Development of antireflective coatings realized by thin film systems requires their characterization and optimization of their properties. Functional properties of such interference devices are determined by optical constants and thicknesses of the individual films and various defects taking place in these systems. In optics industry the characterization of the films is mostly performed in a relatively narrow spectral range using simple dispersion models and, moreover, the defects are not taken into account at all. This manner of characterization fails if applied to real-world non-ideal thin film systems because the measured data do not contain sufficient information about all the parameters describing the system including imperfections. Reliable characterization requires the following changes: extension of spectral range of measurements, combination of spectrophotometry and ellipsometry, utilization of physically correct dispersion models (Kramers-Kronig consistency, sum rules), inclusion of structural defects instrument imperfection into the models and simultaneous processing of all experimental data. This enables us to remove or reduce a correlation among the parameters searched so that correct and sufficiently precise determination of parameter values is achieved. Since the presence and properties of the defects are difficult to control independently by tuning of the deposition conditions, the optimization does not in general involve the elimination of defects. Instead they are taken into account in the design of the film systems. The outlined approach is demonstrated on the characterization and optimization of ultraviolet antireflective coating formed by double layer of Al<sub>2</sub>O<sub>3</sub> and MgF<sub>2</sub> deposited on fused silica.
Authors introduce several results of high quality mirror segments testing. The segments were designed for the large-area
light-weight mirror systems for UV detectors of weak optical signals. For this category of the optical components, an increasing of demands on technology production is typical. This is caused by various reasons: 1) Thickness to diameter ratio is 1:100 for this type of segments. For astronomical mirrors, this ratio is about 1:10. This is the reason why the manufacturing technology of the light-weight segment surfaces was changed. Similarly, usually used testing methods of the shape of the optical surfaces are changed. The shapes of the
surfaces are evaluated by the minimal spot diameter of the reflected beam, which contains 95% of the incident light; 2) Processing technology of working surfaces was enhanced because of the UV light wavelength. The technology must respect the fact that the amount of diffused light in the short UV wavelength region is increasing in the
dependence on the surface roughness of the mirror; 3) Surface reflectivity is not the only important parameter of the optical reflecting thin film systems in this kind of applications. Surface roughness and homogeneity of thin films are taken into account of testing methods too.