Orientador: Luís Eduardo Evangelista de Araujo / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Física Gleb Wataghin / Made available in DSpace on 2018-08-28T09:20:19Z (GMT). No. of bitstreams: 1 Santana_OctavioJoseSantosde_M.pdf: 3448778 bytes, checksum: 7217b98747d619e694a9b6c7b21e9f36 (MD5) Previous issue date: 2015 / Resumo: Um feixe de luz, após sofrer uma reexão em uma interface plana, é deslocado do caminho previsto pela ótica geométrica. Este desvio do feixe é muito pequeno, tipicamente da ordem do comprimento de onda da luz, que é muito menor do que o tamanho do feixe. Portanto, é difícil de se medir. A técnica de medição fraca óptica tem sido utilizada com sucesso em investigações de deslocamentos de feixe, tal como o deslocamento Goos-Hänchen (GH) e o deslocamento Imbert-Fedorov (IF) (esses são deslocamentos longitudinal e perpendicular ao plano de incidência, respectivamente). Em uma medição fraca, o sistema de medida é projetado sobre um determinado estado nal (pós-seleção), quase ortogonal ao estado inicial (pré-seleção), dando origem a um valor fraco que pode assumir valores muito grandes (amplicados). Nesta dissertação estudamos experimentalmente o deslocamento Goos-Hänchen de um feixe gaussiano focalizado ao sofrer reexão interna total em um prisma via medida fraca. Investigamos este efeito em torno do ângulo crítico c para reflexão interna total. Nosso experimento demonstra pela primeira vez que há uma dependência axial do valor fraco que tem que ser levado em conta, além de um fator fenomenológico na equação de correção do valor fraco. Comparações entre os dados com e sem o fator fenomenológico foi feita, mostrando que ao utilizar esse fator, nossos resultados experimentais mostram um excelente acordo com a previsão teórica / Abstract: A beam of light, after reection from a planar interface, is shifted from the path predicted by ray optics. Such a beam shift is very small, typically of the order of the wavelength of light, which is much smaller than the physical size of the beam. Therefore, it is dicult to measure. The optical weak measurement technique has been successfully used in investigations of beam displacements such as the Goos-Hänchen (GH) and the Imbert-Fedorov (IF) shifts (these are longitudinal and perpendicular to the plane of incidence, respectively). In a weak measurement, the measured system is projected onto a certain nal state (postselected), nearly orthogonal to the inicial state (preselected), giving rise to a measured weak value that may take on very large (amplied). In this dissertation, we study experimentally the Goos-Hänchen shift of a focused Gaussian ligth beam undergoing total internal reection in a prism via weak measurement. We investigate this eect near the critical angle c for total internal reection. Our experiment demonstrates for the rst time that there is an axial dependence of the weak value that has to be taken into account, plus a phenomenological factor in the correction equation of the weak value. Comparisons between the data with and without the phenomenological factor was made, showing that by using this factor, our experimental results show an excellent agreement with the theoretical predication / Mestrado / Física / Mestre em Física / 1186838/2013 / CAPES

In this thesis I discuss the nature of ‘measurement’ in quantum theory. ‘Measurement’ is associated with several different processes: the gradual imprinting of information about one system onto another, which is well understood; the collapse of the wavefunction, which is ill-defined and troublesome; and finally, the means by which inferences about unknown experimental parameters are made. I present a theoretical extension to an experimental proposal from Leggett and Garg, who suggested that the quantum-or-classical reality of a macroscopic system may be probed with successive measurements arrayed in time. The extension allows for a finite level of imperfection in the protocol, and makes use of Leggett’s ‘null result’ measurement scheme. I present the results of an experiment conducted in Oxford that, up to certain loopholes, defies a non-quantum interpretation of the dynamics of phosphorous nuclei embedded in silicon. I also present the theory of statistical parameter estimation, and discover that a recent trend to employ time symmetric ‘postselected’ measurements offers no true advantage over standard methods. The technique, known as weak-value amplification, combines a weak transfer of quantum information from system to meter with conditional data rejection, to surprising effect. The Fisher information is a powerful tool for evaluating the performance of any parameter estimation model, and it reveals the technique to be worse than ordinary, preselected only measurements. That this is true despite the presence of noise (including magnetic field fluctuations causing deco- herence, poor resolution detection, and random displacements), casts serious doubt on the utility of the method.