'Optical fibre' made out of thin air | Hacker News Reader

'Optical fibre' made out of thin air

cbc.ca

So, I do a little bit of related research in this area:

This is typically called a "filament", a "plasma channel", and sometimes described as a waveguide. It requires high powers (~10^13 W/cm^2, usually femtoseconds in duration) depending on gas. It essentially works because the index of refraction has two nonlinear terms:

    n(ω) = n_0 + n_2 I - n_plasma

the n_2 I makes it so that higher intensities (which occur closer to the center of the filament for a typical gaussian laser beam), have a higher index of refraction, causing the field to focus, increasing the electric field intensity at the center. However, when the field gets high enough, it ionizes the material, causing a plasma to form, which has a negative contribution, defocusing the filament. These two contributions balance out and cause a "waveguide" where the light essentially travels along a narrow channel, allowing the focusing pulse to travel longer without as much spatial dispersion.

You can't send another light beam along these channels, because after the beam has left, all that is there is a plasma channel, which defocuses and absorbs light instead.

This has been around for a while.

What is NEW here is that they modified their beam profile to have 4 or 8 or even an annular ring of these plasma channels. Accoustic waves propagate toward the center of this filament arrangement causing a change in the density (which changes the index of refraction, the index of refraction increases linearly with density) that makes the channels act like a waveguide for another light source. This source of the density change has a duration of microseconds. When the acoustic waves have dispersed, a thermal gradient has a similar effect, and this has a duration of milliseconds.

They used this method to channel light from a spark source while mostly maintaining the spectra of the spark.

Now, what this allows, or doesn't allow:

LIDAR is a possible use, but whatever you point this source at is going to have to be able to take 10^13 W/cm^2... and not a lot of materials can take that without breaking down. While there is good focusing and guiding of the light, I think single photons are probably going to have a high probability of being absorbed (the mode of the photon will overlap into the plasma channels, which will likely absorb it), The nice thing is that signals could likely travel through one of these channels (though there likely will have to be some dispersion compensation at the receiving end).

I feel like I'm missing something important, but if laser photons naturally scatter over distances then wouldn't that issue apply to the laser being used to create the compressed air channel?

The paper goes into more detail.

Basically, if you shoot a pulsed laser of sufficiently high power, you can generate a phenomenon where the laser's interaction with air will basically cause it to self focus - resulting in a filament of plasma that traces out a length several times longer than the normal distance of the laser to double in cross section area.

By using this effect to generate waveguides, the paper specifically singles out utility in two situations (laser induced breakdown spectroscopy and LIDAR), focusing heavily on the first. In laser induced breakdown spectroscopy, you shoot a laser at a remote target, basically blowing it up, and reading the spectrum it gives off as it blows up. It is sensing this spectrum that the paper demonstrates the technique on. By providing the return signal with a waveguide, it is in theory possible to get a much better signal, at further distances.

As for using this technique to help transmit other lasers? I dunno if it makes sense or not. I kind of expect (speculating out of my ass here), at the crazy interactions that let you create the self-focusing/propagation effect also makes it difficult to encode a signal into the laser, so it still might make sense to use this technique to guide a lower power, normally modulated laser signal... maybe.

LIDAR -- that's related to driverless vehicles. I wonder if this might have implications in that space?

Little or none. LIDAR for road driving applications is hardly limited by range or resolution at this point. The problems are cost (which this does nothing to help with - it requires MORE lasers), and dealing with rain/snow/dust, which this also does nothing about.

If you don't mind the vehicle scorching every surface it passes then it might have applications in that space. Otherwise, not so much.

Anyone has any idea of how long such a structure will stay in thin air ? Because to me it seems like a proof of concept but it wouldn't be viable : you need to keep the "fiber" open long enough to transmit data.

Since it's thin air, the soundwave created at the beginning will just continue and the effect would dissipate.

The article mentioned microseconds... so not long, but possibly long enough, if you could keep the pulses going and not cancel each other out, but it sounds like this technique still has a long way to go. And carrying data may not be the ultimate end-point of this particular technique.

This could have applications in free space quantum encryption links. Signals sent in quantum communications must be at the level of single photons to ensure security. Traditionally, high repetition rates are used to overcome loss. i.e. If you have a 1 in a million chance of getting a single photon through, then trying a million times a second lets you get one photon through per second, on average. Increasing the transmission of a long free-space link could increase the maximum distance at which the link will be usable. This may be especially useful for satellite-to-ground links in less than perfect atmospheric conditions. Very cool research!

It would have limited application in satellite-to-ground links, as the photons would soon be outside the atmosphere. It is a neat idea though.

Loss happens in atmosphere, not vacuum. If it increases the overall transmission the link, what does it matter if the waveguide doesn't go the full distance?

I agree that it will help, to the extent that it counters diffraction of the beam whilst is in the atmosphere, hence it having "limited" application rather than no application.

Diffraction still happens in a vacuum. The beam spreads, so photons are lost by not falling on the detector.

The waveguide effect would counteract scattering to some degree. Even then, wouldn't scattering emit photons in random directions, meaning that the majority of scattered photons would be moving at too great an angle to the high/low density boundary to be refracted back into the beam?

Outside the atmosphere there is no scattering so the laser will stay as focused as it was at the start

> This may be especially useful for satellite-to-ground links in less than perfect atmospheric conditions.

Seems to me that the ionosphere would get in the way of this kind of technique. This relys on creating a plasma in air with controlled properties to function as a waveguide, so I imagine that having a high-energy space plasma in the way would make that sort of thing very difficult.

I would be interested in looking at applications with applied electric/optical fields.

Could you shape the plasma after it was created, shaping the waveguide?

Say they manage to sustain these channels over distance, and figure out the pulse timing to be usable for a relatively sustained link...

What happens on a snowy day? (or other weather types with lots of interference... hazy/dusty days, etc).

What effect would a stiff wind have on this?

What effect would it have on a person walking through the plasma creation beam?

The connection will fail if your destination moves out of sight?

I have wondered if such a thing would be possible in the interstellar medium for a sufficiently powerful laser.

Reply on news.ycombinator.com