NSNI 2010

Energy and time response of CsI(Tl)-PIN detector for -rays and fission fragments
Y.K.Gupta, D.C.Biswas, P.Roy, R.G.Thomas, B.K.Nayak, R.P.Vind, A.Inkar, Bency John, Alok Saxena, and R.K. Choudhury Nuclear Physics DivisionBhabha Atomic Research Centre, Trombay, Mumbai 400085. E-mail: ykgupta@barc.gov.in Abstract
Pulse height and timing characteristics of CsI(Tl)-PIN detector have been investigated for Y-rays and fission fragments. Time resolution of the detector has been measured to be 1343.2 ns using Y-Y coincidence method. The light yield response for fission fragments (from 252Cf source) has been investigated. For the heavy mass fragments the light yield is seen to be more than that of light mass fragments at a given energy. It is observed that light output for both the mass groups increases linearly in energy interval 0.2 to 0.9 MeV/A.

Introduction
The CsI(Tl) PIN detector has been used for many charged particle detector arrays all over the world [1], for Y-ray measurement, X-ray imaging sensors, and in space applications because of following advantages: (i) CsI(Tl) scintilator coupled to photodiode, makes the compact geometry (ii)strong dependence of pulse shape on the type of ionizing radiation, because of different ratio of fast to slow components of the light output for different type of ionizing radiation. (iii) Large size crystal can be grown (iv) radiation damage is much less and (v) they are cheaper in comparison to conventionally used silicon detectors. Light output response of the detector for light charged particles and heavy-ions [2] have been studied in detail. However, the light output response for fission fragments has not been studied so far. In heavy-ion reactions separation of fission fragments from projectile like fragments using conventional methods, pulse shape discrimination (zero cross over) [3] or ballistic deficit is not possible [4]. Therefore, this detector has not been used for the measurement of fission fragments. In the present work, we have investigated the light output response of the detector for fission fragments produced from 252Cf source. The fission fragments were separated from -particles using time-of-flight (TOF) technique, where start was taken from a BaF2 detector triggered by prompt Y-rays. The energy dependence of the light output for fission fragments was studied by degrading the fragment energy by P-10 gas at different pressures. We have measured time resolution of the detector, by detecting two prompt Y-rays from 60Co source in coincidence. One of the Y-rays was detected by BaF2, and the other one by the CsI(Tl)-PIN detector. In the present
109

paper we report the details of these measurements. The CsI(Tl) detectors are planned to be used in 4 charged particle detector array, being developed at BARC-TIFR Pelletron/ Linac facility Mumbai.

Detector Setup detail

The CsI(Tl)PIN detectors have been supplied by M/S SCIONIX, Holland. The CsI(Tl) crystal has entrance surface area of 2525mm2 and of thickness 10.0 mm. Except the back surface, all other faces were covered with 1.2 m reflecting foil of aluminized Mylar. A Si-PIN photodiode manufactured by Hamamatsu Photonics is coupled to the back surface via a 252515 mm3 light guide. The photodiode type S3204-08 is 300 m thick with an active area of 1818 mm2. A rectangular collimator of opening area 2222 mm 2 was kept at the front surface of the crystal to avoid the edge effects. The signal readout was achieved by a charge sensitive pre-amplifier, attached to the photodiode. The low power dissipation ~120 mW of the preamplifier allows it to operate in vacuum without cooling. Signals from preamplifier had a dc off-set of 2-3V, which was eliminated by using a capacitor of 6F. The required +12V to the preamplifier was supplied using a battery which improves the signal to noise ratio in comparison to NIM module. Gain of the preamplifier was 6mV/MeV for alpha particles with output impedance of 50. Preamplifier signals were amplified and shaped using a spectroscopy amplifier (CEAN N968). The leakage current at bias voltage +35V was about 2-5nA. Pulse height and energy resolution for -particles (from 241Am-239Pu source) as a function of applied bias voltage and shaping time of spectroscopic amplifier are shown in Fig.1 and 2, respectively. As it is clear from Fig.1 and 2, applied bias voltage and shaping time should be at least +35V and 3.0s respectively, for full

NSNI 2010

light collection of the crystal. The energy spectrum for -particles (from 241Am-239Pu) and -rays (from 60Co) is shown in Fig 3. The energy resolution (FWHM) is ~180 keV for ~5 MeV - particles, and ~190 keV for ~1 MeV -rays at shaping time 3.0s and bias voltage to the photodiode +35V.

Time Resolution
The time resolution is one of the most important parameters of any nuclear radiation detector. We have measured time resolution of the CsI(Tl)-PIN detector by measuring - coincidence between a BaF2 and CsI(Tl)-PIN detector. The Electronics configuration for this study is depicted in Fig. 4. Distances of the CsI(Tl)-PIN and BaF2 detectors from 60Co source were 4.0 cm. and 1.0 cm. respectively. Shaping time of the spectroscopic amplifiers for both the detectors was kept at 3s. The start signal of TAC was from BaF2
110

NSNI 2010

through constant fraction discriminator (CFD), and stop signal was from CsI(Tl)-PIN through timing filter amplifier (TFA),CFD, and a fixed delay (25ns) as shown in Fig. 4. The differentiation and integration times of the TFA were kept at 200 ns and 500ns, respectively. The range of time to amplitude convertor (TAC) was kept at 1.0s. Data were collected in event-by-event mode using a CAMAC based multi parameter data acquisition system. TAC was calibrated using fixed delays of different magnitudes. A 2-dimensional plot of time (from TAC) versus energy of the -rays (of 60Co) is shown in Fig. 5. In Fig.5, two different group of -rays are observed. These two -groups are identified as the one, higher in time which are detected thorough CsI(Tl) crystal, and the lower in time which are reaching directly to the PIN diode and giving signal in photodiode. In Fig.5, apart from these two identified -groups, few events are scattered in the full range of TAC, due to uncorrelated - coincidence.

BaF2 detector, DtElect. is the electronic time resolution. In similar way we can write FWHM for PIN diode. The resolving time of the BaF2, has been measured in literature, and its value is ~3ns.

The X-projection (energy) of both the -groups are shown in Fig. 6a, and 6b. The energy spectrum for the CsI(Tl)-PIN- group (Fig. 6a), is very similar to the Fig.3, except reduced Compton background, because of - coincidence. The energy spectrum of PIN- group (Fig. 6b), is continuous, except for lower pulse heights where it has some contribution from uncorrelated -events, as discussed above. Fig. 7 shows the Y-projection of the 2-D plot shown in Fig.5. In this Y- projection, the peak corresponding to -rays from CsI(Tl) is broader than the peak corresponding to PIN diode, as expected. The FWHM of the CsI(Tl)-PIN timing peak can be written in the form as;

The electronic time resolution was measured using a pulse generator, which we obtained ~ 3ns. Thus, from eq. (1) the time resolution of the CsI(Tl)-PIN detector is 1343.2 ns. This large time resolution of the detectors is due to large decay time constant of the CsI(Tl) crystal as well as limited signal amplification by the PIN-diode. The decay time constant for -rays can be estimated from the present experiment. The time difference between CsI(Tl)-PIN- and PIN- peaks (Fig. 7) is a measure of average decay time constant of the CsI(Tl) crystal for the -rays. This value is ~205 134 ns, which is consistent with earlier reported values [1].

Light output for fission fragments (FFs)

The relative light yield has been measured earlier for light charged particles (Z<14) for the CsI(Tl) scintillation crystals [5]. It is observed that for same energy, the relative light yield is more, for the lighter mass particles than heavier mass particles which have more dE/dX. It is of interest to study the relative light yield of fission fragments as their dE/dX behavior is different in comparison to the light charged particles [6].
111

Where, t CSI(Tl)-PIN is the resolving time of the CsI(Tl)-PIN detector, tBaF2 is the resolving time of the

NSNI 2010

We have used 252 Cf as the source for fission fragments having two different mass groups. The most probable fragments are 104Mo and 144Ba having energies 104 and 80 MeV respectively. In this experiment, the energies were reduced to 0.2 - 0.5 MeV/A for heavy FFs and 0.4 - 0.9 MeV/A for light FFs. The energies were gradually degraded by using P-10 gas at different pressure. Moreover, the aluminized mylar foil of 1.2 m attached at the front face of the detector, acted as an additional degrader at each gas pressure. We used time of flight technique to get rid of -particles (6.1 MeV) emitted from the 252Cf source. The electronic setup was similar to that of Fig. 4 with flight path length of FFs of 27.9 cm. The Fig. 8 shows the 2-D plot of time versus energy of s, -rays, and FFs in vacuum (~1.0 10-3 mbar). Because of random emission, -particles are emitted in full range of the TAC, as shown in Fig 8. If we collect in singles, the ratio of -particles to FFs is ~20, but because of TOF measurement, this ratio reduces to ~0.2. The reduction in -intensity helps in determining the mean energy of the FFs after passing through P-10 gas. The pulse height spectra of FFs after subtracting -contribution are shown in Fig.9 for three typical gas pressures. The change in spectral shape with gas pressure is consistent with the earlier reported work [6]. In Fig.10, we have shown the measured light yields as a function of calculated incident energy of the FFs after taking into account energy loss in gas and aluminized mylar foil. The light yield varies almost linearly as a function of energy. The light yield is more for heavy FFs than light FFs at the same energy. This difference in light yield is because heavier FFs have less dE/dX than lighter ones. The light output response of the FFs is consistent with the behavior observed for light-charged particles. However, further understanding on the basis of the light emission mechanism in CsI(Tl) crystal would be necessary to interpret these results.

References
1. 2. 3. 4. 5. 6. S. Aiello et al Nucl. Instr. and Meth. A 369 (1996) 50. G. Viesti et al and Nucl. Instr. and Meth. A 252 (1986) 75. M. Moszynski et al Nucl. Instr. and Meth A 336 (1993)587 W. Skulski et al Nucl. Instr. and Meth A 458 (2001)759 C.J.W. Twenhofel et al Nucl. Instr. and Meth B 51 (1990)58 D. C. Biswas et al Nucl. Instr. and Meth B53 (1991)251

112