Study of Study of dama damages ges induce induced d on on A - - PowerPoint PPT Presentation

PIXEL 2018, Taipei, 10-14 December 2018

Study of Study of dama damages ges induce induced d

• n
• n A

ATLAS sili TLAS silico con by f n by fast ast extr xtract acted ed and and inten intense se pr prot

• ton
• n be

beam am ir irrad adia iation tion

• C. Bertella, C. Escobar, J. Fernández, C. Fleta, A. Gaudiello, G. Gariano,
• C. Gemme, S. Katunin, A. Lapertosa, M. Miñano Moya, A. Rovani,
• E. Ruscino, A. Sbrizzi, M. Ullán

Introduction

2

• A. Lapertosa

10-14 December 2018

➢ The ATLAS silicon tracker detectors (Pixels, Strips) are designed to sustain high dose integrated over several years of operations at the LHC ➢ The radiation hardness should also favor the survival in case of accidental beam losses ➢ ATLAS Pixel modules are able to survive to LHC beam loss scenarios [NIM A565 (2006) 50-54] ➢ LHC will be upgraded to higher luminosity (HL-LHC) ➢ To cope with HL-LHC conditions, ATLAS is planning a complete update of the tracker: ➢ The new inner tracker (ITk) will replace the current tracking system ➢ ITk will be an all silicon tracker system (ITk Strip TDR, ITk Pixel TDR) ➢ The damage threshold needs to be measured for the new ITk Pixel and Strip modules ➢ The new damage threshold needs to be compared with beam-loss scenarios in HL-LHC ➢ The comparison will be useful as feedback for the choice of the HL-LHC collimator material

➢ The goal of this study:

➢ Study the effects of accidental beam-loss scenarios for ATLAS Pixels and Strips at HL-LHC: ➢ Provide a realistic estimate of the damage threshold for sensors and electronics ➢ Evaluate the performance degradation due to the radiation damage

HL-LHC conditions

3

• A. Lapertosa

10-14 December 2018

➢ HL-LHC beam loss scenarios: ➢ Asynchronous beam dump ➢ Wrong injection settings ➢ Beam injection and shielding at HL-LHC: ➢ New triplets injection magnets ➢ Larger TAS aperture (from 17 to 30 mm radius)

TAS TAS is a 1.9 m long copper cylinder inside the ATLAS shielding system R = [17 mm, 250 mm], Z = 19.04 m

ITk Pixel TDR

After 4000 fb-1 Pixel layers: up to 1016 1 MeV neq / cm2 fluence Strip layers: up to 1015 1 MeV neq / cm2 fluence ITk layout example (not final layout)

HiRadMat facility

4

• A. Lapertosa

10-14 December 2018

➢ Test beams performed at the HiRadMat facility: ➢ Facility at CERN providing high-intensity pulsed beam ➢ 440 GeV proton beam extracted from CERN SPS ➢ Two separated tunnels: ➢ TNC: beam line with experimental tables ➢ TT61: powering/read-out system ➢ Experimental setup: ➢ Long cables passing through a concrete wall ➢ Cables connect powering and read-out to devices ➢ Operation on the modules must be remotely controlled Beam parameters

TNC TT61

Powering and read-out

2017 and 2018 test beams

5

• A. Lapertosa

10-14 December 2018

➢ Test beams performed at the HiRadMat facility: ➢ 440 GeV proton beam perpendicular to silicon modules ➢ Two irradiation schemes: Global (2 mm radius) and Local (0.5 mm) ➢ Increasing the number of bunches step-by-step ➢ From 1 to 288 bunches ➢ 25 ns spacing between bunches ➢ Up to 1 · 1011 protons per bunch ➢ Experimental setup aligned with the beam ➢ Modules mounted on frames ➢ Frames loaded in experimental box ➢ Box placed on aluminum table in the tunnel Year Beam radius Goal Bunches 2017 2 mm Global irradiation (1, 4, 12, 24, 36, 72, 144, 288) x 5·1010p, 288 x 1011p 2017 0.5 mm Local irradiation (1, 12, 72, 288) x 1011p 2018 2 mm Global irradiation (1, 12, 24, 36, 72, 144, 288, 288, 288) x 1011p 2018 0.5 mm Local irradiation (1, 12) x 1011p

Experimental box (2017 and 2018)

6

• A. Lapertosa

10-14 December 2018

➢ Design and construction of the test-box: ➢ Material: epoxy fiber glass, makrolon and aluminum ➢ Designed to host a maximum of 8 detector modules mounted on dedicated frames

July 2017 test beam

7

• A. Lapertosa

10-14 December 2018

➢ 2017: ➢ 2 IBL Pixel modules (3D sensor + FE-I4) ➢ 1 ITk Strip DAQload (ATLAS12 + ABC130) ➢ Issues with operations: ➢ Not able to fully read-out Strip sensor ➢ Cooling fans not properly working ➢ Temperature drift: from 30 to 60 °C

Module IBL Pixel ITk Strip

Module size 2x2 cm2 1x1 cm2 Chip FE-I4 ABC130 Sensor type 3D ATLAS12 Sensor thickness 230 μm 320 μm Channels (pitch) 26680 (50x250 μm) 104 (74.5 μm) Max design dose 250 Mrad 35 Mrad

Experimental box: 2018 improvements

8

• A. Lapertosa

10-14 December 2018

➢ Design and construction of the test-box: ➢ Material: epoxy fiber glass, makrolon and aluminum ➢ Designed to host a maximum of 8 detector modules mounted on dedicated frames ➢ Improvements in 2018 setup: ➢ Darkening of the box (aluminum tape) ➢ Cooling system working: 4 fans (12x12 cm) with filters ➢ Air flux controlled remotely ➢ Motor system to move box in/out of the beam

May 2018 test beam

9

• A. Lapertosa

10-14 December 2018

➢ 2018: ➢ 1 IBL Pixel module (Planar sensor + FE-I4) ➢ 2 ITk Strip DAQloads (barrel mini petalet / LowR + ABC130) ➢ Improvements on modules: ➢ Fully read-out of one ITk strip module (LowR) ➢ Heat dissipation in Pixel frame (heat sinks) ➢ Stable temperature at 36° C

Module IBL Pixel ITk Strip

Module size 2x4 cm2 0.7x2.6 cm2 Chip FE-I4 ABC130 Sensor type Planar LowR Sensor thickness 200 μm 310 μm Channels (pitch) 2x26680 (50x250 μm) 64 (77 μm) Max design dose 250 Mrad 35 Mrad

2017 setup with 3D Single chip IBL Pixel Module

10

• A. Lapertosa

10-14 December 2018

Pixel measurements (2017)

11

• A. Lapertosa

10-14 December 2018

Two different configurations during the beam transit

Module 2 - IBL 3D Module 1 - IBL 3D Stand-by: Sensor bias Off FrontEnd pre-amplifiers Off Stable beam: Sensor bias On FrontEnd pre-amplifiers On

Noise 2D map extracted from threshold scan (HV On: -20 V) Leakage current increase after irradiation (normalized to 273 K) Reference: before test After 2 mm beam (1-288 b) After 1 week of parasitic irradiation (beam to RotColl)

12

• A. Lapertosa

10-14 December 2018

➢ Global irradiation: 2 mm radius beam ➢ Both modules survived to 288 bunches ➢ Self-triggering scan after beam transit ➢ Beam spot visible due to material activation ➢ Intensity of the spot increasing after each shot ➢ Intensity of the spot decreasing with time ➢ Low residual activity after 1 week ➢ Local irradiation: 0.5 mm radius beam ➢ Readout chip LV in short after 288 bunches ➢ Small bump visible on flexible PCB

After glow (288 b, 2 mm)

4.3 mm 2.8 mm Small bump on flexible PCB

Low residual activity after one week

Pixel measurements (2017)

17 July 2018 25 July 2018 After glow (72 b, 0.5 mm)

2018 setup: IBL DC + metallic contact & heat sinks

13

• A. Lapertosa

10-14 December 2018

Default 1 b 4 b 12 b 24 b 36 b 72 b 144 b 288 b

Pixel measurements (2018)

14

• A. Lapertosa

10-14 December 2018

Bias: -80 V Temperature: 36 °C ➢ Module tested in Stable beam configuration ➢ Monitoring of leakage current after each shot ➢ Each bunch is made of 1011p ➢ Before irradiation: 2 μA at -80 V ➢ Increase after irradiation: 170 μA at -80 V ➢ Bulk and surface damage post-irradiation ➢ Linear increase of the leakage current ➢ agree with expectations (Hamburg model)

Linear fit Linear fit . . .

Pixel measurements (2018)

15

• A. Lapertosa

10-14 December 2018

➢ Correlation between performance degradation and proton fluence ➢ Noise increase per pixel used as figure of merit ➢ Proton beam simulated taking into account beam position and bunches ➢ Estimated fluence at beam center: 14·109 p ➢ Noise starts to increase after 4 bunches ➢ Constantly increases after each pulse Noise from threshold scan: after 288 bunches Noise vs Flux correlation 120 e- 250 e- Beam simulation: after 288 bunches

2018 setup: ITk Strip (LowR + ABC130)

16

• A. Lapertosa

10-14 December 2018

Strips measurements (2017/2018)

17

• A. Lapertosa

10-14 December 2018

➢ Leakage current continuosly recorded ➢ Peaks right after the beam transit ➢ About 10 minutes relax time ➢ Constant increase with radiation ➢ Increment is approximately linear

Bias: -150 V

2017 2018

Bias: -6 V

Preliminary

Strip measurements (2018)

18

• A. Lapertosa

10-14 December 2018

➢ 64 channels ➢ Gradual loss of channels ➢ 50% channels lost after 36 bunches Beam 2 mm radius ➢ Noise before and after shots ➢ Many noisy/damaged channels in region interested by the beam ➢ Mainly around the beam center 36 bunches

Before test 4 bunches 24 bunches 288 bunches

Strip measurements (2018)

19

• A. Lapertosa

10-14 December 2018

Fraction of noisy/good/damaged strips close by or far from the beam

IBL Pixel results @ HiRadMat

20

• A. Lapertosa

10-14 December 2018

➢ Three IBL modules were tested with different configurations with SPS beam @ HiRadMat ➢ Two configurations used to reproduce ATLAS operations: “stand-by” and “stable beam” ➢ Two different IBL structure tested: 3D and Planar ➢ Noise increases around the beam spot in a similar way for the three modules ➢ In 2017, we proved that IBL 3D modules are able to sustain 288 bunches with 2 mm radius ➢ In 2018, we proved the same with IBL Planar module ➢ This time, we shoot 3 times 288 bunches → the module survived ➢ We are confident that IBL modules are able to sustain such an intense energy release ➢ We set a conservative damage threshold limit on IBL modules: 1013 p/cm2 ➢ Estimated taking into account 288 bunches made by 1011 protons on 2x2 cm2 surface

ITk strip results @ HiRadMat

21

• A. Lapertosa

10-14 December 2018

➢ Three different ITk strip modules were tested with SPS beam @ HiRadMat ➢ 2018 setup with full read-out (LowR + ABC130) ➢ Increase in leakage current follows the increase of beam intensity ➢ Noise level increases as a function of the proton flux ➢ Fraction of fully working channels decrease with the increase of the dose received ➢ Large effect observed around the beam center ➢ Sensors presents an acceptable damage when exposed to 1011 p with 2 mm beam radius ➢ Limit on the damage threshold: 8 · 1011 p/cm2 ➢ Not macroscopic or physics damages on sensor and chip are visible ➢ On-going investigation of the sensor 2018

Thanks for your attention!

22

• A. Lapertosa

10-14 December 2018

➢Acknowledgments to: ➢ARIES ➢ATLAS ITk ➢HiRadMat team ➢CERN FLUKA team

→ BACKUP 

Beam profile of 1011 proton bunch with respect to the module Average flux over the module: 1.5 · 1010 p/cm2 24

• A. Lapertosa

10-14 December 2018

➢ ATLAS Pixel module: Planar sensor + FE-I3 chip ➢ Designed to sustain 50 MRad ➢ Protons from CERN Proton Synchrotron: 24 GeV ➢ Gaussian beam, parallel to the module ➢ 1-8 bunches made by 1011 protons ➢ Bunch spacing: 256 ns ➢ Operational temperature: 30°C ➢ Limit: 1.5 · 1010 protons/cm2 ➢ Estimated as the average flux 1.5 · 1010 p/cm2 (213 bunches in total, 4 MeV/cm) ➢ Compatible with LHC scenarios

2006 experiment ATLAS Pixel vs LHC beam loss

3D single chip (2017) & Planar double chip (2018)

25

• A. Lapertosa

10-14 December 2018

➢ In July 2017 testbeam, a 3D single chip IBL module was tested ➢ In May 2018 testbeam, a Planar double chip IBL module was used ➢ Double chip: ➢ 1 chip centered on the beam line ➢ 1 chip lateral, as metallic contact ➢ Metallic contact (Aluminum) connected to heat dissipator ➢ Aluminum heat dissipator in the corner of the frame ➢ Stable temperature at 36 °C

3D single chip Planar double chip

2017 and 2018 IBL modules survived 288 bunches

26

• A. Lapertosa

10-14 December 2018

▪ Noise (from threshold scans with HV On) ▪ Similar results for 3D & Planar sensor ▪ Both survived 288 bunches of 2 mm radius ▪ Visible also effect of few bunches of 0.5 mm radius

IV scans (Operating @ 80V)

27

• A. Lapertosa

10-14 December 2018

Values used for measurement of Leakage Current vs Flux Default 1 b 4 b 12 b 24 b 36 b 72 b 144 b 288 b Each bunch is made of 1011 p

Experimental tunnel & parallel tunnel

28

• A. Lapertosa

10-14 December 2018

PC & Readout Setup in the parallel tunnel Table with box in the experimental tunnel 15m long cables

Bad alignment and 1st shot

29

• A. Lapertosa

10-14 December 2018

Self-triggering Source scan after 1st bunch Bad initial alignment…

Moving the beam in position

30

• A. Lapertosa

10-14 December 2018

Self-triggering Source scan After moving the beam 5 mm right Beam moved 5 mm right

Threshold scan HV -80 V (before)

31

• A. Lapertosa

10-14 December 2018

Mean: Tuned at 3000 e Sigma: Noise peak at 118

Threshold scan HV -80 V (after 1-4-12-24-36-72-144-288 bunches)

32

• A. Lapertosa

10-14 December 2018

Large tail up to 6000 e Noise tail up to 400

Beam configuration at HL-LHC

33

• A. Lapertosa

10-14 December 2018

Beam configuration at HL-LHC

34

• A. Lapertosa

10-14 December 2018

➢ The asynchronous beam dump: ➢ The extraction kicker field switch-on is not synchronised with the abort gap ➢ Unlikely off-orbit protons hit directly the experiments ➢ Possible scenario: ➢ Protons hit the TCT4 collimators (120 m away from the IP) ➢ Shower into the experiments

Experimental box in the tunnel

35

• A. Lapertosa

10-14 December 2018

Fixed 90° impact

ATLAS test-box

Experimental box in the tunnel (2017)

36

• A. Lapertosa

10-14 December 2018

ATLAS test-box RotColl experiment

Modules within the box

37

• A. Lapertosa

10-14 December 2018

ITk Strip: Global and Local irradiation (2017)

38

• A. Lapertosa

10-14 December 2018

Pixel and Strip noise measurement

39

• A. Lapertosa

10-14 December 2018

Correlation: Dose (prot. per pixel) vs Noise (electr.)

40

• A. Lapertosa

10-14 December 2018

Simulated beam: 0 bunch Noise scan (HV 80V) before test Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 0 b Threshold scan after the shot Beam profile (simulated) 2D gaussian X/Y Position from source scans X/Y Width from e-log data

• Num. protons from bunches x 1011

Profile plot: (averaged on X) average of protons received vs noise measured Scatter plot: 1 point per pixel X protons received per pixel Y noise measured on threshold scan Note: X axis scale changes for every plot Note: X axis scale fixed

Correlation: Dose (prot. per pixel) vs Noise (electr.)

41

• A. Lapertosa

10-14 December 2018

Simulated beam: 1 bunch 1 b Noise scan (HV 80V) after 1 bunch Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured

Correlation: Dose (prot. per pixel) vs Noise (electr.)

42

• A. Lapertosa

10-14 December 2018

Noise scan (HV 80V) after 1+1 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured Simulated beam: 1+1 bunches 1 b 1 b

Correlation: Dose (prot. per pixel) vs Noise (electr.)

43

• A. Lapertosa

10-14 December 2018

Simulated beam: 4 bunches Noise scan (HV 80V) after 4 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b

Correlation: Dose (prot. per pixel) vs Noise (electr.)

44

• A. Lapertosa

10-14 December 2018

Simulated beam: 12 bunches Noise scan (HV 80V) after 12 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b 12 b

Correlation: Dose (prot. per pixel) vs Noise (electr.)

45

• A. Lapertosa

10-14 December 2018

Simulated beam: 24 bunches Noise scan (HV 80V) after 24 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b 12 b 24 b

Correlation: Dose (prot. per pixel) vs Noise (electr.)

46

• A. Lapertosa

10-14 December 2018

Simulated beam: 36 bunches Noise scan (HV 80V) after 37 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b 12 b 24 b 36 b

Correlation: Dose (prot. per pixel) vs Noise (electr.)

47

• A. Lapertosa

10-14 December 2018

Simulated beam: 72 bunches Noise scan (HV 80V) after 72 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b 12 b 24 b 36 b 72 b

Correlation: Dose (prot. per pixel) vs Noise (electr.)

48

• A. Lapertosa

10-14 December 2018

Simulated beam: 144 bunches Noise scan (HV 80V) after 144 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b 12 b 24 b 36 b 72 b 144 b

Correlation: Dose (prot. per pixel) vs Noise (electr.)

49

• A. Lapertosa

10-14 December 2018

Simulated beam: 288 bunches Noise scan (HV 80V) after 288 bunches Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b 12 b 24 b 36 b 72 b 144 b 288 b

Final Plot: Each scan superimposed

50

• A. Lapertosa

10-14 December 2018

Scatter plot: protons received per pixel, noise measured on pixel Profile plot: average of protons received vs noise measured 1 b 1 b 4 b 12 b 24 b 36 b 72 b 144 b 288 b