abstract: Nanofluidic devices in which one single-walled carbon nanotube (SWCNT) spans a barrier between two fluid reservoirs were constructed, enabling direct electrical measurement of the transport of ions and molecules. Ion current through these devices is about 2 orders of magnitude larger than that predicted from the bulk resistivity of the electrolyte. Electroosmosis drives excess current, carried by cations, and is found to be the origin of giant ionic current through SWCNT as shown by building an ionic field-effect transistor with a gate electrode embedded in the fluid barrier. Wetting of inside of the semi-conducting SWCNT by water showed the change of its electronic property, turning the electronic SWCNT field-effect transistor to "on" state. These findings provide a new method to investigate and control the ion and molecule behavior at nanoscale. / Dissertation/Thesis / Ph.D. Physics 2011

abstract: Mechanisms for oxygen reduction are proposed for three distinct cases covering two ionic liquids of fundamentally different archetypes and almost thirty orders of magnitude of proton activity. Proton activity is treated both extrinsically by varying the concentration and intrinsically by selecting proton donors with a wide range of aqueous pKa values. The mechanism of oxygen reduction in ionic liquids is introduced by way of the protic ionic liquid (pIL) triethylammonium triflate (TEATf) which shares some similarities with aqueous acid solutions. Oxygen reduction in TEATf begins as the one electron rate limited step to form superoxide, O2*-, which is then rapidly protonated by the pIL cation forming the perhydroxyl radical, HO2*. The perhydroxyl radical is further reduced to peroxidate (HO2-) and hydrogen peroxide in proportions in accordance with their pKa. The reaction does not proceed beyond this point due to the adsorption of the conjugate base triethylammine interfering with the disproportionation of hydrogen peroxide. This work demonstrates that this mechanism is consistent across Pt, Au, Pd, and Ag electrodes. Two related sets of experiments were performed in the inherently aprotic ionic liquid 1-butyl-2,3-dimethylimidazolium triflate (C4dMImTf). The first involved the titration of acidic species of varying aqueous pKa into the IL while monitoring the extent of oxygen reduction as a function of pKa and potential on Pt and glassy carbon (GC) electrodes. These experiments confirmed the greater propensity of Pt to reduce oxygen by its immediate and abrupt transition from one electron reduction to four electron reduction, while oxygen reduction on GC gradually approaches four electron reduction as the potentials were driven more cathodic. The potential at which oxygen reduction initiates shows general agreement with the Nernst equation and the acid's tabulated aqueous pKa value, however at the extremely acidic end, a small deviation is observed. The second set of experiments in C4dMImTf solicited water as the proton donor for oxygen reduction in an approximation of the aqueous alkaline case. The water content was varied between extremely dry (<0.1 mol% H2O) and saturated (approximately 15.8 mol% H2O}). As the water content increased so too did the extent of oxygen reduction eventually approach two electrons on both Pt and GC. However, additional water led to a linear increase in the Tafel slope under enhanced mass transport conditions up to the point of 10 mol% water. This inhibition of oxygen adsorption is the result of the interaction between superoxide and water and more specifically is proposed to be associated with decomposition of theC4dMIm+ cation by hydroxide at the elevated temperatures required for the experiment. Oxygen reduction on both Pt and GC follows Nernstian behavior as the water content is increased. Separate mechanisms for oxygen reduction on Pt and GC are proposed based on the nature of the Nernstian response in these systems. / Dissertation/Thesis / Ph.D. Materials Science and Engineering 2011

No fim do século XX cresceu o interesse por materiais com potencial para aplicação na área da nanotecnologia. Um destes materiais é o grafeno (alótropo do carbono da espessura de um átomo), indicado para aplicações que exploram desde propriedades ópticas até mecânicas. Todavia, os métodos de produção estabelecidos até o momento geram normalmente um produto com alto grau de defeitos estruturais. Por isso, o objetivo deste estudo foi desenvolver um método mais brando para obtenção de grafeno, por meio do desenrolamento de nanotubos de carbono (NTC) de parede múltipla na presença de líquidos iônicos (LI) (que possuem caráter lubrificante e estabilizador) e adicioná-lo à resina epoxídica para avaliar seu efeito nas propriedades do nanocompósito. A metodologia consistiu em definir a rota a que a mistura NTC:LI (bis(trifluorometanosulfonil)imidato de 1-n-butil-3-metilimidazólio – BMImNTf2), na proporção 1:10, deveria ser submetida e depois avaliar variações para validação do método (alteração de temperatura, solventes – LI BMImCl e tolueno e nanocarga – grafite). As misturas de grafeno-NTC/LI produzidas foram adicionadas na proporção de 0,1% em massa de grafeno-NTC à resina epoxídica. A caracterização das misturas se deu pelas técnicas de microscopia eletrônica de transmissão (MET), espectroscopia Raman e microscopia de força atômica (AFM), e dos nanocompósitos, também por microscopia eletrônica de varredura (MEV), análise termogravimétrica (TGA), calorimetria diferencial exploratória (DSC), infravermelho (IV) análise dinâmico-mecânica (DMA) e ensaios mecânicos. A morfologia observada em MET e AFM evidenciou que a rota em que a mistura permaneceu 3 h em aquecimento e agitação magnética sob vácuo e 3 h sob sonificação resultou em grafeno de algumas camadas. Os resultados de Raman mostraram que foi produzido grafeno de boa qualidade, baseando-se na razão ID/IG. Nos nanocompósitos, confirmou-se que BMImNTf2 mantém as folhas de grafeno abertas. Já com BMImCl a morfologia dos NTC permanece inalterada, mas seu grau de emaranhamento é diminuído pela ação lubrificante deste LI, o que reflete em melhoria na resistência ao impacto. Os NTC sem a presença de um LI de forma geral não conferiram melhoria à matriz polimérica (ex: queda de 35% na dureza). / In the late twentieth century, the interest in materials with potential application in the nanotechnology´s field has increased. One of these materials is graphene (carbon allotrope of a single atom thickness), suitable for applications, which exploit properties such as optical or mechanical. However, the methods developed up to now to obtain graphene usually generate a product with high degree of structural defects. Therefore, the aim of this study was to produce graphene sheets through a milder route of unrolling multi-walled carbon nanotubes (CNT) in ionic liquids (IL; which have a lubricating and stabilizing character) and to add it to an epoxy resin to evaluate its effect on the nanocomposite properties. The methodology consisted of defining the route through which the 1:10 mixture of CNT:IL (1-n-butyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imidate - BMImNTf2) should be submitted, and then evaluate possible variations for validating the method (change in temperature, solvents - toluene and IL BMImCl and nanofiller - graphite). The produced graphene-CNT/IL mixtures were added at 0.1% by weight of graphene-CNT to epoxy resin. Characterization of the graphene/IL mixtures was done by transmission electron microscopy (TEM), Raman spectroscopy and atomic force microscopy (AFM), and the nanocomposites also by scanning electron microscopy (SEM), thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), Fourier transform infrared spectroscopy (FTIR), dynamical mechanical analysis (DMA) and mechanical tests. The morphology observed using TEM and AFM showed that the route based on 3 h under heating, stirring and vacuum, followed by 3 h of sonication leads to CNT unrolling. Raman spectra showed that the produced graphene was of good quality, based on the low ID/IG ratio. In the nanocomposites, CNT unrolling was confirmed when IL BMImNTf2 was used, yielding reasonable property improvements. In the case of IL BMImCl, the unrolling of the CNT did not occur, but the lubricant effect of the IL decreased the degree of entanglement, which increased its impact strength. Differently, the non-treated CNT generally reduced the properties of the polymer matrix, e.g. a 35% drop in hardness.

The use of and research on gamification has increased substantially since 2010. Most of the research on gamification has been done within education, specifically higher education. There is a dearth of research on gamification for children with disabilities. There has also been much research on the characteristics of successful instructional and educational games. In this thesis, theory on both gamification and the characteristics of successful instructional and educational games were used to design and implement several game design elements in a mobile sign language learning application to make it more enjoyable for the users. The game design elelemnts implemented were points, appropriate challenge, proximal goals, randomness and control. The target group of the application is children with disabilities making the thesis unique compared to any previous research. A rigorous gamification methodology was followed and a meeting with experts on children with special educational needs was held to get a good understanding of the target group. The development was done iteratively with two-week long iterations. After each iteration a presentation was held for the intern company and feedback was given. A school was willing to participate in an evaluation session and at the end of the thesis and the session was held with 22 children with learning disabilities. Both the original application and the application with the added gamification elements were evaluated and the results were compared. Results found that the children using the new version of the application had less throughput and found learning sign language to be more enjoyable and easier compared to the children using the original version of the application. The results are questionable, however, due tolimitations of the evaluation. Some of the issues were that the ages of the children differed between the groups, the severity of the disabilities was not considered, and the participants only got to use one version of the application for fifteen minutes. / 自2010以来，游戏化技术的使用和研究大幅增加。在教育领域,特别是高等教育领域,有很多很多关于游戏化的研究。关于成功教学和教育游戏的特征也有很多相关研究。然而目前对残疾儿童游戏化技术的研究尚不多见。本文运用游戏化理论和成功教学游戏和教育游戏的特点，设计并实现了移动手语学习应用中的几个游戏设计元素，使移动手语学习应用更加有趣。实施的游戏设计要素有明确的目标、难度的选择、随机性、点和层次。应用的目标群体是残疾儿童，使得本论文与以往研究有所不同。我们遵循严格的游戏化方法，并与特殊儿童教育专家举行会议，以便更好地了解目标群体。用两星期的时间进行迭代开发。每次迭代后，在实习生公司进行演示并得到反馈。一所学校愿意参加评测，在论文结束时，22名有学习障碍儿童参加了评测。对原有应用和添加游戏化元素的应用进行评价，并对结果进行比较。结果发现，使用新版本应用程序的儿童比使用原始版本应用程序的儿童的工作量较少，并且学习手语过程更愉快、更容易。然而由于评估的局限性,可能会影响实验结果,例如各组儿童的年龄不同，残疾的严重程度没有考虑，并且参与者只能使用一个版本的应用程序15分钟。后续研究将评估这些因素的影响。

Os líquidos iônicos são uma das principais apostas para revolucionar a química no século XXI. Em geral, apresentam propriedades como baixa pressão de vapor, alta estabilidade térmica, e ampla existência liquida e outras que os tornam interessantes para aplicações como solventes para extração, meio reacional e como meio catalítico. Estas propriedades permitem a redução da geração de efluentes industriais, o que os torna potenciais solventes verdes. O interesse recente aumentou a quantidade de ânions e cátions disponíveis para formar líquidos iônicos. Essa maior diversidade por um lado aumenta o potencial de aplicação industrial dos líquidos iônicos, mas por outro aumenta a carga de trabalho para obter as propriedades necessárias para projeto industrial destas novas substâncias. Métodos preditivos podem ser utilizados para evitar a grande demanda de ensaios experimentais para obter essas propriedades. Entretanto modelos puramente preditivos, como o COSMO-RS, usualmente não obtém uma acurácia adequada para o uso industrial. Modelos que fazem uso de dados experimentais de misturas para estimar seus parâmetros, como o UNIFAC (Do) e o F-SAC, são predominantes na indústria. Esses modelos, que são modelos de coeficiente de atividade, permitem o cálculo de equilíbrio de fases e outras propriedades termodinâmicas. Este trabalho teve como objetivo o desenvolvimento do modelo F-SAC para líquidos iônicos. Baseado em trabalhos anteriores, foram reestimados os parâmetros dos grupos de líquidos iônicos já existentes e adicionados novos grupos de cátions e ânions. Também foram geradas as primeiras curvas de equilíbrio líquido-vapor de misturas contendo líquidos iônicos com o F-SAC. De forma geral, para os 1-alquil-3-etil-imidazólios, foi obtido um R2 de 0.9794 e o valor de AAD de 0.1721 para 1514 misturas com o modelo F-SAC, superior ao encontrado para o UNIFAC (Do). Para o piridínio e o pirrolidínio o F-SAC obteve, respectivamente, um R2 de 0.9281 e 0.9837 e um valor de AAD de 0.2749 e 0.0962. Com este trabalho é possível formar 24 diferentes líquidos iônicos base com o uso de 33 parâmetros eletrostáticos estimados e 22 parâmetros geométricos no F-SAC. Para o mesmo objetivo o modelo UNIFAC (Do) necessitaria de 180 parâmetros binários e 26 parâmetros individuais – todos estimados. Ou seja, o F-SAC, nesse caso, precisou estimar apenas 27% do número de parâmetros que o UNIFAC (Do). / Ionic Liquids are one of the most promising fields to revolutionize chemistry in the 21st century. In general, they show properties like negligible vapor pressure, high thermal stability, wide liquid existence turn them interesting for applications as solvents for extraction, as reactional and catalytic media. These properties allow a reduction in industrial effluent production, making them eligible green solvents. Recent inteerest expanded the number of anions and cations available to compose Ionic. This great diversity enhances the industrial application potential of Ionic Liquids, while increases the workload needed for obtaining properties of these new substances needed for industrial project. Predictive methods can used to avoid the demand of experiments to obtain these properties. Purely predictive methods, as COSMO-RS, are usually not accurate enough for industrial use. Models use experimental data from mixtures for parameter estimation like UNIFAC (Do) and F-SAC, are predominant in the industry. Activity coefficient models like these allow the solving of phase equilibria and other thermodynamic properties. This work aimed at developing the F-SAC model for Ionic Liquids. Based on previous works, the already existing parameters for Ionic Liquid groups were reestimated and new cation and anion groups were added. In addition, the first vapor-liquid equilibrium curves for mixtures containing Ionic Liquids were created for the F-SAC model. For 1-alkyl-3-ethyl-imidazoliums, an overall R2 of 0.9794 and AAD of 0.1721 for 1514 mixtures was obtained, superior to UNIFAC (Do) values. For pyridinium and pyrrolidinium, F-SAC achieved, respectively, R2 of 0.9281 and 0.9837 and AAD of 0.2749 and 0.0962. With this work is possible to form 24 different base Ionic Liquids with 33 estimated electrostatic parameters and 22 geometric parameters with F-SAC. For the same purpose, UNIFAC (Do) would need 180 binary parameters and 23 individual parameters – all estimated. For this particular case the F-SAC model needed to estimate only 27% of the parameter number UNIFAC (Do) needed.

Doctor of Philosophy / Department of Chemistry / Emily McLaurin / Semiconductors are materials whose conductivity is between metals and insulators. Semiconductor nanocrystals (NCs) have sizes in the range 2 to 10 nm. Because of their unique optical properties like tunable emission wavelength, narrow emission peak, and stability over dyes, they have potential applications in displays. Indium phosphide (InP) is considered a less toxic alternative to commercially used cadmium-based semiconductor NCs. Microwave-assisted (MA) methods using ionic liquids (ILs) afford fast reaction heating rates because of the good MW absorbing capacity of ILs. For tuning size and surface, which are some of the important problems associated with the InP NCs, new synthetic methods are reported herein. In MAIL etching HF generated in the microwave reaction etches the InP NCs surface. Pyridinium and imidazolium based ILs containing tetrafluoroborate (BF₄⁻) and hexafluorophosphate (PF₆⁻) ions yield luminescent NCs. In a silicon carbide (SiC) reaction vessel, which blocks most of the microwaves penetrating into the reaction, bigger NCs form than those from a Pyrex reaction vessel because of the higher reaction temperatures in the SiC vessel. By changing microwave set-power (SP), different reaction times can be achieved. Though a small degree of change in average NC diameter of the NCs is observed at different SPs and reaction temperatures, addition of dodecylamine (DDA) yields NCs with average sizes between 3.2 to 4.2 nm with a broad size distribution. At lower SPs smaller NCs form and at higher SPs bigger NCs form. NC luminescence can be tuned from green (545 nm) to red (630 nm) in the visible region with quantum yields as high as 30%. Rapid heating and InP precursor activation might be responsible for the larger change in NC size. The effect of DDA on NC size is also verified by microwave reactions in SiC vessels. ILs containing PF₆⁻ ions at 280 °C will modify the surface of the NCs so the NC dispersibility changes from non-polar (toluene) to polar (DMSO) as the amount of IL increases. This is due to ligand stripping, which is the removal of large palmitic ligands from the NC surface. These NCs have broad absorption features and emission peaks with QYs of up to 30%. Fourier transform infrared spectroscopy indicates the absence of palmitic acid ligands on the NC surface and zeta potential measurements indicate the presence of anions on the NC surface. From X-ray photoelectron spectroscopy and nuclear magnetic resonance spectroscopy, the inorganic ion PO₂F₂⁻ is identified on the NCs surface.

Les dispositifs à base de GaN et ses alliages sont de plus en plus présents dans notre quotidien avec le développement exponentiel des diodes électroluminescentes (LED). Bien que la majorité des productions commerciales soient pour le moment effectuées sur substrat saphir, le silicium, disponible en de plus grands diamètres et pour un coût moindre, est de plus en plus pressenti comme le substrat d’avenir pour le développement des technologies GaN. L’utilisation de ce substrat devrait aussi permettre le développement du marché de l’électronique de puissance du GaN basée sur les transistors à haute mobilité électronique (HEMT) dont les performances dépassent les limites des technologies silicium. Néanmoins, afin de permettre ou faciliter le développement de dispositifs avancés, certaines briques technologiques sont nécessaires comme le dopage par implantation ionique. L’utilisation du GaN soulève des problématiques nouvelles pour ces briques technologiques.Au cours de cette thèse nous avons donc cherché à implémenter le procédé de dopage par implantation ionique du GaN et son étude au sein du CEA-LETI en nous focalisant principalement sur le dopage p par implantation de Mg. Nous avons identifié les principales problématiques liées aux propriétés intrinsèques du matériau (difficulté du dopage p, instabilité à haute température…) et les solutions les plus prometteuses de la littérature. Nous avons ensuite cherché à mettre en place notre propre procédé en développant des couches de protection déposées in-situ pour permettre les traitements thermiques à haute température des couches implantées. Cela a rendu possible l’étude des cinétiques d’évolution des couches implantées pendant des recuits « conventionnels » (rampes < 10 °C/min, durée de plusieurs dizaines de minutes, T < 1100 °C) en utilisant notamment des caractérisations de photoluminescence (µ-PL) et de diffraction des rayons X (XRD). Nous avons aussi mis en évidence un effet de diffusion et d’agrégation à haute température du Mg implanté. Nous avons ensuite cherché à modifier le procédé d’implantation (implantation canalisée, co-implantation) pour favoriser l’intégration du dopant et limiter la formation de défauts. En parallèle nous avons évalué l’intérêt de recuits secondaires (recuits rapides (RTA), recuit laser, micro-ondes) afin de finaliser l’activation du dopant. Finalement nous avons aussi mis en place un procédé de caractérisation électrique de couche de GaN dopées au sein du laboratoire. / GaN-based devices and their alloys are increasingly present in our daily lives with the exponential development of light-emitting diodes (LEDs). Although the majority of commercial production is currently carried out on sapphire substrates, silicon, available in larger diameters and at a lower cost, is increasingly seen as the substrate of the future for the development of GaN technologies. The use of this substrate should also allow the development of the GaN power electronics market based on high electron mobility transistors (HEMTs) whose performances exceed the limits of silicon technologies. Nevertheless, in order to allow or facilitate the development of advanced devices, specific processes are necessary such as doping by ion implantation. The use of GaN raises new problems for these technological bricks.During this thesis we therefore sought to implement the ion implantation doping process of GaN and its study within the CEA-LETI while focusing mainly on p doping by Mg implantation. We have identified the main issues related to the intrinsic properties of the material (difficulty of p-doping, instability at high temperatures...) and the most promising solutions in the literature. We then sought to implement our own process by developing in-situ protective layers to allow high temperature annealing of the implanted layers. This enabled the study of the evolution kinetics of the implanted layers during "conventional" annealing (ramps < 10 °C/min, duration of several tens of minutes, T < 1100 °C) using photoluminescence (µ-PL) and X-ray diffraction (XRD) characterizations. We also evidenced a diffusion and aggregation effect at high temperature of the implanted Mg. We then sought to modify the implantation process (channeled implantation, co-implantation) to promote the integration of the dopant and limit the formation of defects. In parallel we evaluated the interest of secondary annealing (Rapid thermal annealing (RTA), laser annealing, microwave) in order to finalize the activation of the dopant. Finally we also set up an electrical characterization process for doped GaN layers in the laboratory.

abstract: The electrode-electrolyte interface in electrochemical environments involves the understanding of complex processes relevant for all electrochemical applications. Some of these processes include electronic structure, charge storage, charge transfer, solvent dynamics and structure and surface adsorption. In order to engineer electrochemical systems, no matter the function, requires fundamental intuition of all the processes at the interface. The following work presents different systems in which the electrode-electrolyte interface is highly important. The first is a charge storage electrode utilizing percolation theory to develop an electrode architecture producing high capacities. This is followed by Zn deposition in an ionic liquid in which the deposition morphology is highly dependant on the charge transfer and surface adsorption at the interface. Electrode Architecture: A three-dimensional manganese oxide supercapacitor electrode architecture is synthesized by leveraging percolation theory to develop a hierarchically designed tri-continuous percolated network. The three percolated phases include a faradaically-active material, electrically conductive material and pore-former templated void space. The micropores create pathways for ionic conductivity, while the nanoscale electrically conducting phase provides both bulk conductivity and local electron transfer with the electrochemically active phase. Zn Electrodeposition: Zn redox in air and water stable N-ethyl-N-methylmorpholinium bis(trifluoromethanesulfonyl)imide, [C2nmm][NTf2] is presented. Under various conditions, characterization of overpotential, kinetics and diffusion of Zn species and morphological evolution as a function of overpotential and Zn concentration are analyzed. The surface stress evolution during Zn deposition is examined where grain size and texturing play significant rolls in compressive stress generation. Morphological repeatability in the ILs led to a novel study of purity in ionic liquids where it is found that surface adsorption of residual amine and chloride from the organic synthesis affect growth characteristics. The drivers of this work are to understand the processes occurring at the electrode-electrolyte interface and with that knowledge, engineer systems yielding optimal performance. With this in mind, the design of a bulk supercapacitor electrode architecture with excellent composite specific capacitances, as well as develop conditions producing ideal Zn deposition morphologies was completed. / Dissertation/Thesis / Ph.D. Materials Science and Engineering 2011

La réduction de la résistance à l’avancement par des techniques de contrôle active sera une nécessité d’importance croissante dans les années à venir pour le secteur des transports. Cette thèse fait partie d’une nouvelle voie de recherche consistant à utiliser des décharges plasmas à pression atmosphérique pour modifier les écoulements en vue d’améliorer les performances aérodynamiques (diminution de la traînée et augmentation de la portance). L’actionneur retenu est une décharge à barrière diélectrique surfacique avec un diélectrique composé d’une combinaison de deux matériaux polymères souples. L’écoulement induit par l’actionneur plasma, appelé vent ionique, est d’abord étudié expérimentalement dans l’air au repos en fonction du régime de la décharge. Une loi empirique liant la vitesse du vent ionique à la tension et à la fréquence de l’alimentation est proposée. De cette loi se dégagent une vitesse et une longueur caractéristiques utilisées pour prédire les principales propriétés du jet de paroi induit en aval de la décharge (lois d’échelle). L’efficacité de l’actionneur est ensuite testée en soufflerie subsonique pour trois configurations aérodynamiques de complexité croissante : couche limite laminaire non décollée (plaque plane), couche limite turbulente décollée (profil épais) et couche limite décollée 3D (corps de Ahmed, lunette inclinée à 25°). Des résultats significatifs tant sur le plan de la modification de la topologie de l’écoulement que de l’amélioration des efforts aérodynamiques sont obtenus pour des nombres de Reynolds jusqu’à 1,7.106 (35 m/s). / Active flow control is a key issue to meet the environmental requirements for the next generations of aircrafts and cars. The aim of this work is to study subsonic airflows control by plasma actuators in order to improve aerodynamic performances (drag reduction and lift enhancement). The actuator consists in using a surface dielectric barrier discharge with a thin dielectric made by a combination of two polymer materials. The first part of this work is dedicated to the experimental study of the flow induced by the actuator in quiescent air according to the regime of the discharge. An empirical law is proposed in order to determine the velocity of the ionic wind as a function of the high voltage and the frequency of the actuator. From this law the main properties of the wall jet induced by the discharge are predicted. In the second part of the study the actuator is used for wind tunnel experiments for three configurations of increasing complexity: laminar boundary layer (flat plane), separated turbulent boundary layer (wing profile) and 3D separated boundary layer (Ahmed body with rear window tilted at 25°). Significant results are obtained for Reynolds numbers up to 1.7.106 (35 m/s).

Este estudo utilizou os princípios da glacioquímica para determinar e analisar as variações nas concentrações aniônicas de um testemunho de neve e firn obtido pela perfuração no manto de gelo da antártica ocidental no verão austral de 2004/05. O testemunho IC-6 (81°03'S, 79°51'W), de 34,65 m de profundidade, obtido a 750 m de altitude, foi subamostrado em sala limpa (CLASE 100), usando um sistema de derretimento contínuo desenvolvido pela equipe do Climate Change Institute (CCI) da Universidade do Maine (EUA). Esse processo gerou 1.368 amostras para análises por cromatografia iônica, cerca de 58 amostras por metro, permitindo detalhamento sazonal da variabilidade das concentrações dos íons majoritários. O testemunho representa 66 ± 3 anos de dados ambientais, segundo a datação baseada na variação sazonal dos íons Cl-, Na+, Mg+2 e SO4-2. O testemunho de 23,61 m em equivalente d’água, corrigido para variações em densidade, representa uma acumulação liquida média anual de 0,36 m (em equivalente d’água). Assim, a camada ao fundo foi formada no ano de 1938 (± 3 anos). As concentrações iônicas médias medidas no IC-6, são: [(Na+= 66,92 ± 2,32 μg L-1), (K+= 3,31 ± 0,18 μg L-1); (Mg+2= 10,07 ± 0,25 μg L-1); (Ca+2 = 16,93 ± 0,38 μg L-1); (Cl- = 155,74 ± 4,40 μg L-1); (NO3- = 56,01 ± 0,80 μg L-1); (SO4 2 = 55,65 ± 1,36 μg L-1); e (CH3SO3 (MS) = 14,11 ± 1,19 μg L-1)]. As maiores concentrações de Na+, Cl-, e Mg+2 foram interpretadas como picos de invernos, associadas diretamente ao aerossol dos mares circundantes em respostas, provavelmente, a advecção mais intensa de massas de ar (marinho) sobre as plataformas de gelo, e portanto são também traçadores marinhos. Já o perfil (série) de sulfato está em antifase, em relação às variações nas espécies Na+, Cl- e Mg+2. De origem predominantemente marinha, o sulfato total apresentou maiores concentrações durante a primavera e verão (períodos de maior atividade biológica nos mares circumpolares), possivelmente marcando a variação sazonal da atividade biológica na região. Embora em alguns intervalos essa ―antifase‖ não fique tão clara, é o que ocorre na maior parte do testemunho IC-6, condição que auxiliou na interpretação da variação sazonal observada principalmente na série do cloro. O perfil de excesso de sulfato apresenta perfil similar ao de sulfatos total, com picos concomitantes. Além da forte correlação com o íon SO4-2, também é observada uma correlação fraca a moderada com o íon nitrato. Picos concomitantes deste íon com o excesso de sulfato representam eventos episódicos como é o caso das erupções vulcânicas de grande magnitude. A variabilidade da concentração de nitrato não esta associada ao aerossol marinho, como aponta a falta de correlação entre esse ânion e o Cl-, Na+ e Mg2+. Porém, série de nitrato apresenta muitos períodos bem marcados e correlacionados com as concentrações de excesso de sulfato, devendo representar a ocorrência de eventos episódicos, como erupções vulcânicas. Entretanto, a análise de íons maiores nesse estudo não possibilitou a identificação de eventos específicos, será necessário o uso de técnicas complementares para determinação de elementos traços. Sugere-se que o nitrato seria transportado e depositado por massas de ar provenientes da estratosfera ou da alta troposfera e que grandes concentrações dessa espécie poderiam estar associadas ao registro de ocorrências de eventos vulcânicos. Essa característica parece ser coerente com os picos correlacionáveis nos perfis (séries) de nitrato e sulfatos. Além da variação sazonal (observada principalmente no perfil de cloro), foram identificados outros padrões recorrentes no tempo (ciclos), principalmente nas séries de dos íons Na+, Cl- e Mg2+ (origem marinha) e NO3-. O principal ciclo identificado, de aproximadamente 17,3 anos, necessita melhor investigação. A secundária, em torno de 10 anos, estaria associada ao ciclo solar (de 10,7 anos). Também são observados ciclos com períodos entre 2 a 5 anos, que poderiam estar associados ao fenômeno ENOS (El Niño - Oscilação Sul). Ao comparar as concentrações médias do IC-6 com de outros sítios no interior da Antártica, observa-se uma abrupta redução ao atravessar as montanhas Transantárticas em direção ao Polo Sul geográfico. Sugere-se que cordilheira esteja barrando o transporte dos aerossóis marinhos para o interior do continente devido a um efeito orográfico sobre a precipitação. / This study employed glaciochemical principles to determine and analyze the variation of anionic concentrations of a firn and snow core obtained from the Western Antarctic Ice Sheet, in the summer of 2004/05. The IC-6 core (81°03'S, 79°51'W), reaching 34.65 m in depth, was extracted at 750 m above sea level. This core was subsampled in a Class 100 clean room, employing a discrete continuous melting system developed by the team at the Climate Change Institute (CCI), University of Maine, USA. This process produced 1,368 samples for ionic chromatographic analyses, approximately 58 samples per meter, permitting a seasonal-scale resolution of the main ion concentrations and variabilities. This core represents 66 ± 3 years of environmental data, according to Cl-, Na+, Mg+2 e SO4-2 ion seasonal variations. The 23.61 m core, in water equivalent, corrected for the density variation, represents an annual net accumulation average rate of 0.36. The deepest layer was deposited in 1938 (± 3 ). Core mean ionic concentrations are: [(Na+= 66,92 ± 2,32 μg L-1); (K+= 3,31 ± 0,18 μg L-1); (Mg+2= 10,07 ± 0,25 μg L-1); (Ca+2 = 16,93 ± 0,38 μg L-1); (Cl- = 155,74 ± 4,40 μg L-1); (NO3- = 56,01 ± 0,80 μg L-1); (SO4 2 = 55,65 ± 1,36 μg L-1); and (CH3SO3 (MS) = 14,11 ± 1,19 μg L-1)]. The largest concentrations of Na+, Cl-, e Mg+2 were interpreted as winter peaks, directly associated with the aerosols from the surrounding seas, probably, in response to the intensification of marine air mass advection on the ice shelves, and, thus, also being marine tracers. The sulphate profile (series) presents an antiphase, with relation to Na+, Cl- e Mg+2 species variations. Predominantly of marine origin, total sulphates presented greater concentrations during Spring and Summer (periods of greater biologic activity in the Southern Ocean), possibly marking the seasonal variation of biologic activity in the region. Although in some intervals of this ―antiphase‖ are not clearly evident, they are consistent throughout most of the IC-6 core, assisting with the interpretation of the observed seasonal variations, particularly when related to chlorine data series. The sulphate excess profile is similar to total sulphate profile, showing concomitant spikes. Besides the strong correlation to tSO4-2 ion, a weak to moderate correlation was observed for nitrate ions. Coinciding peaks for this ion with excess sulphate may represent episodic events, such as presented by volcanic events of great magnitude. The nitrate concentration variability is not associated to marine aerosols, as shown by the lack of correlation between this anion and Cl-, Na+ e Mg2+. The nitrate series presents many well marked periods and seem to be correlated to excess sulphate concentrations, possibly representing the occurrence of episodic events, such as volcanic eruptions. Even so, the major ions analyses proposed by this work did not make the identification of such episodic events clear. Such events need to be addressed with complementary techniques to determine the specific trace elements. These results suggest that nitrate is transported and deposited by stratospheric or high tropospheric air masses, and that great concentrations of this species could be associated to the recorded volcanic events. This characteristic appears to be coherent with the spikes in the nitrate and sulphate profiles. Besides the seasonal variation (observed, principally, in the chlorine profile), other time cycle/patterns were identified, mainly those related to Na+, Cl- e Mg2+ ion series (of marine origin) and NO3-. The main identified cycle, approximately 17.3 years, ensues to be better investigated. A second cycle, presenting a 10 year period, is possibly associated to the solar cycle (10.7 years). Shorter cycles of 2 and 5 year periods could possibly be related to the ENSO phenomenon. On comparing average concentrations of the IC-6 core with other sites, farther within the Antarctic continent, an abrupt reduction was observed, from the Trans-Antarctic mountains to the Geographic South Pole, suggesting that this mountain range could be a barrier for marine aerosol transport to the interior of the continent, due to an orographic effect on the precipitation.