Negative Ion Production and Beam Extraction Processes in a Large Ion Source

K. Tsumori1,2, K. Ikeda1, H. Nakano1,2, M. Kisaki1, S. Geng2, M. Wada3,

K. Sasaki4, S. Nishiyama4,G. Serianni5, C. Wimmer6, P. Agosttineti5, E. Sartori5,

M. Brombin5, K. Nagaoka1, M. Osakabe1,2, Y. Takeiri1,2,O. Kaneko1,2

and NIFS NBI group1National Institute for Fusion Science, 322-6 Oroshi Toki Gifu 509-5292, Japan 2The Graduate University for Advanced Studies, Shonan Village, Hayama, Kanagawa 240-0193, Japan3Doshisha University, Kyotanabe, Kyoto 610-0321, Japan4Division of Quantum Science and Engineering, Hokkaido University, Sapporo 060-8628, Japan5Plasma Engineering Group, Consorzio RFX - Corso Stati Uniti, 4, 35127 Padova, Italy6Max-Planck-Institut für Plasmaphysik, Bereich ITER-Technologie & -Diagnostik / N-NBIBoltzmannstr. 2, 85748 Garching, Germany

e-mail address: tsumori@nifs.ac.jp

The 16th International Conference on Ion Sources (ICIS 2015)from 23rd to 28th August, 2015 at Manhattan NY, USA

Outline

• Introduction • Source-plasmas response to electric

field• Measurement of Ho and H- temperatures• Flow measurement of charged particles• Beam extraction• Summary

Introduction

Direction of our research

• Negative-ion (H-) rich plasmas are generated widely in beam extraction region in Cs seeded plasma.

• They are contaminated with electron diffused from driver region by applying beam extraction field.

• To investigate charged-particles dynamics further, we start to measure followings:– Ho and H- temperatures – Detailed distribution of charged particles and its response to electric field.– flow of electron, positive and H- ions Before beam extraction

(Cs seeded plasma)e- e-

H0 H+

H+H- H-

H+

H- H- H-

H- H+H+

e- e-H+

During beam extraction

H- density ele. density

NIFS-R&D Negative Ion Source

Magnet setup

B

A

A side

B side

Magnetic field induced in NIFS R&D source

230

mm

350 mm

Magnetic field in very vicinity of plasma grid

• Multi-cusp source with a pair of filter magnets.• Inner size: 700 mm (Height) x 350 mm (Width)

x 230 mm (Depth).• Beam extraction region from PG to filter

magnets are the target of this research.• Filter and electron deflection fields combines in

the region.

Diagnostic devices

Distribution of extracted H-

Beamlet interactionAperture mask, mini-STRIKE

Extracted H- distributionHa CCD imaging

electron and ion flowsDirectional Langmuir probe

H- flowDirectional PD probe

H- temperatureDirectional PD probe

Saturation CRD

H- beam divergenceBeamlet monitor

Space charge compensation Meniscus oscillation

Meniscus formation

Distribution of plasma potential

Langmuir probe

Bias and plasma potentialsLangmuir probe

Normal and ionic plasmasLangmuir probe and CRD

Cs pressure and H-, ele densitiesCs LAS and Cs-IG

Diffusion of ele and pos. ion

Beam transport

Sheath formation and extraction

Particle transport in extraction region

Ambipolar diffusion and drift

Magnetic structure

experimental setup

• NIFS-R&D negative ion source has been applied for the investigation.

• Plasmas in extraction region are measured with multiple diagnostic system.

Driver region

Extraction region

Source-plasma response to electric field

Distribution of plasma potential (bias dependence in H2 plasma)

• Electric field, slope of potential distribution, does not change in pure H2 plasma metal like character

• The field strength is 20~30 V/m• Sheath gap at PG is proportional to bias voltage

PG

Distribution of plasma potential (bias dep. in Cs seeded plasma)

• Slope of potential distribution changes the by increasing the bias voltage in Cs seeded plasma Simi-conductor like character

• This indicates the applied field penetrates inside Cs seeded plasma.• Sheath gap is less sensitive to bias voltage than the case of H2 plasma

Applied field is relaxed less with H--rich sheath.

PGPG

Change of saturation currents before and during extraction

Change of negative and positive saturation currents near PG with beam extraction.

• 2D distribution of negative saturation current shows the electron concentrates along the magnetic field of electron deflection magnet

2D mapped negative saturation currents near PG with beam extraction.

S. Geng et al., Plasma Fusion Res. 10, 3405016 (2015)

ratios of saturation currents with to w/o beam extraction.

Ho temperatures

Ho(H+/H2+) temperature (OES)

• Collimated window of HR- OES (High Resolution Optical Emission Spectroscopy) is installed at the top of bias insulator.

• Wavelength precision: 1 pm Resolution: 10 pm

• Blue wing of Ha has strong bias dependence.

• THo ≥ 1 eV (5 eV at Max).

• Parent particles are H2+ or H+ in

this measurement.

– H2+ + e H(n=3) + H(n=1)

– H+ + H- H(n=3) + H(n=1)

1.0

0.8

0.6

0.4

0.2

0.0

Re

lative

In

ten

sity

656.40656.35656.30656.25656.20

wavelength (nm)

VB = -10 -> 0

1.0

0.8

0.6

0.4

0.2

0.0N

orm

aliz

ed

In

ten

sity

656.40656.35656.30656.25656.20

Intensity (nm)

VB = 0 -> +10

VB: -10 to 0 V

VB: 0 to 10 V

Higher bias

Higher bias

Collimator

Fiber (to HR-OES)

M. Wada et al., TuePE35: ICIS 2015, this conference

Ho temperature (Ha LAS)• Ha laser absorption spectroscopy (Ha LAS) is

applied to measure Ho temperature, THo.

• The temperature is almost proportional to input arc power.

• THo ~ 0.3 eV at 50 kW of input.

• THo decrease by increasing operational pressure.

• parent particle is H(n=2) in this diagnostic.

Arc dependence

Pressuredependence

Parc: 50 kW

H. Nakano et al., AIP Conference Proceedings 1655, 020018, (2015)

Flow of charged particles

Flow of electron and positive ions (four-pin Langmuir probe)

Four-pin Langmuir probe(directional Langmuir probe)

A-tip

B-tip

C-tip

D-tip Available to scan in 3D

Rotatable around stem axis

Flow directions of electron and positive ion are similar due to ambipolar diffusion

Electrons Positive Ions

Flow of electron and positive ions (estimation)

Estimation of flow velocity at a position 26 mm apart from PG.Bias voltage:

Estimation using saturation current

Estimation with cross-field drift

Estimation using flow direction

Estimation using diffusion

Perpendicular direction to filter field Depth direction (z direction)

= 3.8 x 103 m/s = 4.2 x 103 m/s vz = 4.7 x 102 m/s vz = 2.4 x 102 m/s

H- flow and temperature

H- temperature and flow measured with 4-pin PD

vth = (vA + vc) / 2vflow = (vA - vc) / 2

vth: thermal velocityvflow: flow velocity

vA = vth - vflow vC = vth + vflow

Laser pulse

A-tip C-tip

B-tip

C-tip

H- flow

Using the decay time of photodetachment signal, H- temperature is estimated as 0.12 ± 0.03 eV.

Flow direction is opposite to electron and positive ions.

Saturated CRD

high power laserlow H- temperature

Mirror Mirror

PD

Requirement for density measurement

Density is underestimated

• Using higher power of YAG, over-neutralization occurs in cavity path.

• H- temperature of 0.1 eV is obtained by comparing the saturation.

Beam extraction

Ha CCD Imaging

• Ha CCD imaging, which subtracts image with beam off from beam on phases, shows the H- decrement due to the extraction.

• By increasing the bias voltage, the reduction rate of H- decreases as shown in the left-side figure.

• Next question is how the decreasing H- is extracted from PG apertures.

K. Ikeda et al., AIP Conference Proceedings 1655, 040005 (2015)K. Ikeda et al., MonPE11 ICIS 2015, this conference

Beamlet eclipse: Setup

• PG aperture was shaded with movable ceramic cylinder (eclipse shade) to investigate influences of an obstacle for

(1) H- production and

(2) beamlet interaction.• H- beamlets extracted from ion source was monitored with

beamlet monitor called mini-STRIKE.Movable in 3D directions

Eclipse shade

PG mask

PG apertures

P. Veltri et al., TuePS34 ICIS 2015, this conferenceIdea of Dr. H. Nakano and M. Kisaki

Beamlet eclipse: IR images

Normal extraction with beamlet eclipse shade

• Beamlet-eclipse shade changes the beamlet pattern drastically as shown below.

• Nearest neighbor beamlets shift due to beamlet interaction.

Beamlet eclipse: 2D profiles

• By setting the beamlet-eclipse shade at 8.5 mm apart from PG, shaded beamlet decreases the intensity down to ~30 %.

• This decrease is mainly caused by shading the parent particle of H-.

• Beamlet-eclipse method has possibilities to provide more information on beam extraction and formation.

Green: w/o PG-aperture plugBlue: PG-aperture plug@18.5 mmRed: PG-aperture plug@ 8.5 mm

Summary

• In H2 plasma, potential slope is not affected with bias voltage, and electrons sheath adjusts the slope.

• In negative-ion-rich plasma, sheath cannot adjust the bias voltage completely, and potential gradient changes the polarity due to the bias voltage.

• Near PG surface, electron density is rich on the cusp line of electron deflection magnets.

• Ho temperature has been measured with HR-OES and Ha LAS. Origins of Ha are considered H+/H2

+ and H(n=2), respectively.

• Flow velocity of electron, positive ion and H- are 10 times lower than their thermal velocities of 0.1 to 0.5 eV.

• Beamlet-eclipse method could provide some new information on beam extraction and formation.