Il est possible de modifier en quelques picosecondes les fréquences de résonance d’une microcavité optique semiconductrice en injectant optiquement des porteurs de charge dans le semiconducteur. Dans cette thèse, nous étudions en détail de tels évènements de commutation tout-optique pour des cavités planaires et des cavités en forme de micropilier à base de GaAs/AlAs, en utilisant l’émission de boîtes quantiques intégrées dans ces cavités comme source interne de lumière pour sonder la fréquence des modes résonnants en fonction du temps. Des décalages en fréquence très conséquents, de l’ordre de 34 fois la largeur du mode considéré, sont obtenus après optimisation. Nous réalisons une commutation différentielle des modes d’un micropilier en injectant les porteurs de manière très localisée, et modélisons les comportements observés en prenant en compte la distribution des porteurs injectés ainsi que leur diffusion et leur recombinaison en fonction du temps. Nous étudions par ailleurs deux applications potentielles importantes de la commutation ultrarapide de cavité. D’une part, nous modélisons le changement de couleur qui est induit sur de la lumière piégée dans un mode de cavité lors d’un évènement de commutation. Nous montrons que pour une cavité planaire optimisée, une telle conversion de fréquence peut être réalisée de façon très efficace. D’autre part, la commutation de cavité peut aussi être employée pour contrôler en temps réel l’émission spontanée d’émetteurs intégrés, et plus généralement tous les effets d’électrodynamique quantique en cavité. Nous présentons la génération d’impulsions de lumière incohérente de quelques picosecondes seulement, en utilisant l’émission spontanée de boîtes quantiques dans un micropilier commuté. Nous montrons aussi par une étude théorique qu’il est possible de donner une forme choisie aux impulsions à un photon émises par une boîte quantique, ce qui ouvre des applications intéressantes dans le domaine des liens optiques quantiques et du traitement quantique photonique de l’information. / The resonance wavelengths of semiconductor optical microcavities can be changed within few picoseconds through the optical injection of free charge carriers. In this PhD thesis, we study in detail such “cavity switching” events for GaAs/AlAs planar and micropillar cavities, using the spontaneous emission of embedded QDs as an internal light source to probe the time-dependent frequencies of the cavity modes. Switching amplitudes as large as 34 mode linewidths are observed for optimized pumping conditions. Differential switching of micropillar modes is achieved by performing a localized injection of charge carriers, and modeled by taking into account their injection profile, diffusion and recombination processes. We investigate two important potential applications of cavity switching in the field of quantum optics. On one hand, we model the frequency conversion of light trapped in a cavity mode, which is induced by a switching event, and show that adiabatic and highly efficient frequency conversion can be achieved in properly designed planar cavities. On the other hand, cavity switching appears as a powerful resource to control in real-time the spontaneous emission of embedded emitters and more generally CQED effects. As a first example, we demonstrate the generation of few picosecond short pulses of incoherent light, using the spontaneous emission of switched QD-micropillars. We also show theoretically that cavity switching can be used to shape the time-envelope of single photon pulses emitted by a single QD, which is highly desirable for quantum-optical links and photonic quantum information processing.

Le travail proposé dans cette thèse est basé sur des expérience de corrélation de photons faites sur un émetteur de photon unique semi conducteur: une boite quantique de CdSe dans un nanofil de ZnSe. La première démonstration d'émission de photons unique d'une boite quantique épitaxiée à température ambiante y est présentée. Pour expliquer ce résultat, nous avons étudie expérimentalement et théoriquement l'efficacité de couplage exciton-phonon et ses conséquences sur l'intensité de l'exciton avec la température. Nous présentons également des résultats optiques portant sur la robustesse de cette structure à haute température. La technique de corrélation de photons est également appliquée sur des boites quantiques chargées. la présence du exciton chargé nous a permis de sondé la structure fine du trio excité, de décrire ses processus de relaxations et d'obtenir une mesure direct du temps de spin flip du trou sur l'état p. Des indication sont également données sur la nature possible du dopage. Nous avons aussi étudié la diffusion spectrale de l'émetteur causée par les fluctuations électroniques de son environnement. Par un travail théorique nous montrons comment interpréter l'effet de l'élargissement phonon de la raie homogène, (processus poissonien) combiné avec l'effet de la diffusion spectral (processus markovien) sur la fonction de corrélation de la demi-raie. Grâce à l'expérience, nous concluons sur la statistique de l'énergie d'émission de l'émetteur à haute température. Nous appliquons cette théorie sur les nanofils et interprétons les dépendances en température et en puissance des fluctuations de l'environnement grâce au modèle de Kubo-Anderson. / The work proposed in this thesis is based on photon correlation experiments performed on a semi-conductor single photon emitter: CdSe/ZnSe nanowire quantum dot. Is presented the first demonstration of single photon emission at room temperature from an epitaxied quantum dot. To explain this result we investigated by a theoretical and experimental study, the exciton-phonon coupling efficiency and its consequence on the exciton luminescence intensity with temperature. We also present optical results on the robustness against temperature of this structure. Photon correlations techniques are also applied on charged quantum dots. Presence of the charged biexciton allowed to probe the fine structure of the excited trion, to describe its carrier relaxation processes, and to obtain a direct measurement of the p-shell hole spin flip time. Indications are also given on the possible doping nature. We also investigated spectral diffusion of the emitter caused by electronic fluctuations of the environnement. By a theoretical work, we show how to model the effect of the homogeneous phonon broadening, (poissonian emission energy process) combined with the spectral diffusion effect (markovian emission energy process) on the half line autocorrelation function. Thanks to experiments, We conclude on the statisic of the emission energy of the emitter at high temperature. We apply this theory on CdSe/ZnSe nanowire quantum dots and interpret temperature and power dependance of the environnement fluctuation thanks to the Kubo-Anderson Model.

Les plasmons de surface (SP) sont des modes du champ électromagnétique confinés à l'interface entre un métal et un diélectrique. De par leur nature hybride, les SP permettent de concentrer et manipuler la lumière à des échelles sub-longueur d'onde. Ces propriétés sans précédent suscitent un grand intérêt, en particulier pour le transport et le traitement de l'information quantique mais aussi pour le contrôle de l'émission spontanée d'émetteurs fluorescents. Les études présentées dans ce manuscrit s'intéressent au couplage de nanostructures plasmoniques avec des nanoparticules luminescentes. L'outil utilisé est un microscope optique en champ proche (SNOM) dans lequel la nano-source de lumière est un nano-objet fluorescent attaché en bout de pointe (sonde active). Cette technique permet à la fois d'augmenter la résolution théorique accessible en SNOM mais aussi de positionner la sonde avec une précision nanométrique et de l'exciter directement grâce à la lumière laser injectée dans la fibre optique. En utilisant uniquement la lumière émise par l'objet, ces pointes ouvrent la voie à des études originales en nano-optique et en plasmonique. Dans ce travail de thèse, deux aspects distincts ont été abordés. D'une part, nous avons étudié les propriétés des plasmons de surface dans le régime de la plasmonique quantique en utilisant pour cela une sonde active fabriquée à base d'un émetteur de photons uniques, le centre NV (nitrogen-vacancy) contenu dans les nano-diamants. Les résultats fondamentaux obtenus sur ce système permettent d'envisager de nombreuses expériences en plasmonique quantique. D'autre part, le travail de développement des sondes actives à base de nanocristaux de YAG (yttrium-aluminum garnet) dopés au cérium a été poursuivi. Ces sondes nous ont permis de démarrer de nouvelles études sur les résonances plasmoniques localisées de particules colloïdales en or. / Surface plasmons (SPs) are modes of the electromagnetic field confined at the interface between a metal and a dielectric. Due to their hybrid nature, the SPs can be used to concentrate and handle light on subwavelength scales. These unprecedented properties draw great interest, in particular for quantum information transport and processing and also for the control of spontaneous emission of fluorescent emitters. The studies presented in this manuscript report the coupling of plasmonic nanostructures with luminescent nanoparticles. The tool we use is a scanning near-field optical microscope (SNOM), in which the nano-source of light is a fluorescent nano-object attached at the end of the probe (active tip). This technique allows not only to reach a better optical resolution in SNOM but also to position the nano-emitter with a nanometre precision and to excite it directly thanks to the laser light injected into the optical fibre. By using only the light emitted by the object, these tips open the way to original studies in nano-optics and plasmonics. In this work, two distinct aspects were studied. First, we studied the properties of the SPs in the quantum plasmonics regime. For this purpose, we used an active tip based on single photons emitters which are the NV centres (nitrogen-vacancy centre) hosted in nanodiamonds. The fundamental results obtained on this system make it possible to consider many other quantum plasmonics experiments. In addition, a different type of active tips based on Cerium-doped YAG (yttrium-aluminum garnet) nanoparticules was developed. These probes allow us to start new studies on localised plasmonic resonances in colloidal gold particles.

INTRODUÇÃO: O diagnóstico da doença de Parkinson (DP) pode ser um desafio, principalmente nas fases precoces da doença. O diagnóstico acurado desta condição requer mais que a avaliação clínica isolada. A Tomografia computadorizada do crânio de fóton único (SPECT) e a ultrassonografia transcraniana (USTC) podem ser úteis na diferenciação entre a DP e as síndromes parkinsonianas atípicas ou entre a DP e o tremor essencial. O presente estudo objetivou investigar a acurácia da USTC combinada com o teste de olfação Sniffin\' Sticks (SST-16) para diferenciar pacientes com DP de controles saudáveis e comparar com a acurácia do SPECT com 99mTc- TRODAT-1 (TRODAT). MÉTODOS: Trata-se de um estudo transversal que incluiu pacientes com DP segundo critérios do United Kingdom Parkinson\'s disease Society e um grupo controle de indivíduos saudáveis pareados para idade e gênero. Os pacientes foram examinados por um especialista em distúrbios do movimento e submetidos a SPECT encefálico com TRODAT, USTC e SST-16. Curvas Receiver Operating Characteristic (ROC) foram obtidas para definir os pontos de corte dos métodos avaliados para detecção de DP. RESULTADOS: Vinte indivíduos com DP (13 homens e 7 mulheres) e 9 participantes saudáveis foram admitidos no estudo. A idade mediana de início dos sintomas foi de 56,5 anos e a mediana do tempo de duração da doença foi de 5 anos. Maior área de ecogênica da substância negra (SN) foi observada no grupo com DP (p=0,013). Área ecogênica da SN de 0,22 cm2 foi definida pela curva ROC para detecção de DP, com acurácia de 79%. O ponto de corte do potencial de ligação do TRODAT no striatum foi 0,90, com acurácia de 99% para o diagnóstico de DP. Escore do SST-16 maior ou igual a 10 pontos foi o ponto de corte para detecção de DP, com acurácia de 85,8%. A combinação da USTC com teste da olfação levou à acurácia de 95% para detecção de DP. CONCLUSÃO: A combinação da USTC com SST-16 eleva a capacidade de ! detecção da DP. A acurácia da USTC combinada ao SST-16 para identificar pacientes com DP idiopática aproximou-se da acurácia do SPECT com TRODAT / INTRODUCTION: Diagnosing Parkinson\'s disease (PD) can be challenging, especially in the early stages of the disease. An accurate diagnosis requires more than clinical findings alone. Brain single-photon emission computed tomography (SPECT) and transcranial sonography (TCS) are helpful for diagnosing PD and differentiating it from atypical parkinsonian syndromes as well as essential tremor. This study aimed to investigate the accuracy of TCS combined with the Sniffin\' sticks olfactory test (SST-16) for differentiation between idiopathic PD patients and healthy controls compared to that of 99mTc-TRODAT-1 SPECT (TRODAT). METHODS: A cross-sectional study included PD patients diagnosed in accordance with United Kingdom PD Society Brain Bank criteria and a control group of age and sex-matched healthy subjects. All patients were examined by a movement disorder specialist and underwent brain SPECT using TRODAT, TCS examination and SST-16 test. Receiver Operating Characteristic (ROC) curves were used to calculate cut-off points for TCS, Striatal TRODAT binding potentials and SST-16. The area under the ROC curve determined the accuracy of the method. RESULTS: Twenty patients with PD (13 males and 7 females) and nine healthy subjects were included. Median age of PD onset was 56.5 years with median disease duration of 5 years. A larger substantia nigra (SN) echogenic area was observed in the PD group (p=0.013). SN echogenic area cut-off point of 0.22 cm2 was obtained from a ROC curve for PD diagnosis. Considering this cut-off point, TCS accuracy was estimated at 79.2% for PD diagnosis. The cut-off value of 0.90 for striatal TRODAT binding was associated with 99% accuracy for the diagnosis of PD. SST-16 values equal or greater than 10 points showed a 85.8% accuracy for PD diagnosis. Combination of both SST-16 and TCS improved the accuracy to 95% for PD diagnosis. CONCLUSION: Combined assessment of SST-16 and TCS are reliable and highly accurate for distinguishing PD patients from healthy controls. The accuracy of TCS combined with SST-16 for differentiation between idiopathic PD patients and healthy controls is similar to that of SPECT TRODAT

Abstract In the Nordic countries, event-free survival (EFS) exceeds 80% in certain groups of children treated for acute lymphoblastic leukaemia (ALL). With the improved cure rates, however, there are more children suffering from neurological late effects, especially due to therapy directed at the central nervous system (CNS). The aim of this study is to examine the changes taking place in the nervous system after leukemia treatment and to evaluate the role of treatment in these changes in patients and in an animal model. Twenty-seven ALL survivors and healthy controls were examined by means of motor evoked potentials (MEPs). ALL survivors were also examined clinically. The children with ALL continued to show decreased motor nerve conduction in the peripheral nerves, but not within the CNS, five years after the cessation of treatment. Clinical neurological findings were obtained in 33% of the cases. The MEP results indicated reversibility of the motor injury due to CNS effects. Nineteen patients underwent perfusion magnetic resonance imaging (MRI) at the cessation of treatment or 4-8 years after the treatment. Seventeen of them also underwent single-photon emission computed tomography (SPECT). The studies showed small perfusion defects in SPECT, which were not visible by perfusion MRI. Methotrexate (Mtx) neurotoxicity was studied in a swine model using functional MRI, brain perfusion SPECT, iodine-123 labelled 2β-carbomethoxy-3β-(4-iodophenyl) tropane ([123I]β-CIT) SPECT and whole-hemisphere autoradiography with [125I]β-CIT in ten Mtx-treated animals and five control animals. Mtx-related changes in the brain could be detected as reduced or negative blood-oxygen-level-dependent (BOLD) responses to somatosensory activation in BOLD contrast MRI, which indicates changes in flow metabolism coupling. Perfusion defects in brain SPECT were seen in the Mtx group and the control group, which suggests that the perfusion defects seen in brain SPECT are probably multifactorial. The change in dopamine transporter (DAT) density in the Mtx group was not different from that in the controls. The abnormalities in nerve conduction after treatment in survivors of ALL were partly reversible years after the treatment. The patients had perfusion defects in SPECT imaging which were not seen in perfusion MRI. The clinical significance of these defects remains obscure. The animal model suggested perfusion defects to be multifactorial.

Gegenstand dieser Diplomarbeit ist die Untersuchung der Dynamik in dünnen Polymerfilmen anhand von einzelnen Molekülen. Zu diesem Zweck wurden dünne Filme (kleiner 100nm) hergestellt und mittels Einzelmoleküldetektion und zeitaufgelöster Einzelphotonenzählung analysiert, was eine orts- und zeitaufgelöste Untersuchung einzelner Farbstoffmoleküle ermöglicht. Ziel war es, festzustellen ob, und auf welchem Weg, die Dynamik der Umgebung in Fluktuationen der Fluoreszenzlebensdauer einzelner Moleküle sichtbar wird. Neben der Evaluierung der Untersuchungsmethoden wurden in dieser Arbeit zwei Arten von Sensormolekülen - DiD und Malachit Grün - näher untersucht. / The subject of this diploma thesis is the analysis of dynamics in thin polymer films using single molecules. Thin polymer films (less than 100nm) were produced and analysed by Single Molecule Detection (SMD) and Time Correlated Single Photon Counting (TCSPC). This allows a spatial and time resolved investigation of the single dye molecule. The aim was to ascertain if, and in which way, the dynamics of the environment are reflected by fluctuations of the fluorescence lifetime of the single molecule. In addition to evaluating the investigation methods two kinds of molecules - DiD and Malachite Green - were also analysed.

[pt] Detectores de fótons únicos baseados em fotodiodos de avalanche (SPADs) são essenciais em aplicações que requerem alta resolução, como comunicações quânticas e metrologia quântica. O efeito da largura de janela de detecção temporal de fótons é pouco explorado, e não há estudos para a faixa de comprimentos de onda de interesse em telecomunicações em torno de: 1550 nm. Neste trabalho, apresenta-se uma proposta para análise de impacto da largura de janela de detecção de um SPAD de InGaAs/InP, realizando uma análise da estatística entre detecções consecutivas e da probabilidade de detecção de 0 ou 1 evento em função da atenuação óptica. Variou-se o número médio de fótons por janela medido pelo SPAD, e os resultados foram avaliados para os valores de 4 ns, 8 ns, 12 ns, 16 ns e 20 ns de largura de janela de detecção, sendo estimada a Incerteza de Medição Expandida para cada ensaio. Os resultados obtidos indicam uma faixa adequada de potência óptica para calibração de um SPAD com eficiência de detecção de 15 porcento e um tempo morto de 1 microssegundo, no intervalo de 10 nW a 0,15 nW. Nesta faixa de potência, os respectivos produtos associados ao efetivo número médio de fótons por janela de detecção correspondem aos valores de 190 x 10-(4) a 0,32 x 10(-4) (para 4 ns) e 140 x 10(-4) a 2,9 x 10(-4) (para 8 ns). Foram obtidos comportamentos lineares para os ajustes das curvas de calibração para larguras de janela de 4 ns e 8 ns. / [en] Single photon detectors based on avalanche photodiodes (SPADs) are essential in applications that require high resolution, such as quantum communications and quantum metrology. The effect of the width of photon detection gate is little explored, and there are no studies for the wavelength range of interest in telecommunications around 1550 nm. In this work, a proposal is presented for analyzing the impact of the detection gate width of an InGaAs/InP SPAD, performing a statistical analysis of consecutive detections and the probability detection of 0 or 1 events depending on the optical attenuation. The average number of photons per gate measured by the SPAD was varied, and the results were evaluated for the values of 4 ns, 8 ns, 12 ns, 16 ns and 20 ns of detection gate widths, and Expanded Measurement Uncertainty was estimated for each test. The results obtained indicate an adequate optical power range for calibrating a SPAD with a detection efficiency of 15 percent and dead – time of 1 microssecond, in the range of 10 nW to 0,15 nW. In this power range, the respective products, which are associated with an effective average number of photons per gate window, correspond to the values of 190 x 10(-4) to 0,32 x 10(-4) (for 4 ns) e 140 x 10(-4) to 2,9 x 10(-4) (for 8 ns). Linear behaviors were obtained for the adjustment of the calibration curves for gate widths of 4 ns and 8 ns.

In this project, the fabrication of single-photon detectors based on superconducting nanowires is presented, with great focus on extending their operation range to the mid infrared. In particular, Niobium Titanium Nitride (NbTiN) and Molybdenum Silicide (MoSi), superconducting materials with different properties, are presented, studied and used as fabrication platforms. Different approaches are followed, mainly adjusting the nanowire width and thickness to achieve near unity quantum efficiency at mid infrared wavelengths. With the vision of using these devices for atmospheric LIDAR and sensing experiments, saturation at 2050 nm is studied that corresponds to the absorption peak of CO2. For the best device made on NbTiN thin films, unity quantum efficiency is shown at 2050 nm with a time jitter of 116 ps at 1550 nm. Simulations using the transfer matrix method and the commercial software Lumerical are carried out, concluding that the devices made in NbTiN could have 23.1-26.7% system detection efficiency at 2050 nm on a Silicon SiO2/Si platform. Further improvements show that the detection efficiency could reach between 52-62% (for 0.33 and 0.5 fill factor, respectively calculated with FDTD simulations) by engineering optical cavities. / I detta projekt presenteras en fabrikations process för enstaka foton detektorer baserade på supraledande nanotrådar. Fokuset har legat på att utöka våglängds regionen där detektorernas kan detektera till mid-infrarött ljus. Två specifika supraledande material, Niobium Titan (NbTiN) och Molybdenum Silicide (MoSi), med olika egenskaper har studerats och använts som material. Dimensionerna på nanotrådarna, framför allt tjockleken och bredden, har optimerats för att uppnå nära enhetlig kvant-effektivitet vid mid-infraröda våglängder. Med visionen att detektorerna ska användas för atmosfäriska LiDAR mätningar har de studerats för satruering vid 2050 nm som motsvarar ett absorbtions maximum för CO2. Detektorerna tillverkade med NbTinN uppnådde 100% kvant effektivitet för 2050 nm ljus med ett tids jitter på 116 ps vid 1550 nm ljus. Simuleringar med överförings matrisen metoden och den kommersiella mjukvaran Lumerical visar att NbTiN detektorer placerade på en SiO2/Si platform kan ha en 23.1-26.7% effektivitet vid 2050 nm. Ytterligare simuleringas visar att effektiviteten kan nå upp till 52-62% (för 0.33 och 0.5 fyllnadsfaktor, respektive beräknad med FDTD) genom att inkludera optiska kaviteter.

Using shallow implantation of ions and molecules with masses centred at 27 atomic mass units(amu) in diamond, a new artificial optical centre with unique properties has been created. The centre shows a linearly polarised fluorescence with a main narrow emission line mostly found at 582 nm, together with a weak vibronic sideband at room temperature. The fluorescence lifetime is∼2 ns and the brightest centres are more than three times brighter than the nitrogen-vacancy centres. A majority of the centres shows stable fluorescence whereas some others present a blinking behaviour, at faster or slower rates. Furthermore, a second kind of optical centre has been simultaneously created in the same diamond sample, within the same ion implantation run. This centre has a narrow zero-phonon line (ZPL) at∼546 nm and a broad phonon sideband at room temperature. Interestingly, optically detected magnetic resonance (ODMR) has been measured on several single 546 nm centres and two resonance peaks are found at 0.99 and 1.27 GHz. In view of their very similar ODMR and optical spectra, the 546 nm centre is likely to coincide with the ST1 centre, reported once (with a ZPL at 550 nm), but of still unknown nature. These new kinds of centres are promising for quantum information processing, sub-diffraction optical imaging or use as single-photon sources.

In recent years, the usage of 3D deep learning techniques has seen a surge,mainly driven by advancements in autonomous driving and medical applications.This thesis investigates the applicability of existing state-of-the-art 3Ddeep learning network architectures to dense voxel grids from single photoncounting 3D LiDAR. This work also examine the choice of loss function asa means of dealing with extreme data imbalance, in order to segment peopleand vehicles in outdoor forest scenes. Due to data similarities with volumetricmedical data, such as computer tomography scans, this thesis investigates ifa model for 3D deep learning used for medical applications, the commonlyused 3D U-Net, can be used for photon counting data. The results showthat segmentation of people and vehicles is possible in this type of data butthat performance depends on the segmentation task, light conditions, and theloss function. For people segmentation the final models are able to predictall targets, but with a significant amount of false positives, something that islikely caused by similar LiDAR responses between people and tree trunks.For vehicle detection, the results are more inconsistent and varies greatlybetween different loss functions as well as the position and orientation of thevehicles. Overall, we consider the 3D U-Net model a successful proof-ofconceptregarding the applicability of 3D deep learning techniques to this kindof data. / Under de senaste åren har användningen för djupinlärningstekniker för 3Dsett en kraftig ökning, främst driven av framsteg inom autonoma fordon ochmedicinska tillämpningar. Denna avhandling undersöker befintliga modernadjupinlärningsnätverk för 3D i täta voxelgriddar från fotonräknande 3D LiDARför att segmentera människor och fordon i skogsscener. Vidare undersöksvalet av målfunktion som ett sätt att hantera extrem dataobalans. På grundav datalikheter med volymetriska medicinska data, såsom datortomografi,kommer denna avhandling att undersöka om en modell för 3D-djupinlärningsom används för medicinska applikationer, nämligen 3D U-Net, kan användasför fotonräknande data. Resultaten visar att segmentering av människor ochfordon är möjligt men att prestanda varier avsevärt med segmenteringsuppgiften,ljusförhållanden, och målfunktioner. För segmentering av människorkan de slutgiltiga modellerna segmentera alla mål men med en betydandemängd falska utslag, något som sannolikt orsakas av liknande LiDAR-svarmellan människor och trädstammar. För segmentering av fordon är resultatenmer oberäkneliga och varierar kraftigt mellan olika målfunktioner såväl somfordonens position och orientering. Sammantaget anser vi att 3D U-Netmodellenvisar på en framgångsrik konceptvalidering när det gäller tillämpningav djupinlärningstekniker för 3D på denna typ av data.