Light scattering: From academic research towards industrial applications

摘要

For a substantial part of the optics community, light scattering for a long time has been seen as a mystical phenomenon, difficult to measure, difficult to interpret, and for the most part only relevant for "those high-end applications". This is however not true anymore and those ignorantly neglecting light scattering risk to be left behind in the near future. With today's possibilities to reduce the size of optical systems and to reduce the number of optical elements while at the same time continuously pushing performance, the demands on optical elements are steadily increasing. While the first issue to solve for most optical elements is still shape accuracy or figure, light scattering effects have to be considered right after that or even at the same time. This is not trivial, but it is also not impossible. Part of the problem is the complexity of imperfections of optical components and systems. Another part is the scattering community itself - developing extremely sophisticated and highly accurate theories to model scattering, but sometimes failing to see scattering problems from the perspective of a manufacturer or use the insight into scattering to give the manufacturer hints in their language to improve their products. Interestingly, one of the brightest people in the light scattering community and the first to introduce the so called Rayleigh-Rice vector perturbation to the optics world, Eugene Church, did exactly that. Before other people developed all sorts of new calculations based on Maxwell's equations, he just used a simple formula known in the Radar community linking electromagnetic scattering to surface irregularities (roughness) in order to support the manufacturing of diamond-turned optics. Despite all the discussions among experts on little details, we do know that this simple formula is highly accurate and produces the same results as rigorous numerical calculations. Moreover, in contrast to numerical calculations, the simple expression linking scattering and surface roughness can easily be used to quickly estimate scattering based on roughness data (in the relevant range of course) or even to use light scattering measurements as a tool to retrieve information on the roughness of a surface. Using light scattering to measure surface roughness has some exclusive advantages over all other methods: 1 It is non-contact 2 It is insensitive to sample or sensor vibrations 3 Rms-roughness can be measured directly in the spatial frequency range relevant for the optical application 4 The surface power spectral density function (roughness spectrum) is measured directly without the need to Fourier transform 5 Roughness well below 1 nm rms and particles / features with diameters much smaller than the wavelength can be measured 6 Statistically, one typical scatter-based roughness measurement has a robustness of hundreds or even thousands of topographic roughness measurements 7 Scatter-based metrology can be so fast that even large surfaces can be fully scanned for 100% quality control In our presentation, we will briefly discuss simplified scatter models and how to use them in practical applications. We will also try to illustrate the advantages of scatter-based surface characterization for freeform mirrors by showing our latest results obtained using a robotic scatter sensor. Future work will focus on a simplified general approach for optical design and manufacturing that should include scattering effects by: 1 Analysis of potential imperfections (roughness, defects, material imperfections, coatings, SSD, ...) 2 Estimating the impact of different imperfections onto performance 3 Identify most critical contribution(s) 4 Find pragmatic ways to specify and measure these imperfections 5 Provide reliable measurement procedures and data to characterize the overall performance of the final product.