Was reported here. Of course if this is real it is very exciting. It leads me to the question: given that it took so long to find this resonance at a meager 38 MeV, is it possible that all SUSY particles are hiding down in the MeV or KeV range (or lower)?

  • $\begingroup$ "Non-SM resonance"? That's a religious statement. The SM includes QCD, and one has absolutely no clue, not even from the lattice, what happens at 38 MeV. $\endgroup$ – user11581 Aug 22 '12 at 13:57
  • $\begingroup$ I thought the signal model you considered was a conformally symmetric one (here is the theory paper I think: springerlink.com/content/n22j38w2l6885m75/fulltext.pdf), ie a non-SM one. That said, I'm no expert. Maybe someone else can chime in regarding the possibility of a 38 GeV bosonic particle being consistent with QCD? $\endgroup$ – user1247 Aug 22 '12 at 14:11
  • $\begingroup$ In fact, there are nonperturbative inequalities among the masses of various states (for a review see arxiv.org/abs/hep-ph/9911532), which taken together constitute an extremely strong case that the pion is the lightest state in QCD. $\endgroup$ – Matt Reece Aug 22 '12 at 14:58
  • 4
    $\begingroup$ The "undershoot-then-overshoot" pattern that is consistently present in figures 2a,b and 3a,b is a classic sign that you may have a energy shift between the two bits being subtracted. If a grad student brought me these plots I'd ask them to introduce a couple of artificial shift to see if they could get it to get better and or worse. Then we'd go hunting for possible misunderstandings of the energy of the two distributions. And I would pray that we were doing a blind analysis, because this is exactly the kind of place where human bias can screw things up. $\endgroup$ – dmckee Aug 22 '12 at 15:32
  • 1
    $\begingroup$ @GeorgeRupp: I don't need to ask a guru-- there is no way QCD can have a 38 MeV resonance. The confinement scale is 1GeV, and there is no glue excitation lighter than this, it is almost a theorem. $\endgroup$ – Ron Maimon Aug 25 '12 at 21:30

It is not possible that SUSY particles are hiding in keV or MeV range. In particular, there can't be any new charged particles (and similarly new color-charged particles) that would be this light because they would be easily pair-produced and easily detected.

The first (February 2012) claims by different authors (the original ones, Rupp and van Beveren, who made the conjecture) were refused by the COMPASS collaboration (which was used as one of the main pieces of "evidence") here:


COMPASS says that the patterns that attracted the attention or Rupp and van Beveren are due to $\pi^0$, $\eta$, and secondary interactions in the COMPASS spectrometer. Rupp and van Beveren responded that the COMPASS critique is internally inconsistent. It seems more likely to me that COMPASS is right.

The newest Russian experimental paper looks strange to me. For example, it never quotes any confidence levels, as far as I can see, and instead says that there are "almost no errors" in their measurement, a claim that it easily refuted by looking at their chaotic wiggly charts.

An extended discussion may be found on my blog.

  • 1
    $\begingroup$ There is an error analysis towards the end, they simulated the backgrounds by Monte-Carlo, but I agree that the analysis is wanting. I think they just rushed to confirm. Also, their system is messy. Is it possible to exclude diphoton decay of excited nuclear states at about 38 MeV? I mean, they claim to see the same bump in Cu and C, which is unlikely for a nuclear effect. Is there any scenario where you have an electrically neutral boson at 38MeV? I can't imagine how it would be missed. I am siding with COMPASS like you, but I want to be sure. $\endgroup$ – Ron Maimon Aug 22 '12 at 7:31
  • 1
    $\begingroup$ Agreed and I am not sure, either. $\endgroup$ – Luboš Motl Aug 22 '12 at 7:53
  • 2
    $\begingroup$ @LubošMotl, thanks for the new style on your blog, its more readable now $\endgroup$ – lurscher Aug 22 '12 at 18:41
  • $\begingroup$ It is not a new style. It's a mobile template that's existed for several years. A few months ago, it was upgraded to deal correctly with comments and MathJax. But 90% of the people use the "old template". $\endgroup$ – Luboš Motl Aug 23 '12 at 4:32

I would have loved to put this reply directly under Eef's reply above, but I didn't figure out how this page works quickly enough.

Eef, when you're describing what COMPASS has done, I think that you are misrepresenting our work. We didn't do a MC simulation of the E(38) or of this effect, as you seem to claim. Instead, we ran our usual MC simulation, not including any backgrounds, and then saw a bump appear in exactly the right place. We understand this MC bump as being due to the material in the detector -- basically, $\pi^0$s produced in secondary hadronic interactions downstream of the target will appear as low-mass bumps in this spectrum. Depending on the distance from the target these bumps will have different masses. Therefore we can associate the "mass" of the bump with a specific distance, and lo and behold we find some material in the right place for 38 MeV. The exact height of the bump is of course dependent on the backgrounds present in real data, which are not taken into account in this simulation. For this reason we made the statement that "the COMPASS data do not confirm the existence of this state." (quoting 1204.2349).

In this regard I'm in perfect accord with my co-authors.

  • $\begingroup$ You can flag the moderators if you want this answer to be moved to a comment under eef van beveren's answer. (Comments require 50 reputation pts.) $\endgroup$ – Qmechanic Aug 24 '12 at 9:44

You say that you are siding with COMPASS.

However, that is an ambivalent statement.

COMPASS first released the opinion that the signal at around 40 MeV is an artifact of their methods of analysis and their equipment. Since I have no access to their methods and equipment properties, I had to accept such statement. Although I was not convinced.

Later, COMPASS released a Monte-Carlo simulation of the artifact. I showed that that simulation only explains a very small contribution (of about 10 percent) of the measured signal. The main author of the simulation agreed, in private while at a COMPASS meeting in Lisboa, with my observation.

So, with which COMPASS do you side?

Furthermore, the Dubna group consists of several excellent researchers which, moreover, work on new and unique equipment for photo-photon physics.

  • $\begingroup$ Comparing credentials and going by authority is not why one is cautious, but a dirty diavowed signal from a known noise source and two nuclear bumps at 38 MeV which require deep analysis are not the strongest case in the world. Lubos is not dismissing the claims, he is just saying you need more evidence. $\endgroup$ – Ron Maimon Aug 22 '12 at 16:18
  • $\begingroup$ I have the impression that you have not read our works on the E(38), from past year and from this year. Quite some material is collected there. Moreover, I have no accelerator in my office. Evidence has to come from experiment. In the mentioned works I have given quite some places where evidence can be found. $\endgroup$ – eef van beveren Aug 22 '12 at 16:41
  • $\begingroup$ Moreover, COMPASS has not proven their statement. That does not imply that COMPASS may not be right. But also for proving the contrary of our interpretation of the COMPASS signal, COMPASS needs more evidence! $\endgroup$ – eef van beveren Aug 22 '12 at 16:50
  • $\begingroup$ Nobody is certain here, no need to be so defensive. The issue is that this particle definitely has some sort of electromagnetic couplings, so either charged constituents or strong coupling to charged objects which is pretty much the same thing, and it's degrees of freedom should alter total cross sections for quark-quark collisions, which are calculated relatively precisely from pure QCD, so it's hard to see how it works, but it might work, I don't know. $\endgroup$ – Ron Maimon Aug 22 '12 at 17:06
  • 1
    $\begingroup$ Ok, they might not, but I definitely do. The idea that you could have a glueball at such low energies defies all common sense and lattice QCD predictions. This is just categorically impossible, you need new physics to have this, and I can't imagine what it is. $\endgroup$ – Ron Maimon Aug 23 '12 at 1:01

User1247 pointed out my mistaken reading of the scale in a previous answer, now deleted.

Fortunately I found a enter image description herepi0 mass plot in LHCB which shows that there is gamma gamma mass resolution to clear this point about a 38 MeV diphoton resonance.

By now they could provide us with a definitive plot.


The paper has been retracted.

Due to non ordinariness of the obtained results (standing out of The Standard Model) and at the request of co-authors the first version of the article is withdrawn for further verification and more detailed description of the experiment and data analysis. The second version is being prepared.

(from the replaced abstract on the arXiv)

  • $\begingroup$ For the record, there's a conference paper which shares most of its authors with the retracted paper where data from the retracted paper is studied. It states that "[a]ll these results support the conclusion that the observed structure with an effective mass of about 38 MeV/$c^2$ corresponds to the detection of particles with the given mass." epj-conferences.org/articles/epjconf/abs/2019/09/… $\endgroup$ – tobi_s Apr 8 at 2:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.