PROCEEDINGS
OF THE TOMSK POLYTECHNIC INSTITUTE NAMED AFTER S. M. KIROV,
ORDER OF THE OCTOBER REVOLUTION AND ORDER OF THE RED BANNER OF LABOR

Volume 257          1973

RESEARCH IN THE CHEMISTRY OF CARBAZOLE DERIVATIVES

96. Study of the Mechanism of Interaction of Olefin Oxides with Carbazoles in an Alkaline Medium

I. Kinetics of the Interaction of Propylene Oxide with 3-Nitrocarbazole

I. P. Zherebtsov, V. P. Lopatinskii, S. P. Yudina, Z. A. Michurina

(Submitted by the Scientific-Methodological Seminar of the Faculty of Chemical Technology)

Reactions of substances possessing a labile hydrogen atom (amines, amides, alcohols) with olefin oxides are widely known [1–5]. The authors of many studies indicate that the solvent plays a definite role in these processes, promoting, for example, the opening of the epoxide ring, etc.

In reactions of olefin oxides with substances containing a labile hydrogen atom, under conditions of basic catalysis, the formation of the reagent anion plays an essential role.

In studying the reaction of 3-nitrocarbazole (3-NC) with propylene oxide under conditions of basic catalysis, we attempted to assess the influence of the medium both on the ionization of 3-NC and on the stage of interaction of 3-NC with propylene oxide. The rate-limiting stage of the interaction of 3-NC with propylene oxide, probably like many others [1–5], proceeds through a trimolecular complex consisting of propylene oxide, the 3-NC anion, and the solvent (alcohol):

\[ \text{[structural scheme: carbazole }N^{-}\cdots CH_{2}\!-\!CH\!-\!CH_{3},\quad O\cdots HOR\text{]} \]

However, in studying the rate-limiting stage it was necessary to take into account that the rate of the process may be influenced both by the ionization of 3-NC and by the rate-limiting stage. Although the equilibrium

ionization is established rapidly; the degree of ionization, and consequently the concentration of the active species affecting the magnitude of the process rate, changes with a change in the basicity of the medium. Thus, it becomes evident that, with incomplete ionization of 3-nitrocarbazole, the rate constant calculated using the concentration of 3-nitrocarbazole rather than its anion is an apparent constant \((K_{\mathrm{ef}})\). To estimate the rate of the limiting medium, we used the true rate constant \((K_{\mathrm{true}})\), calculated using the concentration of the 3-nitrocarbazole anion.

The reaction under study in an excess of alcohol may be represented as base-catalyzed, bimolecular, and first order with respect to each of the reactants.

Since one of the tasks was to study the effect of the medium both on the entire process as a whole and on its individual stages from the standpoint of acid–base interaction, the medium basicity function \(H_{-}\), reflecting the tendency of the medium to remove a proton, was chosen to characterize the medium [6–8].

Fig. 1. Dependence \(\lg K_{\mathrm{ef}}\) on \(H_{-}\). \(C_{\mathrm{KOH}} = \mathrm{const}\).

1 — isopropanol,
2 — n-propanol,
3 — n-butanol,
4 — methyl cellosolve,
5 — ethanol,
6 — methanol,
7 — ethylene glycol.

As can be seen from Fig. 1, the dependence of the logarithm of \(K_{\mathrm{ef}}\) on \(H_{-}\) is expressed by a straight line. The effective constant increases with the protophilic activity of the solvent, which is associated with an increase in the con-

concentration of the anion [9]. Thus it becomes evident that the solvent also affects the slow stage, participating as a third component in the transition complex. Since it is not possible to trace this influence from the dependence of \(\lg K_{\mathrm{eff}}\) on \(H_-\), it was of interest to consider the dependence of \(K_{\mathrm{ist}}\), which characterizes the limiting stage, on \(H_-\).

Fig. 2. Dependence of \(\lg K_{\mathrm{ist}}\) on \(H_-\)

1 — isopropanol (1.5),
2 — n-propyl (0.84),
3 — n-butyl (0.75),
4 — monomethyl ether of ethylene glycol (0.5),
5 — ethanol (0.803),
6 — methanol,
7 — ethylene glycol (0.32).

\(H_- - C_{\mathrm{KOH}}=\mathrm{const}\)

Considering the dependence of \(K_{\mathrm{ist}}\) on \(H_-\) (Fig. 2), one can assess the influence of the solvent on the slow stage. As is seen from Fig. 2, alcohols of different structures affect the slowest stage in different ways. This stage in the behavior of alcohols differing in structure can apparently be explained by different degrees of specific solvation of both the anion and propylene oxide.

More acidic alcohols increase the solvation of propylene oxide, weakening the carbon–oxygen bond in the oxide molecule and thereby facilitating attack by the anion. On the other hand, the solvation of the anion is also increased, which reduces its nucleophilic strength and thereby hinders attack on the carbon atom of the oxide.

In studying the influence of mixtures of aprotic solvents—acetone, dimethylformamide (DMFA), dimethyl sulfoxide (DMSO)—with methanol, it was found that \(K_{\mathrm{eff}}\) increases with increasing mole percent of the aprotic solvent (Fig. 3), i.e., with an increase in the tendency of the medium to abstract a proton, which is due to an increase in the degree of ionization of 3-NK with increasing basicity of the medium and, as a consequence, an increase in the concentration of the carbazole anion [9].

Fig. 3.

It was of interest to trace the influence of the mole percent of aprotic solvent on \(K_{\mathrm{ist}}\). As is evident from Fig. 3, \(K_{\mathrm{ist}}\) decreases as the content of aprotic solvent increases. This is evidently associated with a decrease in the concentration of propylene oxide solvated by the proton-donor solvent. At the same time, the steric requirements in the transition state decrease, as evidenced by changes in both \(\Delta S^{\ne}\) and \(\lg K_{0}\) (Table I).

In addition, the increase in activation energy with increasing mole percent of aprotic solvent supports this assumption, since a decrease in the degree of solvation of propylene oxide in the transition state causes a decrease in its activity and an increase in the energy barrier. On the other hand, the concentration of the unsolvated anion increases.

Thus, increasing the concentration of the solvent that participates as the proton-donor component in the transition complex facilitates opening of the epoxide ring, while, however, hindering ionization of 3-nk, which is reflected in a decrease in \(K_{\mathrm{eff}}\). Apparently, the best medium for carrying out the reaction should be considered a mixture

of protophilic and protogenic solvents, which have opposite effects both on the solvation of the oxide and the anion and on the ionization of 3-NK under the action of alkali.

Experimental Part

A reaction mixture with a volume of 25 ml, consisting of 3-NK (0.02 mole/liter), a calculated amount of potassium hydroxide necessary to create a definite basicity of the medium (0.0012–15.8200 mole/liter), and solvent, was placed in a reaction vessel closed with a rubber stopper through which a hollow needle was passed; this made it possible, by means of a syringe, to introduce propylene oxide into the reaction and to take samples for analysis. The reaction mixture was then thermostated at a temperature of \(50^\circ \pm 0.1^\circ\)C for 15 min, after which propylene oxide (0.55 mole/liter) was introduced. Specially designed experiments showed that forced stirring was not necessary.

The course of the reaction was monitored by the change in the concentration of 3-NK. A 1-ml sample was placed in a 50-ml volumetric flask, diluted to the mark with acetone, and thoroughly mixed. The concentration of unreacted 3-NK was determined by the method described in [10].

Variations (up to 10%) in the concentration of propylene oxide had no noticeable effect on the rate, since propylene oxide was introduced in large excess.

The reaction rate constant \((K_{\mathrm{eff}})\) was found graphically from the tangent of the slope angle of the straight line \(\lg C_{\mathrm{n}}\)—time, where \(C_{\mathrm{n}}\) is the concentration of unreacted 3-NK.

The true constant was calculated by the formula

\[ K_{\mathrm{true}}=\frac{K_{\mathrm{eff}}}{I}, \]

where \(I\) is the ionization ratio, equal to the ratio of the anion concentration \(/A^{-}/\) to the concentration of the nonionized molecule \(/HA/\) [9].

By special experiments using GLC we did not find any interaction of propylene oxide with alcohols under the conditions of our experiment that would make it necessary to take it into account when calculating the rate constants.

LITERATURE

1. N. N. Lebedev, V. F. Shvets. Kinetics and Catalysis, 5, 989, 1964.
2. N. N. Lebedev, V. F. Shvets. Kinetics and Catalysis, 6, 782, 1965.
3. N. N. Lebedev, V. F. Shvets. Kinetics and Catalysis, 9, 504, 1968.

Table 1

Kinetic and thermodynamic characteristics of the process

1. F. Patat, B. Wojtech. Makromol. Chem., 37, 1, 1960.

2. J. Jtakura, F. Patat. Makromol. Chem., B. 68, 158, 1963.

3. K. Bowden. Chem. Rev., 66, 119, 1966.

4. C. H. Rochester. Quart. Revs., London, Chem. Soc., 20, No. 4, 511, 1966.

5. G. Yagil, M. Anbar. J. Am. Chem. Soc., 85, 2376, 1963.

6. I. P. Zherebtsov, V. P. Lopatinskii, S. P. Yudina, T. V. Melnikova. Izv. TPI, present collection, 85, 1973.

7. V. P. Lopatinskii, I. P. Zherebtsov, S. P. Yudina. Izv. TPI, Vol. 250.