Using photoelectron spectroscopy, we show measurements of energy level alignment of organic semiconducting layers. The main focus is on the properties and the influence of doped layers. The investigations on the p-doping process in organic semiconductors show typical charge carrier concentrations up to 2*10E20 cm-3. By a variation of the doping concentration, an over proportional influence on the position of the Fermi energy is observed. Comparing the number of charge carriers with the amount of dopants present in the layer, it is found that only 5% of the dopants undergo a full charge transfer. Furthermore, a detailed investigation of the density of states beyond the HOMO onset reveals that an exponentially decaying density of states reaches further into the band gap than commonly assumed. For an increasing amount of doping, the Fermi energy gets pinned on these states which suggests that a significant amount of charge carriers is present there. The investigation of metal top and bottom contacts aims at understanding the asymmetric current-voltage characteristics found for some symmetrically built device stacks. It can be shown that a reaction between the atoms from the top contact with the molecules of the layer leads to a change in energy level alignment that produces a 1.16eV lower electron injection barrier from the top. Further detailed investigations on such contacts show that the formation of a silver top contact is dominated by diffusion processes, leading to a broadened interface. However, upon insertion of a thin aluminum interlayer this diffusion can be stopped and an abrupt interface is achieved. Furthermore, in the case of a thick silver top contact, a monolayer of molecules is found to float on top of the metal layer, almost independent on the metal layer thickness. Finally, several device stacks are investigated, regarding interface dipoles, formation of depletion regions, energy alignment in mixed layers, and the influence of the built-in voltage. We show schematic energy level alignments of pn junctions, pin homojunctions, more complex pin heterojunctions with Zener-diode characteristics, as well as a complete OLED stack. The results allow a deeper insight in the working principle of such devices. / Mit Hilfe der Photoelektronenspektroskopie werden in der vorliegenden Arbeit Energieniveaus an Grenzflächen von organischen Halbleitern untersucht, wobei ein Hauptaugenmerk auf dem Einfluss und den Eigenschaften dotierter Schichten liegt. Bei der Untersuchung grundlegender Eigenschaften eines p-dotierten organischen Halbleiters können Ladungsträgerkonzentrationen bis zu 2*10E20 cm-3 nachgewiesen werden. Eine Variation der Dotierkonzentration zeigt einen überproportionalen Einfluss der Ladungsträger auf die Position des Ferminiveaus verglichen mit Experimenten an anorganischen Schichten. Durch den Vergleich mit der Anzahl Dotanden in der Schicht kann gezeigt werden, dass dabei nur etwa 5% der Dotanden einen vollständigen Ladungstransfer eingehen. Eine detaillierte Untersuchungen der Zustandsdichte jenseits des HOMOs (Highest Occupied Molecular Orbital) zeigt, dass die exponentiell abfallende Flanke der Zustandsdichte weiter in die Bandlücke hineinreicht als üblicherweise angenommen. Das Ferminiveau erfährt bei steigender Dotierung ein Pinning an diesen Zuständen, was für eine signifikante Ladungsträgerkonzentration spricht. Weiterhin wurden Untersuchungen zu Metal Top- und Grundkontakten durchgeführt. Es kann gezeigt werden, dass die Ursache für die Entstehung unsymmetrischer Strom-Spannungskurven, trotz eines symmetrischen Probenaufbaus, an einer Reaktion zwischen dem Molekül und den Metallatomen liegt. Dadurch entsteht eine um 1.16eV reduzierte Injektionsbarriere für Elektronen am Topkontakt. Weitere detaillierte Untersuchungen an diesen Topkontakten zeigen, dass im Falle von Silber als Metall diese Grenzfläche von Diffusionsprozessen dominiert ist. Im Gegensatz dazu zeigt das unedle Metall Aluminium keine Diffusion und führt zu abrupten Grenzflächen. Im ersten Fall kann zudem eine Monolage vom Molekül auf dem Metallkontakt nachgewiesen werden, die unabhängig von der Metalldicke aufschwimmt. Zuletzt werden Bauelemente oder Teile solcher mit Photoelektronenspektroskopie vermessen. Hierbei werden die Grenzflächendipole, die Ausbildung von Verarmungszonen, die Energieangleichung in Mischschichten und der Einfluss der Eingebauten Spannung untersucht. Es können die Banddiagramme von pn-Übergängen, einfachen pin Homoübergängen, komplexeren pin Heteroübergänge mit Zener-Dioden Verhalten sowie eine gesamte OLED gezeigt werden. Die Ergebnisse erlauben einen tieferen Einblick in die Arbeitsweise solcher Bauelemente.

The objective of the proposed research is to establish the technology for material growth by molecular beam epitaxy (MBE) and fabrication of indium gallium nitride/gallium nitride (InxGa1-xN/GaN) heterojunction solar cells. InxGa1-xN solar cell have the potential to span 90% of the solar spectrum, however there has been no success with high indium (In) incorporation and only limited success with low In incorporation InxGa1-xN. Therefore, this present work focuses on 15 - 30% In incorporation leading to a bandgap value of 2.3 - 2.8 eV. This work will exploit the revision of the indium nitride (InN) bandgap value of 0.68 eV, which expands the range of the optical emission of nitride-based devices from ultraviolet to near infrared regions, by developing transparent InxGa1-xN solar cells outside the visible spectrum. Photovoltaic devices with a bandgap greater than 2.0 eV are attractive because over half the available power in the solar spectrum is above the photon energy of 2.0 eV. The ability of InxGa1-xN materials to optimally span the solar spectrum offers a tantalizing solution for high-efficiency photovoltaics. Using the metal modulated epitaxy (MME) technique in a new, ultra-clean refurbished MBE system, an innovative growth regime is established where In and Ga phase separation is diminished by increasing the growth rate for InxGa1-xN. The MME technique modulates the metal shutters with a fixed duty cycle while maintaining a constant nitrogen flux and proves effective for improving crystal quality and p-type doping. We demonstrate the ability to repeatedly grow high hole concentration Mg-doped GaN films using the MME technique. The highest hole concentration obtained is equal to 4.26 e19 cm-3, resistivity of 0.5 Ω-cm, and mobility of 0.28 cm2/V-s. We have achieved hole concentrations significantly higher than recorded in the literature, proving that our growth parameters and the MME technique is feasible, repeatable, and beneficial. The high hole concentration p-GaN is used as the emitter in our InxGa1-xN solar cell devices.

Thesis (M.S.) -- Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Made available in DSpace on 2014-06-11T19:23:28Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-05-28Bitstream added on 2014-06-13T19:29:32Z : No. of bitstreams: 1 yonamine_ah_me_bauru.pdf: 4381396 bytes, checksum: 41612d38a875d111ee395a774fd55178 (MD5) / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / Após a descoberta dos novos supercondutores óxidos no ano de 1986, houve uma intensa busca por outros materiais que apresentassem o fenômeno da supercondutividade. Dentre as novas descobertas encontra-se o diboreto de magnésio, material intermetálico cuja supercondutividade era desconhecida até 2001. Atualmente, este é considerado ter um grande potencial de aplicação real devido a sua simplicidade de confecção e ao baixo custo de resfriamento em cryocoolers, visto que o fenômeno ocorre próximo de 39K. Ao longo de quase uma década muitas pesquisas foram realizadas com os objetivos de melhor conhecer o mecanismo de supercondutividade presente no material, assim como de aperfeiçoar a sua capacidade de suportar correntes em campos magnéticos altos e desta forma ampliar suas possibilidades de aplicação industrial. Hoje, encontra-se claro que o mecanismo que rege a supercondutividade no Mg'B IND. 2' é o mesmo descrito em 1958 por Bardeen, Cooper e Schriefer, na teoria BCS. Mas, por outro lado, as questões práticas permanecem incompletas, pois a queda da capacidade de conduzir supercorrente sob campos magnéticos altos (acima de 6T) ainda não foi superada satisfatoriamente. Com este objetivo, estuda-se a introdução de diversos compostos na matriz do diboreto de magnésio tais como elementos metálicos ou compostos de carbono. Mais recentemente, os estudos se voltaram para a utilização das técnicas de processamento cerâmico a fim de aperfeiçoar a microestrutura e melhor densificar o material. Neste trabalho o estudo da adição de um novo dopante, o hidrocarboneto 'C IND. 8''H IND. 18', é realizado comparativamente à outros dopantes extensivamente citados na literatura como o carbeto de silício e o óleo de silicone. O processo de sinterização de pastilhas ex situ é também investigado para todas... / After the discovery of new superconducting oxides in the year 1986, there was an intense search for other materials that show the phenomenon of superconductivity. Among the new discoveries is magnesium diboride, an intermetallic material whose superconductivity was unknown until 2001. Currently, it is considered to have great potential for real application due to its easy fabrication and low cost cooling as cryocoolers, since the phenomenon occurs around 39K. For almost a decade surveys have been conducted with the objective of better understanding the mechanism of superconductivity in this material, as well as to improve its ability to withstand current in high magnetic fields and thus broaden their scope of industrial application. Today it is clear that the mechanism governing the superconductivity in Mg'B IND. 2' is the same as described in 1958 by Bardeen, Cooper and Schriefer, in the BCS theory. But on the other hand, practical issues remain incomplete, since the fall of the ability to conduct supercurrent in high magnetic fields (up to 6T) has not been satisfactorily overcome. To this end, it has been studied the introduction of several compounds in the matrix of magnesium diboride elements such as metal or carbon compounds. More recently, studies have turned to the use of ceramic processing techniques to improve the microstructure and better densification of the material. In this work the study of the addition of a new doping, 'C IND. 8''H IND. 18' hydrocarbon, is performed compared to other doping compounds that are extensively cited in the literature as silicon carbide and silicon oil. The process of sintering pellets ex situ is also investigated for all these additions in order to obtain denser material and therefore better performance in the transport properties. The used techniques include X ray diffraction, thermal analysis,... (Complete abstract click electronic access below)

Orientador: Cláudio Luiz Carvalho / Banca: Rafael Zadorosny / Banca: Carlos Yujiro Shigue / Resumo: Com a descoberta dos supercondutores de alta temperatura crítica na década de 80, as propriedades elétricas desses materiais passaram a ser intensamente estudadas, principalmente por meio do processo de dopagem. Desta forma, este trabalho teve por objetivo estudar os efeitos da dopagem nas propriedades elétricas do sistema supercondutor BSCCO com fórmula estequiométrica Bi1,6Pb0,4Sr2-xRExCa2Cu3O10+δ, por meio da substituição do elemento terra rara (RE), lantânio (La), em sítios de Sr, onde 0≤x≤2,0, em intervalos de 0,5. As soluções precursoras foram preparadas baseadas no método de Pechini, obtendo-se uma resina polimérica submetida a tratamento térmico de 200oC/10h, resultando em um material na forma de pó, o qual foi levado novamente a tratamento térmico entre 400oC a 810oC. Com o pó, foram preparadas pastilhas, submetendo-as a um novo tratamento térmico de 810oC/+31h. Para se conhecer as principais características das amostras, foram feitas a caracterização estrutural por meio da técnica de Difratometria de Raios X (DRX) para todos os tratamentos térmicos realizados; a caracterização elétrica, pelo método de quatro pontas dc; para a caracterização morfológica e química, a técnica de microscopia eletrônica de varredura (MEV), juntamente com a técnica de Espectroscopia por Energia Dispersiva de Raios X (EDX); e medidas magnéticas ac e dc para a caracterização magnética das amostras. Os resultados apontaram uma degradação das propriedades de supercondução, além de mudanças estruturais, morfológicas e magnéticas com o aumento da concentração de dopante / Abstract: With the discovery of high critical temperature superconductors in the 80, the electrical properties of these materials became intensively studied, mainly through the doping process. Thus, the objective of this work was the study of the effects of doping on the electrical properties of BSCCO superconducting system with stoichiometric formula Bi1,6Pb0,4Sr2-xRExCa2Cu3O10+δ, by the substitution of rare earth element (RE), lanthanum (La) in the sites of Sr, where 0 ≤ x ≤ 2.0, in interval of 0.5. The precursor solutions were prepared based by Pechini method, obtaining a polymer resin subjected to heat treatment at 200° 10h, which results in a material in C/ a powder. In sequence, futher heat treatments between 400oC to 810oC were made. With the powder were prepared bulks by subjecting them to a new heat treatment at 810° C/+31h. To know the main characteristics of the samples, structural characterization were made by the technique of X-Ray Diffraction (XRD) for all heat treated samples; the electrical characterization was made by the dc four probe method; for the morphological and chemical characterization by the technique of Scanning Electron Microscopy (SEM), together with the technique of Energy Dispersive X-ray Spectroscopy (EDX); and ac and dc magnetic measurements to magnetic characterization of the sample. The results shown a degradation of superconducting properties, as a consequence of structural, morphology and magnetic changes with the dopant concentration increase / Mestre

Depuis plus de dix ans les capacités MIM (Métal Isolant Métal) sont des composants passifs largement intégrés au niveau des interconnections des puces de microélectronique. A cause de la miniaturisation et de la réduction de la surface des puces, la densité de capacité des capacités MIM doit être constamment augmentée. Une solution est l'utilisation d'un isolant avec une constante diélectrique élevée dit « high-k ». Pour les prochaines générations de condensateurs, des densités de capacité supérieur à 30 fF/µm² sont visées. L'oxyde de zirconium (ou zircone) a été sélectionné pour remplacer de l'oxyde de tantale actuellement utilisé. Il possède une constante diélectrique qui dépend de sa structure cristalline. Elle est respectivement de 17, 47 et 37 dans les phases monoclinique, quadratique et cubique. Il est donc nécessaire de déposer la zircone dans la phase quadratique. Cependant, les couches minces de zircone ne sont pas entièrement cristallisées dans la phase quadratique. De plus, elles ne répondent pas aux critères de fiabilité requis par la microélectronique. L'objectif de cette thèse est la stabilisation de la zircone dans la phase quadratique par le dopage. Le tantale et le germanium sont les deux dopants choisis grâce à une étude de sélection de matériaux. Des couches minces d'environ 8 nm de zircone dopée à différentes concentrations ont été réalisées par PEALD (Plasma Enhanced Atomic Layer Deposition). Après les dépôts, des recuits à 400°C pendant 30 min ont été effectués afin de reproduire les traitements thermiques subis par les couches lorsqu'elles sont intégrées dans des puces de microélectronique. Plusieurs types de caractérisations ont été effectuées afin d'étudier l'influence des dopants sur la structure cristalline de la zircone mais aussi sur ses propriétés physico-chimiques. Des tests électriques sur des capacités MIM intégrées ont permis de mesurer les propriétés électriques des couches et la fiabilité de la zircone dopée a été évaluée. Ce travail a pour but la fabrication d'une capacité MIM planaire à base d'oxyde de zirconium dopée dont la densité de capacité sera supérieure à 30 fF/µm² pour des applications de découplage. / For more than ten years Metal-Insulator-Metal capacitors (MIM) have been integrated at the level of copper interconnections. All new technology nodes have led to a decrease of the surface of chips; capacitance density must be thus enhanced. The best solution is to use a material with a high dielectric constant commonly named “high-k”. For the next MIM capacitor generation, capacitance density has to be higher than 30 fF/µm². Tantalum oxide, currently used, has reached its limits and it must be replaced. Zirconium dioxide has a high dielectric constant of 47 in the tetragonal phase with a sufficient band gap for MIM applications. When deposited in thin films, zirconia is not fully crystalized in the tetragonal phase. Moreover, this pure zirconium oxide does not fulfill the reliability criteria. The aim of this work is to stabilize zirconia in its tetragonal phase by alloying it with other elements. Tantalum and Germanium are the two dopants selected thanks to a bibliographic study. Thin layers of zirconia of 8 nm alloyed with Tantalum and Germanium have been deposited by Plasma Enhanced Atomic Layer Deposition (PEALD). Samples were annealed at 400°C during 30 minutes after deposition to reproduce the thermal conditions that microelectronic chips are submitted to. Different characterization technics have been used to study the effect of dopants on zirconia's crystalline structure and its physic-chemical properties. Tests have been made on integrated MIM capacitors with Titanium Nitride electrodes to determine the electrical properties of the layers. Reliability of zirconia doped layers was also evaluated. The purpose of this work is the production of zirconia based planar MIM capacitor with a capacitance density of 30 fF/µm².

Le carbure de silicium est un semi-conducteur à grand gap qui s’est récemment imposé comme un matériau clé pour l’électronique de puissance. Les cristaux massifs ainsi que les couches épitaxiales actives sont aujourd’hui obtenus par des procédés en phase gazeuse, comme la croissance par sublimation (ou PVT) et le dépôt chimique en phase gazeuse (CVD), respectivement. Le procédé de croissance en solution à haute température est actuellement revisité en raison de sa capacité à atteindre des qualités cristallines exceptionnelles. Ce travail est une contribution au développement du procédé de croissance en solution à partir d’un germe (TSSG), avec comme objectif principal l’accès à des cristaux de 4H-SiC fortement dopés de type p. Le dopant p le plus utilisé est l’Aluminium. Différentes étapes élémentaires du procédé sont étudiées, avec pour chaque étape l’évaluation de l’effet de l’Al. Après un bref rappel historique sur le SiC, les données fondamentales du SiC sont introduites dans le chapitre 1 et discutées par rapport aux applications en électronique de puissance. Dans le chapitre 2, le réacteur de croissance est détaillé. Les trois principaux aspects techniques du procédé sont exposés : i) l’apport en carbone par dissolution à l’interface entre le creuset en graphite et le liquide, ii) le transport du carbone de la zone de dissolution à la zone de cristallisation, et iii) la cristallisation sur le germe. Ces trois aspects ont été étudiés et améliorés par l’ajout de métaux de transition (Fe ou Cr) au solvant de façon à augmenter la solubilité en carbone, en favorisant le transport du carbone par l’optimisation de la convection forcée (i.e. la rotation du cristal) et en stabilisant le front de croissance. Après optimisation, un cristal de 4H-SiC a pu être obtenu à une vitesse supérieure à 300 µm/hr et avec un élargissement du diamètre d’environ 41% par rapport au diamètre initial du germe. Le chapitre 3 porte sur l’étude de l’interaction entre le solvant et la surface du 4H-SiC à l’équilibre, sans croissance, en utilisant la méthode de la goutte posée. L’effet du temps, de la température et de l’ajout d’Al ont été étudiés. L’interface liquide/solide présente une évolution en trois étapes : i) dissolution, ii) step-bunching et iii) facettage, la surface initiale en marches et terrasses se décomposant en facettes de type (0001), (10-1n) et (01-1n). L’augmentation de la température de 1600°C à 1800°C provoque le même effet que l’ajout d’aluminium : une accélération de la deuxième étape ainsi qu’une limitation de la troisième étape. Dans le chapitre 4, des phénomènes transitoires ont été étudiés lorsque le substrat touche la surface du liquide. A l’instant du contact, il a été démontré par simulation numérique que le liquide au voisinage du substrat est sujet à de très fortes fluctuations de températures et donc à de fortes fluctuations de sursaturation. Ceci est à l’origine d’une germination transitoire de 3C-SiC sur la surface du cristal et ce, même à très haute température. Ce phénomène peut être évité soit en préchauffant le cristal avant le contact soit en ajoutant de l’aluminium dans le liquide. L’amélioration de la convection forcée est un moyen efficace pour augmenter la vitesse de croissance. Cependant, au-delà d’une certaine vitesse de rotation du cristal, un type d’instabilité spécifique se développe. Elle est basée sur l’interaction entre la direction d’avancée de marches à la surface du cristal et la direction locale du flux de liquide au voisinage de la surface. Ceci fait l’objet du chapitre 5. Finalement, la concentration de porteurs ainsi que la concentration totale en azote (N) et en aluminium (Al) sont étudiées en fonction de différents paramètres de croissance dans le chapitre 6. Une concentration en Al aussi élevée que 5E+20 at/cm3 a pu être obtenue à 1850°C. Cette valeur est très prometteuse pour le futur développement de substrats de 4H-SiC de type p+. / Silicon Carbide is a wide band gap semiconductor which has recently imposed as a key material for modern power electronics. Bulk single crystals and active epilayers are industrially produced by vapor phase processes, namely seeded sublimation growth (PVT) and chemical vapor deposition (CVD) respectively. The high temperature solution growth is currently being revisited due to its potential for achieving high structural quality. This work is a contribution to the development of the top seeded solution growth (TSSG) process, with a special focus on heavily p-type doped 4H-SiC crystals. Aluminum (Al) is the most commonly used acceptor in SiC. Different elementary steps of the process are studied, and for every cases, the effect of Al is considered and discussed. After a brief history of SiC material, basic structural and physical properties of silicon carbide are introduced in chapter 1 and discussed with respect to power electronics applications. In chapter 2, the crystal growth puller is detailed and the three most important technical issues of the SiC solution growth process are discussed : i) carbon supply by dissolution at the graphite crucible/liquid interface, ii) carbon transport from the dissolution area to the growth front, and iii) crystallization on the seed substrate. These three steps are studied and improved by adding transition metals (Fe or Cr) to the solvent in order to increase the carbon solubility, by increasing the carbon transport with the optimization of the forced convection (i.e. rotation of the crystal) and by stabilizing the growth front. After optimization, a 4H-SiC crystal is demonstrated with a growth rate of over 300 µm/h and a diameter enlargement of about 41% compared to the original seed size. Chapter 3 is dedicated to the investigation of the interaction between the liquid solvent and the 4H-SiC surface under equilibrium conditions, i.e. without any growth, using a sessile drop method. Effect of time, temperature and the addition of Al to pure liquid silicon are investigated. It is shown that the liquid/solid exhibits a three stages evolution: i) dissolution, ii) step bunching and iii) faceting, the original step and terrace structure being decomposed into (0001), (10-1n) and (01-1n) facets. Increasing the temperature from 1600°C to 1800°C or adding Al drastically enhances the second stage, but reduces the third one. In chapter 4, transient phenomena during the seeding stage of the growth process on the seed crystal are investigated. With the help of numerical modeling, it is shown that strong temperature fluctuations during the contact between the seed and the liquid can give rise to transient 3C-SiC nucleation on the crystal surface, even at high temperatures. This phenomenon can be avoided by either pre-heating the seed or by adding Al. Increasing forced convection (rotation rate of the crystal) is a good way to increase the growth rate. However, above a critical rotation rate, a special surface instability develops. It is based on the interaction between the step flow at the growing surface and the local fluid flow directions close to the surface. This is investigated in Chapter 5. Finally, carrier concentrations and total dopant (nitrogen and aluminum) concentrations are investigated as a function of different process parameters in chapter 6. Al incorporation as high as 5E+20 at/cm3 has been achieved in layers grown at 1850°C. This value is very promising for the future development of p+ 4H-SiC substrates.

A redistribuição e ativação elétrica dos dopantes tipo n (As e Sb) e tipo p (Ga e In) em Si com excesso de vacâncias foram analisadas. As vacâncias foram geradas por implantação iônica de altas doses de oxigênio ou nitrogênio em alta temperatura, de acordo com procedimentos já estudados. Em seguida foram implantados os dopantes com dose de 5x1014 cm-2 a 20 keV na região rica em vacâncias. Dopagens idênticas foram realizadas em amostras de Si sem vacâncias e em SIMOX. Em seguida foram feitos recozimentos a 1000ºC por 10 s ou 15 min. Os perfis atômicos dos dopantes foram medidos com Medium Energy Ion Scattering e os perfis dos dopantes ativados, com Hall diferencial. A redistribuição e as propriedades elétricas de cada um dos dopantes no Si sem vacâncias foram bastante similares às observadas no SIMOX, porém várias diferenças foram observadas em relação às amostras com excesso de vacâncias. As vacâncias reduziram a ativação elétrica do As e do Sb, mas proporcionaram maior estabilidade da ativação após recozimentos longos. A redistribuição destes dopantes foi infuenciada pelo íon usado na geração das vacâncias, ou seja, nitrogênio ou oxigênio. O oxigênio proporcionou maior dose retida de As e o nitrogênio, maior dose retida de Sb. Já para o Ga e o In, as vacâncias tiveram papel fundamental na sua redistribuição, diminuindo a difusão para fora das amostras e garantindo maior dose retida. A ativação elétrica do Ga e especialmente a do In foram baixas, onde observamos forte influência do íon pré-implantado, principalmente o oxigênio. / The redistribution and electrical activation of n type (As and Sb) and p type (Ga and In) dopants in Si with excess vacancy concentration were analyzed. The vacancies were formed by high dose ion implantation of oxygen or nitrogen at high temperature, following previously studied procedures. Dopants were implanted to a dose of 5x1014 cm-2 at 20 keV in the vacancy rich regions of the samples. Identical doping implantations were performed in bulk Si and SIMOX. Samples were then submitted to thermal annealing at 1000ºC for 10 s or 15 min. The dopants atomic profiles were obtained by Medium Energy Ion Scattering and the active dopant profiles, by differential Hall measurements. The redistribution and the electrical properties of each dopant in bulk Si were similar to those observed in SIMOX, but several differences were observed in the vacancy-rich samples. Vacancies reduced the electrical activation of As and Sb, although the activation was maintained stable after long annealing times. The redistribution of these dopants was, otherwise, dominated by the ion used in the vacancy generation, i.e., nitrogen or oxygen. The presence of oxygen resulted in larger As retained dose, while the presence of nitrogen, in larger Sb retained dose. Regarding the p type dopants, Ga and In, the vacancies played an important role in their redistribution, reducing their out-diffusion and allowing larger retained doses. Ga and especially In electrical activation was low, where strong influence of the pre-implanted ions was observed, especially oxygen.

Cette thèse porte sur l’optimisation d’un procédé de croissance, reproductible et contrôlé, d’une monocouche de graphène sur la face –Si du carbure de silicium (SiC (0001)) par sublimation sous faible pression d’argon (10 mbars). Au vue de la littérature, cette croissance à faible pression reste un challenge. Différentes techniques complémentaires telles que la spectroscopie Raman, la microscopie à force atomique, la microscopie à effet tunnel et des mesures d’effet Hall ont été menées afin de valider la croissance de la monocouche et d’en étudier sa morphologie de surface ainsi que ses propriétés structurales et électroniques. L’ensemble des résultats obtenus démontre le contrôle de la croissance d’une monocouche de graphène homogène, continue et de grande taille (6x6mm²). Plus de 50 échantillons monocouches ont été synthétisés pendant la thèse démontrant ainsi un procédé reproductible dans un bâti de croissance prototype de la société montpelliéraine Annealsys. Un mécanisme de croissance en bord de marche et la présence de marches et de terrasses a pu être mis en évidence alors que la littérature rapporte des difficultés à optimiser des procédés de croissance à basse pression d’argon. L’effet de la vitesse de montée en température a également été étudié dans le but de contrôler la morphologie du SiC de façon à pouvoir évaluer l’impact de la largeur des marches sur les propriétés électroniques du graphène. La largeur des marches obtenue (10 µm) permettront des mesures originales de transport, localisées sur une marche.Le procédé robuste et reproductible développé a permis différentes études approfondies sur ce graphène épitaxié. Sur la face-Si du SiC croît d’abord une couche tampon liée de manière covalente au SiC. Une deuxième couche tampon croît sous la première qui devient alors du graphène. Le peu de résultats présents dans la littérature nous a conduit à étudier cette couche d’interface entre le graphène et le SiC. A partir d’un nombre important de mesures par spectroscopie Raman, la signature de cette couche tampon a pu être obtenue. Un spectre Raman inhomogène de celle-ci a été mis en évidence. Pour aller plus loin, nous avons mis en œuvre deux techniques d’exfoliation du graphène pour avoir accès à la couche tampon sur SiC. Les signatures Raman des couches tampon couvertes ou non de graphène ont été analysées et comparées. Deux résultats majeurs sont à souligner : (i) l’aire du signal Raman de la couche tampon augmente après le retrait du graphène et (ii) deux pics fins sont observés seulement sur le spectre du graphène épitaxié. Ces résultats démontrent l’existence d’un couplage entre le graphène et la couche tampon.La dernière partie de ce travail de thèse concerne les propriétés électriques de ces monocouches de graphène sur SiC. Contrairement au classique dopage n du graphène épitaxié sur SiC (0001), un dopage résiduel de type p a été mesuré et attribué à un effet de l’environnement. Les impuretés chargées présentes à la surface des échantillons pourraient être à l’origine de flaques d’électrons et de trous (puddles) réparties à la surface des échantillons et responsables de leur dopage inhomogène. Ces fluctuations de potentiel ont été estimées en ajustant les données expérimentales à partir d‘un modèle mettant en jeu deux types de porteurs. De plus, nous avons pu mettre en évidence un changement de dopage d’un type p à n sous vide et sous illumination UV. La désorption d’absorbants chargés pourrait expliquer ce changement. Ces résultats démontrent une possible modulation des propriétés électriques de nos échantillons par un facteur externe tel que l’exposition aux UV. / This manuscript presents a work aiming to optimize a reproducible and controlled growth process of a monolayer graphene on Si-face of SiC (SiC (0001)) by sublimation under low argon pressure, i.e. 10 mbar. This low pressure process is challenging regarding the results in the literature. Various complementary techniques as optical microscopy, Raman spectroscopy, atomic force microscope, scanning tunneling microscope, and Hall Effect measurements have been performed on the samples in order to validate the monolayer graphene growth and investigate their surface morphology, their structural and electronic properties. All the results obtained from these measurements confirm the control of homogeneous, continuous and large-size (6×6 mm²) monolayer graphene from our optimized growth process. More than 50 monolayers graphene were produced during this thesis, validating a reproducible process in a prototype furnace developed by Annealsys, local company in Montpellier. The step-flow growth mode which encourages the formation of step-terrace surface structures is obtained under this unclassical growth condition contrary as established in the literature. Moreover, we have investigated the effect of the temperature ramp on the SiC morphology to evaluate the impact of the width of the terraces on electronic properties of graphene. Samples with terraces larger than 10 µm have been obtained allowing original transport measurements localized on only one terrace.Thanks to the reproducibility of our optimized growth process, further characterization studies on epitaxial graphene were investigated. The first carbon layer grown on SiC (0001) is a buffer layer covalently linked to SiC. Then a second buffer layer grows under the first one that becomes graphene. This well-known buffer layer at graphene / SiC (0001) interface has been investigated in this thesis to complete the poor literature on this topic. Statistically buffer Raman signatures have been obtained and compared to the literature demonstrating an inhomogeneous buffer layer. Furthermore, we have developed two graphene transfer techniques aiming to exfoliate graphene layer and leave behind only the buffer layer on the sample surface. The Raman signatures of buffer layer in these two cases (with or without graphene coverage) have been compared. We believe the evidenced evolution could be related to the coupling between graphene and buffer layer. Two major results illustrate this coupling: (i) the Raman signature of buffer layer increases in integrated intensity after the graphene transfer and (ii) two fines peaks are observed only in epitaxial graphene spectra and not in uncovered buffer layer spectra.The last part of this work concerns the electrical properties of monolayer graphene on SiC (0001). Contrary to the typical n-type doping of epitaxial graphene, the low p-type residual Hall concentration observed in our samples has been related to the atmospheric effect. More precisely, the charged impurities deposited on the sample surface could lead to the formation of electron-hole puddles, resulting in an inhomogeneous doping. The potential fluctuation has been estimated by fitting the experimental data using a model of two types of charges. Moreover, we have shown that the doping type change from p-type to n-type under vacuum condition or under UV illumination. This could be explained by desorption of the charged absorbents during the pumping or UV illumination. These results demonstrate the possibility of tuning the electrical properties of our samples by external factor such as UV light.

The Cadmium Telluride thin film solar cell is one of the leading photovoltaic technologies. Efficiency improvements in the past decade made it a very attractive and practical source of renewable energy. Considering the theoretical limit, there is still room for improvement, especially the cell’s open circuit voltage (VOC). To improve VOC, the p-type carrier concentration and minority carrier lifetime of the CdTe absorber needs to be improved. Both these parameters are directly related to the point defect distribution of the semiconductor, which is a function of deposition stoichiometry, dopant incorporation and post-deposition treatments. CdTe films were deposited by the Elemental Vapor Transport (EVT) deposition method, which allowed in situ control of the vapor phase stoichiometry (Cd/Te ratio). Extrinsic doping of polycrystalline CdTe by in situ incorporation of antimony (Sb) and phosphorus (P) was investigated. The structural and electrical properties of CdTe thin films and solar cells were studied. Sb and P incorporation were found to increase the net p-doping concentration. Cl and Sb improved the minority carrier lifetime of polycrystalline CdTe, while lower lifetime with Cu and P doped films were indicated. Deep Level Transient Spectroscopy (DLTS) was performed on devices fabricated with different deposition stoichiometry, post-deposition treatments, and phosphorus dopant dose. Several majority and minority carrier traps were identified, and assigned to different point defects based on first principle studies in the literature and experimental conditions used for the deposition and processing of the thin films.