A team of researchers at MIT produced strongly coupled superconductivity in a tri-layer graphene structure, creating a path for exploring a new family of superconducting materials with potential applications in quantum information science. Another group of MIT researchers fabricated a silicon chip with integrated LEDs, evoking visions of a future in which LED sensors no longer need to be manufactured separately from silicon processors.

 Physicists Create Tunable Superconductivity in Twisted Graphene ‘Nanosandwich’

This artist’s rendition shows magic-angle twisted tri-layer graphene, composed of three honeycomb lattices. The tightly bound electrons (yellow spheres connected by blue halos) indicate the new structure’s strongly coupled superconducting state. (Image courtesy Ella Maru Studio)

Structure may reveal conditions needed for high-temperature superconductivity.

By Jennifer Chu | MIT News

Publication Date: February 1, 2021

When two sheets of graphene are stacked atop each other at just the right angle, the layered structure morphs into an unconventional superconductor, allowing electric currents to pass through without resistance or wasted energy.

This “magic-angle” transformation in bilayer graphene was observed for the first time in 2018 in the group of Pablo Jarillo-Herrero, the Cecil and Ida Green Professor of Physics at MIT. Since then, scientists have searched for other materials that can be similarly twisted into superconductivity, in the emerging field of “twistronics.” For the most part, no other twisted material has exhibited superconductivity other than the original twisted bilayer graphene, until now.

In a paper appearing [February 1, 2021], in Nature, Jarillo-Herrero and his group report observing superconductivity in a sandwich of three graphene sheets, the middle layer of which is twisted at a new angle with respect to the outer layers. This new tri-layer configuration exhibits superconductivity that is more robust than its bi-layer counterpart.

The researchers can also tune the structure’s superconductivity by applying and varying the strength of an external electric field. By tuning the tri-layer structure, the researchers were able to produce ultra-strongly coupled superconductivity, an exotic type of electrical behavior that has rarely been seen in any other material.

“It wasn’t clear if magic-angle bi-layer graphene was an exceptional thing, but now we know it’s not alone; it has a cousin in the tri-layer case,” Jarillo-Herrero said. “The discovery of this hypertunable superconductor extends the twistronics field into entirely new directions, with potential applications in quantum information and sensing technologies.”

His co-authors are lead author Jeong Min Park and Yuan Cao at MIT, and Kenji Watanabe and Takashi Taniguchi of the National Institute of Materials Science in Japan.

A New Super Family

Shortly after Jarillo-Herrero and his colleagues discovered that superconductivity could be generated in twisted bi-layer graphene, theorists proposed that the same phenomenon might be seen in three or more layers of graphene.

A sheet of graphene is an atom-thin layer of graphite, made entirely of carbon atoms arranged in a honeycomb lattice, like the thinnest, sturdiest chicken wire. The theorists proposed that if three sheets of graphene were stacked like a sandwich, with the middle layer rotated by 1.56 degrees with respect to the outer layers, the twisted configuration would create a kind of symmetry that would encourage electrons in the material to pair up and flow without resistance — the hallmark of superconductivity.

“We thought, why not, let’s give it a try and test this idea,” Jarillo-Herrero said.

Park and Cao engineered tri-layer graphene structures by carefully slicing a single gossamer sheet of graphene into three sections and stacking each section on top of each other at the precise angles predicted by the theorists.

They made several tri-layer structures, each measuring a few micrometers across (about 1/100 the diameter of a human hair), and three atoms tall.

“Our structure is a nanosandwich,” Jarillo-Herrero said.

The team then attached electrodes to either end of the structures and ran an electric current through, while measuring the amount of energy lost or dissipated in the material.

“We saw no energy dissipated, meaning it was a superconductor,” Jarillo-Herrero said. “We have to give credit to the theorists — they got the angle right.”

He added that the exact cause of the structure’s superconductivity — whether due to its symmetry, as the theorists proposed, or not — remains to be seen, and is something that the researchers plan to test in future experiments.

“For the moment, we have a correlation, not a causation,” he said. “Now at least we have a path to possibly explore a large family of new superconductors based on this symmetry idea.”

‘The Biggest Bang’

 In exploring their new tri-layer structure, the team found they could control its superconductivity in two ways. With their previous bi-layer design, the researchers could tune its superconductivity by applying an external gate voltage to change the number of electrons flowing through the material. As they dialed the gate voltage up and down, they measured the critical temperature at which the material stopped dissipating energy and became superconductive. In this way, the team was able to tune bi-layer graphene’s superconductivity on and off, similar to a transistor.

The team used the same method to tune tri-layer graphene. They also discovered a second way to control the material’s superconductivity that has not been possible in bi-layer graphene and other twisted structures. By using an additional electrode, the researchers could apply an electric field to change the distribution of electrons between the structure’s three layers, without changing the structure’s overall electron density.

“These two independent knobs now give us a lot of information about the conditions where superconductivity appears, which can provide insight into the key physics critical to the formation of such an unusual superconducting state,” Park said.

Using both methods to tune the tri-layer structure, the team observed superconductivity under a range of conditions, including at a relatively high critical temperature of 3 kelvins, even when the material had a low density of electrons. In comparison, aluminum, which is being explored as a superconductor for quantum computing, has a much higher density of electrons and only becomes superconductive at about 1 kelvin.

“We found magic-angle tri-layer graphene can be the strongest coupled superconductor, meaning it superconducts at a relatively high temperature, given how few electrons it can have,” Jarillo-Herrero said. “It gives the biggest bang for your buck.”

“The work is a meaningful step up in structural complexity of a twistronic system that can be faithfully reproduced in several samples,” said David Goldhaber-Gordon, a professor of physics at Stanford University who was not involved in the study. “That structural complexity is not just pursued for its own sake but rather aims to make the effect of electronic interactions tunable. Applications of such sophisticated multi-layer structures will likely be in quantum information science where the exquisite control of electronic structure will be important.”

The researchers plan to fabricate twisted graphene structures with more than three layers to see whether such configurations, with higher electron densities, can exhibit superconductivity at higher temperatures, even approaching room temperature.

“Our main goal is to figure out the fundamental nature of what underlies strongly coupled superconductivity,” Park said. “Tri-layer graphene is not only the strongest-coupled superconductor ever found, but also the most tunable. With that tunability, we can really explore superconductivity, everywhere in the phase space.”

This research was supported, in part, by the Department of Energy, the National Science Foundation, the Gordon and Betty Moore Foundation, and the Ramon Areces Foundation.

 Reprinted with permission of MIT News (http://news.mit.edu/).


An LED That Can Be Integrated Directly into Computer Chips

MIT researchers have developed a bright, efficient silicon LED, pictured, that can be integrated directly onto computer chips. The advance could reduce cost and improve performance of microelectronics that use LEDs for sensing or communication. (Image courtesy of the researchers)

The advance could cut production costs and reduce the size of microelectronics for sensing and communication.

By Daniel Ackerman | MIT News

Publication Date: December 14, 2020

Light-emitting diodes — LEDs — can do way more than illuminate your living room. These light sources are useful microelectronics, too.

Smartphones, for example, can use an LED proximity sensor to determine if you’re holding the phone next to your face (in which case the screen turns off). The LED sends a pulse of light toward your face, and a timer in the phone measures how long it takes that light to reflect back to the phone, a proxy for how close the phone is to your face. LEDs are also handy for distance measurement in autofocus cameras and gesture recognition.

One problem with LEDs: It’s tough to make them from silicon. That means LED sensors must be manufactured separately from their device’s silicon-based processing chip, often at a hefty price. But that could one day change, thanks to new research from MIT’s Research Laboratory of Electronics (RLE).

Researchers have fabricated a silicon chip with fully integrated LEDs, bright enough to enable state-of-the-art sensor and communication technologies. The advance could lead to not only streamlined manufacturing, but also better performance for nanoscale electronics.

Jin Xue, a Ph.D. student in RLE, led the research, which was to be presented at the IEDM conference in December 2020. MIT co-authors included Professor Rajeev Ram, who leads the Physical Optics and Electronics Group in RLE, as well as Jaehwan Kim, Alexandra Mestre, Dodd Gray, Danielus Kramnik, and Amir Atabaki. Other co-authors included Kian Ming Tan, Daniel Chong, Sandipta Roy, H. Nong, Khee Yong Lim, and Elgin Quek, from the company GlobalFoundries.

Silicon is widely used in computer chips because it’s abundant, cheap, and a semiconductor, meaning it can alternately block and allow the flow of electrons. This capacity to switch between “off” and “on” underlies a computer’s ability to perform calculations. But despite silicon’s excellent electronic properties, it doesn’t quite shine when it comes to optical properties — silicon makes for a poor light source. So electrical engineers often turn away from the material when they need to connect LED technologies to a device’s computer chip.

The LED in your smartphone’s proximity sensor, for example, is made from III-V semiconductors, so called because they contain elements from the third and fifth columns of the periodic table. (Silicon is in the fourth column.) These semiconductors are more optically efficient than silicon — they produce more light from a given amount of energy. (You don’t see the light produced from the proximity sensor because it is infrared, not visible.)

These two images show the silicon LED switched on (left) and off. (Image courtesy of the researchers)

And while the proximity sensor is a fraction of the size of the phone’s silicon processor, it adds significantly to the phone’s overall cost. “There’s an entirely different fabrication process that’s needed, and it’s a separate factory that manufactures that one part,” said Ram. “So, the goal would be: Can you put all this together in one system?” Ram’s team did just that.

Xue designed a silicon-based LED with specially engineered junctions — the contacts between different zones of the diode — to enhance brightness. This boosted efficiency: The LED operates at low voltage, but it still produces enough light to transmit a signal through 5 meters of fiber optic cable. Plus, GlobalFoundries manufactured the LEDs right alongside other silicon microelectronic components, including transistors and photon detectors. While Xue’s LED didn’t quite outshine a traditional III-V semiconductor LED, it easily beat out prior attempts at silicon-based LEDs.

“Our optimization process of how to make a better silicon LED had quite an improvement over past reports,” said Xue. He added that the silicon LED could also switch on and off faster than expected. The team used the LED to send signals at frequencies up to 250 megahertz, indicating that the technology could potentially be used not only for sensing applications, but also for efficient data transmission. Xue’s team plans to continue developing the technology. But, he said, “it’s already great progress.”

Ram envisions a day when LED technology can be built right onto a device’s silicon processor — no separate factory needed. “This is designed in a standard microelectronics process,” he said. “It’s a really integrated solution.”

Ram envisions a day when LED technology can be built right onto a device’s silicon processor — no separate factory needed. “This is designed in a standard microelectronics process,” he said. “It’s a really integrated solution.”

In addition to cheaper manufacturing, the advance could also improve LED performance and efficiency as electronics shrink to ever smaller scales. That’s because, at a microscopic scale, III-V semiconductors have nonideal surfaces, riddled with “dangling bonds” that allow energy to be lost as heat rather than as light, according to Ram. In contrast, silicon forms a cleaner crystal surface. “We can take advantage of those very clean surfaces,” said Ram. “It’s useful enough to be competitive for these microscale applications.”

“This is an important development,” said Ming Wu, an electrical engineer at the University of California at Berkeley, who was not involved with the research. “It allows silicon integrated circuits to communicate with one another directly with light instead of electric wires. This is somewhat surprising as silicon has an indirect bandgap and does not normally emit light.”

Silicon “occupies the crown in electronic devices” and will continue its reign “without a doubt,” said Chang-Won Lee, an applied optics researcher at Hanbat National University, who also was not involved in the work. However, he agreed with Wu that this advance represents a step toward silicon-based computers that are less reliant on electronic communication. “For example, there is an optical CPU architecture that the semiconductor industry has been dreaming of. The report of silicon-based micro-LEDs shows significant progress in these attempts.”

Ram is confident that his team can continue fine tuning the technology, so that one day LEDs will be cheaply and efficiently integrated into silicon chips as the industry standard. “We don’t think we’re anywhere close to the end of the line here,” said Ram. “We have ideas and results pointing to significant improvements.”

This research was supported by Singapore’s Agency for Science, Technology and Research, and by Kwanjeong Educational Foundation.

Reprinted with permission of MIT News (http://news.mit.edu/).

Subscribe Now

Design-2-Part Magazine

Get the manufacturing industry news and features you need for free in a format you like.

FREE Print, Digital, or Both »

You have Successfully Subscribed!