Fakultät für Physik - Digitale Hochschulschriften der LMU - Teil 04/05 Archiv

Knowledge about the kinetics of chemical reactions in cells is important for an understanding of signaling pathways and regulation. Even though there are many kinetic measurements of in vitro reactions in literature, methods for in vivo measurements are sparse. With help of Temperature Oscillation Optical Lock-in (TOOL) microscopy we measure the kinetics of DNA hybridization inside cells and detect signicant acceleration or deceleration compared to in vitro measurements, dependent on the DNA sample. The dierences can not be explained by molecular crowding eects. Only models that take the background interactions with genomic DNA and RNA as well as the activity of single stranded and double stranded binding proteins into account, can be tted to data. The results imply that the biological relevance of kinetic rates measured in vitro has to be rejudged carefully. The RNA world hypothesis predicts catalytic molecules based on RNA, as for example early replicators, as precursor of modern biology. But how can a pool of appropriate RNA molecules arise under early earth conditions? In a Gillespie-model, we observe the length distribution, secondary structure and sequences of a pool of RNA molecules in porous rocks like they appear near sites of volcanic activity. We assume a monomer in ux, a length dependent out ux, a random, non-templated polymerisation and a degradation that is much stronger for single stranded than for double stranded RNA. After equilibrium is reached, the pool is populated with many hairpin-like structures due to the selection pressure for hybridized strands that can be bricks for RNA machines. Once sequence motifs and their complements appear in the reactor, they protect each other and are present longer than statistically expected. This "protection by hybridization" has the same ngerprint as a weak replication. As a consequence, the pool does not cover the full sequence space but includes more similar sequences, which is an important condition for chemical reactions. Replication of genetic information by RNA molecules is considered to be a key process in the beginning of evolution. It is so crucial that traces of this early replication are expected to be present in key processes of modern biology. We present a replication scheme based on hairpins derived from the sequence of tRNA that replicates the genetic information about a succession of sequence snippets. The replication is driven by temperature oscillations as they occur naturally inside of porous rocks in presence of temperature gradients, and independent on external chemical energy sources. It is selective for correct information and shows exponential growth rates with doubling times in the range of seconds to minutes and is thereby the fastest early replicator in the literature. The replication scheme can naturally be expanded to longer successions by using double hairpins derived from full tRNA sequences by only few mutations. By charging double hairpins with amino acids or peptides, the proposed replication bridges the gap from the RNA world to modern biology by oering a rudimentary translation mechanism, that sorts amino acids to chains according to genetic information.

Die vorliegende Dissertation befasst sich mit verschiedenen Aspekten und Techniken zur Konstruktion von String-Modellen. In diesem Kontext ist es nötig die Topologie von Calabi-Yau Mannigfaltigkeiten zu verstehen, da diese ausschlaggebend für die Nullmodenstruktur des entsprechenden Differenzialoperators und damit für das Teilchenspektrum der kompaktifizierten Niederenergietheorie ist. Für diejenigen Calabi-Yau Räume, die als Unterräume torischer Varietäten definiert werden, sind alle topologischen Größen in der Kohomololgie von Linienbündeln über der entsprechenden torischen Varietät verschlüsselt. Aus diesem Grund umfasst ein Teil dieser Dissertation die Entwicklung eines effizienten Algorithmus’ für ihre Berechnung. Nach der mathematischen Vorbereitung widmen wir uns der Herleitung und dem Beweis des auf diese Weise entstandenen mathematischen Theorems. Wir untersuchen zudem eine Verallgemeinerung auf Räume, die durch das Herausteilen einer Zn-Symmetrie konstruiert werden. Anschließend demonstrieren wir die zahlreichen Anwendungen dieser Methoden zur Konstruktion von String-Modellen. Außerdem finden wir einen Zusammenhang zwischen Kohomologiegruppen von Linienbündeln und getwisteten Sektoren von Landau-Ginzburg Modellen. Als nächstes nutzen wir die entwickelten Methoden um so genannte Zielraum Dualitäten zwischen heterotischen Modellen zu untersuchen. Diese Modelle weisen eine asymmetrische (0,2)-Weltflächensupersymmetrie auf und können über geeichte lineare Sigma-Modelle formuliert werden, in welchen sie eine Phasenstruktur ausbilden. Es lässt sich nun zeigen, dass die Phasenräume verschiedener physikalischer Modelle durch nicht-geometrische Phasen miteinander verbunden sind, was eine hochgradig nicht-triviale Dualität der entsprechenden Geometrien implizieren könnte. Unser Beitrag ist nun die Untersuchung der hierdurch verbundenen und daher potentiell dualen Modelle. Wir entwickeln ein Verfahren, welches die Konstruktion aller dualer Modelle zu einem beliebigen (0,2) Modell erlaubt und finden Evidenz dafür, dass es sich hierbei um eine echte Dualität und nicht bloß um einen Übergang verschiedener physikalischer Modelle ineinander handelt. In diesem Kontext untersuchen wir verschiedenste Szenarien, u.A. Modelle mit den Eichgruppen E6, SO(10) und SU(5), sowie mit Kompaktifizierungsräumen der Kodimension eins und zwei. In einer Untersuchung der Stringlandschaft werden dazu über 80.000 Räume auf diese Dualität untersucht.

Galaxy clusters are optimal laboratories to test cosmology as well as models for physical processes acting on smaller scales. X–ray observations of the hot gas filling their dark matter potential well, i.e. the intra–cluster medium (ICM), still provides one of the best ways to investigate the intrinsic properties of clusters. Methods based on X–ray observations of the ICM are commonly used to estimate the total mass, assuming that the gas traces the underlying potential well and satisfies spherical symmetry, and thermal motions dominate the total pressure support. However, non–thermal motions are likely to establish in the ICM, hence, contribute to the total pressure and have to be taken into account in the mass estimate. In this thesis I study the ICM thermo–dynamical structure by combining hydrodynamical simulations and synthetic X–ray observations of galaxy clusters. The main goal is to study their gas velocity field and the implications due to non–thermal motions: first, by analysing directly the velocity patterns in simulated clusters and, secondly, by reconstructing the internal ICM structure from mock X–ray spectra. To this aim, I developed and applied an X–ray photon simulator to obtain synthetic X–ray spectra from the gas component in hydrodynamical simulations of galaxy clusters. The main findings of this work are as follows. (i) Ordered, rotational patterns in the gas velocity field in cluster cores can establish during the mass assembly process, but are found to be transient phenomena, easily destroyed by passages of gas–rich subhaloes. This suggests that in smoothly growing haloes the phenomenon is in general of minor effect. Nonetheless, major mergers or highly disturbed systems can indeed develop significant ordered motions and rotation, which contribute up to 20% to the total mass. (ii) It is indeed possible to reconstruct the thermal structure of the ICM in clusters from X–ray spectral analysis, by recovering the emission measure (EM) distribution of the gas as a function of temperature. This is possible with current X–ray telescopes (e.g. Suzaku) via multi–temperature fitting of X– ray spectra. (iii) High–precision X–ray spectrometers, such as ATHENA, will allow us to measure velocity amplitudes of ICM non–thermal motions, from the velocity broadening of heavy–ion (e.g. iron) emission lines. In this work, these achievements are obtained by applying the virtual X–ray simulator to generate ATHENA synthetic spectra of simulated clusters. The non–thermal velocity of the ICM in the central region is used to further characterise the cluster and the level of deviation from the expected self–similarity. By excluding the clusters with the highest non–thermal velocity dispersion, the scatter of the LX −T relation for the sample is significantly reduced, which will allow for a more precise comparison between observations and simulations.

Galaxies have been found to divide into two families: one dominated by late-type, star forming, blue objects, which are rich in cold gas and have a low stellar mass surface density (mu*); the other is made of early-type, red and passive galaxies with higher mu* and on average low gas content. The physical mechanisms responsible for the galaxy transition between the active and passive regime are still debated. In the high mass range, mechanisms proposed to quench the star formation (SF) through cold gas heating or depletion are not efficient enough to reproduce the correct red sequence of passive systems, when implemented in models of galaxy evolution. Input for a better understanding of the physics of quenching mechanisms, and of their relative importance and efficiency, can come from a comparison of the cold atomic neutral hydrogen (HI) content and SF for a statistically significant sample of massive systems where quenching is at work. However, existing surveys do not sample this high mass, gas poor regime well enough. In this work, we study the HI properties of a volume-limited sample of ~5000 nearby galaxies with stellar mass M*>10^10 Msun, selected from the state-of-the-art blind HI survey ALFALFA to have optical and ultraviolet data so that star formation and galaxy properties can be derived. As ALFALFA does not sample with sufficient sensitivity the high mass, gas poorest range, we developed a software tool to co-add its data, in order to obtain average gas properties of galaxy classes which individually may be largely undetected. Using this technique, we study three types of quenching processes, namely the presence of a bulge component, feedback from an active galactic nucleus (AGN), and environmental mechanisms acting on the interstellar medium. Simulations of early-type galaxies with non star-forming HI disks have suggested that the presence of a bulge can stabilize the gas, thus preventing star formation, but on average we do not observe this. We find that, once mu* and NUV-r colours are fixed, the HI gas fraction in massive bulge- and disk-dominated galaxies is the same. A similar negative result is obtained if we compare M_HI/M* of AGN hosts and control galaxies, despite simulations that invoke feedback from AGN to heat or deplete cold gas in massive systems. The relation we observe between the cold gas content and the accretion rate in the red population actually points towards a co-evolution of SF and AGN activity, both driven by the amount of gas available. The last class of quenching mechanisms studied in this work includes environmental processes, which are known to affect the SF properties of galaxies and, at least in rich clusters, their cold gas content. For the first time, though, we study the effect of the environment on the HI content as a continuous function of local density, comparing it with global and inner specific star formation rate. The gradual increase in the suppression of SF from the inner to the outer regions that we observe, and the even stronger HI deficiency as a function of increasing local density, can be explained by a mechanism acting on the disk from the outside-in, like ram-pressure stripping of the HI. A comparison with mock catalogs from models, which include only removal of the hot gas, shows how models underestimate environmental effects, especially on the cold gas component of galaxies. We therefore suggest that, in order to improve our understanding of the galaxy bimodality in the local Universe, observations and models should particularly focus on environmental mechanisms acting on the cold interstellar medium. These processes are efficient over a broader range of local densities than previously thought, and could solve parts of the puzzle in the formation of massive and passive systems.

Der erste Teil dieser Arbeit stellt die Konstruktion, Inbetriebnahme und ersteDaten eines neuen optischen, faserbasierten Feldspektrographen namens VIRUS-Wvor. Die Entwicklung wurde durch den Bau des neuen 2 m Fraunhofer Teleskops aufdem Berg Wendelstein in den Bayrischen Alpen motiviert. Die besondereEigenschaft dieses Instruments liegt zum einen in der Fähigkeit, Spektren ineinem zweidimensionalen Sichtfeld aufzuzeichnen, und zum anderen in derKombina- tion eines grossen Sichtfeldes von 105 ′′ × 55 ′′ mit einer relativgrossen instrumentellen Auflösung von R = 8700. Diese Auflösung erlaubt es,Geschwindigkeitsdispersionen von Sternen und Gas bis hinunter zu 15 kms−1aufzulösen. Es wird im Allgemeinen erwartet, dass dies dem Regime entspricht,in dem sich Sterne in Scheiben bilden. Der abgedeckte Spektralbereich diesesDynamikmodus beträgt 4850 A bis 5480 A. Zusätzlich bietet VIRUS-W einen zweiten Modus, welcher der Studie der Zusammensetzung von stellarenPopulationen, also z.B. deren Alter und Metal- lizität, gewidmet ist. Diespektrale Auflösung ist hier geringer (R = 3300), aber der abgedeckteSpektralbereich ist weiter und beträgt 4340 A bis 6040 A. Dies deckt einegrössere Zahl von Absorptionsbanden im stellaren Spektrum ab. Insbesondere isthier Hβ enthalten, welches eine wichtige Grundlage zur Bestimmung des Alterseiner Stellaren Population liefert. Bis zur Fertigstellung des WendelsteinTeleskops wird sich der Spektrograph an dem 2.7m Harlan J. Smith Teleskop des McDonald Observatoriums in Texas befinden. Dort nahm er im November 2010 erfolgreich seinen Betrieb auf. Die bei der Inbetriebnahme gewonnenen Daten erlauben uns, die genaue spektrale Auflösung und die Effizienz zu bestimmen. Durch den Vergle- ich der aufgezeichneten Daten eines spektrophotometrischen Standardsterns mit veröffentlichten Werten errechnen wir fu ̈r das gesamte Atmosphären-Teleskop-Spektrographen-System eine Ef- fizient von 37% im hochauflösenden Dynamikmodus und von 40% im niedriger auflösenden Modus. Wir schliessen diesen Teil mit der Diskussion der Beobachtungen der drei Galaxien NGC2903, NGC205 und NGC3091 aus den November- und Dezember- Kampagnen ab und präsentieren erste Geschwindigkeitsfelder und Messungen von Absorptionen. Der Vergleich mit Literaturdaten für NGC205 zeigt, dass wir in der Lage sind, diese sehr niedrigen Dispersionen von ≃ 20km/s zuverlässig wiederzugeben. Der zweite Teil dieser Arbeit befasst sich mit einer Beobachtungsreihe von Bulgeregionen be- nachbarter Spiralgalaxien, die wir am Hobby-Eberly-Teleskop des McDonald Observatoriums mit dem LRS Langspaltspektrographen durchgefu ̈rht haben. Wir präsentieren kinematische Profile entlang der Hauptachse von 46 Galaxien der Hubbletypen S0 bis Sc. Für 28 dieser Objekte stellen wir ausserdem Profile entlang der kleineren Halbachse vor. Ein systematischer Vergleich der gewonnenen Daten mit photometrischen Dekompositionen erlaubt es uns zu zeigen, dass die Unterscheide, die verschiedene Bulgetypen in ihrem so genan- nten Sersic-Index zeigen, auch in der Kinematik widergespiegelt werden. Pseudobulges zeigen oft eine scheibenartige Morphologie und haben niedrigere Sersic-Indizes — also ein Leuchtkraft- profil, das eher demjenigen einer Scheibe entspricht. Klassische Bulges haben im Allgemeinen grössere Sersic-Indizes (n > 2). Wir zeigen, dass Pseudobulges auch schwächere Gradienten der Geschwindigkeitsdispersion als Funktion des Radius aufweisen — ihre Dispersionsprofile sind also vorwiegend flach. Höhere Sersic-Indizes hingegen treten in Galaxien mit steiler abfallenden Dispersionprofilen auf. Ausserdem beobachten wir, dass Pseudobulges vorwiegend in Galaxien mit nierigeren zentralen Geschwindigkeitsdispersionen auftreten und einen grösseren Grad von Rotation im Verhältnis zu ihrer mittleren Geschwindigkeitsdispersion zeigen.

Solar cells generate clean electricity from sunlight. However, they remain significantly more expensive than other, less environmentally-friendly, energy generation technologies. Although the emergence of thin-film solar cells, low-cost alternatives to the prevailing crystalline silicon solar cells, has been a significant advance in photovoltaic technology, these devices typically suffer from low absorption. If this absorption could be enhanced, it would enable an increase in power conversion efficiency and hence a reduction in cost/kW of generating capacity. This is the motivation of the work presented in this doctoral thesis. Metallic nanostructures are used to trap light within the semiconductor film in organic solar cells. By increasing the optical path length, the probability that photons are absorbed before exiting the film is increased. A novel process is developed to fabricate nanostructured metallic electrode organic solar cells. These devices feature a nanovoid array interface between the metallic electrode and the semiconduc- tor film. Absorption enhancements over conventional, planar architectures as high as 45% are demonstrated. This light-trapping is found to be largely enabled by localized void plasmons. The experimental investigations are supported by finite element simulations of absorption in solar cells, which display very good agreement with experimental results. It is found that light trapped in organic solar cell architectures is very efficiently absorbed by the organic film - in- creases in the exciton generation rate per unit volume of semiconductor material of up to 17% are observed. The simulation routine is additionally used to compare and contrast common plasmonic architectures in organic solar cells. The role of the metallic nanostructure geometry on the dominant light-trapping mechanism is assessed for various size domains and optimum architectures are identified. When implemented according to the findings of this thesis, light- trapping will have the potential to vastly increase the efficiency and hence decrease the price of thin-film solar cells.

This thesis is focused on detector concepts and analyses investigated at a future linear electron positron collider. For precision measurements at such a collider, the CALICE collaboration develops imaging calorimeters, which are characterized by a fine granularity. CALICE has constructed prototypes of several design options for electromagnetic and hadronic calorimeters and has successfully operated these detectors during combined test beam programs at DESY, CERN and Fermilab. To improve the hadronic energy reconstruction and energy resolution of a hadron calorimeter prototype with analog readout three software compensation techniques are presented in this thesis, of which one is a local and two are global software compensation approaches. One method is based on a neural network to optimize the energy reconstruction, while two are energy weighting techniques, depending on the energy density. Weight factors are extracted from and applied to simulated and test beam data and result in an average energy resolution improvement of 15 - 25 % compared to a reconstruction without software compensation. Whether such software compensation techniques are also applicable to a detector concept for a future linear electron positron collider is studied in the second part of this thesis. Simulated data, two different hadronic detector models and a local software compensation technique are used for this study. The energy resolutions for single hadrons and for jets are presented with and without software compensation. In the third part of this thesis, a study on top quark pair production at a center-of-mass energy of 500 GeV at the proposed electron positron collider CLIC is presented. The analysis is based on full detector simulations, including realistic background contributions dominated by two photon processes. The mass and width of the top quark are studied in fully-hadronic and semi-leptonic decays of top quark pairs using event samples of signal and Standard Model background processes, corresponding to an integrated luminosity of 100 fb^-1. Statistical uncertainties of the top mass of 0.08 GeV and 0.09 GeV are obtained for the fully-hadronic and the semi-leptonic channels, respectively.