This thesis reports on the development of glass-based femtosecond laser sources around 2 µm wavelength. In order to be able to produce 2 µm radiation the dopants used were trivalent Thulium (Tm³⁺) and trivalent Holmium (Ho³⁺) that could be optically pumped with Ti:Sapphire radiation at 0.8 µm and semiconductor disk lasers (SDL) at 1.2 µm. The samples were produced at Leeds University and polished in-house in bulk form and deployed in free space laser cavities. Tellurite compounds doped with Tm³⁺ produced stable continuous wave 1.94 µm radiation when pumped at 800 nm with a maximum efficiency of 28.4% with respect to the absorbed power and maximum output power around 120 mW when pumped using a Ti:Sapphire operating around 0.8 µm. The radiation was broadly tunable across 130 nm. Tm³⁺-Ho³⁺ doubly doped tellurite samples lased around 2.02 µm with maximum efficiency of 25.9% and with P[subscript(OUT)]=75 mW and a smooth tunability of 125 nm. The fluorogermanate glass doped with Tm³⁺ gave an absorbed to output power efficiency of 50%. The maximum continuous wave output powers obtained were around 190 mW and limited by the available pump power at 0.8 µm. These results together with a very low threshold of 60 mW of incident power were comparable to the crystalline counterparts to this gain medium. The Tm3+ tellurite and the Tm³⁺-Ho³⁺ tellurite compounds were also pumped by an SDL operating at 1215 nm to obtain an indication of the viability of such a pump scheme. The results were a maximum internal slope efficiency of 22.4% with a highest output power of 60 mW. The comparison demonstrated that 1.2 µm pumping was competitive with using 0.8 µm wavelength. The use of semiconductor saturable absorbing mirror (SESAM) technology was used for the modelocking of these lasers. The SESAM was produced in Canada and implanted with As⁺ ions in order to reduce the relaxation time. Trains of transform-limited laser pulses at 222 MHz as short as 410 fs centred at 1.99 µm were produced for the first time with a bulk Tm³⁺:Fluorogermanate glass. The maximum average output power obtained was of 84 mW. The same SESAM deployed on the Tm³⁺-Ho³⁺ Tellurite compounds gave trains of transform-limited pulses as short as 630 fs at 2.01 µm with a repetition rate of 143 MHz and a maximum averaged output power of 43 mW. The regime of propagation obtained was soliton-like and the modelocking was self-starting. The results obtained with bulk glass were very promising and open interesting research pathways within the realm of amorphous bulk gain media.

In this thesis two previously separate fields of study are brought together: optical micromanipulation and ultrashort laser research. Here, the benefits of combining the high peak powers of ultrashort pulsed lasers and conventional optical micromanipulation techniques are explored. As optical trapping has been studied extensively, the focus of this research is on optical guiding. Moreover, the emphasis is on the use of Bessel beams as these have been shown to offer greater guiding distances than comparable Gaussian beams. The studies within this thesis show that optical guiding in Bessel and Gaussian beams is governed by the average power of the laser. However, the benefits of guiding with ultrashort pulsed lasers to exploit multi-photon processes become evident as the demonstration of simultaneous optical guiding and second harmonic generation in microscopic nonlinear crystal fragments is detailed. This work is developed by using ultrashort pulses to induce two-photon excitation-induced fluorescence in the guiding medium. This allows direct visualisation of the beam-particle interaction and measurement of the reconstruction of the Bessel beam around an object. Some studies using two-photon excitation to investigate Bessel beam penetration through turbid media are discussed. Finally, the work is concluded by exploring the use of pulsed white-light lasers in optical guiding. The wavelength-dependent propagation and reconstruction properties of the white-light Bessel beam are studied before some preliminary optical guiding experiments are discussed. From this, the broad bandwidth of the supercontinuum source is found to offer extended guiding distances in Gaussian beams thereby negating the need for Bessel beams.

This dissertation elucidates a greater understanding of protein folding and unfolding processes during the lifetimes of electrospray and nano-spray droplets in laser electrospray mass spectrometry (LEMS) and nano-laser electrospray mass spectrometry (nano-LEMS) measurements, respectively. The similarity in mass spectral features obtained from conventional electrospray measurements for supercharged proteins with those of LEMS measurements suggested that supercharging phenomena occurs in the electrospray droplets during the droplet desolvation process. It was observed that the laser vaporization of protein from condensed phase into the electrospray droplets containing denaturing electrospray solution and a supercharging reagent resulted in the increase in ion abundance of higher charge states in comparison with electrospray measurements. Conversely, the addition of solution additives with varying gas phase basicity in the electrospray solvent resulted in charge reduction for unfolded protein upon laser vaporization from condensed phase into the charged electrospray droplets. The extent of charge reduction and the fraction of folded protein within the electrospray droplets was found to be dependent upon both the extent of protein denaturation in the solution prior to laser vaporization and the gas phase basicity of solution additives. The ability of the LEMS technique to analyze molecules from solution with high matrix effects was established by the successful detection of protein molecules from solution with high salt concentration. Experiments with LEMS enabled the detection of a protonated protein feature as the dominating peak in the mass spectra for up to 250 mM sodium chloride while conventional electrospray resulted in predominantly salt-adducted features, with suppression of the protonated protein ions for the salt concentration of 5 mM. This dissertation also expanded upon the use of a reaction system to measure the lifetimes of laser vaporized liquid droplets coupled with electrospray and nano-spray postionization mass spectrometry. Electrospray and nanospray droplet lifetimes were measured to be 4.5±0.6 ms and 1.4±0.3 ms using LEMS and nano-LEMS measurements, respectively. Time dependent protein folding measurements using LEMS revealed intermediate states during protein folding processes which are often limited in conventional electrospray measurements where bulk solution in manipulated (change in pH) to achieve protein folding. / Chemistry

Les caméras à balayage de fente sont les instruments de détection directe de la lumière les plus précis en termes de résolution temporelle. Ces instruments sont capables de capturer des évènements de l’ordre de la picoseconde à un taux de répétition d’une centaine de mégahertz. Cependant, les performances de la caméra sont limitées par de nombreux phénomènes propres au fonctionnement de cette dernière mais aussi au système l’implémentant. Plusieurs effets dégradant la résolution temporelle sont étudiés. Le premier axe exploré concerne la synchronisation de la caméra avec l’évènement lumineux capturé. Cette investigation débouche sur le développement d’un discriminateur à fraction constante permettant de déclencher la caméra avec un jitter inférieur à 200 fs RMS. Une autre étude présente l’impact qu’ont le bruit d’amplitude et le bruit de phase des lasers usuellement utilisés avec la caméra sur sa synchronisation. Enfin une analyse des phénomènes intrinsèques à la photocathode de la caméra permet d’évaluer la variation du temps de transit des électrons dans celle-ci. / Streak cameras are the direct light detection instruments that are the best in terms of temporal resolution. Those instruments can capture picosecond light events at a hundred megahertz repetition rate. However their characteristics are limited by various phenomena specific to the camera and the implementing system. Several effects that affect the temporal resolution are studied. The first examined line deals with the synchronization of the camera with the studied light event. This inquiry led to the design of a constant fraction discriminator allowing a sub 200 fs RMS jitter triggering. Another study shows the impact of the usually used laser amplitude noise and phase noise on the system’s synchronization. Finally, an analysis of the camera’s photocathode intrinsic phenomena allows estimating the transit time variation of the electrons within the vacuum tube.

Ce travail porte sur l’étude expérimentale de la réponse optique linéaire et non-linéaire de nanoparticules métalliques individuelles. Pour cela une technique optique originale en champ lointain, basée sur la modulation spatiale de l’échantillon, a été mise en place et combinée avec une technique pompe-sonde résolue en temps à l'échelle femtoseconde. Dans un premier temps nous nous sommes intéressés à la caractérisation optique de nanoparticules modèles formées par une nanosphère d’argent encapsulée dans une couronne de silice. Nous avons mis en évidence la conséquence principale du confinement quantique électronique sur la réponse optique des nanoparticules : l'élargissement de leur résonance plasmon de surface avec la réduction de la taille (proportionnel à l’inverse du rayon, pour des nanosphères). Cette étude détaillée n’avait jamais été réalisée, car elle nécessite des mesures sur des nanoparticules uniques (pour s'affranchir des effets d’élargissement inhomogène de la résonance dus aux dispersions en taille, géométrie et environnement) et de taille connue. Ceci a été possible grâce à la mesure quantitative de leur section efficace d'extinction et/ou à la corrélation directe de l'image optique du nano-objet avec son image par microscopie électronique. La deuxième partie de ce travail a été consacrée à l'étude de la réponse optique linéaire et non-linéaire résolue en temps de nanobâtonnets d’or. Ces systèmes sont particulièrement intéressants car leurs propriétés optiques peuvent être contrôlées en modifiant leur géométrie (rapport d’aspect en particulier). Leur spectre d'extinction et sa dépendance en polarisation ont tout d'abord été étudiés au niveau de l'objet individuel. La comparaison de cette réponse optique linéaire avec des modèles théoriques a permis d'obtenir des informations sur leur orientation et géométrie. La réponse optique non-linéaire ultrarapide du même nanobâtonnet a ensuite été mesurée par une technique pompe-sonde femtoseconde à très haute sensibilité. La nanoparticule étudiée étant totalement caractérisée optiquement, une mesure quantitative de sa non-linéarité a ainsi pu être réalisée, et son origine physique déterminée à partir d'un modèle théorique. / This work focuses on the experimental study of linear and nonlinear optical responses of single metal nanoparticles. For this, an original far field optical technique, based on the spatial modulation of the sample, was set-up and combined with a femtosecond time-resolved pump-probe technique. We have first investigated the optical properties of model nanoparticles, formed by a silver nanosphere coated with a silica shell. We have highlighted the main consequence of electronic quantum confinement on the optical response of a nanoparticle: the broadening of its localized surface plasmon resonance with size reduction (proportional to the inverse of the radius, for nanospheres). This detailed study had never been realized, because it requires measurements on single nanoparticles (to eliminate the influence of inhomogeneous effects on the resonance due to size, geometry and environment dispersions) and of known size. This was possible thanks to the quantitative measurement of extinction cross-section and / or correlation of the optical image of the nano-object with its electron microscopy image. The second part of this work has been devoted to the study of the linear and time-resolved non-linear optical responses of gold nanorods. These systems are particularly interesting because their optical properties can be controlled by changing their geometry (in particular their aspect ratio). First, the extinction spectrum and its polarization dependence have been studied on individual objects. Comparison of the linear optical response with theoretical models provides information on orientation and geometry of the nanorod. The ultrafast nonlinear optical response of the nanorod was then measured with a high sensitivity femtosecond pump-probe technique. The investigated nanoparticles being fully characterized optically, a quantitative measurement of their non-linearity could be achieved, and its physical origin determined using a theoretical model.

Dans ce travail nous présentons nos travaux réalisés sur le contrôle de la générationd’harmoniques d’ordres élevés dans les gaz. Dans un premiers temps nous montronscomment, en générant les harmoniques hors du foyer du faisceau IR, il est possiblede contrôler la phase spatiale des harmoniques dans le milieu générateur permettantd’obtenir un front d’onde divergent, collimaté ou convergent. Par cette méthode nousmontrons qu’il est possible de focaliser les harmoniques à des distances pouvant atteindresix longueur de Rayleigh après le point focal du faisceau IR. Nous avons ensuiteétudié des faisceaux harmoniques XUV présentant des distributions spatio-spectralesen champ lointain structurées. Dans cette étude nous observons l’influence d’un irisde diamètre variable positionné avant la focalisation de l’IR. Dans un troisième tempsnous étudions des méthodes de contrôle du spectre harmonique. Tout d’abord nousavons contrôlé finement la longueur d’onde centrale des harmoniques par modificationdu contenu spectral de l’IR en superposant deux impulsions IR retardées. Puis nousavons utilisé les effets collectifs de la génération d’harmoniques afin de favoriser uneharmonique spécifique ou un groupe d’harmoniques en champ lointain. Enfin, nousprésentons une méthode de caractérisation de la durée d’impulsions attosecondes dansle domaine temporel. Cette méthode, appelée ionisation par paliers, utilise l’ionisationcomme sonde pour mesurer des durée d’impulsions pouvant atteindre la centained’attoseconde. / We present our work on the control of high order harmonic generation in gases.We first show how, by generating the harmonics outside the focus of the IR beam,it is possible to control the spatial phase of the harmonics in the generating mediumallowing to obtain a divergent, collimated or convergent wavefront. With this methodwe show that it is possible to focus the harmonics up to six Rayleigh length after thefocal point of the IR beam. Then we study XUV harmonic beams presenting structuredspacio-spectral distributions in the far field. In this study, we observe the influence ofthe diameter of an iris positioned before the focusing of the IR. In a third step we studymethods for controlling the harmonic spectrum. First, we finely control the harmonicscentral wavelength by modifiying the spectral content of the IR by adding two delayedIR pulses. Then we used the collective effects of the high order harmonic generationin order to foster a specific harmonic or a group of harmonics in the far field. Finally,we present a method for characterizing the duration of attosecond pulses in the timedomain. This method, called ionization ladder, uses ionization as a probe to measurepulse duration of up to hundreds of attosecond.

In this study we applied ultra-short laser pulses on a biological sample (Arabidopsis), in order to cut it precisely in a square pattern and subsequently use it for studying stress generation mechanisms. For this purpose, we utilized femtosecond laser pulses at 100 fs pulse width and 80 MHz repetition rate. We took two processing parameters into consideration such as laser power, laser exposure time which is related to the stage speed. Therefore, we were able to find the laser optimum conditions for ablation of biological tissues. The mutant and wildtype (control) obtained from laser cutting with a size of 500 µm × 500 µm were directly transferred (in-situ with laser cutting) into a microfabricated chamber containing ~500 nanoliters deionized water for measuring ion conductivity. The ion conductivity is a signature of cell-death mechanisms caused by various stresses. A light with intensity of 100 µmol was exposed to the samples for 2 hours and 20 minutes as a source of stress. A quantitative electrical analysis with high accuracy was assured by utilizing a microchamber, which enables a measurement in nanoliter volume. We measured the impedance which is reciprocal of conductivity using a lock-in amplifier and a precise current source at frequency of 130 Hz. Initially high impedance of mutant sample tended to drop within 2 hours and finally approached the constant value which signified that the cell death mechanism was complete. However, the wildtype sample demonstrated approximately constant impedance (conductivity) during the experiment.

Nous avons étudié la spectroscopie et la dynamique d’excitation d’un système hybride constitué de molécules et de nanoparticules (NP) sur couche isolante, qui est aussi un catalyseur modèle (NP de Pd épitaxiées sur une couche mince de MgO sur Ag(100)). Nous avons mis en évidence le rôle de la forme et de la taille des NP ainsi que de l’épaisseur de la couche d’oxyde sur l’interaction entre NP et molécule de CO, par des expériences fondamentales capables de différentier les sites d’adsorption (spectroscopie laser vibrationnelle par somme de fréquences (SFG)). De plus, des expériences pompe-sonde nous ont permis de sonder la dynamique d’interaction des électrons excités dans les NP avec les molécules. Une analyse combinée par LEED et STM nous a permis de déterminer les meilleures conditions de croissance du film de MgO. Par la suite des NP de palladium ont été épitaxiées sur ce film avec une densité et une distribution de taille satisfaisantes. Les spectres SFG montre une forte dépendance de la fréquence de vibration avec la taille des NP et le taux de couverture en CO. Le modèle d’interaction dipolaire que nous avons développé met en évidence le fait que le déplacement de la fréquence de CO avec le taux de couverture et la taille des NP a deux causes : l’interaction dipolaire entre molécules d’une part, qui est modélisable, et d’autre part la variation de la liaison chimique molécule-substrat quand la couverture en adsorbats varie. Le modèle nous a permis de montrer que la polarisabilité vibrationnelle de CO varie d’environ 40 % dans notre gamme de taille de nanoparticule. La diminution de la force de la liaison chimique se traduit par la décroissance de la fréquence à couverture nulle avec la taille des NP. Ces variations vont dans le sens de la littérature : quand la taille des NP diminue, la densité d’états électroniques diminue, les liaisons Pd-Pd se contractent et l’énergie d’adsorption des molécules de CO diminue. L’excitation des électrons des NP et du substrat d’Ag se manifeste par une réponse spectroscopique et par la photodésorption de CO. On observe le découplage de l’excitation produite dans l’argent quand l’épaisseur de la couche d’oxyde dépasse quelques plans atomiques. On observe clairement un effet de taille sur l’efficacité de l’excitation électronique des NP sur les molécules de CO, qui diminue avec la taille des NP. Ceci montre que le confinement des électrons dans la particule a davantage pour effet d’augmenter la vitesse de relaxation électronique vers les phonons, que d’exciter plus efficacement les molécules adsorbées. Un modèle optique de double couche (NP/oxyde) sur un substrat d’argent et un modèle à trois températures (électrons, phonons et adsorbats) ont été développés dans le but d’interpréter quantitativement ces observations. Le modèle optique fait apparaître des variations très fortes de l’absorption par la couche de nanoparticules avec son épaisseur équivalente : les variations du coefficient de réflexion et l’effet des interférences multiples ne sont pas négligeables. Les résultats du modèle à trois températures montrent que la température électronique d’une couche continue est deux fois plus importante que celle des NP de taille finie. / We have studied the spectroscopy and excitation dynamics of a hybrid system consisting of molecules and nanoparticles (NPs) on insulating layer, which is also a model catalyst (Pd NPs grown on a thin layer of MgO on Ag(100)). We have highlighted the role of the shape and the NPs size as well as the thickness of the oxide layer in the interaction between NPs and CO molecules, by fundamental experiments allowing to differentiate the adsorption sites (Sum Frequency Generation (SFG)). In addition, pump-probe experiments allowed us to probe the dynamics of interaction between the photoexcited electrons bath in the NPs and the molecules. A combined study by LEED and STM allowed us to determine the growing conditions of MgO film. Subsequently palladium NPs were grown on this film with satisfying density and distribution size. The SFG spectra show a strong dependence of the vibrational frequency with the NPs size and the CO coverage. A dipole interaction model was developed showing that the CO frequency shift with the coverage and the NPs size has two causes: the dipolar interaction between molecules on the one hand, which is modeled, and on the other hand the variation of the molecule-substrate chemical bond when the adsorbate coverage varies. The model has allowed us to show that the vibrational polarizability of CO changes by 40 % in our range of NPs size. The decrease in the strength of the chemical bond results in the decrease of the frequency at zero coverage when the NPs size decreases. These variations are consistent with the literature: upon a decrease in the particle size, the electronic DOS decreases, the Pd-Pd bonds contract and the adsorption energy of CO molecules decreases. The electronic excitation of the NPs and the silver substrate manifested by the spectroscopic response and the CO photodesorption. We observed the decoupling of the electronic excitation produced in silver when the thickness of the oxide layer exceeds a few atomic planes. There is a clear size effect on the efficiency of electronic excitation of NPs on the CO molecules, which decreases as the NPs size decreases. This shows that the effect of electrons confinement in the particle rather consists in an increase in the relaxation rate of electrons to phonons, than in a more efficient excitation of the adsorbed molecules. A double-layer optical model (NP/oxide) on a silver substrate and a three-temperature model (electrons, phonons and adsorbates) have been developed to quantitatively interpret these observations. The optical model results show a very strong variation of the absorption intensity by the layer with its equivalent thickness: variation of the reflection and the effect of multiple interference are not negligible. The three temperatures model results show that the electronic temperature of a continuous layer is two times more important than in the case of finite size NPs.

Recently, laser light sources of different regimes have emerged as an essential tool in the biophotonics research area. Classic applications include, for example: manipulating single cells and their subcellular organelles, sorting cells in microfluidic channels and the cytoplasmic delivery of both genetic and non-genetic matter of varying sizes into mammalian cells. In this thesis several new findings specifically in the optical cell sorting as well as in the photo-transfection study fields are presented. In my optical cell sorting and guiding investigations, a new technique for enhancing the dielectric contrast of mammalian cells, which is a result of cells naturally engulfing polymer microspheres from their environment, is introduced. I explore how these intracellular dielectric tags influence the scattering and gradient forces upon these cells from an externally applied optical field. I show that intracellular polymer microspheres can serve as highly directional optical scatterers and that the scattering force can enable sorting through axial guiding onto laminin coated glass coverslips upon which the selected cells adhere. Following this, I report on transient photo-transfection of mammalian cells including neuroblastomas (rat/mouse and human), embryonic kidney, Chinese hamster ovary as well as pluripotent stem cells using a tightly focused titanium sapphire femtosecond pulsed laser beam spot. These investigations permitted advanced biological studies in femtosecond laser transfection: firstly, the influence of cell passage number on the transfection efficiency; secondly, the possibility to enhance the transfection efficiency via whole culture treatments of cells thereby, synchronizing them at the mitotic (M phase) as well as the synthesis phases (S phase) of the cell cycle; thirdly, this methodology can activate the up-regulation of the protective heat shock protein 70 (hsp70). Finally, I show that this novel technology can also be used to transfect mouse embryonic stem (mES) cell colonies and the ability of differentiating these cells into the extraembryonic endoderm.

Nous présentons nos travaux sur la génération d’harmoniques élevée (HHG) avec des impulsions térawatt femtosecondes. Nous avons effectué l’HHG avec les impulsions fondamentales de haute énergie et caractérisé l’émission de XUV spatialement et spectralement de manière monocoup et nous avons observé plusieurs structures dans le champ lointain. Ces structures sont très robustes et reproductibles et ont été observés dans de nombreux types de gaz et des géométries de génération. Sans caractérisation monocoup spatiale et spectrale les structures ne sont pas visibles. Nous avons développé des simulations simples pour identifier ces structures et nous avons pu observer des structures similaires. Nous les avons identifiés comme étant liées à la cohérence spatiale de la source XUV et la diffraction dans le champ lointain. Dans le champ proche, la phase et l’amplitude harmonique évoluent spatialement et temporellement et leurs profils sont fortement modulés. Ces profils modulés diffractent et créent des spectres structurés dans le champ lointain. Nous avons observé que la propagation dans un milieu générateur fin a peu d’influence sur les structures. Nous démontrons une mise en forme spatiale de l’impulsion fondamentale via l’optique adaptative et leur avantage pour HHG. Une optique adaptative nous permet d’avoir un faisceau à profil de phase régulier. Un tel faisceau est utilisé pour HHG avec un faisceau de grand diamètre et contrôle du faisceau XUV est démontrée. Nous avons développé une technique de post-compression de haute énergie, et nous avons obtenu des impulsionsde 10 fs et 10 mJ dans un profil quasi gaussien. La technique repose sur l’automodulation de phase induite par l’ionisation et est compatible avec des impulsions niveau TW de haute énergie. Nous avons effectué des HHG avec ces impulsions et obtenu des spectres XUV quasi continu avec des structures spatiales et spectrales. Nous avons effectué des simulations simples etdes simulations de SFA et nous avons observé des structures similaires même sans prendre en compte la propagation dans le milieu. / We present our work on high harmonic generation with TW femtosecond pulses. We performed HHG with high energy femtosecond pulses and characterize the generated XUV emission spatially and spectrally at the single-shot basis and we observed many structures in spatially resolved XUV spectra in the far field. Those structures are very robust and reproducible and have been observed in many different gases and generation geometries. Without spatial and spectral characterization on the single-shot basis the structures are not visible. We developed simple simulations to identify those structures and we observed similar structures as experimentally. We identified them as a result of spatial coherence of the XUV source and the diffraction to the far field. In the near field, the harmonic amplitude and phase are spatially and temporally dependent and their profiles are strongly modulated. Such modulated profiles diffract to structured spatially resolved spectra in the far field. We observed that propagation of the XUV in the generating medium has little influence on the structures. We demonstrate spatial shaping of the driving pulses via adaptive optics and their advantage for HHG. An adaptive optics allows us to have the driving beam of regular spatial profile and phase even out of focus. Such a beam is used for HHG with a large diameter driving beam and control of the XUV beam is demonstrated. We developed a high energy TW post-compression technique and we obtained pulses of 10 fs and 10 mJ in a quasi-Gaussian spatial profile. The technique is based on ionization-inducedself-phase-modulation and is compatible with high energy TW level pulses. We performed HHG with such TW pulses and obtained XUV quasi-continuum spectra with spectral and spatial structures. We performed simple simulations and SFAsimulations and we observed similar structures even without considering the XUV propagation in the medium.