EFFECT OF IMPURITIES ON THE THERMAL DECOMPOSITION OF AMMONIUM PERCHLORATE

R. N. Isaev, Yu. A. Zakharov, V. V. Bordachev

(Presented by the scientific seminar of the Department of Radiation Chemistry and the ChTT Laboratory)

The thermal decomposition of $\mathrm{NH_4ClO_4}$ has been studied by many authors [1—3]; a considerable number of publications are also devoted to the question of the effect of impurities on the regularities of thermolysis of this salt [5—8], reviews—[19, 20]. However, up to the present time there has been no single, agreed-upon opinion either on the stoichiometry of the decomposition of $\mathrm{NH_4ClO_4}$ or on the causes and mechanism of the effect on the rate of thermolysis of certain additives studied as thermolysis catalysts. The formation of an opinion on the latter question is hindered to a considerable extent by the absence of data on the influence of impurities on the qualitative and quantitative composition of the gaseous products of thermolysis of $\mathrm{NH_4ClO_4}$.

In the present work the first results obtained by us in this direction are reported.

Experimental procedure

In the experiments, crystals of $\mathrm{NH_4ClO_4}$ of “pure” grade, with a fraction $<60\mu$, were used; the impurities were the oxides $\mathrm{Co_2O_3}$, $\mathrm{ZnO}$, $\mathrm{CdO}$, and $\mathrm{CaO}$ of chemically pure grade. The impurities were added to AP in an amount of 10 wt.% by mechanical mixing in a test tube with a ground-glass stopper. Thermal decomposition was carried out in a specially assembled apparatus [9], which made it possible to study the kinetics of isothermal decomposition or to perform DTA of samples in vacuum.

At the same time, the reaction vessel of the apparatus was part of the gas inlet system to an MX-1302 mass spectrometer. Thermal decomposition was studied in the temperature interval 211—297°C at an initial pressure of $5 \times 10^{-3}$ mm Hg. DTA was performed at a rate of 8—10° per min at the same initial vacuum. Analysis of the gaseous decomposition products was carried out on the MX-1302 instrument at an ionizing voltage of 70 V, an accelerating voltage of 1.5 kV, and an emission current of 2 mA.

Experimental results and discussion

The effect of impurities on the thermal decomposition of ammonium perchlorate is shown in Fig. 1, from which it is seen that the impurities $\mathrm{Co_2O_3}$, $\mathrm{CdO}$, and $\mathrm{ZnO}$ accelerate thermolysis. Similar results were obtained when DTA was carried out (Fig. 2). Using the dependence of the transformation coefficient

Fig. 1. Effect of impurities on the thermal decomposition of \(NH_4ClO_4\) at \(211^\circ C\). 1—pure \(NH_4ClO_4\); 2—\(NH_4ClO_4 + 10\) wt.% \(CdO\); 3—\(NH_4ClO_4 + 10\) wt.% \(Co_2O_3\); 4—\(NH_4ClO_4 + 10\) wt.% \(ZnO\).

Fig. 2. Differential-thermal analysis of \(NH_4ClO_4\) with impurities: 1—pure, 2—with 10% \(Co_2O_3\), 3—with 10% \(ZnO\), 4—with 10% \(CdO\), 5—with 10% \(CaO\).

from the kinetic curves on temperature \(K=K_0e^{-E/RT}\) [10], we found values of the apparent activation energy for the thermolysis process of \(\mathrm{NH_4ClO_4}\). It proved to be equal, in kcal/mole: for the pure substance, 19.6; with addition of \(\mathrm{ZnO}\), 20.1; with addition of \(\mathrm{CdO}\), 25.8; with addition of \(\mathrm{Co_2O_3}\), 21.8. The thermogram of pure \(\mathrm{NH_4ClO_4}\) has the same form as that obtained in air. The DTA curves of \(\mathrm{NH_4ClO_4}\) containing impurities differ from the thermogram of the pure perchlorate. Thus, on the DTA curves of mixtures \(\mathrm{NH_4ClO_4}+10\%\ \mathrm{ZnO}\), immediately after the endothermic peak corresponding to the modification transition, there follows a flash. For the mixture of \(\mathrm{NH_4ClO_4}\) with \(\mathrm{CdO}\), the endothermic peak has an indistinct bifurcation, while for the mixture with \(\mathrm{CaO}\) even 2 additional endothermic peaks are observed. The presence of such peaks may be due to chemical interaction of the impurity with \(\mathrm{NH_4ClO_4}\).

Visual observations of the course of the isothermal decomposition of \(\mathrm{NH_4ClO_4}\) showed that, during thermolysis of the mixture \(\mathrm{NH_4ClO_4}+10\%\ \mathrm{Co_2O_3}\), the color of the solid residue after completion of thermolysis is black, as is the color of \(\mathrm{Co_2O_3}\); its weight is equal to the amount of the introduced \(\mathrm{Co_2O_3}\) impurity; the residue is insoluble in water. In the case of the mixture of perchlorate with \(\mathrm{CdO}\), the decomposition residue is decolorized in comparison with the color of the initial mixture of \(\mathrm{NH_4ClO_4}\) with the additive (cocoa-colored) and dissolves completely in water. Similarly, the residues after thermolysis of mixtures of \(\mathrm{NH_4ClO_4}\) with \(\mathrm{ZnO}\) and \(\mathrm{CaO}\) are soluble in water. All the foregoing shows that impurities \(\mathrm{CdO}\), \(\mathrm{ZnO}\), and \(\mathrm{CaO}\), when heated with \(\mathrm{NH_4ClO_4}\), give soluble salts. In the case of \(\mathrm{CaO}\), this salt is definitely calcium perchlorate, since one of the methods for obtaining it is the reaction of \(\mathrm{CaO}\) with \(\mathrm{NH_4ClO_4}\) [11]. The impurities \(\mathrm{ZnO}\) and \(\mathrm{CdO}\) must also give the corresponding perchlorates; it remains unclear, however, why \(\mathrm{CaO}\) slows the rate of thermolysis of \(\mathrm{NH_4ClO_4}\) [8], whereas \(\mathrm{ZnO}\) and \(\mathrm{CdO}\) accelerate it. In addition, in the case of \(\mathrm{ZnO}\) it is unclear why the effect of doping \(\mathrm{ZnO}\) with \(\mathrm{Li^+}\) and \(\mathrm{Al^{+3}}\) ions has a significant effect on the catalytic activity of the additive.

To some extent, answers to these questions may be obtained from the results of mass-spectrometric analysis of the decomposition products of \(\mathrm{NH_4ClO_4}\). These data are summarized in Table 1, where fragment ions are not given and the background and additional peaks from the interaction of oxygen with vapors and gases adsorbed on the walls of the MX-1302 system have been subtracted [12].

Table 1

Principal decomposition products of pure \(\mathrm{NH_4ClO_4}\) and with impurities according to the results of MSA. Decomposition temperature 211° C (for the CaO impurity, 230° C)

To take the latter effect into account, we admitted pure oxygen obtained by thermal decomposition of \(\mathrm{KMnO_4}\), after preliminary evacuation of the salt to \(5\cdot10^{-3}\) mm Hg while heating it to \(70^\circ\mathrm{C}\). Upon admission of oxygen, peaks of \(\mathrm{C}\), \(\mathrm{CO}\), and \(\mathrm{CO_2}\) appear in the mass spectrum, the magnitudes of which were proportional to the amount of \(\mathrm{O_2}\). The magnitudes of the fragment peaks that give overlap with the main peaks,

17 (OH, NH\(_3\)), 28 (N\(_2\) from N\(_2\)O), 30 (NO from N\(_2\)O) were taken from [12–13] and from the results of our calibration of the system when admitting N\(_2\)O [14]. The relative contents of the components given in the table were calculated with respect to O\(_2\), whose peak magnitude was taken as 100%.

The results of the analysis presented in Table 1 show that the products of thermolysis of NH\(_4\)ClO\(_4\) contain HCl, which previously had not been detected by the authors (with the exception of [16]) and did not appear in the equations for the decomposition of NH\(_4\)ClO\(_4\) [15]. Usually it was found in the gaseous products after the explosion of AP in a bomb [15]. It should be noted that chloride ion has also been detected in the solid residue from the decomposition of NH\(_4\)ClO\(_4\) [17]. Our analyses of the decomposition of NH\(_4\)ClO\(_4\) at higher temperatures (\(>350^\circ\)C) showed that HCl is present at all temperatures; the ratio of NO and N\(_2\)O also changes: with increasing temperature the concentration of NO increases, while that of N\(_2\)O decreases. On the basis of the foregoing, the following approximate equations for the decomposition of NH\(_4\)ClO\(_4\) may be given:

\[ 10\mathrm{NH_4ClO_4}\xrightarrow{<300^\circ\mathrm{C}}8\mathrm{HCl}+\mathrm{Cl_2}+10\mathrm{O_2}+2\mathrm{N_2O}+2\mathrm{N_2}+2\mathrm{NO}+16\mathrm{H_2O}; \]

\[ 12\mathrm{NH_4ClO_4}\xrightarrow{>300^\circ\mathrm{C}}10\mathrm{HCl}+\mathrm{Cl_2}+10\mathrm{O_2}+8\mathrm{NO}+\mathrm{N_2O}+\mathrm{N_2}+19\mathrm{H_2O}. \]

Table 1 also gives the results of mass-spectrometric analysis of NH\(_4\)ClO\(_4\) with impurities. Comparison of the analysis data for the products of thermolysis of pure NH\(_4\)ClO\(_4\) and of thermolysis of a mixture with Co\(_2\)O\(_3\) shows a significant increase in the thermolysis products of the mixture in the relative content of all components except O\(_2\), with the NO:N\(_2\)O ratio being considerably larger than for pure NH\(_4\)ClO\(_4\). These data, together with the results of the thermal analysis presented above, make it possible to conclude that the Co\(_2\)O\(_3\) impurity acts as a typical catalyst for the thermal decomposition of NH\(_4\)ClO\(_4\) and, possibly, as a catalyst for the oxidation of NH\(_3\) and N\(_2\)O to NO. In the case of the CaO impurity, the analysis showed the absence of NH\(_3\), whose release had earlier been used to explain the retardation of the thermal decomposition reaction of NH\(_4\)ClO\(_4\). In view of the untenability of this assumption shown by us, it becomes necessary to elucidate the probable causes of the inhibition of decomposition by the CaO impurity.

As such causes, first, one should point to the possible influence of an increase in the thermolysis products of N\(_2\)O, which retards the decomposition process, and, second, to a decrease in the oxygen content, which accelerates the thermolysis of perchlorate (see table).

The cause of the retardation of decomposition may also be connected with the noticeable endothermicity of the reaction of NH\(_4\)ClO\(_4\) with CaO.

Thus, on derivatograms of a mixture of NH\(_4\)ClO\(_4\) + 10% CaO, recorded at a heating rate of \(4^\circ\)C/min, a sharp decrease in temperature is noticeable in the region of the onset of the reaction of CaO with NH\(_4\)ClO\(_4\). The CdO impurity also enters into reaction with NH\(_4\)ClO\(_4\), giving Cd(ClO\(_4\))\(_2\), which is confirmed by mass-spectrometric analysis: in the products of thermolysis of this mixture HCl and Cl\(_2\) are absent, and the O\(_2\) content decreases. In the work of Voldyreva [18] it was shown that, in the solid residue after thermolysis of this mixture, there are absorption bands in the IR region characteristic of Cd(ClO\(_4\))\(_2\). The acceleration of the thermolysis reaction of NH\(_4\)ClO\(_4\) by the CdO impurity is explained by them by the formation of the eutectic Cd(ClO\(_4\))\(_2\) + NH\(_4\)ClO\(_4\). Apparently, this conclusion is correct. In the case of the ZnO impurity, whose influence was also studied by them, a similar conclusion is also valid; however, formation of zinc perchlorate

is probably due not to the solid-phase reaction \(\mathrm{ZnO} + \mathrm{NH_4ClO_4}\), but to the interaction of ZnO with the decomposition products of \(\mathrm{NH_4ClO_4}\), since MSA showed the presence of the same products as in the decomposition of pure AP (except for NO), but in smaller amounts. The possibility cannot be excluded that, in the case of CaO, CdO, and ZnO impurities, along with perchlorates, chlorides of the corresponding metals are formed.

REFERENCES

1. L. Z. Bircumshaw, B. H. Newman, Proc. Roy. Soc., A227, No. 1168, 115–32 (1954).
2. L. Z. Bircumshaw, B. H. Newman, Proc. Roy. Soc., A277, No. 1169, 228–41 (1955).
3. G. B. Manelis, Yu. I. Rubtsov, Zh. fiz. khimii 40, No. 7, 770–4 (1966).
4. A. Galwey, P. Jacobs, Trans. Farad. Soc., 55, 837–44 (1959).
5. F. Solimosi, L. Revezs, Z. anorg. allg. Chem., 322, No. 1–2, 86–100 (1963).
6. F. Sholmoshi, L. Reves. Kinetics and Catalysis. 4, 88 (1963).
7. F. Solimosi, L. Revezs, Nature, 192, No. 4797, 64 (1961).
8. K. K. Andreev. Thermal Decomposition and Combustion of Explosives. Nauka, Moscow, 84 (1966).
9. R. N. Isaev, Yu. A. Zakharov, V. V. Bordachev. In: Chemical Reactions Involving Solids. The present collection, p. 50.
10. I. M. Emanuel, D. G. Knorre. Course of Chemical Kinetics. Moscow (1961).
11. I. Schumacher. Perchlorates. Properties, Production, and Application. Moscow (1961).
12. R. A. Friedel, A. G. Sharkey et al., Anal. Chem., 25, No. 9, 1314–20 (1953).
13. S. M. Rock, Anal. Chem., 23, No. 2, 261 (1951).
14. Yu. V. Karyakin. Pure Chemical Reagents. Goskhimizdat, Moscow–Leningrad (1947).
15. A. E. Simchen, J. Appl. Chem., 13, 369 (1963).
16. G. A. Heath, J. R. Majer, Trans. Farad. Soc., 60, No. 502, 783–91 (1964).
17. Yu. A. Zakharov, R. N. Isaev, V. A. Nevostruev, L. A. Vakhovskaya. High Energy Chemistry (in press).
18. A. V. Boldyreva, V. N. Mozzhova. Kinetics and Catalysis. 7, No. 4, 734–6 (1966).
19. Yu. A. Zakharov, G. G. Savel’ev. Kinetics and Catalysis. 7, 54 (1966).
20. Yu. A. Zakharov, G. G. Savel’ev. Russian Chemical Reviews (in press).

NOTED MISPRINTS