Quantum Efficiency measurement system for large area CsI photodetector

1
Quantum Efficiency measurement system for large
area CsI photodetector
Francesco Cusanno INFN Roma I Gruppo Sanita on
behalf of Hall A RICH collaboration

• TJNAF - Hall A RICH
• Evaporation system
• The Q.E. measurement system
• Measure principles and procedure
• UV source deuterium lamp and filters
  
• Reference PMT
• Collection charge chamber
• Results
• Thickness dependence
• Contamination and reconditioning
• Conclusion and future work

2
TJNAF - Hall A RICH
We built a proximity focusing RICH for Hall A at
Thomas Jefferson National Accelerator Facility
(TJNAF or Jefferson Lab)
Rich in Hall A
RICH specifications
190 cm
RICH Phase space in Hall A
190 cm
CsI PhotoCathode
PC1
PC2
PC3
64.0 cm
Cu Ni Au pad layers
40.3 cm
3
(No Transcript)
4
Evaporation system
10-6 mbar vacuum, 2 nm/s CsI deposition at T 60
ºC (CERN experts indications). Vacuum - heating
conditions start 15 24 h before evaporation. A
post-evaporation heat treatment is done for 12
hours.
Rotating mirror (CaF2)
120 cm
UV source box
110 cm
Photocathode
PMT
Collection chamber
Movement system
Crucible bars
5
Evaporation layout
Crucibles positions
64,40

• PhotoCathode crucibles plane distance 42 cm
• 4 mm Ni 1 mm Au support
• crucible quantity 0.8 g weight each one,
  corresponding to 320 nm thickness (expected and
  measured)

Expected thickness
Thickness (nm)
6
Measurement principle
The ratio A2/A1 Q.E.(CsI)/Q.E. (PMT), indeed
the wire chamber is in the vacuum (no charge
amplification) and the anode and grid voltage
allow to work in full collection regime for the
chamber (and the PMT dynodes and anode are
connected together to ground, so no charge
amplification for PMT too PMT supply is 78
V). A3 current (PhotoDiode in the optic box)
monitors the UV source stability.
7
Measurement procedure

• flow N2 to purge optical box (stable, gt 5 Torr
  overpressure, starting at least 12 h, before
  measurement)
• use batteries to reduce instruments electric
  ground floating
• we use a supply-meter box, connected to the PD,
  PMT and the collection chamber and to the
  picoamp-meter KEITHLEY 485
• Current measurements for any position
• Select the PMT position
• Read the PMT current
• Change the mirror position toward the CsI plane
• Read the chamber current
• Change the position on the CsI plane

8
UV source
Hamamatsu L2D2 lamp (C7860 power supply) 161 nm
spectral peak 3 filters (Acton Research
Corporation), 20 nm FWHM wide, centered at 160
185 200 nm)
Optical box
Filter rotating switch system, adjustable iris,
(6 position, one is Al disk)
Evaporator wall
N2 flow tube
9
Hamamatsu L2D2 Deuterium lamp
10
UV filters (Acton Research Co.)
Peak 158.80 nm spread 25.20 nm FWHM
Peak 198.40 nm spread 23.40 nm FWHM
11
Peak 185.80 nm spread 21.60 nm FWHM
CsI Q.E.
Q. E. ()
PMT- source convolution
CsI - source convolution
UV source spectrum (lamp filter)
PMT Q.E.
12
Q. E. ()
Q. E. ()
13
PMT
We use Electron Tubes Limited (ETL) 9402B PMT,
Q.E. is known by ETL (single PMT) datasheets.
ETL 9402B characteristics
Photocathode type Effective Cathode size Q. E. peak Entrance window
CsTe 23 mm 10 MgF
9402 S/No. 38 PMT Q.E.
Wavelength 160 nm 180 nm 190 nm 200 nm
Q.E. 7.3 8.8 8.9 8.3
14
Collection charge chamber
CsI Photocathode (0 V)
vacuum
2 mm
e-
133 V
Anode wires
2 mm
Grid (0 V)
Light spot
Anode wires collect electrons from the CsI plane
The grid wires stop electrons on the anode
Anode wires are 20 mm diameter, grid wires are 50
mm diameter, anode and grid wires are
crossed. Wire distance is 4 mm. Chamber
collecting area is 50 mm x 50 mm (light spot is
smaller, 10 mm).
15
Results

• Usally 24 row (1.5 cm distance), some cross
  checks

64.0 cm
Measurement layout
40.3 cm
We get a complete Q. E. map in this case 1 bad
spot ( 1 cm2). If bad results we can repeat the
evaporation (soon).
Good uniformity on all the PC (total spread
21.7 -24.4 , apart the bad spot average Q.
E. 23.7 ).
16
Correspondent pattern for all the wavelenghts
Average Q. E. at 200 nm 5.5 Same pattern at
185 nm too, Average Q. E. at 185 nm 11.6
17
Comparison of the results

• On 3 different PCs we have similar results, Q.
  E. total spread lt 10
• Results may be specific to the substrate -
  support

15 error bars are plotted

• Fit using Jlab experimental RICH data will
  supply a check on the direct Q. E. measurement

18
Thickness dependence
Crucibles weight 0 0 1.2 g 1.2 g
Expected thickness

• Higher range was expected, probably we lost
  partially CsI due to a mismatching between
  crucible volume and CsI weight

19
Air exposure and reconditioning
22 h. air exposure (19.5 ºC, 41 relative
humidity), 27 h. pumping has a small effect 12
h. heating restore about 1/3 of the loss.
Outgassing and reconditioning
25 not-pumping d., 0.25 mbar, reconditioning 12
h. pumping 14 h. heating 60ºC second (longer)
heating has no effect. 6 not-pumping d., 0.013
mbar, has similar effect than after pumping
reconditioning in the previous case (50 loss)
Possible interpretation is the outgassing of
organic particles from the substrate (it is
cleaned by organic solvent before evaporation).
20
Conclusion and future work

• The system performs reliable Q.E. evaluation and
  indicative absolute measurements on large
  photosensitive area, it can perform the measure
  since immediately after evaporation until
  immediately before the assembling of the
  PhotoCathode in the RICH chamber.
• Therefore it can monitor the eventual decrease
  of the Q.E., in case of delay between evaporation
  and detector assembling.
• Also the system can perform thickness, air
  exposure, post heat treatment dependence study.
• Preliminary thickness and air exposure tests
  seem to show that a thicker CsI has better Q.E.
  and it is less sensitive to air exposure. The
  Q.E. loss due to air exposure can be (at least
  partially) recovered by heating or pumping (dry
  gas flow).

21

• Extrapolating the thickness dependence results
  (adding the standard 300 nm result) it possible
  to expect higher Q. E. at higher thickness

Expected thickness
Crucibles weight 0 0 1.6 g 1.6 g
22
Support and temperature dependence

• Support thickness 7 mm Ni 2 mm Au

• Also we are interesting on T dependence study of
  CsI performances

23
Different T (and P) produces different growth