B. V. Tronov and L. P. Kulev

Determination of Hydrogen Activity in Organic Compounds from the Electromotive Force of the Reaction with Sodium

The purpose of the present work is to compare the mobility of hydrogen in various organic compounds by the method of determining the electromotive force arising in the reactions of these compounds with metallic sodium. We tested compounds with mobile hydrogen, first of all and most fully alcohols; then carbonyl compounds that give the enol form and, preliminarily, some hydrocarbons, amines, nitriles, and nitro compounds were examined.

Of the works in this field, the one closest to our investigations is the paper “Relative Mobility of Hydroxyl Hydrogen in Certain Alcohols”*). The authors measured the rate of reaction of alcohols with \(p-\mathrm{NO_2}-\mathrm{C_6H_4COCl}\). The experiments were carried out in ethereal solution. The rate constant was calculated at concentrations of 1 mol KOH + 1 mol \(\mathrm{NO_2C_6H_4COCl}\) in 1000 g of ether. The titration of the HCl formed was carried out after preliminary treatment of the reaction products in the cold with \(\mathrm{C_6H_5N(CH_3)}\), water, and carbon disulfide (to extract organic substances). The results obtained by the authors are presented in the following table:

Thus, both in the fatty and in the aromatic series, the authors established periodicity of activity. Butyl alcohol proved to be more active than propyl alcohol, and phenylethyl alcohol more active than benzyl and phenylpropyl alcohols.

The esterification reaction is also one of the indirect methods for determining the activity of the hydroxyl hydrogen of alcohols. It is known that

* J. F. Norris, A. A. Ashdown, J. Am. Ch. S. 47, 837.

the interpretation of the mechanism of esterification and saponification, which has more than once been proposed in the literature and which best satisfies the experimental data, corresponds to the scheme:

\[ \mathrm{RC\!\left(\begin{array}{c} \|O\\[-2pt] OH \end{array}\right)} +\mathrm{HOR^1} \rightleftarrows \mathrm{RC\!\left(\begin{array}{c} OR^1\\[-2pt] OH\\[-2pt] OH \end{array}\right)} \rightleftarrows \mathrm{RC\!\left(\begin{array}{c} OR^1\\[-2pt] \|O \end{array}\right)} +\mathrm{HOH} \]

according to which the alcohol reacts not with the aqueous residue, but with the hydroxylic hydrogen. Consequently, by comparing the rates of formation of esters, one can also compare the mobility of the hydrogen of alcohols in these reactions.\(^{1}\)

The method we propose is as follows: into a vessel of capacity “22.5 cc” were immersed electrodes—a platinum electrode, in the form of a hollow cylinder, and metallic sodium. The size of the platinum plate was: width “40” mm, length “66.” To avoid oxidation, the metallic sodium was pressed out of a glass cartridge at the very moment of determining the EMF and was immersed in the solution of the compound under investigation. The EMF was determined from the scale of a sensitive galvanometer (Keiser Schmidt No. 690) and a voltmeter with a scale up to 2.5 V.

Experimental procedure.

5 cubic centimeters of the substance under investigation were placed in vessel A. Platinum electrode B was immersed in the liquid. From a piece of metallic sodium, 9 mm thick, a cylinder C was cut out with a glass cartridge and, with the aid of rod D, was pressed out into the vessel containing the substance. At the moment the sodium was immersed in the liquid, the electromotive force was noted. In order to make the results comparable and to equalize the resistances of the substances tested, the latter were taken in equimolecular concentrations in benzene solution. Octyl alcohols, as compounds with the greatest molecular volume, were tested without solvent; the remaining compounds were diluted to 5 cc with benzene, in amounts corresponding to the molecular weight of octyl alcohol. Thus, the weighed portions of certain alcohols, diluted to 5 cc with benzene, were expressed, for example, by the following figures:

and so forth, calculated for the same concentration of CH.

The experiments were carried out under approximately identical temperature conditions (within 18–20°C).

Thus, the heat liberated in the reaction with sodium influenced the magnitude of the electromotive force and, moreover, the alcoholate formed did not make it possible to judge the true magnitude of the EMF of the basic

\(^{1}\) Zh. R. Kh. O. 18, 355; 19, 623; 23, 263; 38, 980. N. Menshutkin.

reaction, we took into account only the maximum readings of the voltmeter at the beginning of the reaction. In all, we tested 52 compounds. Of these, the alcohols are represented more fully.

Fatty series:
Primary . . . . . . . . 10
Secondary . . . . . . . 5
Tertiary . . . . . . . . 2
Undetermined . . . . . . 1

Aromatic series:
Primary . . . . . . . . 3
Phenols . . . . . . . . 4
Unsaturated . . . . . . 1
Cyclic . . . . . . . . . 1

The results obtained are presented in the following table: the maximum EMF for alcohols is expressed in millivolts.

For all the alcohols listed here, experiments were also carried out to determine the electromotive force without solvent. These experiments give less comparable results, since the concentrations of hydroxyl groups are different here; nevertheless, in this way it proved possible to compare alcohols that gave no EMF at all in benzene solution (tertiary alcohols), and also to investigate glycols not miscible with benzene.

In addition to alcohols, as a preliminary experiment, we investigated other compounds that also give an EMF in reaction with metallic sodium.

The experiments were carried out under conditions analogous to those for alcohols, i.e., in benzene solution and without a solvent.

In benzene solution:

Without solvent:

The results obtained make it possible to draw the following conclusions:

I. In the series of saturated monohydric alcohols:

With increasing molecular weight, the activity of hydrogen decreases. In the first three members it decreases nonuniformly; further along the homologous series, on average, by about 2.

II. Branching of the chain (isobutyl and isoamyl) decreases the activity by a factor of 4 (propyl and isobutyl; butyl and isoamyl).

III. Secondary alcohols, in general, are 20–30 times less active than primary ones (the exception is \(CH_3CHOHCH_3\), which is only 6 times less active than propyl). This is evidently explained by the fact that propyl alcohol, with the shortest chain length, is more accessible to reactions with hydroxyl hydrogen than all other secondary alcohols.

IV. For secondary alcohols with a branched chain, the coefficient of weakening of activity is approximately equal to the product of the corresponding coefficients for secondary alcohols and alcohols with a branched chain.

V. In normal primary alcohols we also established a periodicity in the change of hydrogen activity. But this periodicity is not expressed as clearly as in the experiments with \(C_6H_5COCl\). Activity decreases continuously with lengthening of the chain, but the difference between an even member of the series and the preceding odd one is greater than with the subsequent odd one (thus propyl—120, butyl—60, amyl—45).

Tertiary.

I. The activity of hydrogen is negligibly small. Comparison of the results obtained without solvent leads to an interesting conclusion. Tertiary butyl alcohol proved to be considerably more active than tertiary amyl alcohol. This is fully in accord with the fact that trimethylcarbinol reacts with acetic anhydride, although slowly, forming an ester, whereas dimethylethylcarbinol cannot be esterified by this route at all\(^{2}\).

Aromatic compounds with hydroxyl in the side chain:

1. Their activity is considerably less than that of saturated alcohols with the same chain, despite dilution of the latter with benzene.

2. In aromatic alcohols a periodicity has been established in the change of hydrogen activity, depending on the length of the chain (\(C_6H_5CH_2CH_2OH\)—greater than the others).

3. With lengthening of the chain, the activity decreases.

Unsaturated alcohols.

1. The double bond increases the activity of hydrogen in the fatty series in alcohols with an open chain.

\(^{1}\) See J. F. Norris, A. A. Ashdown

\(^{2}\) B. Tronov and N. Sibgatullin: “On the preparation of esters by the action of acetic anhydride on alcohols.”

Alcohols of the cyclane series

1. Ring closure has almost no effect on activity (hexyl—cyclohexanol—secondary isoheptyl).

Phenols

1. In their activity phenols approach primary saturated alcohols with the same number of carbon atoms (hexyl and phenol, heptyl and cresols).

2. In cresols the activity changes in the following order: meta \(>\) para \(>\) ortho. Meta-cresol (EMF—8.70) is very close to primary heptyl (EMF—8.25) and benzyl (EMF—8.30), despite the fact that phenols possess a definite acidic character, which saturated and aromatic alcohols do not have.

Thus, no definite dependence can be observed between the activity measured by our method and the constant of electrolytic dissociation of alcohols. Apparently, this must be explained by the fact that the dissociation constant is determined in aqueous solution. Dissolution experiments in alkalis are carried out under the same conditions. We, however, worked with benzene solutions. Water forms complex compounds with alcohols, in which the activity of hydrogen must be strongly changed.

Other compounds

Aldehydes and ketones

No definite dependence between the activity of aldehydes and ketones and that of the corresponding alcohols can be established.

Examples: Propionic aldehyde is 200 times less active than propyl alcohol. Benzaldehyde is 2 times less active than benzyl alcohol. Cyclohexanone is 8 times more active than cyclohexanol. Acetone and isopropyl alcohol are almost equal. Methyl propyl ketone and diethyl ketone are about 3 times more active than secondary butyl alcohol.

Comparison of aldehydes with ketones

1. Acetone is 40 times more active than the propionic aldehyde isomeric with it. With increasing complexity of the ketone molecules, the activity decreases more slowly than in alcohols.

Amines

1. Of the amines, only aromatic and heterocyclic amines exhibit activity; moreover, this activity is not always determined by the mobility of the hydrogens attached to nitrogen.

1) In benzene solution, however, there are complex aggregates only of alcohol molecules.

Nitro compounds.

1. They exhibit a somewhat greater activity than the corresponding amines.

Nitriles.

1. They have, comparatively, very great activity. Thus acetonitrile is more active than acetone. Benzyl cyanide is more active even than benzyl alcohol. Since nitriles, under the action of sodium, polymerize, it must be thought that the EMF here arises not owing to the replacement of hydrogen by sodium, but corresponds to the reaction of addition of the metal to nitrogen.

Esters.

1. Acetoacetic ester proved to be the most active, as the one most capable of enol grouping. It is more active than lactic ester, where a hydroxyl group is present.