Abstract

Lithium niobate (LN) is an extremely interesting material for the realization of optical devices and circuits, thanks to its good optical and electro-optical properties. At the state of the art, the use of LN crystals for nonlinear optical functions is limited, despite the very large d₃₃ nonlinear coefficient exhibited by this crystal, because LN is severely affected by the so called “optical damage” due to the photorefractive effect [1]. When a high intensity, spatially inhomogeneous, visible light beam impinges on the LN crystal, a significant and non uniform electric field is produced in the crystal, because of the low crystal photoconductivity, and this yields (through the electro-optic effect) a non-uniform modification of the crystal refractive indices. As a consequence, the efficiency of nonlinear interactions in LN can be greatly reduced, and it becomes necessary to operate the device at high temperature (in some cases even > 200°C), in order to get rid of photorefractivity. In order to operate the nonlinear device at room temperature (or close to it), it is possible to use LN crystals doped with a small quantity of appropriate elements able to significantly increase crystal photoconductivity, like Mg, In, or Sc [1], but it is difficult to obtain high-quality crystals when LN is doped with such ions. Recently it was shown that by using tetravalent ions, like Hf or Zr, the dopant concentration required to significantly reduce the crystal photorefractivity is considerably lower than that required for bivalent and trivalent dopants, even if some uncertainty about the exact required concentration is still present due to the discrepancies between the values reported in the literature. The low dopant-concentration, in combination with the fact that the segregation coefficient of the Zr ion is reported to be very close to 1, should allow growing more easily large homogeneous high-optical-quality Zr:LN crystals.

© 2011 Optical Society of America