UBC Theses and Dissertations
The photorefractive effect in Lithium Niobate Cornish, William D
Exposure of the insulating ferroelectric crystal, lithium niobate, to light of the appropriate wavelength causes small changes in the refractive indices. This phenomenon which has recently been named the photorefractive effect allows phase holograms to be stored in the crystal. The work described in this thesis was undertaken to obtain an understanding of the mechanisms of the photorefractive effect in connection with possible engineering applications. The process is thought to involve the spatial redistribution of photo-excited electrons among traps. Space charge fields develop which modulate the refractive indices through the electro-optic effect. Initially, the mechanisms proposed for charge transport were diffusion and drift in an internal field of pyroelectric origin. Using these mechanisms, Amodei had treated the initial development of phase holograms on the assumption that the electron transport length was short. A theoretical treatment without the restriction of short transport length is presented which shows that the efficiency of hologram writing increases for increased transport length up to a certain limit. In addition, it is shown that the resolution of the recording medium is not limited by increased transport length. More recently, Glass, von der Linde and Negran have proposed a new phenomenon, the bulk photovoltaic effect, as being responsible for charge transport. Photocurrent measurements are presented which provide further evidence for the existence of this effect. The relative contributions of drift, diffusion and the bulk photovoltaic effect to the photorefractive process are investigated by applying a field during hologram formation. It is found that the effects of positive and negative applied fields are not symmetric. The degree of asymmetry depends on what fraction of the crystal is illuminated. It is also found that both the voltages applied during previous exposures and voltages applied during the current exposure influence the diffraction efficiency. It is thought that these effects are caused by large scale space charge fields which are produced by exposure to light. The development of these space charge fields is discussed. It is concluded that the holograms were written by a combination of diffusion, drift in applied and space charge fields, and the bulk photovoltaic effect. The importance of multiple internal reflections between the faces of the crystal had not previously been considered. It is shown that in measuring the photorefractive sensitivity by holographic, ellipsometric and adjustable-compensator techniques, neglecting multiple reflections may cause serious errors. Two methods of probing large scale changes in the refractive indices are outlined. In the first method an automated ellipsometer was modified and programmed to measure the birefringence of the lithium niobate crystals and the change in birefringence due to illumination. This method has yielded information on the extent of optically-induced space charge fields, the uniformity of the crystal and the effects of heating crystals under different conditions. A second and more rapid method of inspecting large scale changes in the refractive indices is based on making the crystal act as a Fabry-Perot interferometer. The sensitivity of photorefractive crystals to light has been associated with impurities and defects in the crystal lattice. Methods of modifying the valence state of iron impurities are important since the photorefractive sensitivity is dependent on the amount of Fe²⁺ present in the crystal and on the ratio of Fe³⁺ to Fe²⁺. One of the methods used to convert Fe³⁺ to Fe²⁺ is heating the crystal in lithium carbonate. It is shown that this treatment, in addition to reducing iron impurities, changes the birefringence of the crystal and reduces the rate at which space charge fields decay. It is suggested that the last effect is caused by the destruction of shallow traps. A further charge transport mechanism has. been suggested in which intervalence transfer of electrons between Fe²⁺ and Fe³⁺ impurity states occurs. It is argued that if, instead, electrons enter the conduction band, luminescence should be observable. It is shown that a photoluminescence band which is associated with iron impurities in the crystal may be observed in the region of 770 nm.
Item Citations and Data