US2890337A - Ion beam focusing means for calutron - Google Patents

Description

June 1959 J. G. BACKUS 2,890,337

ION BEAM FOCUSING MEANS CALUTRON Filed June 28, 1946 3 Sheets-Sheet 1 A/PC SUP/ L Y F/ZAME/VT SUP/ L Y HEA 75/? SUP/ Z Y 1 A W A INVEN T01? JOHN 6 546x05 AITORNE Y June 9, 1959 J. G. BACKUS ION BEAM FOCUSING MEANS CALUTRON 3 Sheets-Sheet 2 Filed June 28, 1946 Dam/v05 4 INVENTOI? JOHN G. fi/zc/ws ATTOHNEY J. G. BACKUS 'xou BEAM FOCUSING MEANS CALUTRON June 9, 1959 3 Sheets-Sheet 35 Filed June 28, 1946 INVENTOR JOHN 6 5401 05 ATTORNEY United States atent O ION BEAM FOCUSING MEANS FOR 'CALUTRON John G. Backus, Los Angeles, Calif., assignor to the United States of America as represented by the United States Atomic Energy Commission Application June 28, 1946, Serial No. 679,884

4 Claims. (Cl. 250-413) This invention relates to commercial isotope separators,

and'more particularly to calutrons.

practice in mass spectrography of withdrawing ions from an ion generator through a narrow elongated slit. The result is a tremendously greater ion beam current that has an accuracy of resolution of the same order as a mass spectrograph.

In view of the fact that the mass spectrograph and the calutron employ the same basic scientific principle; namely, the curvature of the paths of charged particles in a magnetic field, the calutron is usually described with reference tothe mass spectrograph. There is little real comparison between the two devices, however, since the mass spectrograph or mass spectrometer is a laboratory instrument and the calutron is a massive production ma- :chine. Thep-resent invention will be described, neverthe less, with reference to the mass spectrograph and the mass spectrometer, hereinafter referred to as mass spectrodevices, as this facilitates the exposition of the problems and principles involved.

Calutrons are described generally in Atomic Energy for Military Purposes by H. D. Smyth, and in great detail in US. Patent No. 2,709,222, issued to E. 0. Lawrence, on May 24, 1955. In general, a polyisotopic beam of ions is projected into a magnetic field wherein the ions travel in curved paths, the paths of the lighter ions being 'of smaller radius than the heavier ones, enabling collectors to be placed across the different paths at a convenient point such as the 180 point of travel along their arcuate paths. Cal'utrons have been chiefly employed, and very successfully employed, in the separation of the isotopes of the metal uranium, the isotope of mass 235 being separated from the isotope of mass 238. Calutrons, however, can be employed for separating commercial quantities of isotopes of practically all of the elements and probably also in the separation of ionized molecules as well as ionized atoms.

As pointed out in the Smyth report, supra, at the outset of the work in separating commercial quantities of U in his patent, noted above.

In spite of the fact that the supposed limiting-current of the ion beams was exceeded and means found for preventing the beam from blowing up, this current was Patented June 9, 1959 still much too small for separation purposes. Even with these currents that were beyond the realm of mass spectrograph it would require many years to accumulate commercial quantities of material with any reasonable number of separating machines. The next problem was therefore the provision of a prodigious source of ions. The solution of this problem was the design and operation of an ion generator that involved an intense arc struck in a vapor or gas containing the desired element, namely uranium.

While the intense arc did give a much larger current, still greater currents were desired, and it was found necessary to expose to a strong accelerating electric field, a considerable area of the surface of the arc. This, however, involved another serious problem of beam resolution at the ion collectors since the practical arc length is limited, necessitating the use of a wide area of are compared to the arc length. The present invention solved this problem of a wide arc width by causing the ions to travel paths such that they seemingly came from an arc of narrow width, whereas, in fact, the arc width was great. The invention therefore results in a much larger ion beam than is otherwise possible, with little or no sacrifice of resolution at the collector.

It is therefore an object of the invention to provide a virtual focus of the ion beam of a calutron near the ion source.

Another object of the invention is the provision of a calutron having an ion source with a large ion exit, but at the same time permitting resolution of the ion beam comparable to that from an ion source with a small ion exit.

Another object of the invention is to provide a calutron having an arc as an ion source in which the surface of the arc is curved to provide avirtual focus of ions emanating from the surface of the arc.

Still another object is to provide a calutron with an ion source that has a magnetically transverse dimension of its ion exit that is substantially as large as the separation distance at the collector of the isotopic ions being separated.

Other objects and advantages of the invention will be apparent in the following description and claims, together with the accompanying drawings in which:

Figure 1 is a schematic diagram of a calutron as viewed in a plane transverse to the magnetic field,

Figure 2 is a view taken along the line 22, of Figure 1,

Figure 3 is a diagram of the paths taken by ions of a different mass that were initially projected into the magnetic field along parallel paths,

Figure 4 is a diagram of the paths taken by ions of different mass that were initially projected into the magnetic field from a point source along divergent paths, and

Figure 5 is a diagram of the ion generator and the accelerating and decelerating electrodes together with the equipotential lines of the electric fields thereabout and with an indication of the paths of the ions from the ion generator through the electric fields.

Referring to Figures 1 and 2, a magnetic field is established between upper and lower magnet pole faces 10 and 11 respectively, the direction of the magnetic field in the air gap being shown by the flux lines 12. A vacuum tank 13 is positioned in the magnetic field and has a removable face plate 14. Suitable pumping means 16 is connected to the tank to maintain the pressure therein at any suitable subatmospheric pressure such as 10* to 10' millimeters of mercury. Secured to the face plate 14 is an insulator 17 that supports an ion generator 18, the ion generator including a heaterchamher 19 in which is placed a heater 20, a reservoir chamber 21, and an arc chamber 22. A plate 23 is placed over the arc chamber in which is formed an arcuate electron stream defining slot 24.

vAn electron emissive filament 26 is spaced from and placed over the defining slot 24, but for purposes of clarity is shown at one side thereof. The forward walls of the ion generator 18 define an ion exit slit or arc slit 27, elongated in the direction of the magnetic field. A pair of accelerating electrodes 28 are positioned ad-- jacent the arc slit 27, the opening formed between the electrodes being opposite the arc slit opening. When an accelerating electric field is maintained between the ion generator 18 and the accelerating electrodes 28, an ion beam 29 is projected in the magnetic field 12, and curves as indicated in Figure 1. The beam separates into two parts 290 and 29b corresponding to the masses 235 +238 respectively of the two isotopes present. Positioned in the paths are two collectors 31 and 32 mounted on a pedestal 33 secured to the removable face plate 14.

In the operation of the cal utron of Figures 1 and 2, a source of accelerating DC. voltage 34 is energized, its negative terminal being grounded and connected to the accelerating electrodes 28, grounding them. The positive terminal is connected through a variable resistor 36 to the insulated ion generator 18. Thus an electric field, which may be of the order of 35 kilovolts, is set up between the accelerating electrodes 28 and the ion generator 18, and since the electrodes are negative relative to the generator, positive ions are withdrawn from the generator 18 and projected into the magnetic field 12 to form the ion beam 29 of the calutron. Inasmuch as thearc slit 27 is elongated in the direction of the magnetic field, the ion beam is in the shape of a ribbon.

Vapor is supplied by a solid charge of a compound that vaporizes at a suitable temperature, and for uranium separation, uranium tetrachloride is commonly used. Heat to vaporize the solid charge is supplied by the heater energized by a heater supply 37 maintained at ion generator potential and varied by a resistor 38. The vapor passes from the reservoir chamber 21 to the arc chamber 22 where it has a suitable pressure such as 10' mm. of mercury, and where it is bombarded by an electron stream from the thermally emissive filament 26, the stream passing through the stream defining slit 24. The electrons are attracted to the ion generator 18 by an electric field established by an arc supply 39 also maintained at generator potential, and the output of which may be varied by a resistor 41. In practice an electric field of several hundred volts between the filament 26 and the generator 18 is used, the electrons passing through the defining slot 24 because of the collimating effect of the magnetic field. As mentioned previously, the filament 26 is placed over the slot 24 so that the same portion of the magnetic field that passes through the slot 24 passes through the filament 26 also. The filament 26 is rendered thermally emissive by resistance heating from a filament supply 42 varied by a resistor 43, the filament commonly being made of tungsten.

In summary therefore, a vapor is generated from a solid salt by the application of heat. The vapor is bombarded by electrons which break up the molecules, and in the case of uranium, generate positive ions, the majority of which are singly ionized. The pressure and electrical conditions are carefully maintained to create an intense arc discharge, the plasma of which provides a copious supply of ions. drawn from the surface of the plasma by the strong negative field of the electrodes 28, projecting the ion beam 29 into the magnetic field.

The ion beam of Figure l is shown as simple traces of the paths of the two uranium isotopes U and U Inthe actual operation of a commercial separator it will The positive ions are withbe readily realized, however, that such a situation could not exist and the showing of Figure 1 is diagrammatic only. Instead, Figures 3 and 4 give a more adequate picture of the possibilities of ion beams from an arc plasma of appreciable width such as width a of Figure 3. The two obvious possibilities of beam paths are, first, initial projection of ions into a magnetic field along parallel paths, and second, projection of ions along divergent paths. Figure 3 shows the first possibility of initially parallel paths. Figure 4 shows the paths of ions that were initially projected along divergent paths. In this latter case there is shown two divergent rays that started from an are surface at approximately the same point and a central ray is also supplied to furnish the median line for the two divergent rays. Obviously, when using a plasma source of width a (Fig. 3) the diagram of Figure 4 would be duplicated many times within the space a, but for the sake of simplicity and clarity, divergent rays from one point only are shown in Figure 4. The parallel rays may also be considered as a special case of divergent rays, namely, when the divergence is zero.

The paths of Figures 3 and 4 are simple geometric constructions of semicircles, using different points for centers, and using two different radii, one corresponding to mass 235 and the other to mass 238. This geometric construction is based on the physical principle that if charged particles are projected or shot into a uniform magnetic field and suffer no collisions or other retarding forces, they will describe perfect circles that return to the point of origin and thereafter continue circling along the initial circle. When charged particles are of difierent mass and projected into the magnetic field with substantially the same energy, the heavier particles will describe circles of larger radius than the lighter particles. When both light and heavy particles are projected into a magnetic field from the same point source, both circular paths will return to the same point of origin. It is obvious in this case, that the paths of two particles of different mass will be farthest away from each other at the point of travel. The mass spectro devices, the mass spectrograph and the mass spectrometer, utilize this phenomenon to get traces or readings of isotopes. This same principle is utilized in the calutron also, resulting in the placement of collectors at the 180 point of travel.

Viewing Figs. 3 and 4 in the light of the foregoing explanation, it will be noted that the radius of the circles for U and U are different, and further that the radius for either isotope is the same in Figures 3 and 4. Therefore, to get a true picture of a beam having random ion paths emanating from a slit width a, and remembering that Fig. 4 should be duplicated many times over a distance a, one drawing should be superimposed on the other. It is plainly evident, that although isotopes might be separated with a beam composed only of initially parallel rays, there could be no separation when the two are combined. An ion beam of random paths is characteristic of mass spectro devices, and in order to get resolution at the 180 point, the source slit distance a is made very small compared to the separation distance at the 180 point. In such a system the initially parallel rays are reduced to such a thin ribbon as to be shown by a single line compared to Figure 3. The divergent rays have then what is essentially a point or line source as represented in Figure 4. Only then is separation possible in mass spectrometers and mass spectrographs.

This inability of mass spectro devices to give separation except when the ion slit is narrow is best demonstrated mathematically. The separation distance at the 180 point for two isotopes each represented by a single ray, is expressed in terms of the mass difference of the two isotopes. This distance is sometimes referred to as slugs, a slug being the distance between two paths and also may be represented by mass units.

.able zone 2 by the slit width a.

of isotopes varying by one mass unit in weight. Thus in separating U from U there is a mass difference of three units resulting in three slug separation at the 180 point. This distance can be measured between similar rays of the two isotopes and is indicated in Figures 3 and 4 as 3s. This distance is expsessed mathematically in terms of the radius of the two diflierent paths,

l2 He wherein c is the velocity of light, M the weight in mass units, v the velocity in the same units as c, H is the magnetic field strength, and e is the charge on the particle. In separation work, H and 'e are the same for both isotopes, and v is slightly different. Since 0 is a constant, these constant items can be replaced with K, the formula reading Therefore, the radius and separation distance are directlly proportional to the mass and directly related to each other.

Referring to Figure 4, it will be noted therein that the divergent rays strike fairly close to the central ray for any selected isotope at the 180 position. This distance z is strikingly small compared to the spread at the 90 point and has thus become a named phenomenon, magnetic focusing, and is well described in the literature. There is no sharp focus as in optics, but rather, a narrow zone of magnetic focus 2. It will be noted that both marginal components intersect each other on the 180 line, so that the spread or zone is the same whether the divergence from a central ray is on both sides or only one side. The dimension of this zone 2 is in accord-- ance with the formula z=2R(l-cos 0) which inits working form is wherein 6 represents the angular divergence in radians from a central ray. It may be calculated from this formula that a uranium beam ray that diverges from a central ray by 6 /2 will create a zone of focus of about 3 slugs wide. units difference, such as U and U 6 /2" is therefore the upper limit of divergence, since otherwise the two zones would overlap.

This calculation is as follows and is based on the linear relationship of radius of curvature to the mass of a particle as described previously. The zone 1 will be set equal to three slugs or in terms of mass, to three mass units. The radius is a linear distance Also the formula for z may have both sides divided by R, the

radius, taking the form,

Replacing z by 3 mass units and taking the R of one of the isotopes, for example 238, we get 0:.124 radians=6.4 or 6 in round figures.

This 6 /2 limit for three slug separation is based upon a point source.

as in mass spectro devices. This then reduces the allow- Thus for any appreciable width of source slit, for example 2 slugs, the divergent angles must be so small that they reduce the ion current greatly. In this connection, it should be Suppose now that the source is .of finite width a and rays diverge from its entire width noted that the wider the divergence permissible in an ion beam, the greater the ion current.

The zone of focus 1 is now reduced by slit width which will be taken as two slugs, for purposes of demonstration. Thus z-=32=1 and using this in the formula just mentioned,

%=0 or 0: .065 radians=3.8

Therefore, with a source slit of 2 slugs in mass spectro devices, the allowable angular divergence for separating U and U has been almost cut in half, reducing the ion current by a corresponding amount. Thus the greater the slit width in mass spectro devices the smaller is the allowable divergence, and instead of an increase in ion beam current, a decrease will be encountered.

This is in contrast to the present invention wherein good separation of the receiver is obtained with slit widths a that are equal to the separation distance at the receiver. In fact, one of the most successful calutrons employs a source slit width a of 3 /2 slugs and give excellent resolution at the collector in separating U from U Unlike the mass spectro devices, every fraction of an inch increase in slit width gives a corresponding increase in ion beam current since it exposes a greater area of are surface. This is accomplished by shaping the surface of miniscus of the arc plasmas so that the ions leaving the plasma are accelerated normal to this surface and tend to converge in an area quite small compared to the arc slit width.

Referring now to Figure 5, there is shown an arc chamber block 50 preferably formed from a single block of carbon and secured by metal brackets 52 disposed on either side." A sheet 53 of carbon or equivalent material is inserted in grooves in the arc block and has a slot 54 formed therein that is elongated in the direction of the magnetic field which is perpendicular to the plane of the drawing. 'In this embodiment, the slit width is about 3 slugs wide. Vapor is supplied to the arc chamber thru an aperture 51 in the back of the carbon arc block. The vapor is ionized by an electron stream 56 as shown, the shape of which is carefully controlled by a defining slot such as the slot 24 of Figure l. Electrical conditions and vapor density is maintained such that an intense arc discharge takes place in the arc chamber 'iilling'the entire arc chamber with a plasma, that is, an atmosphere containingapproximately equal quantities of unlike charges. Thus when U01 is used as a chargematerial, its vapor is dissociated into negative ohlorinc ions and these together with the electrons equal the charges present on the positive uranium ions. There are also present neutral atoms and molecules.

The positive ions are withdrawn from the arc plasmas by a strong electric field of 65 kv. or stronger set up by two pairs of electrodes 57 and 58, and in this respect the showing of Figure 5 differs from that of the schematic showing'of Figure 1 which shows only one pair of electrodes. The purpose of the double pair is" to create a strong electric field to withdraw ions from the arc plasma, and since these ions would have too much energy to describe their arcuate paths within desired limits, the second electrodes 58 are supplied to slow down the ions to a desired level. The result is a greater ion beam for the same radius, as disclosed in US. Patent No. 2,725,478, which issued to Byron T. Wright, on November 29, 1955. The net electric field is the same as in Figure 1, namely 35 kv.

The negative electric field of 65 kv. is set up by maintaining the arc block 50 at a positive potential of 35 kv. and anaintaining the accelerating electrodes 57 at a negative potential of 30 kv. The second electrodes 58 slow down the'ions and are therefore referred to as decelerating electrodes and are maintained at ground potential. The important equipotential lines are drawn in and their 7 voltages indicated. The paths of the ions are designated by five selected lines emanating from the arc plasma surface and passing thru the accelerating and decelerating electrodes 57 and 58 respectively.

Considering now the are plasma in the arc block 50, it will be noted that a curved surface 60 is indicated for the plasma and touches near the two edges of the ion exit or are slit 54. This surface is defined by the predominance of positive ions over negative ions and is therefore subject to electric fields. In this connection it should be noted that an arc plasma is not subject to the influence by electric fields since by definition it consists of equal quantities of unlike charges and if this relation ship is substantially changed it is no longer a plasma. The number of ions present is greatest where the bombardment of the charge vapor is the greatest; namely, in the electron stream. Therefore the electron defining slot 24 (Figure l) is positioned to bring the electron stream 56 as close to the curved arc plasma surface 60 as possible, much closer than indicated in Figure where the separation is great for purposes of clarity. The shape and positioning of the aperture 24 is disclosed in my prior US. Patent No. 2,850,636, which issued on September 2, 1958.

The positive ions are accelerated forward from the plasma surface 60 and travel at right angles to the equipotential lines. The concave curvature of the surface 60 causes the paths to tend to converge but space charge and a straightening of the equipotential lines in the region of the zero voltage line keeps them apart. A small part of the ion paths, however, not only converge but actually cross each other, and are not shown, the path lines of Figure 5 indicating the behavior of the majority of the particles. As the paths approach the region of the accelerating electrode 57, the equipotential lines are bowed in the opposite direction and tend to disperse the beam. The ions at this point have gained so much velocity and energy, however, that they are but slightly affected by these defocusing equipotential lines. As the ions travel along their paths between the electrodes 57 and 58 they experience a repelling or decelerating electric field which not only slows them, but also tends to diverge the paths. This divergence is due to the tendency of a charged particle to assume a path other than normal to the equipotential lines in a decelerating field. Also, when the ions lose their initial high energy in a decelerating field, they come more under the influence of the magnetic field and tend to curve, and this tendency is clearly evident in the decelerating gap of Figure 5. Once the ions emerge from the decelerating electrodes 58, they are completely space charge neutralized and are thereafter free of electric fields. Their paths thereafter are influenced only by the magnetic field and in a uniform magnetic field form a perfect circle or part thereof if intercepted sooner.

The width of the narrowest part of the ion beam between the accelerating electrode 57 and the arc slit 54, is the actual focus of the beam, but this is not the measure of the virtual focus width. Rather, the virtual focus is much smaller. The point of virtual focus is obtained by projecting the curved path that the ions travel beyond the electrode 58, back toward the arc block 50. This is indicated by the broken lines 62, and when thus projected they meet at a common point 64 which is the point of virtual focus. Thus the place from which all ion paths apparently emanate after they leave the electrode structure is a point, and this point is known as the virtual focus of origin of the ion beam. Even the ions whose paths cross have curved paths that have the same virtual focus. While the virtual focus thus is indicated as a point in Figure 5, it will be remembered that the slit 54 is elongated in the direction of the magnetic field and thus the focus is a line.

In production calutrons the magnetic field is not uniform, but is purposely distorted to permit the use of a beam that diverges more than would be permissible with a uniform magnetic field. The principle of location, and the definition of the virtual focus, nevertheless remains the same. Also it should be noted that while the virtual focus is preferably a line or point, in actual practice it does approach a pencil or area, the dimension of which isa small part of the ion beam at its narrowest part, for example, less than one fourth as large. This narrowest part or actual focus is about a third to a fourth of the arc slit width. Thus the actual focus may be as large as a fifteenth or twentieth of the arc slit width.

The formation of a concave plasma surface has been found to depend upon the accelerating voltage relative to plasma ion density. The plasma density is controlled by the amount of vapor introduced into the arc chamber, and the strength of the electron stream 56, and these in turn are controlled by the heater 20 (Fig. l) and the conductive current thru the filament 26. The exact equation for the control of the concavity of the plasma surface 60 has not been determined. However, the space charge equation gives the proper order of magnitude and may be used to control the shape of the surface 60. The equation relates the ion current to the voltage on the accelerating electrode 57 and the distance from the arc slit plate 53 to the electrode 57 where a' is the distance from are slit plate 53 to electrode 57. It has been found that if the arc plasma and the accelerating voltage are so regulated that the accelerating voltage is exceeded by 15% or more than would be expected by the equation for the given distance and the measured ion current, a concave surface on the plasma is produced that gives rise to the virtual focus phenomena. In actual practice, however, the accelerating voltages are fixed, and the electron stream and vapor supply are varied to give maximum separation at the collectors 31 and 32 as determined by current readings made there.

An important result of the establishment of the point of virtual focus is that the collector must be placed at the point or otherwise as determined by the location of the virtual focus, and not related to the arc slit.

While the invention has been described relative to specific structure, it is not limited to the structure shown as other structure could obviously be employed. Nor is the invention limited in any fashion other than by the terms of the following claims.

What is claimed is:

1. A calutron comprising means for establishing a magnetic field, means forming a chamber having an exit opening elongated in the direction of the magnetic field, means for creating an arc discharge in the chamber in a charge gas, means for varying the ion density in the arc, electrodes disposed opposite the exit opening, means for applying a potential to the electrodes with respect to the arc chamber for accelerating ions therefrom, and a collector, the magnetically transverse dimension of the opening being substantially as large as the zone of magnetic focusing of the calutron ion beam at the collector.

2. A calutron for separating isotopes of a selected element comprising means for establishing a magnetic field, means for generating ions of the selected element and having an ion exit opening elongated along the magnetic field, electrodes disposed opposite the opening for projecting the ions into the magnetic field after which they describe arcuate paths, a collector disposed across the paths substantially at the 180 point of travel along the arcuate paths, the opening having a magnetically transverse dimension substantially as large as the separation distance at the collector of two different isotopic ions having initially parallel paths.

3. A calutron for separating isotopes of a selected element comprising means for establishing a magnetic field, means for generating ions of the selected element and having an ion exit opening elongated along the magnetic field, electric field establishing means disposed opposite the opening for Withdrawing ions in a beam from said generating means and contoured for establishing a virtual focus of origin outside said generating means, the ions of said beam thereafter traversing arcuate paths under the influence of said magnetic field, a collector disposed across the paths substantially at the 180 point of travel along the arcuate paths, the opening having a magnetically transverse dimension substantially as large as the separation distance at the collector of two different isotopic ions having initially parallel paths.

4. A calutron for separating isotopes of a selected element comprising means for establishing a magnetic field, means for generating ions of the selected element and having an ion exit opening elongated along the magnetic field, first electric field establishing means disposed opposite the opening for withdrawing a converging beam of ions transversely of said magnetic field, second electric 1G field establishing means disposed opposite said first electric field establishing means for establishing a diverging force upon the ions of said converging beam, the ions of said beam thereafter traversing arcuate paths under the influence of said magnetic field, a collector disposed across the paths substantially at the 180 point of travel along the arcuate paths, the opening having a magnetically transverse dimension substantially as large as the separation distance at the collector of two different isotopic ions having initially parallel paths.

References Cited in the file of this patent The Production and Focusing of Intense Positive Ion Beams, in Physical Review of August 1, 1935, by Tuve et al., pp. 241-256, volume 48.

High Current Ion Sources for Nuclear Investigations, in Physical Review of December 1, 1935, by Lamar et al., pp. 886-892, volume 48.

Images