ABSTRACTS

A NEW METHOD OF PHOTOGRAPHY IN INFRARED RAYS

The infrared part of the spectrum of optical sources presently accessible to experimental investigation extends to approximately \(0.4\) mm. By origin, these are predominantly rotational and rotational-vibrational spectra of molecules; the former give wavelengths of the order of tens and hundreds of microns, the latter of the order of one or several microns. For the theory of molecular structure and for determining the nature of intramolecular bonds, the study of infrared spectra is of exceptional interest. It is carried out chiefly by thermal methods: with a bolometer, thermoelement, or microradiometer. Recently photoelements have also begun to be used.

Thus, the methods for investigating the infrared part of the spectrum are rather varied and have already attained great refinement. The photographic method is less developed than the others; here, progress by each tenth of a micron toward longer wavelengths is achieved with great difficulty, and before the appearance of the works under review by Czerny \(^{1}\) and Villenberger \(^{2}\), matters had not advanced beyond \(1.7\,\mu\). Meanwhile, the photographic method gives investigations a documentary character and, if sufficiently developed, to a considerable degree frees them from the use of cumbersome measuring instruments.

In addition to the study of spectra, photography in infrared rays is also important for landscape photography. In dusty air and other turbid media, short visible rays are strongly scattered; distant views appear veiled, and their details cannot be captured by photography. However, according to the well-known Rayleigh law, scattering is inversely proportional to the fourth power of the wavelength and, for example, for infrared rays with \(\lambda = 1200\,\mathrm{m}\mu\) will be 81 times less than for violet rays with \(\lambda = 400\,\mathrm{m}\mu\): the former pass almost completely where, for the latter, the medium is practically entirely opaque. Figures 1 and 2 show two photographs taken from Angerer’s booklet \(^{3}\), the first of them taken in visible rays, the second in infrared. The clarity of distant views in the second photograph is incomparably higher; the distribution of light-and-shade intensities in the pictures is, of course, somewhat different.

Fig. 1. Photograph in visible rays. The distance is hazy.

Fig. 2. Photograph in infrared rays. Mountains are visible in the distance.

In photographing in infrared rays, up to the present time one of the following methods has been used: a) sensitization of photographic plates, b) the method of removing a veil, c) the method of quenching the glow of phosphorescent screens.

In the first of the cases listed, plates sensitized with dicyanine are chiefly used. Photographs obtained are of high quality; however, the plates are fragile (the dye sensitizers decompose quickly), exposures are long, on the order of several hours, and the active spectral region is limited to the nearest infrared portion, breaking off at about \(\lambda = 1 \mu\).

In the second method,* a sensitive photographic plate is treated with a weak solution of one of the dyes that strongly absorbs the active blue and violet rays and transmits infrared comparatively well, for example a solution of malachite green; then the plate is veiled by keeping it for 10–12 sec. under illumination of the order of tenths of a lux, as a result of which it is exposed, but, thanks to the dye, only in the surface layer. When an infrared spectrum is superposed on such a plate, in the illuminated places the veil is removed, and a positive image of the spectrum is formed. The exposure time is calculated in several tens of minutes; the region of the spectrum reproducible by photography extends to \(\lambda = 1.1 \mu\).

The third method is the method of quenching phosphorescence. An excited phosphor, luminescing under ordinary conditions for several hours, changes the time course of its emission under the influence of rays that do not belong to the excitation region (including under the influence of infrared rays). In some cases a short-term flash of phosphorescence is observed, during which the phosphor gives off almost all the energy of excitation; as a result, after the end of the flash, against the bright background of the phosphor there is formed a more or less contrasty darkening of the places illuminated by infrared rays (the process is called bleaching). In other cases, direct quenching of the phosphorescence by infrared rays takes place, with the shadow contrast appearing without a preliminary flash.

When photographing by this method, the phosphorescent substance is applied to a screen having the dimensions of the photographic plate. The phosphor, excited by preliminary illumination, is placed in the spectrograph in place of the photographic plate. After a short exposure, lasting from a few seconds to several minutes, darkened places corresponding to the infrared bands of the spectrum under study appear on the phosphorescent screen. By contact printing from the screen, a positive photograph is obtained on the plate. The method is more sensitive than the two preceding ones and is suitable up to \(\lambda = 1.7 \mu\), but it requires further improvement.

Thus even the most sensitive third method gives photographs of spectra no farther than \(1.7 \mu\). Further advance toward

one runs, as Czerny points out, into difficulties of a, to a certain extent, fundamental nature. Relying on the law of black radiation, Czerny calculated the following table:

TABLE I

The second column of the table gives the number of quanta belonging to the spectral intervals indicated in the first column, falling, at a laboratory temperature of $20^\circ$ C, from both sides in the course of one second onto $1\ \mathrm{cm}^2$ of a photographic plate.

The third column is included for the sake of clarity; it indicates at what distance from the plate a Nernst burner should be placed so that the number of quanta of visible light falling per second from the burner onto $1\ \mathrm{cm}^2$ of the plate would be equal to the number of quanta of black radiation indicated in the second column. From comparison of the figures in the second and third columns it is evident that even at room temperature a photographic plate sensitive to radiation with wavelength $2\,\mu$ receives such a powerful flux of quanta that the plate must blacken by itself within a very insignificant interval of time.

Using freshly prepared plates, we nevertheless may hope to obtain a photograph of the comparatively distant infrared part of the spectra of strong sources. In the fifth column of the table is given the number of quanta falling each second through the slit of a spectrograph onto $1\ \mathrm{cm}^2$ of a photographic plate from a Nernst burner under average conditions of exposure; these numbers considerably exceed the corresponding numbers of the second column.

Finally, in the fourth column are given numbers showing the quantity of molecules of atmospheric air which, under normal conditions, strike every second $1\ \mathrm{cm}^2$ of the surface of a photographic plate, possessing energy corresponding to the magnitudes of the quanta of the spectral regions indicated in the first column. From consideration of Table 1 it follows that, for photographing regions lying beyond $2\,\mu$, one should choose such a process in which darkening of the plate would be excluded. Such a method was indicated by Czerny and subsequently developed by his co-worker Willenberg.

The essence of the method is that, under the influence of infrared rays, part of the liquid covering a very thin celluloid membrane (on the order of tenths of a micron) evaporates from the illuminated places. The membrane is in a closed evacuated vessel filled with vapors of the liquid, saturating the space; consequently, evaporation of the liquid layer does not occur in the absence of illumination; likewise, the nonuniformity in the thickness of the layer that has arisen as a result of illumination is not smoothed out. Figure 3 shows the arrangement of the apparatus and the disposition of its parts. Here \(Z\) is a metal cylinder, \(P\) a round plate with a square window covered by a plate \(S\) of rock salt (\(5 \times 5\) cm), \(R\) a ring onto which the protective cover \(K\) is inserted. The infrared spectrum under investigation enters the chamber from the left in the drawing, through the window. \(S.M\) is the ring on which the celluloid membrane, obtained from capon varnish, is stretched. On the left side the membrane is blackened with a thin layer of soot, which almost completely absorbs infrared rays; on the right side it is covered with a layer of kerosene (a fraction boiling from \(260\) to \(280^\circ\) C). Owing to the extraordinary thinness of the kerosene layer, its flow proceeds very slowly, so that it may practically be neglected. \(H\) is a movable heating coil, by means of which the required portion of kerosene, settling on the membrane, is evaporated from a small test tube. When evaporation is completed, the coil is moved out of the field of view by means of device \(T\). \(W\) is a resistance manometer, \(B\) a broad glass plate, and \(G\) a gasket. The glass window \(B\) serves for observation and also for photographing the membrane. At the lower right in Fig. 4 an ordinary photographic camera is shown. When the membrane was photographed, it was illuminated by a \(200\)-W lamp \(L\), whose thermal radiation was blocked by water and by a Schott light filter BG 9.

Fig. 3.

Changes in the thickness of the kerosene layer are followed by the change in its interference colors. Another method consists in observing the darkening of those places from which evaporation has occurred, resulting from the exposure of microscopic irregularities of the membrane.

A very important factor for successful photography proved to be the choice of the evaporating substance. In his original work Czerny used for this purpose a layer of anthracene crystals that had been subjected to sublimation. However, the use of a solid substance does not give sufficient sensitivity—

ness. In addition, over time recrystallization of the substance occurs (small crystals are sublimated, large ones grow), as a result of which the sensitive layer becomes coarse-grained and incapable of reproducing details.

The heterogeneous composition of kerosene also presents certain inconveniences, owing to the nonuniform evaporation of its various fractions; however, this drawback proved to be the least of the evils, since numerous tests of a whole series of other substances, for a great variety of reasons, invariably led to negative results.

Fig. 4.

Fig. 5.

As was indicated above, because of the slowness with which the liquid flows down, the vertical position of the membrane does not lead to distortion of the image. Much more harmful to the results of photography is the action of surface tension, which, when the surface contracts, tends to level out the depressions formed in the liquid layer as a result of evaporation.

Fig. 6.

Fig. 7.

To reduce the harmful influence of this factor, the liquid layer should be made as thin as possible; the dispersion of the spectrograph, on the contrary, as large as possible.

To determine the ability of sensitive membranes to reproduce details, an image of a grating with slits of variable width from 0.20 to 0.01 mm was projected onto the membrane. Fig. 4 gives a photograph of the grating enlarged by a factor of 2.7 in reproduction. On the left is an ordinary photograph; on the right, a photograph in infrared rays. It turned out that in the latter case two strokes were reproduced as separate if

the distance between them reached 0.09 mm. Thus the sensitive membrane acts like a coarse-grained plate with a grain size of about 50 μ.

Figures 5, 6, and 7 present photographs obtained by the method described. Fig. 5 is the spectrum of HCl bands; the spectrum is reproduced up to 3.6 μ. Fig. 6 is the absorption spectrum of naphthalene; in the photograph the spectrum extends to 6.5 μ; the exposure is several seconds. Fig. 7 shows the possibility of obtaining photographs of weakly heated objects; it gives a photograph of a flask with water at 90° C; exposure 30 sec.

The results presented show that the new method surpasses, many times over, all methods practiced up to the present time, both in sensitivity and in the breadth of the spectral region covered.

Literature

1. M. Czerny, Zs. f. Phys. 53, 1, 1929.
2. H. Willenberg, Zs. f. Phys., 74, 663, 1932.
3. Augerer, Grundlagen der wissenschaftlichen Photographie.
4. Mc. Lennan, Proc. Roy. Soc. (A), 100, 200, 1921.
5. A. Terenin, Zs. f. Phys. 23, 294, 1924.
6. P. Lenard, Handb. d. Experimentalphysik, 23, 2nd half, 751, 1928.

V. Levshin, Moscow.