tried various ways of engineering surfaces to reduce this effect, but now a team from France, Switzerland and Ireland has gone a step further by doing away with solid channel walls altogether. Instead, the researchers confine their fluid within a conduit that itself consists of a liquid.

Writing in Nature, Peter Dunne (University of Strasbourg), Takuji Adachi (University of Strasbourg and University of Geneva) and colleagues describe how they form the walls of their conduit using ferrofluid – a colloidal suspension of magnetic nanoparticles in oil.

The researchers outlined the shape of their desired channel using long, rod-shaped neodymium magnets held in a 3D-printed framework. With four such magnets arranged with alternating polarities around the channel, they created a quadrupolar magnetic field whose strength fell to zero at the centre. This meant that the ferrofluid stuck to the inner edges of the framework, while the water that they channelled was confined to a narrow stream at the conduit’s centre.

By varying the properties of the ferrofluid and the distance between the magnets, the researchers created water “antitubes” approximately 14 µm across. They calculated that the right combination of parameters could yield antitubes thinner than 1 µm, though this was beyond the detection limit of the team’s equipment.

Glycerol flows freely through a fluid-walled channel (right), while a solid tube of the same length and diameter becomes nearly completely clogged. (Courtesy: Lucas Giachetti)

The ease with which a fluid flows over a surface is described by the slip length. In fluids that interact with the channel wall, flow is fastest at the centre and decreases along an approximately parabolic trend towards the solid–liquid interface. The slip length defines the distance beyond this interface at which the flow velocity would fall to zero. A fluid that is held stationary at the channel edge has a slip length of zero, while longer slip lengths indicate more freely flowing fluids.

That, at least, is the case when the sides of the channel are solid and immovable. In the experiments reported by Dunne and Adachi, the magnetic fluid making up the conduit walls moves too, flowing along with the water near the centre of the channel, and recirculating back to the start along the outside edge, next to the magnets. This results in an effective slip length much longer than could be achieved if the conduit walls were static.

“What is amazing is that we get a slip length of millimetres, whereas normally this is measured in nanometres or micrometres,” says study co-author Thomas Hermans, of the University of Strasbourg.

Magnetic control

While this very-low-friction fluid conduit was achieved with a static arrangement of magnets, the researchers found that they could manipulate the flow by altering the magnetic field configuration. Bringing an external magnet near to the conduit, for example, severed the water antitube, stopping its flow and functioning as a magnetic valve. When the external magnet was removed, the antitube repaired itself and the flow of water resumed.

The team used a similar effect to create a “magnetostaltic” pump, which they call the Qpump. In this device, concentric rings of magnets define a circular conduit, except that the water antitube is periodically disrupted by magnetic polarity reversals along the length of the inner ring. When the inner ring rotates, these localized blockages move along the channel’s circumference, driving water as they go.

Damage-free pumping — Platelet-rich plasma is transparent after pumping with the Qpump, but red after peristaltic pumping, due to increased rupture of red blood cells. (Courtesy: *Nature* 10.1038/s41586-020-2254-4)

When they tested the Qpump with blood instead of water in the central channel, Dunne and colleagues found it to be a much gentler way to drive fluid flow than conventional peristaltic pumps. Peristaltic pumps work by squeezing the walls of the conduit, but this induces shear forces that can rupture blood cells. The Qpump avoids such forces, meaning it could be used to drive a heart–lung machine, for example, without damaging the blood in the process. First, though, the team intends to prove the principle by applying it outside of the clinic.

“Our start-up company, Qfluidics, just entered the Merck global accelerator programme, where we will see if Qpumps can be used to pump delicate biologicals other than blood,” says Hermans. “The regulations for such uses are very much reduced as compared to medical applications, so this will be the first real application we will pursue.”

Artificial eye has the potential to outperform human vision

An artificial device that closely mimics the structure and function of the human eye has been unveiled by Leilei Gu and colleagues at The Hong Kong University of Science and Technology. The team based its design around a hemispherical arrangement of light-sensitive nanowires, which imitate photoreceptor cells in the human retina. Their device has the potential to produce images at higher resolutions than the human eye and could lead to significant new advances in robotics.

Our eyes provide us with around 80% of the information we perceive from our environment. They give us a field of view as wide as 160°; quickly adapt to different lighting and visibility conditions; and resolve details as small as 30 cm from 1 km away. This is possible thanks to our retinas – concave hemispheres that contain 10 million photoreceptor cells in every square centimetre.

While some other parts of the body have been successfully mimicked by technology, the eye has proven far more difficult to copy. At the heart of the problem are the flat configurations of today’s most advanced image sensors, which make them all but impossible to integrate into hemispherical structures to create artificial retinas.

Light-sensitive electrodes

Gu’s team circumvented this issue by mimicking the eye’s photoreceptor cells using perovskite nanowires, which served as light-sensitive electrodes. The wires were deposited onto the inside surface of a porous, hemispherical shell of aluminium oxide, and attached to liquid-metal wires which simulated the nerve fibres behind the retina.

An opposite-facing, tungsten-coated aluminium hemisphere then acted as the countering electrode to the nanowires; while the space between the hemispheres was filled with an ionic fluid electrolyte, comparable to the vitreous fluid found in human eyes. Finally, a lens and adjustable aperture were placed over a hole in the second hemisphere.

Gu and colleagues have built a prototype of their design with a nanowire arrangement that creates 100 pixels – enough to reconstruct images of different letters projected onto the device’s lens. Although this is far from the capabilities of the human eye, the researchers calculate that the nanowires they use could be packed in densely enough to allow for images with resolutions ten times higher than those that our eyes produce.

Applications for a fully functional artificial eye could include human-like robots that carry out autonomous tasks and excel at interacting with humans. In addition, the design could inspire the development of prosthetic eyes that entirely reconstruct the sight of visually impaired people. Gu’s team now aims to realize the full potential of their device, and hope to produce biomimetic eyes that outcompete our own eyes in the near future.

The artificial eye is described in Nature.

Saving The Scream

Edvard Munch’s painting The Scream (ca. 1910) is often described as the ultimate expression of the anxiety-ridden existence of modern man, but after more than 110 years, this evocative artwork is showing its age. The degradation is especially severe in areas where Munch used cadmium-sulphide (CdS)-based pigments, and the painting has become so delicate that it is rarely exhibited, remaining instead in a protected storage area in the Munch Museum in Oslo, Norway. An international team led by researchers at the National Research Council (CNR) in Italy have now used a combination of in-situ non-invasive spectroscopic and synchrotron X-ray techniques to show that moisture is the main cause of the degradation. According to the Munch Museum’s Irina Crina Anca Sandu, the team’s work could help conservation experts develop new conservation strategies to better preserve this and other works of art.

Munch created several versions of his masterpiece: two paintings, two pastels, a series of lithographic prints and several drawings and sketches. The most familiar are two paintings, created in 1893 and around 1910, which belong to the National Gallery and the Munch Museum, respectively.

The Scream is considered Munch’s most central work of art, and its impact comes from his intensive use of rhythmic wavy lines and contrasting straight bands typical of the Art Nouveau period. In The Masterworks of Edvard Munch (Museum of Modern Art, 1979), Munch is quoted as saying: “I walked one evening on a road—on the one side was the town and the fjord below me. I was tired and ill—I stood looking out across the fjord—the sun was setting—the clouds were coloured red—like blood—I felt as though a scream went through nature—I thought I heard a scream—I painted this picture—painted the clouds like real blood. The colours were screaming.”

Some screaming colours have chemically transformed

To make the “screaming” colours, Munch experimented with combinations of diverse binding media (tempera, oil and pastel) and brilliant and bold synthetic pigments such as zinc white, Prussian blue, synthetic ultramarine blue, chrome yellow and green, and cadmium orange and yellow. He did not know, however, that these novel materials would chemically transform over time, altering in colour or becoming structurally damaged. Today, some yellow areas of the sunset cloudy sky, as well as the neck area of the central figure in the ca. 1910 painting, show clear signs of degradation: the cadmium yellow brushstrokes have become off-white, and the lake water, which Munch thickly painted with opaque cadmium yellow, is flaking.

Earlier studies that applied scanning electron microscopy-energy dispersive X-ray and Fourier transform infrared (FTIR) techniques to microsamples of The Scream revealed that cadmium carbonate makes up most of the paler yellow tones of the sky and the main subject’s neck. These studies also showed that the cadmium carbonate had been mixed with varying amounts of sulphur, chlorine and sodium compounds in the lake region of the painting.

These observations, however, left the CNR-led team with several unanswered questions. Was the extent of the degradation in the CdS-based paint surface linked to its chemical composition? Into which compounds had the cadmium yellow compounds degraded? And finally, what caused these paints to deteriorate?

Non-invasive spectroscopy and synchrotron radiation X-ray techniques

To answer these questions, the researchers studied selected CdS-based areas of the painting using a series of spectroscopic analyses through the European MOLAB platform – a network of facilities from Italy, France, Poland, Greece and Germany that provides portable equipment for in-situ non-invasive measurements on artworks. They combined these analyses with the study of micron-sized samples from the painting that they obtained by scraping off an area from a spot of the flaking yellow surface of the lake region. They analysed these minute samples using micro X-ray diffraction, micro- X-ray fluorescence and micro X-ray absorption near-edge structure spectroscopy, mainly at the ID21 beamline at the ESRF (the European Synchrotron) in Grenoble, France. “This beamline is one of the few in the world where we can perform imaging X-ray absorption and fluorescence spectroscopy analysis of the entire sample at low energy and with sub-micrometre spatial resolution,” explains team member Koen Janssens of the University of Antwerp.

Annalisa Chieli, Letizia Monico and Gert Nuys — Three members of the team during measurements of cadmium yellow microflakes of The Scream (ca. 1910) at the beamline ID21 of the European Synchrotron. Courtesy: ESRF

The team compared their results to those obtained on artificially-aged oil paint mock-ups that had a similar composition to the lake material. They prepared the latter using an early 20th century cadmium yellow pigment powder and a cadmium yellow oil paint (labelled as Jaune de cadmium citron) that once belonged to Munch himself. They also obtained another set of oil paint mock-ups by mixing powders of cadmium sulphide with equal amounts of sodium sulphate and cadmium chloride, explains study lead author Letizia Monico of the CNR.

To artificially age the samples, the researchers exposed them first to UVA-visible light and a relative humidity (RH) of 45% and then a RH of more than 95% at 40°C for up to 100 days in the absence of light. “The goal of these experiments was to extrapolate the causes that can lead to deterioration,” Monico says.

Moisture is the culprit

The results of these experiments, which are detailed in Science Advances, reveal that the original CdS transforms into cadmium sulphate (CaSO₄) in the presence of chloride-containing compounds in high moisture conditions (a RH of 95% and above). This occurs even in the absence of light. The results also show that exposure to moisture causes (Cd,Cl) species to migrate through the paint along with the oxidation of the original CdS to CdSO₄. This phenomenon does not occur on Cl-free oil paint mock-ups aged under similar conditions.

To mitigate further degradation of the cadmium yellow pigment in The Scream (ca. 1910), Monico says the painting shouldn’t be exposed to moisture levels higher than 45% RH, while lighting conditions should be kept at “normal values for lightfast painting materials”. Currently, the Munch Museum stores and exhibits paintings at a RH of about 50% and a temperature of around 20°C.

Since Munch’s contemporaries, including Henri Matisse and Vincent van Gogh, also used cadmium-sulphide-based yellows, the findings could aid the development of preservation strategies for works by these artists too, explains MOLAB coordinator Costanza Miliani.

“This kind of work shows that art and science are intrinsically linked and that science can help preserve pieces of art so that the world can continue admiring them for years to come,” she states.

Exploring the philosophical, historical and sociological dimensions of physics, researchers ponder their return to the lab as lockdowns lift

This episode features the philosopher of science and Physics World columnist Bob Crease in conversation with Matin Durrani – who has edited Crease’s Critical Point column since its inception 20 years ago.

Crease explains how he finds inspiration for his columns from both his personal experiences and by interacting with Physics World readers. He also talks about the highs and lows of writing the column, including an embarrassing early-morning appearance on BBC radio.

Thankfully, it looks like some countries are at the point of relaxing COVID-19 restrictions and allowing some people to go back to work. But how will experimental physicists go back to their labs, and what will working conditions be like with some social distancing rules still in place?

Physics World’s Margaret Harris is on hand to chat about the challenges of returning to the lab and explains how one group in France is managing the transition.

Two members of the French team have written a Physics in the Pandemic blog for Physics World and Harris also talks about how this series of articles has chronicled how physicists around the globe have adapted to the pandemic.

This podcast is sponsored by Teledyne Hastings Instruments.

Browse articles by content type