Physicists in the United States have just announced a major breakthrough after they literally put a new spin on one of the greatest inventions in history: the transistor. The scientists made an entirely novel switching device called a magneto-electric transistor that uses 5% less energy than conventional semiconductor transistors, while potentially reducing the number of transistors needed to store data by as much as 75%.
A new age of computing may be upon us
A semiconductor transistor consists of three critical terminals that need to be connected to an electrical circuit. The “source” and the “drain” serve as the starting and end points, respectively, allowing electrons to flow through the circuit. Above these two terminals lies the “gate”, which aptly named allows or restricts the flow of electricity through the transistor. The buildup or absence of charge is encoded as a “1” or “0”, or “yes” or “no”, the basic binary language that can be used to compute anything. With enough transistors, like the trillions crammed inside your typical modern computer chip, you can perform all sorts of amazing things from streaming your favorite sitcom on your phone to guiding a robotic rover on Mars from 35 million miles away.
Since they were first invented in 1947, the number of transistors packed onto a single silicon chip has steadily grown in density, leading to the exponential computing growth that has enabled the production of increasingly powerful computers. This progress has been summed up by Moore’s Law, which states that the number of transistors in an integrated circuit doubles about every two years.
Moore’s Law has been remarkably accurate and, to this day, it hasn’t abated from its stated trendline. But this doesn’t mean it can go on forever. A transistor is a physical object and consequently, it has hard constraints determined by the laws of nature.
“The traditional integrated circuit is facing some serious problems,” Peter Dowben, Charles Bessey Professor of physics and astronomy at Nebraska, said in a statement. “There is a limit to how much smaller it can get. We’re basically down to the range where we’re talking about 25 or fewer silicon atoms wide. And you generate heat with every device on an (integrated circuit), so you can’t any longer carry away enough heat to make everything work, either.”
The problem is compounded by a double demand for not only faster transistors but more energy-efficient ones too. In 2020 alone, the world manufactured a trillion computer chips, each crammed with billions of tiny transistors. In the future, many more chips will have to be built and shipped in order to satisfy the growing needs of the digital age.
It’s not just computers and phones that need computers. A modern car can easily have more than 3,000 chips, for instance. In fact, you’ll find computer chips in almost anything right now from TVs to toasters. All of these electronic devices, while useful, require an increasingly large amount of energy to power them. Bitcoin mining alone uses as much power as the Philippines, a country of 110 million people. At this rate, we’ll soon use as much energy as the entire United States just for memory alone, Dowben warns.
“So you need something that you can shrink smaller, if possible. But above all, you need something that works differently than a silicon transistor, so that you can drop the power consumption – a lot,” he added.
A new spin on the transistor
It was by looking at these demand problems and the physical constraints of the conventional transistor that Dowben and colleagues reckoned that they had to come up with something that works fundamentally differently. Eventually, they figured out how to make an electric-magnetic transistor.
Here’s how it works. Instead of leveraging the switching of the flow of electrons through a circuit, the electric-magnetic transistor uses a fundamental property of electrons called spin, which can point either up or down. The orientation of a particle’s spin can be manipulated using, you’ve guessed it, magnetism.
To make their spintronic-based transistor, the researchers started off with an atom-thick layer of graphene, which has the appealing property of allowing electrons flowing through it to maintain their initial spin orientation for relatively long distances. But in order to produce a switching effect, the graphene had to be paired with the right material. In this case, the physicists used chromium oxide, a magneto-electric material where spins of atoms at its surface can be flipped up or down by applying just a tiny amount of voltage.
When the applied voltage is positive, the spin of the underlying chromium oxide atoms points up. This forces the spin orientation of the electrons flowing through the graphene underneath to veer left, producing a detectable signal in the process. Negative voltage flips the chromium oxide atoms down and the electrons in the graphe to the right. The two signals are clearly distinguishable from each other — that’s your 1 or 0.
The resulting transistor not only uses 5% less energy than a conventional modern transistor, but can also be used as a much more efficient memory device. Random-access memory, or RAM, requires a constant supply of power to maintain its binary states. However, microelectronics based on a magneto-electric transistor will be able to remember exactly where a user left off even after the device has been powered off since switching doesn’t involve changing the current.
“The implications of this most recent demonstration are profound,” said Dowben.
While this proof of concept is impressive, the researchers claim this is merely scratching the surface. Now that this demonstration has been completed, other labs and research groups across the world can use this framework and pick up where Dowben and colleagues left off. There are many other 2-D materials that are better suited to magneto-electric applications than graphene.
“Now that it works, the fun begins, because everybody’s going to have their own favorite 2D material, and they’re going to try it out,” Dowben said. “Some of them will work a lot, lot better, and some won’t. But now that you know it works, it’s worth investing in those other, more sophisticated materials that could [work].”
“Now everybody can get into the game, figuring out how to make the transistor really good and competitive and, indeed, exceed silicon.”