ACCELERATION OF CHARGED PARTICLES BY INTENSE ...

© Gordon and Breach, Science Publishers Ltd.

Particle Accelerators 1973, Vol. 5, pp. 81-91

Printed in Glasgow, Scotland

ACCELERATION OF CHARGED PARTICLES BY INTENSE ELECTRON BEAMStt Sand~a

GEROLD YONAS Laboratories, Albuquerque, New Mexico 87115, USA

Several groups ha~e observed that. energetic ions are accelerated in the direction of the electron flow in vacuum ~~d pl~malfilled dIodes, as well as In ele~tron beams propagating in low pressure neutral gases. Ion energies many Imes e e ectro~ energy and acceleratIng fields of 1 MV/cm have been measured. In all of these exeriments

~aCkgr~und gas IOns a~e accelerated ~t the sa~e time that the electron beam self pinches.

Several

m~e1s hav~

een ~ vanced to e~plaln t.he acceleratIon of an Ion bunch in such high intensity electron beams. The results of the expenlme~ts are revle~ed I~ terms of co?erent and collective acceleration models, with particular emphasis placed on co lectIve acceleratIon In a propagatIng potential well.

1. INTRODUCTION The two distinct types of collective ion acceleration being considered at this symposium differ greatly, not only in the details of the mechanism producing the requisite electron concentration, but also in the manner the subject is being pursued. In the case of the electron ring accelerator (ERA), several rather comprehensive analyses were carried out prior to detailed design of the experimental hardware. Only then, were extensi~e experimental programs begun. Even with this extensive analytical preparation, several surprises still have occurred as evidenced by the ring stability problems experienced by the Berkeley group. Ion acceleration using linear electron beams, on the other hand has been characterized from the very beginnin~ by surprises. This is undoubtedly because there has been no comprehensive program,to study the basic principles of this acceleration technique. Virtually all of the results obtained thus far have come as unpredicted observations from experiments oriented toward understanding complex beam-plasma phenomena. This, in itself, represents a turn of events compared to what might have been expected almost 20 years ago when Veksler created much of the original interest in the use of linear beams for coherent and collective acceleration. 1 At that time, he suggested that coherent acceleration of a t Th~s ~ork was supported by the U.S. Atomic Energy CommIssIon. t Invited paper presented at the Symposium on Collective Methods of Acceleration, Dubna, USSR, September 1972.

cluster of ions by inverse Cherenkov drag from a linearly propagating stream of electrons could provide accelerating fields of many megavolts per centimeter. The major problem associated with accomplishing this kind of acceleration was that the required intense electron beams did not exist at that time. Although Veksler's early proposal resulted in an effort which has continued here at Dubna in the form of ~RA, there have apparently been no extensive programs oriented toward applying either Veksler's original coherent acceleration suggestion or any of the other linear collective approaches which were discussed in a review article by Rabinovich. 2 One of the most intriguing concepts discussed in the review paper by Rabinovich was an idea credited to L. V. Kovrizhnyk. He suggested the application of traveling localized constriction in a propagating electron beam to create a potential well. The beam concentration would be caused by an accelerating magnetic mirror field which would be externally driven to achieve a balance between the well acceleration and the Coulomb field between the electron concentration and the ion bunch lagging behind. Although he presented no suggestion as to a method of providing this accelerating magnetic constriction, it was pointed out that fields of 1.0 MV/cm could be achieved with a 50 kA, 1 MeV beam. Rabinovich did not discuss the experimental results of Plyutto and his colleagues who had already accomplished what was probably the first successful demonstration of collective acceleration. 3 ,4 These results as well as

GEROLD YONAS

82

region, providing further ion acceleration. They discovered that ions were accelerated when the diode impedance suddenly increased during a time of a few nanoseconds within a much more slowly rising current pulse. The resulting current decrease and inductive voltage spike apparently occurred only when a critical ratio of the beam electron density to plasma density was reached. (Unfortunately this ratio was not defined.) They also found that the higher the background gas density the weaker the impedance change and voltage increase. At the time of the rapid current decrease, an intense electron beam was accelerated in the direction of the' applied field, but with a peak energy of three times the applied voltage. Simultaneously, an ion pulse, with peak energy as much as 60 times the applied voltage, was extracted from the vacuum region. Although the ion spectrum was nonreproducible from shot to shot, there was a distinct upper energy cutoff in the energy, and the ions from a single pulse tended to be monoenergetic (20 per cent spread). At the time of the rapid decrease in the total current, the current density on axis increased to greater than' 10 kAJcm 2 in.a time of 4 nsec. If this intense beam was allowed to pass through a hole in the accelerating electrode into the vacuum region, then the ions received roughly twice as high an energy as was the case when the electron beam stopped and only ions were extracted. The energetic electron beam extracted from the plasma had a pulse duration of approximately

other examples of collective acceleration, will be reviewed in the next section. As we shall suggest, the acceleration process could have been a result of a traveling constriction similar to that suggested by Kovrizhnyk, except that it was created naturally through an instability in the beam envelope. 2. EXPERIMENTAL OBSERVATIONS Without specifically addressing the problem of collective acceleration, the first reported observation of ions accelerated in an electron beam was made by Plyutto in 1961. 3 The essential feature of his experiments involved the extraction of an electron beam and an ion bunch from a plasma filled diode. In 1967, he reported observing proton energies as high as 4-5 MeV from a 200-300 kV discharge,4 and a more detailed exposition of the earliest work appeared in 1969. 5 One of the recent and most comprehensive studies of ion acceleration in plasma filled diodes was reported by Mkheidze, Plyutto and Korop6,7 (M-P-K). They emphasized the relation between the temporal and spatial behavior of the ion acceleration mechanism and the electrical characteristics of the plasma filled discharge. Their apparatus (Fig. 1) consisted basically of a plasma source, provided by a spark discharge, which injected a plasma into an accelerating gap. They provided an aperture in the accelerating electrode so that the beam could propagate in a field free vacuum

RADIAL ELECTRON CURRENT COLLECTORS

ION COLLECTOR AND ENERGY ANALYZER

LONGITUDINAL ELECTRON CURRENT COLLECTOR

r---I

60VQ@

I I I IL

_

(~

~-~~~~---I DRIFT REGION

PLASMA FILLED DIODE

PLASMA INJECTOR

FIG. 1. SchematIc of plasma diode apparatus (J. P. Mkheidze, A. A. Plyutto, and E.

o. Korop).

83

ION ACCELERATION BY INTENSE ELECTRON BEAMS

50

r--------------------------------~}

Proton Arrival Time n sec.

__ -

25

-__-----

0=100 ~

~O

__ -- --X

__~ ~~15

X=150,u. 6= 200 JL

----

6

T= _1_

/2

I

n

r 20

40 Oistance

60

80

100

em

FIG. 2. Proton time of arrival (S. Graybill and J. Uglum).

10 nsec, and propagated through the vacuum chamber with a beam perturbation velocity which was equal to the ion velocity. This velocity was detected by the radial spray of electrons reaching detectors outside the beam channel. In the plasma filled diode, energies of 0.3-0.6 MeV wefe reached within a 1.0 cm acceleration distance. Without realizing that such work had been under way in the USSR, Graybill and Uglum (G-U) discovered that when an intense pulsed electron beam (1.5 MeV, 40 kA, 25 nsec risetime) was injected into a low pressure gas, ionization and acceleration of the background gas atoms occurred.S Using ion time of flight, detected with current collectors, and total ion yield based on neutron production, G-U concluded that up to 10 13 protons could be accelerated reproducibly in a fairly monoenergetic and spatially compact bunch to energies of 5 MeV. They showed that the process could be extended to deuterium, helium, and nitrogen with ion energies proportional to the accelerated ion charge state; they also found that the ion energy was independent of background gas pressure and proportional to the square of the beam current. The yield, however, was very dependent on pressure; in deuterium, for instance, there was a very sharp maximum in the ion yield at pressures of 200 /l, which then fell off to a negligible yield at 300 /l. They were able to correct the arrival time of the ions at the current collector for the time of flight, and found that the ions were accelerated at the P.A. A3

instant in the beam pulse that sufficient collisional ionization of the background gas had taken place to achieve radial force neutralization (Fig. 2). Since the beam parameters (v(y-1)-1 < l)t indicate that the beam could penetrate its own space charge potential barrier prior to radial force neutralization, it is not surprising that the beam front was able to propagate to the end of the drift chamber long before the ion acceleration process occurred. By varying the length of this chamber, G-U determined that the ions were accelerated over a length of 30 cm giving an average field of 0.2 MeV/ cm. Graybill has most recently reconfirmed much of the original G-U workS with only slight modification. 9 ,10 With a slightly higher impedance diode (1.7 MeV, 30 kA) than that used originally, he found that although helium and nitrogen energies showed no pressure dependence, proton and deuteron energies increased by almost a factor of two in the 100-200/l pressure range. For the lowest pressures at which ion acceleration in each gas was found, ion energy was proportional to charge and independent of mass (Fig. 3), but this behaviour did not extend to the upper pressure cut-off in deuterium and hydrogen. Graybill also reported less well-supported data for argon, giving a charge state of + 12 and energies as high as 14 MeV. The argon currents were only 1-2 A but lasting for 1'01

t v=I/17,OOOP (I in amperes); P=v/c.

84

GEROLD YONAS

24 22

20 18

Ion 16 Energy 14

;100.0

>10 10 1011-1012

2.0 0.4 0.1