Diffuse scattering of optical one-dimensional gratings becomes increasingly critical as it constrains the performance, e.g., of grating spectrometers. In particular, stochastic disturbances of the ideal grating structure provoke straylight. In this paper, the straylight spectrum of stochastically disturbed gratings is examined. First, a 1D-method is presented that allows to calculate 2D-diffuse scattering of arbitrarily polarized light originating from stochastic disturbances of the grating geometry on the basis of standard optical simulation tools. Within the scope of this method an enormous reduction of computational effort is achieved compared to the full 2D-simulation approach, i.e., the computation time can be reduced by several orders of magnitude. Hence, the method also allows to address even large period gratings that are not possible to calculate within a full 2D-approach. In analogy to scattering theories for surface roughness the method relies on typical characteristics of straylight originating from small disturbances, that the angle resolved scattering (ARS) can be separated into a product of the power spectral density describing the 2D stochastic process and additional factors depending on the undisturbed 1D grating structure. In a second part, an analytical model within Fourier optics utilizing thin element approximation (TEA) describing the wide angle scattering of lamellar gratings disturbed by line edge roughness (LER) for TE-polarized light is derived and verified by applying the 1D-simulation method. For shallow gratings, we find an excellent agreement between simulation and TEA over the whole transmission half space. In addition, this model allows a descriptive understanding of the underlying physical effects and, accordingly, the influence of relevant parameters (grating geometry, refractive indices, illumination) onto the scattering spectra is discussed. Further, it is shown that LER-scattering can be described within a modified Rayleigh-Rice-ARS usually found within the frame of surface roughness.
Read full abstract