Review of Scaled Penning H" Surface Plasma Source with
Slit Emitters for High Duty Factor Linacs
Joseph D. Sherman, William B. Ingalls, Gary Rouleau, and H. Vernon Smith, Jr.
Los Alamos National Laboratory, Los Alamos, NM87545
Abstract. The Penning H" surface plasma source (SPS) has been used at Rutherford Appleton Laboratory (RAL) for 20
years to provide the required H" beams for charge-exchange injection into the 800-MeV proton synchrotron on the ISIS
spallation neutron source. The RAL source is based on the first H" Penning SPS operated at Los Alamos. Since that
original technology exchange, Los Alamos has developed scaled-up versions of the Penning H" SPS with the goal of
extending the H" beam duty factor (df) while maintaining high beam brightness. A 250-mA H" beam with rms
normalized emittance of <0.3 (Tcmm-mrad) in both transverse planes has been extracted from a 4X scaled Penning
source at a discharge df of 0.5%. Using discharge scaling laws and the 250 mA H" current results, it is predicted that a
4X Penning H" SPS with a slit emitter would be capable of producing >100-mA, low emittance H" beams in the 5% df
range. A source with these parameters would be suitable for the European Spallation Source (ESS) and other highpower proton driver projects.
LANL Penning and the Rutherford-Appleton Labs
(RAL) Penning SPS H" source performances are made.
The final section summarizes conclusions concerning
the H" Penning SPS scaling laws, and the possibility of
attaining a slit-based Penning H" SPS with beam
characteristics of > 100 mA, transverse plane emittance
< 0.3 (Tcmm-mrad), and 5% df. For background
literature for H" production using cesiated cathode
surfaces see[6], and other publications contained in
this series.
INTRODUCTION
Proposed accelerator drivers for short and long
pulse spallation targets for neutron beam production
place demanding requirements on the injector
section[l]. The European Spallation Source (ESS)
proposal, for instance, asks in the short-pulsed mode
for 100-mA H" beams in a linac operating at 5% duty
factor (50 Hz, 1.0 ms)[2]. The purpose of this paper is
to review the dimensionally and discharge-power
scaled H" Penning surface-plasma sources (SPS) using
slit emitters in terms of high H" current and
comparatively long duty factor (df) requirements (5%).
Previously published work on the 4X scaled Penning
source at Los Alamos showed that 250-mA H" current
could be extracted at 29 keV from a slit emitter[3] at
0.5% df (5 Hz, 1 ms). This attractive beam current
was somewhat overshadowed by asymmetric
emittances of 0.15 X 0.29 (Tcmm-mrad) rms,
normalized, which was thought to be difficult to match
to a radio-frequency quadrupole (RFQ). However,
over the past ten years RFQ design activities have
shown that RFQs with a beam acceptance of 0.3
(Tcmm-mrad) are possible while still maintaining a
high-quality linac design [2]. Further, design studies of
matching beams with asymmetric emittances in the
transverse planes to an RFQ using solenoid-lens [4]
and quadrupole magnet focusing[5] low-energy beam
transport (LEBT) systems indicate successful matches
are possible. Thus there is a renewed interest and
acceptance of the use of slit beams as injectors for
RFQs.
The following section summarizes the Penning H"
discharge scaling laws as developed at Los Alamos
National Lab(LANL)[3]. Historical comparison of the
PENNING H SPS SCALING LAWS AND
HISTORICAL PERSPECTIVE
Application of discharge scaling laws to the SPS H"
Penning discharge for scaling source dimensional
parameters up by the factor 4 in two of the three
spatial dimensions has led to relationships among the
discharge parameters summarized in the equations of
Table 1. Here the subscript 4X refers to the Penning
SPS H" source with a cathode area four times the IX
source. These equations are based on the general
theory of surface H" production [6] and discharge
scaling laws contained in standard texts [7] on
discharge physics. The prospect for increased 4X
source df (df4X) is related to the decreased cathode
power load, F4X. A fundamental limit to the stability
of the Penning discharge and hence stable H" beam
formation is the cesium equilibrium concentration in
the discharge volume and cathode surfaces[8]. The
pulsed temperature rise of a surface[9] is given by
ATnX = 2FnXsqrt(At/(7iKpC)) where ATnX=pulsed
temperature rise,
FnX=cathode power flux
(W/cmz)=VnXInX/(3CAnX), At=pulse
length(s),
CP642, High Intensity and High Brightness Hadron Beams: 20th ICFA Advanced Beam Dynamics Workshop on
High Intensity and High Brightness Hadron Beams, edited by W. Chou, Y. Mori, D. Neuffer, and J.-F. Ostiguy
2002 American Institute of Physics 0-7354-0097-0
288
TABLE 2.
2. Comparison
Comparison of
ofthe
theLANL
LANLand
andRAL
RAL1X
IXSPS
SPS
TABLE
dimensions
with
the
4X
source.
dimensions with the 4X source.
K=thermal
conductivity
constant(W/cm°C),
K=thermal
conductivity
constant(W/cmoC),
3
3), and C=specific heat (J/gm°C).
p=density(gm/cm
ρ=density(gm/cm ), and C=specific heat (J/gmoC).
TABLE
of the
the 4X
and 1X
IX ion
ion
TABLE 1.
1. Predicted
Predicted relationships
relationships of
4X and
source
discharge
parameters.
The
lower
case
parameters
source discharge parameters. The lower case parameters ii
and
case characters
characters
and jj refer
refer to
to the
the extracted
extracted H"
H- beam.
beam. Upper
Upper case
refer
to
the
ion
source
discharge
parameters.
Typical
units
refer to the ion source discharge parameters. Typical units
are
indicated.
are indicated.
DISCHARGE
DISCHARGE PARAMETER
PARAMETER
11
22
33
44
55
66
77
88
99
10
10
11
11
12
12
Discharge
Discharge pressure
pressure (Torr)
(Torr)
H
gas
mass
2
H2 gas mass flow
flow (seem)
(sccm)
Discharge
Discharge magnetic
magnetic field
field (G)
(G)
Discharge
Voltage
(V)
Discharge Voltage (V)
Discharge
Discharge Current
Current (A)
(A)
Cathode
Cathode area
area (cm
(cm2))
Cathode
Cathode2power
power load
load
(kW/cm
(kW/cm2))
Discharge
current
Discharge
current densities
densities
2
(A/cm
(A/cm2))
22
Emission
Emission area
area (cm
(cm ))
Extracted
Extracted H"
H- beam
beam current
current
(mA)
(mA)
Extracted
Extracted H"
H- beam
beam current
2
density
density (mA/cm
(mA/cm2))
H"
H- beam
beam production
production
efficiency
efficiency (mA/kW)
(mA/kW)
COMPONENT
COMPONENT
Ion Source
Source
Ion
(mm)
LL (mm)
Cath.-cath. gap
gap
Cath.-cath.
W(mm)
W (mm)
Discharge depth
depth
Discharge
(mm)
TT (mm)
Discharge length
length
Discharge
Emission slit
slit
Emission
(mm)
yy (mm)
Perp. to
to BB field
field
Perp.
(mm)
xx (mm)
Parall. to
to BB field
field
Parall.
4X-1X
4X
– 1X
RELATIONSHIP
RELATIONSHIP
P4X
4 x=Pix/4
P
= P1X/4
Q4X
=Q
Qix
4X =
Q
1X
B
=B
Bix/4
4X =
B4X
1X/4
V4X
=V
4X=
V
V1X
1X
= IlX
II4X
4X = I1X
CA4x
= 4CAlx
CA
4X = 4CA1X
F4X
=F
Fix/4
4x =
F
1X/4
RAL
RAL
IX
1X
LANL IX
LANL 1X
LANL 4X
LANL 4X
55
4.3
4.3
1717
22
33
1717
10
10
12
12
1616
10
10
10
10
11.4
11.4
0.6
0.6
0.5
0.5
2.8
2.8
Figure 11 shows
shows the
the measured
measured cathode
cathode power
power load
load
Figure
for the
the 1X
IX and
and 4X
4X sources
sources plotted
plotted versus
versus
nX for
FFnX
performance publication
publication date.
date. The
The RAL
RAL 1X
IX source
source
data
are
from
ref.[ll,12],
and
the
LANL
IX
source
data are from ref.[11,12], and the LANL 1X source
data are
are from
from ref.[13,14].
ref.[13,14]. The
TheiinX
anddf
dffor
forthe
thevarious
various
data
nxand
sources are
are shown
shown in
in Fig.
Fig. 1.1. An
An observation
observation for
for the
the
sources
IX sources
sources isis that
that HH"
current isis inversely
inverselyrated
ratedtotothe
the
1X
current
80
duty factor
factor with
with the
theproduct
product of
of(i(ii
)(dfi
(mA%)
duty
1Xx)(df
1Xx)) ==80 (mA%)
being nearly
nearly constant.
constant. The
The exploratory
exploratory experiment
experiment
being
completed for
for the
the 4X
4X slit
slit emitter
emitter which
which produced
produced 250
250
completed
mA H
H"
did not
not probe
probe the
the df
df4X
limit, although
although the
the 4X
4X
did
4X limit,
source has
has been
been operated
operated atat 2.3%
2.3% df
df4X
with
circular
source
with
circular
4X
emission apertures
apertures with
with the
the instantaneous
instantaneous cathode
cathode
emission
power load
load of
of 2.5kW/cm
2.5kW/cm22[3],
[3], slightly
slightly more
more than
than the
the
power
2.2kW/cm22 for
for the
the slit
slit beam
beam result
result shown
shown in
in Fig.
Fig. 1.1. ItIt
2.2kW/cm
is
is readily
readily apparent
apparent from
from previous
previous 4X
4X circular
circularaperture
aperture
measurements
measurements and
and the
the data
data contained
containedininFig.
Fig.11that
thatthe
the
4X slit
slit source
source df
df4X
could be
beincreased.
increased.
4X could
= JJix/4
JJ44Xx =
1X/4
EA4x
4EA1X
EA4X = 4EA
lx
ii4x
4X = iiix
1X
J4X
= jlX/4
j4X =
j1X/4
^1X
ξ^4X
4X = ξ
1X
The
The index
index nn equals
equals one
one or four for the discussion in
this
temperature
this paper.
paper. The
The factor
factor (1/3) in the surface temperature
increase
increase equation
equation comes from
from equipartitioning the
total
total discharge
discharge power
power to each of the two cathodes and
to
to the
the anode.
anode. The
The fundamental
fundamental hypothesis of the 4X
technology is
is that
that by decreasing the pulsed surface
technology
temperature rise
rise the
the proper
proper cesium balance can be
temperature
maintained at
at longer
longer duty factors. The beam current
maintained
density defined
defined as
as entry
entry 11 in Table 1 is defined as jnnX
density
x
/EAnX
== iinnX
x/EA
nX. The emission area EAnX is equal to the
emission slit
slit length
length (ynX
emission
nX) multiplied by its width (xnX)
please see Table 2. The
-– please
The x4X
= 2.8mm
2.8mm is
is somewhat
somewhat
4X =
larger than
than the
the value
value of 2.0mm desired from the factor
larger
4X scaling,
scaling, and
and was the
the result
result of a fabrication error.
4X
Cathode Power Load, F (kW/cm2)
20
20
Table 22 shows
shows dimensional comparisons of the
Table
LANL
1X
source,
the RAL
RAL source, and the 4X source
LANL IX source, the
as developed
developed at
at LANL.
LANL. It is noted the LANL 1X
as
IX and
RAL Penning
Penning sources have nearly identical
RAL
identical
dimensions, reflecting
reflecting the
the earlier
earlier technology
technology exchange
dimensions,
exchange
between the
the two
two labs[10].
labs[10]. Thus,
Thus, both
both the
the LANL
LANL 1X
between
IX
and RAL
RAL sources
sources may
may be
be referred
referred to
to as
and
as 1X
IX sources.
sources.
Further definition
definition of
of the
the ion
ion source
source and
and emission
emission slit
Further
slit
components
are
described
in
Fig.
2
of
ref.[3].
components are described in Fig. 2 of ref.[3].
LANL 4X, 0.5% df
•
15
15 -
ISIS 1X, 2.2% df
160
mA H160mAH-
LANL 1X, 0.75% df
LANL 1X, 0.5% df
10-
10
104
mA H104mAH55
35
35mA
mAHH-
250
250mA
mAHHA
0
1975
1975
1980
1980
1985
1985
1990
1990
1995
1995
2000
2000
Year
Year
FIGURE
FIGURE 1.
1.- Measured
Measured cathode
cathodepower
powerloads
loadsfor
forthe
the1X
IXand
and
4X
scaled
H
Penning
SPS
.
4X scaled H" Penning SPS.
Another
inin scaled
Another useful
useful factor
factor to
to consider
consider
scaled HH"
Penning
Penning SPS
SPS technology
technology isis the
the HH" beam
beam production
production
efficiency
ξ
shown
as
entry
12
in
Table
efficiency ^nXx shown as entry 12 in Table1[3].
1[3]. The
The
definition
of
ξ
=
j
/F
,
thus
ξ
=
ξ
is
nX = jnXx/FnXx, thus ^
4X = ^
1X isexpected
definition of ^
expected
x
n
n
x
x
from
from the
the scaling
scaling laws.
laws. The
The publication
publication data
data used
used inin
289
H- Beam Production Efficiency (mA/kW)
Fig. 11 is
is used
used to
to construct
construct Fig.
Fig.2,2,where
whereξ^
x is plotted
Fig.
nX is plotted
vs
year
for
the
IX
and
4X
sources.
^
can
bethought
thought
x
vs year for the 1X and 4X sources. ξnX can be
- ions from
of as
as the
the global
global efficiency
efficiency of
of moving
moving HH"
of
ions from
their cathode
cathode birthplace
birthplaceto
tothe
the emission
emissionslit
slitaperture
apertureby
by
their
the
resonant
charge-exchange
process
on
slow
atomic
the resonant charge-exchange process on slow atomic
- attenuation as
hydrogen ions,
ions, and
and the
the subsequent
subsequent HH"
hydrogen
attenuation as
they
move
to
the
emission
slit[6,15].
TheLANL
LANL1X
IX
they move to the emission slit[6,15]. The
source
^
is
somewhat
greater
than
the
RAL
IX.
This
source ξ1Xx is somewhat greater than the RAL 1X. This
difference may
may be
be due
due to
to slightly
slightly different
different anode
anode
difference
geometries, where
where iii
are
exponentially
dependent
on
x
geometries,
are
exponentially
dependent
on
1X
anode length
length in
in the
the direction
direction of
of the
the emission
emission slit[15].
slit[15].
anode
outstanding feature
feature observed
observedininFig.
Fig.2,2,however,
however,isis
The outstanding
the factor two
two increased
increased cathode
cathode power
power efficiency
efficiency ofof
the 4X
4X source
source compared
comparedto
tothe
the1X
IXsources.
sources.
400
400
These phase-space
phase-space measurements
measurements were
were made
made
These
approximately
10
cm
after
the
extraction
electrodes
approximately 10 cm after the extraction electrodes
with high
high resolution
resolution electric
electric sweep
sweep phase-space
phase-space
with
scanners
[16].
Corresponding
x-y
phase-space
scanners[16].
Corresponding x-y phase-space
emittancesatatthe
the250-mA
250-mAHH"
currentareareεxeXx Xεye=y =
0.15
emittances
current
0.15
X0.29
(Tcmm-mrad)
rms,
normalized.
The
emittances
X0.29 (πmm-mrad) rms, normalized. The emittances
are derived
derived byby extrapolating
extrapolating the
the measured
measured rms
rms
are
emittances
at
beam
fraction
F
(£rms(F))
to
100%
beam
emittances at beam fraction F (εrms(F)) to 100% beam
fractionassuming
assuminga aGaussian
Gaussianbeam
beammodel[17].
model [17].For
For
fraction
thethe
caseofofthe
thex-plane
x-planephase-space
phase-spacedistribution,
distribution,thethebeam
beam
case
appearstotobebeGaussian
GaussiantotoF F= =0.95.
0.95. For
Forthethey-plane
y-plane
appears
distribution,deviations
deviationsfrom
fromthe
theGaussian
Gaussianmodel
modeloccur
occur
distribution,
0.63. The
They-plane
y-planenon-Gaussian
non-Gaussianbehavior
behaviorat at
atatFF==0.63.
higher beam
beam fractions
fractions isisattributed
attributedtotothetheslit-end
slit-end
higher
aberrationsseen
seenininFig.
Fig.4(B).
4(B).These
Theseaberrations
aberrations
could
aberrations
could
ISIS 1X, 2.2% df
(A) x-plane
LANL 4X, 0.5% df
300
. (B) y-plane
LANL 1X, 0.75% df
LANL 1X, 0.5% df
; 200
200 -
-20
100
100 -
0
1975
1975
1980
1980
1985
1985
1990
1990
1995
1995
2000
2000
Year
Year
FIGURE
FIGURE 2.
2. H
H"beam
beam production
productionefficiency
efficiency for
forthe
the1X
IXand
and
4X
4X sources
sources plotted
plotted vs
vsdata
datapublication
publicationyear.
year.
4X
4X SCALING
SCALING LAW
LAWRESULTS
RESULTS
The
The 250-mA
250-mA HH"beam
beam pulse
pulse atat 29
29keV
keVextraction
extraction
energy
is
shown
in
Fig.
3.
The
beam
noise
energy is shown in Fig. 3. The beam noiseofof+/+/-1%
1%
FIGURE. 3. 250-mA H- beam pulse obtained from the 4X
FIGURE. 3. 250-mA H" beam pulse obtained from the 4X
slit source.
slit source.
is compatible with linear accelerator applications.
is compatible with linear accelerator applications.
Figure 4(A) shows the measured x-plane (narrow slit
Figure
4(A) shows the measured x-plane (narrow slit
dimension) phase space and Fig. 4(B) shows the
dimension)
phase space and Fig. 4(B) shows the
measured y-plane (long slit dimension) phase space
measured
y-plane
(long slittodimension)
space
measurement corresponding
the 250-mAphase
H- current.
measurement corresponding to the 250-mA H" current.
-10
0
10
Position (mm)
FIGURE
FIGURE4.4. AAsetsetofofx x(A)
(A)and
andy y(B)(B)phase-space
phase-space
measurements
beam
extracted
from
measurementsonona a29-keV,
29-keV,250-mA
250-mAH H"
beam
extracted
from
a aslit
.
slitemitter
emitter.
bebereduced
reducedwith
withimproved
improvedextractor
extractordesign.
design.The
The250250mA
mA HH"current
current and
and emittance
emittancemeasurements
measurementswere
were
made
made inin two
two days
days inin 1987,
1987, and
and nono further
further
characterization
characterizationofofthe
thesource
sourcewas
wasmade
madeasasnonoapparent
apparent
application
applicationwas
wasenvisioned.
envisioned.
Table
results[3]
Table3 3summarizes
summarizesthe
the1XIXtoto4X4Xscaling
scaling
results[3]
for
for the
theslit
slitbeam.
beam. Column
Column1 1gives
givesthe
theion
ionsource
source
parameter,
parameter,column
column2 2gives
givesthe
the1XIXactual
actualdesign
designand
and
column
column3 3gives
givesthe
the4X
4Xdesign
designbased
basedononthethescaling
scaling
laws
lawsofofTable
Table1.1.Column
Column4 4shows
showsthetheactual
actualresult
resultforfor
the
optimized
4X
source,
and
column
5
shows
the optimized 4X source, and column 5 showsresults
results
for
forananaggressively
aggressivelycooled
cooled4X
4Xcathode
cathodeoperating
operatingininthethe
discharge-only
discharge-onlymode[18].
mode[18]. There
Thereis isgenerally
generallygood
good
agreement
agreementbetween
betweencolumn
column3 3(4X
(4Xdesign)
design)and
andcolumn
column
44(4X
increased
4X was
(4Xactual).
actual).The
Thedischarge
dischargedepth
depthWW
4X was increased
from
12
to
17
mm
to
obtain
a
quiescent
discharge
that
from 12 to 17 mm to obtain a quiescent discharge
that
produces
a
low-noise
beam.
To
reach
the
j
4X design a
produces a low-noise beam. To reach the j4X design
a
2
cathode
cathode power
power density
density ofof only
only2.2
2.2kW/cm
kW/cm2was
was
required, about a factor of two lower than the
required, about a factor of two lower than the
4.2kW/cm2 2 scaling law prediction. This reduction
4.2kW/cm scaling law prediction. This reduction
results in a factor 2 increase in the 4X H- beam
results in a factor 2 increase in the 4X H" beam
production efficiency (cf. Fig. 2) compared to that of
production efficiency (cf. Fig. 2) compared to that of
the 1X sources. Operating at j4X = 783 mA/cm2 2
the IX sources. Operating at j = 783 mA/cm
resulted in i4X=250mA. One may 4X
expect a factor of
resulted in i4X=250mA. One may expect a factor of
5.6 increase in the 4X source x-plane emittance[17]
5.6 increase in the 4X source x-plane emittance[17]
based on the increase of slit width from 0.5mm to
based
on the increase of slit width from 0.5mm to
2.8mm, but only a factor of 2.5 emittance increase is
2.8mm, but only a factor of 2.5 emittance increase is
observed (0.15 πmm-mrad vs. 0.06 πmm-mrad).
observed (0.15 Tcmm-mrad vs. 0.06 Tcmm-mrad).
290
There is a 70% increase in the y-plane emittance,
which may be decreased by optimized extractor
design.
This paper is concluded by a discussion of the
application of the 4X technology to 100-mA, 5% df4X
H" beams. First, from the fifth column of Table 3 the
4X source has already operated at 6% df4X at LiX =
125A of arc discharge current[18]. Thus, to reach
nominal 6% df at 100-mA H" from the measured 4X
power efficiency in column 4, only I4X = 80A of arc
discharge current would be required, as
I4X(i4X=100mA) = (100mA/250mA)197A = 80A. This
scaling implies that the H" beam production is a linear
function of the Penning discharge current. This
scaling has been demonstrated from 10"2 to 102 A arc
current in the IX source[13]. Another argument is the
observation that a circular emitter beam on the 4X
source operated at 180A discharge with 2.3% df4X[3].
Assuming the product of i4X and discharge df4X(%) is
constant (as observed in the LANL and RAL IX
sources - see discussion above), one can write
(i4x)(df4X) = 250mA(180A/197A)(2.3%) = 525(mA%),
or at 5% df4X a 105 mA H" beam could be obtained.
Using the linear H" beam/arc discharge current scaling
the 105 mA H" would require I4X = 83A of discharge
current,
as
given
by
I4X(i4X= 105mA)
=
(105mA/250mA)*197A = 83A. These two rather
independent scalings indicate order 100 mA H" beam
may be available from the 4X slit source at 80A
discharge current for df4X = 5%. If such performance
could be demonstrated, the Penning 4X technology
could be considered for application to high-power
linac drivers [2].
After completion of the high-current slit beam
work on the 4X source in 1987, addition of 1-3% of N2
gas to the Penning H2 discharge was found to give
beneficial effects for quieting the extracted beam
current. This effect is shown in Fig. 5, where the
quieting effect of N2 gas in the Penning 4X discharge
is demonstrated. All source parameters were the same
in Figs. 5(A) and 5(B), except that approximately 1%
TABLE 3. Experimental results for the application of the
scaling laws to the 4X H" Penning SPS using slit emitters.
No slit aspect ratio study has been made on the 4X source to
optimize H" beam brightness.
ION SOURCE
PARAMETER
IX
ACTU
AL
4.3-5
2-3
12
2200
0.5
4X
DESI
GN
17
12
16
550
2-3
Disch. voltage (V)
I (A)
i (mA)
JH- (mA/cm2)
F (kW/cm2)
$ (mA/kW)
100
180
160
3200
16.7
192
100
240
160
800
4.2
192
Emission slit (y,x)
(mm)
Emittance, rms norm
(Timm-mrad)
10, 0.5
10,
2.0
L (mm)
W (mm)
T(mm)
B-field (G)
Duty factor (%)
0.17yO
.06x
4X
ACTU
AL
17
17
16
460
0.5
meas
93
197
250
783
2.24
349
11.4,
2.8
0.29y
0.15x
4X
DISCH.
ONLY
17
17
16
500
6%
115
125
Not meas
Not meas
1.76
Not meas
There is an approximate factor of 6 increase in
(inxXdfnx ) when the 4X (i4Xdf4X = 525mA%) is
compared to the IX source (iixdfix = 80mA%). A
factor of 4 increase may arise from the decrease in F4X
when compared to Fix. The other unexpected factor of
2 may arise from the increased H" production
efficiency of the 4X compared to the IX. The
enhancement factors are apparently additive. One may
conclude that this empirical observation of times 6
enhancement of (i4X)(df4X) compared to (iix)(dfix)
represents a confirmation of the original 4X
technology hypothesis: reducing the pulsed cathode
temperature rise would allow extended df operation of
the H" Penning SPS.
Assuming ion source development confirms the
Penning H" discharge scaling law assertions contained
in this paper and in the earlier work[3], substantial
issues for a 100-mA H" source and injector remain.
The source must be consistent with reliability and
lifetime requirements of the total accelerator facility.
Given the I4X = 80A for 100-mA H" is realized for the
4X source and assuming V4X = 100V, the cathode
power
loading
F4X(i4X=100mA)
=
(100V)(80A)/(3*CA4X) = 0.99kW/cm2, where CA4X =
W4X*T4X = 2.7cm2 (cf Table 2.). This scaled F4X is
Not meas.
N2 gas was introduced in the discharge of Fig. 5(B).
The N2 effects were mostly studied for circular
aperture beams[3], and the effect on the beam noise
JME (W&
FIGURE 5. Photos of beam extracted from a circular
aperture Penning 4X source. The discharge parameters are
identical, except for the addition of 1% N2 gas in (B).
and phase space of H" beams extracted from slit
emitters remains largely unknown.
Emission
spectroscopy measurements[19] showed that the Cs+
and Mo+ ion densities were increased when the 4X
Penning discharge was operated with small (1%) N2
concentrations.
291
9. H. S. Carslaw and J. C. Jaeger, Conduction of Heat
in Solids, Second Edition, Oxford University Press
(1959), 401.
10. P. E. Gear and R Sidlow, Proc. of the Second
International Conference on Low-Energy Ion Beams,
Institute of Physics Conf. Series No 54(1980), 284.
11. R Sidlow and N D West, "Performance of the
Penning H- Ion Source at ISIS", in the Proc. of the
Third European Particle Accelerator Conference,
Berlin, Editions Frontieres, (March, 1992), 1005.
12. R. Sidlow, et. al., "Operational Experience of
Penning H" Ion Sources at ISIS", in the Proc. of the
Fifth European Particle Accelerator Conference,
Sitges (Barcelona), Institiute of Physics Publishing,
(June, 1996), 1525.
13. Paul W. Allison, "A Direct Extraction H" Ion
Source", IEEE Transactions on Nuclear Science, Vol.
NS-24, No. 3 (June 1977), 1594.
14. Paul Allison and Joseph D. Sherman, "Operating
Experience with a 100-keV, 100-mA H" Injector", in
the Proc. of the Third International Symposium on the
Production and Neutralization of Negative Hydrogen
Ions and Beams, Brookhaven New York, AIP
Conference Proceedings No. Ill, (1983), 511.
15. P. W. Allison, et. al., "H" Ion Source Research at
Los Alamos", in the Proc. of the Second International
Symposium on the Production and Neutralization of
Negative Hydrogen Ions and Beams, Brookhaven
New York, BNL 51304, (Oct., 1980), 171.
16. Paul W. Allison, et. al., "An Emittance Scanner for
Intense Low-Energy Ion Beams", IEEE Transactions
on Nuclear Science, Vol. NS-30, No. 4, 2204 (1983).
17. Paul Allison, et al., "Comparison of Measured
Emittance of an H" Ion Beam with a Simple Theory",
Los Alamos Technical Report, LA-8808-MS (1981).
18. H. Vernon Smith, Jr., et. al., "H" Ion source with
High Duty Factor", in Proc. of the 1987 IEEE Particle
Accelerator Conf., (March, 1987), 301.
19. H. Vernon Smith, Jr., "Emission Spectroscopy of
the 4X Source Discharge with and without N2 Gas",
AT-10 Technical Note: 89-07, (April, 1989).
20. J. W. G. Thomason and R. Sidlow, "ISIS Ion
Source Operational Experience", in the Proc. of the
7th European Particle Accelerator Conference,
Vienna, Austria, Austrian Academy of Sciences Press,
(2000), 1625.
21. D. G. Dzahbarrov and A. P. Naida, "Motion of
Positive Ions in the Plasma Produced by a Dense
Beam of Negative Ions in a Gas", Sov. J. Plasma Phys.
6(3) (1980), 316.
already lower than the cathode power loads shown in
Fig. 1. The IX Penning SPS H" source is continuing to
demonstrate reliable operation on the ISIS facility at
RAL[20], with Fix in the range of 4kW/cm2, so there
is hope the 4X would offer improved lifetime, at least
from the viewpoint of cathode erosion. If a magnetic
focusing, beam space-charge neutralized LEBT were
chosen for RFQ matching of a 100-mA H" beam, little
emittance growth would be allowed to maintain an
overall 0.3 (Tcmm-mrad) injector
emittance
specification. The extracted ion beam energy would
have to be carefully chosen to minimize emittance
growth in the transport elements, emittance growth
from beam plasma effects[ 14,21], and then satisfying
minimal emittance growth in the extraction process.
The beam plasma effects can be controlled by
operating at higher H" beam energies in the LEBT,
introduction of xenon gas in the LEBT, and keeping
the LEBT length short[14].
REFERENCES
1. G. I. Dimov, "Use of Hydrogen Negative Ions in
Particle Accelerators", Rev. Sci. Instrum. 67,
(October, 1996), 3393.
2. R. Duperrier, et. al., "The ESS Front End
Associated with the SC Linac", ESS TAG Meeting n°
1, FZ Juelich, January 7-9, 2002.
3. H. Vernon Smith, Jr., et. al., "H" and D" Scaling
Laws for Penning Surface-Plasma Sources", Rev. Sci.
Instrum. 65, (January, 1994), 123.
4. C. W. Planner, "Matching Unequal Transverse
Emittances from an H" Ion Source into a RFQ",
Particle Accelerators 48, (1995), 243.
5. O. R. Sander, et. al., "Transverse Emittance of a
2.0-MeV RFQ Beam with High Brightness", in the
Proc. of the 1985 Particle Accelerator Conference,
Vancouver, BC, Canada, IEEE Trans. on Nuclear Sci.,
(Oct., 1985), 2588.
6. Yu I. Belchenko, et. al., "Physical Principles of the
Surface Plasma Method for Producing Beams of
Negative Ions", in the Proc. of the Symposium on the
Production and Neutralization of Negative Hydrogen
Ions and Beams, Brookhaven New York, BNL 50727,
(1977), 79.
7. A. von Engel, Ionized Gases, Second Edition,
Oxford University Press (1965), 288.
8. G. E. Derevyankin and Vadim Dudnikov,
"Production of High-Brightness H" Beams in Surface
Plasma Sources", in the Proc. of the Third
International Symposium on the Production and
Neutralization of Negative Hydrogen Ions and Beams,
Brookhaven New York, AIP Conference Proceedings
No. Ill, (1983), 376.
292