Coherent Raman scattering (CRS) is a class of third-order non-linear optical microscopy techniques for non-invasive, label-free and non-destructive imaging, with ever growing applications in biology, medicine and material science. CRS is based on the Raman scattering effect and in particular it is possible to exploit the vibrational modes of molecules, relative to chemical bonds and chemical groups, as the contrast mechanism to identify them, just like the fingerprint for humans. In CRS two input narrowband synchronized pulses are required, one at frequency OMEGAp (pump) and one at frequency OMEGAs (Stokes), and when the frequency detuning (OMEGAp - OMEGAs) between the input fields equals the resonance frequency OMEGAnu of a vibrational mode of the molecules in the sample, it is possible to make all the molecules in the focal volume to oscillate in phase in a vibrational coherence which enhances of many orders of magnitude the radiated signal with respect to the spontaneous Raman effect. The simplest implementation schemes for CRS are coherent anti-Stokes Raman scattering (CARS) and stimulated Raman scattering (SRS). This doctoral thesis work focuses on two main themes: first, the development of a multimodal CRS imaging system (Chapters 3 and 4) and, second, the implementation of novel strategies for broadband CRS (Chapters 6 and 7). The multimodal CRS imaging system developed during the PhD activity consists in a fiber-format tunable laser coupled to a home-made microscope, which has been designed and built with only off-the-shelf components to ensure easy reconfigurability and access to the optical path. The presented system has been tested in the SRS, CARS and two-photon excitation fluorescence (TPEF) imaging modalities on a variety of samples, from polymers mixture to plants, cells and tissues, demonstrating state-of-the-art performances comparable to commercial systems (Chapter 3). The use of a low-cost, compact, turn-key fiber laser source, instead of a bulk solid-state laser system, is motivated by the idea of creating a compact, user-friendly CRS imaging system that can be translated, in the near future, to clinical studies. Since fiber laser systems are intrinsically more noisy than the bulky solid-state counterparts, in terms of intensity noise, a novel and simple balanced detection system, called in-line balanced detection (IBD), to remove the laser excess noise is here presented and shot-noise limited measurements in the SRS modality are demonstrated (Chapter 4). The presented imaging system is fully controlled by a LabView software and it is able to automatically perform multimodal imaging and laser wavelength tuning, for the desired experimental requirements. The second part of the thesis is devoted to the development and implementation of novel approaches for broadband/multiplex CRS and, in particular, for the SRS modality. In fact SRS modality has two main advantages with respect to CARS. First, it does not suffer from the so-called non-resonant background (NRB), which deviates CARS spectra from the spontaneous Raman ones, and thus molecule recognition in SRS is straightforward. Second, the signal in SRS is proportional to the number of molecules in the focal volume, and not to the square of this number like in CARS, making possible quantitative microscopy. In the literature there are only few implementations of broadband/multiplex SRS, due to complexity of the detection system which is based on sophisticated electronics, like lock-in amplifiers, to detect the small SRS signal on the large linear background of the Stokes/pump with high sensitivity. Here two new methodologies are presented, both based on time-domain measurements and single-pixel detection, which strongly simplifies the overall system architecture with respect to previous implementations. The first technique is called Fourier-transform SRS (FT-SRS) (Chapter 6) and it is based on a home-made interferometer, while the second one is called photonic time-stretch SRS (PTS-SRS) and it is based on time-domain dispersive Fourier transform (Chapter 7).
Lo scattering Raman coerente è una tecnica di microscopia ottica non lineare del terzo ordine che permette di effettuare misure non invasive, senza marcatori e non distruttive su qualsiasi tipo di campione; essa ha diversi ambiti di applicazione che spaziano dalla biologia alla medicina, fino alle scienze dei materiali. Lo scattering Raman coerente è basato sull'effetto Raman e, in particolare, è possibile utilizzare i modi vibrazionali delle molecole, derivanti dai legami chimici presenti, come meccanismo di contrasto per identificare le molecole stesse in modo univoco, proprio come accade con le impronte digitali per gli esseri umani. In questa tecnica vengono utilizzati due impulsi a banda stretta sincronizzati, il primo a frequenza OMEGAp (pompa) e il secondo a frequenza OMEGAs (Stokes), e quando la differenza tra queste due frequenze (OMEGAp - OMEGAs) eguaglia la frequenza di risonanza OMEGAnu associata ad un modo vibrazionale del campione è possibile fare oscillare in fase coerentemente tutte le molecole presenti nel campione stesso generando un segnale che è molti ordini di grandezza più intenso di quello che si otterrebbe con l'effetto Raman spontaneo. I due metodi di implementazione più semplici per questa tecnica sono lo scattering coerente anti-Stokes (CARS) e lo scattering Raman stimolato (SRS). Questa tesi di dottorato si occupa dei seguenti due temi: primo, lo sviluppo di un sistema di microscopia multimodale basato sullo scattering Raman coerente (Capitoli 3 e 4) e, secondo, l'implementazione di nuove strategie per ottenere lo scattering Raman coerente su una banda larga (Capitoli 6 e 7). Il sistema di microscopia multimodale per lo scattering Raman coerente sviluppato in questo progetto di tesi di dottorato è costituito da un laser in fibra accordabile in frequenza accoppiato ad un microscopio artigianale, che è stato sviluppato e costruito esclusivamente con componenti disponibili in commercio per assicurare un facile adattamento alle richieste specifiche dell'esperimento. Il sistema è stato testato con le tecniche SRS, CARS e con la fluorescenza a due fotoni (TPEF) su diversi tipi di campioni, tra cui polimeri, piante, cellule e tessuti (Capitolo 3). I risultati ottenuti sono paragonabili a quelli ottenuti da sistemi commerciali. La scelta di utilizzare un laser in fibra, al posto di un laser a stato solido, è una conseguenza diretta della volontà di realizzare un sistema a basso costo, chiavi in mano e compatto, che potrà essere utilizzato in studi clinici a breve. Dal momento che i laser in fibra sono intrinsecamente più rumorosi rispetto a quelli a stato solido, è stato sviluppato un nuovo e semplice sistema di misura bilanciata, chiamato in-line balanced detection (IBD), per rimuovere il rumore in eccesso presente e ottenere così misure SRS limitate dal solo rumore shot (Capitolo 4). Il sistema di microscopia qui descritto è totalmente controllato con un software LabView che permette di automatizzare le misure. La seconda parte della tesi si focalizza sullo sviluppo di nuove tecniche per ottenere lo scattering Raman coerente su una larga banda di frequenze vibrazionali e, in particolare, è stato applicato nella modalità SRS. Infatti la modalità SRS offre diversi vantaggi rispetto al CARS. Primo, la tecnica SRS non soffre della presenza del rumore di background non-risonante (NRB) che devia gli spettri CARS da quelli ottenuti con il Raman spontaneo, permettendo, in questo modo, di riconoscere facilmente le molecole presenti. Secondo, il segnale SRS è direttamente proporzionale al numero di molecole presenti nel volume focale, e non al quadrato di queste come nel CARS, permettendo di essere quantitativi nelle misure. In letteratura si trovano poche realizzazioni per la tecnica SRS a banda larga, dal momento che la attrezzatura elettronica richiesta per effettuare tali misure è molto sofisticata e costosa, come gli amplificatori lock-in. Infatti è richiesta la capacità di distinguere con alta sensitività un piccolo segnale, cioè lo SRS, su un background molto intenso, cioè la pompa o lo Stokes. In questo lavoro vengono presentate due nuove strategie, per ottenere misure SRS a banda larga, basate su tecniche nel dominio del tempo e sull'utilizzo di un singolo fotodiodo. Ciò permette di semplificare estremamente l'architettura del sistema rispetto a proposte precedenti. La prima tecnica presentata si chiama Fourier-transform SRS (FT-SRS) (Capitolo 6) e si basa sull'utilizzo di uno speciale interferometro. La seconda tecnica invece si chiama photonic time-stretch SRS (PTS-SRS) (Capitolo 7) e si basa sulla trasformata di Fourier dispersiva nel dominio del tempo.
Innovative techniques for coherent Raman microscopy
CRISAFI, FRANCESCO
Abstract
Coherent Raman scattering (CRS) is a class of third-order non-linear optical microscopy techniques for non-invasive, label-free and non-destructive imaging, with ever growing applications in biology, medicine and material science. CRS is based on the Raman scattering effect and in particular it is possible to exploit the vibrational modes of molecules, relative to chemical bonds and chemical groups, as the contrast mechanism to identify them, just like the fingerprint for humans. In CRS two input narrowband synchronized pulses are required, one at frequency OMEGAp (pump) and one at frequency OMEGAs (Stokes), and when the frequency detuning (OMEGAp - OMEGAs) between the input fields equals the resonance frequency OMEGAnu of a vibrational mode of the molecules in the sample, it is possible to make all the molecules in the focal volume to oscillate in phase in a vibrational coherence which enhances of many orders of magnitude the radiated signal with respect to the spontaneous Raman effect. The simplest implementation schemes for CRS are coherent anti-Stokes Raman scattering (CARS) and stimulated Raman scattering (SRS). This doctoral thesis work focuses on two main themes: first, the development of a multimodal CRS imaging system (Chapters 3 and 4) and, second, the implementation of novel strategies for broadband CRS (Chapters 6 and 7). The multimodal CRS imaging system developed during the PhD activity consists in a fiber-format tunable laser coupled to a home-made microscope, which has been designed and built with only off-the-shelf components to ensure easy reconfigurability and access to the optical path. The presented system has been tested in the SRS, CARS and two-photon excitation fluorescence (TPEF) imaging modalities on a variety of samples, from polymers mixture to plants, cells and tissues, demonstrating state-of-the-art performances comparable to commercial systems (Chapter 3). The use of a low-cost, compact, turn-key fiber laser source, instead of a bulk solid-state laser system, is motivated by the idea of creating a compact, user-friendly CRS imaging system that can be translated, in the near future, to clinical studies. Since fiber laser systems are intrinsically more noisy than the bulky solid-state counterparts, in terms of intensity noise, a novel and simple balanced detection system, called in-line balanced detection (IBD), to remove the laser excess noise is here presented and shot-noise limited measurements in the SRS modality are demonstrated (Chapter 4). The presented imaging system is fully controlled by a LabView software and it is able to automatically perform multimodal imaging and laser wavelength tuning, for the desired experimental requirements. The second part of the thesis is devoted to the development and implementation of novel approaches for broadband/multiplex CRS and, in particular, for the SRS modality. In fact SRS modality has two main advantages with respect to CARS. First, it does not suffer from the so-called non-resonant background (NRB), which deviates CARS spectra from the spontaneous Raman ones, and thus molecule recognition in SRS is straightforward. Second, the signal in SRS is proportional to the number of molecules in the focal volume, and not to the square of this number like in CARS, making possible quantitative microscopy. In the literature there are only few implementations of broadband/multiplex SRS, due to complexity of the detection system which is based on sophisticated electronics, like lock-in amplifiers, to detect the small SRS signal on the large linear background of the Stokes/pump with high sensitivity. Here two new methodologies are presented, both based on time-domain measurements and single-pixel detection, which strongly simplifies the overall system architecture with respect to previous implementations. The first technique is called Fourier-transform SRS (FT-SRS) (Chapter 6) and it is based on a home-made interferometer, while the second one is called photonic time-stretch SRS (PTS-SRS) and it is based on time-domain dispersive Fourier transform (Chapter 7).File | Dimensione | Formato | |
---|---|---|---|
2018_02_PhD_Crisafi.pdf
accessibile in internet per tutti
Descrizione: Testo della tesi
Dimensione
24.19 MB
Formato
Adobe PDF
|
24.19 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/137887