Photography and Spectrography of the Night Sky in the Infrared Region of the Spectrum Using an Electron-Optical Converter

It is well known that the sensitivity range of photographic emulsions extends into the infrared region of the spectrum only relatively slightly—in the best cases up to 11,000 Å. Moreover, even at the maximum, the sensitivity to infrared rays is very low, barely reaching the sensitivity of diapositive plates in the visible region of the spectrum. This in itself almost entirely rules out the possibility of using photographic plates for photographing, and still more for spectrographing, weak infrared radiations, such, for example, as the glow of the night sky.

Modern photocells (as well as photoelectronic multipliers) possess, in the infrared region of the spectrum, a considerably higher sensitivity, making them suitable for measuring very weak light fluxes; moreover, the spectral region accessible for measurement extends somewhat farther toward long wavelengths than in photographic emulsions, approximately up to 12,000 Å. However, with photocells it is possible to measure only the values of the total flux of radiant energy falling on the photosensitive layer. Therefore, investigation of an entire spectrum or photometric study of an extended object (say, the region of the sky) entails highly painstaking and laborious work, not to mention the possibility of changes in the object being measured itself during the time required for the measurements.

In addition, in ordinary arrangements a photocell, unlike a photographic emulsion, records not the quantity of radiant energy incident upon it, but the instantaneous value of its intensity, which deprives it of one of the principal advantages of the photographic emulsion—the possibility of a significant increase in sensitivity by increasing the exposure time. This drawback can easily be eliminated by using special integrating circuits, which, however, by increasing the time required for a single measurement, introduce new serious complications.

in the investigation of extended objects. Finally, photoelectric methods of measurement lack the documentary quality inherent in photographic methods.

All that has been said makes the idea of combining the merits of photocells with the advantages of a photographic emulsion very attractive. This proves possible by using, as the light receiver, an electron-optical image converter in combination with a photographic plate.

In principle, the scheme for using an electron-optical image converter is extremely simple¹ (Fig. 1). An optical image (say, a spectrum or the image of an object) is projected onto a semitransparent photosensitive cathode. Photoelectrons emitted from this cathode are accelerated by an electric field and fall on a transparent fluorescent screen, forming on it a visible image which, with a greater or lesser degree of accuracy, reproduces the image projected onto the cathode. (Instead of the accelerating field, a more refined electron-optical system may also be used.) Since the spectral composition of the screen’s luminescence does not depend on the spectral composition of the radiation incident on the cathode, it can be chosen so that the screen’s luminescence is as actinic as possible for the photographic emulsion. The image is photographed by direct contact printing from the fluorescent screen. Photometric processing of the negative obtained is then carried out by the methods customary for photographic photometry, with additional allowance for the spectral sensitivity of the photocathode.

Fig. 1. Diagram of the use of an electron-optical converter with a photographic plate.

It should be noted that, in addition to the advantages already mentioned above, arising from the combination of photoelectric and photographic methods, and also the absence of any need for amplifiers, the described scheme makes it possible to raise the threshold of sensitivity through the relative reduction of the background produced by thermoelectrons, and also because the quantum yield of the screen can be made greater than the quantum yield of the receiving photocathode.¹

Thus, by using an electron-optical converter with a photographic plate, we obtain the possibility of increasing many times the sensitivity of the receiving system (i.e., ultimately, of the photographic plate) to infrared radiation. This opens up entirely new possibilities for the study of weak infrared radiations, including spectroscopy.

The works reviewed²˒³ are, so far as we know, the first serious attempt to use the device described for scientific purposes, and they brilliantly illustrate the broad possibilities of this method.

Infrared radiation of the night sky was first discovered in 1939 by L. A. Kubetskii⁴ with the aid of a photoelectronic multiplier. In the same year, Slipher⁵ detected photographically, in the spectrum of the night sky, radiation with wavelength \(\lambda = 8600\) Å. In 1944–1945, Stebbins, Whitford, and Swings⁶, using a set of light filters, discovered extremely intense radiation with wavelength about \(10\,440\) Å. Finally, in 1948, S. F. Rodionov and E. N. Pavlova⁷, using

photoelectric multiplier with a set of filters, confirmed the presence in the spectrum of the night-sky glow of a line with wavelengths close to \(\lambda = 8660\) and \(\lambda = 10\,440\) Å, and also obtained an indication of the possible existence of a line with \(\lambda = 9100\) Å.

It turned out that the intensity of the night-sky glow in the infrared region of the spectrum undergoes characteristic changes during the night, namely, it has a sharply pronounced maximum around local midnight\(^7\) (see also\(^8\)). Let us recall that an analogous maximum, but only around 1 a.m., is exhibited by the intensity of the green line of the night-sky glow\(^9\).

The intense infrared radiation of the atmosphere is of interest not only because it is a serious hindrance for a number of astrophysical investigations, but also, chiefly, because it is closely connected with the problem of the structure of the upper layers of the atmosphere (see\(^4,8\)). Of still greater interest are the spectra of the night-sky glow in the infrared region of the spectrum, first obtained by V. I. Krasovskii with the aid of an electron-optical converter with a photographic plate. The author used a converter with an oxide-antimony-cesium cathode, possessing sensitivity up to 12,000 Å.

The dispersion of the spectrograph on which the spectra were obtained was, at \(\lambda \sim 1\,\mu\), \(7000\) Å/mm.

During 1948 a number of spectra were obtained both near Moscow and at the Crimean Astrophysical Observatory.

A reproduction of one of them is given by the author (Fig. 2). (In the upper part is the spectrum of the night sky; in the lower part, the spectrum of the helium glow.) Lines (or bands?) near 10,400 and 8600 Å are distinctly visible. The author notes that on the negative one can see the lines 6300 and 5577 Å, and also one more infrared line. (In\(^2\), probably erroneously, the wavelength indicated for it is “about \(\lambda = 8600\) Å,” i.e., coinciding with the wavelength of one of the bright lines.) In addition, a continuous spectrum is regularly observed.

Fig. 2. Infrared spectrum of the night sky (above) and of helium (below).

The distribution of energy over the spectrum varied from photograph to photograph. In one of the photographs the energy illumination at the maximum of the 10,400 Å line turned out to be five times greater than for the 8600 Å line, in satisfactory agreement with the data of Rodionov and Pavlova\(^7\) (by 2–4 times, see\(^8\)).

According to the author’s measurements, the brightness of the infrared glow of the night sky at the horizon is higher than at the zenith, which is characteristic of the glow of the upper layers of the atmosphere.

The author does not give exact values of the wavelengths and intensities of the observed lines, which are of primary interest from the standpoint of the problem of atmospheric nitrogen\(^4\). However, there is no doubt that further application of the described method will make it possible in the very near future to carry out thorough measurements of the spectrum of the infrared radiation of the night sky. In any case, before us for the first time opens the possibility of detailed spectroscopic study of weak radiations in this region of the spectrum.

The second of the papers under review is devoted to photographing, in infrared rays, the region of the sky in the direction toward the center of the Galaxy. It is well known that the center of the Galaxy is hidden from us by an absorbing layer of dark

matter, and only south of the galactic center, through gaps, a bright stellar cloud is observed. Astronomical data indicate in favor of the fact that this stellar cloud is the outer part of the galactic nucleus. Since the dimensions of the nucleus of the Galaxy may prove to be its essential cosmological characteristic, a study of the entire region of the galactic center appears to be important. Such a study can be carried out in infrared rays, where the absorption of light by dark matter should be considerably less. Stebbins and Whitford[^10], who investigated this region on 60- and 100-inch reflectors of the Mount Wilson Observatory with the aid of a photoelectric cell with a light filter (maximum sensitivity about \(1.03\,\mu\)), did indeed detect the presence of a “luminous body” not observed in visible rays. However, their measurements, extremely laborious and covering only a small region of the sky, could not give a complete picture of the whole region of the galactic nucleus.

Fig. 3. Photograph of the region of the galactic center, obtained in infrared rays with an electron-optical converter. (At the left is the well-known bright stellar cloud in the constellation Sagittarius.)

Fig. 4. Ordinary photograph of the same region as in Fig. 3, in visible rays.

A. A. Kalinyak, V. I. Krasovsky and V. B. Nikonov[^3], with the aid of an electron-optical converter and a photographic camera with an appropriate light filter, obtained a small-scale photograph of the region of interest in infrared rays with a wavelength of the order of \(0.97\,\mu\). The use of still longer-wavelength radiation seemed inadvisable in view of the intense glow of the night sky in the spectral region close to \(1.04\,\mu\) (see above).

The light-sensitive camera produced on the screen of the converter an image with an area of more than 200 square degrees, at a scale of \(0.5\,\mathrm{mm}^3\) per 1 square degree. The observations were made at the Crimean Astrophysical Observatory (Simeiz). Exposures were 20–40 minutes.

In Fig. 3 one of the photographs obtained by the authors is shown. For comparison the authors present a photograph obtained with an ordinary photographic camera in visible rays (Fig. 4). The scales of the photographs are the same. In infrared rays the “luminous body” is clearly revealed, which is not detected in visi-

region of the spectrum (and only partially coinciding with that detected by Stebbins and Whitford). The authors carried out photometric processing of two photographs, on the basis of which it proved possible to formulate a number of conclusions of a preliminary character. In particular, the authors were able to estimate the dimensions of the galactic nucleus. According to their data, the angular size of the galactic nucleus exceeds \(9^\circ\), which corresponds to linear dimensions of the order of 1200 parsecs.

Undoubtedly, the continuation of analogous studies, both small-scale and large-scale, with the use of advanced astronomical optics, will make it possible to advance considerably in a number of branches of astrophysics and astronomy.

The works reviewed are, of course, preliminary in character. However, they convincingly show how fruitful are the applications developed by the authors of the above-described principle of combining an electro-optical image converter with a photographic plate. There is no doubt that this principle will find the widest application not only in astrophysical and geophysical investigations, where the chief experimental difficulty that has to be overcome is the weakness of the measured light fluxes, but also in the practice of infrared spectroscopy under laboratory conditions.

G. Rozenberg.

CITED LITERATURE

1. N. S. Khlebnikov, UFN 29, 201 (1946).
2. V. I. Krasovskii, DAN 66, 53 (1949).
3. A. A. Kalinyak, V. I. Krasovskii and V. B. Nikonov, DAN 66, 25 (1949).
4. I. A. Khvostikov, UFN 33, 570 (1947).
5. V. M. Slipher, Monthly Not. Roy. Ast. Soc. 93, 666 (1939).
6. J. Stebbins, A. E. Whitford and P. Swings, Phys. Rev. 66, 225 (1944); Astrophys. Journ. 101, 39 (1945).
7. S. F. Rodionov and E. N. Pavlova, DAN 65, 831 (1949).
8. M. V. Shishkina, Abstract in this issue of UFN.
9. I. A. Khvostikov, Glow of the Night Sky, Publishing House of the Academy of Sciences of the USSR (1947). See also S. F. Rodionov, E. N. Pavlova and E. V. Rdultovskaya, DAN 66, 55 (1949).
10. J. Stebbins and A. Whitford, Astrophys. Journ. 106, 235 (1947).