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

Fast and efficient detection of single atoms is a universal requirement concerning modern experiments in atom physics, quantum optics, and precision spectroscopy. In particular for future quantum information and quantum communication technologies, the efficient readout of qubit states encoded in single atoms or ions is an elementary prerequisite. The rapid development in the eld of quantum optics and atom optics in the recent years has enabled to prepare individual atoms as quantum memories or arrays of single atoms as qubit registers. With such systems, the implementation of quantum computation or quantum communication protocols seems feasible.

This thesis describes a novel detection scheme which enables fast and efficient state analysis of single neutral atoms. The detection scheme is based on photoionisation and consists of two parts: the hyperfine-state selective photoionisation of single atoms and the registration of the

generated photoion-electron pairs via two channel electron multipliers (CEMs). In this work, both parts were investigated in two separate experiments. For the first step, a photoionisation probability of p_ion = 0.991 within an ionisation time of t_ion = 386 ns is achieved for a single 87Rb-atom in an optical dipole trap. For the second part, a compact detection system for the ionisation fragments was developed consisting of two opposing CEM detectors. Measurements show that single neutral atoms can be detected via their ionisation fragments with a detection

efficiency of eta_atom = 0.991 within a detection time of t_det = 415.5 ns. In a future combined setup, this will allow the state-selective readout of optically trapped, single neutral 87Rb-atoms via photoionisation detection with an estimated detection efficiency eta = 0.982 and a

detection time of t_tot = 802 ns.

Although initially developed for single 87Rb-atoms, the concept of photoionisation detection is in principle generally applicable to any atomic or molecular species. As efficient readout unit for single atoms or even ions, it might represent a considerable alternative to conventional detection methods due to the high optical access and the large sensitive volume of the CEM detection system. Additionally, its spatial selectivity makes it particularly suited for the readout of single atomic qubit sites in arrays of neutral atoms as required in future applications

such as the quantum-repeater or quantum computation with neutral atoms.

The obtained high detection efficiency eta and fast detection time t_tot of the new detection method fullfill the demanding detector requirements for a future loophole-free test of Bell's inequality under strict Einstein locality conditions using two optically trapped, entangled 87Rb-atoms at remote locations. In such a configuration, the locality and the detection loophole can be simultaneously closed in one experiment.

Gekoppelte Klima-Chemie-Modelle erweitern die in einem System wirkenden Rückkopplungen, indem neben den physikalischen auch chemische Rückkopplungen berücksichtigt werden. Die vorliegende Arbeit hat zum Ziel chemische Rückkopplungen erstmals zu quantifizieren und ihren Einfluss auf die Klimasensitivität zu bestimmen. Durch die Kopplung des Deckschichtozeanmodells MLO an das Klima-Chemie-Modell EMAC wird es ermöglicht, Klimagleichgewichtssimulationen mit interaktiver Chemie durchzuführen. Für Gleichgewichtssimulationen, die durch eine Erhöhung der CO2-Konzentrationen angetrieben werden, zeigt sich eine signifikante Dämpfung der Klimasensitivität unter Berücksichtigung chemischer Rückkopplungen. Hierfür verantwortlich sind die negative Rückkopplung über stratosphärisches Ozon und die negative Rückkopplungsänderung über stratosphärischen Wasserdampf. Im Falle von Gleichgewichtssimulationen, die durch eine Erhöhung der anthropogenen NOx- und CO-Emissionen angetrieben werden, ist die Klimasensitivität infolge interaktiver Chemie nicht signifikant erhöht. Der Vergleich mit dem CO2-Experiment zeigt, dass die Variation der Antriebsart unterschiedliche dominierende Rückkopplungs-prozesse auslöst und somit die Klimasensitivität in verschiedenartiger Weise beeinflusst wird.

Dark energy, which constitutes about 73% of the total energy density and causes the Universe expansion to accelerate, is one of the most important open problems in physics. The nature of dark energy can be inferred from its effects on the evolution of the Universe and the growth of structures as it influences the distribution of galaxies and gas at cosmological scales at low and intermediate redshifts. To extract unbiased information from the large scale galaxy power spectrum, accurate models, encoding the distortions due to non-linear evolution, bias and redshift space distortions, are needed.

In this thesis I present a model for the full shape of the power spectrum and test its validity against a suite of 50 large volume, moderate resolution N-body dark matter simulations. My results indicate that this simple model provides an accurate description of the full shape of the dark matter and halo power spectrum, both in real and redshift space, for k<0.15 h/Mpc. Even though its application is restricted to large scales, this prescription can provide tighter constraints on the dark energy equation of state parameter w_{DE} than those obtained by modelling the baryonic acoustic oscillations signal only, where the information of the broad-band shape of the power spectrum is discarded.

I then apply this model to a measurement of the power spectrum of the distribution of about 90000 luminous red galaxies (LRGs) across 7646 deg^2 in the Northern Galactic Cap from the seventh data release of the Sloan Sky Digital Survey. The errors and mode correlations are estimated from the 160 LasDamas mock catalogues, created in order to simulate the LRG galaxies and the survey geometry. Using the galaxy distribution, in combination with the latest measurement of the temperature and polarisation anisotropy in the cosmic microwave background, the luminosity-distance relation from the largest available type 1a supernovae dataset and a precise determination of the local Hubble parameter, I obtain cosmological constraints for five different parameter spaces. The analysis performed in this thesis shows that the Universe is geometrically flat and composed by about 4% of baryons, 23% of cold dark matter and 73% of dark energy. I measure w_{DE} to be -1.025+-0.065 without any evidence of time dependence, which is compatible with a cosmological constant.

In the next years new experiments will allow to measure the clustering of galaxies with a precision much higher than today, and models like the one used here will provide valuable tools in order to achieve the full potentials of the observations and obtain unbiased constraints on the cosmological parameters.

Future Extremely Large Telescopes (ELTs) will be segmented. To achieve the desired image quality, in particular for high contrast imaging, the segments of the primary mirror

must be phased to an accuracy which corresponds to a small fraction of the observing wavelength.

Information on the misalignments of the segments of the primary mirror will be provided with a phasing camera. The Zernike phase contrast method is a novel technique to phase

the primary mirrors of segmented telescopes.

The present thesis aims at studying the phasing of segmented mirrors using the Zernike phase contrast method. The theoretical background of this sensor and the algorithm used to retrieve the piston, tip, and tilt information are described.

The Zernike Phase Contrast Sensor has been tested in the laboratory and on-sky on a Unit Telescope of the ESO Very Large Telescope with a segmented mirror conjugated to

its primary mirror to emulate a segmented telescope. The performance of the sensor as a function of various parameters such as star magnitude, seeing, and integration time is discussed.

The phasing accuracy obtained on sky has always been below 15 nm root mean square wavefront error under normal conditions of operation. Equivalent limiting star magnitudes

of 18 and 15.7 achieved in the laboratory and on-sky, respectively, with this sensor would be sufficient to phase segmented telescopes in closed-loop during observations.