An ECR ion source-based low-energy ion accelerator: development and performance

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2011 Phys. Scr. 2011 014038

(http://iopscience.iop.org/1402-4896/2011/T144/014038)

Download details:

IP Address: 158.144.57.8

The article was downloaded on 27/12/2011 at 06:27

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING PHYSICA SCRIPTA

Phys. Scr. T144 (2011) 014038 (4pp) doi:10.1088/0031-8949/2011/T144/014038

An ECR ion source-based low-energyion accelerator: development andperformanceA N Agnihotri1, A H Kelkar1,3, S Kasthurirangan1,2, K V Thulasiram1,C A Desai1, W A Fernandez1 and L C Tribedi1

1 Tata Institute of Fundamental Research, Mumbai 400005, India2 Institute of Chemical Technology, Matunga, Mumbai 400019, India

E-mail: lokesh@tifr.res.in (L C Tribedi)

Received 16 May 2011Accepted for publication 18 May 2011Published 20 June 2011Online at stacks.iop.org/PhysScr/T144/014038

AbstractElectron cyclotron resonance (ECR) ion sources produce low-energy, highly charged ions.A new 14.5 GHz ECR-based low-energy ion accelerator facility has been developed. The ionsource involves a plasma chamber (‘supernanogan’) surrounded by permanent magnets thatprovide a suitable magnetic field. The entire assembly including the ion source and theanalyzing magnet is mounted on a 400 kV deck. A LabVIEW-based command and controlsystem has been developed for the beamline. In addition, wireless communication has beeninstalled to operate the machine in high voltage. The charge state distribution of several ions(He, N2, O2, Ne, Ar and Xe) has been measured. For Ar and Xe, the maximum charge statesmeasured were 16+ and 29+, respectively. A direct x-ray measurement for plasma diagnosticswas also initiated.

PACS numbers: 52.50.Sw, 41.75.Ak

(Some figures in this article are in colour only in the electronic version.)

1. Introduction

Electron cyclotron resonance (ECR) ion sources are wellknown to produce low-energy, highly charged ions. ECR ionsources are used to perform a wide variety of experimentsinvolving collisions at low velocities probing the processeswith low cross-sections. ECR ion sources use a magneticfield to confine the ions and electrons. Electrons moving inthe magnetic field perform cyclotron motion with the Larmorfrequency. Electrons are heated by resonant transfer of energyfrom microwaves (MWs). The impact of energetic electronson neutral atoms (gas or metal) produces ions. ECR ionsources can provide a huge ion current (of the order of tensof µA) with good stability for continuous operations. ECRion sources are used as an injector in various acceleratorsfor experiments and for radiation therapy [1–5]. An ECRion-source-based ion-accelerator facility has been installed inTIFR (TIFR-ECRIA).

3 Present address: MPIKP, Heidelberg, Germany.

2. Details of the TIFR-ECRIA beamline

2.1. Design details

The ECR ion source (‘supernanogan’)4 at TIFR(TIFR-ECRIA) consists of a Cu plasma chamber. It is apermanent magnet (SmCo Permanent) ECR ion source withdipolar axial and closed structure hexapolar radial magneticfields. Field maxima at the ends are 0.8 and 1.1 T. Themagnetic field has a minimum at the center of the plasmachamber; therefore it is known as the min-B configuration.The MW frequency is 14.5 GHz with 500 W maximumoutput5. At this frequency, the resonance magnetic fieldturns out to be 0.51 tesla. A plasma chamber contains aninsulated metal rod (called the bias rod). The maximumextraction voltage is 30 kV. Gases are introduced in theplasma chamber by two electrically controlled gas inlet

4 Designed and manufactured by Ms Pantechnik, France.5 TWT amplifier from ETM Electomatic Inc.

0031-8949/11/014038+04$33.00 1 © 2011 The Royal Swedish Academy of Sciences Printed in the UK

Phys. Scr. T144 (2011) 014038 A N Agnihotri et al

Figure 1. Schematic diagram of the TIFR-ECRIA beamline.

valves. Extracted ions are focused by an electrostatic Einzellens. A 90◦ bending dipole magnet (0.3 tesla) analyzesthe charge state of ions6. Figure 1 shows a schematicdiagram of the ECR ion source and beamline. The beamlineassembly from the ion source to the analyzing magnet ismounted on the isolated deck. The voltage of the deck canbe raised up to 400 kV to accelerate the ions further7. Theanalyzing magnet (on the HV deck) is joined to the restof the beam line (on ground) by an accelerating columnmade of ceramic insulators. Accelerated ions are focusedby a set of electrostatic quadruplole lens triplets. A set ofX–Y deflectors, two pneumatically controlled Faraday cups(FC1 and FC2) and one beam-profile monitor (BPM) arealso mounted. The position and dimensions of differentbeamline elements were frozen after performing intensivebeam transport simulations. GWBASIC code was used tosimulate the beamline trajectory with all active elements.ECR ion sources produce high dosage of x-ray radiation andhence, to avoid radiation exposure, the high-voltage (HV)deck has been covered with aluminum-covered lead shields(of thickness 6–10 mm). Residual radiation outside theshielding was measured near the ECRIA and was found to bewell within the safe limit. Vacuum in the plasma chamber is∼8E-8 mbar or better and that in the beamline is ∼8E-9 mbar.A new motor alternator–generator assembly provides therequired power to the electronic modules and power suppliesfor the MW system, dipole magnet, ion source, etc, on theisolated deck.

2.2. Command and control system

A LabVIEW-based command and control system has beendeveloped. The control system consists of a PC, connected6 Designed and manufactured by DANFYSIK, Denmark.7 Regulated High Voltage Power Supply (400 kV) from Glassman HighVoltage Inc.

with a PXI 1002 chassis, PXI 6115 multi-function DAQcard, a PCI card and PXI 4071 digital multimeter DMMmodule at the ground potential. Field-point modules alongwith the controller are mounted on the HV deck. The dipolemagnet, extraction voltage, Einzel lens, bias rod and gasflow controller channels were controlled through the NIFP AO-200 analogue output module channels. The magnetcurrent, extraction and Einzel lens voltages, Faraday cupcurrent, ECR body temperature and MW tuner positionwere monitored with the FP AI-110 analogue input modulechannels. One channel of FP RLY-420 was used as aninterlock for the MW amplifier and the other channel wasutilized to power the MW amplifier serial communicationunit. All the field-point modules are cascaded to the FP-2000real-time controller8 and connected to the wireless accesspoint kept on the HV deck. The PC is also connectedthrough the wireless router. All the control and monitoringparameters of the source are transmitted and received throughthis wireless mode. There are two separate wireless serialcommunication channels to control and monitor the MWamplifier.

3. Initial measurements

3.1. Charge state distribution

Performance of the machine was tested by measuring thecharge state distributions for different ions. The outputcurrents were measured at the Faraday cup (FC2 in figure 1).Such distributions were obtained for several gases such asnitrogen, oxygen, neon, argon and xenon. Figure 2 shows thetypical distributions obtained. Maximum input MW power isindicated in the figures. For nitrogen, oxygen and neon fullystripped ions were obtained. For Ar and Xe gases, maximum

8 National Instruments, www.ni.com.

2

Phys. Scr. T144 (2011) 014038 A N Agnihotri et al

1 2 3 4 5 6 7 8 9 100.1

1

10

100

1000

10000

Charge State

NeonInput Power: 180 W

(d)

1 2 3 4 5 6 7 81

10

100

1000

10000

Charge State

OxygenInput Power: 180 W

(c)

Cur

rent

(nA

mps

)

1 3 5 7 9 11 13 15 17

10

100

1000

10000

Cur

rent

(nA

mps

)

Charge State

ArgonInput Power: 180 Watt

(a)

5 8 11 14 17 20 23 26 290.1

1

10

100

1000

10000

Charge State

XenonInput Power: 190 W

(b)

Figure 2. Charge state distribution for output current obtained from TIFR-ECRIA for the gases (a) argon, (b) xenon, (c) oxygen and(d) neon.

charge states obtained were 16+ and 29+, respectively. Toincrease the high-charge state yields, mixing gas and biasrod techniques were used. For argon and xenon, nitrogen andargon gases were used as the mixing gas, respectively. In thecase of Xe (figure 2(b)), the sudden increase in currents atq = 13, 16, 26 and 28 is due to the mixing of Ar charge stateswith Xe charge states.

3.2. Effect of gas mixing, pressure and input power variation

The yield of higher-charge states can be increased by mixinga second gas (of lower atomic number) with the main gas.One possible explanation for this effect is that the ions arecooled by the loss of kinetic energy caused by collisionswith the mixed gas atoms. Cold ions are better confinedand get highly charged by repeated collisions with electrons.Other explanations are also presented in the literature ([6]and references therein). In order to optimize the yields of thehigher-charge-state ions, we have used different mixing gases.The yields of higher-charge states (>18+) were found to benegligible (at ∼190 W) in the case of pure Xe, i.e. without anymixing gas. It was also found that Ar is a better mixing gas forXe when compared with O2 or Ne. The output current dependson the feed gas pressure in the plasma chamber (not shownhere). The input MW power also plays an important role inthe output current. For example, the yield of Ar3+ initiallyincreases with the input power and then decreases. However,

for higher-charge states such as 6+ and 9+ ions, the currentkeeps increasing before it falls at higher power (∼150–160 W)again. Applying a negative bias of a few hundreds of volts tothe bias rod helps to increase the beam current. This effect hasalso been observed by other groups [6] and possible reasonsare discussed there although an exact explanation awaits.

3.3. Direct x-ray measurements for plasma diagnostics

We have measured the x-ray spectra of electronbremsstrahlung from the ion source using a NaI (Tl)scintillation detector with suitable absorbers to cut downthe low-energy x-rays. The x-ray detector was mountedjust outside the additional alignment port (facing the plasmachamber) on the dipole magnet as shown in figure 1. The x-rayintensities were investigated as a function of input powerand gas pressure. From the x-ray spectra, we have estimatedthe electron temperature prevalent in the ECR plasma [7, 8].Figure 3(a) shows a typical x-ray spectrum obtained at verylow MW power (10 W), for which the temperature estimatedwas 32.5 (±0.4) keV for gas pressure ∼1.5E-6 mbar. Thetemperature is estimated by fitting an exponential to theefficiency- and absorption-corrected bremsstrahlung spectra.It is well known that ECR plasmas exhibit two distinctelectron temperatures [9]. Our technique is sensitive onlyto the higher-energy component. Figure 3(b) shows thedependence of electron temperature on gas pressure for

3

Phys. Scr. T144 (2011) 014038 A N Agnihotri et al

1 1026

28

30

32

34

36

MW power 20 W

Ele

ctro

nT

empe

ratu

rekT

e(k

eV)

Pressure (10-6mBar)

(b)100 125 150 175 200 225 250

0

500

1000

1500

2000

2500

3000N

orm

alis

edY

ield

(arb

.uni

ts)

Energy (keV)

Oxygen plasmaMW Power = 10 WGas pressure = 1.5 x 10-6 mBarEstimated kT

e= 32.5 keV ± 0.4 keV

DataExponential fit

(a)

Figure 3. (a) X-ray spectrum obtained at 10 W input MW power; (b) dependence of electron temperature on gas pressure for 20 W inputMW power.

20 W MF power. The electron temperature of the plasma isseen to decrease steadily with an increase in gas pressure.Similar experimental investigations for other gases arereported in [8] and references therein, and the electrontemperature for Ar plasma is found to be approximately afactor of two lower compared with our results for O2.

4. Conclusions

In conclusion, we have installed a 14.5 GHz ECR ion-source-based highly charged ion accelerator for collisionexperiments with low-energy ions. The voltage of the entireECR assembly, the MW system and the dipole magnet forcharge analysis was raised to 350 kV by using an HV deck.A LabVIEW-based command and control system has beeninstalled. The charge state distributions of different gaseousions were measured as a function of the MW power, gaspressure and mixing gas. A direct x-ray measurement forthe plasma diagnostics was used to measure the plasmatemperature at low MW power.

The pressure dependence of the electron temperaturewas also investigated. Experiments on the ionization andfragmentation of the large molecules, surface nanostructureformation, ion implantation and irradiation experiments werealso initiated. In the near future it is planned to install aswitching magnet having several ports in order to increase thenumber of beamlines.

References

[1] Tinschert K et al 2008 Rev. Sci. Instrum. 79 02C505[2] Winkelmann T et al 2008 Rev. Sci. Instrum. 79 02A331[3] Haberer Th et al 1993 Nucl. Instrum. Methods Phys. Res. A

330 296–305[4] Sortais P et al 1998 Rev. Sci. Instrum. 69 656–8[5] Geller R 1996 Electron Cyclotron Resonance Ion Sources and

ECR Plasmas (Bristol: Institute of Physics Publishing)[6] Drentje A G 2003 Rev. Sci. Instrum. 74 2631–45[7] Lamoureux M 2000 Radiat. Phys. Chem. 59 171–80[8] Gumberidze A et al 2010 Rev. Sci. Instrum. 81 033303[9] Barue C et al 1994 J. Appl. Phys. 76 2662

4