Chemistry

M. Yu. Lukina, S. V. Zotova, and Academician B. A. Kazanskii

The Influence of the Nature of Supports on the Direction of the Catalytic Hydrogenation Reaction of Hydrocarbons of the Cyclopropane Series

Earlier ((^1)) we had occasion to note that in the literature on the catalytic hydrogenation of cyclopropane hydrocarbons there are many contradictory data concerning the direction of rupture of the three-membered ring. Most authors believe that in this reaction, with addition of hydrogen, the bond between the two most hydrogenated carbon atoms is cleaved ((^{1-6})), for example:

[
\ce{
\begin{array}{c}
\mathrm{CH_2}\[-2pt]
\diagup\quad \diagdown\[-2pt]
\mathrm{CH_2}\quad \mathrm{CH{-}CH_2{-}CH_3}
\end{array}
+ H2 -> CH3-CH(CH3)-CH2-CH3
}
\tag{I}
]

However, several cases have also been described of ring cleavage in other directions ((^7,^8)). In particular, Slobodin and co-workers ((^9)) consider that, in the hydrogenation of gem-dimethylcyclopropane in the presence of nickel deposited on kieselguhr, the reaction proceeds as follows:

[
\ce{
\begin{array}{c}
\mathrm{CH_2}\quad \mathrm{CH_3}\[-2pt]
| \ \diagdown \quad |\[-2pt]
\mathrm{CH_2}\quad \mathrm{C}\quad \mathrm{CH_3}
\end{array}
+ H2 -> CH3-CH(CH3)-CH2-CH3
}
\tag{II}
]

These authors come to the conclusion that in the given case the hydrogenolysis of the cyclopropane ring proceeds through an intermediate stage of isomerization of the alkylcyclopropane into an olefin, which is then hydrogenated; the presence of olefin in the products of incomplete hydrogenation was proved by them with the aid of combination-scattering spectra.

The authors express no considerations regarding the reasons for the different direction of rupture of the three-membered ring and, apparently, are inclined to attribute it to features of the structure of one or another hydrocarbon.

There is, however, one important factor capable of influencing the direction of cleavage of the three-membered ring during hydrogenation, namely, the support of the hydrogenating metal, since the catalysts employed are usually supported catalysts. As we showed earlier, silica gel ((^{10})) and activated charcoal ((^{11})) catalyze the isomerization reaction of cyclopropane hydrocarbons into olefins, though under different conditions: silica gel even at a temperature close to (0^\circ), activated charcoal only at a temperature of about (200^\circ). Isomerization occurs through rupture of the C—C bond located at the least hydrogenated carbon atom of the ring:

[
\ce{
\begin{array}{c}
\mathrm{CH_2}\[-2pt]
| \ \diagdown\[-2pt]
\mathrm{CH_2}\quad \mathrm{CH{-}R}
\end{array}
-> CH2=CH-CH-R
}
\tag{III}
]

It is natural to suppose that if hydrogenating metals are deposited on supports capable, like silica gel and activated charcoal, of bringing about the isomerization of cyclopropane hydrocarbons into olefins, then in the presence of such catalysts and under suitable conditions the direction of cleav-

cleavage of the ring will be different than when inert supports are used. Indeed, in the first case the olefin that will be hydrogenated is the one formed during isomerization of the alkylcyclopropane (equation (III)), whereas in the latter case it is the cyclopropane hydrocarbon itself (equation (I)). The relative isomerizing activity of such supported catalysts will determine the direction of cleavage of the three-membered ring.

In the present work we describe experiments on the isomerization of certain cyclopropane hydrocarbons to olefins in the presence of such substances as aluminosilicate, kieselguhr, and pumice, which are often used as supports for preparing hydrogenating catalysts. It proved that, in the presence of aluminosilicate, ethylcyclopropane was almost completely isomerized to a mixture of normal olefins already at (50^\circ); in the presence of kieselguhr, by 75% at (120^\circ); whereas on pumice at (120^\circ) isomerization did not occur at all, at (170^\circ) it proceeded to the extent of 20%, and at (220^\circ), to 45%.

If the hydrogenation of cyclopropane hydrocarbons is carried out in the presence of platinized carbon and platinized kieselguhr at (150^\circ), i.e., at a temperature at which activated carbon does not isomerize hydrocarbons with a three-membered ring to olefins, while kieselguhr isomerizes them rather intensively, then in the case of platinized carbon only cleavage of the ring along the C—C bond between the most hydrogenated carbon atoms takes place:

[
\ce{
(CH3)-CH(-CH2)-C(CH3)2
->[\ce{+H2}][\ce{Pt/C}]
CH3-C(CH3)2-CH2-CH3
}
\qquad (100\%)
\tag{IV}
]

In the presence of platinized kieselguhr, however, chiefly other bonds of the three-membered ring are cleaved, and 2,2-dimethylbutane is formed only in an amount of 15%.

[
\ce{
(CH3)-CH(-CH2)-C(CH3)2
->[\ce{+H2}][\ce{Pt/kieselguhr}]
\begin{cases}
\ce{CH3-C(CH3)2-CH2-CH3} & 15\% \
\ce{CH3-CH(CH3)-CH(CH3)-CH3} & \
\ce{CH3-CH2-CH2-CH(CH3)-CH3} &
\end{cases}
}
\left.\begin{array}{c}
\
\
\end{array}\right} (85\%)
\tag{V}
]

Thus, the influence of a support active in the sense of isomerization on the direction of hydrogenolysis of the three-membered ring is beyond doubt. This can explain the different results obtained in those cases where the supports are kieselguhr or aluminosilicate, on the one hand, and pumice or activated carbon, on the other, although the hydrogenating metal is one and the same.

This, apparently, can also explain the differences in the kinetics of hydrogenation of cyclopropane that were observed by Bond and coauthors ((^{12,13})) in the presence of nickel on pumice and by Benson and Kwan ((^{14})) in the presence of nickel on aluminosilicate. It should be noted that both groups of authors believe that the mechanism of hydrogenation of cyclopropane consists in collisions of molecules of gaseous cyclopropane with hydrogen atoms adsorbed on the catalyst, which leads to the formation of cyclopropyl or propyl radicals. However, apparently, in Bond’s experiments such a reaction did indeed occur, while in Benson’s experiments a different reaction occurred, since there hydrogenation was preceded by isomerization of cyclopropane to propylene at the expense of the aluminosilicate included in the catalyst. The same reaction occurred in the work cited above by Slobodina ((^9)) and coworkers, where nickel on kieselguhr was used as the catalyst.

Experimental Part

Isomerization of ethylcyclopropane in the presence of aluminosilicate. Ethylcyclopropane with b.p. (35.9^\circ) (760 mm),

$n_D^{20}$ 1.3786; $d_4^{20}$ 0.6841 and an aniline point of $+17.9^\circ$ was passed through a tube with 10 ml of aluminosilicate catalyst at $50^\circ$ with a space velocity of 0.2 hr$^{-1}$. The resulting catalyzate instantaneously decolorized bromine water and, judging by the bromine numbers, contained 88.4% unsaturated compounds.

Isomerization of ethylcyclopropane in the presence of kieselguhr. “Kisatibi” kieselguhr was finely ground, moistened with water, formed into pieces (3 × 3 mm), and dried at $100^\circ$.

Ethylcyclopropane was passed through a tube with 10 ml of kieselguhr at $75^\circ$ with a space velocity of 0.2 hr$^{-1}$. The catalyzate decolorized bromine water with difficulty. The catalyzate obtained at $120^\circ$ instantaneously decolorized bromine water and had $n_D^{20}$ 1.3798 and $d_4^{20}$ 0.6659. The product of its hydrogenation with platinum black had b.p. 34.0–36.0°, $n_D^{20}$ 1.3624 and $d_4^{20}$ 0.6398. On a 40-t.p. column the following fractions were obtained (Table 1).

Table 1

From the properties of the fractions it is evident that the catalyzate consists of n-pentane and ethylcyclopropane. Calculation of the composition of the fractions from the specific volumes and aniline points shows that the catalyzate contains 74.2% n-pentane and 25.8% ethylcyclopropane (for the properties of n-pentane see Table 4).

Isomerization of 1,1,2-trimethylcyclopropane in the presence of pumice. Pumice in pieces 6 × 3 mm was thoroughly purified from iron by boiling with HCl and washing with water until a negative reaction for CNS′ was obtained. After drying at $120^\circ$, 10 ml of pumice were placed in the catalytic tube and heated to $300^\circ$ in a nitrogen atmosphere. 1,1,2-Trimethylcyclopropane (b.p. 52.6° (760 mm), $n_D^{20}$ 1.3862, $d_4^0$ 0.6948) was passed over the pumice at temperatures of 120–220° with a space velocity of 0.2 hr$^{-1}$. The results of the experiments are given in Table 2.

Table 2

| Experiment No. | Temp., °C | \multicolumn{3}{c}{Properties of the catalyzate} |
|---:|---:|---:|---:|---:|
| Experiment No. | Temp., °C | $n_D^{20}$ | bromine number | content of unsaturated compounds, % |
| 1 | 120 | 1.3862 | — | 0 |
| 2 | 170 | 1.3889 | 38.4 | 20.1 |
| 3 | 220 | 1.3915 | 85.6 | 45.0 |

Thus, 1,1,2-trimethylcyclopropane is unchanged at $120^\circ$, while at 170 and 220° it is isomerized by 20% and 45%, respectively.

Hydrogenation of 1,1,2-trimethylcyclopropane in the presence of platinized kieselguhr. 1,1,2-Trimethylcyclopropane was passed through a tube with 10 ml of platinized “Kisatibi” kieselguhr (20% Pt) in a stream of hydrogen at $150^\circ$ and a space velocity of 0.2 hr$^{-1}$.

The saturated catalyzate ($n_D^{20}$ 1.3740, $d_4^{20}$ 0.6583) was fractionated on a 40-t.p. column (Table 3).

The constants of the hydrocarbons that could be formed from ethylcyclopropane and 1,1,2-trimethylcyclopropane are given in Table 4.

On the basis of comparison of the data in Tables 3 and 4, it may be concluded that fraction 1 consists of 2,2-dimethylbutane, which is also present in fraction 2. Fractions 3, 4, and 5 consist mainly of 2,3-dimethylbutane. Fractions 6 and 7 also contain 2-methylpentane. Thus, the catalyzate consists approximately of 15% of the product of direct hydrogenolysis (2,2-dimethylbutane) and 55% of the product of isomerization to olefins followed by hydrogenation (2,3-dimethylbutane and 2-methylpentane).

Hydrogenation of ethylcyclopropane in the presence of platinized kieselguhr was carried out

Table 3

Table 4

Table 5

as also in the case of 1,1,2-trimethylcyclopropane. The fractionation of the limiting catalyst ($n_D^{20}$ 1.3569; $d_4^{20}$ 0.6242) is given in Table 5.

From comparison of the data in Tables 4 and 5 it is seen that fractions 1 and 2 contain n-pentane with a small admixture of 2-methylbutane. Fractions 3 and 4 consist of pure n-pentane. The composition of the catalyst is approximately as follows: 87% n-pentane and 15% 2-methylbutane.

Hydrogenation of 1,1,2-trimethylcyclopropane in the presence of platinized charcoal. 1,1,2-Trimethylcyclopropane was passed through a tube with platinized charcoal (20% Pt) at 150° and a space velocity of 0.2 h$^{-1}$. The limiting catalyst (3.8 g) had b.p. 50.1° (760 mm), $n_D^{20}$ 1.3685 and $d_4^{20}$ 0.6495, i.e., it was 2,2-dimethylbutane.

Zelinskii Institute of Organic Chemistry,
Academy of Sciences of the USSR

Received
18 VII 1958

REFERENCES

1. M. Yu. Lukina, V. A. Ovodova, B. A. Kazanskii, DAN, 97, 683 (1954).
2. N. M. Kizhner, ZhRKhO, 44, 165 (1912).
3. N. A. Rozanov, ZhRKhO, 48, 175 (1916).
4. R. W. Shortridge, R. A. Craig et al., J. Am. Chem. Soc., 70, 946 (1948).
5. R. V. Volkenburgh, K. W. Greenlee et al., J. Am. Chem. Soc., 71, 172 (1949).
6. B. A. Kazanskii, M. Yu. Lukina, V. A. Ovodova, Izv. AN SSSR, OKhN, 1954, 878.
7. R. G. Kelso, K. W. Greenlee et al., J. Am. Chem. Soc., 74, 287 (1952).
8. N. M. Kizhner, ZhRKhO, 45, 987 (1913).
9. Ya. M. Slobodin, V. I. Grigor’eva, Ya. E. Shmulyakovskii, ZhOKh, 23, 1480 (1953).
10. B. A. Kazanskii, V. T. Aleksanyan et al., Izv. AN SSSR, OKhN, 1955, 1118.
11. M. Yu. Lukina, S. V. Zotova, B. A. Kazanskii, DAN, 116, 793 (1957).
12. G. C. Bond, J. Sheridan, Trans. Farad. Soc., 48, 713 (1952).
13. G. C. Bond, J. Turkevich, Trans. Farad. Soc., 50, 1335 (1954).
14. J. E. Benson, T. Kwan, J. Phys. Chem., 60, 1601 (1956).
15. F. D. Rossini et al., Selected Values of Physical and Thermodynamic Properties of Hydrocarbons and Related Compounds, Pittsburgh, 1953.