B. V. Tronov and L. P. Kulev.

Complexes of Nitro Compounds of the Benzene Series with Acids

Complex compounds have long been known\(^{1}\) that are formed by picric acid and other similar nitro derivatives with many organic substances, even with certain hydrocarbons (naphthalene, anthracene, etc.). In the formation of these complexes of picric acid, the oxygen of the nitro group takes part. The more nitro groups there are, the more readily, in general, the given compound enters into reaction. Thus, tetranitromethane reacts not only with those comparatively complex compounds that can give picrates, but also with certain simpler substances of the aliphatic series, and has been proposed as a reagent for the double bond between carbon atoms.

In our laboratory a new class of ternary complex compounds has been found and studied,\(^{2}\) which cannot be prepared with hydrocarbons, but which can be obtained with very many substances containing a carbonyl group, a double or triple bond of carbon with nitrogen, etc. One component here is, as before, picric acid or another polynitro compound (it may also be a quinone, since picrates are constructed similarly to quinhydrones), the second component is an aldehyde, ketone (from the simplest upward), ester, nitrile, heterocyclic amine of the pyridine type, etc., and the third is certain salts, among which sodium iodide proved especially suitable. The ternary compound of trinitrobenzene with acetone and sodium iodide has, for example, the composition

\[ \mathrm{C_6H_3(NO_2)_3 + NaI + 3(CH_3)_2CO.} \]

Such complexes (whose formation is apparent from the red, orange, or yellow coloration) are readily obtained with trinitro compounds of the benzene series. With two nitro groups the coloration is weaker and the complex compounds are very unstable (they are readily decomposed, for example, by a small amount of alcohol); finally, nitrobenzene under these conditions gives nothing.

For nitrobenzene a complex compound with sulfuric acid of composition

\[ \mathrm{C_6H_5NO_2H_2SO_4} \]

is known.\(^{3}\) It was obtained in the form of crystals with a definite melting point.

Determination of the rate of nitration of various compounds of the benzene series showed\(^{4}\) that nitrobenzene reacts with nitric acid thousands of times more slowly than benzene itself. It is oxidized (by acetic-chromic anhydride) only twice as slowly.\(^{5}\) Such a sharp difference suggested

\(^{1}\) For reviews of them, see A. Werner, Ber. 42, 4324–28. 1909. P. Pfeiffer, Lieb. Ann. 404, 1–20. 1914. P. Pfeiffer, “Organische Molekülverbindungen” (Stuttgart, 1922).

\(^{2}\) B. V. Tronov, A. N. D’yakonova-Shul’ts, E. A. Zonova, ZhRKhO 59, 333–345. 1927.

\(^{3}\) V. Chembuliez, Chem. Zbl. 1923. I/II. 1491.

\(^{4}\) B. V. Tronov and I. Ya. Ber, ZhRKhO 62, 2335–2345. 1930.

\(^{5}\) B. V. Tronov and A. A. Grigor’eva, ZhRKhO 60, 1014, 1928.

suggests that the special difficulty of nitrating \(C_6H_5NO_2\) is due not only to a decrease in the mobility of the hydrogen atoms, but also to the attachment of the acid to the nitro group. Having become complexly attracted to the oxygen of the \(NO_2\) group, nitric acid can no longer (or, more accurately, almost no longer, since the complex nevertheless cannot be very stable) act on the benzene ring. To test this supposition, experiments were carried out\(^{1}\) on the action of various amounts of \(HNO_3\) on nitrobenzene. The result proved quite characteristic.

Table No. 1

Ratio of the rates of nitration of nitrobenzene.

The figures in this table can be explained only by the fact that \(HNO_3\) forms two complex compounds with \(C_6H_5NO_2\): \(C_6H_5NO_2 + HNO_3\) and \(C_6H_5NO_2 + 2HNO_3\), the first of which is evidently the stronger. Therefore, within the limits of the first gram-molecule of nitric acid the rate of nitration increases comparatively little (according to the law of mass action). Addition of the second molecule leads to a sharp jump; at the third the jump is somewhat smaller, and only here does the acid already begin to act directly on the benzene ring, since thereafter the increase in the rate again becomes normal.

In the present work we set ourselves the aim of determining whether nitro derivatives of benzenoid hydrocarbons form complex compounds with carboxylic acids. To detect the complexes formed, we used the method of determining the activity of hydrogen (in the present case, the hydrogen of the carboxyl group) from the electromotive force of the reaction with sodium. In our previous works\(^{2}\) this method gave good results when applied to individual compounds with mobile hydrogen (chiefly alcohols were studied) in the pure state or in a solvent not forming complexes here (benzene); subsequently systems of two active components were taken: carboxylic acid (sometimes water, alcohol, phenol) + amine. Many of the systems studied gave a sharp increase in the electromotive force in comparison with that obtained for the separate components. The presence of such a maximum evidently indicates the formation of a complex in which the hydrogen becomes considerably more active.

We have now studied from this standpoint nitrobenzene, meta-dinitrobenzene, and \(\alpha\)-nitronaphthalene, which take the place of the amine in the complex, and the acids: acetic, propionic, butyric, isobutyric, and isovaleric. The experiments were carried out under the same conditions as before, i.e., in the same apparatus, without solvent or with the addition of benzene. The temperature at the beginning of the experiments was \(16^\circ\)—\(20^\circ\), but sometimes the reaction was accompanied by noticeable self-heating, at times even reaching ignition of the sodium, especially in the absence of solvent. Therefore the results obtained in benzene solution must be considered more reliable. The most comparable results were obtained at a concentration of 0.025 g-mol. of the sum of the acid and the nitro compound, i.e. 0.025 g-mol. per 6 cu. cm of acid solution, or less acid, but with a corresponding addition of the nitro compound. Artificial cool—

\(^{1}\) B. V. Tronov and A. P. Grigor’eva, Proceedings of the Siberian Physico-Technical Institute, issue 1, 1932.

\(^{2}\) B. V. Tronov and L. P. Kulev, “Determination of the Activity of Hydrogen in Organic Compounds from the Electromotive Force of the Reaction with Sodium.”

B. V. Tronov and L. P. Kulev, “Comparison of the Activity of Alcohols and Carboxylic Acids under Various Conditions.”

we have not yet applied. With acetic acid, which (anhydrous) gave a particularly small EMF, experiments were also carried out with doubled concentrations.

For the work, the purest possible Kahlbaum preparations were taken, which were further purified by distillation. Special attention, as it turned out, had to be paid to the absence of water, since even a small admixture of it sometimes increases the electromotive force by tens of times. For some substances (e.g., isovaleric acid), separate measurements were made with different fractions obtained during distillation.

The experiments gave the following results:

Table No. 1.

Acetic acid with nitrobenzene.

That is, the maximum EMF corresponds to 60% acetic acid; then a decrease is observed, and at 80% a second maximum arises.

With α-nitronaphthalene and metadinitrobenzene such oscillations were not observed, and the magnitude of the EMF approached that given by pure \(C_2H_4O_2\). The acid for this experiment was prepared by a single freezing-out of commercial glacial acetic acid.

Table No. 2.

Acetic acid with α-nitronaphthalene.

Here the maximum shifts toward 70% acid. The same is also observed with metadinitrobenzene (dilution to 6 cc with benzene):

Table No. 3.

Note. Owing to the poor solubility of \(\alpha\text{–}C_{10}H_7NO_2\) and \(m\text{–}C_6H_4(NO_2)_2\), it was necessary to limit ourselves to a small number of experiments for these compounds.

For comparison, experiments were also carried out with acetic acid alone in benzene solution:

Table No. 4.

From the tables presented it is evident that the EMF increases in parallel with the change in the concentration of the acid in the solution. At the same time, the maximum corresponding to 100% \(CH_3COOH\) is considerably smaller than in the preceding case (Table No. 1). On detailed investigation, the reason for this phenomenon became clear. It turned out that an insignificant amount of water in acetic acid changes the magnitude of the EMF very substantially; thus experiments with especially carefully prepared anhydrous acid gave the following results:

Table No. 5.

Experiments with doubled concentrations.

Table No. 6.

The addition of water to absolute acetic acid changes the EMF in the following order:

Propionic acid with the same nitro compounds gives an analogous picture; thus, for example, with \(C_6H_5NO_2\) the maximum EMF also corresponds to 60% acid:

Moreover, with nitrobenzene \(C_6H_5O_2\) gives a somewhat greater EMF than acetic acid under the same conditions.

At other ratios of acid and \(C_6H_5NO_2\), a vigorous reaction with strong heating was observed, especially at the points corresponding to the maximum EMF:

The addition of a small amount of benzene noticeably slows the reaction and lowers the EMF:

With \(\alpha\)-nitronaphthalene, at the ratios of acid and nitrobenzene indicated in Table No. 8, the following results are obtained:

That is, the maximum EMF, as for \(CH_3COOH\), is observed at 70%.

With metadinitrobenzene, propionic acid gives analogous results, but the magnitude of the EMF is in this case considerably smaller than with acetic acid:

With normal butyric acid in a mixture with nitrobenzene, the same regularity is observed as for \(C_2H_4O_2\) and \(C_3H_5O_2\), i.e. the maximum EMF corresponds to 60% of this acid in the solution:

At high acid concentrations, along with an increase in EMF, strong heating is also observed, as with propionic acid:

With \(\alpha\text{-}C_{10}H_7NO_2\), the maximum EMF is attained at the same concentration as for the other acids:

Significantly more active under the same conditions is isovaleric acid:

Experiments with nitrobenzene

The most active in complex with nitrobenzene proved to be isovaleric acid, the maximum EMF corresponding to 50% acid. Slight heating is observed even at minimal acid concentrations in benzene solution.

Experiments with \(C_6H_5NO_2\) (to 6 cc with benzene):

Experiments with separate fractions of isovaleric acid

That is, both fractions gave almost the same electromotive force. In these experiments, as in both cases, the benzene solution became turbid (the mixture of the acid with nitrobenzene was diluted to 6 cc with benzene). With \(\alpha\)-\(C_{10}H_7NO_2\) and \(m\)-\(C_6H_4(NO_2)_2\) slight heating was also observed. The EMF of isovaleric acid with metadinitrobenzene is less than for all the other acids we studied.

Experiments with $\alpha$-$\mathrm{C}_{10}\mathrm{H}_7\mathrm{NO}_2$ (up to 6 cc with benzene).

For comparison, experiments were also carried out in nitrobenzene solution for acetic and propionic acids:

The remaining acids in nitrobenzene solution react very vigorously, with strong heating, and cannot be compared under these conditions. In addition, in order to determine the dependence of complex formation of acids with nitrobenzene on time, experiments were carried out under the following conditions:

Mixtures of acids with nitrobenzene and benzene were placed in sealed test tubes for one day (at room temperature). After this time had elapsed, the test tubes were opened and the test was carried out in the usual manner. The results are presented in the following table.

That is, no noticeable change in the magnitude of the EMF occurs.

Conclusions.

The following table gives the maximum values of the electromotive force in experiments in benzene solution at the same total concentration of the nitro compound and acid (0.025 gram-mol. per 6 cubic cm).

Experiments with nitrobenzene

Experiments with metadinitrobenzene

Experiments with $\alpha$-nitronaphthalene

From the numbers in this table, and also from the results of all the other experiments, the following conclusions may be drawn:

1. All the carboxylic acids we have taken give, with nitro compounds of the benzene series, complexes that are considerably more active with respect to metallic sodium than the acids themselves.

2. The composition of the complexes, so far as can be seen at present, ranges from 1 mole of acid to 2 molecules per 1 mole of nitro derivative. The irrational (at least for the most part) character of the curves (according to N. S. Kurnakov’s classification) indicates the low stability of the complexes.

3. Of the nitro compounds, nitrobenzene generally gives the most active complexes. Nitronaphthalene is close to it, while dinitrobenzene deviates noticeably, both in the activity of its complexes and in its relation to the individual acids.

4. Acids with a branched chain of carbon atoms give, in complexes with nitrobenzene and nitronaphthalene, a much greater electromotive force than acids of normal structure. The difference in the magnitude of the electromotive force of the reaction of nitrobenzene complexes with sodium seems to exceed the difference in the magnitudes of all the other constants of these acids. It may be hoped that this feature can be used to determine the structure of acids.

5. Dilution of the mixtures with benzene decreases the electromotive force appreciably more rapidly than would follow from a simple decrease in concentration.

6. The addition of water, already in an amount of 1%–3%, increases the electromotive force tens of times in the case of acetic acid and has almost no effect in the case of isovalerianic acid. We think that this method can be applied for determining small amounts of water in acids (acetic, and perhaps some others).

A theoretical explanation of the regularities found, connected with elucidating the structure of the complexes studied, naturally requires additional investigation. Something can be inferred from the following facts:

a) The complex of nitrobenzene with isovaleric acid, which gives the greatest electromotive force, at the same time has the most definite composition, \( \mathrm{C_6H_5NO_2} + 1\,\mathrm{C_5H_{10}O_2} \). Complexes with a small electromotive force approach a component ratio of \(1:2\).

b) With the addition of water (which is apparently a catalyst), the electromotive force of the complex with acetic acid becomes close to that given by isovaleric acid.

Comparison of the indicated facts suggests that complexes with a small electromotive force contain dimeric acid molecules, which attach to the nitro group without dissociating, for example:

\[ \mathrm{C_6H_5NO_2} + (\mathrm{C_2H_4O_2})_2 . \]

These dimeric molecules are apparently more readily formed by acids of normal structure. In branched chains steric hindrance may have an effect. As everything known about carboxylic acids leads one to think (for example, the fact that in pure form and in benzene solution they conduct almost no current, whereas with water they become rather good electrolytes), in dimers (and, perhaps, in the complexes containing them) the activity of hydrogen is greatly reduced. Water is probably a catalyst that splits the dimer into separate molecules, and these, possibly (if the composition of the complex compound still proves to be \( \mathrm{C_6H_5NO_2} + 2\mathrm{C_2H_4O_2} \)), attach to both oxygens of the nitro group. In the case of isobutyric and isovaleric acids, this again may be hindered by steric obstacles.