Talk:Quantum entanglement/Archive 3

Experimental proof / scientific consensus

So a few weeks ago, I was listening to MIT Prof. Seth Lloyd on NPR explaining the experiment behind this promo text:

"New research into quantum entanglement by a group of researchers in Austria reveals that two photons can be entangled -- in a strange linked state --without a connection between them, even when the photons in question are on opposite sides of the Danube."

The paper was published in *Science* in 2003 - Aspelmeyer et al., Long-Distance Free-Space Distribution of Quantum Entanglement. You can find it on sciencemag.org, but registration is required. Prof. Lloyd was certainly under the impression that quantum entanglement is real. A free popular-press summary: http://physicsweb.org/article/news/7/6/20/1

Digging a bit further, I also found this press release

February 26, 2003

Michigan researchers achieve quantum entanglement of three electrons

ANN ARBOR, Mich. - The quantum entanglement of three electrons, using an ultrafast optical pulse and a quantum well of a magnetic semiconductor material, has been demonstrated in a laboratory at the University of Michigan, marking another step toward the realization of a practical quantum computer. While several experiments in recent years have succeeded in entangling pairs of particles, few researchers have managed to correlate three or more particles in a predictable fashion.

The results were presented in an article on Nature Materials' web site on February 23 and will appear in the March 4 issue of Nature Materials, titled "Optically induced multispin entanglement in a semiconductor quantum well." Authors of the paper are Jiming Bao, Andrea V. Bragas, Jacek K. Furdyna (University of Notre Dame), and Roberto Merlin.

You can find this article on nature.com, but again, registration is required.

These purported demonstrations of quantum entanglement are more recent (2003) than the experiments in the 1970s and 1980s which the article Clauser and Horne's 1974 Bell test currently discredits. My preliminary sampling of professional opinion suggests to me that the peer-reviewed physics community is of the opinion that quantum entanglement has been convincingly demonstrated. I therefore dispute the claim made on Talk:Clauser and Horne's 1974 Bell test that "[t]here is as yet no convincing evidence that quantum entanglement occurs." I invite discussion of the 2003 experiments and further investigation of peer-reviewed articles and the opinions of professional scientists and engineers, to resolve the dispute. -- Beland 02:41, 17 Aug 2004 (UTC)

I'm not sure that there is not much dispute over the experimental results. It's pretty clear that there no generally experimentally accepted results that contradict QM. You can find lots of loopholes by which you can argue that local realist models are still correct. However, it is also the case that any local realist models are going to have to deal with explaining the phenonmenon that looks like quantum entanglement.

Roadrunner 04:43, 17 Aug 2004 (UTC)

Except for the conflict between QM and General Relativity, I think you mean, no? -- Beland 01:23, 19 Aug 2004 (UTC)

That's theory and not experiment. The current state of place is that there are no known experimental results that conflict with QM. There are also no known experimental results that conflict with GR. Now it is true that at high enough energies, QM and GR conflict with each other. Unfortunately those energies are so high that we can't do an experiment.

Roadrunner 03:51, 21 Aug 2004 (UTC)

This is a difficult situation. I know full well that the physics community will tell you that quantum entanglement has been experimentally confirmed, but I have more complete information on the subject than most.

I think its necessary to distinguish the phenonmenon of "quantum entanglement" from quantum mechanics If you define it as the the phenonmenon by which the observed state of one particle is correlated with the observed state of a distant particle, that is definitely been observed. What isn't completely solid is that QM or something mathematically equivalent is the only possible explanation for the phenonmenon.

Roadrunner 03:51, 21 Aug 2004 (UTC)

I know, for instance, that the tests used in these modern experiments have the same loopholes as the CHSH one. They are not valid since they depend on the assumption of "fair sampling". To see why this assumption is not reasonable, I do urgently suggest that you look at my 1996 paper or one of several on my web site or the quant-ph archive, e.g. http://freespace.virgin.net/ch.thompson1/Papers/TheRec/TheRecord.htm/ or http://arxiv.org/abs/quant-ph/0210150 . I need to put my basic "Chaotic Ball" diagrams in wikipedia.

Meantime, as Roadrunner says, local realist models need to explain what is observed, which does, I admit, look remarkably like the text book "entanglement". This can be done, but requires a different model of light and a different way of modelling the output from a nonlinear crystal in "parametric down-conversion" (PDC). The only paper I have on the subject, though, is my own unpublished one, at http://arxiv.org/abs/quant-ph/9912082.

I have been discussing these issues with the professional community for the past 10 years! If you notice, Shimony (of CHSH fame) references my 1996 paper in his new article at http://plato.stanford.edu/entries/bell-theorem/ . He does not anywhere claim that there is convincing evidence. He is currently placing his money on Fry et al's ideas for a "loophole-free" experiment. Tittel referenced one of my quant-ph papers in one of his articles on long-distance experiments in Geneva. But my ideas on PDC do come under the heading of original research, so I realise I can't mention them in the articles here. A pity! Some day, when the community eventually realises what the basic fair sampling loophole is all about, they will start looking again at what really happens and classical ideas on the behaviour of light resurrected and updated.

Actually, I think that what will turn heads is an experiment which produces an unambigious result that cannot be explained by QM. Since local realist models are incompatible with QM, then there should be a way of producing an experiment which conflicts with QM. I have a suspicion that this experiment probably has nothing to do with Bell's test. Also, I have a strong suspicion based on history that if this happens it will start with something weird experimental result that no one has any idea is connected with this issue.

What I was planning to do was after you get done with the Bell test loopholes, to point out why despite them, the physics community still largely thinks that QM is correct.

Roadrunner 03:51, 21 Aug 2004 (UTC)

Roadrunner 03:51, 21 Aug 2004 (UTC)

Caroline Thompson 08:52, 17 Aug 2004 (UTC)

Bell test loopholes

I've added a couple of links re Bell test loopholes recently. Too much has been written about quantum entanglement on the assumption that the experiments have confirmed it convincingly. Re the apparent (limited) success of applications of entanglement, some readers might be interested in an article I've archived [1] that has yet to find a publisher. It seems likely that the real world is producing something useful by way of real correlations of phase and frequency but not otherwise taking much notice of accepted theory. The net result is better described by classical wave theory, adapted to allow for the behaviour of modern beamsplitters and photodetectors. Caroline Thompson 08:42, 1 Jul 2004 (UTC)

Are you saying that quantum mechanics is incorrect? -- Tim Starling 08:50, Jul 1, 2004 (UTC)

As far as the experimental evidence is concerned, there is at present nothing to choose between QM and local realism. Popular accounts that claim local realism has been ruled out by, e.g. Aspect's experiments, are wrong. Which theory is actually right is best tackled, I think, not by searching for "loophole-free" experiments (which several teams are doing) but by investigating more carefully which model fits the facts best in existing setups, when the conditions are modified slightly. If the local realist model (the basic form of which has been known all along) turns out to be superior, correctly predicting the way the coincidence curves will change when you vary, for instance, detector efficiency, then it will have been shown that QM does not work for these experiments. However, the determined quantum theorist will be able to continue to believe that it works for smaller systems. The experiments are in fact macroscopic. (Caroline Thompson 17:21, 1 Jul 2004 (UTC))

---

I toned down Reeh-Schlieder remark, as I see no much beef in the Reeh-Schlieder theorem for quantum entanglement. RS simply states that the non-local effects aren't exactly zero but vanish exponentially with distance. And it is a very general statement about all states and local operators. QE in QFT would need larger correlations for a smaller set of states. Pjacobi 09:33, 9 Jul 2004 (UTC)

General Comments

Perhaps some kind soul could write a non-formal section explaining how particles become entangled and some of the interesting effects of entanglement. I am sorry to say that this article is incomprehensible to those of us who are not students of physics or who have had but a single year of college physics. I am also sorry to say that it seems to have gotten less comprehensible over the last year or so.

I realize that part of the reason I find this article so difficult to understand is the result of precision. Unfortunately, that precision has become excruciating.

I'm afraid the reason the page is incomprehensible is that it tries to explain something that does not actually happen! It reproduces some of the QM formalism, which, at the end of the day, is not based on any physical assumptions about how the entangled particles could possibly stay entangled. The fact is that there is no experimental evidence for entanglement that holds water. No actual experiment has ever violated a Bell test in "loophole-free" conditions, and every experiment I've looked at (which is quite a number) has had a fairly obvious alternative explanation in terms of ordinary variables, such as polarisation direction or phase difference, that are simply shared by both photons. Yes, lots of experiments reproduce the QM prediction, but, since local realist models can do the same (within the limits of experimental error), does this prove QM right? See Bell's theorem and linked pages, especially Bell test loopholes, Bell test experiments and Local hidden variable theory.

Some of the applications of entanglement may well bear fruit, but this does not prove the theory behind them is correct. They in fact all depend on some of the classical properties of light such as interference, together with some interesting correlations in the light output by Spontaneous parametric down conversion (the process whereby you get two lower-frequency photons output from a nonlinear crystal pumped by a laser). The apparently wonderful correlations seen are the effect of the combination of these physical properties with some algebraic, logical and geometrical facts related to the behaviour of ratios of counts when the things being counted have been selected by chosing only coincident pairs.

I'm not surprised that a year of a physics course is no help in understanding the page! A lifetime would not be enough to truly understand entanglement. The human mind is only capable of understanding something logical. It can learn the "facts" of QM by rote, but Bohr was merely stating the truth when he said that, effectively, if you think you've understood it you're wrong.

Anyway, I'm afraid that as things stand the only way to find out (some of) what really happens is by reading some of the articles on my web site or on the quantum physics archive. Several now cover the subject of the |Chaotic Ball model, which is, as even some experts agree, probably the best way to illustrate the above-mentioned logic, geometry etc.. No physics course covers the local realist approach. You will be lucky if they even mention loopholes.

Caroline Thompson 09:02, 1 Sep 2004 (UTC)

It's my understanding that the current mainstream scientific consensus is that quantum entanglement does happen. Have you looked at the 2003 experiments I referenced? -- Beland 01:18, 3 Sep 2004 (UTC)

I agree, BTW, that a layperson-friendly explanation is both possible and needed. -- Beland 01:20, 3 Sep 2004 (UTC)

---

I've looked at one of the press releases you gave, re

Jiming Bao, Andrea V. Bragas, Jacek K. Furdyna and Roberto Merlin, "Optically induced multispin entanglement in a semiconductor quantum well", March 4 issue of Nature Materials.

The experiment looks fascinating, but they seem to be using the word "entanglement" to mean just ordinary correlation. There is no mention of having done a Bell test to establish it. Instead the press release says:

"In the experiments, the signature of entanglement involving m electrons is the detection of the mth-harmonic of the fundamental Zeeman frequency in the differential reflectivity data."

But at the end of the day it will be just such ordinary correlations that will be used in what they will call "quantum computing", which will never really use the kind of entanglement that can infringe Bell inequalities since this kind simply does not happen. We're at an impasse, I'm afraid. I should love to re-write the page but my version would not be about "quantum entanglement" as quantum theorists understand it. Their version exists only on paper and in their formulae.

Incidentally, I've read: M. Riebe et al, “Deterministic quantum teleportation with atoms”, Nature 429, 734-737 (2004) which was in the news recently. Here they establish entanglement by measuring "fidelity" which, they say, cannot exceed 66.7% under local realism. What, however, is one to make of this?

" ... obtained fidelities range from 73% to 76%. Teleportation based on a completely classical resource ... yields a maximum fidelity of 66.7% ... Note, however, that to rule out hidden variable theories, a fidelity in excess of 0.87 is required ..."

So they've established entanglement but not ruled out local realism! That's nonsense. I have not been able to find out quite what "fidelity" is, but judging from the context it seems to be much the same as "visibility", and tests based this depend on the assumption that the system obeys Malus' Law. I suspect that they never check this thoroughy. Caroline Thompson 09:13, 3 Sep 2004 (UTC)