Aucun titre de diapositive

1
D.Berst, J.Bol, W. de Boer M. Caccia, G.Claus,
C.Colledani, G.Deptuch, M. Deveaux, W.Dulinski,
G.Gaycken, D. Grandjean, L.Jungermann, J.L.
Riester, M.Winter
Radiation hardness of Monolithic Active Pixel
Sensors (MAPS)

• Outline
• Operation principle of MAPS
• Features of MAPS
• Radiation hardness tests on MAPS

2
What stands MAPS for?
Monolithic Readout-electronics and sensors are
integrated on the same substrate. Active Pixel
Signal processing microcircuits are integrated
in each pixel. Sensor
MAPS were developed for visible light
applications by industry. MAPS are produced with
standard CMOS-processes. RD for International
Linear Collider VD since 1999 _at_ IReS/LEPSI
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
3

The operation principle of MAPS
Sensor design
P-Well
Diode (N-Well)

• A MIP creates 80e/h pairs per µm in Si
• The Active volume (Epitaxial layer) is not
  depleted.
• Charge gets collected via thermal diffusion.

Epitaxial Layer
Substrate
Diode
1-2µm
4-14µm
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
4
The operation principle of MAPS
Particle trajectory
Preamplifier (one per pixel)
Diffusing free electrons
20-40µm
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
5
Some simple preamplifiers
3 Transistor Pixel
Self Bias Pixel
Amplifier (Source Follower)
High resistivity diode
Amplifier (Source Follower)
Classical MAPS-design leakage current gt
Pedestals after CDS Regular RESET is required
Leakage currents get compensated No RESET is
required
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
6
The MIMOSA - Technology
Minimum Ionizing Particle MOS Active Pixel Sensor

• Features of the MIMOSA (I VI) detectors
• Single point resolution 1.5µm - 2.5µm
• Typical Pixel pitch 20µm
• Thinning achieved to 120µm (1 Wafer to 50µm...)
• S/N for MIPs 20 40
• Detection efficiency gt 99
• 1MPixel sensor serial readout in 10ms
• Produced in various commercial CMOS-Processes
• Radiation hardness several 1011 up to
  1012neq/cm²

MIMOSA IV
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
7
Studies on radiation hardness
Key parameter Charge collection
MAPS were calibrated with 5,9 keV (55Fe)
Photons. Produce 1640 free electrons (same order
of magnitude as MIPs) Very local interaction
Hit in the depleted area. 100 collection
efficiency.
Hit in the epitaxial layer. Collection efficiency?
Hit in the substrate close to epitaxial layer.
Sensitive with limited efficiency.
Build a charge collection spectrum
8
First studies Mimosa 2
Peak from epitaxial layer. Substantial drop in
charge collection after irradiation
400kRad
Peak from depleted area. No change gt Readout
electro-nics ok.
Underground from substrate

• Conclusion on Mimosa 2 after 400kRad
• Leakage currents increases by a factor 5.
• Noise increases by some percent.
• Readout electronics OK.
• Charge collection drops by 50 (kills the chip)

9
Two preamplifiers, two different results
Mimosa 2 (MIETEC 0.35)
Mimosa 4 (AMS 0.35)
High resistivity diode
Amplifier (Source Follower)
Amplifier (Source Follower)
3 transistor pixel
Self bias pixel

• Conclusion
• Charge loss observed
• leakage current can be measured

• Indication
• No Charge loss
• Leakage current cannot be measured

CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
10
A working hypothesis
Main difference Presence of an enclosed reset
transistor
SiO2 non irrad.
SiO2 irradiated

• Reset transistor positive charge create
  locally high fields.
• P-Well barrier gets depleted.
• Parasitic charge collection by the
  Reset-Transistor.

CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
11
Successor1, a chip to study radiation hardness
SUCCESSOR Sucima Cmos ChargE SenSOR

• Designed within the SUCIMA project (FP5).
• Tests IReS, SUCIMA and GSI.
• Process AMIS 0.35 (Should be similar to MIETEC
  0.35)
• Design goal Radiation hard prototype for a
  medical dosimeter with high spatial resolution.
• 8 different pixel designs, only 1 discussed here
• 33 x 32 pixels
• 3.2 x 3.2 µm² diode size.
• Source and drain of reset transistor swapped
• Non uniform pixel pitch (25 and 35 µm steps)
• Irradiated Up to 1MRad X-Rays

Silicon ultra fast cameras for electron and gamma
sources in medical applications
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
12
Successor1 leakage current before and after
irradiation
I
y
x

• Observation
• Leakage current increases by a factor 200-300

Leakage currents are very different for identical
pixels depending on their position on the
chip. gt Induced heat due to nearby output
buffer?
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
13
Successor1 noise
(Lines to guide the eye)
(Lines to guide the eye)
Noise increase depends on the running
conditions More than factor 2 for 20C and
2.5MHz sampling (1ms integration time) Less than
20 for 15C and 10MHz sampling (0.2ms
integration time)
Even lower integration time is required for most
tracking applications
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
14
Successor1 charge collection
Successor1 before and after 1MRad X-Rays
Observations (Priliminary) No significant
influence on gain
T -15C t 200µs
Charge collected in 1 pixel ADC
No significant charge loss
Charge collected in 9 pixels ADC
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)
15

• Conclusion
• Charge loss, being the main problem so far,
  seems stopped.
• Noise increase can probably be handled.
• Evidence that a 1MRad resistant MAPS-detector
  can be build.
• Outlook
• Analysis on Successor1-data has to be refined and
  completed
• Beamtests with similar tracking chip were done.
• Results underway.

Long term Find a way to reduce leakage currents
CBM collaboration meeting, GSI Darmstadt, 6-8 Oct
2004, Michael Deveaux (michael.deveaux_at_ires.in2p3
.fr)