Hacker News new | comments | show | ask | jobs | submit login
Physicists ‘squeeze’ light to cool microscopic drum below quantum limit (sciencebulletin.org)
88 points by upen on Jan 12, 2017 | hide | past | web | favorite | 27 comments

arXiv: https://arxiv.org/pdf/1606.08795.pdf

By the way, this group achieved ground-state cooling of a very similar mechanical oscillator back in 2011.

"Sideband cooling of micromechanical motion to the quantum ground state" http://www.nature.com/nature/journal/v475/n7356/abs/nature10... arXiv: https://arxiv.org/abs/1103.2144

Both oscillators were of roughly the same frequency (10 MHz) and size (~15 nm diameter aluminum disc). In terms of phonon occupation number, things improved from 0.34 to 0.19.

They emphasize in this new paper that they were able to get colder by using a squeezed vacuum state of laser light. I'm no expert, but I think the use of this technique is much more important than the improvement in the mechanical state they were able to achieve.

What does this mean to the layman? Is physics broken now?

Physics is not broken.

--not 100% accurate, but a close approximation in lay terms--

If you want to measure the position of something precisely, you can bounce a laser off of it and time how long it takes the light to return.

If you want take this experiment to it's extreme in sensitivity, you eventually end up measuring tiny changes in the phase of the light bouncing off of the object. Your ability to measure position is limited by your ability to measure phase.

Quantum mechanics however requires there to be noise both in the amplitude and phase of the light. This is heisenberg's uncertainty principle. So measuring at the "quantum limit" means measuring fluctuations in phase equal to half of the uncertainty limit (half noise in amplitude and half in phase).

There is no law that said half the noise has to be in phase, just that the product of amplitude and phase noise is greater than or equal to the heisenberg limit. You can take extra noise out of phase and put it into amplitude to measure "beyond the quantum limit" without breaking any rules. This is called a "squeezed state."

That is exactly what they did. They created "squeezed light" and used it to measure a degree of freedom (like phase) "beyond the quantum limit"

This was a very challenging and impressive measurement. John Teufel has been doing work leading up to this for over 7 years (2 post doctorates and a professorship). He worked on quantum computers as an undergraduate too so you may even be able to add a few more years onto that.

Wow, he's a devil of a scientist!

No, it's a pretty traditional technique to "beat" quantum mechanics by using squeezed light states. Short version: you don't really beat quantum mechanics, you just shift where the uncertainty resides.

So where does the uncertainty move to?

To a different subset of the Hilbert space. See here for beginners: http://www.squeezed-light.de

I've often read that absolute zero (0 K) is impossible to obtain, but never really understood whether it's just for practical reasons or if there's actually a theoretical background.

As far as I understand, there's no reason why a particle having 0 kinetic energy can't exist. Also, achieving it seems theoretically possible when a particle has kinetic energy equal to one photon, then emits a photon. Obviously it's hard to experimentally produce such a situation, but not theoretically impossible.

So, would 0 K theoretically be possible or am I missing something?

Temperature doesn't really make sense for one particle. In fact any one particle will have 0 kinetic energy in some frame of reference.

The third law of thermodynamics actually states that there's no systematic way that you can reach 0K, although I haven't found a particularly thorough explanation why. In general it seems to boil down to the fact that no matter which parameter you try to tune there will always be some entropy remaining, so you can't reach T=0 (zero entropy) in a finite number of steps.

And then you have quantum mechanical effects that prevent a particle from actually having exactly 0 kinetic energy.

Well, Heisenburg uncertainty becomes more relevant the colder you get. In this case, [uncertainty in position] * [uncertainty in momentum] >= ħ/2, so something with mass but no kinetic energy would be everywhere in an infinite universe with equal probability.

Ah, this definitely makes sense, so it's at the least not observable.

Another thing crossing my mind is whether Heisenberg uncertainity becomes relevant for these kind of experiments. I think we're talking about ~10^-7 K, thus ~10^-30 J mean energy. For a hydrogen atom (1 u), that would be roughly p = ~10^28 Ns. Hence, uncertainity of location would be h/p ~= 10^-6 m, i.e. in micrometer range, which seems quite large. Does this make sense?

Edit: It's aluminium in the experiment (27 u), but the numbers should be in the same ballpark.

Sounds about right to me. Wikipedia's list of Orders of magnitude (temperature)* shows atomic waves being coherent over centimetres at 10^-15 K, and a 10^8 difference in temperature is a 10^4 difference in momentum — which fits μm to cm.

* https://en.wikipedia.org/wiki/Orders_of_magnitude_(temperatu...

For those curious of possible applications:

Sensors would become more sensitive. You can store information longer. If you were using it in a quantum computer, then you would compute without distortion, and you would actually get the answer you want.

Fun fact: LIGO's next big upgrade will be to install a squeezed light source to reduce quantum noise at higher frequencies.

This all sounds very delicate and refined but doesn't LIGO use a 1MW laser? Will this be equally as powerful?

Squeezing is independent of the laser power used, so that's not really relevant here. Squeezed light is injected at the output port, where the photodetectors are. It is injected in the opposite direction to the classical light leaving the interferometer, where it mixes with the light field in the interferometer and has the effect of reducing quantum shot noise of the classical light that leaves.

In any case, I think you've confused the circulating light power with the input laser power. The circulating light power will eventually reach something like 1MW (currently it's lower) thanks to optical cavities, but the input laser power is actually only 50W (currently), eventually 125W.

It's a beautiful image, I hope it's not an illustration.

From the article: "squeezed light removes the generally accepted cooling limit". Generally removing a limit in physics is a huge deal. What was this limit based on? Has some fundamental quantum physical property been disproven? Is this a breakthrough more significant than setting a coldness record? Was the generally accepted limit based on what would be "really tough" or something else? Anyone know?

> What was this limit based on?

It's base on an assumption that the cooling apparatus isn't using squeezed light. If you can squeeze light, then the limit no longer applies.

> Was the generally accepted limit based on what would be "really tough" or something else?

Squeezed light is experimentally difficult to produce. But the definition of "squeezed" is defined as a particular, precise mathematical property of the state of light generated. So it was not some arbitrarily level of "toughness".

An analogy: Historically, there weren't many practical ways to cool things besides putting it along side things that were already cold and allowing them to transfer heat. (This is why ice harvesting existed: https://en.wikipedia.org/wiki/Ice_trade ). Therefore, there was a "limit" to how cold you could easily make something, namely, the temperature of the coldest natural stuff you had access to. But once you can build cheap refrigerators, the limit disappeared, and you could easily cool things much colder than ice.

> Has some fundamental quantum physical property been disproven?

No. Squeezing light doesn't violate any laws of physics.

> Is this a breakthrough more significant than setting a coldness record?

Coldness records themselves are almost never intrinsically important. They just make good popular press releases. So the technique is more important than the coldness record, but it's not necessarily that important.

> Coldness records themselves are almost never intrinsically important.

Temperatures below 0K (i.e. you need to heat a system to have 0K back) are very important to physics. They will allow to understand vacuum much better than today.

I don't think you know what a negative temperatures is. The wikipedia page is pretty good.


I clearly understand what negative temperature is: if somebody squeezed an amount of energy out of a system, it's temperature drops. If somebody squeezed energy from a system at arbitrary zero point at temperature scale, then system temperature is below zero: system needs energy to return back to zero. Zero point at Kelvin scale is chosen when a system has no kinetic motion. But system at 0°K still has lot of energy, which can be released. I'm talking about energy of "hot" vacuum, which causes Casimir effect. This energy can do some mechanical work even at 0°K, so system still have some energy.

Your comment makes many confusing/wrong statements, but in particular this

> Zero point at Kelvin scale is chosen when a system has no kinetic motion.

is unambiguously wrong. The idea that temperature is monotonic function of kinetic energy is only an approximation to the fundamental definition of temperature, and that approximation isn't valid for systems that have negative temperature.

Definition of Absolute Zero from Wikipedia:

> Kelvin scale is an absolute, thermodynamic temperature scale using as its null point absolute zero, the temperature at which all thermal motion ceases in the classical description of thermodynamics.

> Absolute zero is the lower limit of the thermodynamic temperature scale, a state at which the enthalpy and entropy of a cooled ideal gas reaches its minimum value, taken as 0.

> In the quantum-mechanical description, matter (solid) at absolute zero is in its ground state, the point of lowest internal energy.

It's still possible to squeeze some energy from a system past Absolute Zero as defined using Casimir effect, so it's possible to reach temperature below 0K.

I know, that current interpretation of negative temperatures at Kelvin scale is different, but definition of temperature is different in that interpretation too, so let ignore it.

> in the classical description of thermodynamics.

The "classical description of thermodynamics" is only an approximation to the fundamental definition I mentioned.

I'm sorry, but I won't be able to continue this conversation.

How can something have a fraction of a quantum of energy? I thought energy levels had to be discreet integer multiples of Planck's constant? The more I read about Quantum Mechanics, the less I understand!

(I don't know a lot about QM)

I think that only applies for particular bounded situations (like electrons). There's nothing inherent in the wave function that says energy is discrete multiples; it's only due to the boundary conditions in the electron (i.e. the energy at r=0 and r=inf IIRC) that forces particular solutions.

The analogy (it's actually quite similar in the differential equations) is a string. If you hold both ends only certain solutions are possible; standing waves where the length of the string is a multiple of half the wavelength. But if you don't, anything is possible.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact