1 Results from Silicon Photo-Multiplier neutron irradiation test R. Faccini Sapienza Università di Roma and INFN, Sezione di Roma, Italy D. Pinci INFN, Sezione di Roma, Italy E-mail: davide.pinci@roma1.infn.it W. Baldini, R. Calabrese, G. Cibinetto, A. Cotta Ramusino, R. Malaguti and M. Pozzati. Unversitá degli Studi and INFN Ferrara, Italy M. Angelone and M. Pillon ENEA, Frascati, Italy Silicon photo-multipliers, often called SiPM, are semiconductor photon detectors built from a square matrix of avalanche photo-diodes on common silicon substrate. SiPM have been proposed for several different applications in High Energy Physics, in particular where a large detection granularity and a high photon detection efficiency is needed. In this paper the results of a radiation hardness test performed at the Frascati Neutron Generator are presented. Several SiPMs of different manufacturers have been irradiated integrating up to 7 10 10 1-MeV-equivalent neutrons per cm 2. For the first time, their performance have been recorded during the neutron irradiation and a gradual deterioration of their properties was found to happen already after an integrated dose of the order of 10 8 1-MeV-equivalent neutrons per cm 2. Keywords: Silicon Photmultipliers; Neutron irradiation. 1. Introduction We are studying the possibility of using SiPM 1 as light detectors for the extruded scintillators of the Instrumented Flux Return of the experiment running on the super flavor factory proposed at Laboratori Nazionali di Frascati. 3 Four meter long scintillator bars, equipped with wavelength
2 shifting fibers are foreseen that have to provide 1 ns time resolution in order to measure the track position along the bar. One of the main issues in using SiPMs is the high neutron flux expected in the IFR region. From very preliminary calculations one can expect 10 4 10 5 n/s cm 2. 2. The Frascati Neutron Generator The Frascati Neutron Generator 2 (FNG) uses a deuteron beam accelerated up to 300 kev impinging on a deuteron target to produce a nearly isotropic 2.5 MeV neutron output via the D(d,n) 3 He fusion reaction. The beam current at the target can be regulated up to 1 ma resulting in a maximum neutron production rate of 5 10 8 neutrons on the whole solid angle per second. During our tests, the neutron flux was monitored online by acquiring the rate of a calibrated liquid scintillator (NE213) provided by the FNG-staff. The NE213 was used with pulse shape discrimination to reject gamma-ray events. 3. Measurement set-up Six devices (five 1 1 mm 2 and one 2 2 mm 2 ) produced by the IRST 4 (SiPM in the following) and four produced by the Hamamatsu 5 (MPPC in the following) have been tested with neutrons. Depending on the distance from the production point, in four days of test they integrated the total 1-MeV-equivalent doses shown in Tab.1. Table 1. Device positions and total integrated doses. Device x y z tot. int. dose a (mm) (mm) (mm) (10 10 n eq/cm 2 ) SiPM #4 5.0 0.0 3.0 1.25 SiPM #6 3.0 1.3 3.0 2.71 MPPC #2 1.0 0.0 3.0 7.32 SiPM #8-1.0 0.0 3.0 7.32 MPPC #1-3.0 0.0 3.0 3.07 SiPM 2 2-5.0 0.0 3.0 1.25 Note: a The total integrated Dose equivalent to 1 MeV neutrons on silicon was evaluated with a MC and from. 6 4. Online Measurements The drawn currents and the dark rates of the SiPMs supplied with a voltage of 33 V (within the working region) were recorded during the neutron irradi-
3 ation. Figure 1a shows that the current drawn by the SiPM increases up to Fig. 1. For SiPMs in different positions : increase factors of the current drawn by the SiPM as a function of the accumulated dose and zoom of the drawn current increase factors for low integrated doses. a factor 30 for the most irradiated device (7.32 10 10 n eq /cm 2 ). The effects of the different neutron fluences are not visible at the level we operated. No significant recovery effects appeared after a whole night without neutron irradiation: the absolute value of the current and the increase rate, once the flux was back on, didn t change. A zoom for low doses, Fig. 1b, shows that the currents are quite stable for doses of the order of 10 8 n eq /cm 2 and then, after a jump of a factor 2 or 4, they start to increase. By means of a custom board, the dark counting rate of each device was continuously monitored. The increase factor of the SiPM dark counts, Fig. 2a, reaches the value of Fig. 2. For SiPMs in different positions: increase factors of the dark counting rate as a function of the accumulated dose and zoom of the dark counting rate increase factors for low integrated doses. 300 for the most irradiated device. The behavior of the dark count rates of the SiPM exposed to a lower irradiation seem to show a slower increase (Fig. 2b). The neutron beam has been paused several times in order to per-
4 form low voltage scans and to measure the effects on the dark currents and on the dark counting rates at different bias values for different integrated doses. The current behavior in the low voltage scan changed rapidly with!+#.!"#!#!'& Drawn current (ma)!+# &,'$ &,' &,&- &,&* &,&( &,&$ & :5<985 4567894/:5;3/94 /&,(%/'&1'&23 /',-%/'&1'&23 /#,&$/'&1'&23 /#,&$/'&1'&23 /),'0/'&1'&23 After 3.0 10 10 neq/cm 2 After 1.9 10 10 neq/cm 2 After 0. After 5.2 10 10 neq/cm 2 $% #& #' #$ ## #( #) #* Supply voltage!"# (V) -i2:-t2! Drawn Current -i2 (ma) 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0-3 10 before 1.85 10^10cm^-2 4.26 10^10cm^-2 After 4.3 10 10 neq/cm 2 After 1.8 10 10 neq/cm 2 68 68.5 69 69.5 70 70.5 71 71.5 72 Supply Voltage -T2 (V) Fig. 3. Behavior of the drawn current as a function of the low voltage applied for a SiPM and a MPPC for different integrated doses. the integrated dose as it is shown in Fig. 3. 5. Off-line Measurements The SiPM and the MPPC have been tested with cosmic rays before and after the neutron irradiation and the charge spectra obtained are shown in Fig. 4 and Fig. 5. After the neutron irradiation, a lower gain was found 7 10 10 neq/cm 2 Pedestals 7 10 10 neq/cm 2 Fig. 4. Pedestal (top) and cosmic signal charge spectra (bottom) for a SiPM downstream of a 150 cm long WLS fiber for different integrated doses. and the noise pedestals are much broader. Both these effects lead to an important reduction of the detection efficiency from more than 95% to about 70%. A non evident dependence of the performance deterioration on the integrated dose was found.
5 Pedestals Fig. 5. Pedestal (top) and cosmic signal charge spectra (bottom) for a MPPC downstream of a 150 cm long WLS fiber for different integrated doses. 6. Conclusion Several Silicon Photo-Multipliers have been exposed to an intense neutron flux integrating up to a total dose of 7.32 10 10 n eq /cm 2. Their performance were studied before, during and after the irradiation. The drawn currents has been found to increase up to a factor 30 while the dark counts up to 300. The detection efficiency measured with cosmic rays, higher than 95% before the irradiation, was evaluated to be around 75% after. From the measurements shown we conclude that Silicon Photo-Multipliers performance would start deteriorating after few hours of running on a super flavor factory. References 1. B. Dolgoshein et al., Nucl. Instr. and Meth. A563 (2006) 368, and references therein. 2. M. Martone et al., Frascati Neutron Generator Proc. SPIE Vol. 2339, p. 208-212, International Conference on Neutrons and Their Applications, George Vourvopoulos; Themis Paradellis; Eds. 3. M. Bona et al., SuperB A High-Luminosity Heavy Flavour Factory Conceptual Design Report INFN/AE - 07/2, SLAC-R-856, LAL 07-15. 4. Information can be found at http://www.itc.it/irst/renderer.aspx?targetid=111. 5. Information can be found at http://www.hamamatsu.com. 6. Standard practice for Characterizing Neutron Energy Fluence Spectra in Terms of an Equivalent Mono-energetic Neutron Fluence for Radiation Hardness Testing of Electronics, ASTM E 722-93.