|Thesis abstract: |
My Ph.D. research was focused on the study, the design, the fabrication and the characterization of new Single-photon Avalanche Diodes (SPADs), made in InGaAs/InP, for the photon detection in the near-infrared range, up to 1700nm. This research exploited the collaboration with the Canadian National Research Council (NRC), where the devices were fabricated.
In the last years, single photon detection has become more and more important when very faint or ultra-fast optical signals must be measured. Advanced applications are found in many fields, such as measurement of fluorescent decays, characterization of new materials, VLSI testing, quantum cryptography, etc... In order to be able to detect single photons, the sensor must have a very high sensitivity. Different kinds of single-photon detectors have been developed. Among the others, solid-state detectors, like SPADs, are reliable and robust and very portable.
A detailed simulation activity was carried out to obtain the optimal structure for InGaAs/InP SPADs, analyzing the effect of all structural parameters on SPAD performance. A custom ¿SPAD simulator¿ was developed, able to simulate the detection efficiency at different wavelength, the dark count rate and afterpulsing, in different operating conditions. Many different SPAD structures were designed with different diffusion masks, in order to experimentally study the effect of diffusion geometries on SPAD performances and to extract unknown parameters. Moreover, some specific new structures were introduced, with special patterns in the diffusion masks, to improve SPAD performances. The manufacturing processes involved in SPAD production were analyzed in depth and, thanks to the collaboration with NRC, it was possible to slightly adapt it to our specific requirements. Another important part of the present research activity was the characterization of commercially-available InGaAs/InP SPADs (with state-of-the-art performances), not just to acquire data on their performances, but also to obtain information about their behaviors and problems related to their design and fabrication. The main contributions to dark count rate were identified and it was also deeply studied the negative effect of electric field uniformity on the time resolution of the detector.
Concerning the PoliMi InGaAs/InP SPADs, the performances of the first produced devices were very promising. Afterpulsing, which is the main bottleneck of InGaAs/InP SPADs is significantly lower in PoliMi SPADs than in commercial SPADs. Moreover, the time resolution (lower than 80ps) and the electric field uniformity of PoliMi SPADs are better than commercial SPADs and their timing response distribution is sharper, without tail or second peaks.