The tandem solar cell concept is a promising approach to improve the efficiency of photovoltaic devices. However, characterization of tandem solar cell devices is challenging since correct efficiency determination demands special experimental infrastructure as well as suitable characterization procedures. Even though the appropriate IEC and ASTM measurement standards define all that very precisely, they cannot be applied without special care to organic photovoltaics (OPV) because they were originally developed for inorganic devices. As a consequence, nowadays almost all tandem organic solar cell publications are not using correct characterization procedures, often resulting in questionable efficiency values. The aim of this work is developing a measurement procedure for tandem organic solar cells assuring their correct characterization. Therefore, at first the existing standards and measurement procedures for tandem solar cells are reviewed and challenges when applying these standards to organic solar cells are identified. As main challenges the relatively low fill factors and distinct nonlinearities of organic solar cells are identified. As preliminary experiments, single junction organic solar cells are investigated to analyze the influence of measurement parameters like bias irradiance, bias voltage, and chopper frequency on the external quantum efficiency (EQE) of organic solar cells. This results in parameter sets assuring minimized artifacts for the subsequent EQE determination of the subcells of tandem organic solar cells. The main part of this thesis presents the detailed characterization of a tandem OPV example device. First, EQE is measured and validated by two independent institutes. The EQE results are used to calculate the illumination conditions to reach AM1.5g conditions for both subcells with a multi-source sun simulator. The resulting efficiency value under standard reporting conditions (SRC) is found to be 5% lower than the efficiency measured with a single-source sun simulator. A full spectrometric characterization shows that differing fill factors of the subcells are the reason for this behavior. To overcome the main reason for the complicated measurement procedure of tandem solar cells, the inaccessibility of the individual subcells, three different approaches for the jV-characteristics determination of the subcells are presented. The so-called Bias Voltage Approach is based on EQE-measurements under varying bias voltage and needs no additional electrical contacts. Therefore, it can be applied to existing devices. The Voltage Contact Approach as well as the Current Contact Approach require in changed stack designs. Therefore, they cannot be applied to existing devices but give more accurate results. Finally, a procedure for characterizing tandem organic solar cells is formulated. This procedures aims at giving practical advice how to characterize tandem organic solar cells to achieve results conforming to the measurement standards and being as accurate and reproducible as possible. Hence, this thesis attempts to establish standards for a correct measurement of tandem organic solar cells of which other emerging solar cell technologies can profit as well.

Nous possédons la faculté de reconnaître individuellement des centaines d’individus. Ceci nous permet d’évoluer dans une société complexe dont l’organisation est en partie forgée par les relations interindividuelles. La reconnaissance individuelle peut être réalisée par l'identification de divers éléments distincts, comme le visage ou la voix, qui forment chez l’Homme une seule représentation conceptuelle de l'identité de la personne. Nous avons démontré que les singes rhésus, comme les humains, reconnaissent individuellement leurs congénères familiers, mais également les individus humains connus. Ceci montre que la reconnaissance fine est une compétence partagée par un éventail d'espèces de primates pouvant servir de fondement à la vie en réseaux sociaux sophistiqués, et également que le cerveau s’adapte de façon flexible pour reconnaître les individus d'autres espèces lorsque ceux-ci ont une importance socioécologique. Par la suite, au niveau neuronal, ce projet a mis en lumière que les connaissances sociales concernant autrui sont représentées par les neurones hippocampiques ainsi que par les neurones inférotemporaux. Ainsi nous avons observé l’existence de neurones sélectifs aux visages non seulement dans le cortex inferotemporal, comme ceci a été décrit précédemment, mais également dans l’hippocampe. La comparaison des propriétés de ces neurones au sein de ces deux structures, suggère que les deux régions joueraient des rôles complémentaires au cours de la reconnaissance individuelle. Enfin, parce que l'hippocampe est une structure qui a évolué à des degrés divers chez différents mammifères pour soutenir la mémoire autobiographique et les représentations spatiales, la caractérisation des différents types de neurones et de leur connectivité a fourni un cadre commun pour comparer les fonctions de l’hippocampe à travers les espèces / Humans can individually recognize some hundreds of persons and therefore operate within a rich and complex society. Individual recognition can be achieved by identifying distinct elements such as the face or voice as belonging to one individual. In humans, those different cues are linked into one conceptual representation of individual identity. I demonstrated that rhesus monkeys, like humans, recognize familiarpeers but also familiar humans individually and that they match their voice to their corresponding memorized face. Thus it shows that fine individual recognition is a skill shared across a range of primate species, which may serve as the basis of a sophisticated social network. It also suggests that animals’ brains flexibly adapt to recognize individuals of other species when socio-ecologically relevant. Following at the neuronal level, this project put in light that social knowledge about other individuals is represented by hippocampal neurons as well as by inferotemporal neurons. For instance I observed the existence of face preferring neurons not only in the inferotemporal cortex as previously described but also in the hippocampus. Comparison of their properties across both structures, suggests that they could play complementary roles in recognition of individuals. Finally, because the hippocampus is a structure that evolved in various degrees to support autobiographical memory and spatial information in different mammals, I characterized the different subtypes of neurons and their network connectivity in the monkey hippocampus to provide a common anatomical framework to discuss hippocampal functions across species

Les travaux présentés dans ce manuscrit de thèse concernent le développement d’une méthode de caractérisation électrique de cellules biologiques, sans marquage, basée sur la diélectrophorèse Ultra Haute Fréquence (DEP-UHF). Sous l’action d’un champ électrique alternatif non uniforme, les cellules biologiques sont soumises à des forces de déplacement essentiellement liées à leurs propriétés diélectriques. En particulier, aux hautes fréquences, le champ électrique pénètre à l’intérieur de la cellule et interagit donc avec son contenu intracellulaire. Il est donc possible d’accéder à une «signature diélectrophorétique» de la cellule représentative de ses propriétés biologiques internes mais aussi de mécanismes physiologiques tels que l’apoptose ou encore la différenciation. Ce manuscrit présente le développement d’un microsystème innovant, implémenté à partir des couches passives d’une puce BiCMOS et couplé à un réseau microfluidique, pour la caractérisation, à l’échelle cellulaire, par DEP-UHF. Le microsystème développé permet une analyse fine et précise du comportement DEP haute fréquence d’une cellule. Un banc expérimental dédié aux caractérisations cellulaires, capable de générer des signaux hautes fréquences dans la gamme 10 MHz – 1 GHz pour des amplitudes allant jusqu’à 18 Vpp, a été développé. Ces travaux exploratoires ont pour but de démontrer le potentiel de discrimination de cette méthode entre différentes lignées cellulaires cancéreuses humaines à des stades tumoraux différents, dans l’objectif de développer de nouveaux outils d’aide au diagnostic. L’existence de différences significatives entre les signatures de certains types cellulaires ouvre des perspectives très intéressantes notamment pour le développement d’outils de tri cellulaire originaux basés uniquement sur les propriétés diélectriques intracellulaires. / The work presented in this dissertation concerns the development of an original label-free electrical characterization method dedicated to biological cells based on Ultra High Frequency dielectrophoresis (DEP-UHF). Under the action of a non-uniform alternative electric field, the biological cells are subjected to displacement forces related to their own dielectric properties. In particular, at high frequencies, the electric field penetrates inside the cell and thus interacts with its intracellular content. Therefore, it is possible to access to a «dielectrophoretic signature» of the cell that it is representative of its internal biological properties but also of physiological mechanisms such as apoptosis or differentiation. This dissertation presents the development of an innovative microsystem, implemented in the passive layer stack of a BiCMOS chip and associated with microfluidic, dedicated to biological characterization, at the cellular level. The developed microsystem allows an accurate analysis of a single cell DEP-UHF behaviour. An experimental bench, dedicated to cell characterization, and able to generate high frequency signals from 10 MHz to 1 GHz up to 18 Vpp magnitude, has been also developed accordingly. Actually, the led exploratory work achieved was focused on evaluating the discrimination potential of this method between different human cancer cells at different tumor stages with the objective to envision new kind of diagnostic tools. Finally, the existence of significant differences between the signatures of different cell types leads to very interesting perspectives, particularly for the development of new cell sorting tools based especially on the intracellular dielectric properties.