CLIC Vertex Detector R&D Sophie Redford on behalf of the CLIC - - PowerPoint PPT Presentation

CLIC Vertex Detector R&D

Sophie Redford

• n behalf of the CLIC Detector and Physics Collaboration

CLIC - a collider for the future

2

156 ns 312 bunches per train

• Linear electron-positron collider
• √s = 3 TeV (staged construction)
• High luminosity: 1034 cm-2s-1
• Small bunch size: σxyz(40 nm, 1 nm, 44 µm)
• Beam structure:

Detector environment

3

• Beamstrahlung creates high particle rate

‘beam induced backgrounds’

• most at low angle, low pT, constrained

by B field

• Inner radius of vertex detector restricted

by particle density

Maximum occupancy including safety factor 5: 1.9% per pixel in the barrel layers 2.9% per pixel in the forward layers

The CLIC detector

4

Detector challenges at CLIC,

• S. Lukic, Monday 17:00

Precision physics in a challenging environment: broad programme of R&D Highly granular particle flow calorimetry, using tungsten absorber 5.5 m diameter cryostat for superconducting solenoid, B field 4-5 T Instrumented steel return yoke Complex forward region

Vertex detector requirements

5

G

• a

l

Efficient tagging of heavy quarks through a precise determination of displaced vertices

• Single point resolution of 3 µm
• Material budget of < 0.2% of a

radiation length per layer

• No active cooling elements -

use forced air flow cooling

• Limit the power dissipation to

50 mW/cm2 in sensor area

• Hit time slicing of 10 ns

Multi-layer barrel and endcap pixel detectors

• 560 mm in length
• Barrel radius from 30 mm to 60 mm

Olympic programme of R&D

6

Thin assemblies Powering Cooling Readout Supports

7

Geometry optimisation studies

Single-sided layers Double-sided layers

• Similar flavour tag performance for two considered layouts
• Increasing the material has a larger impact than the layout

] ° [ !

20 40 60 80

Average number of layers

2 4 6 8

CDR spirals double_spirals

0.5 0.6 0.7 0.8 0.9 1

Misidentification eff.

• 3

10

• 2

10

• 1

10 1

Charm Background double_spirals spirals LF Background double_spirals spirals Dijets at 200 GeV

Beauty eff.

0.5 0.6 0.7 0.8 0.9 1 double_spirals/spirals 0.8 0.9 1 1.1

Charm Background LF Background

0.5 0.6 0.7 0.8 0.9 1

Misidentification eff.

• 3

10

• 2

10

• 1

10 1

Charm Background double_spirals_v2 double_spirals LF Background double_spirals_v2 double_spirals Dijets at 200 GeV

Beauty eff.

0.5 0.6 0.7 0.8 0.9 1

double_spirals_v2/double_spirals

1 1.2 1.4

Charm Background LF Background

Comparison of 5 single-sided layers and 3 double-sided layers

First Medipix3 Image using TSV

Thin sensor assemblies

8

• Hybrid planar pixel technology
• Ultimate goal: 50 µm sensor on 50 µm ASIC
• 25 µm pitch
• Thin edge sensors using Through-Silicon-Vias

50 µm thick silicon wafer

TSVs:

• Vertical electrical connection - no wire bonds
• Sensors buttable on all sides - better tiling
• 60 µm hole diameter
• Wafer thinned to 120 µm
• 5 µm copper layer for TSV

Testbeam analysis

9

• Thin sensors (50 - 300 µm) bonded to normal Timepix chips
• One 100-on-100 µm assembly
• Data recorded at DESY: 5.6 GeV electron beam

50 µm thick sensor Efficiency 99.2% at

• perating threshold

100 µm thick sensor - low charge sharing Track position: cluster size 2 Track position: cluster size 4 50 µm sensor 750 µm Timepix

Threshold 380 385 390 395 400 405 410 415 Efficiency 0.955 0.96 0.965 0.97 0.975 0.98 0.985 0.99 0.995

C04-W0110 Oct13

Preliminary

Constant 5.90e+01 ± 1.28e+04 Mean 0.0000182 ±

• 0.0002745

Sigma 0.000016 ± 0.005334 Y residual (mm)

• 0.1
• 0.05

0.05 0.1 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 Constant 5.90e+01 ± 1.28e+04 Mean 0.0000182 ±

• 0.0002745

Sigma 0.000016 ± 0.005334

2 hit clusters no calibration DigitalCentroid

Preliminary Constant 6.570e+01 ± 1.412e+04 Mean 0.000016 ±

• 0.000255

Sigma 0.000015 ± 0.004809 Y residual (mm)

• 0.1
• 0.05

0.05 0.1 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 Constant 6.570e+01 ± 1.412e+04 Mean 0.000016 ±

• 0.000255

Sigma 0.000015 ± 0.004809

2 hit clusters no calibration EtaCorrection

Preliminary

Two-hit cluster resolutions

10

100 µm thick sensor Digital centroid Charge weighting (eta corr.) Resolutions include tracking resolution

• Tracking resolution of telescope ~3.5 µm
• Digital centroid method effective due to low charge sharing

Constant 29.9 ± 3897 Sigma 0.000026 ± 0.005214 Track y position (mm) 0.01 0.02 0.03 0.04 0.05 0.06 Number of hits 1000 2000 3000 4000 5000 6000 Constant 29.9 ± 3897 Sigma 0.000026 ± 0.005214

Projection of (1,2)

Project

• Eta correction method of charge weighting best

resolution: take into account non-linearities in charge sharing

• Unfolding the tracking resolution gives a single

point hit resolution of 3.3 µm for 2 hit clusters

Energy (keV) 20 40 60 80 100 120 140 0.01 0.02 0.03 0.04 0.05 0.06 0.07 Cluster size 1 Cluster size 2 Cluster size 3 Cluster size 4

Global calibration

241Am 60 keV 241Am 26.2 keV 55Fe: 5.8 keV 09Cd: 22.9 keV

Indium 24 keV Brass 8.1 keV Sources and X-ray fluorescence

Sensor calibration

11 Resolutions include tracking resolution

a b c t

y = ax + b − c x − t

• Calibrate TOT values by

measuring response to photons of known energy

• Accounts for non-linearities
• Calibration aligns Landau’s

and improves the resolution: 4.8 µm → 4.7 µm

TOT value 1000 2000 3000 4000 5000 0.02 0.04 0.06 0.08 0.1 0.12 0.14 Cluster size 1 Cluster size 2 Cluster size 3 Cluster size 4

No calibration

Calibration

Constant 6.718e+01 ± 1.441e+04 Mean 0.0000162 ±

• 0.0002513

Sigma 0.00001 ± 0.00471 Y residual (mm)

• 0.1
• 0.05

0.05 0.1 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 22000 Constant 6.718e+01 ± 1.441e+04 Mean 0.0000162 ±

• 0.0002513

Sigma 0.00001 ± 0.00471

2 hit clusters global calibration EtaCorrection

Preliminary

Readout ASIC: CLICpix

12 1.85 mm 3 mm 64 x 64 25 µm pixels

Chip Board FPGA Board

• The CLICpix ASIC: a fast, low power readout chip with 25 µm pitch
• Implemented in 65 nm CMOS technology
• 4-bit time and energy measurements for each pixel
• Supports power-pulsing and data compression

CLICpix characterisation

13 Pixels 1-64 Pixels 1-64 0.35 0.3 0.25

• Time Over Threshold gain distribution
• Uniform gain across the whole matrix
• Gain variation is 4.2% r.m.s. (for

nominal feedback current)

• Matrix equilisation
• Calibrated spread is 0.89 mV (about

22 e-) across the whole matrix

• (Expect a signal of ~thousands of

electrons in 50 µm sensor)

TOT gain distribution

Power-pulsing strategy

14

• Power pulse CLICpix ASIC to achieve dissipation <50 mW/cm2 in the sensor area
• Analog electronics can be turned off: 2 W/cm2 → 2 mW/cm2
• Digital electronics in idle except during readout: 100 mW/cm2 → 13 mW/cm2

20µs ON OFF ON 20µs

Analog Chip [1:12] Train Bunch

ON ON

2 W/cm2

ON

100 mW/cm2

OFF

Turned OFF

ON ON Idle Read Out Idle ON Idle

8 mW/cm2

ON Idle Read Out Idle 20/12 ms Read Out 360 mW/cm2 20/12 ms

Digital Chip [1] Digital Chip [2] Digital Chip [12]

ON Read Out Idle 20/12 ms Idle

Power delivery

15

1 cm 1 cm 24 cm

CLICpix Power each half a ladder Ladders Vertex barrel P

• w

e r i n P

• w

e r i n

• Power ladders from each end of the barrel:
• constant current sources, low dropout regulators, silicon capacitors

Material budget:

• Aluminium flex cables and silicon capacitors reduce material
• Powering adds 0.1% X0 per layer. Projected: < 0.05% X0

Power-pulsing lab tests

16

Back-end cables Al Flex Cable Dummy load Controlled current source

Iload for 1 ASIC Vload VCap

Particular case ton = 20μs 2 A per chip

5.3 V 1.2 V ∆V = 16 mV 1.4 V

Analogue results Analogue:

• Voltage drop < 20 mV
• Measured average power

dissipation < 10 mW/cm2 Digital:

• Measured average power

dissipation < 35 mW/cm2 Total dissipation: < 50 mW/cm2

17

Air-flow cooling

• Total heat load after power-pulsing ~470 W
• Cooling provided by forced air-flow:
• Dry air cooling at 0°C
• Low material: radiation length of air ~310m

Cool air Warm air

Air-flow simulations

18

• Mass flow: 19.9 g/s
• Avg. velocity in barrel: 6.3 m/s
• Silicon temperature below 40oC
• Conduction not taken into account

Mechanical support structures

19

• Develop and characterise low-mass carbon-fibre structures
• Stave dimension 1.8 mm*26 mm*280 mm
• Goal material per layer: 0.05% X0

Skin stave Full sandwich stave Cross braced staves

Honeycomb core (Nomex and Carbone) Rohacell core (PMMA) Mass 3.74 g 3.45 g 3.08 g 2.74 g 1.76 g X/X0 0.121% 0.112% 0.118% 0.068% 0.051%

Stave mechanical characterisation

20

Next: measure the amplitude of stave vibration in air flow

• Measure the flexural stiffness (resistance to bending) of the staves
• ver span of 180 mm

Measurements 6.95 N/mm 2.24 N/mm 3.3 N/mm 2.96 N/mm 2.23 N/mm Simulations 6.95 N/mm 2.35 N/mm

• 2.30 N/mm

Thermo-mechanical test bench

21 Air flow

• Measure wind speeds, stave temperatures, stave vibrations
• Allows validation of simulations

90 deg. 0 deg.

Thermal camera Simulation

Test bench results

7 3 67 63 27 23 87 2 8 9 3 : ; <

!"#$"%&'(%")*+,%"&-")./01) 23%)4"5/,3'6).#7-1) 0/+-'&+')$/8"%)9):;)#<7,#

=7&>%,&?&@6 =7&>%,&?&@2 =7&>%,&?&@8 =7&>%,&?&@9 93&>%,&?&@6 93&>%,&?&@2 93&>%,&?&@8 93&>%,&?&@9 7&>%,&?&@6 7&>%,&?&@2 7&>%,&?&@8 7&>%,&?&@9

7&>%,& 93&>%,& =7&>%,&

=7&>%,&?&@6 =7&>%,&?&@2 =7&>%,&?&@8 =7&>%,&?&@9 93&>%,&?&@6 93&>%,&?&@2 93&>%,&?&@8 93&>%,&?&@9 7&>%,&?&@6 7&>%,&?&@2 7&>%,&?&@8 7&>%,&?&@9

Constant power 50 mW/cm2

• Temperatures decrease

asymptotically for increased air flow

• Perpendicular flow gives

lowest and most homogeneous temperature

• Difference between 90°

and 45° is small

22

• 50
• Simulation in good

agreement with measurements

• 4

1 54 51 04 01 64 61 34 31 3 7 8 54 50

!"#$"%&'(%")*+,%"&-")./01) 23--3$&'"4)5/6"%).71)

94&:%,&;&<5 94&:%,&;&<0 94&:%,&;&<6 94&:%,&;&<3 4&:%,&;&<5 4&:%,&;&<0 4&:%,&;&<6 4&:%,&;&<3 =>?&94&:%,&;&<5 =>?&94&:%,&;&<0 =>?&94&:%,&;&<6 =>?&94&:%,&;&<3 =>?&4&:%,&;&<5 =>?&4&:%,&;&<0 =>?&4&:%,&;&<6 =>?&4&:%,&;&<3

14&.@A'.0& 4&:%,& 94&:%,&

• 94&:%,&;&<5

94&:%,&;&<0 94&:%,&;&<6 94&:%,&;&<3 4&:%,&;&<5 4&:%,&;&<0 4&:%,&;&<6 4&:%,&;&<3 =>?&94&:%,&;&<5 =>?&94&:%,&;&<0 =>?&94&:%,&;&<6 =>?&94&:%,&;&<3 =>?&4&:%,&;&<5 =>?&4&:%,&;&<0 =>?&4&:%,&;&<6 =>?&4&:%,&;&<3

&

Constant velocity 5 m/s

Summary

23

Thanks for your attention!

• CLIC machine and physics requirements place challenging demands
• n the vertex detector
• Initial layouts are being refined
• Active R&D into thin sensor assemblies and readout chips
• Powering, cooling and mechanical supports under design and test

The CLIC vertex detector: precision at high energy