Frequently Asked Questions

This page is a collection of questions that were asked at conferences.

Neutrino candidates

Q1: About the 8 neutrino candidates observed: is this number good or bad? Why is so low? Is it compatible with what we expect?

A1: These candidates remain after a strong selection, limited to a section of the detector. In this case, we expected a total of 5 neutrino events. However, the aim of this analysis was to observe neutrino interactions with high significance, not yet to compare them with theoretical expectations of neutrino production at LHC. Further analysis is ongoing to address systematic and compare events with expectations.

Neutrino flux from charm vs pion + kaon

Q1: How well can you constrain electron neutrino flux from charmed hadron decays from pp collisions?

A1: Taking into account uncertainties in the correlation between the yield of charmed hadrons in a given eta region with the neutrinos in the measured eta region, it was evaluated that SND@LHC can measure the charmed-hadron production in pp collision with a statistical uncertainty of about 5%, while the leading contribution to the uncertainty is the systematic error of 35%.

Q2: What is the source of the neutrinos?  How many events are expected from pion + kaon vs charm?

A2:  Muon neutrinos are mainly produced in pion/Kaon decays for both FASERv and SND@LHC. About 50% of electron neutrinos interacting in FASERv come from charmed hadron decays. This fraction amounts to 90% in SND@LHC.

Normalisation

Q: How do you get the normalisation for the cross section ?

A: We can not get an absolute cross-section using the Run 3 setup, as we do not have reliable normalisation. With AdvSND's near detector, we aim to get the normalisation from the near detector.

Neutrino energy range

Q: What is the expected range of neutrino energies ?

A: Between a few hundred GeV and a few TeV.

ν vertex resolution

Q. What is the error in electron neutrino vertex position, as predicted from the associated electromagnetic shower.?

A.  We are still completing the development of the algorithm . The track position errors in the SciFi are of the order of 200 μm.

Energy resolution

Q. Why is the fractional energy resolution (of SciFi and HCAL) estimated at the level of 20%? What is the main source of uncertainty?

A. The reported value (22%) is obtained using a counting method based on the correlation between the true energy and the number of registered hits. The application of CNN-based algorithms showed to further improve the resolution. This needs to be verified once the calibration is in place.

Timing performance

Q. What timing performance do you expect?

A. O(100ps) for both SciFi and Veto/Muon tagger. See also New Physics, A2.

CC vs NC

Q1. How do you distinguish νe NC and νμ NC ?

A1. We do not. All we do is to check the ration between all flavour NC interactions and all flavour CC interactions to understand the detector systematics in separating NC and CC.

Q2. In this boosted regime of high nu energies, in a νe interaction electromagnetic and hadron showers mix up. How well do you separate the NC interaction from νe CC interactions?

A2.  Because of the micrometric accuracy of the emulsion you can topologically separate electrons from neutral pions by observing the displaced vertex associated with photon conversion. This information can be combined with the response of the electronic detectors, where simulations indicate there is room for distinguishing nueCC from NC events using the hit distribution in both SciFi and HCAL (see preliminary study reported by Cris (https://indico.cern.ch/event/1116258/contributions/4689322/attachments/2409891/4123463/cv_sndlhc_cc_nc_sep_20220317.pdf).

But perhaps what this person was asking about was how to discriminate neutrino scattering events  from neutral hadron backgrounds. This was studied by FASERnu, they have determined the energy spectra of both cases and use a neural network to do the discrimination  (https://journals.aps.org/prd/pdf/10.1103/PhysRevD.103.056014).

Neutrino flavour identification

Q1. How are the three neutrino flavours identified?

A1. The detector is composed of a hybrid system based on an 800 kg target mass of tungsten plates, interleaved with emulsion and electronic trackers, followed downstream by a muon system.
The configuration allows efficiently distinguishing between all three neutrino flavours.
The identification of the neutrino flavour is done in charged current interactions by identifying the charged lepton produced at the primary vertex.

  • Electrons will be clearly separated from π0’s thanks to the micrometric accuracy, which will enable photon conversions downstream of the neutrino interaction vertex to be identified.
  • Muons will be identified by the electronic detectors as the most penetrating particle, beyond the hadronic shower.
  • Tau leptons will be identified topologically in the emulsion, through the observation of the tau decay vertex, together with the absence of any electron or muon at the primary vertex.

Muon flux in the acceptance

Q1. Even after several 100 ms of rock, maybe too many muons will be passing through?

A1. The integrated rate of muons  in the SND@LHC acceptance is estimated to be about 350 Hz, equivalent to about 2×104 /cm2/fb1.
This value accounts for the effect of multiple scattering in the rock upstream of the detector.

A1. The preliminary estimates from data are: SciFi (1.94±0.01)x104fb/cm2, DS (2.10±0.01)x104fb/cm2 (see Simona 14/12/2022), Run 0 emulsion ( 1.5 ± 0.1) x 104 fb/cm2 (see Antonio 14/12/2022)

Q2. Why is there a vertical muon flux gradient?

A2. It is due to the muon trajectory bending in the B fields of various LHC magnets. We observe the gradient also in the simulations.

Background

Q. Do we get background coming from the back of the detector (i.e. muons coming from the ALICE direction)?

A. We have detected during the scrubbing runs that LHC beam losses cause background, at -110 rad in horizontal plane. They are possibly also the source of muons hitting the SND@LHC detector at a wide angle range centerred
at about -0.54 rad. During stable beam no excess of muons at such angles have been observed.

New physics

Q1. How can you distinguish between the elastic scattering of a new physics particle and a neutrino off protons?

A1. A scattered proton will have a secondary vertex at the start of the hadron shower which is O(9 cm, one lambda) away from the primary vertex. A neutrino DIS event produces a cascade at the primary vertex. If one of the products in this cascade is a proton which produces another cascade, this event can be excluded by looking at the multiplicity at the primary vertex, which should be 1 in the case of the scattering of a new particle. For inelastic scattering one has to look at the energy spectrum of the neutrino, which is separated from that of a mediator for energies from 1 GeV (see TP page 98, fig. 73).

Q2. How confident are you that you can use TOF to distinguish between neutrino and new physics interactions? Is the time resolution sufficient to take into account the 200ps LHC bunch size?

A2. In the TP, the TOF was not used. It is an extra handle that can be applied to reject background. The significance of the use of the TOF depends on the particle mass and is given in Figure 72 on page 96 of the TP. The timing resolution of both the SciFi and  the scintillator bars is O(100 ps). However, TOF is not efficient for all scenarios, see arxiv:2104.09688.pdf, page 10.

Q3. If you had a particle decaying to muon pairs, would you potentially be able to separate it from backgrounds if the decay occurred (a) before the initial scintillator plane, or (b) within the iron blocks?

A3. These events have not been studied in detail, because the background has not been estimated. One source of background is a neutrino interaction when you produce a muon and a pion in the final state and you mis-identify the pion as a muon. Normally, at the LHC energies, this background is suppressed by the multiplicity of events being higher than 2 charged particles. It needs to be studied. We don't have a muon spectrometer so we won't be able to measure the invariant mass. If the decay happens in the emulsion brick we can at least precisely determine the multiplicity (in the iron we won't see nuclear fragments and other soft particles emerging) and the vertex with the muon slopes. Intuitively, the only chance we have to reduce the background is with events originated in the emulsion target. Even with these events, the lack of a muon momentum is a problem. This is one reason to go for an upgrade.

Q4. What is the accuracy of your Lepton Flavour Universality test?

A4. The unique tests of lepton flavour universality with neutrino interactions can reach 30% statistical and 22% systematic uncertainty in the ratio between 𝜈e and 𝜈𝜏, and about 10% for both statistical and systematic uncertainties in the ratio between 𝜈e and 𝜈𝜇 at high energies.

Q5. How is a dark photon interaction different from a neutrino NC?

A5. The background rejection in the FIPs search depends on the process. If you look for elastic scattering on electrons, then the background mainly comes from elastic scattering of electron neutrinos, which is almost negligible at SND@LHC energies. If you look for the scattering on nucleons, then the background from neutrino NC DIS may be not negligible, and a ToF measurement is crucial for the rejection, as mentioned in Q2.

Emulsion Cloud Chambers

Q1. Why does SND@LHC use Tungsten instead of Lead?

A1. The ECC technology alternates emulsion films, acting as tracking devices with micrometric accuracy, with passive material acting as the neutrino target.Tungsten is used as a passive material to maximize the mass within the available volume.

Q2. The tau neutrino was discovered by the DONUT experiment, and the emulsion analysis was especially challenging. After twenty years, what are the main improvements in emulsion scanning and analysis?

A2. The main improvements are on the scanning speed, and on the online processing employing GPU processors. The main challenges are currently related to track density and accelerator intensities, and we are working into addressing the

Q3. How does the scanning work?

A3. One film is scanned at a time, and micro tracks, clusters are reconstructed online to reduce disk usage.

 
 

Comparison with FASERnu

Q1. How is SND related or compared to FASERnu? Why are two experiments necessary?

A1. SND and FASERnu are complimentary in the sense FASER will observe neutrinos with eta>9, while SND looks at 7.2<eta<8.4. Actually, SND is making use of the background measurements performed for FASER, so this is already a nice example of cross-collaboration. There are relevant differences also in the detector layout.

Q2. And concerning the energy spectrum? Are they different at all?

A2. In the eta range explored by SND a large component of neutrinos reaching the detector come from the decay of charmed hadrons. This is not the case of FASERnu, which cannot perform charm production measurements.

Comparison with DONUT

Q.  Is the detector similar to DONUT? is the purpose to detect tau neutrinos?

A. No, although DONUT was conceived to be hybrid, at the end it could only rely on the emulsion part because of the unexpected background. DONUT was designed to observe for the first time tau neutrinos. We see also tau neutrinos but actually the most important sample for us is νe because we use them to derive the charm production cross-section. And the physics case is much wider. 

Syncing events with ATLAS

Q. Is it possible to sync SND@LHC events with ATLAS?

A. At the moment it is not possible, but there is work in progress to match event timestamps in ATLAS and SND@LHC.  However, since our eta range is out of reach for ATLAS, they will not trigger on particles that we detect in SND@LHC.

AdvSND

Q1. Why don't you contribute to pushing the advancement of silicon detectors instead of reusing CMS's sensors?

A1. It is a question of money. The CMS sensors are worth 3M CHF and we are getting them for free, plus a certain amount of support.

Q2. What's the spacial resolution you achieve with those sensors?

A2. The track residual and resulting resolution is compatible with the expectation of 122 μm/√12 ≈ 35 μm.

Q3. Can we see tau neutrinos with the silicon strip detector?

A3. It will be more difficult than with emulsion due to the coarser segmentation along Z and the worse time resolution. Studies are ongoing to understand the expecting tagging efficiency.

Q4. Is the magnet using superconducting technology?

A4. No.

Q5. Are the proposed civil engineering works agreed with CERN?

A5. Discussion is ongoing with the LHCC and a ECR is being prepared. We have been asked to study in more detail the 'minimal' option (now called 'compact').

Q6. How much luminosity do we intend to accumulate in Run 4 and Run 5?

A6. We intend to take date during the entire Run 4 and 5 periods (250 fb-1 per year?).

Q7. How off axis would the upgraded detector be?

A7. Up to eta ~8.

Veto replacement in 2023/2024 YETS

Q. Why did the original veto didn't already have a plane with vertical bars (instead of two with horizontal bars)?

A.  The vertical bars of plane 1 are shifted by 2cm wrt plane 2, and so  there are no overlapping inefficient regions. Having one of the planes vertical would mean that there would be horizontal as well as vertical inefficient regions. However, since a major source of inefficiency is the dead time of a bar, it could be that a second muon within this dead time would hit the same bar, and this muon would then escape undetected through the gap in the second plane. To remove this source of inefficiency, a third, vertical, plane was installed.