US1906963A - Impregnated silicon carbide article and the manufacture thereof - Google Patents

Description

y 2, 1933- I I A. H. HEYROTH 1,906,963

IMPREGNATED SILICON CARBIDE ARTICLE AND THE MANUFACTURE THEREOF Filed Nov. 17, 1931 ATTORNEYS Patented May 2, 1933 UNITED STATES PATENT OFFICE ALBERT H. HEYROTH, OF NIAGARA FALLS, NEW YORK, ASSIGNOR TO GLOIBAB CORPORATION, OF NIAGARA FALLS, NEW YORK, A CORPORATION OF NEW YORK IMPREGNATED SILICON CARBIDE ARTICLE AND THE MANUFACTURE THEREOF Application filed November .17, 1981, Serial No. 575,578, and in Canada April 26, 1980.

This invention relates to non-metallic articles composed principally of silicon carbide, which are impregnated with a metal, or combination of metals, or with a material such as silicon having metallic characteristics. The invention also relates to the treatment of silicon carbide resistors, and especially to the treatment of the terminal portions of such resistors to increase their electrical conductivity.

In the use of silicon carbide resistors for electrical heating purposes, one of the principal difliculties encountered is that of making satisfactory electrical contact. This is especially true of high temperature electric furnace elements, which often operate at temperatures as high as 1400 to 1500 C. If the metal terminal is brought into direct contact with thesilicon carbide at these tempera tures, the excessive heating of the terminal metal and the oxidation of the end portion of the rod due to local overheating and arcing at the points of contact soon cause the electrical connection to become insulating and the metal terminal is often completely disintegrated.

Various efforts have been made to increase the electrical'conductivity of the terminal portions of the resistor so as to afford alower operating temperature at the contact. Al-

though the methods heretofore used have been moderately successful for the smaller elements operating at comparatively low temperatures, theyhavenot proved entirely satisfactory for the larger elements operating in high temperature electric furnaces, be-

cause they have not afforded a terminal portion of considerable length and uniformly low electrical conductivity throughout the entire cross section of the element.

In my process I impregnate the entire end portion of, the resistor with molten metal so as to produce a dense substantially impermeable end of high electrical conductivity. The end of the rod after treatment is uniformly metallized throughout its cross section, and the length of the treated portion can be readily controlled. By a further treatment of the impregnated portion of the resistor after impregnation is complete, I have terminal portion of the rod.

found it possible to produce an end of even lower electrical resistance than one in which the impregnating material merely fills the pores of the silicon carbide.

The complete impregnation of the terminal portion of a silicon carbide resistor with molten metal has heretofore been considered impractical because of the difliculty in causing the metal to flow into the pores of the article. It is well known that molten metals, even when heated considerably above their melting points, will not wet nor permeate nonmetallic objects, owing to the high surface tension of the metal. Silicon, which I have found to be the most suitable material for impregnation of the ends of the resistor, offers even greater difliculties than most of the common metals when used as a molten impregnating material. Silicon melts at approximately 1450 C. and even when heated to approximately 17 00 C. it will not permeate the end of the resistor. At this temperature any silicon which may adhere to the element is merely in the form of a surface crust. If the silicon is heated to a higher temperature as for example about 2000 C. and the element immersed in the molten bath, a vigorous reaction seems to take place, and the greater part of the immersed rod is disrupted. Even when the immersed portion can be recovered intact, the junction between the body of the rod and the portion immersed is so weak mechanically that the element is not suitable for commercial use.

Another diiiiculty arises from the extreme reactivity of the silicon. It is almost impossible to maintain a silicon bath suitable for impregnation for any prolonged period of time, for the only practicable container is carbon, and the silicon oxidizes rapidly if exposed to the air and is converted to silicon carbide if covered or protected. This =-is especially true when small quantities of silicon are used.

Immersing of the end of the resistor in other molten metals at any ordinary temperature such as would be used for pouring or casting does not result in the impregnation of the I have found that silicon and most of the common metalswill completely impregnate entire cross section. When the critical temthe rod.

perature is reached the im regnation is almost instantaneous, and I ave found that there is suflicient capillary attraction between the silicon carbide and the metal at these temperatures that the molten metal will be absorbed throu bout the silicon carbide body even when the metal is present in only a very small exces over that ry to impregnate the immersed portion. Because of the great reduction in interfacial tension between the silicon carbide and the impregnating material at the critical temperature, I am able to heat together the end of the rod and a small execs of the impregnating material until the critical temperature is reached, .whereupon a considerable JH'OPOI'tlOIl of the total amount of metal use is absorbed into By heating the impregnating material and the end of the resistor together in this manner from a comparatively low temperature, I am able to secure an end which is completely impregnated throughout its cross section without the disruption and mechanical weakening which occur with the 1 usual dipping process. By further heating to a higher temperature I am able to remove by vaporization the small excess of impregnating material not absorbed into the rod, so that the original contour of the element is retained without any crust or globules of excess metal adhering to the surface of the element.

Upon still further heating of the end of the element to a very high temperature after the companying drawing.

In the drawing:

Figure 1 is a diagrammatic view showing a silicon carbide resistor positioned in a type of crucible suitable'for impregnation and surrounded by granular impregnating material;

Figure 2 represents a deeper crucible than that shown in Figure l, the crucible being refilOVflblG from the high temperature source of Figure 3 is a sectional view showing an impregnated resistor in which the junction between the impregnated and unimpregnated uminum, nicke cohalt and- I 3 Y i,ooe,eea

portions presents a conical rather than a plane surface. I The resistors treated can be of the ty commonly known and sold under the tra name of Globar, although other types of silicon carbide resistors can also be used. The Globar resistors are self-bonded by a process of recrystallization, and I have found that the previously recrystallized material is readily impregnated and gives an end of high electrical conductivit without being weakened at the juncture between the imregnated and non-impregnated portions. e rec stallized material does not contain a bond w ich is fusible at the temperature of impregnation, and it retains its stren h up to very high temperatures, so that during the impre ation process there is no difliculty from t e material losing its shape.

The impregnating material which I prefer to use is silicon or an alloy of silicon, because of the. tou hness and resistance to oxidation of the resu tin product. In the case of silicon the critica impre ation tem rature is approximately 1900 although it should be realized that the exact measurement of the critical temperature is diflicult because of the fumes evolved in the process.

1 The use of silicon for impregnation pses a'number of advantages in com arison with most of the common metals. e siliconized end produced by my process is very tough, and being non-porous it can be operated at tem ratures up to 1400 C. or even higher, with with very little change in electrical resistance.

"In carrying out my process, the resistor can be immersed in a slight excess of the impregnating material at a temperature only sli htly above the melting point of the latter, an the end of the resistor and the surrounding material heated together to the critical temperature. With the higher meltin metals, and es ially with silicon, I have ound, however, t at a much stronger end can be produced by positioning the resistor in a crucible and surrounding it with the impregnating material in solid, and preferably in granular form. The crucible is then heated rapidly through the melting point of the impregnating material to the critical tem rature, at which temperature the metal is a sorbed into the rod almost instantaneously. The resid- .perature is suilicient to vaporize silicon, as

well as most of the common metals and alloys.

Onemethod of carrying out my invention is illustrated diagrammatically in Figure 1. The crucible 2 is preferably a carbon block,

only a surface oxidation and which is adapted to be heated almost instantaneously to very high temperatures, as for i example, by resistance heating using a very L heavy current. The capacity of the crucible. 5 is only slightly reater than the volume of by making the electrodes of suflicient size it is possible to conduct a currentof from 10,000 to 20,000 amperes through the crucible if desired. The resistor 4 or its end portion is preferably preheated before positioning it in the crucible. Preheating is especially desirable with elements of large diameter in order to reduce thermal shock, but it is not absolutely necessary in all cases. After positionin the end of the resistor in the crucible, the end is surrounded with a predetermined quantity of impregnating material, preferably silicon or an alloy of silicon, the amount added being only slightly in excess of that neces- Tary to impregnate the resistor to the desired en h.

shall describe the heating process as adapted for impregnation with silicon, although it will be realized that various modifications can be made, especially in the case of other impregnating materials. With silicon the heating should be comparatively rapid, owing to the tendency of the material to oxidize, vaporize-and form silicon carbide by reaction withthe carbon crucible before impregnation takes place. It is very difiicult to melt a small quantity of silicon in a carbon crucible and subsequently immerse the end of the resistor in the melt because of the extreme reactivity of the silicon and I have found that by surroundin the end of the rod with solid silicon and rapidly heating the crucible to the impregnation temperature, the silicon is absorbed into the rod, and the absorption is so nearly instantaneous that there is practically no difiiculty from contamination of the silicon with oxide or silicon carbide before impregnation takes place.

The crucible and its contents, as illustrated in Figure 1, are heated rapidly until the excess metal is removed from the crucible by vaporization. The crucible and the end of the rod are still further heated for a very short interval, as for example, from 5 to I 30 seconds, depending upon the rate at which heat is applied, in order to produce a further decrease in electrical resistance of the impregnated material. The final crucible temperature. as nearly as can be measured, is from 2500 to 2800 C., and care must be taken during the final heating period to control the process so as not to completely decompose the end of the rod to form graphite. The entire heating period can be from one to five minutes, and should not be extended over a period greatly exceeding ten minutes. This extremely rapid heating can be efiected either by passing a hea current directly through the crucible, or y insertin a removable crucible into a previously eated high tem rature furnace, as indicated b Figure 2 in which the removable crucible 18 inserted in a furnace 7, the furnace being represented conventionally. High frequency induction can also be used for rapid heating if desired.

Bv conducting the final vaporizin process while the element is positioned in t e crucible,'the crucible can be boiled completely dry, and the rod upon removal possesses a contour identical with that of the original element before treatment. If the element is removed after heating to about 1900 to 2000 C. impregnation is com late, but sufiicient excess sillcon usually ad eres to the rod so as to give it a somewhat irregular contour and an unsightly appearance.

' The additional heating of the resistor to a very high temperature after impregnation is complete produces a marked increase in electrical conductivity of the impregnated portion of the rod. For example, with a silicon carbide body having an original spe-' cific resistance of about 0.4 ohm per centimeter cube at the operating tem rature of the resistor, the impregnated and before the high temperature treatment will have a specific resistance of .02 to .03 ohm per centimeter cube at the operating temperature of the resistor, whereas after high temperature heating this value can be reduced to approximately .01 ohm per centimeter cube.

loying action between the silicon and the silicon carbide, and possibly to a partial decomposition of the silicon carbide into silicon and graphite. Under the microscope the silicon in the pores of the article seems to dissolve the smaller particles of the silicon carbide and also the surface of the larger particles, and the further the heating process s continued. the nearer the product resembles a true alloy instead of an aggregate in which the silicon merely fills the pores. As the decomposition temperature of the silicon carbide is approached or possibly slightly exceeded, the end becomes darker in color, and Ibelieve that this material contains some graphite either in solid solution or otherwise alloyed with the silicon or the silicon carbide. Although the end may be partially decomposed it is still very resistant to oxidation. and at the same time it possesses a very high electrical conductivity. I have found that this material will retain its high conductivitv and afford aivery satisfactory cold end throughout the'life of the element.

The alloying or solvent action of molten silicon on silicon carbide is further evidenced by the vigorous reaction and disruption of "I believe this change is due to further althe end of the rod when itisdipped into molten silicon at very high temperatures. With a large execs of silicon, such an action may cause a weakening of the-bond between the individual soastocanse a artial dissolution or I ption of the arti e. In m process any allo or solvent'action es place in situ a the silicon has been absorbed into the rod. 4 19 Although in out my proce a the heating may be of veg rapidly, there is no abrupt change in e temperature gradient at any one point, as is the case if an attempt be made to dip the silicon carbide rod into a superheated bath of molten silicon, which causes a very abrutgt change in the temperature gradient of e rod at the level of the molten bath. By m process I am able to obtain a high degree mechanical strength at the junction between the impregnated and unimpregnated portions of the rod. This junction is the chief source of dificulty with an end of this type, and by certain modifications in the rocess I am able to still further improve t e mechanical properties of the treated resistor.

In the case of silicon the addition of aluminum, preferably in quantities of from approximately 2 to 20 percent, increases the toughness of the junction and minimizes breakage at this int. The addition of aluminum as an al oyin'g ingredient to the silicon also seems to lower the temperature of final heating necesary to produce the high conductivity end, and if a partially decomposed end is desired, the proces need not be subject to as accurate control to prevent complete decom tion of the end of the rod as is the case w en silicon alone is used as the impregnating material, and the end of the rod heated to a higher temperature. heThe telidency to fracituredat the junctigg tween t e impregna an unimpregna portions can also be minimized by changing the shape of the junction so that it does not present a plane surface. The usual juncture is fiat, and represents approximately the level of the impregnating li uid. By using a comparatively deep crucib e, as shown in Figure 2. so that the element itself becomes heated above the critical impregnation temperature for a considerable distance beyond the level of the liquid, the impregnating material rises rod than near the outer surface, so that the juncture is approximatelv cone shaped, as indicated in Figure 3. Such a juncture is not only less readily fractured than the usual to a higher level in the central portion of the,

means oestrus characteristics of the material impregnated by my procem is that it may be rendered complete y impermeable to gases.-

The resofthearticlecanbefilledcomplete y with nmlten metal, which is not the case when the metal is incorporated into the onglhnal mix. When measured by the usual me ods of determining the gas rmeability of refractories, the gas permea ility of m material is zero, whereas the usual recrysta lised silicon carbide shows every high permeability when subjected to such a test. A method of measuring the permeability of refractories, together wi the experimental determination of the permeability of rec tallized silicon carbide, is 'ven in the Bu letin of the University of inois, volume 23, No. 50, August 17, 1926, (Circular No. 14, enginee experiment station).

With a solid material oxidation is limited tothe outer surface, whereas with the porous material resulting from the incor ration of the metal into the original mix, rapidly oxidized throughout the body of the article. a

My method of impregnation afiords a means for producing a si icon carbide bod in which the pores are completely filled with metal, but in which the original shape of the article is retained. It is 'ble to'obtain a very smooth surface a r impregnation, especially when the body is composed of fine grit silicon carbide. For this reason my process is very advantageous for the imp nation of entire shapes of silicon carbi e when the preservation of shape and surface is important, as is the case with bricks and other articles which must be made to size.

It is thus possible to obtain a non-porous refractory article of accurately controlled shape which w: veliy high thermal and electrical conductivity. n carrying out such a process it is desirable to recrystallize the silicon carbide before impregnation by'heating the article in a non-oxidizing atm here to a very high temperature at which t e individual crystals grow together or become self-bonded. The formed article is then surrounded with a small excess of the impregnating material, heated to the critical temperaturenry for rapid impregnation, and the excess'of the impregnating material then removed by vaporization.

By the term metal in the present specification and claims. I mean bot the elements e metal is its classified chemically as metals, and also the so-called metalloids? such as silicon, which are commercially designated as metals and are commonly 'used as such for alloying pur- This application is a continuation-in-part ofmy earlier filed application Serial No. 882,- 004, filed July 29, 1929.

I claim:

1. The method of impregnating a silicon carbide-article with metal, which comprises surrounding the article with metal at a temperature substantially below the tempera into ture of rapid impregnation of the meta the article, and heating the article in contact with the surrounding metal to a temperature sufficient to cause a rapid flow of molten metal throughout the pores of the article;

2. The method of impregnating a silicon carbide article with metal, which comprises the article and completely permeate the said article.

3. The method of impregnating a -silicon carbide article with metal which comprises surrounding the article with a small excess of metal over the amount necessary to impregnate the article, the surrounding metal being at a temperature substantially below the temperature of rapid impregnation of the metal into the article, heating the article and the surrounding metal to a temperature at which the metal in the molten state flows rapidly into the pores of the article, and further heating the article until the excess metal is removed by vaporization.

4. The method of impregnating a silicon carbide body with metal which comprises surrounding the article with solid metal in excess of the amount necessary to impregnate the article, heating the surrounding metal in contact with the article to a temperature sufficient to liquefy the said metal and cause it to flow rapidly into the pores of the article, and further heating the article until the excess metal is removed by vaporization.

5. The method of making a shaped body of siliconized. silicon carbide which comprises surrounding a previously formed silicon carbide article with solid silicon, and heating the article and the surrounding silicon to a temperature above approximately 1900 C. s

6. The method of making a shaped body of siliconized silicon carbide which comprises surrounding a previously formed silicon carbide article with a small excess of solid silicon over the mount necessary to completely impregnate t e article, heating the article and the surrounding silicon until the silicon in a molten state rapidly permeates the pores of the article, and further heating the article until the excess silicon is vaporized and the contour of the original shaped article is restored.

7. The method of increasing the electrical conductivity of the terminal portions of a silicon carbide resistor, which comprises surrounding the end portionof the resistor with metal while the metal is at a temperature substantially below the temperature of rapid impregnation of the metal into the pores of the resistor, and heating the resistor and the surrounding metal until the metal in a molten condition completely permeates the end portion of the resistor throughout its entire cross section.

8. The method of increasing the electrical conductivity of the terminal portion of a sili-, con carbide resistor, which comprises surrounding the end portion of the resistor with solid silicon and rapidly heating the end of the resistor and the surrounding silicon to a temperature above approximately 1900 C.

9. The method of increasing the electrical conductivity of the terminal portion of a silicon carbide resistor which comprises sur-' rounding the terminal portion of the resistor with a small excess of solid silicon over that necessary to impregnate the end of the resistor to a predetermined length and rapidly heating the resistor in contact-with the silicon to a temperature of approximately 2500 to 2800 C.

10. The process described in claim 9 in which the period of heating is limited to less than ten minutes.

11. The method of increasing the electrical conductivity of the terminal portion of a silicon carbide resistor which comprises surrounding the end of the resistor with silicon at a temperature substantially below the temperature of rapid impregnation of silicon into a porous silicon carbide body, heating the end of the resistor and the surrounding silicon until the silicon in a molten condition flows into the pores of the end portion of the resistor, removing the excess silicon by vaporization, and further heating the resistor after the excess silicon has been driven off.

12. In the impregnation of silicon carbide with a molten impregnating material consisting substantially of silicon, the step comprising adding a minor proportion of aluminum to the silicon.

13. The method of increasing the electrical conductivity of the terminal portion of a silicon carbide resistor which comprises surrounding the end of the resistor with solid metal, and rapidly heating the end of the resistor and the surrounding metal to a very high temperature until incipient decomposition of the silicon carbide contained in the end portion of the resistor takes place.

14. The method of increasing the electrical conductivity of the terminal portion of a silicon carbide resistor which comprises surrounding the end of the resistor with solid silicon, and rapidly heating the end of the resistor and the surrounding silicon to a very high temperature until incipient decomposiof the silicon; carbide contained in the end portion of the resistor takes place.

15. A silicon carbide resistor having terminal portions impregnated with aluminum silicon allo 16. A 'con carbide resistor hav' terminal portions im regnated with me the juncture between e impregnated and the unimpregnated portions of e resistor being substantially conesha 17. A sihcon carbi e resistor having terminal portions im regnsted with silicon, the juncture between t e impregnated and unimpregnated portions being substantially cone s ram my hand.

ALBERT H. HEYROTH.

Images