HARBIN, CHINA
For sixty years, humanity's plan for reaching another star has been beautifully, brutally simple. Build a mirror the size of a city block. Make it thinner than a soap bubble. Point the most powerful laser array ever assembled at it. Then hold the trigger down and wait.
No fuel tanks. No engine. No exhaust. Just light, hammering away at a sheet of reflective film, until that film is moving at a serious fraction of the speed of light and the stars stop being scenery and start being destinations.
This week, that plan slammed into a wall nobody had properly accounted for. And in the cruellest twist physics has served up in years, the wall is made of light.
A new paper posted to the arXiv preprint server by Chao Shen and Jiaze Li of the Harbin Institute of Technology — picked up this week by Universe Today and syndicated across the science press — claims that once a laser-driven lightsail passes roughly 75 percent of the speed of light, part of the light hitting it stops pushing the craft forward and starts dragging it back. Not metaphorically. Actually. A braking force, generated by the same beam that is supposed to be the engine.
THE DREAM THAT REFUSES TO DIE
To understand why this matters, you have to understand how badly humanity wants this to work.
Our fastest spacecraft are pathetic on interstellar scales. Voyager 1, launched in 1977 and now the most distant human object in existence, is crawling along at about 17 kilometres per second. At that pace, a trip to Proxima Centauri — our nearest stellar neighbour at 4.24 light years — would take something in the region of 75,000 years. Chemical rockets are not a solution to interstellar travel. They are the reason interstellar travel is impossible.
The lightsail is the escape hatch. Because a sail carries no propellant, it dodges the tyranny of the rocket equation entirely: every gram of fuel a conventional rocket carries is a gram it also has to accelerate. Strip out the fuel, leave the energy source parked on Earth in the form of a gigantic laser bank, and the arithmetic suddenly stops being absurd.
The concept isn't science fiction. Japan's IKAROS spacecraft unfurled a working solar sail in deep space in 2010. The Planetary Society flew LightSail 2 in 2019. NASA's Advanced Composite Solar Sail System launched in April 2024 and deployed a sail roughly the size of a small apartment from a spacecraft the size of a toaster oven — a genuine milestone, even though the craft ended up with a bent boom and a persistent tumble. NASA has also funded studies of exotic "diffractive" sails that bend light rather than simply bouncing it, work associated with researcher Amber Dubill and the agency's Innovative Advanced Concepts programme.
And then there was Breakthrough Starshot. Announced in 2016 with $100 million of billionaire funding and Stephen Hawking's endorsement, it proposed a fleet of a thousand gram-scale probes, each attached to a sail a few metres across, blasted by a ground-based laser array to 20 percent of light speed. Twenty years of cruise. Four years for the photos to come home. A postcard from Proxima b within a single human lifetime.
Starshot has since gone quiet — Scientific American memorably described its "quiet demise" — but the idea outlived the organisation. A 2026 paper revived the concept of a near-relativistic swarm capable of imaging Proxima b at 20-metre resolution and sniffing for biosignatures. Every serious roadmap to another star still runs through a sail. Which is exactly why Shen and Li's result stings.
THREE WAYS LIGHT CAN SHOVE YOU
The Harbin researchers did something that sounds mundane and turns out not to be: they took the forces that photons exert on a sail and split them into three separate categories, then tracked each one all the way up to relativistic speed.
The first and strongest is incident light — the raw momentum of photons slamming into the sail's surface. Photons have no mass but they do carry momentum, and when billions of trillions of them arrive every second, the sail feels a shove.
The second is specular reflection — the extra kick delivered when a photon bounces cleanly off the mirror surface and heads back the way it came. Think of the difference between catching a ball and hitting it back with a bat. Bouncing gives you roughly twice the punch of absorbing.
The third and weakest is diffuse scattering — photons that get absorbed by the sail material and then re-emitted in essentially random directions, the way a sheet of paper glows evenly under a lamp rather than reflecting a sharp image. At everyday speeds, all three effects point the same way: forward. The sail accelerates. The physics is friendly. Then relativity turns up.
THE DOPPLER TAX
The first problem is one physicists have understood for a long time, even if the public hasn't: as the sail runs away from the laser, the light chasing it gets stretched.
This is the Doppler effect, the same phenomenon that drops the pitch of an ambulance siren as it speeds past you. Applied to a lightsail, it means the laser beam arriving at the sail is progressively redshifted — lower frequency, lower energy per photon, less punch.
The faster you go, the weaker your engine becomes. And the Harbin paper finds this tax applies to all three force components at once. Incident pressure, specular reflection, diffuse scattering — every one of them fades as velocity climbs.
That alone is a brutal design constraint. It means acceleration isn't just difficult to sustain; it actively degrades the harder you push. Engineers have known they'd have to build laser arrays of frankly terrifying power to compensate. Starshot's own numbers called for something in the region of 100 gigawatts — comparable to the entire electrical output of a mid-sized country, focused on a target a few metres across, for several minutes.
But the Doppler tax is at least intuitive. What Shen and Li found at 75 percent of light speed is not.
THE WALL AT 75 PERCENT
The culprit is a phenomenon called relativistic light aberration.
At ordinary speeds, when the sail re-emits scattered photons in random directions, roughly as many go forward as backward, and the net effect is a small forward push from the geometry of the beam. But relativity does strange things to angles. When you are moving at a substantial fraction of light speed, light that you emit in all directions gets swept forward — from the perspective of a stationary observer back on Earth, the scattered glow is beamed ahead of the craft like a headlight.
And here is where Newton's third law turns traitor. Every action has an equal and opposite reaction. If your spacecraft is throwing photons forward, those photons are pushing your spacecraft backward.
Above roughly 75 percent of the speed of light, according to the paper, diffuse scattering flips sign. The weakest of the three forces stops being thrust and becomes drag. The researchers describe identifying, for the first time, a critical velocity beyond which diffuse-scattering thrust "transforms into drag."
To be clear about the scale of the problem: the total force does not go negative. The laser is still winning. A sail at 75 percent of light speed does not slam on the brakes and reverse. But the efficiency of the whole propulsion system takes a substantial hit precisely at the point in the journey where you can least afford it, because the Doppler tax has already gutted your thrust.
You are trying to squeeze the last few percent of speed out of a system that is simultaneously being weakened by redshift and actively resisted by its own glow.
WHAT THE PAPER LEAVES OUT — AND WHY THAT'S WORSE
Here is the genuinely uncomfortable part. Shen and Li deliberately modelled an idealised case. Their sail is a perfect mirror. Their analysis covers radiative dynamics only.
That means the paper excludes: collisions with interstellar gas and dust, which at 75 percent of light speed turn individual hydrogen atoms into something closer to particle-accelerator projectiles; the thermal limits of real materials, which matter enormously when you are pointing a 100-gigawatt beam at a film thinner than kitchen foil and hoping it does not simply vaporise; and the curvature of spacetime across interstellar distances, which the authors also simplify away.
In other words, the drag effect is what remains after you have removed every other problem. It is a floor, not a ceiling.
THE EXPERT VIEW: NOT A DEAL-BREAKER, A DESIGN BRIEF
The propulsion community's response to results like this is rarely despair. It is a shopping list.
Andy Tomaswick, writing for Universe Today in the piece that carried the paper to a general audience, noted that engineers are already investigating advanced metamaterials and photonic crystals tuned to specific laser wavelengths — surfaces that do not behave like the naive mirror in the Harbin model at all. Crucially, some researchers think the aberration effects the paper describes could be turned into an asset rather than a liability: a sail engineered with the right optical response might use them to self-correct its orientation and stay centred in the beam, solving the notorious "beam-riding" stability problem that has haunted lightsail design for decades.
That is not spin. It is how this field advances. A 2024 arXiv paper on all-optical damping forces enhanced by metasurfaces was already exploring exactly this territory. You cannot engineer around a force you have not characterised — and until this week, nobody had characterised this one.
There is also a pragmatic escape route. Almost nothing in the near-term interstellar playbook actually requires 75 percent of light speed. Starshot targeted 20 percent. The 2026 swarm proposal for Proxima works in the same neighbourhood. At those velocities the drag term never activates. The wall exists, but the first generation of missions would not reach it.
WHAT THIS MEANS FOR YOU
Realistically, nobody reading this will board a lightsail. So why should anyone outside a physics department care?
Three reasons.
First, the technology has near-term uses that have nothing to do with other stars. Solar sails are being seriously studied for solar weather monitoring — a sail can hover at an orbital position no conventional spacecraft can hold, giving forecasters earlier warning of the coronal mass ejections that can knock out power grids and satellites. That is infrastructure protection, and it is coming this decade.
Second, this is a rare public look at how physics actually gets done. Not a triumphant discovery but a careful, unglamorous accounting exercise that found something inconvenient and published it anyway. The Harbin team could have modelled only the two big forces, ignored the weakest one, and produced a cheerier paper. They didn't.
Third, and least comfortably: this is another data point suggesting the universe is not merely large but structurally hostile to being crossed. Relativity does not just cap your speed at c. It builds in a series of escalating penalties as you approach it — time dilation, mass-energy costs that climb toward infinity, redshifted engines, and now a self-generated drag from your own scattered light. The cosmic speed limit is not a single line. It is a gradient that gets steeper the harder you push.
CONCLUSION
The lightsail is not dead. The number to remember is 75 percent — and almost every credible mission concept on the table today tops out at a quarter of that.
But the Harbin paper has quietly moved the goalposts for anyone dreaming of the truly fast crossing, the one that gets a probe to Alpha Centauri in a decade rather than two. That mission now has to be designed around a force that, until this summer, was not in anyone's spreadsheet.
Somewhere in a lab, a materials scientist is already reading arXiv:2606.04052 and sketching a metasurface that turns the problem into a steering system. That is the optimistic reading, and it is a reasonable one.
The pessimistic reading is simpler. We have spent a century learning that the speed of light is not a wall you hit. It is a wall that pushes back — and it pushes hardest exactly when you are closest.
