3

update: The 2020 Jan/Feb issue of the CERN Courier News analysis: Rekindled Atomki anomaly merits closer scrutiny

edit: @annav has noted an informative blog post summarizing the theoretical (but not experimental error) issues related to this measurement.

There is some excitement about a bump in an electron-positron opening angle spectrum from 8Be internal pair conversion that it might be a new particle - mediator of a possible new force. eg. Feng et al. 2016.

In Krasznahorkay et al. 2016 (ref. 6 of Feng 2016) the bump can be seen in Fig. 4:

enter image description here

However, Gulyás et al (2015) (ref. 30 of Feng et al. 2016) shows a measure of the sensitivity as a function of the opening angle, and the improved design with 5 lepton telescopes is still fairly bumpy:

The resulted angular correlation for the uncorrelated events gave us the experimental response curve. Reasonably good agreement was obtained with the results of the MC simulations as presented in Fig. 5.

enter image description here enter image description here

My Question: Recalling another electron positron "bump" experiment (see below), and it's final conclusion after three different experiments by three different groups at two laboratories finally got together with sufficient wine, what is the evidence so far that this bump is not just a calibration error, an incorrect efficiency correction, or something similar?

From this review or "retrospective" presentation - refer to the original document for more background.

enter image description here

above: slide #4 from here.

enter image description here

above: slide #26 from here.

uhoh
  • 6,308
  • 2
    Most particle physicists would give a high probability for one of the explanations you offer. The crux is that if such a particle exists, it should show up in e+e- laboratory experiments – anna v Aug 24 '16 at 13:49
  • 2
    All these big pictures make the question a little hard to handle. Not that it's bad, really, but it would certainly be an improvement if you can trim it down a bit. (I think the screenshots of the slides could be safely omitted, for example.) – David Z Aug 24 '16 at 13:50
  • @DavidZ I see what you mean - I'll trim them down now. – uhoh Aug 24 '16 at 13:55
  • 1
    i changed the imgur to medium – anna v Aug 24 '16 at 13:57
  • @annav wow! Please teach me how to do that! Of course now the captions are small and blurry and difficult to read. – uhoh Aug 24 '16 at 13:59
  • 1
    Related, related. – rob Aug 24 '16 at 14:00
  • @rob yep, but I'm asking specifically about the data analysis, and not about the theory or any possible particle or force. "what is the evidence so far that this bump is not just a calibration error, an incorrect efficiency correction, or something similar?" Is the bump an artifact of the analysis of the data? It's a different question. This is why I wanted the plots of the data and their captions to be large and readable, not small and fuzzy. – uhoh Aug 24 '16 at 14:05
  • @Anna's comment is the real sticking point. SLC and LEP should have seen a thing like that easily. But running down systematic problems with someone else's experiment is very hard, and many experimenters won't make strong statements in public unless and until they are gotten some details and thought for a while. – dmckee --- ex-moderator kitten Aug 24 '16 at 14:10
  • @uhoh Yes, I agree. I didn't discuss it specifically in my answer to the other question, but deconvolving the angular sensitivity of the detectors from the observed angular distribution was probably the murkiest part of the Krasznahorkay et al. paper. There are several right ways and several wrong ways it could have been done. One promising observation is that Krasznahorkay et al. cite previous observations of this effect going back about 20 years; however the apparent mass is different in the older papers. A proper analysis is more work than I can put into Phys.SE posts right now. – rob Aug 24 '16 at 14:13
  • @dmckee I understand but there may be published information out there that I haven't seen or can't access which is better than the one now out-of-focus monte carlo simulation with the caption that is now too fuzzy to read. – uhoh Aug 24 '16 at 14:16
  • @dmckee That's what's interesting about the Feng et al. analysis. Their approach is: suppose this state exists, and it hasn't shown up in other experiments; what does that say about its couplings? I don't know that much about SLC and LEP, but I can imagine a 20 MeV particle being far, far below the floor of their search regions. – rob Aug 24 '16 at 14:16
  • @rob Thanks! I'm just trying to keep my question from being closed or marked as duplicate - at least too quickly. This SE can be a tough room sometimes and I feel awfully defensive. AnnaV just defocused my figures via a poor sampling/compression feature and won't tell me how :) – uhoh Aug 24 '16 at 14:21
  • 2
    @rob Remember that you are not looking at the boson but at the scattering mediated by it. Momentum transfer is only limited by the boson mass if you don't have enough energy. Both SLC and LEP spent a lot of time and effort on weak interaction (80-90 GeV carriers) and weak coupling. A circa 20 MeV boson gives much larger 'geometric' component to the cross-section and there was plenty of energy. So the couple constant would have to be tiny for them to miss it. But how does it effect low energy decays strongly enough to show? Not impossible with a running coupling constant, but ... – dmckee --- ex-moderator kitten Aug 24 '16 at 14:31
  • 1
    @uhoh "Please teach me how to do that!" Look at the edit history. It's easiest to see in the "side-by-side markdown" view. – dmckee --- ex-moderator kitten Aug 24 '16 at 15:09
  • What type of "evidence" do you expect for something not being an error? We generally take "no one has shown it to be wrong" to be the best evidence of such results not being wrong we can get. – ACuriousMind Aug 24 '16 at 16:29
  • 1
    If you look at the link given for the picture you have downloaded, if it is .png you just insert an "m" ...m.png for medium,( or an ...s.png for small ). The only defocussed one is theone before the last one, the others are clear enough for people wearing the glasses they should. I deleted the m an lo, the original – anna v Aug 24 '16 at 18:36
  • @ACuriousMind I'm keeping an open mind & not "pre-expecting" the answer. If you read a number of extended experimental papers that contain some very surprising and unexplained results, there is usually a careful discussion of all possible sources of error that could produce the unexpected feature. I don't see it here - Fig. 4 of Gulyás et al (2015) doesn't seem strong enough on its own to show that the variation of sensitivity with opening angle is well-enough characterized to show the "bump" isn't an artifact. Does (for example) the bump appearing in the |y| <0.5 but not |y|>0.5 data help? – uhoh Aug 24 '16 at 23:12
  • Have you seen this blog entry? http://motls.blogspot.com/2016/08/the-delirium-over-beryllium.html#more – anna v Oct 29 '16 at 04:19
  • @annav If I answer now, "no", but if I answer after I look, then "yes" :) I'll take a look now, thanks! (I hate blogs with embedded avertisements that keep moving my window scroll up and down while I'm trying to read, also hate blogs that are very long and incredibly skinny with teeny tiny font). – uhoh Oct 29 '16 at 04:24
  • when it stops loading the motion should stop. It is a serious blog for the physics part – anna v Oct 29 '16 at 04:26
  • @annav I think this result has been so welcomed into the theoretical community because it simply provides fodder for bored theorists who need a bump to theorize about. Like the 750 GeV di-photon that disappeared after the 500 papers were written :) – uhoh Oct 29 '16 at 04:39
  • @annav that advertisements eventually stop is not a vindication of irritating embedded advertising. I don't like the sentence "When the protons have around 1.03MeV of kinetic energy, they excite lithium into the 18.15MeV beryllium state." One does not "excite" 7Li into 8Be. OK I'm done complaining now, this looks like a great review of the theoretical situation - thanks!!! I still would like to see a better discussion of the more mundate calibration of the angular sensitivity of the detector Fig. 5, which so far appears poor - no vert. errorsbars, low granularity & statistics, etc. – uhoh Oct 29 '16 at 04:40

0 Answers0