Status of the light signal simulation for ProtoDUNE-DP€¦ · Simulation: the 3 productions are...

1/23

Status of the light signal simulation for ProtoDUNE-DP

Anne CHAPPUIS – Isabelle DE BONIS

Dual-Phase Photon Detection System Consortium Meeting

September 21th 2017

2/23Anne CHAPPUIS, September 21th 2017 ProtoDUNE-DP

The light signal simulation can be divided in 2 distinct parts:

Scintillation (production of the photons) Light propagation in the detector

This talk focuses on the light propagation simulation: Using pre-calculated maps For 6x6x6m3 and 3x1x1m3 detectors

Simulations performed using the LightSim software based on GEANT4

Introduction

36 PMTs

Outline: Light signal in ProtoDUNE-DP Light map production procedure Light map characteristics Examples of map utilization

6x6x6m3 (fiducial) DLAr TPC @CERN

3x1x1m3 (fiducial) DLAr TPC @CERN

5 PMTs


Light Signal in ProtoDUNE-DP


Light production in Liquid Argon

Scintillation (S1) photon characteristics: λ = 128nm (E = 9.69eV)

Isotropic emission

2 contributions with different lifetimes: τFast = 6ns τSlow = 1600ns

LAr

GAr

e-S1

Direct excitation

scintillation photon(S1 Signal)

Ionisatione- / ion pair creation

electron drifted toward the anode

recombination

Charge particle crossing LAr

Time [ns]0 50 100 150 200 250 300 350 400 450 500

S1 S

igna

l [ph

/5ns

]

0

10

20

30

40

50

60

S1 signal induced by a 10GeV muonFast contribution to S1Slow contribution to S1

ProtoDUNE-DP

Simulation based on the NEST approach(arXiv:1106.1613v1)Not detailed here


For the time being, G is estimated ~300ph/e → Has to be determined more precisely

Light production in Argon Gas Not recombined electrons are drifted toward the top of the detector.

Dual-Phase technology : e- travel through Ar gas.

→ Production of S2 photons:

S2 photon production by electroluminescence:

Simulation: the 3 productions are taken into account via an electroluminescence gain G

LEMs(e- amplification)

Extraction Grid

e-

Anode3

1

2

λ = 128nm (E = 9.69eV) Lifetimes:

τFast = 7ns

τSlow = 3200ns

G = Number of photons produced in GAr / drifted electron

From the LEMs during e- amplification: Main contribution Need additional work and simulations to be precisely

determined

Between LAr surface and LEMs

Between LEMs and anode

1

2

3

Need to estimate the number of photons emitted toward the PMT array

LAr

GAr

e-S1

S2


Light propagation simulation : light maps

7/23Anne CHAPPUIS, 18 May 2017 ProtoDUNE-DP

Implementation of the detailed geometry

Light collection: 36 TPB-coated PMTs

2 options for the PMT positioning:

Non-uniformly spaced (to cover the full fiducial area)

Uniformly spaced (more dense configuration)

LEM Plates

Extraction grid

Field cage

Ground grid

Cathode pipes + supporting structure

Z

X

Y

Implementation of all the components that have an impact on the photon trajectories

Visualization from LightSim

8/23Anne CHAPPUIS, 18 May 2017 ProtoDUNE-DP

Light Propagation in ProtoDUNE-DPPropagation of scintillation photons (128nm):

Absorption in LAr Rayleigh scattering on LAr molecules Absorption on different detector components (field cage, cathode supporting structure, tank…)

Problematic: Very large amount of photons (ex: for a 5-GeV muon track, production of 65.106 scintillation photons) Tracking each scintillation photon takes a lot of time Less than 1% of photons finally reach the PMTs The exact knowledge of all the photon tracks is not needed

Solution: simulate the tracks only once, and store the useful information in Light Maps

Note: these maps describe the photon propagation → Independent from the scintillation parameters

λRayleigh =200m λRayleigh = 55cm(arXiv:1502.04213v2)


Light map production procedureSolution: simulate the tracks only once, and store the useful information in Light Maps

For each photon production point in the detector, and each PMT, the map gives:

Probability to reach the PMT Travel time distribution





LAr and GAr volumes are split in voxels

voxel

Production point





voxel


Generation of N photons in each voxel

Photons are tracked across the detector

Production point Tracking





voxel




Travel time distribution obtained for each PMT

Production point Tracking Travel time distribution

Entries 1566

Mean 40.98

Travel time [ns]0 50 100 150 200 250 300 350 400

Phot

ons

reac

hing

PM

T21

0 20 40 60 80

100 120 140 160 180 200 220 240

Entries 1566

Mean 40.98





voxel




Travel time distribution obtained for each PMT

Extraction of time distribution parameters Light Maps

Production point Tracking Travel time distribution

Entries 1566

Mean 40.98

Travel time [ns]0 50 100 150 200 250 300 350 400

Phot

ons

reac

hing

PM

T21

0 20 40 60 80

100 120 140 160 180 200 220 240

Entries 1566

Mean 40.98


Light map characteristics


LAr maps: Large voxels definition: 250mmx250mmx250mm Number of generated photons per voxel: 107 over 4π

GAr maps: Voxel definition: 250mmx250mmx5mm Only 1 voxel in Z Number of generated photons per voxel: 5.108 over 4π

✔ To save time, photons are generated in ~1/8 of the detector, then we use the X-Y symmetry of the detector to reconstruct the whole map.

→ Gain of a factor 8 in the execution time.

Propagation parameters in LAr

LAr absorption process is not included in the map generation

→ Will be parametrized when using the maps

ProtoDUNE-DP light map characteristics

Z

X

Y

X


Travel time distribution characteristics Probability to reach the PMT:

Travel time distribution shape: strongly depends on the distance to the PMT

weight = w0 =Number of photons reaching the PMT

Number of generated photons

Travel time [ns]0 50 100 150 200 250 300

Phot

ons

hitti

ng th

e PM

T

1

10

210

PMT14

PMT17

Photons produced 1m above the cathode pipes

Travel time [ns]0 50 100 150 200 250 300

Phot

ons

hitti

ng P

MT1

4

1

10

210Photons produced:

~1m above the cathode pipes~2.5m above the cathode pipes

Z

→ Find a general parametrisation with the minimum number of parameters

ProtoDUNE-DP

ProtoDUNE-DP

12 28

7 17 23

11 16 22 27

10 15 21 26

6 14 20

9 25

31

30

8 18 24 32

5 13 19 29

4

3

2

1

36

35

34

33

production point

PMT Grid


Travel time distribution characteristics

→ Satisfactory in most cases

→ Finally, 4 parameters are stored in the maps (ROOT TH3 objects)

Travel time [ns]0 50 100 150 200 250

Phot

ons

hitti

ng P

MT

14

0

20

40

60

80

100

LightSim simulationReconstructed with Laundau fit

b) Reconstruction for photons generated at

(-1125, -1375, -2080)mm

χ2=1.08

ProtoDUNE-DP

The time distribution is reconstructed using a landau fit and 3 parameters: t0 (the first bin Nentries > 0) The MPV and σ of the landau fit Voxels with Nentries<50 are not taken into account

Mean 0.9387

RMS 0.9731

/DoF2χ0 1 2 3 4 5 6 7 8 9 100

200

400

600

800

1000

1200

Mean 0.9387

RMS 0.9731

χ2 distribution

6.2% of voxels with χ2 > 2.5: voxels close to the PMT

ProtoDUNE-DP

Peak at χ2<0.4: voxels with a small number of collected photons

LAr light maps (ProtoDUNE-DP)


Light maps for the (3x1x1)m3 prototypeSame work as been done for the (3x1x1)m3 prototype

108 photons generated in each voxel(~5h per voxel)

Same voxel definition: (25x25x25)cm3

Implementation of the detailed geometry

3 TPB-coated PMTs 2 PMT with TPB-coated plate above them

Same LAr propagation parameters

Adaptation to a smaller volume:

Narrow travel time distributions

Use of a different travel time parametrization: Double-exponential parametrization 5 parameters: w0, tstart, τ1, τ2 and relative

normalization (between the two exponential)

Entries 11373

Travel time [ns]0 10 20 30 40 50 60

Phot

ons

reac

hing

the

PMT

0

500

1000

1500

2000

2500

Entries 11373

Example of travel time distribution (3x1x1)


Scintillation light signal from light maps


Light signal using light maps

LAr

GAr

e-S1

S2Generation of cosmic muons or beam particles.

At each step: simulation of the scintillation process

→ Number of scintillation photons and drifted electrons

Tracking of the particles across the detector

Number of detected S1 photons:

w0 from the maps LAr absorption length

(PMT quantum efficiency)

Number of detected S2 photons:

w0 from the maps LAr absorption length

Electroluminescence gain (PMT quantum efficiency)

Hit time of each photon = vertex time +scintillation time (+ drifted time) + travel time

Parameters from the maps

Lifetimes τFast and τSlow

exp(−ttravel⋅c

λAbs⋅nLAr

)

Abs = Absorption lengthnLAr = Refractive index

Parametrization of the LAr absorption:

if S2 photons


Map values interpolation

Interpolation between the 8 nearest voxels using the weighted average method (ROOT interpolation)

Interpolation of the map values to the true coordinates of the photon production point

→ The interpolation is satisfactory

→ The voxelisation is suitable

MPV map for PMT11 at z=-1080mm Interpolated map

(6x6x6)m3(6x6x6)m3


Studies using light maps

Time [ns]4000− 3000− 2000− 1000− 0 1000 2000 3000 4000

310×

Lig

ht s

igna

l [ph

/400

ns]

0

200

400

600

800

1000

1200

1400Light signalLAr contribution (prompt signal)

Study of the PMT positioning impact on light signal

Study and rejection of the cosmic muon background

These simulations and the map generation allow different studies for the collaboration: Impact of different designs on the light collection Tag and rejection of the cosmic muon background

Xcoordinate [mm]3000 2500 2000 1500 1000 500 0

Num

ber o

f ph

oton

s re

achi

ng th

e PM

T ar

ray

410

PMTs Uniformly spaced (65cm)

PMTs Nonuniformly spaced

For 107 photons generated 1m above the cathode pipes

→ Ongoing studies

λAbs = 30m (arXiv:1306.4605V2)


The light propagation is simulated using pre-calculated “light maps”

Maps have been produced for (6x6x6)m3 and (3x1x1)m3 detectors Maps for LAr (S1) and GAr (S2) light signal Containing the detailed geometry of the detector For the ProtoDUNE-DP map: 2 options for the PMT positioning

2 different travel time distribution parametrizations (Landau and double-exponential)

→ The detector scale have a non-negligible impact on the travel time distribution shapes

Given the map format (TH3 objects), the maps can be used independently from the library that produce them

Maps implemented in QScan, and the implementation in LArSoft is ongoing (CIEMAT people)

Maps are currently used to perform light signal studies (cosmic background, impact of LAr propagation parameters…)

Foreseen: comparison with (3x1x1)m3 data

Difference between coated-PMT and coated-plate above PMTs S2 estimation (w.r.t to extraction and LEM amplification fields) ...

Summary

Status of the light signal simulation for ProtoDUNE-DP€¦ · Simulation: the 3 productions are...

Documents

Transcript of Status of the light signal simulation for ProtoDUNE-DP€¦ · Simulation: the 3 productions are...