Silicon Pixel and Strip Detectors for LHC Experiments

1
Silicon Pixel and Strip Detectors for LHC
Experiments

• 1st Coordination Meeting of the CBM Experiment at
  the future GSI facility
• GSI, Nov. 15-16, 2002

P. Riedler ALICE Silicon Pixel Team CERN
2
Acknowledgements M. Campbell, P. Collins, H.
Dijkstra, F. Faccio, H. Pernegger, G. Stefanini
and the ALICE SPD Team
3
Outline
- The LHC and its experiments

• Radiation damage in silicon
• Electronics
• Detectors

• A closer look at the ALICE SPD

ALICE Silicon Pixel Telescope Reconstructed
event Testbeam 2002
GSI - 15/11/2002
P.Riedler - CERN
3
4
(No Transcript)
5
The LHC and its Experiments

• head-on collisions of protons (7TeV on 7 TeV)
• and heavy ions
• Lmax1034cm-2 s-1
• f(4cm)3 1015 (neq) cm-2 in 10 years
• (gt85 charged hadrons)
• ! RADIATION DAMAGE !

• Detectors for LHC under full construction now
• Installation 2006, First Beam 2007
• gt RD groups (e.g. RD48, now RD50) already work
  on solutions for next generation of detectors

6
(No Transcript)
7
2 general purpose detectors Higgs in SM and in
MSSM, supersymmetric Particles, B physics (CP
violation, ...),
CMS
ATLAS
Strips 61m2, 6.3 x 106 channels Pixels 2m2,
80 x 106 channels
210m2, 9.6 x 106 channels 2m2, 33 x 106 channels
8
CP violation and rare decays
Heavy ion physics
ALICE
LHCb
Strips 4.9m2, 2.6 x 106 channels Drifts
1.3m2, 1.33 x 105 channels Pixels 0.2m2, 9.83
x 106 channels
VELO 0.32m2, 2 x 105 channels Tracker 14m2, 8
x 105 channels HPD 0.02m2, 1 x 106 channels
9
(No Transcript)
10
Radiation Damage in Silicon
Surface Damage
Bulk Damage
e.g. ATLAS Pixel Detector
11
Electronics
Single Event Effects (SEE)
Cumulative Effects
Permanent (e.g. single event gate rupture
SEGR) Static (e.g. single event upset
SEU) Transient SEEs
Total Ionizing Dose (TID) Ionisation in the SiO2
and SiO2-Si interface creating fixed charges
(all devices can be affected)
Displacement Defects (bipolar devices,
opto-components)
In the following the effects of TID only will be
discussed
12
Total Ionizing Dose

• Ionization due to charged hadrons, g, electrons,
  in the SiO2 layer and SiO2-Si interface
• Fixed positive oxide charge
• Accumulation of electrons at the interface
• Additional interface states are created at the
  SiO2-Si border

R. Wunstorf, PhD thesis 1992
13
Effects of TID in CMOS devices
Threshold voltage shift, transconductance and
noise degradation, source drain leakage, leakage
between devices
E.g. transistor level leakage and threshold
voltage shift
F. Faccio, ELEC2002
Parasitic channel between source and drain
14
Radiation Levels in some LHC experiments
total dose fluence 1MeV n eq. cm-2
after 10 years ATLAS Pixels 50 Mrad 1.5 x
1015 ATLAS Strips 7.9 Mrad 2 x 1014 CMS
Pixels 24Mrad 6 x 1014 CMS
Strips 7.5Mrad 1.6 x 1014 ALICE
Pixel 500krad 2 x 1013 LHCb VELO - 1.3 x
1014/year Set as limit, inner layer
reaches this value after 2 years inner part
of detector (inhomogeneous irradiation )
A radiation tolerant design is important to
ensure the functionality of the read out over the
full life-time!
15
Solution - Technology Hardening
Flatband-voltage shift as function of the oxide
thickness

• Tunneling of trapped charge in thin oxides
• D VT 1/tox2 for tox gt 10nm
• D VT 1/tox3 for tox lt 10nm

After N.S. Saks, M.G. Ancona, and J.A. Modolo,
IEEE Trans.Nucl.Sci., Vol. NS-31 (1984) 1249
16
Using a 0.25µm CMOS process reduces th-shift
significantly
17
Enclosed geometrie to avoid leakage
18
F. Faccio, ELEC2002
19
Front end technology choices of the different
experiments
Technology Chip ALICE Pixel 0.25µm
CMOS ALICE1 ALICE Strips 0.25µm
CMOS HAL25 ALICE Drift 0.25µm
CMOS PASCAL ATLAS Strips DMILL ABCD ATLAS
Pixel DMILL-gt0.25µm CMOS FE-D25 CMS
Pixel DMILL-gt0.25µm CMOS PSI CMS Strips 0.25µm
CMOS APV25 LHCb VELO DMILL/0.25µm
CMOS SCTA/Beetle LHCb Tracker 0.25µm
CMOS Beetle
Deep sub-µm means also speed, low power, low
yield, high cost
20
Radiation Damage in Detectors

• Surface Damage
• Creation of positive charges in the oxide and
  additional interface states.
• Electron accumulation layer.

21
Macroscopic Effects

• Surface Damage
• Increase of interstrip
• capacitance (strips!)
• Pin-holes (strips!)

• Bulk Damage
• Increase of leakage current
• Increase of depletion voltage
• Charge trapping

Effects signal, noise, stability (thermal
run-away!)

• Annealing effects will not be discussed here.
• But Do not neglect these effects, esp. for long
  term running!
• All experiments have set up annealing scenarios
  to simulate the damage after 10 years.

22
Leakage current
23
Depletion Voltage
Type-Inversion n-type bulk starts to behave like
p-type bulk -gt depletion from the backside of the
diode! Vdep increases with fluence (after
inversion)
If depletion voltage has increased too so much
that underdepleted operation is necessary-gt
charge loss and charge spread!
24
Possible Solutions

• n-in-n detectors
• Underdepleted operation is possible!

25
2. Oxygenated Silicon
Defect engineering (RD48) - to reduce reverse
annealing gt Lower depletion voltage can be
expected after several years sunning (including
warm-up times)
But improvement only for charged hadrons and g.
No effect for neutrons observed. Also spread
of depletion voltage of detectors from different
suppliers can reduce the beneficial effect
ATLAS pixel uses oxygenated Si
26

• Further solutions to allow a reasonable operating
  voltages even after high fluences and annealing
• Low resistive silicon
• Thin detectors (also intersting for material
  budget reasons)
• CZ starting material (under investigation)
• lt100gt to reduce interstrip capacitance

Choice of LHC experiments ALICE pixel p-in-n s
tandard FZ ATLAS pixel n-in-n oxygenated ATLAS
strips p-in-n standard FZ CMS
pixel n-in-n standard FZ CMS strips p-in-n sta
ndard FZ lt100gt LHCb VELO n-in-n standard FZ
27
A closer look at the ALICE Silicon Pixel Detector
(SPD)
2 barrel layers
D z 28.3 cm r 3.9 cm 7.6 cm
INFN Padova
28
The two barrels will be built of 10 sectors, each
equipped with 6 staves
stave
INFN Padova
INFN Padova
Material budget(each layer) 0.9 X0 (Si 0.37,
cooling 0.3, bus 0.17, support
0.1) (lowest material budget of all pixel
detectors!)
29
Each Stave is built of two HALF-STAVES, read out
on the two sides of the barrel, respectively.
193 mm long
30

• Bus
• 7 layer Al-Kapton flex
• Wire bonds to the ALICE1LHCb chip

240µm
200µm
goal150µm
M.Morel
31
ALICE1LHCb chip
Multi Chip Module (MCM)
AP
DP
GOL
Laser and pin diode

• Analog Pilot
• Reference bias
• ADC (T, V and I monitor)

Data out
JTAG
Clock

• Digital Pilot
• Timing, Control and Readout

32
ALICE1LHCb chip

• Mixed signal (analogue, digital)
• Produced in a commercial
• 0.25µm CMOS process
• Radiation tolerant design
• (enclosed gates, guard rings)
• 8192 pixel cells
• 50 µm x 425 µm pixel cell
• 100 µW/channel

33
Low minimum threshold 1000 electrons Low
individual pixel noise100 electrons
34
Fully developed test system for wafers
Class I - Mean Threshold
Class I 42-75 Class II 6-12 Class III
17-42 (sample 4 wafer, 750µm)
Production testing will start this autumn
35
Ladders and Assemblies
Detector

• Detectors
• single chip detectors
• 5 chip detectors for ladders
• p-in-n
• 300 µm thick(tests) -
• final thickness 200µm
• Chips
• single chips
• 750 µm thick (tests) - 150µm final

Chip

• Bump-bonding
• VTT/Finland
• Pb-Sn solder bumps
• AMS/Italy
• In bumps

36
First testbeam with full size ladder - July 2002
chip0
chip1
chip2
chip3
chip4
37
Detector
Chips
38
(No Transcript)
39
Sr-source measurement of thin ladder (300µm chip,
200µm detector)
Missing Pixels missing working
63 28 0.34 99.66
53 21 0.26 99.74
50 44 0.53 99.47
43 3 0.04 99.96
33 61 0.74 99.26
Chip 33
Chip 43
Chip 50
Chip 53
Chip 63
matrices
40
Summary

• All LHC experiments use silicon detectors to
  improve their tracking capabilities (up to
  gt200m2!).
• Installation foreseen in 2006.
• The high radiation environment demands radiation
  tolerant technologies for front end chips and
  detectors.
• Almost all silicon detectors use 0.25µm CMOS
  chips (future?).
• P-in-n and n-in-n detectors are used depending
  on the expected fluences and the annealing
  damage.
• The current challenges are the actual
  construction and integration of the detectors.