EM Digital calorimeter with MAPS Elena Rocco Nikhef

EM Digital calorimeter with MAPS Elena Rocco Nikhef – Utrecht University CALICE week @ CIEMAT 2014

Physics motivations At large Q 2 and small x the high gluon density should lead to a nonlinear evolution and saturation -> new state of QCD matter, crucial to understand initial state of high energy collisions Main goal is the measurement of gluon density at small x -> direct photons at forward rapidity p. QCD with shadowing Expected suppression of forward direct photons in p -Pb compared to pp relative to p. QCD expectations CGC* *Jalilian-Marian, Rezaeian PRD 86 (2012) 034016 CALICE week at CIEMAT 2014 erocco@nikhef. nl 2

Forward Calorimeter (Fo. Cal) Proposed detector as ALICE upgrade Needs to discriminate direct photons from the π0 decay photons p 0 → 2γ 24 layers silicon-tungsten calorimeter • 3. 5 mm of W absorber (≈1 X 0) • 2 types of Si layers : • Low granularity (1 cm 2 pad) and analog r/o • High granularity (30 micron pixel FULL DIGITAL PROTOTYPE AS summed in 1 mm 2 macro-pixels) PROOF OF CONCEPT and digital r/o CALICE week at CIEMAT 2014 erocco@nikhef. nl 3

Fully digital prototype as proof of concept Extra W block 20 mm Assembly w/o cooling y, direction W blocks, 3. 0 mm (<X 0) 24 layers 28 X 0 x, d i rec tion c e r di m a e z, B 6 mm 11 beam W absorber PS MAPS A M 100 micron air gap MIMOSA 23 CALICE week at CIEMAT 2014 W absorber PCB for chip operation 1 mm Half layer tungsten erocco@nikhef. nl 4 Sensor positioning within the plane (not to scale)

MAPS - MIMOSA 23* *A. Brogna et al. , Manual PHASE 1, IPHC Strasbourg Working principle Monolithic Active Pixel Sensor It is CMOS integrated circuit On the prototype: 96 sensors -> 39 M pixels 19. 2 mm 640 x 640 pixels 19. 2 mm MIMOSA 23 30 micron pitch CALICE week at CIEMAT 2014 • On-sensor digitisation: sensor level threshold setting, 1 bit per pixel • Sequential row readout (“rolling shutter”) • →pixel integration time: 642 µs (very slow, BUT a lot of R&D on going!) erocco@nikhef. nl 5

Raw data rate 61 Gb/s Readout system One half r/o system!! Data transfer to DAQ PC Demuxing data at the DAQ PC VIRTEX FPGA 2 GB BUFFER SPARTAN FPGA CHIPS SYNC ~39 Mpixel to be read every 640 µs → buffering required!!! CALICE week at CIEMAT 2014 erocco@nikhef. nl At that time larger MAPS application! 6

Beam tests performed and planned Facility Particle Beam Energies (Gev) DESY Electron beam 2, 5 2012 Settings Not optimised PS/SPS Mixed beam (e-, π ) 2, 3, 5, 6, 8/30, 50, 100, 150, 200, 250 2012 Settings Not optimised Utrecht lab Cosmic rays (μ) Large spectra 2013 -2014 Optimised settings DESY Electron beam 2, 3, 4, 5. 4 2014 Optimised settings PS/SPS Mixed beam (e-, π ) 2÷ 10/30, 50, 10 0, 150, 200+ Sept/Nov 2014 Optimised settings CALICE week at CIEMAT 2014 Year Comments erocco@nikhef. nl 7

ot n gs d n i t t Se timise op Linearity • reasonably good linearity o no obvious effect of saturation 2 mm • absolute calibration: o needs charge sharing in simulation o calibration with MIPs not consistent CALICE week at CIEMAT 2014 erocco@nikhef. nl 8

Energy resolution Observed resolution about 2 x worse than simulation (GEANT 3) WORK IN PROGRESS 4. 1 % 3. 4 % 2. 7 % Settings not optimised 8. 3 % Simulated data 6. 5% 5. 7% In the simulation the sensor response is assumed to be equal -> to be improved CALICE week at CIEMAT 2014 erocco@nikhef. nl 9

Energy Resolution Se t op tings tim no ise t d • resolution ≈2 x worse compared to simulation o different sensor sensitivities not fully corrected o relative calibration: how? • so far: single correction factor per sensor to minimize resolution CALICE week at CIEMAT 2014 erocco@nikhef. nl 10

Longitudinal profile Average over thousands of showers • Confirmation of the expected longitudinal shape , but instrumental issue (missing channels/dead sensors) 0 2 X 0 4 6 8 CALICE week at CIEMAT 2014 10 12 14 16 18 20 27 28 erocco@nikhef. nl 11

Moliere radius RM=11 mm r to tec e As ICHEP 2014 -VALENCIA d the rom n! f ted esig c e d xp e erocco@nikhef. nl - 05/07/14 12

Alignment 1 entry -> 1 sensor (96 entries in total) Not aligned Aligned Mean residuals (μm) Precision <30 micron (pixel pitch): important for the measurement of shower core position CALICE week at CIEMAT 2014 erocco@nikhef. nl 13

5. 4 Ge. V electron Lateral profile Average over thousands of showers Layer 0 is a tracking plane and it is used for the shower core determination (aligned data) Different sensitivity from sensor to sensor -> Calibration? CALICE week at CIEMAT 2014 erocco@nikhef. nl 14

Sensor Sensitivities • cumulative transverse distribution per sensor o clear sensitivity difference: calibration method? CALICE week at CIEMAT 2014 erocco@nikhef. nl 15

Shower Profiles • significant differences to GEANT 3 simulations o under study (GEANT 4? ) CALICE week at CIEMAT 2014 erocco@nikhef. nl 16

Application: PID Se t op tings ti Two events with same number of hits, taken with different mise not d Number of hits beams 50 Ge. V/c (mostly electrons) and 200 Ge. V/c (mostly pions) Mostly e- Mostly π How to discriminate different events with the same deposited energy? CALICE week at CIEMAT 2014 erocco@nikhef. nl 17

Application: PID Different shower evolution!! EM shower CALICE week at CIEMAT 2014 Se t op tings tim no ise t d Hadronic shower erocco@nikhef. nl 18

Proton therapy instrumentation (a non HEP application) CALICE week at CIEMAT 2014 erocco@nikhef. nl 19

Proton therapy instrumentation • The final goal is providing a detector that allows the precise dose verification on-line (180 -250 Me. V) and to be used in proton radiography (400 Me. V) Dose calculation Sketch for dose verification • Additional tracker planes to strengthen the particle reconstruction • Bragg peak measurement in December at KVI (Groningen. NL) CALICE week at CIEMAT 2014 erocco@nikhef. nl 20

Outlook & perspective • Investigation of new QCD matter at large Q 2 and small x -> unique measurement possible with ALICE (Fo. Cal Upgrade) o Stringent hardware requirements: high granularity and small RM; • Full digital (MAPS) prototype: o o o Good linearity Measured RM of 11 mm Particle identification (shower discrimination) Energy resolution not yet satisfactory Discrepancy between data and simulation (trying GEANT 4) Calibration procedure • Interesting/appealing technology for future projects like FCC CALICE week at CIEMAT 2014 erocco@nikhef. nl 21

Fo. Cal ALICE group University of São Paulo Czech Technical Institute, Prague University of Jyväskyla VECC, Kolkata Bhabha Atomic Research Center (BARC), Mumbai Jammu University Tsukuba University of Tokyo (CNS) Nikhef; Utrecht University Bergen University ORNL University of Tennessee Wayne State University Thank you for you attention! CALICE week at CIEMAT 2014 erocco@nikhef. nl 22

Any interest in closer collaboration? CALICE week at CIEMAT 2014 erocco@nikhef. nl 23