Study of the systematic errors in HAUP measurements

A phenomenological model is proposed to account for the systematic errors characteristic of the NAUP (high-accuracy universal polarimeter) technique. The model is based on the assumption that the sample surface possesses effective dichroic properties dde to polishing, inhomogeneities or differential Fresnel reflection. It is found that there is a sample contribution both to the parasitic ellipticities of the polarizers and to the so-called delta Y error. This contribution adds to those which are intrinsic to the optical device. Measurements are presented in three test materials: LiNbO3, Rb2ZnCl4 and SiO2. The results are interpreted in the light of the proposed model. For the first two materials the surface effects are visible in the form of spurious linear dichroism and residual ellipticity. For the third case, the problem of multiple reflections arises, since the sample had a high degree of plane parallelism. The influence of these additional contributions is eliminated experimentally. Finally, in view of all the results, some criteria are established for the sample optimization in HAUP measurements.