Science 2 publishers2 min readPublished September 22, 2026 at 2:46 PM PDT

One-sided photon interactions pushed a light beam upstream through a quantum fluid of light

Yi Hu's group at Nankai University sent a narrow beam against the flow of a broader optical fluid. The swimmer felt attraction while the fluid pushed back, and the net force pointed upstream.

The Scientist · Science desk

Photograph accompanying One-sided photon interactions pushed a light beam upstream through a quantum fluid of light — Photo: nature.com

What happened

Physicists sent two timed laser beams through a nonlinear crystal. The broad beam stood in for a fluid and the narrow one for a swimmer, and tilting the fluid beam set its flow direction and speed.
The narrow beam reshaped the fluid as it passed, leaving a higher intensity on one side than the other. That asymmetric distribution produced the force that carried it against the flow.
In the usual reciprocal setup, run as a direct comparison, the same narrow beam moved downstream instead.
Sweeping fluid speed and density showed the upstream motion was strongest at intermediate values of both, between the highest and lowest settings the team tried.
Earlier upstream motion came from vortex shedding, whose net recoil momentum still obeys Newton's third law. Phys.org reports this is the first demonstration to use non-reciprocal interactions.

Compiled by The ScientistSomething wrong?How this is made

Why it matters

capability Designers of quantum-light hardware gain a way to bias where light goes without adding an energy-consuming element; Albert said such steering could eventually be useful inside devices for communications or quantum information processing.
constraint The platform was an optical analog of a quantum fluid, so anyone working with ultracold atoms or liquid helium gets a hypothesis to test here.
precedent New Scientist frames the work as a possible new type of active matter in a quantum system. The swimmer consumes no energy at all.

According to New Scientist, the swimmer experienced an attractive interaction with the river of light, while the river experienced a repulsive force on the swimmer ^[7]. Hu said the net effect of that mismatch was an upstream force on the swimmer ^[8]. The crystal sat under an electric voltage that made the photons interact, and the swimmer beam was shaped into a single solitary wave ^[4].

Siyu Li and colleagues describe the same asymmetry from the fluid's side. "The swimmer tends to reshape the fluid, passing through it asymmetrically, which in turn exerts on the swimmer a force opposite to the fluid flow direction. Such a counterintuitive process stems from the nonreciprocity of their mutual interactions, which results in an inverted recoil compared to that in equilibrium systems," they wrote in Physical Review A ^[9]^[1].

The experiment measures a reversal of direction between the two settings ^[24]. Alongside the bench work, the team ran a theoretical scattering analysis and computer simulations ^[12].

Phys.org notes that action-reaction symmetry survives plenty of non-equilibrium situations and breaks when the interactions are non-reciprocal. That is the point at which internal interactions can be converted into net momentum ^[18]. Mathias Albert at Cote d'Azur University in France told New Scientist that earlier upstream motion in quantum fluids of light always relied on tiny vortices created behind the swimmer. The recoil from creating them pushed it along ^[14].

This was an optical analog of a quantum fluid ^[17]. Neither account gives a momentum budget for the crystal and the pump light. The non-reciprocity they report is the one between the two beams ^[25].

"This question is very much in the air at the moment, with several groups exploring how the concepts of active matter can be combined with quantum fluids and photonic systems," Albert said ^[15]. The comparison class in that field is systems where objects consume energy to move, such as bacterial mixtures and flocks of birds ^[22]. The beam here consumed none and still showed behavior of the active kind ^[21]. "For me, the biggest takeaway is that active behaviour does not necessarily require an intrinsically active particle or swimmer," Hu said ^[19].

What to watch

Hu says the team wants more complex experiments with tighter control of the swimmer, or several swimmers in the same fluid at once.
Whether the same non-reciprocal route produces upstream motion in atomic gases or engineered quantum materials, as the authors suggest testing.
An independent replication that reaches upstream motion without any vortices behind the swimmer.

Claim ledger

Ranked by verification strength, evidence, and original report placement.

[1]
The study, "Active upstream motion in a quantum fluid of light arising from nonreciprocal interactions" by Siyu Li et al, was published in Physical Review A (2026), DOI 10.1103/lwyj-7m5f.
ReportedSupported2 sources— sign in to open View cited source
[2]
Yi Hu at Nankai University in China and his colleagues carried out the work.
ReportedSupported2 sources— sign in to open View cited source
[3]
The team sent two timed laser beams through a nonlinear crystal, one representing the fluid and the other the swimmer; tilting the fluid beam let them control its transverse flow direction and speed.
ReportedSupported2 sources— sign in to open View cited source
[4]
The quantum light fluid was created by shooting lasers into a special crystal exposed to an electric voltage, which made the photons interact and form a river-like system; the swimmer was another beam of quantum light shaped to form a single, solitary wave.
ReportedSupported2 sources— sign in to open View cited source
[13]
Upstream motion had been demonstrated in previous experiments, but the mechanism involved vortex shedding, which imparts a net upstream recoil momentum and still obeys Newton's third law; the new study is the first to use non-reciprocal interactions, violating action-reaction symmetry.
ReportedSupported2 sources— sign in to open View cited source
[14]
Mathias Albert at Cote d'Azur University in France says upstream motion has previously been achieved in other quantum fluids of light, but it always relied on creating tiny vortices behind the swimmer in the fluid, with the recoil from their creation pushing the swimmer, so the new experiment demonstrates a different mechanism for motion.
ReportedSupportedSource: Mathias Albert, Cote d'Azur University, via New Scientist2 sources— sign in to open View cited source
[5]
In the experiments, the narrow laser beam acting as the swimmer moved opposite to the flow of a broader optical fluid.
ReportedSupportedView cited source
[6]
In the usual reciprocal setup, the beam instead moved downstream, providing a direct comparison.
ReportedSupportedView cited source
[7]
In the experiment the swimmer experienced an attractive interaction with the river, but the river experienced a repulsive force on the swimmer, a non-reciprocal interaction that breaks the action-reaction principle.
ReportedSupportedView cited source
[8]
The net effect of the non-reciprocal interaction was an upstream force on the swimmer, says Hu.
ReportedSupportedSource: Yi Hu, Nankai University, via New ScientistView cited source
[9]
The study authors write: "The swimmer tends to reshape the fluid, passing through it asymmetrically, which in turn exerts on the swimmer a force opposite to the fluid flow direction. Such a counterintuitive process stems from the nonreciprocity of their mutual interactions, which results in an inverted recoil compared to that in equilibrium systems."
ReportedSupportedView cited source
[10]
The upstream motion is due to the asymmetric distribution of the fluid caused by the swimmer: the intensity of the fluid on one side of the swimmer is higher than on the other.
ReportedSupportedView cited source
[11]
By adjusting fluid speeds and densities, the team found the upstream motion was strongest at intermediate fluid speeds and densities, rather than at the highest or lowest values.
ReportedSupportedView cited source
[12]
The team combined theoretical scattering analysis, computer simulations and laboratory experiments.
ReportedSupportedView cited source
[15]
Albert says: "This question is very much in the air at the moment, with several groups exploring how the concepts of active matter can be combined with quantum fluids and photonic systems."
ReportedSupportedSource: Mathias Albert, Cote d'Azur University, via New ScientistView cited source
[17]
The experiments involved an optical analog of a quantum fluid, not a demonstration using material quantum fluids such as ultracold atoms or liquid helium.
ReportedSupportedView cited source
[18]
Action-reaction symmetry is not broken just because something is out of equilibrium, but it can be broken when particles exhibit non-reciprocal interactions, in which internal interactions can be converted into net momentum.
ReportedSupportedView cited source
[19]
Hu says: "For me, the biggest takeaway is that active behaviour does not necessarily require an intrinsically active particle or swimmer."
ReportedSupportedSource: Yi Hu, Nankai University, via New ScientistView cited source
[20]
Hu says the team now wants to construct more complex experiments where they could control the swimmer's motion more precisely or add several swimmers to the fluid at once.
ReportedSupportedSource: Yi Hu, Nankai University, via New ScientistView cited source
[22]
Non-reciprocal interactions are known to be crucial for shaping the motion of active matter systems where objects consume energy to move, such as bacterial mixtures, flocks of birds or swarms of tiny robots.
ReportedSupportedView cited source
[23]
Future research could test whether related effects occur in other quantum platforms, such as atomic gases or engineered quantum materials.
ReportedSupportedView cited source
[25]
Neither published account reports a momentum budget for the crystal or the pump light; the non-reciprocity described is between the swimmer beam and the fluid beam.
DerivedSupported2 sources— sign in to open
[24]
The reported contrast between the two configurations is a reversal of travel direction: upstream when the interaction is non-reciprocal, downstream when it is reciprocal.
DerivedSupported1 source— sign in to open
[16]
Albert says that, eventually, the ability to control chunks of light so they move in some preferred direction could be useful within devices that use quantum light, such as those used in communications or quantum information processing.
ReportedInsufficientSource: Mathias Albert, Cote d'Azur University, via New ScientistView cited source
[21]
The experiments by Hu and his colleagues focused on an object that does not consume energy yet still benefited from non-reciprocal interactions, and their work could usher in the creation of a new type of active matter in a quantum system.
ReportedInsufficientView cited source

Sources

Publishers with included, body-backed reporting in this cluster.

newscientist.com
1 article · September 21, 2026
Swimmer made of quantum light breaks Newton’s third law
phys.org
1 article · September 22, 2026
Light beam 'swims' upstream through a quantum fluid by violating Newton's third law

Ask Clarity

Have a question this piece left open? The answer comes from its reporting and claim ledger — nothing else.

Try

Clarity analysis

How the coverage leans and what the record actually supports. Read the methodology

Perspective Coverage

How much of the coverage speaks to builders, operators and investors. It measures emphasis, not whether the story is right.

The usable detail sits in the phys.org account: two timed beams pass through a nonlinear crystal under an applied voltage. The broad beam is tilted to set the transverse flow, and the narrow one is shaped into a solitary wave. Flip the interaction from reciprocal to non-reciprocal and the drift flips with it, from downstream to upstream. The effect was strongest at middling flow speeds and densities, not at the extremes, which is where anyone reproducing it would start. No speed, density or force figure appears in either account, so the DOI phys.org prints is the only route to magnitudes.

Reality

Evidence, adoption, hype gap, incentives and confidence, each read independently from the current record on the scale shown beneath it. How these are measured.

Evidence70

0100

Single peer-reviewed paper with a reciprocal control run

Both accounts trace to one paper in Physical Review A, and phys.org prints its DOI. The strongest thing in it is the control: with reciprocal interactions the same beam drifted downstream, giving the upstream result a direct comparison to check against. Against that, no numbers were published, so the claim that the push peaks at intermediate speeds and densities rests on the authors' word. Mathias Albert, who is not on the paper, told New Scientist that earlier upstream swimming always came from vortices behind the swimmer.

Adoption10

0100

One group, one paper, no replication

What exists is a journal publication and two write-ups of it. Albert said several groups are exploring how active matter combines with quantum fluids and photonic systems, which is interest in the question, not use of this result. Hu's own follow-up is more swimmers in the same crystal.

Hype gap+25

−100 understated0 aligned+100 overstated

Headlines harder than the paper's own caveat

Both headlines say Newton's third law was violated, and New Scientist's opening goes further, promising new ways of controlling quantum light in devices. The phys.org body is more careful: the fluid is an optical analog, not ultracold atoms or liquid helium, and the non-reciprocity described holds between two beams inside a voltage-biased crystal driven by external lasers. Neither write-up reports a momentum budget for the crystal or the pump light.

Incentives35

0100

Mild novelty pressure, plus a disclosed funding appeal

A first-of-its-kind framing is what gets a paper covered, and the priority claim reaches us from phys.org's write-up of the paper rather than from anyone who went back through the prior literature. Phys.org also asks readers for money at the end of its report and names its editor and fact-checker. New Scientist's one outside voice, Albert, works on the same question he is assessing, as he says himself.

Confidence68

0100

Core result firm, the priority claim thinner

Peer review, a printed DOI, a reciprocal control run and an outside physicist put the central fact on solid ground: the beam went upstream where the reciprocal configuration sent it downstream. I would hold the reach more loosely. That this is the first non-reciprocal case rests on phys.org alone, the device uses are Albert's own "eventually", and no magnitude for the effect appears in either account.

Topics and entities

Follow any of these and your For You feed starts watching them — no settings page required.

Topics

Entities

science
Leipzig and Prague physicists switch a microswimmer's propulsion mode with patterned laser light
1 publisher
science
Swimmers in the glass: activity, not preparation, may decide how amorphous solids break
1 publisher
science
Unequal colloid sizes turned mutual attraction into self-propelled pairs at Tokyo University of Science
1 publisher
science
A 0.3-degree tilt restored the second-harmonic light this metasurface had cancelled
1 publisher

What happened

Why it matters

What to watch

Claim ledger

Sources

Let Clarity write the post for you.

Topics and entities

Topics

Entities

Related stories

Claim ledger

Let Clarity write the post for you.

Topics and entities

Topics

Entities

Perspective Coverage

Reality

Related stories

Sources