The PANDA Barrel DIRC Detector at FAIR

Roman Dzhygadlofor the PANDA Cherenkov Group

IEEE 2018 conference

PANDA experiment Barrel DIRC design Expected performance Validation in beam tests Summary & outlook

The PANDA Cherenkov Group:

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 2/19

The PANDA Experiment at FAIR

High Energy Storage Ring 5 x 1010 stored cooled antiprotons 1.5 to 15 GeV/c momentum Cluster jet / pellet target High luminosity mode

Δp/p ≈ 10-4 (stochastic cooling) L = 1.6 x 1032 cm-2s-1

High resolution modeΔp/p ≈ 5 x 10-5 (electron cooling) L = 1.6 x 1031 cm-2s-1

Facility for Antiproton and Ion Research at GSI (German national lab)

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 3/19

Study of QCD with Antiprotons

PANDA Physics Program

Excellent PID requiredExcellent PID required

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 4/19

DIRCs in PANDATwo DIRC detectors for hadronic PID: Barrel DIRC

Goal: 3 s.d. π/K separation up to 3.5 GeV/c Endcap Disc DIRC

Goal: 3 s.d. π/K separation up to 4 GeV/c

Barrel DIRC

Disc DIRC

antiproton momentum: 7 GeV/cEvtGen kaon phase space example

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 5/19

Detection of Internally Reflected Cherenkov Light

Charged particle traversing radiator with refractive index (n

1 ≈ 1.47) and β = v/c > 1/n

emits Cherenkov photons on cone with half opening angle cos θc = 1/βn(λ)

Some photons are always totally internally reflected for β≈1 tracks

Radiator and light guide: polished, long rectangular bar made from Synthetic Fused Silica (“Quartz”)

Proven to work (BABAR-DIRC)

Novel type of Ring Imaging CHerenkov detector based on total

internal reflection of Cherenkov light

DIRC Principle

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 6/19

Barrel DIRC Design

Conservative design: similar to BABAR DIRC, baseline design for TDRExcellent performance, robust, little sensitivity to backgrounds and timing deterioration

Based on BABAR DIRC with key improvements (compact fused silica prisms, spherical lenses)

48 radiator bars (16 sectors), synthetic fused silica17mm (T) x 53mm (W) x 2400mm (L)

Focusing optics: triplet spherical lens system Compact expansion volume:

30cm-deep solid fused silical prisms Photon detectors:

Micro-Channel Photo Multipliers, ~11k channels Fast FPGA-based photon detection

~100ps per photon timing resolution Expected performance (simulation and particle beams):

better than 3 s.d. π/K separation for entire acceptance

Flat mirrorRadiator bars

Expansion volume

Focusingoptics

Photondetectors

Electronics

240cm

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 7/19

SimulationQuantum efficiency of different MCP-PMTs

Geant4 simulations includes: Realistic materials properties Photons transport efficiency Single photon time resolution Quantum and collection efficiency Dark counts

Accumulated hit pattern from 1000 K+

at 3.5 GeV/c and 25o polar angle

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 8/19

Reconstruction MethodsGeometrical BABAR-like Needs Look-Up Tables

pixe

l

pixe

l

detection timepion candidate kaon candidate

PDFs of kaon and pion

for given pixelTime imaging Belle II TOP-like Needs Probability Density

Functions of the propagation time

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 9/19

Expected Performance

(using geometrical reconstruction)

Photon yield Single Photon Cherenkov angle

resolution (SPR)GEANT simulationBaseline design, 3 bars per bar box, 3-layer spherical lens

Cherenkov track resolution:

tracking resolution 2-3 mrad

Yield and SPR reach performance goal

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 10/19

Expected Performancetrack-by-track max. likelihood fit

Design meets and exceeds PID requirements

from earlier: kaon phase space for 7 GeV/c

Geant simulation

Barrel DIRC PID 3 s.d. goal

Time imaging reconstruction

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 11/19

Key components Focusing lens Designed several spherical and cylindrical lenses, several prototypes build by industry

prism

simulation:

focal plane

of 3-layer lens

bar

pris

m

Micro-channel Plate Photomultipliersexcellent timing and magnetic field performance; used to have issues with rate capability and aging, now solved

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 12/19

Beam Tests at GSI and CERN 2008-2014 – test of different focusing systems, expansion volumes

2015 – validation of the design with narrow bar, 3 layer spherical lens and 3x5 MCP-PMTs

2016 – validation of the alternative design with a wide plate radiator (cost saving)

Cherenkov angle resolution per track:

~3.5 s.d. π/K

@ 3.5 GeV/c

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 13/19

TOF particle identification

π/p

beam

TO

F2

TO

F1

T3

T2

T1DIRC

Beam profiling Trigger

CERN PS/T9 area

HO

DO

SC

OP

E

Beam Test at CERN 2017-2018External TOF PID

7 GeV/c

Most of the data taken at 7 GeV/c(7 GeV/c π/p sep. ≈ 3.5 GeV/c π/K)

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 14/19

Beam Test at CERN 2017-2018PHOTONIS PlanaconMCP-PMT array

Spherical 3 layer lens

narrow bar (35 mm width)33O prism

2017 prototype

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 15/19

Beam Test at CERN 2017-2018 Goal: validate near final design Narrow bar (17x32x1250 mm3 ) Fused silica prism

Hit patterns, pion tag:4x2 MCP-PMTs4x3 MCP-PMTs

4x3 and 4x2 MCP-PMT array Focusing with 3-layer spherical lens ~200 ps time resolution

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 16/19

Beam Test at CERN 2018 pions protons

SPR = 8.6 mrad

beam data

π/p @ 7 GeV/c

π/p separation power @ 7 GeV/c

Good performance

Good agreement with Geant simulations

photon multiplicitysingle photon resolution

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 17/19

Summary The PANDA Barrel DIRC is a key component

of the PANDA PID system

Design features narrow bars and 3-layer spherical lens and compact expansion volume

Simulations predict more than 3 s.d. π/K separation up to 3.5 GeV/c

Successfully validated PID performance in particle beams

TDR available JINST 13 C03004, DOI:10.1088/1748-0221/13/03/C03004

Design with 4x2 MCP-PMTs layout considered as a cost saving option

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 18/19

Outlook2018-2024:Component Fabrication, Assembly, Installation

2018: Finalize specifications, call for tenders and contracts

2019-2021: Industrial fabrication of fused silica bars and prisms

Industrial production of MCP-PMTs

2019-2020: Production and QA of readout electronics

2019-2022: Industrial fabrication of bar containers and mechanical support frame,

gluing of bars/plates, construction of complete bar boxes, detailed scans of all sensors

Assembly of readout units

2023-2024: Installation of the Barrel DIRC into the PANDA detector

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 19/19

Thank you for the attention

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 20/19

Backup slides

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 21/19

Simulation and Reconstruction

Reconstruction

Geometry/material

Front-coated

mirror

Eljen EJ-550 optical grease

Epotek 301-2 glue

Expansion volume (Tank/Prism)

Radiator (narrow bars/plate)

Focusing (different lenses)

MCP-PMT

Gean3, Geant4

Hit Finder

Digitization

Particle

transport

Event

generation

IEEE NSS | Sydney 13.11.18 Roman Dzhygadlo 22/19

Readout and Mechanical Design

Mechanical Design● Light-weight and modular, allows staged bar

box installation, access to inner detectors.

● Mechanical support elements made from

aluminum alloy or carbon fiber (CFRP)

● Boil-off nitrogen flush for optical surfaces

Readout Electronics● ~100ps timing per photon for small MCP-PMT

pulses – amplification and bandwith optimization

● 20MHz average interaction, trigger-less DAQ

● Current approach: HADES TRBv3 board with

PADIWA amplifier/discriminator

● Near future: DiRICH, integrated backplane,

joint development with HADES/CBM RICH

DiRICH, PADIWA,and TRB3