La décharge contrôlée par barrière diélectrique est un type de décharge hors-équilibre, fonctionnant à la pression atmosphérique. Normalement, elle est générée en état filamentaire qui se caractérise par une multitude de micro-décharges. Par contre, dans certaines conditions, nous pouvons obtenir une décharge homogène. Par exemple, dans notre étude, une décharge homogène est obtenue en atmosphère principale d’azote à la pression atmosphérique et comme ses caractéristiques électriques sont similaires à celles d’une décharge sombre de Townsend à basse pression, elle est appelée décharge de Townsend à la pression atmosphérique (DTPA). Pour maintenir un claquage de Townsend, un effet mémoire entre deux décharges est nécessaire. Cet effet mémoire conduit à la création d’électrons germes sous faible champ qui, lors de l’inversion de polarité permettent l’obtention d’une décharge homogène. Un marqueur de cet effet mémoire observable par les caractéristiques électriques de la décharge est le saut de courant quand la tension du gaz passe par zéro: plus le saut de courant est grand, plus l’effet mémoire est important. Des études précédentes ont montré l’importance des métastables de l’azote N2(A), qui produisent des électrons par émission secondaire entre deux décharges lors du bombardement des surfaces diélectriques. Néanmoins, nous observons que l’ajout d’une faible quantité de gaz oxydant (ici de l’oxygène) permet d’obtenir une décharge homogène plus stable, malgré la destruction considérable de N2(A) par quenching par des espèces oxydantes. En conséquence, nous proposons un autre processus pour la production des électrons germes en volume, basé sur la réaction d’ionisation associative suivante: N(2P)+O(3P) -->NO++e- où N(2P) est créé par la réaction: N(4S)+N2(A)-->N(2P)+N2(X). Pour vérifier cette hypothèse, nous utilisons la technique de la fluorescence induite par laser (LIF/TALIF), afin de déterminer les densités absolues de N(4S), O(3P) et NO. Les mesures sont faites pour différentes conditions expérimentales pour étudier l’influence du flux de gaz, de la puissance de la décharge et tout particulièrement de la concentration d’oxygène. Avec une augmentation de la concentration en oxygène jusqu’à 200ppm, la densité de N(4S) diminue à cause de sa destruction par les espèces oxydantes. Les densités de O(3P) et NO(X) augmentent puis saturent. Ceci peut être expliqué par le fait que la production de O(3P) et NO(X) est liée à la densité de N2(A). Ainsi pour de faibles concentrations en oxygène, l’ajout d’oxygène favorise la production de O(3P) et NO(X), mais pour des concentrations plus fortes, la destruction de N2(A) par quenching par les espèces oxydantes devient plus importante, limitant ainsi la production de O(3P) et NO(X). Avec les densités de N(4S), O(3P) et NO(X) mesurées expérimentalement, et la densité de N2(A) déterminée par Dilecce et al, la densité de N(2P) entre deux décharges peut être estimée par un modèle simple. Il est alors possible d’estimer la production d’électrons germes par les réactions d’ionisation associative et finalement le saut de courant qui en résulte. Un bon accord est observé entre les évolutions du saut de courant mesuré et calculé, même si des écarts quantitatifs subsistent. En conclusion, l’ionisation associative peut être considérée comme une bonne candidate pour expliquer l’augmentation de la création d’électrons germe entre deux décharges lorsqu’une faible quantité d’oxygène est introduite dans l’azote. / Dielectric barrier discharge is a type of non-equilibrium discharge, operating at atmospheric pressure. Normally, it is generated in filamentary mode which is characterized by a multitude of micro-discharges. Nevertheless, under certain conditions, it is possible to obtain a homogeneous discharge. In our study, the discharge is ignited in a nitrogen based atmosphere at atmospheric pressure and since its electrical characteristics are similar to that of a Townsend discharge at low pressure, it is called atmospheric pressure Townsend discharge (APTD). To maintain a Townsend discharge, a memory effect between two successive discharges is necessary. This memory effect is characterized by the creation of seed electrons under low electric field. A marker of this memory effect can be observed on the electrical characteristics: a current jump is observed when the gas voltage polarity reverses. The larger the current jump, the more important the memory effect. Previous investigations showed the importance of the N2(A) metastable molecules, which produce electrons by secondary emission on the dielectrics. Nevertheless, we observe that the addition of a small amount of oxidizing gas (in this case oxygen) results in a more stable homogeneous discharge, despite the considerable destruction of N2(A) by quenching by the oxidizing species. Therefore, we propose another process for the production of seed electrons, based on the following associative ionization reaction: N(2P)+O(3P)-->NO++e- where N(2P) is created by: N(4S)+N2(A)-->N(2P)+N2(X). To verify this hypothesis, laser-induced fluorescence (LIF/TALIF) measurements were done to determine the absolute densities of N(4S), O(3P) and NO(X) between two discharges. The measurements were performed under different experimental conditions to study the influence of the gas flow, the discharge power and more specifically the concentration of oxygen. For increasing oxygen concentration up to 200ppm, the density of N(4S) decreases because of its higher destruction by the oxidizing species. The densities of O(3P) and NO(X) increase and then become nearly constant. It can be explained by the fact that the production mechanisms of O(3P) and NO(X) involve N2(A) molecules. Then, whereas the addition of a small amount of oxygen favors the production of O(3P) and NO(X), a higher oxygen concentration induces a larger destruction of N2(A) by quenching due to the oxidizing species, which finally limits the production of O(3P) and NO(X). Knowing the densities of N(4S), O(3P) and NO(X) from experimental measurements, and the density of N2(A) from the work of Dilecce et al, the density of N(2P) can be estimated using a simple model, as well as the production of electrons due to associative ionizations. Finally the current jump can be calculated. The evolutions of the measured and calculated current jump have the same tendency even if the calculated values are much higher. In conclusion, associative ionization can be considered as a serious candidate to explain the increase of the memory effect and discharge stability when a small amount of oxygen is added to the nitrogen atmosphere of an APTD.

Neste trabalho estudamos o processo de foto-ionização associativa(PAI) em uma amostra fria de átomos de sódio com o objetivo entender os efeitos dos estados repulsivos e dimensão da colisão. Realizamos experimentos de PAI com duas cores em uma armadilha magneto-óptica adicionando um feixe de prova com intensidade, frequências e polarização ajustáveis. O formato dos átomos aprisionados também foi uma das variáveis no estudo da PAI. Para os átomos em formação esférica, observamos uma mudança marcante no comportamento da constante de taxa de formação da foto-ionização associativa(K) para um determinado domínio de frequências, e essa mudança no comportamento pode ser atribuída a participação de estados moleculares repulsivos na PAI e a formação de um possível cruzamento evitado entre os níveis moleculares. No atomotron ,armadilha atômica em forma de anel, variamos a polarização do laser de prova e constatamos que a razão entre K das polarizações paralela e perpendicular ao movimento dos átomos é igual a 4. Uma comparação entre K do atomotron e o da armadilha esférica em função da intensidade do feixe de prova, nos mostrou uma diferença no comportamento e no valor da constante de taxa. / Photoassociative ionization (PAI) in a cold sample of sodium atoms was the main subject of our studies as a way to understand the effects of repulsive states and collision dimensions. Two-color PAI experiment were preformed in a magneto-optical trap (MOT) trough the addition of a probe laser beam, the intensity, polarization and frequency of that probe laser were tunable. The shape of the trapped atoms also could be changed. In a spherical shape MOT, we observed a marked change in the PAI rate constant (K) for a definite frequency range, and that change can be attributed to the influence of repulsive molecular states and the a possible formation of an avoided crossing between molecular levels. In atomotron, ring shaped mot, we changed the polarization of the probe beam, and saw that the ratio between K for a polarization parallel to the atoms motion and a perpendicular one is 4. Comparing the K as a function of the intensity between a spherical shaped mot and atomotron showed us a difference in the behavior and the value of the rate constant.

Neste trabalho estudamos o processo de foto-ionização associativa(PAI) em uma amostra fria de átomos de sódio com o objetivo entender os efeitos dos estados repulsivos e dimensão da colisão. Realizamos experimentos de PAI com duas cores em uma armadilha magneto-óptica adicionando um feixe de prova com intensidade, frequências e polarização ajustáveis. O formato dos átomos aprisionados também foi uma das variáveis no estudo da PAI. Para os átomos em formação esférica, observamos uma mudança marcante no comportamento da constante de taxa de formação da foto-ionização associativa(K) para um determinado domínio de frequências, e essa mudança no comportamento pode ser atribuída a participação de estados moleculares repulsivos na PAI e a formação de um possível cruzamento evitado entre os níveis moleculares. No atomotron ,armadilha atômica em forma de anel, variamos a polarização do laser de prova e constatamos que a razão entre K das polarizações paralela e perpendicular ao movimento dos átomos é igual a 4. Uma comparação entre K do atomotron e o da armadilha esférica em função da intensidade do feixe de prova, nos mostrou uma diferença no comportamento e no valor da constante de taxa. / Photoassociative ionization (PAI) in a cold sample of sodium atoms was the main subject of our studies as a way to understand the effects of repulsive states and collision dimensions. Two-color PAI experiment were preformed in a magneto-optical trap (MOT) trough the addition of a probe laser beam, the intensity, polarization and frequency of that probe laser were tunable. The shape of the trapped atoms also could be changed. In a spherical shape MOT, we observed a marked change in the PAI rate constant (K) for a definite frequency range, and that change can be attributed to the influence of repulsive molecular states and the a possible formation of an avoided crossing between molecular levels. In atomotron, ring shaped mot, we changed the polarization of the probe beam, and saw that the ratio between K for a polarization parallel to the atoms motion and a perpendicular one is 4. Comparing the K as a function of the intensity between a spherical shaped mot and atomotron showed us a difference in the behavior and the value of the rate constant.