EFFECT OF D-LU A h l T I O N ON CORROSION Em. 2605 CCCLVIII.-L4n Investigation of the Efject of Digerential &ration on Corrosion by means of Electlrode Potential Neusurements. By A. L. MCAUUY and F. P. BOWDEN. h a recent paper (American Chemical Society Corrosion Sym-posium 1925; Id. Eng. Chem. 1925 17 363) Evans discnsses oxygen distribution as a factor in the corrosion of metals. The present investigation was undertaken with the object of measuring the potentials between solution and metal in the various zones recognised by Evam on a metal corroded by a partly azjrated solu-tion. The potential diilerences in merent zones were very clerrrly distinguished and on attempting to co-ordinate the results by measurements on standard surfaces it was found fhat the method provided a sensitive meam of detecting the commencement of corrosion and recognising the state of the surface.E x P E R I M E N T AL. In the first set of experiments the method of preparing the specimens described by Evans wm adopted. Sheets of zinc and iron were partly immersed in N/lO-sodium chloride and after 24 hours their d a m s were studied. Following Evans four distinct zone8 were recognised on the corroded epecimen (1) A zone unwetfed by the solution. (2) A zone over which the eolution hm crept necessarily well aerated. (3) A zone beneath but close to the surface of the solution also well aerated. (4) A zone of deficient ahtion at greater depths 2606 MCAUIAY AND BOWDEN AN INVESTIGATION OF THE EFFECT Zones 1,2 and 3 were bright and zone 4 waa more or less M y corroded.Measurements were made of the single electrode poten-tial between N/lO-sodium chloride and the various zones. In the second set of experiments the single electrode potentials between N/lO-sodium chloride and standard surfaces were meamred under mering conditions. These surfaces were extremely sensi-tive to exposure either to air or to electrolyte; e.g. to obtain the true potential of freshly deposited electrolytic zinc it was necessary carefully to plate the whole of the exposed surface wash with water, and measure the electrode potential before the zinc had completely dried. If the surface were allowed to become dry for more than a few minuw its electrical character changed completely. Broadly classxed all the surfaces were electrically in one of two standard conditions and the rapid changes were from one to the other of these standard conditions.Method of Neasurement and Appardt~.-Two distinct methods of mewurement were used. In the “ A ” method the specimen was dry and the single electrode potential was measured between the metal and a drop of N/lO-sdium chloride of about Q mm. diameter placed on the metal surface. It is necessary that the drop be very small as otherwise the effects of corrosion by the drop are marked owing to differential aeration. This type of measurement gives the true single electrode potential between solution and metal. In the ‘‘ B ” method of measurement the specimen or the region under investigation was flooded with sodium chloride solution or with water or occasionally was immersed to a depth of 0-5 cm.in the solution and the potential Merence between the electrolyte and the metal surface was measured at different places. This type of measurement gives the electrical conditions while corrosion is in progress. In the case of iron the specimens prepared by Evans’s method were not allowed to dry but were kept wet with water; this was necessary on account of the rapid decrease in the electrode potential of the corroded portion on exposure to air. The measurements were made against a normal calomel electrode, A (Fig. 1). This communicated with a tube By filled with N/10-sodium chloride the end of which was drawn out to a hollow tip with a diameter of about Q mm. This tip made contact with the solution above the metal plate C .The potential difference between the mercury of the calomel electrode and the metal plate was measured by a potentiometer. The apparatus was sensitive to a millivolt. Experirnents.-In the first series of experiments specimens of zinc and iron were corroded by partial immersion in XT,ilO-sodiu OF D- A&LATION ON CORROSION EN. 2m chloride for a period varying from 12 hours to 2 clap. The results of measnrements made on them by methods A and B are d d t with in the next section. In the second series of experiments measurements were made by method A on the following standard surface8 : Zinc freshly deposited electrolytically. Zinc freshly plished with sandpaper. Zinc exposed to a WeU-aGrated solution of NjlO-sodium chloride for 2 days.Zinc the surface of which had been exposed to air for a con-Zinc which had been heady oxidised by heating. Polished and oxidised iron surfaces were also examined. siderable time. As above mentioned all these surfaces fell into one of two broad classes one characteristic of freshly deposited metal which has never been exposed to air and the other characteristic of aiimted surfaces. Under the influence of air or differentially aerated electrolyte one surface would paas rapidly from one class to the other. In measuring the potential between standard surfaces and N/10-sodium chloride it was found that the potential Werence varied with the size of the drop. A small drop on a dry surface (with th 2608 MCAULBP AND BOWDEN AX INVESTIGATION OF TEE EFFECT exception of dried zone 4 on zinc) gave the potential associrtted with one class but if this drop were enlarged the potential immediately began to acquire the characteristic of that of the other class.This was found to be due to electrolytic action taking place between the well-aerated edge of the drop and its poorly aerated centre. The following measurements were made to confirm this result. A large drop from 1 to 2 cm. in diameter was placed on a zinc surface for about a minute and then blotted off and the potentials where the centre and the edge of the drop had rested were measured with a drop about Q mm. in diameter. The edge showed the potential characteristic of an aerated surface and the centre that characteristic of freshly deposited zinc. The latter very rapidly returned to its initial condition and when measured was usually between the two conditions and falling rapidly.To c o h further this result two zinc surfaces were prepared to simulate under very different experimental arrangements the conditions existing in large and small drops on the zinc surface. The specimen corresponding to a large drop was partly immersed in electrolyte and corroded under conditions of differential aeration. The potential of zone 4 was the same as the highest value obtained for the centre of the drop but the zone when visibly corroded did not return to the aerated state for several days. The potential of zone 3 was that of the outside of the drop. The specimen corre-sponding to a small drop was immersed in electrolyte which was well aerated throughout by bubbling air through it.It remained uncorroded and ita surface after treatment for more than 24 hours showed the potential characteristic of the small drop on a well-agrated surface. Discw&m of Results. In the corrosion of a metal immersed in an electrolyte the metal shows an equi-potential surface owing to the small currents flowing. There is however a fall of potential down the solution the positive regioas being situated where the metal is most negative to the solution. Such a region corresponds exactly to the negative plate of a simple primary cell and is called anodic. Figs. 2 and 3 summarise diagrammatically the general nature of the process that takes place. Chlorine ions move towards the anodic region under the influence of the electric field in the solution produced by the grater solution pressure in the corroded region.They there neutralise metallic ions lowering the potential difference between solution and metal and enabling more metallic ions to leave the metal. These ions on emerging restore the potential difference. Sodium ions migrate to the ennobled region and are neutralised by electrons drawn from the metal OF DIFFERENTIBL dRATION ON CORBOSION ETC. 2609 As a b d generahtion two types of d a c e may be recognised c l d e d by the potential between them and N/lO-sodium chloride. In the case of zinc the electrode potential of the first class against the normal calomel electrode is generally within 15 millivolts of -1.075 volts but may be more negative still where the corrosion is heavy.h h electrolytically deposited zinc and zone 4 of the corroded metal are in this class. They are the less noble regions on the corrosion specimens or the more anodic. The second class com-prises zones 1 2 and 3 on corroded specimens surfaces freshly polished with sandpaper specimens exposed to uniformly &mated electrolyte and all surfaces which have been exposed to air for more than a few minutes with the exception of zone 4 on corroded specimens which remains in the h t class even after prolonged FIG. 2. _- ++ 2c1A -2Na $ ++ ++ J-+ ++ ++ fi L+ Zn Zn Zn Zn Zn Zn Zn Oxidised surface. Corroded &ace. Potential diflerence -1.000 volt. Potential difference - 1.075 volt. FIG. 3. ++ ++-- -2Na ZnCG k ++ ++ ++ ++ -!+ ++ ++ Zn Zn Zn Zn Zn Zn Zn Oxidised d a c e .corroded surface. exposure to air. This second class of surface gives an electrode potential against the calomel electrode within about 15 millivolts of - 1.O00 volt. Values considerably less negative than this are obtained with specimens oxidised by heating in air. Surfaces in this second class are the more noble regions on the corroded specimens. In the case of iron few measurements against standard surfaces have been made. Measurements on corroded specimens and on polished specimens show the same general effects as with zinc the two classes of surface described above being recognised and occur-ring under similar conditions. The first class has a potential differ-ence against the calomel electrode of about - 0-540 volt the second class has a potential against the calomel electrode of - 0.340 volt.Iron surfaces of the h t class on exposure to air very rapidly chanse to the second class. This is true for zone 4 in iron which in zinc remains true t o type when exposed to air. VOL. CXXVII. 4 2610 EFFECT OF DIFFERENTLBL A.%RATION ON CORROSION ETC. Figs. 4 and 5 show diagrammatically corroded specimens of zinc and iron respectively. The potential differences as measured by method A against the calomel electrode are shown on the figures. All these potentials given against the calomel electrode include the contact E.M.F. due to the liquid junction N-KCl-N/lO-sodium chloride. No attempt wits made to standardise this contact and capillary tubes and plugs of filter-paper were in-cluded in the circuit.It is the relative rather than the absolute potentials of these zones which are of value. The principal results of the investigation may therefore be sum-marised as follows There are two normal states in which iron and zinc surfaces tend to exist one a more electro-negative state characteristic of pure metal and corroded regions the other a less electro-negative state characteristic of aerated regions. For zinc, the difference in single electrode potential between these states is about 75 millivolts; for iron it is about 200 millivolts. By drastic treatment (heavy corrosion in the first case heavy oxidation in the second) the potentials in these statestbecome more negative and more positive respectively. When drastically treated the surfaces are visibly heavily corroded and in the case of zinc do not rapidly change their condition without further drastic treatment. Experi-ments with drops have shown that surfaces having a clean bright appearance may be in either of the two s t a h and then very rapidly change from one state to the other with change of conditions. UNIVERSITY OF TASMANIA [Received August 7th 1925.1 HOBART