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

The outer halos of galaxies are known to store vital information about the formation history and merger-induced evolution of their central galaxies, since the relaxation timescales are much larger than in the innermost parts and thus the memory of the events is conserved over a long period. This information provides fundamental insights into the processes of mass growth and morphological changes, broadening our understanding of the different mechanisms of structure formation. Additionally, the radius regime where the stellar component starts to dominate over the dark matter component is the perfect place to study the interplay between dark matter and stars. This interaction between the collisionless components of a galaxy, although much slower than the gas-induced processes, significantly alters the appearance of a galaxy in the long term. A better understanding of those processes can help to shed light on the dark sides of the galaxies.

In this work, we use the combined strength of idealized high-resolution simulations of individual galaxies and large cosmological simulations to unveil some of the information encoded in the outer halos of galaxies. The high resolution simulations allow us to disentangle the impact of selected physics on the formation and evolution of galaxies in particular, while the large cosmological simulations provide a statistically meaningful sample of galaxies covering a large range in masses and environments. The first part of this thesis focuses on the interplay between dark matter and stars, revealing that both parts actually do interact through their common potential by re-ordering into a stable state where the total halo is isothermal and its density distribution follows a $\rho \propto r^{-2}$ profile. The gas, which dissipates energy and sinks towards the center on much shorter timescales, disturbs this process, forcing the total halo into a more compact state with approximately $\rho \propto r^{-3}$. Therefore, as long as gas is present, the collisionless attractor state can not be reached, but every dry merger evolves the system towards it. This is also apparent by the fact that more evolved halos have higher central dark matter fractions and smaller amounts of stars formed in situ, and that the slopes are generally steeper at high redshifts. We conclude that the equilibrium attractor state of dry merging systems provides a new test case for $\Lambda$CDM and prove that, if $\Lambda$CDM is correct, the dark matter and the stars do communicate through their common gravitational potential.

The second part of this thesis deals with the information provided by the stellar halo. We show that the radial density profiles of all stellar halos have a universal shape which can be described by a curved exponential, independent of the morphology of their central galaxy. The strength of the curvature appears to be an indication for the amount of merging a galaxy suffered, since the stellar halo mostly grows through merging as the cold gas density in the halo region is much too low to cause a significant amount of star formation at all redshifts. With such a universal shape at hand, it is possible to study the deviations from this shape to learn about the details of the stellar accretion history of a galaxy, since different types of events leave distinct signatures. We suggest that more emphasis on the understanding of those different signatures is needed in the future to fully exploit the rich information contained in the outer halos, to learn more about the accretion driven but also the secular evolution of galaxies.

Turbulence is one of the most widespread phenomena in nature occurring in fluids and

plasmas at all scales - from the blood flow in the human body, via the Earth's atmosphere

to the remnants of supernovas at astrophysical scales. Despite its frequent occurrence,

constructing a theory of turbulent motion, which provides reliable quantitative predictions,

represents one of the unsolved problems of classical physics. Most of the research efforts

in the past have been focused on studying the Navier-Stokes model of simple fluids and trying to understand the Fourier spectrum of velocity fluctuations. Due to this common restriction to the Navier-Stokes equations, turbulence is usually associated with power-law spectra of universal form which arise only at scales where both driving and dissipation mechanisms are inactive. However, recent studies reveal that many active systems which do not possess a true inertial range can, nevertheless, exhibit power-law spectra. Furthermore, those spectra are not of universal form which contradicts the classical theory of turbulence.

One of the turbulent models we shall consider in this work derives from the Kuramoto-Sivashinsky equation. It describes a simple one-dimensional active system where energy is injected at large scales and dissipated at small scales. Based on observations from plasma physics we modify the linear part of the equation such that the large scales remain practically intact but the damping rate at high wave numbers approaches a constant.

We construct a semi-analytical approximation for the modified equation which predicts a power-law form for the energy spectrum in the range where the ratio between the characteristic linear and nonlinear frequencies is scale-independent. Furthermore, we conclude that the steepness of this power law is not universal but depends on the frequency ratio. These results are confirmed by numerical simulations. Our analysis could also be relevant for kinetic Alfven-wave turbulence in the solar wind where similar conditions might occur.

Further in this work we present the first systematic study of another active system which

provides a continuum model aimed at the coarse-grained description of the dynamics observed

in dense bacterial suspensions. The model extends the framework of the familiar Navier-Stokes equations by including additional linear and nonlinear terms in order to emulate energy injection and dissipation as well as the flocking tendency of bacteria. The resulting dynamics has been described as 'low-Reynolds-number turbulence' and the corresponding energy spectrum exhibits nonuniversal power laws at large scales. With the aid of extensive numerical simulations we study the scale-to-scale energy flow in spectral space. The physical insight gained this way helped us to develop an approximation for the

spectral energy balance equation. Its solution provides an energy spectrum of a power-law form at small wave numbers. Furthermore, we derive a functional dependence of the steepness of this power law on the system parameters. A comparison with data from numerical simulations verifies our results.

Diese Arbeit befasst sich mit der Entwicklung und dem Test optischer Frequenzkämme zur Kalibration astronomischer Spektrographen. Die Genauigkeit der besten Spektrographen war bisher durch ihre Kalibration begrenzt. Die Benutzung von Frequenzkämmen als hochgenaue optische Frequenzreferenz verspricht die Überwindung dieser Limitierung, und damit die Bestimmung von Linienpositionen in astronomischen Spektren mit nie dagewesener Genauigkeit. Dies eröffnet faszinierende neue Möglichkeiten in der Astronomie, wie die Entdeckung erdähnlicher extrasolarer Planeten über Radialgeschwindigkeitsmessungen, die direkte Messung der Beschleunigung der kosmischen Expansion, oder eine genauere kosmologische Suche nach Veränderlichkeit von Naturkonstanten. Auf Basis der vorliegenden Arbeit wurde eine kommerzielle Version des astronomischen Frequenzkamms entwickelt, die derzeit für die Installation und den Routinebetrieb an mehreren Observatorien vorbereitet wird.

Um die Kammstruktur mit astronomischen Spektrographen ausreichend gut aufzulösen, werden Frequenzkämme mit extrem großen Modenabständen von typischerweise >10 GHz benötigt. Zur Erzeugung von Frequenzkämmen mit derart hohen Modenabständen verfolgt diese Arbeit einen Ytterbium-Faserlaser-basierten Ansatz, der auf der Unterdrückung ungewollter Moden eines Frequenzkamms mit ursprünglich geringerem Modenabstand beruht. Zur breitbandigen Kalibration von Spektrographen muss das erzeugte Kammspektrum über einen großen Teil des sichtbaren Spektralbereichs hinweg verbreitert werden. Bei Pulswiederholraten von >10 GHz erweist sich dies als sehr herausfordernd, und bringt bis dahin unbekannte Effekte hervor.

Die vorliegende Arbeit entwickelt Strategien zur spektralen Verbreiterung astronomischer Frequenzkämme, und untersucht hiermit verbundene Fragen wie Farbzentren-Bildung im Kern photonischer Kristallfasern. Des Weiteren wird theoretisch und experimentell nachgewiesen, dass spektrale Verbreiterung mit einer drastischen Verstärkung unterdrückter Kammmoden einhergeht, und es wird gezeigt, wie hierdurch bedingte Kalibrations-Ungenauigkeiten begrenzt werden können. Da die Einhüllende des verbreiterten Spektrums stark strukturiert ist, ist es von Nutzen diese abzuflachen. Hierbei werden die Signalpegel aller Kalibrationslinien auf dem Spektrographen angeglichen, was deren Signal-zu-Rausch-Verhältnis maximiert und dadurch die Kalibrationsgenauigkeit erhöht. Mehrere Konzepte zur adaptiven spektralen Abflachung werden entwickelt, wobei über einen Bereich von >200 nm abgeflachte Spektren erzeugt werden.

Der astronomische Frequenzkamm wird an HARPS getestet, dem bis heute führenden Spektrographen zur Exoplanetensuche, der sich am La Silla Observatorium in Chile befindet. Über kurze Zeitspannen wird hier eine Wiederholbarkeit der Kalibration von 2,5 cm/s erreicht – einen Faktor 4 besser als mit einer Thorium-Lampe, der bis dahin besten Kalibrationsquelle. Erstmals wird der Orbit eines extrasolaren Planeten mit Hilfe eines Frequenzkamms rekonstruiert, und ein Frequenzkamm-kalibrierter Atlas solarer Linien wird aus Beobachtungen von Mondlicht erstellt. Instrumentelle Effekte werden gründlich untersucht, insbesondere Kalibrationsverschiebungen, die von den Signalpegeln auf der Spektrographen-CCD abhängen.

Hinsichtlich seiner Anwendungen in der Sonnenastronomie wird der Frequenzkamm am VTT Sonnenspektrographen auf Teneriffa getestet. Hier wird eine Technik eingesetzt, die den Spektrographen über eine monomode Glasfaser gleichzeitig mit Kalibrationslicht und Sonnenlicht versorgt. Dadurch wird Modenrauschen des Faserkanals als Ursache für Ungenauigkeiten ausgeschlossen, und die Kalibrationswiederholbarkeit verbessert sich um ca. 2 Größenordnungen gegenüber einer zeitlich getrennten Übertragung. Dieses Konzept wird zur Vermessung globaler Sonnenoszillationen und zur Bestimmung der Stabilität von Absorptionslinien aus der Erdatmosphäre angewandt.

Turbulence is a ubiquitous phenomenon in astrophysical plasmas. Most of these systems exhibit a property called cross helicity, a non-zero correlation between velocity fluctuations and magnetic-field fluctuations. In the presence of a magnetic mean-field, such as in the solar wind or in the interstellar medium, cross helicity is equivalent to an imbalance between Alfven waves co- and counter-propagating with respect to the mean-field direction. Although this imbalance can have a dramatic influence on the heating and scattering rate of charged particles which propagate through the plasma, it is often neglected in computational studies of turbulent particle transport.

In an effort to remedy this situation, we present numerical simulations of magnetohydrodynamic turbulence in which we can control the energy and the cross helicity of the system, without injecting kinetic or magnetic helicity as an unwanted side effect. Varying the strength of a magnetic guide-field allows us to determine the degree of anisotropy that the system assumes as a steady-state configuration. Detailed analysis proves that these simulations conform to theoretical models of realistic turbulence.

The diffusion of cosmic-ray particles in turbulent plasmas is often calculated using quasilinear theory and a simplified description of the electromagnetic-field spectra. By computing the trajectories of test-particles in dynamically evolving turbulence simulations with non-zero cross helicity, we study whether such quasilinear predictions of the heating rate of charged particles are valid under realistic conditions. Theory and numerical results agree well for particles propagating at the Alfven velocity, unless resistive effects play a dominant role.

Furthermore, strongly anisotropic field configurations are used to compare quasilinear pitch-angle diffusion coefficients with measurements of test-particle scattering after one gyroperiod. In particular, we focus on the scaling of the scattering rate with cross helicity. We observe excellent agreement in simulations of both balanced and imbalanced turbulence and explain the role of the magnetic moment, an approximate invariant of charged-particle motion, for pitch-angle scattering on timescales of several gyroperiods.

Up to now, short-wavelength free-electron lasers (FEL) have been systems on the scale of hundreds of meters up to multiple kilometers. Due to the advancements in laser-plasma acceleration in the recent years, these accelerators have become a promising candidate for driving a fifth-generation synchrotron light source – a lab-scale free-electron laser.

So far, demonstration experiments have been hindered by the broad energy spread typical for this type of accelerator. This thesis addresses the most important challenges of the conceptual design for a first lab-scale FEL demonstration experiment using analytical considerations as well as simulations.

The broad energy spread reduces the FEL performance directly by weakening the microbunching and indirectly via chromatic emittance growth, caused by the focusing system. Both issues can be mitigated by decompressing the electron bunch in a magnetic chicane, resulting in a sorting by energies. This reduces the local energy spread as well as the local chromatic emittance growth and also lowers performance degradations caused by the short bunch length. Moreover, the energy dependent focus position leads to a focus motion within the bunch, which can be synchronized with the radiation pulse, maximizing the current density in the interaction region. This concept is termed chromatic focus matching. A comparison shows the advantages of the longitudinal decompression concept compared to the alternative approach of transverse dispersion.

When using typical laser-plasma based electron bunches, coherent synchrotron radiation and space-charge contribute in equal measure to the emittance growth during decompression. It is shown that a chicane for this purpose must not be as weak and long as affordable to reduce coherent synchrotron radiation, but that an intermediate length is required.

Furthermore, the interplay of the individual concepts and components is assessed in a start-to-end simulation, confirming the feasibility of the envisioned experiment. Moreover, the setup tolerances for a first demonstration experiment are determined, confirming the general practicability. The revealed challenges, besides the energy spread, especially concern the source stability and the precision of the beam optics setup.

Over the past few decades, the generation of high energetic ion beams by relativistic intense laser pulses has attracted great attentions. Starting from the pioneering endeavors around 2000, several groups have demonstrated muliti-MeV (up to 58 MeV for proton by then) ion beams along with low transverse emittance and ps-scale pulse duration emitted from solid targets. Owing to those superior characteristics, laser driven ion beam is ideally suitable for many applications. However, the laser driven ion beam typically exhibits a large angular spread as well as a broad energy spectrum which for many applications is disadvantageous.

The utilization of nano-targets as ion source provides a number of advantages over micrometer thick foils. The presented PhD work was intended to investigate laser driven ion acceleration from carbon nano-targets and demonstrate the potential feasibility for biological studies. Two novel nano-targets are employed: nm thin diamond-like-carbon (DLC) foil and carbon nanotubes foam (CNF). Both are self-produced in the technological laboratory at Ludwig-Maximilians-Universität München. Well-collimated proton beams with extremely small divergence (half angle) of 2 degrees are observed from DLC foils, one order

of magnitude lower as compared to micrometer thick targets. Two-dimensional particle-in-cellsimulations indicate a strong influence from the electron density distribution on the divergence of protons. This interpretation is supported by an analytical model. In the same studies, the highest maximum proton energy was observed with a moderate laser intensity as low as 5*10^18W/cm^2. Parallel measurements of laser transmission and reflection are used to determine laser absorption in the nano-plasma, showing a strong correlation to the maximum proton energy. This observation indicates significance of absorbed laser energy rather than incident laser intensity and is supported by an analytical model. The ion energy also depends on pulse duration, a reduced optimum pulse duration is found as compared to micrometer thick targets. This behavior is attributed to a reduction of transverse electron spread due to the reduction of thickness from micrometer to nanometer. These remarkable proton bunch characteristics enabled irradiating living cells with a single shot dose of up to 7 Gray in one nanosecond, utilizing the Advanced Titanium: sapphire LASer (ATLAS)system at Max-Planck-Institut of Quantum Optics (MPQ). The experiments represent the first feasibility demonstration of a very compact laser driven nanosecond proton source for radiobiological studies by using a table-top laser system and advanced nano-targets.

For the purpose of providing better ion sources for practical application, particularly in terms of energy increase, subsequent experiments were performed with the Astra Gemini laser system in the UK. The experiments demonstrate for the first time that ion acceleration can be enhanced by exploiting relativistic nonlinearities enabled by micrometer-thick CNF targets. When the CNF is attached to a nm-thick DLC foil, a significant increase of maximum carbon energy (up to threefold) is observed with circularly polarized laser pulses. A preferable enhancement of the carbon energy is observed with non-exponential spectral shape, indicating a strong contribution of the radiation pressure to the overall acceleration. In contrast, the linear polarization give rise to a more prominent proton acceleration. Proton energies could be increased by a factor of 2.4, inline with a stronger accelerating potential due to higher electron temperatures. Three-dimensional (3D) particle-in-cell (PIC) simulations reveal that the improved performance of the double-layer targets (CNF+DLC) can be attributed to relativistic self-focusing in near-critical density plasma. Interestingly, the nature of relativistic non-linearities, that plays a major role in laserwakefield-acceleration of electrons, can also apply to the benefit o

The work presented in this thesis studies the role that accretion

events play in the evolution of galaxies in dense environments, such as

galaxy clusters. Cosmological simulations allow us to study in detail

the evolution of galaxies' halos in cluster environments and have shown

that the formation of extended halos around central cluster

galaxies and intracluster light (ICL) is closely correlated to the

morphological transformation of galaxies in clusters. However, the

extremely low surface brightness of these components makes it

difficult to gather observational constraints.

This thesis studies the light and stellar motion in the halo of the

giant elliptical galaxy M87 and its surrounding IC component at the

centre of the Virgo cluster. Virgo is the nearest ($\sim 15$~Mpc away)

large scale structure, a young cluster characterised by both spatial

and kinematic substructures. M87 has one of the oldest stellar

populations in the local Universe and a stellar halo that contains

$\sim$ 70\% of the galaxy light down to $\rm{\mu_{V}=27\, mag\,

arcsec^{-2}}$. Moreover, deep images of the Virgo cluster core have

revealed an extended network of tidal features suggesting that

accretion events characterise its mass assembly. Thus, M87 and its host

environment are prime targets to shed light on the hierarchical

assembly of structure in the Universe.

This work uses new Suprime-Cam@Subaru photometry and FLAMES@VLT

spectroscopy to study a $\sim 0.5\, \rm{deg^2}$ area around

M87, in the transition region between galaxy halo and ICL. We use

planetary nebulas (PNs) as tracers, whose strong [OIII] $\lambda$5007

\AA\ emission line makes them excellent photometric and kinematic

probes, also at large distances from the galaxy's centre.

The photometric analysis of the PN sample shows the superposition of

two stellar populations, both halo and ICL. This is confirmed by

spectroscopically detected PNs, whose velocity phase-space also

reveals that halo and ICL split into two different kinematic

components. They have very different spatial distributions and parent

stars, as indicated by the properties of the PN populations they are

associated with, such as the $\alpha$-parameter and the slope of the

planetary nebula luminosity function (PNLF).

In this thesis I give the observational proof that in Virgo

the central galaxy and the ICL both evolve through the

ongoing accretion of smaller systems. However, stellar halo and ICL

are dynamically distinct components with different velocity and

density distributions, and parent stellar populations. Whether or not

these conclusions are true for different galaxies in different Virgo

subclusters is still an open question and the topic of one of

my planned future studies.

Time-resolved electron diffraction is a powerful tool to observe ultrafast structural dynamics in materials and molecules with atomic spatial as well as temporal resolution. Due to Coulomb repulsion, however, the use of only single-electrons or few-electrons per pulse is inevitable to reach the shortest pulse durations. Electrons have rather high scattering cross sections and thus experiments in transmission require ultrathin samples in the nanometer-range, making sample preparation very challenging. Up to now, ultrafast single-electron diffraction was only demonstrated at an electron energy of 30 keV; these measurements were performed in our group at the “UED1-beamline”.

This work introduces our second-generation beamline, “UED2”, where the electron

acceleration voltage is upgraded from 30 to 100 kV, which allows the investigation of significantly thicker samples. This is decisively widening the range of complex materials that can be studied. In the experiment, electron pulses are generated by a two-photon photoemission process and the long-term stability of the source is shown. The samples can be placed in transmission as well as grazing-incidence geometry. To achieve phase-matching between the optical and electron

pulses, tilted optical pulses can be applied. We figured out that to avoid temporal distortions in tilted pulses, a geometry must be chosen in which the propagation direction of the tilted pulses is perpendicular to the grating’s surface. Furthermore, temporal distortions for ultrashort electron pulses caused by misaligned magnetic lenses are examined. It is found that a displacement or tilt of the lens causes significant temporal aberrations on a femtosecond time scale and pulse-lengthening is only minimized if the beam travels precisely on the symmetry axis. An experimental procedure detailed here for aligning lens-position and -tilt reduces the aberrations to less than one femtosecond.

For the “UED2-beamline”, a new laboratory was established and a first time-resolved electron diffraction experiment at this beamline performed. Anisotropic ultrafast atomic motion in carbon-nanotubes was observed, revealing the nature of the system’s chemical bonds, which vary from relatively weak van der Waals to strong covalent interactions.

In summary, it is thus shown that ultrafast electron diffraction at 100 keV with single/few electrons per pulse is an excellent method to study ultrafast atomic-scale dynamics even in complex solid samples with the highest possible resolution in space and time.