Simulation for Daya Bay Detectors Liangjian Wen Institute

Simulation for Daya. Bay Detectors Liangjian Wen Institute of High Energy Physics Sep 19, 2008 Topical Seminar on Frontier of Particle Physics 2008: Neutrino Physics and Astrophysics 1

Daya. Bay Neutrino Experiment • Daya. Bay Neutrino Experiment is a precise measurement to with a near-far detector configuration. Small-amplitude oscillation due to 13 Determine with a sensitivity to 0. 01 (90% C. L) Large-amplitude oscillation due to 12 2

3

• Detector layout (Far site) Water Pool Inner Water Shield Outer Water Shield (separated by tyvek) 3 -zone AD module Target: Gd-doped scintillator g-catcher: normal scintillator Buffer: Oil 4

Daya. Bay simulation • Our simulation is based on Geant 4 (G 4 dyb), but with modifications and extensions to accommodate the specific requirements of the Daya. Bay Experiment. • We did many validations for our simulation by Ø Comparison with other simulations (FLUKA, MCNPX, Geant 3) Ø Comparison with previous measurement data Ø Comparison with our prototype experiment data 5

Neutrino Detection at Daya. Bay Inverse reaction in Gadolinium-doped Liquid Scintillator (Gd. LS) Two critical things in simulation ØNeutron capture processes ØOptical model of the detector. 6

Neutron Capture Process • We validate the neutron capture cross section for Gd/H/C targets in Geant 4 with Geant 3 simulation and experimental data. • Geant 4 is incorrect for the ncapture final state with multiple gammas. It is hard to give a general solution for all ncapture targets in Geant 4. • We modified the G 4 neutron capture processes for Gd/C/H targets, based on experimental spectrum. Simulated and measured gamma energy spectrum for n-Gd capture 7

Scintillation Process: Quenching • No quenching effect in G 4 Scintillation • We write our own scintillation process with the quenching effect be considered (J. B. Birks’ law): • We measured the quenching factor for the Daya. Bay Gd. LS and LS: Gd. LS (ppo, bis-MSB, LAB, 0. 1% Gd) 6. 49(± 1. 06) LS (ppo, bis-MSB, LAB) 8. 21(± 1. 23) unit : 8

Scintillation Process: Re-emission • Use the measured Gd. LS/LS emission spectrum in scintillation processes simulation • Re-emission of Cerenkov light and scintillation light are very important. • Currently an assumed re-emission possibility spectrum is used and we are doing the measurement for our Gd. LS/LS. 9

Other Optical Properties in AD • Detector simulation needs inputs of the optical properties of the detector components from measurements. Ø Ø Ø Light yield Emission spectrum Absorption length for Gd. LS/LS/Oil Refractive index of acrylic vessel Reflectivity of top/bottom reflectors PMT Quantum Efficiency spectrum 10

• Requirements on uncertainties necessary to achieve the 0. 01 sensitivity goal. 11

Positron & Neutron Efficiency • Positron event：Evis>1 Me. V Ø Efficiency = 99. 8% Ø Error ~ 0. 05% assuming 2% energy scale error • Neutron event：Evis>6 Me. V Ø Efficiency ~ 90. 7% (overall) Ø Error ~ 0. 2% assuming 1% energy scale error 12

3 Zone detector Neutron efficiency v. s LS thickness 42. 5 cm, 91% Sensitivity v. s target mass 4 x 20 ton 15 cm Detector response for different e - positions in detector 13

Reflector Simulation & Event Reconstruction no reflector simulation with reflector s. E/E = 12%/ E sr = 13 cm reconstruction 14

• Prototype simulation v. s data 60 Co 137 Cs 15

Muon Simulation Modified Gaisser Formula + MUSIC Modified Gaisser Formula: More reasonable muon flux parameterization at sea level MUSIC (MUon SImulation Code) : A three-dimensional code transports muons through the rock to underground lab 16

Muon detection efficiency OWS: threshold 13 PMT, efficiency 97. 7% IWS: threshold 11 PMT, efficiency 98. 1% 17

Neutron Simulation • Muon-induced fast neutron background is an important background. Its rates calculated using Geant 4 are consistent with earlier Geant 3 calculations that used empirical parametrization of measurements, accurate to ~20% (Y. Wang et al. , PRD 64, 013012 (2001)). n n Energy spectrum of fast neutron backgrounds 18

Muon capture • Muons ( ) stop in water pool/antineutrino detector can Ø Decay -> Michel electron (lifetime 2. 2 ) Ø Capture on C, O, Fe and emit a neutron (fast neutron background) Ø Capture on C and form a 12 B (delayed signal) • Muon stopping ratio from Geant 4 simulation is consistent with FLUKA simulation. • Its capture rate on C, O, Fe from Geant 4 simulation is consistent with experimental data. • Yet the neutron energy spectra given by Geant 4 is not in good agreement with measurements. So we implemented new neutron spectrum according to experimental data. 19

Radioactivity Generator • Full decay chains of U, Th, K simulated to provide HEPevt input to G 4 dyb ( U, Th using programs of A. Piepke) • 60 Co simulation for calibration source and background • 68 Ge, 252 Cf and Pu-C simulation for calibration sources 252 Cf neutron from Pu-C prompt signal 252 Cf delayed signal 20

• Natural radioactivity simulation, give material specification for detector construction Ø Ø From Stainless Steel tank From Gd. LS (radioactivity accompany with Gd) From PMT From rocks or water (Radon) Ø… 21

Software Framework Nu. Wa : our Gaudi-based offline software 22

Summary • Daya. Bay simulation is based on Geant 4, and we made extensions and modifications to accommodate the specific requirements of the Daya. Bay Experiment. • Many validations have been done for current simulation. • Important simulations are based on G 4 dyb Ø Positron and neutron efficiency and its error estimation Ø Detector design related issues 23