Available on CMS information serverCMS NOTE 2003/000
October 2003
Spatial efficiency of the CMS Vacuum Phototriodes
N.Godinovic, I. Puljak, I. Soric,
Technical University of Split, Croatia
Z. Antunovic, M. Dzelalija
University of Split, Croatia
D J A Cockerill
Particle Physics Department, CLRC Rutherford Appleton Laboratory, Chilton, UK
Abstract
Specially developed vacuum phototriodes (VPTs) will be used to detect scintillation light produced in the lead tungstate (PbWO4) crystals in the end-cap sub-system of the electromagnetic calorimeter (ECAL) of the Compact Muon Solenoid (CMS) detector. The spatial uniformity of the VPTs photocathode quantum efficiency was measured by scanning their photosensitive area with collimated light source.
1Introduction
The operating conditions at the Large Hadron Collider (LHC) and the CMS design require dedicated photodetectors for the barrel part and for the end-cap part of the CMS electromagnetic calorimeter (ECAL) 1. In close collaboration with industry CMS has developed photodetectors for the ECAL readout. Avalanche photodiodes (APDs) were developed in collaboration with Hamamatsu Photonics for the ECAL barrel part. They are described in detail elsewhere 2. However, silicon devices cannot be sued in the end caps since the radiation dose and neutron fluences are much higher there than in the barrel section 1. Vacuum phototriodes, which are essentially single stage photomultiplier tubes have therefore been chosen for the end caps. Vacuum phototriodes (VPTs) were developed in collaboration with the Research Institute Electron (RIE) in St. Petersburg, Russia. A detailed description of the VPT characteristics can be found elsewhere 3. Both photodetectors have an internal gain and an optimized area to cope with the modest light yield of the PbWO4 crystal. VPTs with an external diameter of 22 mm have about the same order of total efficiency (quantum efficiency sensitive photocathode area) for the detection of PbWO4 light as 50 mm2 APDs with several times higher quantum efficiency. Spatial uniformity of the APDs quantum efficiency and gains is given in 4. Here we report about the measurements of the spatial uniformity of VPT photocathode performed at Technical University of Split.
Figure 1. illustrates schematically the VPT construction (left) and shows a photography of VPT (right). A planar, bialkali, semitransparent photocathode is deposited on the inner surface of glass faceplate. The anode is a fine metallic mesh located between the photocathode and the dynode. The dynode is the solid metal coated with the same material as the photocathode. In operation the photocathode is grounded, the dynode is at VD= 800 V to ensure a high electron secondary emission factor, and the anode voltage is approximately 200 V higher (VA=1000 V) to optimize the collection of secondary electrons emitted by the dynode 3.
Figure 1. A schematic view of a vacuum phototriode (right) and photography of a typical VPT (left)
Measurements of the photocurrent as a function of the position of the collimated light across the VPT photocathode quantify the level of (non)uniformity and allow to estimate its influence on the detector performances. They also provide information about the quality and stability of the production process. The execss noise factor F which quantify the contribution of a random nature of multiplication process to the width of the anode signal produced by the light pulse could be also related to the spatial (non) uniformity of the photocathode quantum efficiency.
2The setup for uniformity measurements
Photography of the setup for the VPT uniformity measurements is shown in Figure 2.
Figure 2. Photography of the setup for measurements of the VPT uniformity
The setup for VPT uniformity measurements is built at Technical University of Split. It consists of a moveable table with two degrees of freedom with maximum ranges of 25 mm. Stepper motors are mounted on micrometers screws, which move the table along two perpendicular directions. Specially developed LabView based software via a parallel port controls the step motors. LED light is transported by an optical fiber to the VPT photocathode. Fiber support is mounted upon the moveable table. The VPT is fixed and the collimated light spot scans across its faceplate. LabView based software allows to choose the different ways of scanning procedures and to vary the step size as a multiple of a minimum of 2.5 m. Voltage source (ORTEC 451) was used for the VPT biases. The photocurrent is measured by Multimeter Keithley 2000 as a voltage drop across a 1M resistor (0.1% precision) in series with the photocathode. The same software which control the moveable table, via an RS232 port readouts multimeter, and through another RS232 port the temperature is measured using a PT 100 temperature probe (Pico Technology 0.1 oC precision).
3 Spatial efficiency of the VPT photocathode
The measurement of the VPT photocathode spatial efficiency is performed operating VPTs in diode mode (no gain). The photocathode is grounded and the anode and dynode are operated at VAD=250 V. The photocurrent is calculated from the measured voltage drop across 1 M resistor in series with photocathode. The light spot of 500 microns, provided by blue LED with peak at 480 nm, scans the VPT faceplate. All VPTs undergo a photocathode visual inspection upon receipt from the producer. The majority of VPTs are produced with photocathodes which appear uniform under visual inspection. However, approximately two percent of VPTs have faceplates that appear to have missing photocathode regions, usually crescent shaped, at the edge of the faceplate and involving up to 10% of the faceplate surface area. A further two percent have up to 40% missing. Four VPTs (one good (137), three bad (354, 864, 728), under visual inspection) have been scanned to evaluate whether the observation from the visual inspection were correlated with non-uniformity in photocathode efficiency across the faceplate. The results of photocathode efficiency scans are shown in Figure 3. In Figure 4. the same data for VPT 137 and VPT 864 are presented as a contour plots to emphasize the difference between the VPT 137 with high level of uniformity and VPT 864 with very low level of uniformity.
Figure 3. Photocathode efficiency scan in step of 500 microns with light spot of 500 microns provided by blue LED with peak emission at 480 nm, a.) VPT 354, b.) VPT 137, c.) VPT 728, d.) VPT 864.
Figure 4. Photocathode efficiency scan in step of 500 microns with light spot of 500 microns provided by blue LED with peak emission at 480 nm, left VPT 137, right VPT 864.
Figure 5. Photography of the VPT 864 faceplate (right), the measured spatial efficiency by light spot of 500 microns scans across the VPT 864 faceplate (left).
The shape of the faceplate (non)uniformity measured by a light spot scan has the same shape as it is seen under visual inspection, Figure 5.
Distribution of the photocurrent coming from the photosensitive region (22 mm diameter) are shown in Figure 12. The mean value of the photocurrent and r.m.s. of its distribution are indicated on each plot. The level of (non)uniformity defined as the r.m.s. in percent of the mean value of the photocurrent distributions is given in Table 1, together with the percent of missing photocathode area estimated by visual inspection.
Figure 6. Distributions of the photocurrents across the photosensitive area of the VPTs faceplate and their mean value and r.m.s.
VPT barcode / missinig photocathode(visual inspection) / / /
137 / 10 % / 3.09 / 0.22 / 7.25
354 / 20 % / 3.7 / 0.54 / 14.6
864 / 25 % / 35.36 / 6.46 / 18.27
728 / 40 % / 26.9 / 10.15 / 37.73
Table 1. The percent of missing photocathode area estimated by visual inspection and spatial (non)uniformity of VPT photocathode measured by the light spot scan.
The measured level of the spatial efficiency of the VPT photocathode is correlated with the “missing” photocathode region estimated by visual inspection as can be seen in Figure 7. The measurement results demonstrate that a visual inspection of the VPTs photocathode is a reliable method to select devices whose spatial non-uniformity of the quantum efficiency will be bellow the level which could affect performance of the ECAL end-cap part.
Figure 7. The Correlation between the “missing” photocathode region under visual inspection and the measured (non)uniformity by the collimated light spot scan across the photocathode, line is a linear fit to the data points (circles).
It has been found that increasing the illuminated photocathode area increases execs nose factor in zero magnetic field 5. A possible explanation for this phenomena could be a contribution of the spatial non-uniformity of the photocathode efficiency. Thus it would be interesting to investigate the correlation between the spatial efficiency of the photocathode and the excess noise factor.
5Conclusion
Visual inspection is sufficiently sensitive method to select VPTs whose spatial non-uniformity of the phtocathode efficiency is bellow 10 %.
References
[1] CMS Collaboration, The electromagnetic calorimeter project, Technical Design Report, CERN/LHC 97-33, December 1997.
2] K. Deiters et al, Nucl. Instrum. Meth. A442 (2000) 193
3] B. Brown et al, Nucl. Instrum. Meth.A469 (2001) 29.
[4] N. Godinovic et al., Uniformity Measurements Across the area of CMS ECAL Avalanche Photodiode. (to be publishe as CMS note)
[5] N. A. Bayanov et al., CMS Note 1998/080
1