Electron movements form the basis of electronics as they enable storage, processing and transfer of information. State-of-the-art electronic circuits have reached maximum clock rates of several billion switching cycles per second, limited by the heat accumulated in the process of switching power on and off.

The electric field of light changes its direction a trillion times per second and is able to move electrons in solids at this speed. Thus light waves could form the basis for future electronic switching once the induced electron motion and its influence on heat accumulation is precisely understood.

Physicists from the Laboratory for Attosecond Physics have already found out that it is possible to manipulate the electronic properties of matter at optical frequencies. In a follow-up experiment the researchers, in a manner similar to their previous approach, shot extremely strong, femtosecond-laser pulses (one femtosecond is a millionth of a billionth of a second) onto silicon dioxide glass.

A single oscillation

The light pulse comprises only a single strong oscillation cycle of the field, hence the electrons are moved left and right only once. The full temporal characterization of the light field after transmission through the thin glass plate now, for the first time, provides direct insight into the attosecond electron dynamics, induced by the light pulse in the solid.

This measurement technique reveals that electrons react with a delay of only some ten attoseconds (one attosecond is a billionth of a billionth of a second) to the incoming light. This time delay in the reaction determines the energy transferred between light and matter.

Since it is now possible to measure this energy exchange within one light cycle, the parameters of the light-matter interaction can be understood and optimized to reach out for the ultimate speed in signal processing. The more reversible the exchange and the smaller the residual energy left in the medium after the light pulse has passed, the more suitable the interaction for future light field-driven electronics.

Cool relationship

To understand the observed phenomena and identify the best set of experimental parameters to that end, the experiments were backed up by a novel simulation method based on first principles developed at the Center for Computational Sciences at University of Tsukuba. The theorists there used the K computer, currently the fourth fastest supercomputer in the world, to compute electron movement inside solids with unprecedented accuracy.

Finally, the researchers succeeded in optimizing the energy consumption by adapting the amplitude of the light field. At certain field strengths energy is transferred from the field to the solid during the first half of the pulse cycle and is almost completely emitted back in the second half of the light.

These findings verify that a potential switching medium for future light-driven electronics need not overheat. Thus the ‘cool relationship’ between glass and light might provide an opportunity for dramatically accelerating electronic signal and data processing, up to the ultimate limit.

Hokkaido University scientists are testing the development of solar cells made of solid materials to improve their ability to function under harsh environmental conditions.

Gold nanoparticles harvest light and provide a visible light response to the cell. Credit: Tomoya OSHIKIRI

Scientists at Hokkaido University in Japan are making leeway in the fabrication of all-solid-state solar cells that are highly durable and can efficiently convert sunlight into energy. The team employed a method called “atomic layer deposition”, which allows scientists to control the deposit of very thin, uniform layers of materials on top of each other. Using this method, they deposited a thin film of nickel oxide onto a single crystal of titanium dioxide. Gold nanoparticles were introduced between the two layers to act like an antenna that harvests visible light.

The team tested the properties of these fabricated devices with and without an intermediary step following the deposition of nickel oxide that involves heating it to very high temperatures and then allowing it to slowly cool – a process called “annealing”.

Photocurrent generation was successfully observed on the all-solid-state photoelectric conversion device. The device was found to be highly durable and stable because, unlike some solar cells, it does not contain organic components, which have a tendency to degrade over time and under harsh conditions.

The researchers also found that annealing affected the properties of the device by changing the interfacial structure of the layers. For example, it increased the voltage available from the device but also increased the resistance within it. It also decreased the device’s efficiency in converting light to electricity. The results suggest that the structural changes caused by annealing prevent the layer of gold nanoparticles from injecting electrons into the titanium dioxide layer.

The team’s fabrication process is inexpensive and can be scaled up easily but the resultant device’s properties are still insufficient for practical use and its efficiency in converting light to energy needs to be improved. Further research is needed to understand the roles of each layer in conducting energy to improve the device’s efficiency.

New fundamental research by UT Dallas physicists may accelerate the drive toward more advanced electronics and more powerful computers.

The scientists are investigating materials called topological insulators, whose surface electrical properties are essentially the opposite of the properties inside.

“These materials are made of the same thing throughout, from the interior to the exterior,” said Dr. Fan Zhang, assistant professor of physics at UT Dallas. “But, the interior does not conduct electrons — it’s an insulator — while the electrons on the surface are free to move around. The surface is therefore a conductor, like a metal, but it is in fact more robust than a metal.”

There are two types of topological insulators: strong and weak. The difference between them is subtle and involves complex physics, but is critically important.

“If you had a cube of material that is a strong topological insulator, all six faces can conduct electrons,” Zhang said. “For the weak one, only the four sides are conducting, while the top and bottom surfaces remain insulating.”

Strong topological insulators were made experimentally shortly after they were theoretically proposed. Zhang said they are common in nature, and several dozen variations have been identified and experimentally confirmed.

On the other hand, weak topological insulators have been more elusive. Scientists have proposed various ways to construct a weak topological insulator, but because of its distinctive properties, researchers have not been able to say definitively that they have experimentally produced one.

Zhang, a theoretical physicist, has devised a new way to make a weak topological insulator, one that involves a relatively simple mix of two chemical elements: a crystal composed of bismuth combined with either iodine or bromine. He and his colleagues published the research recently in the journal Physical Review Letters and presented their work at the March meeting of the American Physical Society.

In the 1970s, German scientists grew bismuth iodides and bismuth bromides, but they didn’t understand their potential as weak topological insulators, Zhang said.

“This class of materials we are proposing is a unique platform for exploring exotic physics with fairly simple chemistry,” he said. “With further research and experimentation, our findings could lead to significant advances in technology, especially in electronics and quantum computing.”

Electrically conductive materials are the fundamental building blocks of the traditional transistors that power electronic devices including cellphones and computers. Researchers are developing new theories and experiments with innovative physics and materials to create new transistor-like technologies that run devices and make computers more powerful.

With such exotic electrical properties, topological insulators offer a potential option, Zhang said.

“Our lives have been modified over time by our understanding of the conduction of electrons and the exploitation of this physics for use in electronic devices,” he said. “We now need to revolutionize transistors. One possible substitution is a so-called topological field effect transistor, which could be made of a thin film of a weak topological insulator.”

Computers also are heading for a fundamental redesign, and those efforts might be aided by Zhang’s research.

“The fundamental computing scale is now very limited,” he said. “For many applications, like weather forecasting and information encoding and decoding, today’s computers are way too slow. However, quantum computers have been proposed that would use the principles of quantum physics to compute exponentially faster than today’s computers.

“Weak topological insulators could make quantum computing feasible.”

As a theorist, Zhang used old-fashioned pencil and paper to construct the basis of his theory about the bismuth compounds. His postdoctoral researcher Dr. Cheng-Cheng Liu, the study’s lead author and now an assistant professor at Beijing Institute of Technology, then crunched specific numbers using high-speed supercomputers at the Texas Advanced Computing Center based at UT Austin.

Zhang’s UT Dallas colleague, Dr. Bing Lv, assistant professor of physics, has made samples of bismuth iodide.

“The next step will be to characterize the material to explore the unique properties that a weak topological insulator can offer to fundamental physics and to our everyday lives,” Zhang said.

A family of compounds known as perovskites, which can be made into thin films with many promising electronic and optical properties, has been a hot research topic in recent years. But although these materials could potentially be highly useful in applications such as solar cells, some limitations still hamper their efficiency and consistency.

Now, a team of researchers at MIT and elsewhere say they have made significant inroads toward understanding a process for improving perovskites’ performance, by modifying the material using intense light. The new findings are being reported in the journal Nature Communications, in a paper by Samuel Stranks, a researcher at MIT; Vladimir Bulovic, the Fariborz Maseeh (1990) Professor of Emerging Technology and associate dean for innovation; and eight colleagues at other institutions in the U.S. and the U.K. The work is part of a major research effort on perovskite materials being led by Stranks, within MIT’s Organic and Nanostructured Electronics Laboratory.

Tiny defects in perovskite’s crystalline structure can hamper the conversion of light into electricity in a solar cell, but “what we’re finding is that there are some defects that can be healed under light,” says Stranks, who is a Marie Curie Fellow jointly at MIT and Cambridge University in the U.K. The tiny defects, called traps, can cause electrons to recombine with atoms before the electrons can reach a place in the crystal where their motion can be harnessed.

Under intense illumination, the researchers found that iodide ions — atoms stripped of an electron so they carry an electric charge — migrated away from the illuminated region, and in the process apparently swept away most of the defects in that region along with them.

“This is the first time this has been shown,” Stranks says, “where just under illumination, where no [electric or magnetic] field has been applied, we see this ion migration that helps to clean the film. It reduces the defect density.” While the effect had been observed before, this work is the first to show that the improvement was caused by the ions moving as a result of the illumination.

This work is focused on particular types of the material, known as organic-inorganic metal halide perovskites, which are considered promising for applications including solar cells, light-emitting diodes (LEDs), lasers, and light detectors. They excel in a property called the photoluminescence quantum efficiency, which is key to maximizing the efficiency of solar cells. But in practice, the performance of different batches of these materials, or even different spots on the same film, has been highly variable and unpredictable. The new work was aimed at figuring out what caused these discrepancies and how to reduce or eliminate them.

Stranks explains that “the ultimate aim is to make defect-free films,” and the resulting improvements in efficiency could also be useful for applications in light emission as well as light capture.

Previous work reducing defects in thin-film perovskite materials has focused on electrical or chemical treatments, but “we find we can do the same with light,” Stranks says. One advantage of that is that the same technique used to improve the material’s properties can at the same time be used as a sensitive probe to observe and better understand the behavior of these promising materials.

Another advantage of this light-based processing is it doesn’t require anything to come in physical contact with the film being treated — for example, there is no need to attach electrical contacts or to bathe the material in a chemical solution. Instead, the treatment can simply be applied by turning on the source of illumination. The process, which they call photo-induced cleaning, could be “a way forward” for the development of useful perovskite-based devices, Stranks says.

The effects of the illumination tend to diminish over time, Stranks says, so “the challenge now is to maintain the effect” long enough to make it practical. Some forms of perovskites are already “looking to be commercialized by next year,” he says, and this research “raises questions that need to be addressed, but it also shows there are ways to address” the phenomena that have been limiting this material’s performance.

In 2010 the Nobel Prize in physics was awarded for the discovery of the exceptional material graphene, which consists of a single layer of carbon atoms arranged in a honeycomb lattice. But graphene research did not stop there. New interesting properties of this material are still being found. An international team of researchers has now explained the peculiar behaviour of electrons moving through narrow constrictions in a graphene layer. The results have been published in the journal Nature Communications.

Electron wave passing through a narrow constriction. Credit: TU Wien

The electron is a wave

“When electrical current flows through graphene, we should not imagine the electrons as little balls rolling through the material,” says Florian Libisch from TU Wien (Vienna), who led the theoretical part of the research project. The electrons swash through the graphene as a long wave front, the wavelength can be a hundred times larger than the space between two adjacent carbon atoms. “The electron is not confined to one particular carbon atom, in some sense it is located everywhere at the same time,” says Libisch.

The team studied the behaviour of electrons squeezing through a narrow constriction in a graphene sheet. “The wider the constriction, the larger the electron flux – but as it turns out, the relationship between the width of the constriction, the energy of the electrons and the electric current is quite complex,” says Florian Libisch. “When we make the constriction wider, the electric current does not increase gradually, it jumps at certain points. This is a clear indication of quantum effects.”

If the wavelength of the electron is so large that it does not fit through the constriction, the electron flux is very low. “When the energy of the electron is increased, its wavelength decreases,” explains Libisch. “At some point, one wavelength fits through the constriction, then two wavelengths, then three – this way the electron flux increases in characteristic steps.” The electric current is not a continuous quantity, it is quantized.

Theory and experiment

This effect can also be observed in other materials. Detecting it in graphene was much more difficult, because its complex electronic properties lead to a multitude of additional effects, interfering with each other. The experiments were performed at the group of Christoph Stampfer at the RWTH Aachen (Germany), theoretical calculations and computer simulations were performed in Vienna by Larisa Chizhova and Florian Libisch at the group of Joachim Burgdörfer.

For the experiments, the graphene sheets hat to be etched into shape with nanometre precision. “Protecting the graphene layer by sandwiching it between atomic layers of hexagonal boron nitride is critical for demonstrating the quantized nature of current in graphene,” explains Christoph Stampfer. Current through the devices is then measured at extremely low temperatures. “We use liquid helium to cool our samples, otherwise the fragile quantum effects are washed out by thermal fluctuations,” says Stampfer. Simulating the experiment poses just as much of a challenge. “A freely moving electron in the graphene sheet can occupy as many quantum states as there are carbon atoms,” says Florian Libisch, “more than ten million, in our case.” This makes the calculations extremely demanding. An electron in a hydrogen atom can be described using just a few quantum states. The team at TU Wien (Vienna) developed a large scale computer simulation and calculated the behaviour of the electrons in graphene on the Vienna Scientific Cluster VSC, using hundreds of processor cores in parallel.

Edge states

As it turns out, the edge of the graphene sheet plays a crucial role. “As the atoms are arranged in a hexagonal pattern, the edge can never be a completely straight line. On an atomic scale, the edge is always jagged,” says Florian Libisch. In these regions, the electrons can occupy special edge states, which have an important influence on the electronic properties of the material. “Only with large scale computer simulations using the most powerful scientific computer clusters available today, we can find out how these edge states affect the electrical current,” says Libisch. “The excellent agreement between the experimental results and our theoretical calculations shows that we have been very successful.”

The discovery of graphene opened the door to a new research area: ultrathin materials which only consist of very few atomic layers are attracting a lot of attention. Especially the combination of graphene and other materials – such as boron nitride, as in this case – is expected to yield interesting results. “One thing is for sure: whoever wants to understand tomorrow’s electronics has to know a lot about quantum physics,” says Florian Libisch.

Scientists have developed a new type of graphene-based transistor and using modelling they have demonstrated that it has ultralow power consumption compared with other similar transistor devices. The findings have been published in a paper in the journal Scientific Reports. The most important effect of reducing power consumption is that it enables the clock speed of processors to be increased. According to calculations, the increase could be as high as two orders of magnitude.

(A) Electron spectrum E(p) in bilayer graphene (left) and energy dependence of its density of states, DoS (right). At energy levels corresponding to the edge of the “Mexican hat” the DoS tends to infinity. (B) The red areas show the states of electrons that participate in tunneling in bilayer graphene (left) and in a conventional semiconductor with “ordinary” parabolic bands (right). Electrons that are capable of tunneling at low voltages are found in the ring in graphene, but in the semiconductor they are only found at the single point. The dotted lines indicate the tunneling transitions. The red lines indicate the trajectories of the tunneling electrons in the valence band.

“The point is not so much about saving electricity – we have plenty of electrical energy. At a lower power, electronic components heat up less, and that means that they are able to operate at a higher clock speed – not one gigahertz, but ten for example, or even one hundred,” says the corresponding author of the study, the head of MIPT’s Laboratory of Optoelectronics and Two-Dimensional Materials, Dmitry Svintsov.

Building transistors that are capable of switching at low voltages (less than 0.5 volts) is one of the greatest challenges of modern electronics. Tunnel transistors are the most promising candidates to solve this problem. Unlike in conventional transistors, where electrons “jump” through the energy barrier, in tunnel transistors the electrons “filter” through the barrier due to the quantum tunneling effect. However, in most semiconductors the tunneling current is very small and this prevents transistors that are based on these materials from being used in real circuits.

The authors of the article, scientists from the Moscow Institute of Physics and Technology (MIPT), the Institute of Physics and Technology RAS, and Tohoku University (Japan), proposed a new design for a tunnel transistor based on bilayer graphene, and using modelling, they proved that this material is an ideal platform for low-voltage electronics.

Graphene, which was created by MIPT alumni Sir Andre Geim and Sir Konstantin Novoselov, is a sheet of carbon that is one atom thick. As it has only two dimensions, the properties of graphene, including its electronic properties, are radically different to three-dimensional carbon – graphite.

“Bilayer graphene is two sheets of graphene that are attached to one another with ordinary covalent bonds. It is as easy to make as monolayer graphene, but due to the unique structure of its electronic bands, it is a highly promising material for low-voltage tunneling switches,” says Svintsov.

Bands of bilayer graphene, i.e. the allowed energy levels of an electron at a given value of momentum, are in the shape of a “Mexican hat” (fig. 1A, compare this to the bands of most semiconductors which form a parabolic shape). It turns out that the density of electrons that can occupy spaces close to the edges of the “Mexican hat” tends to infinity – this is called a van Hove singularity. With the application of even a very small voltage to the gate of a transistor, a huge number of electrons at the edges of the “Mexican hat” begin to tunnel at the same time. This causes a sharp change in current from the application of a small voltage, and this low voltage is the reason for the record low power consumption.

In their paper, the researchers point out that until recently, van Hove singularity was barely noticeable in bilayer graphene – the edges of the “Mexican hat” were indistinct due to the low quality of the samples. Modern graphene samples on hexagonal boron nitride (hBN) substrates are of much better quality, and pronounced van Hove singularities have been experimentally confirmed in the samples using scanning probe microscopy and infrared absorption spectroscopy.

An important feature of the proposed transistor is the use of “electrical doping” (the field effect) to create a tunneling p-n junction. The complex process of chemical doping, which is required when building transistors on three-dimensional semiconductors, is not needed (and can even be damaging) for bilayer graphene. In electrical doping, additional electrons (or holes) occur in graphene due to the attraction towards closely positioned doping gates.

Under optimum conditions, a graphene transistor can change the current in a circuit ten thousand times with a gate voltage swing of only 150 millivolts.

“This means that the transistor requires less energy for switching, chips will require less energy, less heat will be generated, less powerful cooling systems will be needed, and clock speeds can be increased without the worry that the excess heat will destroy the chip,” says Svintsov.