PROCEEDINGS
OF THE TOMSK POLYTECHNIC INSTITUTE
ORDER OF THE RED BANNER OF LABOR
named after S. M. Kirov

Volume 151　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　1966

ON THE MODIFICATION OF OXIDE CATALYSTS FOR THE CONVERSION OF AMMONIA TO NITRIC OXIDE

P. E. BOGDANOV, R. K. TARANENKO

(Presented by the Department of Technology of Silicates and Inorganic Substances)

Cobalt oxide catalysts, under certain conditions, can be highly effective contacts for the process of oxidizing ammonia to nitric oxide [1, 2]. According to [2], active catalysts are obtained by preparing contacts by the method of precipitating cobaltous hydroxide from a solution of cobalt nitrate with caustic potash, followed by treatment of the cobaltous hydroxide with hydrogen peroxide and calcination of cobalt oxide hydrate to Co₃O₄.

The higher activity of Co₃O₄ prepared with the aid of KOH, in comparison with catalysts obtained by calcining cobalt nitrate [2], may be explained by the presence of potassium impurities in the contact. In the present work, using X-ray and spectral methods and the electronic theory of catalysis, we attempted to determine the reasons for the increased activity of catalyst samples promoted with potassium. Figure 1 presents three X-ray diffraction patterns of cobaltous-cobaltic oxide prepared by the precipitation method with KOH. X-ray pattern I corresponds to freshly prepared cobaltous-cobaltic oxide treated at 105°. X-ray pattern II was obtained for a sample calcined at 550°, and, finally, X-ray pattern III was taken for a sample of cobaltous-cobaltic oxide after testing it for catalytic activity at temperatures of 600–850°.

A comparison of the interplanar spacings of these catalyst samples with the data of the X-ray determinant [9] shows that the phase composition of the freshly prepared cobalt oxide catalyst and of the catalyst that had been in operation did not change (Table 1). All the principal lines of the three X-ray diffraction patterns under consideration correspond to the composition Co₃O₄, with the exception of lines 1 and 7, corresponding to the composition Co₂O₃. The CoO phase is absent in the samples; likewise, no lines corresponding to the K₂O phase were detected. However, spectral analysis of the catalyst samples, carried out on a KSA-1 apparatus by burning the powdered sample in copper electrodes, showed that the catalyst contains potassium. The concentration of the latter does not exceed 0.0001%.

The absence of lines for CoO and of lines characteristic of potassium in the X-ray diffraction patterns under consideration suggests that the catalyst samples studied are spinel-type compounds (CoCo₂O₄) with incorporation of potassium ions into the crystal lattice of cobaltous-cobaltic oxide in the form of impurity ions. As was shown in [2], the activity of the catalyst in the process of ammonia oxidation is preser-

is high, despite the considerable change in the specific surface area of the contact during the reaction. Obviously, the reason for the high activity of cobalt oxide–cobalt sesquioxide, prepared by precipitation, lies in the electronic properties of the catalysts. It is known that cobalt oxide–cobalt sesquioxide has hole conductivity, which can be intensified in the process of preparing the catalyst through the capture of ions by the main mass of $\mathrm{Co_3O_4}$. It may be expected that, during the heat treatment of cobalt oxide hydrate, either a solid solution of $\mathrm{K_2O}$ in $\mathrm{Co_3O_4}$ is formed with the appearance of point defects (ionic vacancies, impurity ions), or defects are formed that lead to a change in the external and internal surface of the catalyst as a consequence of changes in the crystal habit, the appearance of dislocations and their groupings, etc. [3, 4]. From the standpoint of semiconductor catalysis, $\mathrm{Co_3O_4}$ should be classed among semiconductors with double valence.

Table 1

Characteristics of X-ray pattern III for Fig. 1

To obtain a more complete picture of the structural constitution of such a semiconductor, let us consider the crystal lattice of $\mathrm{Fe_3O_4}$, which is itself a semiconductor with double valence. The structure of $\mathrm{Fe_3O_4}$ has been studied rather fully by Verwey and co-workers [5], who believe that at ordinary temperature all $\mathrm{Fe^{++}}$ ions and half of the $\mathrm{Fe^{+++}}$ ions are distributed statistically in the unit cell over cation sites of one type—octahedral. The remaining $\mathrm{Fe^{+++}}$ ions occupy tetrahedral positions. The electronic conductivity, magnetic, optical, and other properties of ferrous–ferric oxide are due to the ease of electron transfer from $\mathrm{Fe^{++}}$ ions to $\mathrm{Fe^{+++}}$ ions. The appearance of defects in the crystal lattice of the oxide during heat treatment is reflected in a decrease in conductivity. This indicates the appearance of additional holes in the crystal; consequently, the $p$-conductivity of $\mathrm{Fe_3O_4}$ in this case increases.

It should be noted that, in the process of formation of the crystal lattice, an ordered occupation of octahedral or tetrahedral interstices does not always occur. Ions of the same element with different valence, ions of different elements, can enter these interstices.

tions, etc. In this case, the formation of a spinel of a real crystal is associated with a substantial change in the electron density in the lattice.

Analogously to the structure of \(\mathrm{Fe_3O_4}\), one can represent the structure of the crystal lattice of cobalt oxide–oxide with defects of impurity potassium ions. The capture of potassium ions by the crystal lattice of cobalt oxide–oxide has been confirmed by us by the above X-ray and spectral methods. If one starts from the values of the ionic radii, then the direct introduction of potassium ions into a cation vacancy of the \(\mathrm{Co_3O_4}\) crystal lattice is difficult, since the ionic radius of \(\mathrm{K^+}\) is \(1.33\ \text{\AA}\), while the ionic radii of \(\mathrm{Co^{++}}\) and \(\mathrm{Co^{+++}}\) are respectively \(0.82\) and \(0.64\ \text{\AA}\). Apparently, inclusion of potassium as an impurity ion may take place if, in the crystal—

Fig. 1. X-ray pattern I—for a freshly prepared cobalt oxide catalyst treated at \(105^\circ\mathrm{C}\). X-ray pattern II—for a \(\mathrm{Co_3O_4}\) sample calcined at \(550^\circ\mathrm{C}\). X-ray pattern III—for a \(\mathrm{Co_3O_4}\) sample after testing for catalytic activity in the temperature range \(600\text{–}850^\circ\mathrm{C}\).

—lattice of \(\mathrm{Co_3O_4}\) there is a defect of a pair of vacancies: “cation–anion,” according to Zeitz [7]. The ionic radius of the oxygen anion is \(1.36\ \text{\AA}\). Taking into account that the potassium ion also has a polarizability coefficient equal to \(0.87\), one may assume its inclusion in the vacant site of the “cation–anion” pair. Replacement of the principal ion in the lattice \((\mathrm{Co^{++}})\) by an ion with a lower charge \((\mathrm{K^+})\) will lead to the formation of a certain number of trivalent cobalt ions \((\mathrm{Co^{+++}})\) in accordance with the co-

...by the number of incorporated monovalent potassium ions. In this case, the electroneutrality of the crystal will be compensated by oxygen from the gas phase. Such a process may be represented as follows:

\[ 2\,\mathrm{Co}^{++}+\mathrm{K}_2\mathrm{O}+\frac{3}{2}\mathrm{O}_2 \longrightarrow 2\,\mathrm{Co}^{+++}+2\mathrm{K}^{+}+4\mathrm{O}_2^{--} \]

The capture of the potassium ion by the surface layers or by the bulk of the crystal lattice of \(\mathrm{Co}_3\mathrm{O}_4\) leads to a shift of the Fermi level, on whose position the catalytic activity of substances depends significantly [6]. If cobaltous-cobaltic oxide is represented in the form of the compound \(\mathrm{CoO}\cdot\mathrm{Co}_2\mathrm{O}_3\), then, to a first approximation, one may admit the incorporation of potassium ions into the \(\mathrm{Co}_3\mathrm{O}_4\) lattice according to the scheme:

\[ \begin{array}{ccccccccc} \mathrm{Co}^{++} & \mathrm{O}^{--} & \mathrm{Co}^{++} & \vdots & & \vdots & \mathrm{Co}^{++} & \mathrm{O}^{--} & \mathrm{Co}^{++} \\ \mathrm{O}^{--} & \mathrm{Co}^{++} & \mathrm{O}^{--} & \vdots & \boxed{\mathrm{K}^{+}} & \vdots & \mathrm{O}^{--} & \mathrm{Co}^{++} & \mathrm{O}^{--} \\ \mathrm{Co}^{++} & \mathrm{O}^{--} & \mathrm{Co}^{++}\oplus & & \mathrm{O}^{--} & \mathrm{Co}^{++} & \mathrm{O}^{--} & \mathrm{Co}^{++} \\ \mathrm{O}^{--} & \mathrm{Co}^{++} & \mathrm{O}^{--} & & \mathrm{Co}^{++} & \mathrm{O}^{--} & \mathrm{Co}^{++} & \mathrm{O}^{--} \end{array} \]

Thus, the appearance of a potassium ion in the crystal lattice leads to the displacement of an electron from the \(\mathrm{Co}^{++}\) ion, converting the latter into a trivalent ion. In other words, an additional hole has appeared in the lattice and, consequently, the positive conductivity, caused as it were by the displacement of holes owing to electron transitions, will increase.

An increase in hole conductivity should lead to an improvement in the process of ammonia oxidation to nitric oxide in the presence of \(\mathrm{Co}_3\mathrm{O}_4\) with potassium ions admixed, since the holes will serve as a “trap” for capturing electrons of oxygen adsorbed from the gas phase. The considerations presented confirm the mechanism of ammonia oxidation to nitric oxide according to N. P. Kurin and M. S. Zakharov [8].

REFERENCES

1. M. I. Temkin et al. Problems of Physical Chemistry (Transactions of the L. Ya. Karpov Physicochemical Institute) 2, 14, 1959.

2. N. P. Kurin, P. E. Bogdanov. Catalysis in Higher Education, Part II, p. 215, Moscow State University Press, 1962.

3. A. Ries. Chemistry of Crystals with Defects, Foreign Literature Publishing House, 1956.

4. V. V. Boldyrev. The Influence of Defects in Crystals on the Rate of Thermal Decomposition of Solids, Tomsk University Press, 1963.

5. E. J. Weywey, J. H. Boer, Proc. Phys. Soc., 49, 59, 1937.

6. F. F. Volkenshtein. Semiconductors as Catalysts of Chemical Reactions, Znanie Publishing House, Leningrad, 1956.

7. Cited after F. Stoun. Collection Chemistry of the Solid State, Foreign Literature Publishing House, p. 60, Moscow, 1961.

8. N. P. Kurin, M. S. Zakharov. Catalysis in Higher Education. Part II, p. 234, Moscow State University Press, 1962.

9. American X-ray Powder Data File card index; key to it: Index to the X-ray Powder Data File, 1960; XRDC, 1945.