No description available.

No description available.

No description available.

La technique Storing Matter a pour objectif d’améliorer la sensibilité et la quantification des analyses par SIMS. Elle consiste à découpler la phase de pulvérisation de l’échantillon de la phase d’analyse. Pour cela, la matière pulvérisée par bombardement ionique est déposée en sous-monocouche sur un collecteur optimisé. Le dépôt est ensuite analysé par SIMS. La probabilité d’ionisation de la matière ne dépend plus de son environnement initial (“effet de matrice”), mais de la surface du collecteur. Le choix du collecteur permet un gain de sensibilité et la quantification des concentrations de l’échantillon initial. L’efficacité de la technique dépend du choix du collecteur et d’un facteur de collection ? caractérisant la phase de pulvérisation-dépôt. Dans ce travail, nous avons étudié la pulvérisation et l’émission de la matière sous bombardement ionique pour optimiser ce facteur ?. Nous avons mis au point un dispositif expérimental ainsi qu’un protocole d’analyse par SIMS qui nous a permis d’étudier la distribution angulaire sous un bombardement d’ions Cs+ avec une incidence oblique pour différents paramètres d’impact. L’étude menée sur quatre cibles (Si, Ge, InP et GaAs) a montré que la distribution angulaire est de forme cosn (?-?Max) pour une énergie et un angle d’impact de respectivement 2 à 10 keV et 30° à 60°. L’exposant n est ~2 tandis que la direction d’émission préférentielle ?Max varie de la normale à la surface (0°) jusqu’à un angle d’émission de 35° dans la direction spéculaire au faisceau en fonction de l’énergie d’impact et de l’angle d’incidence. Ces résultats appliqués à Storing Matter ont permis de déterminer la configuration optimum pour une collection maîtrisée en fonction du bombardement / The Storing Matter technique aims at optimising the sensitivity and quantitativeness of SIMS analysis. It consists in decoupling the sputtering of the specimen from the subsequent analysis step. The specimen is sputtered by means of an ion beam. The emitted particles are deposited at a sub-monolayer level on an optimised collector. The deposit is subsequently analysed in a SIMS instrument. The ionisation probability in SIMS does not depend anymore on the initial sample composition (“Matrix effect”), but on the collector surface chemistry. The collector is chosen in order to increase the sensitivity and to quantify the specimen. The efficiency of this new technique depends on the collector choice and on the collection factor ? characterising the sputter-deposition step. In this work, the sputtering and emission processes under ionic bombardment have been studied in order to optimise this factor ?. We developed an experimental set-up and an analysis protocol based on SIMS that allows us to study the angular distribution under Cs+ bombardment with an oblique incidence for different impact parameters. Four targets (Si, Ge, InP and GaAs) were studied. The results show that the angular distribution is shaped as a cosine function cosn (?-?Max) for impact energies between 2 and 10 keV and for incidence angles from 30 to 60°. Under these conditions, the exponent n is ~2 and the preferential direction of emission ?Max varies from the normal to the surface to 35° in the specular direction in function of the impact energy and the incidence angle. The results allowed to find the best settings for the Storing Matter technique to control the sputtered matter collection in function of the bombardment parameters

Das künstliche Radionuklid 137Cs wurde seit über einem halben Jahrhundert in die Natur eingebracht. Sein erstes Erscheinen in Sedimenten der zentraleuropäischen Seen korrespondiert mit den Nuklearwaffentests in den 60er-Jahren des 20. Jahrhunderts. Die stärkste Kontaminierung der europäischen Seen und Flüsse entstand als Folge des radioaktiven Niederschlags nach dem Unfall in Tschernobyl im Frühjahr 1986. In dieser Arbeit wurde das Migrationsverhalten des künstlichen 137Cs im Lago Maggiore und anderen Seen im Alpenvorland als Folge dieses Niederschlags untersucht. Der Lago Maggiore zählt zu den größten Trinkwasserreservoiren südlich der Alpen. Nach dem Unfall in Tschernobyl gingen ungefähr 20 kBq•m-2 des 137Cs auf die Oberfläche des Sees nieder. Im Zeitraum von 2003 bis 2005 wurden Sedimentkerne und Wasserproben an sieben unterschiedlichen Stellen des Lago Maggiore entnommen. Diese Arbeit beschäftigt sich mit Daten über die Verteilung des 137Cs in den Zuflüssen, dem Wasser des Sees, in Schwebstoffen und Bodensedimenten und mit der Assoziierung von 137Cs mit verschiedenen geochemischen Fraktionen. Um den Abfluss des 137Cs aus dem Wassereinzugsgebiet in den Fluss zu modellieren, wurde ein Compartmentmodell verwendet. Zur Modellierung der Aufnahmemenge von 137Cs im Sediment und der vertikalen Verteilung innerhalb dessen wurde ein Diffusions-Konvektions-Modell entwickelt. Dieses Modell berücksichtigt die Aufnahme von Aktivität durch Sedimentation, Fixierung und Rücklösung, retardierte Diffusion, die Verdichtung des Sediments sowie den Einfluss konkurrierender Ionen auf die retardierte Diffusion innerhalb des Sediments. Die Ergebnisse der Parameteroptimierung – im Wesentlichen die Sedimentationsrate und der 137Cs-Verteilungskoeffizient Kd, welcher die Aufnahme der Aktivität in das Sediment determiniert – werden erörtert und mit denen anderer europäischer Seen, die eine ähnliche 137Cs-Deposition, aber unterschiedliche limnologische Eigenschaften aufweisen, verglichen. Zur Beurteilung der Bioverfügbarkeit von 137Cs wurde die 137Cs-Aktivitätskonzentrationen von Fischproben aus dem Lago Maggiore gemessen. Aus bereits existierenden Daten und unseren Messergebnissen wurden 137Cs Fisch-Wasser Konzentrations-Verhältnisse errechnet und mit denen für andere Seen, die von ähnlichen 137Cs-Kontaminationen betroffen sind, verglichen. / Artificial Cs-137 has been introduced into the environment for more than half a century. Its first appearance in central European lake sediments corresponds to the nuclear weapons testing in the 1960s. The largest contamination of European lakes and rivers occurred as a consequence of the fallout after the Chernobyl accident in spring 1986. In this work the migration behaviour of artificial Cs-137 in Lago Maggiore and other pre-alpine lakes as a consequence of these fallouts was studied. Lago Maggiore is one of the largest drinking water reservoirs in the south of the Alps. After the Chernobyl accident roughly 20 kBq•m-2 of 137Cs were deposited onto the lake surface. From 2003 to 2005 bottom sediment cores and water samples were collected at 7 different locations of Lago Maggiore. Data on the 137Cs distribution in tributaries, lake water, suspended matter, bottom sediments, and the 137Cs association to different geochemical fractions are presented in this work. To model the run-off of 137Cs from the watershed into the lake a compartment model was used. For modeling the input of 137Cs into and the vertical distribution within the sediment a diffusion–convection type model was developed. This model takes into account the uptake of activity by sedimentation, fixation and redissolution, retarded diffusion, the influence of competing ions on the retarded diffusion within the sediments, and compaction of sediments. The results of the parameter optimization – mainly the sedimentation rate and the 137Cs distribution coefficient Kd, which determines the uptake of activity into the sediment – are discussed and compared with those of other European lakes characterized by similar 137Cs deposition levels but different limnological properties. To estimate the bioavailability of 137Cs, its activity concentrations in fish samples from Lago Maggiore were measured. Combining the existing data with our measurements, 137Cs fish–water concentration ratios were calculated and compared with those for other lakes which were affected by similar 137Cs contamination.

Submitted by ANA CAROLINE SONCIN DA SILVA null (col_zinha@hotmail.com) on 2016-05-21T03:08:22Z No. of bitstreams: 1 Dissertação_Mestrado.pdf: 2242269 bytes, checksum: d87570a91a3ceb91484c2ca2920007e1 (MD5) / Approved for entry into archive by Felipe Augusto Arakaki (arakaki@reitoria.unesp.br) on 2016-05-24T13:46:03Z (GMT) No. of bitstreams: 1 silva_acs_me_ilha.pdf: 2242269 bytes, checksum: d87570a91a3ceb91484c2ca2920007e1 (MD5) / Made available in DSpace on 2016-05-24T13:46:03Z (GMT). No. of bitstreams: 1 silva_acs_me_ilha.pdf: 2242269 bytes, checksum: d87570a91a3ceb91484c2ca2920007e1 (MD5) Previous issue date: 2016-03-29 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / O mexilhão dourado (Limnoperna fortunei) tem provocado transtornos ambientais e econômicos nas localidades onde foi introduzido. Dessa forma, o uso do seu resíduo moído como reparador da defasagem de nutrientes e de acidez em solos agrícolas pode vir a ser uma boa aplicação para esse molusco bioinvasor. Entretanto, apesar dessa ação reparadora, é necessária uma averiguação prévia da ausência de elementos potencialmente tóxicos no resíduo. Neste contexto, o presente trabalho foi realizado com o intuito de quantificar, em amostras do Limnoperna fortunei, os elementos chumbo, cádmio, alumínio, ferro e manganês, os quais dependo de suas quantidades podem ser tóxicos aos vegetais e aos demais seres vivos ao longo da cadeia alimentar. Para quantificação desses analitos foi utilizada a Espectrometria de absorção atômica com fonte contínua e de alta resolução. Foram desenvolvidos e otimizados programas de aquecimento para a determinação dos elementos traço chumbo e cádmio, empregando amostragem direta de sólidos e atomização com forno de grafite. A determinação direta de chumbo só foi possível com a utilização da calibração com o material de referência certificado para tecido de mexilhão (Mussel Tissue – 2976), empregando 5 µL da mistura 0,1% (m/v) Pd (NO3)2 + 0,05% (m/v) Mg (NO3)2 em presença de 0,05% (m/v) de Triton X-100 como modificador químico. As temperaturas de pirólise e atomização foram 1000°C e 2200°C, respectivamente. O estudo de massa permitiu a utilização de massas de amostras dentro da faixa de 0,150 a 0,850 mg. Para o cádmio só foi realizada calibração com solução padrão 12,5 µg/L, também com o uso de 5 µL do modificador químico Pd+Mg. Assim, foram alcançadas temperaturas de pirólise de 900 °C e de 1600 °C para a atomização. O estudo de massa precisou ser desenvolvido com uma das amostras coletadas e demonstrou que puderam ser usadas, para as quantificações, massas próximas a 0,250 mg. Os limites de detecção e quantificação foram, nessa ordem, 0,009 mg/kg e 0,03 mg/kg para o chumbo, e 0,0004 e 0,001 mg/kg para o cádmio. No caso dos elementos alumínio, ferro e manganês, as quantificações foram realizadas utilizando a atomização com chama. Os limites de detecção e quantificação foram, respectivamente, 0,15 mg/kg e 0,50 mg/kg para o alumínio, 0,050 mg/kg e 0,15 mg/kg para o ferro, e 0,00067 mg/kg e 0,0023 mg/kg para o manganês. Os resultados para todos os elementos quantificados nesse estudo apresentaram concordância de 95% com o test-t (Student) não pareado, o que valida e viabiliza os métodos desenvolvidos para as quantificações de chumbo e cádmio, bem como comprova a exatidão dos demais resultados obtidos. Portanto, as técnicas analíticas selecionadas foram eficientes para as quantificações dos elementos desejados no resíduo moído desse mexilhão dourado. / The golden mussel (Limnoperna fortunei) has caused environmental and economic disruption in the locations where it was introduced. Thus, the use of their ground waste as repairer of the gap of nutrients and acidity in soils may prove to be a good application for this mollusk bio invader. However, despite of this remedial action, a prior test of the absence of potentially toxic elements in the residue is required. In this context, the present study was performed in order to quantify, in samples of the Limnoperna fortunei, the elements lead, cadmium, aluminum, iron and manganese, which depending on their quantities they can be toxic to plants and other living beings throughout the food chain. To quantify these analytes was used high-resolution continuum source atomic absorption spectrometry. Heating programs were developed and optimized for the determination of lead and cadmium trace elements, using direct solid and graphite furnace atomization. The direct determination of lead was only possible with the use of calibration with certified reference material for mussel tissue (Mussel Tissue - 2976), using 5 uL of the mixture 0.1% (m/v) Pd (NO3)2 + 0.05% (w/v) Mg (NO3)2 in the presence of 0.05% (w/v) Triton X-100 as a chemical modifier. The temperatures of pyrolysis and atomization were 1000 ° C and 2200 ° C respectively. The mass study allowed the use of samples of mass within the range of 0.150 to 0.850 mg. For cadmium it was only performed calibration with standard 12.5 g/L solution, also using 5 uL chemical modifier Pd + Mg. Thus, pyrolysis temperatures were reached 900 ° C and 1600 ° C for atomization. The mass study needed to be developed with one of the samples collected and it showed that they could be used for the measurements, masses close to 0,250 mg. The limits of detection and quantification were, in this order, 0.009 mg/kg and 0.03 mg/kg to lead, and 0.0004 and 0.001 mg/kg to cadmium. In case of the elements aluminum, iron and manganese, the measurements were carried out using atomization with a flame. The limits of detection and quantification were, respectively, 0.15 mg/kg and 0.50 mg/kg to aluminum, 0.050 mg/kg and 0.15 mg/kg for iron, and 0.00067 mg/kg and 0.0023 mg/kg to manganese. The results for all quantified elements of this study showed agreement of 95% with the t-test (Student) unpaired, which validates and enables the methods developed for lead and cadmium quantifications and proves the accuracy of the other results. Therefore, the selected analytical techniques were efficient to the quantifications of the desired elements in the ground waste this golden mussel.

Orientador: Mirian Cristina dos Santos / Resumo: O mexilhão dourado (Limnoperna fortunei) tem provocado transtornos ambientais e econômicos nas localidades onde foi introduzido. Dessa forma, o uso do seu resíduo moído como reparador da defasagem de nutrientes e de acidez em solos agrícolas pode vir a ser uma boa aplicação para esse molusco bioinvasor. Entretanto, apesar dessa ação reparadora, é necessária uma averiguação prévia da ausência de elementos potencialmente tóxicos no resíduo. Neste contexto, o presente trabalho foi realizado com o intuito de quantificar, em amostras do Limnoperna fortunei, os elementos chumbo, cádmio, alumínio, ferro e manganês, os quais dependo de suas quantidades podem ser tóxicos aos vegetais e aos demais seres vivos ao longo da cadeia alimentar. Para quantificação desses analitos foi utilizada a Espectrometria de absorção atômica com fonte contínua e de alta resolução. Foram desenvolvidos e otimizados programas de aquecimento para a determinação dos elementos traço chumbo e cádmio, empregando amostragem direta de sólidos e atomização com forno de grafite. A determinação direta de chumbo só foi possível com a utilização da calibração com o material de referência certificado para tecido de mexilhão (Mussel Tissue – 2976), empregando 5 µL da mistura 0,1% (m/v) Pd (NO3)2 + 0,05% (m/v) Mg (NO3)2 em presença de 0,05% (m/v) de Triton X-100 como modificador químico. As temperaturas de pirólise e atomização foram 1000°C e 2200°C, respectivamente. O estudo de massa permitiu a utilização de massas de am... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The golden mussel (Limnoperna fortunei) has caused environmental and economic disruption in the locations where it was introduced. Thus, the use of their ground waste as repairer of the gap of nutrients and acidity in soils may prove to be a good application for this mollusk bio invader. However, despite of this remedial action, a prior test of the absence of potentially toxic elements in the residue is required. In this context, the present study was performed in order to quantify, in samples of the Limnoperna fortunei, the elements lead, cadmium, aluminum, iron and manganese, which depending on their quantities they can be toxic to plants and other living beings throughout the food chain. To quantify these analytes was used high-resolution continuum source atomic absorption spectrometry. Heating programs were developed and optimized for the determination of lead and cadmium trace elements, using direct solid and graphite furnace atomization. The direct determination of lead was only possible with the use of calibration with certified reference material for mussel tissue (Mussel Tissue - 2976), using 5 uL of the mixture 0.1% (m/v) Pd (NO3)2 + 0.05% (w/v) Mg (NO3)2 in the presence of 0.05% (w/v) Triton X-100 as a chemical modifier. The temperatures of pyrolysis and atomization were 1000 ° C and 2200 ° C respectively. The mass study allowed the use of samples of mass within the range of 0.150 to 0.850 mg. For cadmium it was only performed calibration with standard 12.5 g/L sol... (Complete abstract click electronic access below) / Mestre

Suite à la catastrophe nucléaire de Fukushima Dai-Ichi, des milliers de tonnes d’eau douce et d’eau de mer ont été utilisées pour le refroidissement des réacteurs ou contaminées du fait des infiltrations souterraines. La décontamination de ces eaux est rendue difficile par la présence d’autres cations (Na+, K+, Ca2+, Mg2+) présents naturellement dans ces eaux. Un procédé de décontamination en colonne garnie de deux types adsorbants minéraux, le TERMOXID 35 et le SORBMATECH® 202, a été étudié dans ce contexte. Le premier est un adsorbant commercial constitué du ferrocyanure mixte K/Ni imprégnés sur une matrice solide Zr(OH)4. Le second, synthétisé au CEA, est composé de ferrocyanure K/Cu imprégnés sur une matrice solide SiO2. Ces deux matériaux se sont révélés extrêmement efficaces pour décontaminer le Cs dans l’eau de mer avec des Kd,Cs de l’ordre de 105 mL/g.Les études menées en batch dans différentes solutions (eau pure, eau douce et eau de mer) ont permis de mettre en évidence les cinétiques de sorption ainsi que les mécanismes d’échange d’ions responsables de la sorption du Cs+ en tenant compte des effets compétitifs des cations des eaux naturelles. La modélisation des batchs a été menée avec le code géochimique CHESS en prenant en compte ces effets compétitifs selon le formalisme de Vanselow et les coefficients de sélectivité en constituant une base de données thermodynamiques spécifique. Les performances de ces matériaux ont ensuite été testées en colonne. Les paramètres opératoires tels que la vitesse de Darcy et le ratio H/D ont été étudiés pour définir les conditions de bon fonctionnement de ce procédé. Le T35 s’est révélé être moins performant du fait notamment de la diffusion lente du Cs dans les pores de ce matériau. Le S202 s’est révélé être un bon candidat pour l’application de forts débits de traitement. Les courbes de percée ainsi obtenues dans l’eau douce ont par ailleurs fait l’objet de modélisation avec les codes de transport réactif HYTEC et OPTIPUR couplés à la base CHESS de données thermodynamiques. Cette démarche permettra de mieux dimensionner les unités de décontamination définies par l’exploitant. / Following the nuclear disaster in Fukushima Dai-Ichi, thousands of tons of fresh water and seawater were used for cooling the reactors or contaminated as a result of groundwater seepage. Decontamination of these waters is complicated by the presence of other cations (Na+, K+, Ca2+, Mg2+) naturally present in these waters. Decontamination process in columns packed was studied in this context with two types of mineral adsorbents: the TERMOXID 35 and the SORBMATECH® 202. The first one is a commercial adsorbent and consists of mixed ferrocyanide K/Ni impregnated over a solid matrix Zr(OH)4. The second one was synthesized in CEA and is composed of ferrocyanide K/Cu impregnated over a solid matrix SiO2. Both materials have shown a high efficiency for Cs decontamination in seawater with Kd,Cs of about 105 mL/g.Batch studies conducted in different solutions (pure water, fresh water and seawater) allowed determining sorption kinetics and ion exchange mechanisms responsible for the sorption of Cs+, taking into account competitive effects of the natural water cations (Na+, K+, Ca2+, Mg2+). Modelling of batchs was performed with the geochemical code CHESS considering competitive effects according to the Vanselow formalism and selectivity coefficients, developing a specific thermodynamic database. The performances of these materials were then tested in column. The operating parameters such as Darcy’s velocity and the H/D ratio were studied for a proper functioning of this process. The T35 has proven to be less efficient mainly because of the slow diffusion of Cs in the pores of the material. The S202 has proven to be a good candidate for the application of high flow rates. The breakthrough curves obtained in fresh water have been modelled with the reactive transport codes HYTEC and OPTIPUR using the CHESS thermodynamic database. This approach will eventually help to support the design of a decontamination unit by the operator.

Diese Dissertation umfasst drei voneinander unabhängige Artikel. In diesen werden neue Prozess-modelle vorgestellt, die von der entscheidungspsychologischen Forschung inspiriert wurden. Im ersten Artikel werden Entscheidungsprozesse mit mehreren Entscheidungsmerkmalen als gesteuerte Suchprozesse modelliert. Zunächst wird ein theoretischer Rahmen vorgestellt, in dem ökonomische Modelle Entscheidungen mit Suche mit Modellen des subjektiven Nutzens aus dem Bereich der psychologischen Forschung integriert wird. In den so modellierten Entscheidungsprozessen wird angenommen, dass Individuen ihre Entscheidungsalternativen nach deren abnehmenden Nutzen ordnen und dann so lange durchsuchen, bis die erwarteten Suchkosten höher als die entsprechenden Gewinne sind. Anschliessend wird die Güte dreier Entscheidungsmodelle an zwölf realen Datensätzen überprüft. Im zweiten Artikel werden die Ergebnisse zweier Experimente vorgestellt, in denen untersucht wurde, wie Personen ihre Urteile verändern, wenn sie den Urteilen und dem der Konfidenzniveau anderer Personen ausgesetzt sind. Ein Baummodell wird eingeführt, welches abbildet, wie Urteile aufgrund solcher Informationen revidiert werden. Dieses Modell basiert auf den Ergebnissen der beiden Experimente: Indem soziale Informationen berücksichtigt werden, kann es zeigen, wie Urteile in einer Gruppe interagierender Personen zusammenlaufen oder polarisieren. Im dritten Artikel wird kollektives Verhalten in Märkten für kulturelle Produkte untersucht. Personen ordnen die Optionen entsprechend ihrer Popularität an und entscheiden sich dann für diejenige, die einen Nutzen hat, der über einer bestimmten ausreichend guten Schwelle liegt. Nach jeder individuellen Entscheidung wird die Rangfolge revidiert. Innerhalb dieses einfachen Rahmens wird demonstriert, dass solche Märkte durch eine sogenannte rich get richer-Dynamik charakterisiert sind. Diese führt zu Ungleichheiten in den Marktanteilen und ungewissen finanziellen Erlösen. / This dissertation comprises of three independent essays which introduce novel psychologically inspired process models and examine their implications for individual, collective or market behavior. The first essay studies multi-attribute choice as a guided process of search. It puts forward a theoretical framework which integrates work on search and stopping with partial information from economics with psychological subjective utility models from the field of judgment and decision making. The alternatives are searched in order of decreasing estimated utility, until the expected cost of search exceeds the relevant benefits; The essay presents the results of a performance comparison of three well-studied multi-attribute choice models.The second essay reports the results of two experiments designed to understand how people revise their judgments of factual questions after being exposed to the opinion and confidence levels of others. It introduces a tree model of judgment revision which is directly derived from the empirical observations. The model demonstrates how opinions in a group of interacting people can converge or polarize over repeated interactions. The third essay, studies collective behavior in markets for search products. The decision makers consider the alternatives in order of decreasing popularity and choose the first alternative with utility higher than a certain satisficing threshold. The popularity order is updated after each individual choice. The presented framework illustrates that such markets are characterized by rich-get-richer dynamics which lead to inequality in the market-share distribution and unpredictability in regard to the final outcome.

La mémoire de travail est un système cognitif essentiel à notre vie quotidienne. Elle nous permet de conserver momentanément des informations dans le but de réaliser une tâche cognitive. Une des caractéristiques principales de ce type de mémoire est d’être limitée en capacité. Les raisons de cette limitation sont largement débattues dans la littérature. Certains modèles considèrent qu'une cause principale de l'oubli en mémoire de travail est l'existence d'un déclin temporel passif de l'activation des représentations mnésiques alors que d'autres modèles supposent que les interférences entre les informations suffisent à expliquer la capacité limitée de cette mémoire. Deux modèles computationnels ont été proposés récemment (TBRS* et SOB-CS) et illustrent parfaitement ce débat. En effet, ils décrivent de manière très différente ce qui se passe au cours d’une tâche de mémoire de travail impliquant à la fois la mémorisation et le traitement d’informations. En plus de s'opposer sur les causes de l’oubli, ils proposent des processus de maintien en mémoire de travail distincts : le rafraîchissement des informations pertinentes selon TBRS* versus la suppression des informations non pertinentes selon SOB-CS. Les travaux de cette thèse se sont organisés autour de deux objectifs principaux. Premièrement, cette thèse a porté sur l’étude de ces deux modèles et leur mécanisme de maintien. Pour cela, nous avons réalisé des expériences comportementales utilisant la tâche d’empan complexe afin de tester des hypothèses précises de ces modèles. Deuxièmement, nous avons étudié, à l'aide des modèles computationnels, les causes des déficits de mémoire de travail observés chez les personnes âgées, dans le but, à long terme, de créer ou d'améliorer les outils de remédiation. Concernant le premier objectif, les différents résultats d’études ont montré une discordance entre le comportement humain et les simulations. En effet, TBRS* et SOB-CS ne permettent pas de reproduire un effet positif du nombre de distracteurs contrairement à ce qui a été observé expérimentalement. Nous proposons que cet effet positif, non prédit par les modèles, est relié à la mémorisation à long terme non prise en compte dans ces deux modèles. Concernant le deuxième objectif, les résultats comportementaux suggèrent que les personnes âgées auraient principalement des difficultés à rafraîchir les traces mnésiques et à stabiliser les informations à long terme au cours d’une tâche d’empan complexe. Dans l’ensemble, les résultats de cette thèse suggèrent d'approfondir les recherches concernant les liens entre les mécanismes de maintien en mémoire de travail et la mémorisation à long terme, par exemple en proposant un nouveau modèle computationnel permettant de rendre compte de nos résultats. Au-delà des avancées concernant la compréhension du fonctionnement de la mémoire de travail, cette thèse montre également que l’utilisation de modèles computationnels revêt un caractère particulièrement pertinent pour l'étude d'une théorie ainsi que pour la comparaison de différentes populations. / Working memory is a cognitive system essential to our daily life. It allows us to temporarily store information in order to perform a cognitive task. One of the main features of this type of memory is to be limited in capacity. The reasons for this limitation are widely debated in the literature. Some models consider that a main cause of forgetting in working memory is the existence of a passive temporal decay in the activation of memory representations whereas other models assume that interference between information are sufficient to explain the limited capacity of this memory. Two computational models have recently been proposed (TBRS* and SOB-CS) and they perfectly illustrate this debate. Indeed, they both describe differently what happens during a working memory task involving both storage and information processing. In addition to opposing the causes of forgetting, they propose separate maintenance processes: refreshing relevant information according to TBRS* versus removing irrelevant information according to SOB-CS. This thesis was organized around two main objectives. First, we focused on the study of these two models and their maintenance mechanisms. To do so, we performed behavioral experiments using the complex span task to test specific hypotheses of these models. Second, using computational models, we investigated the causes of working memory deficits observed in the elderly, with the aim, in the long term, of creating or improving remediation tools. Regarding the first objective, results showed a discrepancy between human behavior and simulations. Indeed, TBRS* and SOB-CS did not reproduce a positive effect of the number of distractors contrary to what has been observed experimentally. We propose that this positive effect, not predicted by the models, is related to the long-term storage not taken into account in these two models. Regarding the second objective, the behavioral results suggest that older people would have difficulty mainly in refreshing memory traces and in stabilizing information in the long term during a complex task. Overall, the results of this thesis suggest to deepen the research on the links between the maintenance mechanisms and the long-term storage, for example by proposing a new computational model accounting for our results. Beyond advances in understanding the functioning of working memory, this thesis also shows that the use of computational models is of particular relevance for the study of a theory as well as for the comparison of different populations.