ON THE POSSIBILITY OF THERMAL EXCITATION OF CANDOLUMINESCENCE IN CRYSTALLINE PHOSPHORS

V. A. Sokolov

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

The capacity for candoluminescence is usually associated with pure metallic oxides and sulfides of white color. However, Nichols and his collaborators [1] also tested, qualitatively, certain Lenard and Klatt preparations for high-temperature excitation in a flame; these were crystalline phosphors specially prepared for luminescence purposes, with an activator. It was established that a number of such luminophors, at high temperatures, give a glow similar in color to the photo- or cathodoluminescence of these same preparations.

The quenching action of thermal heating on already excited crystalline phosphors is generally known in the case of ordinary types of luminescence. However, the question of the possibility of exciting luminescent centers by transfer to them, from the crystal lattice, of energy stored during heating has until now scarcely been discussed. It seems to us that, in principle, such excitation is entirely possible. Indeed, in recent times voices have increasingly been heard to the effect that even in the case of photoluminescence the crystal lattice plays an important role in the transfer of energy to the luminescence centers.

There is direct evidence that, in a number of cases, a large part of the energy of the exciting light is absorbed by the lattice of the host substance and is then transferred to the activator [2]. As for such kinds of luminescence as are excited by cathode bombardment or by the action of other corpuscular radiations, in these cases the role of the lattice of the host substance in transferring energy to the activator centers of the crystalline phosphor appears still more distinctly. In modern works on the luminescence of crystals, many authors already inevitably arrive at the conclusion that, in luminescent processes, the crystal as a whole acts as a system perturbing the emission centers [3].

Since this is so, the energy stored in any way by the crystal lattice can be transferred to the emission centers of the given crystal. In principle, therefore, energy stored by the lattice as a result of thermal excitation can also be transferred under certain conditions, which in no way contradicts theoretical considerations.

When certain crystalline phosphors are heated to a high temperature—for example, in the flame of a Bunsen burner—one can indeed observe, as in Nichols’s experiments, a glow whose color resembles the luminescence of these same substances. Moreover, the same effect can be obtained both by direct heating in the flame and in isolation from it (for example, by heating in a quartz tube), which indicates the purely thermal nature of the effect. In some cases this kind of effect appears quite distinctly. This applies above all to the ZnS—Mn phosphor. As is known, zinc sulfide gives a temperature glow of a light-green shade. However, the same zinc sulfide, activated with manganese in an amount of 0.01%, i.e., the ZnS—Mn phosphor, shows, under temperature excitation, a bright orange glow characteristic of the photoluminescence of this luminophor. As for the green color characteristic of the temperature glow of pure ZnS, in this case not even a hint of it can be detected by eye, despite the fact that zinc sulfide makes up 99.99% of this crystalline phosphor. In other crystalline phosphors, although less distinctly, one can also notice by eye, in their temperature glow, color shades corresponding to the photoluminescence of these same phosphors. Moreover, the closer to the ultraviolet region the luminescence spectrum of a given crystalline phosphor lies, the more difficult it is to excite the same glow by thermal means.

In light of the question of thermal excitation of luminescence centers, it would be of interest to compare the spectra of temperature glow and the photoluminescence spectra of the same phosphors.

We obtained, in parallel, both kinds of spectra for certain phosphors and carried out such a comparison of them with one another. The photoluminescence spectra were excited by ultraviolet light from a mercury arc, filtered with Wood’s glass. The spectra of temperature glow were obtained either by direct excitation of the phosphor in the flame of illuminating gas, or by heating, with the same flame, a transparent quartz tube containing the luminophor under investigation. As already noted, the character of the phosphor’s temperature glow was the same in both cases. The spectra were photographed with an ISP-51 glass spectrograph. Photometry of the photographs was carried out on an MF-2 microphotometer.

Figure 1 shows the photoluminescence spectrum (a) and the temperature-glow spectrum (b) of the ZnS—Mn phosphor that we obtained. It is easy to see that, first, the two spectra overlap to a significant extent and, second, in the region of the temperature spectrum corresponding approximately to the greatest intensity of the phosphor’s photoluminescence, a secondary maximum is clearly outlined, the appearance of which is due to thermal excitation of the luminescence centers of the crystalline phosphor.

Figure 2 gives the spectra of photoluminescence and temperature glow of the ZnS, CdS—Cu phosphor. In this case the secondary maximum on the curve of the temperature glow (b), corresponding to excitation of the luminescing centers, is likewise indicated quite clearly. Curve (a), as in the preceding case, represents the photoluminescence spectrum of the phosphor.

Approximately the same picture is also observed for the ZnS—Cu phosphor, with the sole difference that the secondary maximum on the temperature-glow curve is shifted somewhat to the right relative to the main maximum of the photoluminescence of this same luminophor.

Phosphors with blue luminescence are thermally excited with great difficulty. Thus, in the ZnS—Ag phosphor the blue glow, charac-

the region for its photoluminescence, in the case of thermal excitation is not visually manifested in an obvious way, and only at comparatively high temperatures (above 1500°C) of the exciting flame is a slight secondary maximum of blue luminescence indicated in the short-wavelength part of the visible spectrum of the phosphor’s thermal radiation.

Fig. 1

The more difficult excitability of the short-wavelength part of the radiation is readily explained on the basis of energetic considerations; however, even in these cases thermal excitation of the luminescent component of the phosphor is entirely possible, with the sole difference that higher temperatures are required for this.

Fig. 2

All this indicates that, by thermal means, the energy stored by the lattice can be transferred to the luminescing centers in such a way that, as a result, they become the source of emission of the same quanta as are emitted, for example, in the photoluminescence of the given crystalline phosphor.

As already mentioned, the excitation of crystalline phosphors in hot flames, as observed by eye, was regarded in Nichols’ school as qualitative proof of the presence of candoluminescence in these preparations. Our experiments on the investigation of the spectra of the temperature radiation of crystalline phosphors, it would seem, could already serve to some extent as quantitative proof of this. However, purely thermal excitation of true luminescence would fundamentally contradict the second law of thermodynamics. Indeed, luminescence is rad—

an excess over the thermal radiation of the body at a given temperature. Therefore, if there existed any body capable, when heated by thermal conduction, of producing luminescence¹, then radiative exchange between it and a normally radiating body would lead to the realization of a perpetual-motion machine of the second kind. Thus, while acknowledging the fact of thermal excitation of the luminescent centers of a crystalline phosphor, we must nevertheless regard the radiation thereby produced as a component of equilibrium thermal radiation.

As is well known, a precise and exhaustive definition of luminescence was given by S. I. Vavilov [4] on the basis of a refinement of Wiedemann’s idea of opposing luminescence to thermal radiation. However, proceeding from this opposition, S. I. Vavilov at the same time emphasized that it is based only on a thermodynamic criterion, and not on the principle of a difference between the elementary acts, which may be the same in either case. Thermal radiation necessarily includes both forced and intrinsic, i.e., luminescent, acts of radiation, but the mutual influences of the elementary processes under conditions of thermal equilibrium are such that the individual properties of the particles are effaced and the general thermodynamic properties of the ensemble appear.

Consequently, in the case of thermal heating as well, we are in principle entitled to apply to a crystalline phosphor the semiconductor model of activator excitation (transfer of an electron, owing to thermal vibrations of the lattice, into the conduction band and subsequent recombination with the ionized center), but we must only remember that the radiation component arising in this process will represent merely a constituent of the equilibrium thermal radiation, which must not exceed, in this region of the spectrum, the radiation intensity of an absolutely black body.

If this is so, then the subsidiary maxima observed by us on the curves of temperature radiation of crystalline phosphors must arise in full accordance with Kirchhoff’s law.

M. N. Alentsev [5] drew our attention to the fact that these maxima can quite well be explained on the basis of Kirchhoff’s law if Stokes–Lommel’s rule is taken into account. Indeed, according to this rule, the absorption spectra and the photoluminescence spectra of a crystalline phosphor overlap one another as indicated in Fig. 3. Consequently, in the region of photoluminescence with which we compared the spectrum of temperature radiation, there is additional absorption, which may even extend somewhat beyond the luminescence maximum. According to Kirchhoff’s law, the emissive power of a body \(E\) is equal to its absorptive power \(A\) in the given spectral region and at the given temperature, multiplied by the emissive power of an absolutely black body \(\varepsilon_0\) under the same conditions.

If the character of the absorption of the crystalline phosphor according to Fig. 3 is taken into account, as well as the fact that the function \(\varepsilon_0\) increases considerably more rapidly than \(A\), then the appearance of a subsidiary maximum on the curve of temperature radiation of the crystalline phosphor in the region corresponding to its photoluminescence proves to be regular.

Summarizing all the foregoing, the following may be said. Thermal excitation of crystalline phosphors is possible. This is supported both by theoretical considerations and by directly carried out—

… experiments. However, the luminescence component arising in this case, corresponding to the photoluminescence of the crystalline phosphor, is in itself only a constituent of the thermal radiation that arises in accordance with Kirchhoff’s law. Thus, the so-called candoluminescence of crystalline phosphors observed at high temperatures is not

Fig. 3

true luminescence and is explained only by the peculiarities of thermal radiation. The existence of luminescence in its generally accepted sense in the case of purely thermal excitation of a substance would contradict the second law of thermodynamics.

REFERENCES

1. Nichols E. L., Howes H. L. and Wilber D. T., Cathodoluminescence and the Luminescence of Incandescent Solids, Washington, 1928.

2. Morgenstern Z. L., DAN SSSR, vol. 105, 250, 1955.

3. Alperovich E. I., Some Problems in the Theory of Luminescence of Crystals, Moscow—Leningrad, 1951.

4. Vavilov S. I., Luminescence and Light Duration. Jubilee collection of the Academy of Sciences of the USSR dedicated to the 30th anniversary of the October Socialist Revolution, p. 377, Moscow—Leningrad, 1947.

5. Alentsev M. N., Remarks in the discussion of the report by V. A. Sokolov “On the Candoluminescence of Crystalline Phosphors,” Izv. AN SSSR, ser. fiz., vol. 21, 528, 1957.

ERRATA

Izv. TPI, vol. 95.

1. In the present case thermoluminescence is excluded, since it is caused by heating only under the condition of preliminary excitation of the phosphor by some other agents. ↩