Nuclear Instruments and Methods in Physics Research A 718 (2013) 295–296
Contents lists available at SciVerse ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
Development of thin pixel detectors on epitaxial silicon for HEP experiments Maurizio Boscardin a,n, Daniela Calvo b, Gabriele Giacomini a, Richard Wheadon b, Sabina Ronchin a, Nicola Zorzi a a b
FBK, CMM, Via Sommarive 18, I-38123 Povo, Trento, Italy INFN and Dipartimento di Fisica, Universita di Torino, Via Pietro Giuria, I-10125 Torino, Italy
a r t i c l e i n f o
abstract
Available online 10 November 2012
The foreseen luminosity of the new experiments in High Energy Physics will require that the innermost layer of vertex detectors will be able to sustain fluencies up to 1016 neq/cm2. Moreover, in many experiments there is a demand for the minimization of the material budget of the detectors. Therefore, thin pixel devices fabricated on n-type silicon are a natural choice to fulfill these requirements due to their rad-hard performances and low active volume. We present an R&D activity aimed at developing a new thin hybrid pixel device in the framework of PANDA experiments. The detector of this new device is a p-on-n pixel sensor realized starting from epitaxial silicon wafers and back thinned up to 50–100 mm after process completion. We present the main technological steps and some electrical characterization on the fabricated devices before and after back thinning and after bump bonding to the front-end electronics. & 2012 Elsevier B.V. All rights reserved.
Keywords: Particle physics detectors Pixel hybrid detector Thin detectors
1. Introduction The realization of hybrid pixel sensors on epitaxial material is a very attractive method to get thin [1] and radiation hard [2] detectors. Indeed, it is possible to reduce the sensor thickness by mechanical or chemical techniques, by removing the portion of the heavily doped material, which acts as a mere mechanical support, and leaving only the portion of the epitaxial silicon. This manufacturing process is simple and inexpensive since it does not require the use of a support wafer. This paper reports on preliminary results for epitaxial pixel sensors compatible with the PANDA ToPix_v3 readout chip.
2. Detector fabrication Detectors were processed at FBK on 100 mm diameter, n-type epi wafers. Starting substrates have a resistivity of 0.02 and 3600 O cm, for bulk and epitaxial layers, respectively. Different epi thicknesses have been used, ranging from 50 to 100 mm. With reference to the schematic flow chart of Fig. 1, the fabrication process can be outlined as follows: a) isolation oxide growth, p-plus and n-plus regions doping by ion implantation, dopants drive-in and silicon nitride deposition; b) contact holes opening, n Correspondence to: FBK-irst, Micro Technologies Lab, Via Sommarive 18, I-38123 Povo, Trento, Italy. Tel.: þ39 0461 314458. E-mail address:
[email protected] (M. Boscardin).
0168-9002/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.nima.2012.10.100
deposition and definition of metal and passivation; c) bump bonding and detector thinning. The sensor fabrication process (excluding the bumps deposition) required 5 mask steps. After the bump bonding, the wafers have been thinned to partially remove the heavy doped substrate by grinding, so obtaining a thin sensor.
3. Electrical characterization 3.1. Test structures After fabrication, static measurements on test structures (i.e. 1D large area diodes, MOS capacitors, gated diodes, Van der Pauw resistors) allowed to assess the good quality of the fabrication process. In particular, bulk leakage currents as low as few nA/cm2 at full depletion (typically few volt) were measured. Regardless of the epitaxial layer thickness, generation times t in the order of 3–10 ms were extracted (to be compared with t 20 ms for FZ substrates, which were processed in parallel, as a reference). Surface generation velocities s0 have been found in the order of 2 cm/s, at the same level of FZ wafers. On multiguard diodes with the same termination of the pixel sensors, breakdown voltages in excess of 500 V have been generally measured. This assures the high-voltage handling capability required for operation after irradiation. After substrate thinning, such test structures did not show degradation of any of the electrical performances.
296
M. Boscardin et al. / Nuclear Instruments and Methods in Physics Research A 718 (2013) 295–296
silicon nitride
1.E-05
silicon oxide
epi
p+
p+
n+
Si-bulk
Current [A]
1.E-06
1.E-07 100 epi+20 Cz S1: 100epi+525Cz S2: 100epi+525Cz S3: 100epi+525Cz S3: 100epi+525Cz
1.E-08
1.E-09 epi
p+
p+
n+
Si-bulk
passivation
Si-bulk
20
40 60 Bias voltage [V]
80
100
Fig. 2. Total current of bump bonded pixel sensors: thinned detectors (continuous line) and not thinned detector (dashed line).
FE chip
epi
0
p+
p+
substrate do not show a worsening of the electrical parameters after bumping and thinning processes. This technology is a possible solution for the fabrication of thin silicon detectors for HEP application.
n+ Acknowledgments
bump
Fig. 1. Schematic process flow for the fabrication of pixel detectors on epitaxial wafer.
3.2. Pixel sensors Some pixel sensors composed of 32 20 pixels (pitch 100 mm 100 mm) have been thinned and bump-bonded by IZM. The readout chip is ToPix_v3 (developed at INFN-Torino) [3]. Total-current measurements of bump bonded sensors do not show breakdown up to 100 V (Fig. 2). The observed difference of the leakage currents in thinned vs. not thinned sensors is in the range of the typical spread evaluated by test structures. New thinned detectors will be delivered in June and additional measurements are planned.
4. Conclusions We have shown the technology to get thinned detectors starting from epitaxial silicon wafers. The devices realized on thin silicon
We acknowledge the support from the MEMS2 joint project of the Istituto Nazionale di Fisica Nucleare (INFN), Fondazione Bruno Kessler (FBK) and Provincia Autonoma Trento (PAT). References [1] D. Calvo, et al., Nuclear Instruments and Methods A 624 (2010) 290. [2] E. Fretwurst, et al., Nuclear Instruments and Methods A 552 (2005) 7. [3] G. Mazza, et al., Journal of Instrumentation 7 (2012) C02015.