US2464922A - Prevention of pencilling corrosion of metallic salt bath electrodes - Google Patents

Description

Patented Mar. 22, 1949 PREVENTION OF PENCILLING CORROSION OF METALLIC SALT BATH ELECTRODES Harry Rollason Copson, Cranford, N. 1., assignor to The International Nickel Company, Inc., New York, N. Y., a corporation of Delaware No'Drawing. Application November 13, 1947, Serial No. 785,794

8 Claims.

The present invention relates to a method for improving operational characteristics of molten neutral salt baths used in heat treating furnaces of the type heated by an alternating current passing through the molten salt between metallic electrodes; and, more particularly, to the prevention of a peculiar type of electrolytic corrosion which has been found to occur on the surfaces of the metallic electrodes used to introduce the heating current into the bath.

As is known to those skilled in the art, neutral heat treating salt baths of the type heated by alternating current passing therethrough between metallic electrodes are those which do not strongly tend to either carburize or decarburize steel or other ferrous base alloys 'heat treated therein. The term neutral salt baths" is used in the art to denote baths comprised predominantly of chlorides of the alkali and alkaline earth metals, and to mixtures of these salts, although alkali and alkaline earth metal fluorides and carbonates which are stable in molten salt baths within the temperature range of 1300" F; to about 2200 F. may also be present in the baths. Neutral salt baths based on the alkali metal chlorides are employed in heat treating steel in the medium temperature range of from about 1300 F. to about 1800 F., while neutral salt baths comprised chiefly of barium chloride are used in heat treating steel at temperatures above about 1800 F., e. g., temperatures up to about 2200 F. The neutral salt baths are usually composed of two or more salts mixed in proportions approximating the eutectic or lowest melting composition. The operating temperature range for a particular bath usually will be between a temperature sufficiently above the melting point of the bath to provide fluidity and a temperature beyond which excessive fuming or thermal instability is encountered. Thus, the operating tem-- perature range for each bath will vary according to the melting point of the bath.

In neutral salt bath furnaces of the type heated by alternating current passing through the salt between metallic electrodes, it has been found that a peculiar and unpredictable type of electrolytic corrosion attacks the surfaces of the metallic electrodes exposed to the neutral salt bath when current isflowing. This type of corrosion results in a comparatively uniform consumption of the electrode, and is frequently of such severity that electrodes have been known to fail after a very short period in service, e. g.. about two' weeks or even less, whereas the normal life expectancy of electrodes in this service is on the order of six to twelve months or more. Because corrosion of this type usually tends to attack the exposed electrode surface uniformly. to cause the electrodes to become rounded in cross-section and to have a smooth surface, it has been been called pencilling corrosion. It has been found that when the electrodes are suffering severe electrolytic corrosive attack (or pencilling) and rapid failure therefrom, thermocouple protection tubes, metal pots, etc., of the same metal immersed coincidentally with the electrodes, but not subjected to alternating electric current, suifered no attack or a very limited type of attack which was confined chiefly to the metal area exposed to the salt-air interface.

It is an object of the present invention to substantially preventthe severe electrolytic corrosion, such as pencilling corrosion, of metallic electrodes exposed to the action of alternating current in neutral salt baths.

It is a further object of the present invention to prevent pencilling corrosion of metallic electrodes in neutral salt baths heated by alternating current passing therethrough without deleteriously affecting steel work or other ferrousbase alloy articles being heat treated in such baths.

It is another object of the present invention to provide a method whereby neutral salt baths may be operated at average current densities above about 7 amperes per square inch without the electrodes undergoing pencilling corros on,

or other similar electrolytic corrosion, resulting" the present invention which contemplates maintaining in the aforementioned molten neutral salt baths a small amount up to about 2500 parts per million of chromate radical while employing electrodes, for the purpose or introducing the heating current into the bath, which are made of a chromium-containing alloy of one or more metals of the iron group and operating the said neutral salt baths at average current densities of at least about 7 amperes per equare inch and preferably between about 7 amperes per square inch and about 36 amperes per square inch. In speaking of chromate it is to be understood that any chromate soluble in the molten salt bath may be employed. Compounds convertible to chromates soluble in the bath, e. g., dichromates, may also be employed. I prefer to employ an alkali metal chromate.

The neutral salt baths employed in the process of the invention are preferably those comprised principally of alkali or alkaline earth metal chlorides or mixtures of these chlorides. Particularly preferred baths are those based on the chlorides of sodium, potassium and barium or mixtures of these chlorides. For example, I have obtained good results in baths comprising 56% potassium chlorides and 44% sodium chloride, and 77% barium chloride and 23% sodium chloride.

The electrodes employed in the process of the invention are made of an alloy containing at least one metal of the iron group and about to about 35% of chromium, particularly nickelchromium-iron alloys containing about to about 22% chromium, about 7% to about 82% nickel and about 0.1% to about 80% iron. Preferred alloys are those nickel-chromium-iron alloys containing about 7% to about 80% nickel, about 12% to about 20% chromium and about 5% to about 75% iron. In addition, these alloys may contain up to about 2% molybdenum, up to about 0.3% copper, up to about 2% manganese, up to about 3% silicon, up to about 0.3% carbon, up to a total of about 4% of titanium, columbium and/or aluminum, and small amounts of impurities and deoxidizers such as phosphorus, sulfur, lead, zirconium, etc. Illustrative electrode compositions are set forth in Table I.

Table I Alloy Per Per Per Per Per Per No Class Cent Cent Cent Cent Cent Cent Ni Cr Mn Si 0 Fe 1 80% Ni, 14% Cr, 77.5 1 1.4 0.2 0.1 0.1 Ba].

2 65% Ni, Cr.-- 62 13. 6 2 0.1 0. 2 B81.

3 18% Cr, 8% Ni 9. 8 17. 9 0. 2 0. 2 0. 05 B81.

It has been found that pencilling does not occur in fresh neutral salt baths. However, as is well known in the art, molten salt baths soon become contaminated in use. For example, decomposition of the salts, impurities introduced from work going through the bath and from the atmosphere surrounding the bath, corrosion products from the electrodes and/or from the metal pot (if one be used to hold the salt) etc., may contaminate the bath composition markedly and rapidly in use. It is in neutral salt baths which have become impure or contaminated through use that the destructive electrolytic corrosive attack upon the electrodes becomes marked, and in which the process of the invention effectively prevents this attack. Since the electro-chemical reactions occurring at the interface between electrode and salt when an alternating current is flowing are very complex, it is not proposed herein to advance any general theory relating to the cause of electrolytic corrosive attack, such as pencilling corrosion.

However, it is known in the art that salts such as the alkali metal chlorides employed in making up neutral salt baths tend to decompose to oxides such as the alkali metal oxides during the course of operation of the baths. It has been found that the presence of a small amount of alkali in a neutral salt bath, e. g., as alkali metal carbonate, hydroxide or oxide, will produce pencilling of chromium-containing alloy electrodes under conditions which will produce very little or no pencilling in the fresh neutral salt bath. Thus, when two electrodes made of a nickel-chromium-iron alloy containing about 60% nickel, 14% chromium and the balance essentially iron (such as is sold under the trade mark Inconel) were employed to pass alternating current at an average current density of about amperes per square inch of immersed electrode area for about 18 hours, through a fresh, molten, neutral salt bath comprised of about 56% potassium chloride and about 44% sodium chloride when the bath was at a temperature of about 1500 F., the electrodes were corroded at the low rate of only 0.5 gram per square inch of exposed surface per day. However, when 1% sodium carbonate (equivalent to about 0.59% sodium oxide) was added to a similar fresh bath operated at the same temperature and exposed to about the same current density for 12 hours, similar electrodes were corroded at the high rate of about 3.2 grams per square inch of exposed surface per day. The weight loss in the second test in a neutral salt bath having only a small added alkali content was more than six times as great as in the fresh bath, and was at a rate sufliciently high to produce rapid electrode failure. Additions of even smaller amounts of sodium carbonate (e. g., down to 0.2% or even less) have been found to accelerate electrode corrosion in neutral salt baths to about the same extent as the 1% sodium carbonate addition. When the small amount of only 490 parts per million of chromate radical in the form of potassium chromate was added, in accordance with the invention, to the aforementioned bath containing 1% sodium carbonate, the corrosion rate upon similar electrodes was reduced to only 0.1 grams per square inch per day in a test conducted for 18 hours under similar conditions to those in the test which had produced the excessively high corrosion rate of 3.2 grams per square inch per day when no chromate was added to the bath.

In order that those skilled in the art may have a better understanding of the process of the present invention, data relating to specific examples of the process of the invention have been set forth in Table II. In the tests on which these data are based, the electrodes employed in tests 1 to 3 and 7 to 9 were made of alloy 1, while those employed in tests 4 to 6 were made of alloy 2. In tests 1 to 6 the bath was comprised of 56% potassium chloride and 44% soduim chloride while in tests 7 to 9 the bath was comprised of 77% barium chloride and 23% sodium chloride. In all these tests an addition of 1% sodium carbonate was made before the start of the test to provide a salt bath which would normally produce severe electrolytic corrosion and the baths were operated at about 1500 F. at the current densities and with the various chromate additions shown in Table II. The test data were obtained using pairs of square, closely-spaced electrodes and Table II sets forth the average corrosion rate (in terms of weight loss in grams per square inch per day) of the electrodes under the various test conditions with all tests being run for about 18 hours except test 2 which was run for 6 hours.

Table II Aver. O. D.-average current density (alternating current).

P. P. M.parts per million oi chromate (added as potassium chromate).

The data in Table II illustrate how electrode corrosion is greatly reduced by additions of only small amounts of chromate radical, in accordance with the invention, to the neutral salt bath. This protective effect is obtained at current densities ranging from about 7 up to at least 36 amperes per square inch. A comparison of the data of tests 1 and 2 illustrates the general rule that increasing the current density in the absence of chromate radical rapidly increases the corrosion rate. When these tests are compared with test 4 and tests 3, 5 and 6, it may be seen that, at a given current density level (e. g., about 20 amperes per square inch as in tests 1, 3, 4 and 5) the chromate radical is very effective in reducing the electrode corrosion rate; and that, by comparing the results of tests 2 and 6, this protective effect is still pronounced even though the average current density in these tests was approximately double that of the previously described tests. Tests 7, 8 and 9 are illustrative of the fact that the process of the invention is eiiective in a neutral salt bath in which barium chloride is the main component.

The greatly reduced electrode corrosion rate that is obtained in accordance with the process of the present invention in salt baths at various operating temperatures is illustrated by the data set forth in Table III. The data are based on tests conducted using electrodes made of alloy 1 carrying current at an average current density of about 20 amperes per square inch. The electrodes were immersed in a neutral salt bath containing about 55.5% potassium chloride, 43.5% sodium chloride and 1% sodium carbonate and operated at the temperatures indicated in the table. It will be noted that while the rate of corrosion, e. g., pencilling, increases as the operating temperature of the neutral salt bath increases, the addition of chromate in accordance with the invention greatly reduces the electrode corrosion rate at each particular temperature.

It is preferred that the amounts of chromate radical employed in the process of the present invention be maintained in the salt bath in quantities between about 100 and 500 parts per million.

I have found that a chromate radical content substantially less than about parts per million is not effective in preventing pencilling and amounts in excess of about 500 parts per million may occasionally cause excessive decarburization in some applications. It is to be understood that larger quantities of chromate up to about 2500 parts per million or even more can be employed without harmfully affecting the life of nickelchromium-iron alloy electrodes as the upper limit upon the amount of chromate which is permissible in a particular application is the point beyond which excessive decarburization for that application is encountered in the steel or other carboncontaining alloy being treated. It is known in the art that chromates are oxidizing agents and that they would be expected to be decarburizing to steel. Nevertheless, it has now been discovered that chromates can be employed in molten neutral salt baths in amounts suflicient to substantially inhibit or prevent pencilling of nickelchromium-iron alloy electrodes without encountering objectionable decarburization of the steel work or other work being treated coincidentally in the bath. This discovery is of great importance to the heat treater who is concerned, in operating salt baths, with preventing or inhibiting decarburization of the steel or other alloy he is treating. Of course, the amount of decarburization permissible varies with the operation. As is known to the art, the extent of decarburization encountered in any molten salt bath is dependent upon the type of steel being treated. Thus, other conditions remaining the same, more decarburization will be produced in a high-carbon steel than in a low-carbon steel when both steels are treated in a bath having a decarburizing tendency. For example, a test piece of SAE 1065 steel sufiered practically no decarburization at a depth of about 0.008 inch under the exposed surface when exposed for the excessively long time of one hour at 1500 F. to a salt bath comprised of about 56% KCl and 44% NaCl and to which 500 parts per million of chromate had been added and under similar conditions another test piece of like material lost about 0.05% carbon at a depth of about 0.008 inch when exposed to a bath initially containing 2500 parts per million of chromate. However, pieces of SAE 1095 steel showed a carbon loss of about 0.04% and about 0.05% at adepth of about 0.008" when exposed under like conditions to similar baths initially containin about 500 parts per million and 2500 parts per million of chromate respectively. It has been considered in the art that a carbon loss in heat treating of about 0.05% at a depth of about 0.010 inch is ordinarily permissible in many applications. On this basis, the presence of about 2500 parts per million of chromate appears to be a practical upper limit. Of course, in cases where more decarburization would be permissible, e. g., where the work is to be ground after heat treating or where low or medium-carbon steels, such as steels containing less than about 0.5% of carbon, are being treated, the practical upper limit of the chromate radical concentration employed can be higher than about 2500 parts per million. The larger chromate radical concentrations do not deleteriously afiect electrode life, and the use of these larger concentrations of chromate is governed by the amount of decarburization permissible in the work being processed.

As pointed out hereinbei'ore, it is preferred that the chromate content be maintained at least as high as about 100 parts per million as a chromate content substantially less than this amount has been found not to be effective in protecting electrodes made of alloys of chromium with metals of the iron group. Thus, any condition in the bath which tends to consume chromate will usually tend to promote rapid pencilling of the electrodes unless chromate is added to compensate therefor. For example, it has been found that electrode sludge, which is largely metallic and has a high surface area, tends to consume chromate rapidly and to promote pencilling if the chromate is not replenished. Thus, it is preferred that electrode sludge not be permitted to accumulate in the bath.

It is not to be inferred that because of the use of chromates in accordance with the invention the usual deoxidation or recitification procedures, known to those skilled in the art, and employed heretofore for the purposes of deoxidizing the baths, of removing sludge from the baths and of preventing decarburization of steel work processed in molten neutral salt baths, can be dispensed with. On the contrary, these prior art practices may still be required to prevent an excessive decarburizing effect upon steel work being processed in baths wherein the process of the invention is carried out. The present invention is directed toward preventing the vexatious problem of pencilling corrosion and similar corrosion upon metallic electrodes exposed to the bath and not to the entirely different consideration of preventing decarburization of steel or other work. In carrying out the invention, it is much preferred that approved rectification practices be observed. Certain rectification practices have the effect of removing chromate from the bath. In such cases, it will be necessary to reestablish the desired chromate radical content after rectification. In addition, since the small chromate addition contemplated by the invention is oxidizing in nature, other additions to the bath having a reducing effect are not compatible therewith. Thus, cyanide additions, etc., are not compatible with the process of the invention.

The chromate content of the molten bath can be determined colorimetrically by matching the color of the bath with known color standards. In the case of baths comprised of mixed potassium and sodium chlorides, the molten chromate-containing bath is yellow; while, in the case of baths based on barium chloride, the color tends to be greenish in character. Thus, loss of chromate from the molten salt bath is easily detected, giving ample warning that adjustment of the bath composition should be made either by adding chromate, or removing chromate-consuming sludge from the bath, or rectification and removal of the products thereof, etc. This is of great importance in operation because of the discovery that, once conditions favorable to pencilling have become established in the bath, electrode pencilling proceeds at a rapid and accelerating rate.

Under certain conditions in neutral salt baths, it has been found that reactions of an obscure nature take place which cause rapid consumption of chromate or rapid conversion of chromate to a chromium compound ineffective for purposes of the present invention. It has been found that this type of reaction occurs in baths containing calcium chloride, c. g., a bath containing 81% calcium chloride and 19% potassium chloride. Consequently, baths containing large amounts of calcium chloride are not as preferred as other neutral salt baths disclosed herein. However,

baths containing calcium chloride are not widely used in the art because of the fact that this salt is hygroscopic. This property of the salt makes it difiicult to add to molten baths and leads to pitting of steel work processed'ln such baths from which the salt is incompletely removed.

As indicated hereinbefore, the mechanism which produces or promotes pencilling of electrodes used for heating molten salt baths with alternating electric current is so complex in nature that it is not fully understood. However, it is believed that under operating conditions leading to pencilling, metal tends to dissolve from the electrode during the anodic part of the AC cycle and to plate back on to the electrode as a non-adherent powder during the cathodic part of the cycle. It has been found that a metallic sludge forms in the bath beneath the electrodes when pencilling is taking place. Should any chromate be present in the bath when this metallic sludge forms, it will be rapidly consumed. Thus, once electrode pencilling commences, it is accompanied by rapid decrease in the chromate content of the bath which in turn leads to a rapidly increasing rate of pencilling corrosion when the chromate content becomes too low. It has been found that the pencilling corrosion rate tends to increase with the length of time the bath has been used. It is believed that the increase in the corrosion rate with time can be partly explained on the basis of compositional changes which have been found to take place in the bath with the length of time in use. Thus, when baths which initially contained known amounts of sodium carbonate were analyzed after test runs, it was found that the ratio of the sodium oxide equivalent to the carbon dioxide equivalent had changed in such a manner that the ratio was no longer that corresponding to sodium carbonate. The effect of this change was to provide an excess sodium oxide content in the bath not balanced by a carbon dioxide content, as in sodium carbonate. The proportionate increase in sodium oxide appeared to increase as the bath containing the constituents was run, and it is believed that the increase in the excess sodium oxide content is a factor tending to increase the electrode pencilling rate. Another factor tending to increase the average electrode pencilling rate upon continual exposure of electrodes to a particular bath with a constant current flowing is the fact that as pencilling proceeds, the current density automatically increases and thereby accelerates the pencilling corrosion rate. Other factors and conditions may influence the rate of pencilling corrosion. It is believed that the function of chromate in the process of t e inventon is to establish an adherent protective film on the surface of the electrodes which substan ally reven s penc ll ng.

It will be apparent to those skilled in the art that the current density applied upon the exposed electrode surface can be controlled in a number of ways. For example, in furnaces of the Hultgren type which are heated by alternating current passing between pairs of closely spaced electrodes, such expedients as merely lengthening the electrodes, increasing the number of pairs of electrodes, etc. can be employed in reducing the current density effective upon the faces of the electrodes exposed to the salt. Inasmuch as there is a concentration of current upon the corners and facing surfaces of the electrodes, the maximum current that can be employed in the invention can be increased in some cases by merely changing the shape of the electrodes so as to increase the area of the facing surfaces without increasing the total area of the electrodes. The term "average current density" is used herein to mean the current density equivalent to that obtained upon an electrode having a square crosssection and is computed by dividing the current passing through the bath by the average total exposed area of one of the electrodes in the molten salt bath.

Although the present invention has been described in conjunction with certain preferred embodiments, it is to be understood that modiilcations and variations may be resorted to without departing from the spirit and scope of the invention, as those skilled in the art will readily understand. Such variations and modifications are to be considered within the purview of the specification and scope of the appended claims.

I claim:

1. The process for protecting from corrosion the metallic electrodes employed in molten salt bath furnaces of the type heated within the range of about 1300 F. to about 2200 F. by an alternating electric current passing through the bath between said electrodes which comprises maintaining in a molten neutral salt bath about 100 to about 2500 parts per million of chromate while said bath is being heated within the range of about 1300 F. to 2200 F. by passing alternating electric current through said bath between said electrodes at an average current density of about 7 to about 36 amperes per square inch of electrode area exposed to said bath, the electrodes being made of an alloy comprised predominantly of at least one metal of the iron group and containing about to about 35% chromium.

2. The process for protecting from pencilling corrosion the metallic electrodes employed in molten salt bath furnaces of the type heated within the range of about 1300 F. to about 2200 F. by an alternating electric current passing through the bath between said electrodes which comprises maintaining in a molten neutral salt bath containing at least one of the chlorides of the alkali and alkaline earth metals about 100 to about 2500 parts per million of chromate while said bath is being heated within the range of about 1300 F. to 2200 F. by passing an alternating current through said bath between said electrodes at an average current density of about 7 to about 36 amperes per square inch of electrode area exposed to said bath, the electrodes being made of an alloy containing about 0.1% to about 80% iron, about 10% to about 22% chromium and about 7% to about 82% nickel.

3. The process for protecting from corrosion the metallic electrodes employed in molten salt bath furnaces of the type heated within the range of about 1300 F. to about 2200 F. by an alternating electric current passing through the bath between said electrodes which comprises maintaining in a molten neutral salt bath containing at least one of the chlorides of the alkali and alkaline earth metals about 100 to about 2500 parts per million of chromate while said bath is being heated within the range of about 1300 F. to about 2200 F. by passing an alternating current through said bath between said electrodes at an average current density of about 7 to about 36 amperes per square inch of electrode area exposed to said bath, the electrodes being made of an alloy containing about 5% to about 75% iron, about 12% to about 22% chromium and about 7% to about 80% nickel.

at least one of the chlorides of the alkali and alkaline earth metals about 100 to about 2500 parts per million of chromate while said bath is being heated within the range of about 1300 F. to about 2200 F. by passing an alternating current through said bath between said electrode" at an average current density of about 20 amperes per square inch of electrode area exposed to said bath, the electrodes being made of an alloy containing about 5% to about 75% iron, about 12% to about 22% chromium, and about 7% to about 80% nickel.

5. The process for protecting from corrosion the metallic electrodes employed in molten salt bath furnaces of the type heated within the range of about 1300 F. to about 2200 F. by an alternating electric current passing through the bath between said electrodes which comprises maintaining in a molten neutral salt bath containing at least one of the chlorides of the alkali and alkaline earth metals about 100 to about 500 parts per million of chromate while said bath is being heated within the range of about 1300 F. to 2200 F. by passing an alternating current through said bath between said electrodes at an average current density of about 20 amperes per square inch of electrod area exposed to said bath, the electrodes being made of an alloy conta ning about 5% to about 75% iron, about 12% to about 22% chromium and about 7% to about 80% nickel.

6. The process for protecting from corrosion the metallic electrodes employed in molten salt bath furnaces of the type heated within the range of about 1300 F. to about 2200 F. by an alternating electric current passing through the bath between said electrodes which comprises maintaining in a molten neutral salt bath conta ning at least one of the chlorides of the alkali metals and barium about 100 to about 500 parts per million of chromate while said bath is being heated within the range of about 1300 F. to 2200 ,F. by passing an alternating current through said bath between said electrodes at an average current density of about 20 amperes per square inch of electrode area exposed to said bath, the electrodes being made of an alloy containing about'5% to about 75% iron, about 12% to about 22% chromium and about 7% to about nickel.

7. The process for protecting from pencilling corrosion the metallic electrodes employed in molten salt bath furnaces of the type heated within the range of about 1300 F. to about 1800 F. by an alternating electric current passing through the bath between said electrodes which comprises maintaining in a molten neutral salt bath containing at least one of the chlorides of the alkali metals and barium about to about 500 parts per million of chromate while said bath is being heated within the range of about 1300 F. to 1800" F. by passing an alternating current through said bath between said electrodes at an average current density of about 20 amperes per square inch of electrode area exposed to said bath, the electrodes being made of an alloy containing about 0.5% to 75% iron, about 12% to about 22% chromium and about 7% to about 80% nickel.

8. In the heat treatment of metals by means of salt bath iurnaces or the type heated by an alternating current passing through the molten salt between metallic electrodes, the improvement for prevention 01' pencilling corrosion oi said metallic electrodes which comprises introducing into a molten neutral salt bath heated by the passage of alternating current therethrough at an average current density or about 7 to about 36 amperes per square inch of electrode area exposed to said bath between electrodes made of an alloy containing chromium and-at least One metal of the iron group about 100 to about, 2500 parts per million 01! chromate, and operating the chromate-containing molten neutral salt bath at temperature between about 1300 F. and 2290 F. with said electrodes immersed therein and with alternating current passing through said bath at said average current density.

HARRY ROLLASON COPSON.

REFERENCES orrnn The following references are of record in the iile of this patent: I

OTHER REFERENCES Doremus, Heating Piping and Air Conditionmg." May 1931, pages 376, 377.

Roetheli et aL, Industrial and Engineering Chemistrylf October 1931, vol. 23, No. 10, pages 1084-1090.