Design and characterization of single photon avalanche diodes arrays L. Neri a,b , S. Tudisco a,Ã , L. Lanzan  o a,b , F. Musumeci a,b , S. Privitera a,b , A. Scordino a,b , G. Condorelli c , G. Fallica c , M. Mazzillo c , D. Sanfilippo c , G. Valvo c a INFN-Laboratori Nazionali del Sud, via S.Sofia 62, I-95125 Catania, Italy b Universita’ di Catania, via S.Sofia 64, I-95123 Catania, Italy c ST-Microelectronics, Stradale Primosole 50, I-95100 Catania, Italy article info Available online 3 July 2009 Keywords: SPAD Single photon counting Bi-dimensional array Timing Imaging abstract During the last years, in collaboration with ST-Microelectronics, we developed a new avalanche photo sensor, single photon avalanche diode (SPAD) see Ref.[S. Privitera, et al., Sensors 8 (2008) 4636[1];S. Tudisco et al., IEEE Sensors Journal 8 (2008) 1324 [2]]. Such sensor is able to detect and count, with excellent performance, single photons. Today, design and monolithic integration of many elements represents the most challenging goal in this field. In the present contribution we report on design and realization of a bi-dimensional array of SPADs. Such array is realized in standard planar technology and is the first prototype designed for imaging applications. The final aim is to identify the position and the arrival time of the impinging photons. Lay-out, readout strategy, characterization test are discussed. & 2009 Elsevier B.V. All rights reserved. 1. Introduction Bi-dimensional single photon avalanche diode (SPAD) array design needs a full electric description of the single element and of bi-dimensional solutions. As presented in Ref. [3,4], we plan to develop a new electrical SPAD connection concept by performing the signal readout from the cathode and the anode, and sharing of anode’s and cathode’s contact between several diodes. In this work, we show the description of the single diodes, the simulation of the grid connections, and the comparison with real sensor behaviour. 2. Electric SPAD model calibration Every pixel of our matrix sensor is constituted by a SPAD diode, for which we will take into account the electric model proposed by Cova in 1996 [5]. Starting from such model we measured the maximum amplitude and the area of the output signal at various source voltages. The area of the signal is proportional to the charge collected from the total capacitance of the system. Due to the linear relationship between the charge and the power source voltage, the total capacity of the system C, and the breakdown voltage, V b of the diode can be determined via linear fit of data. It resulted C ¼ 1.3570.05 pF and V b ¼ 29.671.6 V. The signal rise with the avalanche creation time cannot be measurable with our readout electronics. The signal slope can be described as the discharge of the total capacity through the internal diode resistance, like a normal RC circuit. The exponential fit of the quenching part of the signal provides the evaluation of the diode resistance. It appears that at voltages 20% of overvoltage the value of the resistance reaches a constant value of 1.4270.06 kO. This behaviour is important for our simulations and is not trivial. By approximating capacities in parallel, and using the Eq. (9) of Ref. [5] we evaluated the capacitances of the model as C s ¼ V S V E R d R s C ð1Þ where V S is the signal maximum, V E the excess bias voltage, R d the diode resistance, R s the readout resistance, C the total capacity and C s the stray capacity. The difference between the total and the stray capacity is the diode capacity. ARTICLE IN PRESS Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/nima Nuclear Instruments and Methods in Physics Research A Fig. 1. Comparison between calibrated electric model simulation, and the real signal of a single SPAD. 0168-9002/$ - see front matter & 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.nima.2009.06.085 Ã Corresponding author. Tel./fax: +39 095 542 263 E-mail address: tudisco@lns.infn.it (S. Tudisco). Nuclear Instruments and Methods in Physics Research A 617 (2010) 432–433