US3520977A - Arc plasma heating device - Google Patents

Description

D. H. PETERSEN ET Al. 3,520,917

ARC PLASMA HEATING DEVICE 2 Sheets-Sheet 1 I r l I IO 6O so iOO PLASMA RESISTANCE (OHMS/cmf DONALD H. PETERSEN WARREN C. SCHWEMER, INVENTORS July 21, 1970 o w w w o 4 2 Q 8 3 3 2 2 Y L mmnhdmmnzzmh 32m 4m July 21, 1970 D. H. PETERSEN EFAI- 3,520,977

ARC PLASMA HEATING DEVICE Filed Dec. 23, 1968 2 Sheets-Sheet 3 IIIIIIIII: IIIIIIIIIA ARGON-NITROGEN L, (M BY VOLUME) I 04- z a. 5

U1 0 @2- HELIUM 8 NITROGEN J m l 1 1 I 20 3o 40 so so 10 ELECTRODE POTENTlAL, VOLTS DONALD H.PETERSEN WARREN c; SCHWEMER, IINVENTORS BY M V K- 2 71 4 AGENT United States Patent 3,520,977 ARC PLASMA HEATING DEVICE Donald H. Petersen, Dallas, and Warren C. Schwemer, Arlington, Tex., assignors to Ling-Temco-Vought, Inc., Dallas, Tex., a corporation of Delaware Filed Dec. 23, 1968, Ser. No. 786,140 Int. Cl. Hb 7/18 US. Cl. 13-1 12 Claims ABSTRACT OF THE DISCLOSURE An electric furnace adapted to be heated with a gaseous plasma, and including within the furnace a quantity of low work function material which promotes the establishment and sustenance of the gaseous plasma, said quantity being on the order of about onemilligram per square centimeter of furnace area.

This invention relates to electric heating means and more particularly to means for heating an enclosure with a gaseous plasma.

The use of a plasma generator to provide a gaseous stream at very high temperatures is well known, and even relatively sophisticated devices such as that described in U.S. Pat. No. 2,972,969 to A. R. Kantrowitz, et al. have been designed. Devices such as the Kantrowitz generator, however, suffer from the difiiculty encountered in reestablishing an are between two spaced electrodes if the arc is deliberately interrupted (as a means of temperature control) or is inadvertently interrupted due to a momentary power failure, etc. That is, plasma generators of the type describedby Kantrowitz et al. are generally serviceable only when a continuous electrical discharge between electrodes is tolerable or desirable.

While the US. Pat. Re. 25,925 to J. M. Beasely et al. discloses a method of heating an enclosure with a gaseous plasma in which an arc can be re-struck with ease if it is ever interrupted, the Beasley et al. furnace also suffered from a liability, namely, a temperature limitation. That is, the Beasely et al. furnace is re-startable without bringing the hot electrodes back into very close proximity only if the temperature in the furnace is never allowed to drop below2800 F. Thus, as stated by Beasely et al. at column 8, lines 38-47, the argon-enriched atmosphere in the furnace interior must be kept at or above 2800 F. in order to realize the advantageous re-starting characteristics of the furnace.

Hence, it has been shown that plasma heating devices in general are old. Furthermore, it has been shown that,

such devices can be adapted to heat the interior of the type of furnace that might be used for melting materials having relatively high melting temperatures; It has also been shown that an arc in a plasma furnace can be re struck without bringing the electrodes together as long as the atmosphere in the furnace is enriched with a noble gas and the furnace interior'is maintained at or above 2800 F. But, heretofore, a method of achieving the advantages of a re-sta'rtable plasma furnace has not been disclosed which obviated the need to maintain the furnace at a-relatively high temperature of 2800 F.

Accordingly, it is a major object of this invention to provide a plasma furnace which is re-startable at temperatures as low as 1650 F.

It is a further object to provide a plasma heating device which is stable at temperatures as low as 1300 F.

Another object is to provide a low-temperature plasma furnace which is operable with a wide variety of gases.

A further object is to provide a method of maintaining a furnace at a desired temperature in a manner such that the furnace is characterized as being self-regulating.

3,520,977 Patented July 21, 1970 "ice Yet another object is to provide a means for promoting the establishment of a gaseous plasma in an arc-heated enclosure.

These and other objects and advantages will be apparent from the specification and claims and from the accompanying drawing illustrative of the invention.

In the drawing,

FIG. 1 is a sectional view in elevation of a furnace in which the present invention can be practiced;

FIG. 2 is a plot of plasma temperature versus plasma resistance in a furnace similar to that shown in FIG. 1, and wherein the gas in the furnace cavity is helium;

FIG. 3 is a sectional view in elevation of an alternate configuration of a furnace cavity, in which a horizontal ledge is provided on which material may be placed to separate it from the melt;

FIG. 4 is a sectional view in elevation of an alternate furnace configuration in which the material is to be melted can be isolated from the gaseous plasma;

FIG. 5 is a plot of operating voltage versus electrode gap for several gases; and

FIG. 6 is a sectional view in elevation of a furnace having holes in the walls which are filled with solid rods of low work function material.

With initial reference to FIG. 1, the furnace generally indicated-by the numeral 10 has a housing 11 with a cavity 12 therein which is to be heated. Such a furnace 10 is described in great detail in US. Pat. Re. 25,958 to Beasely et al., and therefore only those parts which are truly essential for an understanding of the present invention will be repeated here. It should be noted, however, that while this invention is advantageously used with a furnace which will hereinafter be referred to as a Beasleytype furnace, it is also useable with any other furnace which is capable of supporting a gaseous plasma. That is, any furnace which has two or more electrodes that are connectable to a source of AC or DC power, and also has a connection for admitting a gas which is capable of forming a gaseous plasma, will serve adequately as the nucleus for practicing the instant invention.

Extending into the cavity 2 are two electrodes 13, 14 which are separated by a gap 15. The pair of electrodes 13, 14 quite obviously are surrounded by the gas in the cavity, so that the gas will be heated when an arc is struck between the electrodes. In those instances where it is desirable to establish an are between a first electrode and the charge or melt (which is in electrical contact with a second electrode), it should be apparent to those skilled in the art that functionally the charge/melt constitutes a physical (and electrical) extension of the second electrode. Hence, it is appropriate to refer to two electrodes extending into the cavity even if only one of the electrodes actually has the typical elongate shape illustrated in the drawing. The initial heating of the gas to a plasma-forming temperature is conveniently accomplished with the very same electrodes that operate to sustain the gaseous plasma, but initial heating could be achieved in some alternate manner without departing from the concept of the invention.

A source of AC or DC electrical power 16 is connected through suitable leads and insulated connectors to the two electrodes 13, 14. As will be explained more fully hereinafter, the power source 16 is preferably a constant-current power source such as an SCR-type power supply. Such a power supply is available from the Hewlett-Packard Company, 1501 Page Mill Road, Palo Alto, Calif, in their SCR-lO series of DC regulated power supplies which are characterized as providing constant voltage/ constant current operation with automatic crossover. A switch 17 is provided to selectively permit the discharge of electrical energy across the ga 15. The showing of the electrodes 13, 14 by broken lines in FIG. 1 is for the purpose of indicating that the electrodes are movable from positions where they may be very close, or even in contact, to other positions where they are widely spaced. As is taught in the aforementioned Beasley et al. patent, if an arc is struck between the two electrodes 13, 14 while they are very close, and a suitable atmosphere is maintained -between the electrodes at a proper temperature, the gas between the electrodes will become ionized and will act like commonly known conductors of electricity. When this condition is reached, the electrodes 13, 14 may be gradually separated until essentially all of the gas in the furnace is integrated into the heating circuit as an ionized, gaseous resistance element or plasma.

While operation of the Beasley furnace in accordance with the Beasley et al. patent is restricted to operation with a noble gas (such as argon), or at least an atmosphere that is enriched with a noble gas, no such restriction is necessary with the present invention. That is, it has been found that when certain materials are present in the furnace (even in extremely small amounts), that a stable plasma condition can be established in a variety of gases that heretofore had been .avoided for one reason or another. Accordingly, a gas source 18 which is initiallyif not continuouslyin communication with the cavity 12 through valve 19 is provided so as to establish a desired atmosphere in the cavity. Gases that are capable of forming a gaseous plasma in the present invention include argon, helium, carbon monoxide, nitrogen, hydrogen, air, neon, and mixtures of all these gases.

Of all of the structure thus far described, there is relatively little that is novel. What is novel about the furnace and its manner of operation, however, cannot be readily seen in a drawing. Thus, the essence of the invention consists of the combination of the furnace described thus far (or with any other furnace) with a means for promoting the establishment of a gaseous plasma and for stabilizing or sustaining such a plasma once it is formed. Broadly speaking, this means consists of a quantity of low work function material which is placed within the cavity where, quite naturally, it is heated when the furnace is heated. Such material can be in a particulate form, and can be placed in the bottom of the furnace; or, it can be applied to the wall of the cavity 12 as a coating. Furthermore, it is not necessary that the low work function material initially cover all of the wall or walls of the cavity 12, although this may be the eventual result if some of the material vaporizes when it is heated and then condenses on what may be relatively cool walls of the cavity. Thus, it has been determined through experimentation that a coverage of the cavity wall of as little as 10% of the surface area of the wall is sufficient to promote stabilization of'the plasma, with the coated spots being substantially uniformly distributed throughout the cavity. The distribution contemplated is somewhat like the manner in which squares of one color are distributed on a checkerboard.

Examples of what are referred to herein as low work function materials are the oxides of barium, calcium, strontium, radium, uranium, thorium, and materials therrnally decomposable to form these oxides. However, any material which has a work function of 3 or less is believed to be serviceable and should be understood to fall within the scope of the appended claims. By work function, it is meant the voltage that is required to remove an electron from the surface of a material. The respective work functions of the preferred materials are as follows:

Material: I Work function Barium oxide 1.5-2.0

Calcium oxide 2.2 Strontium oxide 2.0 Uranium oxide a 3.0 Thorium oxide 2.9

Although a value for radium oxide has apparently not been reported, it is predicted that it will be found to have a work function between 1 and 2, and should perform nicely in this invention, if certain radiation characteristics can be tolerated.

To explain why the preferred materials are properly denominated as low work function materials (with low being defined as no greater than 3), it should be recalled that, almost invariably, the materials from which furnace housings have been fabricated in the past have been magnesium oxide, silicon oxide, zirconium oxide and aluminum oxide, and mixtures thereof. These materials have work functions of 3.2 for magnesium oxide, 4.6 for silicon oxide, 3.4 for zirconium oxide and 3.8 for aluminum oxide. Hence, all plasma furnaces heretofore have had furnace walls made of material whose work functions are considerably above 3.

If the low work function material is to be applied as a coating to the walls of the cavity 12, it might seem that the thickness of the coating would be important to a person planning to practice the invention. It should be reassuring, then, to know that the thickness of the coating does not appear to be very critical, because re-starts have been obtained as readily at the beginning of a test run (when the coating is fresh and relatively thick) as they have the conclusion of a test run (when a great deal of the coating has perhaps vaporized and leaked out of the furnace or has been absorbed by the porous walls of the housing). Even the initial thickness of the coating is not so critical as to be much of a limiting factor, because it is believed to be practically impossible to apply the low work function material as a coating which is too thin to promote stability of the gaseous plasma. For example, a coating which is so thin as to be nearly transparent has proven satisfactory, with such a coating being on the order of about 1 milligram per square centimeter of wall area. For those who find it difficult to visualize exactly how thin I this is, such a thickness is roughly equivalent to the thickness of that small amount of chalk which invariably re- "mains on a conventional blackboard after the board has been thoroughly cleaned with a dry eraser.

As suggested earlier, the low work function material need not necessarily be afiixed to the cavity wall; it can also be loosely placed in the bottom of the cavity, as long as the quantity of loose material is approximately equal to that quantity which would suitably cover the walls if placed thereon. Since placing the plasma-stabilizing material in the very bottom of the furnace cavity 12 could cause it to be at least temporarily shielded by the material to be melted as heating begins, it is sometimes advantageous to provide a trap or horizontal surface within the cavity and above the level of the materials to be heated. The plasma-stabilizing material is then placed on this horizontal surface instead of on the bottom of the cavity. Such a furnace configurationis shown in FIG. 3.

It may sometimes be desirable not only to separate the cavities 22, 23 therein is shown in FIG. 4. The first and outer cavity 22 is adapted to contain the gaseous plasma,

while the surrounded, inner cavity 23 is adapted tocontain the material to be heated. The heat which is generated in cavity 22 when electrical energy is discharged across the space separating the two electrodes is transferred through the common wall 24 between the cavities 22, 23 so as to raise the temperature within the cavity 23 to a wanted level. If desired, the walls of the housing 21 can be made of the low work function material, so that any labor which might be associated with the preparation of desired coating can be avoided. I

Operation of an apparatus in accordance with the invention begins with the step of providing a suitable gaseous atmosphere in the cavity which is to be heated. For example, if nitriding is to be accomplished with the furnace, the atmosphere should be rich in nitrogen (e.g.,

ammonia). Carburizing can be accomplished by exposing metal parts in the furnace to a carbon-containing gaseous plasma (as, for example, methane, acetylene, etc.). Sintering refractory oxides is conveniently accomplished by using an atmosphere of air or oxygen. If the furnace is to be used as a reactor for chemical synthesis, the atmosphere is naturally selected so as to be compatible with the synthesis to be accomplished. In this respect, the instant invention is markedly different from the Beasley furnace, since a wide variety of gases are usable, and not just the noble gases recited by Beasley et al.

When the desired atmosphere within the cavity 12 has been achieved, it is heated to a temperature at which a plasma condition can exist. Assuming that the electrodes 13, 14 are to be used to strike an are which will initially heat the gas in the cavity, the electrodes are brought sulficiently close to each other to permit an arc to span any gap that may exist between them. Once the arc is struck, the electrodes are then separated so that more and more of the gas between them is brought to very high temperatures. Eventually, the gas between the electrodes 13, 14 becomes ionized and a plasma condition exists which can encompass all of the gas in the cavity, not merely that gas which lies directly between the electrodes. The ionized gas acts like a heating element, as is described in the article entitled, Plasma Resistance Furnace, in the July 1967, issue of Research/ Development.

The temperature at which the plasma will become stabilized is primarily a function of the current input, and an equilibrium temperature is most conveniently established by controlling the current rather than some other parameter. For a given gas, the stabilization temperature is also affected by any low work function material which is present, and may also be affected by the particular furnace construction, i.e., the furnace size, shape, and the insulation (which will affect the temperature of the cavity wall). If the walls of the cavity "are cool, the furnace will not usually operate as efficiently as it will when the walls are hot. An explanation of the effect of cold walls in a plasma furnace is that they induce thermalization and de-ionization of the plasma, which naturally affects plasma stabilization adversely. Also, the plasma-stabilizing activity of the low work function material has been found to increase with increasing temperatures.

It has also been found that the plasma resistance is dependent on the plasma temperature in such a way that the furnace can be essentially self-regulating if the source of electrical power is constant-current power source. To illustrate this, let it be assumed that a given temperature of, say, 3,000 F. has been selected as the desired temperature at which the furnace is to be stabilized. A power supply is then selected based on empirical data or general experience Wtih similar furnaces to achieve this temperature. For example, a person skilled in the art will recognize that a power supply having a rating of about, .kw. would likely be adequate for a small furnace with a spherical cavity having a' diameter of, say, 4 inches. Typical operating values of current and voltage for such a power supply would likely be 450 amps and volts. If the furnace is larger than 4 inches in diameter, all the other items that affect the furnace would be scaled up roughly in direct proportion to the furnace size. It is reassuring that electrode spacing has not been found to be a limiting factor, and it is only required that the power supply be rated at about volts per foot of maximum electrode gap in order to accommodate any of the gases described herein. If the gas to be used and the maximum electrode gap are known, the voltage requirement can be ascertained with greater certainty by reference to a curve like that shown in FIG. 5, which is a plot of operating voltages versus electrode gap for several gases. The slope of these curves indicates that very large gaps can be sustained with practical operating voltages. For example, with argon, a twelve-foot gap would require only about 440 volts.

Once the needed current level has ben identified and the power supply has been appropriately set, the plasma, once established, will tend to sustain itself without a change in current level. Thus, as shown in FIG. 4, the resistance of the plasma (helium in this case) increases with decreasing temperature. Assuming that the furnace is at a temperature represented by point A on the curve, and assuming the furnace cools off somewhat, the plasma resistance would go up, and the product of PR (which is the power consumed by the furnace) would go up, too. This increase in power consumption quite naturally will tend to make the temperature within the furnace move back towards point A. If the plasma temperature moves above the point A, the plasma resistance decreases, and the power consumption (1 R) goes down, such that the furnace will usually tend to cool off (because of inherent heat losses). The overall result is to stabilize the plasma at a temperature which is directly dependent on the current level set by the constant current power supply.

The curve drawn in FIG. 2 is the curve for helium in a furnace of the Beasley type, and without the benefit of any low work function material being present to artificially increase the supply of electrons throughout the cavity. By merely adding a small quantity of low work function material (such as barium oxide) to the cavity and allowing it to become heated, the entire curve is capable of being transposed downward by 200500 F. and frequently by as much as 1000 F. correspondingly dramatic reductions in critical temperatures have been observed with other gases, both pure and mixed. Some of the other gases may not be as favorable for routine operation of the furnace, however, because of their relatively high cost, toxicity, etc. The quantity of low work function material referred to above is that quantity which would be required to cover the walls of the cavity with about 1 milligram of BaO per square centimeter of cavity wall. This quantity is so small that it has been truly difficult to measure it directly. The required quantity has been calculated after experimental results have shown that a quantity of about one gram of BaO placed in a cavity having nearly 700 square centimeters of surface area is sufficient to affect the plasma as described herein. In one experiment, the barium oxide in particulate form was initially placed in the bottom of the cavity; but at the conclusion of the test, most of the barium oxide was found to have vaporized and subsequently to have been deposited on the walls of the cavity. Because of the difficulty of observing the inte- 'rior of a furnace operating at these high temperatures, exactly when the barium oxide is vaporized and subsequently is captured at the cavity walls has not been determined. Thus, whether the barium oxide performs its highly beneficial function before it is vaporized, after some of it has vaporized but is still suspended with the plasma, or after it has vaporized and has subsequently condensed on relatively cool cavity walls, has not been determined. But, in spite of the fact that when it has not been conclusively identified, the fact is known that the material does perform a valuable function.

It is worthy of note, however, that the phenomenon reported herein cannnot be summarily dismissed as obvious in view of some theory related to electron tube operation. That is, the specific low work function materials referred to herein admittedly have been used in electron tubes for years, for the purpose of promoting electron emission. Thus, it is known that filaments coated with alkaline earth oxides are copious sources of electrons.

But, a knowledge of filament coverage requirements and the deactivation of coatings when they react with impurities in an electron tube, would likely lead one to expect just the opposite of the phenomenon that has been observed in the furnaces as will now be explained.

Oxide-coated electron emitters are usually prepared by coating the surface of a filament (normally nickel or tungsten) with a few mils of an alkaline earth hydroxide, nitrate or carbonate. Next, this coating is activated by a process which converts the metallic salt to an oxide which contains some free metal atoms. For example, if the electrode is nickel and barium carbonate is the initial coating, the following reactions occur:

1650 F. BaCO 130.0 CO2 2325 F. BaO Ni MD Ba If the electrode is tungsten and barium carbonate is the initial coating, the following reactions occur:

1650 F. B2100 BaO 002 2370 F. 6139.0 W Ba WOa 3133 Since it is the free barium atoms that produce the improved electron emissions from these filaments, the presence of anything that will restore Ba to BaO will deactivate the electron emitter. Thus, filaments that are activated by alkaline earth oxides are not stable in the presence of oxygen. A pressure of oxygen as low as mm. of mercury is reportedly enough to decrease electron emission by a factor of 100 as compared with an oxygenfree environment. A pressure of 10- mm. mercury of oxygen is reportedly sufficient to poison the coating completely. In the plasma furnace, however, with a basically argon atmosphere, the low work function material performed With no apparent loss in effectiveness even though the oxygen impurity was on the order of 0.1 mm. Hg. It will be recognized that oxygen at pressures of 0.1 mm. Hg is a hundred times greater than that reported to cause de-activation of oxide-coated electron emitters. Accordingly, it is believed that a person skilled in the art of electron tube emission sources would not have predicted what has been observed in the plasma furnace. In fact, such a person would probably have predicted the opposite of what has been observed.

In addition to deactivation of oxide coating on filaments due to the presence of oxygen, prolonged exposure to CO has been reported to completely de-activate such electron emitters. Furthermore, besides chemical deactivation, deterioriation of the coating through evaporation or structural failure (with subsequent separation of the coating from the cathode) due to mechanical shock, etc., will cause the eificiency of the electron emission to fall off rapidly. One example reported at p. 838 of the Journal of Applied Physics, in an article entitled, The Properties of Oxide-Coated Cathodes, by John P. Blewett, volume 10, December 1939, reveals that removal of the activated coating from 50% of the metal surface can result in a thousand-fold decrease in electron emis sion. In contrast to electron tube experience, the plasma furnace has performed satisfactorily when as little as 10% of the cavity walls are coated with the low work function material, and it is possible to even sustain a plasma with gas that is (at least initially) pure C0 Since some low work function materials do vaporize, and in their vapor state they can leak out of the furnace, it is possible for an initially adequate quantity of material to eventually become less than adequate to support stability of the plasma, etc. This deficiency can be readily corrected, however, by the simple expedient of opening a port or the like in a cavity wall and dropping in some replacement material. The entire operation can be about as casual as that of a cook dropping a pinch of salt into a pot of bland stew. Alternatively, the low work function material may be prepared in the shape of a solid rod and utilized like a probe which extends into the furnace cavity, and which is periodically pushed a little farther into the cavity at a rate which is commensurate with the rate at which that end of the probe which is exposed to the plasma is gradually consumed through vaporization. Such a configuration is illustrated in FIG. 6.

While the low Work function material is beneficial in that it lowers the temperature at which a plasma can be stabilized, it is also particularly beneficial in lowering the minimum re-start temperature, i.e., the temperature at which a plasma can be re-established in a hot furnace by merely applying voltage between two widely separated electrodes, and without the requirement of bringing the electrodes together and creating a conventional arc therebetween, etc. For example, as stated in the Beasley et al. patent, re-starts could not be accomplished below about 2800 F. However, with the same furnace and gas, but with the low work function material in the cavity, restarts have been obtained as low as approximately 1650 F. This temperature is about 350 F. above the minimum temperature (roughly 1300 F.) at which a stable plasma can be maintained indefinitely after the furnace temperature has been lowered from an operating temperature of at least 1650 F. This dilferential of about 350 F. with argon is understandable when one considers that re-establishing a plasma is more difiicult than merely sustaining a plasma, just as it takes more energy to initiate an arc discharge across an air gap than it does to sustain the discharge after it is established. Correspondingly dramatic reductions in critical temperatures have been observed in other gaseous plasmas, both in pure and in mixed gases.

As suggested in an earlier paragraph, the marked decrease in the minimum temperatures that are required for establishing a plasma, permitting re-starts, and sustaining a plasma after it has once been established, are perhaps best explained on some theory of artificially increasing the supply of free electrons throughout the cavity, i.e., providing electrons from a source other than the cavity wall, the electrodes, and the gaseous atmosphere in the cavity. Accordingly, this theory may be said to be the theory which is adopted to explain the surprising results reported herein; that is, this is the theory which is adopted insofar as there is a compelling necessity to adopt some theory for the purpose of promoting and understanding of the invention, etc. If it develops, however, that another theory is later found to be more plausible as to the manner of operation of the low work function material, it will be understood that it is not intended to limit the appended claims by reference to a theory of operation that is later found to be misleading to any extent.

What is claimed is:

1. Apparatus, comprising:

a housing having a'cavity therein which is filled with a gas capable of forming a gaseous plasma;

at least two electrodes, each having an end extending into the cavity such that the ends are separated by a p; a source of electrical power connected to the electrodes for discharging electrical energy across the gap; and means for promoting the establishment and sustenance of a gaseous plasma in the cavity when electrical energy is being discharged across the gap, said means consisting of a quantity of low work function material placed in the cavity, said material having a work consisting of the oxides of uranium, and thorium and materials thermally decomposable to form these oxides.

4. The apparatus claimed in claim 1 wherein the gas is argon, helium, or a mixture thereof.

5. The apparatus claimed in claim 1 wherein the gas is carbon monoxide, nitrogen, hydrogen, air, neon, or a mixture of any of these gases.

6. The apparatus claimed in claim 1 wherein the low work function material is uniformly applied to the Wall of the cavity as a coating.

7. The apparatus claimed in claim 1 wherein the low work function material is initially applied to the wall of the cavity as a coating which covers at least 10% of the surface area of the cavity wall and is substantially uniformly distributed around the wall.

8. The apparatus claimed in claim 1 wherein the low work function material is in a particulate form and is loosely placed within the cavity on any substantially horizontal surface.

9. The apparatus claimed in claim 1 wherein the source of electrical power is a constant current power source, whereby the apparatus is characterized as selfregulating in maintaining a desired temperature.

10. The apparatus claimed in claim 1 wherein the housing is made of a material having a work function no greater than 3.

11. The apparatus as claimed in claim 1 wherein the low work function material is prepared in the shape of a solid rod, and wherein the housing has an opening through which the solid rod is extendable into the cavity.

12. The apparatus as claimed in claim 1 wherein at least one wall of the housing constitutes a wall in common with a container for matter to be heated.

References Cited UNITED STATES PATENTS 3,106,594 10/1963 Beasley et al. 131 3,300,561 1/1967 Foex.

H. B. GILSON, Primary Examiner US. Cl. X.R. 139

Images