Relationship Between the Catalytic Properties and Dissociation Elasticity of Metal Oxides in the Oxidation of Ammonia Under Pressure

N. P. Kurin, M. S. Zakharov

(Presented by Prof. A. G. Stromberg, Doctor of Chemical Sciences)

Of great interest is a comparison of the catalytic properties of oxides with respect to the reaction of oxidation of $\mathrm{NH_3}$ to $\mathrm{NO}$ with their dissociation elasticity.

In order to make it possible to trace the indicated dependence more clearly, Fig. 1 is given. In it, the degrees of oxidation ($\alpha$) of $\mathrm{NH_3}$ to $\mathrm{NO}$, in percent, at $700^\circ\mathrm{C}$ [6], are plotted along the ordinate axis, while along the abscissa axis is plotted the logarithm of the dissociation elasticity of the oxides at the same temperature [1].

Fig. 1. Dependence of the degree of oxidation of $\mathrm{NH_3}$ to $\mathrm{NO}$ on the logarithm of the dissociation elasticity of oxides:
1—$\mathrm{Al_2O_3}$; 2—$\mathrm{Cr_2O_3}$; 3—$\mathrm{WO_3}$; 4—$\mathrm{MoO_3}$; 5—$\mathrm{ZnO}$; 6—$\mathrm{SnO_2}$; 7—$\mathrm{NiO}$;
8—$\mathrm{PbO}$ $(760^\circ)$; 9—$\mathrm{Fe_2O_3}$; 10—$\mathrm{CuO}$; 11—$\mathrm{MnO_2}$; 12—$\mathrm{Ag_2O}$.

It is evident from the figure that there is a relationship between the dissociation elasticity of oxides and their catalytic activity; moreover, what is characteristic is that the activity of catalysts with respect to the ammonia oxidation reaction, as the logarithm of the dissociation elasticity changes from $-50$ to $-23$, is practically absent (the exception is $\mathrm{Cr_2O_3}$), whereas—

therefore increases sharply. Thus, for example, in the presence of SnO₂, which has \(P_{\mathrm{O}_2}=10^{-20}\) atm at \(700^\circ\mathrm{C}\), oxidation of NH₃ does not occur, whereas in the presence of NiO, which has \(P_{\mathrm{O}_2}=10^{-19.5}\), the degree of oxidation at \(700^\circ\mathrm{C}\) reaches 34%. On other oxides (PbO, Fe₂O₃, CuO, MnO₂), the logarithm of the dissociation elasticity of which is equal to \(-15.5\), \(-8\), \(-2.6\), and \(+1\), the degree of oxidation also reaches considerable values (46.7; 45; 38.9; 58%).

In the presence of Ag₂O, however, for which the logarithm of the dissociation elasticity at \(700^\circ\mathrm{C}\) reaches 2.5 (which corresponds to 317 atm), oxidation of ammonia occurs only at the initial moment.

Thus, with an increase of \(P_{\mathrm{O}_2}\) from 10 (\(\lg P_{\mathrm{O}_2}=1\), MnO₂) to 317 atm, there is a sharp decrease in the catalytic activity of the oxides.

It is evident from Fig. 1 that oxidation of ammonia occurs only in the presence of those oxides whose dissociation elasticity lies within definite limits. Thus, at \(700^\circ\mathrm{C}\) this limit is \(10^{-20}\)—10 atm. The only exception is chromium oxide, whose dissociation elasticity is less than the indicated value, but which possesses fairly considerable activity. This is apparently explained by its transition, during contact, into other oxides having a higher dissociation elasticity.

Since the catalytic activity of oxides depends on their dissociation elasticity, it may be concluded that, in the process of ammonia oxidation on oxide catalysts, the oxygen of the crystal lattice of the oxide on the catalyst surface plays a major role. Moreover, in the process of catalysis there occur successively the reduction of the higher oxide to a lower one, or even to the metal, and oxidation of the metal or lower oxide to a higher one. The latter process proceeds more slowly for those oxides that have a high dissociation elasticity of oxygen. The decrease in the degree of oxidation on Ag₂O can apparently also be explained by the slow course of the process of oxidation of Ag to Ag₂O by oxygen. Hence it follows that active oxides must have such a value of \(P_{\mathrm{O}_2}\) at which both the oxidation processes and the reduction processes will proceed equally well.

As examinations of the catalyst mass after investigation of its catalytic activity showed, all oxides having a high oxygen elasticity underwent chemical change as a result of the catalytic process. Thus, MnO₂ was reduced to Mn₂O₃, Bi₂O₅ to Bi₂O₃, Ni₂O₃ to NiO, etc. The presence of these transitions indicates that, as the catalyst for the given reaction, a substance has been taken that is unstable under the existing conditions. This leads to the fact that, along with the catalytic processes, there also proceeds the reaction of interaction of the catalyst with the gas mixture and its transition into a new stable phase. The existence of two phases during the reaction is impossible, as Wagner and Hauffe [2] have shown. Consequently, some phase will be stable only at a definite ratio of the reacting components. The correctness of these propositions was demonstrated by Boreskov [3] using the example of oxidation of SO₂ to SO₃. Moreover, the completeness of the transition of the oxide taken into the phase stable under the given conditions will depend on the duration of the process and on the permeability of the newly forming phase to the reacting gas.

In these cases the catalytic properties of the catalyst will be determined not by the initially taken form of the substance, but by the newly formed compound [4].

Hence it becomes clear to us why, with time, the activity

$\mathrm{Ni_2O_3}$ or $\mathrm{PbO_2}$, falls correspondingly to the activity of freshly charged $\mathrm{NiO}$ and $\mathrm{PbO}$.

The dependence of the degree of oxidation of $\mathrm{NH_3}$ to $\mathrm{NO}$ on the dissociation elasticity of oxides, etc., allows us to assert that the process of oxidation of $\mathrm{NH_3}$ is carried out mainly at the expense of the oxygen of the surface layers of the oxide crystal lattice, and not at the expense of oxygen adsorbed by the catalyst surface. Consequently, the oxidation process in the presence of oxides is accompanied by more profound chemical transformations than this process on platinum, on which oxidation occurs at the expense of adsorbed oxygen.

Taking into account the fact that, in the process of ammonia oxidation, oxygen of the crystal lattice of the surface layer of the catalyst participates directly, it may be said that in this process the principal role is played by the forces of chemical affinity, whose action causes the formation and subsequent destruction, on the catalyst surface, of unstable intermediate compounds.

Proceeding from this, and also using Markov’s hypothesis [5], we assume that the process proceeds according to a scheme in which the oxidation of ammonia takes place through hydroxylamine (an intermediate compound):

\[ \mathrm{Fe_2O_3 + NH_3 \longrightarrow 2FeO + NH_2OH\ ads.} \]

\[ \mathrm{NH_2OH \longrightarrow NO + 3H} \qquad \text{(on the catalyst)} \]

\[ \mathrm{2(3H) + 3O \longrightarrow 3H_2O} \]

\[ \mathrm{NH_2OH \longrightarrow H_2O + NH \longrightarrow H_2O + N + H \overset{+O}{\longrightarrow} H_2O + N_2} \qquad \text{(outside the catalyst)} \]

\[ \mathrm{4FeO + O_2\ ads. \longrightarrow 2Fe_2O_3} \]

Undoubtedly, physical adsorption is also necessary for this process, but it plays a secondary role. By means of the latter, fresh molecules of $\mathrm{NH_3}$ and $\mathrm{O_2}$ are merely supplied to the active centers, at which the chemical processes leading to the formation of the final products take place.

When comparing the mechanism of $\mathrm{NH_3}$ oxidation on oxide and platinum catalysts, it is evident that their difference is as follows. On a platinum catalyst, $\mathrm{NH_3}$ molecules are oxidized by adsorbed oxygen atoms (not included in the crystal lattice of the catalyst), and replenishment of oxygen occurs through activated adsorption of the latter; on an oxide catalyst, however, the oxidation of $\mathrm{NH_3}$ occurs by oxygen included in the crystal lattice of the oxide. As a result of this, the higher oxide is reduced to a lower one, which is again oxidized by the oxygen of the gas mixture. In our opinion, the increase in contact time on oxide catalysts by a factor of 520–1300 in comparison with that on a platinum catalyst is due to the occurrence of this process.

Having noted the regularities in the dependence of the degree of oxidation of $\mathrm{NH_3}$ on various factors and having analyzed the proposed mechanism of oxidation of the latter, we shall give an explanation of the phenomena that we observed in our study.

A general regularity for all the oxides we investigated is that the isotherms $\alpha = f(V)$ pass through a maximum. At slight

at higher space velocities the lower yield of nitrogen oxide is due to the diffusion of nitrogen oxide counter to the flow of the ammonia–air mixture. A reaction proceeds between ammonia and nitrogen oxide, leading to the formation of elemental nitrogen. At high temperatures, the effect of oxidation of NH$_3$ in the gas phase and on the walls of the reaction vessel is added to the indicated effect.

An increase in the flow rate of the ammonia–air mixture leads to a decrease in the counter-diffusion of NO, to a decrease in the fraction of ammonia oxidized in the precatalyst space (owing to the decrease in residence time in it) in the gas phase and on the walls of the reaction vessel, and consequently to a decrease in nitrogen deficit.

The fall in the degree of oxidation of NH$_3$ after the maximum is due, first, to an insufficient residence time of NH$_3$ and O$_2$ in the catalyst zone; as a result, NH$_3$ and O$_2$ do not have time to be adsorbed on the contact surface and slip past the catalyst zone, where they can react with each other with nitrogen deficit; second, an increase in the space velocity of the gas leads to rapid removal of NH$_2$OH from the catalyst–gas-phase interface. The latter leads to a decrease in the partial pressure of hydroxylamine in the gas volume and, consequently, to an increase in the rate of its diffusion from the catalyst surface into the gas volume, where it decomposes to elemental nitrogen according to the equation

\[ \mathrm{NH_2OH \rightarrow N + H_2O + H.} \]

For the catalysts studied, the dependence of the degree of oxidation of NH$_3$ on its concentration in the gas mixture passes through a maximum; moreover, for each catalyst there exists a specific optimum ratio at which the greatest degree of oxidation is observed.

Table 1 gives the values of the optimum O$_2$ : NH$_3$ ratios for various oxides under the conditions indicated in this table.

Table 1

Optimum O$_2$ : NH$_3$ ratio for oxides at a pressure of 8 kg/cm$^2$

The passage of the curves $\alpha = f(X_{NH_3})$ through a maximum can be explained as follows. At high oxygen concentrations under elevated pressure, the driving force of oxygen sorption on the catalyst will exceed that of hydroxylamine; as a result, hydroxylamine will be strongly desorbed from the contact surface into the gas volume, where it decomposes to N$_2$. With a decrease in the oxygen concentration, the driving force of oxygen sorption toward the contact will decrease, which leads to an increase in the residence time of NH$_2$OH on the catalyst, where it decomposes to NO. At some definite O$_2$ : NH$_3$ ratio, specific for each cata-

catalyst, under which the best conditions are created both for adsorption of \(O_2\) and for conversion of \(NH_2OH\) into nitric oxide, the greatest degree of oxidation of \(NH_3\) to NO is observed. A further decrease in the oxygen concentration leads to the fact that, because of the lack of oxygen, the entire catalyst surface will not have time to be oxidized to the more active form. As a result, the catalytic activity of the catalyst decreases and, consequently, the degree of oxidation of \(NH_3\) to NO will also decrease.

The dependence of the degree of oxidation of \(NH_3\) to NO on temperature for most oxides also passes through a maximum.

The lower yields of nitric oxide at lower temperatures can be explained by the low rate of decomposition of hydroxylamine formed on the contact surface. For this reason, hydroxylamine is retained on the catalyst for a long time and hinders access of fresh portions of \(O_2\) and \(NH_3\) to it. As a result, in the zone and beyond the catalyst zone, interaction of \(NH_3\) with \(O_2\) and NO will occur with formation of \(N_2\).

When the temperature is increased to certain limits, the rate of decomposition of \(NH_2OH\) on the catalyst increases; consequently, the fraction of the catalyst surface freed from previously formed \(NH_2OH\) and oxidized to the more active form also increases, which leads to an increase in the adsorption rate of fresh portions of \(NH_3\) and \(O_2\) and in the yield of NO.

The decrease in the yield of NO when the temperature is raised above a certain limit (specific for each catalyst) is due to an excessively high rate of desorption of \(NH_2OH\) from the contact surface, while decomposition outside the contact occurs with formation of nitrogen.

The passage of the dependence \(\alpha = f(t)\) through a minimum for the catalysts \(Ni_2O_3\) and \(Bi_2O_5\) is due to decomposition of these oxides to lower ones, namely to \(NiO\) and \(Bi_2O_3\), which are less active than the former. It is quite realistic to suppose that at \(660^\circ C\) and \(740^\circ C\) the decomposition of \(Ni_2O_3\) and \(Bi_2O_5\) to lower oxides was completed, and that the increase in the degree of oxidation of \(NH_3\) with a further rise in temperature was caused by an increase in the catalytic activity of \(NiO\) and \(Bi_2O_3\).

At high temperatures overheating of the catalysts may also occur. This circumstance leads to a decrease in the yield of NO.

For all the oxide catalysts investigated, a sharp decrease in the yield of NO is observed with an increase in the pressure of the gas mixture above two atmospheres. This can apparently be explained by the fact that, as the pressure increases, the partial pressures of \(O_2\) and \(NH_3\) in the gas mixture also increase. The latter leads to an increase in the driving force of sorption of \(NH_3\) and \(O_2\) to the catalyst surface. As a result, the residence time of hydroxylamine on the contact will decrease, and an ever greater part of it will pass into the bulk, where it decomposes to \(N_2\). Of course, the decrease in NO yield with increasing pressure may also occur because of an increase in the amount of \(NH_3\) oxidized in the gas volume, as Markov suggests, but, as our studies with an empty contact apparatus have shown, this effect does not exert a decisive influence on the NO yield when the pressure is increased.

Conclusions

1. The highest activity is possessed by those oxides whose dissociation elasticity lies within certain limits, \(10^{-20}—10\) atmo-

spheres. Oxides whose dissociation elasticity lies above or below these limits prove to be catalytically only weakly active.

1. In this work an assumption is made concerning the mechanism of oxidation of \(\mathrm{NH}_3\), according to which the reaction involves the oxygen of the surface layers of the oxide.

2. The data presented provide some guidance in developing more active non-platinum catalysts.

References

1. F. M. Loskutov, Metallurgy of Zinc, Metallurgizdat, 1945.
2. C. Wagner, K. Hauffe, Z. Elektrochemie. 44, 172, 1938.
3. G. Boreskov, ZhFKh, 9–10, 1937, 1940.
4. P. G. Bruns, ZhFKh, 21, 1011, 1947.
5. V. P. Markov, Oxidation of Ammonia, GIAP Publishing House, 1948.
6. N. P. Kurin, M. S. Zakharov, “Relationship between the electronic and catalytic properties of metal oxides in ammonia oxidation under pressure” (the article is included in this collection).