CMS SLHC Workshop The CMS ECAL Detector at

CMS SLHC Workshop The CMS ECAL Detector at SLHC D Cockerill RAL 26. 2. 2004 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 1

CMS ECAL at SLHC Contents SLHC Radiation environment EE, EB, Preshower Detector performance EE, EB Conclusions 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 2

SLHC – terms of reference CERN-TH/2002 -078 Physics Potential for LHC (107 s/year) 3 years at 1034 cm-2 s-1, 100 fb-1 /y, 300 fb-1 3 years at 1035 cm-2 s-1, 1000 fb-1 /y, 3000 fb-1 Total 3300 fb-1 ECAL TDR , 1997 Radiation levels for 10 years LHC to 5. 105 pb-1 = 500 fb-1 Maximum luminosity 1034 cm-2 s-1 SLHC Integrated dose/fluence Factor 6. 6 wrt ECAL TDR Dose and neutron rates Factor 10 wrt ECAL TDR, for 1035 cm-2 s-1 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 3

SLHC – upgrades LHC Luminosity and Energy upgrade LHC Project Report 626 Phase 0 Phase 1 No hardware upgrades 1 2. 3 3. 6 1034 cm-2 s-1 Hardware upgrades: insertion, injector 3. 3 4. 6 6 7 1034 cm-2 s-1 Phase 1 Superbunch, ib 1 A, bunch length 300 m to avoid electron cloud effects ~ 9. 1034 cm-2 s-1 Phase 2 Major hardware changes for 2020 Equip SPS with superconducting magnets New dipoles in LHC arcs E C of M 25 Te. V 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 4

SLHC – radiation load 1) SLHC design study calculations Assumes each fill to nominal luminosity Assumes turnaround time between fills of 1 h Caveats: Integrated luminosity drops by ~40% if LHC turnaround 6 h Fill to fill variations: <Luminosity> a factor ~0. 7 -0. 8 less Early beam aborts, factor 2? on integrated luminosity 2) ECAL TDR radiation calculations A safety factor of 2 -3 advised on simulation results A further factor of 2 -3 advised for cables and capacitors Radiation loads for tests, balance 1) with 2) ? ECAL TDR radiation levels, scaled to 3300 fb-1, used as the reference point in this talk 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 5

EE at SLHC Unshielded dose rate 0. 2 m. Sv/h =1. 48 Supercrystals and their internal components are inaccessible and cannot be replaced. Components: VPTs, HV pcbs, capacitors, resistors Signal & HV cable, quartz monitoring fibres =3 5 m. Sv/h Repair of SC array would require the dismounting of EE readout electronics on rear of backplate High activation levels, access time limited Qualify SC components for SLHC before EE build 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 6

EE Integrated Dose for 3300 fb-1 k. Gy 400 300 200 Inner radial limit of active electronics 100 EE radial distance from beam pipe (mm) Maximum Dose at = 3 350 k. Gy (35 MRad) SCs, VPTs, HV pcbs (capacitors, resistors), HV/LV cables, monitoring fibres Maximum Dose at = 2. 6 Active ECAL readout electronics 26. 2. 02 150 k. Gy (15 MRad) CMS SLHC workshop, D. J. A. Cockerill (RAL) 7

Neutrons/cm 2/1014 EE Integrated Neutron Fluence for 3300 fb-1 50 Inner radial limit for active electronics 40 30 20 Active electronics behind polyethylene moderator 10 EE radial distance from beam pipe (mm) Maximum fluence at = 3 5. 1015/cm 2 SCs, VPTs, HV pcbs (capacitors, resistors), HV/LV cables, monitoring fibres Maximum fluence at = 2. 6 5. 1014 /cm 2 Active ECAL readout electronics 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 8

Supercrystal items, Co 60 Irradiation tests All tests so far OK – no show stoppers, capacitors (unbiased) 9% change To do in 2004: VPTs, faceplates, capacitors and resistors to 500 k. Gy Brunel University source, 1 k. Gy/h, ~ 21 days 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 9

Supercrystal items, Neutron Irradiation tests All neutron irradiation tests so far OK – no show stoppers 1 capacitor, measured under irradiation, long cables, -17% To do: VPTs, faceplates, capacitors and resistors to 50. 1014 cm-2 Tests carried out at Minnesota, 252 Cf source, 2. 14 Me. V neutrons Neutron rate 107 cm-2 s-1 rate at = 3 at 1034 cm-2 s-1 Noise induced in VPT from local activation ~ 3200 e- 10000 e- at 1035 Compton electrons, from s s, enter VPT faceplate Light, from electrons above Cerenkov threshold, yield VPT photo-electrons 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 10

EE induced activation ECAL TDR Induced activation at = 3 ~0. 25 m. Sv/h <L> = 0. 5. 1033 cm-2 s-1, cooling time 1 day A further drop by ~0. 7 after some weeks Dose regulations/advice Dose limit 1 m. Sv/week Annual dose limit 5 m. Sv SLHC at 1035 cm-2 s-1 factor 20 on ECAL TDR Time to Annual dose = 3. 0 5 m. Sv/h 1 hour = 2. 6 2 m. Sv/h 2. 5 hours = 2. 0 0. 4 m. Sv/h 12 hours = 1. 48 0. 2 m. Sv/h 25 hours ↪ for dismounting EE from HE. Done at outer radius. Repairs on EE: need shielding, remote handling (if indeed repairs actually permitted!) 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 11

EE Readout for 3300 fb-1 Unshielded access time 25 hours Beam Active readout electronics PE moderator to reduce neutron fluence Set of 100 readout channels Inner radial limit r = 50 cm, = 2. 6 LV regulators to 5. 1014 /cm 2 1 hour Access constraints severe at inner radii Require robust LV regulators on EE from outset 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 12

EB at SLHC for 3300 fb-1 = 1. 48 at APDs Dose 5 k. Gy Neutrons 1. 3. 1014 cm-2 Dose 2 k. Gy Neutrons 7. 1013 cm-2 APD certification All screened to 5 k. Gy (some have received 10 k. Gy) – most OK (some have significant change in breakdown voltage - rejected most change by only ~1 V, vs. 40 V breakdown margin) Other tests 2001, Karlsruhe, 48 APDs, 20 k. Gy, 2. 1013 n/cm 2 Minnesota, >1000 APDs, 1 -2. 1013 n/cm 2 – all OK Need a programme of APD neutron tests to ~2. 1014 n/cm 2 and annealing tests at 18 o. C 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 13

Preshower at SLHC for 3300 fb-1 Preshower 1. 65 < | | < 2. 6 Silicon sensors at – 5 o. C Beam Neutrons from EE Protected by 4 cm of moderator. Further 4 cm, upstream, gives 8 cm of protection for Tracker EE Silicon at = 2. 6 Neutrons 1. 3. 1015 cm-2 Dose 700 k. Gy (70 MRad) Dismounting from inner cone Activation at = 2. 8 ~3 m. Sv/h 1. 7 hours for annual dose (EE dominated? ) Need simulation for isolated Preshower, to determine repair accessibility. 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 14

Preshower at SLHC for 3300 fb-1 Silicon sensors to = 2. 6 1. 3. 1015 n/cm 2, 700 k. Gy (70 MRad) Increased leakage current Increased voltage required to full depletion, <500 V for TDR levels Leakage current compensation tested to 6 x. TDR ( ~SLHC) If depletion voltages of 1000 V needed, likely that even best sensors will break down Will be at limit of HV supply components Complete replacement of inner sensors on a fairly regular basis Electronics Expect big trouble with ST LV regulators 0. 25 m chips (front end, ADC, control system etc) “should” survive but no guarantees or tests to SLHC levels PACE 0. 25 m chip – not tested under irradiation yet (PACE DMILL was tested to 6 x 1014 n/cm 2, 100 k. Gy, and was ok) 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 15

ECAL Crystal Performance Crystal LY loss from Co 60 dose rate studies At SLHC, =3, at shower max Dose rate = 10 x 15 = 150 Gy/h Data rate, Cantonal Irradiation 240 Gy/h, 2 h Representative of SLHC worst case Densely ionising hadron shower effects not included LY loss calculated from measured induced absorption LY loss distribution for 677 xtals Assume all colour centres activated – gives worst case 20 30 40 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) % LY loss 16

ECAL LY during LHC fills - SLHC =0 Crystal light yield =2. 5 LHC luminosity fill by fill Colour centre creation dependent on dose rate Dose rate changes during fill and with eta More changes in EB! EE saturates to constant level 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 17

Crystal light yield at LHC Light Yield % Startup 1033 Low 2. 1033 High 1034 SLHC 1035 100 80 60 40 0 < < 3. 0 At SLHC, see significant changes in crystal LY <LY> drops by ~25% EB, 30% EE. 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 18

Crystal LY changes at SLHC 10% RMS LY changes during fills 5% 0% 0 1. 5 Eta 3. 0 Barrel LY changes ~3% through the period of a fill Endcap LY changes ~1% (crystals saturated) LY monitoring – main challenge in EB 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 19

EE performance at SLHC Initial performance 50 Me. V ET, preamp noise 3500 e. Activation noise, SLHC = 2. 5, 10000 e- 140 Me. V ET per channel Losses Xtal LY loss 0. 7 0. 2 Induced abs data VPT faceplate 0. 8 ? Guess, 10% to 20 k. Gy VPT Q. E. (burn-in study) 0. 4 ? 60% loss, 6 days at Ik = 1 A 18 y at 1034 at = 2. 5 VPT gain 1. 0 No change observed Reduced HV 0. 9 Working margin Resultant factor 0. 2 (Hadron damage to xtals, another factor 0. 5? ) Resultant noise 250 (700 with activation) Me. V ET per channel - excluding pileup contributions & other electronics issues Charged hadron effects on xtal LY need to be taken into account 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 20

EB Performance at SLHC Leakage Noise equiv Current/xtal Comment APD current (TDR) 20 A 60 Me. V With annealing, single sampling? APD current (SLHC) 130 A 150 Me. V As (leakage current) Annealing not included 140 Me. V 50 Me. V in quadrature 190 Me. V LY loss in crystals Add EB preamp noise Losses Crystal factor 0. 75 APD - Xtal glue ? Measured to 5 k. Gy? APD Q. E. , Gain ? Reduce gain, leakage? EB noise likely to be ~190 Me. V per channel - excluding pileup contributions & other electronics issues Charged hadron effects on xtal LY need to be taken into account 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 21

ECAL at SLHC - Conclusions EE Repairs very difficult if not impossible, activation Qualify all components to SLHC levels before EE build VPT and component irradiation tests in 2004 to 350 k. Gy Induced activity noise could be important limitation Charged hadron effects on Xtal LY, tests to be completed Detector Noise/channel ET 250 Me. V or greater (excl. pileup) EB APD studies to ~2. 1014 n/cm 2 needed Detector Noise/channel 190 Me. V or greater (excl. pileup) Preshower Replacement of inner silicon likely to be needed – very difficult 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 22

Backup slides 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 23

Simulation of crystal behaviour at LHC Simulation of crystal LY loss Colour centre creation and recovery ØLHC luminosity according to beam lifetime during fill ØFill of 20 h, turnaround 4 h (old regime) ØRelative fill to fill variations, 0. 2 1. 0 ØDose rate calculated at 1 cm steps along each xtal ØColour centres and LY loss calculated for each cm along xtal ØCrystal data from GIF for creation and annealing time constants ØLY loss along full crystal iterated in 1 h intervals ØLY losses calculated for 0< <3. 0 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 24

SLHC – running time Fill lifetime 15 h, L 0=1034 6. 5 h, L 0=4. 5. 1034 Turnaround (h) 1 6 1 T run (h) 5 12 3 Lint fb-1/y 122 78 524 6. 5 h, L 0=4. 5. 1034 6 6 286 ~45% less Lint if turnaround is 6 h and not 1 h 26. 2. 02 CMS SLHC workshop, D. J. A. Cockerill (RAL) 25