On the Catalysis of the Thermal Decomposition of Heavy-Metal Azides by Heterophase Additives

Yu. A. Zakharov, E. S. Kurochkin, G. I. Savel’ev

(Presented by the scientific seminar of the Department of Radiation Chemistry)

A considerable number of works have been devoted to questions of the catalysis of the thermal decomposition of various inorganic compounds by heterophase additives. In them, several variants of the mechanism of catalysis and inhibition of the thermal decomposition of solid substances are considered. The models of catalysis discussed in the literature, however, are constructed with allowance for only certain physicochemical phenomena occurring at the interface between the thermostable additive and the decomposing compound. At the same time, special studies on the physics and chemistry of contacts between two solid bodies show that the range of phenomena and processes taking place at a contact is broad. Some of these phenomena have to date not been taken into account in explanations of heterophase catalysis and have not been specially investigated. These include, first of all, the phenomena of mutual diffusion of components at the contact boundary and the melting of substances in contact at a temperature lower than the melting temperature of the pure compounds. The latter is known in the literature as “contact melting.” The mechanism of contact melting, however, has not been definitively established. Savintsev and coauthors [1, 2] believe that melting is preceded by mutual diffusion and the formation of low-melting eutectics. Zalkin [3], however, considers contact melting as the result of the interaction of [[unclear: word cut off at page edge]]

surface force fields of the substances in contact.

For the development of ideas concerning heterogeneous catalysis of thermolysis, it is necessary, first, to study in each specific case the processes and phenomena occurring at the contact and, second, to take phenomena of this kind into account jointly and to consider the question of their role in catalysis.

In the present work, a variant of a broader consideration of the phenomena of catalysis is set forth on the basis of experiments we have carried out to observe the processes occurring at the contact for a significant group of inorganic semiconductors with heavy-metal azides.

METHODS OF INVESTIGATION

In connection with the task posed, along with studying the kinetics of the thermal decomposition of azides in the presence of catalysts by means of gravimetric and manometric methods, it was necessary to carry out visual observations of the decomposition process using a microscope. The question of mutual diffusion at the azide–additive contact was examined by means of chemical analysis. Experiments in differential thermal analysis (DTA) and thermogravimetric analysis (TGA) were also performed using a derivatograph and a MacBain balance.

In order to discover general regularities and, possibly, a larger number of phenomena in solid–solid catalysis, a wide range of catalysts was selected, and more than 120 systems were investigated using the methods described above. However, in presenting the experimental material in this article, we shall confine ourselves to describing the most general, typical, and characteristic results.

Silver, lead, and thallium azides were obtained by precipitating them upon rapid mixing of 0.2 N solutions of sodium azide, twice recrystallized from distilled water, and the corresponding nitrate. In the experiments, azide powders with a dispersity of 0.03 mm were used. As additives mechanically admixed with the azides, we used oxides, sulfides, and selenides of metals. All the catalysts we studied were, as a rule, of “chemically pure” grade.

The methods for carrying out microscopic observations of the decomposition of azides, both pure and with additives, and the DTA experiments are described in [4].

Chemical analysis of silver and lead azides for the content of the additive cation was performed by extracting the dithizonates of the corresponding metals with carbon tetrachloride from the inorganic phase. For this purpose, three layers of 20 mg of azide were applied to a catalyst tablet and heated in vacuum at 120–140°C for four hours. After heating, the azide was separated from the tablet and dissolved in the corresponding solvent: silver azide in an aqueous ammonia solution; lead azide in acetic acid. The pH of the solution was adjusted to a value corresponding to the complete extraction of ions of the elements being analyzed [5], and a dithizone solution in CCl₄ was added. Thus, the compositions we investigated were $AgN_3 + CuO$, $CoO$, $NiO$, $Cu_2Se$, and $PbN_6 + CuO$, $NiO$. Coloration of the organic phase in a color characteristic of the given ion indicated the presence of the ion being detected.

RESULTS AND DISCUSSION

From the data we obtained it follows that the greatest catalytic activity in the thermal decomposition of silver and thallium azides is exhibited by sulfides and oxides of transition—

metals—copper, cobalt, nickel, manganese, and also copper selenide. Some results for $AgN_3$, obtained by means of a volumetric technique, are presented in Fig. 1. With copper(I) sulfide, zinc and copper(I) selenides, and a number of other catalysts, under the experimental conditions ($175^\circ C$, vacuum) the azide explodes. Bismuth(III), chromium(III), and strontium oxides and lead iodide are inhibitors of the slow thermal decomposition of $AgN_3$.

Our relatively qualitative observations, made with an MIS-2 microscope, of the decomposition process showed that, in the presence of catalysts (oxides of nickel, copper, cobalt, manganese, sulfides), more intense changes in the color of the azides occur. Characteristic changes in silver azide also occur with additives that, according to kinetic data, retard thermal decomposition. Microscopic studies in vacuum of mixtures of silver azide with chromium(III) and bismuth(III) oxides and with lead iodide showed the general picture. The sequence of color change (violet/gray-white) during heating is the same as for pure $AgN_3$, but these changes occur at relatively higher temperatures than in pure azide. Consequently, intense changes in the color of azides, as well as their retardation in time and temperature, can qualitatively indicate the catalytic and inhibiting character of the influence of additives on the thermal decomposition of azides.

Another characteristic feature we found in the course of microscopic studies is a change in the melting temperature of silver and thallium azides in the presence of heterophase additives. According to the literature data, the melting temperatures of silver and thallium azides are $250^\circ$ and $334^\circ C$, respectively [6]. In our experiments,

of porous silver azide melts in vacuum do not melt even at \(330^\circ\mathrm{C}\), whereas in air melting always occurs at \(310^\circ\mathrm{C}\). The latter is confirmed also by DTA data (Fig. 2), where \(AgN_3\) is endothermic—the peak of melting at \(310^\circ\mathrm{C}\). Single crystals of \(AgN_3\), however, melted in air at \(250^\circ\), and in vacuum their melting temperature was somewhat higher, \(\sim 255\text{–}260^\circ\mathrm{C}\). Some data on the melting temperature of single crystals and powder of silver azide in vacuum and in air, as well as with additives—mainly in vacuum—are presented in Table I. Data on the melting temperature in air for pure \(TlN_3\) powder and with additions of \(MnS\), \(CoS\), \(Cu_2S\) are also presented there. As follows from the table, the additives substantially lower the melting temperature of the azides.

In all cases melting is accompanied by decomposition of thallium azide, as indicated by more intense gas evolution in the presence of catalysts. In the case of silver azide, melting is as a rule accompanied by an explosion. These qualitative changes are consistent with the DTA data. From the DTA data it follows that the phase changes of azides mixed with catalysts correspond to exothermic effects on the DTA curves. This indicates decomposition of the azides already at early stages of heating; confirmation is also provided by TA experiments. Melting of azides, accompanied by an intense exothermic reaction, leads to explosion. Some DTA results for silver and thallium azides are presented in Figures 2 and 3. The additives catalyze decomposition of the azides also in the liquid phase, which is most distinctly manifested for thallium azide. In cases where melting of \(TlN_3\) is not accompanied by explosion, strongly increased exothermic peaks are observed on the DTA curves in comparison with pure \(TlN_3\) (cf. curves 1 and 3, Fig. 3).

The presence of copper, cobalt, and nickel ions in \(AgN_3\) and copper and nickel ions in \(PbN_6\), detected by us with the aid of chemical

qualitative analysis, as well as the presence of silver and lead ions in the catalyst, indicate mutual diffusion of the cations in the solid state during the process of thermal decomposition. The result of such interaction, along with other causes, may be a change in the thermal stability of heavy-metal azides.

In connection with these experiments, the so-called “memory” effect observed in catalysis of the thermal decomposition of azides can be explained by diffusion; this effect consists in the fact that the influence of additives is preserved up to high degrees of decomposition, although this should not occur because of shielding of the additive by metal products of the azide thermolysis reaction. The presence of a “memory” effect was shown by us in a direct experiment carried out in a “sandwich” geometry. Removing the catalyst from the surface of an azide tablet after the azide had decomposed by 4% (Fig. 4), followed by its subsequent decomposition under the same conditions (the interrupted experiment), does not affect the catalytic effect in the thermolysis reaction.

Thus, the results presented indicate that catalysis of the thermal decomposition reactions of solid ionic compounds, in particular azides, can proceed by several mechanisms.

First of all, it is probably necessary to consider the melting effects of azides in contact with additives at temperatures below the melting point of the pure compounds—“contact melting.” The mechanism of this phenomenon, as we have already noted, has not yet been definitively established; nevertheless, in any case, the transition of the substance to a more reactive liquid state will lead to its more rapid decomposition.

On the other hand, mutual diffusion of ions should lead to the formation of solid solutions in azides. In this case, the same effects should be expected as in the case of a melt of salts containing similar impurities. Indeed—

however, in the case of azides there exists a good qualitative correlation between the action of homo- and heterophase additives (compare [7–10]).

Catalysis may also be associated with electronic processes at the gas–solid interface as a result of the appearance of a contact potential difference [11]. For example, in mixtures of silver azide with \(Bi_2O_3\), \(SiO\), \(PbI_2\), the inhibition of the decomposition of \(AgN_3\) is most likely associated with electronic transitions, since, as shown in [12], \(Pb^{++}\), which changes the concentration of ionic defects in \(AgN_3\), does not affect its decomposition and, consequently, this effect is not connected with diffusion of ions and formation of solid solutions. The semiconducting properties of lead iodide with respect to silver azide are used by the authors of [13] to explain the lower sensitivity to a light flash that leads to explosion of the azide.

On the basis of our own data, and of data available in the literature on the catalysis of the thermal decomposition of solid inorganic substances by heterophase additives, we carried out a preliminary classification of catalysis mechanisms according to three features:

1) the presence or absence of mass transfer in catalysis,

2) the aggregate state in which catalysis takes place,

3) the subtle mechanism of catalysis.

This classification may be presented in the form of a scheme (Table II).

From this table it is evident that 8 different cases of catalysis of the thermal decomposition of a solid by a solid are possible. It should be noted that the occurrence of one type of catalysis does not exclude all the others, so that a change in the reaction rate by several mechanisms at once is possible.

According to the scheme presented, depending on the presence of mass transfer through the interface during the thermal decomposition of contacting systems, catalytic reactions are divided into two large groups—

its groups. Regardless of mass (cationic or anionic components of the crystal lattice of the constituent compounds), reactions leading to the elimination of the solid phase of the azide substance by formation of eutectic systems of variable composition—solid solutions, double salts, etc.—occur.

The ratio between the melting temperature of such systems and the temperature at which the azide decomposes is significant, since the liquid or solid phase undergoes the thermal decomposition.

In the solid state, the nature of changes in the rate of the process is evidently determined by causes and phenomena analogous to those considered in studying the question of the effect of homogeneous additives on the thermal decomposition of salts.

The principal factor here is the change in the concentrations and mobility of electronic and ionic defects of the lattice.

If diffusion through the phase boundary leads to the formation of sufficiently low-melting eutectics, then the transfer of the mixture into the melt should in itself accelerate its thermal decomposition. In addition, since according to some models one may expect the presence in molten salts of short-range order, it is reasonable to assume that introduced effective acceptors and donors of charges may influence the rate of the process for electronic reasons as well.

According to the proposed scheme, the second group consists of catalytic reactions proceeding without diffusion of ions across the phase boundary. Included here, however, are cases involving transfer of electrons through the contact. Melting in such systems, occurring by a mechanism analogous, for example, to that proposed in [3], may lead to a change in the rate of thermal decomposition for the reasons already set forth.

Deformation interactions of contacting crystals [3] may in general lead to acceleration of reactions that also proceed without melting. And, finally, depletion or enrichment of near-contact regions with electron–hole current carriers should lead to directed changes in thermal stability for the reasons set forth in [11].

Table I

Flash temperatures of \(AgN_3\), \(TlN_3\), and mixtures based on them

Classification of thermal-decomposition reactions called by E. Turofazin additives

Table 2.

Fig. 1. Decomposition of \(AgN_3\) at \(175^\circ C\) in vacuum: pure (1) and with additions of \(CuO\) (2), \(MnS\) (3), \(Cu_2O\) (4). Azide–additive ratio \(1:5\).

Fig. 2. DTA curves of \(AgN_3\): pure (1) and with additions of \(40\%\) by weight \(CoO\) (2), \(Cu_2O\) (3), \(Cu_2Se\) (5), \(MnS\) (4)—\(20\%\) by weight. The sign \(!\), here and in Fig. 3, indicates the explosion temperature.

Fig. 3. DTA curves of \(TlN_3\): pure (1) and with additions of \(35\%\) by weight \(Ni_3O_4\) (2), \(CuO\) (3).

Fig. 4. Effect of an interruption in the experiment on the subsequent decomposition of \(AgN_3\) after removal of the additive at \(\alpha = 4\%\). 1—pure \(AgN_3\); 2, 3, 4, 5—\(AgN_3 + NiO\), \(Cu_2O\), \(Ni_3O_4\), \(CoO\), respectively. \(t = 240^\circ C\), vacuum.

REFERENCES

1. D. D. Saratovkin, P. A. Savintsev, DAN SSSR, Ser. Physics, 58, 1943, 1947.

2. A. A. Shobuzov, P. A. Savintsev, Izv. VUZov, Ser. Physics, no. 10, 99, 1969, TSU, Tomsk.

3. V. I. Zalkin, ZhFKh, 43, 299, 1969.

4. ZhFKh (in press).

5. G. Charlot, Analytical Chemistry, Moscow–Leningrad, “Khimiya,” 1965.

6. B. L. Evans, A. D. Yoffe, Proc. Roy. Soc., A-250, 1262, 1959.

7. D. A. Zakharov, E. S. Kurochkin, ZhNKh, 13, 1762, 1968.

8. G. G. Savel’ev, D. A. Zakharov, G. G. Shch[[unclear: ending of surname]], A. A. Vasil’ev, Izv. TPI, 151, 40, 1966.

9. V. R. Pai Verneker, A. G. Forsyth, J. Inorg. Nucl. Chem., Letters, 3, 257, 1967.

10. E. S. Kurochkin, D. A. Zakharov, G. G. Savel’ev, present collection.

11. D. A. Zakharov, G. G. Savel’ev, Kinetics and Catalysis, 2, 55, 1966.

12. D. A. Zakharov, G. G. Savel’ev, V. L. Boldyrev, D. N. Votinova, Kinetics and Catalysis, 5, 807, 1964.

13. S. K. Deb, B. L. Evans, A. D. Yoffe, VII Symp. on Comb., Baltimore, 829, 1962.