Thermal poling of multioxide silicate glasses and ion-exchanged waveguides
This Thesis reports a theoretical and experimental study of thermal poling of glasses in which a second order optical nonlinearity is introduced into the glass by applying a potential across the glass at elevated temperature. Thermal poling is the most reported and reproducible procedure for the introduction of the nonlinear susceptibility of 2nd order, c(2), in glasses. The attainment of c(2) of the order of 1pm/V has been reported in a wide range of silica based glasses including glasses suitable for ion exchange, UV-writing and rare earth doping. Effective c(2) of 0.1 pm/V has been demonstrated in poled channel waveguides in silica indicating a poor overlap between c(2) and the waveguide modes. These reported values of c(2) must be increased at least one order of magnitude for glasses and two orders of magnitude for waveguides for practical use. Therefore a good understanding of the poling mechanism which was unclear at the beginning of this work was required for poling optimisation. In this Thesis a poling model based on electrostatics and ion transport theory is developed yielding a method for the evaluation of glasses for poling. A new technique for simultaneous poling and waveguide fabrication by differential ionic drift in glasses that contains more than one mobile ion is demonstrated and poling of K+ ion-exchanged waveguides is achieved. c(2) of the order of 1 pm/V was verified in poled soda lime glass and in found in poled potassium ion-exchanged soda lime glass. A poling time for multioxide glasses some 5 times shorter and minimum temperature 50oC lower than reported in the literature was achieved with constant-current thermal poling in vacuum. A procedure to evaluate the average value, thickness and the location of the c(2) region is established. This study provides an improved understanding of the poling mechanism and may contribute to the achievement of higher c(2) in poled glasses and poled waveguides in glasses.