METALS, CORROSION OF, the spontaneous chemical transformation of a useful metal into useless chemical compounds. Most environments, whether liquid or gaseous, cause metals to corrode, and since these natural corrosive forces are constantly at work, rusted steel structures, ``rotted out'' automobile bodies, pitted chromium plate, and the like, are common sights. In these examples the surface of the metal is visibly attacked, but a definition of corrosion must also recognize cases in which the attack is highly localized, for example, in the boundaries between metal crystals. This so-called structural corrosion represents a more insidious form and can cause unforeseen accidents. Such unexpected failure of the metallic component is often compounded by applied stresses, particularly fatigue stresses.
A few examples of corrosion are not of the destructive type. For example, the green patina frequently observed on bronze statuary is an alkaline oxide of copper that effectively protects the underlying metal from further atmospheric attack. This has accounted for the excellent state of preservation of many ancient bronze and copper pieces.
Corrosion is combated by the application of protection methods involving well-established scientific principles, but in many respects it continues to be an important general engineering problem. Corrosion of iron and steel is naturally the most important, and it is estimated that about two percent of the total tonnage in use is lost annually.
Electrochemical Nature of Corrosion.
Michael Faraday's work in the period 1830-1840 proved the connection between chemical reactions and electric currents, and provided the basis for electrochemical theories of corrosion. Faraday explained certain corrosive properties of iron on an electrochemical basis, but no detailed understanding of corrosion was obtained until the early twentieth century.
Electrochemistry itself dates from the discovery in the eighteenth century by Alessandro Volta of the voltaic cell, which gives a continuous current of electricity through the transformation of chemical energy into electrical energy. Such a cell consists essentially of two different metals (the ``electrodes'') partly immersed in an aqueous solution (the ``electrolyte'') capable of conducting electricity. The electrodes are connected together outside the electrolyte by an electrical conductor (a wire). One electrode (the ``anode'') dissolves (corrodes) in the electrolyte forming positive ions that flow into the solution, while hydrogen ions accumulate at the other electrode (the ``cathode''). This positive ion flow is compensated in the external part of the circuit by an electron current flowing from anode to cathode (see Fig. 1). The metal ions entering the solution frequently react with substances in solution to give corrosion products. These products are often soluble and do not hinder further corrosion. In view of the fact that the basic mechanism involves only an electrolyte and two dissimilar electrodes, it is apparent that if adjacent areas on, for example, a steel surface differ slightly from one another with respect to composition or structure, a corrosion cell will form in a suitable environment. One area will be anodic with respect to the other, and will be subjected to corrosion. Since all small local inhomogeneities result in anode-cathode areas, the metal surface contains numerous potential corrosion sites. If the steel is immersed in ordinary water or almost any aqueous liquid, a ready supply of electrolyte is available. Even a moderately humid atmosphere will provide sufficient condensation on the metal surface to complete the electrochemical cell.
FIG. 1. THE VOLTAIC CELL, which illustrates the natual corrosion process. In metals, this process is basically electrochemical. Just as in the voltaic cell, natural corrosion occurs with the passage of electrons from one type of metal to another.
The voltaic cell just described consists of two half-cells, an individual electrode immersed in an electrolyte constituting a half-cell. The potential (electromotive force) of the voltaic cell is equal to the difference in potential of the two half-cells. Electrode potentials are measured by comparison with that of hydrogen. The measured electrode potentials of metals are tabulated in a potential series, in which the ``noble'' metals, gold, platinum, silver, etc., appear at the positive end. These metals are said to be electropositive, that is, their electrode potentials are strongly positive with respect to that of hydrogen. The ``base'' metals, magnesium, aluminum, etc., are strongly electronegative. To a first approximation the position of a metal in the potential series is an indication of its susceptibility to corrosion, the more electropositive metals being more immune. (See Electrochemistry; Electrolytes.)
Movement of positive (e.g., hydrogen) ions toward the cathode results in hydrogen gas forming at the cathode. This changes the potential of this electrode, and a reverse potential is set up which reduces the over-all cell voltage. The cell current falls quite rapidly to a very low value, and the cell is said to be ``polarized.'' This implies a reduction or even cessation of corrosion, but oxygen in the electrolyte can combine with the hydrogen and nullify its effect. The oxygen is called a ``depolarizer.'' In stagnant waters, the shortage of oxygen may result in reduced corrosion rates because of polarization effects. Such cases are rather uncommon because turbulence in the liquid environment is usually sufficient to provide adequate oxygen at the cathode surface.
An important type of corrosion results when the distribution of the depolarizer, usually oxygen, is not uniform over the metal surface. This differential oxygenation results in the formation of an oxygen concentration cell, and causes corrosion in the same way as will any type of voltaic cell.
Passivity and Other Anodic Effects.
The term passivity was originally used with reference to the immunity to corrosion developed by iron on immersion in strong nitric acid. However, the effect is quite general and many metals become passive under appropriate conditions. The phenomenon was correctly explained by Faraday in 1836, and is due to an extremely thin oxide film that is formed by a chemical reaction on the metal surface. Such a film can be reduced (chemically altered) and the metal made active by contact with a more electronegative metal, for example zinc in the case of passive iron. This creates a bimetal couple in which the passive metal is cathodic. The hydrogen liberated at the cathode reduces the oxide film to metal.
Aluminum forms an oxide film which, under many conditions, is impervious to corrosive attack, and anodized aluminum, formed by an anodic oxidation process, is much used for both decorative and utilitarian purposes. In the widest chemical sense all anodic processes undergone by metals are oxidations, but the term ``anodic oxidation'' implies the deliberate formation of a substantial amount of solid oxide. This quite thick film is produced by making the aluminum the anode in a cell that usually contains sulfuric acid or phosphoric acid as the electrolyte. Numerous patents describe various modifications to the process. The anodized surface is originally porous and can be dyed to any desired color. Immersion in a solution of potassium dichromate gives a bright orange-yellow, while potassium ferrocyanide, lead permanganate, and cobalt sulfide give hues of blue, reddish-brown, and black, respectively. In many instances water-soluble organic dyes are also used, and these impart an attractive metallic luster to the colored surface. In all cases, the surface must be subsequently sealed; this can be accomplished by treatment with boiling water alone, although boiling solutions of nickel acetate or cobalt acetate are also used.
Various alloys, particularly many aluminum alloys, develop their maximum strength by a heat treatment known as ``age-hardening.'' This involves the formation of submicroscopic precipitate particles. Under some circumstances the precipitate forms along crystal boundaries in the alloy so that the region immediately adjacent to the boundary becomes anodic with respect to the interior of the crystal. In a corrosive environment the crystal boundaries will be preferentially attacked, and corrosive fissures are likely to extend deeply into the metal structure. This ``structural corrosion'' seriously affects mechanical properties. It can be prevented either by correct heat-treatment methods or by protecting the metal surface with a corrosion-proof coating. ``Cladding'' is a process in which a high-strength alloy is rolled between thin sheets of pure aluminum, thus sealing the alloy. In this way the composite metal is made free of corrosion, while the cladding results in only a small reduction in mechanical properties.
Prevention of Corrosion.
Various methods of corrosion prevention are available, and the choice of any method depends on consideration concerning the particular structure or component to be protected. It has been previously noted that electrochemical corrosion is so dangerous because the corrosion products are often soluble in the solution and do not hinder further corrosive attack. But it may sometimes be possible to add chemical substances to the solution which react with the primary corrosion products to form insoluble, and therefore protective, compounds at either the anode or the cathode. For example, iron corrodes readily in a dilute solution of common salt, but addition of zinc sulfate to the solution results in the formation of zinc hydroxide at the cathode, and this is mostly insoluble. Alternatively, addition of sodium phosphate forms insoluble ferric phosphate at the anode. These are examples of cathodic and anodic inhibitors respectively. Such protection methods can only be applied when the structure is wholly or partly immersed in a liquid corrosive medium.
Cathodic protection is often used to reduce the rate of corrosion. In this method, an electric potential is applied in such a way that the entire structure to be protected is made cathodic. This is accomplished by connecting the structure to one pole of a rectifier or a generator, while the other pole is connected to an external, chemically inert anode such as graphite. As applied to pipe-line protection, for example, the unattackable anode is buried in the ground a short distance away from the pipe line. Separate auxiliary anodes are an occasional addition to this process. Examples include the protective anodes suspended inside water-storage tanks or buried adjacent to underground pipes. In these examples the water in the tank and the moist earth surrounding the pipe act as the respective electrolytes.
Other methods of cathodic protection do not involve impressed currents. The problem, as before, is to supply sufficient current from some source to the structure (which becomes completely cathodic) so as to hold local anodes and cathodes at the same potential. This can be achieved by attaching a metal that is high in the electrochemical series to the metal to be protected, establishing a galvanic couple. ``Sacrificial'' zinc anodes have been used since 1825 when Sir Humphry Davy suggested they would protect the copper-sheathed wooden hulls of ships. Magnesium alloy anodes are currently used to protect ships from sea-water corrosion. Sacrificial anodes offer an advantage over impressed current methods in that they do not require an extra source of current to energize them.
Paints are of course widely used for corrosion protection, particularly when the structure is not actually immersed in a liquid. Here again the purpose is to inhibit either the anodic or the cathodic reaction, and so inhibit the over-all chemical reaction.
The use of sprayed metal coatings and electroplated surfaces must also be mentioned. The purpose may be ornamental as well as protective, as in chromium plate.
SPECIFIC CORROSION TYPES
as the name suggests, involves the deterioration of a metal under the combined action of static stress and corrosion. The basic mechanism seems to be one of the initial formation of corrosion pits and crevices, and subsequent fracture of the component due to stress concentrations associated with these crevices. The detailed mechanism is complicated and not well understood. If the stress is a residual stress, perhaps resulting from a metal-forming process, failure due to the combined action of such stress and corrosion is termed ``season-cracking.'' Pure metals are not susceptible to stress corrosion. However, this is quite common in brass components. In the case of alloys, stress corrosion cracks frequently follow crystal boundaries. This suggests that solute atoms tend to move to crystal boundaries, making them anodic with respect to the grain interior; this enhances crystal boundary corrosive attack and facilitates subsequent crystal boundary cracking.
also involves the simultaneous action of stress and corrosion. However, cyclic stresses, or fatigue stresses, are far more damaging than static stresses. Even in the absence of corrosion, fatigue failure often occurs unexpectedly, and the damaging effect of corrosion crevices, which represent stress concentration sites, under fatigue loading is obvious. Probably all so-called fatigue failures involve corrosion to a certain extent, since it is almost impossible to completely exclude superficial corrosion.
Corrosion by Liquid Metals
is rather a special form of corrosion in that it does not involve an electrochemical mechanism. It is of considerable importance in liquid-metal heat transfer systems, and therefore in nuclear reactor technology. Liquid metals useful as heat transfer media include sodium, potassium, and their alloys; also lead, bismuth, and lead-bismuth alloys. Most structural metals and alloys are attacked to at least a slight extent by these liquids, and several quite different corrosion mechanisms can be distinguished.
First, the container or pipes in the flow system may be relatively uniformly dissolved in the liquid metal. The actual solubility may be quite small, but, as it usually varies with temperature, material may be deposited out of solution in cooler parts of the system and may eventually block the flow channels, or valves. This is known as thermal gradient transfer of material.
Secondly, intergranular penetration might occur if there is a selective reaction with minor constituents in the solid. As in electrochemical intergranular corrosion, the mechanical properties are affected without the appearance or weight being changed. Such cases of attack are fortunately rare.
Thirdly, the liquid and solid metals might interact to form an alloy-surface scale which may or may not act as a diffusion barrier against further attack.
``Corrosion erosion'' refers to a mechanical attack by turbulent liquid metal. In an extreme case this leads to actual cavitation, the process then being known as ``cavitation erosion.'' See Cavitation.
Irradiation Effects in Corrosion.
This subject has been intensively investigated since the inception of the atomic energy program, but there is very little information in the open literature. The commonly used term ``radiation damage'' refers to all changes of a mechanical, physical, or chemical nature in solid materials, which are brought about by irradiation with one or more of the following: ionizing radiation (X rays or gamma rays), light charged particles (electrons), heavy charged particles (alpha particles), or heavy uncharged particles (neutrons). The bombardment of a metal with high-energy, heavy charged or uncharged particles is known to result in atomic-scale imperfections being introduced into the metal, and in appropriate circumstances these can be sites for electrochemical reactions. But changes in the metal's environment, rather than in the metal itself, are more important. These indirect effects are due to ionizing radiations such as gamma rays that induce no permanent changes in metals. For example, in aqueous solutions both highly reactive free radicals and hydrogen peroxide may be produced, and it is to be expected that such materials will tend to increase corrosion rates. In addition, corrosion inhibitors such as sodium dochromate undergo reduction and lose their effectiveness. Another source of corrosion is through the effect of ionizing radiation on oxide films, which are ionic and sensitive to gamma rays. It is evident that the over-all behavior is very dependent on the precise circumstances attending the corrosion.
Oxidation of Metals.
Most metals combine with oxygen gas at all temperatures to form stable oxides of the metal in question. The rate at which the oxidation proceeds is very dependent on the temperature, and at normal temperatures usually only a thin film of oxide will be formed on the metal surface. (On copper, for example, this will be evident as a superficial tarnishing.) At higher temperatures oxidation is likely to proceed more rapidly. The noble metals on the whole represent exceptions to this behavior, and possess little affinity for oxygen. It is believed that gold does not oxidize at all when heated in air or oxygen, while any slight oxidation that platinum may undergo at temperatures up to 450°C is removed by heating to higher temperatures. However, the structurally useful metals oxidize to an extent which depends on the mechanism of the process. Four classifications of oxide formation can be distinguished: volatile, dense, protective, and nonporous.
Volatile oxide formation.
A few high-melting-point metals such as tungsten and molybdenum become brittle at high temperatures and form oxides which are volatile. Therefore a protective oxide layer is not formed, and the metals must be protected by a nonoxidizing atmosphere when used at high temperatures.
Dense oxide formation.
The ultra-light metals have, as a rule, oxides which are too dense to occupy the whole volume of the metal destroyed. Such oxide layers are likely to be porous and will not protect the underlying metal. For this reason, magnesium oxidizes very readily.
Protective oxide formation.
Some metals, not necessarily chemically related, form oxides which are moderately protective under some conditions. The oxide film on aluminum completely covers the metal, but seems to be in a state of compressive stress since cracks develop, possibly on account of changes in temperature and humidity. This type of behavior is confined to relatively low temperatures.
Nonporous oxide formation.
Many ``heavy metals'' form oxides which are not inherently porous and do not tend to crack. The oxides formed on copper, iron, and nickel are of this type, but yet offer no protection to the underlying metal. From the theoretical viewpoint these oxides are most interesting and have been thoroughly investigated. It is established that they contain less than the stoichiometric amount of metal, which is to say that some of the metal atoms are missing, leaving holes in the oxide lattice. Consequently, atoms can diffuse through the lattice, and the oxide layer is continually able to increase in thickness.
Use of Alloys.
Since all the common structural metals are subject to oxidation, components for use at high temperatures in oxidizing environments must be made of alloys which contain an oxidation-resistant metal as a major alloying constituent. Chromium is a reasonably low-cost metal (available as ferrochromium) which meets the requirement, and is found in almost all high-temperature alloys that are required to resist oxidation. Therefore, in addition to their stainless characteristics, all stainless steels have good resistance to oxidation and are used in numerous domestic and industrial appliances where such behavior is necessary. The alloy nichrome, which is extensively used for electric furnace windings, consists of 80 per cent nickel and 20 per cent chromium, and is reasonably oxidation-free up to about 1000°C.
Resistance to oxidation is not necessarily the only property of interest. Mechanical properties are often of importance, and it so happens that certain alloying elements such as chromium serve a dual purpose in giving both high-temperature strength and resistance to oxidation. As a consequence of this, the problem of high-temperature oxidation did not introduce serious difficulties in the development of the gas turbine engine until attempts were made to use residual fuel oils containing vanadium or sodium. These impurities, together with the sulfur in the fuel, produce liquid products which have proved to be very corrosive. In attempting to solve this problem, the approach has been to develop addition compounds to be added to the oil, which will form harmless volatile substances with vanadium and sodium.
does not involve electrochemical attack or direct gaseous oxidation, but is partly a mechanical effect. It might therefore be described as pseudo-corrosion, but it has important practical implications. The term ``fretting corrosion'' originates in its application to contact surfaces which are subjected to minute vibrations, typical examples being the damage of gas turbine blade root grooves by vibration and rubbing in the compressor disk, and the wearing away of gear teeth, etc. If there is no rubbing of surfaces, no fretting occurs, but the term does not apply to large-scale rubbing motions which give predominantly typical wear. The damage seems to be increasingly severe the closer the two surfaces originally were, and this means that there is a tendency to loosen needed tight fits. In addition, notches formed during the process are possible locations for the start of fatigue failures. The fretting corrosion of steels is easily recognized by a brownish-red debris, evidently an oxide, but the over-all mechanism of the process is uncertain at present. See Metals; Metals, Heat Treatment of; Metals, Mechanical Properties of.
Copyright © 1996 P.F. Collier, A Division of Newfield Publications, Inc.
Richard A. Dodd, METALS, CORROSION OF,, Colliers Encyclopedia CD-ROM, 28 Feb 1996.
Our LifeChem Web is a Next Generation Technology in Chemical Maintenance that puts you in control of your maintenance needs! FREE Information, Free use, get it right now by clicking right here! our other users can't be wrong.