Validation and comparison of nonlinear negative ion extraction theory for two experimental configurations

HAL Id: jpa-00246130

https://hal.archives-ouvertes.fr/jpa-00246130

Submitted on 1 Jan 1989

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Validation and comparison of nonlinear negative ion extraction theory for two experimental configurations

J.H. Whealton, P.S. Meszaros, R.J. Raridon, K.E. Rothe, M. Bacal, J.

Bruneteau, P. Devynck

To cite this version:

J.H. Whealton, P.S. Meszaros, R.J. Raridon, K.E. Rothe, M. Bacal, et al.. Validation and comparison of nonlinear negative ion extraction theory for two experimental configurations. Re- vue de Physique Appliquée, Société française de physique / EDP, 1989, 24 (9), pp.945-949.

�10.1051/rphysap:01989002409094500�. �jpa-00246130�

Validation and comparison ^{of nonlinear} negative ion extraction theory

for two experimental configurations

J. H. Whealton (1), ^P. S. Meszaros (1), R. J. Raridon (1), ^{K. E. Rothe} (1), ^{M. Bacal} (2), ^{J. Brune-}

teau (2) ^{and P.} Devynck (2)

(1) ^Oak Ridge ^National Laboratory, ^Oak Ridge, ^TN 37830, ^U.S.A.

(2) Laboratoire de Physique des Milieux Ionisés, Laboratoire du C.N.R.S., 91128 Palaiseau Cedex, ^France

(Reçu ^{le 5} janvier 1989, révisé le 29

1989, accepté ^{le 31 mai} 1989)

Résumé.

L’article

pour objet l’examen des propriétés ^{de deux}

^d’ions négatifs

volume, ^l’une

étant caractérisée par

charge d’espace importante ^{au niveau de} l’extraction, l’autre par

importante pénétration ^du champ électrique d’extraction dans la source. Les résultats expérimentaux ^obtenus

^les

deux sources sont comparés à

^modèle théorique, qui suppose la température ionique ^nulle.

Abstract.

This paper analyzes ^the properties ^{of two volume} negative ^ion sources, ^one being characterized by

a

large space charge ^{in the extractor} while the other is dominated by ^field penetration into the ion source. The

experimental results obtained with these two sources

compared ^to

theoretical model which

ion temperature.

Classification

Physics ^Abstracts

32.80F

29.25C - 52.70N

Negative ^ion sources are being developed actively

in many laboratories. Some of the applications require ^{the use of low emittance} negative ^{ion beams} emanating ^{from such sources.} The characteristics of the plasma edge ^at extraction sometime determine the optimal characteristics of the negative ^{ion source} production ^itself. Sources based on volume pro- duction _appear to provide ^low temperature ^H- /D-

ions and are therefore of particular ^interest (for ^a

review see Ref. [1]). The purpose of this paper is ^to examine the properties ^{of two ion} sources, one, which is characterized by significant space charge ⁱⁿ

the accelerator while the other one is characterized

by significant ^field penetration ^{into the ion source.}

The extraction plasma edge ^as computed ^{from the}

model for the negative ^ion plasma [2] ^{will be com-} pared ^with experiments ^{for these two ion sources.}

Negative ^{ion beam} formation from volume sources

is significantly ^{different from} positive ^{ion beam}

formation. For positive ^ions the formation plasma

can be thought ^to consist of positive ^{ions and} relatively ^{fast and confined} electrons. The ions see a

downhill run to the accelerator column (presheath)

and become extracted. The electrons want no part ^of the accelerator and stay away.

For volume-produced negative ions, however, ^the formation plasma consists of three species : positive

ions, ^{electrons and} negative ions. In the case of

negative ^ion extraction, the electrons are no longer repelled by the accelerator fields and, ^{in the absence} of a transverse magnetic ^field, will rocket out of the

source. The confined species ^{in this case,} positive ions, ^{is slow and less able to} respond ^{in the} relatively rapid ^fashion expected ^of the electrons.

Since negative ^ion extraction behaves differently

from positive ^ion extraction, ^we embarked upon ^an

investigation ^{of the} phenomena by studying ^{a model}

of the negative ^ion presheath plasma ⁱⁿ conjunction

with a fully self-consistant nonlinear multi-dimen- sional analysis ^{of the} extraction sheath with all space

charge ^and image ^{forces included} [2, 3] (for ^{a review}

see Ref. [4]). ^The results of modelling ^are compared

to experimental ^results reported by Stevens et al. [5],

as well as with measurements effected in a hybrid multipole ^source [6] ^{at Ecole} Polytechnique.

The hybrid multipole ^{source has been} described in detail elsewhere [6, 7]. ^A cylindrical stainless-steel chamber is surrounded by ^ten samarium-cobalt _mag- nets, with the south and north poles altematively facing ^the plasma (Fig. 1). ^The endplates ^{are not} magnetized. The bottom end is bound by ^the plasma

electrode (PE) ^{of the} extractor, which is 9 cm in diameter, ^and by ^{an annular} grid G, ^{which is} grounded. ^The primary ^{electrons are} produced by

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/rphysap:01989002409094500

946

Fig. 1. - Hybrid magnetic multicusp ^ion

ten thoriated tungsten filaments biased 50 V negative

with respect ^{to the} chamber walls. These filaments

are located in the upper part ^{of the chamber in the} multicusp magnetic field ; ^{each filament is located in}

the radial plane passing through ^{the saddle} point ^of

the multicusp magnetic ^field (in ^{between two wall} magnets). ^{As a} result of this choice of the filament

position, ^the primary ^{electrons are} trapped ^and

confined in the neighbourhood ^{of the} cylindrical

sidewall. Few energetic electrons escape into the

central, magnetic field-free region.

The extraction system ^consists essentially ^{of three}

electrodes as shown in figure ^{2. The first PE in}

contact with the plasma ^{has a circular} aperture 0.8 cm in diameter. The second electrode, ^called

«

separator », ^is located 0.62 cm from the PE and also has an opening ^{0.8 cm} in diameter. A pair ^of

Sm-Co magnets ^are located in the separatpr just

behind the opening ^{and create a} transverse magnetic

field (B > ³⁰⁰ G) strong enough ^to deflect the accelerated electrons ^onto the separator, ^as long ^as

the extraction voltage ^{does not exceed 5 kV. The}

H- ions are barely ^affected by the presence of this

field, ^{which causes} only ^{a small latéral} displacement

of the H- ion flux reaching ^the collector. The

Fig. ^2.

Extraction system. PE, plasma electrode ; SP, separator ; G, grid ; C, collector ; PM, permanent mag- nets, V,,,, extraction voltage ; V,,, ^bias applied ^{to the} plasma electrode. Vspc, ^{bias applied to} the collector with respect ^{to the} separator.

plasma electrode PE bias has an important ^{role in} optimizing ^{the extracted} negative ^{ion current} [1].

The experimental apparatus ^used by ^{Stevens et}

al. [5] (described ^{in Refs.} [5] ^and [8]) ^{is a conven-}

tional magnetically ^filtered multicusp ^{ion source} [1].

The extraction tests employed ^a high perveance accel-decel extraction system. ^{The emission} aperture

was 3.15 mm in diameter. The negative ^{ion beam}

current was measured using ^a magnetically suppres- sed Faraday cup.

The negative ^{ion current measured} by ^the Faraday

cup ^{-in the} experiment ^of Stevens et al. [5] ^was plotted ^{as a} function of extraction voltage ^{for five}

different arc currents (see ^Ref. [8]). ^{Each of the}

curves, for a given ^{value of the arc current, has a}

horizontal plateau ^whose height ^is proportional ^to

the arc current. It is apparent ^{in this} experimental

situation that the plasma density ^{and the} negative

ion density ^is proportional ^{to the arc current. These}

data can be perveance scaled as shown in figure ³ whereupon ^{the whole} family ^{can be} compressed ^into

a single ^{curve as} illustrated. Thus the scaling ^has meaning ; positive ion extraction is also scalable in the same way. A phenomenological presheath plasma model has been considered to explain ^these

results which are observed to be quantitatively

Fig. 3. - Data ^{from the}

^described by ^{Stevens et}

al. [5] (see ^{also Ref.} [8]) showing negative ^{ion current}

function of accelerating voltage for several different arc currents. These families of

scaled

shown :

dividing the ordinate by ^the plasma density ^{and the}

abscissa by the two-thirds power of the plasma density.

The solid line is the theoretical result.

significantly ^{different from} positive ^{ion extraction}

from a plasma [6, 9, 10]. ^{In it there is an excess} space

charge ^imbalance which reflects the situation that the positive ^ions equilibrate ^less rapidly ^and thoroughly ^{in a} negative ^{ion source than} do electrons in a positive ^{ion source.} This relative _space charge

imbalance causes the plasma ^{to be more forcefull in} pushing against ^the applied ^{fields than is the case}

with positive ^{ions, which is in} agreement ^{with the} observed behavior [4, ^9, 10].

Theoretical results using ^this model, ^but assuming

zero ion temperature, ^{are shown in} figure ³ by ^a

solid line. Significant features of agreement ^include

a horizontal plateau, ^a drastic falloff at low extrac- tion potential, ^{and a mild falloff at} high ^values. By inspection of the ion motion it is readily apparent that the falloff at low values of extraction potential ^is

due to a transverse space charge limit, and the mild falloff at high values of extraction potential ^{is a} highly underdense situation with ion crossover focus-

ing resulting ⁱⁿ significant vignetting ^{on the ac-}

celerator electrodes causing ^{loss of} the ion beam to the Faraday cup. A significant point ^of departure ^of

the theoretical results in figure ^{3 with the} experimen-

tal results is at the transition region (denoted ^{B in} Fig. 3). ^{At this value of the} extraction potential, negative ^{ions start} intercepting the accelerator elec- trode and fail to contribute to the current exiting ^the

accelerator. Since the ions in the computation ^have

zero transverse « temperature ^{», this onset} is very sudden. In the actual experiment, however, ^{due to} finite temperature, ^{there is a} precurser ^at lower than the critical extraction potential whereupon ^{some of}

the ions intercept. ^The transition is more gradual

and is a measure of nonlaminar ion transverse

velocity distribution. Since aberrational fields are not contributing ^{to this nonlinear behaviour} (or ^else they ^would have been observed theoretically), ^the

ion temperature ^{is the} expected ^cause. Therefore,

the curvature of ion current at the critical extraction

potential could allow a measure of ion temperature.

This deduction of the ion temperature would be free of errors introduced by long transport ^{channels and} the pittfalls ^{of emittance scanner} interpretations. ^It

can complement ^{these direct} experimental ^measure-

ments.

A minor disagreement between the theory ^{and the} , experiment of Stevens et al. [5] ^{occurs at the} highest

extraction potentials ^{where the} experimental ^values

appear ^to drop ^{off a little faster than the} theoretical values. There is relatively ^little penetration ^{of the} accelerating ^{electric field in these} experiments except in this region ^of highest ^field.

We turn now to the experimental configuration

studied at Ecole Polytechnique ^{where field} pen- etration into the source plasma ^{is dominant over}

much of the range ^of examination. In figure ^{4a are}

shown the experimental ^{results obtained with the}

Fig. ^4a.

Data from Ecole Polytechnique showing nega- tive ion current

function of accelerating potential ^for

several different

currents.

Fig. ^{4b. - The} family ^{of curves in} figure ^4a

^scaled

shown : dividing the ordinate by ^the plasma density ^and

the abscissa by the two-thirds power of the plasma density.

The solid line is the theoretical result.

hybrid multipole ^{and its} extractor, ^{shown in} figures ¹

and 2. Using ^{the same} scaling principles ^enumerated above ^{we obtained} the results illustrated in

figure 4b. The results for this experimental configur-

ation are significantly ^different from those of Stevens

et al. of figure 3, ^both qualitatively ^and quantitat-

ively. They ^{both exhibit the} rapid ^{falloff at low}

948

extraction potential but there the qualitative ^similari-

ties end ; ^{instead of a} plateau ^at intermediate values of extraction potential, ^{the ion current continues to} climb, although ^more slowly ^{then at low} potential.

Fig. ^5a.

Calculated potentials and H- ion orbits for the perveance marked A in figure ^{4b. PE,} plasma electrode ; SP, separator.

Fig. ^5b.

Calculated potentials and H- ion orbits for the perveance marked B in figure ^{4b. PE,} plasma electrode ; SP, separator.

The modelling reproduces ^this general feature, ^as

well as the space charge ^dominated region ^{as it did}

for the configuration of Stevens et al. [5]. ^{The sheath}

fields for the perveance marked A, B, C, D, ⁱⁿ

Fig. ^5c.

Calculated potentials and H- ion orbits for the perveance marked C in figure ^{4b. PE,} plasma electrode ; SP, separator.

Fig. ^5d.

Calculated potentials and H- ion orbits for the

perveance marked D in figure ^{4b. PE,} plasma electrode ;

SP, separator.

figure ^{4b are} calculated as shown in figure ^{5. The}

reason of the slow rise in current at high ^extraction potential is that the degree ^{of field} penetration ^is increasing ^with increasing extraction potential. ^In

the plateau region ^of figure ^{3 there was no} significant

field penetration.

We remarked that for the data of Stevens et al. [5]

the negative ^{ion current was} proportional ^{to the arc}

current [8] where there was a plateau. ^{This was}

found not to be the case for the Ecole Polytechnique

data where there is not a plateau. ^The dependence ^of

the extracted ion current on arc current is shown in

figure ^6. The deduction of the current dependence

follows immediately from direct observation of the Ecole Polytechnique data. Deduction of the current

density ^{in front of the extractor however} required

the conditions to be the same (e.g. ^field penetration,

sheath position). This is achieved by looking ^{at the}

current extracted at fixed perveance, when the sheath and the field penetration ^{are constant. As can}

be seen the extracted current varies rather strongly

with arc current ; ^the theory predicts ^{most of the}

behavior as shown. One can deduce from this agreement that the ion flux density ^to the extraction sheath is approximately proportional ^{to the arc}

current (assumed ^{in the} theoretical treatment).

In conclusion we have observed that field pen- etration dominates the hybrid multipole ^{studied at}

Ecole Polytechnique, ^{but not} the data of Stevens et al. Transverse space charge ^{limits occur for both}

cases in the domain considered. These studies indi- cate two specific configurations ^{where the} theory ^has

Fig. ^6.

Negative ^{ion current}

function of

current for (a) ^constant acceleration potential, ^denoted by ^the

square symbol, ^and (b) ^constant perveance, denoted by

the round symbol ; theory ^{for constant} perveance is shown

by the dotted line. At constant perveance the sheath and the field penetration

^constant.

produced results in agreement ^{with the} experimental

data.

Acknowledgements.

We thank W. R. Becraft and H. H. Haselton for their help ^and support in this research.

References

[1] ^BACAL M., Nucl. Instrum. Methods in Phys. ^Res.

B 37/38 (1989) ^28.

[2] WHEALTON J. H., ^{BELL M.} A., RARIDON R. J., ROTHE K. E. and RYAN P. M., ^J. Appl. Phys.

64 (1988) ^6210.

[3] WHEALTON J. H., Nucl. Instrum. Methods 189 (1981)

55.

[4] ^LEJEUNE C., Adv. Electron. Electron. Phys. Suppl.

Ser. 13C, ^A. Septier ^Ed. (Academic Press) 1983,

p. 207.

[5] ^STEVENS R., ^{YORK R.} L., LEUNG K. N. and EHLERS K. W., Fourth Intern. Symp.

Production and Neutralization of Negative ^{Ions and Beams} (Brookhaven, ^N. Y.) 1986 : American Institute of Physics Conference Proceedings ^n° 158,

J. G. Alessi Ed., p. 271.

[6] ^{BACAL M.,} BRUNETEAU A. M. and NACHMAN M., ^J.

Appl. Phys. ⁵⁵ (1984) ^15.

[7] ^BACAL M., BRUNETEAU J. and DEVYNCK P., Rev.

Sci. Instrum. 59 (1988) ^2152.

[8] WHEALTON J. H., ROTHE K. E. and MESZAROS P. S., IAEA Symposium

Negative ^{Ion Heat-} ing, ^Culham Laboratory (1987).

[9] McADAMS R., ^HOLMES A. J. T., NIGHTINGALE M. P. S., ^{LEA L.} M., ^{HINTON M.} D., ^NEWMAN A. F. and GREEN T. S., Fourth International

Symposium

Production and Neutralization of

Negative ^Ions and Beams (Brookhaven, NY)

1986 : American Institute of Physics ^Conference Proceedings ^{No. 158,} J. G. Alessi Ed., p. 298.

[10] ^SHERMAN J., Private Communication (1987).