In a classical (i.e. Type I) pyroelectric detector, a crystal plate is coated with two metal electrodes. In order to achieve an absorbing power as close as possible to 100%, different structures have been used in the past: (i) a metal-black coating on the front electrode and (ii) a very thin front electrode having a square resistance matching the impedance of vacuum, with the back electrode making a quarter wavelength structure. However, the quarter wavelength structure becomes inefficient when the absorption length becomes smaller than the plate thickness. The simpler solution is to use a transparent antireflective layer, so that the whole radiant energy would enter the pyroelectric plate and be absorbed. This can be a perfect solution when the double path through the layer matches the phase shift on reflection at the pyroelectric material, and a nearly perfect one for a broad band centered at that wavelength. Transformation of radiant power into heat occurs directly within the pyroelectric. It is shown that a number of semiconductors can be used to make such antireflective layers, and practical solutions are feasible for many pyroelectrics. The advantages of this simple solution over λ/4 structures stuck to the pyroelectric plate, as proposed by Parsons et al., are a negligible added heat capacity as the antireflective coating can generally have enough electrical conductivity to be used as an electrode, and a negligible lateral heat conductivity. When such a solution is not feasible (i.e. reflectivity cannot be cancelled completely), we can still use a very thin metal electrode. It is shown that the electrical conductivity may be 2 orders of magnitude smaller than for the bulk. This leads to small indexes in the IR (e.g. n3≅k3≅ 6), very convenient to give the structure a null reflectivity, when the granular metal layer is covered by a suitable, experimentally available coating (index of refraction n2 = √2n3. Additionally, the determination of the wavelengths and values of the absorption maxima give useful information on the complex refractive index of the coated crystal at wavelengths where reflectivity and transmission measurements are often inaccurate or impossible (T0̃). © 1990.
