Measurement of ambient neutrons in an underground laboratory at - PowerPoint PPT Presentation

Measurement of ambient neutrons in an underground laboratory at Kamioka Observatory Keita Mizukoshi Kobe University TAUP2019 at Toyama International Conference Center 9 Sep. 2019

Introduction • Neutron is the one of the most serious backgrounds (BG) for experiments in underground. • Direct Dark matter search • Neutrino-less double beta decay search • To evaluate and shield neutron BG, 
 it is very important to evaluate ambient neutron flux • Such neutron BG has not measured systematically. • Low rate in underground i → Required high efficiency • Unknown generated points → Energy unknown • Our goal is quantitative neutron flux in the underground � 2 Keita Mizukoshi Kobe Univ.

Detector (He-3 proportional counter) K. Mizukoshi et al., PTEP 123C01 3 He proportional counter • We used a 3 He proportional counter. 3 He + n → 3 H + p + 0 . 76 MeV • The energy of the exothermal reaction in the neutron capture can be obtained. ~5 k barn at thermal DAQ PC • This detector is thermal fast sensitive to thermal SUS φ 52mm neutrons (~0.025 eV), 3 He - 10 atm and cannot measure an initial neutron energy. 380mm “Setup A” � 3 Keita Mizukoshi Kobe Univ.

4 Setup for fast neutron K. Mizukoshi et al., PTEP 123C01 • To measure high energy neutron, 
 fast we used a moderator (polyethylene). thermal • Boron sheet captured thermal neutrons and reduce its effect. 3 He Thermal neutron “Setup A” ~0.025eV 160 ) 2 Counts/(neutron/cm Setup A 140 Setup B thermal 120 fast Setup B w/o B-sheet 100 80 Boron sheet 60 t5mm Polyethylene 40 3 He 20 0 510mm 9 8 6 5 3 − − − 7 − − − 4 − − 2 − 1 2 10 10 10 10 10 10 10 10 10 1 10 10 Generated neutron energy (MeV) Efficiency estimated by Geant4 “Setup B” Keita Mizukoshi Kobe Univ.

noise peak Wall effect Electric 
 Counts/bins Results 100 80 60 40 3 He + n → 3 H + p + 0 . 76 MeV 20 • Full energy peak is 0.76 MeV. 
 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Measured energy (MeV) If 3H or p escapes, continuum region Measured spectrum in setup B will be made in a low energy (Wall Counts/bins effect). 1200 • Low energy region below 0.3 MeV is 1000 dominated by electric noise for 800 ambient neutron measurement. 600 • We counted events up to 0.85 MeV 400 200 and down to 0.5 MeV, then the number 0 of total events was estimated by a 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Measured energy (MeV) clear spectrum observed using 252Cf. Spectrum of Source ( 252 Cf) Count rate in each setup • The count rate of Setup A A B Setup (R A ) and B (R B ) involves a 1.295 ± 0.034 0.446 ± 0.018 Count rate ( × 10 -3 cps) detection of thermal and 
 14.03 19.27 Live time (day) fast neutron, respectively. K. Mizukoshi et al., PTEP 123C01 � 5 Keita Mizukoshi Kobe Univ.

Simulation K. Mizukoshi et al., PTEP 123C01 Main components in each samples (wt. %) sample1 sample2 sample3 • To convert from the count rates (RA, RB) to O 40.5 37.9 35.6 ambient neutron flux, the spectral shape was Ca 28.0 24.3 29.7 required. The shape cannot measured by He-3 Si 16.6 15.6 12.0 detector thus estimated by simulation. Fe 7.6 16.6 13.5 Al 5.2 0.3 0.1 • We considered the source of the neutrons Mn 0.8 3.5 2.9 made from ( α ,n) reaction of U/Th series decay. − 6 × 10 Number of nentrons /chain decay/0.1 MeV • Neutron induced by cosmic muon is negligible. 2.2 Bin width 0.1 MeV JR-1 ( α , n) 2 JA-3 ( α , n) Sample 1 ( α , n) 1.8 Sample 1 + 3% of hydrogen • We picked three types of rocks as samples, 1.6 Sample 2 ( α , n) Sample 3 ( α , n) 1.4 238 Sample 1 ( U fission) they had much different abundance of 1.2 Sample 1 (Th series) 1 chemical compositions. 0.8 Much 
 0.6 difference! 0.4 • The difference affects much the yield of 0.2 0 0 1 2 3 4 5 6 7 neutrons. Neutron energy (MeV) Generated neutron in vary rocks JR-1 and JA-3 are geometrical reference database � 6 Keita Mizukoshi Kobe Univ.

K. Mizukoshi et al., PTEP 123C01 Data driven analysis Thermalization parameter 
 obtained by measurement • We cannot investigate the all wall rock B /R 4 components in details. A Ratio of count rates R 3.5 • Especially amount of water contents in the 3 rock and chemical composition including 2.5 Hydrogen will much affect thermalization of 2 Experimental ratio Error band fast neutrons. 1.5 Sample 1 Sample 2 1 Sample 3 • Thus, thermalization in the rock was JR-1 JA-3 0.5 unknown. 0 1 2 3 4 5 6 7 • We regarded the percentage of hydrogen % of h. e. Experimental ratio v.s. parameter (%of h. e.) in simulation as a thermalization parameter. /s 2 10 2 Counts /MeV /cm 10 Sample 1 with 3% of h.e. • %of h.e. was derived by the experimental JR-1 with 1% of h.e. 1 JA-3 with 1% of h.e. 1/E result (the ratio between setups A and B) 1 − 10 2 − 10 in each rock component. 3 − 10 4 − 10 • The most likely spectra (made from 5 − 10 experimental data) in each sample are 6 − 10 − 7 10 almost same. − 8 10 • This is not affected by uncertainty of 9 8 6 5 3 − − 7 − − 4 − 2 1 2 − − − − 10 10 10 10 10 10 10 10 10 1 10 10 Neutron energy (MeV) Simulation. The most likely spectrum � 7 Keita Mizukoshi Kobe Univ.

Obtained spectrum K. Mizukoshi et al., PTEP 123C01 • We obtained the most likely /s 2 10 2 Counts /MeV /cm The most likely spectrum 10 2% of hydrogen spectrum of the ambient neutron. 1 4% of hydrogen 1/E − 1 10 • We compared the fluxes 
 2 − 10 − 3 10 (the previous study fluxes in other 4 − 10 − 5 10 − 6 10 underground laboratories). Thermal neutron 7 ~0.025eV − 10 − 8 10 • They are the same order of 9 8 6 5 3 − − 7 − − 4 − 2 1 2 − − − − 10 10 10 10 10 10 10 10 10 1 10 10 magnitude. 
 Neutron energy (MeV) The most likely spectrum Neutron fluxes in previous researches It is difficult to compare 
 • Flux ( × 10 -6 cm -2 s -1) Thermal Non-thermal the result simply because 
 there are many difference 
 Kamioka +0.7 +1.2 7.9 ± 0.23 15.6 ± 0.5 (This result, Mizukoshi) -0.7 -1.4 in these measurement 
 Kamioka 8.26 ± 0.58 11.5 ± 1.2 (e.g., detector, assumption 
 (Minamino 2004) Gran Sasso of spectral shape, and 
 13.3 ± 1.5 10.2 ± 1.1 (A. Lindi 1988) ※ definition of flux) LSM 14.3 ± 1.3 4.2 ± 2.8 (K. Eitel 2012) ※ ※ They used the different definition of flux. 
 � 8 We adjusted the same definition of us. Keita Mizukoshi Kobe Univ.

New interest /s 2 10 2 Counts /MeV /cm The most likely spectrum 10 2% of hydrogen 1 4% of hydrogen 1/E 1 − 10 2 − 10 Excess − 3 10 In the previous research, rough • 4 − 10 spectral shape was assumed 
 − 5 10 6 − 10 (e.g., Boltzmann distribution and 1/E). 7 − 10 8 − 10 • The most likely spectrum 
 9 8 6 5 3 − − − 7 − − − 4 − − 2 − 1 2 10 10 10 10 10 10 10 10 10 1 10 10 Neutron energy (MeV) suggests the excess in a The most likely spectrum few MeV. • Since the cross section of • The excess is interesting 
 high energy neutrons is for direct dark matter search. small, it continues to be a • The excess should be high energy neutron. • Once it lose energy, the confirmed by a liquid cross section increases. it scintillator which has a continues to lose energy. sensitivity for the neutron. • Therefore, the excess will • Even such basic information remain at several MeV. has not confirmed… K. Mizukoshi et al., PTEP 123C01 � 9 Keita Mizukoshi Kobe Univ.

Summary • We evaluated an ambient neutron spectrum and + 1.9 obtained the flux (23.5 ± 0.7 stat. sys. × 10 -6 cm -2 s -1 ) - 2.1 at the Kamioka Observatory. using 3 He proportional counter and moderator effectively • with data-driven analysis and simulation • considering systematic errors • • Spectral excess around a few MeV was suggested. It should be confirmed by a sensitive detector for non-thermal neutron. • We are preparing a low BG liquid scintillator. � 10 Keita Mizukoshi Kobe Univ.

Backup slides

Main backgrounds of our experiment cosmic muon • Rate events search experiments are Muon Veto counter placed in underground to reduce BGs ~ Cosmic muon • For Others Shield Detector • Remaining high-energy muon 
 (Pb) ← active veto by scintillator neutron Fiducial 
 • Ambient gamma 
 External 
 Volume (FV) for events ← shield / self-shielding / PSD gamma • Alpha from U/Th chain 
 ← very careful washing • Neutron 
 ← It’s difficult to reduce • Shield with materials which have large cross-section for General ways to reduce BGs neutron in underground experiment • Sometimes neutron makes other BGs (gamma in detector) (Conceptional fig) � 12 Keita Mizukoshi Kobe Univ.

How much neutron is in underground Photo CANDLES Collaboration • I worked for a neutrino-less double beta decay experiment (CANDLES 3+ Experiment) at Kamioka Observatory. • This experiment reduces ambient neutron with Boron shield (~5000 barn for thermal neutron). • Ambient neutron (flux, spectrum) was not understood well. CANDLES Experiment • I would like to show how to for neutrinoless double beta decay measure ambient neutron to using 48 CaF 2 Scintillator demonstrate how difficult to handle neutron as a rare BG. � 13 Keita Mizukoshi Kobe Univ.