1 Improved Non-Ionizing Radiation Tolerance of CMOS Sensors Dennis Doering 1 *, Michael Deveaux 1, Melissa Domachowski 1, Michal Koziel 1, Christian Müntz.

1

Improved Non-Ionizing Radiation Tolerance of CMOS Sensors

Dennis Doering1*, Michael Deveaux1, Melissa Domachowski1, Michal Koziel1,

Christian Müntz1, Paul Scharrer1, Joachim Stroth1,2

1Institut für Kernphysik, Goethe University Frankfurt/M, Germany

2GSI Helmholtzzentrum für Schwerionenforschung, Darmstadt, Germany

Outline- CMOS Monolithic Active Pixel Sensors - Non-ionizing radiation damage effects- MAPS with high-resistivity epitaxial layer - Radiation tolerance- Triangle of non-ionizing radiation

hardness

/17/23

CBMSIS300

MAPS*(2003)

MAPS* (2012)

MIMOSA-26

Binary, 0

Single point res. ~ 5 µm 1.5 µm 1 µm 4 µm

Material budget < 0.3% X0 ~ 0.1% X0 ~ 0.05% X0 ~ 0.05% X0

Rad. hard. non-io. >1013 neq 1012 neq/cm² >3·1014 neq >1013 neq

Rad. hard. io > 3 Mrad 200 krad > 1 Mrad > 500 krad

Time resolution < 30 µs ~ 1 ms ~ 25 µs 110 µsDennis Doering: Improved Non-Ionizing Radiation tolerance of CMOS sensors RESMDD Florence Oct 2012

CMOS Monolithic Active Pixel Sensors

2Optimized for one parameter Current compromise

• Used by industry (digital camera)• Have been modified for charged particle

detection since 1999 by IPHC Strasbourg

• Foreseen for STAR, CBM, ALICE, ILC… => Sharing of R&D costs.

/17/23

Design of a MAPS

Dennis Doering: Improved Non-Ionizing Radiation tolerance of CMOS sensors RESMDD Florence Oct 2012 3

SiO2

N+ P+

P-

P+

Sensing diode

Epitaxial Layer

P-Well

Substrate

N+

50 µm

~50 µm thin sensors ⇒ Low material budget High granularity ⇒ Good spatial resolution

10-40 µm => a few µm resolution

/17/23

Charge spectrum


e-e-

Particle

Epitaxial Layer

Diode

0 250 500 750 1000 1250 1500 1750 20000

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Number of collected electrons

Yie

ld Photons (Fe-55)Summed cluster charge

Energy of Fe-55-K-Photon

/17/23

Non-ionizing radiation damage effect


Charge losses due to recombination at radiation-induced defects

e-

ParticleDefects due toradiation

Epitaxial Layer

Diode

0 250 500 750 1000 1250 1500 1750 20000

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000


Yie

ld

Unirradiated Irradiated

Photons (Fe-55)Summed cluster charge


/17/23

Influence of radiation damage and pixel pitch


e-

e-

Large pixel pitch

Small pixel pitch

Defects due to radiation

Small pixel pitch

Large pixel pitch

Epitaxial Layer

Diode

Epitaxial Layer

Diode

/17/23

Pixel pitch effect


In most of the cases, the whole charge can be detected.Little losses for larger pixel pitch even unirradiated

0 250 500 750 1000 1250 1500 1750 20000

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000


unirradiated

Yie

ld Small pitch Large pitch



/17/23

Pixel pitch effect


Signal losses depend on the pixel pitch.

0 250 500 750 1000 1250 1500 1750 20000

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000Energy of Fe-55-K-Photon


irradiated

unirradiated

Yie

ld Small pitch Large pitch Small pitch Large pitch


/17/23

Pixel pitch effect


0 5 10 15 20 25 30 35 401011

1012

1013

1014

1015

MIMOSA-9

MIMOSA-9

MIMOSA-15 (2006)MIMOSA-18 (2008)

Uncertainty rangeR

ad

iatio

n t

ole

ran

ce [

neq

/cm

²]

Pixel pitch [µm]

Smaller pixel pitch improves radiation tolerance. Drawback: Number of pixel Readout time Power consumption

/17/23

High-resistivity


Larger depleted volumes: ⇒ Accelerated charge collection ⇒ Improved non-ionizing radiation tolerance

SiO2

N+ P+

P-

P+

Epitaxial Layer

P-Well

Substrate

depleted volume

Low-resistivity ~ 30 Ωcm High-resistivity ~1k Ωcm

New CMOS process available:High-resistivity: Decrease of doping concentration in epitaxial layer.

Sensing diode

/17/23

MIMOSA-18 AHR


MIMOSA-18 AHR: Analog sensor with high-resistivity Epitaxial Layer.

Main features: a) High-resistivity ~1k Ωcm EPI layer b) 2.5V-3V depletion voltagec) Pixel pitch from 25µm down to 10µm

Irradiation up to 3·1014neq/cm² @ TRIGA reactor (Ljubljana/Slovenia)

/17/23

Signal response


0 200 400 600 800 1000 1200 1400 1600 1800 20000

500

1000

1500

2000

2500

3000

3500

4000

4500

Yie

ld


Low-resistivity unirradiated High-resistivity unirradiated

Landau-MPV: (293 ± 5) e(591 ± 4) e

More drift, less diffusion: ⇒Signal charge focused to seed pixel ⇒ Signal amplitude doubled

MIP-like β (Ru-106)Seed pixel

/17/23

Signal response


Radiation damage effect after 3·1014neq/cm² : Signal losses due to recombinations observed.However, the irradiated high-resistivity sensor exhibits a higher signal amplitude than the unirradiated low-resistivity sensor.

0 200 400 600 800 1000 1200 1400 1600 1800 20000

500

1000

1500

2000

2500

3000

3500

4000

4500

Yie

ld


Low-resistivity unirradiated High-resistivity unirradiated

High-resistivity 3·1014neq

/cm2

Landau-MPV: (293 ± 5) e(591 ± 4) e(491 ± 20) e

MIP-like β (Ru-106)Seed pixel

/17/23

5

10

15

20

25

30

35

40

45

50

55

60N

oise

[e]

Temperature [°C]

Unirradiated

1014neq

/cm2

3·1014neq

/cm2

-3

Noise


Rad

iatio

nda

mag

e

Substantial increase in the bulk noise is observed.

/17/23

5

10

15

20

25

30

35

40

45

50

55

60N

oise

[e]

Temperature [°C]

Unirradiated

1014neq

/cm2

3·1014neq

/cm2

-3

Noise


Use of temperature dependence of bulk noise

Rad

iatio

nda

mag

e

/17/23

5

10

15

20

25

30

35

40

45

50

55

60

-34-27-15

Noi

se [e

]

Temperature [°C]

Unirradiated

1014neq

/cm2

3·1014neq

/cm2

-3

Noise


Cooling

Noise is alleviated to a factor of 2 with decreasing temperature.Expect further noise reduction in case of faster readout.

Rad

iatio

nda

mag

e

/17/23

Signal to Noise ratio


0 5 10 15 20 25 300

10

20

30

40

50

60

70

80

10µm (Low-resistivity)

Sig

na

l to

No

ise

Radiation dose [1013neq

/cm2]

S/N limit (MIPS)

/17/23



0 5 10 15 20 25 300

10

20

30

40

50

60

70

80 Ru-106

25 µm


T= -34°C

Sig

na

l to

No

ise


/cm2]

Error bars: Fit uncertainty + 10% noise uncertainty

S/N limit (MIPS)

(High-resistivity)

/17/23



0 5 10 15 20 25 300

10

20

30

40

50

60

70

80 Ru-106 12.5 µm 25 µm


T= -34°C

Sig

na

l to

No

ise


/cm2]


S/N limit (MIPS)

High-resistivity

/17/23



0 5 10 15 20 25 300

10

20

30

40

50

60

70

80 Ru-106 10 µm 12.5 µm 25 µm


T= -34°C

Sig

na

l to

No

ise


/cm2]


S/N limit (MIPS)

High-resistivity

/17/23

Radiation tolerance


0 5 10 15 20 25 30 35 401011

1012

1013

1014

1015

MIMOSA-18AHR* (2011)

MIMOSA-9

MIMOSA-9

MIMOSA-15 (2006)MIMOSA-18 (2008)

Sensor based on low-resistivity EPI layerR

ad

iatio

n t

ole

ran

ce [

neq

/cm

²]

Pixel pitch [µm]*operated at -34°C

Preliminary

Sensor based on high-resistivity EPI layer

/17/23

Summary


Radiation tolerance studies up to 3·1014neq/cm²

Results:• CMOS sensors based on high-resistivity epitaxial layer exhibit

substantially improved performance, (S/N ~ doubled) • Radiation tolerance depends on the pixel pitch• Laboratory test indicates:

Radiation tolerance is beyond 3·1014neq/cm² (Pitch 10µm, cooled sensor)

Outlook: • Ionizing radiation hardness addressed by MIMOSA-32 – fabricated in an

0.18µm process -> S. Senyukov (next talk)

/17/23

Triangle of non-ionizing radiation tolerant MAPS


Pitch

Res

istiv

ityTem

perature

/17/23Dennis Doering: Improved Non-Ionizing Radiation tolerance of CMOS sensors RESMDD Florence Oct 2012 24

BACKUP

/17/23

10

20

30

40

50

60

-34-27-15

No

ise

[e

]

Temperature [°C]-3

Radiation tolerance

10

20

30

40

50

60

-34-27-15

No

ise

[e

]

Temperature [°C]-3

Noise increases

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)

0 50 100 150 200 250 300 350 4000

1000

2000

3000

4000

5000

6000

7000

En

trie

s [1

bin

=4

AD

C]

Charge collected [ADC]

<20% less entriesThinner active vol.?

CCE ok

Gain okFe-55 (X-rays)

0 500 1000 1500 2000 2500 3000 35000

500

1000

1500

2000

2500

3000

3500

En

trie

s [1

bin

=4

AD

C]

Charge collected [e]

Ru-106 (b-rays)

99% det. eff.after irrad.

620e (MPV)

490e (MPV) <20% less signalThinner act. vol.?

Noise increases =>Compensate with cooling.

3 · 1014neq/cm² + O(3 MRad)Not irradiated

Dennis Doering: Improved Non-Ionizing Radiation tolerance of CMOS sensors RESMDD Florence Oct 2012 25Preliminary conclusion: Sensor tolerates 3 · 1014neq/cm², to be confirmed in beam test


0 50 100 150 200 250 3000

500

1000

1500

2000

2500

3000

3500

4000

En

trie

s [1

bin

=4

AD

C]


T= -34°C

0 50 100 150 200 250 3000

500

1000

1500

2000

2500

3000

3500

4000

En

trie

s [1

bin

=4

AD

C]


0

3·1013neq

/cm²

1014neq

/cm²

3·1014neq

/cm²

T= -34°C

CCE ok

Fe-55 X-ray CCE shiftedMIMOSA-18 AHR 25µm (A2)MIMOSA-18 AHR 12µm (A1)

gain ok

12 µm pitch: Average CCE is constant. 25µm pitch: Shift to lower values as observed in sensors based on low resistivity EPI layer.

Pixel with 12µm and 25µm pitch (Fe-55)

Seed pixel cluster


12 µm pitch: Shift of cluster peak now visible ⇒ some signal charges lost which diffuse longer distance (to neighbor pixels)

25µm pitch: Dramatic shift ⇒ 12µm pixel pitch: in a part of Epitaxial layer signal lost due to recombination

can be neglected

Pixel with 12µm and 25µm pitch (Fe-55)

0 50 100 150 200 250 3000

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Entrie

s [1

bin

=4A

DC

]


MIMOSA-18 AHR 12.5µm (A1)Fe-55 X-ray

T= -34°C

Cluster of 25 pixel => Larger diffusion paths

0 50 100 150 200 250 3000

1000

2000

3000

4000

5000

6000

7000

0

3·1013neq

/cm²

1014neq

/cm²

3·1014neq

/cm²

Entrie

s [1

bin

=4A

DC

]


MIMOSA-18 AHR 25µm (A2)Fe-55 X-ray

T= -34°C

/17/23

Temperature measurement with infrared camera


Huber system CC 815-70°C - + 35°C

MIMOSA-18

Camera paintSensor is operating

Infrared

-3°C /-20°C

T of cooling liquidT of sensor

/17/23

High resistivity


Doping concentrationSize of the depleted zone

Idea: Decreasing the doping concentration from 1015 should increase the size of the depleted zone: Þ Improved performance expected.

Standard: 13 Ω cm; NA≈1015 cm-3

High resistivity: 400 Ω cm; NA≈3.3 1013 cm-3

/17Dennis Doering: Improved Non-Ionizing Radiation tolerance of CMOS sensors RESMDD Florence Oct 2012 30/23

Applications of MAPS

MAPS are developed for applications as vertex detector since 1999 at IPHC (Strasbourg).

CBM-Experiment (FAIR, GSI)STAR-Experiment

Possible ITS-Upgrade ALICEInternational Linear Collider

/17Dennis Doering: Improved Non-Ionizing Radiation tolerance of CMOS sensors RESMDD Florence Oct 2012 31/23

MIMOSA-26 - Working horse since several years

JTAG slow control On-chip voltageregulators

1152 discriminators

zero suppr. logic

Output memories

Read out time: ~100 µs => ~ 10k frames/second

Radiation tolerance: >300 kRad; >

21.2 x 10.6 mm²

Pixe

l colu

mn

Digital part

Sensing part

/17/23

0,0 0,2 0,4 0,6 0,8 1,0 1,20

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Charge Collection Efficiency

irradiated

unirradiated

Ent

ries

Small pitch Large pitch Small pitch Large pitch

Pixel pitch effect


Signal losses depend on the pixel pitch.


1 Improved Non-Ionizing Radiation Tolerance of CMOS Sensors Dennis Doering 1 *, Michael Deveaux 1, Melissa Domachowski 1, Michal Koziel 1, Christian Müntz.

Documents

pixel pitchdennis doering

influence of radiation

larger pixel pitch

smaller pixel pitch

mapsdennis doering

new cmos process available

s110 sdennis doering

charge losses