Development of A Scintillation Simulation for Carleton EXO Project

Rick UenoUnder supervision of Dr. Kevin Graham

Outline Introduction Theory Detector Design Monte Carlo Simulation Empirical Position and Energy

Reconstruction Algorithm Results Conclusion

Introduction

EXO: Enriched Xenon Observatory Neutrinoless double beta decay (0υββ)

Massive neutrino = Majorana particle? Neutrino hierarchy? Effective Majorana neutrino mass?

Enriched 136Xe gas Both the source and detector Produces scintillation signals

Theory: Neutrino Neutrino = neutrally

charged lepton Suggested by Pauli to

explain continuous spectrum of beta decay

Neutrally charged third “ghost” particle carries some energy away

epn e

Theory: Neutrino Oscillation If neutrinos have mass, then weak eigenstate can be

written as a linear superposition of mass eigenstates

Where Uli is a 3 x 3 neutrino mixing matrix. If tau neutrino is neglected for simplicity:

Transition Probability in the vacuum

ilil U

cossin

sincosU

E

LmP

4sin2sin

222

Theory: 0υββ decay 0υββ decay occur only if

massive neutrinos are Majorana particles

Effective Majorana mass

Measured quantity is half-life of 0υββ decay

j

ejjUmm 2

22

2

2

01

2/1 , mMg

gMZEGT F

A

VGT

SNO and Super-K measures Δm2, but hierarchy is still unknown

Theory: Neutrino Mass

Theory: Xenon Gas Acts as both the source

(produces electrons by the decay process) and a detector (produces scintillation light)

Scintillation process Incoming particle loses

energy to form dimers The de-excitation of dimer

emits photons at wavelength centred around 178nm

http://hepwww.rl.ac.uk/ukdmc/iop98njts/index.htm

Detector Design A simple scintillation counter was designed to

study the scintillation process alone Motivation:

Predicting total light yield of gaseous xenon Reconstruction of position and energy for a

better energy resolution when coupled with the existing TPC (Time Projection Chamber) component

Study of how response varies with different gas mixtures (such as addition of quenching gases)

Detector Design: Overall Design Consists of a stainless steel “T”, two

PMTs on either side, wavelength shifter (WLS) on the PMT window

Detector Design: PMT 136Xe produces UV photon of

178nm Possibility: Regular PMT

with WLS or UV-sensitive PMT

We already have equipment to coat materials with WLS (Tetraphenyl Butadiene, TPB)

MC Simulation: Detector Construction MC Simulation using Geant4 was developed The detector design is simplified to a cylindrical

geometry PMT is represented

by a cylindrical tube with a photo-cathode at the end of a glass plate

y

x

z

12”

MC Simulation:Default Initial Values

Property Values

Photon energy 7.07 eV (≈ 178nm)

Scintillation yield 29000 photons / MeV

Absorption length 100 cm

Prompt scintillation timing constant

2.2 ns

Late scintillation timing constant

4.5 ns

Initial Particle Alpha particle

Initial Momentum Random direction

MC Simulation:Event Detection and Outputs

PMT and WLS has some wavelength-dependent efficiency The simulation should be as realistic as

possible The program reads an input data file

containing efficiency data corresponding to a wavelength

Result is outputted into a data file to be analysed

Empirical Position and Energy Reconstruction Algorithm

Reconstruction of initial position and energy Input: Signal output of two PMTs Output: Reconstructed position and energy of

the particle

ReconstructionProgram

PMT1 Hits, PMT2 Hits

Reconstructed Position and Energy

Empirical Algorithm: Position Reconstruction Looking at the

distribution of ratio between PMT1 and total signal as a function of z position

Gives a smooth curve

Can be readily used to reconstruct the initial position in z direction

11 exp1 zahits

hits

total

PMT

1

1ln

1

az

Empirical Algorithm: Energy Reconstruction Looking at the total

signal normalized by the signal at z = 0 as a function of z position

Can approximate to a 4th order polynomial

Used to estimate the hits if the event occurred at the centre of the detector

432 54321 zczczczccratio

ratio

hitshits total

centre

Empirical Algorithm: Energy Reconstruction

Looking at the total signal at the centre as a function of energy

Gives a linear relation

But a0=0 Rearranging the

equation, initial energy is reconstructed

Eaahitscentre 10

ratioa

hits

a

hitsE totalcentre

11

Empirical Algorithm:Radial dependency

The detected signal as a function of position across the diameter of detector shows deficiency up to ~40% near the wall of the detector

Empirical Algorithm:Radial dependency

Predict that the z-position reconstruction has smaller effect than energy reconstruction

Results of independent test simulations

Three test scenarios were simulated with alpha particles with initial energy of 5.4 MeV at various positions

Results:Test Scenario 1

Starting position at (0,0,0) cm Both reconstructed position and energy

agrees nicely

Results:Test Scenario 2

Starting position at (0,0,-10) cm Both reconstructed position and energy are

fairly consistent

Results:Test Scenario 3

Starting position at (5,0,5) cm Reconstructed energy is much lower than

expected

Results:Summary

Test Scenario

1 2 3

Reconstructed Position (cm)

-0.02354 -10.35 4.571

Sigma 0.3619 0.2947 0.3687

Reconstructed Energy (MeV)

5.41 5.065 4.284

Sigma 0.214 0.2825 0.2134

Conclusion Baseline simulation was developed using

Geant4 The reconstruction algorithm was

developed Works well if the event occurs at the centre Problem when the initial event is off-centre

Future plans Xenon gas and additives Implement into existing TPC system