Les pièges radio-fréquences (rf) permettent de confiner des particules chargées (atomes ou molécules) dans un domaine restreint de l'espace. Alliés aux techniques de refroidissement laser, ils permettent l'étude et le contrôle de la matière au niveau atomique. Dans cette thèse sont présentés différents résultats obtenus au moyen de simulations numériques, utilisant la technique de la dynamique moléculaire. Cette thèse se compose de quatre parties. Une première partie rassemble les connaissances indispensables pour appréhender la dynamique des ions en pièges rf. La deuxième partie présente les résultats d'une étude sur l'influence du couplage coulombien sur le refroidissement laser Doppler d'un ensemble d'ions confinés en quadrupôle linéaire. La troisième partie s'intéresse aux propriétés des structures en anneau d'ions, obtenues à basse température au sein des multipôles linéaires d'ordres supérieurs au quadrupôle. Une dernière partie présente une proposition permettant l'obtention de sites supplémentaires où le champ rf s'annule dans les multipôles linéaires d'ordres supérieurs, offrant la possibilité d'obtenir des structures non soumises au forçage rf. / Radio-frequency (rf) traps allows one to confine charged particles (atoms or molecules) in a small region of space. These traps, with laser cooling technics, allows one to study and to control matter at the atomic level.In this thesis are shown different results obtained by numerical simulations using molecular dynamics technics. This thesis is formed of four parts. A first part introduce the necessary knowledge to comprehend the ion dynamics in rf traps. The second part present the results from a study on the influence of coulomb coupling on the doppler cooling of trapped ions in a linear quadrupole rf trap. The third part is a study of the properties of ring structures of ions, obtained for low temperatures in linear higher order rf traps. A last part of the thesis is a proposal allowing one to obtain additional rf field-free regions in linear higher order rf traps.

Cette thèse décrit le travail accompli au cours des trois dernières années sur la nouvelle expérience d’Hélium métastable de l'Institut d'Optique à Palaiseau.Le premier chapitre décrit une étude visant à améliorer à la fois l'efficacité et la précision du système de détection par galettes à micro-canaux (MCP). Nous avons fait des mesures avec des galettes recouvertes d'une couche d'or sur la face avant, obtenant un accroissement de l'efficacité de détection mais également une réduction de la précision. L'ajout d'une tension intermédiaires entre les deux galettes empilées a au contraire améliorer à la fois l'efficacité et la précision des MCP, en l’absence d’une couche d’or.Le deuxième chapitre est consacré à la construction de l'appareil expérimental pour le refroidissement et le piégeage d’atomes. L’excitation de l'état fondamental de l’Hélium à l'état métastable est décrite, ainsi que la collimation et le refroidissement ultérieur par Zeeman lent du faisceau atomique chaud résultant. Le faisceau ralenti est alors capturé dans un piège magnéto-optique (PMO), dans lequel nous avons capturé 8x108 atomes.Dans le troisième chapitre une étude originale du refroidissement Doppler tridimensionnel dans un PMO et une mélasse désaccordée vers le rouge de la transition atomique est discutée. L’atome d’Hélium métastable est unique et ses propriétés ont permis une telle étude. En effet, les faibles densités atomiques impliquent qu’il n'y a pas de diffusion multiple de photons d’une part, et la faible masse et la faible largeur de la transition 23S1 -> 23P2 rend inefficace les processus de refroidissement sous la limite Doppler. Ces conditions nous ont permis d’observer pour la première fois à trois dimensions un gaz refroidit dans le régime Doppler.Le quatrième chapitre présente une étude sur les collisions dans un piège magnéto-optique d’Hélium métastable. Les collisions Penning induites par la lumière, en particulier à des intensités élevées et à des fréquences proches de la fréquence de transition, sont responsables de pertes élevées d’atomes piégés. Nous mesurons le coefficient de taux associé à ces pertes, Ksp = 2,8 ± 0,4 x 10-7cm3/ s. / This thesis describes the work done over the past three and a half years on the new metastable helium experiment at the Institut d'Optique in Palaiseau. In the first chapter it describes a study to improve both the efficiency and the accuracy of the Microchannel Plate (MCP) detection system. We have experimented with adding a gold layer on the top of the input plate, something that we have found increases the efficiency but also decreases the accuracy. The addition of a voltage between the two stacked plates has been shown to both raise the efficiency and improve the accuracy in non-coated MCPs.The second chapter is devoted to the construction of the experimental apparatus. Here the excitation of ground state helium to its metastable state is described, as well as the subsequent collimation and cooling by Zeeman slower of the resulting hot atomic beam. The slowed beam is then captured in a Magneto-Optical Trap, in which we have captured 8x108 atoms.In the third chapter an original study on three-dimensional Doppler cooling in a red-detuned molasses and in the Magneto-Optical trap is presented. The metastable helium system is unique as there is no multiple scattering of photons and there are no sub-Doppler effects. This allows for a never before seen experimental realisation of pure Doppler cooling theory. The fourth chapter describes a study on collisions in a magneto-optical trap of metastable helium. Light-induced Penning collisions are responsible for high trap losses at high intensities and at frequencies close to the transition frequency. We measure the constant rate coefficient to Ksp = 2.8 ± 0.4 x 10-7cm3/s.

Nós demonstramos, pela primeira vez, o resfriamento via laser de um feixe atômico duplo. Ao se carregar um forno a 650oC com uma liga de Li179Na obtêm-se uma destilação efusiva que produz um feixe com fluxos idênticos de Li e de Na. Este feixe formado por duas espécies é desacelerado pela técnica Zeeman utilizando-se dessintonias iguais a Li = -850 MHz e Na = +150 MHz das respectivas transições atômicas 2S1/2 F = 2 2P3/2 F = 3. Em um segundo experimento, mostramos uma nova armadilha magneto-óptica bicromática, em cela de vapor. Obtivemos um incremento de ~10 vezes no número de átomos resfriados utilizando uma armadilha de Na do tipo I circundada por uma armadilha do tipo II, que não se superpõe à primeira. Isto ocorre quando temos o laser de captura a 12 MHz para o vermelho da transição 32S1/2 F = 2 32P3/2 F = 2. Em um terceiro experimento, observamos a primeira armadilha magneto-óptica para alcalinos operando na transição 2S1/2 2P1/2 (linha D1). Para o caso escolhido de átomos de sódio, obtêm-se quatro armadilhas diferentes dentro desta linha D1, contendo de 105 a 107 átomos resfriados. Para duas delas, é necessário inverter-se o sinal das polarizações circulares usuais dos lasers de resfriamento, devido aos desvios Zeeman negativos dos níveis hiperfinos inferiores / We demonstrate, for the first time, a laser cooling of a double atomic beam. By loading an oven at 650oC with a Li179Na alloy we get an effusive distillation producing a beam with equal fluxes of Li and Na. This two-species beam was laser cooled by Zeeman technique with Li = -850 MHz and Na = +150 MHz detunings from the corresponding 2S1/2 F = 2 2P3/2 F = 3 atomic transitions. In a second experiment, we show a new two-color vapor-cell magneto-optical trap. By surrounding type-I Na trap with a non-overlaping type-II trap we achieve a ~10-fold improvement in the number of cooled atoms. This is achieved when the capture laser is detuned 12 MHz to the red of the 32S1/2 F = 2 32P3/2 F = 2 transition. In a third experiment, we observe the first alkaline magneto-optical trap operating on 2S1/2 2P1/2 transition (D1 line). For the choosen sodium atoms, we have four different traps within D1 line, with 105 to 107 cooled atoms. For two of them, one has to revert the usual sign of circular polarizations of the cooling lasers, due to negative Zeeman shift of the lower hyperfine levels

Nós demonstramos, pela primeira vez, o resfriamento via laser de um feixe atômico duplo. Ao se carregar um forno a 650oC com uma liga de Li179Na obtêm-se uma destilação efusiva que produz um feixe com fluxos idênticos de Li e de Na. Este feixe formado por duas espécies é desacelerado pela técnica Zeeman utilizando-se dessintonias iguais a Li = -850 MHz e Na = +150 MHz das respectivas transições atômicas 2S1/2 F = 2 2P3/2 F = 3. Em um segundo experimento, mostramos uma nova armadilha magneto-óptica bicromática, em cela de vapor. Obtivemos um incremento de ~10 vezes no número de átomos resfriados utilizando uma armadilha de Na do tipo I circundada por uma armadilha do tipo II, que não se superpõe à primeira. Isto ocorre quando temos o laser de captura a 12 MHz para o vermelho da transição 32S1/2 F = 2 32P3/2 F = 2. Em um terceiro experimento, observamos a primeira armadilha magneto-óptica para alcalinos operando na transição 2S1/2 2P1/2 (linha D1). Para o caso escolhido de átomos de sódio, obtêm-se quatro armadilhas diferentes dentro desta linha D1, contendo de 105 a 107 átomos resfriados. Para duas delas, é necessário inverter-se o sinal das polarizações circulares usuais dos lasers de resfriamento, devido aos desvios Zeeman negativos dos níveis hiperfinos inferiores / We demonstrate, for the first time, a laser cooling of a double atomic beam. By loading an oven at 650oC with a Li179Na alloy we get an effusive distillation producing a beam with equal fluxes of Li and Na. This two-species beam was laser cooled by Zeeman technique with Li = -850 MHz and Na = +150 MHz detunings from the corresponding 2S1/2 F = 2 2P3/2 F = 3 atomic transitions. In a second experiment, we show a new two-color vapor-cell magneto-optical trap. By surrounding type-I Na trap with a non-overlaping type-II trap we achieve a ~10-fold improvement in the number of cooled atoms. This is achieved when the capture laser is detuned 12 MHz to the red of the 32S1/2 F = 2 32P3/2 F = 2 transition. In a third experiment, we observe the first alkaline magneto-optical trap operating on 2S1/2 2P1/2 transition (D1 line). For the choosen sodium atoms, we have four different traps within D1 line, with 105 to 107 cooled atoms. For two of them, one has to revert the usual sign of circular polarizations of the cooling lasers, due to negative Zeeman shift of the lower hyperfine levels