The interaction of active molybdenum field emission microcathode arrays with oxygen, water, carbon dioxide, methane, hydrogen and helium gases was studied. Experiments were setup to measure the emission characteristics as a function of gas exposures. The resulting changes in the surface work function of the tips were determined from the Fowler-Nordheim plots. The kinetics of the FEA-gas interaction were studied by observing the ion species originating from the array during and after gas exposures with a high resolution quadrupole mass spectrometer. With the work function data and the mass spectrometry information, the mechanisms responsible for emission degradation and subsequent device recovery after exposures have been determined. The data obtained was used in estimating the device lifetimes under various vacuum environments. Also it was found that the gas exposure effects are similar in dc and pulsed modes of operation of the arrays, thus permitting the use of dc mode testing as an effective acceleration method in establishing the device lifetimes under various vacuum conditions. The vacuum conditions required for the long term emission current stability and reliability of vacuum microelectronic devices employing FEAs are established. Exposure of Mo field emitter arrays to oxygen bearing species like oxygen, water and carbon dioxide resulted in serious emission current degradation. Whereas, exposure to methane and hydrogen caused a significant increase in emission current. The control of residual gases like 02, C02 and H20 in the vacuum envelope is essential for the emission current stability and long term reliability of vacuum microelectronic devices employing field emission microcathode technology.

<div>Amongst the various classes of mass analyzers, the quadrupole ion trap (QIT) is by far the most versatile. Although it can achieve only modest resolution (unit) and mass accuracy (101-102 ppm), it has high sensitivity and selectivity, can operate at pressures exceeding 10-3 torr, is tolerant to various electrode imperfections, and has single analyzer tandem mass spectrometry (MS/MS) capabilities in the form of product ion scans. These characteristics make the QIT ideal for mass spectrometer miniaturization, as most of the fundamental performance metrics of the QIT do not depend on device size. As such, the current drive in miniature systems is to adopt miniature ion traps in various forms – 3D, linear, toroidal, rectilinear, cylindrical, arrays, etc.</div><div><br></div><div>Despite being one of the two common mass analyzers with inherent MS/MS capabilities (the other being the Fourier transform ion cyclotron resonance mass spectrometer), it is commonly accepted that the QIT cannot perform one-dimensional precursor ion scans and neutral loss scans - the other two main MS/MS scan modes - or two-dimensional MS/MS scans. The former two are usually conducted in triple quadrupole instruments in which a first and third quadrupole are used to mass select precursor and product ions while fragmentation occurs in an intermediate collision cell. The third scan can be accomplished by acquiring a product ion scan of every precursor ion, thus revealing the entire 2D MS/MS data domain (precursor ion m/z vs. product ion m/z). This, however, is not one scan but a set of scans. Because the ion trap is a tandem-in-time instrument rather than a tandem-in-space analyzer, precursor ion scans, neutral loss scans, and 2D MS/MS are, at best, difficult.</div><div><br></div><div>Yet miniature mass spectrometers utilizing quadrupole ion traps for mass analysis would perhaps benefit the most from precursor scans, neutral loss scans, and 2D MS/MS because they generally have acquisition rates (# scans/s) an order of magnitude lower than their benchtop counterparts. This is because they usually use a discontinuous atmospheric pressure interface (DAPI) to reduce the gas load on the backing pumps, resulting in a ~1 scan/s acquisition rate and making the commonly-used data-dependent acquisition method (i.e. obtaining a product ion scan for every abundant precursor ion) inefficient in terms of sample consumption, time, and instrument power. Precursor and neutral loss scans targeting specific molecular functionality of interest - as well as 2D MS/MS – are more efficient ways of moving through the MS/MS data domain and thus pair quite readily with miniature ion traps.</div><div><br></div><div>Herein we demonstrate that precursor ion scans, neutral loss scans, and 2D MS/MS are all possible in a linear quadrupole ion trap operated in the orthogonal double resonance mode on both benchtop and portable mass spectrometers. Through application of multiple resonance frequencies matching the secular frequencies of precursor and/or product ions of interest, we show that precursor ions can be fragmented mass-selectively and product ions ejected simultaneously, preserving their relationship, precursor ion -> product ion + neutral, in the time domain and hence allowing the correlation between precursor and product ions without prior isolation. By fixing or scanning the resonance frequencies corresponding to the targeted precursor and product ions, a precursor ion scan or neutral loss scan can be conducted in a single mass analyzer. We further show that 2D MS/MS - acquisition of all precursor ion m/z values and a product ion mass spectrum for every precursor ion, all in a single scan - is possible using similar methodology. These scan modes are particularly valuable for origin-of-life and forensic applications for which the value of miniature mass spectrometers is readily evident.</div>

The use of short columns, higher carrier gas velocity and fast temperature programs in Comprehensive Two-Dimensional Gas Chromatography coupled to Time-of- Flight Mass Spectrometry (GC x GC/TOFMS) technique is expected to increase the speed of analysis up to several orders of magnitude when compared to conventional gas chromatography (GC) or gas chromatography/mass spectrometry (GC/MS). A systematic evaluation of the GC x GC/TOFMS configuration for high-speed applications has received little attention in the literature. The feasibility of High Speed Comprehensive Two-Dimensional Gas Chromatography coupled to Mass Spectrometry (High speed GC x GC/MS) for complex mixtures has been investigated in this thesis. A particular focus was placed on comparing conventional scanning quadrupole mass spectrometry (qMS) with a newly available non-scanning time-of-flight instruments (TOFMS). Experiments were carried out using GC/qMS, GC x GC/qMS, GC/TOFMS and GC x GC/TOFMS both in normal (slow) and fast temperature rates coupled with high frequency modulation in GC x GC. Initially a complex mixture consists of 24 semivolatile compounds was used as the analyte for the above purpose. In the initial experiments parameters like acquisition rate and duty cycle for qMS were determined to evaluate the effectiveness of the instrument for fast analysis. The practical duty cycle value obtained for the qMS was only about 18 % for single ion and one compound at a dwell time of 10 ms in SIM mode. In both high-speed GC/qMS and high-speed GC x GC/qMS techniques only about 40 % of the components in the complex mixture were found to be well separated. The acquisition rate of scanning instruments like qMS is incompatible for fast eluting peaks in high speed GC. TOFMS that has an acquisition rate of several hundred spectra per second offer the potential to define the fast GC peaks accurately. The high quality spectra from TOFMS also enable deconvolution of coeluting peaks in the complex mixtures. The advantage of the automated spectral deconvolution is demonstrated for the identification of the coeluting peaks in the complex mixtures. Coelution of peaks is also observed with highspeed GC/TOFMS technique. The high-speed GC x GC/TOFMS was also tested with two different analyte system ??? A pesticide mixture and platformate (an aromatic mixture) to evaluate the suitability for high-speed analysis of complex mixtures. A poor resolution was observed for the pesticide mixture in the two-dimensional plane and it appeared, as there was nearly no orthogonal separation in the second dimension. The platformate mixture displayed a better two-dimensional separation. Chromatographic peak resolution is not really a primary requirement for locating and identifying the coeluting compounds in high-speed GC x GC/TOFMS technique. However, it was observed that the high-speed GC x GC/TOFMS too faced problem to unscramble the mass spectra of those compounds with similar structure and sharing the same unique masses.

<p>&#65279;<p></p><p>Giant multipole resonances are a fundamental property of nuclei and</p><p>arise from the collective motion of the nucleons inside</p><p>the nucleus. Careful studies of these resonances and their properties provides</p><p>insight into the nature of nuclear matter and constraints</p><p>which can be used to test our theories. </p><p></p></p><p><p></p><p>An investigation of the Isovector Giant Quadrupole Resonance (IVGQR)</p><p>in ²⁰⁹Bi has been preformed using the High Intensity &gamma;-ray</p><p>Source (HI&gamma;S) facility. Intense nearly monochromatic</p><p>polarized &gamma;-rays were incident upon a ²⁰⁹Bi target producing</p><p>nuclear Compton scattered &gamma;-rays that were detected using the HI&gamma;S</p><p>NaI(Tl) Detector Array (HINDA). The HINDA array consists of six</p><p>large (10''x10'') NaI(Tl) core crystals, each surrounded by an</p><p>optically segmented 3'' thick NaI(Tl) annulus. The scattered &gamma;-rays</p><p>both parallel and perpendicular to the plane of polarization were</p><p>detected at scattering angles of 55&deg; and 125&deg; with</p><p>respect to the beam axis. This was motivated by the realization that</p><p>the term representing the interference between the electric dipole</p><p>(E1) and electric quadrupole (E2) amplitudes, which appears in the</p><p>theoretical expression for the ratio of the polarized cross sections,</p><p>has a sign difference between the forward and backward angles and also</p><p>changes sign as the incident &gamma;-ray energy is scanned over the E2</p><p>resonance energy. The ratio of cross sections perpendicular and</p><p>parallel to the plane of polarization of the incident &gamma;-ray were</p><p>measured for thirteen different incident &gamma;-ray energies between 15 and</p><p>26 MeV at these two angles and used to extract the parameters of the</p><p>IVGQR in ²⁰⁹Bi.</p><p></p></p><p><p></p><p>The polarization ratio was calculated at 55&deg; and</p><p>125&deg; using a model consisting of E1 and E2 giant resonances as</p><p>well as a modified Thomson scattering amplitude. The parameters of the E1 giant</p><p>resonance came from previous measurements of the Giant Dipole</p><p>Resonance (GDR) </p><p>in ²⁰⁹Bi. The finite size of the nucleus was</p><p>accounted for by introducing a charge form factor in the (modified)</p><p>Thomson amplitude. This form factor was obtained from</p><p>measurements of the charge density in inelastic electron scattering</p><p>experiments. </p><p></p></p><p><p></p><p>The resulting curves were fit to the data by varying the</p><p>E2 parameters until a minimum value of the &chi;² was found.</p><p>The resulting parameters from the fit yield an IVGQR in ²⁰⁹Bi</p><p>located at E_res=23.0&plusmn;0.13(stat)&plusmn;0.25(sys) MeV</p><p>with a width of &Gamma;=3.9&plusmn;0.7(stat)&plusmn;1.3(sys) MeV and a</p><p>strength of 0.56&plusmn;0.04(stat)&plusmn;0.10(sys) Isovector Giant</p><p>Quadrupole Energy Weighted Sum Rules (IVQEWSRs).</p><p></p></p><p><p></p><p>The ability to make precise measurements of the parameters of the</p><p>IVGQR demonstrated by this work opens up new challenges to both</p><p>experimental and theoretical work in nuclear structure. A detailed</p><p>search for the missing sum rule strength in the case of ²⁰⁹Bi should</p><p>be performed. In addition, a systematic study of a number of nuclei</p><p>should be studied with this technique in order to carefully examine</p><p>the A dependence of the energy, width and sum rule strength of the</p><p>IVGQR as a function of the mass number A. The unique properties of</p><p>the HI&gamma;S facility makes it the ideal laboratory at which to perform</p><p>these studies.</p><p></p></p><p><p></p><p>Such a data base will provide more stringent tests of nuclear</p><p>theory. The effective parameters of collective models can be fine</p><p>tuned to account for such precision data. This should lead to new</p><p>insights into the underlying interactions responsible for the nature</p><p>of the IVGQR. Furthermore, with the recent advances in computational</p><p>power and techniques, microscopic shell model based calculations</p><p>should be possible and could lead to new insights into the underlying</p><p>properties of nuclear matter which are responsible for the collective</p><p>behavior evidenced by the existence and properties of the IVGQR.</p><p></p></p> / Dissertation

The use of short columns, higher carrier gas velocity and fast temperature programs in Comprehensive Two-Dimensional Gas Chromatography coupled to Time-of- Flight Mass Spectrometry (GC x GC/TOFMS) technique is expected to increase the speed of analysis up to several orders of magnitude when compared to conventional gas chromatography (GC) or gas chromatography/mass spectrometry (GC/MS). A systematic evaluation of the GC x GC/TOFMS configuration for high-speed applications has received little attention in the literature. The feasibility of High Speed Comprehensive Two-Dimensional Gas Chromatography coupled to Mass Spectrometry (High speed GC x GC/MS) for complex mixtures has been investigated in this thesis. A particular focus was placed on comparing conventional scanning quadrupole mass spectrometry (qMS) with a newly available non-scanning time-of-flight instruments (TOFMS). Experiments were carried out using GC/qMS, GC x GC/qMS, GC/TOFMS and GC x GC/TOFMS both in normal (slow) and fast temperature rates coupled with high frequency modulation in GC x GC. Initially a complex mixture consists of 24 semivolatile compounds was used as the analyte for the above purpose. In the initial experiments parameters like acquisition rate and duty cycle for qMS were determined to evaluate the effectiveness of the instrument for fast analysis. The practical duty cycle value obtained for the qMS was only about 18 % for single ion and one compound at a dwell time of 10 ms in SIM mode. In both high-speed GC/qMS and high-speed GC x GC/qMS techniques only about 40 % of the components in the complex mixture were found to be well separated. The acquisition rate of scanning instruments like qMS is incompatible for fast eluting peaks in high speed GC. TOFMS that has an acquisition rate of several hundred spectra per second offer the potential to define the fast GC peaks accurately. The high quality spectra from TOFMS also enable deconvolution of coeluting peaks in the complex mixtures. The advantage of the automated spectral deconvolution is demonstrated for the identification of the coeluting peaks in the complex mixtures. Coelution of peaks is also observed with highspeed GC/TOFMS technique. The high-speed GC x GC/TOFMS was also tested with two different analyte system ??? A pesticide mixture and platformate (an aromatic mixture) to evaluate the suitability for high-speed analysis of complex mixtures. A poor resolution was observed for the pesticide mixture in the two-dimensional plane and it appeared, as there was nearly no orthogonal separation in the second dimension. The platformate mixture displayed a better two-dimensional separation. Chromatographic peak resolution is not really a primary requirement for locating and identifying the coeluting compounds in high-speed GC x GC/TOFMS technique. However, it was observed that the high-speed GC x GC/TOFMS too faced problem to unscramble the mass spectra of those compounds with similar structure and sharing the same unique masses.

The present thesis is focused on study of dust grain charging. The experimental part covers interaction between dust grains and high-energy electrons and self-discharging of grains by both field electron emission and field ionization. The second part of the thesis describes construction and evaluation of a linear electrodynamic trap of the novel design. We have observed charging of small dust grains towards high positive electric potentials when bombarded by the high-energy electron beam. We have described an increase of the secondary electron- electron emission yield from negatively charged grains due to the surface field. Further, self-discharging characteristics for both positively and negatively charged grains were measured. The relationship between discharging rate and the rate of the flow of atoms leaving the grain surface due to diffusion was observed for positively charged grains. This suggests significantly lower surface field necessary for ionizing such atoms compared to the ionization of atoms of surrounding gas and compared to the typically published field ionization thresholds. Based on the design published in the master thesis, a linear quadrupole trap of novel design was constructed. Testing measurements confirmed functionality and expected characteristics of the trap. In addition, effects of...

Abstract In this thesis nuclear magnetic resonance (NMR) and magneto-optical rotation (MOR) parameters are investigated for water, paying special attention to the effect of solvation from gaseous to liquid phase. Nuclear magnetic shielding and quadrupole coupling tensors of NMR spectroscopy are studied for gaseous and liquid water. Liquid state is modelled by a 32-molecule Car-Parrinello molecular dynamics simulation, followed by property calculations for the central molecules in clusters cut out from the simulation trajectory. Gaseous state is similarly represented by a one-molecule simulation. Gas-to-liquid shifts for shielding constants obtained this way are in good agreement with experiments. To get insight into the local environment and its effect on the properties the clusters are divided into groups of distinct local features, namely the number of hydrogen bonds. The analysis shows in detail how the NMR tensors evolve as the environment changes gradually from the gas to liquid upon increasing the number of hydrogen bonds to the molecule of interest. The study sheds light on the usefulness of NMR experiments in investigating the local coordination of liquid water. To go a bit further, the above mentioned NMR parameters along with the spin-spin coupling constant are examined for water dimer in various geometries to have insight into solvation and hydrogen bonding phenomena from bottom to top. Characteristic changes in the properties are monitored as the geometry of the dimer is systematically varied from very close encounter of the monomers to distances and orientations where hydrogen bonding between monomers ceases to exist. No rapid changes during the hydrogen bond breaking are observed indicating that the hydrogen bonding is a continuous phenomenon rather than an on-off situation. However, for analysis purposes we provide an NMR-based hydrogen bond definition, expressed geometrically, based on the behaviour of the NMR properties as a function of dimer geometry. Our definition closely resembles widely used definitions and thus reinforces their validity. Magneto-optical rotation parameters, the nuclear spin optical rotation (NSOR) and the Verdet constant, are computed for gaseous and liquid water, in the same manner as the NMR properties above. Recent pioneering experiments including NSOR for hydrogen nuclei in liquid water and liquid xenon have demonstrated that this technique has a potential to be a useful new probe of molecular structure. We reproduce computationally, applying a first-principles theory developed recently in the group, the experimental NSOR for hydrogen nuclei in liquid water, and predict hydrogen NSOR in gaseous water along with the oxygen NSOR in liquid and gaseous water. NSOR is an emerging experimental technique that needs interplay between theory and computation for validation, steering and insight.

La consommation énergétique en France a pour origine principale le secteur résidentiel et tertiaire. En environnement urbain, l’habitat est encore principalement ancien, avec des déperditions importantes. L’amélioration des performances énergétiques passe par la quantification des pertes, basée sur une méthode globale de mesure par caméra infrarouge à haute résolution spatiale.L’interprétation des images obtenues nécessite une description des termes radiatifs composant le signal, pour cela un simulateur original est mis au point. A partir d’une scène tridimensionnelle maillée, le champ de température est calculé pour les parois 1D et pour les ponts thermiques 2D, par le programme thermique développé à cet effet, SUSHI qui s’appuie sur un pré-calcul d’éclairement solaire et un pré-calcul de réponse indicielle 2D. Le signal du capteur infrarouge est alors modélisé en adaptant le code radiatif MOHICANS.Cette chaîne logicielle a l’originalité de proposer une fusion efficace des simulations de la réponse dynamique en température et en luminance de zones présentant un transfert 1Dà travers la paroi et de zones présentant un transfert 2D.La mise en œuvre de la campagne expérimentale BATIR a permis de mesurer le comportement thermique d’une façade de bâtiment et de son environnement radiato-convectif. Une validation ponctuelle des températures calculées par SUSHI a été réalisée par confrontation à une mesure par thermocouple. Des caméras infrarouges ont été mises en œuvre afin de collecter la luminance issue de la façade étudiée en bande II et III. Les luminances calculées par MOHICANS sont comparées à ces acquisitions, et valident la chaîne logicielle à ce niveau. / The main origin of the energy consumption in France is the residential and commercial sector. In urban environment, housing is mostly old, which means high heat losses. The improvement of energy performances requires the quantification of heat losses. This quantification may be based on a global measurement by an infrared camera with high spatial resolution.The infrared image interpretation requires a description of the radiative terms that make up the signal. For that purpose, a novel simulator is developed. The temperature field is calculated from a meshed three-dimensional scene composed of 1D walls and 2D thermal bridges. This operation is realized by the developed thermal software SUSHI, which is based on solar irradiance pre-computation and on 2D unitary response pre-computation. The software uses as input environmental data measured in the field. The infrared sensor signal is then modelled by adapting the radiative program MOHICANS. This software chain has the distinct advantage of an efficient fusion of dynamic response simulations of temperature and radiance, for areas with unidirectional and 2D heat transfer.The experimental campaign BATIR was set-up for measuring the thermal behavior of a building façade and its convective and radiative environment. A local validation of temperature calculation by SUSHI was realized through a comparison with thermocouple measurement results. Infrared cameras were operated in order to collect the radiance coming from the analyzed façade in band II and III. The radiances calculated by MOHICANS were compared with these acquisitions in order to validate the software chain at this level.

No description available.

Les lasers à électrons libres (LEL) X sont aujourd'hui des sources lumineuses cohérentes et intenses utilisées pour des investigations multidisciplinaires de la matière. Un nouveau schéma d'accélération, l'accélérateur laser plasma (LPA), est maintenant capable de produire une accélération de quelques GeV/cm, bien supérieure à celle des linacs radiofréquence. Ce travail de thèse a été mené dans le cadre des programmes de R&D du projet LUNEX5 (laser à électrons libres utilisant un nouvel accélérateur pour l’exploitation du rayonnement X de 5e génération) de démonstrateur LEL avancé et compact avec applications utilisatrices pilotes. Il comprend un linac supraconducteur de 400 MeV de haute cadence (10 kHz) pour l’étude de schémas LEL avancés, et LPA pour sa qualification par une application LEL. La ligne LEL utilise une configuration d’injection avancée dans la plage spectrale 40-4 nm par génération d’harmoniques à gain élevé (HGHG) et schéma d’écho (EEHG) avec des onduleurs compacts cryogéniques à champ élevé de courte période courte. L'étude de solutions adaptées aux applications LEL compactes et avancées est donc examinée. Un premier aspect concerne la réduction du milieu de gain du LEL (électrons dans l'onduleur), le raccourcissement de la période se faisant au détriment du champ magnétique. Les onduleurs cryogéniques compacts à base d'aimants permanents cryogéniques (CPMU), dans lesquels les performances de l'aimant sont améliorées à la température cryogénique sont étudiés. Une deuxième partie du travail développée dans le cadre l’expérience de R&D COXINEL visant à démontrer l’amplification LEL à l’aide d’un LPA. La ligne permet de manipuler les propriétés des faisceaux d’électrons produits (dispersion en énergie, divergence, variation de pointé) avant d’être utilisées pour des applications de sources lumineuses. Le faisceau d'électrons généré est très divergent et nécessite une bonne manipulation juste après la source avec des quadrupôles forts placés immédiatement après la génération d'électrons. Ainsi, des quadrupôles innovants à aimants permanents de gradient élevé réglable appelés «QUAPEVA», sont développés. Ils sont optimisés avec le code RADIA et caractérisées avec trois mesures magnétiques. Un gradient de 200 T/m avec une variabilité de 50 % est obtenu tout en maintenant une excursion du centre magnétique réduite à ± 10 µm, qui a permis un alignement par compensation de pointé du faisceau dans COXINEL grâce au centre magnétique variable des systèmes, avec un faisceau bien focalisé sans dispersion. Les QUAPEVA constituent des systèmes originaux dans le paysage des quadrupôles à de gradient élevé et variable développés jusqu'à présent. Une troisième partie des travaux concerne l’observation du rayonnement d’onduleur monochromatique ajustable sur la ligne COXINEL. Le faisceau d'électrons d'énergie de 170 MeV est transporté et focalisé dans un CPMU de 2 m et de période de 18 mm émettant à 200 nm. Le flux spectral est caractérisé à l'aide d'un spectromètre UV et le flux angulaire mesuré par une caméra CCD. La longueur d'onde est accordée avec l’entrefer. Les distributions spatio-spectrales mesurées en forme de lune du rayonnement de l'onduleur sont bien reproduites par les simulations de rayonnement utilisant les distributions d’électrons mesurées et transportées le long de la ligne. Elles permettent aussi de renseigner sur la qualité du faisceau d’électrons, de son transport et d'en estimer les paramètres tels que la dispersion en énergie et la divergence. Le dernier aspect du travail est lié à la comparaison entre la génération des harmoniques en gain élevé et le schéma d’écho, dans le cadre de ma participation à une expérience réalisée à FERMI @ ELETTRA. Nous avons pu démontrer un LEL de type écho à 5,9 nm, avec spectres plus étroits et une meilleure reproductibilité que le schéma HGHG à deux étages. Cette thèse constitue un pas en avant vers les lasers à électrons libres compacts et avancés. / X-ray Free Electron Lasers (FEL) are nowadays unique intense coherent fs light sources used for multi-disciplinary investigations of matter. A new acceleration scheme such as Laser Plasma Accelerator (LPA) is now capable of producing an accelerating gradient of few GeV/cm far superior to that of conventional RF linacs. This PhD work has been conducted in the framework of R&D programs of the LUNEX5 (free electron Laser Using a New accelerator for the Exploitation of X-ray radiation of 5th generation) project of advanced and compact Free Electron laser demonstrator with pilot user applications. It comprises a 400 MeV superconducting linac for studies of advanced FEL schemes, high repetition rate operation (10 kHz), multi-FEL lines, a Laser Wake Field Accelerator (LWFA) for its qualification by a FEL application. The FEL lines comports enables advanced seeding in the 40-4 nm spectral range using high gain harmonic generation (HGHG) and echo-enabled harmonic generation (EEHG) with compact short period high field cryogenic undulators. The study of compact devices suitable for compact FEL applications is thus examined. One first aspect concerns the reduction of the Free Electron Laser gain medium (electrons in undulator) where shortening of the period is on the expense of the magnetic field leading to an intensity reduction at high harmonics. Compact cryogenic permanent magnet based undulators (CPMUs), where the magnet performance is increased at cryogenic temperature making them suitable for compact applications, are studied. Three CPMUs of period 18 mm have been built: two are installed at SOLEIL storage ring and one at COXINEL experiment. A second part of the work is developed in the frame of the R&D programs is the COXINEL experiment with an aim at demonstrating FEL amplification using an LPA source. The line enables to manipulate the properties of the produced electron beams (as energy spread, divergence, induced dispersion due) before being used for light source applications. The electron beam generated is highly divergent and requires a good handling at an early stage with strong quadrupoles, to be installed immediately after the electron generation source. Hence, the development of the so-called QUAPEVAs, innovative permanent magnet quadrupoles with high tunable gradient, is presented. The QUAPEVAs are optimized with RADIA code and characterized with three magnetic measurements. High tunable gradient is achieved while maintaining a rather good magnetic center excursion that allowed for beam pointing alignment compensation at COXINEL, where the beam is well-focused with zero dispersion at any location along the line. The QUAPEVAs constitute original systems in the landscape of variable high gradient quadrupoles developed so far. A third part of the work concerns the observation of tunable monochromatic undulator radiation on the COXINEL line. The electron beam of energy of 170 MeV is transported and focused in a 2-m long CPMU with a period of 18 mm emitting radiation light at 200 nm. The spectral flux is characterized using a UV spectrometer and the angular flux is captured by a CCD camera. The wavelength is tuned with the undulator gap variation. The spatio-spectral moon shape type pattern of the undulator radiation provided an insight on the electron beam quality and its transport enabling the estimation of the electron beam parameters such as energy spread and divergence. The final aspect of the work is related to the comparison between the echo and high gain harmonic generation, in the frame of my participation to an experiment carried out at FERMI@ELETTRA. At FERMI, we have demonstrated a high gain lasing using EEHG at a wavelength of 5.9 nm where it showed a narrower spectra and better reproducibility compared to a two-stage HGHG. This PhD work constitutes a step forward towards advanced compact Free Electron Lasers.