Selektivität zu erreichen ist eines der Hauptziele der Chemie. In der Katalyse wird allgemein von einem engen Zusammenhang zwischen der Selektivität und der Katalysatorstruktur ausgegangen - allerdings erschwert die komplexe mikroskopische Struktur realer Katalysatoren ein tiefergehendes Verständnis; daher wird versucht, weitergehende Erkenntnisse an strukturell vereinfachten Materialien zu gewinnen. Für diese Arbeit wurde hierzu ein Pd/Fe3O4-Modellkatalysator verwendet. Auf diesem System wurde die Selektivität in zwei katalytische Modellreaktionen untersucht. Die Reaktantenexposition erfolgte dabei über Molekularstrahlen im Hochvakuum, und die Reaktionsraten wurden massenspektrometrisch gemessen; Adsorbate wurden IR-spektroskopisch detektiert. - Zersetzung von Methanol: Es wird gezeigt, dass Methanol auf dem Oxid Fe3O4 sehr selektiv durch Reaktion mit Oberflächensauerstoff (Mars-van-Krevelen-Mechanismus) zu Formaldehyd und Wasser dehydrogeniert wird. Auf Pd-Metall zersetzt sich Methanol im wesentlichen sehr schnell zu Kohlenstoffmonoxid und Wasserstoff (bzw. zu Kohlenstoffablagerungen in einer Nebenreaktion). Es werden Experimente gezeigt, die darauf hindeuten, dass Diffusion von oxidgebundenem Methanol/Methoxy auf die Pd-Metallpartikel signifikant zur Gesamtaktivität des Modellkatalysators beiträgt. - Umsetzung von 2-Buten mit Deuterium: Zunächst wird gezeigt, dass die Erzielung katalytischer Aktivität kritisch von der dissoziativen Adsorption des Reaktanden Deuterium abhängt, die durch Kohlenwasserstoffadsorbate stark inhibiert wird; es war allerdings möglich, diese Limitierung experimentell zu umgehen. Darüberhinaus wird gezeigt, dass die Hydrierungsreaktion durch die Anwesenheit stark zersetzter Kohlenwasserstoffablagerungen selektiv induziert werden kann, während die alternative Reaktion (H/D-Austausch/Isomerisierung) auch in Abwesenheit dieser Spezies abläuft; mögliche Erklärungsmodelle werden diskutiert. Schließlich wird die mögliche Ursache für die unter bestimmten Reaktionsbedingungen beobachteten unterschiedlichen Reaktionsraten mit cis- und trans-2-Buten als Reaktanten diskutiert. / The achievement of selectivity is one of the main objectives in chemistry. For catalysis, selectivity is generally seen to be closely linked with catalyst structure; the complex microscopic structure of real catalysts, however, obstructs to obtain a deeper understanding; for this reason, structurally simplified materials are studied. For the current work, studies have been conducted on a Pd/Fe3O4 model catayst. On this system, the selectivity in two catalytic reactions has been examined. The exposure of the reactants was effected by molecular beams in high vacuum, and the reaction rates have been measured mass spectrometrically; additionally, adsorbates were detected by IR-spectroscopy. - Decomposition of Methanol: It is shown that on the oxide Fe3O4 methanol is dehydrogenated very selectively to formaldehyde and water by reaction with surface oxygen of the oxide (Mars-van-Krevelen mechanism). On Pd metal it is mainly decomposed very quickly to carbon monoxide and hydrogen (and, in a side reaction, to carbonaceous deposits). Experiments are shown indicating that the diffusion of oxide-related methanol and methoxy to the Pd metal-particles contributes significantly to the overall activity of the model catalyst. - Conversion of 2-Butene with Deuterium: At first it is shown that the catalytic activity depends critically on the dissociative adsorption of the reactant deuterium, which is strongly inhibited by hydrocarbon adsorbates; it was, however, possible to overcome this limitation experimentally. In addition, it is shown that the hydrogenation reaction can be selectively induced in the presence of strongly dehydrogenated carbonaceous deposits, whereas the alternative reaction (H/D-exchange/isomerisation) can proceed also without the presence of those species; possible models for explanation are discussed. Finally, the possible origin of the different reaction rates with cis- and trans-2-butene that were observed only under certain reaction conditions is discussed.

Der Einfluss der Verspannung während des Wachstums von InAs auf GaAs(001) mittels Molekularstrahlepitaxie wird in dieser Arbeit untersucht. Eine Biegebalkenapparatur wurde benutzt, um den Verlauf der Filmkraft während des Wachstums und dem nachfolgenden Anlassen bei Wachstumstemperaturen zu messen. Die Steigung in einer Darstellung von Filmkraft gegen Filmdicke ist gleich der Verspannung, die sich während des heteroepitaktischen Wachstums bildet. Während des Wachstums von InAs auf GaAs(001) unter As-reichen Bedingungen zeigt die Filmkraft zuerst eine lineare Steigung. Dieser lineare Verlauf entspricht dem Aufdampfen der Benetzungsschicht (WL). Nach Erreichen der kritischen Schichtdicke verläuft die Filmkraft mit geringerer Steigung, was auf einen Abbau der Verspannung durch das Auftreten von Quantenpunkten (QP) hindeutet. Werden die QP nachfolgend angelassen, nimmt die Filmkraft wieder ab was durch Reifung der QDs und durch Desorption von InAs hervorgerufen wird. Modelle wurden entwickelt um die Filmkraft-Kurven, die während des Anlassens gemessen wurden, anzupassen. Die QP reifen unter Standard-Ostwald-Bedingungen für Temperaturen unterhalb 470°C. Verschiedene Mechanismen bestimmen den Reifungsprozess. Beim Anlassen bei höheren Temperaturen zeigt sich ein anderes Verhalten. Die Verspannung der QP baut sich auf Werte unterhalb der Verspannung ab, die durch das Aufbringen der Benetzungsschicht entstanden ist. Rasterkraftmikroskop-Aufnahmen zeigen, dass die QP zuerst reifen und sich dann nach ca. 450s bis 600s wieder auflösen. Im Unterschied zum Wachstum unter As-reichen Bedingungen führt das Wachstum unter In-reichen Bedingungen nicht zur Ausbildung von QP sondern verläuft im Lagenwachstumsmodus. Filmkraft-Kurven wurden ebenfalls unter diesen Bedingungen gemessen und zeigen, wie erwartet, eine deutliche Abweichungen von Kurven, die während des Stranski-Krastanov-Wachstums gemessen wurden. Eine erste vorläufige Analyse dieser Filmkraftkurven wird beschrieben. / The influence of stress on the growth of InAs on GaAs(001) by molecular beam epitaxy (MBE) is investigated in this thesis. An in-situ cantilever beam measurement (CBM) setup was used to measure the evolution of the film force during deposition and subsequent annealing at the growth temperature. The slope in a plot of film force versus film thickness is equal to the stress that builds up during heteroepitaxial growth. During the growth of InAs on GaAs(001) under As-rich conditions, the film force shows a linear slope up to a value of 2.3 N/m. This linear increase in film force corresponds to the deposition of the wetting layer. Beyond the critical thickness of 1.5-1.6 monolayers, the film force proceeds with a decreasing slope, indicating a strain release by the formation of quantum dots. When the samples are subsequently annealed, the film force decreases again due to the ripening of the quantum dots and the desorption of InAs. Models were developed to fit and explain the relaxation of the film force measured during the annealing of InAs quantum dots. At temperatures lower than 470°C, quantum dots undergo standard Ostwald ripening. Different mechanisms determine the ripening process. Fits of the models based on these mechanisms were made to the film force relaxation curves. Annealing of quantum dots at temperatures higher than 500°C shows a very different behavior. The film force accumulated during the quantum dot formation relaxes below the value which was built-up by the wetting layer growth. Atomic force microscopy images reveal that the quantum dots ripen first and then dissolve after 450s to 600s annealing. In contrast to the growth under As-rich conditions, the growth under In-rich conditions does not lead to the formation of quantum dots but proceeds rather in a layer-by-layer growth mode. The film force curves were also measured during this deposition mode. A preliminary analysis of the film force curves is presented.

When approaching a metal surface, the electronic structure of Rydberg atoms or molecules is perturbed by the surface potential and at close enough distances resonant ionisation of the Rydberg electron into the conduction band of the surface can occur. It is possible to interfere in this process and steer the ionisation distance by making use of the polarisability of the Rydberg orbital in the presence of electric fields. The resulting ions from the surface can extracted via electric fields and subsequently detected via well established ion detection schemes. The question of how this charge-transfer process is affected by different properties of the surface (both electronic and structural) represents the main aspect of the work presented in this thesis. At first, the charge transfer of atomic hydrogen Rydberg atoms with a flat gold metal surface is investigated. While such a surface might appear homogeneous, stray fields are present in its vicinity due to local variations in the surface work function. The surface ionisation process as a function of applied electric field is therefore measured experimentally and the results are compared with classical Monte-Carlo simulations (which include stray field effects). This way the possibility to utilize Rydberg states as a probe of the magnitude of such stray fields is demonstrated. To investigate the effect the surface structure can have on the ionisation process, the interaction of Rydberg atoms with surfaces covered by nanoparticles is investigated. Surface ionisation is measured at a 5 nm nanoparticle monolayer surface and it is shown that population transfer between surface- and vacuum-oriented Rydberg states occurs. In addition, results are presented, which suggest a dependence of the ionisation process on the relative size of Rydberg orbital and nanoparticle. Furthermore, charge transfer between a Rydberg state and discrete electronic states at the surface vacuum interface are investigated by performing experiments with a Cu(100) band-gap semiconductor surface. By analysing surface ionisation as a function of collisional velocity ionisation rates can be determined and are subsequently compared with theoretical predictions. The potential of identifying resonant ionisation is thereby demonstrated. Last, a new method to produce 2s atomic hydrogen via mixing of the 2s and 2p state in an electric field is proposed and first experimental results are presented, thus demonstrating viability of the idea. The experiments presented in this thesis represent the most in depth analysis of the charge-transfer process between atomic hydrogen Rydberg states and a range of different surfaces to date. As such, they demonstrate the potential of utilizing the unique properties of Rydberg states and their applicability as surface probes. In addition, these results pave the way for further experiments involving thin films or the phenomenon of quantum reflectivity.

No description available.

No description available.

No description available.

No description available.

Les nanofils (NFs), de par leur propriétés opto et nanoélectroniques sont devenus des éléments indispensable à la fabrication des dispositifs de la nanoélectronique. Le problème principal reste la reproductibilité en terme de densité de NFs, de diamètre... Cette thèse a pour but de développer grâce à la technologie FIB, un procédé permettant l'élaboration de NFs organisés et homogène en taille. / Due to their ease of fabrication and unique physical properties, semiconductor nanowires (NWs) have been proposed as building blocks for new nanoelectronic and photonic devices. Various processes have been developed to obtain large density of ultra-small NWs but naturally forms nanowires often lack reproducibility. We propose to develop a bottom-up (B-U)processes which is based on naturally formed NWs grown on a patterned substrate resulting from self-assembly of metallic clusters or exposition to a focused ion beam (FIB). The major goal consist to obtain organized and homogeneous NWs.

Ce mémoire présente une étude de la synthèse par la méthode Vapeur-Liquide-Solide (VLS) de nanofils de silicium et de germanium par Epitaxie par Jets Moléculaires ainsi que de l'effet de la présence de manganèse sur leur croissance. La croissance des nanofils est fortement modifiée par la présence de manganèse. Les nanofils de silicium élaborés sous un faible flux de manganèse présentent des propriétés morphologiques et structurales remarquables. La présence de manganèse modifie le diamètre d'équilibre des gouttes AuSi utilisées pour la croissance par voie VLS et permet l'élaboration de nanofils de silicium de longueurs élevées et de faibles diamètres. De plus, leur qualité cristalline est considérablement améliorée par rapport aux nanofils de silicium formés sans apport de manganèse. Dans ce mémoire nous proposons quelques explications à ce phénomène. Dans le cas des nanofils de germanium, l'incorporation de manganèse n'a pu être obtenue par codépôt. Aussi, (i) le dopage par implantation ionique de nanofils de germanium et (ii) la fonctionnalisation de nanofils de germanium par la formation d'hétérostructures type cœur/coquille Ge/GeMn ont été considérés : - les mesures d'aimantation effectuées sur des nanofils de germanium implantés au manganèse démontrent l'existence de propriétés ferromagnétiques avec des températures de Curie supérieures à 400K. Il s'agit d'un résultat très prometteur en vue d'applications utilisant des nanofils de germanium ferromagnétiques à température ambiante ; - pour accéder aux propriétés magnétiques des nanofils de germanium fonctionnalisés par dépôt de GeMn, nous avons mis au point une procédure de prises de contacts adaptée à la mesure de leurs propriétés de magnétotransport. Les caractéristiques électriques de ces dispositifs montrent que les propriétés de transport sont dominées par la présence de la couche coquille de GeMn, surtout à basse température. Des mesures de magnétotransport effectuées à 100K indiquent l'existence d'effets de magnétorésistance liés aux propriétés ferromagnétiques des nanofils de Ge ainsi fonctionnalisés. / This thesis presents a study of the Vapour-Liquid-Solid (VLS) synthesis of silicon and germanium nanowires by Molecular Beam Epitaxy and the effect of the presence of manganese on the growth properties. The presence of manganese strongly modifies the growth of nanowires and observed behaviours are very different for AuSi and AuGe systems. Silicon nanowires grown in the presence of manganese exhibit very interesting morphological and structural properties. The presence of manganese modifies AuSi droplets' diameter and allows manufacturing long nanowires with relatively small diameters. Moreover, the crystalline quality is dramatically improved as compared to that of silicon nanowires grown without manganese. In this manuscript we propose some explanation for the growth phenomena. In the case of germanium nanowires, manganese incorporation could not be obtained by concomitant deposition of germanium and manganese. Consequently, (i) the doping of germanium nanowires by ion implantation as well as (ii) germanium nanowires functionalization by core/shell Ge/GeMn heterostructures formation were considered: - magnetization measurements performed on implanted germanium nanowires demonstrate ferromagnetic properties with Curie temperatures above 400K. This result is very promising for the processing of devices using room-temperature ferromagnetic germanium nanowires ; - in order to access Ge/GeMn nanowires magnetic properties, we processed samples to probe nanowires magnetotransport properties. Electrical resistivities of devices show that transport properties are dominated by GeMn shell layer even more at low temperature. Magnetotransport measurements done at 100K indicate magnetoresistance effects linked with nanowires ferromagnetic properties.

Efficiency droop is a critical issue for the Group III-nitride based light-emitting diodes (LEDs) to be competitive in the general lighting application. Carrier spill-over have been suggested as an origin of the efficiency droop, and an InAlN electron-blocking layer (EBL) is suggested as a replacement of the conventional AlGaN EBL for improved performance of LED. Optimum growth condition of InAlN layer was developed, and high quality InAlN layer was grown by using metalorganic chemical vapor deposition (MOCVD). A LED structure employing an InAlN EBL was grown and its efficiency droop performance was compared with a LED with an AlGaN EBL. Characterization results suggested that the InAlN EBL delivers more effective electron blocking over AlGaN EBL. Hole-injection performance of the InAlN EBL was examined by growing and testing a series of LEDs with different InAlN EBL thickness. Analysis results by using extended quantum efficiency model shows that further improvement in the performance of LED requires better hole-injection performance of the InAlN EBL. Advanced EBL structures such as strain-engineered InAlN EBL and compositionally-graded InAlN EBLs for the delivery of higher hole-injection efficiency were also grown and tested.