New Applications of Infrared Photography

J. Eggert, Leipzig *

The importance of infrared photography as an auxiliary tool in scientific research and in the technical testing of materials is based chiefly on two remarkable properties of infrared rays: 1) because of their long wavelengths, they are scattered less in turbid media than visible light, and 2) substances that appear quite identical in color and brightness in visible light often absorb infrared light to different degrees, as a result of which their images in pictures obtained in infrared rays show unequal darkening. The ability of infrared rays to penetrate an atmosphere saturated with vapor has already found practical application for a number of years in telephotography. It is also known that very many objects, on photographic plates sensitized to infrared rays, can give completely different intensity ratios than they do to the human eye (an example is the extremely varied rendering of green foliage in landscape photographs in infrared light). However, this latter property has so far been used very rarely. In the following sections, where new applications of infrared photography are described, this specific rendering of various kinds of colors in infrared photographs will be examined in detail.

I. Application of Infrared Photography in the Graphic Industry

In Fig. 1 are shown photographs of several specimens of pigment dyes, printed in the form of “wedges” according to the density of their coloration, i.e. each print begins with the densest coloration at the top, which gradually passes into the white color of the backing at the bottom. These “wedges” were photographed in white light on plates “Agfa—Superpan” and in infrared light in two regions of its wavelengths. The first photograph was obtained on Agfa “750 rapid” plates with a black Agfa No. 83 filter and covers the region from 720 to 780 mµ. For the second photograph infrared Agfa “850 rapid” plates were used with a black light filter Agfa No. 85, which made it possible to cover the region 840–900 mµ. The dye specimens were selected in such a way that pairs of adjacent prints, which appear completely different in infrared light, would appear as similar as possible in visible light. It may be noted that dyes 2, 5, 7, 9, and 11 already partly appear in the first infrared photograph, and even more so in the second, considerably weaker (lighter) than in the photograph made in visible light, and therefore differ sharply from their partners, which seem equally colored in visible light. This means that some dyes transmit infrared rays more strongly with increasing wavelength, since we

* Naturwiss. 23, 281, 1935, translated by A. A. Ilyina.

we see that the white paper on which the prints are located appears much more strongly in these photographs. This increase in the transmission of infrared light with increasing wavelength was also observed in spectroscopic studies of dyed fabrics, clothing, wallpaper, and so on. The opposite case, i.e., an increase in absorption with increasing wavelength, has so far occurred only for the coloring matter of blood, which will be considered below.

The peculiar changes in the absorbing capacity of dyes in infrared light may be of importance in the production of securities. It is conceivable, for example, to make a design on paper money partly in dyes that absorb infrared light and partly in dyes that are transparent

Orthochromatic photograph

Wavelength range 720–780 mµ

Wavelength range 840–900 mµ

Fig. 1. Dyes:

1. Cinnabar
2. Mummy
3. Hansa yellow
4. Hansa yellow + traces of Milori blue
5. Hansa green + Hansa yellow

6. Milori blue + yellow chrome

7. Reflex blue

8. Indanthrene blue
9. Fanale + Bremer blue B
10. Milori blue
11. Chloramine black
12. Soot

Prints of pigment dyes photographed in visible and infrared light.

for it, but which appear completely identical to the ordinary eye. In the same way, hidden images can be placed on paper money, appearing only in infrared light. Further, in specimens transparent to infrared light one may imprint designs that are identically colored in white light, but made with dyes that absorb infrared rays; or one may print a design from a substance that absorbs infrared light together with a design equally colored in visible light but transparent to infrared light, and so on. In all these cases the forgery becomes extremely difficult, since both the harmlessness of the genuine design and the harmlessness of the genuine drawing can easily be established by means of an infrared photograph.

Of particular significance here is the circumstance that a quite insignificant admixture of a substance absorbing infrared rays, imperceptible to the eye, to dyes transparent to this light is easily detected in an infrared photograph. In this case, substantially less than 1% of an admixture of the absorbing substance to a substance similarly colored in visible light but transparent to infrared rays is sufficient for the absorption of infrared light by the resulting mixture to change noticeably.

II. New Possibilities for the Use of Infrared Photography in Medicine

Haxthausen[^3] established from photographs of the vascular system that its reproduction in infrared photographs is obtained much more fully and in greater detail than in photographs made in visible light. This is explained by the fact that, for infrared light, the skin is more transparent than for visible light. At the same time, the ability of blood to absorb infrared rays plays no role, since it is almost the same for visible and for infrared light. Kretz, in his communication[^4], asserts that veins filled with blood poisoned by carbon monoxide appear stronger in infrared photographs.

Fig. 2. Photographs of blood (in cuvettes, dilution 1:3) under different illumination. 1—reduced blood, 2—blood saturated with CO, 3—oxygenated blood, 4—regenerated blood.

For a more precise investigation of these relations, in addition to the indicated photographs made in vivo, we set up corresponding in vitro experiments.[^5] The arrangement of the experiment can be seen from Fig. 2. Small cuvettes (liquid layer thickness 10 mm), placed against a background of white paper with printed black text, were illuminated and photographed.*

* For the shooting conditions, see the end of the article.

The more transparent to the light used in the photography was the liquid in the cuvette, the more clearly the letters of the underlying paper appeared. In the cuvettes there was blood of a pig, diluted with water in a ratio of 1:3 and treated in various ways. The blood solutions were filtered and were quite transparent.

The reduced blood in cuvette 1 was obtained by adding several drops of ammonium sulfide to 20 cm³ of diluted blood; the blood saturated with carbon monoxide in cuvette 2 was obtained by passing illuminating gas through it for several minutes; the oxidized blood (cuvette 3) by passing oxygen and, finally, the preparation designated as regenerated blood (cuvette 4) was obtained by passing air for 30 minutes through a solution of blood saturated with carbon monoxide.

To the eye, reduced blood (Fig. 2, cuvette 1) is almost black, whereas the blood samples in the following cuvettes are colored cherry-red and do not differ from one another. A photograph made in visible light does not convey these rather slight differences in brightness, and all four cuvettes appeared equally dark in the photograph. By contrast, the blood samples in infrared photographs differ markedly from one another. The transmission capacity of the blood dye (hemoglobin) for infrared rays, moreover, depends on their wavelength. CO-hemoglobin and oxyhemoglobin, regenerated from it with the aid of air, appear in infrared light of three different wavelength regions to be almost equally transparent. In the case of regenerated oxyhemoglobin, one may perhaps establish a slight increase in absorptive capacity with increasing wavelength of infrared light. In this respect regenerated oxyhemoglobin is similar to normal oxyhemoglobin.

Reduced hemoglobin gives an entirely different picture than oxyhemoglobin. For short wavelengths of infrared light, oxyhemoglobin is almost transparent, and, conversely, reduced hemoglobin under the same conditions absorbs almost completely. In long-wave infrared light this relation will be the reverse, namely—reduced hemoglobin transmits, while oxyhemoglobin absorbs more strongly than in light of short wavelengths.

We extended our investigations to the blood of human beings and guinea pigs, and our previous experiments were confirmed. The same research method, applied in pathological cases, proved extraordinarily sensitive, in which respect it markedly surpassed the other spectroscopic methods, for example, the method of obtaining absorption curves in the visible part of the spectrum and of verifying with their aid the presence of reducing substances in the form of admixtures to blood solutions. Taking a drop of venous blood, slightly poisoned with carbon monoxide, and photographing it in infrared light by the former method in comparison with a drop of normal blood, we obtained a positive on which the poisoned blood looked transparent and the normal blood dark. Even after the patient had remained for 15 minutes in the fresh air and by spectroscopic means we no longer found carbon monoxide in his blood, the infrared photograph nevertheless gave a positive result.

Especially striking in all these experiments is the fact that blood subjected to the action of carbon monoxide, even in the case when spectroscopically it is no longer possible to find traces of carbon monoxide in it, for a long time still retains a transparency for infrared rays characteristic of it, differing noticeably in the infrared photograph from normal blood. Further experiments must clarify whether this peculiar behavior of blood solutions arises from stubbornly retained traces of carbon monoxide in them, which can be detected only with the aid of infrared photography, or from some other causes.

Further, with the aid of infrared photography we were able directly to observe the well-known beneficial effect of injections of methylene blue in carbon-monoxide poisoning. Of two cases of poisoning

carbon monoxide; in one, an injection of methylene blue was made, in the other—not. Blood samples 45 min. after the injection were photographed in infrared rays, and it turned out that blood subjected to the action of methylene blue does not give deviations from the norm, whereas poisoned blood still noticeably transmitted infrared light. However, up to now we have still not succeeded in reproducing in cuvettes these experiments conducted in vivo.

In order to make this effect more noticeable, it is recommended in mild cases of poisoning to work with undiluted blood and to use, for obtaining photographs, small drops of blood placed on some object. In the most severe cases the blood may be diluted fourfold with water.

In order to establish more precisely the different absorptive capacities of blood solutions obtained in one way or another, we measured absorption in the visible region (Fig. 3) in infrared light from 650 to 1000 mμ (Fig. 4*). The objects of observation were the same blood samples as in Fig. 2; however, here it was necessary to work with different concentrations of blood solutions for different spectral regions. The curves were obtained by the spectrographic method. For the photographs, “Agfa-Superpan” plates were used for the visible region, and “Agfa 950” infrared plates in combination with light filter No. 42—for the infrared region. For spectrograms in the visible light, the blood solutions were diluted with water 50 times. With such dilution, all blood samples in infrared light prove almost equally transparent. For taking the infrared spectrogram, a dilution of the blood samples in the ratio 1 : 2 proved most convenient. In Fig. 3 one can clearly see the well-known shift of the maximum of absorption of hemoglobin from 579 to 570 mμ for blood poisoned with carbon monoxide. In Fig. 4 one can note the characteristic increase in the absorption of oxyhemoglobin, which is also observed for CO-hemoglobin in the region 900—1000 mμ, but already to a considerably lesser degree.

Fig. 3. Absorption curves of pig blood in the visible region (dilution 1 : 50). I—reduced blood, II—blood saturated with CO, III—oxygenated blood, IV—regenerated blood.

Fig. 4. Absorption curves of pig blood in the infrared region (dilution 1 : 2). I—reduced blood, II—blood saturated with CO, III—oxygenated blood, IV—regenerated blood.

* Measurements were carried out by M. Biltem.

III. Application of Infrared Photography in Phytopathology

The green coloring substance of the leaf—chlorophyll—is transparent to red light, beginning at 680 mμ. Taking into account the chemical proximity of chlorophyll to hemoglobin, one may say that the absorption curve of chlorophyll in infrared light corresponds rather to CO-hemoglobin than to oxyhemoglobin of normal blood. Owing to these optical properties of chlorophyll, the positive image of a green leaf photographed in infrared rays should come out almost white, since the colorless substance of the cell no longer contains chlorophyll. Of particular interest is the circumstance that the transparency of chlorophyll to infrared light is preserved even after 30,000,000 years, as is evident from Fig. 5, which shows photographs of a fossil leaf found in Geiseltal. 7

These specific properties of chlorophyll make it possible usefully to apply infrared photography for the study and diagnosis of those plant diseases which destroy chlorophyll or have this as a consequence, since only the cellular contents of a healthy leaf are transparent to infrared light. 8 The destruction of chlorophyll by parasites or in some other way should likewise be detected in an infrared photograph, which can indeed be seen in the photograph of a drying leaf (see Fig. 6). In those places where the leaf has been affected, there are no longer viable cells and, consequently, the chlorophyll in these places has been destroyed, as a result of which in the infrared photograph they should differ from the untouched areas of the green leaf. An orthochromatic photograph does not show this.

IV. Application of Infrared Photography in Microscopy

In the preparation of microscopic specimens, especially such as specimens of insects and other small animals, one often has to deal with pigmented membranes, which must be decolorized with the aid of appropriate chemical agents in order to make them more or less transparent. It should also be mentioned that the methods of such treatment are not always harmless, not only in relation to accidental mechanical injuries to the specimen, which is especially unpleasant when the specimen is a rarity, but also because of the difficulty of regulating the bleaching process. These obstacles are completely eliminated by the use of infrared photography. 9 In Fig. 7* the head of a mealworm is shown in the ordinary and in the bleached state, photographed in visible and in infrared light. It may be noted that the infrared photograph of the unbleached specimen makes it possible to examine at least all the same details that are also present on the bleached specimen.

In addition, by selecting appropriate plates and filters one can achieve the isolation of individual parts of the observed object.

Some difficulty is caused by focusing the microscope in infrared light. Naumann 10 gives the following practical advice: the microscope is first focused in yellow-green light (for example, with the aid of Agfa filter No. 70) and the position of the micrometer screw is noted; then sharp focus is obtained in red light (for example, with a red Agfa light filter No. 42) and the position of the screw is again noted. Subtracting the second reading from the first and turning the screw back from its position in red light by twice the difference of these readings, one can obtain sharpness in infrared light approximately for 820 mμ. With regard to photographs at greater wavelengths one may add the following: the photographs in Fig. 7 were obtained with a Leitz—Panphot microscope and a “3.8×” objective, a 10× eyepiece* on

* The specimen was kindly provided to us by Prof. Frisch (Munich), for which we express our gratitude to him.

Fresh leaf

Fossil leaf

(Geiseltal)

Fig. 5. Fresh and fossil leaf, photographed in visible and infrared light. Left: orthochromatic photograph; right: infrared photograph.

Fig. 6. Photograph of a leaf damaged by drought. Left: orthochromatic photograph; right: infrared photograph.

Fig. 7a. Microphotograms of the head of a mealworm (scale 63:1). Unstained specimen. On the left, an orthochromatic photograph; on the right, an infrared photograph.

Fig. 7b. Microphotographs of the head of a mealworm (scale 63:1). Illuminated specimen. Left: orthochromatic image; right: infrared image.

infrared plates “Agfa 850 rapid” with a black Agfa No. 87 light filter (transmitting the region 850–900 mµ). After focusing in red light the screw had to be shifted by 130 µ. To obtain a sharp image in infrared light, the screw must be moved by 130 µ from its position for sharpness in red. The exposure in this case, when using this microscope, a nitro lamp, and an open diaphragm, was 1 sec.

In conclusion we shall give some additional photographic data for the infrared photographs illustrating the present article. The second series of photographs (Figs. 1 and 2) was obtained on plates for infrared light, “Agfa 750 rapid,” with a black Agfa No. 83 light filter. The exposure with diaphragm 11, under illumination by three nitro photo lamps of 500 W each, at a distance of 1 m, was 1 sec. The photographs of the third series (Figs. 1 and 2), as well as the infrared photographs (Figs. 5 and 6), were made on “Agfa 850 rapid” plates with Agfa No. 85 light filter. The illumination time was the same, with diaphragm 5.6 and under the same illumination conditions. The photographs of the fourth series (Fig. 2) were made on “Agfa 950 rapid” plates with Agfa No. 889 light filter. The exposure under the same illumination conditions and with diaphragm 4.5 was 13 sec. An “Agfa-Solinar” \(f = 1 : 4.5\) objective in an Agfa-isolar camera was used as the lens. The photographing conditions (Fig. 7) have already been given. Development was carried out in a dilution of 1:20 for 5 min. under illumination by a glow-discharge lamp through an Agfa No. 114 protective light filter.

LITERATURE

1. A. Fröhlich, Dtsch. Drucker 40, 291, 1934.
2. A. Fröhlich, Illustrierte Textilztg. 1933, No. 41/42; Dtsch. opt. Wschr. 19, 477, 1933; Dtsch. Wirkerztg. 1934, 39, 11.
3. Haxthausen, Dermat. Wschr. 35, 1289, 1933; R. T. Payne, Lancet 1934, H. Febr. 3; Brit. J. Photography 81, 69, 1934.
4. Krötz, Vortrag am 9 Januar 1934 in der Med. Ges. in Hamburg.
5. A. Fröhlich u. G. Rodenacker, Münch. med. Wschr. 82, 146, 1935.
6. L. Lewin, A. Miethe u. E. Stenger, Pflügers Arch. 118, 80, 1907.
7. A. Fröhlich u. F. Luft, Umschau 38, 534, 1934.
8. F. C. Bawden, Nature 1933 (29/VII); Brit. J. Photography 80, 2822, 449, 1933.
9. G. Reinert, Zeiss-Nachr. 1933, H. 4, 13; P. Kraft, Ber. über den 8 Internationalen Kongress für Photographie in Dresden 1931, S. 341, Lpz. 1932; F. Pax, Zool. Anz. 106, 15, 1934.
10. H. Naumann, Z. Instrumentenkunde 54, 276, 1934; Blätter für Untersuchungs- u. Forschungsinstrumente, 38, 1933.