Spallation background in the Super-Kamiokande experiment Laura Bernard, on behalf of the Super-Kamiokande Collaboration ICHEP conference, neutrino session July 29, 2020 Low energy searches in the SK experiment The Super-Kamiokande experiment has played a major role in astrophysics by investigating low energy O(10) MeV neutrinos, notably : → characterization of 8B solar neutrino spectrum → currently exhibits the best sensitivity to relic neutrinos from distant supernovae → See talk by S. El Hedri 30/07 Signal & backgrounds for supernovae relic ν¯e : Solar neutrinos : elastic scattering of νe Supernova relic ν¯e : Inverse β-decay 5-20 events / year are expected → Time & space correlated signal 2.2 MeV γ n thermalisation νe diffusion p e+ H(n,γ) Prompt signal Delayed signal (~200 mus) [12 - 80] MeV Left : Neutrino-induced background PhysRevLett.93.171101 Right : Muon spallation background ZHANG Yang PhD thesis 2015 A dominant background for SRN and solar searches is caused by muon spallation in water → Well caracterized in liquid scintillators but not yet in water detectors → Need simulations Laura Bernard, July 29, 2020 2/14 The SK detector ◦ Located in the Kamioka Mine, Japon Mountain ◦ 1000 m underground (∼ 2 Hz muon rate) 2700 m.w.e. ◦ 50 kton Water Cerenkov detector ◦ 11129 ID PMTs (3 ns resolution, 50 cm) ◦ Energy coverage : 4 MeV to ∼ TeV ◦ Operational since 1996, now in phase VI ◦ Gd loading just started ! Rock Fiducial γ ~8 MeV γ-cascade γ Volume ν n thermalisation e 22.5 kton Rock diffusion p Inner Detector Inner e+ Outer Detector γ Gd(n,γ) Prompt signal Delayed signal (~100 mus) [12 - 80] MeV → improved n-tagging → reduce spallation background Rock Laura Bernard, July 29, 2020 3/14 Muon-induced spallation processes Illustration of muon-induced spallation in the Super-Kamiokande water tank Muons abundantly produce daughter particles that Hadronic showers are the dominant process initiate electromagnetic and hadronic in the production of unstable isotopes, showers (e, π, p, n, γ) through which decay products (e, γ, n) can be a successive spallation processes background at low energy μ (< 20 MeV) From Nikhil Vittal Shetty, M.Sc. PhD thesis The shower and isotopes can be located by the neutron captures 2-3 meters isotopes in the medium (H, Gd), which is easier in LS than in water detectors neutron cloud Illustration of muon-induced spallation in the Super-Kamiokande water tank Laura Bernard, July 29, 2020 4/14 How do we deal with spallation in the data ? K.Bays et al. Phys. Rev. D 85, 052007 (2012) S.Locke, poster at Neutrino2020 Muon track Entry point ID (x, y, z) ◦ Extract information directly from the data : Correlations between low Distance to μ entry energy events and muon track, neutron cloud to locate hadronic shower Transverse Distance to μ → Use WIT (Wideband intelligent Maximum trigger) system for a lower threshold Neutron captures dE/dx : Qpeak to see 2.2 MeV γ’s from n-capt Isotope candidate ◦ Eliminates ∼ 90% of the background Longitudinal Transverse for [50-90]% signal acceptance Distance to Qpeak Distance to μ ◦ Need to reduce even more this background Exit point ID (x, y, z) Laura Bernard, July 29, 2020 5/14 Simulation of the muon flux at SK with MUSIC MUSIC is a 3D Monte-Carlo code for muon propagation through the rock. Antonioli et al. Astropart.Phys.7:357-368,1997 To reach the SK tank, a muon must have an energy > 1.3 TeV Angular deviation and lateral displacement of µ’s (multiple scattering, bremsstrahlung, pair production and inelastic scattering) are taken into account ◦ Mean energy of muons reaching SK tank : 258 GeV ◦ Spectrum extends up to several TeV Kenoyama topological profile Spectra provided by A. Coffani, PhD at LLR, Polytechnique Phys.Rev.C81:025807,2010, initially from Geographical Survey Institute of Japan (1997), unpublished. Laura Bernard, July 29, 2020 6/14 Spallation simulation with FLUKA FLUKA is a fully integrated particle physics MonteCarlo simulation package. Fluka version: 2011.2x.7, flair version: 2.3-0 ◦ A first FLUKA-based simulation of vertical µ− in the SK tank was done in 2014 by Shirley Weishi Li and John F. Beacom. Physical Review C 89.4. ◦ We reproduced this simulation and improved it by: ◦ Using the real MUSIC flux + ◦ Adding the µ+ component (ratio at sea level µ = 1.26) → reduces isotopes yields (16N) µ− ◦ Interfacing FLUKA with the official simulation of Super-K PhD work of A. Coffani, LLR, Polytechnique 45 FLUKA All 40 FLUKA E > 3.5 MeV FLUKA E > 6 MeV 35 −→ Comparison of the isotope production SK data Events/kton/day 30 ◦ In the data, the yields have been extracted using a fit of the decay times of each isotope 25 from µ track ZHANG Yang PhD thesis 2015 20 ◦ The comparison is within an factor of a 15 X few, consistent with uncertainties from 10 nuclear physics models 5 18 17 16 16 15 14 13 13 12 12 12 11 11 9 9 8 8 8 N N N C C B O B N B Be Be Li C Li B Li He Laura Bernard, July 29, 2020 7/14 Spallation simulation with GEANT4 ◦ Developped a Geant4-based muon-induced spallation in water simulation and compared to FLUKA → 4 lists tested : QGSP_BERT_HP, QGSP_BIC_HP, FTFP_BIC_HP, FTFP_BERT_HP Master 1 thesis from G.De La Valle, Polytechnique ◦ X Overwall observables are very well reproduced (nature of isotopes’ parents as a function of the energy, path lengths, muon deposited energy, ...) 16 ◦ X Yield of dominant isotopes ( N, 12B) within a factor 4, consistent with uncertainties from nuclear physics, Path lengths : Left : Geant 4 – Right : FLUKA stronger discrepency for 8Li (3d most important) → Simulations are about to be implemented in a Geant-4 Super-K simulation (WCsim) to better understand impact of reconstruction, and will be of great interest for Hyper-K studies Distance of particles to muon track : Left : Geant 4 – Right : FLUKA Laura Bernard, July 29, 2020 8/14 Interfacing FLUKA with Super-K detector simulation (SKDetSim) SKdetSim is Geant3 based → Need to evaluate how to propagate the FLUKA events to SKDetSim. ◦ Unstable isotopes are mainly produced in hadronic showers → Need to save hadronic showers from FLUKA : π+/−/0, neutrons , γ ... isotope decays ◦ Turn off the creation of secondaries in hadronic processes in SKDetSim ◦ Turn off muon nuclear interactions in SKDetSim ◦ Let SKDetSim take care of the electromagnetic part ◦ Event 1 Generated muon track From FLUKA simulations (t = 0) MIP muon c (t = d ) Particles from shower Reconstructed muon track With MuBoy fitter Secondaries : pions (~ 35/μ) , gammas (~ 40/μ) dedx scale ◦ Event 2 : Neutron captures on H (~ 42/μ) (t = tn−capt ) 2.2 MeV γ μ Isotopes (~ 2.2/ ) from n-H capture Qpeak = maximum deposited charge ◦ Event 3 : (t = tdecay ) γ/β from isotopes decay The tricky part is the muon + shower energy deposition and reconstruction Laura Bernard, July 29, 2020 9/14 First Data/MC comparison of spallation background in water Spallation data used : may 2020, supernovae relic (SRN) analysis. Contain pairs of muons and relic candidates allowing for stat. random subtraction. Comparison done for through-going muons only. Muons directions and track length in tank: 0.025 0.06 Simulation 0.025 Simulation Simulation 0.02 Data Data 0.05 Data 0.02 0.04 0.015 0.015 0.03 0.01 0.01 0.02 Normalized to integral 0.005 Normalized to integral 0.005 Normalized to integral 0.01 0 0 0 −0.005 −0.005 2 −1 −0.5 0 0.5 1 2 −1 −0.5 0 0.5 1 2 −1 −0.5 0 0.5 1 1.5 Muon direction in X 1.5 Muon direction in Y 1.5 Muon direction in Z Ratio Ratio Ratio 1 1 1 0.5 0.5 0.5 0 0 0 −1 −0.5 0 0.5 1 −1 −0.5 0 0.5 1 −1 −0.5 0 0.5 1 Muon direction in X Muon direction in Y Muon direction in Z Simulation 0.05 Data 0.04 0.03 0.02 Normalized to integral 0.01 X Validation of MUSIC simulation 0 20 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 ◦ Slight discrepencies for track length 1.5 Length (cm) Ratio 1 0.5 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Length (cm) Laura Bernard, July 29, 2020 10/14 First Data/MC comparison of spallation background in water Energy deposition of muon with shower : 0.05 0.08 Simulation Simulation Simulation Data 0.1 Data Data 0.07 0.04 0.06 0.08 0.05 0.03 0.06 0.04 0.02 0.03 0.04 Normalized to integral Normalized to integral Normalized to integral 0.02 0.01 0.02 0.01 0 0 3 3 3 ×10 0 ×10 ×10 20 100 200 300 400 500 600 700 800 20 100 200 300 400 500 600 700 800 20 20 40 60 80 100 120 140 160 180 200 1.5 Qtot (PE) (muqismsk) 1.5 resQ (PE) (muqismsk - peprcm*length) 1.5 Qpeak corrected Ratio Ratio Ratio 1 1 1 0.5 0.5 0.5 × 3 × 3 × 3 0 10 0 10 0 10 0 100 200 300 400 500 600 700 800 0 100 200 300 400 500 600 700 800 0 20 40 60 80 100 120 140 160 180 200 Qtot (PE) (muqismsk) resQ (PE) (muqismsk - peprcm*length) Qpeak corrected Total deposited charge (Qtot) Residual charge (Qtot - MIP) Charge in higher segment (Qpeak) Simulation 0.04 Data 0.03 0.02 Normalized to integral ◦ Total charge : ∼ 120000 PE ∼ 20 GeV deposited in the 0.01 tank (using 6 PE/MeV conversion) 0 20 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 1.5 Qpeak position (cm) Ratio Validation of FLUKA shower model and interfacing 1 X 0.5 0 with SKDetSim 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Qpeak position (cm) Location of Qpeak Laura Bernard, July 29, 2020 11/14 First Data/MC comparison of spallation background in water Transverse distance from isotope muon track : Muon track 0.05 Simulation Data 0.04 0.03 Ll (longitudinal distance to Qpeak) 0.02 Isotope candidate Normalized to integral Qpeak 0.01 (maximum Lt (transverse distance) 0 dEdx) 20 50 100 150 200 250 300 350 400 450 500 1.5 Transverse distance to muon track (cm) Ratio 1 0.5 0 0 50 100 150 200 250