Une impulsion laser ultra-courte et ultra-intense se propageant dans un gaz de faible densité est capable d'accélérer une partie des électrons de ce gaz à des énergies relativistes, de l'ordre de quelques centaines de MeV, sur des distances de seulement quelques millimètres. Pendant leur accélération et dû à leur mouvement transverse, ces électrons émettent de plus un rayonnement X fortement collimaté et dirigé vers l'avant appelé rayonnement bétatron. Les caractéristiques de cette source la rendent intéressante pour son utilisation en imagerie à ultra-haute résolution.Dans ce manuscrit, nous explorons trois axes de travail autour de cette source à l'aide de simulations réalisées avec les codes Particle-In-Cell CALDER et CALDER-Circ. Nous commençons ainsi par étudier la création d'une source bétatron avec des impulsions laser de durée picoseconde et d'énergie kilojoule, donc plus longues et plus puissantes que celles habituellement utilisées par la communauté. Nous montrons que malgré les paramètres inhabituels de ces impulsions lasers il est toujours possibles de générer des sources X, et ce dans deux régimes différents.Ensuite, afin de comprendre une partie des différences généralement observées entre expériences et simulations, nous montrons dans une autre étude que l'utilisation dans les simulations de profils lasers réalistes au lieu de profils parfaitement Gaussiens dégrade fortement les performances de l'accélérateur laser-plasma et de la source bétatron. De plus, ceci conduit à un meilleur accord qualitatif et quantitatif avec l'expérience.Enfin nous explorons plusieurs techniques pour augmenter l'émission X basées sur une manipulation des profils de plasmas utilisés pour l'accélération. Nous trouvons que l'utilisation d'un gradient transverse ou d'une marche de densité conduisent tous deux à une augmentation de l'amplitude du mouvement transverse des électrons, et donc de l'énergie émise par la source bétatron. Alternativement, nous montrons que cet objectif peut-être atteint par la transition d'un régime de sillage laser vers un régime d'accélération par sillage plasma induit par une augmentation de la densité. L'accélération des électrons est optimisée dans le premier régime, tandis que l'émission X est fortement favorisée dans le second. / An ultra-short and ultra-intense laser pulse propagating in a low-density gas can accelerate in its wake a part of the electrons ionized from the gas to relativistic energies of a few hundreds of MeV over distances of a few millimeters only. During their acceleration, as a consequence of their transverse motion, these electrons emit strongly collimated X-rays in the forward direction, which are called betatron radiations. The characteristics of this source turn it into an interesting tool for high-resolution imagery.In this thesis, we explore three different axis to work on this source using simulations on the Particles-In-Cells codes CALDER and CALDER-Circ. We first study the creation of a betatron X-ray source with kilojoule and picosecond laser pulses, for which duration and energy are then much higher than usual in this domain. In spite of the unusual laser parameters, we show that X-ray sources can still be generated, furthermore in two different regimes.In a second study, the generally observed discrepancies between experiments and simulations are investigated. We show that the use of realistic laser profiles instead of Gaussian ones in the simulations strongly degrades the performances of the laser-plasma accelerator and of the betatron source. Additionally, this leads to a better qualitative and quantitative agreement with the experiment.Finally, with the aim of improving the X-ray emission, we explore several techniques based on the manipulation of the plasma density profile used for acceleration. We find that both the use of a transverse gradient and of a density step increases the amplitude of the electrons transverse motions, and then increases the radiated energy. Alternatively, we show that this goal can also be achieved through the transition from a laser wakefield regime to a plasma wakefield regime induced by an increase of the density. The laser wakefield optimizes the electron acceleration whereas the plasma wakefield favours the X-ray emission.

This thesis investigates the use of high-power lasers for synchrotron radiation sources with high brilliance, from the EUV to the hard X-ray spectral range. Hereby lasers accelerate electrons by laser-wakefield acceleration (LWFA), act as optical undulators, or both. Experimental evidence shows for the first time that LWFA electron bunches are shorter than the driving laser and have a length scale comparable to the plasma wavelength. Furthermore, a first proof of principle experiment demonstrates that LWFA electrons can be exploited to generate undulator radiation. Building upon these experimental findings, as well as extensive numerical simulations of Thomson scattering, the theoretical foundations of a novel interaction geometry for laser-matter interaction are developed. This new method is very general and when tailored towards relativistically moving targets not being limited by the focusability (Rayleigh length) of the laser, while it does not require a waveguide. In a theoretical investigation of Thomson scattering, the optical analogue of undulator radiation, the limits of Thomson sources in scaling towards higher peak brilliances are highlighted. This leads to a novel method for generating brilliant, highly tunable X-ray sources, which is highly energy efficient by circumventing the laser Rayleigh limit through a novel traveling-wave Thomson scattering (TWTS) geometry. This new method suggests increases in X-ray photon yields of 2-3 orders of magnitudes using existing lasers and a way towards efficient, optical undulators to drive a free-electron laser. The results presented here extend far beyond the scope of this work. The possibility to use lasers as particle accelerators, as well as optical undulators, leads to very compact and energy efficient synchrotron sources. The resulting monoenergetic radiation of high brilliance in a range from extreme ultraviolet (EUV) to hard X-ray radiation is of fundamental importance for basic research, medical applications, material and life sciences and is going to significantly contribute to a new generation of radiation sources and free-electron lasers (FELs).

Die vorliegende Arbeit beschäftigt sich in Experiment und Theorie mit Laser-Plasma beschleunigten Elektronen und optischen Undulatoren zur Erzeugung von brillianter Synchrotronstrahlung. Zum ersten Mal wird experimentell nachgewießen, dass laserbeschleunigte Elektronenpulse kürzer als 30 fs sind. Ferner werden solche Elektronenpulse erstmalig in einem Demonstrationsexperiment durch einen magnetischen Undulator als Synchrotronstrahlenquelle genutzt. Aufbauend auf diesen experimentellen Erkenntnissen, sowie umfangreichen numerischen Simulationen zur Thomsonstreuung, werden die theoretischen Grundlagen einer neuartigen Interaktionsgeometrie für Laser-Materie Wechselwirkungen entwickelt. Diese neue, in der Anwendbarkeit sehr allgemeine Methode basiert auf raum-zeitlicher Laserpulsformung durch nichtlineare Winkeldispersion wie diese durch VLS- (varied-line spacing) Gitter erzeugt werden kann und hat den Vorteil nicht durch die Fokussierbarkeit des Lasers (Rayleighlänge) begrenzt zu sein. Zusammen mit laserbeschleunigten Elektronen ermöglicht dieser traveling-wave Thomson scattering (TWTS) benannte Ansatz neuartige, nur auf optischer Technologie basierende Synchrotronstrahlenquellen mit Zentimeter bis Meter langen optische Undulatoren. Die hierbei mit existierenden Lasern erzielbaren Brillianzen übersteigen diese bestehender Thomsonquellen-Designs um 2-3 Größenordnungen. Die hier vorgestellten Ergebnisse weisen weit über die Grenzen der vorliegenden Arbeit hinaus. Die Möglichkeit Laser als Teilchenbeschleuniger und auch optischen Undulator zu verwenden führt zu bauartbedingt sehr kompakten und energieeffizienten Synchrotronstrahlungsquellen. Die hieraus resultierende monochromatische Strahlung hoher Brillianz in einem Wellenlängenbereich von extremen ultraviolett (EUV) zu harten Röntgenstrahlen ist für die Grundlagenforschung, medizinische Anwendungen, Material- und Lebenswissenschaften von fundamentaler Bedeutung und wird maßgeblich zu einer neuen Generation ultrakurzer Strahlungsquellen und freien Elektronenlasern (FELs) beitragen. / This thesis investigates the use of high-power lasers for synchrotron radiation sources with high brilliance, from the EUV to the hard X-ray spectral range. Hereby lasers accelerate electrons by laser-wakefield acceleration (LWFA), act as optical undulators, or both. Experimental evidence shows for the first time that LWFA electron bunches are shorter than the driving laser and have a length scale comparable to the plasma wavelength. Furthermore, a first proof of principle experiment demonstrates that LWFA electrons can be exploited to generate undulator radiation. Building upon these experimental findings, as well as extensive numerical simulations of Thomson scattering, the theoretical foundations of a novel interaction geometry for laser-matter interaction are developed. This new method is very general and when tailored towards relativistically moving targets not being limited by the focusability (Rayleigh length) of the laser, while it does not require a waveguide. In a theoretical investigation of Thomson scattering, the optical analogue of undulator radiation, the limits of Thomson sources in scaling towards higher peak brilliances are highlighted. This leads to a novel method for generating brilliant, highly tunable X-ray sources, which is highly energy efficient by circumventing the laser Rayleigh limit through a novel traveling-wave Thomson scattering (TWTS) geometry. This new method suggests increases in X-ray photon yields of 2-3 orders of magnitudes using existing lasers and a way towards efficient, optical undulators to drive a free-electron laser. The results presented here extend far beyond the scope of this work. The possibility to use lasers as particle accelerators, as well as optical undulators, leads to very compact and energy efficient synchrotron sources. The resulting monoenergetic radiation of high brilliance in a range from extreme ultraviolet (EUV) to hard X-ray radiation is of fundamental importance for basic research, medical applications, material and life sciences and is going to significantly contribute to a new generation of radiation sources and free-electron lasers (FELs).

This thesis investigates the use of high-power lasers for synchrotron radiation sources with high brilliance, from the EUV to the hard X-ray spectral range. Hereby lasers accelerate electrons by laser-wakefield acceleration (LWFA), act as optical undulators, or both. Experimental evidence shows for the first time that LWFA electron bunches are shorter than the driving laser and have a length scale comparable to the plasma wavelength. Furthermore, a first proof of principle experiment demonstrates that LWFA electrons can be exploited to generate undulator radiation. Building upon these experimental findings, as well as extensive numerical simulations of Thomson scattering, the theoretical foundations of a novel interaction geometry for laser-matter interaction are developed. This new method is very general and when tailored towards relativistically moving targets not being limited by the focusability (Rayleigh length) of the laser, while it does not require a waveguide. In a theoretical investigation of Thomson scattering, the optical analogue of undulator radiation, the limits of Thomson sources in scaling towards higher peak brilliances are highlighted. This leads to a novel method for generating brilliant, highly tunable X-ray sources, which is highly energy efficient by circumventing the laser Rayleigh limit through a novel traveling-wave Thomson scattering (TWTS) geometry. This new method suggests increases in X-ray photon yields of 2-3 orders of magnitudes using existing lasers and a way towards efficient, optical undulators to drive a free-electron laser. The results presented here extend far beyond the scope of this work. The possibility to use lasers as particle accelerators, as well as optical undulators, leads to very compact and energy efficient synchrotron sources. The resulting monoenergetic radiation of high brilliance in a range from extreme ultraviolet (EUV) to hard X-ray radiation is of fundamental importance for basic research, medical applications, material and life sciences and is going to significantly contribute to a new generation of radiation sources and free-electron lasers (FELs).

Die vorliegende Arbeit beschäftigt sich in Experiment und Theorie mit Laser-Plasma beschleunigten Elektronen und optischen Undulatoren zur Erzeugung von brillianter Synchrotronstrahlung. Zum ersten Mal wird experimentell nachgewießen, dass laserbeschleunigte Elektronenpulse kürzer als 30 fs sind. Ferner werden solche Elektronenpulse erstmalig in einem Demonstrationsexperiment durch einen magnetischen Undulator als Synchrotronstrahlenquelle genutzt. Aufbauend auf diesen experimentellen Erkenntnissen, sowie umfangreichen numerischen Simulationen zur Thomsonstreuung, werden die theoretischen Grundlagen einer neuartigen Interaktionsgeometrie für Laser-Materie Wechselwirkungen entwickelt. Diese neue, in der Anwendbarkeit sehr allgemeine Methode basiert auf raum-zeitlicher Laserpulsformung durch nichtlineare Winkeldispersion wie diese durch VLS- (varied-line spacing) Gitter erzeugt werden kann und hat den Vorteil nicht durch die Fokussierbarkeit des Lasers (Rayleighlänge) begrenzt zu sein. Zusammen mit laserbeschleunigten Elektronen ermöglicht dieser traveling-wave Thomson scattering (TWTS) benannte Ansatz neuartige, nur auf optischer Technologie basierende Synchrotronstrahlenquellen mit Zentimeter bis Meter langen optische Undulatoren. Die hierbei mit existierenden Lasern erzielbaren Brillianzen übersteigen diese bestehender Thomsonquellen-Designs um 2-3 Größenordnungen. Die hier vorgestellten Ergebnisse weisen weit über die Grenzen der vorliegenden Arbeit hinaus. Die Möglichkeit Laser als Teilchenbeschleuniger und auch optischen Undulator zu verwenden führt zu bauartbedingt sehr kompakten und energieeffizienten Synchrotronstrahlungsquellen. Die hieraus resultierende monochromatische Strahlung hoher Brillianz in einem Wellenlängenbereich von extremen ultraviolett (EUV) zu harten Röntgenstrahlen ist für die Grundlagenforschung, medizinische Anwendungen, Material- und Lebenswissenschaften von fundamentaler Bedeutung und wird maßgeblich zu einer neuen Generation ultrakurzer Strahlungsquellen und freien Elektronenlasern (FELs) beitragen. / This thesis investigates the use of high-power lasers for synchrotron radiation sources with high brilliance, from the EUV to the hard X-ray spectral range. Hereby lasers accelerate electrons by laser-wakefield acceleration (LWFA), act as optical undulators, or both. Experimental evidence shows for the first time that LWFA electron bunches are shorter than the driving laser and have a length scale comparable to the plasma wavelength. Furthermore, a first proof of principle experiment demonstrates that LWFA electrons can be exploited to generate undulator radiation. Building upon these experimental findings, as well as extensive numerical simulations of Thomson scattering, the theoretical foundations of a novel interaction geometry for laser-matter interaction are developed. This new method is very general and when tailored towards relativistically moving targets not being limited by the focusability (Rayleigh length) of the laser, while it does not require a waveguide. In a theoretical investigation of Thomson scattering, the optical analogue of undulator radiation, the limits of Thomson sources in scaling towards higher peak brilliances are highlighted. This leads to a novel method for generating brilliant, highly tunable X-ray sources, which is highly energy efficient by circumventing the laser Rayleigh limit through a novel traveling-wave Thomson scattering (TWTS) geometry. This new method suggests increases in X-ray photon yields of 2-3 orders of magnitudes using existing lasers and a way towards efficient, optical undulators to drive a free-electron laser. The results presented here extend far beyond the scope of this work. The possibility to use lasers as particle accelerators, as well as optical undulators, leads to very compact and energy efficient synchrotron sources. The resulting monoenergetic radiation of high brilliance in a range from extreme ultraviolet (EUV) to hard X-ray radiation is of fundamental importance for basic research, medical applications, material and life sciences and is going to significantly contribute to a new generation of radiation sources and free-electron lasers (FELs).