In a first, physicists spot elusive 'free-range' atoms — confirming a century-old theory about quantum mechanics
When you buy through links on our articles, Future and its syndication partners may earn a commission.
An illustration of atoms floating freely in the air. | Credit: Stanislaw Pytel via Getty Images
For the first time, scientists have observed solo atoms floating freely and interacting in space. The discovery helps to confirm some of the most basic principles of quantum mechanics that were first predicted more than a century ago but were never directly verified.
Individual atoms are notoriously difficult to observe due to their quantum nature. Researchers cannot, for example, know both an atom's position and its velocity at the same time, due to quantum weirdness. But using certain laser techniques, they have captured images of clouds of atoms .
"It's like seeing a cloud in the sky, but not the individual water molecules that make up the cloud," Martin Zwierlein , a physicist at MIT and co-author of the new research, said in a statement .
The new method goes one step further, allowing scientists to capture images of "free-range" atoms in free space. First, Zwierlein and his colleagues corralled a cloud of sodium atoms in a loose trap at ultracold temperatures. Then, they shot a lattice of laser light through the cloud to temporarily freeze the atoms in place. A second, fluorescent laser then illuminated the individual atoms' positions.
Related: There may be a 'dark mirror' universe within ours where atoms failed to form, new study suggests
The observed atoms belong to a group called bosons. These particles share the same quantum mechanical state and, as a result, behave like a wave, bunching together. This concept was first proposed by French physicist Louis de Broglie in 1924 and has subsequently become known as a "de Broglie wave."
Top: Two illustrations show how atoms in an atom trap (red) are suddenly frozen in place via an optical lattice. Bottom: Three microscope images show (left to right) bosonic 23Na forming a Bose-Einstein condensate; a single spin state in a weakly interacting 6Li Fermi mixture; and both spin states of a strongly interacting Fermi mixture, directly revealing pair formation. | Credit: Yao et al.
Sure enough, the bosons Zwierlein and his team observed displayed de Broglie wave behavior. The researchers also captured images of lithium fermions — a type of particle that repels similar particles rather than bunching together.
RELATED STORIES
—Physicists create hottest Schrödinger's cat ever in quantum technology breakthrough
—Scientists claim to find 'first observational evidence supporting string theory,' which could finally reveal the nature of dark energy
—'Einstein's equations need to be refined': Tweaks to general relativity could finally explain what lies at the heart of a black hole
The results were published May 5 in the journal Physical Review Letters . Two other groups reported using a similar technique to observe pairs of bosons and fermions in the same issue of the journal.
"We are able to see single atoms in these interesting clouds of atoms and what they are doing in relation to each other, which is beautiful," Zwierlein said.
In the future, the team plans to use the new technique — called "atom-resolved microscopy" — to investigate other quantum mechanical phenomena. For example, they may use it to try observing the "quantum Hall effect," in which electrons sync up under the influence of a strong magnetic field.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
22 minutes ago
- Yahoo
A 'Strawberry Moon' is coming to June's skies — and it holds a special record for 2025
When you buy through links on our articles, Future and its syndication partners may earn a commission. Stargazers, get ready for some low-hanging fruit: The full Strawberry Moon is about to rise! In addition to being the final full moon of spring in the Northern Hemisphere, June's Strawberry Moon will be the lowest full moon of the year as seen from north of the equator, as well as one of the farthest from the sun. The moon will become full at 3:45 a.m. EDT Wednesday (June 11), but because that's so early in the day, the best time to see it will be Tuesday (June 10) evening, when it rises during dusk. As you watch the moon appear, look to its upper right to spot Antares, a bright star 550 light-years away, in the constellation Scorpius. June's full moon is always one of the lowest-hanging of the year, as seen from the Northern Hemisphere. That's because a full moon is, by definition, opposite the sun, so it mirrors our star's position in the sky. Because the summer solstice occurs on the night of June 20 (or June 21 GMT), when the sun is as high in the sky as it can get, the closest full moon is the lowest of the year. That means it will rise during dusk in the southeastern sky; drift across the southern sky, never getting too far above the southern horizon; and set in the southwest at dawn. RELATED STORIES —The 10 best stargazing events of 2025 —Best telescopes 2025: Explore planets, galaxies and beyond —Best binoculars for stargazing 2025: Take a walk through space The Strawberry Moon is one of the farthest full moons from the sun simply because Earth's orbit of the sun is slightly elliptical. That means there's a closest point and a farthest point in its orbit. The farthest point, called aphelion, occurs on July 3 this year. Because a full moon is opposite the sun, this is when the moon is at its farthest from the sun all year. June's full moon will be approximately 94,600 miles (152,200 kilometers) from the sun. The Strawberry Moon is named for the wild strawberries that ripen in some areas of the Northern Hemisphere this month, according to Other Native American names for this month's full moon include the Berries Ripen Moon, Green Corn Moon, Hot Moon and Blooming Moon. English names include the Flower Moon, Planting Moon and Mead Moon, while some Celtic names for it are the Horse Moon, Dyan Moon and Rose Moon.
Washington Post
42 minutes ago
- Washington Post
Little-known cells might be key to human brain's massive memory
A new model of memory — and a little-heralded type of brain cell — might explain why the human brain has such a huge storage capacity, researchers reported in the journal PNAS in May. The study looks at astrocytes, star-shaped cells that interact with neurons in the brain. The brain contains billions of astrocytes, and scientists have long known they play a part in cleaning up molecules within brain synapses, the junctions where neurons come together. More recent research, though, suggests that astrocytes do more. The new model suggests the astrocytes could also be used for computation, coordinating with neurons and connecting synapses in networks. These complex connections might allow the brain to encode memories in dense networks that expand the brain's capacity beyond its neurons alone, the researchers write. 'This makes neuron-astrocyte networks an exciting candidate for biological 'hardware' implementing Dense Associative Memory,' they add. The model runs counter to the prevailing theory that memory storage occurs in synapses, and implies that the brain is capable of storing even more memories than once thought possible. The researchers also describe how the theory could be validated in the lab. 'We hope that one of the consequences of this work could be that experimentalists would consider this idea seriously and perform some experiments testing this hypothesis,' said Dmitry Krotov, a research staff member at the MIT-IBM Watson AI Lab and IBM Research and the paper's senior author, in a news release. The model could also be used as a 'fresh source of inspiration' for future artificial intelligence technology, the researchers conclude.
Forbes
44 minutes ago
- Forbes
Prosthetics Startups' Biggest Market Isn't Humans. They're Building Hands For Robots Instead.
Aadeel Akhtar, the founder of Psyonic, has a PhD from the University of Illinois Urbana-Champain. Matt Carney was good at building robots — he just didn't want to. While earning his PhD at MIT, he'd spent years studying mechanical engineering and biomechatronics in service of developing bionic prosthetics that could help people who'd lost limbs. He hoped to build robotic legs that could pick up on the phantom signals sent by a body's muscles or function autonomously so it could move naturally, unlike the plastic, unmoving prostheses that are common now. But as he began talking to venture capitalists about funding a company that would develop these so-called bionics, he quickly discovered that the market didn't want robotic devices that could replace human limbs, it wanted robots that could replace human beings, the sort of humanoids championed by sci fi laureates like Isaac Asimov and self-appointed tech visionaries like Elon Musk. Investors cautioned him against venturing into the cost-heavy medical world and regularly asked if he might be more interested in building humanoids or exoskeleton suits. Even trusted experts building bionic limbs told him the technology currently available wasn't advanced enough to be truly helpful. Without a breakthrough in AI, let alone an addressable market, building futuristic robot body parts wasn't something investors would bankroll. So Carney looked elsewhere, taking a job as chief engineer at Persona AI, an early-stage contender in the field of humanoid robotics that has raised $27 million in funding. It's facing off against much larger startups like Boston Dynamics, Foundation Robotics and Figure AI, which have achieved valuations greater than $1 billion for their human-like bots. Then there's the elephant in the room: Musk's Tesla, which the world's richest man has pivoted toward building humanoid robots, with the idea that they'll generate trillions in revenue for the company and someday outnumber humans. Musk has touted a lofty vision of the future where he sees these robots as a catch-all method for labor, posting on X about replacing surgeons, being a 'personal C-3PO or R2-D2' to perform the bidding of an owner, and, of course, driving a car. Much of the tech world seems sold on them, too. The Information reported Wednesday that Amazon will begin testing out package delivery with humanoids. Speaking onstage at an event in late May for the Ronald Reagan Presidential Foundation and Institute, billionaire venture capitalist Marc Andreessen said that humanoid robotics provide a solution to regaining manufacturing dominance in the United States, with these robots performing tedious tasks such as installing screws on an assembly line. Carney plans to return to the prosthetics space one day. But he isn't the only entrepreneur who originally wanted to focus on building so-called bionic limbs to help real people, and instead have been pushed to follow the money: why not build for robots too? Take San Diego-based Psyonic, which officially launched in 2021 solely as a prosthetics company, building a waterproof robotic hand with built-in sensors to register touch sensations (alongside perks like the ability to charge a cell phone). Psyonic found some early traction, with Medicare covering the cost of its hands for patients. It was last valued at $65 million in 2024 and has raised a total of $8 million, according to PitchBook. Not long after launching, CEO and founder Aadeel Akhtar started to realize that there could be an entirely new market for his startup's hands, signing on Meta as an early robotics customer for an AI project. By 2023, humanoid robotics were booming, and demand skyrocketed. Now, the majority of Psyonic's business comes from selling its hands to manufacturing firms and robotics companies like Apptronik, which has raised more than $400 million to build humanoid, general purpose robots. The robotics side of his business 'is growing exponentially right now,' Akhtar told Forbes, later adding, 'The big draw there is being able to have one generalized robot do many tasks as opposed to activity specific robots.' Shifting toward catering to humanoid robotics startups comes with advantages. The hundreds of millions being poured into the space is helping commoditize the cost of the core technologies — the actuators, sensors, control methods and carrying capabilities needed to build both humanoid robots and bionic limbs. Akhtar said the shift has helped it further develop its prosthetics technology, and lower its prices. For San Antonio-based Alt-Bionics founder and CEO Ryan Saavedra, who was inspired to build inexpensive robotic hands after his hand was injured in a rock climbing accident, Apptronik is also a major customer. Now, the company sells more robotic hands than prosthetics to 'tier one humanoid companies.' It's a form of 'cross-pollination,' Saavedra said. And Leeds, England-based COVVI, which makes prosthetic hands with the ability to move individual fingers, recently launched a robotic hand tailormade for humanoid robots. 'Every humanoid robot needs robotic arms and legs,' said Connor Glass, the CEO and founder of Phantom Neuro, which is developing a non-invasive implant that enables a person to control a prosthetic device and works closely with prosthetics companies. 'Now [startups] are able to pivot in a way, and try to generate revenue by working with these humanoid robot companies.' It looks like a savvy business move. A Goldman Sachs report from 2024 projects that the total addressable market for humanoid robots will reach $38 billion by 2035. It projects that millions of general purpose human-shaped robots will be produced by then, with a 40% reduction in costs. (Right now, they're often prohibitively expensive; Apptronik is targeting a $50,000 price tag but hasn't reached this unit price, and Boston Dynamics' Atlas robot reportedly costs hundreds of thousands of dollars.) By contrast, there are only 5.6 million U.S. citizens with limb loss or limb difference, according to the Amputee Coalition. That makes this a very small market, and one that's accessed through the U.S. healthcare system. It can take years and millions of dollars to get a more advanced product through the FDA or navigate a complicated insurance system. Robots don't have such hurdles. Not all prosthetics startups are going after this new opportunity. Joel Gibbard, who cofounded Bristol, U.K.-based Open Bionics in 2014 and has raised $18 million to date, has decided to stay focused on making its 3D-printed bionic arms for humans, not robots. 'Everyone in this space is probably having those ideas, thinking about it as a growth opportunity,' Gibbard said. 'I don't know if we've made the right judgment, but I can tell you that we've made a conscious decision.' Bionics needs a watershed moment to get investors' attention, said Tyler Hayes, founder of Atom Limbs, which is building a complex AI-powered arm. 'A company is going to need to demonstrate a pretty significant breakthrough, as the public would see it, for bionic limbs to get that kind of traction,' he said. Building robot limbs for bots is far from a sure bet. Ken Goldberg, the cofounder of robotics firms Ambi Robotics and Jacobi Robotics and a University of California-Berkeley professor, told Forbes that while advances in humanoid robotics are possible in the (very) long term, the timeline and hype surrounding them are exaggerated. Videos promoting these humanoids' capabilities are often misleading, and there's always a wizard behind the curtain, he said. 'Robots are getting very good at locomotion, walking, and so there are all these robots out there that look like humans, walk like humans, and people think 'Well then, they are humans',' he said. 'Manipulation, the hands, is where the challenge is.' (That's why his company, Ambi Robotics, is building purpose-built AI-powered robots that can grip and sort packages, while his newer startup, Jacobi Robotics, is developing purpose-built robots for moving items on and off industrial pallets.) Simple tasks like folding a box or clothes, anything that requires dexterity, are challenging and harder than an activity like walking. It's no wonder humanoid startups are looking toward bionic prosthetics for help. 'We're nowhere near being able to automate those things,' Goldberg said. 'Hands on a robot is a surprisingly big leap.'
