P-STOP isolation technique

Download Report

Transcript P-STOP isolation technique

List of Authors:
Maria Rita Coluccia,
J. A. Appel, G. Chiodini,
D. C. Christian, S. W. Kwan, G. Sellberg
with Fermi National Accelerator Laboratory
L. Uplegger
with INFN Milano (Italy)
ABSTRACT
We present IV and CV curves for irradiated
prototype n+/n/p+ silicon pixel sensors, intended
for use in the BTeV experiment at Fermilab. We
tested pixel sensors from different vendors and
with two pixel isolation techniques: p-stop and pspray. Results are based on exposure with 200
MeV protons up to 6 x 1014 protons/cm2.
• BTeV [1] is a new heavy quark experiment that
will run at the TEVATRON collider at Fermi
National Accelerator Laboratory.
• It is designed to cover the “forward” region of the
proton-antiproton interaction point running at a
luminosity of 2 x 1032 cm-2s-1.
• The experiment will employ a silicon pixel vertex
detector to provide high precision space points for an
on-line lowest-level trigger based on track impact
parameters [2].
THE VERTEX DETECTOR
ONE STATION
The baseline BTeV silicon pixel detector has rectangular 50 μm x
400 μm pixel elements and consists of a regular array of 30
“stations” of “planar” pixel detectors distributed along the
interaction region.
L-SHAPE
HALF PLANE
Each station contains one
plane with the narrow pixel
dimension vertical, and one
with the narrow dimension
horizontal, and is split in
order to allow the sensors to
be moved away from the
beam during acceleration
and other unstable beam
conditions.
HYBRID ASSEMBLY
The basic building block of the detector is a hybrid assembly
consisting of a sensor, a number of readout chips and a flexible
printed circuit (a high-density interconnect: HDI).
SENSOR DESIGN
pixel sensors have n+/n/p+
The BTeV
configuration and
therefore it is necessary to provide explicit electrical isolation
between neighboring n+ electrodes. We explore two techniques.
P-STOP ISOLATION TECHNIQUE
P-STOP isolation technique [3] in which a high dose
p-implant surrounds each n+-type region.
P-SPRAY ISOLATION TECHNIQUE
P-SPRAY isolation technique [4], developed by the ATLAS
collaboration, that consists of a medium dose p-implant that is applied
to the entire n-side and is overcompensated by the high dose n+ pixel
implants. To increase radiation hardness and the breakdown voltage
before irradiation, a “grading” of p-spray implantation (moderated pspray) [5] is used, that leads to a step in the effective p-spray dose
along the gap between two n+-implants.
DEVICES TESTED
P-STOP sensors
•
•
•
From SINTEF (Norway).
Low resistivity (1.0-1.5 KΩcm) <100> silicon, 270μm thick.
Two arrays tested:
1. test-sized sensors 12 x 92 cells that, except for four,
are all connected together.
2. FPIX1-sized sensors 18 x 160 cells (same size as the
readout chip FPIX1 [6], no bias grid structure).
P-SPRAY sensors
•
•
•
•
From TESLA (Czech Republic) and CiS (Germany).
High resistivity (2-5 KΩ cm) <111> silicon, 250 μm thick.
One array tested (FPIX1-sized sensors).
Bias grid structure for biasing all the pixel simultaneously.
P-STOP SENSOR PERFORMANCES
I-V CURVES BEFORE IRRADIATION
Fig. 1
Typical I-V curves for both devices before irradiation. The test-sized
sensors show very good performance. The FPIX1-sized sensors have lower
breakdown voltage. This appears to be related to the fact that it is difficult to bias all
of the pixel cells during wafer testing.
I-V CURVES BEFORE IRRADIATION
Fig. 2
The breakdown voltage of the FPIX1-sized sensors increases considerably
after bump bonding to a readout chip. We tested 5 sensors that were bump
bonded to a readout chip and 3 sensors, with indium bumps deposited on the n+ side,
that were glued with conductive silver epoxy to a piece of silicon in order to mimic
the presence of the readout chip.
I-V CURVES AFTER IRRADIATION
Fig. 3
I-V curves for a test-sized p-stop sensor irradiated up to 1.5 E14 p/cm-2
The current after irradiation increases by a few orders of magnitude. However,
operating at lower temperature can alleviate this problem.
Leakage Current: Temperature Dependence
I leak  T e
2
Fig. 4
(
Eg
2 k BT
)
We repeated the measurements at various temperatures (10 oC, 0 oC and -10
oC). As expected, we observed that the current decreases exponentially with temperature
(Ileak  T2 exp (-E / 2kBT) [7]). The figure shows the comparison between data and the
predicted dependence of the leakage current vs temperature. There is good agreement
between the fit and the data.
C-V CURVES AND DEPLETION VOLTAGE
Fig. 5
Depletion voltage as a function of proton fluence. Note that for 1
year of BTeV running at nominal luminosity the fluence will be 1 x 1014 p/ cm2
C-V CURVES AND DEPLETION VOLTAGE
Fig. 6
Here a typical example of bulk capacitance versus bias potential for a
test-sized p-stop irradiated up to 1.2 x 1014 p/cm2 . We use this method to determine
the depletion voltage.
P-SPRAY SENSORS PERFORMANCE
I-V CURVES BEFORE AND AFTER IRRADIATION
Fig. 7
We irradiated
two of these sensors, one
up to 8  1013 p/cm2 and
the other one up to 1.2 
1014 p/cm2. The figure
shows the increase in the
leakage current due to the
irradiation for the most
irradiated sensor.
C-V CURVE AFTER IRRADIATION
Fig. 8
CV measurement for sensor irradiate to 1.2  1014 p/cm2 . Vdep ~70 V.
Before irradiation was Vdep ~60 V . Type inversion already occurred.
CONCLUSIONS
Two different pixel isolation techniques were studied:
• P-stop isolation:
1. Most of the tested sensors meet specifications.
2. Problem to measure the breakdown voltage before
bump bonding to readout chip.
• P-spray isolation:
1. Results obtained are promising.
2. It is possible to determine the breakdown voltage due
to a bias grid.
3. Need more irradiation studies.
We will study the performance of these sensors, bump bonded
to a readout chip, in a test beam.
REFERENCES
[1] A. Kulyavtsev, M. Procario, J. Russ, J. You, J. Cumalat, J. Appel, et al.
"BTeV proposal," Fermilab, May 2000.
[2] E. E. Gottschalk, "BTeV detached vertex trigger," Fermilab-CONF-01-088-
E,
June 2001.
[3] G. Batagnani, F. Bosi, L. Bosisio, A. Conti, E. Focardi, F. Forti, et al..
"Double-sidedreadout silicon strip detectors for the ALEPH minivertex,"
Nucle. Instr. Meth., vol. A 277, pp 147-153, 1989.
[4] R. H. Richter, L. Andricek, T. Gebhart, D. Hauff, J. Kemmer, G. Lutz, et al.,
“Strip detector design for ATLAS and HERA-B using two-dimensional
device
simulation," Nucle. Instr. Meth., vol. A 377, pp 412-421, 1996.
[5] G. Lutz, et al., Streifendetektor, Patentoffenlegungsschrift-OS 196 20 081
A1
21.11.97, Munich, 1997
[6] D. C. Christian, et al.. “Development of a pixel readout chip for BTeV,“
Nucle. Instr. Meth., vol. A 335, pp 144, 1999.
[7] M. R. Coluccia, et al., “Characterization of prototype BTeV silicon pixel
sensors before and after irradiation," accepted for publication in IEEE Trans.