US2897093A - Process of siliconizing - Google Patents

Description

July 28, 1959 H. A. ECKMAN PROCESS OF SILICONIZING Filed March 2, 1955 United States Patent PROCESS or SILICONIZING Harry A. Eckman, Chicago, lIlL, assignor to Crane Co., Chicago, Ill., a corporation of Illinois Application March 2, 1955, Serial No. 491,777

7 Claims. (Cl. 117-533) This invention relates to a metallurgical process for siliconizing or effecting the silicon cementation of metal articles, particularly those of ferrous metal.

As has been adequately covered by the prior art on this subject, the desirability of forming articles in their final size and shape from machinable or formable metal, as, for example, low carbon steel, and then causing them to become impregnated with silicon at least along the outer surfaces, thus acquiring the characteristics of silicon metal at those surfaces for high resistance to Wear, corrosion and acid attack while preserving the strength and resiliency of a non-siliconized core has long been recognized.

Also previously covered are the various and many attempts to accomplish the above goal and the many difficulties encountered by predecessors and their failure to produce satisfactory results.

Extensive investigation and tests I have made substantiate these difficulties and failures. For instance, considering the only process that has seen even limited commercial use so far as is known, the articles produced by this process are marked by spalling and breaking away of the siliconized case or outer layer from the base metal or core, especially when subjected to thermal or mechanical shock or even when reheated. Moreover, the articles are susceptible to acid attack and corrosion because of a rather high degree of porosity of the case. The case is also subject to swelling and distortion prior to actual spalling and the article undergoes a substantial loss of surface metal and consequently weight during the siliconizing process. The process also lacks uniformity in the matter of thickness and quality of the case and its adherence to the base metal. Further, the siliconizing takes place at a rather slow rate requiring an appreciable amount of time for the process. In addition, materials with appreciable amounts of alloying components, particularly chromium, do not lend themselves to siliconizing under this process because of a reduction of the chromium content of the metal for instance under the action of the processing atmosphere. Further, yet, the siliconized case cannot be thickened by additional processing as the original case tends to peel 01f when reheated and also the case formed in the second processing remains separated from the first resulting in what is termed double-case. The process is therefore marked with difficulties greatly hampering its commercial use, the products produced thereunder being subject to very limited use.

The poor bond between the siliconized portion and core which is responsible for much of the difliculty and the poor resistance to corrosion and acid attack also signified by peeling or loosening of the case is believed largely the result of by-product formation and corrosive attack during the siliconizing process.

This prior art process has been carried out by heating the article being treated to a cementing temperature in the presence of a silicon bearing material, such as silicon carbide or ferro-silicon, and then passing a current of chlorine gas over the silicon material at the elevated temperature, or the chlorine can be brought into contact with the silicon material in a separate container from which the gases are permitted to flow on into contact with the article. In either case, the chlorine gas reacts with the silicon material to form silicon tetrachloride gas which is the active reagent in the siliconizing process. In actual practice, the articles are placed in a retort with ample silicon carbide and brought up to heat (around 1850 F.) in either a reducing or oxidizing atmosphere (see Patent No. 2,157,902 to Ihrig of May 9, 1939, as to the latter), and then chlorine gas is flowed through the retort until the desired depth of siliconized metal is produced, usually requiring three hours or longer for a relatively thick case. The articles are thus exposed to the siliconizing atmosphere for a rather prolonged period of time.

It has been found that the siliconizing atmosphere in this process including the products of the reactions taking place is not conducive to the formation of a desirable case. To gain an understanding of this, consider what takes place in the reaction chamber.

In the drawing which is generally schematic in character, a heated container 1 for holding the liquid silicon chloride 2 is provided with a heating unit 3, producing in due course the vapor of silicon chloride as indicated at 4. A suitable fluid tight cover 5 is applied at the container upper portion, while upon a side wall 6 a vapor discharge conduit 7 connects the said container with a processing vessel 8, which in actual practice is a cylindrical rotating retort.

The material to be processed designated by the dot and dash lines with the silicon carbide powder is loaded into the processing vessel by any suitable door (not shown) and unloaded in the same manner.

The vessel 8 is heated by an electric heating unit or coil as indicated at 9 or by any other convenient method. A suitable exhaust at 10 provides for the escape of gases and other gaseous products.

A control valve, as designated at 11, is mounted on the conduit 7 to regulate the passage of the vaporous silicon tetrachloride into the vessel 8.

Screens 12 define the end limits of the chamber in which the material to be processed is confined.

The chlorine gas reacts with the SiC to form gaseous SiCl as mentioned above, and also with carbon which remains on the inside surface of the retort. The silicon tetrachloride which is the active compound in the siliconizing process is broken up by reaction with the metal being treated in some manner or other into silicon which is taken up or absorbed by the metal presumably in solid solution therein and free chlorine which is released into the processing atmosphere. Continued action between the SiCL; and the metal surface in time produces a surface layer of silicon bearing metal.

The liberated chlorine, however, reacts with the metal and forms chlorides, as, for instance, iron chloride in the case of ferrous metal. These chlorides are gaseous at process temperature, and the effect of this reaction is to progressively remove or eat away the iron from the surface or from under the surface of the article treated, thus causing a loss in weight and a certain porosity of the surface. Also, some of the iron chloride remains entrapped in cavities and fissures of the case and solidifies there when the article cools. The porosity facilitates acid or corrosion attack and undermining of the case especially along the soft core while the iron chlorides not only weaken the bond between the case and core, but also tend to absorb moisture, since it is hygroscopic, resulting in further weakening of the bond by rust formation, swelling and loosening of the case. It should be pointed out, however, that as the metal takes on silicon, it becomes more resistant to the reaction of the liberated chlorine. Nevertheless, a considerable loss in weight and porosity takes place before the chlorine attack reduces in intensity. This is particularly manifest in this method of siliconizing, since a high concentration of chlorine gas is used to initiate and carry out the process, this gas being continuously supplied throughout the processing. It is, therefore, evident that the articles are subject to serious corrosive attack at the very inception of the siliconizing action. Such concentration of chlorine, which only partly reacts with the silicon carbide to form the silicon tetrachloride, together with the chlorine liberated in the course of the siliconizing action also produces an excessively high concentration of iron-chloride vapor which besides saturating the atmosphere and becoming lodged within the cavities and fissures of the porous case also tends .to react with the silicon tetrachloride present and apparently form some complex iron-silicon-chloride compound which tends to deposit on the surface of the material treated, inhibiting the desired silicon impregnation.

High chlorine content in the processing vessel further tends to counteract the siliconizing. Obviously, if a siliconized material were exposed to a chlorine atmosphere at processing temperature, the chlorine gas would attack the siliconized surface forming silicontetrachloride in the same manner as it reacts with silicon carbide, ferro-silicon or other silicon bearing material.

In order to avoid the above difliculties and to produce a satisfactory case, a second process has previously been proposed. This latter process which has not been employed commercially or to any substantial extent so far as is known makes use of a mixture of silicon tetrachloride gas and hydrogen which is flowed over the articles at a cementing temperature for the siliconizing action. The reactions as explained in Patent No. 2,501,051, issued March 21, 1950, are as follows:

SiCl +2Fe=Si+2FeCl (Replacement) SiCl +2H =Si+HCl (Deposition) It is thus seen, according to the theory set forth, that there are two siliconizing reactions.

The ratio of the hydrogen to the silicon tetrachloride gas and the extents to which the two reactions are permitted to take place, however, are very critical to the achievement of proper results. As disclosed in column 14, lines 12 to 16 of the patent, hydrogen percentages (based on silicon tetrachloride) of from 0.1 to 1.4 are ordinarily most consistently effective in producing articles that are satisfactory from all around performance, and according to one of the claims, namely claim 15, the percentage is between 0.35 and 0.45.

Also, it is imperative that the hydrogen employed in the process, have an extremely low moisture content not exceeding a particular stated value, and that it also be substantially free of such oxidizing agents as carbon dioxide (CO and carbon monoxide (CO). Since commercially available hydrogen according to the patent contains more than the permissible amount of oxygen, free or combined, a special purifying treatment is necessary in conjunction with the siliconizing process.

Of special importance in this process is the rate of flow of the siliconizing gases through the reaction chamber. These gases must not only be supplied at a sufiicient rate to produce the desired silicon impregnation, but must also be supplied at a high enough rate to effect a sweeping out of the by-products of the reactions before they can have any detrimental effect on the articles. Especially recognized is the danger of occlusion of ferrous chloride in the siliconized case. It also appears that at least some detrimental effect to the articles treated would result from attack by chlorinations resulting from thermal dissociation of hydrogen chloride formed in the processing.

It is, therefore, seen that the process must be rather accurately adjusted and controlled in at least three respects, namely, the proportion of hydrogen to silicon tetrachloride gas, the moisture content of and amount of oxidizing agent present in the hydrogen, and the rate of flow of the mixed hydrogen and silicon tetrachloride through the reaction chamber to effect the blowing out of the reaction gases without undue waste. Concerning the first of these, in order to produce satisfactory siliconized articles consistently, it is essential that the progress of the respective reactions aforesaid be constantly con-trolled and coordinately, by maintaining the ratio of silicon chloride and hydrogen, one to the other, within definite limiting values as aforesaid" to quote the top of column 6 of the patent.

The process also requires rather elaborate apparatus. In the terms of the patent and starting with line 54 of column 9 thereof, this apparatus includes means for supplying and metering the hydrogen or hydrogen-containing gaseous medium; means for purifying the same to eliminate or sufliciently reduce its content of objectionable components; means for testing and indicating the degree of purification from at least one of such components, as a control measure; a source of silicon chloride F vapors or gas and means for metering, directly or indirectly the quantity thereof to be mixed with the purilied hydrogen containing medium; provision for effecting such mixture of the hydrogen-containing gas and silicon chloride vapors; a muffle furnace arranged to receive such gaseous mixture and providing a high temperature reaction zone in which the articles to be siliconized are treated under controlled conditions; and means for trapping out and recovering any excess silicon chloride contained in the gases exiting from the furnace, and for leading away and discharging from the system any finally remaining gases; the apparatus system as a whole being protected against accidental ingress of air into any part thereof, the latter measure being necessary to safeguard the equipment and personnel against explosions with the hydrogen. Delivering the hydrogen in the proper form alone requires the equipment indicated by numerals 2 through 16 and 29 through 37, including an additional mufiie furnace 7, drying towers 11, 12 and a suitably cooled dew-point meter 34.

Besides the elaborate equipment and careful control that is necessary to carry out the process, the latter is also inherently dangerous because of the use of hydrogen which may produce explosions and fires if brought out into contact with air. That this is specially recognized as a real danger note the final clause of the above quoted recitation of apparatus. To the best of applicants knowledge, explosions and fires have actually occurred in connection with this process. Because of the use of hydrogen, there is also the danger of producing hydrogen embrittlement in the siliconized case or in the metal immediately below it, particularly if the hydrogen content shouldexceed the maximum amount specified.

Table I of this prior art patent also shows what poor results are had from the standpoint of acid resistance and weight loss during siliconizing of articles when the amount of hydrogen used is reduced below the prescribed hydrogen silicon tetrachloride ratios. To completely eliminate the hydrogen in thesiliconizing process (which is proposed in the instant application) would apparently result in a product which is entirely unsatisfactory. As

Expressed in the patent, commencing at line 5 3, column 6, any attempt to carry out a gaseous siliconizing process of the general type in question, employing silicon tetrachloride and hydrogen in a ratio, one to the other, that is outside the definite range of values hereinafter specified as the Widest permissible, although it may produce some siliconiZing of the article, will result in failure to produce a durable coating or case that will satisfactorily meet the generally accepted commercial standards or requirements as regards resistance to thermal and mechanical shock and protection of the coated metal against corrosion, such protection being dependent not only upon the density and integrity of the case itself, but also upon the integrity of its bond or union to the underlying metal core. It would follow from this and the data of Table I that the total absence of hydrogen would result in a case of most inferior character and bond to the base metal. Further testifying to this and paraphrasing lines 57 to 69 of column 4, it has been proposed by prior art literature to deposit silicon on iron by the use of a silicon chloride, specifically silicon tetrachloride (SiCl employed in vapor or gaseous form and, optionally, in the presence of another gas, such as nitrogen or hydrogen, but the results obtained were so erratic that such gas treatment has never been put into practical use.

It has been discovered by applicant, however, that an excellent siliconized case can be produced without the use of hydrogen and employing silicon tetrachloride gas.

It is accordingly an object of the present invention to avoid the difiiculties and shortcomings experienced up until now through the use of an improved process of siliconizing.

Other objects and advantages will become more readily apparent and better understood upon proceeding with the specification.

The use of silicon tetrachloride gas alone was met with by unsatisfactory results. It was discovered, however, that the addition of even a small amount of a silicon bearing substance, such as silicon carbide (SiC) to the reaction chamber and the bringing into contact with the silicon carbide and the article being treated of silicon chloride gas, particularly silicon tetrachloride, did result in siliconizing of the article. Tests of cases produced in this manner showed that they were of a superior quality and bond and that the defects and shortcomings of prior art cases had quite surprisingly been eliminated. The present invention is therefore based on this discovery and the startling results flowing therefrom.

Besides the notable success of this process from the standpoint of the high quality of the siliconized product obtained, the process is also characterized by being extremely simple and easy to perform, requiring little control and apparatus as compared particularly to the hydrogen process of Patent No. 2,501,051. The adjustments and control that are necessary to carry out the present process are not critical in nature as they are in that patent. Since hydrogen is not employed in the present process, the explosion and fire hazard has been eliminated. Also significant, the time required for achieving siliconization under the present process is appreciably less than that of any other known process.

As pointed out above, the process involves bringing silicon chloride gas, particularly silicon tetrachloride, into contact with the article being treated in the presence of silicon bearing material such as silicon carbide.

Considerable study on the part of this inventor indicates that a regenerative process takes place during the siliconizing treatment that is effective in doing several things.

Silicon tetrachloride gas which is brought into contact with the article enters into a reaction with the said article being processed whereby the gas is broken up into silicon which is taken up or absorbed by the metal of the article apparently in solid solution therein and free chlorine which is released into the processing atmosphere. The chlorine is in turn reactable with the iron, if the article soon reach a danger point productive of a poor quality of si'liconized article. This is particularly so inasmuch as for every atom of silicon released for impregnation into the metal, four atoms of chlorine are liberated.

According to the patented process employing hydrogen, control of by-product concentration depends on a sweeping out of the gases by the introduction of fresh, uncontaminated siliconizing gases, of course, of the proper critical proportioning, which is quite wasteful if'it is to elfective. The other prior art processes do not recognize the need of this type of control at all. Nevertheless, experience based on production processing utilizing chlorine gas shows that the rate of flow of the chlorine must be kept high enough to eliminate iron chloride vapor from the processing atmosphere, otherwise very poor or unsatisfactory silicon impregnation results. This likewise is wasteful.

Although preferably there is some introduction of new silicon tetrachloride in the present process, control of the chlorine and by-product concentration is primarily accomplished in an entirely new way. As has already been mentioned, silicon terachloride is employed rather than chlorine which eliminates the presence of any chlorine to start with. It has been found that after chlorine has begun to be produced in the course of the siliconizing action and if a silicon bearing substance, which is preferably iron free, such as silicon carbide (SiC) is added to the reaction chamber much of this chlorine will be taken out of the processing atmosphere before it can do any harm by reaction of the same with the silicon carbide as follows:

This reaction is responsible for eliminating large amounts of chlorine in the course of the processing, and to this extent serves as a means of controlling the chlorine concentration. Such control or reduction of the chlorine content is of great significance in producing siliconized articles of the desired characteristics according to the present process.

Besides reducing the amount of chlorine available for attack on the iron resulting in a certain weight loss and porosity, keeping the chlorine concentration reduced in the first instance also results in the production of less iron chloride that may become entrapped within the metal with its bad effect thereon as previously noted. The smaller amount of iron chloride produced in this process, however, is largely if not entirely eliminated from the atmosphere thereof by reacting with the silicon carbide in the following manner:

It should be noted that silicon tetrachloride is produced by both this reaction'and the one previously disclosed between the chlorine and silicon carbide. This is of the greatest importance inasmuch as this chemical is the active ingredient in the siliconizing. process. The above reactions therefore se1ve not only to reduce the chlorine and iron chloride present and, therefore, the by-product concentration which is of signal importance, but also to reconstitute or regenerate the active siliconizing reagent, silicon tetrachloride. The concentration of this reagent is thus maintained or substantially maintained so long as there is silicon carbide present. The introduction of a continuous supply of fresh reactant is therefore either not necessary or not nearly as necessary as it is in the case of the patented process above described in which this is the only means of expelling or eliminating the undesirable by-products and of keeping the silicon chloride concentration at a proper level for efiective treatment. As pointed out before, the latter type of control is rather wasteful.

Besides eliminating the use of hydrogen with its critical aspects and requirement of additional apparatus and control and the ever present explosion hazard, the use of silicon bearing material, such as silicon carbide provides a certain self regulation of by-product concentrations and also preservation of the strength of the siliconizing reactant which results in a process of much greater efiiciency, simplicity and economy.

Although the active ingredient, silicon tetrachloride, is regenerated and the by-products pretty well eliminated, some resupply from time to time of the silicon tetrachloride during the course of the processing appears desirable for best results, preferably in the form of a slow continuous inflowing of the gas. Also, it is preferred to produce the siliconizing in an atmosphere which is substantially free from moisture and oxygen in any form so as to prevent even small amounts of iron chloride that may become entrapped within the case from becoming hydrous in nature and consequently hygroscopic and deliquescent with the undesirable effects previously noted as well as to prevent any oxidation of the metal at the process temperature.

The process can be carried out in various ways, but the following has been effective in producing desirably good results. The articles receiving the siliconizing treatment are placed in a horizontally disposed drum type retort along with an ample supply of silicon carbide in granular or powdered form which may be spread, if desired, along the bottom of the retort before the articles are placed therein. The retort is then connected to a source of substantially dry inert gas, such as nitrogen or argon, which is allowed to fill the retort replacing the air originally present. After this preliminary flushing, the flow of nitrogen, for instance, is turned quite low so as to be just sufficient to make up for any loss that may occur. The retort is then caused to rotate and heat is applied thereto by suitable means as, for instance, passing an electric current through resistance wiring to bring the contents thereof up to a proper siliconizing temperature. This temperature is not critical, but generally best results are had at a temperature of approximately 1.850 F. for ordinary low carbon steel with different temperatures for other consistencies, alloys and non-ferrous metals.

When the siliconizing temperature is reached, the flow of nitrogen is shut off or disconnected and silicon tetrachloride gas is caused to flow into the retort filling the same. This gas may be easily generated by applying a little heat to a receptacle to which a sufiicient amount of liquid silicon tetrachloride for the entire process has been added. Although the vapor pressure created above the liquid is sufiicient to cause the gas to flow into the retort of its own accord, the movement of the gas may be substantially accelerated by connecting or opening the source of nitrogen under pressure in bottle form, so as to expel it across the top of the silicon chloride vaporizer and thus serve as an inert carrier. The nitrogen may also be bubbled through the liquid silicon chloride by discharging it from the end of a tube submerged in the liquid, if desired, to help vaporize the liquid besides acting as a carrier. After the retort has had time to become filled with the siliconizing gas, with or without the presence of a carrier, the flow of the gas and carrier, if used, is reduced to a slow continuous inflow for the actual siliconizing process.

The usual time of exposure to the siliconizing atmosphere after the retort and the articles therein have been brought up to siliconizing temperature is two and a half hours for a relatively case. This is appreciably less time than is required in any other known process even aside from the practicability of those processes and is one of the important contributions of the present :invention.

It was previously noted that theprocess employing chlorine gas requires three hours or longer for a similar thickness of case; lines 57 and 58 of column 7 of the patent above mentioned state that three hours or more are necessary for the process employing hydrogen.

After receiving treatment, the siliconized articles may be allowed to cool in the retort after the means for heating the latter has been turned off or withdrawn to perhaps another retort in a battery or the articles may be taken from the heated retort immediately or allowed to cool to a more desirable temperature for handling and either air cooled or quenched without any ill effects. The latter is quite unique inasmuch as quenching or rapid or uncontrolled cooling in other processes have produced serious spalling and breaking away of the case. Being able to carry out this type of cooling constitutes another important feature of the invention. Besides facilitating quicker handling of the articles and considerably shortening the time required to carry out the entire process, more eflicient use of the retort can be had by the immediate unloading and recharging that is permitted. This permits the retort to be kept at a relatively high temperature and eliminates time consurning and costly preheating and cooling of the retort between runs contributing to overall greater efficiency and economy.

it should be noted from the above description that the apparatus required to carry out this process is simple and that much less of it is necessary than in the case of the patented process employing hydrogen gas. By way of summary, the present apparatus includes a rotatable drum-like retort and means for heating same, an exhaust connection at one end and a removable plate or cap at the other through which an inlet tube passes, a

cooker or pressure vessel of simple pot-like form con nected to the inlet tube and a valve therebetween, a burner or resistance wiring for heating the cooker, means permitting the introduction of liquid silicon chloride into the cooker, and a tank of nitrogen and connections alternatively through the cooker and through the end plate or cap of the retort. The cooker should preferably also have a liquid level indicator and pressure gage. Also, as evident from the above description, the present process is not critical and requires very little control as compared to the patented process.

In spite of the simplicity of this process and the relatively short exposure time required, the same is productive of a very excellent quality of siliconized article to be demonstrated now.

First of all and concerning one of the most serious defects of the prior art, that is the failure to withstand heat or to sustain thermal and mechanical shock, articles siliconized according to the present process for a period of two and one-half hours were subjected to the following tests.

The articles were reheated to 1150 F. and held at this temperature for a period of 72 hours after which the articles were air cooled. When examined, the articles were found to be perfectly sound and free from any signs of spalling or breaking away of the siliconized case. Production articles siliconized under the previously noted process employing chlorine and for the same exposure time of two and one-half hours for processing were subjected to the same test except for a less severe time duration of 48 hours instead of the 72 hours of the other test, since they would not stand up to such prolonged time, but with far difierent results. In the latter instance, the entire siliconized case separated from the base metal or core in spite of the reduced time exposure and broke off over several extensive areas exposing the base metal. The portions of the case around these broken areas were found cracked and loosened assrzgooe from the base metal, the exposed edges of the case showing a condition of swelling or expansion of the case away from the base metal. Other articles produced under the present siliconizing process and of the same process time as before were subjected to reheating at a more severe temperature of 1550" F. although at a somewhat reduced exposure time after which the articles were further subjected to a much more severe means of cooling or thermal shock by being quenched in water. Again, the articles showed no sign of ill effect or unsoundness in any way. A third and yet more severe test was performed by reheating articles siliconized under the present process and for the same process time to a temperature of 1800 F., the articles being held at this temperature for the sustained period of four hours after which they were again taken out and quenched in water. Still there were no signs of spalling or unsoundness of the case. Articles of the other process mentioned were not subjected to the more severe tests inasmuch as the upper limits had already been exceeded even in the first test. It should also be mentioned that these same tests were run with articles exposed to siliconizing under the two processes for a period of one and one-half hours to represent less thickly cased articles, with substantially identical results as before.

Samples of the articles were also submitted to a drop hammer test as a measure of their ability to withstand mechanical shock. In this case, a twenty-five pound hammer weight was allowed to drop four inches onto chamfered ends of hollow cylindrical samples of about an inch in diameter, said chamfers being of a substantial angle and presenting relatively sharp outer exposed edges for contact by the hammer. Of repeated tests on samples siliconized according to the present process, there was no visible effect whatsoever of any injury to the samples. On the other hand, identical samples siliconized under the process employing chlorine and for the same exposure time of two and one-half hours were put to the same test, but with far different results. In every instance, the case split off on the outside of the samples, in some instances the chipped portions extending over very considerable areas ofthe outer surfaces of the samples.

Also indicating the superior bond of the siliconized case to the base metal under the instant method of processing are the results of compression tests run on samples similar to those above described in connection with the mechanical shock tests. Comparative loadings to destruction of the siliconized case being identical samples produced under the present process and that employing chlorine and for the same process time show that samples under the present process withstand approximately twice the loading before chipping or spalling than can samples under the other process.

Comparative tests were also made as to ability to withstand rusting. Samples siliconized according to the present process were kept under sealed jars in a moist atmosphere for a prolonged period of eight months with no rust formation being observed at the end of this time. Samples produced under the process employing chlorine were tested under identical conditions, but became rusty in a few days. Under prolonged exposure in certain instances, the siliconized case completely separated from the base metal.

Ability to resist rusting and acid attack, the latter to be covered below, seems to be a function of and depend largely on the density or lack of the porosity of the outer siliconized case. Reflecting the better quality of case obtainable from the present process are test results showing 30% to 50% smaller weight loss of articles siliconized under the present process as compared to those produced under the process employing chlorine.

Also of great significance in the matter of commercial use of articles produced under the present process is resistance of the same to acid. Ability to withstand attack by sulphuric and hydrochloric acids is indicated in the following table.

001d H spec.

Hot H 804, spec. gravity 1.21

Cold H01, spec. gravity 1.21

gravity 1.09

Change in weight, percent Exposure time Change in Weight, percent Change in weight, percent Exposure Exposure time time Gain 1.34 Loss .13 Loss .15.

2 weeks 4 weeks 6 weeks Gain .27 Gain .29 Gain .33.

2 weeks. 4 weeks- 6 weeks Loss .15. Loss 1.04. Loss 1.56.

2 Weeksn 4 weeks 6 Weeks.

The above results show that processed articles under the procedure of the instant invention remain practically unaffected by either cold or hot sulphuric acid of dilute highly corrosive concentration and the attack of cold hydrochloric acid of similar highly active concentration is very slow. Although results with hot hydrochloric acid and nitric acid, both cold and hot, are not as favorable as those above set forth, articles processed according to the present method of siliconizing are much more highly resistant to these acids than those produced under the process employing chlorine.

Concerning the silicon content of the case a .007 of an inch cut of a sample produced according to the present process and for a process time of two and one-half hours disclosed just under 14%, or 13.62% silicon. Successive cuts of .007 of an inch showed that the silicon content remained approximately the same for the first three cuts or to a depth of .021 of an inch with the lowest value of 13.33% silicon in these cuts, after which the value fell comparatively slowly to 12.00% on the fifth cut or depth of .035 of an inch and more rapidly to a value of 6.28% on the sixth cut or overall depth of .042 of an inch.

Tests also show that products according to the present invention may be thickened or further cased by additional processing at any time interval after the original processing without any ill effects or breaking away of the original case, which occurs in the case of articles produced under the process employing chlorine. This ability to withstand reheating up to processing temperature at any subsequent time for purposes of adding to the original case, or for any other purpose for that matter in which an elevated temperature is desired, gives a flexibility which is quite desirable. Reprocessing may be desired not only for the thickening or adding to the original case but for purposes of refinishing or rejuvenation after prolonged exposure to corrosive media, for instance.

Also of interest in the matter of reheating, it was found that when samples were brought up to a temperature of 2250 F. in an inert atmosphere and held there for four hours that appreciable diffusion of silicon from the case into the underlying base metal had taken place. Such diffusion not only increases the depth of the silicon impregnation and further refines the case, but also increases the strength of the bond between the case and base metal. Obviously, such improved bond and other properties of the product may be of great importance in certain types of service of a severe nature.

It should be evident from the above description and consideration of the problems involved in this field that the applicant herein has discovered and produced a practical and effective process for siliconizing wherein others have failed.

Although the description has primarily been concerned with the siliconizing of metal articles, both ferrous and non-ferrous, it may be desired to siliconize articles containing metal either near the surface or throughout, and, for this purpose referred to as metallic articles. Obviously, also, certain variations of the process may be had within the underlying spirit of the invention, it accordingly being desired that the appended claims be limited only by their terms within that spirit.

11 I claim: 1. The process of siliconizing an article comprising exposing at least a portion of the article while at an elevated temperature and in an atmosphere substantially free from oxygen both free and combined to a source of only a single reagent consisting of SiCl said reagent being brought into contact with the article in the presence of SiC whereby to effect regeneration of the reagent SiCL after reaction in the siliconizing process.

2. The process of siliconizing an article comprising the steps of raising at least a portion of the article to an elevated temperature in a substantially inert atmosphere and then causing a reagent of the class silicon chloride in the absence of any other reagent to be admitted for contact with at least that portion of the article for siliconizing action in the presence of SiC whereby to effect regeneration of the reagent of the class silicon chloride after reaction in the siliconizing process.

3. The process of siliconizing an article comprising traising at least a portion of the article to an elevated temperature and causing liquid silicon chloride to be vaporized and brought into contact with the article for siliconizing of at least that portion of the article in the presence of SiC whereby to effect regeneration of the reagent silicon chloride after reaction in the siliconizing process and to effect reduction of undesired by-products,

4. The subject matter of claim 3, said process including the use of substantially dry inert gas to help direct the vaporized silicon chloride towards at least the portion of said article being treated.

5. The subject matter of claim 3, said process including passing inert gas through the liquid silicon chloride, said inert gas thereafter helping to carry the silicon chloride vapor into the region adjacent at least the portion of the article being treated for the siliconizing action.

substitution of substantially moisture free inert gas for the original atmosphere in the region adjacent at least a portion of the article to receive treatment followed by raising at least such portion to an elevated temperature and then admitting to said region silicon chloride without any substantial amount of free chlorine gas, said sili con chloride being brought into contact with the article in the presence of SiC whereby to eflect regeneration of the silicon chloride after reaction in the siliconizing process and to efiect reduction of undesired by-products.

7. The process of siliconizing a metallic article comprising exposing' the article to gaseous silicon chloride while at an elevated temperature, said process being characterized by keeping the free chlorine content of the atmosphere adjacent the article to a minimum through the supply of silicon chloride gas which is substantially free from uncombined chlorine together with the use of a sub-- stantially iron free SiC in the presence of the gaseous silicon chloride and the article eflective to eliminate iron chloride produced in the siliconizing action by reaction of such chloride with the substantially iron free silicon containing substance, said latter reaction besides eliminating iron chloride also resulting in regeneration of the silicon chloride, thus maintaining the concentration of the siliconizing reagent at an effective level.

References Cited in the file of this patent UNITED STATES PATENTS

Claims (1)

1. THE PROCESS OF SILICONIZING AN ARTICLE COMPRISING EXPOSING AT LEAST A PORTION OF THE ARTICLE WHILE AT AN ELEVATED TEMPERATURE AND IN AN ATMOSPHERE SUBSTANTIALLY FREE FROM SOXYGEN BOTH FREE AND COMBINED TO A SOURCE OF ONLY A SINGLE REAGENT CONSISTING OF SICL4, SAID REAGENT BEING BROUGHT INTO CONTACT WITH THE ARTICLE IN THE PRESENCE OF SIC WHEREBY TO EFFECT REGENERATION OF THE REAGENT SICL4 AFTER REACTION IN THE SILICONIZING PROCESS.

Images