PHYSICAL CHEMISTRY

Corresponding Member of the Academy of Sciences of the USSR A. I. BRODSKII, I. F. FRANCHUK,
and V. A. LUNENOK-BURMAKINA

ISOTOPIC STUDY OF THE MECHANISM OF ELECTROLYTIC FORMATION AND HYDROLYSIS OF PERSULFATE

The various mechanisms that have been proposed for the anodic formation of persulfates during the electrolysis of sulfates may be assigned to two types. According to the widely held views \((^{1,11})\), persulfate is formed by direct recombination of discharging sulfate ions (or bisulfate ions), for example,

\(2\mathrm{SO}_4^{--} \to \mathrm{S}_2\mathrm{O}_8^{--} + 2e\),

or in two steps:

\(\mathrm{SO}_4^{--} \to \mathrm{SO}_4^{-} + e;\quad 2\mathrm{SO}_4^{-} \to \mathrm{S}_2\mathrm{O}_8^{--}\).

According to other proposals \((^{2,3})\), at the anode or in the near-anode layer of the electrolyte the primary products formed are products of water oxidation \((\mathrm{H}_2\mathrm{O}_2,\ \mathrm{OH},\ \mathrm{OH}^{-}\), surface oxides, etc.), which then oxidize sulfate by transfer of electrons or oxygen atoms. In particular, the mechanism of Glestoun and Hickling \((^3)\), with intermediate formation of hydrogen peroxide, has gained wide acceptance:

\[ 2\mathrm{HO}^{-} \to \mathrm{H}_2\mathrm{O}_2 + 2e;\quad 2\mathrm{SO}_4^{--} + \mathrm{H}_2\mathrm{O}_2 \to \mathrm{S}_2\mathrm{O}_8^{--} + 2\mathrm{OH}^{-}. \]

These authors extend it to processes of anodic oxidation in general. Most of the other proposed mechanisms belong to one of the two types, with different degrees of detail in the intermediate stages.

A. N. Frumkin and co-workers \((^4)\) found that during the electrolysis of a solution of \(\mathrm{K}_2\mathrm{SO}_4\) in \(\mathrm{H}_2\mathrm{O}^{18}\), in an acidic, neutral, or weakly alkaline medium, the oxygen of the persulfate obtained is free of excess heavy oxygen. This makes it possible to exclude from consideration all mechanisms in which the participation of water oxygen in persulfate formation is assumed.

We used the heavy isotope of oxygen to clarify the possible participation of hydrogen peroxide in the anodic formation of persulfate and to study the mechanism of its hydrolysis. From earlier data \((^{4,5,10})\) it is known that \(\mathrm{H}_2\mathrm{O}_2\) and \(\mathrm{K}_2\mathrm{S}_2\mathrm{O}_8\) do not exchange oxygen with water.

Solutions of 40 g of \(\mathrm{KHSO}_4\) in 200 ml of water were subjected to electrolysis with a current of 3 A between platinum electrodes at \(10\text{–}15^\circ\). To take samples of anodic oxygen and to improve mixing of the electrolyte, a wire anode of \(0.8\ \mathrm{cm}^2\) was placed in an open-bottom tube between two vertical cathode plates. The content of \(\mathrm{O}^{18}\) was determined with an MS-2 mass spectrometer: in anodic oxygen directly, and in water—by a method developed in our laboratory \((^6)\), in the form of \(\mathrm{CO}_2\) after its exchange with vapors of the water under investigation. Sulfate was precipitated as \(\mathrm{PbSO}_4\), which was converted into \(\mathrm{CO}_2\) by ignition with dehydrated hydrogen charcoal. Oxygen was liberated from persulfate by heating. For isotopic analysis of hydrogen peroxide, it was decomposed with a solution of \(\mathrm{KMnO}_4\) directly in the electrolyte after its degassing by pumping, or, at low concentrations, after extraction with ether. All these procedures were checked by control experiments. Hydrogen peroxide was determined by titration with permanganate, and persulfate iodometrically in the presence of \(\mathrm{Cu}^{++}\) as catalyst \((^7)\). During electrolysis no significant quantities of Caro’s acid were formed. The results of the final experiments are reported below.

1. In experiments without the use of isotopic indicators, to the electrolyte do-

10–20 g/liter of H₂O₂ was added. In this case the yield of persulfate fell sharply in comparison with the experiment without addition of H₂O₂ (curve 1) and then increased as the undecomposed residue of H₂O₂ decreased. In Fig. 1, curves 2 show the increase in the concentration of K₂S₂O₈, while curves 3 show the decrease in the concentration of H₂O₂ in the electrolyte in two experiments, a and b, with initial H₂O₂ contents of 12 and 20 g/liter. Electrolysis in all three experiments was carried out until solid K₂S₂O₈ precipitated. From these data it is evident that both anodic processes—the decomposition of H₂O₂ and the formation of K₂S₂O₈—apparently proceed independently of one another, with a large predominance of the former if the H₂O₂ concentration is sufficiently high, so that the presence of peroxide interferes with the formation of persulfate. This in itself casts doubt on the theory of the intermediate formation of hydrogen peroxide. Qualitatively the same results and the same conclusions were obtained by Gaisinskii (8) in the formation of percarbonate by electrolysis of K₂CO₃ in the presence of H₂O₂.

Fig. 1. Dependence of the persulfate yield on the concentration of hydrogen peroxide.
1—without addition of H₂O₂; 2a and 2b—with initial H₂O₂ contents of 12 and 20 g/liter; 3a and 3b—decrease in the H₂O₂ content in the electrolyte in the same experiments

Table 1

Isotopic composition of the anodic oxygen during electrolysis of KHSO₄ + H₂O₂ in H₂O¹⁸ with 0.890% O¹⁸

The independence of the two anodic processes is confirmed by the following experiments on the electrolysis of a solution of KHSO₄ + H₂O₂ in H₂O¹⁸ with isotopic analysis of the anodic oxygen (Table 1). In it the O¹⁸ content at the beginning was close to that which the peroxide had (natural: 0.20%), and gradually approached its content in water in accordance with the curves in Fig. 1. From these data it is also evident that the anodic oxidation of hydrogen peroxide takes place without participation of the oxygen of water.

Table 2

O¹⁸ content in H₂O₂ during electrolysis of KHSO₄ + H₂O₂ in H₂O¹⁸

1. For final direct proof that hydrogen peroxide does not participate in the anodic formation of persulfate, we used the isotope dilution method. In experiments analogous to those just described, isotopic analysis was performed on the residue of H₂O₂ in the electrolyte after it had fallen below 1 g/liter. If H₂O₂ were formed at the anode from water, it would have to contain an excess of O¹⁸ and introduce it into the residue of the H₂O₂ added beforehand, mixing with the latter. At a cathodic current density of 0.05–0.17 A/cm² we did in fact find in the residue of H₂O₂ several times more

increased content of \(O^{18}\), which was the greater the lower the current density (Table 2). It was caused by cathodic formation of peroxide by the Brutto reaction \(O_2 + 2H^+ + 2e \to H_2O_2\), which is unrelated to the anodic process and, as is known, proceeds with a better yield the lower the cathodic current density. Indeed, when we increased the current density to \(0.75—1.0\ \mathrm{A/cm^2}\), even the last residues of \(H_2O_2\) contained no excess \(O^{18}\), whereas according to the mechanism of Gleyston and Hickling the \(O^{18}\) content in them should have approached its content in water.

Table 3

Hydrolysis of persulfate in the presence of \(HClO_4\) at \(70^\circ\) (content of \(O^{18}\) in percent)

From all these data it is clear that hydrogen peroxide cannot be an intermediate product in the formation of persulfate at the anode. Apparently, neither can OH radicals be such products (according to the scheme \(2SO_4^- + 2OH \to S_2O_8^{--} + 2OH^-\)), since they rapidly exchange oxygen with water \((^9)\) and easily recombine to \(H_2O_2\).

1. A mixture of \(1.5—4\) g \(K_2S_2O_8\) with \(1—3\) g of 70% \(HClO_4\) or 50% \(H_2SO_4\) was subjected to hydrolysis at \(70^\circ\), by passing water vapor through it at a pressure of 30 mm Hg. The isotopic composition of oxygen from \(H_2O_2\) in the distillate and of bisulfate in the residue was determined as indicated above. In experiments with \(K_2S_2O_8 + H_2O^{18}\) and \(K_2S_2O_8^{18} + H_2O\) (the heavy persulfate was obtained by electrolysis of \(KHSO_4^{18}\), prepared from \(H_2SO_4 + H_2O^{18}\)), the hydrogen peroxide had the composition of the water, as is evident from Table 3. The same result was given by hydrolysis under conditions of heating a solution of \(K_2S_2O_8\) with acid without distillation of the peroxide. Thus, all the oxygen of the peroxide originates from the oxygen of the persulfate without participation of the oxygen of water. These data agree with earlier studies on the decomposition of \(H_2O_2\) and several other peroxides in water with a different isotopic composition of oxygen \((^{11})\). In all cases it was found that the peroxide bridge is not ruptured and that the oxygen of water is not incorporated into the decomposition products of these peroxides (other peroxide compounds or \(O_2\)). This was found, in particular, by Banton and Llewellyn \((^{10})\) for the decomposition of Caro’s acid.

Comparison of these data with ours shows that in the sequence of transformations
\(S_2O_8^{--} \to SO_5^{--} \to H_2O_2 \to O_2\)
the peroxide group —O—O— passes, without rupture, from persulfate into the final product of its decomposition—oxygen. Since, under the conditions of our experiments, persulfate hydrolysis proceeds with the intermediate formation of Caro’s acid \((^{12})\), its simplest mechanism is as follows:

\[ {}^{-}OS(O_2)OO\boxed{S(O_2)O^- + HO}H = {}^{-}OS(O_2)OO^- + H^+ + H\overset{*}{O}S(O_2)O^- \]

\[ H\boxed{\overset{*}{O}H + {}^{-}OS(O_2)}OO^- + H^+ = H_2O_2 + H\overset{*}{O}S(O_2)O^- . \]

According to this mechanism, bisulfate should contain \(3/4\) of its oxygen from persulfate and \(1/4\) from water. The oxygen from persulfate decomposition had an isotopic composition corresponding to this scheme, but in the bisulfate its composition was close to that of water (Table 3), evidently because of fairly rapid exchange between \(HSO_4^-\), or the \(H_2SO_4\) formed from it, and water. In order to eliminate this side exchange as far as possible, we carried out experiments with addition of \(Pb(ClO_4)_2\), so that the sulfate formed immediately precipitated as \(PbSO_4\). This greatly decreased the exchange, but did not eliminate it completely. In two experiments at \(70^\circ\) and an initial content of 1.17% \(O^{18}\) in the water, the sulfate contained 0.412% (with an experiment duration of 1 hour) and 0.492% (2 hours) \(O^{18}\), instead of the 0.44% required by the scheme, or somewhat less, taking into account the initial—

initial dilution with light water introduced with \(\mathrm{HClO_4}\). When the hydrolysis temperature was increased to \(100^\circ\), the \(\mathrm{O}^{18}\) content increased to 0.635%. These data undoubtedly indicate that a significant amount of \(\mathrm{O}^{18}\) is introduced into the bisulfate formed by secondary exchange. They are consistent with the scheme presented, although they do not provide its final quantitative confirmation.

L. V. Pisarzhevsky Institute of Physical Chemistry
Academy of Sciences of the Ukrainian SSR

Received
13 III 1957

REFERENCES

1. K. Elbs, O. Schönherr, Zs. Elektrochem., 1, 417, 468 (1894); O. Essin, Zs. Elektrochem., 32, 267 (1926); W. D. Bancroft, Trans. Elektrochem. Soc., 71, 195 (1937).
2. M. Le Blanc, Zs. phys. Chem., 8, 299 (1891); 12, 333 (1893); F. Foerster, Electrochemie nichtwässeriger Lösungen, Leipzig, 1922; O. A. Esin, E. Alfimova, ZhFKh, 3, 439 (1932).
3. S. Glasstone, A. Hickling, Electrolytic Oxydation and Reduction, London, 1935; Chem. Rev., 25, 407 (1939).
4. A. N. Frumkin, R. I. Kaganovich, M. A. Gerovich, R. N. Vasil’ev, DAN, 102, 981 (1955).
5. E. R. S. Winter, H. V. A. Briscoe, J. Am. Chem. Soc., 75, 496 (1953); A. E. Cahill, H. Taube, J. Am. Chem. Soc., 74, 2312 (1952); P. Baertschi, Experientia, 7, 215 (1952).
6. A. I. Brodsky, S. G. Demidenko, L. L. Strizhak, V. R. Lechekhleb, Zhurn. anal. khim., 10, 256 (1955).
7. M. A. Bodin, Zav. lab., 11, 1248 (1938).
8. M. Haissinsky, Trans. Farad. Soc., Discuss. No. 1, 254 (1947).
9. O. R. Forscheimer, H. Taube, J. Am. Chem. Soc., 74, 3705 (1952); 76, 2099 (1954); I. A. Kazarnovskii, N. P. Lipikhin, M. V. Tikhomirov, ZhFKh, 30, 1429 (1956).
10. C. A. Bunton, D. R. Llewellyn, Res., 5, 142 (1952).
11. W. C. Schumb, Ch. N. Satterfield, R. L. Wentworth, Hydrogen Peroxyde, N. Y., 1955.
12. H. Palme, Zs. anorg. Chem., 112, 97 (1920).
13. I. M. Kolthoff, I. K. Miller, J. Am. Chem. Soc., 73, 3055 (1951); L. S. Lewitt, Canad. J. Chem., 31, 915 (1953).