United States Patent 1 1 3,571,642
[72] Inventor Carl H. Westcott 2,770,755 11/1956 Good 315/542 Deep River, Ontario, Canada 2,813,996 1 1/1957 Chodorow 315/542 [21] App], No. 698,644 2,904,720 9/1959 Bell 315/5.42UX [22] Filed Jan. 17, 1968 2,925,522 2/1960 Kelliher 315/542 [45] Patented Mar. 23, 1971 2,979,635 4/ 1961 Burleigh..... 315/5.42X [73] Assignee Atomic Energy of Canada Limited 3,319,109 5/1967 Haimson 313/63X Ottawa, Ontario, Canada 3,331,961 7/1967 Leboutet et a1 313/63X 3,403,346 9/1968 Giordano 3 1 3/ 63X I 4 METHOD AND APPARATUS FOR INTERLEAVED Primary Examiner-James W. Lawrence CHARGED PARTICLE ACCELERATION Assistant Examiner-Palmer C. Demeo 9 (Jaims, 3 Drawing Figs Attorney-Graham and Baker [52] U.S. Cl 313/63,
250/419, 315/542 328/233 ABSTRACT: A method and apparatus for accelerating [5 1 Int. harged anides is described in which charged particles of [50] Field of Search 313/63' 0 pposite sign but of similar charge to mass ratio are ac- 328/233, 256; 315/541 (11191-11199); celerated through the same structure by means of an altemat- 250/41-9 q ing electric field. The two sets of particles can be accommodated in the same apparatus by grouping particles of the [56] References Cited same charge in bunches which are spaced by a phase dif UNITED STATES PATENTS ference of approximately 11' radians with respect to the ac- 2,545,595 3/ 1951 Alvarez 3l5/5.42X celerating field frequency from the bunches of the opposite 2,633,539 3/1953 Altar 315/5.41ux chargesign.
PATENTED HAR23 I97! 3,571, 642
sum 1 0F 2I 750 kw Mlil'lllilltlllll Ahll) AlPllAlilA'llllS Milli llN'lEhLEAl/ED QilAHZiGEll) lAll't'llltCLlE MICELERATIIGN This invention relates to a method and apparatus for the acceleration of charged particles in radio frequency wave structures. More particularly it has reference to an apparatus and method for increasing the output of a linear accelerator structure designed for charged particles of a given sign by using charged particles of similar charge to mass ratio but of opposite sign.
For the making of an intense neutron generator structure Atomic Energy of Canada Limited has suggested an accelerator system which would provide thermal neutron fluxes of the order of neutrons cm. sec. which it is proposed to achieve by the spallation reaction produced by protons of energy of the order of 1 Gev. in a thick heavy element target such as lead bismuth eutectic. Such an accelerator is however useful, not only for the production of neutrons, but for producing mesons and other high energy particles and also for other experiments and tests in connection with the developrnent of nuclear research, which require high energy particles for their production in sufficient quantity and at sufficient energy. This whole situation has been discussed in Atomic Energy of Canada Limited Publication AECL 2600 dated Jul. 1966 and the research and technical applications of a high energy of this type are discussed in Chapter V of this publication.
Of particular significance is that the protons produced from a proton source are accelerated by first bunching them into groups which can then be further accelerated by a linear accelerator. Details of the linear accelerator types of structures are given in Chapter XIVB of AECL publication 2600.
in general these accelerating structures also include magnetic focusing elements which are preferably quadrupole mag nets, but may be solenoids, grids or foils.
A quadrupole magnet system can only focus charged particles in one plane so that two or more are necessary for focusing in both of two planes at right angles. Therefore quadrupoles may be present in singlets (i.e., spaced pairs) or grouped in pairs or triplets for example. These systems would be equally suitable for focusing particles of the same charge to mass ratio but of opposite charge to those for which the system is designed even though the planes of focusing would be reversed for the oppositely charged particles for a given member of a system.
i have found that if bunches of oppositely charged particles of approximately the same charge to mass ratio are introduced substantially l80 out of phase with the bunches of charged particles for which the linear accelerator is designed, those oppositely charged particles will also be accelerated. The bunches of particles can be separated from one another into separate beams when this is desired, by suitable steering magnets or electrostatic fields.
The present invention although exemplified by a proton accelerator is not limited to it, but more particularly provides in the method of accelerating charged particles of one charge sign which are grouped into first bunches and then accelerated by an alternating electric field, the improvement which comprises, bunching charged particles of opposite sign and of substantially the same charge to mass ratio as said particles of said one sign into second bunches spacing the second bunches from the first bunches by a phase difference of approximately rrradians with respect to the accelerating field frequency, an accelerating the second bunches in said field.
The present invention also provides apparatus for accelerating charged particles of opposite charge sign but of substantially the same charge to mass ratio which comprises, a pair of respective particle sources, an alternating electric field particle accelerator, means for directing particles from said sources into said accelerator and including means for bunching said particles from said sources into discrete bunches of particles of one charge sign alternately spaced between discrete bunches of particles of opposite charge sign before introduction to said accelerator, said bunching means spacing said bunches for encountering the alternating electric field of said accelerator for acceleration in the same respective relative phase for each bunch of the one sign and in the same respective relative phase for each bunch of the said opposite sign, said relative phases being substantially 'n'radians difi'erent from one another with respect to the alternating electric field.
in the description which follows reference will be made to the accompanying drawings in which:
FIG. 1 shows one scheme for introducing positively and negatively charged particles into a single accelerating structure;
FlG. 2 shows an alternative method should there be excessive interaction between the two types of particles before acceleration and;
FIG. 3 shows in more detail the arrangement of FIG. 1.
The first embodiment of the invention is exemplified in FIG. 1. In this a source l of positive ions (suitably generating the ions from an ionized plasma from which protons are stripped) delivers the ions into astatic preaccelerator 2 which they leave with an energy of some 750 Kev. They are then deflected by amagiet 3, passing for a second deflection through amagnet 4 into abuncher 5 and from which they are delivered to a wave type accelerator such as a linac at 6. Although no details of the buncher are given here it can for instance be of the travelling wave type using either a simple gap or cavity. The bunched protons are acceptable to the accelerator and the ion current into the accelerator can be regulated by variation of the amplitude of the accelerating alternating signal applied to the buncher.
it will be appreciated that if negatively charged ions are now injected into the buncher there will be a tendency for them to be collected in bunches which are displaced or 1r radians in phase with respect to the buncher signal frequency from the bunches of protons from source 1. if therefore a source of negatively charged protons, or H ions (that is hydrogen atoms with one excess electron) is placed at 10 the ill ions whose mass differs by less than 0.2 percent from that of the protons, can be allowed to pass through the poles of magnet lli. which is unenergized, and will be bent by the magnet 4i in the opposite direction to the protons from source l and will enter thebuncher 5. To allow for overhaul of the sources which operate at fairly high currents and are subject to erosion, alternativenegative ion source 12 and proton source l3 are provided. When thesource 13 is in operation magnet ii is energized, and as this will be in conjunction with thesource i2 magnet 3 is deenergized at this time.
As a second alternative FlG. 2 shows a system which may be employed if it is found that proper bunching cannot be obtained by using asingle unit 5 for both positive and negative ions. Twoproton sources 20 and 21 and two negatively chargedhydrogen ion sources 22 and 23 are free to deliver their ions torespective bunchers 24 and 25. The bunched ions from 24 and 25 are deflected by amagnet 26 into theaccelerator 27 as before. it will be clear that whensource 20 is in operation the magnet 2b is energized and whensource 23 is usedmagnet 29 is energized. FlG. 2 shows amagnetic splitter 40 for separating the bunches of opposite charge after acceleration.
For the sake of greater clarity FIG. 3 shows details of the system which would be needed for a layout such as that of FIG. 1. Thegenerator 30 provides a high voltage of 750 ltV. for which operate in general at much higher currents have their own independenthigh voltage units 31 and 32. Magnetic focusing units 33 (which may be quadrupoles) are provided at the output of each of the preaccelerating columns for each source to reduce beam blowup and general dispersion of the ion stream, After passing throughmagnet 35 further magnetic focusing is provided at 34 and minor adjustments in beam direction are achieved by thesteering magnet 35. The ions then meet the switching magnet d which will deflect the negatively and positively charged ion beams in opposite senses. and form a composite beam, following which further focusing is carried out bymagnetic system 35 belfore the beam of mixed IOI1S 1S delivered to thebuncher gap 5, After bunching the ions are focused by themagnetic system 37 and are then captured by the first section of the linear accelerator at 6.
Negative ion sources for producing H ions are available commercially and may suitably be a Duo-Plasmatron, or type which is formed as a tube. at one end of which is a source of electrons such as a heated filament, and at the other end a pierced anode. When an axial magnetic field is applied to such a structure and hydrogen gas is led into the space between cathode and anode, electrons are accelerated along the axis towards the anode and strike the hydrogen molecules to produce the reaction It is found that there is a tendency for greater quantities of the H ions to be produced in certain regions of the plasma than in others and if the hole in the anode is suitably placed some of the H- ions will drift through the hole from which they may be picked up by a second anode.
One point to bear in mind in the system of interleaving the negative and positive ions is that, in a linear accelerator which consists of two sections operating at different frequencies, it is necessary that the second section be driven at a radio frequency which is an odd multiple of the frequency of the first section. This is necessary to maintain 1r radians phase difference between the bunches of positive and negative ions.
The particular linear accelerator chosen by Atomic E Energy of Canada Limited, consists of two sections. The first part is an Alvarez section in which a standing wave pattern is developed in what is essentially a long cylindrical cavity. The cavity is excited in the mode of oscillation which is designated as TM and which provides an alternating axial electric field which has its maximum value on the axis of the cylinder and zero at the walls, with circumferential magnetic field lines of zero intensity at the axis and a maximum near the walls. Ions are accelerated by being injected along the axis at a time when the electric field is in the direction of motion. Since an ion cannot traverse the entire length of the cavity during one-half of an RF cycle it is necessary to put a series of electrostatic screens (in practice copper tubes) along the axis. The spacings and lengths of the tubes are chosen so that the ions will be drifting through a tube screened from the electric field when the field is in the reverse direction for that charge of particle in that position. The length of each succeeding drift tube is increased so that a particular ion (the synchronous particle) will arrive at each successive gap in the same relative phase. Focusing quadrupole magnets are provided in the drift tubes to preserve the shape of the bunch and keep the particles on the axis.
In principle, acceleration to any given energy is possible by making the cylindrical cavity long enough or increasing the electric field strength. In practice, however, certain limitations are imposed and it is found that a cavity length should not exceed 20A (where )t is the wave length corresponding to the resonant frequency of the cavity), in order to limit excitation of harmonics which would lead to a nonuniform field on the axis.
The second part of the linear accelerator is made as a wave guide structure which allows the charged particles to pass through a series of excited cavities which are contiguous with one another. The particles are now at an energy at which a fixed design velocity can be used for each tank M, and problems are eased in that the particles can be considered to travel at a constant speed through each tank M. A discussion of the types of wave guide structures to be used appears in AECL 2600, Chapter XIVpages 4 to 1 1.
Let us now examine the theoretical and practical considerations of this scheme with particular reference to the apparatus of AECL 2600. In many instances the most convenient plan for experiments which it is wished to perform involves the use of several high energy beams which may be produced in an accelerator and later separated spatially by suitable means. In particular in the Atomic Energy of Canada Limited scheme described in AECL 2600 there is a need for certain meson experiments in which a beam of high energy protons having an intensity of approximately km 1 percent of the main beam is called for. Additionally the particles used for these meson experiments should preferably have a continuous characteristic except for microstructure, (that is, structure within times of the order of 1 1.) so as to minimize counting loses in the experimental arrangements likely to be used.
The interleaving of the negatively charged ions described in this application is particularly suitable for providing the beam requirements of this intensity although with better negative ion sources there is no doubt that higher intensities could be produced. The splitting of the H- ions at the output of the linear accelerator by means such assplitter 40 is a simple operation involving only a bending magnet of suitable design. Care must of course be taken not to exceed a critical field strength above which electron stripping may occur.
It is important that the linear accelerator be straight because although the axial focusing systems will work for both the I-I and theH- ions, if any change in direction of the beam is required, the magnetic field for achieving this will act in opposite directions on the two sets of particles.
The I-I ion sources for the AECL intense neutron generator give about 250 ma. of which some 65 ma. are accelerated to l Gev. If we have a 12 ma. source of r]- ions, then in conditions similar to those for the H ions %/zma. will be accelerated. These are injected, suitably bunched, by the sine wave buncher into the linac at 11' radian phase difference from the H" ions.
A requirement for this method is that the ratio of the radio frequency of the later sections of the linac to that of the Alvarez section be an odd integer, (in particular suitably 268.3 MHz. and 805 MHz.) so that a 11' phase difference between H and H- ions in the Alvarez portion leads to a 1r phase separation in the final section.
The phase stability factors are preserved for the H- ions because the mass difference of -0. 109 percent from the I-I ions is trivial. The action of quadrupole magnets will be inverted, but since a quadrupole converging in one plane is divergent in a perpendicular plane, this is equally satisfactory, at least for a cylindrically symmetrical system, although care may be needed in the injection region which may depart from symmetry. From the point where the H- and I-I beams are merged, until the first bending magnet after acceleration, there must, however, be (i) no DC (E or B) correcting fields, i.e., the linac must be straight, (ii) no windows or foils in the beam, which would strip the H- ions, and (iii) a sufiicient vacuum. After the H" beam is split from the I-I beam by the magnet at the end of the linac, foils may be used in either beam.
Two important conditions are involved in the feasibility of usingI-I-beams, those of electric dissociation and gas-collision dissociation.
Richardson (N.I.M.24, p. 493, Nov. 1963) quotes sources from which a l millisec life for an I I ionis assuredjatleast within a factor i3) if the fields do not exceed 1.9 Mv.7cm. or el te .f 9.efo Z-fLMyl.L Ihe8t field in the quadrupoles at 0.5 cm. from the axis for a 6,000 fiiQsiHF gradient is safe, giving for [3 0875 a field of only 1.48 Mv./cm. (and this for only a small fraction of the transit-time), while the main linac accelerating field is very much smaller, averaging only 11 kv./cm.
The second difficulty, gas collision dissociation, is avoided if an adequate vacuum can be maintained. For calculation purposes we assume 1 X 10- Torr of air to remain in the path, ex-
The H- beam can be (magnetically) separated and after only 18 m., stripped by a thin foil into 995 percent protons,
and 0.5 percent neutral H atoms. In the 154 m. long Alvarez section the accelerating gradient is less nearly uniform, and the calculated loss of I-I- ions by gas dissociation at [0" Torr air pressure is 0.13 percent.
The problems of beam spill are. for a given fraction spilled, less serious at injection than later. since the particles then have a low energy, but the cross sections are high and the question of loss of ions by collision must be carefully examined. Not only can gas dissociation occur (and the gas pressure near the ion sources is relatively high) but a loss of H ions by the reaction Hd'HifiZH" must be considered. After bunching, the H and H- ions are in separate bunches, but since the buncher is less than 100 percent efficient, I-I ions may still be in excess in the region of the H- bunch. Since there are relatively few H- ions, loss of the I-l beam intensity due to this reaction can never by appreciable.
Assuming for the moment that this loss by charge exchange can be tolerated, we may merge the beams before bunching in the simple arrangement indicated in FIG. 1.
On the other hand, if tests show that too many H" ions are neutralized with this arrangement, the alternative arrangement of FIG. 2, which uses two bunchers and a mixing of the I-l and H" beams after bunching, can be adopted. It is true in either case that l-I ions lost in the buncher will travel a certain distance into the linear accelerator and may therefore neutralizeH ions there, but as the latter become accelerated this loss decreases to a negligible fraction. It may also be noted that, if the scheme of FIG. 1 is adopted, the H- ions after passingmagnet 4 will undergo space-charge forces tending to condense the beam, whereas the H ions in the region of the Hions are subject to space-charge repulsion losses.
In conclusion therefore the production of a separate H- beam is feasible and its intensity is limited by available negative ion sources and by dissociation due to residual gas. Other negative ions can be used but they must have approximately an equal charge to mass ratio to the ions for which the acceleration system is designed. In practice allowable tolerancies in charge to mass ratio will probably be found to be about percent (for example allowing the use of Ne and Ne car rying equal but opposite charges).
Iclaim:
1. In the method of accelerating charged particles of one charge sign which are grouped into first bunches and then accelerated by an alternating electric field, the improvement which comprises, bunching charged particles of opposite sign and of substantially the same charge to mass ratio as said particles of said one sign into second bunches, spacing the second bunches from the first bunches by a phase difference of approximately ar radians with respect to the accelerating field frequency, and accelerating the second bunches in said field.
2. The method as defined in claim 1 including the further step of separating said first and second bunches after acceleration in said field.
3. Apparatus for accelerating charged particles of opposite charge sign but of substantially the same charge to mass ratio which comprises, a pair of respective: particle sources, an alternating electric field particle accelerator, means for directing particles from said sources into said accelerator, and including means for bunching said particles from said sources into discrete bunches of particles of one charge sign alternately spaced between discrete bunches of particles of opposite charge sign before introduction to said accelerator, said bunching means spacing said bunches for encountering the alternating electric field of said accelerator for acceleration in the same respective relative phase for each bunch of the one sign and in the same respective relative phase for each bunch of the said opposite sign, said relative phases being substantially 1r radians different from one another with respect to the alternating electric field.
4. Apparatus as defined inclaim 3 comprising means for separating said bunches of said one charge from those of said other charge leaving said accelerator.
5. Apparatus as defined inclaim 3 said accelerator being constructed for accelerating said particles in a straight line.
. Apparatus for providing a stream of negative y charged ions interleaved with a stream of positively charged ions of substantially the same charge to mass ratio which comprises a positive ion source, a negative ion source, a sine wave buncher and an alternating electric field accelerator, means coupling the buncher to the accelerator, means for directing positively charged ions and negatively charged ions from said respective sources into said buncher, and means for separating bunches of positively and negatively charged ions leaving said accelera- 01.
7. Apparatus as defined inclaim 6 comprising a pair of positive ion sources; a pair of negative ion sources; and means for directing positive ions from the first. of said positive ion sources and negative ions from the second of said negative ion sources into said buncher, and alternatively directing ions from the second of said positive ion sources and ions from the first of said negative ion sources into said buncher.
8. Apparatus as defined in claim 7 said means for directing said ions from said first sources comprising a first deflector magnet, and said means for directing said ions from said second sources comprising a second deflector magnet, said sources being aligned for directing ions into said magnets so that ions are directed into said buncher firstly by energizing one of said magnets and deenergizing the other, and alternatively by energizing said other magnet and deenergizing said one.
9. Apparatus as defined in claim 7 said alternating field accelerator comprising an Alvarez linear accelerator.