NEWS

of the Tomsk Polytechnic Institute named after S. M. Kirov, holder of the Order of the October Revolution and the Order of the Red Banner of Labor

Volume 257

1973

RESEARCH IN THE CHEMISTRY OF CARBAZOLE DERIVATIVES

94. Determination of the Epoxide Number in 9-Propenoxides of Carbazoles

I. P. Zherebtsov, N. M. Rovkina, V. P. Lopatińskii, R. V. Kotlyarova

(Presented by the scientific-methodological seminar of the departments of organic chemistry)

In studying the synthesis and transformations of 1,2-epoxy-3-carbazolylpropanes (9-propenoxides of carbazoles), we required a procedure for determining the epoxide number in the named compounds, as well as in their mixtures with other carbazole derivatives. The present communication describes the development of such a procedure.

The extensive literature \[1–9\] shows that there is no universal method that would be equally well suited for the analysis of different epoxide compounds. This is connected with the diverse reactivity of the epoxide group and with the very strong influence of the structure of the epoxide compound on both the rate and the direction of the reactions occurring during analytical procedures. Nevertheless, the method most often recommended is one consisting in the addition of hydrogen chloride to the epoxide compound, followed by titration of the excess hydrogen chloride. It is indicated \[3, 9\] that the addition of hydrogen chloride is often not the only reaction, and the process may be accompanied by hydrolysis and isomerization of the epoxide group to an aldehyde. In addition, the stability of the chlorohydrin formed as a result of the reaction of the epoxide compound with hydrogen chloride is often insufficient to allow the acid to be accurately titrated with alkali in its presence, since, depending on the nature of the radical bonded to the chlorohydrin group, in weakly alkaline or even neutral media the chlorohydrin is capable of reversibly

be converted into an epoxide. All of the foregoing usually lowers the result of the analysis. Obtaining correct results, therefore, requires considerable experimental work to determine the most suitable analytical conditions for the given epoxy compound or for a group of compounds of similar structure.

In the analysis of carbazole 9-propenoxide by the action of hydrochloric acid in pyridine, followed by determination of the excess hydrochloric acid by titration, a satisfactory result was obtained [12]. A similar result (19.02%) was also obtained by us (from six determinations), and this value does not differ statistically from the calculated value (19.28%). However, for routine and production analyses such a method is of little use because of the use of toxic pyridine, which is inconvenient to handle. A wide variety of solvents are recommended: acetone, diethyl ether, etc. [1, 2, 4]. For reasons of availability and convenience of operation we chose acetone. Moreover, the use of acetone as a solvent is also recommended for certain technical determinations of epoxy numbers [6].

As the basis for developing the procedure we took the data recommended in [6], using phenolphthalein as the indicator for titration.

The initial experiments were carried out by the following procedure. Accurate weighed portions of carbazole 9-propenoxide were placed in two 150–200 ml flasks and dissolved in 75 ml of acetone. In parallel, a blank experiment was set up (without a weighed portion). To all three flasks, 20.0 ml of hydrochloric acid of known concentration was added, and the mixture was boiled under a reflux condenser for the specified time. After cooling to room temperature, the contents of the flasks were titrated with sodium hydroxide solution in the presence of phenolphthalein. The mean of the results of two titrations of the sample with the weighed portion was taken. The epoxy number was determined by the formula

\[ \text{E.N.} = c \frac{(V_1 - V_2)\cdot K \cdot 0.0043 \cdot 100}{g}, \]

where

\[ V_1 \]
— volume of alkali used to titrate the blank sample, ml;

\[ V_2 \]
— volume of alkali used to titrate the sample with the weighed portion, ml;

\[ K \]
— correction factor for the normality of the 0.1 N sodium hydroxide solution;

0.0043 — number of epoxy groups in grams corresponding to 1 ml of 0.1 N sodium hydroxide solution;

\[ g \]
— weighed portion, g.

Three series of experiments were set up, differing as follows:

the first series—0.5 N hydrochloric acid was added, titration was carried out with a 0.5 N sodium hydroxide solution, the sample weight was 0.200 g, and the boiling time was varied;

the second series—0.5 N hydrochloric acid was added; titration was carried out with a 0.1 N sodium hydroxide solution, the boiling time was 90 minutes, and the sample weight was varied;

the third series—the sample weight was 0.200 g, the boiling time 90 minutes, titration was carried out with a 0.1 N sodium hydroxide solution, 1.0 g of sodium chloride was added to the mixture, and the normality of the hydrochloric acid was varied.

The results of the experiments are presented in Table I.

Table I

Note:

\(\overline{X}\)—mean result,

\(n\)—number of determinations,

\(S^2\)—variance,

\(S_x\)—standard deviation of an individual result,

\(S_{\overline{x}}\)—standard deviation of the mean result,

\(\alpha\)—confidence probability, 0.95.

\[ \varepsilon_a \text{ is the precision of the determination,} \]

\[ \frac{\varepsilon_a \cdot 100}{a} \]
is the relative error of the mean result.

The same designations are used in Tables 2 and 3.

The statistical processing was carried out according to the recommendations of [10, 11].

From the data presented in Table 1 it follows that in none of the cases presented was a satisfactory result obtained, since the epoxide values are greatly underestimated.

Considering the underestimation of the results as a consequence of the possible occurrence of dehydrochlorination of chlorohydrin during titration to a strongly alkaline medium in the presence of phenolphthalein, it was to be assumed that, when indicators whose transition interval lies at lower pH values than that of phenolphthalein are used to fix the end point of titration, better analytical results may be obtained. In addition, as follows from the literature [2, 7] and from the results presented in Table 1, the addition of metal chlorides to the reaction mixture can significantly improve the result by shifting the equilibrium during chlorohydrin dehydrochlorination toward chlorohydrin through the creation of an increased concentration of chloride ion. It is pointed out that, to ensure the favorable effect of a metal-chloride additive, its good solubility is important [7]. Therefore crystalline calcium chloride, like sodium chloride, was added; it dissolves in the reaction mixture better. Thus, further experiments were carried out to determine the most suitable additive and the most suitable indicator for obtaining possibly more correct analytical results.

These experiments were carried out under the following conditions: hydrochloric acid concentration, 0.5 N; titration was carried out with 0.1 N sodium hydroxide solution; boiling time, 90 min; sample weight, 0.200 g; sodium chloride additive, 1.0 g (it is not advisable to add more, since the sample does not dissolve completely); calcium chloride additive, 5.0 g.

As can be seen from the results of the experiments presented in Table 2, the assumption concerning the positive influence of metal-chloride additives is justified and indicates a substantial influence of the reversibility of chlorohydrin dehydrochlorination on the analytical result. On the other hand, the type of indicator is important, which suggests the possibility of dehydrochlorination occurring during the titration process. Here it should—

It should be noted that, when titrating a strong acid with a strong alkali, the transition range of the indicator has no substantial effect on the titration result [13]. Indeed, in our case the results of titration in blank experiments, within the limits of titration error, do not depend on the indicator and additive used. Consequently, the differences in the titration results in experiments with a weighed sample are due entirely to processes associated with transformations of the compound being analyzed.

It also follows from Table 2 that the best result can be obtained when phenol red is used as the indicator, with additions of calcium chloride or sodium chloride. Under these conditions, the result obtained does not differ statistically significantly from the theoretical value at a confidence probability of 0.95.

To determine the applicability of the method developed for determining 9-propenoxide of carbazole in mixtures containing other carbazole derivatives, we studied the influence of various impurities on the analytical result. As impurities we tested: carbazole, 1,3-di-(9-carbazolyl)propanol-2, and 1-(9-carbazolyl)-2-chloropropanol-3. Artificial mixtures were prepared for analysis in such a way that each weighed sample contained 0.200 g of 9-propenoxide of carbazole and an equal amount by weight of impurity.

The results presented in Table 3 indicate that the presence of carbazole and 1-(9-carbazolyl)-2-chloropropanol-3 does not substantially affect the correctness and precision of the analysis. The addition of 1,3-di-(9-carbazolyl)propanol-2 causes a certain lowering of the result, owing to the appearance during titration of turbidity caused by the low solubility of 1,3-di-(9-carbazolyl)propanol-2 in the aqueous–acetone mixture. However, for an approximate judgment of the composition of mixtures of 9-propenoxide of carbazole with 1,3-di-(9-carbazolyl)propanol-2, the method tested can be successfully used.

The method developed was tested for determining epoxy numbers in 9-propenoxides of carbazoles substituted in the nucleus. Thus, for 9-propenoxide of 3-chlorocarbazole an average value of 16.72% was obtained (theoretical value 16.70), and for 9-propenoxide of 3,6-dichlorocarbazole, 14.30 (theoretical number, 14.70). The relative error in both cases (from six parallel determinations) does not exceed 2%.

Table 2

Indicators

Note: For all experiments $n = 6$.

Table 3

I — (9-carbazolyl)-2-chloropropanol-3 and calcium chloride;
II — the same, but without calcium chloride;
III — carbazole and calcium chloride;
IV — 1,3-(9′-carbazolyl)-propanol-2 and calcium chloride;
for all experiments $n = 6$.

Thus, the analysis conditions adopted in the series of experiments with phenol red as indicator and with the addition of sodium or calcium chloride should be regarded as satisfactory for obtaining correct and reproducible results for 9-propenoxides of carbazoles and their mixtures with certain carbazole derivatives.

Reagents and materials:

1,2-epoxy-3-(9′-carbazolyl)-propane. \(T_{\text{m.p.}} = 108–109^\circ\mathrm{C}\) (literature data [12]: \(T_{\text{m.p.}} = 108–109^\circ\mathrm{C}\)).

Found: C — 81.35%, H — 5.76%, \(N\) — 6.27%, \(C_{15}H_{13}ON\).
Calculated: C — 83.10%, H — 5.65%, \(N\) — 7.19%.

1-(9-carbazolyl)-2-chloropropanol-3. \(T_{\text{m.p.}} = 69–70^\circ\mathrm{C}\);

Found: C — 69.40%, H — 5.60%, \(N\) — 5.43%, \(C_{15}H_{14}ONCl\).
Calculated: C — 69.50%, H — 5.40%, \(N\) — 5.40%.

Acetone — “analytical grade.”

Sodium chloride — “chemically pure.”

Crystalline calcium chloride — “analytical grade.”

Hydrochloric acid — prepared by diluting a standard solution.

Sodium hydroxide solution — prepared by diluting a standard solution.

Indicators — reagent grade.

Conclusions

A procedure has been developed for determining the epoxy number in 9-propenoxides of carbazoles and their mixtures with certain carbazole derivatives.

References

1. Guben–Weyl. Methods of Organic Chemistry (Methods of Analysis). Moscow, Goskhimizdat, 1963.

2. A. M. Paken. Epoxy Compounds and Epoxy Resins. Leningrad, Gosnauchtekhizdat, 1962.

3. F. Critchfield. Analysis of the Principal Functional Groups in Organic Compounds. “Mir,” 1965.

4. Analytical Chemistry of Polymers. Vol. I, IL, 1963.

5. M. R. F. Ashworth. Titrimetric Methods of Analysis of Organic Compounds. “Khimiya,” 1968.

6. Collection of Technical Specifications for Paint and Varnish Materials. Vol. 2, “Khimiya,” 1971.

7. P. V. Zimakov. Ethylene Oxide. “Khimiya,” 1967.

8. M. F. Sorokin, K. A. Lyalushko. Laboratory Manual on the Chemistry and Technology of Film-Forming Substances. “Vysshaya shkola,” 1970.

1. A. A. Blagonravova, A. I. Nepomnyashchii. Epoxy lacquer resins. “Khimiya,” 1970.

2. L. P. Adamovich. Rational methods for compiling analytical procedures. Publishing House of Kharkov University, 1966.

3. K. Doerffel. Statistics in analytical chemistry. “Mir,” 1969.

4. I. M. Ezrieleev, N. A. Larin. ZhOKh, 26, 791, (1956).

5. A. K. Babko, I. V. Pyatnitskii. Quantitative analysis. Moscow, 1956.