Abstract

We observe and explain a strong dependence of energy exchange on the orientaion of the crystal. Experiments are performed with a 1.8-mm BSO crystal of holographic cut (001, 110,$1\overline{1}0$). The crystal is mounted such that it can be rotated around the [110] axis by 360°. Two equal-intensity (≃10mW/cm²) beams a and b are incident onto the (110) plane, each with an angle of 35° to the normal. The angle for maximum energy exchange is often different from the values 0° or 90° usually used. For linear polarization the angularly dependent curves have four maxima and four minima. To explain such behavior we solve two-dimensional amplitude equations where G is optical activity, r is an electro-optic factor proportional to r₁₄, ΔE is the space-charge field, and ϕ is the angle of the light grating relative to the [110] axis. Analytical solutions that include optical activity and both anisotropic and isotropic diffraction agree well with linearly polarized curves (four cases: both along [001], both along [110], left along [001] and right ±45° to [001]) of energy exchange versus crystal rotation angle. However, for circular polarization the data follow the theoretically expected sinϕ dependence, but exhibit dips at 90° and 270° that are not fully understood. Anisotropy in the diffusion field might explain this discrepancy.

© 1988 Optical Society of America