Sensor development for the ALICE ITS upgrade in LS3 - Magnus Mager (CERN) on behalf of the ALICE collaboration
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Sensor development for the ALICE ITS upgrade in LS3 Magnus Mager (CERN) on behalf of the ALICE collaboration
Outline
‣ ITS2 (LHC LS2, i.e. now)
TDR Commissioning
- Detector layout
- Sensor technology il
- Highlights from commissioning
‣ ITS3 (LHC LS3, i.e. soon)
Idea Prototypes
- Detector concept + predicted
performance
- Project overview
- Sensor development
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 2
ALICE
Detector and main goals
‣ Study of QGP in heavy-ion collisions at LHC
- i.e. up to O(10k) particles to be tracked in
a single event
‣ Reconstruction of charm and beauty hadrons
‣ reconstruction
Interest in low momentum (≲1 GeV/c) particle
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 3
ALICE today
LS2 upgrades with Monolithic Active Pixel Sensors (MAPS)
Inner Tracking System
ITS2
LS2
6 layers: 7 layers:
2 hybrid silicon pixel all MAPS
2 silicon drift
2 silicon strip
Inner-most layer: Inner-most layer:
radial distance: 39 mm radial distance: 23 mm
material: X/X0 = 1.14% material: X/X0 = 0.3%
pitch: 50 ⨉ 425 μm² pitch: O(30 ⨉ 30 μm²)
rate capability: 1 kHz rate capability: 100 kHz (Pb-Pb)
Muon Forward Tracker
MFT
new detector
5 discs, double sided:
based on same technology as ITS
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 4
ITS2
Overview of the detector Inner Barrel (IB) Flexible PCB
3 Inner Layers: 12+16+20 Staves
9 sensors
Layout 1 Module / Stave
Cold Plate
9 sensors per Module
Space Frame
96 Modules to be produced
(2 Inner Barrels + spares)
40
27 cm
cm
Outer Barrel (OB)
2 Middle Layers: 30+24 Staves
2⨉4 Modules / Stave
2 Outer Layers: 42+48 Staves
m 2⨉7 Modules / Stave
7 c
14
2⨉7 sensors / Module
Total: (Middle and Outer Layers are
- 24k chips equipped with the same Module
- 10 m2 Type)
- 12.5 GPixel
1880 Modules to be produced
(including spares)
Good news: it was all built, assembled and tested!
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 5
ALPIDE
MAPS technology
524 288 pixels
3 c m
1.
5c
m
‣ Process: TowerJazz 180 nm CIS
- deep p-well to allow CMOS circuitry inside matrix
- reverse-substrate bias
‣ Detection layer: 25 μm high-resistive (>1 kΩcm) epitaxial layer
‣ Thickness: 100 μm (OB) or 50 μm (IB)
More details in Maxim’s talk later! Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 6
ITS2 commissioning
Magnus Mager (CERN) | Novel Ideas in Silicon Tracking Detector (ALICE) | IAS HEP2021 | 21.01.2021 | 7
ITS2 commissioning
Cosmics (2)
3 layers:
real time!
Magnus Mager (CERN) | Novel Ideas in Silicon Tracking Detector (ALICE) | IAS HEP2021 | 21.01.2021 | 8
ITS2 commissioning
(No) noise / fake-hit rate
‣ This is the real rate measured on-
detector, including final services
‣ Essentially, apart from a handfull pixels
per chip, the detector is noise-free
‣ Biggest contributor are cosmic rays
(which were not excluded here)
Excellent on-detector performance! Looking forward to recording collisions!
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 9
ITS3 — the golden detector
Magnus Mager (CERN) | Novel Ideas in Silicon Tracking Detector (ALICE) | IAS HEP2021 | 21.01.2021 | 10ITS2 inner barrel
40 ITS2: global layout ITS2: assembled three inner-most half-layers
27 cm
cm
c m
7
14
‣ ITS2 will already have unprecedented performance
‣ The Inner Barrel is ultra-light but rather packed → further improvements seem possible
‣ Key questions: Can we get closer to the IP? Can we reduce the material further?
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 11Motivation for ITS3
‣ Observations:
- Si makes only 1/7th of total material
- irregularities due to support/
cooling
‣ Removal of water cooling
- possible if power consumption
stays below 20 mW/cm2
‣ Removal of the circuit board
(power+data)
- possible if integrated on chip
‣ Removal of mechanical support
- benefit from increased stiffness by
rolling Si wafers
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 12ITS3
detector concept
‣
truly cylindrical
detection layers Key ingredients:
- 300 mm wafer-scale chips, fabricated
using stitching
- thinned down to 20-40 μm (0.02-0.04%
X0), making them flexible
- bent to the target radii
rb o n - mechanically held in place by carbon
a
c rs
c e
n- sp
l l
a c e foam ribs
e
op oam
f
‣ Key benefits:
- extremely low material budget:
0.02-0.04% X0
(beampipe: 500 μm Be: 0.14% X0)
- homogeneous material distribution:
negligible systematic error from material
distribution
The whole detector will comprise six (!) chips – and barely anything else
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 13Performance figures
pointing resolution tracking efficiency
[ALICE-PUBLIC-2018-013]
improvement of factor 2 over all momenta large improvement for low transverse momenta
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 14Lambda-c (Λc)
schematic view of a Λc decay
K ‣ Analysis difficult due to large
combinatorial background:
π
- O(10k) charged particles in a
central Pb-Pb collision
p
‣
)
O
(6
0µ
m
Discrimination of background via:
)=
c
t(Λ Λc
c
- Particle identification (relatively low
“d D yield of protons and Kaons wrt.
C
pions)
is
ta
n
ap ce
A(
pr of
oa c
p)
ch los
” e
- Topology: cut on DCA of single
st
tracks (before making the
combinations) and decay vertex
position (needs combinations)
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 15Lambda-c (Λc) (2)
[ALICE-PUBLIC-2018-013]
‣ Large improvement of S/B + significance due to better separation power of secondary decay
vertex (Λc ≈ 60 µm)
‣ Allows for precision measurements over a wide pT range
Λc yield S/B Significance
ITS3 / ITS2 10 4
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 16ITS3 project startup
EoI, LoI, start of R&D
Sep 2019 Dec 4th 2019
R&D!
>150 people!
gaining momentum!
[ALICE-PUBLIC-2018-013] [CERN-LHCC-2019-018 ; LHCC-I-034] [ITS3 R&D kickoff meeting]
[…] […]
The LHCC is impressed by the new concept for the ITS3 The LHCC congratulates the ITS3 groups for
[…]
the successful start of the project.
The LHCC endorses the plan of ALICE to carry out the […]
necessary R&D [LHCC minutes: CERN-LHCC-2020-008 ; LHCC-142]
[…]
[LHCC minutes: CERN-LHCC-2019-010 ; LHCC-139]
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 17Wafer-scale chip
Principle of photolithography 200 mm ALPIDE prototype wafer
‣ Chip size is traditionally limited by
CMOS manufacturing (“reticle size”)
- typical sizes of few cm2
- modules are tiled with chips
connected to a flexible printed
circuit board
Wafer-scale sensor
‣ New option: stitching, i.e. aligned
Stave design
o l i ng
FPC + chips
Courtesy: R. Turchetta, Rutherford Appleton Laboratory
exposures of a reticle to produce pp or t + co
su
larger circuits 14 c m
27
- actively used in industry ALPID
E chip
cm
FP s
- a 300 mm wafer can house a chip C
(p
ow
to equip a full half-layer er
+d
at
- requires dedicated chip design a)
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 18Sensor dimensions
65 nm process
‣ All the three foreseen (half-) layers can be
equipped with a single chip from a 300 mm
wafer
‣ This is the driving argument for ITS3 to go to
the 65 nm process
‣ At the same time, ITS3 wants to exploit the
technology at its best, potentially:
- lower power
- more logic
- smaller pixels
- faster readout
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 19Architecture
Ideas + Challenges
Possible architecture – many more being explored right now!
‣ Stitching requires a regular, periodic structure of the matrix
‣ Different architectures are under study
- here, the concept laid out in the LoI is mostly inspired by
ALPIDE
‣ ITS3 needs three different chip sizes
- naturally comes with stitching a variable number of
“rows”
- likely adaptable to other geometries (e.g. square-like
matrices)
‣ Electrical interface only on one side
‣ Power consumption needs to stay low in the matrix region,
but can be larger at periphery
- dictated by air cooling capabilities and acceptable
voltage drop along the sensor
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 20Yield
3 Summary of test results & Yield
Figures 8 and 9 show respectively the number and fraction of chips classified as gold , silver , bro
and NOK for each of the previously described testing parameters. The results are shown only fo
chips that can be powered.
ALPIDE results
‣ Producing a working wafer-scale sensor is very
challenging
‣ Dedicated effort on yield optimisation is needed
‣ Different directions are explored:
- layout optimisation
Fig. 9: Classification summary – fraction of chips in each category (respective colors, with red for NOK )
- testing
for single redundancy
parameters.
- fault tolerance
‣ ALPIDE results
Table from several
7: Final ALPIDE thousand wafers
yield summary.
provide valuable input Category
OK NOK
gold silver bronze
30.4 22.9 10.6 36.1
Yield (%)
63.9 36.1 Fig. 8: Classification summary – number of chips in each category (respective colors, with red for N
forMager
Magnus single testing
(CERN) parameters.
| Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 21The next chip
baseline specifications
ALPIDE is a great starting point, samller technology node will open further possibilities!
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 22Sensor prototyping
First 65nm submission “MLR1”
~12 mm
‣ ITS3 plays a leading role in the evaluation of the TowerJazz
65 nm technology for MAPS
‣ First test structures in 65 nm have been just submitted
- transistor test structures for radiation hardness studies
- various diode matrices for charge-collection studies
~16 mm
- analog building blocks
‣ Large number of dies in different processing flavours are
expected
- will provide crucial input into the next design steps
‣ Wafer-scale blocks will be diced out to study mechanical
properties expect O(500) dies per wafer
‣ First chips expected mid 2021: stay tuned!
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 23Mechnical porperties of thin, wafer-scale chips
Bending
Mylar foil
50 μm-thick dummy silicon,
full inner-layer size
R = 30 mm mandrell
Magnus Mager (CERN) | Novel Ideas in Silicon Tracking Detector (ALICE) | IAS HEP2021 | 21.01.2021 | 24Summary
‣ ALICE ITS2 is the first really large-scale (O(10 m2)) application of MAPS in
HEP
- the detector is fully built and ready for installation
- commissioning results verify excellent system-level performance
‣ The development of ALPIDE marks a new generation of MAPS in terms of
functionality and performance figures
‣ ALICE ITS3 will push the technology even further, approaching the massless
detector
- R&D encouraged by LHCC in Sep 2019 and progressing at full steam
- just proved the feasibility of curved silicon detectors, marking the start of a
new chapter on silicon-detector designs
- fully integrated, wafer-scale sensors will allow large-scale minimal
material budget applications
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 25!ank you!
Magnus Mager (CERN) | Sensor development for the ALICE ITS upgrade in LS3 | DAQFEET-2021 | 08.02.2021 | 26You can also read
