Depuis de nombreuses années, l'industrie microélectronique s'est engagée dans une course à l'augmentation des performances et à la diminution des coûts de ses dispositifs grâce à la miniaturisation de ces derniers. La génération de ces structures de petites dimensions repose essentiellement sur l'étape de lithographie. Dans cette optique, plusieurs techniques de lithographie nouvelle génération (NGL) sont en cours de développement afin de pouvoir répondre aux besoins de l'industrie pour les nœuds technologiques inférieurs à 20 nm. Parmi elles, les solutions de lithographie à faisceaux d'électrons multiples semblent très prometteuses grâce à leur écriture directe sans masque (ML2), ainsi que leur coût et encombrement réduits. Le CEA-LETI s'est associé à l'entreprise Mapper Lithography basée aux Pays-Bas afin d'aider au développement d'une technologie de lithographie électronique à faisceaux d'électrons multiples basse énergie (d'énergie 5 keV). Les travaux de thèse de ce manuscrit visent à contribuer au développement de cette technologie qui pourrait à terme permettre de réaliser des dispositifs CMOS pour les nœuds technologiques actuels et futurs. L'intégration d'une nouvelle technique de lithographie dans l'industrie repose sur 3 grands critères du procédé lithographique, la production horaire (sensibilité), la résolution (taille minimale des structures réalisées) et la rugosité de ligne. La rugosité de ligne est devenue l'un des paramètres les plus critiques limitant à l'heure actuelle la miniaturisation et pour cause cette dernière impacte de manière négative les performances des dispositifs. Alors que l'ITRS préconise une rugosité de ligne inférieure à 1.7 nm pour les futurs nœuds technologiques inférieurs à 20 nm, les lithographies actuelles ne permettent pas d'obtenir des rugosités inférieures à 4-5 nm. Les travaux de cette thèse visent la minimisation de la rugosité de ligne de résine imprimée par lithographie électronique en proposant des stratégies alternatives d'écriture ou en modifiant les empilements de matériaux sous-jacents la résine, ou encore par l'introduction de traitements post-lithographiques tels que des recuits thermiques ou des traitements plasma. Les études ont montré qu'en combinant une stratégie d'écriture et un traitement plasma à base de dihydrogène une réduction de 41% du LWR pouvait être obtenue. / For decades, the growth of the Semiconductor Industry (SI) has been driven by the paramount need for faster devices at a controlled cost primarily due to the shrinkage of chip transistors. The performances of future CMOS technology generations still rely on the decrease of the device dimensions. However, the photolithography is, today, the limiting factor for pattern miniaturization and the technology has been at a standstill since the development of 193-nm water-based immersion lithography. Moreover, another parameter limiting further semiconductor scaling is the transistor gate linewidth roughness (LWR), i.e. the standard deviation of the gate critical dimension (CD) along the line. The LWR needs to be controlled at the nanometer range to ensure good electrical performances of the future CMOS device. The lithography step is again identified as the root cause of the gate LWR. Indeed, the significant LWR (4-5 nm) of the patterns printed by photolithography is transferred into the gate during the subsequent plasma etching steps, resulting in a final gate LWR far above the sub-2 nm LWR targeted for the sub-20 nm technological nodes. In order to continue scaling down feature sizes of devices, the semiconductor industry is waiting for the maturity of next generation lithographies (NGL). Among NGL, one can find the promising mask-less direct-write techniques (ML2) in which multiple electron beam lithography (multibeam lithography) is regarded as a serious candidate for providing high resolution structures at a low cost. The firm MAPPER Lithography, associated with CEA-LETI is working on the development of such a technology. The aim of this work is to contribute to the development of a low energy (5 keV) multibeam technology and to focus on the improvement of the LWR of the printed patterns. Several process parameters have been investigated to decrease the LWR: the effect of a specific writing strategy, the influence of the under layers and the introduction of post-lithographic treatments such as plasma treatments or thermal annealing. This work has shown that by combining a biased writing strategy with H2 plasma treatment, a 41% LWR decrease could be obtained. Although this performance is still above the ITRS requirements, this work opens the pace for LWR optimization with multi-beam lithography.

The shape of collision-broadened microwave absorption lines is reviewed, along with a number of other broadening mechanisms. The Anderson and Murphy-Boggs linewidth theories are reviewed in detail. Several published modifications to these theories are reviewed. Computer programs which numerically evaluate linewidths and lineshifts are presented. Approximations are made to reduce the need for extensive use of the modified Bessel functions, thereby reducing computation time. Only dipole-dipole forces are considered.

Contributions to the second order collision efficiency function S ⁽²⁾ (b), used in semiclassical perturbation approaches to pressure broadening of microwave and infrared spectra, due to several leading terms, dipole and quadrupole components, in the expansion of the intermolecular interaction energy are derived by method of irreducible spherical tensor operators for molecules of arbitrary symmetry. Results are given explicitly in terms of dipole and quadrupole line strengths. General expressions for dipole moment line strength in the asymmetric rotor basis as well as quadrupole moment line strength for the special case of molecules with two independent quadrupole moment components are derived. Computer programs for calculating linewidth parameters in the rotational spectrum of ¹⁴NO₂ based on Anderson and Murphy and Boggs theories are presented.

Compared to single-mode VCSELs, multimode VCSELs have not received much attention in models and characterizations for functional parameters, despite making up the majority of commercially available VCSELs [1]. In particular, the extraction of the linewidth enhancement factor for multimode VCSELs has been overlooked, likely due to difficulties in measurement. Additionally, multimode models for VCSELs have, until recently, omitted spectral characteristics such as linewidth [2]. This is the first work to report a measured linewidth enhancement factor value (lower bound) for a multimode VCSEL. A characterization for the functional parameters of a red multimode vertical-cavity surface-emitting laser (VCSEL) is shown herein. The extracted values form a complete working set of parameters for the laser rate equations. The techniques employed for extracting values include frequency responses, power versus current fittings, and optical spectral measurements. From the frequency responses at various bias currents, the relaxation oscillation frequency and damping factor are found. The power versus current curve is fitted to find parameters including the modal spontaneous emission rate and carrier density at threshold. The spectral measurements are used for evaluating the linewidth enhancement factor (LEF) also known as the alpha factor or Henry factor. These 5 methods have been applied previously to characterizing single-mode VCSELs [3]–[5]. The experimentally extracted parameters herein are important for creating accurate models and simulations for multimode VCSELs. Improved multimode VCSEL models are necessary for improving optical communication, especially for short-range optical interconnects [2]. The measured parameters for the characterized VCSEL are comparable to similar single-mode VCSELs characterized in other works. This is promising because multi-mode VCSELs have higher output power than their single-mode counterparts, thus these results may aid in improving short-range optical interconnects.

No description available.

Linewidth is a measure of the frequency stability of any kind of oscillator, and it is a defining characteristic of coherent lasers. Narrow linewidth laser technology, particularly in the field of fiber-based infrared lasers, has progressed to the point where highly stable sources are commercially available with linewidths on the order of 1-100 kHz. In order to achieve a higher level of stability, the laser must be augmented by an external frequency stabilization system. This paper presents the design and operation of a frequency locking system for infrared fiber lasers. Using the Pound-Drever-Hall technique, the system significantly reduces the linewidth of an input laser with an un-stabilized linewidth of 2 kHz. It uses a high-finesse Fabry-Perot cavity, which is mechanically and thermally isolated, as a frequency reference to measure the time-varying frequency of the input laser. An electronic feedback loop works to correct the frequency error and maintain constant optical power. Testing has proven the Pound-Drever-Hall system to be highly stable and capable of operating continuously for several seconds at a time. / Master of Science

Due to the recent success, in terms of accuracy and precision, of a number of strontium optical lattice optical frequency standards, and the classification of the 5s² ¹S₀ to 5s5p ³P₀ transition in neutral strontium as a secondary definition of the SI unit of the second, many new strontium lattice clocks are under development. The strontium optical lattice clock (Sr OLC) at the National Physical Laboratory (NPL) is one such project. This thesis describes the design and build of the NPL Sr OLC, discussing the considerations behind the design. Details of the first cooling stage are given, which includes the characterisation of a novel permanent-magnet Zeeman slower by measurements of the longitudinal velocity distributions and loading of the MOT at 461 nm. Development of a narrow linewidth laser system at 689 nm is described, which is used for initial spectroscopy of the second-stage cooling transition. In particular, this work describes progress towards two independent ultra-narrow linewidth clock lasers. The new generation of strontium lattice clock experiments have focused on characterising the systematic frequency shifts and reducing their associated fractional frequency uncertainties, as well as reducing the fractional frequency instability of the measurement. One focus of the Sr OLC at NPL is to help characterise the frequency shift of the clock transition due to black-body radiation (BBR), which is currently the largest contributor to the uncertainty budget of the measured clock frequency. Our approach, discussed here, is to make a direct, differential measurement of the shift with the atoms housed alternately in environments of differing temperatures. Better characterisation and control of the BBR frequency shift of the strontium clock transition is crucial for the future of the Sr OLC as a leading frequency standard.

This thesis describes investigations into the optical and device design of organic light-emitting diodes (OLEDs) with the aim of exploring the factors controlling the spatial emission pattern of OLEDs and developing novel OLEDs with directional emission by applying wavelength-scale photonic microstructure. The development of directional OLEDs was broken down into two steps: the development of efficient narrow linewidth OLEDs and the integration of wavelength-scale photonic microstructures into narrow linewidth OLEDs. The narrow linewidth OLEDs were developed using europium (Eu) complexes. The electrical optimisation of solution-processed Eu-based OLEDs using commercially available materials was investigated. The optimised Eu-based OLEDs gave an external quantum efficiency of 4.3% at a display brightness of 100 cd/m². To our knowledge, this is the highest efficiency reported for solution-processed Eu-based OLED devices, and the efficiency roll-off has been reduced compared with other reported references. Photonic microstructures were applied to develop directional OLEDs using the efficient Eu-based OLEDs. Two contrasting strategies were used. One was to embed photonic microstructures into Eu-based OLEDs, the other was to couple photonic microstructures externally onto the devices. The microstructured devices developed by the former strategy boosted the emitted power in desired angles in both s- and p-polarisations and doubled the fraction of emission in an angle range of 4⁰. The devices developed by the external coupling strategy achieved even higher directionality and the out-coupled emission was a confined beam with easy control of beam steering. Around 90% of the emitted power was confined in an angular range of 20⁰ in the detection plane. The optical properties can be optimised independently without compromising the electrical properties of devices, which gives major advantages in terms of effectiveness and versatility. Optical models were also developed to investigate the out-coupling mechanism of various trapped modes and develop OLEDs with stronger directionality.

<p>This thesis discusses a new scheme for generating radiation in the mid infrared spectral region, especially the 3.5-5 µm range. The scheme uses established Nd³⁺-lasers at 1.06 µm and down conversion in nonlinear optical crystals. The down conversion is made by two optical parametric oscillators (OPO) in series. The second OPO is a classical OPO using a zink germanium phosphide (ZGP) crystal. ZGP is the best nonlinear material available for the 4-8 µm spectral range, but it is absorbing below 2 µm. The new development presented in this thesis is the OPO used to convert the 1.06 µm laser radiation to a suitable OPO pump near 2 µm.</p><p>The OPO uses a type I quasi phase-matched crystal, which accesses high nonlinearities and avoids walk-off. The problem with type I OPOs close to degeneracy is the broad bandwidth of the generated radiation, which reduces the efficiency of a second OPO. This has been solved with a spectrally selective cavity using a volume Bragg grating output coupler. Unlike other bandwidth limiting schemes this introduces no intracavity losses and thus efficient OPO operation is achievable.</p><p>Narrow linewidth (~0.5 nm) OPO operation has been achieved with periodically poled LiNbO₃ (PPLN) and periodically poled KTiOPO4 (PPKTP) while locking the signal wavelength at 2008 nm and simultaneously generating an idler at 2264 nm. A high average power PPLN OPO with 36 % conversion efficiency and 47 % slope efficiency is reported. Operation very close to degeneracy at 2128 nm with the narrowband signal and idler peaks separated by 0.6 nm was demonstrated in a PPKTP OPO. Both the signal at 2008 nm and the combined signal and idler around 2128 nm from the PPKTP OPOs have been used to show efficient pumping of a ZGP OPO. The maximum conversion efficiency from 1 µm to the mid-IR demonstrated is 7 % with a slope efficiency of 10 %. This is not quite as high as what has been presented by other authors, but the experiments reported here have not shown the optimum efficiency of the new scheme. Relatively simple improvements are expected to give a significant increase in conversion efficiency.</p>

Atomically thin, semiconducting transition metal dichalogenides (TMDs), a special class of layered semiconductors, that can be shaped as a perfect two dimensional material, have garnered a lot of attention owing to their fascinating electronic properties which are achievable at the extreme nanoscale. In contrast to graphene, the most celebrated two-dimensional (2D) material thus far; TMDs exhibit a direct band gap in the monolayer regime. The presence of a non-zero bandgap along with the broken inversion symmetry in the monolayer limit brands semiconducting TMDs as the perfect candidate for future optoelectronic and valleytronics-based device application. These remarkable discoveries demand exploration of different materials that possess similar properties alike TMDs. Recently, III-VI layered semiconducting materials (example: InSe, GaSe etc.) have also emerged as potential materials for optical device based applications as, similar to TMDs, they can be shaped into a perfect two-dimensional form as well as possess a sizable band gap in their nano-regime. The perfect 2D character in layered materials cause enhancement of strong Coulomb interaction. As a result, excitons, a coulomb bound quasiparticle made of electron-hole pair, dominate the optical properties near the bandgap. The basis of development for future optoelectronic-based devices requires accurate characterization of the essential properties of excitons. Two fundamental parameters that characterize the quantum dynamics of excitons are: a) the dephasing rate, 𝛾, which represents the coherence loss due to the interaction of the excitons with their environment (for example- phonons, impurities, other excitons, etc.) and b) excited state population decay rate arising from radiative and non-radiative relaxation processes. The dephasing rate is representative of the time scale over which excitons can be coherently manipulated, therefore accurately probing the source of exciton decoherence is crucial for understanding the basic unexplored science as well as creating technological developments. The dephasing dynamics in semiconductors typically occur in the picosecond to femtosecond timescale, thus the use of ultrafast laser spectroscopy is a potential route to probe such excitonic responses. The focus of this dissertation is two-fold: firstly, to develop the necessary instrumentation to accurately probe the aforementioned parameters and secondly, to explore the quantum dynamics and the underlying many-body interactions in different layered semiconducting materials. A custom-built multidimensional optical non-linear spectrometer was developed in order to perform two-dimensional spectroscopic (2DFT) measurements. The advantages of this technique are multifaceted compared to regular one-dimensional and non-linear incoherent techniques. 2DFT technique is based on an enhanced version of Four wave mixing experiments. This powerful tool is capable of identifying the resonant coupling, probing the coherent pathways, unambiguously extracting the homogeneous linewidth in the presence of inhomogeneity and decomposing a complex spectra into real and imaginary parts. It is not possible to uncover such crucial features by employing one dimensional non-linear technique. Monolayers as well as bulk TMDs and group III-VI bulk layered materials are explored in this dissertation. The exciton quantum dynamics is explored with three pulse four-wave mixing whereas the phase sensitive measurements are obtained by employing two-dimensional Fourier transform spectroscopy. Temperature and excitation density dependent 2DFT experiments unfold the information associated with the many-body interactions in the layered semiconducting samples.