Historical Articles

February, 1953 issue of Plating

 

A Theory of Hydrogen Overvoltage and Its Relation to the Electrodeposition of Metals

---

George Dubpernell, United Chromium Inc., Waterbury, CT and
Roberta Dubpernell, Formerly of the Naugatuck Chemical Division, United States Rubber Company, Naugatuck, CT

 

DISCUSSION
While the Haring Cell has been widely used lor about twenty-five years, it seems to have been employed entirely in studies of metal deposition, especially copper, nickel, and silver. Ferguson describes a special Haring Cell arrangement for studies of polarization both with and without metal deposition, but the authors are not familiar with any published reports on polarization or hydrogen overvoltage alone.

Stareck and Taft give results on plain KCN solution between platinum electrodes, but do not deduct the IR drop measured in the center compartment, and do not report the IR drop measurements. Using Ag CN and KCN they observed a black silver deposit at the cathode at high current densities, which was most stable in alkaline solutions, and which evolved hydrogen gas and turned white upon decomposition. They assumed this to be a sil-ver hydride containing negative hydrogen, and noted that it had strong reducing properties. On the basis of the views in this paper, this black deposit would more simply be regarded as an electronegative metal alloy of silver, and this would explain its reducing properties without the need to assume the existence of negative hydrogen or “nascent hydrogen”.

The term “nascent hydrogen” has been used frequently for over one hundred years to describe reactions not attributable to hydrogen as it is ordinarily known, and yet there does not seem to be closer approach to an understanding of the matter than formerly. The use of the term was questioned by Reed and others, and a review is given in Gmelin. Many workers refer also to such things as “atomic hydrogen”, “overvoltage compounds”, “overvoltage hydrides”, “active material”, and so on, without justifying their use of these terms.

The presence of “atomic hydrogen” at the cathode was first suggested as a pure speculation by Ostwald, who also suggested the rate of combination of the atoms as a possible mechanism for the explanation of overvoltage. This type of explanation has been taken up by many more recent workers with enthusiasm, even though it was not mentioned in a later paper by Ostwald on “Electrolysis and Catalysis”, and the concept of the secondary formation of hydrogen at the cathode after primary deposition of an electronegative metal, was expounded at some length.

The properties of atomic hydrogen are well reviewed by Glockler and Lind, and in Gmelin. Langmuir studied it in some detail, but stated that he felt that it was a mistake to try to account for overvoltage by means of infinitesimal concentrations of atomic hydrogen, and again that slowness of progress in studies of overvoltage and passivity “. . . has probably been largely caused by an undue emphasis on what may be called the thermo-dynamic viewpoint in electrochemistry”. He established that atomic hydrogen is only stable in appreciable quantities at temperatures in the range of 2,000° C to 3,000° C or above. It may be stable for a short time at room temperature in a high vacuum in glass tubes, but recombines instantly to form molecular hydrogen in the presence of any oxygen or water. It has some reducing power, but nothing comparable to cathodic reduction, or that of metallic sodium.

Another reason for assuming the existence of atomic hydrogen at cathodes is the diffusion of hydrogen sometimes noted through thin metal foils, and the high- pressures sometimes built up inside of sealed metal tubes treated cathodically, or inside of crevices in metal so as to cause blistering. Ferguson and Dubpernell have shown that hydrogen does not diffuse through even very thin foils of platinum or palladium, and that the phenomena are due to porosity of the metal, and capillary and surface conductivity effects. These vie-s have not yet received very wide acceptance, but make any assumption of the existence of atomic hydrogen at room temperature unnecessary.

A recent paper by Fischer and Heiling is accepted as demonstrating the diffusion of atomic hydrogen or cathode polarization from the outside into the inside of a thin, annealed carbonyl iron beaker in N H2SO4. However neither the thickness nor the exact method of production of the iron beakers used is given, and apparently they were not tested for porosity. There was probably a direct solution connection between the outside and the inside, since the cathode polarization would increase on the inside as much as 0.3 volt or more in a few seconds following an increase on the outside. Likewise, every detail of polarization on the outside was reproduced on the inside, but with lower values.

The controversy which followed Le Blanc and Ostwald’s suggestion of the primary rather than secondary formation of hydrogen at the cathode, is reviewed in some detail by Sack. Arrhenius reported the results of numerous experiments in which he found a substantial time required to get hydrogen evolution on mercury cathodes from neutral salt solutions, and which he explained as due to the primary deposition - of the alkali metal before hydrogen evolution occurred. Le Blanc replied at some length, and felt that hydrogen evolution occurred immediately and directly, but went unobserved for one reason or another.

Nernst also investigated the matter, and reached the conclusion that while primary hydrogen evolution is possible, in practice it is mostly secondary. A student of Nernst’s, Glaser, published a more detailed study in which it was concluded that while theoretically hydrogen may deposit primarily, in actuality, at moderate current densities it is mainly formed secondarily.

Bancroft first favored the primary deposition of sodium and secondary formation of hydrogen. In discussing this 1905 paper, C. J. Reed favored the direct deposition of the “more easily reducible” hydrogen, while J. W. Richards emphasized that sodium deposits primarily on mercury, with no hydrogen evolution. In 191643 and 192044 Bancroft favored the assumption of the existence of atomic hydrogen at the cathode to explain overvoltage, but in 1929 emphasized the direct deposition of sodium, and the possibility of reproducing any cathode reaction chemically with sodium.

Fichter and co-workers appear to have followed the suggestion of Bancroft, in connection with organic reductions. On page eight of his book, Fichter refers to a number of his publications and states that it has been shown in numerous examples that an alloy of pure lead and sodium gives exactly the same reduction effect as a lead cathode in acid solution. Similarly, Parker and Swann found that reduction of sugars in acid solutions at zinc and lead cathodes only took place if an alkaline film could be maintained at the cathode surface, and that, if the catholyte was free of sodium ions, no reduction was obtained with an acid concentration above 0.1N.

Many phenomena become more understandable in the light of the theory of cathode polarization here presented. The change in surface tension of the mercury in the Lippman capillary electrometer is probably partially associated with the deposition of a minute amount of sodium, as pointed out by Freundlich. Edison discovered in 1874 in connection with the development of telephone and telegraph equipment that friction was decreased on a cathodically polarized metal surface and increased on the anode. The phenomenon was studied by a number of workers including Arons, but no satisfactory explanation was found. The effect might be due to a lubricating action of caustic soda formed at the cathode, on the surface alloyed with alkali metal. The “electrical lubrication” of equipment for handling clay by making the surfaces cathodic might involve a similar explanation, although this lubricating effect is usually ascribed to the production of a layer of water at the cathode by electroendosmosis.

In polarography, using the dropping mercury electrode, the “limiting current” or “diffusion current” which gives the concentration of the substance being determined, is stabilized and made more reproducible by working in the presence of an “indifferent salt” or “supporting electrolyte”, generally 0.1N KCl. This eliminates what is called the “migration current” and leaves the diffusion current as a more measurable quantity. It also tends to eliminate “hydrogen waves” and has other desirable effects. When the alkali or alkaline earth metals are being determined, a different supporting electrolyte is used such as tetramethylammonium hydroxide. The observations in this field should be more consistently understandable on the basis of the concept of hydrogen overvoltage proposed in the present paper.

Potentials frequently seem to involve adsorbed or otherwise held sodium or other alkali metal ions. The glass electrode is a sodium electrode to a considerable extent according to Dolea. MacGillavry, et al found little effect in acid solution, but substantial negative potentials on nickel immersed in alkaline solutions.

The salt effect in connection with hydrogen overvoltage was investigated by de Bethune, who observed many irregularities using a mercury cathode with various concentrations of hydrochloric acid and potassium chloride. When the concentration of acid was low and that of potassium chloride high, he obtained potassium deposition with 100 per cent current efficiency and overvoltages around 1.? volts or higher. By working carefully at low current densities, and with more acid and less salt, he considered that the mercury cathode functions as a hydrogen electrode, and measured overvoltages up to about 1.2 volts. It is felt that it is much more reasonable to regard the cathode potentials as due entirely to potassium rather than hydrogen, and then proceed to ascertain if any of the polarization at low current densities may be ascribed to hydrogen or not. The slight increase in overvoltage observed in the more dilute solutions of acid is probably due to the relatively greater effect of alkali dissolved from the glass, and future work should include the avoidance of glass apparatus.

The importance of avoiding the use of glass in some electrochemical experiments appears to be generally overlooked. Much of modern electrochemistry could be said to be the electrochemistry of glass containers. Lavoisier showed in 1770 that water only left a residue if evaporated in glass containers, and that if the water was distilled and evaporated in metal vessels it was pure and evaporated with no residue.

Sir Humphry Davys showed by using gold containers that no alkali is formed at the cathode when electrolyzing water, and that the alkali is ordinarily due to the glass containers used. By working in a hydrogen atmosphere with the gold containers he was able to show that pure water is decomposed into hydrogen and oxygen alone, and DO alkalinity is formed at the cathode, or acidity in the anode compartment. It ‘would seem essential for modern electrochemists to be familiar with Davy’s work, and to repeat and extend his experiments.

RELATION TO THE ELECTRODEPOSITION OF METALS AND THE STRUCTURE OF DEPOSITS
The electrodeposition of metals from complex salts such as cyanides is sometimes considered to involve the primary deposition of the alkali metal and secondary formation of the heavy metal coating. Hedges & showed that the electrodeposition of metals from cyanide solutions takes place through a secondary reaction, with periodic hydrogen evolution when the cathode becomes more saturated with codeposited alkali metal. Mercury would not deposit from 0.1N mercuric chloride until 0.1N potassium chloride was added. Periodic codeposition was also observed with mercury and’ sodium, barium, strontium, calcium, and aluminum. Tin deposition from caustic soda solution gave poor, slow periods lasting 15 minutes. This is of the same order of magnitude as the periods of 20 minutes which Snavely calculated to correspond to the laminations in certain chromium deposits.

Fink suggested that the primary deposition of sodium and secondary deposition of metals gave fine grained deposits. - In the discussion which followed, W. Blum and A. L. Ferguson protested that no sodium deposition was possible under usual conditions, without, however,’ considering the possibility of depolarization by the metal of the cathode (alloy formation). The ease of depositing sodium on mercury at low potentials was emphasized by Evans’, in connection with the explanation of tungsten deposition from aqueous solutions by the formation of an alloy. In the same discussion, the senior author suggested that periodic gassing in many plating processes, as in bright cadmium plating, is caused by the charging of the deposit with an alkali metal, giving increased polarization, until finally hydrogen is evolved. The agitation due to the hydrogen evolution brings up fresh solution, decreases the polarization, and the cycle sets in all over again.

It is suggested that something like this is involved in the formation of banded, laminated, striated, or layered deposits frequently observed, especially in connection with the production of bright plates. Thus Meyer and Meyer and Phillips found that bright cyanide copper deposits produced by alloying with a small amount of lead or cadmium had a laminated or banded structure when examined in cross section. They also determined that periodic fluctuations of cathode polarization occurred which corresponded to the thickness of the bands in the deposits. They reviewed a number of previous observations of banded or striated deposits of iron, cobalt, nickel, and copper.

It seems noteworthy that cadmium and lead are also brighteners for nickel baths, and are in addition the highest cathodic overvoltage metals, aside from mercury. It is not meant to suggest that bright copper or nickel deposits containing cadmium or lead also contain noteworthy amounts of alkali or other electronegative metals. These electronegative metals probably participate in the formation of the deposits and determine the cathode polarization and pH of the cathode film, without, however, remaining in the deposit in appreciable quantities. Nevertheless, the reducing power and ability to later give off hydrogen sometimes noted in freshly made deposits, is probably an indication of the presence of some residual alkali metal. The banded structure observed may be due to varying grain size, or varying quantities of inclusions associated with varying pH of the cathode film.

Many studies of the banded structure of nickel deposits have been published but the cause of the phenomenon does not seem to be understood. A current paper gives an excellent review. In his paper on the mechanism of bright plating, Henricks discussed periodic deposition at some length. The present paper merely adds to all this the fact that the periodic deposition of an alkali or similar metal is probably also involved, and in fact is probably a controlling factor.

ACKNOWLEDGMENTS
The authors wish to express their appreciation to United Chromium, Inc., for encouragement and assistance with many phases of the investigation, and particularly to the engineering department for help in the design of equipment and preparation of the paper for publication. They also wish to acknowledge their substantial indebtedness to Dr. Paul Doigan and others at the University of Connecticut where the experimental work was done. Flame photometer measurements were made under the supervision of J. Petrocelli and G. Tatoian at the Patent Button Company, in collaboration with M. Orient and P. Mazzamaro of United Chromium, Inc.