James Clerk Maxwell might turn in his grave if he heard about some of the extremely cool things humans have done with his intellectual work. Looking at this paper, I said out loud, “They made a Maxwell’s demon out of lasers.” There isn’t literally a tiny demon controlling an atom-sized door. And they weren’t stopping atoms from moving into one chamber from another trying to violate thermodynamics. But scientists have made a pair of “optical tweezers” by focusing laser beams, and used them to fish individual rubidium-87 atoms out of a cloud and then pass the atoms back and forth between tweezers. The researchers who did it believe this development will be useful for quantum computing.
To trap individual atoms, the researchers first used laser cooling to bring a cloud of rubidium-87 atoms down to just a few degrees shy of absolute zero, slowing the atoms down from their usual, high-speed trajectories. They then directed a different laser beam through an optical splitter. The number of beams that come out, and the angle at which they exit the splitter, are both a function of RF waves applied to the deflector. Panning and pulsing the RF emissions let the researchers control the laser, on the microsecond scale.
After they ran the laser beam through the splitter to turn it into more of a laser comb, the researchers focused the newly split beams so that they would have a “waist” just 90 nm wide: a focal region where the laser beam was brightest. Centering those foci in the cloud of ultracold atoms, the researchers discovered that each beam’s focus attracted a single atom, fishing it out from the cloud and holding it hovering in midair.
“It’s similar to charging up a comb by rubbing it against something woolen, and using it to pick up small pieces of paper,” explains coauthor Vladan Vuletic. “It’s a similar process with atoms, which are attracted to regions of high intensity of the light field.”
While the atoms are being held in the bright focal points of the optical tweezers, they themselves give off light. Using a CCD camera, the researchers could see that changing light and tell which tweezers were holding atoms and which weren’t. Thusly empowered, using the directional RF emitter they could shuffle atoms between tweezers, to create arrays of up to 50 atoms. The arrays persisted for several seconds and were reliably “free of defects” — which is to say, well-behaved and properly populated with compliant single atoms.
What makes this relevant to quantum computing is that the power to manipulate individual atoms makes it possible to induce quantum gates, to see if they’ll scale. Quantum computing will require retaining quantum gates for longer than just a few seconds, but a few seconds is a good start. Just like how Boolean logic gates are made of transistors and process bits, quantum gates are the functional unit of quantum circuits and work with qubits. (I have it on good authority that one standard qubit is the length of Alan Turing’s forearm.) Theoretically quantum computing is a whole lot faster than silicon, but this has yet to be demonstrated at scale. Researchers have induced quantum gates between a single pair of neutrally charged atoms, but not more than two, not yet, and they don’t stick around long.
Mostly, the research on quantum computing has been using ions for their quantum gates, because ions are charged, and charged particles are a little easier to manipulate with electromagnetism. But for quantum computing, these researchers reasoned, ions might be less inclined to act as quantum gates of their charges: Like charges repel, so you can’t very well make a dense array out of them, and unlike charges will form ionic compounds that aren’t charged anymore. These researchers used neutral rubidium-87 atoms, hoping to sidestep the problem of mutually repelling charged particles and so reach the milestone of inducing not just one persistent quantum gate, but many.