On the Mechanism of the Influence of Catalysts on the Combustion Processes of Condensed Mixtures

V. I. Bratishchik, G. V. Ivanov, N. I. Odintsov

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

As is known, small amounts of transition-metal oxides substantially change the burning rate of condensed mixtures, which is explained by heterogeneous catalysis of the chemical reactions occurring during combustion. In particular cases, this method of action can be regarded as a means of obtaining information about the regularities of the combustion process.

It has previously been suggested \([1,2]\) that catalytic additives should lower the surface temperature of \(NH_4ClO_4\). There are almost no data in the literature on experimental verification of this assumption. One can cite only the work \([3]\), in which, however, no effects of the influence of catalysts exceeding the experimental error were found. The latter may be connected with the well-known methodological difficulties of measuring \(t_s^\circ\) at elevated pressures with embedded thermocouples.

A number of methodological difficulties can be avoided by conducting experiments at reduced pressures, since under these conditions, in the immediate vicinity of the surface, there is no zone of high-temperature gas reactions, and it is possible to measure \(t_s^\circ\) with a thermocouple with beads.

In the present work an attempt is made to use this method to determine the effect of the influence of several additives on the value of \(t_s^\circ\) of a burning mixed composition.

EXPERIMENTAL PART

We used a model mixture consisting of ammonium perchlorate (70%) and epoxy resin (30%); the hardener was maleic anhydride. Additives: $CuO$, $MnO_2$, $Fe_2O_3$ were introduced into the composition in excess of 100% in an amount of 2 weight percent. The specimens ($d = 8$ mm, length 50 mm, relative density 0.97–0.98) were produced by free casting under vacuum.

The experiments were carried out under a glass bell jar of 15-liter capacity, at pressures of 50, 150, and 400 mm Hg. A buffer vessel of 40 liters ensured maintenance of constant pressure during combustion of the specimen. At pressures above 400 mm Hg, experiments were not performed because of rapid burnout of the thermocouples.

In order to prevent breakage under the action of the load during the experiments, round thermocouples 100 $\mu$ in diameter (manganin–constantan) were used, as well as the same thermocouples rolled to a thickness of 50 $\mu$. Such a thickness does not allow one to judge rigorously the absolute value of $t_s^\circ$; however, the difference in the temperatures of the near-surface zone during combustion of specimens of a composition containing various additives (see Fig. 1) is recorded quite clearly.

In the first variant the thermocouple was stretched taut between the jaws of a holder and placed on the specimen. As it burned, the thermocouple moved behind the combustion surface under the action of the weight of the holder sliding along guides.

In the second variant the method of $\Pi$-shaped thermocouples with weights, developed at the Institute of Chemical Physics of the Academy of Sciences of the USSR, was used. In this case the branches of the thermocouple were passed through two through channels parallel to the axis of the specimen. The channels were made by inserting thermocouple wires into the unpolymerized specimen; these wires were removed one day before completion of polymerization.

A 5-mm segment of the same specimen, fitted with a thermocouple, was placed on the specimen. The latter was burned through by an electric coil. For this reason, the thermocouple did not record the temperature profile, but only registered the temperature at the burning surface, accompanying the surface for long intervals of time. The thermocouple signal was recorded with an N-105 loop oscillograph. The results of parallel experiments were reproduced well. The arithmetic mean error does not exceed \(\pm 30^\circ\mathrm{C}\) (indicated in Table I). It is evident from Tables I and II that the additives strongly change the temperature near the burning surface; the effect far exceeds the experimental error. Naturally, the use of burned-through thermocouples gives underestimated temperature values in comparison with bead thermocouples. However, in both cases the magnitude of the effect of the change in \(t_s^\circ\) with additives remains approximately the same (cf. Table I and Table 2). Namely, additions of \(\mathrm{CuO}\) and \(\mathrm{MnO_2}\) lower \(t_s^\circ\), while additions of \(\mathrm{Fe_2O_3}\) appreciably increase it.

Table I

Temperature of the near-surface zone of the mixed composition with additives

Variant No. 1, bead thermocouple \(d = 100\,\mu\)

Table 2

U-shaped thermocouple /rolled/

Table 3

Burning rate of mixed compositions

Fig. 1. Typical oscillograms of temperature profiles for the composition without additives and with additives $CuO$ and $Fe_2O_3$.

DISCUSSION OF RESULTS

At the present time it is considered generally accepted (see, for example, [4,5]) that ammonium perchlorate (AP) first dissociates into ammonia and perchloric acid according to the reaction:

\[ NH_4ClO_4 \rightleftarrows NH_3 + HClO_4 . \tag{1} \]

Thus, in studying the combustion of mixtures of AP with combustible substances, the reactions in the gas phase between perchloric acid, ammonia, and the pyrolysis products of the fuels should be regarded as the principal ones.

They are preceded by reaction (1) and by the process of gasification of the fuel. The combustion rate may be increased as a result of the catalysis of one or several of the reactions mentioned.

The rate of reaction (1) can be changed if the equilibrium between AP and the products of its dissociation is shifted; that is, removal of \(NH_3\) and \(HClO_4\) should increase the rate of decomposition, while their additional introduction should inhibit it. It is known [6] that the introduction into the reaction zone, by one method or another, of small amounts of ammonia strongly inhibits the rate of AP decomposition. Removal of the dissociation products may, in particular, be achieved by catalysis of the reaction of ammonia oxidation and of perchloric acid decomposition.

From [8,9,10] it is known that oxidation of ammonia by oxygen proceeds at temperatures above \(600^\circ C\); however, in the presence of transition-metal oxides this reaction occurs already at \(300^\circ C\) [ibid.], with the oxides of cobalt, manganese, and copper catalyzing it most actively. Iron oxide in this case is a relatively weak catalyst [11].

In [12,13] it was shown that additions of \(CuO\) and \(MnO_2\) lower the flash temperature of ammonium perchlorate from \(430^\circ C\) to \(280\text{–}250^\circ C\), whereas \(Fe_2O_3\) practically does not change it. Apparently, the lowering of the flash temperature is a consequence of the catalysis, by oxides of copper and manganese, of the reaction of ammonia oxidation by perchloric acid, more precisely by the products of its decomposition.

In [14] this catalysis is considered to be the reason for the decrease in the ignition temperature of APC under the pulsed action of a heat flux.

In our case, during combustion in vacuum, the additives \(\mathrm{CuO}\) and \(\mathrm{MnO_2}\) also, apparently, catalyze the oxidation of ammonia; this process begins to proceed at lower temperatures, as a result of which the decomposition temperature of APC decreases, which is recorded by the thermocouple located near the burning surface.

The increase in the temperature of the near-surface zone with the additives \(\mathrm{Fe_2O_3}\) and \(\mathrm{Cr_2O_3}\) can be explained as follows.

Since \(\mathrm{Fe_2O_3}\) weakly catalyzes the oxidation of ammonia and does not change the flash temperature of APC, during combustion in vacuum no decrease in \(t^\circ_s\) is observed. It is known, however [15,16], that iron oxide very effectively catalyzes the oxidation of various hydrocarbon gases. In our case, an increase in the rate of oxidation of the pyrolysis products of the fuel in its presence raises the heat release per unit volume, and the temperature in the near-surface zone increases.

It is evident from Table 3 that the additives \(\mathrm{CuO}\), \(\mathrm{MnO_2}\), and \(\mathrm{Fe_2O_3}\) increase the burning rate to approximately the same extent, although they affect the temperature near the surface differently. Thus, the burning rate does not depend directly on this temperature, but is determined by the nature of the chemical reactions proceeding in the near-surface zone, and reactions involving catalysts may play a decisive role. The chemistry of the process can be substantially altered by replacing the catalysts, since the latter are characterized by a certain selectivity of action.

Data on the selectivity of the action of catalysts are confirmed by the following observations:

At pressures up to 400 mm Hg, combustion of compositions without additives and with the additive \(\mathrm{Fe_2O_3}\) is accompanied by the release of large amounts of white smoke, which settles on the walls of the chamber. Chemical analysis of the deposit showed that it contains 92–95% \(\mathrm{NH_4Cl}\), the re-

metallic soot and resinous organic substances.

The $ClO_4^-$ ion is not detected.

Thus, in the indicated cases, combustion in vacuum is reduced to oxidation of the pyrolysis products of the binder by chloric acid. Ammonia, however, is not oxidized and is released in the form of $NH_4Cl$.

When a composition with additions of $CuO$ and $MnO_2$ is burned, the tube walls remain clean, and $NH_4Cl$ is not released; i.e., these additives, unlike iron oxide, catalyze the oxidation of ammonia and increase the completeness of combustion in vacuum.

In conclusion it should be noted that in the region of elevated pressures the ammonia oxidation reaction will apparently precede the oxidation reaction of the binder pyrolysis products; nevertheless, the mechanism of the influence of the catalysts may be preserved. However, the nature of their influence on the magnitude of $\chi'_s$ in this case is unclear and requires special investigation.

Conclusions

1. Using a thermocouple with weights, the temperature of the near-surface zone during combustion in vacuum of model mixtures based on ammonium perchlorate was measured.

2. It has been established that additions of $CuO$ and $MnO_2$ lower the temperature near the surface, whereas $Fe_2O_3$ and $Cr_2O_3$ raise it.

3. On the basis of the results obtained it is concluded that $CuO$ and $MnO_2$ catalyze predominantly the ammonia oxidation reaction, whereas $Fe_2O_3$ and $Cr_2O_3$ catalyze the oxidation reaction of the pyrolysis products of the fuel.

Literature

1. Ed. by Summerfield. Investigation of Solid-Propellant Rocket Engines. Mir Publishing House, Moscow, 1963.

2. N. N. Bakhman. Fizika goreniya i vzryva, 1965, No. 3, p. 98.

3. Sobolelli, Veivograd, Summerfield. Raketnaya tekhnika i kosmonavtika, 1965, No. 9, p. 16.

1. Pearson, Sutton, Raketnaya tekhnika i kosmonavtika [Rocket Engineering and Cosmonautics], 1966, No. 5, p. 241.

2. Bakhman N. N. Combustion of Heterogeneous Condensed Systems. Moscow: Nauka, 1967.

3. Biccumshaw L. L., Newman B. H. Proc. Roy. Soc. A-227, p. 228.

4. Inami S. H., Rosser W. A., Wise H. J. Phys. Chem., 1963, 67, No. 5, 1077.

5. Stetens E., Rease R. J. Amer. Chem. Soc., 72, 1950, 1138.

6. M. I. Temkin, I. I. Morozov, V. M. Pykhov et al. Problems of Physical Chemistry. Issue 2, p. 14. Moscow: Goskhimizdat, 1959.

7. Molinari Ettore. Patent, 1966, 145735.

8. I. I. Morozov, L. I. Lukyanova, M. I. Temkin. Kinetics and Catalysis, 1966, vol. 7, issue 1, p. 172.

9. Katatani Kenji. Report Aeronaut. Res. Inst., Tokyo, 1962, No. 4, p. 79.

10. A. Hermony (Makovky), A. Salmon. Bull. Res. Council Israel, 1960, A-9, No. 3–4, 206–207.

11. Evans M. J. Chem. Phys., 1964, 40, No. 9, 2431–2438.

12. G. F. Bolshakov, B. I. Gubin, I. N. Torchaeva. Physicochemical Principles of the Application of Motor, Jet, and Rocket Fuels. Moscow–Leningrad: Khimiya, 1965, p. 29.

13. G. D. Andrushkevich, V. V. Popovskii, G. K. Boreskov. Kinetics and Catalysis, 1963, vol. 6, issue 6.