平面回折格子による偏光強度分布

抄録

The intensity distribution of diffracted light by Echelette grating which is irradiated by natural light has been theoretically discussed by many workers. Their scaler theories can approximately explain the experimental values, but not so in the case of polarized light.
Calling the polarized light P-polarized light when the electric vector of the incident light is perpendicular to the groove direction and S-polarized light when it is parallel to the groove, we know that the values of intensity distribution of P-polarized light measured and calculated from the scaler theory are nearly equal, but in the case of S-polarization they are fairly different which means that the scaler theory is no longer valid for S-polarized light, whereas the intensity distri bution formulae, derived on the assumption that the material of the groove surface is a perfect: conductor and that there is an apparent phase change in proportion to the depth of the groove, give the values which agree well with the experimental values.
Originally, Echelette grating is provided with a fairly large blaze angle to separate the concerned diffracted light from the direction of zero order, and hence the grooves are made deep. If the width of the groove is much larger than the wavelength of incident light, the incident and reflected lights will interfere and produce the standing wave which will be localized only in the groove when the width of the groove becomes small. Since the antinode of the standing wave becomes the source of wavelets and positions of antinodes differ in the two polarizations, so in S-polarization a phase difference proportional to the depth of the groove is expected. When the substance of the groove surface is of dielectric, standing waves will not be produced on account of small amplitude of reflected waves therefore, the difference in intensity distribution between the two polarizations will not be found.
This paper proves that the formulae derived from the above consideration explains well the experimental values reported hitherto, supplemented with remarks on the grating efficiency and the grating anomaly.