This article was originally published in Electrochemical and Metallurgical Industry Publication of September 1907. Information within this article is therefore correct as of 1907. The publication of this material aims to provide historical insight on the subject and its place in industry.
The two papers of Dr. Walker and Dr. Cushman, published elsewhere in this issue, give us an opportunity to add some further comment to our remarks in the July issue. We have now three important hypotheses on the causes of corrosion of iron. The carbon dioxide theory, originally proposed by Grace Calvert and Crum Brown, and now strongly defended by Mr. Moody, assumes that the corrosion of iron proceeds in two steps, the first of which assumes the presence of carbon dioxide.
The equations of the two reactions are: (1) Fe +H2O + CO2 = FeCO3 + H2 and (2) 4FeCO3 + 6H2O + O2 = 2Fe2O3(H2O)3 + 4CO2. Hence, in the second step carbon dioxide is reformed, which then acts on fresh iron, and so on. The hydrogen peroxide theory also assumes that the corrosion of iron proceeds in two steps, of which the first is the formation of hydrogen peroxide by the reaction (3) Fe + O2+ H2O = FeO + H2O2, while the second step is the action of H2O2 on Fe and FeO, according to the equations (4a) Fe + H2O2 = FeO + H2O and (4b) 2FeO + H2O2 = Fe2O3 (H2O). The electrolytic theory, originally proposed by Dr. W. R. Whitney and now so ably defended and further developed by Dr. Cushman and Dr. Walker, assumes that when an iron surface is in contact with moisture, which may act as an electrolyte, differences in the composition of the iron will result in making some parts of the iron surface anodes and other parts cathodes. Thus, we have short-circuited galvanic cells. At the anode the iron will pass into solution, assuming the ferrous ionic state, while hydrogen is liberated at the cathode, the equation being most easily expressed in the notation of the ionic theory (5) Fe + 2H = Fe + H2. For those who object to the electrolytic dissociation theory, this equation will mean nothing more than that one free iron atom gains two bonds, while simultaneously two monovalent hydrogen ions lose each one bond; expressed in this way the equation cannot be objectionable to anybody. The second step is the oxidation of the ferrous ions to ferric by oxygen of the air. The equation may tentatively be written (for argument's sake, although it is not the most general form) (6) 2FeA + (x + 2) H2O + O = Fe2O3(H2O)x + 2H2A, where A stands for any bivalent anion. Here the two bivalent Fe atoms on the left-hand gain each one positive bond and become three-valent, while the one free ion on the left-hand gains two negative bonds. The loose colloidal ferric hydroxide moves towards the cathode under the influence of the current and piles up there in form of rust.
It is unnecessary to emphasize the differences between the' three theories. It seems more useful to point out how far they agree. All three theories agree that the oxidation of iron by oxygen of the air is not direct, and in this point they agree clearly with all experimentally established facts. Further, all three theories agree that the formation of rust proceeds in essentially two steps, of which the first step is the passing of the free metallic iron into the bivalent state, and that the formation of rust is a subsequent second step, involving the change from the ferrous to the ferric state.
According to the carbon dioxide theory, the first step is the change of Fe to FeCO3, a bivalent compound; according to the hydrogen peroxide theory the change is from Fe to FeO, also a bivalent compound; while the electrolytic theory simply claims the metallic iron to pass into solution and get bivalent. Evidently the electrolytic theory- which leaves the question of the anion entirely open is the more general theory, and in this respect it comprises the other two theories as special cases, representing possible, but not necessary, reactions. The adherents of the electrolytic theory cannot deny that when iron is in contact with carbon dioxide, water and oxygen, the reaction claimed by the carbon dioxide theory will occur; but they deny that it is by means of these special reactions that rust is formed in general, and Dr. Cushman's and Dr. Walker’s repetition of Mr. Moody's experiments appear to offer conclusive evidence in favor of their theory. Viewed from the standpoint of the electrolytic theory, the carbon dioxide theory emphasizes one single possible anodic reaction, namely the change of iron into ferrous carbonate. On the other hand, the hydrogen peroxide theory seems to emphasize to a certain extent one single possible cathodic reaction, if we follow the interesting suggestion of Dr. Walker. It was found by Richarz and Lonnes that the electrolysis of water saturated with air is possible at a relatively low electromotive force, because the air in the water acts as a depolarizer at the cathode, so that the cathodic product is not hydrogen gas but hydrogen peroxide. If hydrogen peroxide can exist at an iron surface, we should expect it to be formed at such spot of the iron which act as cathodes, and this formation (on account of the lower potential compared with hydrogen gas development) would accelerate the passing of iron from the areas into the ferrous ionic state.
While the subject is of highest theoretical interest, it is of no less practical importance. What can be done to overcome the innumerable troubles due to corrosion of iron? Surely Dr. Cushman's suggestion to try treatment with bichromate solutions on a large scale deserves fullest attention. It should be noticed that Dr. Cushman is exceedingly careful and conservative in his predictions, yet he concludes: "Although it is true that laboratory tests are frequently unsuccessful in imitating the conditions in service, it, nevertheless, appears that chromic acid and its salts should under certain circumstances come into use to inhibit extremely rapid corrosion by electrolysis." However, we should not overlook that such treatment is somewhat like treating a sick person continually with medicines and nursing him along in such a way without restoring him really to health. The fundamental trouble in the corrosion of iron lies in the iron itself, in its non-uniformity. Treatment with chromic acid has only a temporary superficial effect and does not change the composition of the iron.
The radical cure to prevent corrosion of steel would be to make better steel, "as free as possible from certain impurities, such as manganese, and so homogeneous as not to retain localized positive and negative nodes for a long time without change." This is quoted from Dr. Cushman, with whom we fully agree, but we may add again what we already pointed out in the July issue, that according to the present state of art it is quite possible that the cheapest way to do this might be to treat open-hearth or Bessemer steel subsequently in the electric furnace. If the corrosion of iron is an electrolytic process and electrochemistry is at the bottom of the corrosion trouble, it would be only proper that another branch of electrochemistry - the electric furnace - should solve the problem.