PHYSICAL CHEMISTRY

R. E. MARDALЕISHVILI, A. G. POPOV, V. V. NIKISHA, and F. S. YAKUSHIN

ON TWO TYPES OF ELEMENTARY REACTIONS IN THE CATALYTIC HYDROGENATION OF OLEFINS

(Presented by Academician N. N. Semenov, 14 VII 1960)

Several years ago, N. N. Semenov, V. V. Voevodskii, and F. F. Vol’kenshtein \((^1)\) advanced ideas concerning the presence of free valences on solid surfaces and the possibility that, upon adsorption of various substances on them, so-called surface radicals are formed. According to the authors, a number of heterogeneous catalytic processes proceed with the participation of these surface radicals. The authors consider, in particular, that during adsorption and in the course of hydrogenation of olefins on the catalyst surface, radicals of two types are formed; for example, in the case of ethylene,

\[ \mathrm{C_2H_4 + 2\dot{k} \rightleftarrows CH_2-\dot{CH}_2 + \dot{k} \rightleftarrows CH_2-CH_2,} \tag{1} \]

\[ \begin{array}{c} \phantom{\mathrm{C_2H_4 + 2\dot{k} \rightleftarrows }}\Big| \\ \mathrm{k} \end{array} \qquad\qquad \begin{array}{c} \phantom{\mathrm{CH_2-}}\Big| \quad \Big| \\ \mathrm{k \quad k} \end{array} \]

\[ (a) \]

\[ \mathrm{CH_2-\dot{CH}_2 + \dot{H} \rightleftarrows CH_2-CH_3 \rightleftarrows \dot{CH}_2-CH_3} \tag{2} \]

\[ \begin{array}{cccc} \Big| & \dot{} & \Big| & \dot{} \\ \mathrm{k} & \mathrm{k} & \mathrm{k} & \mathrm{k} \end{array} \]

\[ (b) \qquad\qquad\qquad\qquad (b) \]

radicals (a), bound to the catalyst by a two-electron bond, and (b)—by a one-electron bond. The recombination reaction of the latter with hydrogen atoms leads to formation of the final product, in this case ethane,

\[ \mathrm{\dot{CH}_2-CH_3 + \dot{H} \rightarrow C_2H_6 + 2k.} \tag{3} \]

\[ \begin{array}{cc} \dot{} & \dot{} \\ \mathrm{k} & \mathrm{k} \end{array} \]

Chemical equations (1)—(3), in the symbols of the radical concepts under consideration, express the generally accepted scheme for the mechanism of catalytic hydrogenation of olefins \((^2)\). It should be noted that, if one adopts the idea of the reality of surface radicals, then it should be expected that, during catalytic hydrogenation of olefins, alongside the recombination processes (2)—(3), disproportionation reactions of the following type should also occur under certain conditions:

\[ \mathrm{CH_2-\dot{CH}_2 + CH_2-\dot{CH}_2 \rightarrow CH_2-CH_3 + CH=CH_2,} \tag{4} \]

\[ \begin{array}{ccc} \Big| & \Big| & \Big| \\ \mathrm{k} & \mathrm{k} & \mathrm{k} \end{array} \qquad \begin{array}{c} \Big| \\ \mathrm{k} \end{array} \]

\[ \mathrm{\dot{CH}_2-CH_3 + \dot{CH}_2-CH_3 \rightarrow CH_2-\dot{CH}_2 + C_2H_6 + \dot{k},} \tag{5} \]

\[ \begin{array}{cccc} \dot{} & \dot{} & \Big| & \\ \mathrm{k} & \mathrm{k} & \mathrm{k} & \end{array} \]

\[ \mathrm{\dot{CH}_2-CH_3 + CH_2-\dot{CH}_2 \rightarrow CH=CH_2 + C_2H_6 + \dot{k},} \tag{6} \]

\[ \begin{array}{ccc} \dot{} & \Big| & \Big| \\ \mathrm{k} & \mathrm{k} & \mathrm{k} \end{array} \]

with transfer of a hydrogen atom from one surface radical to another by a mechanism known for homogeneous free radicals \((^3)\).

Similar ideas regarding the elementary act of disproportionation of so-called “semi-hydrogenated complexes” \((^4)\) have been expressed by a number of authors \((^5)\) in explaining experimental data on the distribution of deuterium in the products of catalytic hydrogenation of olefins with deuterium.

Since, however, the interpretation of these data, in the opinion of other authors \((^{2,4,6})\), can also be carried out without invoking the concept of disproportionation of “semihydrogenated complexes,” the question of the reality of this process remains open.

Determining the shares of the recombination processes (2)—(3) and disproportionation (4)—(6) in the course of olefin hydrogenation by ordinary methods presents well-known difficulties, since both lead to the formation of one and the same final product. Meanwhile, it is obvious that the ratio of the rates of these processes should be determined by the ratio of the concentrations of radicals and hydrogen atoms on the catalyst surface. It is precisely this circumstance that we have used in the present work, carried out with the aim of experimentally verifying the concept of surface radicals, from which follows the conclusion that disproportionation processes of the type (4)—(6) may exist. We studied the hydrogenation of ethylene, propylene, and their mixtures at various hydrogen pressures and temperatures. If the assumption of the reality of surface radicals and their disproportionation is correct, then during the hydrogenation of mixtures of two olefins, for example ethylene and propylene, a new reaction should appear that does not occur during the hydrogenation of each of the olefins separately. Such a reaction may be the disproportionation of two radicals formed from different olefins, for example

\[ \mathrm{CH_3-\dot{C}H-CH_2} + \mathrm{\dot{C}H_2-CH_2} \longrightarrow \mathrm{CH_2=CH-CH_2} + \mathrm{CH_3-CH_2}, \tag{7} \]

\[ \begin{array}{cccc} \big| & \big| & \big| & \big|\\ \mathrm{K} & \mathrm{K} & \mathrm{K} & \mathrm{K} \end{array} \]

in this case one may expect that, owing to the different activity of C3 and C2 radicals, transfer of a hydrogen atom will occur predominantly from a radical of one structure to another, and, in particular, by analogy with homologous ethyl and propyl radicals, from adsorbed propylene to adsorbed ethylene (reaction (7)). This circumstance should lead to the fact that the values of the ratio of the initial rates \(w_{32}/w_{23}\) of hydrogenation of propylene and ethylene in their mixture will be smaller than the values of the ratio of the initial rates \(w_3/w_2\), found during hydrogenation of each of these olefins separately. This difference in the rate ratios \(w_{32}/w_{23}\) and \(w_3/w_2\) should decrease with increasing hydrogen pressure, since under these conditions the recombination processes of the corresponding surface radicals with hydrogen atoms should become predominant. The ratio of the rates of the recombination processes, however, should not depend on whether each olefin is hydrogenated separately or in mixture with the other (provided that the adsorption of these olefins is the same, in an excess of hydrogen both ratios must be equal).

An analogous picture should occur for the temperature dependence of the ratio of the rates of formation of propane and ethane during hydrogenation of the corresponding olefins separately and in their mixtures. The greatest difference in the ratios \(w_{32}/w_{23}\) and \(w_3/w_2\) should occur at low temperatures, at which the concentration of olefin on the surface, and consequently the share of the disproportionation process, are significant. With increasing temperature, especially up to the temperature corresponding to the maximum hydrogenation rate and higher, the olefin concentration as a result of desorption decreases so much that the disproportionation processes may be neglected in comparison with their recombination with hydrogen atoms. Therefore, with increasing temperature the values of \(w_{32}/w_3\) and \(w_2/w_2\) will converge, and since the difference in the adsorptions of ethylene and propylene decreases at the same time, this should lead to coincidence of the values of the indicated rate ratios.

In our work, experiments to establish the predicted regularities were carried out in a vacuum apparatus with circulation of the reacting gases through a reactor with a catalyst, for which pla-

platinum wire (\(d=0.1\) mm, \(l=150\) mm), heated by an electric current. The course of the reaction was followed both manometrically, from the change in pressure in the system, and by mass-spectrometric analysis of samples taken at different degrees of conversion. The reaction was carried out at pressures up to 525 mm Hg, olefin : hydrogen component ratios from 1 : 1 to 1 : 20, in the temperature interval 0–280°, and at the maximum catalyst activity, which was attained by heating the platinum wire in oxygen and then in hydrogen at 300° for 30 min before each experiment. To prevent poisoning of the catalyst during the experiment by mercury vapors and grease, the circulating gases before and after the reactor were passed through a trap cooled to \(-70^\circ\). The reactor was separated from the rest of the apparatus by metal stopcocks and between experiments was evacuated to a pressure of \(10^{-6}\) mm Hg.

Table 1

Table 1 gives the values of the initial rates of formation of ethane \(w_2\) and propane \(w_3\) in the hydrogenation of each of the corresponding olefins, and also of ethane \(w_{23}\) and propane \(w_{32}\) in the hydrogenation of a mixture of ethylene and propylene at different hydrogen pressures. The ratio of the rate \(w\) of hydrogenation of the ethylene–propylene mixture to the mean hydrogenation rate of each of the olefins separately is close to unity, which may serve as a characteristic of the reproducible action of the catalyst.

Considering the value of \(w_3/w_2\) at different hydrogen pressures in the system (Fig. 1), one can see that at first it decreases sharply and then reaches a limiting value. This change in the ratio \(w_3/w_2\) is due to the higher order with respect to hydrogen for the hydrogenation of ethylene (\(w_2 \sim p_{\mathrm{H}_2}\)) in comparison with propylene (\(w_3 \sim \sqrt{p_{\mathrm{H}_2}}\)), which is in agreement with literature data. However, in the hydrogenation of a mixture of these olefins, with increasing hydrogen pressure the opposite dependence is observed, namely an increase in the ratio \(w_{32}/w_{23}\) up to a certain value, which agrees with the picture expected by us. If these data are plotted in the coordinates \(w_{23}/w_2\) and \(w_{32}/w_3\) as functions of \(p_{\mathrm{H}_2}\), then, provided the adsorption of ethylene and propylene is the same, both values should tend to 0.5, since in this case it may be assumed that each of the olefins is hydrogenated on half of the catalyst area present in the system, whereas in the hydrogenation of the individual olefins the entire surface participates. The experimental values of \(w_{23}/w_2\) and \(w_{32}/w_3\), with increasing hydrogen pressure, tend respectively to 0.43 and 0.63, which is apparently due mainly to greater adsorption of propylene than of ethylene.

Figure 2 presents the experimental data obtained in studying the temperature dependence of \(w_3/w_2\) and \(w_{32}/w_{23}\). Curve 1, corresponding to the case of hydrogenation of individual propylene and ethylene on Pt, confirms the temperature behavior of the ratio of the hydrogenation rates of these olefins found by Toyama for a nickel catalyst \((^7)\). Curve 2 expresses the temperature dependence of the ratio of the hydrogenation rates of propylene and ethylene in their mixture.

Fig. 1. Ratio of the rates of formation of propane and ethane as a function of hydrogen pressure during hydrogenation. Designations here and in Fig. 2:
1 — \((w_3/w_2)\)—for individual \(\mathrm{C_3H_6}\) and \(\mathrm{C_2H_4}\);
2 — \((w_{32}/w_{23})\)—for their equimolar mixture;
3 — \((2w/w_2+w_3)\)—a curve characterizing the reproducibility of the experimental data.

ethylene in their mixture. From a comparison of curves 1 and 2 it is seen that, in the mixture of olefins studied by us, the relative rate of hydrogenation of ethylene is higher than in the case of hydrogenation of each of these hydrocarbons separately.

Thus, the ratios of the initial rates of formation of propane and ethane during hydrogenation of the corresponding olefins and of their mixtures fully confirm the dependences of these quantities on temperature and hydrogen pressure predicted by us. Therefore these results may be regarded as confirmation of the correctness of the conclusion concerning two types of elementary acts—recombination and disproportionation of intermediate labile products formed in the course of olefin hydrogenation. At the same time, these data may serve as an argument in favor of the existence of surface radicals.

Fig. 2.

Moscow State University
named after M. V. Lomonosov

Received
7 VII 1960

REFERENCES

1. V. V. Voevodskii, F. F. Vol'kenshtein, N. N. Semenov, in: Problems of Chemical Kinetics, Catalysis, and Reactivity, Publishing House of the Academy of Sciences of the USSR, 1955, p. 423; N. N. Semenov, On Some Problems of Chemical Kinetics and Reactivity, Publishing House of the Academy of Sciences of the USSR, 1959, p. 272.
2. D. D. Eley, Catalysis, 3, N. Y., 1955, p. 49; H. A. Taylor, Catalysis, 5, N. Y., 1957, p. 257; J. Horiuti, Catalysis 6, 250 (1958).
3. E. W. R. Steacie, Atomic and Free Radical Reactions, N. Y., 1954.
4. M. Polanyi, J. Horiuti, Trans. Farad. Soc., 30, 1164 (1934); M. Polanyi, R. K. Greenhalgh, Trans. Farad. Soc., 35, 520 (1939); E. K. Rideal, G. H. Twigg, Proc. Roy. Soc., London, A 171, 55 (1939); D. D. Eley, Advances in Catalysis, 1, N. Y., 1948, p. 157; D. D. Eley, Quart. Rev. Chem. Soc., 3, 209 (1949).
5. J. Turkevich, D. Schissler, P. Irsa, J. Coll. and Phys. Chem., 55, 1078 (1951); G. G. Bond, Trans. Farad. Soc., 52, 1235 (1956).
6. T. Keii, J. Chem. Phys., 22, 144 (1954).
7. I. Jenkins, E. K. Rideal, J. Chem. Soc., 1955, 2496.
8. O. Toyama, Rev. Phys. Chem., Japan, 14, 86 (1940).