1
2
3
4
5
6
7
8
Performance of the RICH detectors of LHCb
Antonis PapanestisSTFC - RAL
for the LHCb Collaboration
Outline
Detector description: See previous talk by Davide
Perego.
Cherenkov angle resolution Alignment Corrections for the
magnetic field distortions. Particle identification
PID algorithm. Small selection of physics
results that depend on particle ID.
9
Monte Carlo simulation of the invariant mass for B->hh decays
Cherenkov angle resolution (i)
10
Typical run
σ= 1.62 mrad σ= 0.68 mrad
Cherenkov angle resolution (ii)
11
Resolution distribution of all 2010 runsVery good stability in time
1.63 mrad
0.68 mrad
Resolution in time
RICH1
RICH2
Using isolated rings
12
Cherenkov angle vs momentum in RICH1
PID algorithm
Global event likelihood algorithm. Likelihood function includes
expected contributions from signal plus background for every pixel.
Signal photons come from the track that generated the ring, background can be noise or Cherenkov light from other tracks.
The whole event is considered as a whole.
Performs better than methods that treat every track separately, especially at high occupancies with overlapping rings.
13
Pure particle samples
14
†Nucl. Inst. & Meth. A 555 (2005) 356-369
σ=3.0 MeVσ=1.5 MeV
σ=1.0 MeV
PID performance
15
Improvement expected at the next reprocessing
K/ p/
RICH physics potential
16
Physics analysis (i)
17
Measurement of direct CP violation in charmless charged two-body B decays at LHCb (LHCb-CONF-2011-011)
KB
B KKB
Without PID dominated by:
with kinematic cuts onlyhhB
Physics Analysis (ii)
18
Measurement of the relative yields of the decay modes B0→D−π+ ,B0 → D−K+ , B0s → D−
s π + , and determination of fs/fd for 7 TeV pp collisions (LHCb-CONF-2011-013)
DBd DKBd
No combinatorial background
Philips Digital Photon Counting(PDPC)Dr. York HämischSenior DirectorPhilips Corporate Technologies
SiPM’s: from analog to digital, from laboratory to application
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
20
Digital Silicon Photomultipliers (dSiPM) -The next solid state (digital) revolution?
Transistor Television
X-Ray imaging
Digital Camera
Digital Photon Counting
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
21
PDPC: short history
2004 Research project: “Novel technologies for future PET systems”
2005 Dr.Thomas Frach invented the digital SiPM
2006 Research project: “Integrated digital light sensor”; first test chip with promising results (inkubator)
2006 Disentanglement of Philips Semiconductors – now NXP
2007 Start of the PDPC venture
2008 Proof of concept
2009 Technology launch at IEEE NSS-MIC sensor V1.0
2010 PDPC separate unit of Philips Corporate Technologies
2011 Introduction of PDPC-TEK (Technology Evaluation Kit) and sensor version 2.0
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Initial Motivation: A better detector for Positron- Emission-Tomography (PET) with TOF and DOI
Graphics courtesy of Spanoudaki & Levin, Stanford,in: Phys. Med. Biol. (56) 2011
TOF – time-of-flight
DOI – depth of interaction
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
23
From single APD to multiple SPAD’s:Single Geiger-mode APD
Geiger-mode APD Array (SPAD’s):“Silicon Photomultiplier”
B. Dolgoshein, V. Saveliev, V. GolovineFirst idea: Russia, early 80‘s
• fully analog• poor timing• high bias voltage, but
below breakdown
• single photon resolution• binary, but still analog• better timing• low bias voltage, above breakdown
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
(S)APD in Geiger-ModeE
xt.
Vo
ltag
eC
urr
ent
Time
VBD
Meta-stable
Breakdown
Quenching
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
25
SiPM-SPAD’s: sensitivity vs. dynamic range
Photographs courtesy of Spanoudaki & Levin, StanfordSensors, 10, 2010
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital Photon Counting: the concept
Intrinsically, the SiPM is a digital device:
A single cell (SPAD) breaks down or not.
It works like a switch:
no breakdown: „0“ – no photonbreakdown: „1“ – single photon
Therefore, while the APD is a linear amplifier for the input optical signal with limited gain, the SPAD is a trigger device so the gain concept is meaningless.(source: Wikipedia)
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital Photon Counting: realization
analog SiPM
www.hamamatsu.com
Summing all cell outputs leads to an analog output signal and limited performance
TDC andphoton counter
Digital output of• Number of photons• Time-stamp
Digital Cells
digital SiPM (dSiPM)
Integrated readout electronics is the key element to superior detector performance
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Analog vs. Digital SiPM
28
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital SiPM: Principle
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital SiPM: Principle
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital SiPM: Principle
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital SiPM: Principle
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital SiPM: architecture and pixel diagram
33
Single pixel
Die (2 x 2pixel)
(smart) tile (8x8 pixel)
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital SiPM array (tile 2.0)advanced integration
34
FPGA/Flash:
• tile firmware• data collection/concentration• Skew correction• Saturation correction• configuration• temperature measurement • dark count maps
200 MHz ref. clock
Power (1.2V, 1.8V, 2.5V, 3.3V, 30V)
SPI interface
Serial Dataoutput(x2)
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
Digital SiPM: Typical acquisition sequence (example)
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
PDPC dSiPM: intrinsic timing resolution
PDPC array
USB
Coincidencedetection
PDPC array
Clock,Config,Data
Clock,Config,
Data
PCFPGA board FPGA board
electronic trigger
36
psec-laser
-100 -80 -60 -40 -20 0 20 40 60 80 1000
10000
20000
30000
40000
50000
60000
70000
Co
un
ts (
arb
. u
.)
Timestamp difference (ps)
(44 ± 1) ps FWHM
-120 -80 -40 0 40 80 1200
10000
20000
30000
Co
un
ts
Timestamp difference (ps)
Timing jitter: 44 ps FWHM 59 ps FWHM
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
PDPC: scintillator coincidence setup
37
Recently achieved 120 ps FWHM using LSO:Ca
(Schaart et.al., TU Delft,submitted to IEEE-MIC)
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
PDPC dSiPM: Dark count mapWorst cells are switched off
• Dark counts per second at 20°C and 3.3V excess voltage
• ~ 95% good diodes (dark count rate close to average)
• Typical dark count rate at 20°C and 3.3V excess voltage: ~150Hz / diode
• Dark count rate drops to ~1-2Hz per diode at -40°C
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
PDPC dSiPM: slow scan imaging mode
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
PDPC dSiPM: crosstalk reduction by trenches
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
41
Comparison of light detectors – which one is going to win?
Table courtesy of Spanoudaki & Levin, Stanfordin: Sensors, 10, 2010
PDPCdSiPM
meaninglessfirst photon
25-75< 350.33No
mm²-m²Currently: ~25
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
A digital light sensor might beuseful beyond PET
?
?
ClinicalPET
imaging
Pre-clinical
PET&SPECT
PET/MR
ClinicalSPECTimaging
SpectralCT
Focus Nuclear Imaging
Adjacent opportunities
Low dose CT
Other Medical Imaging Analytical Instrumentation
High Energy Physics
Night Vision / Surveillance / Security
DNA Sequencing
Microscopy
Microarrays
Antineutrino Detection
Particle Accelerators
Automotive Night Vision
LIDAR
Facility/Homeland
Security
Cherenkov Detectors
?
LoC
Medical Imaging
Intra-operativeprobes
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
4343
Key parameters of light detectors
λ
sens.
res.
speed
timingdyn. range
size
priceKey parameters:
1. Wavelength2. Sensitivity3. Spatial resolution4. Speed5. Timing6. Dynamic range7. Size8. Price
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
4444
What does the ideal spider web for HEP look like?
Key parameters:
1. Wavelength2. Sensitivity3. Spatial resolution4. Speed5. Timing6. Dynamic range7. Size8. Price
λ
sens.
res.
speed
timingdyn. range
size
price
YH ©Philips Digital Photon Counting, February 2011
Philips Digital Photon Counting
www.philips.com/digitalphotoncounting
Thank you for your attention!
•Makoto Tabata•Japan Aerospace Exploration Agency (JAXA)•TIPP 2011 in Chicago
Recent Progress in Silica Aerogel Cherenkov Radiator
Makoto TabataJapan Aerospace Exploration Agency (JAXA)
TIPP 2011 in Chicago
Recent Progress in Silica Aerogel Cherenkov
Radiator
Silica Aerogel
1
Silica aerogel is a 3D structural solid of SiO2 particles.
Scanning electron microscope image 10 mg/cc silica aerogel (magnification: x 100,000)
• porous (>99% air)low bulk densitythermal insulator
• transparentΟ(10) nm SiO2 particlesRayleigh scattering
• feel like styrofoam
30 mg/ccsilica aerogel
100 nm
1st method: Single-step method2nd method: Two-step method world standard3rd method: KEK method our original
1.Wet-gel synthesis (sol-gel step) We have various preparation recipes for chemicals.
2.Aging3.Hydrophobic treatment our original4.CO2 supercritical drying 1 month in total
Conventional Production Method
3
CO2 autoclave
extract liquid component of wet-
gel through supercritical drying
Wet-gel
Hydrophobic
Refractive index (density) can be controlled in the sol-gel step.
1.Wet-gel synthesis (1st density control)2.Aging3.Pin-drying (2nd density control)
4.Hydrophobic treatment5.CO2 supercritical drying
Pin-drying Production Method
7
1.00
1.05
1.10
1.15
1.20
1.25
1.30
0.5 0.6 0.7 0.8 0.9 1.0
length contraction
refr
activ
e in
dex
n0 = 1.06
shrink
original non-shrink aerogel
Semi-sealed container with some pin-holes
several weeks
Pin-drying (PD) method is 4th method to produce aerogelwith high refractive index. need additional time
for the pin-drying process
methanol solvent
Partial evaporation of solventpin-
hole
Reproducibility in PD Method
8
1.10
1.15
1.20
1.25
1.30
20% 30% 40% 50%
wet-gel weight
refr
activ
e in
dex
Target n = 1.25 (20 tiles)Target n = 1.20 (10 tiles)
Target n = 1.12 (10 tiles)0.01
0.01
0.005
Target refractive index is well-controlledby monitoring wet-gel weight.
40 tiles were produced forn = 1.25, 1.20 and 1.12.
(Wet-gels were synthesizedfor n0 = 1.06 usingmethanol solvent.)
Refractive index fluctuations were evaluated at each
target index.
0
10
20
30
40
50
60
70
1.00 1.05 1.10 1.15 1.20 1.25 1.30
refractive index
tran
smis
sion
leng
th [m
m]
Conventional (Ethanol)
Conventional (Methanol)
Conventional (DMF)
PD (Methanol)
Expansion of High Index Range in 2005
10
Original (non-shrink) aerogeln = 1.06 (methanol)
Ultra-high refractive index (n > 1.10) aerogels with sufficienttransparency (Λ > 20 mm) were developed.
Entirely newrefractive index range
Methanol solvent
0
10
20
30
40
50
60
70
1.00 1.05 1.10 1.15 1.20 1.25 1.30
refractive index
tran
smis
sion
leng
th [m
m]
Conventional (Ethanol)Conventional (Methanol)Conventional (DMF)PD (Methanol)PD (DMF)
Improvement of Transparency in 2008
11
Original (non-shrink) aerogeln = 1.066 (DMF)
The transmission length was improved in n > 1.10.
It takes long pin-drying process because DMF is difficult to evaporate.
DMF solvent
0
10
20
30
40
50
60
70
1.00 1.05 1.10 1.15 1.20 1.25 1.30
refractive index
tran
smis
sion
leng
th [m
m]
Conventional (Ethanol)Conventional (Methanol)Conventional (DMF)PD (Methanol)PD (DMF)
The Most Transparent Sample in 2008
12
Original (non-shrink) aerogeln = 1.049 (DMF)
The highest transparency (Λ > 50 mm) was obtained in n ~ 1.06.
DMF solvent
1.003 < n < 1.26 available
• n = 1.05 + 1.06, 2 cm thick each (total 4 cm thick)
Np.e. = 10.6 (conventional) 13.6 (improved)60% ring acceptance
• 1.10 < n < 1.231 cm thick each Np.e. = 5-10 (new data)50% ring acceptance
Photoelectron Yield
14
0
2
4
6
8
10
12
1.00 1.05 1.10 1.15 1.20 1.25refractive index
Np.
e./t
rack
MethanolDMFClear Cherenkov
rings were observed.
Standard method
PD method
TransparencyDMF > Methanol
Sufficient photoelectrons were detected.
1 cm thick