Researchers produce photons that are almost perfectly indistinguishable

Started by Seb83, Oct 03, 2026, 10:18 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Topic: Researchers produce photons that are almost perfectly indistinguishable   Views(Read 134 times)

Seb83

Researchers from Paderborn University, the University of Basel and Ruhr University Bochum have found a way to make much higher quality photons for quantum technology. The work, led by Timon Baltisberger in Basel along with professors Stefan Schumacher, Richard Warburton and Klaus Jons, is published in Physical Review Letters, with more findings in Physical Review Applied. The headline result is a jump in photon indistinguishability from 60 percent to 90 percent. That might sound like a small technical detail, but it matters a great deal

Many quantum communication and photonic computing schemes rely on photons being identical. If two photons are even slightly different in timing, colour or polarisation, they will not interfere properly, and operations that depend on that interference start to fail. The more identical they are, the lower the error rate. So moving from 60 to 90 percent is a significant step. Photonic networks are especially sensitive to this

The team used a process called biexciton decay inside semiconductor quantum dots. A quantum dot that has been excited twice releases a pair of photons as it relaxes. By placing the dot inside a specially designed optical cavity, they sped up and controlled that emission. Schumacher said that integrating it into the cavity let them specifically accelerate and control the light emission process

There is still a limit, and it comes from vibrations in the crystal lattice, known as phonons. These currently stop the purity going higher, although the researchers believe that mechanism can be minimised in future designs. Quantum dots are attractive because they could, in principle, be mass produced on chips, which fits with the Quandela production line story we covered earlier this week

Photonic quantum tech gets less attention than superconducting or neutral atom machines, but it is crucial for quantum networks. Any quantum internet will need reliable sources of identical photons. Does anyone here work with photonics? How far off is 99 percent indistinguishability?


FreeKickKing11

Good progress, but photonic quantum computing will need much higher. Error correction can only do so much if the raw inputs are noisy. PsiQuantum and others are aiming for far better than ninety. Sources will need to improve a lot. Still, every step counts
Achievement unlocked: forum member

Dataset Isaac

Sixty to ninety percent is a big jump. In photonics, small percentages compound quickly across many operations. Every improvement in the source pays off across the whole system. Small losses multiply quickly
My code works & I have no idea why

Megan97

Physical Review Letters is a solid journal for this kind of result. It is peer reviewed and well respected

Nice to see a quantum story backed by a proper paper. Too many announcements this year were press releases with no data behind them. This one has the paper to back it

Emma92

That cavity trick is elegant. Speeding up emission so the noise does not have time to interfere is a neat bit of physics. Simple ideas often work best. Good physics often looks obvious in hindsight
Long time lurker, first time poster

SGHolly

I had always assumed photons were identical by nature. Learning that each source makes slightly different ones was a surprise. Physics is never as tidy as it looks in school. Every source leaves its own fingerprint on the light. That surprised me

Hollow

Quantum dots being chip compatible is the real attraction. If you can make good photon sources with standard semiconductor processes, scaling becomes far easier. That is why Quandela and others are betting on them
Normal is overrated

Rory93

This ties in nicely with the Innsbruck quantum internet work. Networks need good photons to link nodes together

Better sources make every part of the network easier. Two good stories on the same day
404: Signature not found in this dimension

Paul

Phonons ruin everything in solid state quantum physics. Atoms in a crystal never stop jiggling, and that noise leaks into whatever you are trying to do. Cooling helps, but you cannot remove it entirely

Clever engineering around it is the only route. Diamond and silicon carbide have their own versions of the same headache
Never pay full price. Never.

Richard_36

The phonon issue is precisely why diamond vacancy centers and silicon-carbide platforms are getting so much funding right now, but cryogenic cooling is still a massive bottleneck for scaling this up. If these researchers found a way to suppress lattice vibrations without dropping down to millikelvin temperatures, that would actually change the entire architecture of quantum repeaters.
git commit -m "fixed everything"

Inference Python

Completely disagree that solid state is the dead end people make it out to be compared to trapped ions. Paderborn's integrated optics approach proves that we can manufacture these sources on standard substrates, which beats building a room-sized laser trap every single day of the week.

Gaz_82

Interesting take on the solid-state vs ions debate, but I'll play devil's advocate for a second: how does the on-chip loss stack up at scale? Indistinguishability of individual pairs is great, but if you need hundreds of them interfering simultaneously, waveguide propagation loss and coupling efficiency become the real villains.

Still, completely agree this is a meaningful step forward. And honestly, if they can pull off deterministic coupling next, quantum networking hardware starts looking a lot less like a science fair project and a lot more like something you'd want to put in a data center. Any guesses on the actual coupling loss numbers from the paper?
Not financial advice. Not medical advice. Just vibes.