Strahlungskorrekturen zur Hadronproduktion bei niedrigen Energien, wie man sie an den e+e- -Beschleunigern DAFNE und VEPP-2M misst, werden untersucht. Ziel dieser Arbeit ist es, die von QED--Korrekturen befreiten hadronischen Wirkungsquerschnitte aus den hadronischen Daten mit einer Praezision auf Promille-Niveau zu extrahieren. Hadronische Praezisionsdaten werden benoetigt, um den theoretischen Fehler zur laufenden Feinstrukturkonstanten alpha(s) und zum anomalen magnetischen Moment des Myons a(mu) zu senken und sie stellen daher einen Schluessel zur moeglichen Entdeckung ``neuer Physik'' dar. Insbesondere die Paarproduktion geladener Pionen e+e- -> pi+ pi- unterhalb einer Energie von 1 GeV ist von grosser Wichtigkeit. Zu diesem Prozess werden die vollstaendigen O(alpha)-QED-Korrekturen zum Anfangszustand, Endzustand sowie die Interferenzkorrekturen berechnet. Analytische Formeln zu den virtuellen und reellen photonischen Korrekturen werden angegeben. Der totale Wirkungsquerschnitt sigma, die Pion-Winkelverteilung dsigma/dcos(theta) und die Invariantemasseverteilung des Pionpaars dsigma/ds' werden fuer den Fall realistischer kinematischer Schnitte untersucht. Es wird gezeigt, dass zusaetzlich zu den vollstaendigen O(alpha)-Korrekturen zusaetzlich die photonischen Anfangszustandskorrekturen der Ordnung O(alpha^2) und fuehrende Photonbeitraege der Ordnung O(alpha^3) sowie Beitraege zur e+ e- -Paarabstrahlung vom Anfangszustand beruecksichtigt werden muessen wenn mindestens eine Genauigkeit auf Prozent-Niveau verlangt wird. Fuer die Datenanalyse wird der Schwerpunkt auf eine inklusive Behandlung aller Photonen gelegt. Die Messung sowohl des totalen Wirkungsquerschnitts als auch der pi+ pi- Invariantemasseverteilung betreffend wird der theoretische Fehler dieser Behandlung der Strahlungskorrekturen mit 2 Promille abgeschaetzt. Ausserdem wird die Modellunsicherheit als Folge der Pion-Substruktur diskutiert. Um den Formfaktor mit der gewuenschten Praezision aus den experimentellen Daten extrahieren zu koennen, wurde ein auf diese Fragestellung zugeschnittenes Fortran-Programm geschrieben, welches die Beruecksichtigung realistischer kinematischer Schnitte erlaubt. Insgesamt erfuellt die Genauigkeit der theoretischen Vorhersagen die Erfordernisse der Niedrigenergie- e+ e- -Experimente wie diejenigen bei DAFNE oder VEPP-2M. / Radiative corrections to low energy hadron production as measured at the e+ e- colliders DAFNE and VEPP-2M are investigated. The goal of this work is to provide the theoretical condition for extracting hadronic cross sections undressed from QED corrections from the measured data with a precision of per mill level. High precision hadronic data are required to reduce the theoretical error of the running fine structure constant alpha(s) and the muon anomalous magnetic moment a(mu) and therefore represent a key to a possible discovery of ``new physics''. Especially the channel of charged pion pair production e+e- -> pi+ pi- below 1 GeV appears to be of great importance. To this process the complete O(alpha) QED initial state, final state and initial-final state interference corrections are calculated. Analytic formulae are given for the virtual and for the real photon corrections. The total cross section sigma, the pion angular distribution dsigma/dcos(theta) and the pi+ pi- invariant mass distribution dsigma/ds' are investigated in the regime of experimentally realistic kinematical cuts. It is shown that in addition to the full O(alpha) corrections also the O(alpha^2) and leading log O(alpha^3) photonic corrections as well as the contributions from IS e+ e- pair production have to be taken into account if at least per cent accuracy is required. For the data analysis I focus on an inclusive treatment of all photons. The theoretical error concerning this treatment of radiative corrections is then estimated to be 2 per mill for both the measurement of the total cross section and the pi+ pi- invariant mass distribution. In addition the model uncertainty due to the pion substructure is discussed. To be able to extract the pion form factor from the experimental data with the desired accuracy a dedicated Fortran program was written which allows to take into account experimentally realistic kinematical cuts. Altogether the precision of the theoretical prediction matches the requirements of low energy e+ e- experiments like the ones at DAFNE and VEPP-2M.

Die Berechnung von Tensorintegralen ist eines der komplizierteren Probleme bei der Berechnung von Einschleifen-Feynmandiagrammen. In dieser Arbeit wird die Computerprogrammbibliothek PJFry entwickelt, mit der Tensorintegrale mit bis zu fünf äusseren Beinen und unter Zugrundelegung beliebiger Kinematik numerisch ausgewertet werden können. Im Programm PJFry sind Algorithmen implementiert, mit denen bei der Reduktion von Pentagon-Tensoren inverse Potenzen der Gramdeterminanten vermieden werden können. Gramdeterminanten der Boxdiagramme werden unter Verwendung von Rekursionsrelationen mit variabler Raum-Zeit-Dimension in einem Satz neuer Basisintegrale isoliert. Die neuen Basisintegrale werden ebenfalls durch Rekursionsrelationen mit variabler Raum-Zeit-Dimension oder durch Entwicklung in kleinen Gramdeterminanten ausgewertet. Die Konvergenz letzterer wird durch Padé-Extrapolation erheblich beschleunigt. Ein Cache-System erlaubt die mehrfache Verwendung von numerischen Bausteinen und erhöht zusätzlich die Effizienz des Programmpakets. Ausser ausführlichen Tests von Struktur und Genauigkeit der Algorithmen wird eine nichtriviale Beispielanwendung ausgearbeitet und mit dem Programm NGluon verglichen: die Berechnung von fünf-Gluon-Helizitätsamplituden. Schließlich werden die virtuellen Einschleifenkorrekturen zur Myonpaarproduktion mit Emission energiereicher ("harter") Photonen berechnet. Die Methode wird erläutert, wie auch Renormierung und Behandlung der Polstruktur in dimensionaler Regularisierung. Numerische Vorhersagen für differentielle Wirkungsquerschnitte werden berechnet, unter Zugrundelegung der kinematischen Situationen, wie sie bei den Detektoren KLOE (DAFNE, Frascati) und BaBar (SLAC) typisch sind. / In this thesis we discuss the problem of evaluation of tensor integrals appearing in a typical one-loop Feynman diagram calculation. We present a computer library for the numerical evaluation of tensor integrals with up to 5 legs and arbitrary kinematics. The code implements algorithms based on the formalism which avoids the appearance of inverse Gram determinants in the reduction of pentagon diagrams. The Gram determinants of box integrals are isolated in the set of new basis integrals by using dimensional recurrence relations. These integrals are then evaluated by dimensional recurrence or expansion in small Gram determinant, which is improved by Padé extrapolation. A cache system allows reuse of identical building blocks and increases the efficiency. After describing the cross checks and accuracy tests, we show a sample application to the evaluation of five gluon helicity amplitudes, which is compared with the output of the program NGluon. In the last part the program is applied to the calculation of the one-loop virtual corrections to the muon pair production with hard photon emission. The computation method is explained, followed by a discussion of renormalization and pole structure. Finally, we present numerical results for differential cross sections with kinematics of the KLOE and BaBar detectors.

Die Arbeit untersucht theoretisch die Wechselwirkung neutraler Teilchen (Atome, Moleküle) mit Oberflächen, soweit sie durch das elektromagnetische Feld vermittelt wird. Spektrale Energiedichten und Kohärenzfunktionen werden hergeleitet und liefern eine umfassende Charakterisierung des Felds auf der sub-Wellenlängen-Skala. Die Ergebnisse finden auf zwei Teilgebieten Anwendung: in der integrierten Atomoptik, wo ultrakalte Atome an thermische Oberflächen koppeln, und in der Nahfeldoptik, wo eine Auflösung unterhalb der Beugungsbegrenzung mit einzelnen Molekülen als Sonden und Detektoren erzielt werden kann. / We theoretically discuss the interaction of neutral particles (atoms, molecules) with surfaces in the regime where it is mediated by the electromagnetic field. A thorough characterization of the field at sub-wavelength distances is worked out, including energy density spectra and coherence functions. The results are applied to typical situations in integrated atom optics, where ultracold atoms are coupled to a thermal surface, and to single molecule probes in near field optics, where sub-wavelength resolution can be achieved.

The purpose of this thesis is to advance the understanding of strong-field QED processes in short laser pulses. The processes of non-linear one-photon and two-photon Compton scattering are studied, that is the scattering of photons in the interaction of relativistic electrons with ultra-short high-intensity laser pulses. These investigations are done in view of the present and next generation of ultra-high intensity optical lasers which are supposed to achieve unprecedented intensities of the order of 10^24 W/cm^2 and beyond, with pulse lengths in the order of some femtoseconds. The ultra-high laser intensity requires a non-perturbative description of the interaction of charged particles with the laser field to allow for multi-photon interactions, which is beyond the usual perturbative expansion of QED organized in powers of the fine structure constant. This is achieved in strong-field QED by employing the Furry picture and non-perturbative solutions of the Dirac equation in the presence of a background laser field as initial and final state wave functions, as well as the laser dressed Dirac-Volkov propagator. The primary objective is a realistic description of scattering processes with regard to the finite laser pulse duration beyond the common approximation of infinite plane waves, which is made necessary by the ultra-short pulse length of modern high-intensity lasers. Non-linear finite size effects are identified, which are a result of the interplay between the ultra-high intensity and the ultra-short pulse length. In particular, the frequency spectra and azimuthal photon emission spectra are studied emphasizing the differences between pulsed and infinite laser fields. The proper description of the finite temporal duration of the laser pulse leads to a regularization of unphysical infinities (due to the infinite plane-wave description) of the laser-dressed Dirac-Volkov propagator and in the second-order strong-field process of two-photon Compton scattering. An enhancement of the two-photon process is found in strong laser pulses as compared to the corresponding weak-field process in perturbative QED.

Le phénomène de blocage de photon est observé lorsqu'un système à deux niveaux est fortement couplé à une cavité, limitant ainsi le nombre d'occupation des photons dans le mode de la cavité à zéro ou un. Ce phénomène est analogue au blocage de Coulomb en physique mésoscopique et a été observé en optique en couplant un atome unique ou une boîte quantique à une cavité. L'efficacité du blocage, mesurée par la fonction d'autocorrélation du deuxième ordre g2(0) augmente d'autant plus que la non-linéarité du système est grande devant la largeur de la cavité. Ce travail de thèse présente l'étude théorique et expérimentale d'un nouveau phénomène de blocage appelé "phénomène de blocage anormal de photon", dans le régime micro-onde. Ce phénomène apparaît dans un système photonique comprenant deux modes couplés, où au moins un des modes est non-linéaire. Par contraste avec le blocage de photon standard, le blocage anormal permet d'obtenir un blocage parfait (g2(0)=0) pour une non-linéarité arbitrairement faible. Nous présentons les propriétés théoriques du blocage anormal, et notamment sa formulation en termes d'états gaussiens. Afin de mettre en évidence ce phénomène, nous avons réalisé deux résonateurs supraconducteurs couplés, dont l'un est rendu non-linéaire et ajustable en fréquence par l'ajout d'un SQUID. Nous montrons les techniques de fabrication employées et la caractérisation des paramètres de notre échantillon. Enfin, nous présentons les techniques mises en œuvre pour mesurer g2(t). Cette mesure nous a permis de mettre en évidence le phénomène de blocage anormal et d'en étudier les propriétés en fonction des différents paramètres expérimentaux. / Photon blockade is observed when a two level system is strongly coupled to a cavity thus limiting the occupation number of the cavity mode to zero or one photon. This phenomenon is analogous to the Coulomb blockade effect in mesoscopic physics and has been observed in optics by coupling a single atom or a quantum dot to a cavity. The efficiency of the blockade, as measured by the second order auto-correlation function g2(0) increases with the non-linearity of the system in comparison to the cavity width. In this thesis, we present a theoretical and experimental study of a new blockade mechanism, called "anomalous photon blockade effect", in the microwave domain. This effect appears in photonics systems consisting of two coupled modes, where at least one of the mode is non-linear. In contrast to the standard blockade effect, perfect blockade (g2(0)=0) can be achieved with an arbitrary weak non-linearity strength. In the first part, we present a theoretical study of the anomalous blockade, and we use, in particular, a description in terms of gaussian states. To experimentally observe this effect, we have micro-fabricated two coupled superconductive resonators, where one of the resonator is frequency tunable and non-linear thanks to a SQUID. In the second part, we present the fabrication process and the characterization of our sample. Finally, we present the different techniques that we use to measure the auto-correlation function g2(τ). This measurement allowed us to demonstrate the anomalous blockade effect and to study its various properties as a function of the experimental parameters.

Ce travail de thèse apporte une contribution au domaine émergent de la photonique de Josephson en étudiant des corrélations entre photons émis par effet tunnel inélastique à travers une jonction Josephson soumise à une différence de potentiel électrique. Nous démontrons la possibilité de modifier fortement la statistique de ces photons en incorporant la jonction dans un environnement électromagnétique soigneusement conçu. Dans ce contexte, nous avons élaboré et mesuré une source de rayonnement micro-onde de forte intensité, capable d'émettre des photons dont les statistiques dites « de groupement » et « de dégroupement » ne dépendent que d'un seul paramètre modifiable in situ.Afin de réaliser cette expérience, nous avons mis en place un montage de type Hanbury-Brown & Twiss pour mesurer les corrélations entre photons à l'aide d'amplificateurs linéaires dans un cryostat à dilution. De plus, nous avons conçu des circuits micro-onde où la jonction est exposée à des impédances dépendant spécifiquement de la fréquence. Pour réaliser ces dispositifs, nous avons développé un procédé de nano-fabrication de jonctions Josephson verticales à base de nitrure de niobium utilisant de l'oxyde de magnésium comme barrière tunnel. Enfin, en vue d'une meilleure compréhension de nos échantillons, nous avons contribué aux avancées théoriques liées à l'extension de la théorie dite P(E) de l'effet tunnel inélastique de paires de Cooper, dans le but de décrire les corrélations entre événements tunnel.Ces résultats ouvrent la voie, non seulement à une évolution de ces sources de lumière vers le domaine fréquentiel du THz, mais aussi à l'élaboration d'autres dispositifs basés sur la même physique, tels que des détecteurs et des amplificateurs proches de la limite quantique. / This thesis contributes to the emerging field of Josephson photonics through the study of correlations between microwave photons emitted by inelastic Cooper pair tunneling across a voltage-biased Josephson junction. We show that the photon statistics can be strongly modified by embedding the junction into a carefully engineered electromagnetic environment. Doing so, we have elaborated and measured a bright on-demand radiation source, capable of emitting bunched and anti-bunched microwave photons depending only on a single in-situ tunable parameter.In order to conduct this experiment, we have implemented a Hanbury-Brown & Twiss setup for photon correlation measurements using linear amplifiers in a dilution refrigerator. Furthermore, we have designed microwave circuits presenting specific frequency-dependent impedances to the junction. To build these devices we have developed a nano-fabrication process for vertical Josephson junctions made from niobium nitride and using magnesium oxide as a tunnel barrier. Finally, we have contributed to the theoretical advances associated with the understanding of these devices, which extend the so-called P(E) theory of inelastic Cooper pair tunneling to include correlations between tunneling events.These results pave the way for further developments, notably with the possibility to extend the frequency range of these radiation sources to the THz domain but also in view of other devices based on the same physics, such as detectors and amplifiers close to the quantum limit.

Es gibt im Allgemeinen zwei Paradigmen für Entdeckungen in der Teilchenphysik: direkte und indirekte Suchen. Direkte Suchen zielen darauf ab, klare Signale für vermutete Phänomene zu finden, während indirekte Suchen nach Abweichungen zwischen theoretischen Vorhersagen und experimentellen Messungen suchen. Nach dem Nachweis des Higgs-Bosons, wodurch das Standardmodell der Teilchenphysik vervollständigt wurde, haben sich indirekte Suchen als besonders relevant erwiesen, da direkte Nachweise von Physik jenseits des Standardmodells bei aktuellen Energiebereichen unwahrscheinlich sind. Die Herausforderung besteht darin, die Präzision der theoretischen Vorhersagen zu erhöhen, um mögliche Diskrepanzen zu erkennen. Hierbei spielen Gitter-QCD Simulationen für die Berechnung nichtperturbativer hadronischer Observablen eine zentrale Rolle. Für Vorhersagen mit subprozentualer Genauigkeit sind Korrekturen durch Strahlungseffekte und Isospin-Brechung zunehmend wichtig, was durch die Simulation von QCD+QED erreicht wird. Die Einbeziehung von QED stellt neue technische Herausforderungen dar. Diese Arbeit fokussiert sich auf einen Ansatz, der QED-Probleme in endlichen Volumina löst und dabei Eichinvarianz, Lokalität und Translationssymmetrie wahrt, bekannt als QED mit C-Paritäts-Randbedingungen (QED$_C$). Es werden erste umfangreiche QCD+QED$_C$-Simulationen analysiert, darunter acht Eichfeld-Ensembles mit unterschiedlichen Werten der renormierten elektrischen Kopplung, jedoch gleicher Pionenmasse und Gitterabstand. Außerdem wird auf die Einstellung (tuning) der Eingabeparameter für Gittersimulationen eingegangen, um reale physikalische Verhältnisse zu reproduzieren, sowie eine optimierte Strategie mittels Neugewichtung (reweighting) der nackten Quarkmassen im Kontext des RHMC-Algorithmus vorgestellt und evaluiert. / Particle physics research employs two primary approaches for discoveries: direct and indirect searches. Direct searches aim to directly observe phenomena, while indirect searches seek discrepancies between theoretical predictions and experimental results. With the discovery of the Higgs boson, the standard model of particle physics was completed, shifting the focus towards indirect searches due to the lack of compelling evidence for new physics at current energy scales. These searches necessitate highly precise theoretical predictions, particularly for non-perturbative hadronic observables, which are calculated using lattice QCD simulations. The need for sub-percent precision has highlighted the importance of accounting for radiative and isospin-breaking corrections, leading to the simulation of fully dynamical QCD+QED. This thesis addresses the challenges of incorporating QED into lattice QCD, focusing on an approach that maintains gauge invariance, locality, and translational invariance using QED with C-parity boundary conditions (QED$_C$). It presents a comprehensive technical analysis of the first large-scale QCD+QED$_C$ simulations, detailing eight fully dynamical gauge field ensembles with various renormalized electric coupling values ($\alpha_\mathrm{R} \in \{0,1/137,0.04\}$), consistent pion mass ($m_\pi \approx 400$ MeV), and lattice spacing ($a\approx 0.05$ fm). The thesis examines the stability of the simulation algorithm, finite volume effects, and the behavior of different hadron masses. Furthermore, it elaborates on the tuning of input parameters for lattice simulations to replicate real-world physics accurately, focusing on the hadronic renormalization scheme used to fix bare quark masses. It introduces an optimized strategy for tuning QCD+QED parameters via mass reweighting, adapted for simulations using the RHMC algorithm, highlighting its development, implementation, and testing.

Cavity quantum electrodynamics (cavity QED) with single atoms and single photons provides a promising route toward scalable quantum information processing (QIP) and computing. A strongly coupled atom-cavity system should act as a universal quantum interface, allowing the generation and storage of quantum information. This thesis describes the realisation of an atom-cavity system used for the production and manipulation of single photons. These photons are shown to exhibit strong sub-Poissonian statistics and indistinguishability, both prerequisites for their use in realistic quantum systems. Further, the ability to control the temporal shape and internal phase of the photons, as they are generated in the cavity, is demonstrated. This high degree of control presents a novel mechanism enabling the creation of arbitrary photonic quantum bits.

It is possible to reach the radiation-reaction–dominated regime in today’s high-intensity laser facilities, using the collision of a wakefield-accelerated GeV electron beam with a 30 fs laser pulse of intensity 10²² Wcm^-2. This would demonstrate that the yield of high energy gamma rays is increased by the stochastic nature of photon emission: a beam of 10⁹ electrons will emit 6300 photons with energy > 700 MeV, 60 times the number predicted classically. Detecting those photons, or a prominent low energy peak in the electron beam's post-collision energy spectrum, will provide strong evidence of quantum radiation reaction; we place constraints on the accuracy of timing necessary to achieve this. This experiment would provide benchmarking for the simulations that will be used to study the plasmas produced in the next generation of laser facilities. With focused intensities > 10²³ Wcm^-2, these will be powerful enough to generate high fluxes of gamma rays and electron-positron pairs from laser–laser and laser–solid interactions. It will become possible to test the physics of exotic astrophysical phenomena, such as pair cascades in pulsar magnetospheres, and explore fundamental aspects of quantum electrodynamics (QED). To that end we will discuss: classical theories of radiation reaction; QED processes in intense fields; and a Monte Carlo algorithm by which the latter may be included in particle-in-cell codes. The feedback between QED processes and classical plasma dynamics characterises a new regime we call QED-plasma physics.

Oropharyngeal Cancer diagnoses make up three percent of all cancer diagnoses in the United States per year. Recently, there has been an increase in the incidence of HPV-associated oropharyngeal cancer, necessitating updates to prior survival estimation techniques, in order to properly account for this shift in demographic. Clinicians depend on accurate survival prognosis estimates in order to create successful treatment plans that aim to maximize patient life while minimizing adverse treatment side effects. Additionally, recent advances in data analysis have resulted in richer and more complex data, motivating the use of more advanced data analysis techniques. Incorporation of sophisticated survival analysis techniques can leverage complex data, from a variety of sources, resulting in improved personalized prediction. Current survival prognosis prediction methods often rely on summary statistics and underlying assumptions regarding distribution or overall risk. We propose a k-nearest neighbor influenced approach for predicting oropharyngeal survival outcomes. We evaluate our approach for overall survival (OS), recurrence-free survival (RFS), and recurrence-free overall survival (RF+OS). We define two distance functions, not subject to the curse of dimensionality, in order to reconcile heterogeneous features with patient-to-patient similarity scores to produce a meaningful overall measure of distance. Using these distance functions, we obtain the k-nearest neighbors for each patient, forming neighborhoods of similar patients. We leverage these neighborhoods for prediction in two novel ensemble methods. The first ensemble method uses the nearest neighbors for each patient to combine globally trained predictions, weighted by their accuracies within a selected neighborhood. The second ensemble method combines Kaplan-Meier predictions from a variety of neighborhoods. Both proposed methods outperform an ensemble of standard global survival predictive models, with statistically significant calibration.