Slow light

[pendell]

YA sensational article which I'm having a hard time making heads or tails of .

An atom at room temperature moves at high speed, but when it is bombarded from three directions by laser beams it loses energy and slows down: it cools off. In a complicated series of stages, Dr. Hau's apparatus uses lasers to cool the sodium atoms part way, and then evaporates the fastest (and therefore hottest) of them, saving the coolest in the trap. At the end of this cooling operation (which takes 38 seconds), the cloud of atoms in the trap has been reduced to only 50 one-billionths of a degree above absolute zero, a temperature far colder than any in nature, even in the depths of space.

Because of the Uncertainty Principle -- one of the fundamental rules of quantum mechanics -- the more precisely the momentum (or velocity) of a particle is known, the less precisely is it possible to measure its position, and vice versa. At exactly zero degrees (which in practice could never be reached), a particle would have zero momentum, a precise value. This would mean that the particle's position would be highly uncertain; it might be found anywhere within a large volume of possible places.

The volume occupied by each ultra-cold sodium atom in Dr. Hau's trap expands enormously, so much so that atoms in the trap are forced to overlap and merge into what physicists call a Bose-Einstein condensate (named for the theorists Satyendra Nath Bose and Albert Einstein), in which the atoms' quantum wave functions are combined.

Once the condensate is created, a ''coupling'' laser tuned to resonate with the trapped mass of atoms is beamed into the trap chamber so that the atoms and photons of light become ''entangled'' with each other, behaving as if they were a single entity. A pulsed laser probe is then shot into the ''laser dressed'' condensate from a different direction, and some of its light passes through, but at a speed 20 million times as slow as the speed of light in a vacuum.

So ... if I'm reading this correctly we can now create a condensate which will drastically increase the time it takes light to travel through it. The implications are, according to the article, that this will greatly improve our computer speeds, keeping Moore's law alive. I'm still having trouble following exactly how this works.

Anyone have any light to shed? :)

Respectfully,

Brian P.

[SiuiS]

It makes heuristic sense to me, but I'm sure as soon as it's dissected into math I'll lose it.

The upshot for me is that we can now use freezing powers like mordin Solus.

[Radar]

To make a typical computer, you need a basic mechanism resambling a transistor in working: you need to be able to switch some flow on or off by turning on or off another flow. Here we have a mechanism that significantly delays light beam by use of another light beam. If we split the delayed beam before it enters the condensate and then merge the two parts (one going through a constant path outside the condensate), then the interference between the two would depend only on what happens in the condenste. We can tune the whole thing so that the resulting beam is on, only when the steering beam is for example on or, if we want, the other way around. This is a direct copy of a transistor (analog characteristics would be somewhat different, but we won't need those for digital circuits). One of the key problems with transistors is, how fast they can switch from one state to the other. As I understand it, condensate-based optical switches are much faster.

Aside from that, slowing light down to a crawl is being researched for quite a few years with a lot of interesting results.

[Eldan]

How would that increase computing speed? If the light moves slower, wouldn't that mean using it to transmit information would also be slower?

Edit: nevermind, TBAMs explanation makes a lot of sense.

Also, any matter slows down light a bit when it goes through it. That's normal

[Knaight]

Also, any matter slows down light a bit when it goes through it. That's normal

Sure, but the amount involved is an outlier. The speed of light in a material is c/n, where n is the index of refraction. That would be 20,000 here, which is extremely high. It might not be high for Bose-Einstein condensates, but it's extremely high compared to matter in the much more common phases.

[the_druid_droid]

I'm actually kind of surprised this is only being reported now. As far as I recall, the original B-E condensate experiments that measured properties like the speed of light in the medium were years ago (enough so that several of the grad students doing that work are now professors themselves).

I suppose it might be that they're only now proposing a technological application for them, but I suspect it's very much a wait-and-see situation. The condensates aren't exactly trivial to produce, let alone scale up (or down) to real computing applications.

EDIT: And the article itself is a bit odd. Why, for example, does it feel the need to discuss Dr. Hau's marital status?

[Knaight]

EDIT: And the article itself is a bit odd. Why, for example, does it feel the need to discuss Dr. Hau's marital status?

Part of the reason is that the new york times is a bit of a mess, part of it is that the author decided to write a lifestyle article (see: "bit of a mess"), a large part of it is that Dr. Hau is a woman and as such low grade media outlets are going to find a way to insert marital status and prominently feature a few lines on appearance regardless of subject matter.

[Sith_Happens]

So I guess I'm the first one to notice that the article is dated 1999?

[RebelRogue]

As noted this is an old article. I've had the pleasure of seeing Dr. Hau lecture on her research. That was a few years ago, so I'm not up to speed on her results (or an expert on the subject at all, really). But iirc the low speeds are achieved by fine-tuning dispersion relations of the photons. It's actually possible to stop the light completely, which basically means that the BEC can store information about photons. As long as the coherence isn't broken this information could in theory be stored in the BEC indefinitely.

[MLai]

I was wondering how old the article was. Was thinking, "Wait, we've had BEC for a while now, and only now they play around with its properties??"
Question: How does shooting lasers at something cool it down?

part of it is that the author decided to write a lifestyle article (see: "bit of a mess"), a large part of it is that Dr. Hau is a woman and as such low grade media outlets are going to find a way to insert marital status and prominently feature a few lines on appearance regardless of subject matter.

Hey don't knock it. I was curious about her marital status, age, ethnicity, and looks. The article answered 3 out of the 4 for me.
(I was curious about Brian Greene's marital status as well. Intelligence is sexy.)

It's actually possible to stop the light completely, which basically means that the BEC can store information about photons. As long as the coherence isn't broken this information could in theory be stored in the BEC indefinitely.

What, seriously?

[Yora]

From what I understand, the laser apparently cushions the movements of the atoms. Like a small vibrating piece of metal stops vibrating when you touch it.
No idea if that's how it actually works.

An interesting thing about slowing light is, that every photon always moves at the speed c. If light appears to move slower, then it's the result of photons hitting an atom, getting absorbed, the atom returning to its previous state, and a new completely identical photon being emited. Light is always traveling at c, but when it travels through a medium, it makes short stops that make it look like the light goes slower.
I think any photon created in the center of the sun takes several million years of those chain reactions (on average, because quantum physics) before it reaches the surface and flies out into space. Simply because there is such an immense number of atoms at incredible densities inside a star so that a photon can't travel far before it hits something again.

In a Bose-Einstein-condensate, a photon could potentially be captured by a supercool atom and stay there indefinitly without a new photon being emmited. Until you use a laser to give the atom a bump.

[Sith_Happens]

Question: How does shooting lasers at something cool it down?

The trick is to use a laser just below one of the absorption frequencies of the gas, such that any particles moving towards the laser emitter absorb photons due to blue-shifting while the rest are unaffected. Because the absorbing particles are always moving opposite to the photons, conservation of momentum dictates that they slow down.

[Yora]

That is both genius and also very weird that I think I understand it.

[Radar]

Laser cooloing is really neat technique for cooling gases. The key idea relies on the fact that atoms can absorb only photons of specific frequencies. If light is out of tune, atoms won't be interacting with it much. Therefore you tune the laser so the light frequency is slightly lower then that desired by atoms, which you try to cool. Then you use 6 such lasers: two for each cartesian direction pointed toward each other and all focusing on one spot, where the atom gas is traped. When some atom goes in a given direction, it sees the photons coming from that direction as having higher frequency due to Doppler effect. Therefore the atom is more likely to absorb a photon going in an opposite direction then the atom did. This in itself shaves off some of the atom's momentum Soon enough it emits the photon back in a random direction and is again most likely to absorb another photon, which would diminish atom's momentum. In the end, you can almost stop all the atoms up to a momentum of a single photon used (due to the inevitable emission).

tl;dr Laser cooling is often called optical molasses and it works kind of like that.

An interesting thing about slowing light is, that every photon always moves at the speed c. If light appears to move slower, then it's the result of photons hitting an atom, getting absorbed, the atom returning to its previous state, and a new completely identical photon being emited. Light is always traveling at c, but when it travels through a medium, it makes short stops that make it look like the light goes slower.

That's not entirely accurate, since light does truly slow down in various materials even without scattering (you wouldn't get Cherenkov radiation otherwise for example). Typicaly, when photon gets absorbed and then it is emitted, it won't be exactly in the same state as before (different direction and phase). Also due to the fact that that particular photons are absorbed and emitted at random times, so the whole beam loses coherence. Therefore, for optical materials, you want to minimise the likelyhood of absorption.

I don't know the process for such extreme light slowing or stoping, but it has to be a conversion of light into some combined state of atoms and radiation like polaritons or some other quasi-particles.