Dans une société qui peine à renoncer à l'utilisation des énergies fossiles, la production d'hydrogène à partir d'eau par photocatalyse solaire est une alternative que les chimistes se doivent d'explorer. La mise en place de cette solution est conditionnée à la résolution de deux problèmes majeurs : augmenter l'efficacité de la conversion solaire par le développement de nouveaux semi-conducteurs, et améliorer la réactivité en surface par le développement de co-catalyseurs efficaces. Dans un premier temps, nous avons contribué à l'étude d'une nouvelle famille de semi-conducteurs par des calculs ab initio DFT. Différentes propriétés clés ont été calculées et comparées aux valeurs expérimentales. Nous avons montré que pour une même famille de matériaux, l'absorption peut être contrôlée par modification de la composition.Ensuite, nous nous sommes focalisés sur la réaction d'oxydation de l'eau (OER). Pour cette réaction les catalyseurs de type oxyhydroxydes et leurs dérivés sont très prometteurs car à la fois efficaces et contenant des éléments abondants sur Terre, mais la compréhension de son mécanisme reste limitée. Sur le composé modèle CoOOH, nous en avons réalisé une analyse approfondie par une étude exhaustive des intermédiaires ainsi que par modélisation explicite du potentiel électrochimique. Nos études ont montré la nécessite de prendre en compte plusieurs sites réactionnels dans la détermination des mécanismes d'oxydation de l'eau sur la surface de ces catalyseurs. Enfin, une étude préliminaire de l'utilisation de la biomasse a été menée, dans le but de combiner valorisation de la biomasse et production d'hydrogène. Le cas de l'oxydation du glycérol est envisagé. / In a society struggling to waive the use of fossil fuels, hydrogen production from water by solar photocatalysis is a alternative chemists have to consider. Setting up of this solutions asks to tackle two major issues : increase solar energy conversion by developing new semiconductors, and enhance the surface reactivity by developing efficient cocatalysts.First, DFT \emph{ab initio} calculations were carried out on a new family of semiconductor materials. Different key properties were computed and compared to experimental values. We showed that for a same material family, absorption can be controlled by changing the composition.Then, we focused on Oxygen Evolution Reaction (OER). For this reaction oxyhydroxides catalysts and their derivatives seem promising as both efficient and containing earth-abundant elements, but the understanding of its mechanism still remains unclear. On CoOOH model compound, we realised an extended analysis by a comprehensive study of intermediates and by explicit modelling of electrochemical potential. Our studies showed that taking into account several reactive sites is necessary to determine the OER mechanism on these catalysts' surface.Last, a preliminary study of biomass use was carried, in order to combine its valorization along with hydrogen production. The case of glycerol photoreforming is considered.

At the heart of high harmonic generation lies a combination of optical and collision physics entwined by a strong laser field. An electron, initially tunnel-ionized by the field, driven away then back in the continuum, finally recombines back to rest in its initial ground state via a radiative transition. The emitted attosecond (atto=10^-18) XUV light pulse carries all the information (polarization, amplitude and phase) about the photorecombination continuum-to-ground transition dipolar field. Photorecombination is related to the time-reversed photoionization process. In this perspective, high-harmonic spectroscopy extends well-established photoelectron spectroscopy, based on charged particle detection, to a fully coherent one, based on light characterization. The main achievement presented in this thesis is to use high harmonic generation to probe femtosecond (femto=10^-15) chemical dynamics for the first time. Thanks to the coherence imposed by the strong driving laser field, homodyne detection of attosecond pulses from excited molecules undergoing dynamics is achieved, the signal from unexcited molecules acting as the reference local oscillator. First, applying time-resolved high-harmonic spectroscopy to the photodissociation of a diatomic molecule, Br2 to Br + Br, allows us to follow the break of a chemical bond occurring in a few hundreds of femtoseconds. Second, extending it to a triatomic (NO2) lets us observe both the previously unseen (but predicted) early femtosecond conical intersection dynamics followed by the late picosecond statistical photodissociation taking place in the reaction NO2 to NO + O. Another important realization of this thesis is the development of a complementary technique to time-resolved high-harmonic spectroscopy called LAPIN, for Linked Attosecond Phase INterferometry. When combined together, time-resolved high-harmonic spectroscopy and LAPIN give access to the complex photorecombination dipole of aligned excited molecules. These achievements lay the basis for electron recollision tomographic imaging of a chemical reaction with unprecedented angstrom (1 angstrom= 0.1 nanometer) spatial resolution. Other contributions dedicated to the development of attosecond science and the generalization of high-harmonic spectroscopy as a novel, fully coherent molecular spectroscopy will also be presented in this thesis.

L'anomalie isotopique de l'ozone observée au début des années 1980 a été la première manifestation de fractionnement isotopique indépendant de la masse d'origine chimique. Attribuée au départ, essentiellement au processus de formation de l'ozone, les travaux récents mettent en évidence d'autres contributions telles que la photodissociation. Cette thèse utilise une approche théorique basée sur des calculs ab initio de chimie et dynamique quantique pour déterminer les sections efficaces d'absorption et leurs variations isotopiques. Ces sections efficaces permettent d'étudier la photodissociation de l'ozone par irradiation du soleil (flux actinique) ; ce qui permet une évaluation précise du processus de photodissociation à l'enrichissement isotopique de l'ozone. Ces résultats pourront être intégrés dans un modèle global prenant en compte à la fois les processus de formation et de destruction de l'ozone.

At the heart of high harmonic generation lies a combination of optical and collision physics entwined by a strong laser field. An electron, initially tunnel-ionized by the field, driven away then back in the continuum, finally recombines back to rest in its initial ground state via a radiative transition. The emitted attosecond (atto=10^-18) XUV light pulse carries all the information (polarization, amplitude and phase) about the photorecombination continuum-to-ground transition dipolar field. Photorecombination is related to the time-reversed photoionization process. In this perspective, high-harmonic spectroscopy extends well-established photoelectron spectroscopy, based on charged particle detection, to a fully coherent one, based on light characterization. The main achievement presented in this thesis is to use high harmonic generation to probe femtosecond (femto=10^-15) chemical dynamics for the first time. Thanks to the coherence imposed by the strong driving laser field, homodyne detection of attosecond pulses from excited molecules undergoing dynamics is achieved, the signal from unexcited molecules acting as the reference local oscillator. First, applying time-resolved high-harmonic spectroscopy to the photodissociation of a diatomic molecule, Br2 to Br + Br, allows us to follow the break of a chemical bond occurring in a few hundreds of femtoseconds. Second, extending it to a triatomic (NO2) lets us observe both the previously unseen (but predicted) early femtosecond conical intersection dynamics followed by the late picosecond statistical photodissociation taking place in the reaction NO2 to NO + O. Another important realization of this thesis is the development of a complementary technique to time-resolved high-harmonic spectroscopy called LAPIN, for Linked Attosecond Phase INterferometry. When combined together, time-resolved high-harmonic spectroscopy and LAPIN give access to the complex photorecombination dipole of aligned excited molecules. These achievements lay the basis for electron recollision tomographic imaging of a chemical reaction with unprecedented angstrom (1 angstrom= 0.1 nanometer) spatial resolution. Other contributions dedicated to the development of attosecond science and the generalization of high-harmonic spectroscopy as a novel, fully coherent molecular spectroscopy will also be presented in this thesis.

During my master thesis I became familiar with ion imaging and velocity map imaging (VMI) techniques in experiments with molecular beam. My major contribution was writing a computer program for data acquisition and pre-analysis. The program features initial filtering of the data, improving significantly signal to noise ratio. Non-trivial effort was required to make the program compatible with three different CCD cameras implemented on different experiments in the laboratory. I have also simulated ion trajectories in VMI and wrote a program code to imulate VMI image distortions due to imperfections in apparatus geometry. These programs will be useful in aligning and tuning the apparatus. Finally, I wrote a program for data manipulation and conversion to formats compatible with various data analysis programs available in the laboratory. All the above programs I have tested by measurements of 243 nm photodissociation of HBr molecules.

Electronic excitation triggered by the absorption of light enables numerous chemical, physical and biological processes and transformations. Accordingly, full control over the processes involving excited molecules requires an in-depth knowledge of electronic UV/vis spectra and potential energy surfaces. Unsurprisingly, most electronic spectra are acquired in the condensed phase in which molecules are dissolved and most transformations occur. However, our knowledge of excitation, transformations and processes at the level of isolated molecules is still limited, partly because such studies require unconventional experimental approaches and equipment. This Thesis describes experimental methods for recording electronic spectra of isolated molecules in the gas phase by ion spectroscopy, which combines mass spectrometry with optical spectroscopy. Using these methods, experimental factors which affect the electronic excitation and therefore the electronic spectra of ions were determined and evaluated for various fluorescent xanthene dyes, iron-containing complexes and molecular pho- toswitches. Furthermore, factors which govern photochemical processes, such as photo- oxidation, photoreduction and photoisomerization, were also analyzed in detail, with surprisingly different outcomes from previous studies...

L'objectif principal de cette thèse est de rendre compte d'une étude expérimentale sur la dissociation neutre de molécules simples dans un champ laser intense créé par un cristal titane saphir. Ces excitations fortement non linéaires nous indiquent que les molécules peuvent être peuplées dans les états hyperexcités. Dans ce travail, un laser titane saphir femtoseconde est utilisé pour amener les molécules de H₂, 0₂, NO, CH₄, C₂H₄, C₃, H₆, 1 — C₄H₈ et cis — 2 — C₄H₈ dans les états très excités. En utilisant une méthode de spectroscopic, on arrive à détecter des signaux de fluorescence. Le diagramme d'énergie des fragments excités et molécules neutres supporte l'excitation des états hyperexcités dans les molécules que nous avons étudiées. La dépendance hautement non linéaire du rendement produit selon la puissance du laser a été observée. En outre, les résultats soutiennent le mécanisme d'hyperexcitation multiphotonique. En utilisant la technique de pompe (800 nm) et sonde (1338 nm) et aussi de la technique de spectroscopie par fluorescence, nous confirmons le mécanisme de hyperexcitation multiphotonique des molécules en présence d'un champ laser intense. Nous arrivons à déterminer la durée de vie des états hyperexcités en regardant l'atténuation des signaux de fluorescence observée à l'aide du faisceau sonde. Nos observations expérimentales ont été également vérifiées à l'aide des calculs semi—empiriques. Nous constatons que certaines impulsions laser provoquent la dissociation neutre de nombreuses molécules.

L'objectif de ce travail de thèse est d'étudier l'influence d'un environnement de gaz rare sur les propriétés statiques et dynamiques de petites molécules de sodium. Les différentes propriétés physicochimiques de ce système permettent une modélisation à deux niveaux dans laquelle seuls les degrés de liberté associés aux électrons de valence de la molécule sont traités quantiquement. Nous avons développé une approche générale permettant de traiter le problème de la structure électronique de la molécule immergée par une méthode d'interaction de configurations, dans laquelle ses noyaux et les atomes de gaz rare sont traités en dynamique moléculaire classique d'atomes polarisables.<br />En adaptant ce modèle théorique, à l'atome et aux dimères de sodium immergés dans des matrices d'argon, nous avons déterminé la géométrie d'équilibre et les propriétés spectrales de ces systèmes. Le site de piégeage le plus favorable du dimère est différent selon qu'il soit chargé ou pas. Nous retrouvons ce résultat de manière dynamique si l'on procède à l'ionisation du Na2 immergé. <br />En étudiant la dynamique sur le premier état excité de Na2+, nous avons observé l'importance de la taille de la matrice sur la dissociation de ce dimère. Nous avons ainsi déterminé un nombre critique d'argon au-delà duquel la dissociation est empêchée dû à un changement de site du Na2+ au sein de sa rangée d'insertion. Mais en introduisant les couplages non adiabatiques au moyen d'un algorithme de saut de surfaces, ce changement de site, observé pour les systèmes comportant plus d'une centaine d'argon, est avorté grâce à une désexcitation non radiative vers l'état fondamental au bout de quelques picosecondes.

This thesis describes the development of a relatively novel technique for the gen- eration and subsequent trapping of cold species. Molecules in a pulsed supersonic expansion are photolysed, such that the centre-of-mass velocity vector of one of the fragments is equal in magnitude but opposed in orientation to the lab-frame velocity of the precursor molecule. This technique, known as ‘Photostop’, leaves a fraction of the fragments with a narrow velocity distribution, centered around zero velocity in the lab-frame. They can be shown to have zero velocity by changing the time between photodissociation and ionisation; fragments with a high kinetic energy will leave the ionisation volume prior to interrogation. The underlying velocity distribu- tion is uncovered by using the velocity-map imaging technique, and the temperature of the fragments can be determined. The method was originally optimised for the molecular case. Cold NO has been produced from the dissociation of NO₂ molecules, and a single rotational state has been shown to remain in the ionisation volume 10 μs after dissociation, implying a sample temperature of 1.17 K. Using the optimised experimental conditions de- rived from the velocity cancellation of NO, the atomic case is demonstrated for the dissociation of Br₂ to give zero-velocity Br fragments. The Br atoms are seen for delay times in excess of 100 μs, showing the greater applicability of the method to the atomic case. The temperature of the residual atoms is shown to be in the milliKelvin regime, as determined through detailed Monte Carlo simulation of the motion of the stopped atoms. The possibility of trapping the ultracold Br atoms in a magnetic field is explored, and a quadrupolar trap created between two per- manent bar magnets is demonstrated to confine the atoms spatially, within the ion extraction optics, for delays in excess of 1 ms. The Photostop technique is intended to be a stepping stone on the way to widening the number of chemical species available for study in the ultracold regime. The possibility of improvements to the experiment is considered, in order to increase the efficiency of the experiment such that the number density becomes high enough to be viable as a source of atoms for use in cold chemical reactive studies. The possibility of extending the method so as to be used as a tunable velocity source of atoms is also discussed.

This thesis presents the study of a selection of gas phase chemical processes using time-resolved Fourier transform infrared (FTIR) emission spectroscopy. Such processes include molecular energy transfer, chemical reaction and photodissociation. The major focus of this thesis was the investigation of collisional energy transfer from the electronically excited states of NO and OH, with particular attention paid to the fate of the electronic energy. NO A²&Sigma;⁺(v = 0) is prepared by laser excitation, pumping the overlapped Q₁ and P₂₁ band heads of the NO A-X (0,0) transition at 226.257 nm. The quenching of this state by O₂ and CO₂ was studied. Experiments were performed to investigate what channels contribute to the quenching process, the branching ratio of these different channels and the partitioning of energy among the various products. Quenching by O₂ was found to proceed mostly through non-reactive channels. High vibrational excitation of NO X ²&Pi; was observed, with population detected in v = 22, representing 79% of the available energy. The O₂ product was found to be formed in more than one electronic state: the ground state, X ³&Sigma;^-<sub style='position: relative; left: -.3em;'>g</sub>, and a high-lying electronically excited state, such as the A ³&Sigma;⁺<sub style='position: relative; left: -.5em;'>u</sub>, A' ³&Delta;_u or c ¹&Sigma;^-<sub style='position: relative; left: -.5em;'>u</sub> states. A reactive channel producing vibrationally excited NO₂ was observed, but was found to be a minor process with an upper limit of 18% for the branching ratio. In contrast the quenching of NO A ²&Sigma;⁺(v = 0) by CO₂ was found to proceed predominately by reaction, with a branching ratio of 76 %. While emission from NO₂ was observed, it was weak, and therefore it was concluded that the main reaction products were CO, O(³P) and NO X ²&Pi;(v = 0). The nascent strong CO₂ v3 emission band from the non-reactive channel exhibited a large red-shift from its fundamental position. This indicates that the CO₂ vibrational distribution is significantly hotter than statistical. Investigations were then performed studying the quenching of NO A ²&Sigma;⁺(v = 1) by NO and CO₂, with both systems exhibiting similar characteristics to the quenching of the ground vibrational level of NO A ²&Sigma;⁺. From comparison of the emission intensity of the CO fundamental and CO₂ v3 mode following quenching of the v = 0 and 1 levels of the NO A ²&Sigma;⁺ state, it was concluded that the branching ratio for reactive quenching was larger in the latter case. Secondly, experiments were performed to measure the rate constants for the quenching of NO A ²&Sigma;⁺(v = 0) by the noble gases. The noble gases are inefficient quenchers of electronically excited NO and therefore careful experimental design was required to minimise the influence of impurities on the results. All the rate constants were found to be of the order of 10^-14 cm³ molecule^-1 s^-1. The value for Xe was 50 times smaller than reported previously in the literature. In light of this new measurement, a re-analysis of experiments, performed previously in the group, on the electronic quenching of NO A ²&Sigma;⁺(v = 0) by Xe was performed. A very hot vibrational distribution of NO X ²&Pi; was obtained. Next, the collisional quenching of OH A ²&Sigma;⁺(v = 0) by H₂ was investigated. OH radicals were generated in situ by the photolysis of HNO₃ at 193 nm, which were excited to the A ²&Sigma;⁺(v = 0) state on the overlapped Q₁(1) and P₂₁(1) rotational lines at 307.935 nm. Reactive quenching was found to be the major pathway, in agreement with the literature. Copious emission from vibrationally excited water was observed. Comparison of this emission with theoretical calculations revealed a hotter distribution than predicted. It was concluded that the energy channelled into the vibrational modes of H₂O is in excess of 60% of the available energy. Experiments performed with D₂ allowed the non-reactive channel to be studied; a cold vibrational distribution of the OH X ²&Pi; was observed. Finally the reaction between CN radicals and cyclohexane was studied. CN was generated by the photolysis of ICN at 266 nm. Prompt emission from HCN in the C-H stretching region was observed meaning the new bond was formed in a vibrationally excited state. Analysis of the emission revealed HCN was populated up to v3 = 2. Excellent agreement with the results of a theoretical study of the system was found.