On the Nature of Candoluminescence of Calcium Oxide

V. A. Sokolov, I. S. Grozina, A. N. Gorban

(Presented by Professor Doctor A. A. Vorob’ev)

Calcium oxide belongs to the group of oxides whose bright glow under the action of high-temperature flames is associated with the so-called candoluminescence of these oxides. A sign of candoluminescence is that the emissive capacity of a given substance exceeds the radiation of a black body at high temperatures. Thus, when excited in flames containing hydrogen, calcium oxide, according to Nichols [1], can, in the temperature interval from 600 to 1000°C, exceed the radiation of an absolutely black body under the same conditions by tens and even hundreds of times. The principal cause of candoluminescence, according to Nichols’s hypothesis, is seen in reduction–oxidation processes occurring in the oxide under the influence of the flame (reduction of the metal and its subsequent oxidation, accompanied by chemiluminescence).

However, the data of Nichols and his school on the super-emissive capacity of calcium oxide, as well as of other similar oxides, have been disputed. The unreliability of these data follows already from the fact that black uranium oxide or copper oxide was taken as the absolutely black body, whereas under the conditions of one and the same flame these may assume a temperature lower than that of the calcium oxide being compared with them, or of other similar white, selectively radiating oxides [2].

In addition, V. M. Kudryavtseva, G. I. Sinyapkina, and V. A. Sokolov [2, 3] reliably established that the so-called “candoluminescence” of ZnO, ZnS, and TiO₂ obeys the usual laws of thermal radiation and, consequently, does not constitute true luminescence in the sense of the Vavilov–Wiedemann definition.

In the present work we set ourselves the aim of investigating the applicability of the laws of thermal radiation to the emission spectra of CaO under flame excitation, and also to the spectra of Ca oxidation, in order to test Nichols’s oxidation–reduction hypothesis. In addition, we investigated the emissive capacity of CaO in comparison with the emission of a laboratory model of an absolutely black body.

The emission spectra (“candoluminescence”) of CaO were obtained by heating, with an illuminating-gas flame, a solid oxide specimen placed in front of the slit of a spectrograph (ISP-51). The oxidation spectra were obtained by burning pieces of metallic calcium in a quartz flask filled with oxygen and placed in front of the spectrograph. The spectra were photographed on “isopanchrome” film with po-

by subsequent photometry on an MF-2 microphotometer. The spectral curves were constructed with allowance for the contrast factor and the spectral sensitivity of the film.

Figure 1 shows the spectrum of calcium oxidation obtained by us and the emission spectrum of calcium oxide in an illuminating-gas flame for that region of wavelengths which is associated with CaO candoluminescence. It is easy to see that both spectra are continuous in character and, to a considerable extent, repeat one another. The latter indicates that the glow of the flame of burning calcium is due mainly to the radiation of incandescent particles formed as a result of the oxide reaction. In appearance, both curves in Fig. 1 are ordinary spectra of the temperature radiation of solids. If this is so, then according to Kirchhoff’s law

\[ E = AC_1\lambda^{-5} e^{-\frac{C_2}{\lambda T}}, \]

they should vary with frequency \(\left(\frac{1}{\lambda}\right)\) proportionally to the quantity

\[ \lambda^{-5} e^{-\frac{C_2}{T}\frac{1}{\lambda}}. \]

Indeed, as Fig. 2 shows, the experimental points corresponding to the values \((\lg E + 5\lg\lambda)\) lie comparatively well on straight lines representing, for both spectra, the functional dependence of the quantity \((\lg E + 5\lg\lambda)\) on \(\frac{1}{\lambda}\).¹

Fig. 1

Thus, in both cases, for the investigated region of the spectra, to which candoluminescence is attributed, the decisive role is played by ordinary thermal radiation. A second conclusion can be drawn concerning the untenability of the reduction–oxidation hypothesis with respect to CaO. Indeed, if in the oxidative process in its pure form we do not observe chemiluminescence phenomena, then all the more we cannot observe them in oxidative processes that may be supposed to result from partial decomposition of the oxide heated by the flame.

The next stage of our investigations concerns the study of the emissive power of calcium oxide in comparison with the emission of an absolutely black body. As already mentioned, Nichols, comparing through a blue filter (with the aid of an optical pyrometer) the glow of CaO with the radiation of black oxide placed in the same flame, observed a considerable excess of the emissive power of calcium oxide relative to black oxide. The erroneous assumption in Nichols’ investigations was the supposition of the same temperature for both oxides. In our

At the same time, a comparison was made of the brightness of calcium oxide with the brightness of black copper oxide, under the condition that each oxide was excited by the flame of a separate burner, and the rate of combustion in each of them was maintained such that the temperatures of the oxides were, as far as possible, identical. The comparison was carried out by means of a Pulfrich photometer, and the temperatures of the oxides were monitored with thermocouples. As was to be expected, we did not obtain the enormous excesses observed by Nichols,

Fig. 2

but nevertheless a small excess (by 10–20%) in the radiating power of CaO as compared with CuO for the blue region of the spectrum was noted by us. This even led us to doubt the complete absence of candoluminescence in CaO.

M. V. Fok [4] pointed out that such experiments also cannot be considered entirely reliable, since the radiation of the oxides was studied under conditions in which thermal equilibrium could hardly be established (the absence of a closed cavity). Therefore, in another series of experiments, as the black body we used no longer copper oxide, but a cylindrical model of an absolutely black body; moreover, the oxide under investigation was likewise enclosed in a closed cylindrical cavity, in which colorless combustion of illuminating gas was carried out¹ by appropriately regulating the forced supply of air. The oxide heated to incandescence in the cavity could be observed through a specially made small aperture, whose diameter was equal to the diameter of the aperture in the black-body model employed. The temperature of both cavities was maintained the same. Temperature control was carried out with a platinum–platinum-rhodium thermocouple. The comparison of brightnesses was performed by means of a Pulfrich photometer.

Experiments with calcium oxide carried out in this way, in the temperature range from 600 to 900°C (our black-body model did not provide a higher temperature), did not show any noticeable excess in the radiating power of CaO relative to the black body when observed through a blue filter. Although, according to Nichols’s data, it is precisely in the blue

region, in the temperature interval from 665°C to 837°C, the candoluminescence of CaO is especially pronounced.

It should be noted, however, that according to our measurements in the aforementioned part of the spectrum the emissive power of CaO reached, or almost reached, the emissive power of a black body. This, however, is also a peculiarity of the thermal radiation of white oxides [5]. Emitting selectively in the visible region, they emit comparatively little in the infrared part of the spectrum, as a result of which in a flame they acquire a higher temperature and, consequently, glow more brightly than other bodies. It was precisely this property of calcium oxide that was formerly used in beacons, projection lamps, etc., in the form of the so-called “Drummond light,” obtained by heating lime cylinders in an oxyhydrogen flame.

We believe that the set of studies we have carried out on the luminescence of CaO already permits us to state with confidence that the so-called “candoluminescence” of this oxide under the action of a flame has, at high temperatures (above 600°C), a purely thermal nature and is least of all connected with true luminescence in the sense of the Vavilov–Wiedemann definition.

REFERENCES

1. Nichols E. L., Howes H. L., and Wilber D. T. Cathodoluminescence and the Luminescence of Incandescent Bodies. Washington, 1928.
2. Sokolov V. A. UFN, 47, 537, 1952.
3. Kudryavtseva V. M. and Sinyapkina G. I. DAN, 59, 1411, 1948.
4. Fok M. V. Remarks in the discussion of V. A. Sokolov’s report at the Fifth Conference on Luminescence. Izv. AN SSSR, ser. fiz., 21, 528, 1957.
5. Ribo. Optical Pyrometry. GTTI, Moscow, 1934.

1. The oxide under investigation was heated not electrically, but with illuminating gas, because it is precisely the action of the flame that is attributed with producing the candoluminescence of oxides. ↩↩