On the Interaction of Aliphatic, Aromatic, and Heterocyclic Amines with Carboxylic Acids

B. V. Tronov and L. P. Kulev

Amines with acids should give salts. It is known, however, that the basic properties of amines are greatly affected by the radicals entering into them. Thus, in aniline the basic properties are weaker than in ammonia; diphenylamine still forms salts with strong acids, for example with HCl, but these salts are immediately hydrolyzed in water; triphenylamine, which contains in its molecule three electronegative radicals, is already incapable of salt formation. If one takes weak acids, for example monocarboxylic acids of the fatty series, it might be thought that with them appreciable salt formation would be observed by no means for all amines.

The basic properties of amines are usually determined from the constant of electrolytic dissociation of the hydroxides formed by them with water:

\[ \mathrm{R_3N + HOH \rightleftharpoons R_3N{-}H \rightleftharpoons [R_3NH]^+ + [OH]^- .} \]

\[ \phantom{\mathrm{R_3N + HOH \rightleftharpoons R_3N{-}H}} \atop \mathrm{\phantom{R_3N{-}}OH} \]

We decided to use for this purpose a method developed by us—the determination of the electromotive force of the reaction with metals, in particular with sodium. Our apparatus, which included a voltmeter based on the principle of a galvanometer, gives, strictly speaking, not directly the magnitude of the electromotive force, but only the current strength passing through the instrument. This current strength depends both on the electromotive force and on the resistance. If the resistance of the system under study is small in comparison with the resistance of the voltmeter,^1 then the readings shown by the instrument are, in fact, close to the magnitudes of the electromotive force of the reaction; but if the reacting system has a resistance considerably exceeding that of the voltmeter, then we must obtain a number that will express rather the product of the emf by the electrical conductivity.

However, for most of the organic compounds and their mixtures studied by us, the indicated error in the measuring apparatus cannot be of fundamental significance. For example, in compounds with active hydrogen, such as carboxylic acids, alcohols, and some others, the electromotive force of the reaction with metals depends on the mobility of the hydrogen. This same hydrogen is split off during electrolytic dissociation, and the more (at the same concentration) the substance is dissociated, the greater its electrical conductivity. All conditions that promote an increase in the activity of hydrogen (e.g., the addition of certain other substances that form complexes with the compound under study) lead to an increase in both the emf and the electrical conductivity, i.e., they increase their product still more. Thus the numbers obtained by our method, although they are not—

^1 Our millivoltmeter has a resistance of 3700 ohms.

...with exact values of the EMF of the reaction of organic compounds with metals, in the sense of a chemical characterization of these compounds, in particular in the sense of expressing the activity of their hydrogen, may prove no less, and even more, suitable than the results of determinations of the true EMF.

The same applies to the reaction of amines with acids. The salt formed dissociates into ions to a greater degree than the free base and the free acid, which increases the electrical conductivity of the solution. On the other hand, the cation of this salt should more readily remove an electron from a metal atom, in particular sodium, i.e., a larger EMF should arise in the reaction. Hence our measurement method should be a good means (not worse, and rather better, than determination of electrical conductivity) for judging the capacity of amines and acids for salt formation.

To determine the electromotive force of the reaction with sodium, we took the following amines:

Triethylamine $(\mathrm{C_2H_5})_3\mathrm{N}$
Aniline $\mathrm{C_6H_5NH_2}$
Dimethylaniline $\mathrm{C_6H_5N(CH_3)_2}$

Pyridine

\[ \begin{array}{c} \text{pyridine ring structure shown} \end{array} \]

Of the acids, the following were taken:
Propionic $\mathrm{CH_3CH_2COOH}$
Butyric $\mathrm{CH_3CH_2CH_2COOH}$
Isobutyric $(\mathrm{CH_3})_2\mathrm{CHCOOH}$
Isovaleric $(\mathrm{CH_3})_2\mathrm{CHCH_2COOH}$.

The reacting substances were mixed either without solvent or in benzene solution. In experiments with a single substance we took $1/40$ gram-molecule; in experiments with mixtures of amine and acid, the sum of the amounts of the one and the other was $1/40$ gram-mole, for example, $9/400$ gram-mole of amine + $1/400$ gram-mole of acid, $8/400 + 2/400$, etc.; benzene was added to the given amounts of reacting components up to $6$ cu. cm. In experiments without benzene the total volume of the mixture was likewise $6$ cu. cm.

The substances were taken, as far as possible, of the highest purity, carefully dried and distilled. The temperature at the beginning of the experiments was $18$–$20^\circ$; in the course of the reaction with sodium, self-heating sometimes occurred, especially in experiments without solvent. The measurement procedure was the same as in the preceding works on EMF.

The results of the experiments are given in Tables 1 and 2.

Table 1.

Experiments without solvent.

Table 2.

Experiments with benzene solutions.

CONCLUSIONS

1. As was to be expected, the magnitude of the electromotive force of the reaction with sodium, shown by our apparatus for the salt of an amine with a carboxylic acid—where there is no reason to doubt the formation of a salt—is noticeably greater than for the free acid. Free amines either do not react with sodium at all or, in the case of pyridine, give only a negligible emf value.

Thus, our method should be regarded as actually suitable for determining the ability of amines and acids to form salts.

1. A clear parallelism is observed between the electrolytic dissociation constant of the hydroxide formed by the amine in aqueous solution and the magnitude of the emf observed by us. Triethylamine, which has the most basic character (dissociation constant in aqueous solution equal to $6.4 \cdot 10^{-4}$ at $25^\circ$), gave a large increase in emf for all acids. Pyridine ($K = 2.3 \cdot 10^{-9}$ at $25^\circ$) gave, with all the acids taken, a noticeable increase in emf, but already a smaller one. Aniline ($K = 4.6 \cdot 10^{-10}$ at $25^\circ$) and dimethylaniline ($K = 2.42 \cdot 10^{-10}$ at $18^\circ$) nevertheless give a noticeable—although much smaller than with the preceding amines—increase in emf with isovaleric acid; but with propionic and butyric acids there is observed not a maximum, but only a gradual increase in emf from amine to acid. It is possible that there is no strict additivity here (our apparatus is not sufficiently accurate to decide this question), but in any case, if a salt is formed, it is only in a negligible amount. It is possible, however, that the reaction between the amine and the acid proceeds in these cases as well, but that instead of salts some complexes are obtained, which react only weakly with sodium.

2. The acids investigated by us¹—propionic ($K = 1.3 \cdot 10^{-5}$ at $25^\circ$), butyric ($K = 1.48 \cdot 10^{-5}$ at $25^\circ$), isobutyric ($K = 1.5 \cdot 10^{-5}$ at $25^\circ$), and isovaleric ($K = 1.7 \cdot 10^{-5}$ at $25^\circ$)—differ little from one another in the electrolytic dissociation constant in aqueous solution. In our experiments, a sharp difference was observed between propionic and butyric acids on the one hand, and isovaleric (and, apparently, isobutyric, insofar as one can judge from the single determination) on the other. The first two

acids show a very considerable increase in EMF in compounds with triethylamine, a small increase with pyridine, and none with aniline and dimethylamine. Isovaleric acid likewise gave the greatest EMF value in experiments with triethylamine, a somewhat smaller one (but much greater than propionic and butyric acids) with pyridine, and a small, though quite noticeable, maximum with aniline and dimethylaniline. In this respect, carboxylic acids with amines behave in the same way as with nitro compounds: acids with a branched chain prove to be much more active than acids of normal structure.

1. The composition of the salts obtained still remains undetermined. The results of our experiments do not yet make it possible to say whether the maximum of the observed EMF corresponds to a definite compound (for example, a salt or a complex) or to a mixture of it with an excess of one of the reacting components, while the compound may correspond to a minimum or an inflection of the curve. To resolve this question it is necessary to apply other, long-known and better-developed methods: determination of electrical conductivity, thermal analysis, etc. Meanwhile, the systems that have given us the most interesting results for the most part still remain very little studied, or not studied at all, from other points of view. Such systems as: amine–acid, amine + phenol, nitro compound + acid, nitro compound + phenol, etc., must be subjected to comprehensive physicochemical investigation. Only then will it be possible to find general regularities expressing the dependence of properties on composition, and to assign to each method its proper place in solving practical problems of organic chemistry.

1. Formic and acetic acids rapidly gave crystalline salts, which made determination of the emf impossible. ↩