«•
(.-*••
i>t(*ri( ir
at
ili.t
-
HEAVY ION SOURCE SUTIORT GAS MIXINl", EXPERIMENTS
E.
D. 1ludson~'
Oak Ridge National Laboratory,* Oak Ridge, Tennessee 37830
and
M. I. Malloiy
Michigan State University,^ East Lansing, Michigan 48S24
Experiments on mixing an easily ionized support
gas with the prir.ary ion source gas have produced
large beam enhancements tor high charge state light
ior.s (masses
<.
20).
In the Cak Ridqe Isochronous
Cyclotron
(OR1C),
the beam increase has been a factor
of S or greater, depending on ion species and charge
state.
Approximately 0.1 cc/niin of the easily ionized
support gas (argon, krypton, or xenon) is supplied to
the ion source through a separate gas line and the
primary gas flow is reduced by i.
30°°.
The proposed
mechanism for increased intensity is as follows: The
heavier support gas ionizes readily to a higher charge
state,
providing increased cathode heating. The
increased heating permits a reduction in primary gas
flow (lower pressure) and the subsequent beam increase.
Introduction
Improvements in the perfurmance characteristics
of heavy ion sources have a direct impact upon the
performance of heavy ion accelerators. Therefore, a
large effort on positive ion source improvement has
been carried out at ORNL in connection with the Oak
Ridge Isochronous Cyclotron. In the past, mixing
gases in the ion source had been viewei as resulting
in poorer source performance, since the proportion of
ions available for ionization would be diluted by the
mixing ratio. In particular, this effect is obviously
seen in enriched isotope gases (e.g., the intensity of
la
0 versus enrichment
factor).
Another observation is that the quantity of gas
needed to support the arc varies with the element
being ionized, namely, for protons the source requires
a large gas flow, whereas for xenon the source requires'
a small gas flow. Assuming that the energy (E) to
heat the cold cathodes to the thermionic emission
"imit is the sane for all gases, then the following
relation can be written:
E «
V,
(15
where n. is the number of ions required for the cathodes
to reach the thermionic emission limit and is related
to the source gas flow, q is the average charge in the
plasma,
and V is the- arc voltage (the potential the
ions fall through in bombarding the
cathodes).
For
hydrogen, q can be only <. 1. As the ion source gas
mass increases, the q charge increases since the
ioni-
zation energy decreases for ions of the same charge
state.
1
Effects in the source and near the cyclotron
central region that are pressure dependent may then
vary as different gases are used, since less gas flow
is needed for the heavy mass gases.
•Operated by Union Carbide Corp. for the U.S. ERDA.
tKork supported by the National Science Foundation.
Another ion source characteristic that is knoun
but not understood is that different cold cathode ion
source geometries require different amounts of pas
flow for source operation. For cxar.ple, the present
ion source of Oak Ridge
2
requires i 3-<! cc/snin gas
flow for normal operation, whereas the ion source of
Michigan State University and some other laboratories
require a flow of 0.5 - 1 cc/min.
5
One would tl.en
expect central region pressure effects and ion source
gas usage effects would be more readily detectable in
the large gas flow source of Oak Ridge.
Gas Mixing Experiments
Gas mixing experiments have been performed and
large beam enhancements have been detected at OR If.
The experimental arrangement for gas mixing is «clv-
matically shown in Figure 1. The Rases are fed to
the ion source through separate gas lines and are
mixed in the plasma chamber. Experiments of irixirg
the gases external to the ion source and feeding
ORNL-DWG 76- U8M
,-CATHODE
GAS LINE
To INSERT
GAS LINE
ACCELERATING
ELECTRODE
PLASMA CHAMBER
CATHODE
Fig.
1. A schematic view of the cold cathode ion
source showing the gas inlet lines. In the gas mix-
ing mode the primary arc gas is fed throur.li one line
and the support gas through the other. In mixing
xenon with neon, visible observation of the arc
through the extraction slit shows a two-colored arc,
one side green (xenon) and the other side pink
(neon).
OISTRIBUTIUN Ofc THIS DUCLflWtNT IS UNLIMITED
i
'
:j'h
'>i!r ,;.is
liiif I..1V1.' j>iot!:irtil substantially
the
c
n-
ill'
.
i'..ii
iMxint;
of
'rryj'tosi wit!) ncciii
for an
'r.ir-tnl It.i.-i of ^'.Nc
1
'* at Hi3 MeV is shovv-n in Fig. 2.
ORNL-OWG 76-(1905
0.2 0.4 0.6 0.8 1.0
KRYPTON ARC SUPPORT GAS (cmVmin)
Fig.
2. The
2O
.\'e
6+
extracted beam intensity versus
the support gas flow
(krypton).
For each point a
small amount of krypton was supplied to the arc then
the primary gas (neon) was decreased until the arc
voltage started to increase, an indication that the
arc was about to drop out. A maximum beam enhance-
ment occurred at 0.15 cc/min. of krypton.
As krypton gas was added to the ion source, the neon
gas flow was decreased until the arc voltage started
to rise. The neon beam intensity increased by approxi-
mately a factor of five for a krypton gas flow of 0.15
cc/min.
A further increase in krypton flow resulted
in a decrease in the neon beam intensity. In Fig. 3
the cvclotron internal beam intensity with radius
2
of
a '"[« * beam is shown for xenon gas mixing. The in-
ternal beam attenuation slope for gas mixing and
without gas mixing is the same, and indicates that
the internal pressure in the cyclotron is the same
and docs not account for the increase in beam inten-
sity that occurs with gas mixing. In Fig. 4 the
internal cyclotron beam attenuation is shown for a
'"NC
6
*
beam. The large step in the beam attenuation
at 63 cm is due to the 3rd harmonic boaw of *°Nc
2
*
falling out of phase with the 1st harmonic Ne • beam.
7 j r 3
'«~-'
:
: , : jT o
1000 2000
3000 4000 5000 6000
PHOBE RADIUS
2
lcm
?
)
Fig.
3. The '"N
5
* beam attenuation with radius as
measured on a cyclotron probe, with and without
xenon support gas. The slopes of the curves are
equal,
indicating equal pressure attenuation in the
cyclotron.
1—1—1
1
j
1 j
1
:
'.
1000 2000 3000 4000 5000 6000
PROBE RADIUS
2
(cm]
2
extracted beam, with and without krypton support pas.
The step at 63 cm is due to the acceleration cf the
20
Ne
2
*
harmonic beam. The intensity of the
2C
Ne
2t
decreases with krypton support gas, and if the
pressure were improving it would be expected to
increase.
The intensity of the
20
Ne
8+
beam with gas mixing and
and without gas mixing indicates that the charge distri-
bution in the source plasma is shifting to the higher
charge states, since a decrease in pressure" would be
expected to produce less attenuation of the
2c
Ne
2
*
beam.
The effects of different support gases (argon,
krypton, and xenon) are shown in 1
:
iK. 5. The maximum
t
p.i'ii
improvement obtained is about the same for the
three gases; the q of these gases are approximately
c^iial. ^Xenon and krypton gas mixing experiments with
an
l<0
Ar'
beam has not shown an increase in intensity
with gas mixing.
ORNL-OWG 76-M903
A
UK'
f
4
}
\
\
1 V
j
1
63 Me
e
6
*
SUPPORT CAS
• XENON
• KRYPTON
o AR
\
GOM
\
0 O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 (.0
SUPPORT GAS (cm
3
/mm)
Fig.
5. The normalized beam enhancement of
20
Ne
6+
for
different support gases (xenon, krypton, and
argon).
The maxinum enhancement appears to be approximately
the same for all gases.
60 80
TIME Imin)
Fig.
6. The extracted beam intensity for an
u
0
5
*
beam versus time. The intensity of oxygen without
support gas tends to increase slowly over a period
of hours. The beam intensity with xenon support
gas reaches even a higher level in a few minutes,
and is more stable.
Acknowledgement s
We thank M. L. Halbert for his initial interest
that brought about this research and J. P. Rnll for
his continued support and help in acquirinf, the data.
We would also like to thank the ORIC operations crf»
for their excellent running of the cyclotron while we
collected the data.
In Fig. 6, the extracted beam intensity of an
l6
0
5+
beam is recorded as a function of time with and
without mixing. The increase in the beam intensity at
40 minutes was obtained by adjusting the xenon gas
flow. The flata indicate the large beam intensity for
the oxygen beam is obtained from the beginning of the
ion source lifetime. An important side effect is that
the xenon gas mixing improves markedly the stability
of the oxygen beam performance of the ion source.
Gas mixing is now routinely used at ORIC. Gas
mixing enhancement has also been observed with an
16
0
$+
beam at Michigan State. Additional gas mixing
experiments are in progress.
References
1. T.A. Carlson, et al., At. Data 2^ (1970) 63.
2.
E.D. Hudson, et al., IEEE Trans.
Nucl.
Sci.
NS-23,
No. 2 (1976) 175.
3. M.L. Mallory, et al., Proceedings thi? conference.
4.
M.L. Mallory, et al.,
Nucl.
Instr. and Meth. US
(1976) 29.
5.
J.R.J.
Bennett, IEEE Trans.
Nucl.
Sci.
NS-19,
No.
2 (1972) 48.
