NIM B

Beam Interactions
with Materials & Atoms
Nuclear Instruments and Methods in Physics Research B 261 (2007) 588–593
[Link]/locate/nimb

Computer simulation of MC-SNICS for performance improvements

a,*
B.X. Han , J.R. Southon b, M.L. Roberts a, K.F. von Reden a

a
National Ocean Sciences AMS Facility, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA
b
Keck Carbon Cycle AMS Facility, University of California, Irvine, Irvine, CA 92697, USA

Available online 2 April 2007

Abstract

Increasing ion source current output while keeping the beam emittance small is a key for improving measurement throughput and
quality of an accelerator mass spectrometry (AMS) system. The NEC MC-SNICS ion sources have widely been used for AMS around
the world. In this work, a computer simulation study on MC-SNICS ion source was carried out using the code PBGUNS. The primary
Cs+ beam is simulated as space-charge limited emission from the ionizer surface. The secondary negative ion beam is simulated in the
presence of the space-charge induced by both the primary and the secondary beam. Two different configurations of the MC-SNICS have
been studied: the original factory version and the UC Irvine modified version. The latter has a 0.668 in. radius spherical ionizer and an
immersion lens at cathode potential with the original Cs focus lens removed. Simulation showed that the modified version provides more
intense and better focused Cs+ beam at the cathode target, hence higher negative ion production from the sample by sputtering. The new
arrangement of Einzel-preacceleration section of the MC-SNICS at UC Irvine was also simulated and discussed. Simulation results are in
good agreement with the experimental observations made at UC Irvine. Hints for selection of an immersion lens geometry have been
obtained.
 2007 Elsevier B.V. All rights reserved.

PACS: [Link]; 41.75.i; 41.85

Keywords: Sputter ion sources; MC-SNICS; Computer simulation; Accelerator mass spectrometry

1. Introduction by several AMS labs and NEC on ion source performance

improvements. Specifically, the Keck Carbon Cycle AMS
Accelerator mass spectrometry (AMS) has been an (KCCAMS) facility at UC Irvine has made progress in
ultra-sensitive, high-precision and efficient technique for increasing output and improving reliability and serviceabil-
the measurement of long-lived radioisotopes since its ity of the MC-SNICS [4]. In this work, putting the empha-
invention in the late 1970s [1]. The negative ion source is sis on the ion beam production and transport issues, a
an essential part in an AMS system, and to date, AMS computer simulation study on the MC-SNICS ion source
ion sources are almost exclusively Cs sputter ion sources. has been conducted at the National Ocean Sciences AMS
The National Electrostatics Corporation (NEC) multi- (NOSAMS) facility in collaboration with the KCCAMS
cathode source of negative ions by cesium sputtering to help understand and further develop the MC-SNICS.
(MC-SNICS) [2,3] has widely been used for AMS around The original factory version and the UC Irvine modified
the world. To achieve higher throughput and better quality version of the MC-SNICS source were simulated and com-
of AMS measurements, a collaborative effort is being made pared. The new arrangement of Einzel-preacceleration sec-
tion of the UC Irvine modified MC-SNICS was also
*
discussed. Poisson code PBGUNS [5] was chosen for this
Corresponding author. Present address: Oak Ridge National Labora-
tory, P.O. Box 2008, MS 6461, Oak Ridge, TN 37831, USA. Tel.: +1 865
simulation work due to its capabilities of handling space-
414 6437; fax: +1 865 241 9831. charge effects, corrected Child’s Law constraint for curved
E-mail address: hanb@[Link] (B.X. Han). surfaces, treating secondary sputtered beam in the influence

0168-583X/$ - see front matter  2007 Elsevier B.V. All rights reserved.
doi:10.1016/[Link].2007.03.060
 B.X. Han et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 588–593 589

Fig. 1. The configuration of a NEC original factory version MC-SNICS ion source.

of primary sputtering beam, and extending a fine-mesh crit- ciency which correlates inversely with the work function of
ical region simulation for larger scale transport section by the sample surface, and exp(n0Lrd) is a factor by which
saving and restoring beam data. the negative ion current is decreased due to the electron
detachment collisions with the residual gases whose particle
2. Fundamentals and simulation approaches density is denoted as n0. It can be seen that higher Cs+ cur-
rent onto the sample target, better vacuum, and lower work
In a Cs sputter ion source, the secondary negative ion function of the sample surface all contribute to obtaining
emission from the sample surface by the Cs+ beam sputter- higher negative ion output. Assuming other conditions
ing is a complicated process which is affected by many are unchanged, negative ion current increases proportion-
experimental factors and conditions. According to Ishik- ally to the increase of Cs+ current at the sample surface.
awa [6], the negative ion current, I, generated at the sam- Using the code PBGUNS, the maximum emission current
ple surface can be expressed by the following equation. of the Cs+ beam from the ionizer surface could be esti-
mated as a space-charge limited (Child’s Law with Lang-
I  ¼ I þ Ag expðn0 Lrd Þ; muir–Blodgett corrections for curved surfaces) emission
where, I+ is the primary Cs+ beam current for sputtering, A for a given configuration of a sputter source, and the trajec-
is the sputtering yield, g is the negative ion formation effi- tories of Cs+ ions are calculated by taking space-charge

Fig. 2.1. Cs+ beam from a conical ionizer at low focus voltage. Fig. 2.2. Cs+ beam from a conical ionizer at high focus voltage.
590 B.X. Han et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 588–593

Fig. 2.3. Simulation of the NEC factory version MC-SNICS for negative secondary beam emission and extraction.

effects into account. The approach used in PBGUNS for 3. Simulation of negative ion generation and extraction
simulating the sputtered negative ion emission is to assume
that the current density distribution of the emitted 3.1. NEC original factory version MC-SNICS
secondary negative ions follows the current density distri-
bution pattern of the primary sputtering beam at the sam- Fig. 1 shows a cross sectional drawing of a NEC factory
ple surface. Angular spread can also be added to the version MC-SNICS ion source. This configuration is
secondary ion emission. The total value of the negative equipped with a conical shape ionizer, a Cs+ focus elec-
ion current can be set by the user in accordance with exper- trode and a snout-type ‘‘extractor’’ inside the source body
imental measurement. Trajectories of the secondary and is usually followed by an Einzel-preacceleration
negative ions are calculated in the presence of the space- section.
charge induced by both the primary and the secondary To investigate the primary Cs+ beam emission and
beams. focusing feature of this configuration, the cathode was set

Fig. 3.1. Cs+ beam from a spherical ionizer with a plate shape immersion Fig. 3.2. Cs+ beam from a spherical ionizer with a snout shape immersion
lens. lens.
 B.X. Han et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 588–593 591

at a relatively high voltage of 8.0 kV (typical of high out- higher focus voltage results in increased Cs+ emission cur-
put operation), and the Cs+ focus voltage was varied in the rent from the ionizer surface, the useful Cs+ current at the
simulations. As indicated in Figs. 2.1 and 2.2, while the recessed sample surface (1 mm diameter) does not increase

Fig. 3.3. Simulation of the UC Irvine modified version MC-SNICS for negative secondary beam emission and extraction (plate shape immersion lens).

Fig. 3.4. Simulation of the UC Irvine modified version MC-SNICS for negative secondary beam emission and extraction (snout shape immersion lens).

Fig. 4.1. Extended simulation of the C beam within the UC Irvine modified Einzel-preacceleration section.
592 B.X. Han et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 588–593

proportionally. Most of the Cs+ ions are wasted on the

sample holder due to insufficient focusing.
By storing the space charge distribution of the primary
Cs+ beam in the simulation matrix, the negative beam
emission and extraction can be simulated in the presence
of space charge from the primary beam. Shown in
Fig. 2.3 is a simulation output of the negative carbon beam
as consequence of the simulation in Fig. 2.2. The C cur-
rent was set to 100 lA, which was reported by Roberts
et al. using similar settings [7], and three different emission
angles with initial energies of 25 eV were applied in the sim-
ulation. It is seen that the beam almost completely fills the
6.5 mm diameter ionizer hole. The calculated normalized
RMS emittance of the beam after extraction (at
Z = 80 mm) is about 1.1 · 102 p mm mrad.

3.2. UC Irvine modified version MC-SNICS

Since acquiring of a MC-SNICS ion source along with a

Fig. 4.2. Emittance plot at the location (Z = 900 mm) of beam waist.
compact AMS system in mid 2002, the KCCAMS facility
has made many modifications to the factory version of
MC-SNICS for higher output and improved reliability was carried out using the UC Irvine new arrangement of
and serviceability. The development work at UC Irvine the Einzel-preacceleration section. The output is shown in
includes the use of a spherical ionizer, replacement of the Fig. 4.1. After passing through the Einzel-preacceleration
original Cs focus electrode with an immersion lens, open- section, the beam is quite parallel and the beam radii given
ing-up of the snout shape extractor hole, and a new at the two positions (Z = 710 mm and 1010 mm), where
arrangement of the Einzel-preacceleration section by the beam profile monitors (BPM) are placed in KCCAMS
replacing the NEC acceleration tube with a gap lens etc [4]. system, are both 4 mm. These are close to the experimen-
Currently, KCCAMS facility operates the modified MC- tal data of 6 mm and 4 mm measured at KCCAMS.
SNICS with a 0.668 in. radius spherical ionizer and is test- Fig. 4.2 is an emittance plot at the beam waist (the Twiss
ing two different shapes of an immersion lens: simple plate parameter a = 0) position which is at about Z = 900 mm in
shape and snout shape, respectively. A negative carbon the Fig. 4.1. The calculated normalized RMS emittance of
beam current of 150 lA is routinely obtained using either the beam at this position is 0.84 · 102 p m mrad. This
of the immersion lenses. The simulations of the primary value translates to a lab emittance value of 10 p mm mrad
Cs+ beam and the secondary C beam are shown in Figs. for the 55.5 keV carbon beam and corresponds to the area
3.1 and 3.3 for the simple plate immersion lens configura- of the dashed-line ellipse generated by the code in
tion and in Figs. 3.2 and 3.4 for the snout shape immersion Fig. 4.2. In AMS measurements, it is always desired to
lens. Both configurations achieve very good Cs+ beam spot transport the beam through the system without loss. In this
focused at the 1 mm diameter sample surface and the sense, taking the area of the best circumscribing ellipse
150 lA intense C beams are extracted cleanly through drawn by the authors with solid line in Fig. 4.2 as a working
the 4.8 mm diameter ionizer hole. The calculated normal- emittance estimate might be of more interest than using the
ized RMS emittances of the extracted C beams (at RMS calculation value to the AMS community. In Fig. 4.2,
Z = 80 mm) in the two configurations are 0.92 · 102 the area of the solid-line ellipse is  40 p mm mrad. This
p mm mrad and 0.54 · 102 p mm mrad, respectively, and value agrees well with an experimental measurement done
both are smaller than that of the original factory version at KCCAMS using two BPMs and by focusing the beam
MC-SNICS. From the point of standardizing the lens at the second BPM location [8].
geometry with keeping the potential for future source
improvement in other aspects, the snout shape immersion 5. Summary
lens is preferable in that it can produce more intense Cs+
beam and the negative beam emittance is also smaller in this The original factory version and the UC Irvine modified
configuration. version of the NEC MC-SNICS ion source were simulated
and compared. The latter has the capability of providing
4. Simulation of Einzel-preacceleration section intense and better focused Cs+ beam at the sample cathode,
hence producing higher current output of secondary nega-
To make direct comparison between the simulation tive beams. Simulations also showed that even with the
results and the experimental observations, an extended more intense current output, the UC Irvine modified source
simulation from the beam data at Z = 80 mm in Fig. 3.4 has smaller emittance than the original factory version.
 B.X. Han et al. / Nucl. Instr. and Meth. in Phys. Res. B 261 (2007) 588–593 593

While the two different geometries of the immersion lens (134 MC-SNICS), in: Proceedings of the 15th International Conference
used in the UC Irvine source both function satisfactorily, on the Applications of Accelerators in Research and Industry, AIP
Conference Proceedings 475 (1999) p. 640.
the snout shape is preferred for future improvement of [4] J. Southon, G.M. Santos, Life with MC-SNICS. Part II: Further ion
the ion source in other aspects. Simulation results are in source development at the Keck carbon cycle AMS facility, Nucl.
good agreement with the experimental observations made Instr. and Meth. B, in press doi:10.1016/[Link].2007.01.147.
with BPMs at the KCCAMS facility. [5] PBGUNS 5.04, Thunderbird Simulations, <[Link]
[Link]/>.
[6] J. Ishikawa, Negative ion sources, in: I.G. Brown (Ed.), The Physics
References and Technology of Ion Sources, Wiley-VCH Verlag GmbH, Wein-
heim, 2004, p. 285.
[1] L.K. Fifield, Accelerator mass spectrometry, in: R. Hellborg (Ed.), [7] M.L. Roberts, R.A. Culp, D.K. Dvoracek, G.W.L. Hodgins, M.P.
Electrostatic Accelerators, Springer, Herdelborg, Berlin, New York, Neary, J.E. Noakes, The 14C AMS system at The University of
2005, p. 461. Georgia, Nucl. Instr. and Meth. B 223–224 (2004) 1.
[2] G.A. Norton, R.L. Loger, J.E. Raatz, R.D. Rathmell, Multi-cathode [8] John Southon, Guaciara dos Santos, Baoxi Han, Cs feed tests and
SNICS ion source, in: Proceedings of the 25th Symposium of North emittance measurements on a modified MC-SNICS ion source for
Eastern Accelerator Personnel, Santa Fe, NM, USA, October 16–19, radiocarbon AMS, presented at the 19th International Radiocarbon
1991, World Scientific, Singapore, 1992, p. 295. Conference, Radiocarbon, submitted for publication.
[3] R.L. Loger, J.E. Raatz, M.L. Sundquist, R.C. Garner and J. Barker,
134 Sample multi-cathode source of negative ions by cesium sputtering