- Email: [email protected]
Characterization of silicon photomultipliers for new high-energy space telescopes
Nuclear Inst. and Methods in Physics Research, A (
)
–
Contents lists available at ScienceDirect
Nuclear Inst. and Methods in Physics Research, A journal homepage: www.elsevier.com/locate/nima
Characterization of silicon photomultipliers for new high-energy space telescopes K. Lacombe a,b, *, J. Knödlseder a,b , B. Houret a,b , T. Gimenez c , J.-F. Olive a,b , P. Ramon a,b a b c
IRAP, CNRS, 9 avenue du colonel Roche, 31028 Toulouse Cedex 4, France IRAP, University of Toulouse, OMP-UPS, 118 Route de Narbonne, 31062 Toulouse, France CNES, CST, 18 Av. Edouard Belin, 31401 Toulouse Cedex 9, France
a r t i c l e Keywords: SiPM Space telescope Qualification Characterization Performance
i n f o
a b s t r a c t Photon detection is a major issue of high-energy astronomy instrumentation. One classical setup that has proven successful in space missions is the combination of photomultiplier tubes (PMTs) with scintillators, converting incoming high-energy photons into visible light, which is converted in an electrical impulse. Although being extremely sensitive and rapid, PMTs have the drawback of being bulky, fragile, and requiring a high-voltage power supply of thousands volts. The silicon photomultipliers (SiPM) appear to be a promising alternative to PMTs in many applications such as small satellites. We have started a R&D program to assess the possibility of using SiPMs for space-based applications in the high-energy astronomy domain. We present here the results of our characterization of SiPMs coming from several manufacturers. Each detector has been tested at low temperature and pressure to study its performance in a representative space environment. For this, we developed a dedicated vacuum chamber with a specific mechanical and thermal controlled system. Once dark current, dark count rate and PDE were measured, we made irradiation tests on two selected detectors to understand the susceptibility of SiPM to radiation damage. Finally, we aim to perform thermal cycling and mechanical tests on detectors and study their coupling to scintillators, in parallel with their space qualification. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Silicon Photo-Multipliers (SiPMs) appear to be the next generation of detectors to replace traditional Photo Multiplier Tubes (PMTs) in many high-energy astrophysics applications [1]; they can be used either for direct light detection or to readout a scintillator detector that converts incoming high-energy photons into visible light. The technical characteristics of SiPMs are powerful arguments for using them on future space telescopes: their use avoids operating high voltage supplies, ensuring robustness and reliability. Moreover, as they are insensitive to magnetic fields, we can use them in high field environments instead of PMTs. Their high Photon Detection Efficiency (PDE) would enlarge the overlap in cosmic-ray energies detected with ground-based facilities. At last, their low power consumption decreases the thermal dissipation. Consequently, one option would be to associate inorganic scintillators (LaBr3 , CeBr3 , . . . ) to SiPM detectors [2,3], in order to form a calorimeter for a space instrument to determine the energy of the incident photons [4]. The short decay time allows measurement of fast coincidences, which opens the way of measuring the time-of-flight of
the photons, a crucial parameter for reducing instrumental background noise in space environment. In 2015, we started a R&D study with a measurement campaign on three commercial devices; we selected SiPM references from Hamamatsu, SensL and Ketek (respectively S12572-050C, SB30035-SMT and PM3350-B63T75S-P4) to characterize their performance at room temperature under atmospheric pressure, and especially at low temperatures and pressures [5]. Then we studied their susceptibility to radiation, with a particular emphasis on the effects of protons, aboard a satellite, as well as to hostile phenomena such as Van Allen radiation belts. An irradiation test was carried out at room temperature on non-polarized detectors on the UCL cyclotron bench in Louvain (Belgium). Due to their good performance, the Ketek and SensL detectors have been chosen to be irradiated under a fluence of 2.1011 protons/cm2 by protons of 50 MeV [6]. This level would allow a variety of orbits, from LEO to GEO. Afterward, we focused our studies on the effect of annealing after a proton irradiation to evaluate a potential recovery. Lately, we are in
* Corresponding author at: IRAP, CNRS, 9 avenue du colonel Roche, 31028 Toulouse Cedex 4, France.
E-mail address: [email protected] (K. Lacombe). https://doi.org/10.1016/j.nima.2017.11.005 Received 28 September 2017; Accepted 3 November 2017 Available online xxxx 0168-9002/© 2017 Elsevier B.V. All rights reserved.
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
A/cm2 at room temperature. In a nutshell, the breakdown voltage slightly decreases with the temperature. In fact, the breakdown voltage of a junction increases with a higher temperature, in relation to a decrease of the values of the ionization coefficients. 3.2. Dark count rate As the dark current, the dark count rate (DCR) is particularly dependent on temperature; we show here that, in the case of SensL SiPM, it has been reduced by a factor of 3 at the overvoltage of 20%, by making it almost independent of this parameter. The same tendency is observed for the Ketek, which gets a decreasing of 5 by lessening of almost 40 ◦ C at OV = 25% (Fig. 5).
Fig. 1. Pictures of SiPM detectors SensL & Ketek observed under binocular (Left picture : 4774 cells of 35 μm pitch/Right picture : 3600 cells of 50 μm pitch).
3.3. PDE and single photon at low temperature progress to study two new references of SensL and Ketek (Figs. 1 and 2), to rise the statistics and develop a first matrix prototype of several chips on the same substrate [7].
3. Results
Then, to study the PDE, we first measured the pixel capacitance C for each detector. Then, we calibrated the optical chain to know the input photon flux, thanks to a photodiode power sensor in the wavelength range of the R&D from 350 nm to 800 nm [9,10]. Afterward, we measured the SiPM output voltage as a function of the wavelength. Our measurements were made for a bias voltage of 32 V for Ketek and 29 V for SensL SiPM for the temperatures of +23 ◦ C and −22 ◦ C. These results are consistent with the PDE values of Ketek and SensL manufacturers at room temperature. At the low temperature of −20 ◦ C, we observe that the PDE is better than at higher temperature for both SiPM detectors, due to the lower DC, by gaining until 25% at 420 nm peak (Fig. 6). An example of pulse area spectrum is shown in Fig. 7; we can notice that the different peaks are clearly distinguishable thanks to the low dark count rate, which reduced the peak to valley ratio (a diminution of 40 ◦ C seems to be appropriate to obtain such a good quality of peaks discrimination) [11]. The gain of a SiPM corresponds to the mean number of charge carriers that a single charge generates during the avalanche process in the depletion region of the detector. We calculated it by taking the difference between two consecutive peaks of the pulse area distribution (red histogram in Fig. 7).
3.1. Dark current as a function of temperature
3.4. Effect of high proton irradiation on SiPM
In this section, we present some results on dark current (DC), a source of noise intrinsic to the SiPM as a function of temperature. Dark Count Rate measurements supplement it, for both Ketek & SensL detectors. Next, we report Photo Detection Efficiency (PDE) and the pulse area distribution to observe the single photon. Thus, one way to suppress dark current is by cooling the detector to lower temperatures. We are studying this thermal dependence to find a temperature range ensuring the best performance of SiPMs. We measured dark current as a function of overvoltage [8] and evaluated the break down voltage at +23 ◦ C and −22 ◦ C at ∼10−4 mbar for the Ketek and SensL SiPMs (Fig. 4). For the Ketek detector, we found a breakdown voltage temperature coefficient of dVBD /dT = 21.6 mV/◦ C which is very close to the value provided in the manufacturer datasheet. For the SensL detector our measurements let us calculate an identical temperature coefficient of 21.6 mV/◦ C as expected. Moreover, at 20% overvoltage, the dark current density is about 2.5 × 10−7 A/cm2 at −22 ◦ C instead of 7 × 10−6
The dark current is a source of noise intrinsic to the SiPM detector, generated even though the cells are not exposed to light. This dark current is due to the thermal excitation of electrons into the conduction band, and hence has a strong temperature dependence. Cooling the detectors to low temperatures reduces this noise. We measured the dark current and evaluated the breakdown voltage of two detectors before and after irradiation at three temperatures under different pressure settings. As the Ketek detector is plugged on board, the SensL SiPM is directly stuck on it, going through a pico-ammeter connected to the output of the SiPMs without any front-end electronics to measure the dark current. Only the power supply has been applied to polarize the SiPM placed in the dark chamber without any light source. Our measurements show that the DC considerably increased for both SiPMs after irradiation, whereas the breakdown voltage remains quite the same. The temperature dependency is also stable; indeed, the dark current slightly decreases with the temperature with the same
2. Setup and configuration We developed a dedicated thermal vacuum test bench in order to characterize the SiPMs in an environment relevant for space applications. We designed a 30 dm3 vacuum chamber equipped with Peltier coolers to operate at low temperature (−22 ◦ C ± 0.5 ◦ C) and low pressure (∼10−4 mbar). A single mechanical support holds a specific low-noise electronics board containing both four detectors (2 SensL & 2 Ketek references) and three thermal probes mounted as SiPM, to ensure thermal stability and avoid EMC issues (Fig. 3). Firstly, we used a halogen source, covering a 350 to 800 nm wavelength range, going through a monochromator, for PDE measurements. Secondly, to observe the pulse area distribution and then obtain the gain value, a LED driver generates a light pulse at 400 nm, illuminating each SiPM via external and internal optical fibers.
Fig. 2. Table of features of the last SiPM detectors (ongoing study), provided in the manufacturers datasheets.
2
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
Fig. 3. Left picture: Specific test bench developed for R&T SiPM; Right picture: Low noise electronics board to test 4 detectors in the same conditions of pressure and temperature.
Fig. 4. Dark current versus bias voltage at 3 temperatures (+23 ◦ C, −15 ◦ C & −22 ◦ C) of Ketek detector (main graph) and SensL detector (inseted graph). Fig. 6. Photon Detection Efficiency as a function of wavelength at 2 temperatures (+23 ◦ C & −22 ◦ C) of Ketek (green curves) and SensL detector (red curves).
Fig. 5. Dark Count Rate versus Overvoltage at 2 temperatures (+23 ◦ C & −15 ◦ C) of both Ketek and SensL detectors.
coefficient for both curves before and after irradiation (Fig. 8). Besides, the voltage breakdown knees remained identifiable after irradiation
Fig. 7. Pulse area distribution of Ketek SiPM at low temperature.
3
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
an increase in dark current, as the bulk damage that creates also an increase in both after-pulse and cross-talk effect [12]. 4. Conclusion On the one hand, we reached the expected performance at room temperature, and on the other hand, our measurements showed a better stability and decreased values of intrinsic noise of SiPM at low temperature. So far we have identified no issues with operating under lowpressure conditions, and the detectors and its associated electronics we designed, called ‘‘SiPM board ’’, behave as if they were operated under atmospheric pressure. This conclusion is the key for space applications. The first step of our study of three different manufacturers (Ketek, SensL & Hamamatsu), has shown great performance in terms of low dark current and PDE and covers a large variety of parameters such as dark count rate and gain. Afterward, only two SiPMs have been selected to be tested in irradiation campaign to assess their resistance to a harsh space environment and check if their good performance is maintained under space constraints. The results here described showed that the dark current increased substantially after proton irradiation for both detectors. However, despite this overall change in scale, the shapes of I–V curves remain identical and indicate that the breakdown voltage is not really impacted by the proton radiation. We point out that the radiation damage does not impact the detector behavior neither in function of the temperature nor in function of the operating voltage. Then, we investigated whether an annealing test would allow recovering the detector properties measured before irradiation and our first results are promising but not significant. At a later stage, we plan a second radiation test campaign to consolidate our initial observations, with the study of the impact of lower fluence of protons and secondly Xrays. In addition, we plan to expose the SiPMs to rapid temperature cycling in the range of −55 ◦ C to +60 ◦ C and also to conduct a longterm temperature storage test to evaluate any effects of aging on the SiPM characteristics. Besides, we aim to characterize a dozen detectors (SensL and Ketek) to get more statistics and, in parallel, to qualify them via mechanical tests on next phase and to study the coupling with the advantageous LaBr3 scintillator offering a great light output that allows a good determination of the energy of the incident photon. Finally, the selected detector and its associated front-end electronics will undergo extended testing to reach at the end a Technology Readiness Level of 6 (subsystem model demonstration in relevant environment).
Fig. 8. Evolution of dark current after a high proton irradiation as a function of bias voltage at 2 temperatures (+25 ◦ C & −22 ◦ C) of Ketek SiPM (non-polarized detector during the irradiation test).
Acknowledgments We thank CNES staff for it support, Ketek and SensL companies for their great responsiveness, and for the quality of the detectors provided. We acknowledge the UCL staff for their support during the proton irradiation test. We are grateful to all participants at the NDIP conference in 2017 for their presentations and their expertise.
Fig. 9. Effect of annealing on dark current of SensL SiPM after being irradiated (green curve at +140 ◦ C).
References although they became more rounded. We can notice that, in the working range, the dark current at room temperature increases by 15 to 20 times (for SensL and Ketek SiPM respectively). At low temperature, this factor rises strongly to reach about 103 times in the case of Ketek detector whereas the SensL one multiplies by around 300, what would suggest its better irradiation resistance. Moreover, we conducted an annealing test to investigate whether these trends are irreversible (Fig. 9). In conclusion, radiation produce additional defects inside the band gap making it easier for electrons to reach the conduction band and generate increased thermal noise. Two types of radiation damages may happen in silicon detectors: the surface damage due to Ionizing Energy Loss (with the accumulation of charges in the oxide SiO2 and traps at Si/ SiO2 interface) and the bulk damage due to Non-Ionizing Energy Loss (a displacement damage, build-up of crystal defects). Several effects can occur directly observed on SiPM detectors: the surface damage causes
[1] A. Ulyanov, L. Hanlon, S. McBreen, S. Foley, D. Byrne, Study of silicon photomultipliers for the readout of scintillator crystals in the proposed grips 𝛾-ray astronomy mission, Proc. Sci. (2013). [2] W. Drozdowski, P. Dorenbos, A. Bos, G. Bizarri, A. Owens, F. Quarati, CeBr3 scintillator development for possible use in space missions, IEEE Trans. Nucl. Sci. (2007). [3] T. Szcześniak, M. Grodzicka, M. Moszyński, M. Szawlowski, D. Wolski, J. Baszak, Characteristics of scintillation detectors based on inorganic scintillators and sipm light readout, Nucl. Instrum. Methods Phys. Res. A (2012) 91–93. [4] A. Ulyanov, L. Hanlon, S. McBreen, et al., An integral view of the high-energy sky (the first 10 years), in: Proceedings of the 9th INTEGRAL Workshop, Paris, 2012. [5] http://www.ketek.net/products/sipm-technology/device-parameters/. [6] Y. Qiang, C. Zorn, F. Barbosa, E. Smith, Radiation hardness tests of SiPMs for the JLab Hall D Barrel calorimeter, Nucl. Instrum. Methods Phys. Res. A (2012) 234–241. [7] http://sensl.com/request-c-series-ds/. [8] R. Pagano, et al., Dark current in silicon photomultiplier pixels: Data and model, IEEE Trans. Electron Dev. (2012).
4
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
[11] F. Acerbi, et al., Characterization of Single-Photon Time Resolution: From Single SPAD to Silicon Photomultiplier, IEEE, 2014. [12] J. Been, Effects of nuclear radiation on a high-reliability silicon power diode, NASATN-D-7423, 1973.
[9] H. Miyamoto, M. Teshima, B. Dolgoshein, et al., SIPM development and application for astroparticle physics experiments, in: Proceedings of the 31st ICRC, Lodz, 2009, p. 1320. [10] C. Jackson, et al., High-Volume Silicon Photomultiplier Production, Performance, and Reliability, SPIE, 2014.
5
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
)
–
Contents lists available at ScienceDirect
Nuclear Inst. and Methods in Physics Research, A journal homepage: www.elsevier.com/locate/nima
Characterization of silicon photomultipliers for new high-energy space telescopes K. Lacombe a,b, *, J. Knödlseder a,b , B. Houret a,b , T. Gimenez c , J.-F. Olive a,b , P. Ramon a,b a b c
IRAP, CNRS, 9 avenue du colonel Roche, 31028 Toulouse Cedex 4, France IRAP, University of Toulouse, OMP-UPS, 118 Route de Narbonne, 31062 Toulouse, France CNES, CST, 18 Av. Edouard Belin, 31401 Toulouse Cedex 9, France
a r t i c l e Keywords: SiPM Space telescope Qualification Characterization Performance
i n f o
a b s t r a c t Photon detection is a major issue of high-energy astronomy instrumentation. One classical setup that has proven successful in space missions is the combination of photomultiplier tubes (PMTs) with scintillators, converting incoming high-energy photons into visible light, which is converted in an electrical impulse. Although being extremely sensitive and rapid, PMTs have the drawback of being bulky, fragile, and requiring a high-voltage power supply of thousands volts. The silicon photomultipliers (SiPM) appear to be a promising alternative to PMTs in many applications such as small satellites. We have started a R&D program to assess the possibility of using SiPMs for space-based applications in the high-energy astronomy domain. We present here the results of our characterization of SiPMs coming from several manufacturers. Each detector has been tested at low temperature and pressure to study its performance in a representative space environment. For this, we developed a dedicated vacuum chamber with a specific mechanical and thermal controlled system. Once dark current, dark count rate and PDE were measured, we made irradiation tests on two selected detectors to understand the susceptibility of SiPM to radiation damage. Finally, we aim to perform thermal cycling and mechanical tests on detectors and study their coupling to scintillators, in parallel with their space qualification. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Silicon Photo-Multipliers (SiPMs) appear to be the next generation of detectors to replace traditional Photo Multiplier Tubes (PMTs) in many high-energy astrophysics applications [1]; they can be used either for direct light detection or to readout a scintillator detector that converts incoming high-energy photons into visible light. The technical characteristics of SiPMs are powerful arguments for using them on future space telescopes: their use avoids operating high voltage supplies, ensuring robustness and reliability. Moreover, as they are insensitive to magnetic fields, we can use them in high field environments instead of PMTs. Their high Photon Detection Efficiency (PDE) would enlarge the overlap in cosmic-ray energies detected with ground-based facilities. At last, their low power consumption decreases the thermal dissipation. Consequently, one option would be to associate inorganic scintillators (LaBr3 , CeBr3 , . . . ) to SiPM detectors [2,3], in order to form a calorimeter for a space instrument to determine the energy of the incident photons [4]. The short decay time allows measurement of fast coincidences, which opens the way of measuring the time-of-flight of
the photons, a crucial parameter for reducing instrumental background noise in space environment. In 2015, we started a R&D study with a measurement campaign on three commercial devices; we selected SiPM references from Hamamatsu, SensL and Ketek (respectively S12572-050C, SB30035-SMT and PM3350-B63T75S-P4) to characterize their performance at room temperature under atmospheric pressure, and especially at low temperatures and pressures [5]. Then we studied their susceptibility to radiation, with a particular emphasis on the effects of protons, aboard a satellite, as well as to hostile phenomena such as Van Allen radiation belts. An irradiation test was carried out at room temperature on non-polarized detectors on the UCL cyclotron bench in Louvain (Belgium). Due to their good performance, the Ketek and SensL detectors have been chosen to be irradiated under a fluence of 2.1011 protons/cm2 by protons of 50 MeV [6]. This level would allow a variety of orbits, from LEO to GEO. Afterward, we focused our studies on the effect of annealing after a proton irradiation to evaluate a potential recovery. Lately, we are in
* Corresponding author at: IRAP, CNRS, 9 avenue du colonel Roche, 31028 Toulouse Cedex 4, France.
E-mail address: [email protected] (K. Lacombe). https://doi.org/10.1016/j.nima.2017.11.005 Received 28 September 2017; Accepted 3 November 2017 Available online xxxx 0168-9002/© 2017 Elsevier B.V. All rights reserved.
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
A/cm2 at room temperature. In a nutshell, the breakdown voltage slightly decreases with the temperature. In fact, the breakdown voltage of a junction increases with a higher temperature, in relation to a decrease of the values of the ionization coefficients. 3.2. Dark count rate As the dark current, the dark count rate (DCR) is particularly dependent on temperature; we show here that, in the case of SensL SiPM, it has been reduced by a factor of 3 at the overvoltage of 20%, by making it almost independent of this parameter. The same tendency is observed for the Ketek, which gets a decreasing of 5 by lessening of almost 40 ◦ C at OV = 25% (Fig. 5).
Fig. 1. Pictures of SiPM detectors SensL & Ketek observed under binocular (Left picture : 4774 cells of 35 μm pitch/Right picture : 3600 cells of 50 μm pitch).
3.3. PDE and single photon at low temperature progress to study two new references of SensL and Ketek (Figs. 1 and 2), to rise the statistics and develop a first matrix prototype of several chips on the same substrate [7].
3. Results
Then, to study the PDE, we first measured the pixel capacitance C for each detector. Then, we calibrated the optical chain to know the input photon flux, thanks to a photodiode power sensor in the wavelength range of the R&D from 350 nm to 800 nm [9,10]. Afterward, we measured the SiPM output voltage as a function of the wavelength. Our measurements were made for a bias voltage of 32 V for Ketek and 29 V for SensL SiPM for the temperatures of +23 ◦ C and −22 ◦ C. These results are consistent with the PDE values of Ketek and SensL manufacturers at room temperature. At the low temperature of −20 ◦ C, we observe that the PDE is better than at higher temperature for both SiPM detectors, due to the lower DC, by gaining until 25% at 420 nm peak (Fig. 6). An example of pulse area spectrum is shown in Fig. 7; we can notice that the different peaks are clearly distinguishable thanks to the low dark count rate, which reduced the peak to valley ratio (a diminution of 40 ◦ C seems to be appropriate to obtain such a good quality of peaks discrimination) [11]. The gain of a SiPM corresponds to the mean number of charge carriers that a single charge generates during the avalanche process in the depletion region of the detector. We calculated it by taking the difference between two consecutive peaks of the pulse area distribution (red histogram in Fig. 7).
3.1. Dark current as a function of temperature
3.4. Effect of high proton irradiation on SiPM
In this section, we present some results on dark current (DC), a source of noise intrinsic to the SiPM as a function of temperature. Dark Count Rate measurements supplement it, for both Ketek & SensL detectors. Next, we report Photo Detection Efficiency (PDE) and the pulse area distribution to observe the single photon. Thus, one way to suppress dark current is by cooling the detector to lower temperatures. We are studying this thermal dependence to find a temperature range ensuring the best performance of SiPMs. We measured dark current as a function of overvoltage [8] and evaluated the break down voltage at +23 ◦ C and −22 ◦ C at ∼10−4 mbar for the Ketek and SensL SiPMs (Fig. 4). For the Ketek detector, we found a breakdown voltage temperature coefficient of dVBD /dT = 21.6 mV/◦ C which is very close to the value provided in the manufacturer datasheet. For the SensL detector our measurements let us calculate an identical temperature coefficient of 21.6 mV/◦ C as expected. Moreover, at 20% overvoltage, the dark current density is about 2.5 × 10−7 A/cm2 at −22 ◦ C instead of 7 × 10−6
The dark current is a source of noise intrinsic to the SiPM detector, generated even though the cells are not exposed to light. This dark current is due to the thermal excitation of electrons into the conduction band, and hence has a strong temperature dependence. Cooling the detectors to low temperatures reduces this noise. We measured the dark current and evaluated the breakdown voltage of two detectors before and after irradiation at three temperatures under different pressure settings. As the Ketek detector is plugged on board, the SensL SiPM is directly stuck on it, going through a pico-ammeter connected to the output of the SiPMs without any front-end electronics to measure the dark current. Only the power supply has been applied to polarize the SiPM placed in the dark chamber without any light source. Our measurements show that the DC considerably increased for both SiPMs after irradiation, whereas the breakdown voltage remains quite the same. The temperature dependency is also stable; indeed, the dark current slightly decreases with the temperature with the same
2. Setup and configuration We developed a dedicated thermal vacuum test bench in order to characterize the SiPMs in an environment relevant for space applications. We designed a 30 dm3 vacuum chamber equipped with Peltier coolers to operate at low temperature (−22 ◦ C ± 0.5 ◦ C) and low pressure (∼10−4 mbar). A single mechanical support holds a specific low-noise electronics board containing both four detectors (2 SensL & 2 Ketek references) and three thermal probes mounted as SiPM, to ensure thermal stability and avoid EMC issues (Fig. 3). Firstly, we used a halogen source, covering a 350 to 800 nm wavelength range, going through a monochromator, for PDE measurements. Secondly, to observe the pulse area distribution and then obtain the gain value, a LED driver generates a light pulse at 400 nm, illuminating each SiPM via external and internal optical fibers.
Fig. 2. Table of features of the last SiPM detectors (ongoing study), provided in the manufacturers datasheets.
2
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
Fig. 3. Left picture: Specific test bench developed for R&T SiPM; Right picture: Low noise electronics board to test 4 detectors in the same conditions of pressure and temperature.
Fig. 4. Dark current versus bias voltage at 3 temperatures (+23 ◦ C, −15 ◦ C & −22 ◦ C) of Ketek detector (main graph) and SensL detector (inseted graph). Fig. 6. Photon Detection Efficiency as a function of wavelength at 2 temperatures (+23 ◦ C & −22 ◦ C) of Ketek (green curves) and SensL detector (red curves).
Fig. 5. Dark Count Rate versus Overvoltage at 2 temperatures (+23 ◦ C & −15 ◦ C) of both Ketek and SensL detectors.
coefficient for both curves before and after irradiation (Fig. 8). Besides, the voltage breakdown knees remained identifiable after irradiation
Fig. 7. Pulse area distribution of Ketek SiPM at low temperature.
3
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
an increase in dark current, as the bulk damage that creates also an increase in both after-pulse and cross-talk effect [12]. 4. Conclusion On the one hand, we reached the expected performance at room temperature, and on the other hand, our measurements showed a better stability and decreased values of intrinsic noise of SiPM at low temperature. So far we have identified no issues with operating under lowpressure conditions, and the detectors and its associated electronics we designed, called ‘‘SiPM board ’’, behave as if they were operated under atmospheric pressure. This conclusion is the key for space applications. The first step of our study of three different manufacturers (Ketek, SensL & Hamamatsu), has shown great performance in terms of low dark current and PDE and covers a large variety of parameters such as dark count rate and gain. Afterward, only two SiPMs have been selected to be tested in irradiation campaign to assess their resistance to a harsh space environment and check if their good performance is maintained under space constraints. The results here described showed that the dark current increased substantially after proton irradiation for both detectors. However, despite this overall change in scale, the shapes of I–V curves remain identical and indicate that the breakdown voltage is not really impacted by the proton radiation. We point out that the radiation damage does not impact the detector behavior neither in function of the temperature nor in function of the operating voltage. Then, we investigated whether an annealing test would allow recovering the detector properties measured before irradiation and our first results are promising but not significant. At a later stage, we plan a second radiation test campaign to consolidate our initial observations, with the study of the impact of lower fluence of protons and secondly Xrays. In addition, we plan to expose the SiPMs to rapid temperature cycling in the range of −55 ◦ C to +60 ◦ C and also to conduct a longterm temperature storage test to evaluate any effects of aging on the SiPM characteristics. Besides, we aim to characterize a dozen detectors (SensL and Ketek) to get more statistics and, in parallel, to qualify them via mechanical tests on next phase and to study the coupling with the advantageous LaBr3 scintillator offering a great light output that allows a good determination of the energy of the incident photon. Finally, the selected detector and its associated front-end electronics will undergo extended testing to reach at the end a Technology Readiness Level of 6 (subsystem model demonstration in relevant environment).
Fig. 8. Evolution of dark current after a high proton irradiation as a function of bias voltage at 2 temperatures (+25 ◦ C & −22 ◦ C) of Ketek SiPM (non-polarized detector during the irradiation test).
Acknowledgments We thank CNES staff for it support, Ketek and SensL companies for their great responsiveness, and for the quality of the detectors provided. We acknowledge the UCL staff for their support during the proton irradiation test. We are grateful to all participants at the NDIP conference in 2017 for their presentations and their expertise.
Fig. 9. Effect of annealing on dark current of SensL SiPM after being irradiated (green curve at +140 ◦ C).
References although they became more rounded. We can notice that, in the working range, the dark current at room temperature increases by 15 to 20 times (for SensL and Ketek SiPM respectively). At low temperature, this factor rises strongly to reach about 103 times in the case of Ketek detector whereas the SensL one multiplies by around 300, what would suggest its better irradiation resistance. Moreover, we conducted an annealing test to investigate whether these trends are irreversible (Fig. 9). In conclusion, radiation produce additional defects inside the band gap making it easier for electrons to reach the conduction band and generate increased thermal noise. Two types of radiation damages may happen in silicon detectors: the surface damage due to Ionizing Energy Loss (with the accumulation of charges in the oxide SiO2 and traps at Si/ SiO2 interface) and the bulk damage due to Non-Ionizing Energy Loss (a displacement damage, build-up of crystal defects). Several effects can occur directly observed on SiPM detectors: the surface damage causes
[1] A. Ulyanov, L. Hanlon, S. McBreen, S. Foley, D. Byrne, Study of silicon photomultipliers for the readout of scintillator crystals in the proposed grips 𝛾-ray astronomy mission, Proc. Sci. (2013). [2] W. Drozdowski, P. Dorenbos, A. Bos, G. Bizarri, A. Owens, F. Quarati, CeBr3 scintillator development for possible use in space missions, IEEE Trans. Nucl. Sci. (2007). [3] T. Szcześniak, M. Grodzicka, M. Moszyński, M. Szawlowski, D. Wolski, J. Baszak, Characteristics of scintillation detectors based on inorganic scintillators and sipm light readout, Nucl. Instrum. Methods Phys. Res. A (2012) 91–93. [4] A. Ulyanov, L. Hanlon, S. McBreen, et al., An integral view of the high-energy sky (the first 10 years), in: Proceedings of the 9th INTEGRAL Workshop, Paris, 2012. [5] http://www.ketek.net/products/sipm-technology/device-parameters/. [6] Y. Qiang, C. Zorn, F. Barbosa, E. Smith, Radiation hardness tests of SiPMs for the JLab Hall D Barrel calorimeter, Nucl. Instrum. Methods Phys. Res. A (2012) 234–241. [7] http://sensl.com/request-c-series-ds/. [8] R. Pagano, et al., Dark current in silicon photomultiplier pixels: Data and model, IEEE Trans. Electron Dev. (2012).
4
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.
K. Lacombe et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
[11] F. Acerbi, et al., Characterization of Single-Photon Time Resolution: From Single SPAD to Silicon Photomultiplier, IEEE, 2014. [12] J. Been, Effects of nuclear radiation on a high-reliability silicon power diode, NASATN-D-7423, 1973.
[9] H. Miyamoto, M. Teshima, B. Dolgoshein, et al., SIPM development and application for astroparticle physics experiments, in: Proceedings of the 31st ICRC, Lodz, 2009, p. 1320. [10] C. Jackson, et al., High-Volume Silicon Photomultiplier Production, Performance, and Reliability, SPIE, 2014.
5
Please cite this article in press as: K. Lacombe, et al., Characterization of silicon photomultipliers for new high-energy space telescopes, Nuclear Inst. and Methods in Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.11.005.