Powerful and precise multi-color lasers now fit on a single chip
71 points
9 days ago
| 7 comments
| phys.org
| HN
pjc50
4 days ago
[-]
Actual paper: https://www.nature.com/articles/s41566-025-01769-z DOI https://doi.org/10.1038/s41566-025-01769-z

" We show microcombs with total on-chip power levels up to 158 mW and comb lines with an intrinsic linewidth as narrow as 200 kHz."

150mW is a lot for a single-chip laser, given that the eye safety limit for standard red laser pointers is about 5mW.

reply
pppone
4 days ago
[-]
In astronomy, laser frequency combs are horribly expensive (~$0.5M), but fantastic for calibrating high precision spectrographs. It would be interesting to see if this method could be tuned for that application (namely, shifting to the visible), such to enable cheaper spectrographs.
reply
mmmBacon
4 days ago
[-]
Visible will always be expensive because it’s very niche and low volume. So the techniques here are only practical economically for the large volumes of light sources required for communications. This won’t extend to the visible unless there’s a similarly large market.

The cheapest way I’d think to generate a visible frequency comb would be to frequency double the IR comb laser using a nonlinear crystal like BBO.

Also here the accuracy is relative and not absolute which is fine for communications. The absolute accuracy of the comb may not good enough for spectroscopy in the visible.

reply
casparvitch
4 days ago
[-]
reply
_joel
4 days ago
[-]
"Beyond data centers, the same chips could enable portable spectrometers"

Tricorders ftw

reply
IAmBroom
4 days ago
[-]
OMG, genuine tricorders, and not just some kluge of a few common tools!
reply
danw1979
4 days ago
[-]
How would you modulate the individual wavelengths, considering they all come from the same source ?

I had, maybe naively assumed that laser diodes were switched on/off electronically to modulate a signal. With this laser you’d have to modulate after the light source somehow ?

reply
IAmBroom
4 days ago
[-]
There are no "individual wavelengths", ever, anywhere.

Do you mean modulating multiple bands of light, as a TV does with broad-bands of R, G, & B? Do you mean time-band modulation of a single band, like radios do with AM?

reply
khalic
4 days ago
[-]
You can always filter the frequencies you don’t need
reply
jagged-chisel
4 days ago
[-]
The question remains: how to modulate individual wavelengths.
reply
bartlettD
4 days ago
[-]
Likely some kind of Electro-Optic Modulator. You could use their wavelength comb to separate the light and then use a Mach–Zehnder interferometer to perform On-Off-Keying as an example
reply
fuzzfactor
4 days ago
[-]
Notice the breakthrough was accomplished in a lab which is utilizing more square-footage of "bench" space (/"shelf" space) compared to floor space than most other labs you will find.

Almost like a storage room, except with as much operational, calibrated equipment at the fingertips as the working room would possibly fit.

Regardless of the essential auxiliary storage space having at least 5x the square footage of the working lab itself. Where hopefully at least 20% of the equipment there is operational, if not currently calibrated or in use. Which would then equal the amount in operation in the lab.

If the storage area is down the hall, or maybe in the basement, or a convenient nearby building, the same breakthroughs will be possible by the same researchers.

It will just take more time the further the storage area is, and the more pieces of equipment for which there is no backup in storage.

And way more time if at all, when the storage area is too small to get the job done.

Anything less and you're shooting yourself in the "footage" :)

>Cleaning up messy light

Or cleaning up your messy lab, you can have both, you just have to prioritize what you want to accomplish more of in your lifetime.

reply
PaulHoule
4 days ago
[-]
Reminds me in grad school doing the Fizeau experiment

https://en.wikipedia.org/wiki/Fizeau%27s_measurement_of_the_...

in the Physics 510 lab, the idea is that you send light through a system of mirrors through such a long path that a slot in a rotating disc (or a mirror) can move enough to block the light or maybe not block the light if it can rotate all the way to the next opening. Unlike Fizeau we did it entirely indoors and the experiment depends on empty space.

reply
mojomark
4 days ago
[-]
Ha, I wasn't going to read the article, but I had to after reading this comment. Yikes dude - I hope you don't ever happen upon my messy-ass/inefficient lab!
reply
fuzzfactor
4 days ago
[-]
It's supposed to be obvious I would not be in favor of removing what others often refer to as "clutter" :)

The aux storage areas are not where the overlaid layers of equipment on and around the benches would be as immediately useful. Plus, most importantly the storage rooms are supposed to already be full to the gills so you can hardly walk inside them ;) Containing equipment from which thousands of hours of learning has been gleaned beforehand.

In both, you often need as much stuff squeezed into a small space as possible before you can come near the goal line.

I'll take scientific progress that's good enough to emerge from the lab over a "clinical" appearance of the lab itself any day.

reply
hdjfjkremmr
4 days ago
[-]
could you use this in show lasers? currently they use RGB mixing with electro-acoustical crystals for intensity modulation.
reply
mezzman
4 days ago
[-]
Polychromatic acousto-optic modulation hasn't been used in laser shows for quite some time since RGB diode based laser systems came about. Granted, nothing beats a mixed gas ion laser and PCAOM for the beautiful colors you can get but these days nobody misses dragging around water hoses and sorting out 60 Amps of three phase power to run those old beasts.
reply
brookst
4 days ago
[-]
Maybe? Show lasers are much more than 150mw. Lasers can be combined but I’m not sure the practicality of combining 100 chips to get 15w.
reply
bobsmooth
4 days ago
[-]
This seems like the kind of technology that will quietly revolutionize a lot of things in 10 years when manufacturing is figured out.
reply
pjc50
4 days ago
[-]
It sounds like it's already manufacturable - silicon photonics uses the IC manufacturing process, in the same way that MEMS does.
reply
bobsmooth
4 days ago
[-]
Sure, but can they make 10 million of them? I really hope they can. Tiny terabit transceivers sounds awesome.
reply