This work presents a new implementation of the QP algorithm, a computer method to diagonalize the extremely large matrices arising in multimode vibronic problems. Benchmark calculations are included, showing the accuracy of the program. The QP algorithm is extended to treat multiple electronic surfaces for competitive control and this is demonstrated with an Hamiltonian including three electronic states, a model of the benzene radical cation. Finally, the evolution of zeroth-order states in a simple two electronic states, two dimensional model with a conical intersection is explored, towards building a time-dependent view of overlapping resonances coherent control.

No description available.

On considère les paquets d'ondes hydrodynamiques comme mécanismes de génération de bruit des jets subsoniques. Cette approche résulte tout d'abord de l'analyse de données numériques - DNS d'une couche de mélange (Wei et Freund 2006) et LES d'un jet à Mach 0,9 (Daviller 2010) - permettant de déterminer les propriétés des sources en termes de compacité, d'intermittence et de structure azimutale. L'identification d'un rayonnement intermittent associé aux modifications des structures cohérentes des écoulements permet de proposer un modèle de paquet d'onde pour représenter ce phénomène dans l'analogie de Lighthill, dont l'enveloppe présente des variations temporelles d'amplitude et d'étendue spatiale. Celles-ci sont tirées de données de vitesse de simulations numériques de jets subsoniques, et un accord de l'ordre de 1,5dB entre le champ acoustique simulé et le modèle confirme sa pertinence. L'exploration du concept proposé est ensuite poursuivie expérimentalement, avec des mesures de pression acoustique et de vitesse de jets turbulents subsoniques, permettant la décomposition des champs en modes de Fourier azimutaux. On observe l'accord des directivités des modes 0, 1 et 2 du champ acoustique avec le rayonnementd'un paquet d'onde. Les modes 0 et 1 du champ de vitesse correspondent également à des paquets d'onde, modélisés comme des ondes d'instabilité linéaires à partir des équations de stabilité parabolisées. Finalement, des corrélations de l'ordre de 10% entre les modes axisymétriques de vitesse dans le jet et de pression acoustique rayonnée montrent un lien clair entre les paquets d'onde et l'émission acoustique du jet. / Hydrodynamic wavepackets are studied as a sound-source mechanism in subsonic jets. We first analyse numerical simulations to discern properties of acoustic sources such as compactness, intermittency and azimuthal structure. The simulations include a DNS of a two-dimensional mixing layer (Wei and Freund 2006) and an LES of a Mach 0.9 jet (Daviller 2010). In both cases we identify intermittent radiation, which is associated with changes in coherent structures in the flows. A wave-packet model that includes temporal changes in amplitude and axial extension is proposed to represent the identified phenomena using Lighthill's analogy. These parameters are obtained from velocity data of two subsonic jet simulations, and an agreement to within 1.5dB between the model and the acoustic field of the simulations confirms its pertinence. The proposed mechanism is then investigatedexperimentally, with measurements of acoustic pressure and velocity of turbulent subsonic jets, allowing the decomposition of the fields into azimuthal Fourier modes. We find close agreement of the directivities of modes 0, 1 and 2 of the acoustic field with wave-packet radiation. Modes 0 and 1 of the velocity field correspond also to wavepackets, modelled as linear instability waves using parabolised stability equations. Finally, correlations of order of 10% between axisymmetric modes of velocity and far-field pressure show the relationship between wavepackets and sound radiated by the jet.

Few experimental techniques have found such a diverse range of applications as has ion imaging. The field of chemical dynamics is constantly advancing, and new applications of ion imaging are being realised with increasing frequency. This thesis is concerned with the application of a fast pixelated imaging sensor, the Pixel Imaging Mass Spectrometry (PImMS) camera, to ion imaging applications. The experimental possibilities of such a marriage are exceptionally broad in scope, and this thesis is concerned with the development of a selection of velocity-map imaging applications within the field of photoinduced molecular dynamics. The capabilities of the PImMS camera in three-dimensional and slice imaging applications are investigated, in which the product fragment Newton-sphere is temporally stretched along the time-of-flight axis, and time-resolved slices through the product fragment distribution are acquired. Through experimental results following the photodissociation of ethyl iodide (CH₃CH₂I) at around 230 nm, the PImMS camera is demonstrated to be capable of recording well-resolved time slices through the product fragment Newton-sphere in a single experiment, without the requirement to time-gate the acquisition. The various multi-hit capabilities of the device represent a unique and significant advantage over alternative technologies. The details of a new experiment that allows the simultaneous imaging of both photoelectrons and photoions on a single detector for each experimental acquisition cycle using pulsed ion extraction are presented. It is demonstrated that it is possible to maintain a high velocity resolution using this approach through the simultaneous imaging of the photoelectrons and photoions that result from the (3 + 2) resonantly enhanced multi-photon ionisation of Br atoms produced following the photodissociation of Br₂ at 446.41 nm. Pulsed ion extraction represents a substantial simplification in experimental design over conventional photoelectron-photoion coincidence (PEPICO) imaging spectrometers and is an important step towards performing coincidence experiments using a conventional ion imaging apparatus coupled with a fast imaging detector. The performance of the PImMS camera in this application is investigated, and a new method for the determination of the photofragment detection efficiencies based on a statistical fitting of the coincident photoelectron and photoion data is presented. The PImMS camera is applied to laser-induced Coulomb explosion imaging (CEI) of an axially chiral substituted biphenyl molecule. The multi-hit capabilities of the device allow the concurrent detection of individual 2D momentum images of all ionic fragments resulting from the Coulomb explosion of multiple molecules in each acquisition cycle. Correlations between the recoil directions of the fragment ions are determined through a covariance analysis. In combination with the ability to align the molecules in space prior to the Coulomb explosion event, the experimental results demonstrate that it is possible to extract extensive information pertaining to the parent molecular structure and fragmentation dynamics following strong field ionisation. Preliminary simulations of the Coulomb explosion dynamics suggest that such an approach may hold promise for determining elements of molecular structure on a femtosecond timescale, bringing the concept of the `molecular movie' closer to realisation. Finally, the PImMS camera is applied to the imaging of laser-induced torsional motion of axially chiral biphenyl molecules through femtosecond Coulomb explosion imaging. The target molecules are initially aligned in space using a nanosecond laser pulse, and torsional motion induced using a femtosecond 'kick' pulse. Instantaneous measurements of the dihedral angle of the molecules are inferred from the correlated F+ and Br+ ion trajectories following photoinitiated Coulomb explosion at various time delays after the initial kick pulse. The technique is extended to include a second kick pulse, in order to achieve either an increase in the amplitude of the oscillations or to damp the motion, representing a substantial degree of control of the system. Measurements out to long kick-probe delays (200 ps) reveal that the initially prepared torsional wave packet periodically dephases and rephases, in accordance with the predictions of recent theoretical work.

La Méthode de la Trajectoire Adiabatique Contrainte est développée dans le but de résoudre globalementl’équation de Schrödinger. Cette méthode utilise le formalisme de Floquet et une décomposition de Fourier pourdécrire les dépendances temporelles. Elle transforme ainsi un problème dynamique en un problème aux valeurspropres partiel dans un espace de Hilbert étendu au temps. Cette manipulation requiert l’application decontraintes sur les conditions initiales de l’état propre de Floquet recherché. Les contraintes sont appliquées parl’intermédiaire d’un opérateur absorbant artificiel. Cet algorithme est adapté à la description de systèmes dirigéspar des hamiltoniens dépendant explicitement du temps. Il ne souffre pas de l’accumulation d’erreurs au cours dutemps puisqu’il fournit une solution globale ; les erreurs éventuelles proviennent de la non-complétude des basesfinies utilisées pour la description moléculaire ou temporelle et de l’imperfection du potentiel absorbant dépendantdu temps nécessaire pour fixer les conditions initiales. Une forme générale de potentiel absorbant a étédéveloppée pour être en mesure d’intégrer un problème avec une condition initiale quelconque. Des argumentsrelatifs au suivi adiabatique dans le cas de Hamiltoniens non-hermitiens sont également présentés. Nous insistonssur le rôle des facteurs de phase géométrique. Les méthodes développées sont appliquées à des systèmesatomiques ou moléculaires soumis à des impulsions laser intenses, en relation avec la problématique du contrôlemoléculaire. Nous considérons plusieurs exemples : modèles d’atomes à deux ou trois niveaux, ion moléculairehydrogène et molécules froides de sodium. / The Constrained Adiabatic Trajectory Method (CATM) allows us to compute global solutions of the time-dependent Schrödinger equation using the Floquet formalism and Fourier decomposition. The dynamical problem is thustransformed into a “static” problem, in the sense that the time will be included in an extended Hilbert space. Thisapproach requires that suitable constraints are applied to the initial conditions for the relevant Floquet eigenstate.The CATM is well suited to the description of systems driven by Hamiltonians with explicit and complicated timevariations. This method does not have cumulative errors and the only error sources are the non-completeness ofthe finite molecular and temporal basis sets used, and the imperfection of the time-dependent absorbing potentialwhich is essential to impose the correct initial conditions. A general form is derived for the absorbing potential,which can reproduce any dispersed boundary conditions. Arguments on adiabatic tracking in the case of nonhermitianHamiltonians are also presented. We insist on the role of geometric phase factors. The methods areapplied to atomic and molecular systems illuminated by intense laser pulses, in connection with molecular controlproblems. We study several examples : two or three-level atomic models, hydrogen molecular ion, cold sodiummolecules.