Contalbrigo Marco INFN Ferrara On behalf of the

Contalbrigo Marco INFN Ferrara On behalf of the CLAS 12 RICH Group RICH 2013 Workshop, 2 nd December 2013, SVC - Japan

CEBAF Upgrade at Jefferson Lab add Hall D (and beam line) Beam Energy 12 Ge. V Beam current 90 m. A Beam polarization 85 % Upgrade magnets and power supplies CHL-2 Enhance equipment in existing halls Ø Ø Nucleon structure & Nuclear effects Hybrid mesons (gluonic excitations) Low-energy tests of SM Heavy photon search Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 2

The CLAS 12 Spectrometer Ongoing upgrade of the CLAS detector. First beam expected in 2016. Highly polarized 12 Ge. V electron beam Luminosity up to 1035 cm-2 s-1 Beam H and D polarized targets Broad kinematic range coverage (current to target fragmentation) 3 D structure of the nucleon by polarized deep-inelastic scattering Hadron ID wanted for flavor separation Crucial for the study of parton dynamics related to angular momentum and spin-orbit effects with flavor sensitivity. Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 3

The CLAS 12 Spectrometer Ongoing upgrade of the CLAS detector. First beam expected in 2016. CH Luminosity up to 1035 cm-2 s-1 RI Highly polarized 12 Ge. V electron beam Beam H and D polarized targets R I CH Broad kinematic range coverage (current to target fragmentation) 3 D structure of the nucleon by polarized deep-inelastic scattering Hadron ID wanted for flavor separation Crucial for the study of parton dynamics related to angular momentum and spin-orbit effects with flavor sensitivity. Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 4

The CLAS 12 Spectrometer Ongoing upgrade of the CLAS detector. First beam expected in 2016. Highly polarized 12 Ge. V electron beam Luminosity up to 1035 cm-2 s-1 H and D polarized targets Broad kinematic range coverage (current to target fragmentation) 2 sectors to accomplish physics program, 1 st sector by the end of 2016 3 D structure of the nucleon by polarized deep-inelastic scattering Hadron ID wanted for flavor separation Crucial for the study of parton dynamics related to angular momentum and spin-orbit effects with flavor sensitivity. Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 5

The CLAS 12 RICH Requirements u cover 3 -8 Ge. V/c momentum range up to 25 degrees u 1: 10 kaon w. r. t. pion flux p rejection 1: 500 required Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 6

The CLAS 12 RICH Requirements u cover 3 -8 Ge. V/c momentum range up to 25 degrees u 1: 10 kaon w. r. t. pion flux p rejection 1: 500 required u Aerogel to separate hadrons with the required rejection factors collection of visible Cherenkov light u Use of PMTs: challenging project, need to minimize the detector area covered with expensive photo-detectors Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 7

The Hybrid Optics Design Direct rings and best performance for high momentum particles plane mirror 2 cm 6 cm spherical mirror gap aerogel g p photon detector 3. 7 m 4. 2 m p 1. 2 m m Bea Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 8

The Hybrid Optics Design Direct rings and best performance for high momentum particles plane mirror 2 cm 6 cm spherical mirror gap g p aerogel photon detector Reflected rings for less demanding low momentum particles p 6 cm 4. 2 m p 1. 2 m Bea Contalbrigo M. m • • gap 2 cm 3. 7 m plane mirror g spherical mirror aerogel photon detector Minimize active area (cost) to about 1 m 2 Material budget concentrated where TOF is less effective Focalizing mirrors allow thick radiator for good light yield Time resolution < 1 ns to distinguish direct and reflected patterns RICH 2013, 2 nd December 2013, SVC - Japan 9

The Hybrid Optics Design Direct rings and best performance for high momentum particles plane mirror 2 cm 6 cm spherical mirror gap g p aerogel photon detector Reflected rings for less demanding low momentum particles 6 cm Spherical mirrors plane mirror Photodetectors Contalbrigo M. • • g spherical mirror gap 2 cm Planar mirrors p Aerogel aerogel photon detector Minimize active area (cost) to about 1 m 2 Material budget concentrated where TOF is less effective Focalizing mirrors allow thick radiator for good light yield Time resolution < 1 ns to distinguish direct and reflected patterns RICH 2013, 2 nd December 2013, SVC - Japan 10

Aerogel Radiator Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 11

Aerogel Characterization Achieved ~ 0. 00050 mm 4 cm-1 clarity for 115 x 11. 5 cm 2 tiles at n=1. 05* (comparable with LHCb at n=1. 03) *Budker and Boreskov Institutes of Novosibirsk Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 12

Aerogel Characterization Achieved ~ 0. 00050 mm 4 cm-1 clarity for 115 x 11. 5 cm 2 tiles at n=1. 05* (comparable with LHCb at n=1. 03) *Budker and Boreskov Institutes of Novosibirsk Refractive index map Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 13

Aerogel Chromatic Dispersion Measured by prisma method: Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 14

Aerogel Chromatic Dispersion Measured by prisma method: Measured by prototype with optical filters: fit MC Expected value from density: n 2(400 nm) = 1+0. 438 r n(400 nm) = 1. 0492 Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 15

Photon Detectors: MA-PMT The only option to keep the schedule is the use of multi-anode photomultipliers (we consider the promising Si. PM technology as the alternative) 28 H 8500: 14 with standard window (H 8500 C) 14 with UV window (H 8500 C-03) 2 H 12700: za 003 and za 0014 Contalbrigo M. ✓ Mature and reliable technology ✓ Large Area (5 x 5 cm 2) ✓ High packing density (89 %) ✓ 64 6 x 6 mm 2 pixels cost effective device ✓ High sensitivity on visible towards UV light ✓ Fast response Torus fringe fields in the RICH photo-detector area RICH 2013, 2 nd December 2013, SVC - Japan 16

MA-PMT Gain Map H 8500 Gain SPE Loss Up to 1: 4 pixel gain variation can be compensated by the read-out electronics Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 17

MA-PMT Gain Map H 8500 Gain SPE Loss H 12700 1: 2 gain variation Contalbrigo M. Up to 1: 4 pixel gain variation can be compensated by the read-out electronics RICH 2013, 2 nd December 2013, SVC - Japan 18

MA-PMT SPE Loss Gain SPE Loss SPE loss limited to ~15% above 1040 V and almost uniform over 28 MA-PMTs H 8500 Pixel SPE Loss Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 19

Gain x 106 MA-PMT SPE Loss HV = 1000 V MA-PMT Gain SPE Loss SPE loss limited to ~15% above 1040 V and almost uniform over 28 MA-PMTs H 12700 features a significant better SPE resolution than H 8500 at similar gain H 8500 Pixel SPE Loss Contalbrigo M. HV = 1040 V RICH 2013, 2 nd December 2013, SVC - Japan 20

MAROC 3 Front-End Electronics (ATLAS) 64 binary outputs with time jitter ~300 ps Noise (Rate saturated) Self-trigger (dark counts) gain equalization & threshold selection Laser Light SPE Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 21

MAROC 3 Front-End Electronics Multiplexed analog output (ATLAS) 64 binary outputs with time jitter ~300 ps Noise (Rate saturated) Self-trigger (dark counts) gain equalization & threshold selection Laser Light SPE Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 22

Neutron Irradiation Tests 2 109 cm-2 y-1 Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 23

Neutron Irradiation Tests -- Before -- After ~3 years No long-term effect on MA-PMT or MAROC 3, null or negligible effects expected on specific components after ~ 20 years of CLAS 12 2 109 cm-2 y-1 Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 24

The Si. PM Test Prototype 1. 5 m coaxial cables to the electronics Commercial Si. PM matrix Beam Water-cooled Peltier cell for temperature control [-25 : +25 Celsius] Contalbrigo M. Custom Si. PM matrices with a pre-amplification stage RICH 2013, 2 nd December 2013, SVC - Japan 25

The Custom Si. PM Matrix @ +25 o Equalization of the single Si. PM is critical High threshold Medium threshold Low threshold 10 -3 level is challenging Signal occupancy (%) Vbias Peak RMS Background occupancy (%) Vbias Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan Vbias 26

The Custom Si. PM Matrix@-25 o For a 12 cm radius Cherenkov cone and a 3 mm Si. PM pixel, an occupancy of 4 % corresponds to about 24 p. e. In a +/- 3 ns window Comparable with H 8500 Largely insensitivity to Vbias and discriminator threshold High threshold Medium threshold Low threshold Signal occupancy (%) Vbias Peak RMS Background occupancy (%) Vbias Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan Vbias 27

RICH Prototype at CERN-T 9 RICH beam K, p Gas Cherenkov signal p GEM beam scintillators (trigger) Contalbrigo M. RICH threshold gas Cerenkov (p/K separation) RICH 2013, 2 nd December 2013, SVC - Japan pions few % kaons 28

RICH Prototype at CERN-T 9 GEM chamber layout RICH beam K, p Gas Cherenkov signal p GEM beam scintillators (trigger) Contalbrigo M. RICH threshold gas Cerenkov (p/K separation) RICH 2013, 2 nd December 2013, SVC - Japan pions few % kaons 29

RHIC Prototype: Direct Light Case Clear hadron separation up to the CLAS 12 maximum momentum 1 m gap H 8500 p p=8 Ge. V/c p=6 Ge. V/c 8 Ge. V beam K (x 40) Radiator beam P (x 40) Aerogel: n=1. 05 2 cm thickness q. C (mrad) P sq ns (Ge. V/c) (mrad) 6 306. 5 1. 41 6. 9 7 306. 8 1. 40 4. 7 8 307. 6 1. 40 3. 9 Ring Coverage ~ 80% Contalbrigo M. sc RICH 2013, 2 nd December 2013, SVC - Japan 30

RHIC Prototype: Direct Light Case Clear hadron separation up to the CLAS 12 maximum momentum 1 m gap H 8500 p p=8 Ge. V/c p=6 Ge. V/c 8 Ge. V beam K (x 40) Radiator beam P (x 40) Aerogel: n=1. 05 2 cm thickness q. C (mrad) Contalbrigo M. P RICH 2013, 2 nd December 2013, SVC - Japan s 1 pe (Ge. V/c) (mrad) 14 H 8500 C 4. 0 14 H 8500 C-03 5. 1 All H 8500 4. 6 31

RHIC Prototype: Reflected Light Case gap not optimized H 8500 Aerogel: n=1. 05 2 cm thickness mirrors + aerogel plane mirror 6 Ge. V beam curved mirror Radiator Aerogel: n=1. 05 6 cm thickness beam Ring Coverage ~ 60% Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 32

RHIC Prototype: Reflected Light Case gap not optimized H 8500 Aerogel: n=1. 05 2 cm thickness With absorbers: sizeable fraction of light survives <N>=13. 1 <N>=5. 3 mirrors + aerogel plane mirror 6 Ge. V beam curved mirror Radiator Aerogel: n=1. 05 6 cm thickness beam with absorbers Ring Coverage ~ 60% without absorbers Arbitrary units and resolution is not significantly degraded Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 33

RICH Simulations reflected light setup direct light setup without absorbers with absorbers Based on measured optical characteristics and validated with RICH prototype data Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 34

The CLAS 12 Hadron ID One charged particle per sector in average: Non trivial RICH light pattern due to reflections: patter recognition and likelihood ID required pion kaon proton Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 35

The CLAS 12 Hadron ID One charged particle per sector in average: pion >> kaon everywhere TOF + HTCC Non trivial RICH light patter due to reflections: patter recognition and likelihood ID required pion kaon proton + RICH few % pion contamination Even with a not yet optimized tuning of pattern recognition and likelihood ID, the p contamination is of the order of 1% Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 36

The CLAS 12 Hadron ID One charged particle per sector in average: pion >> kaon everywhere TOF + HTCC ~1: 500 rejection from x 5 to ~1% Non trivial RICH light patter due to reflections: patter recognition and likelihood ID required pion kaon proton + RICH few % pion contamination Even with a not yet optimized tuning of pattern recognition and likelihood ID, the p contamination is of the order of 1% Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 37

RICH Project Landscape RICH goal: 4 s p/K/p separation from 3 up to 8 Ge. V/c 2010: Concept of Design and Technology 2011: Tests of components and small prototype 2012: Extensive tests with large-scale prototype 2013: June: Technical Review August: TDR September: Project Review with DOE INSTITUTIONS INFN (Italy) Bari, Ferrara, Genova, L. Frascati, Roma/ISS Jefferson Lab (Newport News, USA) Argonne National Lab (Argonne, USA) Duquesne University (Pittsburgh, USA) Glasgow University (Glasgow, UK) J. Gutenberg Universitat Mainz (Mainz, Germany) Starting the construction phase Kyungpook National University, (Daegu, Korea) University of Connecticut (Storrs, USA) GOAL: 1 st sector ready by the end of 2016 UTFSM (Valparaiso, Chile) Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 38

The CLAS 12 RICH Project RICH goal: p/K/p identification from 3 up to 8 Ge. V/c and 25 degrees ~4 s pion-kaon separation for a pion rejection factor ~ 1: 500 ~ 1 ns time resolution to distinguish direct and reflected light patterns INSTITUTIONS INFN (Italy) Bari, Ferrara, Genova, L. Frascati, Roma/ISS Jefferson Lab (Newport News, USA) Argonne National Lab (Argonne, USA) Duquesne University (Pittsburgh, USA) Glasgow University (Glasgow, UK) J. Gutenberg Universitat Mainz (Mainz, Germany) Kyungpook National University, (Daegu, Korea) University of Connecticut (Storrs, USA) UTFSM (Valparaiso, Chile) Forward RICH: 2 sectors to accomplish physics program, 1 st sector by the end of 2016 Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 39

Mirror Technology Metalized Carbon Fiber substrate for spherical mirror Self-supporting structure with minimal material budget (applications in physics experiments) Thin glass skin on a flat support for planar mirrors Cost-effective technology for precise large area mirrors (applications in terrestrial telescopes) Standard technologies already in use and commercially available LHCB mirror Contalbrigo M. MAGIC-II telescope RICH 2013, 2 nd December 2013, SVC - Japan 40

Mirror Technology Metalized Carbon Fiber substrate for spherical mirror Self-supporting structure with minimal material budget (applications in physics experiments) Thin glass skin on a flat support for planar mirrors Cost-effective technology for precise large area mirrors (applications in terrestrial telescopes) Standard technologies already in use and commercially available LHCB mirror Contalbrigo M. MAGIC-II telescope RICH 2013, 2 nd December 2013, SVC - Japan 41

CFRP Spherical Mirror mm m Back View CFRP Frame 637 m m 507 Front View 1400 m 832 m m Radius tolerance <= 1% Surface accuracy: 5 µm RMS Surface quality: 3 nm RMS D 0 < 5 mm Reflectivity > 90% Contalbrigo M. CFRP mirror can be subdivided into more sub-mirrors (e. g. 8 instead of 4) RICH 2013, 2 nd December 2013, SVC - Japan 42

Planar Mirror & Aerogel Holder Six independent sandwich-glass mirrors. Tiles are constrained by a mesh of stretched nylon wires The stiffening frame acts as holder of the aerogel tiles CFRP Substrate Weight = 17 kg Aerogel Weight 35 kg Glass Mirror 6 kg Each Glass Mirror is bonded to three independent mounting jacks. Total Weight = 58 kg Contalbrigo M. RICH 2013, 2 nd December 2013, SVC - Japan 43