Physics tests of an electron suppressor with variable electric and magnetic

fields
R. McAdams, R. F. King, and A. F. Newman

Citation: AIP Conf. Proc. 210, 255 (1990); doi: 10.1063/1.39595

View online: http://dx.doi.org/10.1063/1.39595
View Table of Contents: http://proceedings.aip.org/dbt/dbt.jsp?KEY=APCPCS&Volume=210&Issue=1
Published by the American Institute of Physics.

Related Articles
Profiles of ion beams and plasma parameters on a multi-frequencies microwaves large bore electron cyclotron
resonance ion source with permanent magnets
Rev. Sci. Instrum. 83, 02A317 (2012)
Production of rare isotope beams at the Texas A&M University Cyclotron Institute
Rev. Sci. Instrum. 83, 02A905 (2012)
Simulation of H ion source extraction systems for the Spallation Neutron Source with Ion Beam Simulator
Rev. Sci. Instrum. 83, 02A705 (2012)
Initial study of the optical spectrum of the ISIS H ion source plasma
Rev. Sci. Instrum. 83, 02A704 (2012)
Angular distribution of plasma in the vacuum arc ion source
Rev. Sci. Instrum. 83, 02A503 (2012)

Additional information on AIP Conf. Proc.

Journal Homepage: http://proceedings.aip.org/
Journal Information: http://proceedings.aip.org/about/about_the_proceedings
Top downloads: http://proceedings.aip.org/dbt/most_downloaded.jsp?KEY=APCPCS
Information for Authors: http://proceedings.aip.org/authors/information_for_authors

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 PHYSICS TESTS OF AN k-fl',;~CTRON SUPPRESSOR WITH
VARIABLE V,T.V,CTRIC AND MA~qETIC F T ~ D S

R. McAdams, R.F. King and A.F. Newman

UKAEA Culham Laboratory
Abingdon, Oxon. OXI4 3DB, England

ABSTRACT

This paper reports tests of an electron suppressor, for volume

negative ion sources with variable electric and magnetic fields. The
experin~ts have been carried out with a source and accelerator
operating d.c. and with an extraction aperture of 16 mm diameter.
The results are compared to those obtained with a permanent magnet
suppressor and to a theoretical model of the transport of electrons
by diffusion across the magnetic field.

INTRODUCTION

Practical applications of negative ion sources require stringent

control of any extracted electrons. For volu~e negative ion sources
operating in hydrogen the electron current can be of the order of 50
times the negative ion current if nothing were done to reduce the
extracted electron current. Lea et al. ~ have recently described an
electron suppressor with variable magnetic and electric fields which
showed some promise. However t_heir experiments used a small probe
accelerator with an extraction aperture of 1.5 mm diameter. We have
installed such a suppressor on the Culham Ion Source Test Stand using
a 16 mm diameter extraction aperture and operating d.c.. The results
are co~oar~i to those obtained u~ing a suppressor with permanent
magnets and to a diffusion model of the electron transport across
the magnetic field.

EXPERIMENTAL DETAILS

The volune sou~c~ and accelerator used in this work have been
described elsewhere J' . The plamna generator is3a copper magnetic
multipole bucket of dimensions 195 x 140 x 85 mm . The source filter
field was configured in two modes, as shown in Figure I, the dipole
filter has a field strength of - 30 Gauss at the extraction aperture
whereas the tent filter has only a field of - 5 Gauss at the
extraction aperture and will produce a more uniform plasma across the
extraction plane. The extraction aperture on the source was 16 mm
diameter. The accelerator was a triode design which enabled the
extracted electrons to be dumped within the accelerator structure at
an energy of about 1/6 of the final ion beam energy. The extracted
ion beam was measured downstream by a d.c. beam transformer. Figure
2 shows the schematic source and accelerator.

1990 American Institute of Physics 255

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 256 Physics Tests of an Electron Suppressor

Source Magnetic Filter Configurations

.I
N

S
CGm~ mI

(i) Dipole Filter (ii) Tent Filter

Figure 1 The source f i l t e r f i e l d configurations.

Sourceand Acceleratoraeamllne
Beam
Qr,O. FOr,021 i rQ.03 Transformer
Source | I I I r'-'c'11 DeflectOr
"~. | I I I|~:,[ Magnet

!/ !!--."
/ tAI I II II
Suppressor II ~ !1

I
Figure 2 The source and accelerator.

Figure 3 shows a schematic diagram of an electron suppressor. A

magnetic field in the region of the extraction aperture magnetises
the electrons which might otherwise have been extracted. These
electrons move along the field lines and are collected on a biasable
electrode threaded by the magnetic field.

Figur~ ~ shows the permanent magnet suppressor used

previously J,~" It consists of two 37 x 4 x 4 mm samarium cobalt
magnets arranged in a quadrupole configuration. The integral field
in the source is - i00 Gauss cms. A portion of the extraction
electrode forms the electron collector.

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 McAdams et al. 257

Schematic Electron Suppressor

electron collection
surface "~,,,L,L,.b, LH

~ ction electrode

> beam
plasma
fie

....... rHf..1,,i

electron depleted zone

Figure 3 A schematic electron suppressor.

Permanent Magnet Electron Suppressor

source

extraction grid

",\ /
rll ms
insert
accelerator
T
1
I
section of extraction aperture

Figure 4 The permanent magnet suppressor.

Figure 5 shows the variable field suppressor. The field is produced

by a flat solenoidal coil made of wires space 5 ran apart. The field
profile has a FWHM of - 5 nTn and an peak intensity of ~ 1 Gauss/
Ampere. Thermal considerations limit the coil current to 150 A. At
each end of the array of wires is the collection electrode.

THEORETICAL MODEL

Green2 has used a diffusion model to describe the electrons

transport across the magnetic field. The electrons diffuse by
electron-molecule collisions and are then extracted. Some move along
the field lines and are collected by the biasable plate. This model

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 258 Physics Tests of an Electron Suppressor

VariableMagneticFieldSuppressor

ElectronCollection~ t Extraction
ctrode

Be~nec~rdni~ng ]i
Figure 5 The variable field suppressor.

gives the following relations for the extracted electron current, I^,
the current deposited on the collection electrode, Iins, the magnetic
field strength B and t_he pressure in the suppressor reglon, p:

Ie = (K/B) exp (-TB) (I)

]-n(Ie/Iins) = a - bIin s (2)

and ~n(Ie/Iin s) = c - d/p~ (3)

where K, 7, a, b, c, d are constants.

These ec~ations can be tested directly against t_he data. The

work of Green ~ has already shown that the permanent magnet suppressor
is described well by the diffusion model through equations (2) and
(3) however because of the fixed value of magnetic field equation (I)
could not be tested for this type of suppressor.

Just as i,portant as the effect on the electrons is the effect

of the suppressor on the extracted ion cuz~ent. The diffusion model
does not address the possible repercussions for the ion current.

EXPERIMENTAL RESULTS

a) Comparison between the suppressors

Figure 6 shows the electron suppression action of the two c~vices

as a function of the collector bias, Vin s.

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 McAdams et al. 259

Hydrogen
Dipole Filter Field

2oo "x I~ =20A

",. V~, - 100V
"%
x V,.,i =11kV x OAonColl ]
", VN~,,8OkV 50AonColl L ~ a g n e { ~
"%. IOOAon C.~ / S u p p ~
15o "%.. Q-0.4Tts" *150AonCoil J
"x. Permanent Magnet S u ~

~ 1oo
%k %~
...,,
50

I i K I I I I

v,,,(v)

Figure 6 The dependence of extracted electron current on collector

voltage for both suppressors.

Both devices do indeed suppress the extracted electron current.

The degree of suppression obtained with the variable field device
does indeed depend on the strength of the magnetic field. Even at
the relatively low arc current used of 20 A, coil currents of > 150 A
arc needed to reduce the extracted electron current as low as that
from permanent field device.

Of course just as important is that t h e control of t h e extracted

electron flux does not have a drastic effect on the extracted
negative ion current. In Figure 7 we show the corresponding data for
the ion current, IB.T., as measured at 0.82 m downstream by the d.c.
beam transformer.
Hydrogen
Dipole Filter Field
8

J
i| i i i
3
IL, - 20A x 0A on Coil 1 Electro-
V~,,c - 100V 50A on Coil I" Magnetic
2 O.O.4Tt$-' IOOA on Coil . j Suppmsa0
V.;~ - 1 1 k V 150AonCoil
1 V~.,a - 80kV Permanent Magnet
Suppressor |
0 I ~ I I I I t ! I I I~
2 3 4 5 6 7 8 9 10 lil 12 13
V,,,,(V)

Figure 7 The dependence of extracted negative ion current on

collector voltage for both suppressors.

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 260 Physics Tests of an Electron Suppressor

Although the ion current does not vary by more than 10% with the
collector voltage for either suppressor, the extracted ion current in
the case of the variable field or electro-magnetic suppressor is -
60% of that for the permanent magnet suppressor. Now the
transparency of the tw~ sets of wires in the variable field
suppressor is - (0.75)= = 0.55. This would be the transparency one
would empect if the negative ions were created before the suppressor
fie~d in the coils. However this is not the case as shown by Lea et
al.~. They r~moved the source filter but still found negative ions
being produced due to the suppressor field acting as a filter field.
It could be that there are further insertion losses associated with
the suppressor leading to a decrease in plasna density and hence
negative ion production.

hhat Lea et al. 1 also found was that as the B-field was
increased from zero then the negative ion yield increased from 40% to
80% of the current density obtained in the absence of the suppressor
and then decreased slowly. This has also be observed by Bacal ~. The
data in Figure 7 does not show this behaviour. The negative ion
current decreases with increasing B-field. This experiment was
carried out with the source filter field in the dipole configuration.
The integrated field from the peak to the extraction aperture due to
be dipole field- 200 Gauss cms and the field strength at the
extraction aperture is - 30 Gauss although 1-2 Gauss of this comes
from the accelerator fields. Thus even with zero coil current the
field is never zero. Thus the results of Lea et al. could not be
observed in this filter configuration.

In order to test this hypothesis we reconfigured the source

filter field into the tent configuration as shown in Figure 1 (b).
The field at the extraction aperture then falls to - 5 Gauss. In
Figure 8 will show a cc~parison of the dependence of the ion negative
current on the coil current (or B-field) for the two filter
configurations.

It is seen that as t_he coil B-field approaches to zero in the

tent mode t_he negative ion current falls hence proving the
hypothesis. The coil current is not set to zero because the electron
current approaches 1 A which poses a difficulty in thermal
management. This large electron current at such a low arc current is
associated with the low value of ion current. This is almost
certainly due to the fact that in the tent configuration the
filaments penetrate the filter field thus increasing the plasma
temperature in the extraction region of the source and so producing a
decrease in the density of negative ions.

That this is so is shown in Figure 9 where we plot 2n (Ie)

versus Ic~i] (or B) for both filter configurations. In the tent
filter case-the electron current is - i0 times that in the dipole
filter case. Also the slope of the line for the tent filter data is
less than that for the dipole filter. The efficiency of the
suppression will decrease with increasing temperature thus we see

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 McAdams et al. 261

Hydrogen
Variable Field Suppressor

6.C IA.c = 20A

V.,c= 100V Dipole Filter Field
Q = 0.6Tts-' x Tent Filter Field
5,C
Vl.. = 13.5V

4.0

le'T" 3.0
(mA)

2.0 x
X X x XX
X X
x X X
1.0

0.0
0 2b 4~ ~ 8~ ,60 1~o 1;o 1~o 1~0 260
I ~ (A)

Figure 8 The dependence of extracted negative ion current on

magnetic field for both source filter configurations.

Hydrogen
Variable Field Suppressor
I,~ - 20A
VA,m=10QV
7 Q-0.6Tts" X Tent Filter
x x x x x X V,.-13.5V - Dipole Filter
E X X XX X.

4
t. (i~
3

I
|
,!
0
I~ (A)

Figure 9 The dependence of extracted electron current on magnetic

field for both source filter field configurations.

that the electrons are hotter in the tent filter mode. These hotter
electrons will lead to a reduction in negative ion yield through
destruction by electron impact.

We are presently redesigning the filaments in the source to

avoid the creation of a hot plasma beyond the nominal filter field
and so increase the yield of negative ions beyond what we have at
present in the tent configuration.

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 262 Physics Tests of an Electron Suppressor

b) Omparison with theory

The dependence of extracted electron current on B-field as given

on equation (i) can be tested for the variable field m/ppressor.
Instead of the functional form in equation (i) we find a purely
exponential dependence as shown in Figure i0 which plots ~n (Ie)
versus Icoil (or B).

Hydrogen
Variable Field Suppressor
Dipole Filter Field
IA.C : 2 0 A Q
6.0 VA,c = 100V 0.2Tts-'
V~,, = 13.5V
x 0.4T~s"
o 0.6Tts"
5.0 + 0.8Tts"

4.0

(I.). 3.0

2.0

1.0

0
4o 8o do ' 26o
Ic~.(A)

Figure I0 The exponential dependence of extracted electron current

on magnetic field strength.

The remain/rig experimental tests given by the diffusion model

can be used for both devices. The dependence of ~n(Ie/Iins) versus
Iins is obtained by varying the electron collector voltage. The
results for tb~ two suppressors are shown in Figures ii and 12.

For the permanent magnet suppressor the linearity of this curve

(and hence agreement with the diffusion model) is much more
pronounced than for the variable field suppressor. The deviation
from linearity at the lower values of I L ~ is not entirely due to a
failure of the model. At these low values of collection voltage,
positive ions are also collected because the plamna potential has not
been exceeded thus leading to an apparent decrease in the collected
positive current.

The diffusion model is based on electron-molecule scattering in

the suppressor region leading to transport across the magnetic field.
Thus the pressure dependence of the suppressor action ought ~o be a
very good test of its validity. This dependence is expressed through
equation (3). Instead of using pressure we use the gas flow, Q, the
two being directly proportional.

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 McAdams et al. 263

Hydrogen
Permanent Magnet Suppressor
Dipole Filter
Insert Voltage Scan
1 2 3 4 5 6 7 I,,,(A)
00

-1
I.,,c =100A
V,A - 105V
-2 Q = 0.54T~s-'
4, -16mm

-3

-4

-5

-6

Figure ii The dependence of ~n (Ie/Iins) on Iins for the permanent

magnet suppressor.

Hydrogen
Variable Field Suppressor
Dipole Filter Field
I,. c =20A
1.0 VA. = lO0V
Q = 0.6Tl%-'
I,.. (mA)
0
2bo ' 660 ' ,060
X
-1.0

~n(i-~ )
-2.0

lcall
-3.0 OA
50A
IOOA
-0.4 150A ,=,,,,,,,,

Figure 12 The dependence of ~n (Ie/Iins) on Iins for the variable

field suppressor.

Figures 13 and 14 show data for ~ t h suppressors plotted in the

form 2n (I~/I4,o) or ~n (Ie) versus Q in accordance with equation
(3). For ~e=~rmanent magnet suppressor ~n (Iins) only changed by 5%
hence the use of ~n (Ie).

It is apparent that the diffusion model is in good agreeae~t

with the data from the permanent magnet suppressor but not at all in
agreement for the variable field suppressor where it is almost
independent of pressure. The deviation frcm linearity at high
pressures (or Q) in the case of t~e permanent magnet device has been
shown previously by McAdams et al~ to be due to the collection, in

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 264 Physics Tests of an Electron Suppressor

Hydrogen
Permanent Magnet Suppressor
Dipole Filter
Pressure Scan
t~ -20A
V~.80V
V~,= 5V

In (I.} 2

i
Ot i i
0.8 1:0 1.2 1:4 116 1.8
O-,* (Tts-,)-~

Figure 13 The dependence of ~n (Ie) versus Q- for the permanent

magnet ~t:tm_~ssor.
Hydrogen
Dipole Filter Reid
IAw- 20A
VA,,- IOOV
V,,,, - 13.5V
~ ~rt m-')-~
0 10 20 30
-1.0 , , ,
Im,
x 0A
GOA
o IOOA
-2.0 * 160A

me

-3~ ~o

-4

Figure 14 The dependence of ~n (Ie/Iins) versus Q' for the variable

f i e l d suppressor.

the accelerator, of electron produced by stripping of the H- in

collisions with the gas flowing frcm the source.

CC~CI~SIONS

We have described a new electron suppressor with variable

electric and magnetic field~ based on an extension of previous ideas
and guided by previous data ~. This device worked well with small
extraction apertures i.e. with a diameter smaller than the wire
spacing but its performance for large extraction aperture diameters

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 McAdams et al. 265

is very poor compared to the permanent magnet device. This loss of

performance appears to be associated mainly with the transparency of
the device. Howe~er what we have shown with this variable field
device, and Bacal ~, magnetic field close to the extraction aperture
leads to an enhancement of the negative ion yield confirming the
results of Lea et al. ~.

In our experiments on this enhancement the performance was

limited by the filaments penetrating the filter field leading to a
hotter plasma and hence lower densities of negative ions. Thus in
our continuing work we will optimise the filament/tent filter design
in order to see if the enhancement persists of higher negative ion
current densities. Until this is done t_he full potential of the
enhancement effect will not be realised.

The characteristics of the suppressor action are well described

by a diffusion model in the case of a permanent magnet device. This
is not the case for the variable field device. The reasons for this
could be in a number of areas. Notably the scale lengths of fields
for each suppressor are different as is the detailed magnetic~
geometry. This problem is being addressed by Holmes and Haas v using
a hydrodynamic model.

ACKNO~7 m X ~ T S

The authors would like to thank Dr A. Holmes and Dr T. Green for

many discussions on electron suppressor action. This work was
supported by USAFOSR under contract number F49620-86-C-0064.

REFEPa~CES

I. L.M. Lea, A.J.T. Holmes, M.F. Thornton and G.O.R. Naylor,

Proceedings Int. Conf. Ion Sources (Berkeley 1989), Rev. Sci.
Instr. to be published.
2. T.S. Green, SPIE Proceedings Vol. 1061, Microwave and Particle
Beam Sources and Directed Energy Concepts, P. 628 (1989).
3. R. McAdams, A.F. Newman, R.F. King and A.J.T. Holmes,
Proceedings Int. Conf. Ion sources (Berkeley 1989), Rev. Sci.
Instr. to be published.
4. R. Ym_Adams, A.J.T. Holmes, A.F. Newman and R.F. King, Production
and Application of Light Negative Ions. 3rd European Workshop
Amersfoort p. 15 (1988).
5. M. Bacal unpublished data.
6. A.J.T. Holmes, private communication.

Downloaded 09 Feb 2012 to 194.81.223.66. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions