Surface treatment of metals

θ (patented May 7,1940

<img file="US2199418A_D0001.tif" />

Cross Reference 2,199,418

UNITED STATES PATENT OFFICE

2.199,418

SURFACE TREATMENT OF METALS

John C. Redmond and Ralph W. Hodil, Youngstown, Ohio

No Drawing. Application September 16,1938

Serial No. 230,210

Claims. (CL 148—8)

The present invention relates to the conditioning of the surface of metal articles which may contain oxide, scale, and the like or even appear apparently clean and unoxidized to the eye, by _e removing all the surface oxides and other contamination and imparting an improved type of metal surface preparatory to the application of a protective metal coating.

It is well known and recognized that the suc1Q cessfulness of protective coatings applied to metal articles is in many cases limited practically by the ability of such coatings to withstand subsequent fabricating operations which are necessary before an object may be considered a fin10 ished product. In many cases, although possible, it is an economic disadvantage to apply protective coatings to metal articles after the last fabricating or forming operation has been performed, whereas, if a protective coating can be „ applied early in the process of fabrication while the material to be used is still in a continuous state, great savings can be effected. In fact, it is a common procedure in industry to test the quality of an applied metal coating by carrying „ out tests designed to gauge how much distortion a certain metal which has been coated with a protective metal will endure before the coating is torn from the base metal. It is also recognized that the protective value of an applied metal .₀ coating on a metal object is dependent to a large extent upon its adherence to the base metal. For this reason, any method of treating a surface so that it will bond, or adhere, in a superior manner to any applied protective coating is of extreme importance and of great economic value ⁸⁸ in the art.

It is well known that when ferrous and many non-ferrous metals are heated and exposed to the atmosphere, or merely subjected to atmospheric exposure for long periods without heat⁴⁰ ing, the surface of th? metal becomes oxidized and the article thus coated with a film comprising the oxides of the metal and sometimes other compounds as well, forming a hard scale which is frequently firmly adherent and extremely dif⁴⁸ ficult to remove. Such scale and other surface contamination usually must be removed for proper performance of subsequent operations such as rolling, forming and the like, or applications of g. a protective coating of some other metal such as zinc or tin, or of enamel, paint, lacquer or the like, while in some instances best practice requires that the clean and descaled surface be etched to permit it to make a satisfactory bond . with the subsequently applied coating material.

Numerous methods have heretofore been utilized for removing scale and other surface contamination from metal articles and oxide scale has generally been removed from ferrous articles by immersing the articles in a bath containing an agent, usually an acid, which reacts with the scale to produce a water soluble compound dissolving in the bath or in a subsequent washing water, leaving the surface of the metal free of the scale and its reaction product, some of the _Mscale being dislodged mechanically from the sur- * face through the action of the solution and the mechanical action of gaseous products of the reaction on the base metal. In the practice of this method commonly known as “pickling, a dilute solution of a suitable acid, for example,₍sulphuric or muriatic acid, is usually employed as the picking agent, and relatively large quantities of it are required to maintain a bath of sufficient volume to permit immersion of the _Marticles to be pickled. Moreover, the method is ⁸⁰a slow one and involves much handling of the articles to be cleaned, thus entailing large expense for labor which, added to the cost of the pickling agent, makes it unduly expensive. Furthermore, if an ordinary etched surface on the ⁸⁸articles is desired for the purpose of obtaining a bonding in a subsequent metal coating operation, it can only be obtained by continued immersion in the pickling bath after the surface has been cleaned and descaled and apart from the in- ⁸⁰creased expense involved by this procedure, the surface then contains inclusions of the gaseous products of the pickling operation, such as hydrogen, which results in a material portion of the articles treated being damaged by the occurrence ⁸⁸of blisters, acid black, and other surface defects.

Other methods have been suggested to facilitate the removal of scale frdm metal articles by injecting acids, either liquid or vaporized in the dilute or concentrated forms, into heated retorts ⁸⁸in the presence, or absence, of air with the object in view of loosening the scale so that the articles may be cleaned more easily in a subsequent sand blast or abrading operation. The surface obtained by these methods is no more desirable ⁸⁸than that obtained with the ordinary pickling practice from the standpoint of a subsequent protective metal coating operation.

Another suggested method refers to the removal of drawing compounds specifically hydro- ⁸⁰carbons, mineral or petroleum oils, lubricants and the like, from clean metals which do not have oxidized surfaces, for the purpose of a subsequent coating operation. This method exposes ,

2,109,418

100₂ the metal to be cleaned alternately to a water wash and then to the action of moist chlorine gas without the application of heat and is inoperable on dry metals or in the absence of moisture.

Furthermore, although it may leave the metal in a cleaner state and consequently more suitable to receive a subsequent coating operation. It does not impart to the surface of the metal any qualities which might enable it to make a better bond 10 to any coating subsequently applied.

' In some instances the deoxidization treatment of metals, sometimes known as “bright annealing,” has been practiced, involving heat treatment of the articles in the presence of a reduc15 ing agent such as finely divided carbon, hydrogen, coke oven gas and other reducing agents which, in the presence of heat, reduce the oxide of which the scale is formed to the unoxidized metal which leaves on the surface of the parent 20 metal a film of the reduced metal in an undesirable physical state known as sponge metal.” Such a surface is not suitable to the reception of a protective coating, does not receive the bright finish which may be imparted to clean metal by 25 rolling and the like, and in general must be removed before any further operation on the article may be performed. The removal of this type of film from the surface of metals can be accomplished only slowly by means of any of 30 the known methods of cleaning or descaling, since it reacts only slowly with most reagents practical for use in removing it.

Our Invention is directed to improvements in the art of conditioning metal surfaces, without ₃₅ the employment of acids, or fluid baths containing corrosive reagents, and without the production of sponge metal by chemical reduction of the contaminating material, by removing scale deposits formed of oxide or other undesirable _4(| compounds contaminating the surface of metal articles, if same are found on the surface of the metal, and subjecting the surface of the base metal to a treatment which gives it an improved type of clean, etched or matted surface distin₄₆ guishable from the ordinary etched surface obtained by the well known pickling methods in its ability to form a superior bond than heretofore known to a subsequently applied metal, or other protective coating. In fact, our method can be 50 utilized to effect the removal of sponge iron metal as well as other contaminating substances at a rate of speed far in excess of those possible by other known methods.

In general it may be said that the present in33 vention comprises subjecting-the articles to be treated to the action of hydrogen chloride gas at a certain minimum elevated temperature to convert the oxides and contaminating materials on the surface of the articles, as well as part of the _eo base metal, to an anhydrous chloride of the base metql and water vapor, the lowest temperature of operation being sufficient to Cause the vaporization of said chloride with resultant removal of the chloride from the surface of the 65 metal; allowing the gas to react further on the surface thus freshly exposed, and then cooling the articles in the gas employed or in some other suitable non-oxidizing atmosphere.

We have discovered that oxide scale and sponge iron are removed from steel sheets, steel pipes and the like in from five to ten seconds after introduction of hydrogen chloride gas into a chamber maintained at a temperature above

800° F. and wherein the articles are contained

7g and that at this temperature the ferrous and ferric chlorides produced vaporize rather rapidly, leaving the articles relatively free from any sort of deposit and with a clean appearance similar to that obtained from pickling. On subsequently coating the articles treated in this maimer with a g zinc coating, that is to say, galvanizing the same, we discovered that the adherence of the zinc coating was far greater than the adherence of a coating of zinc applied by the same method to steel articles which had been pickled in the ordl- io nary maimer.

It will be understood that the principal constituents of scale on ferrous metal articles are usually the ferrous and sesqui oxides (FeO and FesOi). We have discovered that the principal ]j mechanism whereby this scale is removed from the surface of metals, according to our invention, is accomplished by the following fundamental reactions which take place. These are of the type: go (Solid) (Gas) (Vapor) (Vapor) (1) FeaOi + 6HC1 -*2FeCb + 3HaO (2) FeO +2HC1-* FeCla + HaO

As is denoted by the reactions listed above, the 28 scale reacts with the hydrogen chloride gas producing the ferrous and ferric chlorides of iron and water vapor. We have discovered that these reactions take, place readily, rapidly, and efficiently between the temperatures of 800° F. and gg 1350° F., the optimum temperature being approximately 1100° F. Both the ferrous and ferric anhydrous chlorides of iron are readily vaporized and are formed in the vapor state at these temperatures. Thus the iron oxides are 38 converted into gaseous chlorides and water vapor which leave the surface so that further action on the underlying metal or oxide may proceed. If the temperature is much lower, the chlorides do not vaporize and form a coating on the surface of the oxide (or metal) which prevents further action. This constitutes the criterion for the lower temperature limit at which the process may be carried out. If a metal containing an appreciable amount of iron oxide on its surface is inserted into a stagnant atmosphere of pure hydrogen chloride gas, assuming that the oxide present in the system is overly abundant, the reaction will not continue until all the hydrogen chloride gas has been used up in converting the 80 oxide to the chloride and water vapor, but will - only continue until a certain equilibrium mixture has been attained in which will be present hydrogen chloride gas, anhydrous ferric and ferrous chlorides of iron in the vapor state, and water vapor, all in their equilibrium concentrations at the temperature of operation, is so because the chlorides which are in the vapor state, and consequently in intimate contact with the remainder of the atmosphere, soon reach this equilibrium concentration and stop the reaction, Since water vapor is also a product of the reaction, the presence of water vapor in excess amounts at the start of the reaction slows the „ reaction down considerably, and brings the re- ^Waction to a stop at a different equilibrium mi»ture. The concentration of the chloride in this mixture, or the amount of oxide converted infa» chloride, is very much less than it would have 70 been if the system were substantially devoid of water vapor at the start of the reaction. In fact, if there is too much moisture in the atmosphere at the start, no appreciable reaction will take place. This point may be demonstrated most ?·ί id

2,109,418 easily by expressing the equilibrium constant for acid in carrying out reaction (1) ₍ (Ph₂o)³(F».cij)’ ^K“ (ΡβοΫ (Partial pressures P expressed in atmospheres) The equilibrium constant K is given by the product of the cube of the partial Fissure oi ,_n water vapor and the square of the partial pres¹⁰ sure of fSic chloride vapor dividedby tteigth newer of the partial pressure of the hydrogen Sride gas It has a definite value for a gteen temperature and the reaction will proceed in .. such a direction as to satisfy this value andwill ¹⁸ stop when this value has been reached. Tb?J® see that if water vapor is present at toe star less chloride can be formed, toanif no water vapor were present at the start. Of «» 1«8® sa amount of water vapor is P^res?ⁿt?\“i^e start, ⁸⁰ only an infinitesimal amount of chloride will be ^Furthermore, we have found that these reactions are reversible reactions, and.at ,,, _tures higher than approximately 1350 F., tire ²⁸ e^iilibrium constant assumes such a value.that mS less chloride is formed than at lower temSures, that is, much more hydrogen chloride nuist be used for the formation of a given quan- .

onnroximately 1350’ F. whereas the lower limitferric Chlorides of iron which is approximately ⁸°with respect to the water vapor content, we ,_n have pointed out above that the presence of mois⁴⁰ ture at the start seriously impairs and hln^der3toe speed of toe reaction, even so far as to stop toe reaction entirely if there is an nf this substance. Thus while we preferably op eLte our in toe substantial absence of ⁴⁵ water vapor, toe presence of water vapor having a nartial pressure of approximately 15% ο._Λtotal gas pressure can be ^tol®^rat®^d^“vapw at ⁵⁰ ample to 25%, toe reaction will not take place.

- Ess.

locity which is extremely large compared to toe Xitte encountered in acid picWng baths whor» scale is removed by action of adds.

We Se ^otaund and it is evident from the « foXtag «Ssdon that in order to matotato a contomed reaction, obtain the marimum: efficiency of conversion of oxide into °^ω°^® J** TOhune of hydrogen chloride gas expended, and βδ maJteuse of toe maximum rate of reaction,^is β*=-**·*·*κ ner that toe spent atmosphere is removed contact with toe articles as soon as toe eqjdlib7A rium mixture has been attained and Immediately eliminated from toe system. ..... ,.....

The above requirements exclude the.possibility of using acids in vaporized states, such be obtained by spraying, atomizing, or boiling.

For example, it is not possible to use hydrochloric acid Is notttag more than hydrogen chloride gas dissolved in water, m its ^m°st concentered form it contains approximately 65% water and 35% hydrogen chloride gas. When vaporized itispparent that even toe most concenteated hydrochloric acid would contain far too much wrier vapor to carry out the reactions necessary in our process. Hydrogen chloride gas may be com pared to hydrochloric acid for our PW“®_^^® same manner that chlorine would be compared to saturated chlorine water. Similarly such as nitric acid and sulphuric ^acidactually water solutions, cannot be used in car rying out our process.

We have found that chlorine gas itself cann be utilized in the removal of scale from metal faces to produce articles which are clean and have a Pickled appearance. The reasons for- tote may be twofold. First, it.is questionablewhether ⁸⁰iron oxide reacts with chlorine at th® tei^ra tures in question without the presence of some reductag agent to take up the free oxygen^which would necessarily be formed according to toe re actions— (3). (4) In our method toe nyarogeu — --- -dfogen chloride reacts with the to 30 water vapor, whereas it is weH that even ₃₅Furthermore if the reactions did proceed to wards toe right at these metaTsirlihprated would be in contact with toe metai sur . face and at toe lowest temperatures necessary.to : toe chloride, they would «oridtee toe „ metal and form scale again. Thus we cannot use chlorine gas in practicing om process upon artiί'ΐρς which are scaled to start with. __

The forego¹? discussion has been concerned primarily with toe chemistry of seal® «moval α from scaled objects since it¹⁸ «^ious that itJs first necessary to expose toe surface to the action of the gases employed in our process to con- . Son £ surface of toe objects treated However we do not confine our process only to condi- go tioning metals which have scaled surfaces to start Sth b^we may also condition objects whose surfaces are free of scale and apparently clean at toe start. Insofar as we have already discussed thechemistry of toe removal of scale from toe U surface by our process, we may now discuss toe action of the gases on the exposed surface and in so doing discuss simultaneously toe conditiontag of alurface which was originally scaled and of one which to start with was unsealed and _M^C1 When hydrogen chloride gas, substantially dry, actewito . .-------- Fe-j-2 _Hci-»FeCla+Ha (6) 2 Fe+6 HCl-*2 FeCh+3 Ha

We have also found that chlorine gas may be 70 substituted for hydrogen chloride gas in toe practice of our method where toe conditioning of scale fre^ surfaces is to be carried out even.though(it cannot be utilized where scaled surfaces are concerned. Since in this case we do not have

FeaOs+6 Cla-»4 FeCla+3 Oa

FeO+2 Cla-»2 FeCla+Oa method the hydrogen present in the itftAarniner υιυδδ πϋΐϋΐϋΐιυο

2,188,418

100 4 to deal with oxides, the steel reacts with the chlorine gas to produce chlorides of iron without the liberation of oxygen as in the case of scaled surfaces, according to the reactions— ⁵ (7) Fe+Ch-»FeCh (8) 2 Fe+3 Ch-»2 FeCh

Although these reactions (5, 6, 7, and 8) do not take place at ordinary temperatures, except 10 in the presence of water, and then only very slowly (by means of the action of an acid formed by the water absorbing the chlorine or hydrogen chloride), they do occur on the application of heat. Beginning at about 400° F., they have 15 an appreciable reaction velocity at 800° F. and are very rapid above that temperature in the absence of water vapor. However, the lower limit, as in the previously discussed case, at which our process can be carried out, is the temperature ί>0 of the vaporization of the chlorides formed. Furthermore, these reactions do not continue until all the hydrogen chloride gas, or chlorine gas, has been converted into the chlorides of iron but as in the case of all gaseous reactions, continue 25 only until ^a certain equilibrium mixture has been reached. Although water vapor Is not a product of the reaction it is obvious from the previous discussions that if water vapor is present in the system in substantial amounts, the water vapor 30 react with the iron chlorides and the equilibrium mixture then established will be the joint equilibrium, mixture of two reactions. In addition, at the temperatures of operation water vapor itself reacts with steel, converting the steel to the ₃₅ oxide and releasing hydrogen gas. All these considerations make it obvious that in the conditioning of metal surfaces the water vapor content must be kept as low as possible for best operation, whereas if the partial pressure of water vapor sub40 stantially exceeds 15% of the total gas pressure, • we have found the desirable reactions are practi! cally non-existent.

Therefore, the upper limiting temperature at which our process can be carried out for the coni 45 ditioning of clean metal surfaces, whether hydrogen chloride gas or chlorine gas is emί ployed, is entirely dependent upon the amount s of water vapor present in the system. In the i event that the water vapor is so controlled as to ? . go be present only ip negligible quantities the up: per limiting temperature of operation is to be

J chosen only with respect to other factors such j as the metallurgical qualities of the metal, or the limitations of the containers, etc. If the water ί gg vapor content Is of thejorder of 15%, the temperature of operation must belimited to approxii mately 1350° F. and if the water vapor content j appreciably exceeds this amount the process cannot be carried out at all.

It has, therefore, been made clear that in the practice of our invention where we are concerned with metal objects containing oxides, scale, and other contamination upon their surfaces, we can employ hydrogen chloride gas, but not chlorine 55 gas, to remove these substances and, if desired, • also condition the exposed metal surface by continued treatment with hydrogen chloride gas, according to our process. Where we are concerned with clean metal objects, we may con70 dition the surface by the use of either hydrogen chloride gas or chlorine gas in the manner described above.

We have found this method of treatment to have many advantages over the etching of metal 75 by means of liquid acid solutions. First, the reactions take place much more rapidly. Whereas the time necessary to etch metals by means of acid baths is to be measured in periods of minutes and often in hours, our reactions are to be measured in seconds. For this reason, they have a 5 unique advantage in the removal of the undesirable sponge iron from the surface of metals. Since the sponge iron is actually pure metal in an undesirable physical form, it can be removed only very slowly by means of acid baths, whereas in 10 the utilization of our method, sponge iron is removed as quickly as is the oxide of the metal. Second, although hydrogen gas is present in the system, the metal does not absorb the gas at the temperatures of operation and consequently 15 there are no inclusions of hydrogen which often in the ordinary processes result in hydrogen embrittlement and in blisters. Thus a very annoying problem, which causes much waste in industry, is automatically eliminated. Third, we *0 have discovered that metals conditioned in this manner, although they do not present a much different appearance to the eye than do ordinary etched metals, have a superior surface insofar as their ability to form a bond to any subsequently 25 applied metal coating is concerned. Although we do not entirely understand the reason for the occurrence of this preferred type of surface, we believe that at the temperatures of operation, the gases employed by us have a twofold action upon 80 the treated surface in addition to those mentioned. First, at the temperatures of operation the kinetic action of the gases attacking the steel is much more violent than in ordinary cases, and results in a surface which is more minutely un- 35 even than an ordinary etched surface. Such a surface might be compared to an etched surface in the same way that the surface exposed by a pane of glass covered with an even coating of sand might be compared to the glass pane itself. 40 The uneven surface would provide a much larger area for gripping the protective coating than would a comparatively flat surface for the same included area of the metal and thus give much greater strength to the bond of the two metals. 45 For example, a completely flat surface one inch square would provide, a surface of one square inch for gripping, whereas if the surface were minutely uneven it might provide twice or even three times that area for gripping. Further- 50 more, it is believed that our process provides for the removal of other surface impurities which might be injurious to the adherence of a subsequently applied metal coating and which are not removed at ordinary temperatures. In other 55 words, a more chemically clean surface is obtained than with the common methods now in. use. We have found that when a zinc coating is applied by the method of galvanizing to a steel surface conditioned according to our invention, 50 the coating will suffer more distortion than will the base metal itself, that is, the base metal will be torn apart before the protective coating is severed from the base metal.

In the practice of our invention we do not limit ⁸⁸ourselves to the use of pure hydrogen chloride or chlorine gases. Thus we may dilute the active gases by mixing them with some inert gas such as, for example, nitrogen, or a mixture of gases which do not enter into chemical reaction with 70 the active gases or the metal surface at the temperatures of operation such as, for example deoxidizing atmosphere. In fact, we prefe. ,-. practice our method when the active gas has i · so diluted by an inert gas, that the hydrog.::.

9,188,418 chloride content, or chlorine if it is being used, is between thirty and fifty per cent of the total mixture. Or we may vary the amount of dilution according to the requirements of a particular operation.

Of course, to prevent the growth of further scale on the article after completion of the chemical reactions, it is necessary that the cooling of the articles to room temperature take place under circumstances excluding oxygen or other oxidizing gases from their surfaces. This may be accomplished by keeping the articles enveloped by the active gases or inert gases during the operation of cooling.

It will be understood that the extent to which the several reactions proceed is dependent upon the time allowed and that after sufficient time has elapsed for substantial completion of the reaction between the hydrogen chloride gas and the scale or other surface contamination, the articles may be removed. They may be left in the chamber in contact with the active gas for a further period if it is desired that their surfaces be conditioned, the extent of conditioning being determined by the control of the time allowed therefor and of the temperature and of the dilution of the hydrogen chloride gas in the active atmosphere.

No specific reference has been made to the particular type of chamber for use in carrying out the method, as any chamber suitable for the purpose may be employed, but it is frequently of advantage to utilize a continuous furnace when conditioning ferrous articles preparatory to galvanizing, as this process may then form a step in the galvanizing operation. In some instances, however, it may be preferable to carry out the process as a batch method of operation in a suitable heating furnace so equipped as to enable the articles to be cooled after conditioning without removing them from the furnace. When pipes are being cleaned, means may be provided for passing a stream of the active gas through their interiors so as to condition their inside as well as their outside surfaces.

Our invention may also be carried out in combination with heat treatment such as normalizing or box annealing through the introduction of the active gases into the normalizing furnace or annealing box at a suitable point during the course of operation.

This application is a continuation in part of our application filed August 9, 1937, Serial No. 158,153.