B. V. Tronov and A. A. Lukanin

The Effect of Neutral Salts

on the Rate of Oxidation of Organic

Compounds by Permanganate

Tomsk
1929

B. V. Tronov and A. A. Lukanin.

Effect of Neutral Salts on the Rate of Oxidation of Organic Compounds by Permanganate.

In elucidating the mechanism of oxidation-reduction processes, it is necessary to consider one question connected with the theory of electrolytic dissociation.

When we have some system, oxidant + reductant, very often one of the components of this system, and sometimes both, are dissociated to a greater or lesser extent into ions. This is especially the case in aqueous solution, i.e., precisely under the conditions in which such reactions are usually carried out. In textbooks of analytical chemistry, as well as in scientific works on this question, it is usually assumed that oxidation-reduction reactions take place between ions. Thus, Holluta, in a series of valuable papers\(^{1}\) on the study of the process of oxidation of salts of formic acid by \(KMnO_4\) under various conditions, arrives at the following equations:

\[ MnO_4' + HCO_2' + H_2O = H_2MnO_4 + HCO_3' \]

\[ \text{and } 2MnO_4' + 2OH' + HCO_2' = MnO_4'' + HCO_3' + H_2O \]

that is, of all the bodies participating in this reaction, only water is written by him in the form of neutral molecules. \(KMnO_4\), \(HCOONa\), and alkali appear here in the form of ions. Other chemists, on the contrary, studying the action of the same oxidant, use equations containing only neutral molecules. For example, in the article by S. S. Nametkin\(^{2}\) we find the following interpretation of the process of oxidation of unsaturated compounds:

\[ \begin{array}{c} \backslash CH\\ \ \ \Vert\\ /CH \end{array} + 2OMnO_2OK \longrightarrow \begin{array}{c} -CH-OMnO_2OK\\ \ \ \vert\\ CH-OMnO_2OK \end{array} \longrightarrow \begin{array}{c} CHOH\\ CHOH \end{array} + 2MnO_4HK \]

\[ 3MnO_4HK \longrightarrow 2MnO_4K + MnO_2 + H_2O + KOH. \]

In general, physical chemists seem to believe that oxidants and reductants can react only in the form of ions, while organic chemists, on the contrary, do not take dissociation into account at all.

In fact, for example, in a solution of \(KMnO_4\), there exist neutral molecules of the salt and the ions \(K'\) and \(MnO_4'\).\(^{3}\) The ions \(K'\) do not change during the oxidation by permanganate of some substance; the ions \(MnO_4'\) and the molecules \(KMnO_4\) disappear, being converted into other compounds, and what is altered first of all—this requires special investigation.

\(^{1}\) J. Holluta. Zeit. f. phys. Chemie 106, 276–294 (1923); 106, 324–340; 113, 461–482 (1924), and others.

\(^{2}\) Journ. f. pract. Chem. 108, 46–51; Journal of the Russian Physico-Chemical Society 55, 70–74.

\(^{3}\) Recently, the theory of complete dissociation of strong electrolytes has acquired many adherents; it was first expressed by Bjerrum in 1906, with subsequent mathematical formulation by Ghosh (Journ. Chem. Soc. 113, 449, 627, 707, 790 (1918); 117, 823, 1390 (1920); Zeit. phys. Chem. 98, 211 (1921)), and also earlier by Kbelin, Zeit. phys. Chem. 77, and others.

However, on the other hand, works are appearing in which it is definitely indicated that sometimes neutral molecules participate in reactions. Thus, for example, E. A. Ranitedt (Chem. Zbl. 1915, III, 12) found that acids accelerate esterification not only by their hydrogen ion, but also by the whole molecule.

From the electronic point of view, the process of oxidation by some atom (or group of atoms) of another atom (group) consists in the former taking electrons away from the latter. The substance that acquires electrons is the oxidizing agent and, in the given process, is reduced; the other is the reducing agent, which is oxidized. It is clear that any body can acquire a negative charge more easily when it is neutral than when it already has a negative charge, and, having a positive charge, more easily than in the neutral state. Conversely, electrons are lost more easily by a negatively charged body than by a neutral one or one having a positive charge. Hence: if some element (or radical) is capable of existing in the form of a negative ion or of a neutral atom (molecule or radical forming part of a molecule), then it must more readily acquire electrons (oxidizing other substances) in the neutral state than in the form of an ion.

Experience shows that, for example, the group $\mathrm{NO_3}$, being only in the form of negative ions, i.e., in a solution of some salt of an alkali metal $(\mathrm{KNO_3}, \mathrm{NaNO_3})$, possesses extremely weak oxidizing properties. The oxidizing character is strengthened if $\mathrm{NO_3}$ is combined with a weak metal (copper, iron, lead) or with hydrogen, since then dissociation is weaker and there are more neutral molecules in the solution. Concentrated nitric acid consists almost entirely of neutral molecules and oxidizes very vigorously.¹) Nitric anhydride, which does not dissociate at all, acts still more vigorously. Nitrous acid, having less oxygen, is, however, a stronger oxidizing agent than nitric acid—probably precisely because of its weak acidity; likewise nitrogen dioxide $\mathrm{NO_2}$. Exactly the same is repeated in the series of oxygen compounds of chlorine. $\mathrm{HClO}$ is a very weak acid; its salts, even with alkali metals, dissociate little into ions and are oxidizing agents. $\mathrm{KClO_3}$ and especially $\mathrm{KClO_4}$ dissociate very well, and their solutions oxidize very poorly. Chromic anhydride, entering into the composition of the negatively charged ion $\mathrm{CrO_4^{\prime\prime}}$ or $\mathrm{Cr_2O_7^{\prime\prime}}$, loses the ability to take electrons from other bodies, and if, for example, alcohol is added to a solution of potassium chromate or dichromate, no reaction will occur. If alcohol is added to a slightly dissociated solution of $\mathrm{CrO_3}$ in water (especially in the presence of sulfuric acid, which suppresses the dissociation of weak acids), the alcohol is oxidized rather rapidly, and with solid chromic anhydride it even ignites.

With the oxides and acids of sulfur we have the following. The group $\mathrm{SO_3}$ can exist either as the neutral molecule of sulfurous anhydride or in the form of the negative ion $\mathrm{SO_3^{\prime\prime}}$ (the anion of sulfurous acid salts, for example, $\mathrm{Na_2SO_3}$). The negatively charged group $\mathrm{SO_3^{\prime\prime}}$ must readily lose electrons, i.e., be oxidized. And indeed, salts of sulfurous acid, namely the normal salts giving the ion $\mathrm{SO_3^{\prime\prime}}$, find application as reducing agents, for example in photographic development. The free acid $\mathrm{H_2SO_3}$ dissociates little and is oxidized with more difficulty. The neutral molecule $\mathrm{SO_3}$, however, is even an oxidizing agent. Sulfuric acid, which can only be an oxidizing agent, is such, on the contrary, only in the pure state, but by no means in salts.

¹) For this reason, strong and dilute nitric acid occupy different positions in the electrochemical series (their oxidizing potentials are different). See I. M. Silin, Proceedings of the Tomsk Technological Institute, vol. 44 (1923).

A somewhat more complex example from organic chemistry belongs to this same category. The carbonyl group \(C=O\), entering into the composition of the neutral molecules of aldehydes and ketones, is easily reduced (for example, by hydrogen at the moment of its liberation). This same group in acids that form a negative ion (for example, the anion of acetic acid

\[ \mathrm{CH_3-C\!\! \begin{matrix} /O \\ =O' \end{matrix}} \]

) is reduced with extreme difficulty. Likewise (and probably still more difficultly) in salts of these acids with strong metals. However, if the hydrogen of the carboxyl is replaced by a hydrocarbon radical not split off in the form of an ion, then the resulting ester is reduced similarly to ketones and other carbonyl compounds.

The iron ion presents an example of a positive ion which readily changes its charge in one direction or the other and, consequently, can serve both as an oxidizing agent and as a reducing agent. In salts with strong acids, such as \(\mathrm{FeCl_2}\), divalent iron is oxidized rather difficultly, i.e. the \(\mathrm{Fe^{\cdot\cdot}}\) ion, already having two positive charges, loses electrons with difficulty. The weaker the acid combined with the iron, i.e. the more undissociated molecules there are, the more easily oxidation takes place (for example, \(\mathrm{FeCO_3}\)). Finally, ferrous hydroxide hydrate \(\mathrm{Fe(OH)_2}\) in air is very rapidly converted into \(\mathrm{Fe(OH)_3}\). \(\mathrm{FeSO_4}\) is stable in the presence of free sulfuric acid, but in neutral solution slight hydrolysis occurs and the hydrate formed is immediately oxidized. Therefore a solution of ferrous sulfate in air turns brown with time and becomes turbid. As an oxidizing agent, however, salts of trivalent iron with strong acids, such as \(\mathrm{FeCl_3}\), are used.

With regard to \(\mathrm{KMnO_4}\), theoretically it had to be assumed that the group \(\mathrm{MnO_4}\) should also be a stronger oxidizing agent in the neutral molecule than in the form of a negative ion. To test this proposition we carried out a series of experiments intended to trace how the oxidizing capacity of permanganate changes at one and the same total concentration of it, but at different degrees of dissociation. The change in the relative amount of ions and neutral molecules was achieved by adding neutral salts having an ion common with \(\mathrm{KMnO_4}\). In addition, for comparison, neutral salts without an ion common with \(\mathrm{KMnO_4}\) were also taken. The following have so far been investigated: \(\mathrm{KCl}\), \(\mathrm{NaCl}\), \(\mathrm{KNO_3}\), \(\mathrm{NaNO_3}\), \(\mathrm{KClO_4}\), \(\mathrm{K_2SO_4}\), and \(\mathrm{Na_2CO_3}\).

Ethyl alcohol was taken as the reducing agent. The reaction was conducted in alkaline solution, since here the first stage of the deoxidation of manganese from heptavalent to hexavalent is easily effected.

For a more detailed elucidation of the course of the process, experiments were also performed at various concentrations of permanganate (from 0.0001 to 0.0004 mol per liter), and of caustic potash (from 0.009 M to 0.0045 M). A large excess of alcohol was taken (0.35 M per liter), and neutral salts up to 0.3 M.

The greater part of the measurements was made at temperatures from \(16.8^\circ\) to \(18^\circ\), some at \(26.8^\circ\), in order to derive the temperature coefficient. Samples of the solutions were titrated after definite intervals of time (after addition of \(\mathrm{HCl}\), \(\mathrm{KI}\), and starch) with \(\mathrm{Na_2S_2O_3}\).

The results of the experiments are given in Tables Nos. 1–4, where are indicated the concentrations, temperatures, time intervals (in minutes) after which samples were taken, and the amount of active oxygen remaining \((a-x)\), expressed in \(\mathrm{cm^3}\) of 0.005 M \(\mathrm{Na_2S_2O_3 \cdot 5H_2O}\) solution.

In the summary Tables Nos. 5–8 a comparison is made of the reaction rates under various conditions. Since the constants “\(K\),” calculated

according to the equation of a monomolecular reaction do not remain constant, it was necessary separately to calculate the acceleration of the reaction for one or another phase of it. The ratios of the quantities “\(K\)” upon addition of neutral salts and without salt were taken for the same percentage of reacted oxidant.

The reaction in these tables was followed only up to 20% of the reacted oxidant, which corresponds to the conversion of manganese from heptavalent to hexavalent (conditionally), since hexavalent manganese undoubtedly, during this interval of time, was partly oxidized further, although to a lesser extent than it was formed; this is proved (not determined quantitatively) by the sufficiently pure green color of the solution and the corresponding appearance of the absorption spectrum.

TABLE No. 1. Concentration: 0.0002 mol \(KMnO_4\); 0.003 m \(KOH\); 0.35 m \(C_2H_5CH_2OH\).

*) at 26 minutes.

TABLE No. 1. (continued)

TABLE No. 2. Concentration: 0.0002 M KMnO₄; 0.045 M KOH; 0.35 M CH₃CH₂OH.

TABLE No. 3. Concentration: 0.0001 M KMnO₄; 0.30 M CH₃CH₂OH. Temperature 17°–18°

TABLE No. 4. Concentration: 0.0004 M KMnO₄; 0.35 M CH₃CH₂OH.

TABLE (summary) No. 5. Concentration 0.0002 M KMnO₄; 0.027 M KOH.

TABLE 6 (summary) No. 6. Concentration: 0.0001 M \(KMnO_4\); 0.045 M \(KOH\); 0.35 M \(CH_3CH_2OH\).

TABLE (summary) No. 7. Concentration: 0.0001 M KMnO₄; 0.3 M CH₃CH₂OH.

TABLE (summary) No. 8. Concentration: 0.0004 M KMnO₄; 0.3 M CH₃CH₂OH.

T = 17–18°

Conclusions

From the series of experiments with salts of monovalent cations \(K^{\cdot}\) and \(Na^{\cdot}\) presented in this work, the following conclusions may for the present be drawn, relating chiefly to the first stage of the reduction of permanganate, i.e., to the transition from \(Mn^{VII}\) to \(Mn^{VI}\).

1. All the neutral salts of mineral acids taken by us do in fact accelerate the reaction of \(KMnO_4\) with \(C_2H_5OH\) in the presence of caustic potash. The increase in the reaction rate is almost (somewhat less than) proportional to the concentrations of the added salt.

2. The nature of the cation of the neutral salt (\(K^{\cdot}\) or \(Na^{\cdot}\)) has no noticeable effect on the rate.

3. The nature of the anion in the case of salts of monovalent strong acids also has little influence. Sulfate salts accelerate the oxidation, but considerably more weakly. A sharp difference is observed in the influence of salts of carbonic acid. In this case, it would seem that a considerable acceleration should be expected, since soda itself, having an alkaline reaction, ought to intensify the action of caustic potash (its influence is large). Experiment, however, gives the opposite result: upon addition of soda, and also of potash (\(K_2CO_3\) experiments are not given by us in the present article), the reaction is retarded.

4. Changing the concentration of potassium permanganate within the limits from 0.0001 M to 0.0004 M per liter has almost no effect on the change in the reaction rate. The influence of neutral salts (\(KCl\)) at these concentrations of \(KMnO_4\) remains almost constant. At considerable concentrations of \(KOH\) (0.13 M per liter), the action of the salt decreases.

5. The temperature coefficient of the reaction (from \(16.8^\circ\) to \(26.8^\circ\)), calculated for the same percentage of active oxygen consumed, increases noticeably upon addition of \(KCl\), and still more upon addition of \(K_2SO_4\). The ratio of the reaction rates at \(16.8^\circ\) for \(K_2SO_4\) and \(KCl\) is 0.7; at \(26.8^\circ\), greater than 1.

In general, the results obtained, with the exception of the experiments with \(Na_2CO_3\), correspond to the theoretical considerations set forth at the beginning of the work. As can be seen from the experiments carried out, the neutral molecule of permanganate is indeed a stronger oxidizing agent than the negative ion. How many times \(KMnO_4\) exceeds the ion \(MnO_4'\) in its oxidizing capacity we find it difficult to say, since it is extremely difficult to estimate the degree of dissociation of potassium permanganate in the presence of rather considerable amounts of other salts and alkali. If the relative concentration of molecules and ions is taken to be the same as in the added \(KCl\), then for the ratio of rates

\[ \frac{kKMnO_4}{kMnO_4'} \]

numbers from 28 to 33 are obtained.

At first glance, the comparatively small acceleration under the influence of \(K_2SO_4\) seems somewhat strange. This salt acts no more strongly than equimolecular concentrations of \(KCl\), etc.; yet in the sufficiently dilute solutions used by us there should be noticeably more \(K^{\cdot}\) ions, and consequently the dissociation of permanganate should be more suppressed. The explanation must probably be sought in the considerations expressed by Christiansen,¹ developing the theory

¹ Zeit. f. phys. Ch. 113, 35 (1924); 117, 433 (1925); Brönstedt. Zeit. f. phys. Chemie 102, 169 (1922) and earlier.

Bjerrum–Brønsted. He draws attention to the electric fields that arise around ions. These fields are further complicated by those large aggregates of ions and molecules that are formed owing to the forces of electrostatic attraction. Having studied the effect of neutral salts on the rate of oxidation of potassium iodide (iodine ions) by hydrogen peroxide, Christiansen found the following regularities:

1) With a constant cation, the rate changes little with the anions.
2. With uni-univalent salts their effect is always positive.
3. With uni-bivalent salts (a bivalent anion), a slight retardation may be observed.
4. With bi-monovalent salts (a bivalent cation), there is a very strong acceleration.

In the case of salts of univalent cations with likewise univalent anions, both causes—the one indicated in our work and the one adduced by Christiansen—act in the same direction. As a result, a considerable acceleration is observed.

In the case of sulfite salts of potassium and sodium, the decrease in the dissociation of potassium permanganate accelerates the reaction, while the influence of the electric fields of the added ions should retard it somewhat. Since we used salt concentrations considerably lower than Christiansen did, it is natural that the first effect sometimes predominates and again gives an acceleration, though a relatively smaller one. It is interesting that this acceleration increases with increasing temperature. This may be due to the disintegration of those aggregates of molecules and ions of which Brønsted and Christiansen speak.

The sharp difference in the effect of $\mathrm{Na_2CO_3}$, which nevertheless is in agreement with Christiansen’s conclusions, will be studied and explained by us in greater detail in the future. Likewise, the effect of salts of di- and trivalent cations, the study of the course of the reaction under other conditions (for example, without alkali), and experiments for the numerical determination of the relative oxidizing power of neutral molecules and ions will constitute the subject of subsequent work.