Search

The large RF negative hydrogen (deuterium) ion source at the ELISE test facility (half of the ITER-NBI source size) has been equipped with a Cavity Ring-Down Spectroscopy (CRDS) system, in order to measure the negative hydrogen (deuterium) ion density in the region in front of the plasma grid (first grid of the extraction system). The challenge of this diagnostic for ELISE relies on the large size of the source and therefore on the plasma length across which the measurements are performed as well as the long pulses at RF power, which can affect the cavity mirror reliability. A dedicated experiment on the mirror reliability was performed, ensuring the feasibility of measurements for long pulses (several hundred seconds) at high RF power. Two horizontal lines of sight were dedicated to CRDS: the measured density was in the range between 4 × 1016 and 1 × 1017 m-3, with a slightly higher density for the bottom lines of sight, for both the isotope hydrogen and deuterium. Different temporal evolution was observed for the two isotopes, showing a higher instability for the deuterium case: this is in correlation with the extracted negative ion current density and inversely correlated with the coextracted electron current density. The CRDS system allowed performing the first measurements of negative ion density for a long pulse (1000 s) in a large source: the temporal behavior and the effect of the beam extraction will also be discussed.

A compact ion source that produces hydrogen plasma with an electron cyclotron resonance (ECR) configuration combined with a 2-stage extraction system with a single aperture of 6 mm diameter has been designed and built to study the performance of different materials as the plasma electrode (PE) of a negative hydrogen ion source. The source has the capability to electrically bias the PE with respect to the ECR plasma. The first experiment with low ECR power input (less than 40 W) was carried out. The PE of the C12A7 electride showed the largest H- current among aluminum, molybdenum, and the C12A7 electride.

Density distributions of negative hydrogen (H−) ions and negative deuterium (D−) ions were measured with the laser photodetachment method in the extraction region of the negative ion source. The distribution of H− ion density peaks at the center of the ion source, while that of the D− ion shows a flatter profile in the direction parallel to the plasma grid. The positive ion densities of hydrogen and deuterium estimated from the positive saturation current indicate similar profiles with different amounts close to the grid. The difference in the H− ion and D− ion distributions can be explained by the difference in the negative ion yield and the survival probability of the ions due to the isotope effect.

Tantalum was compared with tungsten as filament material for negative ion extraction in the discharge of hydrogen and deuterium. For hydrogen discharge, tantalum filament operation showed about 40 % increase of negative hydrogen ion current and decrease of co-extracted electron current by 10 % against tungsten filament operation. Tantalum filament operation with deuterium indicated lower intensity of co-extracted electron current compared with tungsten while negative deuterium ion current showed slightly higher intensity. Dependence of extracted currents upon the plasma electrode bias voltage was linked to the local plasma parameters measured by a Langmuir probe. The effects due to the filament materials in hydrogen and deuterium plasmas are discussed.Tantalum was compared with tungsten as filament material for negative ion extraction in the discharge of hydrogen and deuterium. For hydrogen discharge, tantalum filament operation showed about 40 % increase of negative hydrogen ion current and decrease of co-extracted electron current by 10 % against tungsten filament operation. Tantalum filament operation with deuterium indicated lower intensity of co-extracted electron current compared with tungsten while negative deuterium ion current showed slightly higher intensity. Dependence of extracted currents upon the plasma electrode bias voltage was linked to the local plasma parameters measured by a Langmuir probe. The effects due to the filament materials in hydrogen and deuterium plasmas are discussed.

The consumption rate of cesium (Cs) for negative hydrogen (H-) ion source increases when the source operation gas is changed from hydrogen to deuterium. There was observed a clear indication that a deuterium discharge erodes Cs atoms on the plasma grid (PG) surface to increase work function and the co-extracted electron current. We have proposed a model, that the enhanced sputtering yield of Cs from the PG due to deuterium ions, which carries more kinetic energy to Cs adsorbed on the surface directly or indirectly, is the main reason for this fast dissipation of Cs. Introduction of helium (He) into discharge can verify the enhanced sputtering effect due to the lager mass ions in the source discharge. Comparing the effect due to seeded Cs before and after the He injection into discharge through Cs OES signals as well as the H- density measured with cavity ring-down method, the sputtering/evaporation enhancement due to He is estimated. Neutral atoms and positive ions in the He discharge should cause enhanced sputtering like deuterium, while the system does not generate neutron under the induction of acceleration voltage to diagnose the extracted negative ion beam. Plasma parameters of the H2 and He plasmas are investigated by diagnostics tools installed on NIFS-RNIS (National Institute for Fusion Science, Research and development Negative Ion Source) together with H- density measurement by cavity ring down. Enhanced consumption rate of Cs is compared with proposed sputtering yield data to predict the rate for deuterium operation of negative ion sources.

Hydrogen ion reflection properties have been investigated following the injection of H+, H2+ and H3+ ions onto a polycrystalline W surface. Angle- and energy-resolved intensity distributions of both scattered H+ and H− ions are measured by a magnetic momentum analyzer. We have detected atomic hydrogen ions reflected from the surface, while molecular hydrogen ions are unobserved within our detection limit. The reflected hydrogen ion energy is approximately less than one-third of the incident beam energy for H3+ ion injection and less than a half of that for H2+ ion injection. Other reflection properties are very similar to those of monoatomic H+ ion injection. Experimental results are compared to the classical trajectory simulations using the ACAT code based on the binary collision approximation.

Proton beams at 0.3, 0.5 and 1 keV energies were injected onto a Mo surface at shallow incidence angle, and the angle- and energy-resolved intensity distributions of reflected negative/positive ions were measured. The negative to positive intensity ratio increased as the incident angle and the reflected angle became shallower and the incident energy became lower. A numerical simulation calculation, Atomic Collision in Amorphous Target (ACAT), has been carried out to get the angular distribution of total reflection particle flux. Experimentally measured angular distributions of reflected H- ion intensity showed maxima at smaller angles than the calculated results, indicating that the negative ion formation has a maximum at v⊥ of 0.03 - 0.04 atomic unit.

The achievements of the deuterium beam operation of a negative-ion-based neutral beam injector (N-NBI) in the large helical device (LHD) are reported. In beam operation in LHD-NBIs, both hydrogen (H) and deuterium (D) neutral beams were generated by changing the operation gas using the same accelerator. The maximum accelerated deuterium negative-ion current () reaches 46.2 A from two beam sources with the averaged current density being 190 A m−2 for 2 s, and the extracted electron to accelerated ion current ratio () increases to 0.39 using 5.6 V high bias voltage in the first deuterium operation in 2017. An increase of electron density in the vicinity of the plasma grid (PG) surface, which is considered the main reason for the increase of co-extracted electrons in a beam, is confirmed by the half-size research negative-ion source in the neutral beam test stand at the National Institute for Fusion Science (NIFS). The deuterium negative-ion density is also larger than the hydrogen negative-ion density in the vicinity of the PG surface using the same discharge conditions. In the latest experimental campaign in 2018, increases to 55.4 A with the averaged current density being 233 A m−2 for 1.5 s using the shot extraction gap length. The low of 0.31 can be maintained by using high discharge power. The various parameters mentioned above are defined in detail below.

To understand the physics of the cesium (Cs) recycling in the large Cs-seeded negative ion sources relevant to ITER and JT-60SA with ion extraction area of 45-60 cm × 110-120 cm, the time evolution of the negative ion profile was precisely measured in JT-60SA where the ion extraction area is longitudinally segmented into 5. The Cs was seeded from the oven at 180 °C to the ion source. After 1 g of Cs input, surface production of the negative ions appeared only in the central segment where a Cs nozzle was located. Up to 2 g of Cs, the negative ion profile was longitudinally expanded over full ion extraction area. The measured time evolution of the negative ion profile has the similar tendency of distribution of the Cs atoms that is calculated. From the results, it is suggested that Cs atom distribution is correlated with the formation of the negative ion profile.

A cylindrical carbon pellet with a size of 1.2L × 1.2 mm to 1.8L × 1.8 mm and a velocity of 100 m/s to 300 m/s was injected into large helical device (LHD) for an efficient fueling based on its deeper deposition instead of hydrogen gas puffing and ice pellet injection. Electron density increment of Δne = 1014 cm−3 is successfully obtained by single carbon pellet injection without plasma collapse. Typical density and temperature of the ablation plasma of the carbon pellet, e.g., 6.5 × 1016 cm−3 and 2.5 eV for CII, are examined respectively by spectroscopic method. A confinement improvement up to 50% compared to ISS-95 stellarator scaling is clearly observed in a relatively low-density regime of ne = 2×1013 cm−3 to 4×1013 cm−3, and high ion temperature Ti(0) of about 6 keV is also observed with an internal transport barrier at ne = 1.2×1013 cm−3. In particular, the improvement in the ion temperature largely exceeds that observed in hydrogen gas-puffed discharges, which typically ranges below 3 keV.

A high production rate of negative hydrogen ions (H−) was observed from a nanoporous 12CaO 7Al2O3 (C12A7) electride surface immersed in hydrogen/deuterium low-pressure plasmas. The target was negatively biased at 20–130 V, and the target surface was bombarded by H3 + ions from the plasma. The production rate was compared with that from a clean molybdenum surface. Using the pseudo-exponential work-function dependence of the H− production rate, the total H− yield from the C12A7 electride surface bombarded at 80 V was evaluated to be 25% of that from a cesiated molybdenum surface with the lowest work-function. The measured H− energy spectrum indicates that the major production mechanism is desorption by sputtering. This material has potential to be used as a production surface of cesium-free negative ion sources for accelerators, heating beams in nuclear fusion, and surface modification for industrial applications.

Ion heating experiments have been intensively carried out in high- and low-Zeff conditions of Large Helical Device plasmas. In high-Zeff plasmas utilizing neon or argon gus puffing, the ion heating...

We measured the characteristics of the reflected particles from a carbon nanowall (CNW) deposited on a W surface following the injection of 1–2 keV H+ and O+ ions. The reflected ion energies and intensities indicated a contribution from multiple scattering in the target. The reflection angular dependence of the reflected ion intensities reached the maximum around the mirror angle and showed a sharp distribution, which may be attributable to the effect due to the aligned structure of the CNW. The energies and intensities of the reflected ions decreased with the time of ion bombardment. The intensities and energies of the reflected ions were, however, recovered to some degree by baking the sample, indicating the surface modification due to retention of the injected particles during the injection. We used the Monte Carlo simulation code ACAT (Atomic Collision in Amorphous Target) to study these processes theoretically and the calculated results supported the experimental results.

The angular distributions and energy spectra of positive and negative ions reflected from a vanadium alloy (V–4Cr–4Ti) bombarded by beams of low energy (1–2 keV) ions of H, He, and O were measured. The intensities of H + or H − produced from reflection of 2 keV H + beam at the surface of vanadium alloy showed the maxima at the angles closer to the surface normal from the mirror reflection angle, compared with Mo and W samples. The time dependence of the angular distributions of the reflected H + or H − ions measured at the sample temperature of about 240–260 °C showed a shift of the peak in the angular distribution to the mirror reflection angle.

Characteristics of ionic plasmas, observed in a high-density hydrogen negative ion source, are investigated with a multi-diagnostics system. The ionic plasma, which consists of hydrogen positive- and negative-ions with a significantly low-density of electrons, is generated in the ion extraction region, from which the negative ions are extracted through the plasma grid. The negative ion density, i.e., the ionic plasma density, as high as the order of 1×1017m−3, is measured with cavity ring-down spectroscopy, while the electron density is lower than 1×1016m−3, which is confirmed with millimeter-wave interferometer. Reduction of the negative ion density is observed at the negative ion extraction, and at that time the electron flow into the ionic plasma region is observed to conserve the charge neutrality. Distribution of the plasma potential is measured in the extraction region in the direction normal to the plasma grid surface with a Langmuir probe, and the results suggest that the sheath is formed at the plasma boundary to the plasma grid to which the bias voltage is applied. The beam extraction should drive the negative ion transport in the ionic plasma across the sheath formed on the extraction surface. Larger reduction of the negative ions at the beam extraction is observed in a region above the extraction aperture on the plasma grid, which is confirmed with 2D image measurement of the Hα emission and cavity ring-down spectroscopy. The electron distribution is also measured near the plasma grid surface. These various properties observed in the ionic plasma are discussed.

Distributions of hydrogen Balmer-α (Hα) intensity and its reduction behavior close to a plasma grid (PG) surface have been observed by a spectrally selective imaging system in an arc discharge type negative hydrogen ion source in National Institute for Fusion Science. Hα reduction indicates a reduction of negative hydrogen ions because the mutual neutralization process between H+ and H− ions causes the dominant excitation process for Hα emission in the rich H− condition such as in ionic plasma. We observed a significant change in Hα reduction distribution due to change in the bias voltage, which is used to suppress the electron influx. Small Hα reduction in higher bias is likely because the production of negative ions is suppressed by the potential difference between the plasma and PG surface.

To investigate the flows of charged particles in front of the plasma grid (PG) in a negative hydrogen ion source, the information of the local densities of electrons and negative hydrogen ions (H-) are necessary. For this purpose, the laser photodetachment is applied for pure hydrogen plasmas and Cs-seeded plasma in a 1/3-size negative hydrogen ion source in NIFS-NBI test stand. The H- density obtained by photodetachment is calibrated by the results from cavity ring-down (CRD). The pressure dependence and PG bias dependence of the local H- density are presented and discussed. The results show that H- density increases significantly by seeding Cs into the plasma. In Cs-seeded plasma, relativity exists between the H- ion density and plasma potential.

The velocity distribution function of hydrogen atoms is one of the useful parameters to understand particle dynamics from negative hydrogen production to extraction in a negative hydrogen ion source. Hydrogen atom temperature is one of the indicators of the velocity distribution function. To find a feasibility of hydrogen atom temperature measurement in large scale filament arc negative hydrogen ion source for fusion, a model calculation of wavelength-modulated laser absorption spectroscopy of the hydrogen Balmer alpha line was performed. By utilizing a wide range tunable diode laser, we successfully obtained the hydrogen atom temperature of ∼3000 K in the vicinity of the plasma grid electrode. The hydrogen atom temperature increases as well as the arc power, and becomes constant after decreasing with the filling of hydrogen gas pressure.

"It is very important to demonstrate the ability to sustain the plasma in a steady state on the Large Helical Device (LHD), which has external helical magnetic coils and is a superconducting device. The long pulse discharge experiment was carried out using the ion cyclotron range of frequencies (ICRF) heating mainly. The plasma discharge of 31 minutes and 45 seconds was achieved by a total injected heating energy of 1.3GJ. Swing of the magnetic axis to scatter the local heat load on the divertor plate was one of the key methods for the steady state operation. The repetitive hydrogen pellet injection was tried successfully to fuel the minority hydrogen ions for long pulse operation."

Isotope effect on H–/D– volume production is studied by measuring both VUV emission and negative ion density in the source. In a double plasma type source, under some discharge conditions, extracted D– currents are nearly the same as H– currents, although VUV emission intensity (corresponding to production of vibrationally excited molecules) in D2 plasmas is slightly lower than that in H2 plasmas. Considering the factor √2 due to mass difference, D– ion density in the extraction region of the source is higher than H– ion density. In another experiment with a rectangular arc chamber, axial distributions of H–/D– ion densities in the source are measured directly using a laser photodetachment method. Relationship between H–/D– production and plasma parameter control with using a magnetic filter (MF) is discussed. Furthermore, relative intensities of extracted negative ion currents are discussed compared with the negative ion densities in the source. Production and control of D2 plasmas are well realized with the MF including good combination between the filament position and field intensity of the MF. Extracted H– and D– currents depend directly on negative ion densities in the source. (© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

A powerful neutral beam system with six high current H− ion sources achieved a total power of 10.3 MW in the large helical device (LHD). We describe the present status of the ion sources, and the results of studies of cesium consumption, tungsten filament lifetime, and the velocity spectrum of the beam. Maximum power and beam energy in those ion sources range from 3.5 to 4.4 MW and 165 to 180 keV (design energy), respectively. The maximum H− current is ∼25 A. Much progress on beam acceleration and plasma uniformity was facilitated by a new accelerator and new arc power supply system with careful Cs seeding. Cs consumption for 1/2 year of operation ranges from 3.9 to 10.6 g per one source (over 2×104 shots). It is converted to the rate of 0.17–1.5 mg per one shot per source. It is considered that Cs deposition on the inside surface of the plasma source could be from 0.11 to 1.0 monolayer per one shot per source. This is of the same order as the expected optimum coverage of ∼0.7. Filament weight loss was me...

Ion heating experiments have been carried out in the large helical device using ECH (82.5, 84.0, 168 GHz, ≤1 MW), ICRF (38.5 MHz, ≤2.7 MW) and NBI (H° beam: 160 keV, ≤9 MW). The central ion temperature has been observed from the Doppler broadening of Ti XXI (2.61 A) and Ar XVII (3.95 A) x-ray lines, which are measured using a newly installed crystal spectrometer with a charge-coupled device. Recently, in ECH discharges, on-axis heating became possible. As a result, a high Te(0) of 6–10 keV and a high ion temperature of 2.2 keV were obtained at ne = 0.6×1013 cm−3. A clear increment of Ti was also observed with the enhancement of the electron–ion energy flow when the ECH pulse was added to the NBI discharge. These results demonstrate the feasibility towards ECH ignition. A clear Ti increment was observed also in ICRF discharges at low density ranges of (0.4–0.6)×1013 cm−3 with appearance of a new operational range of Ti(0) = 2.8 keV > Te(0) = 1.9 keV. In low power ICRF heating (1 MW), the fraction of bulk ion heating is estimated to be 60% of the total ICRF input power, which means Pi>Pe. Higher Ti(0), up to 3.5 keV, was obtained for a combined heating of NBI ( Te(0), whereas the Ti(0) remained at relatively low values of 2 keV in H2 and He NBI discharges due to less Pi. The main reasons for the high Ti achievement in the Ne discharges are: (1) 30% increment of deposition power, (2) increase in Pi/ni (five times, Pi/niPe/ne, Pi

An internal transport barrier (ITB) was observed in the electron temperature profile in the Large Helical Device [O. Motojima et al., Phys. Plasmas 6, 1843 (1999)] with a centrally focused intense electron cyclotron resonance microwave heating. Inside the ITB the core electron transport was improved, and a high electron temperature, exceeding 10 keV in a low density, was achieved in a collisionless regime. The formation of the electron-ITB is correlated with the neoclassical electron root with a strong radial electric field determined by the neoclassical ambipolar flux. The direction of the tangentially injected beam-driven current has an influence on the electron-ITB formation. For the counter-injected target plasma, a steeper temperature gradient, than that for the co-injected one, was observed. As for the ion temperature, high-power NBI (neutral beam injection) heating of 9 MW has realized a central ion temperature of 5 keV with neon injection. By introducing neon gas, the NBI absorption power was increased in low-density plasmas and the direct ion heating power was much enhanced with a reduced number of ions, compared with hydrogen plasmas.

The superconducting machine LHD has conducted long pulse experiments for four years to achieve long-duration plasmas with high performance. The operational regime was largely extended in discharge duration and plasma density. In this paper, the plasma characteristics, in particular, plasma performance and impurity behaviour in long pulse discharges are described. Confinement studies show that global energy confinement times are comparable to those in short pulse discharges. Long sustainment of high performance plasma, which is equivalent to the previous achievement in other devices, was demonstrated. Long pulse discharges enabled us to investigate impurity behaviour in a long timescale. Intrinsic metallic impurity accumulation was observed in a narrow density window (2–3×1019 m−3) only for hydrogen discharges. Impurity transport study by using active impurity pellet injection shows a long impurity confinement time and an inward convection in the impurity accumulation window, which is consistent with the intrinsic impurity behaviour. The pulsed neon gas injection experiment shows that the neon penetration into the plasma core is caused by the inward convection due to radial electric field. Finally, impurity accumulation control with an externally induced magnetic island at the plasma edge was demonstrated.

The study of negative ion production in the multicusp negative ion sources (MS) was done by the directed deposition of well-defined amount of cesium into MS, by the cesium recovery from the polluted layers in the MS and by the plasma grid masking. The data obtained evidences the surface-plasma mechanism of negative ion production in the MS. It is shown that a dynamic cesium-tungsten coverage on a plasma grid surface provides the enhanced H− production in the MS. The Cs+W coverage is produced on the hot plasma grid surface due to coadsorption of cesium and of tungsten, evaporated from filaments. The permanent flux of cesium to plasma grid coverage is produced by the thick Cs+W reservoir with a high (75%) percentage of cesium on the cold anode surface. A relatively high cesium seed with a rate of about 20 mg/1 h/30 shots operation is necessary to support the increased H− production in the regular MS. Electrode processing by an additional discharge recovers the cesium from the aged coverage and enhances the ...

The Large Helical Device-Neutral Beam Injector (LHD-NBI) system uses 40 A×180 keV (10 s) negative hydrogen ion sources. Three studies of the H− ion source are reported, i.e., conditioned status, co-accelerated electron beam component, and weight loss of tungsten filaments. Total injection numbers through experimental cycles sum up to ∼8000 shots with four ion sources on two beam lines. Injection power reached a maximum of 5.2 MW. The maximum energy achieved was 171 keV. A pulse length up to 80 s was reached with 0.5 MW. Electron fraction in the accelerated H− beam was evaluated by measuring the heat load on the electron beam dump. Total fraction of the power in the electron beam component deposited on the dump was evaluated to be 1.5%–6% with up to ∼1 Pa of H2. Stripped electrons which were produced inside acceleration gap comprised the majority compared to electrons extracted from source plasma. The weight loss of tungsten filament was 0.01–0.103 mgr/shot/filament. The total loss was ∼0.5% after ∼6500 sh...

The Large Helical Device (LHD) experiments have started after a construction period of eight years, and two experimental campaigns were performed in 1998. The magnetic field was raised up to 2.75 T at a magnetic axis position of 3.6 m at the end of the second campaign. In the third campaign, started in July in 1999, the plasma production with ECH of 0.9 MW and auxiliary heating with NBI of 3.5 MW have achieved an electron temperature of 3.5 keV and an ion temperature of 2.4 keV. The maximum stored energy has reached 0.75 MJ with an averaged electron density of 7.7×1019 m-3 by hydrogen pellet injection. The ICRF heating has sustained the plasma for longer than 2 s and the initial stored energy of the NBI target plasma has increased from 0.27 MJ to 0.335 MJ. The major characteristic of the LHD plasma is the formation of the temperature pedestal, which leads to some enhancement of energy confinement over the ISS95 scaling law. The confinement characteristic is gyro-Bohm and the maximum energy confinement has reached 0.28 s. The LHD has also shown its high potentiality for steady-state operation by realizing a 22 s discharge in the second campaign.

A compact cesium deposition system was used for direct deposition of cesium atoms and ions onto the inner surface of the 1/3 scale hydrogen negative ion source for the large helical device-neutral beam injection (LHD-NBI), system. A small, well defined amount of cesium deposition in the range of 3–200 mg was tested. Negative ion extraction and acceleration were carried out both in the pure hydrogen operation mode and in the cesium mode. Single Cs deposition of 3–30 mg to the plasma chamber has produced temporary 2–5 times increases of H− yield, but the yield was decreased within several discharge pulses to the previous steady-state value. Two consecutive 30 mg depositions done within a 3–5 h/60 shot interval, produced a similar temporary increase of H− beam, but reached a large H− yield steady-state value. Deposition of larger 0.1–0.2 g Cs portions with a 20–120 h/150–270 shot interval improved the H− yield for a long (2–5 days) period of operation. Directed depositions of Cs to the various walls of the p...

High power negative-ion-based neutral beam injectors for the LHD were constructed and started in operation in September 1998. Two beam lines with a pair of ion sources are installed tangentially in opposing direction. The maximum energy of the beam is 180 keV for hydrogen, and the total injection power is designed to be 15 MW with the pulse length of 10 seconds. One of the specific features of the negative-ion-based NBI system is the requirement on the vacuum pressure in order to avoid the stripping loss of negative ions as well as to realize the optimum gas line density for neutralization, which results in the long beam line. The system was then designed carefully by considering the focal length of the ion sources, neutralizing efficiency, and geometrical and re-ionizing loss of the beam. It is also considered the error magnetic field produced by the ferromagnetic materials that are indispensable for NBI for the magnetic shield over the long neutralizing region.

To understand the plasma characteristics in the extraction region of hydrogen negative ion sources is very important for the optimization of H− extraction from the sources. The profile of plasma density in the extraction region is analyzed with a 2D PIC modeling of the NIFS-R&D H− sources. The simulation results make clear the physics process forming a double-ion plasma layer (which consists of positive H+ and negative H− ions) recently observed in the Cs-seeded experiments of the NIFS-R&D source in the vicinity of the extraction hole and the PG (plasma grid).

The H-mode initiated by the control of the rotational transform profile in CHS exhibits a very rapid transition (in 100 s or less) in NBI heated plasmas. There is no obvious difference in transition behaviour between hydrogen and deuterium. Langmuir probe measurement has revealed that the rapid change in floating potential occurs at the transition, but the change follows the formation of an edge transport barrier. The presence of an surface just inside the last closed flux surface and a sawtooth crash triggered by internal modes play an important role in the H-mode transition of CHS.

Studies of a multicusp source were recently done at the National Institute for Fusion Science by plasma grid masking. The maximal H− ion yield is ∼1.4 times greater for the shutter mask case than that for the standard source. Negative ion current evolution during the cesium feed to the masked plasma grid evidenced that about 60% of negative ions are produced on the shutter mask surface, while about 30% are formed on the plasma grid emission hole edges, exposed by cesium with the mask open.

The Cavity Ringdown Technique (CRD) is applied for negative hydrogen ion (H−) density measurement in H− source for the neutral beam injector. The CRD is one of the laser absorption techniques. Nd:YAG pulse laser was utilized for negative-hydrogen-ion photodetachment. The H− density related to extracted H− beam was successfully observed by a fixed position CRD. A two-dimensional movable CRD has been developed to measure the H− density profile. Measured profiles were consistent with expected profiles from the H− production area in pure hydrogen and cesium seeded plasmas. By applying absorption saturation in the optical cavity, negative hydrogen ion temperature was evaluated and was confirmed as being a similar value measured with other diagnostics.

A large negative hydrogen ion source with a rod-type magnetic filter was investigated for the construction of a neutral beam injection (NBI) system for the large helical device (LHD). A 1 3 - scale device was constructed and investigated. Three main phenomena are observed on seeding a small amount of cesium into the ion source. These are enhancement of the hydrogen negative ion (H−) current, a reduction in the electron beam extracted with the H− beam, and a decrease in the operating gas pressure. The enhancement factor of the H− current is 3–4, and a total H− current of 8.5 A is obtained from 20 × 40 extraction apertures 9 mm in diameter. This current corresponds to a current density of 17 mA cm−2. A multi-wall structure forming a lattice, called an egg-box cell, is installed between the magnetic filter rod and the plasma grid. The electron current with this cell is half as low as the electron current without the cell, but the H− current is hardly affected. This phenomenon is observed when cesium is seeded in the source chamber. The electron temperature and its density increase by a factor of 2 on strengthening the cusp field. Experiments to obtain a higher H− current density were carried out using the improved are plasma. A hydrogen negative ion current of 54 mA cm−2 was extracted from a restricted aperture of 36 extraction holes. The H− current was extracted from 20 × 40 full apertures of the ion source with strong cusp field. A total H− current of 10.4 A was obtained from these apertures. The optimum filling pressure is less than 1.0 Pa.

Wavelength spectra of Balmer-α light from plasmas in the extraction region of the Large Helical Device-R&D negative ion source, or the LHD one-third ion source have exhibited a blue shift as a negative bias voltage was applied to the plasma grid. The blue shift increased as the negative bias voltage with respect to the local plasma potential was increased. The measured spectra were compared with the velocity distributions of surface reflected hydrogen atoms calculated by atomic collisions in amorphous target code. The arc power and the source H2 pressure also affected the shift and broadening in the observed Balmer-α spectra. The possibility of identifying the negative hydrogen ions produced at the low work function plasma grid surface by high resolution spectroscopy is discussed.

The state of hydrogen plasma in the extraction region in a hydrogen negative-ion (H−) source for NBI has been investigated. We clearly observe an improvement of H− density owing to the surface production effect with Cs seeding. H− ions are widely distributed in the extraction region which is obtained by movable cavity ring down (CRD). We confirm a negative ion rich plasma with a few electrons in the extraction region, which state is important for reduction of electron contamination in extracting beam. An extraction area is reached 30mm from the PG surface, which is measured by a 2D imaging diagnostic for Hα emission. We find the insensitive area for H− extraction at the PG surface between the apertures. Negative ions produced at the surface are considered tohave been supplied in the extraction region. The flow velocity of H− ions is obtained by a four-pin Langmuir probe using a photodetachment technique with an Nd:YAG laser. H− ion flows from the plasma grid surface, and its direction drastically changes at 20mm from the production surface. This flow behavior is considered to be an important characteristic for improving H− density in the extraction region.

Hydrogen negative ion density measurements are required to clarify the characteristics of negative ion production and ion source performance. Both of laser photodetachment and cavity ring down (CRD) measurements have been implemented to a field-effect-transistor based radio-frequency ion source. The density ratio of negative hydrogen ions to electrons was successfully measured by laser photodetachment and effect of magnetic filter field on negative ion density was confirmed. The calculated CRD signal showed that CRD mirrors with >99.990% reflectivity are required and loss of reflectivity due to cesium contamination should be minimized.

An energetic helium neutral beam is involved in the beam neutralization measurement system of alpha particles confined in a DT fusion plasma. A full size strong-focusing He(+) ion source (2 A, the beam radius of 11.3 mm, the beam energy less than 20 keV). Present strong-focusing He(+) ion source shows an emittance diagram separated for each beamlet of multiple apertures without phase space mixing, despite the space charge of a beamlet is asymmetric and the beam flow is non-laminar. The emittance of beamlets in the peripheral region was larger than that of center. The heat load to the plasma electrode was studied to estimate the duty factor for the ITER application.

A magnetic multipole plasma source and the accelerator for H− ion source is installed in the distribution of the externally imposed magnetic field which is quasiuniform over the plasma electrode. As the magnetic field of the permanent magnets in this geometry works not only as a so‐called magnetic filter for electron cooling, but also as a deflection field to the accompanied beam electrons, there is a possibility for making the fabrication of the H− source simpler/easier due to no use of small magnets in the accelerator. The characteristics of the ion source such as H− extraction, optimization in the magnetic field, the plasma parameters, etc. are investigated. The integrated magnetic flux of 230–320 G cm should be determined by the condition which optimizes the volume production in the source plasma. Achievable current density of H− is 6 mA/cm2 with 15 keV at the target. Large amounts of the electrons, however, are accelerated altogether to a full energy. They can be stopped with a retarding potential. I...

A cesiated hydrogen hollow cathode (CHC) was tested for plasma injection in the multicusp negative ion source (MS). The CHC arc with hydrogen feed and cesium seeding through the CHC volume was explored. One cathode unit (40 mm length, 19 mm in diameter, emission opening area 1–3 mm2) with no special cooling provided the MS discharge operation with direct current up to 30 A, and up to 60 A in the long-pulse mode. High efficiency of negative ion production in the MS discharge, driven by a CHC plasma injection was recorded.

The production efficiency of negative ions at tungsten surface by particle reflection has been investigated. Angular distributions and energy spectra of reflected hydrogen ions from tungsten surface are calculated with a Monte Carlo simulation code ACAT. The results obtained with ACAT have indicated that angular distributions of reflected hydrogen ions show narrow distributions for low‐energy incidence such as 50 eV, and energy spectra of reflected ions show sharp peaks around 90% of incident energy. These narrow angular distributions and sharp peaks are favorable for the efficient extraction of negative ions from an ion source equipped with tungsten surface as negative ionization converter. The retained hydrogen atoms in tungsten lead to the reduction in extraction efficiency due to boarded angular distributions.

A systematic study of the extraction probability of surface‐produced H− ions from the ion source has been done with H− and neutral Monte‐Carlo transport modeling. Without the effects of the flow velocity of background H+ ions and neutrals, the extraction probability (E.P.) depends strongly on the strength of the magnetic filter field BMF for the low gas pressure condition. On the other hand, the E.P. does not depend so much on BMF for the high gas pressure condition. Neutral transport modeling shows that H2 neutral possibly has a relatively high velocity (∼1 km/s). The effect on the E.P., however, is not significant even for the high gas pressure regime.

The potential distribution near the extraction region in negative ion sources for the plasma with the surface produced negative ions is studied analytically. The potential is derived analytically by using a plasma‐sheath equation, where negative ions produced on the Plasma Grid (PG) surface are considered in addition to positive ions and electrons. A negative potential peak is formed in the sheath region near the PG surface for the case of strong surface production of negative ions or for low energy negative ions. Negative ions are reflected by the negative potential peak near the PG and returned to the PG surface. This reflection mechanism by the negative potential peak possibly becomes a factor in negative ion extraction. It is also indicated that the potential difference between the plasma region and the wall decreases by the surface produced negative ions. This also has the possibility to contribute to the negative ion extraction.

This paper describes the development of the neutral beam (NB) systems on JT‐60SA, where 30–34 MW D0 beams are required to be injected for 100 s. A 30 s operation of the NB injectors suggests that existing beamline components and positive ion sources on JT‐60U can be reused without the modifications on JT‐60 SA. The JT‐60 negative ion source was modified to improve the voltage holding capability, which leads to a successful acceleration of 2.8 A H− ion beam up to 500 keV of the rated acceleration energy for JT‐60SA.

The aim of the Front End Test Stand (FETS) project is to demonstrate that chopped low energy beams of high quality can be produced. FETS consists of a high power Penning Surface Plasma Ion Source, a 3 solenoid LEBT, a 3 MeV RFQ, a chopper and a comprehensive suite of diagnostics. This paper briefly outlines the status of the project, hardware installation and modifications. Results from experiments running the H− ion source at 2 ms pulse length are detailed: the discharge current is varied between 20 A and 50 A. The discharge repetition rate is varied between 12.5 and 50 Hz. Hydrogen and Caesium vapour flow rates are varied. The effect of electrode surface temperature and beam current droop are discussed. Peak beam currents of over 60 mA for 2 ms pulse length can be achieved. Normalised r.m.s emittances of 0.3 πmm.mrads at the exit of the LEBT are presented for different source conditions.

The Japan Proton Accelerator Research Complex (J-PARC) was extensively damaged by the March 2011 earthquake. The ion source suffered some damage, such as dew condensation on its surface due to flooding in the accelerator tunnel, vacuum leakage due to viewing port breakage, and vacuum pump failure due to the strong quake. The restoration work was accomplished on schedule, and the linac beam operation was restarted on December 9, 2011. The ion source supplies the beam to the accelerator without serious earthquake after-effects. The availability of the ion source is one of the important issues because it is directly related to the operation time of the J-PARC. To evaluate the lifetime of the present ion source, we began two months of beam operation. The results of four recent beam runs show that the ion source is capable of continuous operation for approximately 1,200 h with the beam current of 17 mA. To enhance the ion source availability, we attempted to decrease the maintenance time to replace consumable elements. By unitizing the replacement elements and keeping them under the vacuum condition until just before installation, the maintenance was completed within 24 h without any beam performance degradation of the ion source.

Optical emission spectroscopy (OES) and cavity ring‐down spectroscopy (CRDS) systems are installed in a 1/3‐scaled negative hydrogen‐ion source at the National Institute for Fusion Science testbed to investigate the dynamics of H− ions in the extraction region near the plasma grid. The signal form of the H− ion density rapidly drops after beam extraction on applying a low‐bias voltage. A similar signal drop appears in the intensity of the hydrogen Balmer‐line emission measured by OES and is caused by decreasing atomic hydrogen produced by mutual neutralization effects between H− and H+. Shot trend of the beam currents are similar to the H− density and Hα/Hβ in the extraction region, which increases twice as large immediately after Cs seeding. We observe a linear correlation between the H− density and the inclination of Hα/Hβ which allows for experimentally benchmarking the OES measurement with that of CRDS. Thus, this approach is used for estimating the H− density by OES in negative‐ion sources for high‐e...

The effect of the cesium on hydrogen negative ions is discussed using the vacuum ultraviolet emissions (VUV) from vibrational states of hydrogen molecules. The VUV spectrum from 90 to 165 nm is related to the H− production from the dissociative attachment of electrons to hydrogen molecules. The VUV spectra in hydrogen plasmas are measured in the extraction region of a negative ion source for neutral beam injector. Under the pure hydrogen discharge, the cesium vapor is introduced into the ion source to enhance the hydrogen negative ion density. When the same arc power of ∼99 kW is applied in both with and without cesium admixture cases, the ratio of the observed VUV spectrum without to with cesium is not changed clearly. Therefore the production process of H− related to the wall/electrode surface would be enhanced rather than the volume production process.

Characteristics of radio frequency (RF) plasma production were investigated using a FET inverter power supply as a RF generator. High density hydrogen plasma was obtained using an external coil wound a cylindrical ceramic tube (driver region) with RF frequency of lower than 0.5 MHz. When an axial magnetic field around 10 mT was applied to the driver region, an electron density increased drastically and attained to over 10{sup 19} m{sup -3} in the driver region. Effect of the axial magnetic field in driver and expansion region was examined. Lower gas pressure operation below 0.5 Pa was possible with higher RF frequency. H{sup -} density in the expansion region was measured by using laser photo-detachment system. It decreased as the axial magnetic field applied, which was caused by the increase of energetic electron from the driver.