A SELENIUM PILE AS A POLARIZER FOR INFRARED RADIATION

As is known, the region of transparency of Iceland spar is limited on the long-wavelength side to approximately \(2\,\mu\), as a result of which polarizing prisms are no longer suitable for radiation of greater wavelength. Polaroids are also inapplicable for the infrared region, for their polarizing power falls off sharply already at the red boundary of the visible spectrum. The problem of producing and analyzing polarized radiation in the infrared region of the spectrum has not yet been solved by other, quite varied and at times highly ingenious, devices. They either possess insufficient polarizing power, or substantially weaken the inten-

intensity of light, or, finally, prove extremely inconvenient in handling. The authors of the paper being reviewed*) have succeeded in achieving considerable success in this direction. The idea underlying the device they propose is by no means new, although the authors themselves avoid any mention of predecessors. The matter concerns a stack completely analogous to a glass stack, a description of which may be found in any optics textbook. The only difference is that, instead of comparatively thick glass plates, thin films are used here, made of selenium or of other substances transparent in the infrared region of the spectrum.

According to the authors’ measurements, a selenium film does not exhibit absorption of light in the wavelength interval from 2 to 14 μ (only this interval was investigated by the authors) at least up to a thickness of 52 μ. The refractive index of selenium is \(n \simeq 2.54\), and, correspondingly, the angle of complete polarization is \(68.5^\circ\). Using films 4 μ thick, the authors obtained the following results. For a stack composed of five films, the degree of polarization in the spectral region investigated by them (from 2 to 14 μ) was nowhere less than 94%, being considerably higher over most of this region. For a stack composed of six films, the degree of polarization over the entire region was above 98%.

(Diagrams labeled a, b, c, d; label in diagram: Glue.)

The small variations observed in the degree of polarization for different wavelengths are explained by the authors as interference effects occurring in thin films. As an illustration they give the results of an investigation of the polarizing action of a single film, for which these effects appear very distinctly. As the number of films increases, the interference effects rapidly weaken. Variations of the angle of incidence of the light beam on the stack within \(5^\circ\) do not exert a noticeable influence on the degree of polarization.

Another important indicator is the transparency of the stack. A stack of five films transmits 47% of the incident unpolarized light, i.e. 94% of the corresponding polarized component. The transparency of a stack of six films is somewhat lower.

The authors devote much attention to the description of the technology for making the films. Amorphous selenium is used as the starting material, previously melted (in air) in a Pyrex crucible, heated—

*) A. Elliott, E. J. Ambrose and R. Temple, J. Opt. Soc. Am., 38, 212 (1948).

using an electric current. Evaporation is carried out in a high vacuum in an apparatus of the usual type, such as those used for evaporating aluminum. To obtain a good layer, evaporation is carried out slowly. The deposition of selenium is performed on a thin nitrocellulose film stretched over a metal ring 150 mm in diameter. To prevent the reverse surface of the film from being coated with selenium, it is covered with a metal cap.

The nitrocellulose film is made from a 3% solution of motion-picture film in amyl acetate. Two wires are placed on a glass plate (at a distance of 250 mm from one another), and a little solution is poured between them and distributed uniformly over the surface of the plate so that it covers both wires. After drying, the film that has formed is separated from the glass by warming it with moist steam and, with the aid of the wires, is transferred to a metal ring, to which it is glued with the same solution. The result is an even, tightly stretched film.

After the selenium layer has been deposited on it, the film is cut from the ring and transferred to a sheet of paper (selenium layer upward). A previously prepared bronze frame 0.125 mm thick is then placed on it. The edges of the frame are polished with fine emery paper, and one of them is coated on one side with a cold glue of the LePage type (Fig. b), the frame being placed on the film with the side coated with glue. A nitrocellulose varnish is applied to the film at the opposite edge of the frame. When the glue has dried, the frame is hung (with the glued edge upward) in an empty vessel, which is then slowly filled with acetone. In about 15 minutes the nitrocellulose film completely dissolves, after which the acetone is slowly drained from the vessel. Freed from the nitrocellulose backing, the moist selenium film, under the action of surface-tension forces, adheres to the frame and can, together with it, be carefully transferred to the mount of the pile, in which it is set at an angle of 65° (Fig. c). The entire set of films making up the pile goes through this processing stage and is installed in the mount simultaneously. The mount is then placed in a holder (Fig. d), and the pile is ready.

Along with the one described, the authors also tested a pile of a somewhat different type. On a prism made of rock salt, a layer of sodium fluoride ($n \simeq 1.3$) and potassium iodide ($n \simeq 2.5$) are deposited in succession. A second rock-salt prism, similar to the first (Fig. a), is placed on the last layer in optical contact. Rock-salt prisms are necessary because a ray incident on the face between thallium iodide and sodium fluoride at the angle of complete polarization would, in the absence of prisms, undergo total internal reflection at the parallel sodium fluoride–air face.

If the thicknesses of the layers are chosen so that the rays reflected from the various boundary surfaces mutually reinforce one another as a result of interference, then, as experiment shows, it is possible to reduce the number of layers to three without impairing the polarizing action of the pile. The inconvenience of such a device, in contrast to the preceding one, is that it has good qualities only in a comparatively narrow region of the spectrum. Moreover, its practical realization is convenient only for the visible part of the spectrum, since for the infrared region the thickness of the layers increases so much that it becomes difficult to make them sufficiently homogeneous and non-scattering.

G. Rozenberg