University of Chicago
Integration-Level Testing of Sub-Nanosecond Microchannel Plate Detectors for Use in
Time-Of-Flight HEP Applications
M. Wetstein, B. Adams, M. Chollet, S. Jokela, Z. Insepov, V. Ivanov, J. Elam, A. Mane, Q. Peng for the LAPPD Collaboration
TIPP June 13 2011
Sunday, June 19, 2011
TIPP 2011
2
Making MCPs Faster, Bigger, and Cheaper:
• Microchannel Plate (MCP): A high-gain structure consisting of a thin plate with microscopic (typically <50 μm) pores.
• The material in these plates is optimized for secondary electron emission (SEE).• Plates are held at high voltages (typically a few kV) so that electrons will accelerate and
strike the walls, initiating an avalanche of secondary electrons.• Known for good gain (>103), excellent timing resolution (<100 psec) and spatial resolution
(<1 mm).• Unfortunately, they are also typically expensive.
Sunday, June 19, 2011
TIPP 2011
3
Making MCPs Faster, Bigger, and Cheaper:
pore
LAPPD (Large-Area Picosecond Photodetector) Project:Make large-area MCPs with low-cost, bulk materials, applied independently using atomic layer deposition (ALD), an established chemical process used by industry...
borosilicate glass filters
1. Start with a porous, insulating substrate that has appropriate channel structure.
2. Apply a resistive coating (ALD)
3. Apply an emissive coating (ALD)
4. Apply a conductive coating to the top and bottom (thermal evaporation or sputtering)
ALD GroupJ. Elam, A. Mane, Q. Peng
Sunday, June 19, 2011
TIPP 2011
4
Characterization program:
Microchannel plates, themselves, exist within the context of a larger detector system, a microchannel plate photomultiplier tube (MCP-PMT). The goal of the LAPPD collaboration is the development of a complete 8”x8” sealed tube detector.
MCP 1
MCP 2
Incident Light
PhotocathodeWindow
Anode
to readout electronics...
A strong testing program is essential not only to study individual components, but to understand how these parts work together in an integrated system.
The LAPPD collaboration has several testing facilities:
• MCP testing at Berkeley SSL • Material characterization at
Argonne Material Science Division
• Photocathode characterization lab
• MCP characterization at Arradiance
• MCP gain and electrical testing at the ANL-ALD lab
• MCP testing at the Advanced Photon Source (APS)
Sunday, June 19, 2011
MCP 1
MCP 2
Incident Light
PhotocathodeWindow
Anode
TIPP 2011
5
Characterization program:
Gap spacing voltages:
Gap 1: “first strike”Impacts on variability of transit time and amplification
Gap 2: Impact on saturation of MCP pair, spatial spread of signal
Gap 3:spatial and temporal spreading of the charge cloud. Space charge effects. Interface with anode.
Determine optimal operational voltages. How do these optimal voltages depend on particular choice of MCPs? Explore tradeoffs between gain, timing, saturation.
Sunday, June 19, 2011
MCP 1
MCP 2
Incident Light
PhotocathodeWindow
Anode
TIPP 2011
Characterization program:
6
Geometry (pore size, L/D)Chemistry (SEE, resistive layer) Plate quality Uniformity Noise StabilityResistivity Saturation Relaxation time
MCP performance:What impact do each of the electrical, secondary electron yield (SEY) and geometric properties have on the overall timing, gain, and saturation of the MCP?
Sunday, June 19, 2011
MCP 1
MCP 2
Incident Light
PhotocathodeWindow
Anode
TIPP 2011
Characterization program:
7
Anode Structure, Signal Processing
What is the best anode design for a particular application. How does one reduce channel counts and cost without sacrificing timing or spatial resolution? How to maintain multi-GHz analog bandwidth and 50 ohm impedance?
Anode Design:
Sunday, June 19, 2011
TIPP 2011
8
An Opportunity: Goals of the ANL MCP-Characterization Lab
• ALD gives us the unique ability to vary electrical, secondary electron yield (SEY) and geometric properties of MCPs independently.
• Compared with commercial MCPs, which are typically made from a single material (lead-glass), we can produce MCPs with much wider variety of properties, other properties held fixed.
• Can explore limiting cases and place stronger constraints on MCP models.
Improving Fundamental UnderstandingDevelop Working ExperienceProof of PrinciplesGuide Design
33mm samples
8” testing
Understanding scalabilityDeveloping operational experience
sealed-tube testing
ANL MCP-Testing Program
Working out the challenges of a complete systemDeveloping operational experience
A unique collaboration between the HEP division and the Advanced Photon Source (APS)
Sunday, June 19, 2011
TIPP 2011
Facilities and Resources:
9
Sunday, June 19, 2011
TIPP 2011
10
Facilities and Resources:
Sunday, June 19, 2011
TIPP 2011
11
ANL MCP Characterization Lab:
• A fast (sub-psec), pulsed laser with precision optics• 800 nm Ti:Sapph laser• pulse durations O(10) femtoseconds • 1000 Hz repetition rate• non-linear optics to produce UV(266 nm) and blue light
(400nm)• average power ~800 mWatt• optics capable of micron-level translations and potential
to focus on single pores
• Vacuum systems for testing 33 mm photocathode-MCP-anode stacks approximating a complete device• Capable of holding variable stacks of 1-3 MCPs and
simple photocathode• able to accommodate multiple readout designs• capable of 10-7 torr• 2 complete systems with parts for a third
• 8” MCP testing system (now commissioning)
• Fixtures for testing sealed-tube detectors (now commissioning)
• multi-GHz RF electronics• several oscilloscopes with 3-10 Gz analog bandwidth• high gain, low noise RF amplifiers• high-frequency splitters, filters, etc
Facilities and Resources:
Sunday, June 19, 2011
TIPP 2011
12
Methodology
0
0.125
0.250
0.375
0.500
0 1.75 3.50 5.25 7.00
y = 0.0675x + 0.0027
Fraction of Laser Pulses with Signal
• Control the number of photoelectrons (PEs) by attenuating the laser to the point where only a small fraction of pulses produce signal.
• Trigger on laser pulses to achieve very precise measurements of transit time
• Control size and position of beam to isolate individual spots on the MCP
• Record each pulse separately to produce statistical distributions.
• Integrate and fit the pulses to determine arrival time and gain.
• Able to discriminate between signal pulses and dark-current (random firing of the MCP)
4 5 6 7 8 9 10 11 12 13x 10 9
0.005
0
0.005
0.01
0.015
0.02
0.025
time (seconds)
Mean Pulse Shape, MCP 72/78 at 2.6 kV
1000V anode gap800V anode gap500V anode gap200V anode gap
time from trigger = MCP transit time + known optical and electronic delays... Area of pulse = total charge.
When divided by incident charge, this gives the gain...
UV intensity (nW)
Sunday, June 19, 2011
TIPP 2011
13
• Early study of timing characteristics from a Cesium-Iodide Photocathode
• Demonstration of enhanced gain from ALD coating on a commercial plate
• Developed operational experience working with MCPs
• Observation of first signals from ANL-fabricated, ALD-based MCPs
• Design and commissioning of characterization chambers
Year 1 achievements:
1/19/10 Collaboration Meeting 5
LAPPD Collaboration: Large Area Picosecond Photodetectors
Example Signals at 3 kV
“wides pulses”
4 5 6 7 8 9 10 11 12 13x 10 9
0.005
0
0.005
0.01
0.015
0.02
0.025
time (seconds)
Mean Pulse Shape, MCP 72/78 at 2.6 kV
1000V anode gap800V anode gap500V anode gap200V anode gap
Sunday, June 19, 2011
TIPP 2011
14
• Completion of laser characterization lab for systematic MCP testing in the time domain.
• Developed operational experience performing current-based, average gain measurements.
• Demonstrated > 105 amplification on Argonne-made, 33mm ALD functionalized glass plates.
• Demonstrated better than 200 psec time resolutions for single photoelectons in ALD MCPs
• Developed protocol for pulsed, single-photoelectron characterization.
• Close work with simulations and material characterization to improve fundamental understanding of MCP performance.
• Designed system for characterization of 8” MCPs, sealed tubes and lifetime testing
Year 2 achievements:
1
10
100
1000
10000
100000
600 750 900 1050 1200
Comparison of ALD MCP Gains w Commercial MCP
MCP 122Avg Mock Tile MCPCommercial MCP
Gain curves for mock tile MCPs
Voltage Across the Anode Gap (Volts)200 300 400 500 600 700 800 900 1000
Rel
ativ
e A
rriv
al T
ime
of S
igna
l (ps
ec)
0
50
100
150
200
Relative Shift in Mean Arrival Time of Signal Vs Voltge on Anode Gap
field strength (Volts/mm)voltage (volts)
Gain
Average pulse shape for single MgO MCP at 1.5 kV, different photocathode voltages
Relative shift in mean arrival time of signal VS anode-gap voltage
time (seconds) x10-9
Comparison of ALD-MCP gains with commercial MCP
Sunday, June 19, 2011
TIPP 2011
15
• Systematic comparison of gain and timing for MCPs with identical resistance, but three different SEY (secondary electron yield) compositions.
• 20nm Al2O3• 20nm MgO• 2nm MgO
• Testing operation of single plates at high voltages.
• Comparison of MCP stacks with a common bottom plate.
• Systematic tests conducted for many different operational voltages, with the hope of placing strong constraints on models for avalanche formation.
• Plans for direct comparison of data with simulations and an upcoming publication
Current Work:
0.5 1 1.5 2 2.5 3x 106
100
200
300
400
500
600
700
800
# electrons on central (max) strip
PHD fo 20 nm MgO, 1.5 kV single plate, 300 eV first Strike (50.4 nW incident UV)Integrated Charge Per Pulse, 20nm MgO
Single Plate at 1.5 kV, 300V on Photocathode
x 10-9 seconds-1.3 -0.8 -0.3 0.2 0.7 1.2 1.7
Arrival Time Distribution, 20nm MgO Single Plate at 1.5 kV, , 300V on Photocathode
arrival time
arriv
al ti
me
(sec
onds
)
0.55
0.50
0.45
0.45
0.40
0.35
0.30
0.25
Average Arrival Time Per Pulse, 20nm MgO at Different PC and MCP Voltages
8E+05
9E+05
1E+06
1E+06
1E+06
8E+05
9E+05
1E+06
1E+06
1E+06
0 100 200 300 400
Average Gain Above 800k VS First
MCP at 1.5 kV MCP at 1.44 kV MCP at 1.36 KV
Average Integrated Charge Per Pulse, 20nm MgO at Different PC and MCP Voltages
Sunday, June 19, 2011
TIPP 2011
16
0
10000
20000
30000
40000
0 75 150 225 300
1.5 kV
0.5 1 1.5 2 2.5 3x 106
100
200
300
400
500
600
700
800
# electrons on central (max) strip
PHD fo 20 nm MgO, 1.5 kV single plate, 300 eV first Strike (50.4 nW incident UV)
0
50
100
150
200
0 37500 75000 112500 150000
# O
ccur
ence
s
MgO; Umcp=1.5kV; En=300eV
Current Work:
0
1
2
3
4
5
6
0 200 400 600 800
20 nm MgO Slade's SEY data
SEY,
[1/e
lect
ron]
Primary electron energy, eV
Slade: tilt=0 Our fit: tilt=0Fit: tilt=10 Fit: tilt=20Fit: tilt=30 Fit: tilt=40Fit: tilt=50 Fit: tilt=60Fit: tilt=70 Fit: tilt=80
Complete MC-Data Cycle
• 3 MCP samples made with Identical resistance, but different SEY chemistry
• Simulations based on material-level characterization of SEY layers, guided by material-level simulations.
• MCP-level simulations to be tuned to data for 1 of the 3 MCP samples, taken at multiple operational voltages.
• Once tuned, predictions will be made on the performance of the other two samples, to be compared with data, afterwards...
V. Ivanov
S. Jokela, Z. Insepov
B. Adams, M. Chollet, M. Wetstein
Sunday, June 19, 2011
TIPP 2011
17
Current Work:
0
1000
2000
3000
4000
0 750000 1500000 2250000 3000000
# O
ccur
ence
s
PHD. MgO 20nm; Umcp=1.5kV; En=300eV
Exper.L=4
700000
850000
1000000
1150000
1300000
0 100 200 300 400
Average Gain Above 800k Versus First Strike Energy
MCP at 1.50 kV (simulation) MCP at 1.50 kV (data)MCP at 1.36 kV (simulation) MCP at 1.36 kV (data)
Preliminary
• Still tweaking the model. We expect even better agreement
• The model is tuned to the data, but only for this sample with 20nm MgO SEY coating.
• Once tuned, the simulation will be used to make a priori predictions of our other two samples, before the data results are revealed.
• Coming soon!
20 nm MgO SEY layerV. Ivanov
Sunday, June 19, 2011
TIPP 2011
18
• 8” testing chamber is complete and successfully held 10-7 torr.
• Fixtures for mechanical assembly of 8” MCP stack designed to use spare glass parts for the sealed tube body.
• In the commissioning process.
8” MCP Testing:
• Chamber will be used to test, both:• 8”MCPs and• 33mm samples on 8” transmission
lines
• Looking forward to our first 8” MCP samples.
Sunday, June 19, 2011
TIPP 2011
19
Sealed Tube Testing
• Received our first functional, sealed-tube MCP (“mock-tile”), built to the specs of an 8” MCP stack, but with 4 pairs of working, 33mm MCPs
• Constructed a system for mechanical support, electrical connection, vacuum connection, and signal readout from tile.
• Successfully coupled the tile to our vacuum system and achieved a vacuum of ~10-5 torr (as measured just outside the pump-port).
• Working out some technical difficulties.
• In the midst of basic electrical testing.
• Several new sealed-tubes are currently being made.
Sunday, June 19, 2011
TIPP 2011
20
Near Term: Long Term:
• Systematic test of a 12 sample ensemble of MCPs with varying resistive and secondary emissive chemistries.
• Commissioning of the 8” testing system, successful operation of first working 8” MCPs
• Demonstration of first working sealed-tube detector
• Comparison of several anode designs, testing of PSEC chip on MCP signals
• Commissioning of aging/scrubbing experiment
• Systematic batch testing of identical MCPs
• Integration of testing methods with Tile Factory
• Single pore testing, aging and saturation studies (double-pulsed measurements)
• Tests of potential single-MCP detectors
Conclusions and Future Prospects:
Sunday, June 19, 2011
TIPP 2011
21
Thanks!
Sunday, June 19, 2011
TIPP 2011
22
Details in Simulations:
0
1
2
3
4
5
6
0 200 400 600 800
20 nm MgO Slade's SEY data
SEY,
[1/e
lect
ron]
Primary electron energy, eV
Slade: tilt=0 Our fit: tilt=0Fit: tilt=10 Fit: tilt=20Fit: tilt=30 Fit: tilt=40Fit: tilt=50 Fit: tilt=60Fit: tilt=70 Fit: tilt=80
The LAPPD Simulations Program
• Goal to develop a predictive, pseudo-physical MCP model to help guide MCP design.
• Help improve understanding of what is going on inside the pores.
• Takes experimental materials characterization as input.
• Two components:
• true secondary electron yield (SEY)
• specular reflection of incident primary electron, eg backscattering or BS
• SEY at normal incidence is measured.
• SEY at grazing incidence is extrapolated using a theoretical material model
• quasi-elastic reflection of the primary electron is derived from a theory.
• Normalization of the BS probability is a tunable parameter (controls the fraction of highly energetic electrons in the pore).
Sunday, June 19, 2011
TIPP 2011
23
• A concern in using fast timing are the effects of frequency dependent dispersion, scattering and absorption.
• Using a fast toy MC originally developed by J. Felde we study the time of arrival for photons in an spherical detector.
• For a 50m detector with 100% coverage, the rise time (t90-t10) is of the order of 2 ns which cannot be sampled with standard PMT technology.
• For a given detector size, the rise time stays constant and the uncertainty in the position of the leading edge becomes smaller if larger photodetector coverage is considered.
• A combined improvement in photodetector coverage (for reduced
/ ndf 2 63.95 / 29Prob 0.0001943Constant 4.4! 240.5 Mean 0.1! 366.4 Sigma 0.038! 1.756
Time (ns)360 365 370 375
Phot
ons
0
50
100
150
200
250 / ndf 2 63.95 / 29Prob 0.0001943Constant 4.4! 240.5 Mean 0.1! 366.4 Sigma 0.038! 1.756
Photon time of flight
0
150
300
450
600
20 50 80
Uncertainty on Arrival Time
Un
ce
rta
inty
(p
se
c)
Distance from Detector (meters)
100% Coverage
30% Coverage
10% Coverage
Backup Slides:Applications in Water Cherenkov
Sunday, June 19, 2011
TIPP 2011
Backup Slides:Applications in Water Cherenkov
• Collaboration among the hi-res WCh working group has produced a new platform for testing algorithms on WCh detectors with interactively modifiable photodetector properties.
• These efforts have already identified promising features in observables, such as timing residuals, that could potentially be used to improve track reconstruction and better identify pi0 backgrounds.
• GEANT-based studies are being done in less idealized conditions:Including effects of
trackresidualdistEntries 29230Mean 3.184RMS 3.601
/ ndf 2 597.4 / 417Constant 9.7! 1096 MPV 0.009! 1.278 Sigma 0.0048! 0.6612
-10 -5 0 5 10 15 20 250
20
40
60
80
100
120
140
160
180
200
trackresidualdistEntries 29230Mean 3.184RMS 3.601
/ ndf 2 597.4 / 417Constant 9.7! 1096 MPV 0.009! 1.278 Sigma 0.0048! 0.6612
LBNE Collaboration Meeting01/26/11 20
Difficulties in building a Figure of MeritTiming residual distributions are not Gaussian, but conventional “likelihood” test function is.The residual distribution gets less Gaussian as the timing resolution improves:
• How do you line up the timing residual and the test function?• Which features are more sensitive to variations in the hypothesized vertex?
time residual (ns)-30 -20 -10 0 10 20 300
200
400
600
800
1000
1200
1400
1600
1800
Test Functions for Likelihood
5 nsec resolution50 psec resolution
Rising edge driven by chromatic dispersion
Tail driven by scattered light
Wednesday, January 26, 2011
trackresidualdistEntries 29230Mean 3.184RMS 3.601
/ ndf 2 597.4 / 417Constant 9.7! 1096 MPV 0.009! 1.278 Sigma 0.0048! 0.6612
-10 -5 0 5 10 15 20 250
20
40
60
80
100
120
140
160
180
200
trackresidualdistEntries 29230Mean 3.184RMS 3.601
/ ndf 2 597.4 / 417Constant 9.7! 1096 MPV 0.009! 1.278 Sigma 0.0048! 0.6612
LBNE Collaboration Meeting01/26/11 20
Difficulties in building a Figure of MeritTiming residual distributions are not Gaussian, but conventional “likelihood” test function is.The residual distribution gets less Gaussian as the timing resolution improves:
• How do you line up the timing residual and the test function?• Which features are more sensitive to variations in the hypothesized vertex?
time residual (ns)-30 -20 -10 0 10 20 300
200
400
600
800
1000
1200
1400
1600
1800
Test Functions for Likelihood
5 nsec resolution50 psec resolution
Rising edge driven by chromatic dispersion
Tail driven by scattered light
Wednesday, January 26, 2011
M. Wetstein (UC, ANL/HEP)
Z. Djurcic (ANL), G. Davies (Iowa State),
M. Sanchez (Iowa/ANL), M. Wetstein (U Chicago/ANL),
T. Xin (Iowa State)
Sunday, June 19, 2011
TIPP 2011
25
Backup SlidesDoes this fit in with the LBNE timeline?
• LBNE is not the only application we’re interested in:
• Collider physics: time-of-flight to determine flavor.
• Medical PET imaging
• Homeland security
• This project is just starting year 3 of a 3 year time-table. We have no intention or expectation for LBNE waiting for us.
• We’re not likely to be ready for the first detector and don’t want to interfere with any time-tables.
• Could be ready for upgrades or a second detector.
Sunday, June 19, 2011
TIPP 2011
26
Backup SlidesHow much do will these cost
But, keeping cost down is a major objective:
• Made from inexpensive materials.• Use industrial batch processes.• Inexpensive electronics, trying to reduce number of necessary readout channels.
too soon to tell…
In addition to the bottom-line cost of the detectors are secondary effects.
• Market impact.
• Possible savings on civil construction. Detector can be built closer to walls.
Cost/unit area is not the only relevant factor. Physics gains could be worth a little more.
Sunday, June 19, 2011
Top Related