字幕列表 影片播放 列印英文字幕 Sometimes, scientific breakthroughs happen because of a focused effort on a clear goal. Sometimes, they happen because of one little accident, like the discovery of penicillin. And then sometimes, they happen because a series of improbable accidents all fall perfectly into place. This latest research that may change how we approach building quantum computers is one of those series of unlikely events. This story starts in 1961, when the Dutch-American physicist Nicolaas Bloembergen proposed that the nucleus of an atom could be manipulated using an electric field, what's known as nuclear electric resonance. But experiments that tried to demonstrate nuclear electric resonance proved too challenging, and so even though nearly 60 years have passed since Bloembergen predicted its existence, it still hasn't gotten off the chalkboard. In fact, it was mostly forgotten about until a few researchers at the University of New South Wales in Sydney published a paper in 2020 and “rediscovered” it. The researchers were experimenting with another field pioneered by Boembergen: nuclear magnetic resonance. Nuclear magnetic resonance is a well established phenomenon, and is the principle that magnetic resonance imaging, or MRI machines, use to operate. It's also one possible approach to building quantum computers. Leveraging the bizarre nature of the quantum realm, quantum computers can use a single atom to take the place of a silicon transistor, functioning as either a one or zero like a classical bit, or both at the same time, or any combination in-between. This ability to represent multiple values at once makes quantum bits, or qubits, much more suited for solving complex problems. But quantum computers are much harder to make as small as their classical counterparts. Silicon transistors can be packed into devices by the billions. By contrast, a quantum computer developed by Google in 2019 had to keep its 54 qubits made of superconducting metal near absolute zero, so they were kept in a special refrigerator about the size of a phone booth. One dream for future quantum computers is a best-of-both-worlds scenario, where single atoms embedded in silicon can be manipulated with magnetic fields, producing more compact chips with millions of qubits on them. As far back as 1998, researchers imagined a qubit made out of silicon and phosphorus that used a magnetic field to change the phosphorus nucleus' spin. But magnetic fields by their nature do not fit nicely into this dream scenario. It's hard to confine them to a small space, so while they can influence the spin of one nucleus, they will also likely affect the spins of neighboring nuclei, too. One researcher from the latest paper likens nuclear magnetic resonance to moving a billiard ball by shaking the whole table. And this is where their latest work comes in. The scientists decided to experiment with how nuclear magnetic resonance affected a single atom of antimony. They embedded the atom in a silicon chip along with a microscopic antenna, but when they flipped on the power and ran a current through the metal, the antimony didn't respond as they had expected. Its spin responded strongly to some frequencies and not at all to others. It took a month of head scratching before they realized what had happened. The antenna in their device couldn't handle the strong current running through it, and like an electrical fuse, it broke. This changed its nature. No longer was it emitting a strong oscillating magnetic field, instead it had turned into an electrode that was giving off a strong oscillating electric field. However, the field still wouldn't have done anything if it weren't for another lucky break, so to speak. The atom's nucleus happened to be sitting in an uneven static electric field, because the silicon had been distorted by the contraction of aluminum leads on its surface as the chip was cooled to near absolute zero. Without that uneven field, the electric field wouldn't have had any effect. But even with this string of coincidences, the experiment wouldn't have amounted to much had the researchers used a different element. Had they decided to use phosphorus like the 1998 proposal, the small nucleus wouldn't have responded. But the larger nucleus of the antimony atom did. This all adds up to what could be a huge breakthrough for quantum computing. Because electric fields fade sharply over distance, electrodes can be used to affect single qubits precisely, meaning more can be packed into a smaller space on familiar silicon. Going back to our billiards analogy, it's now as though scientists have been handed a pool cue. So to recap: a team of researchers in Australia decided that, just for the fun, they wold test how a magnetic field affected an atom, and because of some accidental breakage, some surprise shrinkage, and their choice of atom... they made a landmark discovery that could make quantum bits smaller, easier to make, and more powerful. Oh, and they solved a problem that hasn't been cracked since it was proposed in 1961. I guess if there's anything to learn from all this, it's that accidents do happen— and sometimes their results are better than anything we could have hoped for. Bloembergen died in 2017, never seeing his idea of nuclear electric resonance demonstrated. But in a crazy coincidence, as though this story didn't have enough of them, these findings were published in Nature on what would have been Boembergen's 100th birthday. If you like learning about quantum computers check out my video on Google's claim to quantum supremacy. Don't forget to subscribe to Seeker. Thanks for watching, and we'll see you next time.