An Interview with Sergey Vinogradov

Written by katrinlink on 11th March 2015. Posted in

24 June 2015

Dr. Sergey Vinogradov

Senior Marie Curie Research Fellow at the Dept. of Physics, University of Liverpool and the Cockcroft Institute of Accelerator Science and Technology, UK

Could you give a brief overview of SiPM applications?

The first project on Silicon Photomultiplier (SiPM) applications were conducted by MEPhI-Pulsar-ITEP-DESY team in 2003-2005 where a few thousand SiPMs were used as readout of the wavelength-shifting fibre tile calorimeter for the CALICE collaboration. It confirmed the possibility of using SiPM in large scale scientific experiments. A first large industrial-scale project of SiPM implementation was an upgrade of the T2K neutrino detection system in Tokai-to-Kamiokande, Japan. About 60,000 SiPMs were supplied by Hamamatsu and have been successfully tested and assembled by the T2K collaboration since 2006. T2K experiment has been started in 2009, and observation of neutrino oscillations has been reported in 2011. There are a number of accelerator upgrade projects under consideration now, including beam loss monitoring systems for the Large Hadron Collider at CERN, future linear colliders and light sources, detector systems of the CMS HCAL, ALICE and NA61/SHINE high energy physics experiments, calorimeter systems at Fermi Lab and Jefferson Lab (USA).

Medical imaging is another very attractive industrial-scale area of SiPM applications. Now, major vendors of Positron Emission Tomography (PET) scanners, including General Electric, Siemens and Philips, as well as a number of academic-industrial collaborations, e.g. within the EU funded projects “Hyper Image”, Endo TOFPET, the AX-PET collaboration and the OpenPET project were being in a progress of developments and prototyping of SiPM-based systems. These developments are expanding in three directions of image quality improvements exploiting the very good energy and time resolution, pixellation, and insensitivity to magnetic field of SiPMs: 1) Time-of-flight PET for full-body clinical scanners, 2) high resolution PET required for small animal imaging in pre-clinical studies, and 3) combined PET/MRI (magnetic resonance imaging).

Astroparticle physicists benefit from SiPM technology since 2005, when MEPhI-Pulsar SiPMs have been sent in space for cosmic ray studies at the International Space Station. In a framework of APPEC (former ASPERΑ), SiPM were being considered as a promising photon detector for MAGIC, EUSO, DARWIN, GERDA, LAGUNA LENA, and CTA projects. In particular, many R&D teams, members of CTA collaboration, are working hard on prototyping of SiPM-based CTA cameras, and one of the most remarkable result of this activity – FACT (First G-APD Cherenkov Telescope in La Palma, Canary Islands, Spain) based on Hamamatsu MPPCs – are operational and stable since 2011. The key advantage of SiPM technology for CTA applications is high sensitivity to and resolution of a few photon short pulse of a Cherenkov flash, unaffected by an intense night sky background light, because time-separated single counts of the background photons can be easily discriminated from multi-photon events. Photon number resolution facilitates self-calibration of SiPM gain, negative feedback allows to overcome over-lighting turning to normal operations in a few tens nanoseconds, power consumption is rather low, operations are stable … what else would be needed for reliable and efficient work in the field for a long term?

What was your personal implication with this research? What can be expected from SiPMs in the future?

Since 1982, I am with the Solid State Photodetectors group of the Lebedev Physical Institute, Moscow, from where SiPM concepts have been originated. Initially I was in R&D of photosensors for image recording and processing, and then refocused on avalanche detectors in the early 2000s, when my colleagues established collaboration on R&D of SSPM between the Lebedev Institute and the Amplification Technologies, USA. In 2013 I won the 7th European Framework Program Marie Curie International Incoming Fellowship grant “SiPM in-depth” and started to work at the host institution – the QUASAR group of the University of Liverpool and the Cockcroft Institute of Accelerator Science and Technology, UK.

The QUASAR (Quantum Systems and advanced Accelerator Research) group led by Prof. Dr. C.P. Welsch is an internationally structured research group focused on the development and optimization of particle accelerators and their diagnostic systems, in particular, on Beam Loss Monitoring (BLM) for the Compact Linear Collider (CLIC) based on Cherenkov fibre with SiPM readout. From my point of view, BLM is the most challenging application of SiPM technology, because it requires detection of extremely variable in intensities and temporal profiles Cherenkov light flashes and then identification of locations of the losses alongside a CLIC beam line from SiPM response. If we are successful in resolving this challenge, accelerator technology would obtain huge benefits from high accuracy, fast response, and low cost of SiPM-based BLM systems. Moreover, it will pave the way to use SiPMs for super-sensitive high dynamic range detection of arbitrary waveform signals, which could dramatically extend application areas of SiPM technology in optical communications, laser ranging, X-ray computer tomography, homeland security, and many others.

Therefore, in the next few years, we plan to advance experimental studies of SiPMs into transient and nonlinear processes at high light intensities, to make a comparative characterization of SiPMs from major developers, and to start experiments at the CLIC test beam line facility (CTF3) in CERN. Theoretical studies and modelling of SiPMs are an inherent part of my work as well. I will try to describe non-stationary nonlinear history-dependent photon detection in SiPMs by a filtered point process technique from the theory of stochastic processes, and I am looking for collaboration for that with professional mathematicians.

After the end of my Marie Curie Fellowship, I will return to the Lebedev Institute in September 2015 to transfer my new knowledge and collaborations with UK and EU accelerator physicists to my home institution. Obviously, being in Moscow I plan to continue ongoing projects within the QUASAR group in UK and in CERN, as well as, to take part and apply my experience in new projects and collaborations in EU and worldwide.

Dr. Shubin (Lebedev Physical Institute) identified SiPM technology as a revolution in photon detection equal to an invention of lasers. And Dr. Mirzoyan (Max Plank Institute) called his talk “SiPM: on the Way of Becoming an Ideal Low Light Level Sensor” (IEEE Nuclear Science Symposium and Medical Imaging Conference, Knoxville, USA, 2010). In fact, since the 2000s we see very fast and significant progress of SiPM in all technical aspects, in applications and commercialization, as well as an almost exponential growth of the number of SiPM-related papers and conference talks. The most remarkable evidence of that progress for me is the world’s first SiPM-based time-of-flight capable, integrated, simultaneous PET/MR tomography scanner New SIGNA™ presented by General Electric in August 2014, because about 10 years ago they started to learn SiPM technology with samples from Amplification Technologies and with my technical support.

I will present a detailed technical analysis of what we can expect from SiPMs in the future in my talk “SiPM performance: key factors and trade-offs, possible improvements and limitations” at the APPEC Technology Forum, in Munich on April 22-23. I have no doubt that we have a lot of space for improvements in the SiPM design, performance, and functionality which will make it possible for SiPMs to solve the problems that at present seem unsolvable and may contribute to making the world a better place in the future.

V. E. Shubin, D. A. Shushakov, Avalanche Photodetectors, ch. 11 in Encyclopedia of Optical Engineering, Marcel Dekker, Ink, pp. 121–141 (2003)
I.M. Zheleznykh, Z.Ya. Sadygov, B.A. Khrenov, L.G. Tkachev, F. Zerrouk, Novel Micro-pixel Avalanche Photo Diodes and their Possible Application in Cosmic Ray / Astrophysical Researches, in Proceedings of the 30th International Cosmic Ray Conference, Vol. 5 (HE part 2), pp. 1589–1592, Mexico City, Mexico, 2008
B. Dolgoshein, V. Balagura, P. Buzhan, M. Danilov, L. Filatov, et al., Status report on silicon photomultiplier development and its applications, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment vol. 563 (2) p. 368-376 (2006)
V. Saveliev, Silicon Photomultipliers – New Era of Photon Detection, ch. 14 in Advances in Optical and Photonic Devices, InTech, pp. 249–272 (2010)

Dr. Sergey Vinogradov is a Senior Marie Curie Research Fellow at the QUASAR group, Department of Physics, University of Liverpool and the Cockcroft Institute of Accelerator Science and Technology, UK, since 2013; Senior Research Scientist at the Solid State Physics Department of the Lebedev Physical Institute of the Russian Academy of Sciences since 1992; Senior Member of SPIE – International Society for Optical Engineering; member of IEEE – Institute of Electric and Electronic Engineers. He received a MS in Physics (1981) from Moscow State University and a PhD in Solid State Physics (1992) from the P.N. Lebedev Physical Institute, Moscow, Russia. Dr. Vinogradov is an expert in R&D of solid state photodetectors including Silicon Photomultipliers (SiPM), methodology of measurements and characterization of photon detectors at a low light level down to single photons, probabilistic modelling of SiPM response and overall performance. His recent activity is focused on application of SiPM technology in high energy physics, accelerators, astrophysics and medical imaging. He is author of several patents and numerous publications in these areas as well as reviewer, session chair at international conferences and workshops on SiPM physics and applications, and organizer of the Brainstorming Workshop on SiPM time performance (Corsica, May 2015).

Submitted by Eleni Chatzichristou
APPEC Communications Officer

The Silicon Photomultiplier (SiPM) Concept and Design Development

Written by katrinlink on 11th March 2015. Posted in

11 March 2015

Typical SiPM design (KETEK)

by Sergey Vinogradov
Senior Marie Curie Research Fellow at the Dept. of Physics, University of Liverpool and the Cockcroft Institute of Accelerator Science and Technology, UK

Detection of optical light pulses with photon number resolution starting from single photons, was an ultimate goal of R&D in photodetectors for more than half a century. Vacuum photomultiplier tubes (PMTs), semiconductor avalanche photodiodes (APDs) and Geiger-mode APDs or so-called single photon avalanche diodes (SPADs) have been invented, developed and successfully utilized for low light level detection approaching to this goal by different technologies. However, all these devices are unable to provide a complete solution to a problem of photon number resolving detection at room temperature. In order to detect one photon, a detector has to produce measurable electrical signal above electronic noise of the acquisition circuit, which is about 10,000 electrons at the bandwidth of 1 GHz. It means that besides conversion of photons to electrons, a multiplication of the electrons with a very high gain (>10,000) is required to get timing certainty of the detection of about 1 ns.

In turn, multiplication mechanisms of these devices are inherently noisy producing from a single electron at the input a random number of electrons at the output. It means that two photons could hardly be resolved if the multiplication noise is comparable with the mean gain; that situation is typical for PMTs and much worse for conventional linear mode APDs. In contrast, an avalanche process in a SPAD with active quenching of a Geiger-mode breakdown could be considered as a practically noiseless multiplication with a pre-defined gain, but as a binary detector SPAD is unable to detect multi-photon pulse. The attempt to mitigate this by using an array of SPAD pixels fails since active quenching circuitry occupies a lot of space around every pixel resulting in a low fill factor of the array (~ 10%) and dramatic losses of incident photons.

SiPM technology represents an unprecedented attempt to create an ideal solid-state photon detector solving a double trade-off of avalanche processes: the higher multiplication gain inevitably results in higher multiplication noise, as well as, in longer multiplication time (the lower bandwidth). The idea on how to harness a high gain avalanche in such a way so as to make it as fast and noiseless as possible, originated from basic studies of over-breakdown avalanche processes in Metal-Insulator-Semiconductor (MIS) structures. These studies were carried out since the mid-1970s in the Lebedev Physical Institute, Moscow, Russia, in the Solid State Photodetectors group led by Dr. V. Shubin. It was found that charge carriers generated in an avalanche filament and accumulated at semiconductor-insulator interface, provide a self-quenching of a breakdown with self-calibrating of an avalanche charge due to screening of an electric field into the avalanche filament region; moreover, the higher over-breakdown voltage the lower charge fluctuations. So, this approach (known as Avalanche with Negative Feedback, ANF) assumes that an avalanche breakdown as a positive feedback process should be self-suppressed by accumulation of an avalanche charge as a negative feedback process, and both processes should be as strong as possible to get fastest and noiseless result.

However, MIS structures could be super-sensitive to photons only during non-stationary depletion of the semiconductor region adjacent to an insulator layer and it requires ramping bias voltage and synchronous gated detection of optical signals. In order to provide more convenient free-running mode of operation at DC bias voltage, a blocking insulator (SiO2) has been substituted to “insulator with leakage” – wide band gap semiconductor (SiC) and other resistive materials. Another key modification with respect to an initial non-structured planar design of MIS detectors was a separation of an active area into elements or pixels which localize and separate avalanche filaments. These improvements have been introduced by Dr. Z. Sadygov and his colleagues at the INR/MELZ collaboration. Thus, applying the ANF concept to DC operating structures of multi-pixel architecture of various designs, a new generation of high-gain low-noise avalanche detectors have been developed in Russia in the late 1980s – early 2000s. They are known by different names as Metal-Resistor-Semiconductor (MRS) APD, ANF APD, Solid State Photomultiplier (SSPM), Micro-pixel APD, and SiPM.

The most popular detector from this generation – SiPM – has been developed by the MEPhI/Pulsar collaboration led by Prof. B. Dolgoshein. The SiPM of this kind was very attractive due to the simplicity of its design and the ability to be reliably produced using semiconductor technology of moderate level. The negative feedback element in SiPM was realized as a thin-film polysilicon strip resistor connecting each individual APD pixel to a common electrode and located between pixels.

Photon number resolving response of SiPM (Hamamatsu MPPC) to short multi-photon light pulse (S. Vinogradov)

Since the mid-2000s, the SiPM design with polysilicon resistors became a conventional ground for an emerging wave of SiPM developments by Hamamatsu Photonics, SensL, FBK, STMicroelectronics, Excelitas Technologies, KETEK, and some others.

So, after all, the SiPM appears to be rather similar to a SPAD array with passive quenching, but with a much faster pixel recovery and much higher fill factor. In turn, the success of SiPM technology has inspired the re-invention of a SPAD array detector, using the achievements of modern CMOS technology and advancing its readout architecture in the so-called digital SiPM (dSiPM) developed by Philips in 2009, where the main drawback of SPAD arrays – low fill factor – was improved to about 50%.

The ANF concept and SiPM technology were also accepted by developers of non-silicon, first of all, the near-infrared (NIR) APDs starting since the mid-2000s in the USA. The designs have been adapted to III-V compound technology utilizing negative feedback based either on thin film resistor (Princeton Lightwave) or on charge accumulation at heterointerfaces of III-V layers (Jet Propulsion Laboratory, UCSD, UCLA).

Recently, an expansion of the SSPM into ultraviolet (UV) wavelengths and harsh environment operation has been possible through the development of a SiC-based UV SSPM (GE Global Research Center, 2014).

The overwhelming success and worldwide recognition of SiPM technology are due to its unprecedented performance at room temperature in a photon number resolution of multi-photon pulses, as well as, to its arrival time resolution. Highly beneficial features are the low operating voltages, insensitivity to magnetic field and to nuclear counter effect, scalability in creating large arrays, and its robustness and reliability for years, unaffected by occasional over-lighting. Now SiPM is approaching to be a real working horse for a variety of scientific and industrial applications, and mass-production pricing facilitates this progress.

However, conventional SiPM design has some drawbacks, for example the trade-off between density of pixels (dynamic range) and fill factor (detection efficiency), multiple trade-off dealt with pixel area – gain – pulse width – recovery time – crosstalk – after pulsing. Some of them are subjected to ongoing countermeasures and improvements, and some are inherent ones. There is room for alternative designs, thus a number of proprietary designs have been developing so far (Zecotek, Max Plank Institute, Novel Device Laboratory, ASTAR Institute of Microelectronics, and some others).

Further reading:
V. Saveliev, “Silicon Photomultipliers – New Era of Photon Detection”, ch. 14 in Advances in Optical and Photonic Devices, InTech, pp. 249–272 (2010)

Submitted by Eleni Chatzichristou
APPEC Communications Officer