STUDY OF AN ILLUMINATING GAS FLAME AS A SOURCE OF CANDOLUMINESCENCE EXCITATION

O. G. Brik

(Submitted by Professor V. A. Sokolov)

In the flame of illuminating gas [1], as also in a hydrogen flame [2] and in an atmosphere of active gases [3, 4], luminescence of crystallophosphors is observed, which has been given the name candoluminescence. This luminescence is due in its origin to recombination on the surface of the luminophor of free atoms and radicals contained in the flame [5, 6, 7] and responsible for the chain character of the combustion reaction [8]. The energy released during recombination is the source of excitation of such radical luminescence.

Candoluminescence of luminophors under the action of an illuminating gas flame is an experimentally established fact [9]. However, up to now only the integral action of this flame has been studied. Therefore the purpose of the present work is to study radical luminescence in the different zones of an illuminating gas flame, to determine which components of the illuminating gas take part in radical-recombination luminescence, and also to clarify the influence on luminescence of foreign impurities introduced into the flame ($J_2$, $N_2$, $CO_2$, $O_2$, KCl, and NaCl).

Experimental Method

A flame with clearly distinguishable zones was obtained on a burner with a section of $28 \times 8$ mm. The illuminating gas was supplied through a nozzle; the air was injected. The luminophor was introduced into the flame by means of a rotating cylinder with a cooling mixture, according to Sokolov’s method [1], and on a surface cooled by water. The luminescence spectrum was recorded on an ISP-51 spectrograph; the spectrogram was photometered on an MF-2 microphotometer.

The temperature of the flame was measured with a platinum–platinum-rhodium thermocouple 0.1 mm in diameter with a quartz test tube 0.8 mm in diameter. The concentration of free atoms and radicals was measured by the method of a differential thermoelectric probe [10]; the wire diameter was 0.05 mm, the thermocouple was chromel–alumel; one junction of the thermocouple was covered with a quartz test tube with a deposited catalyst, the other junction with a quartz test tube washed with nitric acid and distilled water. A mixed oxide $ZnO \cdot Cr_2O_3$ [11] was used as the catalyst.

A flame with separated cones was obtained with the aid of a Smithells–Ingle flame separator [12]; the diameter of the inner tube was 10 mm, the diameter of the outer tube 25 mm. The primary and secondary air were injec-

... was calibrated (Fig. 1). The outer tube was washed with hot nitric acid and distilled water in order to stabilize the surface and obtain reproducible results.

Fig. 1. Burner with separated cones: I — primary air; II — secondary air; G — combustible gas; 1 — preheating zone; 2 — reaction zone; 3 — combustion products; 4 — first diffusion zone; 5 — second diffusion zone (outer cone).

Iodine was obtained by sublimation of crystalline iodine; carbon monoxide, by passing CO₂ over charcoal at a temperature of 1100°C according to the reaction

\[ \mathrm{CO_2 + C \to 2CO}, \]

with Ascarite serving as the CO₂ absorber. Oxygen, nitrogen, and carbon dioxide were taken from cylinders.

Chemical analysis of the gases was carried out on an KhT-2M chromatograph.

As the ultraviolet source for exciting photoluminescence, a PRK-4 mercury-quartz lamp was used.

The phosphors studied were: ZnS·CdS (Cu), ZnS·CdS (Mn), ZnS (Cu), Zn₂SiO₄ (Mn).

According to our analysis, the composition of the gas under study was as follows:

\[ \begin{aligned} \mathrm{H_2} &— 51\% \\ \mathrm{CH_4} &— 30.1\% \\ \mathrm{CO} &— 8\% \\ \mathrm{CO_2} &— 4.1\% \\ \mathrm{N_2} &— 4.3\% \\ \mathrm{O_2} &— 1.1\% \\ \mathrm{C_2H_2} &— 1.0\% \end{aligned} \]

Results of the Study

1. Studies in the Broad Flame of a Flat Burner

a. Spectra of radical-recombination luminescence of phosphors. The broad flame obtained on a flat burner has clearly distinguishable zones: the preheating zone, the reaction zone, the zone of combustion products, and the diffusion reaction zone (outer cone). The reaction zone (inner cone) has a height of 3 cm; the thickness of the zone is 0.2–0.3 mm, and its color is greenish-blue. The height of the outer cone is 13.5 cm; the apex of the cone fluctuates, which indicates a change in equilibrium during the combustion process.

The broad flame made it possible to introduce the phosphor into different zones. Excitation of radical-recombination luminescence was observed only in the outer zone. In the reaction zone and in the zone of combustion products, no luminescence of these phosphors was detected. The spectrum of radical-recombination luminescence coincides with the photoluminescence spectrum of these phosphors. The maximum of the spectral-intensity curve for the phosphor ZnS·CdS (Mn) lies in the region of 600 mμ; for Zn₂SiO₄ (Mn), in the region of 530 mμ; and for ZnS (Cu), at 550 mμ.

b. Measurement of the concentration of active particles. Gorban’ and Sokolov [13, 14] substantiated the applicability of the thermoelectric-probe method [10] for determining the concentration of active particles and the recombination coefficient in an illuminating-gas flame. The relative heating, measured by us, of a thermocouple coated with ZnO · Cr₂O₃ oxide in various zones of the flame is:

in the diffuse reaction zone \(\Delta T = 64\text{--}72^\circ\),

in the luminous zone \(\Delta T = 8\text{--}14^\circ\),

in the combustion-products zone \(\Delta T = 18\text{--}28^\circ\).

According to the formula of Kondrat’eva and Kondrat’ev [10]

\[ \frac{P_{\mathrm{H}}}{P}=\pi\frac{\Delta T}{T} \]

(for illuminating gas \(\Pi = 1000\)), the concentration of atomic hydrogen was calculated for the flame zones, which corresponds to 7.1 and 2.2% of the composition of the combustible mixture.

c. Determination of the recombination coefficient of active particles on the surface of phosphors. The heating of the phosphor Zn₂SiO₄ (Mn) proved to be 1.5 times greater than the heating of ZnS · CdS (Cu), and that of ZnS · ZnSe (Cu) 1.2 times greater.

According to [13], the recombination coefficient \(\varepsilon\) for the phosphor ZnS · CdS (Cu) is 0.33–0.4. Thus, for Zn₂SiO₄ (Mn) \(\varepsilon = 0.49\text{--}0.61\), and for ZnS · ZnSe (Cu) \(\varepsilon = 0.40\text{--}0.48\), relative to ZnO · Cr₂O₃, for which \(\varepsilon = 1\) [10, 11, 13].

d. Effect of impurities. The effect of the following impurities added to the illuminating-gas flame on the intensity of radical-recombination luminescence was investigated: the salts KCl and NaCl, the gases CO₂ and iodine, and water.

The introduction of KCl, NaCl, and CO₂ into the flame weakens the intensity of the radical-recombination luminescence observed visually. But when the action of these impurities ceases, the luminescence intensity is restored at once. The flame temperature upon introduction of KCl and NaCl is lowered by 30°–40°.

An iodine impurity in the flame extinguishes the radical-recombination luminescence. It is restored 2–3 sec after the iodine is removed from the flame.

The flame temperature upon introduction of iodine is lowered by 130°. A moistened phosphor does not luminesce in the flame, but after drying in air the luminescence is restored.

2. Study in a carbon monoxide flame

In a flame of CO + 2% H₂ with air at atmospheric pressure, the luminescence intensity of the phosphor proved to be 4–5 times less than the intensity in an illuminating-gas flame. The heating of a thermocouple coated with ZnO · Cr₂O₃ was 28°–30°, which corresponds to an atomic-hydrogen concentration in the flame of 3%. The flame temperature was \(1050^\circ \pm 10^\circ\mathrm{C}\).

3. Study in a burner with separated cones

a. Luminescence of phosphors by flame zones. Bright luminescence was observed along the line of contact of the inner flame zone with the phosphor applied to the inner surface of the outer tube. After some time the luminescence, decreasing, disappeared. With sufficient cooling of the outer tube, the radiation po-

sources of luminescence continued for a long time. In the intercone gases, luminescence of the phosphors was not observed. The luminescence of the phosphor in the outer zone coincides in spectral composition with the luminescence in the inner zone.

b. The concentration of hydrogen atoms in the inner and outer zones is ~6%; in the combustion products the concentration of active particles decreases, with increasing distance from the reaction zone, from 2% to 0.2% and lower.

c. The temperature distribution in the intercone gases and in the outer zone is given in Fig. 2. The maximum temperature at the wall of the tube is 710°C; in the inner zone it is 980°. The temperature in the outer zone is ~150° higher than in the inner one.

Fig. 2. Graph of the temperature distribution in a burner with separated cones: 1 — along the diffusion zone of reaction of the outer cone; 2 — along the axis of the outer cone; 3 — at the wall of the tube; C — temperature at the point of contact of the first diffusion zone with the wall, \(T_c = 710^\circ\)C.

d. Effect of impurities. The effect of impurities of the following gases was studied: nitrogen, carbon dioxide, carbon monoxide, oxygen, iodine. The gases were supplied in a mixture with secondary air.

An impurity of nitrogen and carbon dioxide decreases the flame-propagation rate; an impurity of carbon monoxide and oxygen increases the flame-propagation rate up to detonation. When oxygen is added, the flame temperature rises by 400–450°. All these impurities somewhat weaken the luminescence, while an impurity of oxygen and iodine extinguishes the luminescence, just as in the broad flame of a flat burner.

Discussion of the results. Since in the illuminating gas used the concentration of hydrogen is 51%, methane 30.1%, and carbon monoxide 8%, in the combustion process the principal role will be played by the oxidation of these gases. Taking into account the mechanism of oxidation of these gases [8, 15, 16, 17, 18], it may be said that the following radicals should be present in the flame: H, O, OH, HO\(_2\), CH\(_3\), HCO, CH\(_2\)O. Authors who have studied the flame of hydrogen, methane, and carbon monoxide point to an anomalously high concentration of H, O, and OH in the flame zone. Thus, in a hydrogen flame, (H) is measured in tens of percent, (OH) = 1%, (O) —

fractions of a percent; in the methane flame the concentration of atoms is \((\mathrm{H}) \gg (\mathrm{OH}) > (\mathrm{O})\); in the carbon monoxide flame the concentration of \((\mathrm{H})\) is comparable with \((\mathrm{O})\), and the ratio of the concentrations of active particles \((\mathrm{H}) : (\mathrm{O}) : (\mathrm{OH}) = 30 : 5 : 1\). And if one takes into account that the energy of recombination of hydrogen atoms into a molecule is very large, then one should assume the leading role of H atoms in the excitation of candoluminescence.

The different intensity of luminescence in different zones of the flame of the wide burner seems to us to be interconnected with the concentration of H atoms in these zones. The maximum intensity in the diffusion zone corresponds to the maximum concentration of H atoms in this same zone. In the inner zone of the flame the process is only beginning; its development is taking place, and the concentration of free atoms and radicals there is small, increasing gradually as equilibrium is established in the combustion process. The same idea is supported by the observed decrease in the luminescence intensity of the phosphor in the carbon monoxide flame with a decrease in the concentration of active particles.

The quenching action of iodine can be explained by the reaction [19]

\[ \mathrm{H} + \mathrm{J}_2 \to \mathrm{HJ} + \mathrm{J}, \tag{1} \]

i.e., iodine molecules, reacting with atomic hydrogen, considerably decrease its concentration in the flame.

Along with this reaction, we considered possible the destruction of active particles in the volume during a triple collision according to the reactions:

\[ \mathrm{H} + \mathrm{H} + \mathrm{J} \to \mathrm{H}_2 + \mathrm{J}^{*}, \tag{2} \]

\[ \mathrm{H} + \mathrm{H} + \mathrm{J}_2 \to \mathrm{H}_2 + \mathrm{J}_2^{*}, \tag{3} \]

\[ \to 2\mathrm{HJ}^{*} \tag{3a} \]

\[ \mathrm{H} + \mathrm{H} + \mathrm{HJ} \to \mathrm{H}_2 + \mathrm{HJ}. \tag{4} \]

The spectroscopic study we carried out of an illuminating-gas flame with an admixture of molecular iodine showed that, in the visible region, emission lines of excited atomic iodine \(\mathrm{J}^{*}\) are absent. Consequently, if reaction (2) occurs, then the energy of recombination of atomic hydrogen into a molecule is insufficient to excite the radiation of atomic iodine in the visible region.

The bands we found, compared with the data of Pearse and Gaydon [20], indicate the existence of excited molecular iodine. Excitation of molecular iodine occurs by reaction (3) or by the reaction

\[ 4\mathrm{HJ} + \mathrm{O}_2 \to 2\mathrm{J}_2 + 2\mathrm{H}_2\mathrm{O}, \tag{5} \]

since it is known [21] that molecular hydrogen iodide burns well with oxygen.

Thermodynamic calculations carried out by us using the data of Lavrov and co-workers [22] indicate the possibility of reaction (3a). But it is known [21] that \(\mathrm{HJ}\) dissociates well already at a temperature of \(300^\circ\mathrm{C}\), while the flame temperature is about \(1050^\circ\mathrm{C}\). Consequently, the hydrogen iodide formed by reactions (1) and (3a) partly decomposes according to the reaction

\[ \mathrm{HJ} \to \mathrm{H} + \mathrm{J} \tag{6} \]

and partly disappears according to reaction (5).

Atomic iodine, formed in the process of dissociation of the iodine molecule in the flame

\[ \mathrm{J}_2 \to \mathrm{J} + \mathrm{J} \tag{7} \]

and, through reactions (1) and (6), is adsorbed by the surface of the phosphor, binding to the free valences of the surface [23], and thus the surface of the phosphor becomes poisoned for the adsorption of active particles. After some time, as a result of the thermal action of the flame, iodine atoms volatilize from the surface, and the surface again becomes active for the adsorption of active flame particles. This is evidenced by the phenomenon we observed: complete quenching of the phosphor glow in a flame containing an iodine admixture, and restoration of the glow 2–3 sec after the iodine is removed from the flame. When KCl and NaCl salts were introduced into the flame, a decrease in the phosphor glow was observed, followed by restoration of the glow immediately after these salts were removed from the flame.

In parallel, the influence of molecular iodine on photoluminescence was studied. It turned out that the surface of the phosphor adsorbs molecular iodine very well. The luminescence disappears. At room temperature, iodine remains on the surface for a long time. When the temperature is raised, the iodine rapidly volatilizes, and the phosphor again luminesces.

The considerable decrease in flame temperature upon introduction of iodine (down to 150°C) can probably be interpreted as a change in the reaction rate (and hence also in the flame velocity) due to the destruction of atomic hydrogen. As is known, the flame velocity depends on the concentration of radicals [12].

The extinguishing action of water can be explained by the fact that water is a strong poison with respect to the recombination reaction of H atoms on the surface [15].

The weakening action of NaCl and KCl salts can be explained by an increase in the recombination of active particles in the volume owing to an increase in the concentration of third particles in the mixture; the active particles perish in the volume according to the reaction

\[ \mathrm{H} + \mathrm{H} + \mathrm{M} \to \mathrm{H}_2 + \mathrm{M}. \tag{8} \]

As our observations have shown, the inner cone in the flame separator is surrounded by a diffusion sheath formed as a result of the diffusion of oxygen from the secondary air into the reaction zone. The luminescence observed in the tube is caused by the energy of recombination of radicals formed in the diffusion reaction zone on the surface of the phosphor.

The weakening of luminescence under the action of CO₂ and N₂ admixtures can be explained by worsened conditions for the diffusion of active particles to the phosphor surface, and under the action of an oxygen admixture—by temperature quenching.

Conclusions

1. The most active region of the flame with respect to excitation of radical luminescence is the outer zone, which contains the greatest number of free atoms and radicals.

2. In a burner with separated combustion, active zones exist in both the inner and the outer flame.

3. Luminescence in the flame is produced only by catalytically active phosphors.

4. Quenching of the luminescence of crystallophosphors in a flame by foreign admixtures confirms its radical-recombination mechanism.

5. The radicals (free atoms) of hydrogen are chiefly responsible for the radical luminescence of phosphors under the conditions of an illuminating-gas flame.

References

1. V. A. Sokolov. Optics and Spectroscopy. 4, 409, 1958.

2. Smith E. C. Inst. Gas. Engr. 3, 237, 1940.

3. Haule W. und Nierman H. Zeitschrift Naturforsch. 11a, 5, 395 (198).

4. Sancier K., Fredericke W., Hatcher J. and Wise H. J. Chem. Phys. 37, 4, 860, 1962.

5. A. N. Gorban’ and V. A. Sokolov. Optics and Spectroscopy. 7, 569, 1959.

6. V. A. Sokolov and A. N. Gorban’. Proceedings of the IX Conference on Luminescence, Izv. AN SSSR, ser. fizich., 25, 424, (1961).

7. F. F. Volkenshtein, A. N. Gorban’ and V. A. Sokolov. Surface Properties of Semiconductors. Publishing House of the Academy of Sciences of the USSR, 179, 1962.

8. N. N. Semenov. Chain Reactions. ONTI, 1934.

9. V. A. Sokolov. Low-temperature candoluminescence. Izv. TPI (in press).

10. E. I. Kondrat’eva and V. N. Kondrat’ev, ZhFKh, XX, 11, 1239 (1946).

11. Smith W. V. J. Chem. Phys. 11, 110 (1943).

12. A. G. Gaydon. Spectroscopy of Flames. IL, 1959.

13. A. N. Gorban’ and V. A. Sokolov. Optics and Spectroscopy. 7, 259, 1959.

14. A. N. Gorban’. Dissertation “On the Question of the Physicochemical Nature of Candoluminescence,” Tomsk, 1959.

15. B. Lewis and G. Elbe. Combustion, Flames and Explosions in Gases, 1948.

16. V. N. Kondrat’ev. Spectroscopic Study of Chemical Gas Reactions. Moscow, 1944.

17. L. V. Karmilova, N. S. Enikolopyan and A. B. Nalbandyan. ZhFKh, XXX, 4, 798, 1963.

18. N. A. Klemenov, I. N. Antonov, A. M. Markevich and A. B. Nalbandyan. ZhFKh, XXX, 796, 1963.

19. N. N. Semenov. On Certain Problems of Chemical Kinetics and Reactivity. Moscow, 1958.

20. R. Pearse and A. Gaydon. The Identification of Molecular Spectra. IL, Moscow, 1949.

21. B. V. Nekrasov. Course of General Chemistry. Moscow, 1960.

22. F. F. Volkenshtein. Electronic Theory of Catalysis on Semiconductors. Moscow, 1961.

23. N. Lavrov, V. V. Korobov and V. I. Filippova. Thermodynamics of Gasification and Synthesis Reactions from Gases. Publishing House of the Academy of Sciences of the USSR, 1960.