1. Radiation tolerance of MAPS
M. Deveaux1 on behalf of
IKF Frankfurt and the IPHC-PICSEL group
1Institut für Kernphysik, Goethe Universität Frankfurt am Main

2. Non-ionizing radiation damage Ionizing radiation damage
Outline
Definitions
Non-ionizing radiation damage
Ionizing radiation damage
Annealing effects
Random Telegraph Signal
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

3. Ionizing and non-ionizing radiation
To be measured and improved: Radiation hardness against…
Ionising radiation:
Energy deposited into the electron cloud
May ionise atoms and destroy molecules
Caused by charged particles and photons
Non-ionising radiation:
Energy deposited into the crystal lattice
Atoms get displaced
Caused by heavy (fast leptons, hadrons)
charged and neutral particles
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

4. The operation principle of MAPS
+3.3V
Reset
+3.3V
Output
SiO2
SiO2
SiO2
N++
N++ N+ P+ P- P+
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

5. Tolerance against non-ionising radiation
+3.3V
Output
+3.3V
GND
SiO2
SiO2
GND
SiO2
SiO2 N+
P++
P++
N++
P++
Bulk damage
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile
Non-ionising radiation
Energy deposit into crystal lattice

6. Tolerance against non-ionising radiation
+3.3V
Output
+3.3V
GND
SiO2
SiO2
GND
SiO2
SiO2 N+
P++
P++
N++
P++
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

7. CCE-Measurements with 55Fe
5,9 keV (55Fe) Photons:
Small, point-like interaction region.
Number of generated e/h pairs similar to MIPS
5,9keV=>1640 e/h pairs
Which fraction of electrons is collected?
=> Charge collection efficiency (CCE)
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

8. How to study charge collection efficiency
(in groups of 4 pixels)
After neutron irradiation
Before neutron irradiation
Hits in Epitaxial-Layer
Hits in Diode
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile 8

9. CCE as function of neutron fluence
Charge collected in groups of 4 pixels
CCE decreases with increasing radiation.
Decrease is slower for thin epi-layers, high diode density, big diodes
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

10. Tolerance against non-ionising radiation
+3.3V
Output
+3.3V
GND
SiO2
SiO2
GND
SiO2
SiO2 N+
P++
P++
N++
P++
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

11. Tolerance against non-ionising radiation
Mi-18
Mi-15
Mi-9
Mi-9
Small pixels seem mandatory to improve radiation hardness
But: Need to readout additional pixels => Reduced speed, higher power etc…
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

12. Tolerance against non-ionising radiation
+3.3V
Output
+3.3V
GND
SiO2
SiO2
GND
SiO2
SiO2 N+
P++
P++
N++
P++ E
Electric field increases the radiation hardness of the sensor
Draw back: Need CMOS-processes with low doping epitaxial layer
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

13. CCE of MIMOSA-26, standard vs. high resistivity
Radiation dose
CCE shrinks after
irradiation to <50%
of initial value.
x 1012 n/cm²
-20°C, Std. epi, seed pixel
CCE remains mostly
stable.
-20°C
400 Ohm cm, 15µm , seed pixel
D. Doering
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

14. Radiation tolerance Sensor: - Mi-18 AHR, SB-Pixel, 10 µm pitch
3 · 1014neq/cm² + O(3 MRad)
Not irradiated
Sensor:
- Mi-18 AHR, SB-Pixel, 10 µm pitch
- Epitaxial layer: 400 W cm, 15 µm Irradiation:
- fast reactor neutrons (Triga, Ljubljana)
- Chip not powered during irradiation
- Dose: 3 · 1014neq/cm² + O(3 MRad)
Noise increases
Noise increases =>
Compensate with cooling.
<20% less entries
Thinner active vol.?
CCE ok
Gain ok
Fe-55 (X-rays)
Ru-106 (b-rays)
99% det. eff.
after irrad.
620e (MPV)
490e (MPV)
<20% less signal
Thinner act. vol.?
Preliminary conclusion: Sensor tolerates 3 · 1014neq/cm², to be confirmed in beam test

15. Radiation tolerance of MIMOSA-HR
Preliminary
T< -30°C!
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

16. Leakage current, an illustrative example
1013neq/cm²
Leakage current:
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

17. Dependencies of (shot-)noise1 8 ms 10 80
-25°C
+10°C
+40°C
MIMOSA-11, 500kRad X-rays
Dose [neq/cm²]
Noise [e ENC]
too high
Mi18, A. Büdenbender, Bachelor Thesis
still ok
2e13 50
Radiation induced noise is reduced by:
Cooling
Fast readout
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

18. Ionizing radiation Ionising radiation:
Energy deposited into the electron cloud
May ionise atoms and destroy molecules
Caused by charged particles and photons
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

19. Response of MAPS to 55Fe-photons before and after irradiation
Signal of MIMOSA-2 before and after 400kRad
Design and test of 6 detector generations
was needed to learn and improve the design:
MIMOSA-2
MIMOSA-4
SUCCESSOR-1
SUCCESSOR-2
MIMOSA-9
MIMOSA-11
Dead after 400 kRad
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

20. Radiation tolerance against ionising radiation
Results
+3.3V
Reset
+3.3V
Output
SiO2
SiO2
Positive Charge
N++
N++ N+ P+ P- P+
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile
Ionising radiation
Energy deposit into electron cloud

21. Improvement of radiation tolerance
Output
SiO2
N++ N+
P++
GND
P++ guard ring cuts
conduction paths
Thicker P+ isolation found in AMS 0.35
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

22. Dependencies of (shot-)noise1 8 ms 10 80
-25°C
+10°C
+40°C
MIMOSA-11, 500kRad X-rays
Radiation induced noise is reduced by:
Cooling
Fast readout
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

23. How to improve radiation hardness
MIMOSA-2 before and after 400kRad
OK after 1000 kRad
Open issue: Radiation tolerance of “column parallel sensors”
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

24. Column parallel sensors
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

25. Radiation tolerance of column parallel sensors
M. Koziel
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

26. “Single weak point” was not spotted => Use 0.18 µm – process.
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

27. Pedestals in an 1 MRad irradiated Mimosa-26
After 1 day
After 4 days
After 19 days
After 31 days
tint = 400µs (fclk = 20MHz), Tcooling = 15°C
Origin of the effect not yet understood.
Effect substantially alleviated at nominal readout frequency.
Currently radiation tolerance to ~300 kRad
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

28. About the benefits of a little fever
Thermal annealing
About the benefits of a little fever
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

29. Annealing: Modify defects by thermal treatment
Beneficial annealing: Radiation damage is reduced
Reverse annealing: Additional damage is observed
(defects group to more dangerous defect clusters)
Used chips:
Mimosa19
73k pixels; 12 µm pixel pitch
Read out time: 3.69 ms
(only analog read-out)
a) ~10 keV X-rays (200 kRad)
b) Fast reactor neutrons: +(~100 kRad γ)
→ ionizing radiation suppressed (sensors not powered)
c) Combined radiation:
→ ; 1 year at room temperature ; 200 kRad X-rays.
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

30. Annealing study, temperature profile
Time
Neutron
radiation
1 year
X-ray
Measurements and
storage at +20°C
(280h)
Heating at +80°C (73h)
Measurements and storage
at +20°C (191h)
2h transport
T[°C]
+80°C
+20°C
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

31. Annealing study, results
D. Doering, RESMDD 2011
Annealing at
+20°C neglected
No reverse annealing observed for up to 70h at +80°C.
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

32. Annealing of noise Noise at -20°C
(combined radiation/no annealing at +80°C)
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

33. Random Telegraph Signal
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

34. Excotic effects: Random Telegraph Signal
Fake hits in MIMOSA-18 (Window of 64x64 pixels)
After 1013 neq/cm²
Fake hits in 5000 Events
D. Doering, Bachelor Thesis
PixelNr.
M. Deveaux, Seminar at IPHC ,  Strasbourg, France

35. Random Telegraph Signal, a radiation damage of MAPS
UCDS [ADC]
Time
Output signal of two MAPS pixels U1 Q IL
Threshold
Irradiate sensor with ~ 1 MeV reactor neutrons
FRM II, Garching)
Observation: RTS may generate many fake hits
>> 1% of all pixels generate fake hits
How to improve? What is the origin?
Working hypothesis: IL(t) is modulated=> RTS in diode.
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

36. Temperature effects on RTS
Hopkins et al. (2000):
Frequency and amplitude of RTS increase with temperature
Output signal of a selected MAPS pixel showing RTS
* G.R. Hopkinson: “Radiation Effects in a CMOS Active
Pixel Sensor”, IEEE-TNS Vol. 47, No. 6, P. 2480
Hypothesis: Cooling => RTS-Amplitude < Threshold => No fake hits?
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

37. Identified RTS - pixels as function of temperature (3T-pixel)
RTS pixel if signature with:
Amplitude >150e ENC
Tmin = ~ 0.5s
Tmax= ~ 45 min
Preliminary
Number of identified RTS - pixels is dramatically reduced by cooling
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

38. Fake hit rates of irradiated MAPS detectors
Identified RTS provides % of all fake hits
Fake hit rate of warm 3T-pixels seems inacceptable
Cooled SB-pixels show acceptable fake hit rate for < 1013 neq /cm²
M. Deveaux, Seminar at IPHC ,  Strasbourg, France

39. Summary and conclusion
MAPS react to non-ionizing radiation with:
Increased leakage current
Increased noise
Random Telegraph Signal
Alleviate with cooling and fast readout
Drop in charge collection efficiency
Alleviate with small pixels, high resistivity epitaxial layer
Limits: Some 1012 neq/cm² (room temperature),
potentially > 1014 neq/cm² (if cooled).
Ionizing radiation:
Strongly increased leakage currents and noise
Pedestal shifts
“odd effects”
Alleviate with cooling and faster readout
Use 0.18µm process for future improvement
Limits: ~ 300kRad (Col || readout, room temperature),
> 1 MRad (analog readout, cooled)
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

40. Backup
Backup
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

41. Radiation tolerance Sensor: - Mi-18 AHR, SB-Pixel, 10 µm pitch
3 · 1014neq/cm² + O(3 MRad)
Not irradiated
Sensor:
- Mi-18 AHR, SB-Pixel, 10 µm pitch
- Epitaxial layer: 400 W cm, 15 µm Irradiation:
- fast reactor neutrons (Triga, Ljubljana)
- Chip not powered during irradiation
- Dose: 3 · 1014neq/cm² + O(3 MRad)
Noise increases
Noise increases =>
Compensate with cooling.
<20% less entries
Thinner active vol.?
CCE ok
Gain ok
Fe-55 (X-rays)
Ru-106 (b-rays)
99% det. eff.
after irrad.
620e (MPV)
490e (MPV)
<20% less signal
Thinner act. vol.?
Preliminary conclusion: Sensor tolerates 3 · 1014neq/cm², to be confirmed in beam test

42. Chips with intrinsic leakage current compensation (SB-pixel)IL t
vdd (+3.3V) IC U2
IC, U2 t IL U1
UCDS t
Threshold
(Output)
RTS should only be visible in UCDS, if new equilibrium is being established
Steps should be visible in U2
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

43. Raw data from SB-pixels with RTS
Uvdd-U2
U1(t) [shifted, a.u.]
UCDS (t) [a.u.]
UCDS
Step within the
lifetime of the
readout system
Steps within dead time
of the readout system
Time [a.u.]
Observation: SB-pixels suppress RTS in output signal.
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

44. Temperature measurement
T of sensor
T of cooling liquid
-3°C (-20°C) 44
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

45. on-line compilation: http://sesam.desy.de/~gunnar/Si-dfuncs
Relative non ionising dose of neutrons and pions
Data from: A. Vasilescu and G. Lindstroem, Displacement damage in Silicon
on-line compilation:
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile

46. Need for high resistivity
Size of the
depleted zone
Doping concentration
Idea: Decreasing the doping concentration from 1015 to 1013 should increase the size of the depleted zone.
Manifacuring processes became only recently available
Sensors:
MIMOSA-26 AHR (low depletion voltage, full readout chain)
MIMOSA-18 AHR (high depletion voltage, analog readout)
M. Deveaux, Workshop on integrating highly granular and light vertex detectors , 6-9th Sept. 2011, Mont Saint Odile