- Email: [email protected]
SiPM development for the imaging Cherenkov and fluorescence telescopes
Nuclear Instruments and Methods in Physics Research A 623 (2010) 198–200
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
SiPM development for the imaging Cherenkov and fluorescence telescopes Hiroko Miyamoto , Masahiro Teshima Max-Planck-Institut f¨ ur Physik, F¨ ohringer Ring 6, D-80805 M¨ unchen, Germany
a r t i c l e in f o
a b s t r a c t
Available online 16 March 2010
A silicon photomultiplier (SiPM) is a novel solid state photodetector which has an outstanding photon counting ability. The device has excellent features such as high quantum efficiency (QE), good charge resolution, fast response ð o 100 psÞ, very compact size, high gain (up to 2 3 106 ), very low power consumption with low bias voltages ð30 70 VÞ, immunity to the magnetic field. In the last few years, UV sensitive SiPMs with a p-on-n structure have been developed by a few companies such as Hamamatsu, Photonique, Zecotek Photonics Inc., and institutes such as the MPI-HLL (Max-PlanckInstitute for Physics—Max-Planck-Institute Semiconductor Laboratory) as well as the MPI-MEPhI (Max-Planck-Institute for Physics—Moscow Engineering Physical Institute). UV sensitive SiPMs are particularly promising for astroparticle physics applications in imaging atmospheric Cherenkov telescopes (IACTs), such as MAGIC/MAGIC-II [1] and CTA [2], and a fluorescence telescope in the space, JEM-EUSO [3]. The current status of the SiPM development by MPI-HLL, MPI-MEPhI, and the study of the application to the telescopes noted above will be reported. & 2010 Elsevier B.V. All rights reserved.
Keywords: Imaging Cherenkov Imaging fluorescence SiPM MPPC G-APD JEM-EUSO MAGIC-II CTA
1. Introduction
2.1. MPPC (Hamamatsu multi-pixel photon counter)
A SiPM is an array of multiple of avalanche photodiodes (APDs), which are operated in Geiger mode. In this case, each APD pixel is operated in binary mode, which outputs a fixed signal current, regardless of the detected number of photons. Eventually, the output signal from a SiPM is the total charge of all the APD signals, which is proportional to the number of the detected incident photons. The photon detection efficiency (PDE) of a SiPM depends on the transmittance of the surface, the fill factor of the sensitive area, the QE, and on the Geiger efficiency of the device:
MPPC produced by Hamamatsu is the first commercial product of UV sensitive SiPM in the world. They achieved relatively low dark rate, 2 kHz=mm2 at 10 1C for the product with 100 mm micro-cell pitch. A special device of 16 ch (4 4) of 3 mm 3 mm MPPC array has been developed for our camera as shown in Fig. 1.
PDE ¼ Transmittance Fill_Factor QE Geiger_Eff :
ð1Þ
Drawbacks of this device currently are a large dark current, optical crosstalk (OC) between micro-cells and relatively low sensitivity to UV and blue light. In the last few years, MPI-MEPhI and MPI-HLL have developed large size SiPMs (3 mm 3 mm) and matrices of them for astroparticle physics applications. Here the status of the SiPM application study to astroparticle physics experiments such as the IACTs MAGIC and CTA, and the space-borne fluorescence telescope JEM-EUSO, and related issues will be discussed. 2. Devices
Fig. 1. Left: Blue print of 16 ch (4 4) of 3 3 mm2 MPPC array device. Right panel: Photo of 16 ch MPPC array device.
We consider following three devices for the application to the experiments described above.
2.2. Dolgoshein SiPM
Corresponding author. Tel.: + 49 89 32354 349; fax: + 49 89 32354 526.
E-mail address: [email protected] (H. Miyamoto). 0168-9002/$ - see front matter & 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.nima.2010.02.194
The SiPM which MPI-MEPhI has developed is called Dolgoshein SiPM after the name of the inventor [4]. This device has a big advantage of OC suppression, additionally to its promising high
H. Miyamoto, M. Teshima / Nuclear Instruments and Methods in Physics Research A 623 (2010) 198–200
UV sensitivity. OC is a phenomena that a photon which enters a micro-cell of SiPM generates more photons and thus fires neighboring micro-cells. There are two, prompt and slow, OC components. The prompt OC can be caused when one or more photons directly goes to a neighboring micro-cell(s) and triggers a breakdown there like any external photons. The slow OC can be caused when a photon induces charge in its bulk and migrate into the avalanche region. A dedicated design which has an additional junction and grooves between micro-cells which act as an optical isolation is applied to this device as shown in Fig. 2 and consequently OC is suppressed to less than 3%.
Fig. 2. Schematic view of OC suppression.
2.3. SiMPl
199
for a reference, a calibrated Hamamatsu R10408-MODULE (1 in SuperBialkali (SBA) PMT) which is the device for the MAGICII telescope. The ambient temperature is fixed to 5 1C. The surface of R10408-MODULE and R8900U-08-M36-MOD are covered by a slit with a 3 mm 3 mm square window to obtain the same aperture as that of Dolgoshein SiPM and MPPC. The Lecroy WavePro 7300A 3 GHz Dual 20 GS/s digital oscilloscope is used for the data acquisition with externally triggering by the synchronized signal from the function generator. Preliminary results of PDE measurements Fig. 4 shows the results of the measured PDE [%] as a function of wavelength [nm]. Each of filled triangle, empty circle, and filled circle data points represent for the distribution of Dolgoshein SiPM, MPPC, and MAPMT. Assuming Poisson distribution, we calculated the ratio of normalized detected number of photoelectrons of each devices and of the reference PMT. A PDE estimated here is derived by the ratio described above multiplying by the QE and the 95% of the collection efficiency (CE) of the reference PMT at each wavelength. Although there are still some systematics for the absolute estimation, you may see the significantly higher efficiency of Dolgoshein SiPM than MAPMT. MPPC has relatively lower efficiency in UV region, but the study of wavelength shifter is also ongoing and we expect we will take an
The SiMPl developed by the MPI-HLL is a SiPM with a ‘‘bulk internal resistor’’ [5]. The device has a novel concept of a quench resistor integrated into the silicon bulk which enables not only to optimize the filling factor but also to simplify the production procedure. Therefore it has the potential to lower the cost and reduce the production time (see Fig. 3). With the concept of this internal bulk resistor and antireflecting coating, the PDE over 60% is expected. The wafer level of tests are already ongoing.
Fig. 4. Preliminary result of PDE measurements. The filled triangles, gray squares, and filled circles show the distribution of the Dolgoshein SiPM, the MPPC, and the MAPMT representatively. MPPC and D-SiPM are operated at overvoltage of 1.2 and 3.5 V representatively. The PDE is calculated from the ratio of the normalized number of detected photoelectrons in the considered device and in a reference PMT, assuming Poisson distributions. Systematic uncertainties are not shown. Other factors to be considered include the SiPM overvoltage and temperature dependence.
Fig. 3. Concept figure of bulk integrated quench resistor type SiPM (SiMPl).
3. Toward the application of SiPMs to astroparticle physics experiments 3.1. Sensitivity to UV and blue light To detect fluorescence and Cherenkov lights, high PDE in UV region (from 300 to 400 nm) is essential. Our PDE measurement setup is following: A light source LED which is allocated outside of the temperature box, is flashed by Agilent function generator at 2 kHz. Exchanging LEDs whose wavelength are from 310 to 450 nm, four test devices described below are equally illuminated by a single photoelectron level light from the LED via UV transparent fiber optics and UV fused silica diffuser. The devices tested in this measurement are Hamamatsu MPPC S10362-33-100C (3 3 mm2), Dolgoshein SiPM (3 3 mm2), Hamamatsu R8900U-08M36-MOD (1 in. 36 ch Ultra Bialkali-Multi Anode PMT (UBAMAPMT) which is the device for the JEM-EUSO basic design, and
Fig. 5. Left: temperature compensation circuit with a thermistor(Rt). Right-top: gain vs time [min]. Right-bottom: temperature of device (MPPC) vs corresponding time to the gain data.
200
H. Miyamoto, M. Teshima / Nuclear Instruments and Methods in Physics Research A 623 (2010) 198–200
Fig. 6. Left: 256 ch prototype camera with 16 ch MPPC device. Right: Air Cherenkov image of 64 ch (8 8) /256 ch MPPC camera. The whole camera size is 0:93 0:93 .
advantage of its significant high PDE in longer wavelength. Also, the PDE of SiPM depends on overvoltage, and therefore even higher PDE is promising with higher voltage. Study of the PDE overvoltage dependence is also ongoing.
with four 20 pin board-to-board connectors, is assembled with 64 ch signal readout, bias voltage with the temperature compensation described above, and differentiator circuit. First image of air Cherenkov by a pure MPPC camera
3.2. Temperature dependence Another critical issue for the SiPM application is its strong temperature dependence of the breakdown voltage, and thus the gain dependence which is roughly 3%/1C. A temperature compensation circuit for the voltage supply with a thermistor can be a solution. Thermistor is a resistor which has a temperature dependence described as following function: R ¼ R0 eBð1=T1=T0 Þ
ð2Þ
where R is resistance of the thermistor at a temperature of T[K], B is a constant parameter, so called B value, and R0 is resistance of the thermistor at a given temperature of T0[K]. The left panel of Fig. 5 shows the schematic view of the compensation circuit where V0 is a fixed initial voltage, Rt is a thermistor. With using this circuit, we measured the gain and temperature of a MPPC with changing the ambient temperature from 18 to 8 1C. As a result, we obtained the gain fluctuation of less than 0.1% as shown in the right panel of Fig. 5.
4. 256 ch prototype camera with 16 ch MPPC We developed a prototype 256 ch MPPC camera consists of 4 4 array of 16 ch MPPC module as shown in the left panel of Fig. 6. It is designed with one mother board and four daughter boards. Only socket pins for the signal readout and bias voltage, and a hole pair for thermistor lead lines are assembled on the mother board. The surface of mother board is covered by a layer of copper for the temperature uniformity for all the devices. A daughter board, which is connected to the mother board vertically
Here the image of air Cherenkov light detected by the camera using 64 channels out of 256 channels are shown in the right panel of Fig. 6. We use a small telescope whose mirror diameter is 60 cm and focal length is 150 cm. A CAEN VME module V1740, a 64 ch, 12 bit, 65 MS/S digitizer is used to readout signals. With the trigger condition of three next neighbors coincidence over the threshold level of about 15 photoelectrons, we recorded about 8 events/h, which corresponds to Cherenkov signals from about 100 TeV cosmic ray air showers. Note that the observation was done on the roof of the institute, located near the center of Munich where strong background light comes from the City. Acknowledgements We acknowledge Prof. B. Dolgoshein from MEPhI, Dr. R. Mirzoyan and Dr. J. Ninkovic from MPI as well as all the technical and electronic staff who contribute to this SiPM developement/application project in MEPhI, MPI, and HLL. We also acknowledge MAGIC collaboration, CTA consortium and JEM-EUSO collaboration. References [1] [2] [3] [4]
J. Albert, et al., Astrophys. J. 674 (2008) 1037. /http://www.cta-observatory.org/S. /http://jemeuso.riken.go.jp/S. P. Buzhan, et al., Proceedings of 5th International Conference on NDIP 2008, Palais des Congres, Aix-les-Bains, France, June 15–20, 2008. [5] J. Ninkovic, et al., Proceedings of 5th International Conference on NDIP 2008, Palais des Congres, Aix-les-Bains, France, June 15–20, 2008.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
SiPM development for the imaging Cherenkov and fluorescence telescopes Hiroko Miyamoto , Masahiro Teshima Max-Planck-Institut f¨ ur Physik, F¨ ohringer Ring 6, D-80805 M¨ unchen, Germany
a r t i c l e in f o
a b s t r a c t
Available online 16 March 2010
A silicon photomultiplier (SiPM) is a novel solid state photodetector which has an outstanding photon counting ability. The device has excellent features such as high quantum efficiency (QE), good charge resolution, fast response ð o 100 psÞ, very compact size, high gain (up to 2 3 106 ), very low power consumption with low bias voltages ð30 70 VÞ, immunity to the magnetic field. In the last few years, UV sensitive SiPMs with a p-on-n structure have been developed by a few companies such as Hamamatsu, Photonique, Zecotek Photonics Inc., and institutes such as the MPI-HLL (Max-PlanckInstitute for Physics—Max-Planck-Institute Semiconductor Laboratory) as well as the MPI-MEPhI (Max-Planck-Institute for Physics—Moscow Engineering Physical Institute). UV sensitive SiPMs are particularly promising for astroparticle physics applications in imaging atmospheric Cherenkov telescopes (IACTs), such as MAGIC/MAGIC-II [1] and CTA [2], and a fluorescence telescope in the space, JEM-EUSO [3]. The current status of the SiPM development by MPI-HLL, MPI-MEPhI, and the study of the application to the telescopes noted above will be reported. & 2010 Elsevier B.V. All rights reserved.
Keywords: Imaging Cherenkov Imaging fluorescence SiPM MPPC G-APD JEM-EUSO MAGIC-II CTA
1. Introduction
2.1. MPPC (Hamamatsu multi-pixel photon counter)
A SiPM is an array of multiple of avalanche photodiodes (APDs), which are operated in Geiger mode. In this case, each APD pixel is operated in binary mode, which outputs a fixed signal current, regardless of the detected number of photons. Eventually, the output signal from a SiPM is the total charge of all the APD signals, which is proportional to the number of the detected incident photons. The photon detection efficiency (PDE) of a SiPM depends on the transmittance of the surface, the fill factor of the sensitive area, the QE, and on the Geiger efficiency of the device:
MPPC produced by Hamamatsu is the first commercial product of UV sensitive SiPM in the world. They achieved relatively low dark rate, 2 kHz=mm2 at 10 1C for the product with 100 mm micro-cell pitch. A special device of 16 ch (4 4) of 3 mm 3 mm MPPC array has been developed for our camera as shown in Fig. 1.
PDE ¼ Transmittance Fill_Factor QE Geiger_Eff :
ð1Þ
Drawbacks of this device currently are a large dark current, optical crosstalk (OC) between micro-cells and relatively low sensitivity to UV and blue light. In the last few years, MPI-MEPhI and MPI-HLL have developed large size SiPMs (3 mm 3 mm) and matrices of them for astroparticle physics applications. Here the status of the SiPM application study to astroparticle physics experiments such as the IACTs MAGIC and CTA, and the space-borne fluorescence telescope JEM-EUSO, and related issues will be discussed. 2. Devices
Fig. 1. Left: Blue print of 16 ch (4 4) of 3 3 mm2 MPPC array device. Right panel: Photo of 16 ch MPPC array device.
We consider following three devices for the application to the experiments described above.
2.2. Dolgoshein SiPM
Corresponding author. Tel.: + 49 89 32354 349; fax: + 49 89 32354 526.
E-mail address: [email protected] (H. Miyamoto). 0168-9002/$ - see front matter & 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.nima.2010.02.194
The SiPM which MPI-MEPhI has developed is called Dolgoshein SiPM after the name of the inventor [4]. This device has a big advantage of OC suppression, additionally to its promising high
H. Miyamoto, M. Teshima / Nuclear Instruments and Methods in Physics Research A 623 (2010) 198–200
UV sensitivity. OC is a phenomena that a photon which enters a micro-cell of SiPM generates more photons and thus fires neighboring micro-cells. There are two, prompt and slow, OC components. The prompt OC can be caused when one or more photons directly goes to a neighboring micro-cell(s) and triggers a breakdown there like any external photons. The slow OC can be caused when a photon induces charge in its bulk and migrate into the avalanche region. A dedicated design which has an additional junction and grooves between micro-cells which act as an optical isolation is applied to this device as shown in Fig. 2 and consequently OC is suppressed to less than 3%.
Fig. 2. Schematic view of OC suppression.
2.3. SiMPl
199
for a reference, a calibrated Hamamatsu R10408-MODULE (1 in SuperBialkali (SBA) PMT) which is the device for the MAGICII telescope. The ambient temperature is fixed to 5 1C. The surface of R10408-MODULE and R8900U-08-M36-MOD are covered by a slit with a 3 mm 3 mm square window to obtain the same aperture as that of Dolgoshein SiPM and MPPC. The Lecroy WavePro 7300A 3 GHz Dual 20 GS/s digital oscilloscope is used for the data acquisition with externally triggering by the synchronized signal from the function generator. Preliminary results of PDE measurements Fig. 4 shows the results of the measured PDE [%] as a function of wavelength [nm]. Each of filled triangle, empty circle, and filled circle data points represent for the distribution of Dolgoshein SiPM, MPPC, and MAPMT. Assuming Poisson distribution, we calculated the ratio of normalized detected number of photoelectrons of each devices and of the reference PMT. A PDE estimated here is derived by the ratio described above multiplying by the QE and the 95% of the collection efficiency (CE) of the reference PMT at each wavelength. Although there are still some systematics for the absolute estimation, you may see the significantly higher efficiency of Dolgoshein SiPM than MAPMT. MPPC has relatively lower efficiency in UV region, but the study of wavelength shifter is also ongoing and we expect we will take an
The SiMPl developed by the MPI-HLL is a SiPM with a ‘‘bulk internal resistor’’ [5]. The device has a novel concept of a quench resistor integrated into the silicon bulk which enables not only to optimize the filling factor but also to simplify the production procedure. Therefore it has the potential to lower the cost and reduce the production time (see Fig. 3). With the concept of this internal bulk resistor and antireflecting coating, the PDE over 60% is expected. The wafer level of tests are already ongoing.
Fig. 4. Preliminary result of PDE measurements. The filled triangles, gray squares, and filled circles show the distribution of the Dolgoshein SiPM, the MPPC, and the MAPMT representatively. MPPC and D-SiPM are operated at overvoltage of 1.2 and 3.5 V representatively. The PDE is calculated from the ratio of the normalized number of detected photoelectrons in the considered device and in a reference PMT, assuming Poisson distributions. Systematic uncertainties are not shown. Other factors to be considered include the SiPM overvoltage and temperature dependence.
Fig. 3. Concept figure of bulk integrated quench resistor type SiPM (SiMPl).
3. Toward the application of SiPMs to astroparticle physics experiments 3.1. Sensitivity to UV and blue light To detect fluorescence and Cherenkov lights, high PDE in UV region (from 300 to 400 nm) is essential. Our PDE measurement setup is following: A light source LED which is allocated outside of the temperature box, is flashed by Agilent function generator at 2 kHz. Exchanging LEDs whose wavelength are from 310 to 450 nm, four test devices described below are equally illuminated by a single photoelectron level light from the LED via UV transparent fiber optics and UV fused silica diffuser. The devices tested in this measurement are Hamamatsu MPPC S10362-33-100C (3 3 mm2), Dolgoshein SiPM (3 3 mm2), Hamamatsu R8900U-08M36-MOD (1 in. 36 ch Ultra Bialkali-Multi Anode PMT (UBAMAPMT) which is the device for the JEM-EUSO basic design, and
Fig. 5. Left: temperature compensation circuit with a thermistor(Rt). Right-top: gain vs time [min]. Right-bottom: temperature of device (MPPC) vs corresponding time to the gain data.
200
H. Miyamoto, M. Teshima / Nuclear Instruments and Methods in Physics Research A 623 (2010) 198–200
Fig. 6. Left: 256 ch prototype camera with 16 ch MPPC device. Right: Air Cherenkov image of 64 ch (8 8) /256 ch MPPC camera. The whole camera size is 0:93 0:93 .
advantage of its significant high PDE in longer wavelength. Also, the PDE of SiPM depends on overvoltage, and therefore even higher PDE is promising with higher voltage. Study of the PDE overvoltage dependence is also ongoing.
with four 20 pin board-to-board connectors, is assembled with 64 ch signal readout, bias voltage with the temperature compensation described above, and differentiator circuit. First image of air Cherenkov by a pure MPPC camera
3.2. Temperature dependence Another critical issue for the SiPM application is its strong temperature dependence of the breakdown voltage, and thus the gain dependence which is roughly 3%/1C. A temperature compensation circuit for the voltage supply with a thermistor can be a solution. Thermistor is a resistor which has a temperature dependence described as following function: R ¼ R0 eBð1=T1=T0 Þ
ð2Þ
where R is resistance of the thermistor at a temperature of T[K], B is a constant parameter, so called B value, and R0 is resistance of the thermistor at a given temperature of T0[K]. The left panel of Fig. 5 shows the schematic view of the compensation circuit where V0 is a fixed initial voltage, Rt is a thermistor. With using this circuit, we measured the gain and temperature of a MPPC with changing the ambient temperature from 18 to 8 1C. As a result, we obtained the gain fluctuation of less than 0.1% as shown in the right panel of Fig. 5.
4. 256 ch prototype camera with 16 ch MPPC We developed a prototype 256 ch MPPC camera consists of 4 4 array of 16 ch MPPC module as shown in the left panel of Fig. 6. It is designed with one mother board and four daughter boards. Only socket pins for the signal readout and bias voltage, and a hole pair for thermistor lead lines are assembled on the mother board. The surface of mother board is covered by a layer of copper for the temperature uniformity for all the devices. A daughter board, which is connected to the mother board vertically
Here the image of air Cherenkov light detected by the camera using 64 channels out of 256 channels are shown in the right panel of Fig. 6. We use a small telescope whose mirror diameter is 60 cm and focal length is 150 cm. A CAEN VME module V1740, a 64 ch, 12 bit, 65 MS/S digitizer is used to readout signals. With the trigger condition of three next neighbors coincidence over the threshold level of about 15 photoelectrons, we recorded about 8 events/h, which corresponds to Cherenkov signals from about 100 TeV cosmic ray air showers. Note that the observation was done on the roof of the institute, located near the center of Munich where strong background light comes from the City. Acknowledgements We acknowledge Prof. B. Dolgoshein from MEPhI, Dr. R. Mirzoyan and Dr. J. Ninkovic from MPI as well as all the technical and electronic staff who contribute to this SiPM developement/application project in MEPhI, MPI, and HLL. We also acknowledge MAGIC collaboration, CTA consortium and JEM-EUSO collaboration. References [1] [2] [3] [4]
J. Albert, et al., Astrophys. J. 674 (2008) 1037. /http://www.cta-observatory.org/S. /http://jemeuso.riken.go.jp/S. P. Buzhan, et al., Proceedings of 5th International Conference on NDIP 2008, Palais des Congres, Aix-les-Bains, France, June 15–20, 2008. [5] J. Ninkovic, et al., Proceedings of 5th International Conference on NDIP 2008, Palais des Congres, Aix-les-Bains, France, June 15–20, 2008.