Status of ion sources at National Institute of Radiological Sciences^a) Open Access

A mighty hydrogen bomb was tested at Bikini Atoll in 1954 and 23 Japanese fishermen were severely affected by radiation sickness from its fallout. At that time, the uses of atomic energy, radiation, and radioactive substances were just starting to be applied in Japan. In this tense situation that put great attention on radiation among the Japanese people—who already carried the history of August 1945 in their minds. The National Institute of Radiological Sciences (NIRS) was founded by the Japanese Government in 1957 as a core research institute concerning the interactions between radiation and human beings. The main aims of NIRS are comprehensive research and development on: (1) the effects of radiation on the human body; (2) protection from radiation, including diagnosis and treatment of radiation injuries; and (3) medical applications of radiation. NIRS has a research hospital, and medical doctors, biologists, chemists, physicists, environmental scientists, engineers, and related staff join in carrying out these research aims at one institute. There are about 800 staff members. The total budget was about 14 000 million Japanese yen in fiscal year 2010, of which 80% was directly funded by the Japanese Government.

In the Great East Japan Earthquake which occurred on 11 March 2011, a nuclear crisis developed at the Tokyo Electric Power Company's Fukushima Dai-ichi Nuclear Power Station (NPS) due to damage caused by tsunami following the earthquake. Radioactive materials have been discharged into the environment. NIRS is playing an important role in handling the serious consequences of the nuclear accident, including the dispatch of specialists, admission to the NIRS hospital of NPS workers suspected of receiving high dose exposure, the measurement and estimation of radiation exposure to Fukushima Prefecture residents, offering a health counseling hot line, and so on. NIRS has reported these activities to among others, the World Health Organization, the United Nations Scientific Committee on the Effects of Atomic Radiation, the Organization for Economic Co-operation and Development, and the International Atomic Energy Agency.

Ion beam technology is a powerful tool for life science studies. It provides various radiation sources for experiments on radiation effects and technical methods for material analysis. Ion beam technology is also utilized for medical treatment and diagnosis. In order to fulfill its research and development aims, NIRS maintains some accelerators: they are two electrostatic tandem accelerators, one large heavy-ion accelerator complex consisting of two synchrotrons and four linacs for heavy-ion radiotherapy, and three cyclotrons. Various types of ion sources supply ion species required by the accelerators, and they are supporting wide applications of ion beams. Their status is reported here.

The Proton-induced x-ray emission Analysis System in Tandem Accelerator facility (PASTA) has been constructed;¹ it is shown schematically in Fig. 1. There are three experimental ports: a conventional port for parallel analysis of up to 12 samples; a micro-beam scanning port with an estimated beam size of 0.1 μm; and a 2 mm diameter in-air port. Elemental analyses in animal and plant samples done using PASTA reveal the behavior for uptake and accumulation of radioactive materials in the environment. The facility can also provide a micro-beam which is 1 μm or less using the Single Particle Irradiation System to Cell (SPICE).² It is utilized for precise irradiation experiments such as investigation of the bystander effect of radiation.

The accelerator 1.7 MV TANDETRON³ was delivered by High Voltage Engineering Europe Corp. (HVE) in 1999 to replace an old 3 MV Van de Graaff accelerator. A conventional, well-established duoplasmatron ion source that provides sufficient beams of H⁻ and He²⁺ ions. Typical beam currents for ¹H⁺ and ⁴He²⁺ are 25 and 1 μA, respectively. Operation time has been 12 h for 240 days per year and the source has provided excellent service in experiments. Its routine maintenance is necessary every 3 months.

A Neutron exposure Accelerator System for Biological Effect Experiments (NASBEE) (Ref. 4) supplies an intense proton beam on a target to introduce neutron beams into a specific pathogen free (SPF) animal experiment room. It can be used to investigate the effects of neutrons on cells or animals with a relative low dose of about 200 mSv.

The 2.0 MV TANDETRON was also delivered by HVE in 2006. It has a multi-cusp ion source originally developed by TRIUMF.⁵ A typical D⁻ output current is 600 μA and the filament lifetime is less than 500 h. The accelerated D⁺ beams impact a Be target (3 mm thickness) producing neutrons through the ⁸Be(d, n) reaction. A typical target current is 500 μA, and the realized dose rate is 2.18 Gy/h with an area of ϕ240 mm at a distance 1170 mm from the target.

The project to develop the world's first medical dedicated heavy-ion synchrotron was started in 1984 as the Heavy-Ion Medical Accelerator in Chiba (HIMAC).⁶ Since HIMAC was fully developed by NIRS as an accelerator complex, it was necessary to establish a new division of accelerator physics and to bring in accelerator researchers from outside NIRS. The construction was completed in 1993 and carbon ions were selected as the ion species for the first clinical trial mainly due to its good biological dose distribution. HIMAC has two different types of ion sources, a penning ion source, named NIRS-PIG⁷ and a 10 GHz electron cyclotron resonance ion source, named NIRS-ECR; although the NIRS-PIG is not suitable in terms of lifetime and maintenance.

NIRS-ECR was developed in collaboration with the Institute for Nuclear Study (INS) at the University of Tokyo and the Tokyo Institute for Technology (TIT). Increasing the beam intensity of C⁴⁺ was studied in order to alleviate the load on the linacs as the injector for the synchrotron. As a result, NIRS-ECR showed better results with hydrocarbon gases such as CH₄ than with carbon oxide gases such as CO. This behavior seems similar to observations in gas mixing effect. However, hydrocarbon gases result in a large deposition of carbon on the surface of everything in the vacuum chamber. This causes heat-up of the walls leading to increasing pressure in the vacuum chamber. This problem was solved by the cooling solution summarized in Ref. 8. Since varying the condition of the deposition leads to instability or bad reproducibility, it is important that the source must be always operated under the same “dirty” condition. Therefore, maintenance is not necessary within a few years. The total beam time for the treatments reached 1700 h in fiscal year 2010 for 770 treatments.

Many countries planned heavy-ion radiotherapy facilities in the 1980s; however, most plans were aborted except for the HIMAC. The successful and promising clinical data obtained at HIMAC^9,10 are awakening again worldwide interest in heavy-ion radiotherapy facilities.¹¹ However, hospitals require a more compact and less expensive model. NIRS has designed a hospital-specific facility and developed prototypes of parts for important key technologies, including ion sources. A series of compact permanent magnet ECR ion sources, named Kei,¹² Kei2,¹³ and KeiGM,¹⁴ were developed to meet the requirements to produce C⁴⁺ for the newly designed facility. The design concept was to make a copy of the successful magnetic field configuration of the NIRS-ECR. The gas mixing effect was studied precisely through collaboration with the Kernfysisch Versneller Instituut (KVI),¹⁵ and a record output of 1 mA beam of C4+ ions at an extraction voltage of 40 kV was obtained after carefully optimization. KeiGM was installed at the Gunma University Heavy Ion Medical Center (GHMC). NIRS-Kei2 fulfilled its role as a prototype, and it was installed into the local backup injector of HIMAC.

HIMAC also has as an essential task to operate as a facility for performing fundamental experiments. It has one biological experiment room with exactly the same conditions as the treatment rooms, and there are three general purpose beam lines and two radioactive isotope beam lines. Three user stations can each simultaneously use a different beam. The total time for experiments was 5300 h in fiscal year 2010.

NIRS-PIG produces light ions from solid materials with the sputtering method such as B, Mg, and Si. In addition, an 18 GHz ECR ion source, named NIRS-HEC, extends the range of available ion species. The most important ion species to be provided are iron ions. An Fe-ion beam is desired by biomedical researchers to study the risks of space exploration due to galactic cosmic rays. The Metal Ions from Volatile Compound method (MIVOC) was adapted to produce Fe and several other ions in collaboration with the Accelerator Engineering Corporation Ltd. (AEC).¹⁶ 

For the production of intermediate charge-state ions, optimization of the extraction configuration is most effective. For HIMAC, first the extraction voltage was increased to a maximum of 60 kV. Then the position of the extraction electrode was optimized. However, NIRS-HEC did not obtain the desired beam intensity. Then the dependence on the radial magnetic field was studied in collaboration with Institute of Nuclear Research at the Hungarian Academy of Science (ATOMKI).¹⁷ From this work, two assumptions were verified: highly charged ions are localized inside the ECR zone; and the ion trajectory is tightly bound to the magnetic flux line from the ECR zone to the extraction aperture. Too strong a concentration of the magnetic flux at the extraction aperture causes beam losses due to the large space charge effect, and too weak a concentration causes beam losses on the plasma electrode and unpredictable local heating. An optimum radial size of ECR plasma may exist for a given size of the extraction aperture as shown in Fig. 2. In the case of the NIRS-HEC, the intensity was increased by a factor of four.

The NIRS-HEC has the function to be a backup for the NIRS-ECR. Both MIVOC and carbon-ion production always deposit carbon on the surface of the vacuum chamber walls. From this viewpoint, any tricks and techniques to improve the highly charged ion production or to produce ions from solid materials have to work under this dirty condition without additional maintenance. Considering the particle losses in the ECRIS, the electron flux dominates the axial losses and the ion flux dominates the radial diffusion; both fluxes compensate in the conducting surface material forming the so-called “Simon short circuit”. Since the loss of electrons is strongly confined near the central axis, the loss of ions is also confined near the electron-dominated region due to the negative potential formed by electrons. Other regions, that is almost all of the side walls, have no flux as shown in Fig. 2. Applying positive and negative repelling potentials against the ion and electron fluxes will lead to breaking of the short circuit. It is theoretically expected that the loss fluxes are reduced and this leads to better performance and beam intensity increases. In addition, this multi-electrode technique is promising even under the dirty condition. This technique was studied in collaboration with KVI, ATOMKI, Toyo University, and Osaka University.

The two-frequency heating technique is also promising under the dirty condition. Observations showed that the two-frequency heating technique improved the beam intensity under the conditions of sufficient power and precise frequency tuning for the additional microwave.⁸ This seemed mainly due to preventing occurrence of plasma instabilities. The effect appeared to be most important for the higher charge states. Fig. 3 shows the improvement of output current of Ar¹³⁺. It should be noted that a frequency of 18.00 GHz was optimized when a single Traveling Wave Tube amplifier (TWT) was used. However, when the mixture of klystron amplifier (KLY) and TWT was used, a TWT frequency of 17.88 GHz was better than 18.00 GHz.

Two commercial compact cyclotrons, the Sumitomo Heavy Industry Ltd. (SHI) model HM-18¹⁸ and the Japan Steel Works Ltd. model BC-2010 have been installed at NIRS. The main use for them is radioisotope production as probes for positron emission tomography (PET). These cyclotrons are connected with NIRS's own radiopharmaceutical system of automatic synthesis and delivery. More than 100 kinds of molecular probes are stably provided to clinical and experimental PET scanners including animal PET scanners. About 60 probes have been approved for clinical research. In order to perform various research activities, NIRS integrates this radiopharmaceutical technology with medical facilities such as the hospital and animal breeding facilities for rodents and monkeys. The use of PET-based molecular imaging to elucidate brain mechanisms of mental disorders such as schizophrenia is expected to contribute to the development and evaluation of new medicines. For cancer diagnosis, PET has also been established as one of the most efficient tools. ¹¹C-methionine is mainly utilized for diagnosis and follow-up of carbon-ion radiotherapy before and after the treatment for the majority of patients.

Both the compact cyclotrons have negative PIGs. Typical beam intensities of H⁻ and D⁻ are 20 and 10 μA, respectively. Routine maintenance for the ion source is necessary every 3 months. The total produced radioactivity was about 4500 GBq in fiscal year 2010. Clinical syntheses totaled over 1000 of which ¹¹C-methionine comprised about 80%.

The K = 110 cyclotron, NIRS-930, was designed by Thomson CFS and manufactured by SHI in 1973. The first aim of NIRS-930 was originally for use in particle radiotherapy. Fast neutron and proton radiotherapies were carried out in 1975 and 1979, respectively. Since the termination of these radiotherapies, radioisotope production still forms an important use of NIRS-930; however, the two compact cyclotrons are taking over that task. Presently three fourths of the operation time is used for experiments on biology, physics, and material science as shown in Fig. 4.

In order to extend the kinds of ion species that can be produced and to increase the beam intensity, an external beam-injection line was installed at NIRS-930.¹⁹ Kei was coupled to this injector after finishing its research and development. The output currents are 200 eμA of H⁺, 200 of H₂⁺, 200 of D⁺, 70 of ¹²C⁴⁺, 25 of ¹³C⁵⁺, 10 O⁵⁺, and 2 of Ne⁶⁺ eμA, respectively. The total operation time was about 1600 h in fiscal year 2010. In addition, for future medical applications of radioactive nuclear beams such as ¹¹C, a high ionization efficiency electron string ion source (ESIS) is being studied in collaboration with the Joint Institute for Nuclear Research (JINR).²⁰ 

The Great East Japan Earthquake of March 11, 2011 caused damage in the PASTA and SPICE setups. The earthquake was observed near NIRS to be a grade 5-lower on the Japan Meteorological Agency seismic intensity scale. The accelerator itself was almost undamaged; however, the alignment of the magnets was distorted at many places. Turbo molecular pumps were broken. A small leakage of insulation gas was found. The scheduled operations were canceled and the system is now being repaired with this work to be completed at the end of 2011. Fortunately, HIMAC and the cyclotron facilities had almost no damage and are operating. It seemed that the weights of the components in PASTA were relatively lighter than in the other facilities and their mounting was not as strong; these factors contributed to the greater damage.

The research and development projects at NIRS are based on collaborations with many organizations and persons. The authors would like to thank AEC, ATOMKI, GHMC, INS, JINR, KVI, Osaka University, SHI, TIT, Toyo University, and all of other collaborators.