Characterization of Hamamatsu 14160 series of Silicon Photo-Multipliers
P.W. Cattaneo, A. Menegolli, M.C. Prata, G.L. Raselli, M. Rossella
PPrepared for submission to JINST
Characterization of Hamamatsu 14160 series of SiliconPhoto-Multipliers
P.W. Cattaneo, a , A. Menegolli, a , b M.C. Prata, a G.L. Raselli, a M. Rossella, a a INFN Sezione di Pavia, Pavia, Italy b Dipartimento di Fisica, University of Pavia, Pavia, Italy
E-mail: [email protected]
Abstract: Silicon Photo-Multipliers (SiPMs) are semiconductor-based photo-detectors with per-formances similar to the traditional Photo-Multiplier Tubes (PMTs). An increasing number ofexperiments dedicated to particle detection in colliders, accelerators, astrophysics, neutrino andrare-event physics involving scintillators are using SiPMs as photodetectors. They are graduallysubstituting PMTs in many applications, especially where low voltages are required and high mag-netic field is present. Hamamatsu Photonics K.K., one of leading producers of photo-detectors, inthe last year introduced the S14160 series of SiPMs with improved performances. In this work, acharacterization of these devices will be presented in terms of breakdown voltages, pulse shape,dark current and gain. Particular attention has been dedicated to the analysis of the parameters asfunction of temperature. Corresponding author. a r X i v : . [ phy s i c s . i n s - d e t ] J un ontents In latest years, solid state photo-detectors have become a staple for many experiments in high energyphysics, either at the accelerators or devoted to neutrino and rare-event physics. In particular, SiliconPhoto-Multipliers (SiPMs) are semiconductor-based photo-detectors adopting the avalanche photo-diode approach to get performance similar to the traditional Photo-Multiplier Tubes (PMTs). Anumber of neutrino and Dark Matter experiments [1] [2] plan to use SiPM also at cryogenictemperatures as scintillation light detectors. Despite the smaller window surface with respect totraditional large area PMTs, SiPMs operating at cryogenic temperature have the advantage of athermal component of the noise decreasing with the reduction of the temperature, thus allowingthe use of these devices at higher overvoltage [3] [4]. Hamamatsu Photonics K.K., one of leadingproducers of photo-detectors, in the last year introduced the S14160 series of SiPMs with improvedperformances. The aim of this work is the test these new devices in terms of breakdown voltages,pulse shape, dark current and gain, and compare them with previous series. An experimentalapparatus based on a climatic chamber allowed to carry out the measurements in a temperaturerange from -40 ◦ C to +40 ◦ C. The adopted experimental setup is based on a climatic chamber (F.lli Galli model Genviro-030LC)with a temperature range from − ◦ C to + ◦ C that can house the SiPM to be tested. The innerpart of the chamber can be connected to the external world by means of electrical and optical(optical fiber) feed-throughs. Fig. 1 shows a picture of the chamber. A Keithly 6487 modelpicoAmmeter/Source-generator is used both to apply the voltage to the SiPM and to measure thecurrent. An Hamamatsu optical pulser, model PLP10-040, with an emission wavelength λ = 405nm, pulse width of 60 ps (FWHM) and peak power of 200 mW, is used to illuminate the device.About 10 photons per pulse are directly injected in a FC optical fiber followed by an optical fiberattenuator that can reduce the number of photons from 10 to zero. The optical pulser can pulseup to 100MHz, but all the measurements of this work have been carried out at 10Hz. Signals areacquired by means of a LeCroy digital Oscilloscope model Waverunner 610Zi, with 8-bit vertical– 1 –esolution, 1 GHz bandwidth, and 20 GSa/s sampling rate. The complete set-up is shown in Fig. 2.Several Hamamatsu devices from 14160 series have been selected to be tested (see Fig. 3):• S14160-3050HS: 3 × active area, 50 µ m cell, 3531 pixels, 74% fill factor, siliconwindow.• S14160-4050HS: 4 × active area, 50 µ m cell, 6331 pixels, 74% fill factor, siliconwindow.• S14160-6050HS: 6 × active area, 50 µ m cell, 14331 pixels, 74% fill factor, siliconwindow.• S14160-1310PS: 1.3 × active area, 10 µ m cell, 16675 pixels, 31% fill factor, siliconwindow.• S14160-1315PS: 1.3 × active area, 15 µ m cell, 7296 pixels, 49% fill factor, siliconwindow.Hamamatsu S14160 series are innovative surface mount SiPMs with an higher PDE, a lowertemperature coefficient and a lower operation (breakdown) voltage in comparison to previousHamamatsu devices. They are immune to effects of magnetic fields, are poorly affected by cross-talk and after-pulses, operate at a typical breakdown voltage of 38V and have an excellent timeresolution. The 3 × , 4 × and 6 × version have 50 µ m cells, a PDE up to 50%(at the peak wavelength and at 2.7V overvoltage) and a typical gain of 10 . They are suited forscintillation detectors and for industrial applications such as for PET, radiation monitor etc.. Thehigher PDE is due to the HWB (Hole Wire bonding) technology allowing for small dead space inthe photosensitive area. The 1.3 × versions have smaller pixel size (10 µ m and 15 µ m), highfill factor and wide dynamic range; they show low cross-talk and after-pulses and a typical gainof 10 . They are suggested for high energy physics experiments, fluorescence measurement, flowcytometry, DNA sequencers and environmental analysis. Figure 1 . Experimental set-up used for the characterization of the S14160 SiPM series. – 2 – igure 2 . Scheme of the experimental set-up.
Figure 3 . Picture of the devices under test.
Considering that all the devices under test are surface-mount technology, we decided to solder themon electronic boards each equipped with a coaxial MCX connector, as shown in Fig. 4. Solderingwas performed after a vacuum baking of 48 hours at 60 ◦ C, by means of Vapor Phase ReflowSoldering process that guarantees a uniform and well controlled temperature of the process andconsequently no damage on the plastic window of the devices.Preliminary results have been obtained from tests on the devices described above. Some ofthem are here reported in terms of:• I-V curve of S14160-3050HS for various temperatures plotting dark current versus biasvoltage (Fig. 5 left); the breakdown voltage is defined as the voltage where the secondderivative of the curves peaks,• breakdown voltage as function of temperature with temperature coefficient displayed for thethree 50 µ m cell SiPMs (Fig. 5 right): all models show a linear behaviour, with similartemperature coefficients ∼
30 mV/ ◦ C. The same measurement performed in the past withprevious S12572 series SiPMs (3 × area, 50 µ m cell) gave a temperature coefficient of ∼
60 mV/ ◦ C [5], thus showing the larger stability of the new series;– 3 – igure 4 . Left: 3D rendering of our custom board for SiPM mounting. Right: the electronic boards housingthe SiPM models under test. • Direct I-V curve for S14160-3050HS at several temperatures (Fig. 6 left);• quenching resistor for S14160-6050HS and 4050HS as function of temperature (Fig. 6 right).
Figure 5 . Left: I-V curve at different temperatures for S14160-3050HS. Right: Breakdown voltage as afunction of temperature.
A further comparison with the S12572 series has been carried out first of all in terms of leadingedge and decay time of SiPM signals under illumination provided by the Hamamatsu optical pulser(see Fig. 2): the typical pulse shapes of all devices under test are shown in Fig. 7-left, while inFig. 7-right the rise time of the S14160-3050HS as function of the overvoltage is shown.Fig. 8 shows the results on the leading edge (left) and on the decay time (right) of the SiPMsignal for the S14160 and S12572 series. The leading edge of the S14160-6050HS SiPM is aboutthe half of the one of the S12572 SiPM series. It can be noted that for both series the leading edgeis poorly affected by the temperature change, being stable in the range from -40 ◦ C to +40 ◦ C definedwithin the climatic chamber. The decay time is instead similar for the S14160 and S21572 serieswhen the SiPMs have the same 3 × area. The S14160-6050HS, with a 6 × area, isaffected by a longer decay time, as expected due to the higher quenching resistor.– 4 – igure 6 . Left: Direct I-V curve for S14160-3050HS at several temperatures. Right: quenching resistor asfunction of temperature for S14160-4050HS and S14160-6050HS. In order to characterize the devices in terms of noise, we measured the dark current of S14160-6050HS as a function of the overvoltage for several temperatures, as shown in Fig. 9-left. Comparingthe new series with the S12572, a large reduction of the noise for a same overvoltage value is seen,as shown in Fig. 9-right for a temperature of 30 ◦ C. Both SiPMs are 3 × area, 50 µ m cell.Finally, a preliminary study was performed on the signal amplitude of several SiPMs, illumi-nated by the Hamamatsu optical pulser with a fixed number of photons per pulse, by evaluatingthe peak directly from the oscilloscope output and varying the overvoltage. Results are shown inFig. 10 in terms of pulse peak normalized to the value measured for an overvoltage of 3 Volts. Inthe figure, the quantity displayed in the vertical axis is proportional to the product PDE × Gain. Ifwe consider a constant PDE as a function of the overvoltage, the curve slope corresponds to therelative variation ∆ G ∆ V of the gain with respect to the supplied voltage. S14160-6050HS SiPM hasan almost linear behaviour from -40 ◦ C to room temperature (Fig. 10-left). The relative variationof the peak amplitude is the same for all the temperatures, being the slopes of all curves identical.In Fig. 10-right the S14160 and S12572 series are compared at room temperature: while the twoS14160 models have the same relative variation, the slope of the S12572 curve (red dots) is slightlylarger.
The Hamamatsu S14160 series Silicon PM has been widely tested in terms of I-V curve, voltagebreakdown temperature coefficient, quenching resistor and pulse response as a function of temper-ature, from -40 ◦ C to +40 ◦ C. The performance on the most significant parameters of the S14160series of SiPM is largely improved with respect to preceding series, in particular in terms of a lowertemperature coefficient, a lower breakdown voltage, better noise and higher PDE and gain.– 5 – igure 7 . Left: typical pulse shape for all devices under test with the indication of the decay time. Right:leading edge of S14160-3050HS as a function of the overvoltage.
Figure 8 . Comparison between S14160 and S12572 leading edge (left) and decay time (right) as a functionof the temperature.
References – 6 – igure 9 . Left: dark current of S14160-6050HS as a function of the overvoltage, for several temperatures.Right: comparison between S14160-3050HS and S12572-050P in terms of Dark Current as a function ofthe overvoltage.
Figure 10 . Left: peak amplitude of S14160-3050HS as a function of the applied voltage, measured at severaltemperature. Right: peak amplitude of S14160 and S12572 series as a function of the overvoltage, measuredat room temperature.. Left: peak amplitude of S14160-3050HS as a function of the applied voltage, measured at severaltemperature. Right: peak amplitude of S14160 and S12572 series as a function of the overvoltage, measuredat room temperature.