ALCPG, UTArlington

1
Preliminary Investigations of Geiger-mode
Avalanche Photodiodes for use in HEP Detectors

• David Warner, Robert J. Wilson
• Department of Physics
• Colorado State University

2
Outline

• Motivation
• Avalanche Photodiodes
• Characteristics
• RD Plans
• Conclusions

3
Motivation

• Scintillating fiber, or WLS readout of
  scintillator strips basic component of several
  existing detectors (MINOS, CMS-HCAL)
• Standard photodetector photomultiplier tubes,
  great devices but
• Expensive (including electronics etc.),
• Bulky, magnetic field sensitive
• For the next generation would like a photon
  detector to be
• Cheaper
• Compact? Low mass? Magnetic field insensitive?
  Radiation hard?
• Future experiments
• BaBar upgrade - endcap?
• Future ee- Linear Collider? LHC?
• Nuclear physics? Space-based (NASA)?

4
Silicon Avalanche Photodiodes (APD)

• Solid state detector with internal gain.
• Avalanche multiplication
• initiated by electron-hole free carriers,
  thermally or optically generated within the APD
• accelerated in the high electric field at the APD
  junction.
• Proportional Mode
• bias voltage below the breakdown voltage, low
  gain
• avalanche photocurrent is proportional to the
  photon flux and the gain
• Geiger Mode
• bias voltage higher than the breakdown voltage,
  gain up to 108 from single carrier
• avalanche triggered either by single photon
  generated carriers or thermally generated
  carriers
• signal is not proportional to the incident photon
  flux.
• high detection efficiency of single carriers ?
  single photon counter
• to quench Geiger mode avalanche bias has to be
  decreased below the breakdown voltage

5
UV Enhanced Avalanche Photodiodes

• Development by Stefan Vasile et al, Radiation
  Monitoring Devices, Inc. Cambridge,
  Massachusetts, USA. (Now at aPeak, Newton, Mass.)
  
• Small Business Innovative Research (SBIR) award
  motivated by an imaging Cerenkov device
  application (focusing DIRC). c. 1996/97-98
• Design and fabrication of silicon micro-APD
  (mAPD) pixels
• 20-180 µm pixels, single photon sensitivity in
  the 200-600 nm wavelength range.
• Q.E. 59 at 254 nm (arsenic doping, thermal
  annealing)
• very high gain gt 108
• Geiger mode APD array with integrated readout
  designed but process/funding problems.

blue-infrared
UV-blue
6
Geiger Avalanche Characteristics

• Thermal carriers trigger avalanche
• dark count rate decreased using small APD space
  charge region generation volume
• Compatible with 5 volt logic
• strong noise rate dependence
• Temperature dependence
• ? factor 3 decrease for 25C to 0C
• ? factor 20 decrease for 25C to -25C
• Size dependence
• roughly linear with effective avalanche region
  area
• at room temp. predict few kHz for 100 mm, ? 100
  kHz for 500 mm
• Characteristics measured on a small number of
  samples

20 mm diameter pixel, room temp.
RMD Inc.
7
Photon Detection Efficiency
RMD Inc.
RMD Inc.
8
(No Transcript)
9
Prototype mAPD Array
RMD Inc.

• APD active area is 150 mm x 150 mm on 300 mm
  pitch
• Compatible with CMOS process ? potential for low
  cost large-scale production
• 70 photon collection efficiency with fused
  silica micro-mirrors (for f-DIRC)
• Fabrication attempt failed 1998/99. RMD claims to
  have solved the problems but no funds for a
  fabrication run.

10
MINOS Scintillation System

• Uses a large volume of cheap co-extruded
  scintillator bars (8m x 4cm x 1cm) with a single
  1.2mmØ Y11-175 multiclad WLS fiber epoxied in
  extruded groove
• WLS fiber is coupled to a long clear fiber and
  readout with a pixelated pmt
• 3-4 pe/fiber at 3.7 m including connections and
  pmt QE
• Several production facilities still operational

Source BaBar IFR Upgrade Status Report III
11
BaBar Modifications (SLAC/CalTech)

• Short (3.7m vrs 8m) version of MINOS system with
  Time to the get the second coordinate
• Replace the pmt with (low gain) APD 4X higher
  QE
• Increase number of fibers to 4 2X more light
• Increase scintillator thickness to 2cm 1.5X
  more light
• Project 50-60 pe at 3.7m for min. ion.

Source BaBar IFR Upgrade Status Report III
12
CSUSLAC Commissioned RD at aPeak

• P.o. placed December 2002
• 3.1. Package GPD pixels
• Wire bonding
• Breadboard passive quenching circuitry and GPD
  pixels.
• 3.2. Reliability evaluation
• Bias several pixels at 1.1V above breakdown for
  1,000 hours, document changes in dark count rate,
  and failure modes, if any.
• 3.3. GPD performance evaluation
• dark count rate vs. T40 to 30 C
• recovery time vs. pixel area determine if one
  microsecond recovery time can be achieved with
  passive quenching
• Gain vs. Temp. and bias Voltage
• Detection Efficiency _at_ Room Temp.
• 3.4. Optical interface fabrication and assembly
• Fab. and evaluate 4x1 beam couplers using GRIN
  and/or tapered fibers
• 3.5. Test GPD in Cosmic Ray Setup

13
50 mm diameter GPD layout
Proprietary. Do not distribute.
14
Recovery Time with Passive Quenching.
1 x 10 mm GPD
10 ms
475 mV

• Simple electronics -limiting resistor
• 10 ms quench time

15
Recovery Time - Active Quenching
1 ms
Design 1
2.75 V
0.5 ms
Design 2
325 mV
Trade off pulse amplitude with pulse width
(quench the avalanche sooner)
16
Active Quenching - New Design
Design 3
100 ns
Preliminary
1.2 V
17
Temperature Dependence
18
T (C)
Detection Efficiency

• 10 mm f gAPD
• 550 nm, 150 ns laser, 10 kHz
• Avg. 7 photons/pulse
• DE (Illuminated Rate - Dark Rate)/10 kHz

DE
-43
Preliminary
-43
-32
-30
-24
Preliminary
-20
-20
-13
2
2
9
23
23
T (C)
Nominal operating voltage
19
Optical coupling to small diameter pixels

• Couple 4 x 1.2 mm WLS fibers to 4 x 1mm glass
  fibers
• Draw 4 glass fiber into single fiber, various
  exit diameters
• Investigate light transmission efficiency

D
d
Concentration Factor, CF Area of input
aperture (A) / Area of photodetector (a) Coupler
Transmission Factor, TF Intensity at input
aperture / Intensity at output aperture
20
Optical couplers area reduction
Transmission Factor
ratio of areas
Concentration Factor, CF
Concentration Factor, CF

• Benefit from tapered fibers compared to ratio of
  areas is not dramatic ? 50-200
• Preliminary measurements at aPeak are in general
  agreement with the model
• We expect to get samples at CSU soon

21
Test Setup at CSU
Portable dark box

• Cosmics rays
• Calibrated with well-understood PMT at CSU
• Measure efficiency with gAPDcouplers

Initial Tests
22
gAPD Progress Summary

• SLACCSU initiated a p.o. to jumpstart further
  gAPD work at aPeak.
• New design from aPeak claims to be a more
  reliable process than the old one.
• Detection efficiency in 10 micron pixels ?15 at
  room temp., ? 25 at 40C (kHz dark count
  rate).
• Only modest dark count reduction with lower
  temperature expected to be better in next batch.
• Active quenching circuitry provides 1ms-0.1ms
  pulse widths, no additional deadtime.
• Successful fabrication of 4x1 tapered couplers
  complexity trade-off unclear.
• 50 mm diameter gAPDs breakdown occurs
  predominantly at the surface. Due to suspected
  design sensitivity to humidity.
• New run, with better control of the surface
  breakdown is being fabricated. Added backup
  design to layout. Larger, 150 mm devices by
  early February, 2003.

23
Motivation for Geiger-mode APDs - Recap

• High gain (109), gt 1 volt pulses
• Minimizes required electronics
• Good detection efficiency in WLS range (gt20? At
  550 nm)
• Efficient for low light output from WLS fibers
• Low supply voltage requirements (10-40V)
• Simplifies wiring harness
• Minimal cooling requirements
• Simplifies mechanical plant
• CMOS process
• simple
• on-chip integration of readout -gt cost-savings

24
Next Steps

• Many unanswered questions. Need to get the
  devices in our own lab!
• Assisting aPeak with SBIR proposal.
• CSU proposal to DoE Advanced Detector RD.
• Hope to provide a real HEP demonstration of
  utility for broad range of fiber applications.