Hyperentanglement demonstrated

[ZapperZ]

In the upcoming issue of PRL:

We experimentally demonstrate the first quantum system entangled in every degree of freedom (hyperentangled). Using pairs of photons produced in spontaneous parametric down-conversion, we verify entanglement by observing a Bell-type inequality violation in each degree of freedom: polarization, spatial mode, and time energy. We also produce and characterize maximally hyperentangled states and novel states simultaneously exhibiting both quantum and classical correlations. Finally, we report the tomography of a 2×2×3×3 system (36-dimensional Hilbert space), which we believe is the first reported photonic entangled system of this size to be so characterized.

J.T. Barreiro et al. PRL 95, 260501 (2005).

Zz.

 

[humanino]

Thank you ZapperZ for this very interesting reference.

I found the article on arXiv :
Hyper-entangled photons

 

[Careful]

The paper must be a landmark in history given the fact that one did not succeed yet in making an exclusive Bell experiment (one which puts local realism away forever) for polarization only. Despite this, the authors come up with pleanty more (at least they claim so) :

I only looked quickly at this paper and it seems they (a) do not even discuss the distance scales on which their experiment was done (b) also they do not even mention the efficiency of their detectors and (c) they do not report how the data is manipulated ! It might be that I missed something (and that it is hidden somewhere), but if not, then this paper is not up to the standards reports about Bell experiments should be. I notice that all respectable Bell papers still discuss - at the end when most people gave up reading already and are convinced by the good news in the abstract that there is no problem at all - how their experiment relates to both the locality and detection loophole. These words are not even mentioned in this paper which makes me wonder.

 

[vanesch]

Maybe the aim of the paper was not, this time, to try to convince the local realist crowd, but just to check a quantum prediction ? In other words, a failed attempt to falsify quantum theory, and a further indication where we can use the formalism. There's nothing wrong with that, is there ?

 

[Careful]

Maybe the aim of the paper was not, this time, to try to convince the local realist crowd, but just to check a quantum prediction ? In other words, a failed attempt to falsify quantum theory, and a further indication where we can use the formalism. There's nothing wrong with that, is there ?

No, the appearent lack of information is also important for the QM crowd. As you can easily imagine, the data is necessary to have a chance of impartial judgment whether this was a real confirmation of QM or not. I assume the QM believers stil demand this at least, otherwise I do not see the purpose of Bell test papers.

 

[humanino]

I only looked quickly at this paper and it seems they (a) do not even discuss the distance scales on which their experiment was done (b) also they do not even mention the efficiency of their detectors and (c) they do not report how the data is manipulated ! It might be that I missed something (and that it is hidden somewhere), but if not, then this paper is not up to the standards reports about Bell experiments should be. I notice that all respectable Bell papers still discuss - at the end when most people gave up reading already and are convinced by the good news in the abstract that there is no problem at all - how their experiment relates to both the locality and detection loophole. These words are not even mentioned in this paper which makes me wonder.

I must say that I have not paid deep attention to the paper as well. I remember that Zz used to follow work on the fondation of QM (among other things) with scrutinity. I only aimed at providing easy access for everybody (I wish I had access to PRs archives at home... only institutes can afford).
Your remarks on the poor descriptions of the experimental set-up and data analysis sure are valid, yet maybe in a Letter this is often the case. A rigorous physicist always complains about those descriptions : indeed they are crucial if one wants to reproduce the experiment, but often disregarded as secondary information compared to the result. This is very unfortunate. Yet I also know that very often this information is available for anyone with further questions. The authors can be contacted.
Is not PRL a good reference ?
I used to trust PRL blindly.
Can they not afford competent referees ? (especially with regards to the price institutes for the access to their archives...)

 

[ZapperZ]

I must say that I have not paid deep attention to the paper as well. I remember that Zz used to follow work on the fondation of QM (among other things) with scrutinity. I only aimed at providing easy access for everybody (I wish I had access to PRs archives at home... only institutes can afford).
Your remarks on the poor descriptions of the experimental set-up and data analysis sure are valid, yet maybe in a Letter this is often the case. A rigorous physicist always complains about those descriptions : indeed they are crucial if one wants to reproduce the experiment, but often disregarded as secondary information compared to the result. This is very unfortunate. Yet I also know that very often this information is available for anyone with further questions. The authors can be contacted.
Is not PRL a good reference ?
I used to trust PRL blindly.
Can they not afford competent referees ? (especially with regards to the price institutes for the access to their archives...)

First of all, you should NEVER trust any journal blindly, no matter how prestigious it is.

Secondly, publications such as this is meant to be part of an on-going development in the the field of study. Something is published so that other peers and experts in that field can also study it and verify/confirm/challenge/comment etc. on it. Often, this process will take a very long time. So you should never accept things that are published in research-front journals. You should look at them as preliminary "evidence" or report and wait for the proper gestation period until the community accepts it as valid.

Thirdly, PRL has a severe length limitations. A common practice is to make the report and then follow up with a lengthier and more in-depth report in one of the Physical Review A/B/C/D/E journals.

Finally (and this isn't directed at you), I wish people who continue to whine about such papers will put their reputation on the line and write a rebuttal to PRL in this case. I don't see them doing such experiments with contradicting results, since ALL of these types of experiments so far have been consistent.

Zz.

 

[Careful]

**
Your remarks on the poor descriptions of the experimental set-up and data analysis sure are valid, yet maybe in a Letter this is often the case. A rigorous physicist always complains about those descriptions : indeed they are crucial if one wants to reproduce the experiment, but often disregarded as secondary information compared to the result.**

Such description usually does not take more than one page (for example see the Tittel et al paper). I am not asking the authors to publish in detail all the technical specifications, but the important ones such as the overall distance scales, detector efficiency, assymetry between different dectors, value dependent (that is relative angle between polarizers) detection efficiencies, in either the differences between the raw data and the one actually used to plug in the CHSH expression and the most important theoretical assumptions - such as fair sampling - made should be stated explicitely.

**
Is not PRL a good reference ?
I used to trust PRL blindly.
Can they not afford competent referees ? (especially with regards to the price institutes for the access to their archives...) **

I only trust my own brain ... and I think that referees get a bit overloaded in these days. All bad consequences of the publication pressure.

Cheers,

Careful

 

[slyboy]

Bell-tests are not just tests of local realism

Careful, there is a big difference between experiments designed to test QM against local realistic theories and experiments that are just aimed at making progress towards scalable quantum information processing. Both types of experiment require the generation of entanglement, but in the latter it is OK to perform more limited types of Bell-test than in the former case, since the aim is not to rule out local realistic theories.

It is worth noting that it is extremely challenging to controllably generate entangled states involving large numbers degrees of freedom, even in the case where the particles involved are close together, and particularly when the degrees of freedom in question are not just the polarizations of photons. Bell-tests can be used as an "entanglement witness" in this case, since unentangled states won't violate a Bell inequality no matter how the experiment is embedded in spacetime. I agree that this says nothing about the violation of local realism, but that was not the point of the experiment in this case.

 

[dlgoff]

"...We also produce and characterize maximally hyperentangled states and novel states simultaneously exhibiting both quantum and classical correlations. ..."

This is what caught my eye. I didn't even consider the experiment was "...designed to test QM against local realistic theories...".

It appears that our understanding of entanglement (by doing these experiments) is growing by leaps and bounds?

Don

 

[Careful]

**Careful, there is a big difference between experiments designed to test QM against local realistic theories and experiments that are just aimed at making progress towards scalable quantum information processing. Both types of experiment require the generation of entanglement, but in the latter it is OK to perform more limited types of Bell-test than in the former case, since the aim is not to rule out local realistic theories. **


Sorry, but this is utter nonsense. The ONLY purpose of Bell tests is to discriminate between local realist theories and QM (and even then it is disputable which test one should use). You seem to have missed my previous point that a violation of a Bell tests is not necessarily a confirmation of QM either. Therefore, a paper in which data is communicated through the use of Bell tests without any further comments is useless. It is entirely misinformative (and incorrect) to state: ``oh, I used a Bell test, but you should not interpret its violation as evidence against LR, but as a confirmation of QM´´. As said: both statements are incorrect under many circumstances.


**It is worth noting that it is extremely challenging to controllably generate entangled states involving large numbers degrees of freedom, even in the case where the particles involved are close together, and particularly when the degrees of freedom in question are not just the polarizations of photons. Bell-tests can be used as an "entanglement witness" in this case,**

No, they can't. Why not simply take the raw data and say how close one comes to the QM predictions ? Moreover, even then you should mention the distance scales on which the experiment has been performed. The reader might get the impression that QM has been confirmed to have universal validity on all length scales Moreover, as you should know, the Bell inequalities are entirely meaningless at distance scales below 1/10'th of a metre since the separability assumption does not apply under these circumstances.

** since unentangled states won't violate a Bell inequality no matter how the experiment is embedded in spacetime. **

That depends on wether you use the raw data or not (and on your prior bias when ``analyzing´´ the data). There are well known methods based on data substraction and the fair sampling hypothesis which can be used to violate the Bell inequalities even for LR theories.

Cheers,

Careful