The principle of cross-correlation. By linking electron spins, charges and orbitals using a strongly correlated electron system, it is possible to achieve cross-correlation, an unusual form of responses of electricity, magnetism, light and heat.
Yoshinori Tokura, Riken Japan, has a vision for a revolution, which he calls ‘Innovation 4’. He believes that four key technological breakthroughs could once again change society as we know it: an increase in solar cell conversion efficiency to 40% or more, an increase in the thermoelectric conversion figure of merit to 4 or more, an increase in the critical temperature of superconductivity to 400 K or well above room temperature, and an in increase in battery energy density to 400 watt-hours per kilogram or more. “These numerical targets represent a tripling of existing performance indexes. Another goal is to achieve electronic information processing with minimal power consumption to conserve energy. If realized, ‘Innovation 4’ will lead to a sustainable society revolution, but it is difficult to achieve these breakthroughs merely by improving existing technologies.
Tokura and his colleagues have been researching electronic technologies based on principles that are totally different from the mainstream semiconductor electronics of today. “It is assumed that electrons are sparse in conventional semiconductor devices, so the entanglement of electrons is weak. A group of many densely packed electrons, however, interact strongly with each other in what is known as ‘a strongly correlated electron system’. In such a system, non-charge properties that are not important in semiconductors, such as electron spin and orbital, also play important roles. We are seeking to create new functions that are not possible using independent electrons alone by utilizing the features of strongly correlated electron systems. High-temperature superconductivity is another phenomenon that occurs in strongly correlated electron systems. Electronic engineering still has infinite potential.
Ferroelectrics and ferromagnetics.
Ferroelectrics and ferromagnetics permit the orientations of electrical polarization and magnetization to be reversed by applying an electric field and magnetic field, respectively.
“The electron state in strongly correlated electron systems can be described as a solid produced by electrons. The electron state is like a dilute gas in semiconductors or a liquid in metals. Just as a liquid flows when the container is inclined, electricity flows when a voltage is applied to a metal. In a strongly correlated electron system, where the electron state is ‘solid’, electrons are unable to move because of mutual electrical repellence due to their dense packing. Even when a voltage is applied, no electricity flows. Hence, a strongly correlated electron system is an insulator, or specifically, a Mott insulator. When a minor stimulus such as heat, light or an electric field is applied from outside, a phase change from solid to liquid occurs instantaneously, allowing the electrons to move. In strongly correlated electron systems, this state can be changed at ultra-high speed on a nanometer scale.”
Low-energy information processing
Riken wants to use cross correlation, the unusual inversion of magnetization using an electric field, rather than the inversion of magnetization using a magnetic field to achieve lower power usage for computing. If it could be achieved, it will be possible to record information without wasting energy and with minimal power consumption. They are working on using multiferroics to achieve this.
By linking the electrical polarization and magnetization using multiferroics exhibiting both ferroelectric and ferromagnetic properties, it is possible to reverse the orientations of electrical polarization and magnetization using a magnetic field and electric field, respectively.
Multiferroics exhibit both ferroelectricity and ferromagnetism. A ferroelectric exhibits polarization, with one end positively charged and the other end negatively charged, even in the absence of an external electric field. When an electric field is applied to a ferroelectric, the two poles (+ and –) reverse themselves, allowing information to be rewritten. This phenomenon is used in some prepaid ‘e-money’ card systems for transport and shopping. A ferromagnetic, on the other hand, exhibits magnetization in the absence of a magnetic field, and the orientation of magnetization can be reversed by applying a magnetic field. Ferromagnetics are utilized in hard disks and other data recording devices. “In multiferroics, it is possible to realize the unusual response of reversing magnetization and simultaneously reversing the electrical polarization using an electric field by linking the orientations of electrical polarization and magnetization.”
Polarization is caused by a bias in the distribution of electrons in a material, whereas magnetization occurs when electron spins line up, which otherwise can have either upward or downward orientations, become aligned in a given orientation. Electron spin thus serves as the origin of magnetization.
In 2009, Tokura and his colleagues succeeded in experimentally changing the orientation of magnetization at temperatures below –271 °C using an electric field. “If we can improve on this and simultaneously reverse the orientations of magnetization and polarization at room temperature using an electric field, then we will be able to create large-capacity memory that consumes almost no electrical power.”
More recently in June 2010, Tokura’s research group became the first in the world to directly observe skyrmion crystallization, the phenomenon by which electron spin vortices are regularly arranged like a crystal. The result attracted worldwide attention.
Simulated spin structure of skyrmion crystal (left). A skyrmion crystal observed in Fe0.5Co0.5Si by Lorentz electron microscopy (right). The skyrmion state, in which electron spins are arranged in vertexes, is estimated to be movable at currents much lower than conventional levels. Skyrmions are expected to be useful for information processing with low power consumption.
“It is thought that these electron spin vortices can be moved with a small amount of electric current. Hence, by merely changing the orientation of electron spins one after another, it is possible to move the electron spin vertexes to achieve information processing. This has potential for information processing with minimal power consumption.”
New principles for highly efficient solar cells
“We also have an idea for dramatically improving the power efficiency of solar cells,” says Tokura. In conventional solar cells, a medium such as a semiconductor absorbs photons and generates a free pair of negative and positive charges. By separating the negative electron and positive ‘hole’ and transporting them to opposite electrodes, a voltage can be produced. The light-to-electricity conversion efficiencies of modern solar cells is just over 10%, but it should be possible to improve on this efficiency. “Solar radiation contains a broad range of wavelengths. Semiconductor solar cells actually achieve nearly 100% conversion efficiency at particular wavelengths of light, because electron–hole pair can be produced from a single photon at a particular wavelength in each semiconductor with nearly 100% probability. However, an electron–hole pair is also produced when a photon with a shorter wavelength and higher energy level is absorbed, in which case the excess energy is wasted as heat. This accounts for the low power efficiency. If we use a strongly correlated electron system, the wasted energy could be used to create a metallic state and produce a large number of electrons and holes by another mechanism, which could dramatically improve conversion efficiency (Fig. 6). Strongly correlated electron systems are being actively studied worldwide, but Tokura and his colleagues are the only group researching their use in highly efficient solar cells.
Principle of high-efficiency solar cells using a strongly correlated electron system.
When absorbed by a strongly correlated electron system (a ‘solid’ electron state), high-energy light ‘melts’ the solid to generate free electrons and holes, which could be separated and transported to opposite electrical to produce an electrical current.
Thermoelectrics - The thermoelectric effect is the direct conversion of temperature differences to electric voltage and vice versa. as the dimensionless figure of merit ZT is used as a shorthand way of comparing the quality of thermoelectric materials. Values of ZT=1 are considered good, and values of at least the 3–4 range are considered to be essential for thermoelectrics to compete with mechanical generation and refrigeration in efficiency. To date, the best reported ZT values have been in the 2–3 range
An April, 2010 interview with Tokura at Science Watch
1 Ultra-low energy consuming electronics
(nearly dissipation-less, non-volatility, ultra-high density)
From heat to electricity ZT > 4
2 Energy conversion
From light to electicity F > 0.4
3 Energy transfer
Above room temperature superconductivity > 400 Kelvin
4 Energy storage
Quantum battery > 400 WH/Kg
In energy transfer, we need 400 Kelvin in order to realize a real room temperature superconductor. At the moment we have 130 or 140 Kelvin superconductors. So we need not only three times the effort to achieve this, but also three times the innovation.
Of course, if you can use liquid nitrogen and then maybe you can make any power transmission line, but it's still very difficult. High-Tc is a main target of our research. We are still struggling and often it's not so successful, but maybe we will one day reach the 400 Kelvin superconductor.
Another case is thermoelectric materials. Our goal is to produce materials for ultra-low energy consuming electronics. The thermoelectric effect tells us we can generate electricity from temperature differences. An example is the air conditioner, which uses a compressor for refrigeration—that's 19th century thermodynamics.
But if we could directly convert electricity to heat conduction, that would be better. We usually measure this by the so-called thermal figure of merit (ZT). At the moment, ZT is typically 1 or a little more. But if this value exceeds 3 or 4, then every compressor can go away and we can immediately replace it with a direct heat-electricity transformation.
And another case, in terms of energy conversion, is solar cells. As you know, the efficiency is now at 10%, but for industrial use 40% would be very important. So I think correlated electron materials may help. I am not sure, but we are working towards that purpose, and maybe with these correlated electron materials we can generate a surprising result. We will need a very new physics. Yes, it's a dream.
In silicon, light pumps out an electron leaving a hole, positively charged which generates an electric current. But with the use of these new materials, a photon of light comes in then we have a sort of metallic state, and the semiconductor or insulator suddenly turns into a metal. Of course, we have to consider the energy conservation rule, but still a lot of the electrons can be generated and extracted, so we may realize a very highly efficient solar cell. This may be 10 years away.
And batteries, too. This is the problem of energy storage. It's another dream of mine. The best performance of the present state-of-the-art batteries is 100 watt-hours per kilogram. If you could increase this performance three or four times, it would make a great difference in our mobile computing society.
If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks