There are many cutting-edge applications that require manipulation and sensing of single-photons in the short-wave infrared range (SWIR), mostly at wavelengths around 1310 and 1550 nm. From quantum computing to quantum optics experiments, from quantum communication to quantum metrology, there is a strong need for detecting single-photons at 1550 nm wavelength, typically travelling in optical fibers. Similarly, innovative imaging applications, such as ghost imaging, non-line-of-sight imaging and low-power LiDAR, all require eye-safe sources and the typical wavelength of choice is 1550 nm. Also, biomedical applications, such as near-infrared spectroscopy (NIRS) and diffuse correlation spectroscopy (DCS), are extending their range beyond 1.1 µm. This drive from application requirements led to the development of single-photon detectors (SPDs) implemented with innovative technologies and materials capable of detecting NIR single-photons. Among them, the most prominent devices nowadays are superconductive nanowire single-photon detectors (SNSPDs) and single-photon avalanche diodes (SPADs). While the former are capable of photon detection efficiency (PDE) above 90% at selected wavelengths, with dark counts as low as few cps, their main disadvantage is the requirement of bulky and expensive cooling systems to operate the detectors at cryogenic temperatures. SPADs, on the other hand, and specifically InGaAs/InP SPADs, can be integrated in compact, portable systems and, at a temperature of about 230 K, reachable with a small, three-stage thermoelectric cooler, have overall good performance in terms of both PDE and noise. The goal of this Ph.D. work is to push the development of InGaAs/InP SPADs in an effort to optimize the main trade-offs, with the goal of: enhancing the PDE, while keeping DCR as low as possible, and enabling the possibility of operating InGaAs/InP SPADs at higher temperatures and higher photon count-rates. This dissertation is divided into four chapters, organized as follows: Chapter 1 introduces the reasons why the detection of single photons in the near-infrared range (NIR) is becoming more and more important, by presenting some of the most advanced technology applications and highlighting their stringent requirements. Then, the main photon counting techniques and common single-photon detectors are described, as well as the key parameters to quantify their performance. The pros and cons of different SPDs are evaluated, with special attention for InGaAs/InP SPADs. Chapter 2 gives a detailed description of the internal structure of InGaAs/InP SPADs and the main design rules that govern it. Specifically, the front-illuminated planar InGaAs/InP SPADs developed at Politecnico di Milano are described, with a step-by-step illustration of the fabrication process. Then, a part of the chapter is devoted to the description of a simulation model that was implemented to accurately estimate the PDE of InGaAs/InP SPADs at different temperatures and excess bias from TCAD simulations. Finally, the other two most critical metrics to take into account when designing InGaAs/InP SPADs, afterpulsing probability and charge persistence, are described, highlighting how they are affected by design choices. Chapter 3 and Chapter 4 describe the design and characterization of the fifth and sixth generation of InGaAs/InP SPADs developed at Politecnico di Milano. Specifically, Chapter 3 deals with the design, TCAD simulation and experimental characterization of the fifth production run, which was aimed at fabricating two main devices: a low-noise 25-μm SPAD for free-space applications, with the same PDE as the previous generation SPADs but lower dark counts; a 10 μm SPAD with enhanced PDE, for fiber-pigtailed applications, with noise as low as possible. Moreover, an experiment was designed to assess the impact of the double zinc diffusion profile on charge persistence, so SPAD variations in this respect were also added to the final wafers. The TCAD simulations performed to support the new SPAD designs are described, and a complete experimental characterization of low-noise SPADs and high-efficiency SPADs is reported, as well as the results of the experimental observations of the charge persistence dependence on the double zinc diffusion profile. Chapter 4 describes the results achieved with the sixth InGaAs/InP SPADs production run. In this case, our objective was to design SPADs with reduced afterpulsing probability and charge persistence, that could be operated at higher temperatures and higher count-rate, without affecting the PDE. Harnessing the experience from the testing of previous generation devices, we designed and simulated new SPAD variants to meet this objective, as well as additional experiments to investigate innovative ideas. The TCAD simulation process and the results that led to our decisions are reported. However, due to some fabrication issues, only some of the designed wafers were delivered, and not all design specifications were met. Nevertheless, the result of the experimental characterization of the devices we received are reported in this chapter, since they were useful to carry-out an internal comparison among the fabricated sixth generation wafers.
Molte applicazioni tecnologiche all'avanguardia richiedono la rivelazione e manipolazione di singoli fotoni, nel range del vicino infrarosso (Short-Wavelength InfraRed, SWIR). Dai computer quantistici agli esperimenti di ottica quantistica, passando dalle telecomunicazioni quantistiche alla metrologia quantistica, c'è una grande richiesta di sensori in grado di rivelare singoli fotoni a una lunghezza d'onda di 1550 nm, tipicamente nei casi in cui si impiegano fibre ottiche. Allo stesso tempo, moderne applicazioni di imaging, come il ghost imaging, il non-line-of-sight imaging o il LiDAR a bassa potenza, richiedono sorgenti più sicure per l’occhio umano, e la lunghezza d'onda tipica scelta è 1550 nm. Infine, applicazioni di stampo biomedicale, come la spettroscopia nell'infrarosso (NIRS) e la spettroscopia di correlazione diffusa (DCS), stanno estendendo il loro campo oltre 1,1 µm. Questa spinta delle applicazioni ha portato allo sviluppo di rivelatori di singoli fotoni (Single-Photon Detectors, SPD) implementati con tecnologie e materiali innovativi in grado di rivelare singoli fotoni nel vicino infrarosso. Tra di essi, i dispositivi più importanti al giorno d'oggi sono i rivelatori a superconduttore (Superconductive Nanowire Single-Photon Detector, SNSPD) e i diodi a valanga (Single-Photon Avalanche Diode, SPAD). Mentre i primi sono in grado di ottenere un'efficienza di rivelazione (PDE) superiore al 90% a determinate lunghezze d'onda (dipendente, tuttavia, dalla polarizzazione della luce), con bassissimi conteggi di buio, il loro principale svantaggio è la necessità di sistemi di raffreddamento ingombranti e costosi per far funzionare i rivelatori a temperature criogeniche. Gli SPAD, d'altra parte, e in particolare gli SPAD InGaAs/InP, possono essere integrati in sistemi compatti e portatili e, a una temperatura di circa 230 K, raggiungibile con un piccolo raffreddatore termoelettrico a tre stadi, hanno buone prestazioni nel complesso, sia in termini di PDE che di rumore. L'obiettivo di questo lavoro di dottorato è far progredire lo sviluppo degli SPAD InGaAs/InP, nel tentativo di superare od ottimizzare i principali limiti: aumentare la PDE, mantenendo il DCR il più basso possibile, e consentire la possibilità di utilizzare gli SPAD InGaAs/InP a temperature più elevate e con count-rate più elevati. Questa tesi è divisa in quattro capitoli, organizzati nel seguente modo: Il Capitolo 1 introduce le ragioni per cui la rivelazione di singoli fotoni nel range SWIR sta diventando sempre più importante, presentando alcune delle applicazioni tecnologiche più avanzate ed evidenziandone i requisiti stringenti. Successivamente, vengono descritte le principali tecniche di conteggio dei fotoni e i rivelatori di singoli fotoni più comuni, così come i principali parametri per quantificarne le prestazioni. Vengono valutati i pro e i contro di diversi SPD, con particolare attenzione agli SPAD InGaAs/InP. Il Capitolo 2 fornisce una descrizione dettagliata della struttura interna degli SPAD InGaAs/InP e dei principali principi da seguire in fase di progettazione. In particolare, vengono descritti gli SPAD planari a illuminazione frontale in InGaAs/InP sviluppati al Politecnico di Milano, con una illustrazione step-by-step del processo di fabbricazione. Successivamente, una parte del capitolo è dedicata alla descrizione di un modello di simulazione implementato dall’autore per stimare la PDE degli SPAD InGaAs/InP a diverse temperature a partire da simulazioni TCAD. Infine, vengono descritte gli altri due parametri critici da prendere in considerazione nella progettazione degli SPAD InGaAs/InP: la probabilità di afterpulsing e la persistenza di carica (o charge persistence), evidenziando come siano influenzati dalle scelte progettuali. Il Capitolo 3 e il Capitolo 4 descrivono la progettazione e la caratterizzazione della quinta e sesta generazione di SPAD InGaAs/InP sviluppati al Politecnico di Milano. In particolare, il Capitolo 3 tratta della progettazione, della simulazione TCAD e della caratterizzazione sperimentale della quinta generazione di dispositivi, che aveva l'obiettivo di realizzare due dispositivi principali: un SPAD a basso rumore da 25 μm di diametro per applicazioni free-space, con la stessa PDE degli SPAD della generazione precedente ma con conteggi di buio ridotti; un SPAD da 10 μm con PDE migliorata, per applicazioni in fibra ottica (pigtail), con rumore il più basso possibile. Inoltre, è stato progettato un esperimento per valutare l'impatto del profilo di doppia diffusione di zinco sulla charge persistence, quindi determinate varianti di SPAD aggiunte ai wafer finali. Sono poi descritte le simulazioni TCAD eseguite per supportare le nuove progettazioni degli SPAD ed è riportata la completa caratterizzazione sperimentale degli SPAD a basso rumore e ad alta efficienza, così come i risultati delle osservazioni sperimentali della dipendenza della charge persistence dal profilo di doppia diffusione di zinco. Il Capitolo 4 descrive i risultati ottenuti con la sesta produzione di SPAD InGaAs/InP. In questo caso, il nostro obiettivo era progettare SPAD con una ridotta probabilità di afterpulsing e meno charge persistence, che potessero essere utilizzati a temperature più elevate e con count-rate più alti, senza influire sulla PDE. Sfruttando l'esperienza derivata dai test sui dispositivi della generazione precedente, sono stati progettate e simulate nuove varianti di SPAD per raggiungere questo obiettivo, così come ulteriori esperimenti per investigare idee innovative. Il processo di simulazione TCAD e i risultati che hanno portato alle decisioni di progettazione sono riportati in maniera dettagliata. Tuttavia, a causa di alcuni problemi di fabbricazione, solo alcuni dei wafer progettati sono stati consegnati e non tutte le specifiche di progettazione sono state soddisfatte. Tuttavia, i risultati della caratterizzazione sperimentale dei dispositivi ricevuti sono comunque riportati in questo capitolo, poiché sono stati utili per effettuare un confronto interno tra i wafer di sesta generazione fabbricati.
Single-photon avalanche diodes in III-V compound semiconductor for near-infrared applications
Fabio, Telesca
2024
Abstract
There are many cutting-edge applications that require manipulation and sensing of single-photons in the short-wave infrared range (SWIR), mostly at wavelengths around 1310 and 1550 nm. From quantum computing to quantum optics experiments, from quantum communication to quantum metrology, there is a strong need for detecting single-photons at 1550 nm wavelength, typically travelling in optical fibers. Similarly, innovative imaging applications, such as ghost imaging, non-line-of-sight imaging and low-power LiDAR, all require eye-safe sources and the typical wavelength of choice is 1550 nm. Also, biomedical applications, such as near-infrared spectroscopy (NIRS) and diffuse correlation spectroscopy (DCS), are extending their range beyond 1.1 µm. This drive from application requirements led to the development of single-photon detectors (SPDs) implemented with innovative technologies and materials capable of detecting NIR single-photons. Among them, the most prominent devices nowadays are superconductive nanowire single-photon detectors (SNSPDs) and single-photon avalanche diodes (SPADs). While the former are capable of photon detection efficiency (PDE) above 90% at selected wavelengths, with dark counts as low as few cps, their main disadvantage is the requirement of bulky and expensive cooling systems to operate the detectors at cryogenic temperatures. SPADs, on the other hand, and specifically InGaAs/InP SPADs, can be integrated in compact, portable systems and, at a temperature of about 230 K, reachable with a small, three-stage thermoelectric cooler, have overall good performance in terms of both PDE and noise. The goal of this Ph.D. work is to push the development of InGaAs/InP SPADs in an effort to optimize the main trade-offs, with the goal of: enhancing the PDE, while keeping DCR as low as possible, and enabling the possibility of operating InGaAs/InP SPADs at higher temperatures and higher photon count-rates. This dissertation is divided into four chapters, organized as follows: Chapter 1 introduces the reasons why the detection of single photons in the near-infrared range (NIR) is becoming more and more important, by presenting some of the most advanced technology applications and highlighting their stringent requirements. Then, the main photon counting techniques and common single-photon detectors are described, as well as the key parameters to quantify their performance. The pros and cons of different SPDs are evaluated, with special attention for InGaAs/InP SPADs. Chapter 2 gives a detailed description of the internal structure of InGaAs/InP SPADs and the main design rules that govern it. Specifically, the front-illuminated planar InGaAs/InP SPADs developed at Politecnico di Milano are described, with a step-by-step illustration of the fabrication process. Then, a part of the chapter is devoted to the description of a simulation model that was implemented to accurately estimate the PDE of InGaAs/InP SPADs at different temperatures and excess bias from TCAD simulations. Finally, the other two most critical metrics to take into account when designing InGaAs/InP SPADs, afterpulsing probability and charge persistence, are described, highlighting how they are affected by design choices. Chapter 3 and Chapter 4 describe the design and characterization of the fifth and sixth generation of InGaAs/InP SPADs developed at Politecnico di Milano. Specifically, Chapter 3 deals with the design, TCAD simulation and experimental characterization of the fifth production run, which was aimed at fabricating two main devices: a low-noise 25-μm SPAD for free-space applications, with the same PDE as the previous generation SPADs but lower dark counts; a 10 μm SPAD with enhanced PDE, for fiber-pigtailed applications, with noise as low as possible. Moreover, an experiment was designed to assess the impact of the double zinc diffusion profile on charge persistence, so SPAD variations in this respect were also added to the final wafers. The TCAD simulations performed to support the new SPAD designs are described, and a complete experimental characterization of low-noise SPADs and high-efficiency SPADs is reported, as well as the results of the experimental observations of the charge persistence dependence on the double zinc diffusion profile. Chapter 4 describes the results achieved with the sixth InGaAs/InP SPADs production run. In this case, our objective was to design SPADs with reduced afterpulsing probability and charge persistence, that could be operated at higher temperatures and higher count-rate, without affecting the PDE. Harnessing the experience from the testing of previous generation devices, we designed and simulated new SPAD variants to meet this objective, as well as additional experiments to investigate innovative ideas. The TCAD simulation process and the results that led to our decisions are reported. However, due to some fabrication issues, only some of the designed wafers were delivered, and not all design specifications were met. Nevertheless, the result of the experimental characterization of the devices we received are reported in this chapter, since they were useful to carry-out an internal comparison among the fabricated sixth generation wafers.File | Dimensione | Formato | |
---|---|---|---|
TELESCAfabio_PhD_Thesis_2024_05.pdf
accesso solo da BNCF e BNCR
Dimensione
15.47 MB
Formato
Adobe PDF
|
15.47 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/206194
URN:NBN:IT:POLIMI-206194