S CIP P
Surface Properties of ATLAS07 SSD
• Voltages
V lt
on the
th surface
f
off n-on-p sensors
• Surface Parameters pre-rad
Interstrip R
Interstrip C
• Gamma and Proton Irradiation
• Surface Parameters post-rad
Interstrip R
Interstrip C
Additional info:
• Breakdown (Hara)
• Gluing (Affolder)
H. F.-W. Sadrozinski
SCIPP, UC Santa Cruz
For the ATLAS Upgrade Strip Sensor Collaboration
1
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
Testing of ATLAS07
H. Chen,, J. Kierstead,, Z. Li,, D. Lynn
y
Brookhaven National Laboratory
J.R. Carter, L.B.A. Hommels, D. Robinson
Cavendish Laboratory, University of Cambridge,
K. Jakobs, M. Köhler, U. Parzefall,
Physikalisches Institut,
Institut Universität Freiburg,
Freiburg
R. Bates, C. Buttar, L.Eklund, V. O'Shea,
Department of Physics and Astronomy, University of Glasgow,
Y. Unno, S. Terada, Y. Ikegami, T. Kohriki
Institute of Particle and Nuclear Study, KEK
A Chili
A.
Chilingarov, H.
H Fox,
F
Physics Department, Lancaster University,
A. Affolder, P. P. Allport, G. Casse, M. Wormald,
Oliver Lodge Laboratory, University of Liverpool
V. Cindro, G. Kramberger,
g I. Mandić, M. Mikuž,
Jožef Stefan Institute/Dep. of Physics, University of Ljubljana,
I. Gorelov, M. Hoeferkamp, J. Metcalfe, S. Seidel, K. Toms,
Dep. of Physics and Astronomy, University of New Mexico
Z. Dolezal, P. Kodys,
Charles University in Prague
C. Betancourt, N. Dawson, V. Fadeyev M. Gerling, A. A. Grillo, F. Hurley, P. Maddock, C. Meyer,
H. F.-W. Sadrozinski, S. Sattari, A. Seiden, A. Tolpygo
SCIPP, UC Santa Cruz
R. French, S. Paganis, D. Tsionou,
D
Department
t
t off Physics
Ph i andd Astronomy,
At
The
Th University
U i
it off Sheffield,
Sh ffi ld
R. Caupto, B. DeWilde, R. McCarthy,
Department of Physics and Astronomy, Stony Brook University
K. Hara, H. Hatano, S. Mitsui, M. Yamada,
University of Tsukuba, Institute of Pure and Applied Sciences
M. Miñano, C. García, C. Lacasta, S. Martí i Garcia,
IFIC (Centro Mixto CSIC-UVEG) Valencia
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Y. Unno, 2008/7/10Hartmut Sadrozinski, SCIPP, UC Santa Cruz
ATLAS Upgrade Silicon Strip Detectors (SSD)
KEK
S CIP P
Tsukuba
Liverpool
Lancaster
Glasgow
Sheffield
•
•
•
•
•
Share expertise and cost within the ATLAS groups
Leverage rad-hard experience with p-type SSD
(RD50 KEK) iincluding
(RD50,
l di many manufacturers
f t
Sensor fabrication with the only viable large-volume
and high-quality manufacturer (Hamamatsu HPK)
Produce proto-type test structures
(radiation damage, isolation, ..)
Produce full-size sensors to support
pp module/stave
program (stereo, bonding, gluing, thermal management,..)
Cambridge
QML
Freiburg
MPI
Ljubljana
Prague
Barcelona
Valencia
UC Santa Cruz
BNL
Stony Brook
3
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
ATLAS07
S CIP P
Purpose
– Full size 10 cm x 10 cm for
Modules/Stave
– Set of test structures for technology
d
development
l
t
Delivery target
– Pre-series Feb. 2008
– 2nd Pre-series Sep.
p 2008
– Production Mar 2009
Wafer
– 150 mm p-FZ(100)
– 320 µm thick
thi k
n-strip isolation
– Individual p-stop
– P-spray
p y
– P-spray + p-stop
Stereo
– 40 mrad
– Integrated in half area
– Dead area: 2 mm
Strip segments
– 4 for Short Strips
p
– LS: segments wire-bonded
4
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
Test Structures in ATLAS07
Collecting electrons promises much better charge collection performance in
irradiated silicon sensors than collecting holes (like in the SCT), yet due to an
accumulation layer below the oxide, introduces sensitivity to the surface condition,
e.g.
g one needs to test breakdown,, strip
p isolation and interstrip
p capacitance,
p
, and
potentially gluing on the surface depending on the surface treatment.
There are six versions (“Zones”) of the ATLAS07 mini-SSDs with different strip
isolation schemes.
Wafers # <20 have p-spray, Wafers # >20 have no p-spray.
Pre-series II has improvement in punch-through protection & p-spray dose variation
No p-stop
5
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
Voltages on strip side of n-on-p SSD
Issue:
Economical single-sided processing of n-on-p sensors
To reduce g
guard ring
g current,, P+ implant
p
brings
g the bias voltage
g ((up
p to ~1000V ) to
the strip side. Same principle is used on p-on-n.
VBias
Micron 6” RD50 sensor
ATLAS07 has onlyy one g
guard ring
g
(~ 1mm inactive area, 8 guard rings,
P+ implant at edge)
and larger distances edge – ring:
No access to measure voltages
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
6
Voltages on the Surface of n-on-p SSD
S CIP P
Guard Ring Voltage vs. Bias Voltage Micron 2551-7 13-1 N-on-P FZ
1200
8
7
6
5
4
3
1
1000
Guard Ring [V]
G
800
Micron 2551-4 9-2 N-on-P 5e14 Neutron
1200.00
Ring 1
Ring 3
Ring 3 high
Ring 4 high
Ring 5 high
Ring 6 high
Ring 7 high
Ring 8 high
1000.00
Guard RingVoltag
ge
For low
low-current
current sensors:
outermost guard ring on the SSD face
is at Vdep, e.g. up to 1000V,
((innermost close to bias ring
g=g
ground))
800.00
600.00
400.00
200.00
0.00
600
0
200
400
600
800
1000
1200
1000
1200
Bias Voltage
g [V]
[ ]
400
Micron 2535-9 3-1 P-on-N FZ 1.5e14 pions/cm^2
1000
200
1
3
4
5
6
7
8
900
0
200
400
600
800
1000
1200
Bias [V]
Pre-rad (n-on-p)
and post
post-rad
rad (n-on-p
(n on p & p
p-on-n)
on n)
guard ring voltages
vary from Vbias to GND
Hybrid designers
(and pixel bump-bonders!) be aware!
800
Guard R
Ring Voltage [V]
0
700
600
500
400
300
200
100
0
0
200
400
600
800
Bias Voltage [V]
7
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
Time Variation of Interstrip Resistance Rint
Interstrip R for Zone 3 sensor w23-bz3-p21 (no p-spray): bias ramps up
and down after 3 hours at 200 V
1600
1400
1200
1000
800
600
400
200
fresh
3h bias
0
1
10
Ub ias, V
100
1000
Interstrip R for Zone 3 sensors with p-spray: bias ramps up
and down after a time @ 200 V
2000
For zone 3 sensors (p-stop) the
interstrip resistance Rint does not
depend on bias in the range 10200V and doesn’t change after
sensor remaining at 200V bias
during 3 hours.
Typical Rint
1000 GOhm
GOhm.
i t value is ~1000
1800
1600
1400
1200
1000
800
600
400
w07-bz3-p3 fresh
A Chilingarov,
A.
Chili
L
Lancaster
t U
U.
w07-bz3-p3
07 b 3 3 3h bi
bias
200
w04-bz3-p3 fresh
w04-bz3-p3 10' bias
0
1
10
Ubia s , V
100
1000
8
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
Time Variation of the Interstrip Resistance Rint
S CIP P
Interstrip R for Zone 1 sensors with p-spray: bias ramps up
and down after 3 hours at 200 V
1400
1200
1000
800
600
400
w02-bz1-p19 fresh
w02-bz1-p19 3h bias
200
w04-bz1-p7 fresh
w04-bz1-p7 3h bias
0
1
10
Ubia s , V
100
1000
Interstrip R for Zone 1 sensors without p-spray: bias ramps up
and down after ~3 hours at 200 V
For zone 1 sensors without pspray the Rint behaviour is more
complicated. Nevertheless above
100V bias the Rint remains above
100 GOhm even after 3 hours at
200V bias.
1.0E+03
1.0E+02
fresh senso rs,
ramp up
1.0E+01
after 3h bias,
ramp do wn
1.0E+00
For fresh zone 1 sensors with pspray the
th Rint also
l d
doesn’t
’t d
depend
d
on bias in the range 10-200V but
after 3 hour biasing by 200V it
decreases and becomes slightly
bias dependent with Rint value of
~500 GOhm above 100V bias.
1.0E-01
10E
1
.0E-02
02
1.0E-03
w28-bz1-p19 fresh
w28-bz1-p19 3h bias
1.0E-04
w27-bz1-p7 fresh
w27-bz1-p7 3.5h bias
10E
1
.0E-05
05
1
10
Ubia s , V
100
1000
A Chilingarov,
A.
Chili
L
Lancaster
t U
U.
9
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
Time Variation of Interstrip Capacitance Cint
Interstrip Capacitance to next neighbour pair:
Total capacitance ~ 30% higher
Interstrip C vs. time at 600V (sensors w01-17 have p-spray)
0.64
0.63
Z
Zone
1 (no
( p-stop)
t )
0.62
0.61
0.6
At 600V bias the Cint
further converges with time
to a common value of
~0.61 pF for all sensor
types. Note that more than
2h
hours may b
be needed
d d ffor
the Cint stabilisation. This
phenomenon was
extensively studied earlier
for the SCT sensors
0.59
Zone 3 (with p-stop)
0.58
w28-bz1-p19
w27-bz1-p7
w02-bz1-p19
w04-bz1-p7
w23-bz3-p21
w04 bz3 p3
w04-bz3-p3
w07-bz3-p3
0.57
0 56
0.56
The present measurements were made at
~+22oC and 35-45%
relative humidity
humidity.
0.55
0
40
80
Time, min
120
160
200
A. Chilingarov, Lancaster U.
10
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
Cint Comparison with the SCT sensors
S CIP P
Interstrip capacitance per unit length versus pitch
0.85
SCT wedges/Lancs
Atlas07 minis
SCT MPI/NIKHEF
0.01 pF abs.uncertainty
Linear (SCT wedges/Lancs)
0.80
NIKHEF
Cis, pF/cm
0.75
0.70
y = -0.007x + 1.2664
2
R = 0.996
0.65
w12
w32
w31 w21
w22
0.60
60
65
70
75
80
85
90
Average pitch, um
The observed Cint/L agrees well with the data measured for the SCT sensors.
An absolute LCR meter uncertainty of 0.01 pF is also shown in the error.
Thus
us the
e Cis in ATLAS07
S0 miniss can
ca be regarded
ega ded as ssimply
p y geo
geometrical
e ca o
one.
e
A. Chilingarov, Lancaster U.
11
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
Gamma Irradiation at BNL
S CIP P
• 8 ATLAS07 mini-SSD were irradiated at BNL to 1 Mrad, 5 biased, 3 unbiased.
• Breakdown behavior post-rad improved slowly, cleaning and N2 flow helped (?)
• All showed sensitivity to surface conditions (biased ones to a lesser extend)
• Suspect charge-up of surfaces as main cause, and that slow improvement of i-V on
bonded sensors is d
due
e to slo
slow bleeding of ssurface
rface charges to gro
grounded
nded cond
conductors
ctors
• Since charge-up can lead to breakdown through oxide, need to limit dose rate for
future gamma irradiations
All Al readout traces are bonded out together
During
D
i iirradiation
di ti bi
bias th
the metal
t l strips
t i tto +0.5
05V
wrt to bias ring. Bias back plane to 200V.
RT annealing for about 10 days.
Take i-V with the Al either “grounded” or
“floating”
Measure
Isolation, Interstrip C, Breakdown
12
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
i-V post Gamma Irradiation
S CIP P
HPK1 W29 Z4 P4 1 MRad Gamma Irradiated IV
2.E-06
3-18 pre-rad no bond
Post-rad 6-24 grounded
Leakag
ge Current (A
A)
Post-rad 6-24 floating
Post-rad 8-20 No Nitro
Post-rad8-20 with nitro
1.E-06
0.E+00
0
100
200
300
400
500
Bi V
Bias
Voltage
lt
(V)
600
700
800
N2 flow needed? Bonded SSD improve slowly:
Charge bleed rate τ = ρ∗ε
13
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
Proton Irradiation of ATLAS07 mini SSD
“Low”-fluence proton runs
Analysis of gamma irradiations has proven to be difficult, presumably related
t the
to
th very high
hi h resistivity
i ti it off th
the passivation
i ti ttrapping
i charges
h
on th
the surface
f
during the high dose-rate irradation. Slow improvement over time has been
observed.
Irradiation of HPK ATLAS07 sensors at Los Alamos Nat
Nat. Lab
Lab. (800 MeV
Protons) organized by U. of New Mexico was stopped at 1.5*1013 neq/cm2
after an operator fault, giving us a sample with low fluence protons. The total
dose is ~ 500 kRad
kRad, an ideal place to study Rint, Cint and breakdown behavior,
behavior
independent of large bulk currents.
Since the charge collection studies after high hadron fluences show us a very
stable picture of the sensors (independent of the surface details), the
decision which technology to use for strip isolation will be influenced by
these low-fluence runs.
14
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
Strip Isolation post Proton/Gamma Irradiation
S CIP P
Rint [Gohm
m]
100
Z1
Z1
Z3
Z3
Z4
Z4
Z5
Z6
p-spray
p-spray2
p-spray
p-stop
p-spray
p-stop
p-spray
p
p y
p-spray
10
Rint Pre- and Post 1 Mrad Gamma
1.E+06
1.E+05
Rint (MOhm)
Rint ATLAS07 Pre-rad
1000
Pre rad W06-BZ4-p-spray
Pre-rad
W06 BZ4 p spray
1.E+04
1 MRad W06-BZ4-p-spray
1 Mrad N2 Flow W06-BZ4-p-spray
1Mard W29 BZ4 no p-spray
1.E+03
1.E+02
1
0
200
400
600
800
1000
Vbias
1.E+01
0
100
200
300
400
Vbias [V]
Gamma irradiation
reduces Interstrip Isolation:
Post-rad
Post
rad Rint =50’s
50 s -100’s
100 s of MΩ
Interstrip
p Resistance
[M
M ]
Pre-rad Rint =10’s -100’s of GΩ
RINT ATLAS07 after 1.5e14 protons (LANL)
1.E+04
P-stop
1.E+03
1.E+02
Proton irradiation:
P-stop
s op see
seems
s to
op
provide
o de more
oe
robust isolation after proton irradiation
P-spray only
1.E+01
1.E
01
0
50
100
150
200
250
Bias Voltage [V]
W16-BZ1-P19
W17-BZ1-P7
W16-BZ3-P15
W13-BZ5-P23
W29-BZ6-P12
W15-BZ6-P12
W28-BZ6-P12
W14-BZ6-P12
300
350
15
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
Inter-strip Capacitance post-rad Protons
S CIP P
)Z3, p-spray (1MHz
)Z1, p-spray (1MHz
Cint ATLAS07 pre-rad
0.9
Cint [pF]
08
0.8
0.7
0.6
0.5
0.4
0
100
200
300
400
Bias Voltage [V]
500
600
700
W16-BZ1-P19
W16-BZ3-P15
W13-BZ5-P23
W15-BZ6-P12
W14-BZ6-P12
W28-BZ6-P12
W29-BZ6-P12
CINT after 1.5e14 protons (LANL)
Inte
erstrip Capacitaance [fF]
1
1000
900
800
700
Wide pitch
600
500
0
50
100
150
200
Bias Voltage [V]
250
300
350
Interstrip capacitance to next neighbor pair (total ~30% higher): close to 1 pF/cm.
Pre-rad:
same value for p-spray with and without implants, “No” bias dependence for Zone 1
and Zone 3
Post-rad:
Somewhat higher after low-fluence proton irradiation. Correlated with reduced Rint?
No dependence on ssurface
rface treatment (p
(p-stop
stop vs.
s p-spray)
p spra )
Important for long strips: can expect C ~ 13 pf !
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
16
Summary of Gluing Results at Liverpool
S CIP P
Breakdown Volta
age
1200
1000
800
600
400
200
0
With Araldite 2011
(HPK)
With kapton
(HPK)
Pre-gluing
With Ceramic
(HPK)
Post-gluing
With Araldite 2011
(non-passivated
MICRON)
With Araldite
(Passivated
MICRON)
With Epolite 5313
(HPK)
Araldite 2011 interacts poorly with ATLAS07 sensors, but no ill-effects seen with
passivated/unpassivated Micron sensors or p-on-n HPK sensors.
ATLAS07 with Epolite 5313 (electronics grade epoxy) has good performance but
some surface sensitivity seen. Evidence so far suggests that issues might be
different than with Araldite
Araldite.
Irradiations are underway.
A. Affolder, Liverpool U.
17
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
IV-CV curves: Epolite ok, Araldite has Vbd=250V
Epolite: W31-BZ3-P13
Epolite: W31-BZ3-P13
Araldite: W31-BZ3-P3
Araldite: W31-BZ3-P3
S. Paganis, Sheffield U.
18
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
Rinterstrip Measurements at Vbias=200Volts
Unglued: W35-BZ3-P15
Rint=239 GOhms
Epolite: W31-BZ3-P13
Rint=328 GOhms
Araldite: W17-BZ3-P6
The Araldite glue presents low Vbd=300V
and also low Rint << Rint-epolite.
More measurements and tests will follow.
f
Rint= 6.3 GOhms
S. Paganis, Sheffield U.
19
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
•
•
•
•
•
•
ATLAS07 shows mature performance
Inter-strip isolation and capacitance ~ same for all strip isolation schemes
Gamma irradiated need much training/N2
Testing of surface properties post-rad reveals differences due to strip isolation
(p-spray only is inferior to p-stop)
Sensors are stable after low-fluence proton irradiation
– Inter-strip isolation reduced post-rad, p-spray worse than p-stop
– Inter-strip capacitance slightly increased after proton irradiation
Gluing with electronics grade epoxy is shown to work pre-rad, irradiations
underway
HPK n-on-p FZ Median Charge
Hadron irradiated sensors
25000
HPK n-on-p FZ pre-rad
show good signal-to-noise ratio :
20000
HPK n-on-p FZ proton 1.3e15 neq/cm^2
robust n-on-p designs by HPK:
Med
d. Q [e]
•
Conclusions
15000
10000
5000
0
0
200
400
600
800
1000
1200
Bias Voltage [V]
20
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz
S CIP P
SSD Performance Specifications
21
ATLAS Upgrade Tracking Workshop NIKHEF 11/5//08
Hartmut Sadrozinski, SCIPP, UC Santa Cruz