Chemistry

G. N. Nechiporenko, G. M. Tabrina, A. K. Shilova, A. E. Shilov

On the Mechanism of Nitrogen Fixation in the Reacting System

$(\mathrm{C}_5\mathrm{H}_5)_2\mathrm{TiCl}_2$—$\mathrm{C}_2\mathrm{H}_5\mathrm{MgBr}$

(Presented by Academician N. N. Semenov, March 22, 1965)

Recently M. E. Vol’pin and V. B. Shur ($^1,{}^2$) showed that in a number of systems formed by the interaction of organometallic compounds with salts of transition elements under mild conditions, nitrogen is fixed. Especially active is the system formed by the interaction of ethylmagnesium bromide with biscyclopentadienyltitanium dichloride in ether ($^3$). In this case, even at atmospheric nitrogen pressure, up to one nitrogen atom from $\mathrm{N}_2$ is absorbed per titanium atom. After hydrolysis of the reaction products, ammonia is evolved. It was proposed ($^1$) that in these systems $\mathrm{NH}_3$ is formed as a result of the addition of hydrogen from the molecules of the reagents or solvent to nitrogen, which forms a complex with a compound containing a transition-element atom. On the other hand, it is not excluded that the fixed nitrogen is liberated in the form of ammonia only upon hydrolysis and is contained in the reaction products, for example, as the nitride of one of the metals participating in the reaction*.

The present work was undertaken in order to clarify certain questions concerning the mechanism of the reaction of $(\mathrm{C}_5\mathrm{H}_5)_2\mathrm{TiCl}_2$ with $\mathrm{C}_2\mathrm{H}_5\mathrm{MgBr}$, and also to determine in what form the nitrogen bound in this system is present.

Examination of changes in the spectra of the solution during the interaction of the reagents shows that the reaction is complex in character. The spectrum of the rapidly formed homogeneous red solution remains almost unchanged for some time (about 1 min), then the color rapidly becomes lighter; subsequently the solution again darkens, and resinous products precipitate from it. After the reaction has proceeded for 2–3 hr at room temperature, ethane and ethylene were found in the gaseous products. When nitrogen is passed through during the same time, its fixation occurs, with no noticeable changes being observed in the composition or quantity of the gaseous reaction products (Table 1).

Table 1

Approximately one third of all gaseous reaction products (more than 1 mole per mole of Ti) is evolved during the first 5 min, when nitrogen fixation is still practically absent. The amount of bound nitrogen decreases sharply if its passage is begun several hours after mix—

* Recently the authors of work ($^2$) also do not exclude this possibility.

of the components. These results indicate reduction of titanocene prior to nitrogen fixation and the intermediate character of the particles interacting with \(N_2\).

When the reaction products are decomposed with water or sulfuric acid, hydrogen formation is observed. Mass-spectrometric analysis of the hydrogen formed upon the action of \(D_2O\) showed that its isotopic composition (mainly HD and small amounts of \(D_2\) and \(H_2\)) corresponds to hydrolysis of metal hydrides \((^4)\).

Table 2

\([C_2H_5MgBr] = 1.7\cdot 10^{-1}\) mole/l in all experiments.

Table 2 gives the results of several experiments in which the amount of hydrogen formed during hydrolysis of the final products was determined. It may be noted that in the presence of nitrogen, hydrogen is evolved in markedly smaller quantities, i.e., less metal hydride is formed.

The considerable amount of gases evolved during the reaction and the formation of hydrides indicate deep reduction of titanocene. The infrared spectra of the reaction products also indicate profound changes: the sandwich structure of the titanium derivative is destroyed.

When the reaction was carried out with \(C_2D_5MgBr\) as the reducing agent, it was found that the ethane contains mainly \(C_2D_6\). In heavy ether \((C_2D_5)_2O\), the ethane formed during reduction of \((C_5H_5)_2TiCl_2\) by light \(C_2H_5MgBr\) contains only about 3% \(C_2H_5D\). Additions of octadeuterotoluene (20%) likewise did not lead to any noticeable increase in the content of \(C_2H_5D\). Thus, it is evident that only an insignificant part of the ethane is formed at the expense of ethyl radicals abstracting H from the solvent. These data confirm a reduction mechanism through intramolecular disproportionation of ethyl groups:

\[ \cdots Me < \begin{matrix} C_2H_5 \\ C_3H_5 \end{matrix} \to \cdots Me + C_2H_4 + C_2H_6, \]

which is characteristic of reactions of this type \((^5)\). The diminished yield of ethylene compared with ethane is apparently connected mainly with further transformations of ethylene in the reacting system.

The hydrogen formed upon decomposition with light water of the reaction products of \((C_5H_5)_2TiCl_2\) with \(C_2D_5MgBr\) contained mainly HD, i.e., the metal hydrides are formed with participation of hydrogen from the alkyl group.

Some results obtained using \(C_2D_5MgBr\) as the reducing agent and \((C_2D_5)_2O\) as the solvent are relevant to the mechanism of nitrogen fixation.

If, during nitrogen fixation, nitrogen were reduced to ammonia at the expense of hydrogen atoms of ethylmagnesium bromide or of the solvent, then upon hydrolysis of the products with light acid the ammonium salt and the excess acid would have to contain a certain amount of D (taking into account dilution of deuterium by protium as a result of exchange). However, the experimentally determined content of exchangeable deuterium both in the case of \(C_2D_5MgBr\) in light ether and in the solution \((C_2D_5)_2O\) proved to be practically at the level of blank experiments carried out without passing nitrogen (Table 3).

In control experiments with addition of \(ND_3\) to the initial reacting mixture in an amount equivalent to the amount of bound nitrogen, an exchangeable D atom is detected in the reaction products. It may be noted that the exchange data indicate that, after the interaction of ammonia in the reacting system (probably with \(C_2H_5MgBr\)), one hydrogen atom remains on nitrogen.

Thus, it is evident that ammonia is not formed in the system either as a final or as an intermediate product.

Apparently, this indicates that the bound nitrogen is present in the reaction products in the form of nitrides, which form ammonia upon hydrolysis.

It is possible that the formation of nitrides occurs through the interaction of $\mathrm{N_2}$ with the metal (Mg or Ti) in statu nascendi. This is consistent with the deep reduction of the titanium compound observed in the system studied, and with the fact that metals become considerably more active toward nitrogen in a finely divided state.

Table 3

Experimental Part

The starting compound for obtaining $\mathrm{C_2D_5MgBr}$ and $(\mathrm{C_2D_5})_2\mathrm{O}$ was ethyl alcohol labeled with deuterium (90% in the alkyl group). The synthesis was carried out according to the schemes:

\[ \mathrm{C_2D_5OH} \ \xrightarrow[\mathrm{KBr}]{\mathrm{H_2SO_4}}\ \mathrm{C_2D_5Br} \ \xrightarrow[\text{ether}]{\mathrm{Mg}}\ \mathrm{C_2D_5MgBr}, \]

\[ \mathrm{C_2D_5Br}+\mathrm{C_2D_5ONa}\rightarrow \mathrm{NaBr}+(\mathrm{C_2D_5})_2\mathrm{O}; \]

$(\mathrm{C_5H_5})_2\mathrm{TiCl_2}$ was synthesized according to the procedure of (6).

The reduction of $(\mathrm{C_5H_5})_2\mathrm{TiCl_2}$ and nitrogen fixation were carried out in a vessel connected to a vacuum system, which made it possible, after the reaction, to collect and analyze the gaseous products. The amount of gases was determined from the pressure in a known volume. Hydrogen was transferred into a known volume by means of a Toepler pump.

Gas analysis was carried out chromatographically (ethane, ethylene) and mass-spectrometrically (labeled compounds). Isotopic analysis of hydrogen and ethane was carried out on MX-1302 and MV-2302 mass spectrometers. To determine the N—D bond in the reaction products, the mixture was hydrolyzed with a small amount of sulfuric acid ($\mathrm{H^+}/\mathrm{Ti}=14.5$); then, after prolonged standing, a solution of $\mathrm{CH_3MgJ}$ or $\mathrm{LiAlH_4}$ was added, and isotopic analysis of the evolved methane or hydrogen was performed. Experiments with the special addition of $\mathrm{ND_3}$ were carried out under identical conditions.

The authors express their gratitude to Yu. G. Borod’ko for carrying out the IR spectral analysis of the reaction products.

Institute of Chemical Physics
Academy of Sciences of the USSR

Received
13 February 1965

Cited Literature

1. M. E. Vol’pin, V. B. Shur, DAN, 156, 1102 (1964).
2. M. E. Vol’pin, V. B. Shur, Vestn. AN SSSR, No. 1, 51 (1965).
3. M. E. Vol’pin, V. B. Shur, M. A. Ilatovskaya, Izv. AN SSSR, ser. khim., 1964, 9.
4. H. Gaylord, Reduction with Complex Metal Hydrides, Moscow, 1959, p. 33.
5. A. K. Zeifirova, A. E. Shilov, DAN, 136, 599, 1961.
6. J. Wilkinson, J. M. Birmingham, J. Am. Chem. Soc., 76, 4281 (1954).