RESEARCH IN THE CHEMISTRY OF CARBAZOLE DERIVATIVES

35. The Acidity of Carbazole and Its Substituted Derivatives

I. P. Zherebtsov, V. P. Lopatinskii

(Presented by the Scientific-Methodological Seminar of the Organic Departments of the Faculty of Chemical Technology)

The manifestation of the acidic properties of the imino group of carbazole has long been known and is used in practice, for example, for isolating carbazole from crude anthracene by the potassium-fusion method.

To carry out many reactions involving replacement of the hydrogen of the imino group of carbazole, it is necessary first to establish a protolytic equilibrium associated with ionization of the imino group as an acid. Practically important examples are the Stevens and Taker alkylation reactions [1], vinylation with acetylene [2], oxyethylation with ethylene oxide [3], etc. Knowledge of the acidity of carbazole is therefore important for choosing the conditions under which such reactions are carried out.

A quantitative characterization of the acidity of carbazole and its substituted derivatives may also serve as a measure of the influence of structural features of the molecule on the properties of the imino group.

In the literature available to us, however, we have not found values for the acid ionization constant of carbazole itself, although values of ionization constants are available for 3-nitrocarbazole [4] and 3,6-dinitrocarbazole [5]—rather acidic carbazole derivatives. In addition, there are a number of indirect estimates, including the study of hydrogen bonding by UV spectroscopy [6], the study of IR spectra [7, 8], and also Shatenshtein’s assessment of the activity of carbazole as an acid catalyst in the ammonolysis of santonin [9].

The last work makes it possible, from the known Brønsted relationship between the ionization constant and catalytic activity, to calculate the ionization constant of carbazole. Taking for ortho-nitrophenol, phenol, and ortho-nitroaniline the pKa values (in water) of 7.23 [10], 9.98 [10], and 17.88 [4], respectively, we obtain for carbazole a pKa value equal to 16.8.

It should be emphasized, however, that this value is based on the assumptions of Brønsted theory and may not reflect the actual situation.

The difficulty of directly determining the true ionization constant of carbazole lies in its small magnitude, which makes it necessary to use nonaqueous solvents, for which deviations from Brønsted theory are inevitable. A known alternative may be the use of an indicator method; in this case the nature of the indicator, which must be

structurally similar to the acid being studied, and the value of the ionization constant should be referred to water. The most suitable system of indicators in this sense for determining the strength of N–H acids, and of carbazoles in particular, may be the series of nitroanilines and nitrodiphenylamines studied by Stewart and co-workers [4].

In the present communication, the results are given of determining the acidity of carbazole and its substituted derivatives by the indicator method using Stewart indicators.

The results of the determinations are summarized in Table 1.

The results obtained indicate that carbazole is a rather weak acid, and the value of its pKa agrees with that calculated from the data of work [9]. Indole is an even weaker acid, i.e., the introduction of an additional phenylene group into indole with formation of carbazole causes a decrease in the pKa value of the amino group. On the other hand, attempts to ionize diphenylamine in the systems we used were unsuccessful, i.e., diphenylamine is a much weaker acid than indole. It is interesting that the pKa value of pyrrole is close to the corresponding value for carbazole, although somewhat smaller. Consequently, in order of increasing acidity, carbazole and its closest structural analogues form the series:

\[ \text{diphenylamine}<\text{indole}<\text{carbazole}<\text{pyrrole}. \]

In studying proton-donor ability in hydrogen-bond formation by UV spectroscopy, the following series can be compiled:

\[ \text{diphenylamine}<\text{pyrrole}<\text{2-methylindole}<\text{carbazole}\ [6]. \]

On the other hand, in work [11], when the degree of proton-donor ability of various compounds was studied by IR spectroscopy, the following series was compiled: pyrrole > indole > carbazole > diphenylamine. In addition, it was found that the capacity of the indicated substances for self-association increases in the reverse order. Approximate pKa values found by MacEwen also agree with the results of the latter work [12]. The position of the characteristic absorption band of the amino group may serve as a measure of its acidity. In work [7], the data of which agree with the results obtained by A. E. Lutskii and E. N. Goncharova [11], it was found that the maximum of the absorption band shifts toward lower frequencies in the series: pyrrole, indole, carbazole, diphenylamine.

Taking into account that the study of proton-donor activity by observation of the characteristic absorption band is more reliable than the use of UV spectroscopy [13], it should be considered that the observed acidity of the N–H group in the compounds studied agrees only approximately with the proton-donor ability in hydrogen-bond formation.

It follows from Table 1 that the pKa values of carbazole differ rather greatly when acetone and dimethylformamide are used as solvents. It should be noted that in acetone, under the action of alkali, the indicator used is ionized only to a small extent [less than 10%], which also causes deviations in the determinations. Therefore, the pKa value of carbazole determined in acetone with the aid of o-nitroaniline should be regarded only as an approximate value.

Regarding the influence of substituents in the carbazole nucleus on the acidity of the imino group, it may be noted that the acidity of carbazole and its substituted derivatives correlates satisfactorily with the Hammett equation using \(\bar{\sigma}\) constants (for substitution in the para position), pri-

Acidity of carbazole and its substituents

Table 1

Notes: 1) From [4]; 2) From [5]; 3) Obtained in processing the data of [9]; 4) From [14]; 5) From [12]; 6) Average of two values—15.65.

than for those disubstituted in the 3 and 6 positions, the influence of substituent constants is additive.

From these data the value of the reaction constant is equal to +2.42. Such a magnitude of the reaction constant indicates a rather high degree of conjugation of the reaction center with the substituent.

Experimental Part

Apparatus and Materials

Measurements were carried out on an FEK—N—57 electrophotocalorimeter, equipped with narrow-band filters that make it possible reliably to isolate the absorption bands of the indicator anion.

The indicators used were reagent-grade, “pure,” and were recrystallized from a suitable solvent until the literature melting points were reached. 2,6-Dichloro-4-nitroaniline was synthesized by chlorination of 4-nitroaniline and purified by successive recrystallization from acetone and alcohol to m.p. 195°C. Carbazole was purified by repeated recrystallizations of reagent-grade carbazole from acetone to m.p. 246—247°C. 3-Chlorocarbazole, 3,6-dichlorocarbazole, and 3,6-dibromocarbazole were synthesized according to [15] and have literature melting points. 3,6-Diacetylcarbazole was provided by M. M. Sukhoroslova, to whom the authors express their gratitude. It was obtained according to [18], m.p. 232°C. Ethyl alcohol was “rectified” (96%), caustic potassium, acetone—“for analysis,” dimethylformamide—“pure.”

Method for pKa Measurements

In the selected solvent, the extinction coefficient of the indicator anion, which is ionized under the action of alcoholic caustic potassium, is determined. The ionization constant of the acid under study is calculated by measuring the optical density of a solution containing known amounts of the acid, indicator, and caustic potassium. When acetone and dimethylformamide are used as solvents, the amount of alcoholic caustic potassium is kept constant, and the concentration of the acid under study is varied, since the composition of the solution affects the basicity at a constant alkali concentration. The ratio of the alcoholic solution of caustic potassium and the other solvent is chosen so as to provide the necessary basicity of the system and its homogeneity; this is determined by preliminary experiments. In other respects the procedure corresponds to that recommended in the literature [10].

As an example, a description is given below of the determination of the pKa of 3,6-diacetylcarbazole.

1. Determination of the molar extinction coefficient of the indicator. As the indicator, 2,4-dinitrodiphenylamine is used (pKa-13.84 [4]). Into a 25 ml volumetric flask are introduced 5 ml of an alcoholic solution of the indicator with a concentration of \(0.2 \cdot 10^{-3}\) mole per liter, variable amounts of an alcoholic solution of caustic potassium (0.5 N), and ethanol to the mark. The optical density is measured on filters 1—4 in a cuvette 1 cm thick. The comparison solution differs by the absence of indicator. The results of the measurements are presented in Table 2. We assume that the optical density measured on the first filter corresponds to the unionized molecule, and on the fourth—to the anion. After treating the obtained data by the method of least squares and obtaining an equation relating the optical density on the first filter (\(y\)) to the optical density on the fourth filter (\(x\)), we find that at \(y \to 0\) (the condition of complete ionization),

\(x \to 1.034\), i.e., the molar coefficient of the anion, \(E_i\), is found from the equation:

\[ E_i=\frac{1.034\cdot 10^4}{4}=25850. \]

1. Determination of pKa.

Into a 25 ml volumetric flask are added varying amounts of an alcoholic solution (concentration \(1\cdot 10^{-2}\) mole/l) of 3,6-diacetylcarbazole, 5 ml of a solution of 2,4-dinitrodiphenylamine (concentration \(0.2\cdot 10^{-3}\) mole/l), and 0.10 ml of 0.5 N alcoholic potassium hydroxide solution, and alcohol is added to the mark. In parallel, a comparison solution is prepared, differing by the absence of the indicator. Measurements are made in a cuvette of thickness 3 cm on 4 filters (the interchangeability of the cuvettes was checked). The calculations are performed according to the formula

Table 2

Determination of the molar extinction coefficient of 2,4-dinitrodiphenylamine

\[ \mathrm{p}K_a=\mathrm{p}K_{a\,\mathrm{ind}}+\lg\frac{[\mathrm{HA}]}{[\mathrm{A}^{-}]}-\lg\alpha, \tag{10} \]

where \(\mathrm{p}K_{a\,\mathrm{ind}}=13.84\)

\([\mathrm{HA}]\) and \([\mathrm{A}^{-}]\) are the stoichiometric concentrations of the nonionized and ionized forms of the acid, respectively.

\(\alpha\) is the ratio of the concentration of the nonionized form of the indicator to the concentration of the ionized form.

The results obtained from two series of measurements are presented in Table 3.

Table 3

Determination of the pKa of 3,6-diacetylcarbazole

Average result: pKa = 12.51

The accuracy of pKa determinations in other cases is also \(\pm 0.25\).

Conclusions

The ionization constants of indole, carbazole, and some of its derivatives as acids have been determined. A satisfactory correlation of the acidity of the carbazoles studied with the Hammett equation was found.

REFERENCES

1. T. S. Stevens, S. H. Tucker. J. Chem. Soc., 123, 2140, 1923.
2. K. Frisch. Monomers. Collection 2, IL, 1953.
3. V. P. Lopatinskii, E. E. Sirotkina, I. P. Zherebtsov. Izv. TPI, 112, 42, 1963.
4. R. Stewart, J. P. O. Donnel. Can. J. Chem., 42, 1681, 1964.
5. K. Bowden. Chem. Rev., 66, 119, 1966.
6. A. B. Sannigrahi, A. K. Chandra. J. Phys. Chem., 67, 1106, 1963.
7. N. Fuson, M. L. Josein, R. L. Powell, E. Utterback. J. Chem. Phys., 20, 145, 1952.
8. O. R. Wulf, U. Liddel. J. Am. Chem. Soc., 57, 1464, 1935.
9. A. I. Shatenshtein. Theory of Acids and Bases. Goskhimizdat, 1949.
10. A. Albert, E. Serjeant. Ionization Constants of Acids and Bases, “Khimiya,” 1964.
11. A. E. Lutskii, E. N. Goncharova. ZhFKh, 41, 538, 1967.
12. W. K. McEwen. J. Am. Chem. Soc., 58, 1124, 1936.
13. D. Pimentel, O. McClellan. The Hydrogen Bond, “Mir,” 1964.
14. Yu. A. Zhdanov, V. I. Minkin. Correlation Analysis in Organic Chemistry. Rostov State University Press, Rostov-on-Don, 1966.
15. V. P. Lopatinskii, I. P. Zherebtsov, E. E. Sirotkina, S. K. Vereshchagina. Izvestiya TPI, 136, 11, 1965.

16. V. P. Lopatinskii, M. M. Anosova, L. P. Konkina. Izvestiya TPI, 136, 15, 1965.

17. Order 5373