Key points are not available for this paper at this time.
Optical-profilometric measurements provide fast, meaningful descriptions of structured surfaces, such as power spectral density and surface roughness parameters. Conventionally, profilometric height data is used to calculate the far-field scatter of a randomly-structured surface. Random nanostructures were monolithically etched on fused silica (FS) sample surfaces, to enhance spectral transmission of light beyond the boundary's Fresnel value, within a ±2.5° projection angle in the axial direction. We fabricated several FS samples with statistically different roughness, but similar optical performance over the VIS to NIR wavelength band and compared their power spectral density and optical scatter performance. Surface height map profiles were measured using UV confocal profilometry, and the data was leveraged to produce conventional (spatial) power spectral densities for each surface. Using surface height and transverse feature size data, scatter was calculated using both Rayleigh-Rice and Harvey-Shack surface scatter models, and then compared to experimental radiometric scatter data. Using experimentally obtained profilometric height data as input, the scatter models were able to accurately predict diffuse scatter from mechanical rough surfaces compared to experimental scatter measurements; however, the models were unable to predict the axial scatter for high-density transmission-enhancing nanostructured surfaces.
González et al. (Fri,) studied this question.