US2708255A - Minute metallic bodies - Google Patents

Description

y 1955 5. Y. WHITE 2,708,255

' MINUTE METALLIC BODIES Filed June 18, 1949 J 7 "l// w a j 4 V v 4 INVENTOR.

Sidney Wh/fg A T. TORNE Y5 MINUTE METALLIC BODIES Sidney Y. White, New York, N. Y., assignor of one-third to Albert C. Nolte, New York, N. Y.

Application June 18, 1949, Serial No. 99,926

7 Claims. (Cl. 317-239) This invention relates to metallic pellets small enough to be formed into pellets by surface tension when molten, and the production of them, and more particularly to pellets of germanium suitable for use as rectifying crystals in the electrical art.

Development of the electrical crystal art has been substantially suspended from 1925 until a recent date, but has been stimulated to great activity by the need for radar crystals which developed during the recent war.

Silicon has proved its worth in radar, but has some weaknesses which it would be desirable to overcome, the main one being the low burn out point. It was suspected that germanium might be advantageously investigated as a substitute for silicon because of its position in the periodic table next to silicon. Such investigation has led to the substantial purification of germanium, to the successful production of diode rectifiers employing germanium, and to the production of more or less satisfactory amplifying triode transistors employing germanium.

Drawbacks to the employment of germanium crystals at the present time reside in the high cost of the crystals and in the fact that the crystals are not available in sulficient quantity. Both drawbacks result in large measure from the expensive and wasteful procedures presently employed in the manufacture of the crystals.

Germanium crystals are made according to present commercial practice by a very precise metallurgical procedure, and they are cast in rather large ingots to which certain modifying ingredients known as dopes are added to give the correct amount and sign of conductivity. By complex heat treatment, these dopes are caused to be distributed as uniformly as possibleyand huge crystals are grown in the ingots.

The ingots are then cut into slabs by diamond gang saws, and the slabs are further similarly sectioned to l I form approximately one millimeter cubes. Only the center of an ingot is active, so that the described procedure involves a very high percentage of waste. The material is extremely hard and brittle, and cannot be cut into pieces having a thickness of substantially less than one millimeter. As a result of the waste and of the large minimum size of the resulting germanium bodies, it is not possible to secure substantially more than .one thousand useful crystals from a pound of material. Treatmentof a large charge followed by sub-division does not favor uniformity of product because of the unequal dope distribution which results from segregations of solids and/or incomplete dispersion of gases. The crystals produced from a single ingot are, therefore, not identical:

The limitation of subdivision is wasteful because a considerably smaller body will operate more satisfactorily h t ompa ly la g bo e ref rre to s produced y pr mmerc a Practices. t is gen r y agr tha o si e ab y less t an o e micrononnd of at ri l i ual y usef l. in a si gle un n that a g ma m u i of he greate t utility would have di- 2,738,255 Patented May 10, 19 55 mensions roughly of .004 long, .002" wide, and .002 thick.

The world production of germanium in 1948 was one ton. Under present commercial practices this means that only between two and three million crystals per year can be produced. Even if, as is anticipated, the rate of germanium production can be gradually increased to six tons per year, the crystals which could be produced by present methods would be far from adequate to satisfy commercial requirements. The price, moreover, would continue to be high.

It is the primary object of the present invention to provide for the manufacture of germanium crystals of smaller dimensions, and with less waste of material than under the hitherto available commercial procedures, so that a much greater number of serviceable crystals can be had from a given quantity of material and at less cost per crystal.

It is a further object of the invention to produce germanium crystals of more nearly identical and predetermined characteristics than has heretofore been possible.

As has been indicated, the method and apparatus can also be used advantageously in the production of steel pellets.

Other objects and advantages will hereinafter appear.

In the drawing forming part of this specification:

Fig. 1 is a plan view, broken away intermediate its ends for compactness of illustration, showing a measuring plate which is employed for measuring predetermined quantities of the pellet forming material and for holding the material during subsequent treatment;

Fig. 2 is a fragmentary sectional view taken upon the line 22 of Fig. 1, and showing measured powder charges in the recesses;

Fig. 3 is a view similar to Fig. 2, but showing balls in the recesses which result from the reducing and melting of the metal containing powder;

Fig. 4 is a sectional view of one of the pellets or microingots set in a holder prior to grinding and lapping thereof; and

Fig. 5 is a sectional view of the finished germanium unit in its holder with a pair of cat whiskers applied.

The starting material for the manufacture of germanium crystals is germanium dioxide powder. This is a very loose and free-flowing white powder. It is derived as a by-product in the production of lead and Zinc. The producer chooses a'point in the production of lead or zinc where the germanium concentrates naturally occur. At this point the concentration of germanium is of the order of one gram of germanium per ton of other material. The material containing the germanium is then processed to produce the germanium dioxide in powdered form. The germanium dioxide as thus supplied has eight isotopes and about ten detectable impurities of total con- Ceniration of two hundred parts per million. These impurities vary somewhat from batch to batch, and some give positive conductivity while others give negative conductivity, the over-all conductivity often being negative. In every case the starting material consists chiefly, and almost entirely, of germanium dioxide. I

The doping of large ingots as practiced in the prior art is usually with solids. such as tin, antimony, bofon, etc, it being possible to mix and distribute fairly uniformly through the ingot the necessary eighty or one hundred parts per million of this alloying material.

It has also been the practice in connection with these large ingots to employ nitrogen, helium or other gas as the dope. In fact, I have secured .the best results, for transistor use, with nitrogen doped crystals made under the old procedure. There is great difiiculty, however, in having the large ingot absorb sufficient nitrogen, as a 3 nitride shell forms on the ingot and germanium nitride is useless. Neither does the size of the ingot favor uniformity of dispersion of the aborbed gas. The dopes are added in such small quantities that the final pellets are still caused to consist chiefly, and almost entirely, of germanium.

In accordance with the present invention, the germaniurn dioxide is first subdivided into predetermined quantities adapted to be formed directly into minute units which are so small as to require no subsequent subdivision.

- These small measured quantities are thereafter simultaneously subjected to a reducing treatment, then to a doping treatment and finally to a heat treatment, a very largenumber of the small units desirably being dealt with simultaneously.

For the purpose of measuring the unit quantities of germanium oxide and containing them during the reducing and doping treatments, a block or plate 1 of a suitable non-Wetting, refractory material is provided, having a large number of recesses 2 in its upper face. The plate may desirably be about one foot square, and the recesses 2 may be provided at the rate of ten to the linear inch or one hundred to the square inch, so that even after leaving a suitable free margin for handling, a total of ten thousand of the recesses or cavities can be readily provided in the upper face of a single plate. The recesses are made as nearly as is practicable of the same size, and are desirably of uniform shape, the shape as shown being hemispherical. The recesses may be made cylindrical, with hemispherical bottoms, but it is important that the radius of the recess be greater than that of the reduced and liquefied charge, so that there will be no adhesion of the finished product to the plate.

Each recess may be made of a size to measure the quantity of material required to produce a sphere small enough to assume a spherical form by surface tension when melted and desirably not more than two millimeters in diameter. Spheres can be thus produced without waste of material, and hence with an enormous gain of efficiency whether the spheres be near the upper limit of size that can be formed by surface tension or considerably smaller. It is recognized, however, that spheres below the upper limit in size would be desirable, and it is contemplated that as skill is gained in dealing with these tiny bodies. the sphere diameter may be reduced sufficiently toenable spheres to be produced at the rate of substantially a million to the pound.

The loading and measuring technique consists of piling the plate with powder to a depth of a half inch or so, and then vibrating the plate vigorously to insure constant density and close correspondence of the amount of powder in each one of the depressions 2.

The loaded plate is desirably subjected successively to vibrations at the rate of two cycles per second, 60 cycles per second, and in excess of ten thousand cycles per second, all of high amplitude.

When this has been done uniform pressure is desirably applied over the top of the powder by means of a weight having a smooth, plane lower surface. It does not appear, however, that this step is essential.

A Wiper blade is then utilized to scrape or wipe off the surplus powder and give a sharply defined upper limit to the filling in each of the recesses 2. The wiper blade may be formed with a straight rubber edge, or with a straight metallic edge.

. Throughout these operations it is important to guard the material and the apparatus against contamination either by dust in the air, or by exposure to high humidity, or by contact with the person or clothing of the operative. These factors may be controlled by careful handling in an air conditioned room.

.A'number of loaded plates can be simultaneously inserted in an electric furnace of small dimensions. In practice five of the plates can be simultaneously inserted in a two cubic foot furnace to treat fifty thousand charges at a single operation. The material placed in the furnace is subjected first to reduction by hydrogen and then (if gas doping is to be employed) to doping by nitrogen or helium, but all with a single heating so that it is not necessary to allow the material to cool and solidify after reduction and then be reheated for doping.

When the plates have been inserted in the furnace a vacuum is developed in the furnace, the furnace is heated to and maintained at the reducing temperature, and the furnace is then supplied with hydrogen under a pressure corresponding approximately to one millimeter of mercury. A flow of hydrogen is established and the powder is reduced in afew hours. The reduction stops automatically when completed, and hence no harm is done by permitting a continuation of the flow of hydrogen after the reduction is complete. The germanium is reduced at an approximate temperature of 1800 F., the melting point of germanium being approximately 1754 F. It is important to avoid substantially higher temperatures than 1800" F. because at higher temperatures germanium monoxide forms and leaves the furnace as a brown gas, taking much of the germanium with it.

In molten form the small germanium charge forms itself into a sphere slightly flattened at the bottom, and this form it retains when solidified.

The next operation after the reduction is complete is the introduction of a doping or modifying gas, a preferred one being nitrogen, although helium is commonly employed, and other gaseous dopes or modifiers are known in the art. The doping is carried into effect while maintaining the charges in a molten state.

Taking nitrogen as an example, it is introduced at a pressure corresponding to a fraction of a millimeter of mercury to be absorbed by the tiny ingots. It is preferred to have this take a very long time (say two to four hours) in order to secure a high order of control and great diffusion of the nitrogen in the ingot. In the very small spheres here under consideraton, the absorption and diffusion of the nitrogen is greatly favored because of the high ratio of surface area to volume, and because of the small radius. It is inevitable, however, that greater concentration will appear in the shell of this sphere than in the interior thereof. Accordingly, the surface area does not have the desirable characteristics of the interior.

After proper dope concentration has been obtained by experiment, (it being readily possible to have too much dope so that the proper doping treatment must be determined by experiment), the furnace is cooled at a slow rate in order to promote the growth of crystals in the center of the sphere. The cooling from the maximum temperature down to room temperature preferably extends over at least four or more hours.

The heat treatment of the tiny germanium ingots has three objects. The first object is to control the crystal growth. As a general rule the largest possible crystals are desired. In special cases, however, very small crystals may be desired. A product embodying small crystals may be obtained by cooling the ingots more rapidly.

The second object of heat treatment is to prevent undesired segregations such as would be likely to occur if the source of heat were suddenly cut off or rapidly diminished.

The third object of heat treatment is to give special properties to the surface, since there will generally exist an outer shell on the sphere of greater dope concentration and often with no crystal structure.

Each resulting sphere can be pressed into a preformed recess 3 which is provided in a holder or slug 4 of soft metal.

If nitrogen is used as the dope, a nitride shell 5 is formed on. the sphere enclosing the crystal center 6. By a grinding and lapping operation the sphere is seetioned down to a plane such as that designated by -the line H- H of Figs. 4 and 5 to exposethe crystal center of the sphere to contact with'cat whiskers 7 or other suitable terminals. A complete assembly is indicated in Fig. 5, the sectioned sphere 8 being embedded in the material of the holder 4, and having on it the cat whiskers 7.

As an alternative the active center of the sphere could be uniformly exposed by the utilization of well known bearing ball grinding techniques. This would give a sphere whose surface is uniformly active to some degree.

A similar result can be dbtaine'd by employing helium as the dope, since helium forms no harmful products in the shell as nitrogen does.- The best cry'stalline formation would still occur inside the sphere and this, in conjunction with the fact that a fiat face is desired for contact would still require that some grinding be done.

The utilization of solid dopes or modifiers is also withn in the scope of the invention. The dope in the form of one or more metallic oxides or one or more metallic powders is first mixed in proper proportion with the germanium oxide. The charges are then measured as described, the hydrogen reduction is carried out as de scribed, and the furnace is slowly cooled as described, the step of gas doping being omitted.

The sphere must be Well supported by the holder, as the grinding and lapping operations are severe and tend to dislodge the sphere.

The sphere can be united with the holder in any one of several ways. For example, the holder may be cold at the time of application and a forced fit may be obtained by forcing the sphere into a prepared undersize hole in the holder. As one alternative, the holder may be heated sufliciently to make it plastic but not fluid. In this case less strain is imposed upon the germanium sphere when embedding the sphere in the metal of the holder.

The third method consists of melting the holder metal, and then permitting it to set with the sphere embedded in it.

It the holder is either heatedor molten, it should have a coetiicient of expansion so related to that of germanium that a firm grip of the holder on sphere will be had after cooling.

The plate must be rigid and must retain its shape when used in the furnace. It must not decompose under the conditions of service. horizontal upper face on which the depressions are formed.

It should not absorb germanium and desirably should not absorb nor adsorb the gases used for reducing and doping. cally and/ or mechanically each time it is used. It should, itself, be non-contaminating. The plate surface must be of a material which is not wetted by the molten metal. The size of sphere is limited by the surface tension or self-attraction of the molten metal. The word sphere as used herein is not intended to mean necessarily a geometrically perfect sphere, but a pellet of general round shape.

The moderate temperature of reduction and the fact that the plate is used in an inert or reducing atmosphere at all times means that ceramic plates can be produced and utilized which will have an indefinite life in the furnace. The plates may, for example, be advantageously made of one of a series of beryllium ceramics developed by Stupikoif, which are commonly fired at 3200 F.

The plate, if of ceramic, should be about one foot square and one inch thick and should be of a composition adapted for firing at not less than 2250 F. It is probably essential that the plate be vitrified, or that it be glazed with dust, and it is desirable that the plate be both vitrified and glazed with dust.

The plate may be made of metal, but if so it should be enamaled or otherwise protected to render it noncontaminating.

I have described what I believe to be the best embodi- It should have a smooth, fiat,

It should be capable of being cleaned chemi- 6 ments of my invention. 1 do not wish, however, to be confined to the embodiments shown but what I desire to cover by Letters Patent is set forth in the appended claims.

I claim:

1. A spherical body consisting almost entirely of germanium, which has assumed a spherical form through surface tension while molten and which is not more than two millimeters in diameter, said sphere including a crystalline center and an amorphous shell.

2. A spherical body consisting almost entirely of germanium, which has assumed a spherical form through surface tension while molten and which is not more than two millimeters in diameter, said sphere including a small proportion of modifying material, the modifying material being included in ncreasing concentration from the interior outward to the circumference.

3. A spherical body consisting almost entirely of germanium, which has assumed a spherical form through surface tension while molten and which is not more than two millimeters in diameter, said sphere having the physical and electrical properties characteristic of the body having been reduced from germanium dioxide and heat treated as a separate unit.

4. A spherical body consisting almost entirely of germanium, which has assumed a spherical form through surface tension while molten and which is not more than two millimeters in diameter, said sphere having the physical and electrical properties characteristic of the body having been reduced from germanium dioxide, modified and heat treated as a separate unit.

5. In combination, a spherical body consisting almost entirely of germanium, which has assumed a spherical form through surface tension while molten and which is not more than 2 mm. in diameter, and a metallic holder in which the spherical body is fixed, said body having the physical and electrical properties characteristic of the body having been reduced from germanium dioxide and heat treated as a separate unit, the body having its protruding side partially ground away to provide a flat face and to expose an interior portion of the body.

6. In combination, a spherical body consisting almost entirely of germanium, which has assumed a spherical form through surface tension While molten and which is not more than 2 mm. in diameter, and a metallic holder in which the spherical body is fixed, the spherical body having the physical and electrical properties characteristic of the body having been reduced from germanium dioxide, modified and heat treated as a separate unit, said body being partially ground away on its protruding side to provide a fiat face in which the inner portion of the sphere is exposed.

7. A spherical body consisting almost entirely of germanium, small enough to assume a substantially spherical shape by surface tension when molten, and having the physical and electrical properties characteristic of the body having been reduced from germanium dioxide into a molten substantially spherical shape as a separate unit, and thereafter slowly cooled from a molten condition to atmospheric temperature over a period of several hours.

References Cited in the file of this patent UNITED STATES PATENTS 1,130,197 Rafn Mar. 2, 1915 1,175,693 Bosch Mar. 14, 1916 1,211,754 Rawls Jan. 9, 1917 1,736,495 Graft Nov. 19, 1929 2,215,215 Garkisch Sept. 17, 1940 2,394,727 Taylor Feb. 12, 1946 2,441,590 Ohl May 18, 1948 2,441,603 Storks et a1. May 18, 1948 2,474,966 Addink et al. July 5, 1949 2,504,628 Benzer Apr. 18, 1950 (Other references on following page) UNITED STATES PATENTS 2,615,965 Amico Oct. 28, 1952 FOREIGN PATENTS 891,064 France Nov. 29, 1943 5 OTHER REFERENCES OSRD Report 14-341, November 1, 1944-Purdue University, preparation of High Voltage Germanium Crystals. Entire report has 28 pages. Pages 3 and 11 10 relied upon.

The Electrochemical Society, Preprint 8918. Technology of Germanium, by Jafiee et al. Received in Patent Ofi'ice Library May 29, 1940. Pages 211-212. Entire report 12 pages.

Further Developments in the Preparation and Heat Treatment of Germanium'Alloys. Division 14, NDRC Report No. 576. October 31, 1945. Page 6. Entire report 11 pages.

Germanium VII, by Dennis et a1. September 1923. Pages 2034, 2036, 2037, 2040. Entire report 14 pages. (Copy in Division 3).

Claims (1)

1. 6. IN COMBINATION, A SPHERICAL BODY CONSISTING ALMOST ENTIRELY OF GERMANIUM, WHICH HAS ASSUMED A SPHERICAL FORM THROUGH SURFACE TENSION WHILE MOLTEN AND WHICH IS NOT MORE THAN 2MM. IN DIAMETER, AND A METALLIC HOLDER IN WHICH THE SPHERICAL BODY IS FIXED, THE SPHERICAL BODY HAVING THE PHYSICAL AND ELECTRICAL PROPERTIES CHARACTERIS-

Images