Checking the ⁸Be Anomaly with a Two-Arm Electron Positron Pair Spectrometer

Round 1

Reviewer 1 Report

The ATOMKI X17 anomaly is a long-standing puzzle, an important result in nuclear and particle physics. Its independent confirmation is highly desirable.

While the abstract claims (L3) independence, the lead ATOMKI physicist is the third author of the paper, providing the methodology, and an other ATOMKI physicist is the main contributor to the software (see also the reference to ATOMKI website in L123). Nonetheless, the experimental apparatus is indeed independent, though the simulation models are most probably related. 

The abstract (L6-7) emphasizes the confirmation of the absence of the signal at Ep= 17.6 and 0.8 MeV.  However a new analysis of the ATOMKI data suggests a signal at these energies as well: https://arxiv.org/abs/2205.07744v1 The 2022 ATOMKI result highlights the importance of the background model which is not at all discussed here. 

In general the paper needs some work to conform to high publication standards in various aspects, regarding citations, quality of figures and their captions, discussion of the results. Line by line comments are given below.

L29: references are missing for the 3H and 11B results

L36-44: statements should contain references to the observations

L44: give range of statistical significance 

L48: missing reference for the constraints from other experiments

L72: would be useful to quantify cosmic radiation background (later on claimed to be taken into account but not explained how)

L79: “much better resolution”

L82: “greatly reduced” : such statements are very vague and should be quantified, or previous published work referenced. In particular the energy resolution and total energy absorption efficiency for the chosen method is not specified 

L89: “completely stopped”

In general if the aim is to have the measurement easily reproducible (L87) based on the information given in the article, then giving the specifics of the applied instrumentation e.g. PMT (P93), DSSD (L102), etc… product codes takes very little space and provides easily accessible info to understand the characteristics and performance of the components. 

L91: “all corners with the same efficiency… specially shaped light guides” – was this measured? what kind of light guides? A photo of the experimental setup could be useful. 

L108-110: “very fast… very good energy and timing resolutions” - be more specific (description comes from the mesytec website that advertises the product) 

L110: “Up to two simultaneously responding channels are identified” - I assume both in x and y directions so that double-hit events could be detected

L119: “were taken into account” - how? Via the Geant4 simulation, I assume but shall be spelled out. 

L128: TDC connected to the MUX output is not on Fig 3 block diagram

L133: what was the time window for the coincidence condition?

L141: Explaining the difference between the simulation and the data around 350 GeV as double hits from the e+e- pair seems to indicate that the simulation was incomplete… as such double hits should be easy to include. How is the MC normalized? Were there any pedestal runs to estimate the noise (which is blamed for the low energy difference)? Why was 50 keV threshold chosen if the signal starts ~120 keV (Fig. 4)?

L142: How was the detection efficiency determined and which signal it belongs to?

Fig 5: x axis units are missing. Is the undecoded address shown?

L145: “correspond to the coordinates of the particles passing through the individual silicon strips” – not entirely… the hit coordinate distribution in reality is continuous but using threshold readout the reconstructed positions have a discrete distribution matching the 3 mm pitch size. Not stated how can there be any positions between the peaks. Address decoding error? How the position uncertainty contribute to the mass reconstruction uncertainty? 

In general from the measurement technology view point it is interesting to understand what are the main contributions: position or energy resolution? Target thickness? 

L147: “width of the DSD” - pitch size? 

L148 states the Fig 5 right should be position in cm but it clearly has the same undecoded position as on the left

L149: “only a few” - provide a fraction. What was the criteria to exclude events?

Fig. 6: What does the measured distribution say about the energy resolution and background conditions of the system? It should be discussed.

L165: how was the gamma detector used in the analysis?

L172: how was the cosmics background subtracted?

Fig 7: how was the effective thickness of the target (or the excitation function) used in the analysis? Was the excitation function compared to simulation? Caption: how is the function normalized? With which target was this data taken? L162 mentions two types of targets and it is not clear for any of the following plots which target is used and whether the results were compared and consistent.

Note in all the captions not only the beam energy but also the used target (or at least make it clear in the text how the two targets are used). 

L186 and Fig 8: a ratio plot of MC/data would illustrate better the agreement

Fig 8: right and left plots would ideally have the same x axis scale. What conclusion is drawn from their comparison (if any)? What kind of beam energy was used to prepare the distributions, especially for the coincidence events? 

L200: how were the E1 and M1 distributions combined? Where do the applied fractions come from? The 2022 ATOMKI paper emphasises the importance of the background model and the target properties. These are really not discussed here and are essential to be clarified.

Fig 10: y axis is not properly chosen as the M1 curve is not visible

Fig 11: E1 and M1 curves are not visible

How was the background shape derived from the M1 and E1 templates? What are the relative contributions? How does the background only assumption look like? What was assumed for the signal in the simulation? How was the mass derived? What were the inputs to the significance calculation? What assumptions does the given significance range correspond to?

L215: What is the expected precision and stability of the beamspot position? 

Were there any other systematic uncertainties (other than the beamspot) considered related to the experimental apparatus, the MC simulation, etc? Are they really negligible so that they are not mentioned in the paper? 

How the authors explain the fact that the reanalysis of the ATOMKI results claim to see the resonance at all proton energies while here it only appears at the 1040 keV measurement? This should be discussed.

L26: e+e- (missing superscript for 3 and -)

L71: fore → for

L94: photomultiplier tube (PMT)

L105: PCB board → printed circuit board (PCB)

L105: PCB boards → PCBs

L107: multiplexer 

L121: introduce abbreviations ADC, TDC, QDC

Fig 3 caption: abbreviations not appearing before need to be introduced: FA

L124: CFD discriminators → constant fraction discriminators (CFDs)

L130: The energy measured by the DSD was calibrated

L140: when both the e+ and e- created during the internal pair production 

L156: The energy spectrum measured by the scintillators for events selected by gating on double

L161, L162, … units should not be be typeset in italics

L173: from from

L176: in the plastic scintillator larger than 

L177: Drop “There are”. Two resonance peaks were…

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors report an experiment at the VNU University of Science that reproduces Krasznahorkay et al., (Phys. Rev. Lett. 116, 042501 (2016)), an experiment executed at the ATOMKI Laboratory (Debrecen, Hungary) in 2016. One important result of the earlier experiment was the observation of a previously unseen state at 16.70 MeV. This repeat of the earlier experiment also observes the excess indicating a new particle of mass 16.70 MeV.

While this paper is quite thorough and interesting, there are a number of questions that the authors should address before the paper is accepted for publication.  The first and foremost question is whether or not the observed state in this new experiment is truly identical in mass with 16.70 MeV previously reported.  This is statistically surprising (although not impossible) since the errors quoted in the ATOMKI Lab experiment are +/-0.35 (stat) +/-0.5 (sys) and the new result has an error of +/-0.47(statistical) +/-0.35(systematic) MeV. With errors of this scale, it is very unlikely that both experiments would observe EXACTLY the same mass to one one-hundredth of an MeV.

This EXACT result is further questioned by the statement on lines 215-217: "The difference in the angle of the anomaly obtained in this study of 135, and the published one of 140 [1], could be explained by a slight difference in the proton beam spot position on the target." If this is different in the two experiments, does the mass of the excess really match to one one-hundredth of an MeV?

Another general comment is that the figure captions are mostly quite short and do not adequately explain what is presented and its origin in each case. It would be helpful to expand the figure captions to make clear what each refers to in detail.

Abstract - the authors say they have used a different type of electron-positron pair spectrometer,
but do not say what facility that is.  They should state in the abstract that they are using a two-arm electron-positron spectrometer at the VNU University of Science.

Introduction - the VNU University of Science facility that is used is not mentioned until the end of the introduction.  I suggest that should be mentioned in the first paragraph of the introduction.

lines 93-94 - sentence is poorly constructed

line 97 - better construction:
   found to be less than 1 ns, as shown in Fig. 2.

Figure 2 caption - expand caption to include more complete description of meaning of distribution, such as role of Co-60 source.

Figure 6 - Should this be compared to Figure 1(a) of reference 1, which was produced with 7Li(p,e+e- )8Be, Ep =441 keV. If yes, why is the ratio of the two peaks (at 6.05 and 17.6 MeV) very different (about 1.3 in this Figure, but more than 5 - after the scale factor - in reference 1.)

3. Experimental results - please provide the beam current used for this experiment.  Reference 1 states typical currents were 1 microAmp.

lines 162-163 - Li2F target indicates the lithium mix is the naturally occurring one, while 7Li2O target indicates the lithium is a pure Li-7 isotope.  Is this what is meant by the different expressions?

  Two target materials are referred to here.  In the results that follow it is not clear which target material resulted in the Lithium results.  It would be helpful to make this clear.

line 168-173 - no mention of the beam energy used to produce the figure.  Why not include that for clarity?

lines 174-176 and Figure 7 - refers to scanning the proton beam energies from 441 keV to 1300 keV. The plot in Figure 7 shows no points between 700 and 900 keV.  However, lines 202-206 and Figure 10 will refer to the operation at 800 keV.  Why doesn't Figure 7 show anything near 800 keV?

line 214 - "With a deviation of 4-5s sigma from the Standard Model hypothesis."
   This is NOT a complete sentence!

   Greek symbol for sigma is used in the line, followed by written "sigma".
   This is redundant.  Only one should be used.

   Also, does reference to "Standard Model hypothesis" mean simply "deviation from no signal"?
   Another way to express this would be simply "The deviation is a 4-5 sigma excess."

Summary - Reference is made to three operating points using different energy references, which is confusing.  The summary refers to the 17.6 MeV transition, the Ep=800 keV transition, and the 18.15 MeV transition.  It would be more clear if the Ep value were included for each of the three energy points. For example the 17.6 MeV transition (Ep=441 keV), the Ep=800 keV transition, and the 18.15 MeV transition (Ep=1040 keV), or some similar clarifying reference.

Final comment - given the poor construction of sentences on lines 93-94, 97, and 214, the authors should review the text to ensure other poor constructions don't exist.  Most of the language is very good, but there are these few anomalies.

Given the poor construction of sentences on lines 93-94, 97, and 214, the authors should review the text to ensure other poor constructions don't exist.  Most of the language is very good, but there are these few anomalies.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper reports interesting results, but includes several critical questions. Comments are in a separate file.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

I thank the authors for the detailed replies and the significantly improved paper presentation. I have a few small remaining comments, mostly reflecting on the authors' responses:

L167: Please, mention the 120 keV lower limit on the electron / positron energy applied in the event selection. 

L206: 5.2% FWHM

Fig 10 caption: Specify the target, please. Li2O?

L240: The combination of the E1+M1 distributions shows a good agreement with the experimental data when the contributing fractions are fitted. (or something similar to say that the e1 and m1 rate is left floating to fit the data.)

L263: By assuming that the deviation is coming from the creation and immediate decay of an intermediate particle to an e+e− pair [3]

L284: agrees well with the simulated one dominated by the M1 transition

L270: I suggest to add: Additional potential systematic uncertainty caused by e+e- pairs induced by external pairing in DSSD detectors by gamma radiation is being evaluated in simulation. (Based on your response to the related question on v1 of the paper.)

L4: has two arms and simpler 

L43: using fundamentally different position sensitive detectors (plural, remove “a”)

L50: remove double full stop from the end of the sentence

L57: The report also gives

Fig 1 caption: spectrometer focusing or OR spectrometer that focuses

L90: The FWHM at 17.6 MeV was below 20 keV OR An FWHM below 20 keV can be achieved at 17.6 MeV 

L94: The ratio between the numbers of full energy events and total recorded events is smaller than 1.5% at an electron energy of 18 MeV for a 3x3 inch^2 LaBr3 detector.

L101: such a way

L131: manage double-hit events (not mentioned before so drop “the”)

L156: The sentence needs improvement as the first part mentions coincidence between the DSSD and the scintillator while the 2nd half gives the max time difference between the two arms, maybe what was meant is: “When creating the spectrum, we required real coincidence between the DSSD detector and the plastic scintillator located behind it by requiring a time difference smaller than 40 ns.”

Fig 6 caption: The total energy deposited in the plastic scintillator by e+e−pairs, which are selected as double-hit events in the corresponding DSSD. The peaks are from the transitions indicated in the figure; the red line is a Gaussian fit with its center and sigma shown in the legend.

Fig 9 caption: when the LiF target is bombarded by a proton beam at 441 keV….. the numbers in the legend give the contributions of each component

L244: missing closing parenthesis after keV

L256: may have  washed out

Fig 11 caption: when the Li2O target is bombarded by a proton beam… so the background line overlaps with the E1 line.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments are given in a file.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf