close

Вход

Забыли?

вход по аккаунту

?

j.nima.2017.10.003

код для вставкиСкачать
Accepted Manuscript
A versatile triple radiofrequency quadrupole system for cooling, mass
separation and bunching of exotic nuclei
Emma Haettner, Wolfgang R. Plaß, Ulrich Czok, Timo Dickel, Hans Geissel,
Wadim Kinsel, Martin Petrick, Thorsten Schäfer, Christoph Scheidenberger
PII:
DOI:
Reference:
S0168-9002(17)31040-9
https://doi.org/10.1016/j.nima.2017.10.003
NIMA 60149
To appear in:
Nuclear Inst. and Methods in Physics Research, A
Received date : 10 August 2017
Revised date : 26 September 2017
Accepted date : 1 October 2017
Please cite this article as: E. Haettner, W.R. Plaß. W.R. Plaß, U. Czok, T. Dickel, H. Geissel, W.
Kinsel, M. Petrick, T. Schäfer, C. Scheidenberger, A versatile triple radiofrequency quadrupole
system for cooling, mass separation and bunching of exotic nuclei, Nuclear Inst. and Methods in
Physics Research, A (2017), https://doi.org/10.1016/j.nima.2017.10.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
*Manuscript
Click here to view linked References
A Versatile Triple Radiofrequency Quadrupole System
for Cooling, Mass Separation and Bunching of Exotic
Nuclei
Emma Haettnera,b , Wolfgang R. Plaßa,b,∗, Ulrich Czoka,b,1 , Timo Dickela,b ,
Hans Geissela,b , Wadim Kinsela,b , Martin Petricka , Thorsten Schäfera ,
Christoph Scheidenbergera,b
a II.
Physikalisches Institut, Justus-Liebig-Universität Gießen, 35392 Gießen, Germany
Helmholtzzentrum für Schwerionenforschung GmbH, 64291 Darmstadt, Germany
b GSI
Abstract
The combination of in-flight separation with a gas-filled stopping cell has opened
a new field for experiments with exotic nuclei. For instance, at the SHIP/SHIPTRAP
facility at GSI in Darmstadt high-precision mass measurements of rare nuclei
have been successfully performed. In order to extend the reach of SHIPTRAP
to exotic nuclei that are produced together with high rates of unwanted reaction
products, a novel compact radio frequency quadrupole (RFQ) system has been
developed. It implements ion cooling, identification and separation according
to mass numbers and bunching capabilities. The system has a total length of
one meter only and consists of an RFQ cooler, an RFQ mass filter and an RFQ
buncher. A mass resolving power (FWHM) of 240 at a transmission efficiency
of 90% has been achieved. The suppression of contaminants from neighboring masses by more than four orders of magnitude has been demonstrated at
rates exceeding 106 ions/s. A longitudinal emittance of 0.45 eVµs has been
achieved with the RFQ buncher, which will enable improved time-of-flight mass
spectrometry downstream of the device. With this triple RFQ system the measurement of e.g. N =Z nuclides in the region up to tin will become possible at
SHIPTRAP. The technology is also well suited for other rare-isotope facilities
∗ Corresponding
author
Email address: Wolfgang.R.Plass@exp2.physik.uni-giessen.de (Wolfgang R. Plaß)
1 Deceased.
Preprint submitted to NIM A
September 26, 2017
with experimental setups behind a stopping cell, such as the fragment separator
FRS with the FRS Ion Catcher at GSI.
Keywords: Radiofrequency quadrupole, RFQ mass filter, RFQ buncher,
Beam purification, Ion cooling
1
1. Introduction
2
Fusion-evaporation reactions are an efficient method for the production of
3
super-heavy nuclei as well as medium-heavy and heavy neutron-deficient nuclei.
4
These reactions thus enable the synthesis and study of the heaviest elements
5
[1], but also open a rich field for experiments at the proton drip line. The
6
latter include the study of the limits of nuclear existence of neutron-deficient
7
nuclei, proton [2] and α radioactivity [3], the doubly-magic nucleus
8
the
9
energy in N =Z nuclei [6, 7], and the rp and νp nucleosynthesis processes [8, 9].
10
Measurements with the Penning trap mass spectrometer SHIPTRAP [10],
11
which is installed at the exit of the velocity filter SHIP [1, 11] at GSI, have
12
made important contributions in this field. The first direct mass measurements
13
of transuranium elements have been made with SHIPTRAP [12, 13]. The first
14
Penning trap mass measurements beyond the proton drip line have been per-
15
formed [14], and the masses of N =Z+1 nuclei in the region of the rp process
16
have been measured [15].
94m
100
Sn [4],
Ag high-spin isomer [5], isospin symmetry, mirror nuclei, and Wigner
17
At the present SHIP/SHIPTRAP setup (Fig. 1a) the nuclides of interest are
18
produced in fusion-evaporation reactions, separated in-flight from the primary
19
beam in the velocity filter SHIP and stopped in a He-filled stopping cell [19]. In
20
the stopping cell, the thermalized ions are guided by DC and RF fields towards
21
a nozzle and extracted by the gas flow into an RF quadrupole (RFQ). In the
22
subsequent RFQ buncher [16] the ions are further cooled and bunched. The
23
ion packets are extracted into an electrostatic beam line and transferred to the
24
first of the two Penning traps installed in the same superconducting 7 T dipole
25
magnet. In the first Penning trap, the ions are cooled in a mass-selective process
2
26
with a mass resolving power high enough to separate isobars [20]. In the second
27
Penning trap, high-accuracy mass measurement are performed on isobarically
28
clean samples using the TOF-ICR [21, 22] or PI-ICR [23] techniques.
29
So far, SHIPTRAP mass measurements have been made in the region of
30
the transuranium elements on nuclides with production cross section as small
31
as 60 nb [13]. The production cross sections of medium-heavy N =Z nuclides,
32
such as
33
two orders of magnitude. However, for heavy and medium-heavy proton-rich
34
nuclides the reach has so far been limited to nuclides with production cross sec-
35
tions on the order of 130 µb [15, 26, 27], i.e. to nuclides with cross sections that
36
are more than three orders of magnitude larger than those for transuranium nu-
37
clides. This apparent discrepancy is caused by two challenges for experiments
38
with lighter nuclei. Firstly, fast unambiguous particle identification is very dif-
39
ficult for nuclei at the low kinetic energy of fusion products. Thus the rate
40
optimization for a selected nuclide is most difficult, with the exception of reac-
41
tion products that decay by α particle emission. The second reason is that for
42
heavy and medium-heavy nuclei the production rate of unwanted reaction prod-
43
ucts is much higher (>> 103 /s) than that in the case of transuranium nuclei.
44
While the velocity filter SHIP yields a superb suppression of the primary beam,
45
it does not separate the different reaction products. Furthermore, molecular
46
contaminants and adducts may be ionized in the gas-filled stopping cell and be
47
extracted together with the thermalized reaction products, such that the overall
48
rate of ions behind the stopping cell is even higher. Consequently, measurements
49
of rare reaction products are hampered by space charge effects and ion losses
50
due to abundant contaminants in the RFQ buncher and in the Penning traps.
51
Substantial progress in experiments with neutron-deficient nuclei will thus be
52
possible with new separation and diagnostics capabilities.
80
Zr and
84
Mo, are about 10 µb [24, 25] and thus larger by more than
3
53
2. Choice of the separation and diagnostics technique
54
The most advantageous location to perform additional separation and diag-
55
nostics is between the extraction RFQ and the RFQ buncher. There the ions
56
can be cooled, such that the highest mass resolving power can be achieved. Also,
57
contaminants originating from the stopping cell can be removed in addition to
58
the unwanted reaction products. Finally, the formation of the ion packets in the
59
buncher can be performed without the deteriorating influence of space charge
60
effects.
61
The following requirements for an additional separator should be considered:
62
Assuming a total cross-section of 1 b, a primary beam intensity of 5 × 1012 , a
63
target thickness of 0.5 mg/cm2 , a transmission through SHIP of 50% and a
64
combined stopping and extraction efficiency of the stopping cell of 10% [28]
65
a total rate of reaction products of about 106 /s behind the stopping cell is
66
obtained. The reaction products are typically distributed over a range of masses
67
of about 10 u [29]. Most contaminants produced in the stopping cell will have
68
a different mass number than the nuclide of interest. Thus a mass separator
69
with a mass resolution of one mass unit can remove most of the contaminants
70
from the stopping cell and suppress the unwanted reaction products by about
71
an order of magnitude. A suppression of ions with neighboring masses numbers
72
by more than one order of magnitude and a rate capability of at least 106 ions/s
73
is required. Furthermore the separator should have a very high transmission
74
efficiency. It should be compatible with operation in the vicinity of a gas-filled
75
stopping cell, and it should fit into the space of the setup.
76
Different techniques are available for mass separation and diagnostics behind
77
the stopping cell: (i) a dipole magnet, (ii) an RFQ mass filter [30, 31], (iii) ion
78
isolation in a linear RF ion trap mass spectrometer [32, 33] using RF-DC apex
79
isolation or dipolar excitation with e.g. SWIFT waveforms, or (iv) operation
80
of the electrostatic beam line between buncher and Penning traps as a linear
81
time-of-flight mass spectrometer in combination with a fast deflector [34, 35].
82
All these methods can achieve unit mass resolution. In contrast to the other
4
83
methods, time-of-flight mass spectrometry is a non-scanning technique and thus
84
offers superior analysis speed and sensitivity. However, both the TOF technique
85
as well as the linear ion trap may suffer from space charge effects for more than
86
104 ions per cycle. At a typical cycle frequency of SHIPTRAP of about 1 Hz
87
this would significantly limit the rate capability. Only the dipole magnet and
88
the RFQ mass filter are compatible with the rate capability. The RFQ mass
89
filter can be operated at higher residual gas pressures than the dipole magnet,
90
is compact and can be integrated easily into the available space with only minor
91
modifications of the existing beam line, while a dipole magnet would require
92
a new low-pressure section between the extraction RFQ and the RFQ buncher
93
and thus much more space and a re-arrangement of the setup. Furthermore
94
an RFQ mass filter can be operated both in a mass-selective mode as well as
95
in a broadband (RF-only) mode, while a dipole magnet always selects a small
96
momentum range. Therefore, an RFQ mass filter retains the option to analyze
97
the ions extracted from the stopping cell by the fast and sensitive time-of-flight
98
mass analysis in the electrostatic beam line. For these reasons, an RFQ mass
99
filter has been chosen as additional separation and diagnostics tool.
100
The preferred implementation is shown in Fig. 1b. A new system of three
101
matched RFQs, viz. an RFQ cooler, an RFQ mass filter and an RFQ buncher
102
replaces the presently installed RFQ buncher [16]. The RFQ cooler cools the
103
ions delivered from the extraction RFQ by collisions with the buffer gas. It is
104
necessary to inject a cold and well-aligned beam into the mass filter to ensure a
105
good performance of the RFQ mass filter. The RFQ buncher re-cools the ions
106
after mass separation and forms short ion packages with small emittance for
107
efficient transfer and capture into the Penning traps. A small longitudinal emit-
108
tance of the buncher is also of importance for good performance of additional
109
TOF mass spectrometry in the beam line. Since both cooler and buncher are
110
operated with buffer gas, pumping barriers between all three quadrupoles and
111
at the end of the buncher towards the drift section are required.
112
This solution follows a new concept for low-energy beam lines [36]. Here,
113
RFQs are used for ion transport at low kinetic energies (∼ eV) instead of conven5
114
tional electrostatic or magnetostatic beam lines. This enables compact setups
115
(∼ 1 m) that are ideally suited to the gas-filled environment (< 10−1 mbar) in
116
the differential pumping section in the vicinity of a stopping cell. They are easy
117
to tune and provide transport efficiencies close to unity. A variety of functional-
118
ities can be implemented in such an RFQ beam line: Ion transport and cooling,
119
formation of ion packets, ion separation and identification using an RFQ mass
120
filter, ion beam merging and distribution using curved RFQs [37, 38] or an RFQ
121
switchyard [39], as well as break-up of molecules [34, 40]. Solutions exist even
122
for the integration of gate valves, ion sources and detectors into the beam line
123
[41, 42]. Thus the RFQ beam line can be tailored to the specific conditions and
124
needs of the individual experiment.
125
This work demonstrates the performance of the cooling, mass separation and
126
bunching capabilities that can be implemented in an RFQ beam line. For an in-
127
depth description of the triple RFQ system, its operation and the performance
128
characteristics see [43].
129
3. Theory of RF quadrupoles
130
The theory of RFQ ion guides and RFQ mass filters has been covered in
131
several reviews [31, 44]. Here, only some basics relevant for the further discussion
132
will be summarized.
133
A linear RFQ consists of four parallel rods, which are arranged symmetrically
134
with respect to each other at a distance of r0 from the central axis of the RFQ.
135
A time-varying electric potential, ϕ0 (t), of the form
ϕ0 (t) = U − V cos(Ωt),
(1)
136
is applied to the rods, such that neighboring rods have potentials of opposite
137
signs. U is a DC voltage, and V and Ω are the amplitude and angular frequency
138
of an RF voltage, respectively. Provided the resulting field is sufficiently linear,
139
the stability of the motion through the RFQ does not depend on the initial
6
140
141
conditions but only on the Mathieu parameters
a=
4eU
mΩ2 r02
(2)
q=
2eV
mΩ2 r02
(3)
and
142
and can be represented in an a-q diagram, see for example [45]. Here, m and e are
143
the mass and the charge of the ion. A mass spectrum is obtained by gradually
144
varying the amplitudes of the voltages at a fixed ratio U/V and measuring the
145
abundance of the transmitted ions. The ratio U/V controls the mass resolving
146
power of the RFQ. For a = 0 the RFQ transports ions of a very broad range of
147
masses (RF-only mode). The latter mode of operation is applied in the RFQ
148
cooler and RFQ buncher.
149
While rods with hyperbolic cross section (theoretically) enable a purely linear
150
field and thus highest mass-resolving power, conventionally rods with round
151
cross section are used for RFQ mass filters. These are easier to machine and
152
lower the performance of the RFQ mass filter only slightly. For ion guides that
153
are not operated in a mass-selective way, rods with other cross sections can be
154
used as well.
155
4. Design and construction of the RFQ system
156
4.1. Design of the RFQ cooler and RFQ buncher
157
Both the RFQ cooler and the RFQ buncher use a buffer gas to cool the ions
158
and thus need an axial field to drag the ions through the RFQs. The axial field
159
can be produced in different ways. One possibility is to use LINAC rods [18], i.e.
160
four additional electrodes mounted with a small inclination between the RFQ
161
rods. LINAC rods have the advantages of a continuous potential gradient and
162
a simple electronic circuit.
163
The RFQ buncher accumulates the ions and thus, for ion trapping, must
164
include segmented RFQ rods with different DC potentials at its end to create an
165
axial potential well in addition to the radial RF pseudopotential well (Fig. 1b).
7
166
Two different positions of the trap have been considered in this work. The first
167
option is to store the ions at the segment S3 just in front of the exit diaphragm.
168
This has the advantage that a very simple electronic circuit can be used; the
169
trapping and extraction can be realized by switching the potential of just a
170
single diaphragm. The second option is to store the ions at the segment S2 in
171
the middle of the segments S1 and S3. Ion optically this option seems to be more
172
advantageous, since the ions are stored spatially well separated from the fringing
173
field of the exit diaphragm. Extraction of the cooled ions is realized by switching
174
the neighbouring segments to a higher and a lower potential, respectively.
175
At the moment, the cycle time of SHIPTRAP is about 1 s, while the cooling
176
time in the buncher is on the order of milliseconds. Hence only a very small
177
fraction of the ions accumulated in the RFQ buncher (< 1%) are not cooled
178
completely before extraction. However, as a future option it should be possible
179
to increase the cycle frequency. Then the number of ions, which have not yet
180
been sufficiently cooled at the time of extraction, increases. In such a case a
181
pre-trap, which stores newly incoming ions during the extraction phase of the
182
already cooled ions, would be beneficial and is foreseen in the design of the
183
RFQ system. Two options are available. One possibility is to use additional
184
segments in the trapping section of the buncher. This, however, leads to a
185
complex electronic circuit. Therefore the possibility to use the cooler RFQ as
186
a pre-trap is favored. In this case a potential minimum can easily be formed if
187
the potential of the diaphragm D1 (Fig. 1b) between the RFQ cooler and RFQ
188
mass filter is raised for the duration of the storage.
189
4.2. Simulations and design of the RFQ mass filter
190
In order to find the best compromise regarding pressures, apertures, lengths
191
and other dimensions of the mass filter, computer studies were carried out using
192
the programs SIMION [46], ITSIM [47, 48] and COMSOL [49]. The electric
193
potential produced by the electrode structure was calculated on a 2-dimensional
194
or 3-dimensional map using either SIMION or COMSOL. The calculation of ion
195
trajectories was performed with ITSIM using these potential distributions. In
8
196
some cases, an ideal quadrupolar field was calculated analytically in ITSIM
197
instead of using the numerical potential distribution. Unless noted otherwise,
198
the simulations were performed for
199
field radius of r0 = 6 mm and a length of 30 cm.
200
4.2.1. Length of the mass filter
133
Cs+ ions in an RFQ mass filter with a
201
In an RFQ mass filter of finite length, ions with unstable motion, which do
202
not experience a sufficient number of RF cycles to hit the electrodes, appear as
203
tails on the mass peaks. In order to investigate this effect and to determine the
204
appropriate length of the mass filter, simulations of the peak shape as a function
205
of the number of RF cycles were carried out for different oscillation amplitudes
206
of the ions. For these simulations, an ideal quadrupolar field was used. Peak
207
tails were still present after 50 RF cycles. Significantly less tail contributions
208
were found after 100 RF cycles. Another enlargement to 150 RF cycles did not
209
significantly improve the peak shape anymore.
210
These results agree well with a similar computer study that observed a big
211
improvement in peak shape up to 100 RF cycles, thereafter a smaller improve-
212
ment in peak shape up to 200 RF cycles, and no improvement beyond 200 RF
213
cycles [50]. Hence an RF frequency of 1 MHz and a length of the mass filter
214
of 30 cm is suitable. A minimum of 100 RF cycles are then obtained for a ion
215
velocity of 3000 m/s. This velocity corresponds to a kinetic energy of 0.9 eV
216
for ions of a mass m = 20 u and 11.7 eV for m = 250 u. The maximum mass
217
resolving power of an RFQ mass filter has been predicted and experimentally
218
confirmed to be of the form [30]
( m )
N2
≤
∆m
Cres
(4)
219
where N is the number of RF cycles that the ion experience. For the constant
220
Cres different values of 5.43, 12.2 and 20 have been given [30, 45, 51]. For a
221
given kinetic energy K, a length L of the mass fiter and an RF frequency f
222
Eq. (4) yields a mass resolution
∆m ≥ 2Cres
9
K
L2 f 2
(5)
223
With Cres = 20, L = 30 cm, f = 1 MHz and K = 5 eV, this estimate gives a
224
mass resolution ∆m ≥ 0.2 u.
225
4.2.2. Vacuum conditions in the mass filter
226
A high gas pressure in the mass filter has two undesired effects; it leads
227
to transmission losses and to peak tails. Simulations have been performed us-
228
ing 133 Cs+ ions with a parameterized, velocity-dependent collision cross section
229
derived from ion mobility measurements [52] in an ideal quadrupolar field for
230
different helium gas pressures. For these simulations the ions were given a ran-
231
domized combination of initial radial positions and velocities, which correspond
232
to different oscillation amplitudes. The encounter of collisions in the RFQ may
233
increase the oscillation amplitude, resulting in the loss of the respective ion.
234
The simulations show that among the ions with stable motion, the ones with
235
large oscillation amplitudes are the ones that are most likely to be ejected. The
236
sensitivity to ejection increases with increasing mass resolving power. At 10−5
237
mbar no or only small (≈ 2%) losses in the transmission are observed behind
238
a 30 cm long mass filter regardless of the oscillation amplitude (Fig. 2). At
239
10−4 mbar significant losses start to appear. About 20% of the stable ions with
240
an oscillation amplitude of r0 and 5% of the ions with an oscillation amplitude
241
of 0.5r0 are lost. At 10−3 mbar substantial losses appear also for ions with
242
an oscillation amplitude of (1/6)r0 . According to the simulations a maximum
243
pressure of 10−4 mbar seems to be appropriate for the RFQ mass filter.
244
4.2.3. Rod shape and rod size
245
For practical reasons it was chosen to use rods with round cross sections for
246
the RFQ system. However, rods with round cross sections create an electric
247
field that contains higher orders, which may deteriorate the peak shape and
248
mass resolving power. If the four rods are positioned perfectly, the first higher
249
order term that will occur is the 12-pole (dodecapole) term. This term can
250
be eliminated if the rod radius R and the field radius r0 are chosen such that
251
R/r0 = 1.14511 [53]. However, numerical simulations have found an optimal
10
252
value R/r0 = 1.12 − 1.13 [54, 55], since at this value of R/r0 the dodecapole
253
term and the next non-vanishing higher order term (20-pole) are of the same
254
magnitude but have different signs [55]. The effect on the electric field between
255
the rods due to a shielding surrounding the RFQ has been studied and found
256
to be very small [54].
257
Figure 3 shows simulations of the peak shape after 100 RF cycles for RFQ
258
mass filters with values on R/r0 of 1.11, 1.125 and 1.14 and ions of different os-
259
cillation amplitudes. The simulations have been performed using a 2-dimension
260
potential distribution calculated for the respective cross section of the RFQ.
261
For these simulations the ions were given a randomized combination of initial
262
radial positions and velocities, which correspond to different oscillation ampli-
263
tudes. At ion oscillation amplitudes up to 0.5r0 no significant transmission
264
losses are observed, except for R/r0 = 1.14. For larger oscillation amplitudes
265
larger losses arise. For R/r0 = 1.14 the transmission efficiency is much worse
266
than for R/r0 = 1.11 and R/r0 = 1.125. For small oscillation amplitudes, es-
267
pecially for the case of R/r0 = 1.11, peak tails start to appear. These tails
268
are harmful since they deteriorate the suppression of neighbouring mass lines
269
and thus the abundance sensitivity. While the transmission efficiency for the
270
ions of choice for large oscillation amplitudes seems to be somewhat larger at
271
R/r0 = 1.11 than at R/r0 = 1.125, the peak tails are much more pronounced
272
at R/r0 = 1.11. Thus rods with R/r0 ≈ 1.125 are a good choice.
273
The higher the mass resolving power of the RFQ mass filter is, the larger
274
are the oscillation amplitudes of the ions with stable motion. Thus a mass filter
275
with large field radius, r0 , is advantageous in order to avoid transmission losses
276
[56]. Furthermore the effect of non-linear fields is much stronger in the outer
277
regions of the RFQ, hence for a given emittance ions in an RFQ with a large
278
field radius will encounter better fields than an RFQ with smaller field radius.
279
Furthermore, an RFQ with a large radius can be built to a required relative
280
tolerance much more readily than an RFQ with a small radius. However the
281
magnitude of the field radius is limited by the RF power supply since according
282
to Eq. (3) the required RF voltage increases quadratically with the radius. This
11
283
can be compensated by choosing a lower RF frequency, but at the cost of fewer
284
ion oscillations in the RFQ. Initial tests showed that with our resonance circuit
285
design a maximal field radius of 8 mm can be used at an RF frequency of 1 MHz
286
.
287
4.2.4. Brubaker lenses
288
Brubaker introduced an RFQ mass filter, which consisted of an RFQ oper-
289
ated in a mass selective mode and which was preceded and followed by two short
290
RFQs (Brubaker lenses) that are mounted collinearly with the mass-selective
291
RFQ and operated in the RF-only mode [17]. The Brubaker lens at the en-
292
trance ensures that the ions are first brought to a working point well inside the
293
stability diagram (a = 0, q = 0.7) before the quadrupole DC component is ap-
294
plied and the ions are brought to the apex. In this way, working points outside
295
the stability diagram are avoided. At the exit the reverse process occurs.
296
In the case of the new RFQ system it is not obvious that Brubaker lenses in
297
the mass filter will be advantageous in the same way as for stand-alone RFQs.
298
Even though the field is distorted by the diaphragms between the RFQs, the
299
ions enter from and exit into the RF fields of the RFQ cooler and RFQ buncher.
300
Furthermore, there is no fringe field from a detector that draws the ions out
301
through the exit diaphragm.
302
Therefore the effect of Brubaker lenses in the RFQ system was investigated
303
by simulations. The electric potential of the mass filter including the exit section
304
of the RFQ cooler and the diaphragm between the cooler and the mass filter
305
was calculated on a 3-dimensional map. An ion population with a thermal
306
distribution of velocities with an offset of 1000 m/s was started in the cooler
307
section. The ions reaching the end of the mass filter were recorded for the two
308
possibilities that a Brubaker lens with a length of 2.5 cm was used and was
309
not used at the entrance and at the exit. The simulations show a dramatic
310
improvement in transmission when a Brubaker lens is used at the entrance of
311
the mass filter (Fig. 4). This finding is in qualitative agreement with those in a
312
recent work, which investigates the effects of Brubaker lenses in detail [57]. The
12
313
improvement of ion transmission for a Brubaker lens at the exit of the mass filter
314
is smaller than at the entrance, however still significantly. As a consequence,
315
the mass filter was chosen to include Brubaker lenses at both the entrance and
316
the exit to obtain optimum transmission.
317
4.3. Design and construction of the complete RFQ system
318
The new RFQ system will replace the present RFQ buncher, which has a
319
length of 106 cm. In order to avoid a re-arrangement of the beam line, the new
320
RFQ system will have the same length. According to the simulations, a length
321
of 40 cm is needed for the mass filter, including two Brubaker lenses with a
322
length of 5 cm each. Since the diameter of a cooled beam is typically smaller
323
than 1 mm, a diaphragm with a diameter of 2 mm at the entrance of the mass
324
filter (D1 in Fig. 1b) was chosen. This avoids ion losses at the entrance to the
325
mass filter, but allows for differential pumping between cooler and mass filter.
326
According to an estimation a minimum length of 20 cm is needed for the RFQ
327
cooler at a buffer gas pressure of 2×10−2 mbar. In addition, some cooling of the
328
ions extracted from the stopping cell will already occur in the extraction RFQ.
329
If the vacuum chamber housing the RFQ mass filter is pumped with a pumping
330
speed of 600 l/s, the buffer gas in the cooler causes a partial pressure in the mass
331
filter of 3 × 10−5 mbar. If the buncher is operated at a pressure of 5 × 10−3 mbar
332
and the total pressure in the mass filter should be kept below 1 × 10−4 mbar, a
333
diaphragm with an aperture diameter of 5 mm or less is necessary at the exit
334
of the mass filter (D2 in Fig. 1b). For best transmission, a diameter of 5 mm
335
was chosen. The remaining length of 46 cm is used for the RFQ buncher. The
336
cooling section of the buncher has a length of 42.6 cm. The trapping region of
337
the buncher consists of three segments, each with a length of 1 cm. The buncher
338
ends with two extraction diaphragms (D3 and D4 in Fig. 1b) with a diameter
339
of 3 mm each, which also serve as pumping barrier towards the beam line.
340
The rods of the RFQ system are built from stainless steel. The RFQ cooler
341
and the RFQ buncher have rods with a radius of 5.5 mm and a field radius of
342
5 mm. The rods of the mass filter have a radius of 9 mm. The mass filter is put
13
Parameter
Cooler
Mass filter
Buncher
Length / mm
200
400
460
Field radius r0 / mm
5.0
7.98
5.0
Rod radius R / mm
5.5
9.0
5.5
RF frequency / MHz
1.5
1.0
1.0
Pressure (He) / mbar
2 × 10−2
< 10−4
5 × 10−3
Table 1: Geometrical parameters and typical operating parameters of the RFQ system.
343
in a grounded shielding with an inner radius of 32 mm. After the components
344
of the RFQ were assembled, the distances between pairs of opposite rods were
345
measured at the entrance and exit of the mass filter. Three of the four distances
346
were found to be close to 15.96 mm corresponding to a R/r0 ratio of 1.128.
347
The distance of one pair of rods at the entrance is about 80 µm larger. The
348
geometrical parameters and typical operating parameters of the RFQ system
349
are summarized in Fig. 1 and in Table 1.
350
5. Experimental setup
351
A photo of the experimental setup is shown in Fig. 5. In the middle of the
352
photograph, four vacuum chambers containing an ion source, the RFQ system
353
and a detector can be seen. Resonance circuits for the generation of high RF
354
voltages are placed in the boxes on top of the setup. Power supplies for the
355
other electrodes are placed in the electronic racks on the left and the right side
356
of the setup. A helium gas bottle supplies the cooler and buncher with buffer
357
gas. A computer controls the timing for the injection and extraction of ions as
358
well as the voltages of the RFQ mass filter.
359
5.1. Vacuum system
360
All three RFQs are located in individual chambers that are differentially
361
pumped. Both, the RFQ mass filter and the RFQ buncher are mounted in cus-
362
tomized CF 160/200 6-way crosses. The lengths of the chambers were matched
14
363
to the lengths of the RFQs. This facilitates a leak-tight mounting of the RFQ
364
system in the flanges. In the test setup, the cooler RFQ is encapsulated in order
365
to keep the pressure in the vacuum chamber low enough for the operation of
366
an ion source. When the system is installed at SHIPTRAP the encapsulation
367
of the cooler will be removed. Each vacuum chamber of the setup is pumped
368
by its own turbomolecular pump backed by scroll vacuum pumps. In the test
369
setup, the helium buffer gas is fed into the RFQ cooler and the RFQ buncher
370
via manually operated needle valves. The residual pressure in the mass filter
371
and in the buncher amounts to 2 × 10−8 mbar.
372
5.2. Ion sources and detectors
373
For commissioning of the RFQ system surface ionization sources (HeatWave
374
Labs Inc., model number 101139) and an electron impact source [58, 59] were
375
used. Ion detection was performed with a secondary electron multiplier (Pho-
376
tonis 5901 Magnum or ETP DM167 MagneTOF). Depending on the study, the
377
output current from the detector was measured with an oscilloscope, with an
378
electrometer, or with a pulse counter.
379
5.3. Electronics
380
The RF voltages needed to operate the RFQs are generated with high-quality
381
resonance circuits that are based on the principles found in the Refs. [59–61].
382
A sinusoidal voltage is amplified by a KL 500 power amplifier (RM Italy) and
383
coupled via a ferrite transformer core to a resonance circuit. Two RF voltages,
384
RF+ and RF-, with a phase shift of 180 degree are generated. Care needs to
385
be taken in order to avoid small asymmetries between the two RF phases. This
386
asymmetry can be caused by differences in the amplitudes of the two phases or
387
by a phase shift between the two RF phases that differs from 180◦ . It results in
388
a non-vanishing RF potential on the axis of the RFQ and hence in a spread in
389
the kinetic energies of the extracted ions, which can amount to several eV. In
390
case of the RFQ cooler, the consequence would be that the mass resolving power
391
of the mass filter may be decreased. For the buncher, such an energy spread
15
392
of the extracted ion packets may decrease the overall transmission efficiency
393
of SHIPTRAP, since the energy acceptance of the Penning traps is limited to
394
about 10 eV. Therefore adjustable capacitances between each of the RF voltages
395
and ground enable precise adjustment of the two RF phases.
396
The RFQ cooler is operated at an RF frequency of 1.5 MHz and RF voltages
397
(peak-peak) of up to 2.5 kV. In the work of others, the DC voltage offset of an
398
RFQ is often applied in the middle of the coil without any resistors. This solu-
399
tion was rejected in this work, since it may inadvertently produce an amplitude
400
asymmetry for the two RF voltages, RF+ and RF-. Instead, in this work the
401
DC voltage offset for the RFQ cooler is coupled to each of the two RF high
402
voltages RF+ and RF- independently.
403
The RFQ mass filter is operated at an RF frequency of 1.0 MHz and RF
404
voltages (peak-peak) of up to 2.5 kV. This allows to bring singly charged ions
405
with a mass of up to 260 u to the apex of the stability diagram. The three
406
segments of the mass filter are all operated with the same RF voltage, but
407
different DC potentials. In addition to the usual DC monopole potential of
408
the rods, a DC quadrupole potential is applied to the filtering section. The DC
409
voltages are delivered by two HV power supplies (HCN 7E-1250, FuG Elektronik
410
GmbH, Schechen, Germany). The amplitude of the RF voltage needs to be
411
stabilized. For this purpose, an RF/DC converter was developed for a precise
412
measurement of the RF amplitude (see also Section 5.4).
413
The RFQ buncher is operated at an RF frequency of 1.0 MHz and RF volt-
414
ages (peak-peak) up to 1.2 kV. Except for the trapping segment of the buncher,
415
the RF voltage is transferred over a capacitance, coupled to the corresponding
416
DC potential and applied to the electrodes. For the trapping segment the DC
417
potential is coupled to the RF+ and RF- without any capacitor between the
418
coil and the segment. This results in an RF amplitude on the trapping segment
419
that is larger by about 10% than that on the other segments.
420
Even though the amplitudes of the RF+ and RF- are adjusted to be equal to
421
better than 1%, small differences in the amplitude may remain. This asymmetry
422
results in a monopole potential on the axis of the RFQs oscillating with the
16
423
RF frequency. Thus the time-of-flight to the detector downstream may also
424
dependent on the phase of the RF voltage at extraction. This drawback is
425
avoided through implementation of a switch, which shortcuts the resonance
426
circuit and switches off the RF voltage before the ions are extracted. Additional
427
switches are connected to S1 and S3. These are used to change the DC potential
428
to a higher and lower value such that the ions stored on segment S2 are extracted.
429
One finds that the RF voltage is switched off within a fraction of one RF cycle
430
and that the rise time of the extraction voltages are on the order of one µs. Note
431
that the development of these switches enables a push-pull extraction with only
432
one resonance circuit. The buncher can also be operated with the ions stored on
433
segment S3. In that case, the DC potentials on the buncher electrodes are kept
434
constant and the ions are extracted by switching of the potential of diaphragm
435
D3 only.
436
5.4. System control and data acquisition software
437
A system control software has been developed to control the RFQ mass
438
filter. The program sets the RF and DC quadrupole voltages of the mass filter,
439
stabilizes the RF voltage and includes data acquisition. The program sets the
440
voltages via a two-channel frequency generator (Tektronix AFG 3022B) and
441
reads out detector signals via either a counter/timer (Ortec 994) or a multimeter
442
(Keithley Model 2002). The stabilization is realized through a feedback loop.
443
Each of the two RF high voltages is connected to an RF/DC converter. The
444
output is read by a 14-bit AD converter (National Instruments NI-USB 6009).
445
The digital output of the AD converter is read by the control software and
446
used to adjust the output amplitude of the frequency generator using a PID
447
algorithm.
448
Three different operating modes of the RFQ mass filter are supported in the
449
software. The first mode allows operation of the mass filter at a fixed working
450
point, i.e. as mass separator. The next possibility is to scan the voltages U and
451
V at a fixed ratio U/V , i.e. operation as mass spectrometer. The third mode
452
performs a two dimensional scan of the voltages U and V , such that the stability
17
453
diagram is mapped.
454
6. Performance of the RFQ system
455
The performance of the RFQ system was investigated for individual com-
456
ponents of the system as well as with the complete system. Unless indicated
457
otherwise, the tests were carried out using a
458
der the following conditions: The RFQ cooler was operated at q = 0.5 and with
459
a buffer gas pressure of 3 × 10−2 mbar. The buncher was operated at q = 0.4
133
Cs surface ionization source un-
460
in the cooling section and with a buffer gas pressure of 5 × 10−3 mbar. The
461
kinetic energies of the ions at injection into the RFQ cooler and RFQ mass filter
462
amounted to 5 eV and 3 eV, respectively, and the DC voltages of the RF rods
463
and the LINAC rods were chosen such that there was a drop of about 1 eV
464
in the electric potential along the axis of the RFQ mass filter and the cooling
465
section of the RFQ buncher.
466
6.1. Performance characteristics of the RFQ cooler and the RFQ mass filter
467
In a first step, the RFQ mass filter and the RFQ cooler were studied. For
468
these studies, the test setup consisted of the ion source, the RFQ cooler, the
469
diaphragm D1, the RFQ mass filter, the diaphragm D2 and a detector mounted
470
directly behind D2.
471
6.1.1. Cooling performance of the RFQ cooler
472
The cooling properties of the RFQ cooler were studied using a retarding
473
technique. In this test, the cooler RFQ was operated with an RF frequency
474
of 1.2 MHz at a value of the Mathieu parameter q of 0.4. The mass filter was
475
operated in RF only mode at values of the Mathieu parameters a = 0 and
476
q = 0.4. A retarding potential was applied to the diaphragm D1 between the
477
RFQ cooler and the RFQ mass filter. The potential was increased in steps and
478
the ion current on the detector was recorded. The derivative of the detector
479
current with respect to the applied retarding potential gives the distribution
18
480
of the axial ion kinetic energy. This measurement was performed for different
481
pressures.
482
The results are shown in Fig. 6. The energy spread ∆K of the ions decreases
483
with increasing pressure p. At 5 × 10−2 mbar the measured spread (FWHM)
484
is 0.15 eV and at higher pressures, the width does not decrease much further.
485
The solid line shows a fit of the function
∆K = ∆K0 + Ck exp(−p/p0 )
(6)
486
to the data, where ∆K0 , Ck and p0 are constants. The fit predicts a minimal
487
energy spread, ∆K0 , of about 0.1 eV. This measured minimum energy spread
488
is an experimental limit caused by field penetration through the diaphragm and
489
small asymmetries in the RF voltages.
490
The measurement shows that the RFQ cooler works as expected. Following
491
the initial slope of the decrease in energy width in Fig. 6 (dashed line), one finds
492
that the thermal energy of 0.0125 eV (dotted line) is reached at a pressure of
493
about 8×10−2 mbar. This is thus the approximate pressure needed for optimum
494
cooling of
495
was sufficient for an optimum performance. When the RFQ system is installed
496
at SHIPTRAP, the cooling of the ions will start already in the extraction RFQ.
497
The only pre-cooled ions will enter the cooler RFQ, therefore only about half
498
the pressure will be required in the cooler RFQ.
499
6.1.2. Brubaker lenses
133
Cs+ ions. In practice, a pressure in the cooler of 3 × 10−2 mbar
500
During the optimization phase of the RFQ system, i.e. under conditions
501
characterized by lower RF frequencies and non-stabilized RF voltages, the pos-
502
sible benefits of individual Brubaker lens sets were investigated experimentally.
503
The cooler was operated using air as buffer gas at a pressure of 1 × 10−2 mbar
504
and an RF frequency of 1.2 MHz at q = 0.5. Mass scans of
505
performed for four different cases: (i) Brubaker lenses at the entrance and at
506
the exit operated at a = 0 and q = 0.7 and mass filtering carried out only in
507
the 30 cm long middle section of the mass filter. (ii) A Brubaker lens at the
19
133
Cs+ ions were
508
entrance of the mass filter. The two following sections, in total 35 cm, were
509
operated close to the apex of the stability diagram. (iii) The first two sections
510
were operated close to the apex and the third section as a Brubaker lens. (iv)
511
All three sections of the mass filter operated close to the apex of the stability
512
diagram.
513
The results (Fig. 7) show that Brubaker lenses improve both the transmission
514
efficiency and the peak shape. While a comparison of the spectra indicates that
515
a Brubaker lens at the exit improves the performance, but only by a certain
516
amount, it clearly demonstrates that a Brubaker lens at the entrance is crucial
517
for a good performance. The best performance was obtained for the case, in
518
which Brubaker lenses at both the entrance and at the exit of the mass filter
519
were used. These findings are good agreement with the results of the simulations
520
(Section 4.2.4).
521
6.1.3. Resolution and relative transmission efficiency
522
An investigation of the transmission efficiency as a function of the mass
39
K+ and
133
Cs+ ions. Mass scans were
523
resolving power was carried out using
524
performed at different U/V ratios. The mass resolving power was evaluated
525
at full width at half maximum (FWHM) and at the full width at 10% of the
526
peak height (“10%”). The relative transmission efficiency is defined as the ratio
527
between the number of counts during filtering (U/V ̸= 0) and the number
528
of counts in RF only mode (U/V = 0). The measured relative transmission
529
efficiencies as a function of the mass resolving power are shown in Fig. 8.
530
In the case of
133
Cs+ , a relative transmission efficiency of unity was mea-
531
sured up to a mass resolving power of about 200 (FWHM) and 130 (“10%”),
532
respectively. Beyond this value, transmission losses start to appear. If minor
533
losses in transmission efficiency of about 10% are tolerated, mass resolving pow-
534
ers of 240 (FWHM) and 160 (“10%”) are obtained. The study shows a similar
535
trend for 39 K+ ions. However, the transmission efficiency starts to drop at lower
536
mass resolving power. At 90% transmission efficiency, a mass resolving power
537
of 80 (FWHM) and 70 (“10%”) was measured.
20
538
There are different limitations to the mass resolving power of RFQ mass
539
filters. Some of them, such as the number of RF cycles that an ion experiences
540
in the mass filter (Eq. (5)), lead to a constant mass resolution ∆m independent
541
of mass and consequently to a mass resolving power (m/∆m) that increases
542
linearly with mass. Others, such as non-linear field contributions, lead to a
543
constant mass resolving power (m/∆m) independent of mass. The measure-
544
ments for
545
mass, however less than proportionally. Therefore, at present the mass resolving
546
power of the RFQ mass filter is most likely limited by several factors that are
547
of similar magnitude.
39
K+ and
133
Cs+ show a mass resolving power that increases with
548
It is important to note that in both cases, for 133 Cs+ and for 39 K+ , the mea-
549
sured mass resolving power shows that the RFQ mass filter fulfils its task, which
550
is to separate nuclides according to their mass number at very high transmission
551
efficiency. As a comparison, the mass resolving power of the dipole magnet at
552
the IGISOL facility in front of JYFLTRAP is about 500 [62]. The mass filter
553
installed at the LEBIT facility has been reported to have a mass resolving power
554
of 50 for nuclides with a mass of about 32 u [34]. In both cases the corresponding
555
transmission efficiencies have not been reported.
556
Some commercial RFQ mass filters have achieved much higher mass resolving
557
powers. The Extrel MAX-50 features an RF frequency of 2.9 MHz and reaches
558
a mass resolving power of 4000 (FWHM) at a mass-to-charge ratio of 40 u/e
559
[63]. The Thermo Scientific TSQ Quantum features hyperbolic rods and reaches
560
a mass resolving power of 7500 (FWHM) at a mass-to-charge ratio of 508 u/e
561
[64]. In both specifications, no absolute transmission efficiencies are reported.
562
These examples indicate that further improvement in the performance of the
563
RFQ mass filter for SHIPTRAP should be possible with further advances in
564
the technology used, such as the machining accuracy and the quality of the RF
565
electronics, which have been improved over decades in commercial mass filters.
21
566
6.1.4. Abundance sensitivity
567
The abundance sensitivity, i.e. the maximum abundance ratio of neighboring
568
mass lines that can be measured with the RFQ mass filter, was investigated
569
with a thermal ion source containing barium isotopes and
570
ratio was tuned to 90% transmission efficiency, which corresponds to a mass
571
resolving power of 240 (FWHM). Repeated mass scans were performed over a
572
mass-to-charge range of 128 u/e0 to 140 u/e0 , where e0 is the elementary charge,
573
using 120 steps and a measurement time per point of 10 s. In total, 26 spectra
574
were recorded. The total rate of detected ions in this mass range was about
575
1.3 × 106 /s.
133
Cs. The U/V
The sum of all spectra is shown in Fig. 9a. The dominant mass line at
576
133
Cs+ . In the mass-to-charge range from 135 u/e0 to
577
133 u/e0 is identified as
578
138 u/e0 , well separated mass lines with an abundance distribution correspond-
579
ing to the natural abundance of
580
fit of the sum of six Gaussian functions to the data. The green curves show the
581
Gaussian function for each individual mass line. One notes the excellent peak
582
shape that can be well described by a Gaussian function over many orders of
583
magnitude.
135−138
Ba+ are found. The red curve shows a
The spectrum provides an estimate of the suppression of neighboring masses.
584
133
Cs+ line shows a suppression by four orders of magnitude
585
The mass line of
586
at the low mass side (132 u/e0 ) and by at least five orders of magnitude at the
587
high mass side (134 u/e0 ). Mass lines separated by 2 u/e0 are suppressed by at
588
least eight orders of magnitude.
589
Figure 9b shows the corresponding calculated abundance distribution of the
590
ion population transmitted through the RFQ mass filter if it is tuned to isolate
591
135
Ba+ . The spectrum has been calculated from the abundance of fitted curves
135
Ba+ . Clearly, the
592
for the individual mass lines at the centroid position for
593
RFQ mass filter is well suited to provide well isolated ion populations even under
594
the presence of strong contaminants.
22
595
6.1.5. Temporal stability of the mass filter
596
In order to investigate the long-term stability of the RFQ mass filter, mass
597
spectra were recorded for 30 minutes with de-activated stabilization of the RF
598
voltages and for 4 hours with activated stabilization. The nominal mass scan line
599
was the same for all cases. The results of the measurements show that the mass
600
resolving power of the mass peaks varies by more than a factor of two without
601
the stabilization, while no change in the mass resolving power is observed with
602
activated stabilization of the RF voltages. In addition, the stabilization of the
603
RF voltages results in a much more stable peak position, which is equivalent to
604
the selected mass-to-charge ratio.
605
Using the stabilization of the RF voltage, the sensitivity of the RF and DC
606
voltages of the mass filter to changes in the room temperature was studied.
607
In order to separate voltage drifts in the applied DC and RF potentials, mass
608
scans were carried out at constant DC voltage. In this mode, a shift in RF
609
voltage results in a shifted peak position, whereas a drift in the DC potential
610
affects the peak width. The study shows a remaining RF amplitude voltage drift
611
that amounts to about 0.1%/K. From an analysis of the apex of the stability
612
diagram, it follows that the drift of the DC voltage does not exceed 0.03%/K.
613
6.1.6. Absolute transmission efficiency
614
The absolute efficiency of the RFQ system was determined by measuring
615
the ratio of the ion current transmitted through the combination of RFQ cooler
616
and RFQ mass filter to the ion current entering the RFQ cooler. For the mea-
617
surement, a beam analysis unit [65, 66] was used. It allows to transmit the
618
incoming ion beam or to measure the ion current by deflection into a detector.
619
The mass filter was operated in RF-only mode at a Mathieu parameter q = 0.7.
620
The absolute transmission efficiency was determined to be (88 ± 5)% with an
621
additional overall systematic uncertainty of 15%. Further investigations show
622
that the absolute transmission efficiency is actually 10% larger (see discussion
623
in Section 6.3.3). The absolute transmission efficiency in filtering mode is ob-
624
tained by multiplying the absolute transmission efficiency in RF-only mode with
23
625
the relative transmission efficiency for the corresponding mass resolving power
626
(Fig. 8).
627
6.2. Performance characteristics of the RFQ buncher
628
The buncher is constructed such that the ions can be trapped either on
629
the second (S2) or third (S3) segment (Fig. 1b). In the following, these two
630
different modes of operation will be referred to as storage on S2 and storage
631
on S3. As discussed in Section 4.1, each operation mode has its advantage.
632
Therefore investigations were carried out to find out which mode of operation
633
of the buncher gives the best performance. For some tests, the buncher was also
634
operated in a mode called transmission. In this case the buncher is operated as
635
an ion guide and the ions are not trapped at all.
636
For the performance characterization of the RFQ buncher, a test setup con-
637
sisting of an ion source, the RFQ buncher, a tube with a variable length (about
638
5-10 cm) and a detector was used. The potential of the drift section was varied
639
dependent on study and is hereafter referred to as drift potential. The inves-
640
tigations were performed in a chamber where the buncher was encapsulated in
641
order to keep the pressure in the chamber low enough for the operation of the
642
ion source and the detector.
643
6.2.1. Cooling time in the buncher
644
The blocking potential on the diaphragm D3 in the RFQ buncher causes the
645
incoming ions to travel back and forth repeatedly through the cooling section
646
of the buncher. Due to collisions with buffer gas atoms, the ions eventually
647
thermalize and are stored in the potential minimum at the trapping segment.
648
The time scale of this cooling process was studied by a measurement of the
649
temporal width of the extracted ion packet. The ions were injected, cooled
650
for a variable time and then ejected from the buncher into the drift tube and
651
detected. The drift potential was tuned such that the position of the time focus
652
coincided with the position of the detector.
24
653
The measurement was performed at a buffer gas pressure of (5 ± 2) × 10−3
654
mbar. The change of the time spread with cooling time for the storage on S3
655
is shown in Fig. 10. The width is limited mainly by the turn-around time [67],
656
i.e. the time an ion with a velocity component in the direction opposite to the
657
direction of extraction needs to reverse its direction of motion. The temporal
658
spread ∆t of the detected ion packet decreases exponentially with cooling time
659
t and asymptotically approaches the thermal limit ∆t0 . A fit of the form
∆t = ∆t0 + Ct exp(−t/τ )
(7)
660
to the data points yields the time constant τ of the cooling process. The cooling
661
constants were found to be (1.2 ± 0.1) ms and (1.3 ± 0.1) ms for storage on S2
662
and S3, respectively. The measurements show that cooling is easily achieved on
663
the time scale of a few milliseconds and that the cooling works equally at both
664
on trapping positions.
665
6.2.2. Ion temperature
666
667
The temperature T of the ions is related to the turn-around time, ∆tta , of
the ions through [68]
∆tta =
√
8 ln(2)
√
mkB T
,
eEextr
(8)
668
where Eextr is the electric field strength during ejection from the buncher. Thus
669
the axial component of the thermal energy can be determined from the temporal
670
width of the extracted ion packet. A corresponding measurement was performed
671
for different extraction field strengths at a pressure of 3 × 10−4 mbar. The ion
672
temperature was determined from a linear regression to the data. A temperature
673
of (423 ± 35) K was obtained for storage on S2 and of (419 ± 46) K for storage
674
on S3. These values are close to the thermal limit of 300 K.
675
6.2.3. Energy spread
676
The energy spread of the ion packets extracted from the RFQ buncher was
677
measured using a retardation technique. In order to measure the energy spread,
678
five grids were installed in the drift section between the diaphragm D4 and the
25
Trap
Field
Energy
Energy
depth
strength
spread
shift
/V
/ (V/mm)
/ eV
/ eV
Storage on S2
9.5
3.1 ± 0.3
3.0 ± 0.3
−1.2 ± 0.5
5.3
12.5 ± 1.3
14.0 ± 1.4
−6.7 ± 0.5
9.5
12.5 ± 1.3
10.5 ± 1.1
−5.8 ± 0.6
18.0
12.5 ± 1.3
8.9 ± 0.9
−6.9 ± 0.5
9.5
24.9 ± 2.6
19.8 ± 2.0
−12.7 ± 0.7
Storage on S3
9.2
3.0 ± 0.3
2.7 ± 0.3
−5.5 ± 0.5
9.2
11.7 ± 1.1
7.3 ± 1.5
−28.3 ± 0.6
Table 2: Measured energy spread (FWHM) and energy shift of the extracted ion packets for
different extraction field strengths and potential depths of the trap. For details see text.
679
detector. The drift potential was applied to the first and the fifth grid whereas
680
a retarding potential was applied to the three central grids. The abundance of
681
ions reaching the detector was recorded as a function of the applied retarding
682
potential. The derivative of the ion current on the detector with respect to the
683
applied retarding potential gives the energy distribution of the ions in the lon-
684
gitudinal direction. This energy spread (FWHM) was determined for different
685
potential depths of the trap, extraction field strengths and trapping segments.
686
The trap depths were derived from a calculation of the electric potential distri-
687
bution on the axis of the RFQ.
688
The results are given in Table 2. The energy spread increases linearly with
689
the field strength. Furthermore it decreases with increasing trap depth. For
690
a given ion kinetic energy during storage an increasing trap depth leads to a
691
decreasing spatial extension. The results are thus consistent with a kinetic
692
energy spread that is given by
∆K = eEextr ∆z
26
(9)
693
where ∆z is the spatial spread of the ion packet in the axial direction during
694
storage.
695
Furthermore, the measurement shows that for ions stored on S3 the measured
696
average kinetic energy of the ion packets is lower than the potential energy of the
697
ion packet during storage. The latter was determined from a calculation of the
698
electric potential distribution on the axis of the RFQ. This energy shift is caused
699
when the trap is switched from storage to extraction. Since only the potential of
700
diaphragm D3 is switched, the potential at the position of storage also changes.
701
For ions stored at S2 such a potential shift is not expected, since the potentials
702
on the segments S1 and S3 are switched anti-symmetrically during extraction,
703
such that the trap potential is not changed between storage and extraction.
704
Nonetheless a potential shift was observed even for storage at S2. This effect is
705
due to the voltage variations on the RFQ rods during the (non-instantaneous)
706
switch-off of the RF voltage and could be reproduced in simulations taking into
707
account the measured time evolution of the voltages on the segements S1, S2 and
708
S3. This potential shift has a linear dependence on the extraction field strength
709
and is independent of the trap depth. It is about a factor of four smaller than
710
the shift observed for ions stored at S3.
711
6.2.4. Longitudinal emittance
712
Figure 11 shows the measured time and energy spreads from the the tem-
713
perature measurements (Section 6.2.2) and the energy spread measurements
714
(Section 6.2.3). The solid curve shows the theoretical dependence according
715
to Eqs. (8) and (9) for a packet of
716
∆z = 0.8 mm and a temperature T = 410 K. For an energy spread of 10 eV,
717
which is considered to be the approximate upper limit for an efficient capture
718
in the Penning trap, a time spread of 40 ns of is achieved.
133
Cs+ ions, which has a spatial spread
719
At the time focus, the longitudinal emittance of the extracted ion packets
720
is given by the product ∆K∆t. The mean of the five measurements shown in
721
Fig. 11 yields a value of the longitudinal emittance of (0.45 ± 0.05) eVµs. This
722
is close to the theoretical (thermal) limit and an improvement by a factor 10
27
723
compared to the buncher presently installed at SHIPTRAP [16] and by a factor
724
20 compared to the ISOLTRAP buncher [69].
725
The performance of time-of-flight measurements in the electrostatic beam
726
line behind the RFQ buncher to be expected with these buncher characteristics
727
was studied using simulations [35]. The ions were stored on S2 (S3) and ex-
728
tracted with ±200 V (−400 V) corresponding to an electrical field strength of
729
12.5 V/mm (11.7 V/mm). The drift potential was set to 980 V lower than the
730
trapping potential. Voltage settings of the einzel lenses were found such that
731
a time focus is achieved at MCP2 (see Fig. 1a) after a time-of-flight of about
732
28 µs. With a peak width of 40 ns of the ion packet this corresponds to a mass
733
resolving power (FWHM) of 350. This is to be compared to the mass resolving
734
power of 100 on MCP2 achieved experimentally with the previous RFQ buncher
735
[35].
736
6.2.5. Ion capacity
737
The maximum number of ions that can be stored in the trap of the RFQ
133
Cs+
738
buncher at the same time, i.e. the ion capacity, was measured using
739
ions. For an increasing number of ions injected into the buncher, the number of
740
extracted ions increases. For up to about 5 × 106 ions the increase is linear and
741
then starts to saturate. This saturation is most likely due to space charge effects
742
that lead to a widening of the ion cloud and consequently to an emittance growth
743
and losses of ions on the diaphragm D3 during extraction. An ion capacity of
744
1.5 × 107 was obtained for storage on segment S2 and more than 3 × 107 on
745
segment S3. The lower measured storage capacity on S2 is most probably caused
746
by ther longer path of the ions during extraction and hence increased losses on
747
the diaphragm D3. In both cases, the capacity exceeds the number of ions that
748
can be handled by the purification trap by several orders of magnitude and will
749
not represent any limitation to the experiment.
28
750
751
6.2.6. Absolute transmission efficiency
The absolute transmission efficiency of the RFQ buncher was measured with
133
Cs+ ions. The di-
752
the same technique as described in Section 6.1.6 using
753
aphragm D4 and the drift tube were set to a drift potential about 1 kV below
754
the trapping potential. This ensures that the efficiencies measured are obtained
755
under conditions relevant for SHIPTRAP. The absolute transmission efficiencies
756
were determined to be (89 ± 6)% (transmission), (85 ± 4)% (storage on S2) and
757
(82 ± 5)% (storage on S3). As in the measurement of the transmission efficiency
758
of the RFQ cooler and mass filter, an additional overall systematic uncertainty
759
of 15% applies, and further investigations show that the absolute transmission
760
efficiency is actually 10% larger (see discussion in Section 6.3.3).
761
6.3. Performance characteristics of the complete RFQ system
762
In a final test series, the complete RFQ system was investigated with the
763
setup described in Section 5.
764
6.3.1. Storage time
765
The storage time of the RFQ buncher, i.e. the period of time until the
766
number of storage ions has decreased to 1/e of the original abundance, was
767
measured for different ion species. The potential of the drift tube installed
768
between the diaphragm D4 and the detector was adjusted such that the time
769
focus coincides with the detector position. The detector signal was recorded
770
with an oscilloscope. In this study the buncher was operated at a Mathieu
771
parameter q = 0.5.
772
Storage times of (27 ± 1.4) s (storage on S2) and (20 ± 0.9) s (storage on
773
S3) were measured for cesium ions at a helium buffer gas pressure of 7 × 10−3
774
mbar. Since the ionization potential of cesium is low compared to that of other
775
atoms and molecules, which may be present in the setup, this storage time
776
mainly reflects losses caused by elastic collisions. Further measurements with
777
molecular oxygen ions, O+
2 , using a buffer gas pressure in the buncher of 1 ×
778
10−3 mbar, resulted in a storage time of (24.2 ± 4.4) s for storage on S3. The
29
779
measured storage time of krypton was about 1.5 s, and most likely limited by
780
water introduced with the buffer gas.
781
The measurements show that long storage times in excess of 20 s are possible.
782
These long storage times are independent of the storage position (S2 or S3) and
783
obtained both for low mass (32 u) and medium-heavy mass (133 u). The long
784
storage time was demonstrated up to an ionization potential of the ion species
785
of 12.1 eV (O2 ). The storage times have to be compared with the cycle time
786
of SHIPTRAP of typically 1 s. They are thus long enough by far and will not
787
influence the storage of singly charged ions of any element except for a few
788
lighter species (N, O, F, Cl, Br) and noble gases (He, Ne, Ar, Kr). However, the
789
storage time of doubly charged ions may be limited. For these, a gas cleaning
790
system will be required at the gas inlet to the RFQ buncher.
791
6.3.2. Resolution and relative transmission efficiency
792
The measurement of transmission efficiency as a function of the mass resolv-
793
ing power (section 6.1.3) was repeated with the complete RFQ system. Verifi-
794
cation was considered to be necessary since in the mass filter the ion have large
795
oscillation amplitude and it is easier to detect all ions passing through the exit
796
diaphragm of the mass filter if a detector is installed behind the mass filter than
797
to capture the ions in the RF field of the buncher and detect them behind the
798
buncher. The buncher was operated with a buffer gas pressure of 7 × 10−3 mbar
799
at a Mathieu parameter q = 0.5. The measurement shows that the transmission
800
efficiency starts to deviate from unity at an only slightly lower mass resolving
801
power.
802
6.3.3. Absolute transmission efficiency
803
The absolute transmission efficiency for the complete RFQ system was de-
804
termined using the method described in Section 6.1.6. The RFQ mass filter was
805
operated in the RF only mode (q = 0.64) and the buffer gas pressure in the
806
buncher was 7 × 10−3 mbar. The absolute transmission efficiencies were deter-
807
mined to be (88 ± 5)% (transmission), (79 ± 4)% (storage on S2) and (82 ± 5)%
30
Cooler +
Buncher
Mass filter
Cooler +
Mass filter +
Buncher
(89 ± 6)%
(88 ± 5)%
Storage on S2
(85 ± 4)%
(79 ± 4)%
Storage on S3
(82 ± 5)%
(82 ± 5)%
Transmission
(88 ± 5)%
Table 3: Measured absolute transmission efficiencies of the individual quadrupoles and for
the complete RFQ system. Note that the true transmission efficiencies are larger by 10% and
that an overall systematic uncertainty of 15% applies, see discussion in the text.
808
(storage on S3). The results of this measurement and those for the individual
809
components are summarized in Table 3.
810
It is interesting to note that the measured individual transmission efficien-
811
cies of the combination of cooler and mass filter one the one hand and of the
812
buncher in transmission mode on the other hand are (88 ± 5)% and (89 ± 6)%,
813
respectively. Thus one would expect the overall efficiency for the whole system
814
in transmission mode to be given by the product of these values, i.e. 78%. How-
815
ever the measured transmission efficiency for the complete system is also close
816
to 90%. Similarly, one would expect the transmission efficiency of the complete
817
system in storage mode to be about 88% of the values measured for the buncher
818
itself. However, almost the same as the transmission efficiencies for the whole
819
system are measured as for the RFQ buncher. This indicates that during the
820
transmission measurements a 10% loss of ions has occurred inadvertently during
821
the passage from the beam analysis unit into the entrance of the respective RFQ
822
entrance. This loss is not related to the transmission efficiency of the RFQs.
823
The true transmission efficiencies are thus in reality larger by about 10% than
824
measured. Hence the transmission efficiencies are close to 100% in the trans-
825
mission mode and about 90% for storage on S2 and S3. The overall systematic
826
uncertainty of these measurements was estimated to 15%.
827
A comparison of the transmission efficiencies between the RFQ system and
828
that of other RFQs used in nuclear physics experiments shows that the figures of
31
829
the new system are highly competitive. For example, the reported transmission
830
efficiencies for other RFQ cooler/bunchers are on the order of 20-40%, 60%,
831
close to 100%, and 40%, respectively [16, 69–71].
832
7. Conclusions and outlook
833
A system of three matched RF quadrupoles has been developed, which will
834
extend the reach of SHIPTRAP to rare nuclides produced together with high
835
rates of unwanted reaction products. This triple RFQ system has a total length
836
of one meter only and provides ion cooling, identification and separation by mass
837
number, and ion packet bunching. The absolute transmission efficiency, which is
838
of utmost importance when dealing with rare exotic nuclides, was measured to
839
be almost 100%. A mass resolving power (FWHM) of 240 at 90% transmission
840
efficiency has been achieved for
841
masses by more than four orders of magnitude has been demonstrated for barium
842
isotopes at rates exceeding 106 ions/s. Cooling times on the time scale of a few
843
milliseconds and storage times of up to 24 s were obtained in the RFQ buncher.
844
The ion capacity was found to be larger than 107 ions. These performance
845
values exceed the requirements for experiments at SHIPTRAP. The longitudinal
846
emittance was measured to be (0.45 ± 0.05) eVµs and is close to the theoretical
847
(thermal) limit. This will allow for efficient injection into the Penning trap and
848
an improved broadband time-of-flight mass spectrometry in the electrostatic
849
beam line.
133
Cs+ ions. A suppression of neighboring
850
These remarkable performance characteristics were enabled by several im-
851
portant technical developments. High quality resonance circuits were developed
852
and care was taken to achieve very symmetric RF voltages. An electronic circuit,
853
which enables to switch off the RF voltages in much less than one RF cycle and
854
to shift the DC potential of the electrodes independently, was developed. This
855
circuit is important in order to avoid a distorted time focus of the extracted ion
856
packet and to keep the energy spread of this ion packet small. A system control
857
and data acquisition software was developed, which includes stabilization of the
32
858
RF voltage of the mass filter.
859
The triple RFQ system will open new research possibilities at SHIPTRAP.
860
For example, the measurement of N =Z nuclides in the region up to tin, such
861
as
862
for access to nuclides with half lives of a few milliseconds, and for even more
863
sensitive measurements a multiple-reflection time-of-flight mass spectrometer
864
(MR-TOF-MS) and isobar separator can be employed [66, 72]. This will then
865
give access to nuclides such as
80
Zr and
84
Mo, will become possible. For further suppression of isobars,
94
Ag and
100
Sn.
866
The RFQ system follows a new concept for low-energy beam lines [36], in
867
which RFQs are used for ion transport at low kinetic energies instead of con-
868
ventional electrostatic beam lines. This enables compact setups that are ideally
869
suited to the gas-filled environment in the differential pumping section in the
870
vicinity of a stopping cell. They are easy to tune, provide transport efficiencies
871
close to unity and offer a variety of functionalities. The present work shows
872
that RFQ mass filters can ideally be implemented in such an RFQ beam line.
873
At the LEBIT facility transmission losses from an RFQ mass filter to an RFQ
874
buncher were observed that increase with increasing mass resolving power of
875
the RFQ mass filter [73], and it was argued that RFQ mass filters are not well
876
suited to perform mass separation in a stopping cell system. This problem can
877
be avoided if the ions are cooled in a gas-filled RFQ directly after mass selec-
878
tion, as implemented in this work, rather than transmitting the ions through an
879
electrostatic beam line first before injection into an RFQ buncher.
880
The technology developed in this work is broadly applicable. The combina-
881
tion of the fragment separator FRS [74] at GSI and the FRS Ion Catcher setup
882
[41] installed at the final focus of the FRS enables precision experiments with
883
thermalized projectile and fission fragments. The present work has served as a
884
basis for the development of the RFQ beam line of the FRS Ion Catcher [42],
885
and the voltage control system for the mass filter is currently used for operation
886
of the mass filter at the FRS Ion Catcher [75]. The technology is also used in
887
the RFQ transport system in the MR-TOF-MS at the TITAN experiment at
888
TRIUMF [76] and it will be employed in the extraction system of the stopping
33
889
cell [77] for the Low Energy Branch of the Super-FRS [78] at FAIR. Further-
890
more, a duplicate of the RFQ mass filter was built and successfully provided
891
the necessary mass number selection in an experiment, in which the decay of
892
the nuclear isomer
893
Acknowledgements
229m
Th was directly detected for the first time [79, 80].
894
We would like to thank R. Weiß and the members of the machine shop of
895
the physics institutes of the Justus Liebig University Gießen for help with the
896
construction of the RFQ system and for the careful machining of the mechanical
897
components. B. Fabian, R. Thöt and J. Werner are acknowledged for their
898
contributions to the simulations, the sytem control software and for studies on
899
the buncher design. We thank M. Block and F. Herfurth for fruitful discussions
900
during the development of the RFQ system. This work was supported by the
901
Helmholtz Association of German Research Centres (HGF) and by GSI under
902
contract no. VH-NG-033 and by the German Federal Ministry for Education
903
and Research (BMBF) under contract no. 06GI185I.
904
References
905
906
[1] S. Hofmann, G. Münzenberg, The discovery of the heaviest elements, Rev.
Mod. Phys. 72 (2000) 733–767.
907
[2] A. Gillitzer, T. Faestermann, K. Hartel, P. Kienle, E. Nolte, Groundstate
908
proton radioactivity of nuclei in the tin region, Z. Phys. A 326 (1987)
909
107–119.
910
[3] R. D. Page, P. J. Woods, R. A. Cunningham, T. Davinson, N. J. Davis,
911
A. N. James, K. Livingston, P. J. Sellin, A. C. Shotter, Alpha radioactivity
912
above 100 Sn including the decay of 108 I, Phys. Rev. C 49 (1994) 3312–3315.
913
[4] C. B. Hinke, M. Böhmer, P. Boutachkov, T. Faestermann, H. Geissel,
914
J. Gerl, R. Gernhäuser, M. Gorska, A. Gottardo, H. Grawe, J. L. Gre-
915
bosz, R. Krücken, N. Kurz, Z. Liu, L. Maier, F. Nowacki, S. Pietri,
34
916
Z. Podolyak, K. Sieja, K. Steiger, K. Straub, H. Weick, H.-J. Woller-
917
sheim, P. J. Woods, N. A.-D. amd N. Alkhomashi, A. Atac, A. Blazhev,
918
N. F. Braun, I. T.Celikovic, T. Davinson, I. Dillmann, C. Domingo-Pardo,
919
P. C. Doornenbal, G. de France, G. F. Farrelly, F. Farinon, N. Goel,
920
T. C. Habermann, R. Hoischen, R. Janik, M. Karny, A. Kaskas, I. M.
921
Kojouharov, T. Kröll, Y. Litvinov, S. Myalski, F. Nebel, S. Nishimura,
922
C. Nociforo, J. Nyberg, A. R. Parikh, A. Prochazka, P. H. Regan, C. Rigol-
923
let, H. Schaffner, C. Scheidenberger, S. Schwertel, P.-A. Söderström, S. J.
924
Steer, A. Stolz, P. Strmen, Superallowed Gamow-Teller decay of the doubly
925
magic nucleus
100
Sn, Nature 486 (2012) 341–345.
926
[5] I. Mukha, E. Roeckl, L. Batist, A. Blazhev, J. Dor̈ing, H. Grawe, L. Grig-
927
orenko, M. Huyse, Z. Janas, R. Kirchner, M. La Commara, C. Mazzocchi,
928
S. L. Tabor, P. Van Duppen, Proton-proton correlations observed in two-
929
proton radioactivity of
94
Ag, Nature 439 (2006) 298–302.
930
[6] P. V. Isacker, D. D. Warner, D. S. Brenner, Test of Wigner’s spin-isospin
931
symmetry from double binding energy differences, Phys. Rev. Lett. 74
932
(1995) 4607–4610.
933
[7] A. S. Lalleman, G. Auger, W. Mittig, M. Chabert, M. Chartier, J. Fermé,
934
A. Gillibert, Mass measurements of exotic nuclei around N =Z=40 with
935
CSS2, Hyperfine Interact. 132 (2001) 315–322.
936
[8] H. Schatz, A. Aprahamian, J. Görres, M. Wiescher, T. Rauscher, J. F.
937
Rembges, F.-K. Thielemann, B. Pfeiffer, P. Möller, K.-L. Kratz, H. Herndl,
938
B. A. Brown, H. Rebel, rp-process nucleosynthesis at extreme temperature
939
and density conditions, Physics Reports 294 (1998) 167–263.
940
[9] C. Fröhlich, G. Martinez-Pinedo, M. Liebendörfer, F.-K. Thielemann,
941
E. Bravo, W. R. Hix, K. Langanke, N. T. Zinner, Neutrino-induced nucle-
942
osynthesis of A > 64 nuclei: The νp process, Phys. Rev. Lett. 96 (2006)
943
142502.
35
944
[10] M. Block, D. Ackermann, K. Blaum, A. Chaudhuri, Z. Di, S. Eliseev,
945
R. Ferrer, D. Habs, F. Herfurth, F. P. Heßberger, S. Hofmann, H.-J.
946
Kluge, G. Maero, A. Martin, G. Marx, M. Mazzocco, M. Mukherjee, J. B.
947
Neumayr, W. R. Plaß, W. Quint, S. Rahaman, C. Rauth, D. Rodriguez,
948
C. Scheidenberger, L. Schweikhard, P. G. Thirolf, G. Vorobjev, C. Weber,
949
Towards direct mass measurements of nobelium at SHIPTRAP, Eur. Phys.
950
J. D 45 (2007) 39–45.
951
[11] G. Münzenberg, W. Faust, S. Hofmann, P. Armbruster, K. Güttner,
952
H. Ewald, The velocity filter SHIP, a separator of unslowed heavy ion
953
fusion products, Nucl. Instrum. Methods 161 (1979) 65–82.
954
[12] M. Block, D. Ackermann, K. Blaum, C. Droese, M. Dworschak, S. Eliseev,
955
T. Fleckenstein, E. Haettner, F. Herfurth, F. P. Heßberger, S. Hofmann,
956
J. Ketelaer, J. Ketter, H.-J. Kluge, G. Marx, M. Mazzocco, Y. N. Novikov,
957
W. R. Plaß, A. Popeko, S. Rahaman, D. Rodriguez, C. Scheidenberger,
958
L. Schweikhard, P. G. Thirolf, G. K. Vorobjev, C. Weber, Direct mass
959
measurements above uranium bridge the gap to the island of stability, Na-
960
ture 463 (2010) 785–788.
961
[13] E. M. Ramirez, D. Ackermann, K. Blaum, M. Block, C. Droese, C. E.
962
Düllmann, M. Dworschak, M. Eibach, S. Eliseev, E. Haettner, F. Her-
963
furth, F. P. Heßberger, S. Hofmann, J. Ketelaer, G. Marx, M. Mazzocco,
964
D. Nesterenko, Y. N. Novikov, W. R. Plaß, D. Rodriguez, C. Scheiden-
965
berger, L. Schweikhard, P. G. Thirolf, C. Weber, Direct Mapping of Shell
966
Effects in the Heaviest Elements, Science 337 (2012) 1207–1210.
967
[14] C. Rauth, D. Ackermann, K. Blaum, M. Block, A. Chaudhuri, Z. Di,
968
S. Eliseev, R. Ferrer, D. Habs, F. Herfurth, F. P. Heßberger, S. Hofmann,
969
H.-J. Kluge, G. Maero, A. Martin, G. Marx, M. Mukherjee, J. B. Neumayr,
970
W. R. Plaß, S. Rahaman, D. Rodriguez, C. Scheidenberger, L. Schweikhard,
971
P. G. Thirolf, G. Vorobjev, C. Weber, First Penning trap mass measure-
972
ments beyond the proton drip line, Phys. Rev. Lett. 100 (2008) 012501.
36
973
[15] E. Haettner, D. Ackermann, G. Audi, K. Blaum, M. Block, S. Eliseev,
974
T. Fleckenstein, F. Herfurth, F. P. Heßberger, S. Hofmann, J. Ketelaer,
975
J. Ketter, H.-J. Kluge, G. Marx, M. Mazzocco, Y. N. Novikov, W. R. Plaß,
976
S. Rahaman, T. Rauscher, D. Rodrı́guez, H. Schatz, C. Scheidenberger,
977
L. Schweikhard, B. Sun, P. G. Thirolf, G. Vorobjev, M. Wang, C. Weber,
978
Mass measurements of very neutron-deficient Mo and Tc isotopes and their
979
impact on rp process nucleosynthesis, Phys. Rev. Lett. 106 (2011) 122501.
980
[16] D. Rodrı́guez Rubiales, An RFQ buncher for accumulation and cooling of
981
heavy radionuclides at SHIPTRAP and high precision mass measurements
982
on unstable Kr isotopes at ISOLTRAP, PhD thesis, Universidad de Valen-
983
cia - CSIC, 2003.
984
985
[17] W. M. Brubaker, An improved quadrupole mass analyser, Advances in
mass spectrometry 4 (1968) 293–299.
986
[18] A. Loboda, A. Krutchinsky, O. Loboda, J. McNabb, V. Spicer, W. Ens,
987
K. Standing, Novel linac II electrode geometry for creating an axial field
988
in a multipole ion guide, Eur. J. Mass Spectrom. 6 (2000) 531–536.
989
[19] J. Neumayr, L. Beck, D. Habs, S. Heinz, J. Szerypo, P. Thirolf,
990
V. Varentsov, F. Voit, D. Ackermann, D. Beck, M. Block, Z. Di,
991
S. Eliseev, H. Geissel, F. Herfurth, F. Heßberger, S. Hofmann, H.-J.
992
Kluge, M. Mukherjee, G. Münzenberg, M. Petrick, W. Quint, S. Rahaman,
993
C. Rauth, D. Rodriguez, C. Scheidenberger, G. Sikler, Z. Wang, C. We-
994
ber, W. Plaß, M. Breitenfeldt, A. Chaudhuri, G. Marx, L. Schweikhard,
995
A. Dodonov, Y. Novikov, M. Suhonen, The ion-catcher device for SHIP-
996
TRAP, Nucl. Instrum. Methods B 244 (2006) 489–500.
997
[20] G. Savard, S. Becker, G. Bollen, H. J. Kluge, R. B. Moore, T. Otto,
998
L. Schweikhard, H. Stolzenberg, U. Wiess, A new cooling technique for
999
heavy ions in a Penning trap, Phys. Lett. A 158 (1991) 247–252.
1000
1001
[21] G. Gräff, H. Kalinowsky, J. Traut, A direct determination of the proton
electron mass ratio, Z. Physik A 297 (1980) 35–39.
37
1002
[22] M. König, G. Bollen, H.-J. Kluge, T. Otto, J. Szerypo, Quadrupole exci-
1003
tation of stored ion motion at the true cyclotron frequency, Int. J. Mass
1004
Spectrom. Ion Proc. 142 (1995) 95–116.
1005
[23] S. Eliseev, K. Blaum, M. Block, C. Droese, M. Goncharov, E. M. Ramirez,
1006
D. A. Nesterenko, Y. N. Novikov, L. Schweikhard, Phase-imaging ion-
1007
cyclotron-resonance measurements for short-lived nuclides,
1008
Lett. 110 (2013) 082501.
1009
Phys. Rev.
[24] W. Gelletly, M. Bentley, H. Price, J. Simpson, C. Gross, J. Durell, B. Var-
1010
ley, O. Skeppstedt, S. Rastikerdar, The observation of
1011
B 253 (1991) 287 – 291.
84
Mo, Phys. Lett.
1012
[25] C. J. Lister, M. Campbell, A. A. Chishti, W. Gelletly, L. Goettig, R. Mo-
1013
scrop, B. J. Varley, A. N. James, T. Morrison, H. G. Price, J. Simpson,
1014
K. Connel, O. Skeppstedt, Gamma radiation from the N=Z nucleus 80
40 Zr40 ,
1015
Phys. Rev. Lett. 59 (1987) 1270–1273.
1016
[26] N. Marginean, C. Rossi Alvarez, D. Bucurescu, Y. Sun, C. A. Ur, A. Gadea,
1017
S. Lunardi, D. Bazzacco, G. de Angelis, M. Axiotis, M. De Poli, E. Farnea,
1018
M. Ionescu-Bujor, A. Iordachescu, S. M. Lenzi, T. Kröll, T. Martinez,
1019
R. Menegazzo, D. R. Napoli, G. Nardelli, P. Pavan, B. Quintana, P. Spo-
1020
laore, First observation of excited states in the Tz = 1/2 nucleus
1021
Phys. Rev. C 65 (2002) 034315.
85
Mo,
1022
[27] D. Rudolph, F. Cristancho, C. J. Gross, A. Jungclaus, K. P. Lieb, M. A.
1023
Bentley, W. Gelletly, J. Simpson, H. Grawe, J. Heese, K.-H. Maier,
1024
J. Eberth, S. Skoda, J. L. Durell, B. J. Varley, D. J. Blumenthal, C. J.
1025
Lister, S. Rastikerdar, First observation of the nuclei
1026
Phys. G 17 (1991) 113–119.
87
Tc and
88
Tc, J.
1027
[28] S. A. Eliseev, M. Block, A. Chaudhuri, Z. Di, D. Habs, F. Herfurth, H.-J.
1028
Kluge, J. B. Neumayr, W. R. Plaß, C. Rauth, P. G. Thirolf, G. Vorobjev,
1029
Z. Wang, Extraction efficiency and extraction time of the SHIPTRAP
1030
gas-filled stopping cell, Nucl. Instrum. Methods B 258 (2007) 479–484.
38
1031
[29] M. Mazzocco, D. Ackermann, M. Block, H. Geissel, F. Herfurth, F. P.
1032
Heßberger, S. Hofmann, N. Iwasa, K. Nishio, W. R. Plaß, C. Scheiden-
1033
berger, H. Weick, M. Winkler, MOCADI-FUSION: Extension of the Monte-
1034
Carlo code MOCADI to heavy-ion fusion-evaporation reactions, Nucl. In-
1035
strum. Methods B 266 (2008) 3467–3480.
1036
1037
1038
1039
[30] W. Paul, H. P. Reinhard, U. von Zahn, Das elektrische Massenfilter als
Massenspektrometer und Isotopentrenner, Z. Phys. 152 (1958) 143–182.
[31] P. H. Dawson, Quadrupole Mass Spectrometry and its applications, Elsevier Scientific Publishing Company, Amsterdam-Oxford-New York, 1976.
1040
[32] J. C. Schwartz, M. W. Senko, J. E. P. Syka, A two-dimensional quadrupole
1041
ion trap mass spectrometer, J. Am. Soc. Mass Spectrom. 13 (2002) 659–
1042
669.
1043
1044
[33] D. J. Douglas, A. J. Frank, D. Mao, Linear ion traps in mass spectrometry,
Mass Spectrom. Rev. 24 (2005) 1–29.
1045
[34] P. Schury, G. Bollen, M. Block, D. Morrissey, R. Ringle, A. Prinke, J. Sa-
1046
vory, S. Schwarz, T. Sun, Beam purification techniques for low energy rare
1047
isotope beams from a gas cell, Hyperfine Interact. 173 (2006) 165–170.
1048
[35] J. Werner, Flugzeitmassenspektrometrie und Designstudie für einen neuen
1049
RF-Quadrupol-Buncher bei SHIPTRAP, Bachelor thesis, Justus Liebig
1050
University Gießen, 2008.
1051
[36] W. R. Plaß, T. Dickel, M. Petrick, D. Boutin, Z. Di, T. Fleckenstein,
1052
H. Geissel, C. Jesch, C. Scheidenberger, Z. Wang, An RF quadrupole-
1053
time-of-flight system for isobar-separation and multiplexed low energy rare-
1054
isotope beam experiments, Eur. Phys. J. Special Topics 150 (2007) 367–368.
1055
[37] W. R. Plaß, M. Petrick, T. Dickel, T. Fleckenstein, H. Geissel, C. Jesch,
1056
C. Scheidenberger, Novel concept for a beam preparation and manipulation
1057
system for high precision experiments at the LEB of the Super-FRS, GSI
1058
Scientific Report 2006 (2007) 39.
39
1059
[38] M. Petrick, Aufbau und Erprobung einer neuartigen Separationsmethode
1060
für exotische Kerne bei relativistischen Energien, PhD thesis, Justus Liebig
1061
University Gießen, 2008.
1062
[39] W. R. Plaß, T. Dickel, S. Ayet San Andres, J. Ebert, F. Greiner, C. Hor-
1063
nung, C. Jesch, J. Lang, W. Lippert, T. Majoros, D. Short, H. Geissel,
1064
E. Haettner, M. P. R. amd A.-K. Rink, C. Scheidenberger, M. I. Yavor,
1065
High-performance multiple-reflection time-of-flight mass spectrometers for
1066
research with exotic nuclei and for analytical mass spectrometry, Phys.
1067
Scr. T 166 (2015) 014069.
1068
[40] F. Greiner, Collision-induced Dissociation - An Effective Method for Re-
1069
moving Molecular Contaminants in Beams of Short-lived Nuclei (in Ger-
1070
man), Master thesis, Justus Liebig University Gießen, 2017.
1071
[41] W. R. Plaß, T. Dickel, S. Purushothaman, P. Dendooven, H. Geissel,
1072
J. Ebert, E. Haettner, C. Jesch, M. Ranjan, M. P. Reiter, H. Weick, F. Am-
1073
jad, S. Ayet, M. Diwisch, A. Estrade, F. Farinon, F. Greiner, N. Kalantar-
1074
Nayestanaki, R. Knöbel, J. Kurcewicz, J. Lang, I. Moore, I. Mukha, C. No-
1075
ciforo, M. Petrick, M. Pfuetzner, S. Pietri, A. Prochazka, A.-K. Rink,
1076
S. Rinta-Antila, D. Schäfer, C. Scheidenberger, M. Takechi, Y. K. Tanaka,
1077
J. S. Winfield, M. I. Yavor, The FRS Ion Catcher - a facility for high-
1078
precision experiments with stopped projectile and fission fragments, Nucl.
1079
Instrum. Methods B 317 (2013) 457–462.
1080
[42] M. P. Reiter, Pilot Experiments with Relativistic Uranium Projectile and
1081
Fission Fragments Thermalized in a Cryogenic Gas-Filled Stopping Cell,
1082
PhD thesis, Justus Liebig University Gießen, 2015.
1083
[43] E. Haettner, A novel radio frequency quadrupole system for SHIPTRAP
1084
& New mass measurements of rp nuclides, Ph.D. thesis, Justus Liebig Uni-
1085
versity Gießen, 2011.
1086
1087
[44] D. J. Douglas, Linear quadrupoles in mass spectrometry, Mass Spectrom.
Rev. 28 (2009) 937–960.
40
1088
[45] P. H. Dawson, Quadrupole Mass Spectrometry and its applications, Else-
1089
vier Scientific Publishing Company, Amsterdam-Oxford-New York, 1976.
1090
Chap. II, Editor: P. H. Dawson.
1091
1092
[46] D. A. Dahl, SIMION for the personal computer in reflection, Int. J. Mass
Spectrom. 200 (2000) 3 – 25.
1093
[47] W. R. Plaß, The Dependence of RF Ion Trap Mass Spectrometer Per-
1094
formance on Electrode Geometry and Collisional Processes, PhD thesis,
1095
Justus Liebig University Gießen, 2001.
1096
[48] W. R. Plaß, T. Dickel, B. Fabian, E. Haettner, An advanced concept for
1097
mass spectrometry simulations – the Ion Trajectory Simulation program
1098
ITSIM 6, in: Proceedings of the 56th ASMS Conference on Mass Spec-
1099
trometry and Allied Topics, Denver, USA, 2008, p. MP 015.
1100
[49] COMSOL Multiphysics, http://www.comsol.de, 2010.
1101
[50] J. J. Tunstall, A. C. C. Voo, S. Taylor, Computer simulation of the mass
1102
filter for a finite length quadrupole, Rapid Commun. Mass Spectrom. 11
1103
(1997) 184–188.
1104
[51] W. E. Austin, A. E. Holme, J. H. Leck, Quadrupole Mass Spectrometry
1105
and its applications, Elsevier Scientific Publishing Company, Amsterdam-
1106
Oxford-New York, 1976. Chap. VI, Editor: P. H. Dawson.
1107
[52] W. Plass, L. Gill, H. Bui, R. Cooks, Ion mobility measurements by Dc
1108
tomography in an Rf quadrupole ion trap, J. Phys. Chem. A 104 (2000)
1109
5059–5065.
1110
[53] A. J. Reuben, G. B. Smith, P. Moses, A. V. Vagov, M. D. Woods, D. B.
1111
Gordon, R. W. Munn, Ion trajectories in exactly determined quadrupole
1112
fields, Int. J. Mass Spectrom. Ion Proc. 154 (1996) 43 – 59.
1113
[54] J. R. Gibson, S. Taylor, Numerical investigation of the effect of electrode
1114
size on the behaviour of quadrupole mass filters, Rapid Commun. Mass
1115
Spectrom. 15 (2001) 1960–1964.
41
1116
[55] D. J. Douglas, N. V. Konenkov, Influence of the 6th and 10th spatial
1117
harmonics on the peak shape of a quadrupole mass filter with round rods,
1118
Rapid Commun. Mass Spectrom. 16 (2002) 1425–1431.
1119
[56] R. E. Pedder, R. A. Schaeffer, Extrel application note RA 2001B: Resolu-
1120
tion and transmission of quadrupoles: RF-DC and RF-Only operation, in:
1121
43rd ASMS Conference on Mass Spectrometry and Allied Topics, Atlanta,
1122
Georgia, 1995.
1123
[57] J. R. Gibson, K. G. Evans, S. Taylor, Predicted behaviour of QMF systems
1124
with and without prefilters using accurate 3D fields, Int. J. Mass Spectrom.
1125
http://dx.doi.org/10.1016/j.ijms.2017.01.015 (2017).
1126
[58] W. R. Plaß, Aufbau eines Flugzeitmassenspektrometers zur Analyse von
1127
flüchtigen organischen Verbindungen bei geringem Partialdruck, Diploma
1128
thesis, Justus Liebig University Gießen, 1997.
1129
[59] T. Dickel, Design and Commissioning of an Ultra-High-Resolution Time-of-
1130
Flight Based Isobar Separator and Mass Spectrometer, PhD thesis, Justus
1131
Liebig University Gießen, 2010.
1132
[60] O. Gianfrancesco, Design principles of a high field RFQ device for ion cool-
1133
ing and confinement, Master thesis, McGill University, Montreal, Canada,
1134
2002.
1135
1136
[61] I. Cermak, Compact radio-frequency power supply for ion and particle
guides and traps, Rev. Sci. Instrum. 76 (2005) 063302.
1137
[62] P. Karvonen, H. Penttilä, J. Äystö, J. Billowes, P. Campbell, V.-V. Elo-
1138
maa, U. Hager, J. Hakala, A. Jokinen, T. Kessler, A. Kankainen, I. Moore,
1139
K. Peräjärvi, S. Rahaman, S. Rinta-Antila, J. Rissanen, J. Ronkainen,
1140
A. Saastamoinen, T. Sonoda, B. Tordoff, C. Weber, Upgrade and yields of
1141
the IGISOL facility, Nucl. Instrum. Methods B 266 (2008) 4454 – 4459.
1142
1143
[63] High Resolution Power at Low Masses, Application Note, Extrel CMS,
2016.
42
1144
1145
[64] TSQ Quantum Ultra/Ultra AM/Ultra EMR, Product Specifications,
Thermo Fisher Scientific, Inc., San Jose, USA, 2010.
1146
[65] A. Becker, Construction and characterization of a system for measuring
1147
the transmission efficiency of ion optical systems, Bachelor thesis, Justus
1148
Liebig University Gießen, 2008.
1149
[66] T. Dickel, W. R. Plaß, A. Becker, U. Czok, H. Geissel, E. Haettner,
1150
C. Jesch, W. Kinsel, M. Petrick, C. Scheidenberger, M. I. Yavor, A high-
1151
performance multiple-reflection time-of-flight mass spectrometer and isobar
1152
separator for the research with exotic nuclei, Nucl. Instrum. Methods A
1153
777 (2015) 172–188.
1154
1155
[67] W. C. Wiley, I. H. McLaren, Time-of-flight mass spectrometer with improved resolution, Rev. Sci. Instrum. 26 (1955) 1150–1157.
1156
[68] J. L. Franklin, P. M. Hierl, D. A. Whan, Measurement of the translational
1157
energy of ions with a time-of-flight mass spectrometer, J. Chem. Phys. 47
1158
(1967) 3148–3153.
1159
[69] F. Herfurth, Ein neuer Ionenstrahlkühler und -pulser für ISOLTRAP und
1160
Massenmessungen an radioaktiven Argon isotopen, PhD thesis, Ruprecht-
1161
Karls-University, Heidelberg, 2001.
1162
[70] V. S. Kolhinen, S. Kopecky, T. Eronen, U. Hager, J. Hakala, J. Huikari,
1163
A. Jokinen, A. Nieminen, S. Rinta-Antila, J. Szerypo, J. Äystö,
1164
JYFLTRAP: a cylindrical Penning trap for isobaric beam purification at
1165
IGISOL, Nucl. Instrum. Methods A 528 (2004) 776 – 787.
1166
1167
[71] S. Schwarz, G. Bollen, R. Ringle, J. Savory, P. Schury, The LEBIT ion
cooler and buncher, Nucl. Instrum. Methods A 816 (2016) 131–141.
1168
[72] W. R. Plaß, T. Dickel, U. Czok, H. Geissel, M. Petrick, K. Reinheimer,
1169
C. Scheidenberger, M. I. Yavor, Isobar separation by time-of-flight mass
1170
spectrometry for low-energy radioactive ion beam facilities, Nucl. Instrum.
1171
Methods B 266 (2008) 4560–4564.
43
1172
[73] R. Ferrer, A. A. Kwiatkowski, G. Bollen, D. L. Lincoln, D. J. Morrissey,
1173
G. K. Pang, R. Ringle, J. Savory, S. Schwarz, Ion beam properties after
1174
mass filtering with a linear radiofrequency quadrupole, Nucl. Inst. Methods
1175
A 735 (2014) 382–389.
1176
[74] H. Geissel, P. Armbruster, K. H. Behr, A. Brünle, K. Burkard, M. Chen,
1177
H. Folger, B. Franczak, H. Keller, O. Klepper, B. Langenbeck, F. Nickel,
1178
E. Pfeng, M. Pfützner, E. Roeckl, K. Rykaczewski, I. Schall, D. Schardt,
1179
C. Scheidenberger, K. H. Schmidt, A. Schröter, T. Schwab, K. Sümmerer,
1180
M. Weber, G. Münzenberg, T. Brohm, H. G. Clerc, M. Fauerbach, J. J.
1181
Gaimard, A. Grewe, E. Hanelt, B. Knödler, M. Steiner, B. Voss, J. Weck-
1182
enmann, C. Ziegler, A. Magel, H. Wollnik, J. P. Dufour, Y. Fujita, D. J.
1183
Vieira, B. Sherrill, The GSI projectile fragment separator (FRS): a versa-
1184
tile magnetic system for relativistic heavy ions, Nucl. Instrum. Methods B
1185
70 (1992) 286 – 297.
1186
[75] I. Miskun, Commissioning, performance and possible applications of RFQ
1187
mass-filter in the diagnostic unit of the FRS Ion Catcher facility, Diploma
1188
thesis, Tomsk Polytechnical University, Russia, 2015.
1189
[76] C. Jesch, T. Dickel, W. R. Plaß, D. Short, S. Ayet, J. Dilling, H. Geissel,
1190
F. Greiner, J. Lang, K. G. Leach, W. Lippert, C. Scheidenberger, M. I.
1191
Yavor, The MR-TOF-MS isobar separator for the TITAN facility at TRI-
1192
UMF, Hyperfine Interact. 235 (2015) 97–106.
1193
[77] T. Dickel, W. R. Plaß, H. Geissel, F. Heiße, I. Miskun, S. Purushothaman,
1194
M. P. Reiter, A.-K. Rink, C. Scheidenberger, Conceptional design of a
1195
novel next-generation cryogenic stopping cell for the Low-Energy Branch
1196
of the Super-FRS, Nucl. Instrum. Methods B 376 (2016) 216–220.
1197
[78] H. Geissel, H. Weick, M. Winkler, G. Münzenberg, V. Chichkine, M. Yavor,
1198
T. Aumann, K. H. Behr, M. Böhmer, A. Brünle, K. Burkard, J. Benlliure,
1199
D. Cortina-Gil, L. Chulkov, A. Dael, J.-E. Ducret, H. Emling, B. Franczak,
1200
J. Friese, B. Gastineau, J. Gerl, R. Gernhäuser, M. Hellström, B. Jonson,
44
1201
J. Kojouharova, R. Kulessa, B. Kindler, N. Kurz, B. Lommel, W. Mit-
1202
tig, G. Moritz, C. Mühle, J. A. Nolen, G. Nyman, P. Roussell-Chomaz,
1203
C. Scheidenberger, K.-H. Schmidt, G. Schrieder, B. Sherrill, H. Simon,
1204
K. Sümmerer, N. A. Tahir, V. Vysotsky, H. Wollnik, A. F. Zeller, The
1205
Super-FRS project at GSI, Nucl. Instrum. Methods B 204 (2003) 71–85.
1206
[79] L. von der Wense, P. G. Thirolf, D. Kalb, M. Laatiaoui, Towards a direct
1207
transition energy measurement of the lowest nuclear excitation in
1208
J. Instrum. 8 (2013) P03005.
229
Th,
1209
[80] L. von der Wense, B. Seiferle, M. Laatiaoui, J. B. Neumayr, H.-J. Maier,
1210
H.-F. Wirth, C. Mokry, J. Runke, K. Eberhardt, C. E. Duellmann, N. G.
1211
Trautmann, P. G. Thirolf, Direct detection of the
1212
transition, Nature 533 (2016) 47–51.
45
229
Th nuclear clock
Calibrants
(a)
7 T magnet
Penning traps
thermalization
Stopping cell
Drift section ~10-7 mbar
50-100 mbar
He gas
(b)
B2 D2
200 mm
50 m m
<10-4 mbar
400 mm
Æ 5 mm
50 mm
5·10-3 mbar
460 mm
Figure 1: (a) Overview of the SHIPTRAP setup. The RFQ system developed in the present
work is shown together with the presently installed RFQ buncher [16] for comparison. Fusionevaporation products separated by the velocity filter SHIP are thermalized in the stopping cell
and extracted into the extraction RFQ. Currently, the ions are then cooled and bunched in
the presently installed RFQ buncher. Next, the ions are transmitted in ion packets through an
electrostatic beam line. In this beam line time-of-flight mass spectrometry can be performed
using MCP detectors (MCP1/2). A quadrupole deflector is installed in the beam line to
inject ions with well-known masses for calibration. Finally, the ions are mass separated in the
purification trap (PT) before a precision mass measurement is performed in the measurement
trap (MT). The RFQ buncher can be replaced by the new RFQ system of this work. (b)
Schematic figure of the new RFQ system. The system has an overall length of about one
meter and consists of an RFQ cooler, an RFQ mass filter with Brubaker lenses [17] at the
entrance (B1) and exit (B2), and an RFQ buncher. The three RFQs are separated by two
diaphragms D1 and D2. The cooler RFQ and the cooling section of the RFQ buncher use
LINAC rods [18] to create an axial field along the RFQs. The cooling section of the RFQ
buncher is followed by a segmented trapping section. The segments are numbered from S1 to
S3 as indicated in the figure. The RFQ system ends with two extraction diaphragms D3 and
D4. The geometrical parameters and typical pressures are indicated for the different sections
of the setup.
46
10 mm
Æ 2 mm
10 mm
2·10-2 mbar
D3 D4
S1 S2 S3
10 mm
D1 B1
Æ 3 mm
Transmission efficiency
1.0
0.8
0.6
Ion oscillation
amplitude
(1/6)r 0
10
-5
mbar
10
-4
mbar
10
-3
mbar
(1/3)r 0
(1/2)r 0
(2/3)r 0
0.4
(5/6)r 0
r0
0.2
0.0
0.700
0.704
q
0.700
0.708
0.704
q
0.708
0.700
0.704
0.708
0.704
0.708
q
Figure 2: Numerical simulation of the pressure dependence of the peak shapes in an ideal
RFQ mass filter (with a purely linear field) for different ion oscillation amplitudes (given in
units of the field radius r0 ). The relative abundance of transmitted ions is plotted versus the
Mathieu parameter q for a fixed ratio of U/V corresponding to a mass resolving power of 200.
The ions pass through a 30 cm long RFQ operated at an RF frequency of 1 MHz with an
axial velocity of 3000 m/s. The simulation was performed using 100 ions.
Transmission efficiency
1.0
0.8
0.6
Ion oscillation
amplitude
(1/6)r 0
R/r0=1.11
R/r0=1.125
R/r0=1.14
(1/3)r 0
(1/2)r 0
(2/3)r 0
0.4
(5/6)r 0
r0
0.2
0.0
0.700
0.704
q
0.708
0.700
0.704
q
0.708
Figure 3: Numerical simulation of the peak shapes in an RFQ mass filter using round rods
for R/r0 ratios of 1.11, 1.125 and 1.14 and for different ion oscillation amplitudes (given in
units of the field radius r0 ). The relative abundance of transmitted ions is plotted versus the
Mathieu parameter q for a fixed ratio of U/V corresponding to a mass resolving power of 170.
The ions pass through a 30 cm long RFQ operated at an RF frequency of 1 MHz with an
axial velocity of 3000 m/s. The simulation was performed using 100 ions.
47
0.700
q
Relative transmission efficiency
with
1.0
without
Brubaker lens
0.8
at entrance
0.6
0.4
0.2
0.0
0.700
0.704
0.708
q
Figure 4: Numerical simulation of the peak shapes in an RFQ mass filter with (solid black
curve) and without (dashed red curve) Brubaker lenses at the entrance. The mass filter has
a length of 30 cm and the Brubaker lenses a length of 2.5 cm. The simulation was performed
using 500 ions.
48
I
II
III
Figure 5: Photograph of the experimental setup. The vacuum chambers contain (from left to
right) (I) the ion source and the RFQ cooler, (II) the RFQ mass filter, (III) the RFQ buncher
and (IV) a detector. Two crates containing the resonance circuits and box with an RF/DC
converter are placed on the top of chambers II and III. Power supplies, frequency generators
and gas supply are placed on both sides of the setup. A computer with control software for
the mass filter and data acquisition is found in the lower right side of the photograph.
49
IV
Energy spread (FWHM) / eV
1
0.1
Experimental limit
(
K )
0
Thermal limit (300 K)
0.01
0
-2
-2
2x10
4x10
-2
6x10
-2
8x10
-1
1x10
Pressure / mbar
Figure 6: Measured energy spread of
133 Cs+
ions as a function of the He buffer gas pressure
in the RFQ cooler. The solid curve is a fit of Eq. (6) to the experimental data points. It
yields a minimum energy spread of about 0.1 eV limited by the field penetration through
the diaphragm D1 and by small asymmetries in the RF voltages. The intersection of the
extrapolation of the initial slope (dashed line) and the thermal limit (dotted line) gives the
pressure required for optimum cooling.
50
Relative transmission efficiency
1.0
Brubaker lenses at:
entrance
+exit
0.8
entrance
only
0.6
exit only
0.4
no
0.2
0.0
805
810
815
820
825
830
835
Nominal RF voltage / V
Figure 7: Measured peak shapes of the mass filter for (i) Brubaker lenses at the entrance and
at the exit, (ii) Brubaker lens at the entrance only, (iii) Brubaker lens at the exit only and
(iv) no Brubaker lenses. The transmission efficiency is given relative to that in RF-only mode.
The measurement was performed in the optimization phase of the RFQ system that may have
lead to peak shifts and broadening.
51
Relative transmission efficiency
(a)
133
+
Cs
1.0
0.8
0.6
0.4
m=
FWHM
0.2
10%
0.0
10
100
1000
Mass resolving power (m/ m)
(b)
Relative transmission efficiency
39
+
K
1.0
0.8
0.6
0.4
m=
FWHM
0.2
10%
0.0
10
100
1000
Mass resolving power (m/ m)
Figure 8: Measured transmission efficiency as a function of the mass resolving power for (a)
133 Cs+
and (b)
39 K+ .
The mass resolving power was evaluated at full width half maximum
(FWHM) and at 10% of the peak height. The transmission efficiency is given relative to that
in RF-only mode. The lines are drawn to guide the eye.
52
(a )
106
133
Cs+
5
Rate / (1/s)
10
104
134,135,136,137,138
Ba+
103
102
101
100
10-1
10-2
Rate / (1/s)
(b)
1
10
10
135
Ba+
0
10 -1
10-2
-3
10
Ba+
136
134
136
134
10-4
128
130
132
Ba+
138
140
Mass-to-charge ratio / (u/e0)
Figure 9: (a) Mass spectrum of ions from a source containing barium isotopes and
133 Cs
obtained using the RFQ mass filter. The mass spectrum was measured at a mass resolving
power of 240 (FWHM) and a transmission efficiency of 90%. The red curve shows a fit of the
sum of six Gaussian functions to the data. The green curves show the Gaussian functions for
each individual mass line. A suppression by four and by more than eight orders of magnitude is
demonstrated for mass lines separated by one and two mass units, respectively. (b) Abundance
distribution calculated from the measured spectrum shown in panel (a) for the case that the
RFQ mass filter is tuned to select
135 Ba+ .
For details see text.
53
400
Time spread (FWHM) / ns
350
300
250
200
150
100
50
0
4
6
8
10
12
Cooling time / ms
Figure 10: Measured time spread of
133 Cs+
ions extracted from the RFQ buncher (squares)
after being cooled for a time, t, at segment S3 in helium buffer gas at a pressure of (5±2)×10−3
mbar. The solid curve shows a fit of Eq. (7) to the data. The cooling constant τ amounts to
(1.3 ± 0.1) ms.
25
Energy spread (FWHM) / eV
Experiment
Theory
20
15
10
Acceptance limit of
Penning trap
5
40 ns
0
50
100
150
200
Time spread (FWHM) / ns
Figure 11: Measured kinetic energy spread as a function of the time spread of the extracted
ion packets. The curve gives the theoretical dependence according to Eqs. (8) and (9), for a
packet of
133 Cs+
ions, which has a spatial spread of 0.8 mm and a temperature of 410 K. The
measured emittance is (0.45 ± 0.05) eVµs, and close to the theoretical (thermal) limit.
54
Highlights (for review)





A novel compact triple RF quadrupole system has been developed.
It extends the reach of rare-isotope facilities with stopping cells.
It implements ion cooling, identification, mass separation and bunching.
A mass resolving power (FWHM) of 240 at a transmission of 90% has been
achieved.
A longitudinal emittance of 0.45 eV µs has been obtained.
1
Документ
Категория
Без категории
Просмотров
7
Размер файла
7 493 Кб
Теги
niman, 003, 2017
1/--страниц
Пожаловаться на содержимое документа