US2752503A

US2752503A - Isotope separators

Info

Publication number: US2752503A
Application number: US469784A
Authority: US
Inventors: Slepian Joseph
Original assignee: Individual
Current assignee: Individual
Priority date: 1954-11-18
Filing date: 1954-11-18
Publication date: 1956-06-26
Anticipated expiration: 1973-06-26
Also published as: CH341145A; FR1139710A; DE1070850B; GB797558A; NL113583C; NL201949A; BE542860A

Description

2 Sheets-Sheet 1 Filed Nov. 18, 1954 INVENTOR Joseph Slepion ATTORNEY June 26, 1956 J. SLEPIAN ISOTOPE SEPARATORS 2 Sheets-Sheet 2 Filed Nov. 18, 1954 Circumferential Electric Field Direction of Magnetic Force on Curved Current HeV I Fig.5.

United States Patent ISOTOPE SEPARATORS' Joseph. Slepian, Pittsburgh, Pa.

Application November 18, 1954, Serial No. 469,784

21 Claims. (Cl. 250-413) My invention relates to isotope separators, and has particular relation to electromagnetic isotope separators, by operation of which substantial quantities of material may be obtained.

The basic problem which has arisen in the development of the so-called atomic or nuclear industry has concerned itself with the separation of isotopes, and particularly the isotopes of uranium and lithium. To be useful, the separation must make available large quantities of the material consisting purely of a certain isotope. To an extent, this problem has been solved by the provision of separators in which the separation is effected by gas difiusion. Gas diffusion separators, however, demand enormous quantities of power.

It is, accordingly, a general object of my invention to provide apparatus for separating isotopes, the power requirements of which shall be relatively moderate as compared with gas difiusion separators.

My invention in its broader aspects is based on the realization that the electromagnetic mass separator, acting on the isotopes in ionized form, offers means for obtaining substantial quantities of separate isotopes while consuming substantially less power than for gas difiFusion separators. Heretofore, the process of electromagnetic mass separation has been encountered in mass spectrographs. But, in mass spectrographs, the quantity of material transmitted per unit time by the current which fiows is very small, there being no space-charge neutralizing electrons, and, while such spectrographs are suitable for measurement purposes, they are not useful for the purpose of obtaining substantial quantities of separated isotopes, because with space-charge neutralizing electrons absent, they are unable to yield appreciable quantitles of material in a reasonable time interval.

It is, accordingly, another general object of my invention to provide an electromagnetic mass separator, acting on the ions of the isotopes, with space-charge neutralizing densities of electrons, for obtaining substantial quantities of separate isotopes of an element in a relatively short time interval.

An electromagnetic mass separator with Whichthis object may be accomplished must meet two basic conditions. First, a su-ificient number of ions must be deposited per unit time on collectors from which the collected mass may be readily removed to provide a substantial mass or quantity of the material when the apparatus operates over a. shorttime interval. Second, the material must be distributed over the collector, so that there is a substantial enrichment during each operation of the apparatus, or the isotope to be separated in a predetermined region of the collector.

It is, accordingly, a specific object of my invention to provide an electromagnetic mass separator including a source of ions capable of delivering a substantial quantity of material per unit time in which a large proportion of the ions and of electrons shall flow to a collector from which the material may be readily removed.

2,752,503 Patented June 26, 195i;

It is an ancillary object of my invention to provide an electric discharge device including a source of ions and electrons in which the flow of the ions shall be so controlled that they are deposited in predetermined regions of the device.

Another specific object of my invention is to provide an electromagnetic mass separator in which the ions and electrons shall be so distributed over the collector that during each operation of the apparatus there shall be a substantial enrichment of the isotope to be separated in predetermined regions of the collector.

An incidental object of my invention is to provide a novel arc discharge device particularly for producing a discharge rich in ions.

The electromagnetic separator to which my invention relates includes a generally cylindrical. vessel having conductive bases and circumferential walls, the bases capable of being insulated from the circumferential walls. This vessel is maintained in a very high vacuum, less than micron, and at its center an arc is produced which carries substantial current, amperes or more, made up principally of electrons from the cathode, and to a smaller extent, of ions of the material from the anode and a space-charge neutralizing density of electrons also from the anode.

The bases of the vessel may be made up of a number of insulated sections which may be at two or more different. potentials. Among these sections are rings of small diameter coaxial with the are. A high magnetic field is impressed longitudinally on the cylindrical vessel and electric fields are impressed within the vessel between electrodes within the vessel. Usually, also, a negative potential is impressed on the rings.

My invention arises from the discoveries first, that with a field of relatively low intensity impressed in the region of, the are, for example by impressing a negative potential on the rings, the ions of the isotopes to be separated are projected with space-charge neutralizing electrons in substantial quantities into the evacuated space surrounding the arc, and, second, that effective separation of the isotopes may be produced by impressing in this space. circumferential fields of constant polarity; that is, fields produced by direct current potentials. The projection of substantial quantities of ions in the evacuated space occurs because under the action of the relatively low electric field and the magnetic field impressed, the discharge from the arc into the space includes'not only ions, but electrons which are charged oppositely, and supply a charge per unit volume equal and-opposite to that of the ions. Thus, the space charge produced by the ions of the isotopes within the space is effectively neutralized, by that of the electrons, and the ions and the equally and oppositely space-charged electrons flow into the space in substantial quantities.

This motion of the electrons accompanying the ions outwardly is in apparent contradiction to the direction ofi the radial electric field, which is set so that it urges the ions outwardly and the electrons inwardly. Nevertheless, the electrons do move against the opposing radial electric field, the energy for this motion being obtained from interactions with the positive ions. These interactions are not completely describable as collisions, but have properties similar to collisions, in that the energy given to the electrons is randomly directed.

The electrons are readily displaced in the direction of the magnetic field, and the random collisionswith the ions produce rapid reversals in the up and down motion (that is, the motion axially to the cylinder) of the electrons. The electrons cannot escape because of thepotential which is applied to the ring-shaped upper and lower electrodes bounding the magnetic field, and they acquire a random energy equal to a large part of the total energy of the positive ions.

Because of the smallness of mass of the electron, the electron will have a random velocity in the direction of the magnetic field several hundred times the velocity of the ion.

The electron velocity in the directions perpendicular to the magnetic field is also large, since these velocities are random to, coming from the same collisions with positive ions. But, in these directions, because of the low mass of the electron, the path of the electron is bent into small nearly closed curves by the action of the magnetic field.

While the electrons move rapidly up and down but do not escape because of the negative potential 'on the ring electrodes, a small number of positive ions escape at the ring electrodes, but their mass is 'so great, they form a space-charge there so that a sheath forms there which separates the negative potential on the ring electrodes, from the lower (less negative) potential of the space adjacent. In this lower (less negative) potential space, the electrons circulate up and down (axially) with high velocities past the ions, and in the directions perpendicular to the magnetic field, they circulate with similar high velocities but in tight smal spirals, and these spirals are open just enough so that the electrons accompany the ions in their motions.

The ions themselves acquire a random energy through this interchange with the electrons. Their random energy is equal to that of the electrons and has a large part of their mean energy. Hence, if the electrical field is largely radial, only a modest enrichment of parts of the deposit is obtainable because of the large proportion which the random motion of the ions is to the mean motion. My observations during more than ten years of the motion of ions and electrons in magnetic fields completely confirm this.

In the practice of my invention, the circumferential electric fields are set up between two or more radial sets of electrodes or slats between which D. C. potentials are impressed. These radial electrodes or slats are set with one end as near to the are as is practicable, and with the other end at the extreme outer end of the vacuum vessel. The initial radial electric field extends from the are out to the inner edge of the radial electrodes, and is applied by impressing the negative potential to the inner rings of the upper and lower plates, the positive end being connected to the anode of the arc. inner edges of the radial slats.

A direct potential is applied between the alternate radial electrodes. This potential is in accordance with the preferred practice of my invention, independent of The rings extend to the positive ions appears just at the surface of the negative electrode, and a space charge of negative electrons appears at the positive electrode, and these thin space charges limit the electric field in the space between the electrodes to a little more than that corresponding to Hv, where v is the mean radial velocity of the ions at any point, and H is the magnitude of the magnetic field.

I shall say more about this circumferential field later. At this stage, it is enough to say that the circumferential electric field acts differentially upon the ions, and that the lighter ions (with electrons) are moved toward the negative plate, and the heavier ions (with electrons) are moved toward the positive plate. They precipitate upon the cylindrical boundary and the upper and lower plate boundary, and on the radial plates. From these various portions of plates, the material may be removed after all the deposition is done. The parts next to the positive plates are richer in the heavier ion, and the parts next to the negative plates are richer in the lighter ion.

The novel features that I consider characteristic of my invention are set forth generally above. The invention itself, both as to its organization and its method of operation, together with additional objects and advantages thereof, will be understood from the following description of specific embodiments when read in connection with the accompanying drawings, in which:

Figure 1 is a view partly in side elevation and partly in section showing an embodiment of my invention;

Fig. 2 is a view in section taken along lines IIII of Fig. 1;

Fig. 3 is a fragmental view in section of the electrodes used in the practice of my invention;

Fig. 4 is a diagrammatic view showing the trajectories of the ions in the practice of my invention;

Fig. 5 is a diagrammatic view showing the relationship between the forces on the ions in the practice of my invention;

Fig. 6 is a diagrammatic view illustrating a modification of my invention;

Fig. 7 is a diagrammatic view showing another modification of my invention;

Fig. 8 is a diagrammatic view showing still another modification of my invention; and

Fig. 9 is a diagrammatic view of a further modification of my invention.

The apparatus shown in Figs. 1 and 2 includes a vacuum chamber 11 which is connected to a system of vacuum pumps (not shown) through an exhaust tube 13 and the potential between the rings and the arc electrode;

that is, the radial electrodes or slats float electrically relative to the arc electrodes. This potential on the radial slats sets up two oppositely directed circumferential fields, which would draw currents through the mass of ions and electrons brought up to the radial electrodes by the initial radial electric field. These currents are of such a sign that they cause the ions and electrons to be accelerated by the uniform magnetic field outwardly on one side of the negative plates, and inwardly on the other side of the negative plates. The ions and electrons then proceed outwardly in the alternate spaces between the plate electrodes. In the other alternate spaces between the plates, the ions and electrons are not projected out: wards and these spaces may, in accordance with the preferred practice of my invention, be reduced to nearly zero thickness, although, in accordance with the broader aspects of my invention, they may be given any finite thickness.

The ions enter the circumferential fields with a mean velocity (which is nearly radial) and a random velocity, which nearly equals the radial mean velocity.- The currents which are drawn to the plates by the D. C. potential between them is limited, since a space charge of is capable of being evacuated so that the pressure within this chamber is less than 1 micron. The chamber 11 is of generally circularly cylindrical form having conducting bases 15 and 17 and a wall 19, the bases being provided with centrally disposed insulator inserts 21 and 23, and the wall 19 being provided with insulator inserts 25 and 27 and with an opening 29 into which the exhaust tube 13 is sealed. The vacuum chamber is preferably grounded.

Through the centrally disposed insulator inserts 21 and 23, electrode holders 31 and 33 are passed. The lower holder 31 is dimensioned to receive the positive electrode 35 which emits the ions, and the space-charge-neutralizing electrons, and the upper holder 33 is dimensioned to receive the negative electrode 37. The holders are provided with suitable facilities (not shown in detail) for connection to a power supply. This supply must be adequate to produce an are between the

electrodes

35 and 37 and must be suitably regulated to assure that the arc is properly maintained.

The negative electrode 37 may be a cored carbon made by manufacturers of such electrodes in the United States, such as National Carbon Company of Cleveland, Ohio. It is a carbon rod 41 (Fig. 3), with its center drilled out, and having within it a core of oxides of the rare metal earths 173. Other electrodes may be used, however. All that is required is that it serve as a source of electrons; that is, as a cathode of an arc supplying large currents of the order of amperes of electrons.

From each of the inserts 21 and 23 in the bases, a conducting ring 45 and 47 is suspended. A conductor 49 and 51 is sealed through each insert and includes facilities (not shown in detail) for impressing a potential between the conducting rings and the electrode holders. From each of the bases 15 and 17, a plurality of

conductors

61 and 63 and 61 and 63, in the form of segments of a circular arc, are suspended by studs 65. The conducting

segments

61 and 63 extending from each half of each base overlap. The studs 65 by which the segments are suspended are of insulating material, so that the segments may be insulated from the bases.

From the circumferential walls, a plurality of overlapping circularly

cylindrical sections

67 and 69 are suspended. These

sections

67 and 69 are also mounted from insulating studs (not shown) but are themselves conducting. The

sections

67 and 69 may float electrically, in which event each

respective plate

67 or 69 nearest the radial plates 71 through 77 takes on a potential near the potential of the nearest radial plate and the other circumferential plates take on gradually chang ing intermediate potentials, or the circumferential plates nearest the radial plate may each be conductively connected to the nearest radial plate.

A pair of conducting

vertical slats

71 and 73, and 75 and 77 extend radially on both sides of the

arc electrodes

35 and 37. These slats are suspended from insulating

brackets

79 and 81 which are suspended from the bases 15 and 17 and are secured to the slats by insulating bolts 83. Conductors 91 and 97 are connected to

slats

71 and 77, and to a common lead through the insert 25, and

conductors

93 and 95 are connected to

slats

73 and 75 and to a common lead through insert 27, these inserts being in the circumferential wall 19 and serve for impressing a potential between the slats. Each. of the slats extends inwardly to a point very near to the

arc electrodes

35 and 37, the edges of the slats adjacent the rings 45 and 47 extending between the rings. Outwardly, the

slats

71, 73, 75, and 77 extend to a region near the

cylindrical sections

67 and 69. The slats are mounted just between the two

segments

61 and 63, and 61' and 63 suspended from each of the bases 15 and 17, respectively.

The chamber 11 is mounted within an electromagnet 101. The north and south poles 103 and 105, respectively, extend over the bases 15 and 17, respectively, and are spaced a short distance from the bases.

Actual apparatus with which my invention may be practiced includes a vacuum chamber 11 having an ex ternal diameter of about 48 inches and somewhat less than 3 feet high. This chamber has a volume of about 50 cubic feet and is evacuated by a system including a 20-inch oil difiusion pump, an 8-inch oil duffusion booster pump, and a 105 cubic foot per minute Kinney mechanical pump. The slats have aheight somewhat less than 3 feet and a width somewhat less than 48 inches. The spacing between the

slats

71, 73, 75, and 77 is about A; inch, and the insulating

brackets

79 and 81 between the slats are of relatively small dimensions. The magnet has pole faces (103 and 105) which are about, 3 feet in diameter and are spaced approximately 3 feet apart. The magnet is 12 /2 feet high, 13 feet long, and 4" feet wide. The iron of the magnet weighs 90 tons, and the magnet is excited by a winding 107 having copper coils weighing 11 tons. The magnet is adapted to be excited by 2,000 amperes, and when so excited, its flux in the center of the gap between the pole faces is about 10,000 gauss.

The energy for the arc is derived from a direct current supply having a voltage of 250 volts, and the current supplied through the arc is usually of the order of 10 amperes throughout most of my experiments, although this current has been as high as amperes, and may be more. The arc voltage fluctuates between 20 volts and 100 volts.

The arc is fired in the usual manner by bringing the

arc electrodes

35 and 37 into contact and separating them. For this purpose, a motor (not shown), properly geared to one of the electrodes, usually the anode, is provided. The movable electrode holder is sealed through a so-called Wilson seal, so that it may be moved backward and forward, and to a small extent sideways in all directions. The motor is controlled from a suitable thyraton circuit to maintain the arc.

The chamber 11 is grounded. A direct current potential of about 30 volts is impressed between the rings 45 and 47 and the anode electrode 35. A direct current potential of about 200 or 300 volts is impressed between each pair of

coextensive plates

71 and 75, and 73 and 77, one

plate

73 and 75 on each side of the arc electrodes being electrically negative, and the

other plate

71 and 77 being electrically positive. The supply of the potentials between the slats 71 through 77 is insulated from the are potential supply and the supply of the potential between the positive electrode 35 and the rings 45 and 47. Circumferential electric fields are, thus, produced between the

slats

71 and 75, and 73 and 77.

In operation, the various potentials are impressed be tween the

arc electrodes

35 and 37, the rings 45 and 47 and the positive electrode 35 and the slats 71 through 77, and an arc is fired between the arc electrodes. This are operates between the cathode electrode 37 and the eroded surface 107 (Fig. 3) of the anode electrode 35, producing ions of the isotopes to be separated and electrons. Under the action of the electric field produced between the rings 45 and 47 and the anode arc electrode 35, and the magnetic field, the ions andv the electrons are projected into the space in which the circumferential fields are produced by the slats 71 through 77.

I have found that because of the interaction between the ions and the electrons, the ions and electrons are projected into the circumferential field space in substantially equal numbers per unit volume, and the spacecharge effect of the electrons and the ions is almost completely neutralized, an extremely small residue giving: the small electric field which is perpendicular to the magnetic field.

Let us assume that the light ions of mass mi are projected into the circumferential field space with. a random velocity v1, and the heavy ions of mass mz with a random Velocity 1 2, and let v0 be the common mean radial velocity of these ions, where my may vary slightly from one point to another. v0 is somewhat larger than 1 1 or Va. Let V be the potential impressed between the slats which results in the circumferential fields. Then there is a radial distance In from the center of the are for which V=rrr0voH, where H is the magnetic field, the 'lrvo be ing included because there are two pairs of electrodes. The circumferential electric field corresponding to the voltage V is given by Eo=voH, plus a small electric field due to the random velocities v1 and 1 2. Now, if v0 is given, the magnitude which it would have as it is projected into the space and H a magnitude of the order of 10,000 gauss, the equation V=-zrrovoH may be solved for re assuming V=200 or 300 volts, and under the circumstances, it appears that re is equal to the distance from the center of the arc to a point near the circumferential remote end of the slat for a vessel of the dimensions described above.

The situation which exists under the above assumptions in the circumferential field space between

slats

73 and 77, where the radial distance from the arc is less than ro, is illustrated in Fig. 5. Considering for the moment an ion of mass m1 or m2, it is seen that in this region the force on the ion, by reason of the electric field E, is eE, which would be larger than the force produced by the magnetic field Hvo, because r is smaller than r0. There would be then a large drift of the ions of masses m1 and m2 to the negative slat 73. At the same time, there is a drift of electrons towards the positive slat 77. There is then the formation of a thin layer of positive space charge upon the negative slat 73, and the formation of a thin layer of negative space charge upon the positive slat 77. The space charges grow in thickness until the electric field is reduced in the vacuum space to a little more than Eo=Hvo. Thus, in the whole vacuum space, a circumferential electric field nearly equal to Eo=Hvo exists. Thin space charges at the electrodes make up the difference between the applied potential, V=7 rHvu and the value of E0=Hv0.

The total current among the ions is the ion current flowing to the slat 73, and the electron current flowing to 77. There is a force moving this current outwards, due to the action upon this current of the magnetic field H. This force then is positive for r less than 1'0, and becomes zero at r=ro. For 1' greater than re, the force of this current reverses so that many ions will not get to places where r is greater than re.

The ions are projected into a space in which there is an electric field acting circumferentially of intensity approximately E0=Hvo, and if m is the outer bound of this field, it is subject there to a potential V: roHvo. Now, each cubic centimeter of this space acts difierentially upon the ions. The lighter ions, as we shall see, are pushed in the direction of the negative electrode, and the heavier ions are pushed in the direction of the positive electrodes. When they reach the surrounding cylindrical electrode, they deposit out, with the deposit at the positive end, the richer in heavy ions, and with the deposit at the negative end the richer in lighter ions.

To understand this, consider the two ions at a given point of the space, with the masses m1 and m2, with the mean velocity of radial motion v0, and the random velocity v1 and v2 in all directions, with m1(vo +v1 =m2(vo +v2 and v1 and vz less than vo.

v'ei dt where V'o is the circumferential velocity acquired by the ion in the time t. Cancelling Hevo against eEo, and integrating Now letting x be a small radial distance travelled by the ion in time t,

For the gain in the circumferential velocity of the lighter ion on the distance x,

M 1 1 He 0 w o'l' i 0-1 1 1 O 1 Similarly, for a number of heavy ions with 1 2 first adding to v0, and then for an equal number with v2 opposing w,

and,

be the general value for the field. We get then the equation x He[ 0 1 m flo -U2 Now suppose E adjusts itself so that via, is equal to 0,

as it does very nearly so near the negative electrode.

mm m

Then

He v 0 Her 0 He v -v "W m2 "2 o 2 2 L n 2 2 e 2 On the other hand, suppose E adjusts itself so that vB,=0, as it must approximately do so near the positive electrode.

1 2 2 2 Then 1) Hev 0 H el v 1 H rog-112 1 1 1 L 0 1 1 L 0 1 1 L 0 1 Thus, near the forward edge of the negative plate 73, the field takes such a value that m1 takes a nearly zero value of total circumferential velocity. Then the circumferential velocity of particle m2 will be positive and equal to 2 o z x On the other hand, near the rear edge of the positive plate, the electric field will be of such value that the particle m2 will have its total circumferential velocity zero. Then the circumferential velocity of particle m1 will be negative and equal to He U -v At inbetween points, the motion will be inbetween, the difference in circumferential velocity of the two ions being between H e[v v m -0 So far, we have shown that there. is a difference between the twoions, the lighter ions moving toward the negative plate, and the heavier toward the positive plate. However, we have not given calculable formulae since the x, and the number of ions considered was not known. We will give an experimental figure for uranium ions. For a magnetic field of 6000 gauss, where the radial plates started an inch from the arc, the enrichment of the light isotope found deposited on the first inch of the plates, with a voltage difference of 7 volts is found to be one-half of one per cent. That is,

and

h n 2m and the interposition of a radial electric field by Larmors theorem. Of course, for the other ion and the electrons H will not be eliminated. But we see that for such a rotation,

h n 2m and the elimination of the magnetic field, the centrifugal force is and so we see that it varies as the radius, and, therefore, the motion of the ions will vary as the radius in the original motion.

We have then at a distance r inches from the center,

The V will be 7 r volts. For 20 inches, we will have It is not surprising that the enrichment increases so rapidly as a function of r. V=1rroHV0 increases proportional to r. But then also the ratio of v1 or vs to V0 is inversely as r, so that we thus get the square of r coming into the exponential formula in our equations.

The modification of my invention shown in Fig. 7 is similar to the preferred embodiment of my invention, except that each of the slats 71 through 77 is subdivided into a plurality of insulated sections 71a through 71d through 77a through 77d on which progressively higher potentials are impressed. In each case, the potentials are such that the field E is sufiicient to satisfy the requirement that eE is greater than Hevo.

In the modification shown in Fig. 6, the electric cirvc,urrrf ererrtial field is produced by separate

radial slats

201 and 203 at an angle rather than by pairs or sets of slats.

In the, modification shown in Fig. 8, the circumferential electric field is produced by impressing potentials of opposite polarity on

adjacent sector plates

61 and 63, and 61 and 63. This modification does not include radial plates such as 71, 73, 75, 77, or 71a through 77d. This modification does include the

circumferential plates

67, 69, but these are omitted in Fig. 8 for clarity.

In the modification shown in Fig. 9, the circumferential plates 67,v 69 are omitted and instead there are radial plates 301. These plates 301 may be suspended from the

end plates

61, 63, 61, 63', the suspensions. being either insulated or conducting. The plates 301 permit the apparatus to be evacuated more readily than

plates

67, 69. In addition, the plates 301 facilitate the separation, since uncharged material will pass between them and only charged material will be collected by them.

While I have shown and described certain specific embodiments of my invention, I am fully aware that many modifications thereof are possible. My invention, therefore, is not to be restricted except insofar as is necessitated by the spirit of the prior art.

I claim asmy invention:

1. In combination, an arc discharge device including an evacuated chamber having therein a first electrode and a second electrode, means for impressing a potential between said electrodes adequate to produce an are therebetween such that said first electrode is electrically negative relative to said second electrode, and means for impressing a magnetic field coaxial with said electrodes, the said combination being characterized by a first electrode consisting of a conducting material surrounding a core of electron emissive material.

2. The combination according to claim 1, wherein the first electrode is composed of cored carbon.

3. The combination according to claim 2, wherein the second electrode is composed of a uranium containing material.

4. The combination according to claim 1, including means for producing circumferential fields about the are between the electrodes.

5. The combination according to claim 4, including additional means for producing an electric field which in the region immediately adjacent the arc is substantially radial with respect to the arc.

6. Apparatus for separating isotopes of a material including means for producing an are including ions of said isotopes and means for impressing a magnetic field along said arc, the said apparatus being characterized by means for impressing circumferential electric fields about said arc, each said field being maintained at the same polarity.

7. Apparatus according to claim 6 characterized by the fact that the circumferential fields are produced by conducting plates extending radially from the are between which direct current potentials are impressed.

8. Apparatus according to claim 7 characterized by the fact that the radial dimension of each plate is large compared to the radial distance of the edge of said plate from the arc.

9. Apparatus according to claim 6 characterized by the fact that the circumferential fields are produced by a plurality of sets of plates each set consisting of a plurality of plates extending edge to edge radially from the arc, said plates being insulated from each other and corresponding plates of adjacent sets having potentials impressed therebetween which increase progressively with the radial distance from the arc.

10. Apparatus according to claim 9 characterized by the fact that the adjacent edges of successive plates of each set overlap.

11. Apparatus according to claim 9 characterized by the fact that the aggregate radial length of the plates of each set is large compared to the distance from the are of the edge of the plate adjacent the arc. 1

12. The combination according to claim 1 character ized by the fact that the second electrode is composed of a material having a plurality of isotopes.

for impressing a potential between a pair of adjacent plates on each said side.

14. Apparatus according to claim 13 characterized by the fact that the segmental plates overlap at their edges.

15. Apparatus according to claim 6 characterized by the fact that space of the arc and the circumferential fields is bounded by a plurality of end-to-end circumferential conducting plates extending in the direction of the magnetic field, said plates being insulated from each other.

16. Apparatus according to claim 6 characterized by the fact that space of the arc and the circumferential fields is bounded by a plurality of radial conducting plates, said plates being insulated from each other.

17. Apparatus according to claim 7 characterized by the fact that the space of the arc and the radial plates is bounded by a plurality of end-to-end circumferential conducting plates extending in the direction of the magnetic field, between the outside edges of the radial plates, said plates being insulated from each other.

18. Apparatus according to claim 17 characterized by means connecting the respective circumferential plates nearest the radial plates to the radial plates to which they are nearest. 7 i

19. Apparatus according to claim 13 characterized by the fact that the space of the arc and the segmental plates is bounded by a plurality of end-to-end circumferential conducting plates extending in the direction of the magnetic field, between the outside edges of the segmental plates, said plates being insulated from each other.

20. Apparatus according to claim 19 characterized by means connecting the respective circumferential plates :nearest the segmental plates to the segmental plates to which they are nearest.

21. Apparatus according to claim 6 characterized by the fact that the space potential with reference to the arc of each point Within the space of the circumferential field is substantially proportional to the radial distance of said point from the arc.

No references cited.

Claims

1. IN COMBINATION, AN ARC DISCHARGE DEVICE INCLUDING AN EVACUATED CHAMBER HAVING THEREIN A FIRST ELECTRODE AND A SECOND ELECTRODE, MEANS FOR IMPRESSING A POTENTIAL BETWEEN SAID ELECTRODES ADEQUATE TO PRODUCE AN ARC THEREBETWEEN SUCH THAT SAID FIRST ELECTRODE IS ELECTRICALLY NEGATIVE RELATIVE TO SAID SECOND ELECTRODE, AND MEANS FOR IMPRESSING A MAGNETIC FIELD COAXIAL WITH SAID ELECTRODES, THE SAID COMBINATION BEING CHARACTERIZED BY A FIRST ELECTRODE CONSISTING OF A CONDUCTING MATERIAL SURROUNDING A CORE OF ELECTRON EMISSIVE MATERIAL.