Automatic detection of a sinusoidal grating shape during a manufacturing process

We develop a technique for assessing the sinusoidal shape of a diffraction grating during its manufacturing process. The originality of this approach is based on the use of the harmonic analysis of the geometrical periodic profile. A neural network is implemented to evaluate the quality of the sinusoidal character. This is essential to ensure the optimal efficiency of specific targeted applications such as plasmonic sensors. The advantage of this technique is highlighted by studies involving theoretical and experimental tests on a diffraction grating with a period of 627 nm. The detection criterion qualifying the quality of the in-process grating is also discussed.