P. Delahaye for the EMILIE collaboration Optimizing charge breeding techniques for ISOL facilities in Europe: Conclusions from the EMILIE project P. Delahaye for the EMILIE collaboration

World status: 2015 TRIUMF SPIRAL1&2 ISOLDE NSCL ReA3 ANL SPES RISP ECRCB ECRCBEBIS ISOLDE EBIS NSCL ReA3 ANL SPES ECRCBEBIS ECRCB RISP EBIS EBIS ECRCB + EBIS Texas A&M ECRCB ANURIB ECRCB + EBIS ECRCB In commissionning or planned EBIS running

Following the suggestion made for EURISOL P. Delahaye, O. Kester, C. Barton, T. Lamy, M. Marie-Jeanne, F. Wenander Eur. Phys. J. A 46(2010)421 Intense beams ECRIS challenges Beam purity and capture efficiency optimizations

Following the suggestion made for EURISOL P. Delahaye, O. Kester, C. Barton, T. Lamy, M. Marie-Jeanne, F. Wenander Eur. Phys. J. A 46(2010)421 High charge states Dead time, pile-up, fake coincidences REX-EBIS and MINIBALL: DAQ problems with intensities as low as 105-106 pps EBIS beam debuncher motivations: Less dead time, piling-up and fake coincidence problems which are presently caused by the EBIS pulses! 1ms every 50 – 500 ms. REX-EBIS and MINIBALL: data acquisition problems with intensities as low as 105-106 pps EBIS debuncher: solving this problems with intensities from 10^5 up to 10^10pps (maximum EBIS capabilities) EBIS challenges For mid-term ISOL facilities time structure is the main issue (before space charge limitations) Recall Stefan’s presentation

EMILIE objectives EBIS debuncher Simulation, Construction and test of a novel concept of EBIS beam debuncher Optimization of the performances of ECR charge breeders of Phoenix type Gaining understanding in the 1+ n+ technique Optimization of performances First tests at LPC Caen Optimization of the SPIRAL and SPES charge breeders See Laurent Maunoury’s poster ThuPE31 See Alessio Galatà’s presentation and poster ThuPE26

EMILIE objectives EBIS debuncher Simulation, Construction and test of a novel concept of EBIS beam debuncher Optimization of the performances of ECR charge breeders of Phoenix type Gaining understanding in the 1+ n+ technique Optimization of performances

No buffer gas, UHV design CW EBIS charge breeder EBIS Charge breeding Mass separation In trap decay CW Buffer trap Slow extraction Pseudo CW Linear RFQ trap Using the energy spread Dead time, pile-up, fake coincidences REX-EBIS and MINIBALL: DAQ problems with intensities as low as 105-106 pps RFQ cooler 1+ N+ Post acceleration A/q or TOF separation CW injection or bunching in a RF trap Pulsed No buffer gas, UHV design Entrance electrode Exit electrode DC for axial + RF for radial trapping Injection from EBIS CW beam

Debuncher prototype design Completed in 2012 at LPC Y. Merrer P. Desrues Downscaled prototype from original trap L: 0.5m  1m P=10-8mbar  10-11mbar RF rods Entrance electrodes Exit DC segments DC RF Focusing electrodes R0 = 15 mm RF rods Ø34.4 mm

Simulations with SIMION – results 2012 20 p.mm.mrad @ 20 kV E. Traykov E [eV] time [µs] extraction injection Cool effect! E [eV] [counts] extraction injection Transverse emittance increase (FWHM) Time flattening average [counts] time [µs] Typical settings: URF = 400 V, fRF = 4 MHz, Tcycle = 8000 µs, A/Q = 3.5, DE = 10.Q eV , DTOF = 50 µs. Can be flattened further by number/size of drawers, DC rise time, overlapping >85 % efficiency for a 45 cm long debuncher A higher energy spread would require  longer debuncher  stronger confinement

Tests bench mechanical adaptation Adaptation flange The first test of emilie debuncher will be made on the SHIRaC test bench, SHIRaC is a SPIRAL2 high intensity RFQ Cooler demonstrator. It is installed at LPC CAEN laboratory. an adapter flange has been machined to adapt the emilie debuncher to the SHIRaC test bench SPIRAL2 high intensity RFQ cooler demonstrator (SHIRaC) at LPC CAEN Emilie debuncher on the adaptation flange Emilie debuncher on the test bench Jean-François CAM Colloque GANIL 2013

@ low power (< 10 Watts) RF system Production of RF voltages on the quadrupole JF Cam Principle of the production system RF Vrf: 450V pic to pic @ 4MHz Frequency range: 2MHz to 6 MHz Voltage range: (pic to pic) : 0 to 1.2kV @ low power (< 10 Watts) Jean-François CAM Colloque GANIL 2013

Debunching voltage ramp generator JF Cam Goal: Produce 8 voltages ramps (few volts/ms to 100 v/ms) Sequencer: gate and delay generator 8 channels voltage Amplifier 500V G=50 & BW=15KHz 2x4 channels AWG LECROY ArbStudio Wired splitter 8 -> 23 channels x2 11 12 optical fiber link To sequence the generators and the lenses switches, we will use a multi-channel gate and delay generator. To produce the differents waveform for the voltage ramp, we will use 2 AWG and to obtain the necessary voltage amplitude we will use a 8 channels voltage amplifier. The ramping voltages will be distributed by a wired spliter. Because the debuncher will be install on the high voltage platform, it’s necessary to connect the computer to generators with an USB optical fiber link RF Computer to set AWG generators Ramping voltage exemple Emilie quadrupole Jean-François CAM Colloque GANIL 2013

Tests at LPC Caen First tests with continuous beams of singly charged ions Injected bunches defined by pulsing the injection electrode voltage Segments are raised to debunch the beam DC DC Focusing electrodes Focusing electrodes RF MCP or FC Incoming beam Kimball source: Na+ and Cs+ ions, 50nA , 5 keV F1094 Hamamatsu First extraction electrode Last injection electrode -1400V MCP signal: linear with the incoming current

Results Transmission though the trap (no bunching): >90% Trapping half-life: >100ms (Na+) > 1s (Cs+) Injected Na+ bunches baseline

Results Transmission though the trap (no bunching): >90% Trapping half-life: >100ms (Na+) > 1s (Cs+) Injected Na+ bunches baseline

Extraction modes from REXEBIS FWHM ~30 ms FWHM ~400 ms D. Voulot et al., NIM B 266 (2008) 4103 400 µs in every 40-400ms! Typical EBIS breeding cycles ~30 to 50% efficiency …to ~6ms From ~400µs…

EMILIE debuncher: Summary and future plans EMILIE debuncher status EBIS debuncher has produced first results 6-7 ms pulse without much optimization 30% - 50% efficiency Next steps Further voltage and timing optimizations Flattening the debunched beam Better efficiencies Observation of pulses over longer times (up to 100ms) Data acquisition allowing the single ion detection Defining the injected bunches with a pulsed deflector (removal of the overshoot) Mimicking the EBIS pulses Energy dispersion Adding a RF voltage on top of the EBIS debuncher voltage No effect for 7V, what happens for higher voltages Future plans Tests with multiply charged ions at GANIL in 2016 Possible tests with TWIN EBIS at ISOLDE Within ENSAR 2!

EMILIE objectives EBIS debuncher Simulation, Construction and test of a novel concept of EBIS beam debuncher Optimization of the performances of ECR charge breeders of Phoenix type Gaining understanding in the 1+ n+ technique Optimization of performances

EMILIE objectives EBIS debuncher Simulation, Construction and test of a novel concept of EBIS beam debuncher Optimization of the performances of ECR charge breeders of Phoenix type Gaining understanding in the 1+ n+ technique Optimization of performances

Comparison of « standard » ECRIS and charge breeder performances Comparison of performances for a 14GHz source and a 14 GHz charge breeder Phoenix at LPSC vs JYFL ECRIS Same conclusions, despite lower magnetic field confinement and no bias disc in Phoenix Double frequency heating  higher charge states, higher efficiencies H. Koivisto et al, RSI 85 (2014) Presented at ICIS 2013

1+ beam capture probability Charge breeding efficiency is determined by Capture efficiency of the 1+ beam Ionization efficiency from 1+ to n+ Extraction efficiency of the n+ ions Slowing down coefficient of the incident 1+ ions in ion-ion collisions R(uab) is at maximum when the incident ion velocity equals the average speed of the plasma ions Experiments with the Phoenix charge breeder at LPSC: breeding efficiency of high charge state ions as a function of incident velocity matches well with the prediction O. Tarvainen et al., Plasma Sources Sci. Technol. 24, 3, 035014 (2015) Delcroix and Bers  Slowing down depends on the ratio of velocities (ISOL ion/ plasma ion) Slowing down coefficient superimposed to the DV curve Capture mechanism is qualitatively well understood

1+ beam capture probability as a diagnostic tool of ECR plasmas Low charge states extracted from the breeder behave differently Incident 1+ ions can propagate through the breeder plasma. This is confirmed by comparing their ’breeding time’ with the ion flight-time without plasma. 2+ is produced via in-flight ionization Measurement of the 1+ and 2+ ’breeding efficiencies’ yields an estimate for ion mean free path, ion-ion collision frequency and plasma density The incident 1+ beam is a perfect diagnostics tool for ECRIS plasmas! Experiments with Na, Rb, Cs have been carried out, showing similar results. Fraction of 1+ / 2+ beams through the ECR plasma  mean free path Ion – ion collision frequency Q>3 ions are collisional ions Q<~3 are magnetized ions Electron density ~1 order of magnitude below the cut-off density 40kHz for 1+ ions 41011 e/cm3 O. Tarvainen et al., Plasma Sources Sci. Technol. 24, 3, 035014 (2015)

EMILIE objectives EBIS debuncher Simulation, Construction and test of a novel concept of EBIS beam debuncher Optimization of the performances of ECR charge breeders of Phoenix type Gaining understanding in the 1+ n+ technique Optimization of performances

Ionization of C beams at LPSC and GANIL PREMAS task 3 + Ionization of C beams at LPSC and GANIL Nanogan @ GANIL Phoenix + COMIC @ LPSC Phoenix @LPSC For Carbon beams direct multi-ionization in Nanogan is prefered compared to charge breeding in Phoenix BUT: Phoenix possibly better for higher charge states and beam purity Nanogan @ GANIL L. Maunoury et al, RSI 85 02A504 (2014) Presented at ICIS 2013

Light ion charge breeding at ANL Nov. 2011 Efficiencies for Na and K improved by a great amount Playing with jaws to optimize efficiency possible effect of the emittance of the 1+ beam? R. Vondrasek, P. Delahaye, S. Kutsaev and L. Maunoury, Rev. Sci. Instr. 83(2012)113303 P. Delahaye, L. Maunoury, R. Vondrasek, Nucl. Instr. Meth. A, 693(2012)104.

Improvement of charge breeding efficiencies 1) Injection of molecules 2) Vaccuum improvement Lower residual pressure 3) Magnetic field improvements: symmetrization at injection axial field optimization 4) Double frequency heating 40Ar8+ 39K9+ 40Ar8+ 40Ar8+ 39K9+ 133Cs27+ LPSC charge breeder ANL charge breeder EMILIE 2012-2014 23Na8+ 84Kr17+ 39K9+ 133Cs27+ 23Na7+ 85Rb19+ 85Rb19+ 85Rb19+ 23Na7+ 23Na7+ 13C4+ (CO+) 13C4+ EMILIE 2015 SPIRAL charge breeder SPES charge breeder P. Delahaye, NIM B 317(2013) , J. Angot et al, THYO02, ECRIS 2012 H. Koivisto et al, RSI 85 (2014) L. Maunoury et al, RSI 85 02A504 (2014) T. Lamy et al, ECRIS 2014 Laurent Maunoury ThuPE31 Alessio Galatà ThuPE26

Charge breeding times

EMILIE: Summary and future plans Phoenix charge breeder optimization Tests at LPSC for understanding the 1+ n+ technique Tests with the CARIBU CB for light ions at SPIRAL Operational upgraded Phoenix charge breeders SPES charge breeder See Alessio’s poster SPIRAL charge breeder See Laurent Maunoury’s poster Dedicated efforts for beam purity UHV /aluminum components and special surface treatments Use of dedicated separators (SPES) and cyclotron (SPIRAL) mass resolving power Simulation work at INFN MW coupling reproducing the 1+ beam capture DV curve see Alessio’s presentation ECR 1+ hot sources being assembled at LPSC 650°C at 5.4GHz should be tested within EMILIE, 1200°C for a later stage

Thanks a lot for your attention! GANIL L. Maunoury E. Traykov P. Jardin M. Dubois P. Chauveau LPC Caen G. Ban J. F. Cam C. Vandamme INFN LNL Galatà G. Patti G. Prete INFN LNS L. Celona JYFL H. Koivisto O. Tarvainen ANL R. Vondrasek ISOLDE F. Wenander Shornikov LPSC T. Lamy J. Angot J. Angot, G. Ban, L. Celona, J. Choinski, , P. Delahaye (GANIL IN2P3, coord.), A. Galata (INFN, deputy coord.), P. Gmaj, A. Jakubowski, P. Jardin, T. Kalvas, H. Koivisto, V. Kolhinen, T. Lamy, D. Lunney, L. Maunoury, A. M. Porcellato, G. F. Prete, O. Steckiewicz, P. Sortais, T. Thuillier, O. Tarvainen, E. Traykov, F. Varenne, and F. Wenander

Phoenix ECRIS Noble gases are recycled Condensable elements are not DV tuning necessary

Capture is very efficient after cooling in REXTRAP REXTRAP - REXEBIS Capture is very efficient after cooling in REXTRAP Operating range REX-ISOLDE

ANL ECRIS High capture efficiency Better 1+ beam optics Very low residual pressure High charge states 2 frequency heating Very low residual pressure

Beam purity from REX-EBIS Extracted beams from REXEBIS as function of A/q showing residual gas peaks and charge bred 129Cs. The blue trace is with and the red trace without 129Cs being injected. F. Wenander, CERN courier, Jan/Feb 2012, p33

SPIRAL 1 & 2 post-accelerator 3 kW 1,5 kW Ne, Ar, Kr, N, O, F He Targets Nanogan ECRIS source B. Jacquot et al, GANIL status report http://arxiv.org/ftp/nucl-ex/papers/0502/0502016.pdf Suppression ~10 for R=m/dm~1/(2pHN) up to 6000 Suppression ~106 for dm/m=5.10-4 CIME cyclotron

ANL ECRIS charge breeder See talk R. C. Pardo Multiple frequency operation Klystron: 10.44 GHz, 2 kW TWTA: 1113 GHz, 0.5 kW Open hexapole structure RF is injected radially Uniform iron in the injection region for symmetrical fields Improved pumping to the plasma chamber region Base pressure: 2x10-8 mbar Operation: 7x10-8 mbar Extraction pressure: 4x10-8 mbar What is so different? Field symmetries Tunable grounded tube Excellent vaccuum 2 Frequency heating Cold 1+ beams …? Movable grounded tube with 2.5 cm of travel 50 kV high voltage isolation Importance of field symmetry R. Vondrasek et al, Rev. Sci. Instrum. 83, 113303 (2012) Loss patterns of ions on plasma chamber and electrodes SC Jeong et al, Rev Sci Instrum. 83, 02A910 (2012) Design value Running condition Binj 1.31 T 1.16 T Bmin 0.31 0.27 Bext 0.85 0.83 B (radial) 0.86 T Last closed surface 0.61 T

ECRIS optimizations Optimizing charge states and capture efficiencies with double frequency heating Gas mixing Using molecular beams (CO) Light ions Studies of the 1+ beam capture process 1+ beam slowing down simulations 1+ beam capture experimental investigations Low charge states as a diagnostic for the capture Influence on the plasma composition Off-line tests of SPES and SPIRAL charge breeders H. Koivisto et al, RSI 85 (2014) L. Maunoury et al, RSI 85 02A504 (2014) R. Vondrasek et al, Rev. Sci. Instr. 83(2012)11330 P. Delahaye et al, NIM A 693(2012)104 Phoenix ECRIS at the LPSC test stand T. Lamy et al., ECRIS2014, JACoW Tests at LPSC until September 2015

The REX-EBIS setup The LaB6 cathode EBIS specifications jcathode<20A/cm2 je=jtrap<200A/cm2 Ie=460mA (normal operation 200mA) E=3.5-6keV 3 drift tubesL=200 to 800 mm Theoretical capacity 5·1010 positive charges Ultra-high vacuum 10-10 - 10-11 mbar The charge state is selected with a mass separator of Nier-Spectrometer type Performances: F. Wenander et al., Rev. Sci. Instrum. 77, 03B104 (2006) ICIS 05 Proceedings

Beam purity from ECRIS charge breeders Troublesome restgas spectrum: C,N,O… + stainless steel components Entrance of the ECR P=5.10-7mbar Ex: Phoenix at ISOLDE Exit P=2.10-7mbar Background >5nA 2<A/q<7 Phoenix test bench @ISOLDE