Compact segmented antineutrino detector

Antonin Vacheret for the SoLid collaboration

High Energy Physics group Blackett Lab

Imperial College London

SCK•CEN BR2 reactor building

Very short baseline experiment• Upcoming experiments require percent level precision in

antineutrino spectrum measurement

2

Δm2=2.35 eV2 sin22θee = 0.165

2 3 4 5 6 7 8

[/25

0 ke

V]

νN

1000

2000

3000

4000

5000

U 235Mueller et al.

SoLid 50 days - DYB flux

Antineutrino Spectrum

Energy [MeV]2 3 4 5 6 7 80.85

0.9

0.95

1

1.05

1.1

[MeV]

Oscillation search Spectrum shape

•detector located close to reactor core •operating on the surface

Sim

Precise measurements at reactor

3

Gd-doped liquid scintillatortechnology

•detection using IBD reaction

⌫̄e + p ! e+ + n

Gd-doped liquid scintillatortechnology

Gd-doped liquid scintillator technology

•underground laboratory

•large external shielding

•homogenous, well contained energy

•achieve percent level antineutrino flux measurement at PWR

~ 100 m - 1 km baseline

Challenges:

•sensitivity in E and L for oscillation search

•rejection of background

•security and safety constraint on site

Precise measurements at reactor

4

Gd-doped liquid scintillatortechnology

Gd-doped liquid scintillator technology

•Underground laboratory

•Large external shielding

•Well contained energy

•achieve percent level antineutrino flux measurement at PWR

~ 100 m - 1 km baseline ~ 10 m baseline

Precise measurements at reactor

5

Gd-doped liquid scintillatortechnology

Gd-doped liquid scintillator technology

•Underground laboratory

•Large external shielding

•Well contained energy

•achieve percent level antineutrino flux measurement at PWR

Highly segmented detector

sufficiently compact to deploy meters away from core

~ 100 m - 1 km baseline ~ 10 m baseline

Challenges:

•sensitivity in E and L for oscillation search

•rejection of background

•security and safety constraint on site

Segmented neutrino detectors

Liquid scintillator

6

Extruded plastic scintillator

Current generation

ILL-81 LS cells 3He tubes Bugey 3

LS + Li bars

Karmen LS cells Gd sheets

MINOS extruded plastic & WLS fibre technique

T2K near detector extruded plastic WLS fibres MPPCs photo-sensors

NOVA Extruded plastic cells + LS APD sensors

DANSS Extruded plastic & WLS fibre Gd coating

STEREO LS+Gd cells

PROSPECT LS+Li bars

3D segmented composite detector• composite /dual scintillator detector

element :

• 5 cm x 5 cm x 5 cm PVT cube segmentation to contain positron energy and localise interaction

• Layer of LiF:ZnS(Ag) for neutron detection close to interaction

• WLS fibre to collect both scintillation light in X and Y direction

• each cube voxel optically separated from each other by reflective coating

• SiPM to read out fibre signal

7

n

6Li4He

3H

Etot = 4.78 MeV

Signal localisation

• Positron energy contained in cube voxel

• Neutron capture efficiency uniform up to the edge of the detector

• Neutron capture one cube away from interaction gives directional sensitivity

8

Cube plane number0 5 10 15 20

Frac

tion

0

0.2

0.4

0.6

0.8

1

1.2n capture

)+(edepE)γ(annih depE

Sim

Sim

Sim

Visible energy reconstruction

9

Positron energy (MeV)0 1 2 3 4 5 6 7 8 9 10

Abso

rbed

ene

rgy

(MeV

)

0

1

2

3

4

5

6

7

8

9

10

1

10

210

310

• Summing all energy visible in 1 m3 detector • gamma-ray leakage affects energy reconstruction

e+ + 2𝛾Sim

Visible energy reconstruction

10

Positron energy (MeV)0 1 2 3 4 5 6 7 8 9 10

Abso

rbed

ene

rgy

(MeV

)

0

1

2

3

4

5

6

7

8

9

10

1

10

210

310e+ only

• Energy estimation recovered by selecting two highest energy cube

Sim

Energy reconstruction

11

Absorbed energy (MeV)0 1 2 3 4 5 6 7 8 9 10

a.u

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

initial energy+e

Sum of 2 cubes energy

Energy (MeV)0 1 2 3 4 5 6 7 8 9 10

Dev

iatio

n

0.2−

0.1−

00.1

Sim

Challenges of the design

• composite scintillator

• ligh collection with WLS fibres

• particle identification

• energy reconstruction

• uniformity

• data size

• efficiency

12

NEMENIX 2013-14

• small prototype: 64 voxels, 32 channels, 8 kg mass

13

Data

Data

n

EM

Real scale system SM1

14

80 cm 20 cm

• NEMENIX 8kg 64 voxels, 32 chan.

• SoLid Module 1 (SM1) 288kg

2 304 voxels, 288 chan. 9 detector planes

2010- 2013

36x

2014-2015

Proof of concept and small prototype

80 cm

Learning how to build it

15

cube filling

Commissioning at Gent November 2014

Frame machining Sensor PCB productionElectronics

Detector assembly

Target uncertainty

• Solid target intrinsically stable in time

• All 2304 cubes weighed

• Extracted Np

• < 1% uncertainty

16

Deployment at BR2

17

Energy response calibration

• Energy response

18

Data

SoLid preliminary

• PVT response intercalibrated using muons

• cube response equalised to better than 1% for majority of channels

• stability over time of energy scale ~ 1%

Neutron ID and capture time

• Validated PID, neutron tranport simulation (MCNP & G4) and Li capture efficiency

19

t (ns)∆

0 100 200 300 400 500310×

norm

aliz

ed c

ount

s

3−10

2−10

1−10

1 G4 simulationMeasured data

prompt to neutron capture time difference (AmBe source)

fitting results (G4):)2τt/∆exp(-

2) + a1τt/∆exp(-1 + a0a

sµ 0.01 ± = 10.82 1τ

sµ 0.02 ± = 91.36 2τ

= 1.09norm2χ

fitting results (Data):sµ 0.28 ± = 11.30 1τ

sµ 0.31 ± = 90.39 2τ

= 27.98norm2χ

SoLid preliminary

SoLid preliminary

Data 1h

AmBe source

n

EM

Signal analysis

• Demonstrated power of segmentation on background rejection

20

ΔRn

IBD candidate

21

prompt

delayed

Data

Signal analysis• Demonstrated power of segmentation on background rejection

• but SM1 had limited shielding and lower absolute neutron efficiency of ~ 2.5% due to high data rate

22

~10x

~100x

SoLid preliminary

23

• SoLid Module 1 (SM1) 288kg

2 304 voxels, 288 chan. 9 Detector planes

limited performance data rate 0.45 TB/day

• 5x modules 1.6 tonnes 12 000 voxels,

3 200 read out channels high performance

data rate max 0.5 TB/day

5x

2014-2015 Phase-I 2016-2017 SoLi∂

2.5 m

0.8 m

Improvements for

• 4 fibre read out

• 37 PA/cube/MeV +66% increase in light yield from SM1

• on target for 14%/sqrt(E) resolution

• 7% total variation of light yield across detector planes

24

Light yield and uniformity of response

Neutron capture efficiency

• Additional LiF:ZnS sheets

•6Li capture efficiency 0.55 to 0.7 +30%

• Reduced capture time 105 to 66 us

• new screens with improved transparency

SoLi∂

Trigger and efficiency

• Neutron signal : large number of photons but distributed in time and large range of light output

• in SM1 direct threshold had to be set to 6.5 PA to limit data rate and required two channel in coincidence

• reduced neutron detection efficiency to 5%

• Neutron trigger implementation

• limit reactor ON/OFF data sizes and rate

• maximise neutron and IBD efficiency

• reduce rate dependence to threshold

25

Time (ns)0 50 100 150 200 250 300 350

Sign

al (P

A)

0

2

4

6

8

10

Maximum Amplitude (PA)0 5 10 15 20 25 30 35 40

Trig

gers

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000Entries 146510Mean 6.8423RMS 8.1426

AmBe test data

EMn

Neutron trigger and data size

26

Number of Peaks0 10 20 30 40 50 60

Trig

gers

0

100

200

300

400

500

600

700

800

900

1000Entries 146510Mean 4.7517RMS 3.0851

AmBe test data

EMn

• Neutron pattern recognition in firmware

• neutron rate is low: Rn ~ 7 Hz

• Buffer time ±500 us and ±2 planes around neutron

• can recover neutron detection efficiency from 5% to 70% !

• remove inefficiencies of forming coincidence

• Zero suppression threshold at 1.5 PA applied to other signals

• limit data size and storage

• Detector cooling to 5 deg to reduce dark counts AmBe test data

Detector calibration

27

off-site XY calibration system (CALIPSO) plane characterisation neutrons and EM like signals precise cube to cube equalisation

in-situ calibration system (CROSS) Energy scale determination (% level) Absolute neutron detection efficiency (a few % level or better)

Conclusion

• upcoming VSBL experiments require a detector technology suited for high precision measurement very close to reactor core

• a highly segmented detector based on solid scintillators provides higher containment, alternative energy reconstruction and background rejection based on topology

• deployment of real scale module shows that percent level measurement is feasible with a highly segmented target

• scaling up to a larger system with high efficiency requires a neutron trigger scheme and zero suppression of data

• detector could achieve 30% IBD efficiency based on latest optical improvements, simple IBD selection and front-end trigger developments

• automated calibration system and remote control to further reduce on-site maintenance and operation

28

Outlook

• SoLid is the first step towards a cost-effective robust segmented detector

• directionality sensitivity still to be explored

• room to increase performance in the next 2-10 years:

• transmission design (Lattice): CHANDLER, NuLAT to improve the energy resolution

• Adding more Lithium or new materials to raise neutron efficiency ( >90% ?)

• low maintenance integrated design would enable easier future deployment:

• applications underground

• network of sensors

29