With a growing demand in industry for cost effective, increased data handling capabilities great attention has been paid to the study of various polymer systems for use in optical telecommunications. Inorganic crystals, currently used in such systems, have high performance, but are more expensive and less obtainable than organic materials. Recent advances in techniques for developing highly efficient and inexpensive organic polymeric electro-optic (EO) devices compatible with current state-of-the-art electronics have created an interest in the commercialization of such electro-optic devices. In light of the many advantages of utilizing organic materials for electro-optic applications, numerous methods have been developed to produce nonlinear optically (NLO)-active polymeric films for such purposes. Ionic self-assembled multilayer (ISAM) films are a recently developed class of materials that allows detailed structural and thickness control at the molecular level, combined with ease of manufacturing and low cost. However, the layer-by-layer deposition technique utilized for this method currently requires lengthy processing times that challenge the feasibility of fabricating a thick film suitable for EO modulator device fabrication. This study focuses on addressing the influence of several pertinent processing variables affecting these challenges for application to electro-optic device fabrication. This study investigated (1) the effect of forced convection, varying deposition time and varying dye concentration on the properties of PAH/Procion Brown films fabricated via the hybrid reactive deposition scheme, (2) the automation and optimization of the fabrication of thick NLO active films and (3) the use of the hybrid covalent-electrostatic deposition scheme to fabricate a polymeric waveguide device with an electro-optic coefficient comparable to that of lithium niobate (LiNbO₃). At fixed deposition time and concentration conditions, the presence of convection had little demonstrated effect on films with deposition times shorter than 2 minutes. For the 5 minute case, the presence of convection correlated with a ~45% increase in Ï (2)zzz values values and a 25% increase in absorbance per bilayer. At a constant dye concentration of 5 mg/ml, the deposition time had little effect on SHG for deposition times less than two minutes. In the presence of convection, the increase in deposition time from 2 minutes to 5 minutes showed a 57% increase in Ï (2)zzz values and a 30% increase in absorbance per bilayer. For a deposition time of 2 minutes in the presence of convection, the dye solution concentration was successfully reduced 5-fold (from 5 mg/ml to 1 mg/ml) with less than a 5% difference in Ï (2)zzz values, less than a 15% decrease in absorbance per bilayer and no detriment to film quality. These results strongly indicate that the deposition conditions remain well outside of the transport-limited regime at a dye concentration of 1 mg/ml. Rather, the surface reaction rate apparently is controlling. Depositing slides at an elevated temperature (~35°C), had an undetermined effect on Ï (2)zzz values, but showed a 15% increase in absorbance per bilayer. An automatic dipper was programmed to replicate the current manual deposition method to fabricate a film suitable for EO modulator devices. Utilizing the optimal conditions for the processing variables, an optically-homogeneous, 100 nm-thick film was fabricated utilizing the automated process, yielding a Ï (2)zzz values~ 23 x 10⁻⁹ esu. A three-layer coplanar electro-optic device was fabricated utilizing the hybrid reactive deposition method. For this device, the presence of added salt was found to increase the electro-optic coefficient r33 by a factor of 3 compared to its value when made with no added salt. The electro-optic coefficient of the added salt case was found to be about 1/2 that of lithium niobate (LiNbO3). / Master of Science

The photorefractive crystals of the sillenite family are very interesting because their large potentiality of applications in devices development, such as non‐destructive interferometer tests. Among the crystals of this family, the Bi12TiO20 (BTO) is more attractive because their relatively fast response time and the lower optical activity, being promising for application in real time image processing. The effective electro‐optic coefficient ( reff ) is a parameter too important for application of the BTO. The wavelength dispersion of this coefficient in the visible spectrum has attracted some discussion in the literature. Another important parameter is the electric screening field coefficient (ξ ). This parameter is necessary to take into account the material response to the applied electric field. Because the material is photoconductor and due to nonuniform illumination, the material responds creating an electric field, called screening field ( E scr ), that is opposed to the applied ones. In this work, we present an alternative procedure to determine ef r and ξ by measuring the optical intensity variation induced by an applied electric field ( E App ), which is transmitted through an undoped photorefractive 12 20 Bi TiO (BTO) crystal in an oblique incidence setup. The transmitted intensity variation (TIV) was modeled taking into account the transmission coefficients for the polarization plane parallel and perpendicular to the incidence plane, as well as the birefringence induced by the E App, which changes the components of the polarization vector of the light beam. The measurements were performed with infrared radiation at 780 nm and provided 5.5 0.2 pm/V reff = ± . It allowed us to conclude, with support of results from the literature, that the region without dispersion of the electro‐optic coefficient in the BTO crystal ranges from 510 nm to at least 780 nm. The results point to the existence of an intensity threshold inside the crystal to create a significant E scr. The procedure proposed here can be used for distinct wavelengths and seems to be suitable for other electrically induced birefringent materials. / Os cristais fotorrefrativos da família das silenitas são de grande interesse tecnológico devido as suas grandes potencialidades de aplicações no desenvolvimento de dispositivos, como em testes interferométricos não‐destrutivos. Dentro dessa família, o Bi12TiO20 (BTO) se destaca por apresentar tempo de resposta relativamente curto e a menor atividade óptica, sendo promissores para aplicação no processamento de imagens em tempo real. Um parâmetro muito importante para aplicação do BTO é o coeficiente eletro‐óptico efetivo ( ref ). A dispersão desse coeficiente na região visível do espectro tem atraído discussão considerável. Outro parâmetro importante é o coeficiente de blindagem do campo elétrico aplicado (ξ ). Esse parâmetro surge devido à resposta do meio a aplicação de campo elétrico. Pelo fato do material ser fotocondutor, o meio responde criando um campo elétrico interno, chamado de campo elétrico de blindagem (E scr), por conta da não‐uniformidade na iluminação e contatos elétricos. Neste trabalho, propomos um procedimento alternativo para a determinação de ref e ξ . A determinação desses parâmetros é realizada medindo a variação da intensidade transmitida (VIT) do feixe óptico pelo cristal em incidência oblíqua. Esta variação é provocada pelo campo elétrico aplicado (E Ap) e se deve a birrefringência induzida no cristal que muda as componentes do vetor polarização do feixe. A VIT foi modelada levando em conta os coeficientes de transmissão para polarização paralela e perpendicular ao plano de incidência, que no caso de incidência oblíqua são diferentes, como também a birrefringência induzida no cristal por E Ap. Foi investigado um cristal BTO e as medidas foram realizadas em 780 nm. O valor obtido para ref foi 5,5±0,2 pm/V o que nos permitiu concluir, com suporte de resultados da literatura, que a região a qual ef r não apresenta dispersão para os cristais BTO é de 510 nm até pelo menos 780 nm. Verificamos que o parâmetro ξ é dependente da intensidade do feixe incidente, do ângulo de incidência e do ângulo de polarização. Os resultados obtidos indicam a existência de um limiar de intensidade dentro do cristal para a criação de E scr significante. Apesar das medidas terem sido realizadas em apenas um cristal e um comprimento de onda, o procedimento proposto pode ser utilizado em outros comprimentos de onda e outros meios onde a birrefringência é induzida por E Ap.