Low-pressure ultraviolet photon detector with TMAE gas photocathode

Nuclear Instruments and Methods in Physics Research A264 (1988) 235-250 North-Holland, Amsterdam

235

LOW-PRESSURE ULTRAVIOLET PHOTON DETECTOR WITH TMAE GAS PHOTOCATHODE Stan M A J E W S K I *, David F. A N D E R S O N , Penelope C O N S T A N T A - F A N O U R A K I S a n d Brian K R O S S Fermi National Accelerator Laboratory, Batavia, IL 60510, USA

George F A N O U R A K I S University of Rochester, Rochester, New York 14627, USA

Received 29 July 1987

Results of the study of the properties of a low pressure multistep avalanche counter with TMAE (tetrakis(dimethylamine)ethylene) as the photosensitive gas are presented. The optimization of parameters is discussed. Absolute gas ampfification, drift velocity and diffusion of drifting photoelectrons in a low pressure gas and the efficiency of single photoelectron detection are measured. The practical multihit capability and spatial resolution are explored in view of possible application in a Cherenkov ring imaging with multipad wedge-and-strip cathode readout with flash ADCs. The present limit in spatial resolution is due to diffusion of drifting electrons in a low-pressure gas. For applications where photons are detected in a high-rate background of charged particles, the low-pressure operation has many advantages over the atmospheric pressure detectors.

1. Introduction In this paper we present the results of the second stage of a study of a low-pressure TMAE photodetector. In preliminary work with this type of a detector, filled with a mixture of isobutane and TMAE at 20 Torr, it was demonstrated that an efficient, stable and photonfeedback-free operation is possible [1]. Also a test of the single wedge-and-strip pad readout proved its usefulness in the simultaneous determination of both X and Y coordinates of a photoelectron in such a detector. Previously Charpak and Sauli [2] demonstrated that the muhistep structure solves, in an efficient and elegant way, the photon-feedback problem when working with TMAE as a photon converter in atmospheric pressure UV photon detectors. They have also shown the multihit imaging capability of such a detector with a CCD readout, while avoiding long drift regions of the TPCtype designs developed for Cherenkov ring imaging detectors at CERN (DELPHI) [3] and SLAC (SLD) [4]. The main advantage of the low-pressure operation is in the relative insensitivity of a detector to charged particles crossing its active volume. This should lead to a much more stable operation in a high-rate a n d / o r high ionization environment such as in the high energy heavy ion experiments. The combination of the muhistep structure plus * Present address: Department of Physics, University of Florida, Gainesville, FL 32611, USA. 0168-9002/88/$03.50 © Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)

low-pressure operation results in many other advantages which will be discussed in the second section of this paper. In an independent continuation of the original idea the group at Weizmann Institute proposed an application of low-pressure Cherenkov ring imaging detectors in the HELIOS (CERN) heavy-ion experiment and actively pursued the development of the technique [5]. Two Cherenkov detectors are under construction at Weizmann and Heidelberg [6].

2. Theoretical aspects of operation 2.1. Detector structure and mode of operation

The schematic structure of the low-pressure multistep chamber which has been tested for use in the detection of UV photons is shown in fig. 1. An UV photon entering through the external window is absorbed by a TMAE molecule and in the photoelectric process an electron is produced with quantum efficiency of more than 50% [7] (fig. 2). This photoelectron drifts in the constant electric field of the conversion region and travels with high efficiency through the wire mesh electrode M1 into the preamplification (PA) region. In the strong electric field of the preamplification region, with the geometry of the parallel-plate type, many secondary electrons are produced in the avalanche multiplication process. A fraction f of the electron cloud,

S. Majewski et aL / Low pressure ultraviolet photon detector

236 UV PHOTON (DIM~3 IN MM)

~ QUARTZ WINDOW

~3,0

i

~,//"//7//// /

:

CONVERSIOI~REGION

1

52 J

5.0

PA

45

I

][

5.0

DRIF~ ~dREGION

Ai

¢

LA2

M3

~

-.. /

1.6

WIRE MESHES

,,~

/

WEDGE AND STRIP ~-A- PRINTE O BOARD ELECTRODE (ANODE)

Fig. 1. The multistep detector structure used in the present study (details in the text). approximately equal to the ratio (about 5%) of the electric fields on both sides of the mesh M2, is transferred to the drift region. On arrival at the double amplification volume A 1 - A 2 , a second avalanche multiplication takes place in the region A1 and the thin region A2 is used mainly as a collection region for the electrons produced in A1, though some additional multiplication also takes place in A2. Because the A2 region

)

I00

z

80

~

60

z < 2

40

z o

20

0

is thin, the induced electron pulse on the last anode, electrode A, is short with a fast rise time. Ions are too slow in the parallel-plate structure to have an important contribution to the fast component of the pulse. However, due to the much higher reduced electric fields E / p at working pressures of 10 20 Torr, ions are swept away much faster (in 2 - 4 ,us) from the active volume than in the traditional multiwire structures at atmospheric pressures, thus preventing space charge buildup. In the case of a high avalanche multiplication factor in the A2 region, it would be difficult to filter out the positive ion component because of high mobility of positive ions in high E / p fields. This is an additional reason to keep the multiplication factor low in the A2 region. Because of the high ratio, about 10-30, of the electric fields in the A1 region (high field) and the drift region (low field), only a correspondingly small fraction (3-5%) of these positive ions get into the drift region. An even smaller fraction (below 3 × 10 -3) of positive ions migrates up into the conversion region. The double-step amplification scheme provides a very elegant and efficient solution to the well-known photon-feedback problem in photon detectors with T M A E as an active photoagent [8]. Because of the low vapor pressure of T M A E at room temperatures [9] the mean free path of UV photons in the gas is about 15-20 mm [8]. This defines the necessary thickness of the absorption/conversion region (52 mm in our case; see fig. 1) and it also results in a long range of "parasitic" UV photons produced during the avalanche multiplication process. By splitting the multiplication region into two regions of much lower multiplication factors, one

\

CH4

L

Z r~ C

m <

0

I

1400

1500

I

1600

1700

1800

1900

2000

'7100

2100

WAVELENGTH (~) Fig. 2. Absorption curves of some common quench gases superimposed on the TMAE quantum efficiency curve and a quartz window effective absorption (with reflection factor included, for the right angle of incidence). The absorption curves of CO 2, 1-C4H10, C 2 H 6 and c a 4 a r e shown for 10 cm thick gas layers at 1 arm gas pressure, with our results for the 8 ¢m layer of dimethylether (DME) at 1 atm pressure and methylal at 100 Torr pressure.

S. Majewski et al. / Low pressure ultraviolet photon detector avoids copious production of photons in both of these regions in the first place, and secondly, only feedback photons reabsorbed in the conversion region get the full double-step multiplication. One advantage is in a built in time delay in the detector allowing for preparation of a trigger pulse. The total multiplication factor can be written in the form: GXOX = L'_pAGp.4fpA DGmFA1-A2GA2

an induced charge is a d (--- In G) times lower than the produced charge Q = Gq. This, in turn, means that the lower the amplification factor in the last region, the smaller is this reduction factor. For example, provided that the mechanical restrictions allow it, by splitting the 6.0 mm amplification region into two regions of the same electric field, with d I = 5 mm and d 2 = 1.0 mm, and for a total effective multiplication factor of 104, one gets a fast electronic component enhancement factor of

= 0.6GpAfGA1GA2 = 0.6GpAfGA, where f ' s are the electron transfer factors between the corresponding regions, fc-pA = 1, fPA-D = f , fM-A2 = 0.6 for about the same electric field in both regions [10], G ' s are the multiplication factors in the two multiplication regions (G A is split between A1 and A2 regions, as was described above). Background photons produced in the second stage are mostly absorbed in the middle buffer drift region (45 mm thick in our case, see fig. 1) and are discriminated against because of a low effective multiplication factor G 2 = 0.6G A. Electrons from photons converted in the drift region will be, on the average, multiplied less by a factor of: G 2 / G T o T = GpAf, which for typical GpA and f values of 2 × 103 and 5%, respectively, is equal to 100. Thus, the electronic discrimination against these photoelectrons is quite easy and they do not contribute to background [2]. On the other hand, assuming that the number of UV photons produced in the avalanche process is proportional to the multiplication factor, as confirmed experimentally in many scintillating gases [11], there are proportionally fewer background photoelectrons produced in the conversion region. In our example, the production will be down by a factor of 25-50, compared with the case of a one step amplifying structure with a total amplification factor of 5 × 104-105. In fact, when comparing with the multiwire chamber structure, there is a further improvement because a larger number of more energetic U V photons is expected to be produced per ionization electron in the region of stronger electric fields around a wire of a multiwire chamber, than in weaker fields of the parallel plate structure. An additional advantage of a double structure such as A 1 - A 2 is in an increased fast electronic signal component compared with the case of either a single parallel-plate structure or a multiwire chamber structure. If an electronic charge q = Ne enters the parallel-plate avalanche region, the induced fast component of the charge pulse on both electrodes is given by:

Q,,,