3D Silicon Detectors for LHC Upgrades 3D Silicon Detectors for LHC Upgrades 2009 Cinzia Da Viá, The University of Manchester For the 3DATLAS project S- 21 st January 2 •Introduction 3D ili i l d For the 3DATLAS project ster-SLAC –ADS •3D silicon pixel detectors: -Radiation Hardness t d dissi ti ersity of Manche -current and power dissipation -C, noise, timewalk and overdrive with AtlasFE-I3 -module design plans for the IBL for Atlas a Via’-The Unive •Status of fabrication facilities •Conclusions -Timescale and plans for 2009 Cinzia D Conclusions -Timescale and plans for 2009 C. Kenney, S. Parker (MBC, Hawaii), J. Hasi, S. Watts, J. Pater, J. Freestone, S. Kolya, S.Snow; R. Thompson (Manchester), M. Mathes M. Cristinziani, N. Wermes (Bonn), S. Stapnes, O. Rohne, E. Bolle (Oslo),G. Darbo, R. Beccherle (Genova), S. Pospisil, V. Linhart, T. Slaviceck (Praha), K. Einsweiler, M. Garcia-Sciveres (LBL) A. Kok, T-E Hansen (Sintef) – More generally the ATLAS-3D project team
57
Embed
3D Silicon Detectors for LHC Upgrades - Stanford University€¦ · ATLAS Pixel ATLAS SCT - barrel hadrons Φ eq (microstrip detectors) [M.Moll, simplified, scaled from ATLAS TDR]
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
3D Silicon Detectors for LHC Upgrades3D Silicon Detectors for LHC Upgrades20
09pgpgCinzia Da Viá,
The University of ManchesterFor the 3DATLAS project
S-21
stJa
nuar
y 2
•Introduction
3D ili i l d
For the 3DATLAS project
ster
-SLA
C –A
DS •3D silicon pixel detectors:
-Radiation Hardness t d dissi ti
ersi
ty o
f M
anch
e -current and power dissipation-C, noise, timewalk and overdrive with AtlasFE-I3-module design plans for the IBL for Atlas
aVi
a’-Th
e U
nive
•Status of fabrication facilities
•Conclusions -Timescale and plans for 2009
Cinz
iaD Conclusions -Timescale and plans for 2009
C. Kenney, S. Parker (MBC, Hawaii), J. Hasi, S. Watts, J. Pater, J. Freestone, S. Kolya, S.Snow; R. Thompson (Manchester), M. Mathes M. Cristinziani, N. Wermes (Bonn), S. Stapnes, O. Rohne, E. Bolle (Oslo),G. Darbo, R. Beccherle (Genova), S. Pospisil, V. Linhart, T. Slaviceck (Praha), K. Einsweiler, M. Garcia-Sciveres (LBL) A. Kok, T-E Hansen (Sintef)– More generally the ATLAS-3D project team
Introduction20
09S-
21st
Janu
ary
2st
er-S
LAC
–AD
Ser
sity
of
Man
che
aVi
a’-Th
e U
nive
Cinz
iaD
Track reconstruction at different luminositiesTrack reconstruction at different luminositiesd j f k kd j f k k
2009
need to reject fake tracksneed to reject fake tracksS-
21st
Janu
ary
2st
er-S
LAC
–AD
Ser
sity
of
Man
che
aVi
a’-Th
e U
nive
Cinz
iaD
CMS simulation-From Collins, VPI 2007
From J. Nash/IC SLHC R&D, Cern 9 April 200820
09S-
21st
Janu
ary
2st
er-S
LAC
–AD
Ser
sity
of
Man
che
aVi
a’-Th
e U
nive
Cinz
iaD
The LHC and SLHC challenge20
09
at full luminosity L=1034 cm-2 s-1:• ~23 overlapping interactions in each bunch crossing every 25
g
1016
SUPER - LHC (5 years, 2500 fb-1)Pixel (?) Ministrip (?)
Some history..3D silicon sensors were originally proposed by Sherwood ParkerSome history..3D silicon sensors were originally proposed by Sherwood Parkerand fabricated at Stanford by C. Kenney (MBC) who proposed “active edges”. and fabricated at Stanford by C. Kenney (MBC) who proposed “active edges”. And since 2000 by J. And since 2000 by J. HasiHasi (Manchester) (Manchester)
detector bulk instead of being implanted on the Wafer's surface.
h l f ll b
ersi
ty o
f M
anch
e The edge is an electrode (following an idea by C. Kenney). Dead volume at the Edge < 5 microns! Essential for forward physics experiments and material budget
aVi
a’-Th
e U
nive
2006 3DC collaboration was formed
1. NIMA 395 (1997) 328 2. IEEE Trans Nucl Sci 46 (1999) 12243. IEEE Trans Nucl Sci 48 (2001) 1894 IEEE Trans Nucl Sci 48 (2001) 1629
Cinz
iaD Core Members: Brunel/Manchester, Hawaii
Oslo University, Sintef and Stanford (MBC)4. IEEE Trans Nucl Sci 48 (2001) 1629 5. IEEE Trans Nucl Sci 48 (2001) 2405 6. Proc. SPIE 4784 (2002)3657. CERN Courier, Vol 43, Jan 2003, pp 23-268. NIM A 509 (2003) 86-919. NIMA 524 (2004) 236-244
2007 3D Atlas R&D approved 10. NIM A 549 (2005) 12211. NIM A 560 (2006) 12712. NIM A 565 (2006) 27213. IEEE TNS 53 (2006) 167614. NIM A 587(2008) 243-249
ProcessingProcessing20
09
ProcessingProcessingS-
21st
Janu
ary
2
Currently performed at the Stanford-Nanofabrication-
ster
-SLA
C –A
DS
Facility (CIS) Stanford USA
C. Kenney (MBC), J. Hasi (Manchester)
ersi
ty o
f M
anch
e (Manchester)
1000m2
aVi
a’-Th
e U
nive
Manchester HAWAII
1000 1000m2
Cinz
iaD Manchester – HAWAII –
STANFORD (Molecular Biology Consortium)
SINTEF, Oslo, Praha form the 3DCConsortium, since February 2006
Irradiation and measurements performed in PragueC. Da Viá, T. Slaviceck, V. Linhart, P. Bem, S. Parker, S. Pospisil, S. Watts (process J. Hasi, C. Kenney)
Irradiation and measurements performed in PragueC. Da Viá, T. Slaviceck, V. Linhart, P. Bem, S. Parker, S. Pospisil, S. Watts (process J. Hasi, C. Kenney)
2E50 μm
S-21
stJa
nuar
y 2
0
20
40
60
Sign
al e
ffici
en
C. Da Viá July 070
40
80
Sign
al A
mpl
it
C. Da Viá July 07
2E9000e-
Vb~130V
45%
threshold
400 μmμ
n
ster
-SLA
C –A
DS 0
0 2 1015 4 1015 6 1015 8 1015 1 1016
Fluence [n/cm2]
100
120
%]
120
140
1603E-NI7.55e142.00e158.81e15[m
V]
00 50 100 150 200
Bias Voltage [V]
3E 51%3E
50
p 103μmVfd ~20V
ersi
ty o
f M
anch
e
20
40
60
80
Sign
al e
ffici
ency
[%
3E
40
60
80
100
120 8.81e15
Sign
al A
mpl
itude
[3E10200e-
Vb~112V
51%400 μm
50 μm
pn
aVi
a’-Th
e U
nive
120
0
20
0 2 1015 4 1015 6 1015 8 1015 1 1016
S
Fluence [n/cm2]
C. Da Viá July 07
140
0
20
0 50 100 150 200Bias Voltage [V]
C. Da Viá July 07 threshold
4E
p71 μm
Vfd ~8V
Cinz
iaD
60
80
100
120
effic
ienc
y [%
]
4E
60
80
100
120
1404E NI7.55e142.00e15 8.81e15
Am
plitu
de [m
V]4E13200e-
Vb~94V
E
400 μm50 μm
0
20
40
0 2 1015 4 1015 6 1015 8 1015 1 1016
Sign
al e
Fluence [n/cm2]
C. Da Viá et al. July 070
20
40
0 50 100 150 200
Sign
al A
Bias Voltage [V]
C.DaVia July 07
Vb 94V66%
thresholdp
n
Vfd ~5V
Signal Efficiency and Signal Charge in 3DSignal Efficiency and Signal Charge in 3D20
09g y g gg y g g
structuresstructuresS-
21st
Janu
ary
2
particle3Dn+p+ n+n+n+ p+ p+p+ n+
⎤⎡λ
)exp(λx
dtdx
dxdVq
dtdS W −=
ster
-SLA
C –A
DS
--
--
++++
++
++
L Δ
⎥⎦⎤
⎢⎣⎡ −−= )exp(1
λλ xL
S
)(22 λλλ LSE ⎟
⎞⎜⎛
⎟⎞
⎜⎛
ersi
ty o
f M
anch
e
d
~50 μm
--
--
++ )exp(λ
λλλ LLLL
SE −⎟⎠⎞
⎜⎝⎛+⎟
⎠⎞
⎜⎝⎛−=
aVi
a’-Th
e U
nive ~ 500 μmActive edge ~4μm
Φ+= KL
SEτ6.01
1
Cinz
iaD
----
++
++++
L=Δi
Dv
L=Inter electrode distanceΔ thi k
PLANARn+
Trapping times from Kramberger et al. NIMA 481 (2002) 100 NIM A 501(2003) 138 (Vertex 2001)
Δ=thickness
Signal efficiency and signal chargeSignal efficiency and signal charge20
09
[9] C. Da Via et al.”, (NIMA-D-08-00587)[10] G. Kramberger at al., Nucl. Instr. Meths. A 554 (2005) 212-219[11] G. Kramberger, Workshop on Defect Analysis in Silicon Detectors, Hamburg, August2006. http://wwwiexp.desy.de/seminare/defect.analysis.workshop.august.2006.html[12] G. Casse et al., Nucl. Instr. Meths. A (2004) 362-365[14] T. Rohe et al. Nucl. Instr. Meths. A 552 (2005) 232-238
S-21
stJa
nuar
y 2
100 56μm 3D - 4E [9]71μm 3D - 3E [9]103μm 3D 2E [9]
25000.0103μm 3D - 2E [9]71 3D 3E [9]
[ ] ( )[16] F. Lemeilleur et al., Nucl. Instr. Meths. A 360 (1995) 438-444
ster
-SLA
C –A
DS
60
80
103μm 3D - 2E [9]75μm epi [11]150μm epi [11]285μm n+n pixels [14]285μm n+p strips[12]300μm p+n strips [16]
ency
[%]
15000.0
20000.0
71μm 3D - 3E [9]56μm 3D - 4E [9]50μm epi [10]75μm epi [11]150μm epi [11]285μm n+p strips[12]285μm n+n pixels [14]
ge [e
- ]
71μm 3D
ersi
ty o
f M
anch
e
20
40
Sign
al E
ffici
e
5000 0
10000.0
Sign
al C
harg
75 m
aVi
a’-Th
e U
nive
0
20
0 2 1015 4 1015 6 1015 8 1015 1 1016
Fluence [1MeV Equivalent n/cm2]
C.DaVia; S. Watts Aug08
0.0
5000.0
0 2 1015 4 1015 6 1015 8 1015 1 1016
Fluence [1 MeV equivelent n/cm-2]
C. Da Via S. Watts April 08
75μmepi
Cinz
iaD Fluence [1 MeV equivelent n/cm ]
S l 80 (λ/L) Δ 80λ 80 30 2400
Example at 1016 ncm2
SMIP planar ~ 80 (λ/L) x Δ ~ 80λ ~ 80x30 ~ 2400e-
SMIP 3D ~ 80λ x (Δ/L) ~ 2400 x 210/(71-22electrode implant) ~ 10290e-
The geometrical dependence of the signal The geometrical dependence of the signal efficiency on the interefficiency on the inter--electrode spacing Lelectrode spacing L
2009 100
efficiency on the interefficiency on the inter electrode spacing Lelectrode spacing L
1
S-21
stJa
nuar
y 2
80
1E153.5E158.8E15
y [%
] LSE 1
∝
ster
-SLA
C –A
DS
40
60
l Effi
cien
cy
L=Δ~300μm
ersi
ty o
f M
anch
e
20Sign
al
L=Δ~50μm epi
aVi
a’-Th
e U
nive
00 50 100 150 200 250 300
Inter-electrode spacing [μm]
C. Da Via' April 08
L=~50μm 3D
Cinz
iaD
[3D- 56-71-103 μm] C. Da Via et al.”, (NIMA-D-08-00587)[epi 25 - 50μm ] G. Kramberger at al., Nucl. Instr. Meths. A 554 (2005) 212-219[epi 75 μm] G. Kramberger, Workshop on Defect Analysis in Silicon Detectors, Hamburg, August2006 http://wwwiexp desy de/seminare/defect analysis workshop august 2006 html2006. http://wwwiexp.desy.de/seminare/defect.analysis.workshop.august.2006.html[planar 285μm] G. Casse et al., Nucl. Instr. Meths. A (2004) 362-365[planar 285μm] T. Rohe et al. Nucl. Instr. Meths. A 552 (2005) 232-238[F planar 300μm]. Lemeilleur et al., Nucl. Instr. Meths. A 360 (1995) 438-444
Test beam at the CERN SPSTest beam at the CERN SPS20
09
Bonn telescope
S-21
stJa
nuar
y 2
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
ea
Via’-
The
Uni
veCi
nzia
D
Noise characterisation using the Noise characterisation using the ATLAS FEATLAS FE--I3 chip I3 chip –– bumpbump--bonding IZM through Bonn Universitybonding IZM through Bonn University
2009
Tests by E. Bolle, O. Rohne (Oslo)
S-21
stJa
nuar
y 2
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
e
Leakage currents 2
aVi
a’-Th
e U
nive
L a ag curr nts before irradiation~325 pA/pixel1.5
2E C (fF)3E C (fF)4E C (fF)
r mic
ron
(fF)
Cinz
iaD
Noise vsBias voltage0.5
1
Cap
acita
nce
pe simulation
10 20 30 40 50 60
Voltage (V)
Noise vscapacitance
Reproducibility -1 Leakage currents after bump-bonding20
09
•1uA correspond to 350pA/pixel
S-21
stJa
nuar
y 2
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
e
3E-F
aVi
a’-Th
e U
nive
Noisy pixelsPossibly bump-bonding
Cinz
iaD
2E A
Possibly bump bonding
2E-A4E-B
Tracking efficiency June 06Tracking efficiency June 06bumpbump bonding at IZM (Bonn)bonding at IZM (Bonn)
2009
bumpbump--bonding at IZM (Bonn)bonding at IZM (Bonn)
M. Mathes1, C. DaVia2, J. Hasi2, S. Parker3, M. Ruspa4,L. Reuen1, J. Velthuis1, S. Watts2, M. Cristinziani1, K.Ei s il 4 M G i S i s4 K K 5 N W m s1 Data analyisis and silulation by M Mathes M Cristinziani (Bonn)
S-21
stJa
nuar
y 2 Einsweiler4, M. Gracia-Sciveres4,K. Kenney5, N. Wermes1
1Bonn, Germany2Manchester University, UK3University of Hawaii, USA4LBL, Berkeley, USA5Molecular Biology Consortium, Stanford, USA
3E 3200 e-
threshold
Data analyisis and silulation by M. Mathes, M. Cristinziani (Bonn)S. Watts (Manchster)
data
ster
-SLA
C –A
DS
0obeam
data
ersi
ty o
f M
anch
e simulation
ε = (95.9 ± 0.1) %
aVi
a’-Th
e U
nive
bdata
ε (95.9 ± 0.1) %
Cinz
iaD 15obeam
simulation
ε = (99.9 ± 0.1) %
Tracking performance 3E configuration 20
09g p g
Correlation with telescope planes
S-21
stJa
nuar
y 2
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
e
50μm direction:
aVi
a’-Th
e U
nive 50μm direction:
width (49.4±0.1)μm, sigma (4.8±0.1)μm
400μm direction:
Cinz
iaD 400μm direction
width (398.0±0.3)μm, sigma (6.4±0.2)μmM. Mathes1, C. DaVia2, J. Hasi2, S. Parker3, M. Ruspa4,L. Reuen1, J. Velthuis1, S. Watts2, M. Cristinziani1, K.Einsweiler4, M. Gracia-Sciveres4,K. Kenney5, N. Wermes11Bonn, Germany2Manchester University, UK3University of Hawaii, USA4LBL, Berkeley, USA5Molecular Biology Consortium, Stanford, USA
4E electrode angular response 4E electrode angular response –– preliminarypreliminary20
Stanford 3E-device bonded to Atlas FE-I3 front end
ster
-SLA
C –A
DS
Irradiated at PS with 24GeV protons under 40V biasTotal fluence 9.8x1014pcm2
Partially annealed during testing
ersi
ty o
f M
anch
e
y g g
Post irradiation:
aVi
a’-Th
e U
nive
Bond wires destroyed by corrosionSeveral repair attempt including replacement on new boardBias voltage limited to ~3V by hot spot (caused by thermal choc)
Cinz
iaD Bias voltage limited to ~3V by hot spot (caused by thermal choc)
Cooling during test beam limited to 0oC
Test beam dataTest beam dataIrradiated sampleIrradiated sample
Comparison to same typeIrradiated with neutrons
2009
140
1603E-NI7 55e14]
Irradiated sampleIrradiated sample Irradiated with neutronsTested with IR light
S-21
stJa
nuar
y 2
80
100
120
140 7.55e142.00e158.81e15
mpl
itude
[mV]
ster
-SLA
C –A
DS
Hot spot 20
40
60
Sign
al A
m
C Da Viá July 0720%
ersi
ty o
f M
anch
e Hot spot 00 50 100 150 200
Bias Voltage [V]
C. Da Viá July 07
3V bias voltage!
20%
aVi
a’-Th
e U
nive 3V bias voltage!
T=0oC
TOT overall
Cinz
iaD
50%TOT overall
TOT
1x1015pcm-2
Power dissipationPower dissipation Processing: J. Hasi Manchesteri, C. Kenney, MBCMeasurements: M.Hoeferkamp UNMT. Slavicek PragueAnalysis and simulation C.DaVia, S.Watts, Manchester20
09 200
4E n1
S-21
stJa
nuar
y 2
140
160
1804E n2E n3E n3E p
g Vo
ltage
[V]
0.1
4E at -10C neutrons3E at -10C neutrons2E at-10C neutronssimulation alpha 53E at -10C protonssimulation alpha 22 ]
ster
-SLA
C –A
DS
80
100
120
y = 88.075 + 7.4264e-15x R= 0.84247
y = 82.219 + 4.5391e-15x R= 0.99043
Ope
ratin
g
C DaVia and S J Watts July 08
0.01
s u at o a p a
wer
cm
-2 [
W/c
m2
ersi
ty o
f M
anch
e
1.0 10-4
protons
600 5x1015 1x1016 1.5x1016 2x1016 2.5x1016
Equivalent neutron fluence [n/cm2]
C. DaVia and S. J. Watts July 08
0.001
Pow
aVi
a’-Th
e U
nive
6.0 10-5
8.0 10-5
pIfd 20C
Ifd 0C
Ifd -10C
Cur
rent
[A]
0.0001
1014 1015 1016 1017
Fluence [ncm-2]
C. DaVia et al. Oct. 08
Cinz
iaD
2.0 10-5
4.0 10-5
Leak
age
C Fluence [ncm ]
Power/cm2 at -10oC
At 1.0x1015 ncm-2 ~ 3 mWcm-2
0.00 5 1015 1 1016 1.5 1016 2 1016 2.5 1016
fluence ncm-2
C. DaVia, M. Hoeferkamp July 08At 5.0x1015 ncm-2 ~ 33 mWcm-2
At 1.0x1016ncm-2 ~ 120 mWcm-2
At 2.1x1016 ncm-2 ~ 443mWcm-2
Timewalk and overdrive with FETimewalk and overdrive with FE--I3+3DI3+3D20
09
3500
4000
3D-4E
3D Overdrive measurements: Bolle, Heinsweiler, Rohne
F t ti lk ( ll
S-21
stJa
nuar
y 2
2000
2500
3000
3500IL68 IP68 standard
IL96 IP128 20%
IL64 IP192 40%
driv
e [e
-]
3D 2E
3D-3E
3D 4E
3D 3E
3D-4E
Faster timewalk (smalleroverdrive) by increasing FE current
ster
-SLA
C –A
DS
500
1000
1500
2000
Ove
rd
diamond
planar Si
3D-2E
3D-2E
3D-3E
3D-2E
3D-3E
3D-4E • Threshold triggered system• Difference between t and t gives the time walk• New bunch crossing every 25ns• Need to cross threshold within 20ns to be accepted
ersi
ty o
f M
anch
e
0
500
0 50 100 150 200 250 300 350
Noise [e-]
C. Da Via Oct.08
3500
p• How much overdrive is needed to cross threshold within
ATLAS groups participating are from Bergen University, Bonn University, Freiburg University, University of Genova, Glasgow University, the
ersi
ty o
f M
anch
e University of Hawaii, Lawrence Berkeley National Laboratory, Manchester University, the University of New Mexico, Oslo University and the Czech Technical University. At present C. Kenney (Molecular Biology Consortium) and Jasmine Hasi (Manchester University), working at the Stanford Nanofabrication Facility, made all of the Full-3D sensors. For this project industrial companies will join the above mentioned institutions to study the feasibility of a large volume production in time for the upgrade.
aVi
a’-Th
e U
nive
for the upgrade. They are: CNM/Valencia Spain, ,ICEMos, Ireland, IRST Italy and SINTEF Norway.
Topic(s) and goal(s) of the R&D proposal
The primary goal is the development fabrication characterization and testing with and without the front end readout
Cinz
iaD The primary goal is the development, fabrication, characterization, and testing, with and without the front-end readout
chip, of Full-3D – active-edge and Mod-3D silicon pixel sensors of extreme radiation hardness and high speed for the the Super-LHC ATLAS upgrade and, possibly, the ATLAS B-layer replacement. A secondary goal is to start design work for a reduced material B-layer detector module using these sensors.FP420 is used as a test bench for the technology
Stanford Full-3D Yes. 2006 2006/2007 Smallest hole diameter is around 17microns with 250 micron thick wafer.Best aspect ratio around 15:1.
ersi
ty o
f M
anch
e SINTEF Full-3D Yes. 2008 Radioactive sources Has new and excellent DRIE machines(Alcatel) with expected aspect rationof 20:1 or better. Long track record ofworking with HEP producing planardetectors. Does not have all polysiliconfilling equipment (n and p doping) in-house Investigating suitable suppliers
aVi
a’-Th
e U
nive house. Investigating suitable suppliers.
Holes filled at Stanford at present.
CNM Double sided 3D (no active edge) Wafer ready Radioactive sources Use Alcatel etcher and reachedaspect ratio of 25:1. Holes are notfully filled. Plans to move to full3D
Cinz
iaD 3D.
IRST Double sided 3D (no active edge) Yes, 2008 2006 using microstrip readout
Use Alcatel etcher. Holes are not fullyfilled but plans to solve the problem.Plans to move to a full 3D design withactive edges in near future.
SINTEF/3DCSINTEF/3DCMounting Bonn - Bump-bonding IZM/BonnMeasurements O. Rohne, H Gjesdal
FE-I320
09S-
21st
Janu
ary
2st
er-S
LAC
–AD
Ser
sity
of
Man
che
Full3D with active edges10 assemblies bump-bondedn-on-n structuresHigh leakage after bump bondingdue to a missing oxide layer
241Am
Alcatel
aVi
a’-Th
e U
nive due to a missing oxide layer
Particles visible
FE-I4Process
etcher
Cinz
iaD ProcessJust started
Sintef layout mask containing FE-I4 chips20
09
Atlas FE I4 chips
Ready by summer 09
S-21
stJa
nuar
y 2
1E, 2E, 3E, 4E, 5E test structuresAtlas FE- I4 chips 2E configuration
ster
-SLA
C –A
DS
Medipix chips CMS type structures, 14.6 % of wafer area
ersi
ty o
f M
anch
ea
Via’-
The
Uni
veCi
nzia
D
ATLAS 1E, 2E, 3E, 4E, 5E chips
FBKFBK--IRST TrentoIRST Trento--ItalyItaly20
09 • 3D-DDTC concept
S-21
stJa
nuar
y 2
(Double-side Double Type Column)
• Expected to have performance comparable
to standard 3D detectors (if d is small enough)
ster
-SLA
C –A
DS
Batch DDTC 1 DDTC 2
• 2 batches under fabrication
ATLAS pixel, single-chip
ersi
ty o
f M
anch
e Batch DDTC 1 DDTC 2
Substrate type n-type p-type
Subst. thickness (μm) 300 205 – 255
Column depth (μm) 200 180 – 200
(2, 3, 4 or 7 columns/pixel)
aVi
a’-Th
e U
nive
Column depth (μm) 200(not optimized)
180 200(optimized)
Strip design and pitch (μm)
AC/DC coupled, 80 – 100
AC/DC coupled, 80 – 100
Cinz
iaD
Pixel design ALICE, MEDIPIX ATLAS, CMS
Due by August September 2007Due by August 2007
September 2007
FBK/IRST FBK/IRST ––GenovaGenova--CERNCERN FE-I3
Measurements. A. La Rosa (Cern)And G. Darbo (Genova)
2009
FBK/IRST FBK/IRST GenovaGenova CERNCERN FE I3S-
21st
Janu
ary
2st
er-S
LAC
–AD
S
needle70V
80V1μA
ersi
ty o
f M
anch
e 1μA
aVi
a’-Th
e U
nive
Cinz
iaD
Preliminary TOT needs calibration
FBK/IRST FBK/IRST ––GenovaGenova--CERNCERN20
09FBK/IRST FBK/IRST GenovaGenova CERNCERN
Measurements. A. La Rosa (Cern)And G. Darbo (Genova)
S-21
stJa
nuar
y 2
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
ea
Via’-
The
Uni
veCi
nzia
D
CNM/ Glasgow : single column 3D sensors fabricated and future CNM/ Glasgow : single column 3D sensors fabricated and future production of partialproduction of partial--and fulland full--double column designdouble column design
Celeste Fleta Richard Bates, Chris Parkes, David Pennicard – University of Glasgow20
09
Manuel Lozano, Giulio Pellegrini – CNM (Barcelona)
2008 batch completedSuccessfully and beingBump-bonded
04/09 full3D double sidedProcessTo start in fall09 full3D withActive edges
Test samples irradiated
Evolution of IB layer design20
09 Current
IBLEvolution of IB-layer designS-
21st
Janu
ary
2 Current
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
ea
Via’-
The
Uni
ve
B-layer envelope
Cinz
iaD
y p45.5mm
From Neal Hartman, Nikef ATUW, 3 Nov. 08
3D active edges 3D active edges E x B = 0 at 0 angle in solenoidfield - Lorentz angle
In 3D20
09f Lor ntz ang
S-21
stJa
nuar
y 2
NOW Big chips
Active edges
Effective Si thickness
680μ 605μ 515μC b
ster
-SLA
C –A
DS
~1.5 mm
thickness Can be pushed to 450u
ersi
ty o
f M
anch
e .5 mm
Present planar 2.2%Xo
aVi
a’-Th
e U
nive
11oWith active edgeCould aim at 1.6%X0
Cinz
iaD
~4μmModule looks the same, but there is no dead margin on the sensor perimeter, allowing less overlap Still 2 pixel radial overlap in phi This could be a good Single modules=>allowing less overlap. Still 2 pixel radial overlap in phi. This could be a good baseline starting point. From M. Garcia-Scieveres talkPresented at the ID ATLAS Upgrade Workshop. Liverpool 6-8 December 06
(Hasi, Kenney)-FEchips LBL, mounting Genova,Testing Oslo.
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
ea
Via’-
The
Uni
veCi
nzia
D
M D d FE
Alignment system in Genova
SCM using 3D and FE-I3 3D stave
ATLASFP – 2011-2012Forward Detectors = use LHC beam-line as a spectrometerP t n n l ss sults in p t n t j ct h i nt l d p tu
2009
Proton energy loss results in proton trajectory horizontal departureS-
21st
Janu
ary
2
IP1
ster
-SLA
C –A
DS
y(m
m)
220m
���� y(m
m)
420m
ersi
ty o
f M
anch
e
x(mm)x(mm)
���
V. Avati/Totem
V. Avati/Totem
aVi
a’-Th
e U
nive
Cinz
iaD
Tracking Detectors: at 420m 25x5mmTracking Detectors: at 420m 25x5mm2220
09
3D silicon with active edges
50 400
S-21
stJa
nuar
y 2 50μmx400μm
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
ea
Via’-
The
Uni
veCi
nzia
D
x,y50 m 14.4 m
12μσ μ= =
LHC EXPERIMENT
DIMENSIONS
RO SIGNAL
TRIGGER BUFFER
ATLAS 50x400 μm2
7.2x8mm2binary andti
Internal fast-OR
2 - 6.4μs40 MHz
R.ThompsonS.Kolya/Manchester
12time over threshold
FullFull--3D sensor speed3D sensor speed3D Tests with 0.13 3D Tests with 0.13 μμm CMOS Amplifier chipm CMOS Amplifier chip(A Kok, S. Parker, C. Da Viá, P. Jarron, M. Depeisse, G. Anelli), fabricated at StanfordBy J Hasi C Kenney
2009
By J. Hasi, C. KenneyS-
21st
Janu
ary
2st
er-S
LAC
–AD
S Short collection distance
High average e-field at low Vbias
Parallel charge collection
ersi
ty o
f M
anch
ea
Via’-
The
Uni
ve
Raw oscilloscope trace
t 1
Cinz
iaD
3D signal simulation
2ns
T 300K
rt 1.5ns≈rt~1ns
3D Inter-electrodespacing = 50 μm
5��
T 300
Conclusions and Plans for 2009Conclusions and Plans for 200920
09
CNM-FBK-SINTEF successfully completed FE-I3 sensorsCERN and SLAC joined 3DAtlas pixel project
S-21
stJa
nuar
y 2
Progress in understanding of 3D behaviour with FE-I3 crucial for FE-I4 well advancedD t i di t d d i b t t i i
ster
-SLA
C –A
DS Data on irradiated sensors during beam test promising
and power budget within specs– Turbodaq Systems being installed in labs to speed up systematic testing
ersi
ty o
f M
anch
e testing
2009 plan:bump-bonding of existing sensors completed- preparation for test
aVi
a’-Th
e U
nive
bump bonding of existing sensors completed preparation for test beam at CERN. Distribution of existing assemblies to testing labs
organisation of systematic testing and database
Si f FBK l d d l d l
Cinz
iaD new runs at Sintef-FBK completed - dual readout test samples
ready for testing at Stanford
irradiation of sensors with protons and neutrons – testirradiation of sensors with protons and neutrons test
Test beams at CERN – May and Octoberdata analysis and report of systematic testing
Always keep an eye for new detector ideas....20
09
borrowed from P. Le Compte at the LiverpoolAtlas Tracker Upgrade Workshop Dec 2006
S-21
stJa
nuar
y 2
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
ea
Via’-
The
Uni
veCi
nzia
D
2009
S-21
stJa
nuar
y 2
Additional slides
ster
-SLA
C –A
DS
ersi
ty o
f M
anch
ea
Via’-
The
Uni
veCi
nzia
D
Production considerations Production considerations 20
09
•B-layer is ~1500-1800 cm2. This is around 4000 FE-I3 ATLAS Pixel sensors. If FE-I4 chips areused (336 x 80 or ~20 x17mm2 for a total area of 3.4cm2), the sensors will be five times as big, but
S-21
stJa
nuar
y 2
gthen the yield may reduce.
•The yield factor needs to be established. A combination of IRST, Sintef CNM and Stanford couldpossibly produce the required number of sensors (for b-layer ~1500 cm2 or ~1800 cm2 including
ster
-SLA
C –A
DS
spares for a total of ~520 FE-I4 sensors).
•IRST, Sintef and CNM would need to use 2009 to establish the process.
ersi
ty o
f M
anch
e •FE-I4 is being included in the next mask layouts but will need testing (Summer 09?).Stanford alone could produce two batches per year of 25 4 inch wafers
•~12 FE-I4 sensors would fit in a 4inch wafer so one would need 43 4 inch wafers divided by the
aVi
a’-Th
e U
nive yield to produce the 520 needed. The yield is not known. If one assumes 50% then 86 wafers or 4
batches. A reasonable timescale is ~2 years.
• SINTEF could produce more batches per year, but would at present need the holes filled at
Cinz
iaD Stanford (which can fill both 4 inch and 6 inch right now and can handle several hundreds wafers
per year for these steps). 100 wafers per year is considered aggressive but not impossible. Two orthree batches of 25 per year is reasonable. Currently they are processing 4 inch wafers for 3Dwork but could process 6 inch. This is a process change and would require further R&D.
•IRST and CNM could also contribute with 4 inch as soon as active edges are established (early09?)
3D signal modelling and data fit 3D signal modelling and data fit using a 0.25using a 0.25μμmmFE chip (Jarron Anelli Cern MIC)FE chip (Jarron Anelli Cern MIC)