1. Improved Design for an ECR Charge Breeder of Radioactive Beams
4/24/2017
Improved Design for an ECR Charge Breeder of Radioactive Beams
Richard Vondrasek
ATLAS, Physics Division
Argonne National Laboratory
Conception of ECR charge breeding
Implementation at Argonne
Charge breeding investigations
Test

2. ATLAS – Argonne Tandem Linac Accelerator System
4/24/2017
ATLAS – Argonne Tandem Linac Accelerator System
New CARIBU Space
18 x 8.7 m
10 years ago a proposal was put forward to expand the capabilities of the ATLAS accelerator into the realm of radioactive beams - without the need for a full blown ISOL facility. Instead, the radioactive fission fragments would be provided by a Cf-252 fission source thus allowing the entire system to be housed in a small addition at the front end of the accelerator.
Test
R. Vondrasek ICIS August 27, 2015

3. CARIBU – Californium Rare Ion Breeder Upgrade
4/24/2017
CARIBU – Californium Rare Ion Breeder Upgrade
GAS CATCHER
ISOBAR SEPARATOR
1:20,000
TO ECR CHARGE BREEDER
TO LOW ENERGY EXPERIMENTS
1+ SOURCE HEAD
ECR CHARGE BREEDER
To low energy area
252Cf source, gas catcher, isobar separator
ECRCB ion source
Stable beam platform
252Cf fission source provides radioactive species
T1/2=2.6 a % fission branch
1.7 Ci source installed May 2014
Helium gas catcher + µRFQ’s
Energy spread of 1 eV
Emittance of 3·π·mm·mrad
Stopped beams and reaccelerated beams up to 15 MeV/u
Highest yields are in the mid-mass species
The project is called CARIBU and it was the motivation for the development of our ECR charge breeder. Fission fragments from a Cf-252 source are thermalized in a large volume helium gas catcher and extracted as 1+ or 2+ ions. The good beam properties allow a typical mass resolution of 1: The ions are directed to either stopped beam experiments or to an ECR source for charge breeding and subsequent acceleration in the ATLAS linac.
The components of the CARIBU system are housed on a 200 kV high voltage platform. This platform is coupled to a second platform which houses the ECR charge breeder together with a stable 1+ source used for charge breeder development as well as providing a guide beams for setting up the accelerator.
252Cf fission distribution
Test
R. Vondrasek ICIS’13 Chiba, Japan

4. 4/24/2017
1+  n+ conception
But of course our ECR development did not start at zero. There already existed a large body of charge breeding work upon which we built. Richard Geller pioneered the 1+/n+ method in The initial scheme had the 1+ ions injected into the ECR via the extraction aperture. This method was later refined to the much easier forward injection that has become the standard configuration. He made observations of the required electrostatic potentials and their effect on the charge breeding efficiency.
Test

5. 4/24/2017
LPSC charge breeder
The Grenoble group continued to be at the forefront of charge breeding development demonstrating high efficiencies for argon and neon and beginning to show indications of the performance disparity between solid and gaseous species.
Test
R. Vondrasek ICIS August 27, 2015

6. 4/24/2017
KEK charge breeder
The KEK group also developed a charge breeder and demonstrated charge breeding with radioactive species. Again showing a disparity in the capture efficiency between solid and gaseous species. They also started addressing the issue of background contamination – this continuum of seemingly endless m/q combinations - which greatly hindered the production of pure radioactive beams. By sandblasting and then high pressure rinsing their plasma chamber, they succeeded in lower the background contamination.
Test
R. Vondrasek ICIS August 27, 2015

7. TRIUMF and ISOLDE charge breeders
4/24/2017
TRIUMF and ISOLDE charge breeders
This ubiquitous background was seen by the TRIUMF and ISOLDE charge breeding groups as well, and led Thierry Lamy from Grenoble to declare we needed people and money to solve this problem.
Test
R. Vondrasek ICIS August 27, 2015

8. ECR charge breeder issues (in 2006)
4/24/2017
ECR charge breeder issues (in 2006)
Performance disparity between gases and solids
Factor of 2 to 3 better breeding efficiency for gaseous species
Low mass species demonstrated very poor efficiency
Background contamination
Obscures weak radioactive species
Possible improvements to charge breeder performance
Move towards an ultra-high vacuum system
‘Clean’ construction materials and techniques
Injection into ECR source
Beam optics and magnetic field optimization
Flexibility in tuning ion source
On-line movable grounded tube
Multiple frequency heating
Tunable RF sources
So by 2006 when we started our charge breeder design, several issues were already well known. There was a performance disparity between the gaseous and solid species. It was assumed this was due to the solid species being lost to the plasma once they hit the wall. The low mass species exhibited particularly low breeding efficiencies and this was attributed to an inadequate capture into the plasma. And then the afore mentioned background which was very troublesome.
Our existing ECR source which was going to be converted into a charge breeder was already operating in the 10-8 mbar regime even though it had a large number of o-rings on it. The chamber was constructed of 6062 aluminum as was the extractor. The one mistake we made was fabricating the transfer tube from stainless steel. This proved to be a source of iron contamination and was later changed to ultra-pure aluminum, as we did to the extractor as well. We modeled the ion injection to ensure good acceptance into the chamber and set up the optics to preserve the very low emittance beam expected from CARIBU. We made the transfer tube movable so that we could test the effect of its position on line. We also designed in two RF waveguides for multiple frequency heating which we were using to good effect in our day to day ECR operation.
Test
R. Vondrasek ICIS August 27, 2015

9. 4/24/2017
ANL charge breeder
Pre-conversion
Removed the central iron plug to allow for grounded tube penetration
Linear motion stage with 3 cm of travel
Reshaped the remaining iron to maximize Binj (1.3 T) and maintain symmetric magnetic field
Open hexapole allowed radial RF injection and provided direct pumping of plasma chamber
Turbo pumps at injection and extraction tanks of ion source
Multiple frequency operation
Klystron: GHz, 2 kW
TWTA: 1113 GHz, 0.5 kW
All 6061 aluminum components
So the changes we made to the existing source…
Radial launching has been used successfully
JYFL 6.4 GHz ion source
Compared axially and radially launched RF at the same power
Same beam production of O6+ ( W) for both cases
16th International Workshop on ECR Ion Sources
26-30 September 2004 Berkeley, CA
Grenoble Test Source (GTS)
Same beam production in both modes
Rev. Sci. Instrum., Vol. 71, No. 2
February 2000
Post-conversion
Test
R. Vondrasek ICIS August 27, 2015

10. 4/24/2017
ANL charge breeder
1+ or 2+
CARIBU beams
1+ Stable beams
65kV power supply
ΔV power supply n+
In order to achieve the 1:20000 design resolution of the isobar separator, we had to operate at 50 kV. So the breeder had to operate at 50 kV as well. The source can operate at 50 kV extraction potential but has more typically run at 36 kV. Stable beam were provided by either a surface ionization source or an RF discharge source with the beam currents measured on fully shielded faraday cups.
For the radioactive species, we utilized silicon barrier detectors with aluminum cover foils to detect the beta decay of the implanted radioactive species.
Stable beams from a surface ionization source or a RF discharge source
50 kV high voltage isolation
Stable beams: Faraday cups
Radioactive beams: Beta decay with silicon barrier detectors covered with aluminum foil
Test
R. Vondrasek ICIS August 27, 2015

11. 4/24/2017
First charge bred beam
First charge bred beam of 85Rb in May Then stopped in September 2008 while other aspects of the CARIBU program were completed in June 2009
Source was under vacuum the entire time and resulted in the operating pressure improving from 2.0x10-7 to 1.0x10-7 mbar
Peak of charge state distribution shifted from 15+ to 17+
Charge State
Efficiency (%)
2.0x10-7
1.0x10-7
10+
0.7 -
11+
0.8
13+
1.8
15+
3.8
17+
5.2
19+
3.2
20+
2.9
Our first charge bred beam of rubidium was in may we continued our charge breeding studies until september when we stopped in order to focus on other aspects of the CARIBU construction. When we returned to the source in June 2009, we found that the pressure in the source had decreased slightly but that this change was accompanied by a dramatic improvement in the breeding efficiency and charge state distribution. This was the first indication of just how critical the source operating pressure was.
Test
R. Vondrasek ICIS August 27, 2015

12. Injected ion penetration
4/24/2017
Injected ion penetration
Symmetric iron
We also saw benefits from the radially launched RF. By bringing the waveguides in between the hexapole bars, we maintained the magnetic field symmetry where the 1+ ions enter the chamber. The upper picture is looking upstream at the transfer tube and an aluminum disc which covers the shaped iron insert. The lower picture is from our other ECR source and demonstrates the cut out in the iron which are required for launching the RF axially.
Asymmetric iron
Looked at two cases
Symmetric iron with no cut outs for waveguides and RF is launched radially
Asymmetric iron which has cut outs for axial RF launch
Test
R. Vondrasek ICIS August 27, 2015

13. Injected ion penetration
4/24/2017
Injected ion penetration
0.98
0.99
1.00
1.0 -1 1
Symmetric iron
Asymmetric iron
We performed 3d calculations with CST studio for these two cases – same coil currents and hexapole field, same electric potentials, we only changed the particulars of the shaped iron insert. We then tracked the particles entering the plasma chamber. We did not take the plasma into account for these simulations. We were only interested in how the particles were affected by the change in the iron and hence the magnetic field.
133Cs+ ions, V1+: 30 kV, ΔV: -10 V
3D field calculations - Computer Simulation Technology Electromagnetic Studio
Simulations utilized running conditions for coils, hexapole, and potentials
No plasma or collision effects included in simulation
Test
R. Vondrasek ICIS August 27, 2015

14. Injected ion penetration
4/24/2017
Injected ion penetration
Symmetric iron cases
Asymmetric iron cases
What we found was that for all the ions we tested – Na, K, and Cs – in the case of the symmetric iron, a higher percentage of the ions penetrated further into the plasma chamber than in the asymmetric case. And that the lower mass species were more sensitive to the field asymmetry.
133Cs+ , 39K+ , 23Na+ ions
V1+: 30 kV, ΔV: -10 V
Test
R. Vondrasek ICIS August 27, 2015

15. Grounded tube position
4/24/2017
Grounded tube position
Efficiency of 85Rb17+ as a function of the grounded tube position
Position indicates the distance from Bmax to the end of the grounded tube
Three different data sets from 2008, 2009 and 2010
Bmax
Test
R. Vondrasek ICIS August 27, 2015

16. Multiple frequency heating
4/24/2017
Multiple frequency heating
Two-frequency heating with Xenon
129Xe+ beam intensity of 65 nA
85Rb20+
TWT frequency shift with Cesium
GHz 53 W W
GHz 73 W W
11.5%
10%
988
320
Kept GHz constant at 331 W, varied GHz power level
Test
R. Vondrasek ICIS August 27, 2015

17. Low-A performance
Potassium and sodium produced from surface ionization source
Base pressure: 3.7 x 10-8 mbar
Helium plasma: 9.0 x 10-8 mbar
Adding oxygen reduced breeding efficiency
Using He-3 vs. He-4 made no difference in efficiency
Helium plasma
5 s
Oxygen plasma
Response of 39K10+ beam current to pulsing an incoming K+ beam with an intensity of 17 enA. Beam is pulsed with electrostatic steerer immediately after the surface ionization source.
R. Vondrasek ICIS August 27, 2015

18. Charge breeding performance
Ion
Charge State
Efficiency (%)
A/Q
23Na 7+
10.1
3.29
39K
10+
17.9
3.90
84Kr
17+
15.6
4.94
85Rb
19+
13.7
4.47
110Ru (1+)
t1/2 = 11.6 s
22+
11.8
5.00
135Te (1+)
t1/2 = 19.0 s
26+
5.0
5.19
129Xe
25+
13.4
5.16
132Xe
27+
14.1
4.89
133Cs
14.7
5.11
13.5
4.93
141Cs (1+)
t1/2 = 24.8 s
12.3
5.22
142Cs (1+)
t1/2 = 1.69 s
7.3
5.26
143Cs (1+)
t1/2 = 1.79 s
11.7
5.30
143Ba (2+)
t1/2 = 14.3 s
144Ba (2+)
t1/2 = 11.5 s
28+
14.3
5.14
146Ba (2+)
t1/2 = 2.22 s
13.3
5.21
R. Vondrasek ICIS August 27, 2015

19. Charge breeding time tbreeding= tcreation + textraction tbreeding
4/24/2017
Charge breeding time
132Xe+
Electrostatic deflector
132Xe29+
tbreeding= tcreation + textraction
90%
1.3 sec
tbreeding
In order to measure the charge breeding time, we pulse the electrostatic steerer, with its trace shown in red, just after the 1+ source. We then record the rise of the n+ beam current, shown in black, take 90% of the saturation value, and define that as the breeding time. For this case of 132Xe29+ that time is 1.33 sec. And what we are defining as the charge breeding time is more accurately the total time it takes to both create the ion and extract it from the plasma. Based upon how the ion source is tuned, this time can vary tremendously.
Test
R. Vondrasek ICIS August 27, 2015

20. Charge breeding time – 132Xe
4/24/2017
Charge breeding time – 132Xe
Oxygen plasma, single-frequency heating – TWTA GHz
24+
23+
Intensity (mV)
Looking at all of the visible xenon charge states, we can see the pattern of step-wise ionization with the charge breeding time of the 27+ being 210 ms. We can also see jumps in the breeding times for the higher charge states which correspond to shell transitions. But if we make a slight change to the ion source, in this case we shift the operating frequency of the TWTA transmitter by 3 MHz, we see a dramatic change in the charge breeding time.
26+
21+
20+
27+
18+
17+
Time (s)
Test
R. Vondrasek ICIS August 27, 2015

21. Charge breeding time – 132Xe
4/24/2017
Charge breeding time – 132Xe
Oxygen plasma, single-frequency heating – TWTA GHz (Δf = 3 MHz)
24+
23+
26+
Intensity (mV)
Again, we are looking at all of the visible xenon charge states, and we see a pronounced change in the slopes of the breeding times and also note that the peak charge state has shifted from 24+ to 26+.
26+
27+
21+
20+
24+
27+
23+
18+
29+
17+
21+
20+
Time (s)
Test
R. Vondrasek ICIS August 27, 2015

22. Breeding efficiency and rise time – RF frequency
Xe-132 from RF discharge source
Oxygen plasma, single-frequency heating – TWTA only
Avg: 6 msec/q
R. Vondrasek ICIS August 27, 2015

23. Breeding efficiency and rise time – RF frequency
Xe-132 from RF discharge source
Oxygen plasma, single-frequency heating – TWTA only
Avg: 6 msec/q
R. Vondrasek ICIS August 27, 2015

24. Breeding efficiency and rise time – RF frequency
Xe-132 from RF discharge source
Oxygen plasma, single-frequency heating – TWTA only
Avg: 36 msec/q
Avg: 6 msec/q
R. Vondrasek ICIS August 27, 2015

25. ECR background beams – A/q region of interest
4/24/2017
ECR background beams – A/q region of interest
25.0 epA
14/3+
40/8+
16/3+
5.333
40/7+
12/2+
6.000
17/3+
5.714
19/4+
4.750
27/5+
5.400
35/6+
5.833
131/26+
129/26+
132/26+
5.077
132/27+
131/25+
5.240
131/27+
4.8518
No measureable background current with picoammeter
SBD – silicon barrier detector
129/25+
5.160
A behavior we understand all too well is the background contamination that plagues ECR charge breeders. Shown here is a mass scan for an oxygen plasma in the A/q region in which we operate. The full scale on the scan is 25 epA, and we see the background constituents of an oxygen plasma – nitrogen, argon, some aluminum, fluorine, and chlorine, as well as some left over xenon. But there are clean regions in the spectrum highlighted in green that show no background as measured on a picoammeter, and these regions map nicely with the rates we see in our energy diagnostics taken after acceleration in the linac. For example, 143/27+ which is very close to the 16/3+ shows a 330 kHz background rate, whereas 98/20+ which is in a clean region has only 970 Hz background.
144/25+
900 Hz background rate in SBD
5.760
143/27+
330,000 Hz background rate in SBD
5.296
143/25+
66,000 Hz background rate in SBD
5.720
146/28+
500 Hz background rate in SBD
5.214
144/26+
10,000 Hz background rate in SBD
5.538
144/28+
2500 Hz background rate in SBD
5.143
98/20+
970 Hz background rate in SBD
4.900
Test
R. Vondrasek ICIS August 27, 2015

26. Sources of ECR contamination
4/24/2017
Sources of ECR contamination
36Ar7+
72Ge14+
180Hf35+
144Ba28+
108Cd21+
During 144Ba28+ run we encountered a background beam of 36Ar7+ with small amounts of 72Ge14+ , 108Cd21+ , 180Hf35+
Possible sources of Ar-36
Air leak
O-ring permeation
Gas contamination
Chamber walls
CARIBU ON
17 Viton o-rings
Calculated permeation rate of 3.0e-5 torr-l/sec with a base pressure of 3.6e-8 Torr
We see in the energy spectrum of the silicon barrier detector that the majority of the rate (99%) is due to ar-36. We also see much smaller constituents or ge, cd, and hf. The top spectrum is with the Ba2+ coming from the helium gas catcher being injected into the charge breeder. And the bottom spectrum is with the 2+ beam swept away. If we could eliminate the ar-36 contaminant, the remaining ones are on the same scale as the Ba. So what are the possible sources? We eliminated the possibility of an air leak with helium leak checking as well as the N/O ratio on an rga not being correct for an air leak. We eliminated gas contamination when we changed the support gas bottle two times eventually using hihgly enriched o-18, and we observed no change in the ar-36 rate. That left the chamber walls and o-ring permeation which for us is a real concern as we have 17 viton o-rings on the source.
CARIBU OFF
Test
R. Vondrasek ICIS August 27, 2015

27. Sources of ECR contamination
4/24/2017
Sources of ECR contamination
Chemical Element
% Present
Magnesium (Mg)
Silicon (Si)
Titanium (Ti)
Chromium (Cr)
Manganese (Mn)
Iron (Fe)
Copper (Cu)
Zinc (Zn)
Aluminium (Al)
Balance
So what are our sources of contamination? If we look at the RGA spectrum, most of the peaks are not coming from the support gas which in this case was O-18. They are due to desorbtion from the chamber and beamline surfaces as well as o-ring permeation - which in our case is the equivalent of a 10^-5 torr-l/sec leak limiting us to an ultimate pressure of 2x10^-8 Torr. We did bake out the chamber with UV lamps since the permanent magnets and the various plastics used for HV isolation preclude a thermal bake-out. This lowered the base pressure by a factor of 2 mostly due to reduced water desorbtion, but had no real effect on the overall contaminant load.
Our other source of contamination, and the one which is a far greater headache, is the plasma chamber itself. The 6061 aluminum alloy has several components, all of which contribute to our beam contamination. And the forces due to the permanent magnet hexapole preclude using pure aluminum for the plasma chamber body.
ECR plasma chamber
Composition of Aluminum 6061-T6 H2 He
Water
18O N2 O2
18O2 Ar
CO2
Have to inject neutral gas to support plasma (in this case O-18)
Test
R. Vondrasek ICIS August 27, 2015

28. ECR contamination for 146Ba28+
4/24/2017
ECR contamination for 146Ba28+
68Zn13+
47Ti10+
94Mo18+
94Zr18+
120Sn23+
136Xe26+
146Ba28+
193Ir37+
198Hg38+
Ba-146: 104 pps
3% of total beam current
But even in the regions which we define as ‘clean’, once the beam is accelerated to higher energies and we can look at it on a silicon barrier detector, we see plenty of stable components. Not as many as the conflict table predicts there could be, but in this case of Ba-146 the stable beam accounted for 97% of the rate into the detector. Due to there being no strong contaminant near mass 146, we could still clearly see the radioactive beam component.
Test
R. Vondrasek ICIS August 27, 2015

29. Background reduction - cleaning
4/24/2017
Background reduction - cleaning
Solid species – surface and bulk contamination
Plasma chamber liners
Sand blasting
High-pressure rinsing
CO2 “snow” cleaning
High-purity aluminum coating
I can do nothing about all of the o-rings on the source, they are integral to its design, but we have been pursuing methods of reducing the contamination due to the plasma chamber itself. The KEK charge breeding group addressed contamination of their plasma chamber with sand blasting and high pressure rinsing. Since we did not wish to disassemble the source, we looked for alternatives. We had quite a bit of experience with quartz liners as a means of reducing contamination in our ECR sources, but the liners do not have a long lifetime when subjected to RF >200W. With that we moved to CO2 snow cleaning of the chamber surfaces as well as the injection and extraction hardware. The CO2 snow removes micron and submicron particulates and hydrocarbon-based contamination. It is nondestructive, nonabrasive, and residue-free. Additionally, this can all be done in situ. The CO2 cleaning addresses the surface contaminants but not the bulk contaminants. For this, we turned to vacuum coating the plasma chamber surfaces with high purity aluminum in an attempt to cover the bulk impurities within the aluminum chamber.
Test
R. Vondrasek ICIS August 27, 2015

30. Background reduction – coating
4/24/2017
Background reduction – coating
Cleaned chamber with CO2
Used W coil with high-purity (99.999%) aluminum wire
Source pressure: 5.0e-7 Torr
Wall covered with 1 micron layer
Vented to oxygen
Not all surfaces were adequately coated
The aluminum coating took place several months after the CO2 cleaning, so we cleaned the source again with the CO2 jet and then set a W coil which had been saturated with high purity aluminum in the middle of the plasma chamber. The source was evacuated and the coil heated over the course of 30 minutes with an average surface layer of 1 micron being deposited, although not all surfaces were adequately coated – namely the aluminum disk and mating piece at the injection side of the source.
Test
R. Vondrasek ICIS August 27, 2015

31. Background reduction – CO2 cleaning
4/24/2017
Background reduction – CO2 cleaning
40Ar12+
35Cl12+
19F5+
56Fe15+
So the results of this work. We had looked at the source output with very tight analyzing slits before the CO2 cleaning with three of the main contaminants (F, Cl, and Fe) which had negatively impacted us in the past shown here. The argon is shown as a means of normalizing the source running condition.
Test
R. Vondrasek ICIS August 27, 2015

32. Background reduction – CO2 cleaning
4/24/2017
Background reduction – CO2 cleaning
40Ar12+
After CO2 cleaning
F: factor of 20 reduction
Cl: factor of 4 reduction
Fe: factor of 50 reduction
Ar: 22% reduction
35Cl12+
19F5+
After the CO2 cleaning, we repeated the scan with the exact same source settings. The argon beam production has been slightly reduced, and we see that we have reduced our three main contaminants between a factor of 4 and 50.
56Fe15+
Test
R. Vondrasek ICIS August 27, 2015

33. Background reduction – Aluminum coating
4/24/2017
Background reduction – Aluminum coating
40Ar12+
After CO2 cleaning
F: factor of 20 reduction
Cl: factor of 4 reduction
Fe: factor of 50 reduction
Ar: 22% reduction
After aluminum coating
F: factor of 160 reduction
Cl: factor of 17 reduction
Fe: Not detectable
Ar: factor of 3 reduction
35Cl12+
After the aluminum coating, we again repeated the mass scan with the same source settings. The argon beam production has come down, mainly in the higher charge states, and our three contaminants have been further reduced with iron no longer being detectable. So this is encouraging for reducing the amount of stable contaminants coming from the ECR.
19F5+
Test
R. Vondrasek ICIS August 27, 2015

34. Reduction on target Some contaminants were eliminated
4/24/2017
Reduction on target
49Ti10+
98Mo20+
181Ta37+
132Xe27+
After coating
186W38+
After coating
Before coating
But what really matters is what the experimenter sees on target. We had produced a Zr-98 beam shortly before we aluminum coated the source, and then after the coating we a Y-98 beam – so essentially the same beam. In our energy spectrum after acceleration in the linac, we see that several of the stable contaminants have either been eliminated or have come down significantly. We still see a large amount of Mo-98 and unfortunately we have introduced new contaminants – Ta-181 and W-186 – most likely due to the heating coil we used for the evaporation. On target, we see that the Mo-98 has been reduced by a factor of 5 in the experimenter’s detector, the Ce-142 is gone, but the Ta and W now dominate the spectrum.
49Ti10+
54Fe11+
98Mo20+
142Ce29+
113Cd23+
171Yb35+
186W38+
Before coating
Some contaminants were eliminated
Mo-98 was reduced by x5 (on target)
New contaminants Ta-181 & W-186
Work remains to be done
Test
R. Vondrasek ICIS August 27, 2015

35. Worldwide ECR charge breeder performance
4/24/2017
Worldwide ECR charge breeder performance
Alessio Galata ThuPE26
Laurent Maunoury ThuPE31
ECR charge breeding performance extends over a wide range of both efficiency and A/q with the radioactive beams produced at Argonne clustering in the 8-15% efficiency range and for A/q. The SPES charge breeder, built by LPSC for the LNL group, has recently been commissioned with its efficiencies and A/q landing in this range. A new Phoenix ECR charge breeder for SPIRAL, which incorporates many of the lessons learned over the last several years, is presently undergoing testing at LPSC.
Test
R. Vondrasek ICIS August 27, 2015

36. Thank you
R. Vondrasek ICIS August 27, 2015