On the Effect of the Distribution of the \(Cu^{++}\) Impurity on the Thermal Stability of \(AgN_3\) and \(Pb(N_3)_2\)

Yu. A. Zakharov, G. G. Shechkov, G. G. Savel’ev, A. P. Bocharov

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

In work [1] we established the inhibition of the thermal decomposition of \(Pb(N_3)_2\) by the addition of \(Cu^{++}\) ions introduced by coprecipitation. This additive increases the rate of thermal decomposition of \(AgN_3\), while changing the decomposition kinetics of \(AgN_3\). For \(Pb(N_3)_2 + Cu^{++}\), an increase in the first maximum of the rate was observed. To explain these effects, we assumed that the \(Cu^{++}\) additive, along with entering the lattice, is partially adsorbed on the surface of \(AgN_3\) and \(Pb(N_3)_2\), and the corresponding changes in the thermal stability of the \(Pb(N_3)_2 - Cu^{++}\) and \(AgN_3 - Cu^{++}\) systems are the result of the action of the decomposition products of basic copper azide [2]. In the present work, in order to study the state of the \(Cu^{++}\) additive in mixed crystals and the cause of the change in the thermal stability of \(Pb(N_3)_2\) and \(AgN_3\), single crystals of pure \(AgN_3\) and \(Pb(N_3)_2\), as well as those containing the \(Cu^{++}\) additive, were investigated. It was assumed that, owing to the thermodynamic equilibrium between the solution and the solid phase during the preparation of single crystals under slow crystallization conditions, and also as a consequence of repeated recrystallization of the crystals [3], the \(Cu^{++}\) additive should be uniformly distributed throughout the volume of the crystals.

EXPERIMENTAL PROCEDURE

Preparations. Polycrystalline samples of pure and copper-doped \(Cu^{++}\) azides of silver and lead with particle size ...

within 0.01 to 0.03 mm were obtained by coprecipitation from aqueous solutions according to the procedure described in [4].

Single crystals of pure \(Pb(N_3)_2\) and containing 0.1 mol% \(Cu^{++}\) were grown by Attiwig’s method [5].

Crystals of pure lead azide usually had the shape of a regular hexagon and reached 4–8 mm in length. Individual crystals reached a length of 15 mm and had a weight of up to 15 mg. The addition of copper (II) reduces the dimensions of \(Pb(N_3)_2\) crystals by a factor of one and a half to two. The coloration of \(Pb(N_3)_2 + 0.1\%\,Cu^{++}\) single crystals is uniform, light pink.

Single crystals of pure \(AgN_3\) and containing additions of divalent copper ions were obtained by recrystallization of powdered preparations of \(AgN_3 + Cu^{++}\) in a 17% solution of \(NH_4OH\) at a temperature of 16–21°C. White transparent crystals of pure silver azide had the form of fused needles, in some cases reaching 10 mm in length.

Single crystals of \(AgN_3 + Cu^{++}\) were nonuniformly colored; their dimensions did not exceed 15 mm in length.

Thermal decomposition was carried out on a differential apparatus [6] at a pressure of \(10^{-2} \div 10^{-3}\) torr, and also on a weighing apparatus with a helix sensitivity of \(10^{-5}\) g per scale division of the diagram tape, at a pressure of \(5 \cdot 10^{-2}\) torr.

For the experiments, 10 mg samples of preparations were used, placed in cups made of thin (0.005 mm) aluminum foil. Experiments on the differential apparatus, except for specially stipulated cases, were carried out with a Dewar flask containing liquid nitrogen in operation.

EPR spectra of silver and lead azides containing paramagnetic copper ions were recorded on an EPR-2 spectrometer [7].

RESULTS AND DISCUSSION

The addition of \(Cu^{++}\) increases the rate of thermal decomposition of single crystals of both lead azide and silver azide—

At the same time, the decomposition kinetics of single crystals of $Pb(N_3)_2 + Cu^{++}$ changes; their decomposition proceeds, in contrast to polycrystalline samples, without an induction period (Fig. 1, curves 2 and 3). In the single crystals, moreover, there is no distinct maximum in the rate after 5 minutes of decomposition, which is apparently due to a substantial decrease in the fraction of the $Cu^{++}$ additive adsorbed on the surface of $Pb(N_3)_2$ and to the predominantly bulk distribution of the impurity in the $Pb(N_3)_2$ lattice. The rate of thermal decomposition of $Pb(N_3)_2$ single crystals at $245^\circ C$ is considerably lower than the rate of thermolysis of powders (Fig. 1, curves 1, 5), which is evidently caused by differences in dispersity and in the degree of perfection of the crystals.

The addition of $Cu^{++}$ ions catalyzes the thermal decomposition of both poly- and single crystals of $AgN_3$. Unlike single crystals of pure $AgN_3$, whose decomposition proceeds with a small but noticeable induction period, most likely caused by heating of the sample to the experimental temperature, the decomposition of $AgN_3 + Cu^{++}$ polycrystals and pure $AgN_3$, as well as of $AgN_3 + Cu^{++}$ single crystals, proceeds without an induction period and at the maximum rate. In the case of pure $AgN_3$ these phenomena are caused by the specific features of this decomposition, while for $AgN_3 + Cu^{++}$ samples they are apparently due to catalytic decomposition of the additive adsorbed on the $AgN_3$ surface. In polycrystalline $AgN_3 + Cu^{++}$ samples, after 40 minutes of isothermal decomposition at $245^\circ C$, the decomposition rate decreases sharply, and the fraction of decomposed substance is $>80\%$. These $AgN_3 + Cu^{++}$ preparations, which before decomposition have the appearance of pink grains, after thermolysis constitute a single sintered spongy gray-black mass, which most likely indicates melting of the samples during decomposition.

Fig. 1. Decomposition rate of polycrystalline
\(Pb(N_3)_2\), specimen — 1; \(Pb(N_3)_2 + 0.07\ \mathrm{mol}\%\,Cu^{++}\)
at \(240^\circ C\), and of single crystals \(Pb(N_3)_2 + 0.1\ \mathrm{mol}\%\,Cu^{++}\) — 3;
\(Pb(N_3)_2\), specimen — 4 at \(255^\circ C\); \(Pb(N_3)_2\), specimen — 5 at
\(240^\circ C\).

Fig. 2. Rate of thermal decomposition of polycrystals
\(AgN_3\), specimen — 4; \(AgN_3 + 0.1\ \mathrm{mol}\%\,Cu^{++}\) — 1
at \(240^\circ C\), and of single crystals \(AgN_3\), specimen — 2;
\(AgN_3 + 0.1\ \mathrm{mol}\%\,Cu^{++}\) — (3) at \(245^\circ C\).

The appearance of some maximum in the rate in poly- and single-crystalline samples of \(AgN_3 + Cu^{++}\) and the increase of it in polycrystalline samples of \(Pb(N_3)_2 + Cu^{++}\), in comparison with pure \(AgN_3\) and \(Pb(N_3)_2\), is, in our opinion, the result either of adsorption or of near-surface distribution of part of the \(Cu^{++}\) impurity in the crystals and, as a consequence of this, its burning out during the course of decomposition of the sample to a given temperature.

The absence of a sharp acceleration (i.e., of a rate maximum) of decomposition in single crystals of \(Pb(N_3)_2 + Cu^{++}\) indicates that it is incorporated into lattice sites and is distributed uniformly up to

throughout the volume of the crystals, as was shown in the works of Rabykh [8]. This is also evidenced by the uniform coloration of the single crystals of \(Pb(N_3)_2\) containing an additive of copper ions. The nonuniform, predominantly surface distribution of \(Cu^{++}\) in polycrystalline samples of \(AgN_3\) and \(Pb(N_3)_2\) was shown by us using a number of methods.

1. The absence, in the EPR spectra, of HFS (hyperfine structure) indicates, according to Kazanskii et al. [9], the occurrence of a dipole–dipole interaction of paramagnetic ions, which is possible with adsorption or surface distribution of the impurity. In the EPR spectra of polycrystalline samples \(AgN_3 - Cu^{++}\) and \(Pb(N_3)_2 - Cu^{++}\) [1], the hyperfine structure was not resolved.

2. The surface distribution of the \(Cu^{++}\) impurity is also evidenced by the invariance of the phase-transition temperature \((180^\circ C)\) in \(AgN_3\) and \(AgN_3-Cu^{++}\) according to DTA data.

Berg reports a change in the phase-transition temperature of a number of compounds when impurities are introduced into the lattice [10].

1. In single crystals of \(AgN_3 + 0.04\) mol.% \(Cu^{++}\), the distribution of the additive is also nonuniform, predominantly surface. Table I gives the distribution of copper ions in \(AgN_3 + Cu^{++}\) single crystals, which were successively dissolved in four steps. The total weight of the single crystals was 50 mg.

Table I

Distribution of the \(Cu^{++}\) additive over the volume of \(AgN_3\) single crystals.

Burnout of the additive during thermolysis of the systems \(Pb(N_3)_2 - Cu^{++}\) and \(AgN_3 - Cu^{++}\) is demonstrated by the absence of an EPR signal after 10-minute heating of the samples at temperatures of 150 and \(240^\circ C\), and by discoloration of the preparations under such heating. The degrees of transformation of the samples in this process are given in Table 2.

In addition, in the course of uniform heating of lead azide containing copper ions, in a differential setup, a maximum in the rate of gas evolution is observed in the temperature interval \(120\text{–}150^\circ C\), increasing with increasing additive concentration (Fig. 3). Such a rate maximum was not observed upon heating single crystals of \(Pb(N_3)_2 + 0.1\) mol% \(Cu^{++}\) (Fig. 3, curve 4); in this case the crystals do not discolor over a long time.

Fig. 3. Changes in the rate of thermal decomposition of \(Pb(N_3)_2\) during heating of samples \((4\ \text{deg/min})\):

1, 2, 3 — \(Pb(N_3)_2 + 0.7;\ 0.3;\ 0.1\) mol% \(Cu\) (polycrystals), respectively;

4 — \(Pb(N_3)_2 + 0.1\) mol% \(Cu\) (single crystals);

5, 6 — pure \(Pb(N_3)_2\) (poly- and single crystals, respectively).

We compared the concentrations of the additive \(Cu^{++}\) (up to \(N_3^{-}\)), introduced during coprecipitation with \(Pb(N_3)_2\) (polycrystals), and the weight loss of these samples decomposed in vacuum for 10 min, which corresponds to the first maximum, at 150 and \(240^\circ C\) at a pressure of \(10^{-4}\) torr (Table 2).

Table 2

Comparative assessment of the weight loss over 10 min of samples at 150 and \(240^\circ C\) and of the weight of \(N_3^{-}\) corresponding to the concentration of \(Cu^{++}\) in \(Pb(N_3)_2\)

It turned out that the weight loss of the preparation \(Pb(N_3)_2 + 0.7\) mol% \(Cu^{++}\) at \(150^\circ C\) corresponds to the content of the \(Cu^{++}\) impurity. The weight loss in samples decomposed at \(240^\circ C\) significantly exceeds the weight loss corresponding to the concentration of the \(Cu^{++}\) additive in \(Pb(N_3)_2\).

In the first case, obviously, burnout of the additive occurs, since decomposition of \(Pb(N_3)_2\) begins at a temperature \(> 200^\circ C\), whereas in the second case decomposition of \(Pb(N_3)_2\), catalyzed by the \(Cu^{++}\) additive, also occurs.

Since the fact of burnout of the \(Cu^{++}\) additive during thermal decomposition of \(AgN_3 + Cu^{++}\) and \(Pb(N_3)_2 + Cu^{++}\), with the exception of single-crys-

Since no similar increase in the activity of pure crystals of $Pb(N_3)_2 + Cu^{++}$ was established, in order to clarify the reasons for the observed changes in the decomposition rate after the first maximum of the rate, we investigated the thermal stability of $Pb(N_3)_2$ with possible products of the thermal decomposition of copper azide under our conditions—$Cu$, $CuO$, $Cu_2O$. In [11] the presence of $H_2O$ in $Pb(N_3)_2$ was shown even after 18 minutes of isothermal decomposition at $240^\circ C$ in a vacuum of $\sim 10^{-7}$ torr; therefore, the formation of copper oxides during the decomposition of $Cu(N_3)_2$, separated into a distinct phase (adsorbed on the surface), is possible. All the oxides used, of chemically pure grade, were sieved through sieve No. 23 and thoroughly mixed with lead azide in a ratio of 1:10 by weight. An excess of oxides, recalculated for the $Cu^{++}$ ion in comparison with the amount of $Cu^{++}$ introduced by coprecipitation, was taken in order to compensate for the smaller contact surface in the oxide–azide mixture in comparison with the homophase impurity. During thermal decomposition of the mixtures on a weighing apparatus in a vacuum of $8 \cdot 10^{-2}$ torr and at a temperature of $240^\circ C$, the effectiveness of the additives is identical, and all of them are inhibitors of the thermal decomposition of $Pb(N_3)_2$ (Fig. 4).

During decomposition of these same mixtures at a lower pressure of $10^{-5}$ torr and removal of the decomposition products at $196^\circ C$ on a differential apparatus, the $CuO$ additive is a catalyst, whereas $Cu$ and $Cu_2O$ remain inhibitors of the thermal decomposition of $Pb(N_3)_2$ (Fig. 5). A similar change in the nature of the effect of certain oxide additives on the thermal stability of $AgN_3$ was also observed in [12]. The unusual behavior of the $Pb(N_3)_2 - CuO$ mixture is explicable when taking into account the concepts we have developed concerning the interaction of adsorbed gases and decomposition products with undecomposed lead azide [11]. Confirmation of this is the increase in the duration of the induction period during decomposition of the $Pb(N_3)_2$ mixture in the absence of a trap with liquid nitrogen (Fig. 5).

Fig. 4. Decomposition of pure \(Pb(N_3)_2\) at \(245^\circ\)—(1), and of \(Pb(N_3)_2\) containing 10 wt.% \(Cu\), \(CuO\), and \(Cu_2O\), respectively—2–4.

Fig. 5. Rate of thermal decomposition of \(Pb(N_3)_2\) at \(245^\circ C\).

1, 6, 5 — \(Pb(N_3)_2 + 10\) wt.% \(CuO\), \(Cu\), \(Cu_2O\), respectively.
2 — pure \(Pb(N_3)_2\).
4, 3 — \(Pb(N_3)_2 + 10\) wt.% \(CuO\) and pure \(Pb(N_3)_2\), respectively; experiments without a trap with liquid \(N_2\).

The thermal stability of heterophase mixtures \(AgN_3 - CuO\), \(AgN_3 - Cu_2O\) was studied in [13], where it was established that the oxides \(CuO\) and \(Cu_2O\) are catalysts of the decomposition of \(AgN_3\). Thus, there is a parallelism in the influence of the homophase impurity \(Cu^{++}\) and heterophase additives \(Cu\), \(CuO\), \(Cu_2O\), which may be products of the decomposition of \(Cu(N_3)_2\), on the thermal stability of powdered \(AgN_3\) and \(Pb(N_3)_2\). These resul-

The results confirm one of the conclusions of work [14], namely that the thermal stability of mixed crystals, where the thermally unstable microcomponent is adsorbed or concentrated in the near-surface layers of the macrocomponent, is determined by the processes occurring after the additive has burned out at the interface between the substrate and the products of thermal decomposition of the microcomponent.

The active influence of the homophase impurity \(Cu^{++}\) on the thermal stability of polycrystalline \(Pb(N_3)_2\) and of its single crystals indicates the need, when studying the thermal stability of mixed crystals, to take into account the distribution of the impurity, which depends on the synthesis conditions [3].

REFERENCES

1. G. T. Shchekov, G. G. Savel'ev, Yu. A. Zakharov, Izv. TPI, 157, 50, 1970.

2. K. Singh, Trans. Farad. Soc., 55, 1, 124, 1959.

3. V. G. Khlopin, Selected Works, Vol. I, Publishing House of the Academy of Sciences of the USSR, 1957.

4. Yu. A. Zakharov, G. G. Savel'ev, V. V. Boldyrev, L. A. Votinova, Kinetics and Catalysis, 5, 807, 1964.

5. A. Ya. Apin, in: Collected Articles on the Theory of Explosives, Oborongiz, Moscow, 1940, p. 106.

6. G. G. Savel'ev, V. V. Bordachev, Izv. TPI, 157, 75, 1970.

7. A. A. Voevodskii, V. V. Blinshtein, L. G. Semenov, Application of Electron Paramagnetic Resonance in Chemistry, Publishing House of the Siberian Branch of the Academy of Sciences of the USSR, Novosibirsk, 1962, p. 35.

8. V. P. Shchukina, S. M. Ryabykh, Izv. TPI, 199, 66, 1969.

9. V. B. Kazanskii, S. I. Bakhova, A. G. Lofarskii, V. V. Voevodskii, I. M. Ioffe, Kinetics and Catalysis, 2, 862, 1961.

10. J. R. Berg, Introduction to Thermography, Nauka Publishing House, Moscow, 1969.

11. G. T. Shchekov, Yu. A. Zakharov, V. A. Kalgin, G. G. Savel'ev, E. N. Svobodin, Kinetics and Catalysis, 1970 (in press).

12. Yu. A. Zakharov, E. S. Kurochkin, G. G. Savel'ev, Yu. N. Rubov, Kinetics and Catalysis, 7, 3, 425, 1966.

13. E. S. Kurochkin, Candidate’s dissertation, Tomsk, 1969.

14. G. T. Shchekov, present collection.