Operation Status of the ECR Ion Source at Gunma Universitya)
H. Souda,1,b) S. Yamada,1 T. Kanai,1 E. Takeshita,2 M. Muramatsu,3 A. Kitagawa,3
M. Kanazawa,4 H. Izumiya,5 Y. Kano,5
1Gunma
University Heavy Ion Medical Center, Maebashi, Gunma, 371-0044, Japan
Cancer Center, Yokohama, 241-8515, Japan
3National Institute of Radiological Sciences, Chiba, 263-8555, Japan
4SAGA Heavy Ion Medical Accelerator in Tosu, Tosu, Saga, 841-0071, Japan
5Accelerator Engineering Corporation, Chiba, 263-0043, Japan
2Kanagawa
(Presented XXXXX; received XXXXX; accepted XXXXX; published online XXXXX)
(Dates appearing here are provided by the Editorial Office)
An ECR ion source of Gunma University Heavy Ion Medical Center (GHMC), so-called KeiGM [1] has been
operated for cancer therapy and physical/biological experiment since 2010. KeiGM produces typically 230 A
of 10 keV/u C4+ ions from CH4 gases. The vacuum pressure is kept between 1.2×10-4 and 1.7×10-4 Pa so as to
suppress the pulse-to-pulse current fluctuation within 10%. The extraction electrode is cleaned every 6-8
months in order to remove deposited carbon, which increases the leak current and discharge. In order to
investigate the possibility of long-term operation without such maintenances, oxygen aging for the cleaning of
the extraction electrode has been tested in the test bench. The same-designed ion sources at NIRS and SAGAHIMAT are also operated with stable C4+ current, which are suitable for the continuous operation for cancer
therapy.
I. INTRODUCTION
In these decades, heavy ion cancer therapy has been
developed with favorable treatment results and less damage
to normal tissues mainly in National Institute of
Radiological Sciences (NIRS). For further widespread use
of heavy ion therapy, a development project of compact
heavy ion accelerator had been promoted by Japanese
government since 2004. In this project, all permanent
magnet type compact ECR ion source, Kei2, was developed
[1][2] in NIRS. After the development of the compact heavy
ion accelerator was competed, Gunma University Heavy
Ion Medical Center (GHMC)[3] was established and the
first model of the compact medical accelerator was built in
the center. As one of the key components of the accelerator,
the first product of Kei2 type ion source, KeiGM [4], was
manufactured for GHMC. After the commissioning, GHMC
has treated 804 patients from March 2010 to August 2013.
KeiGM has been operated as the only one ion source in the
center and has continued providing carbon beam stably.
II. Operation of KeiGM
5 GyE/min (1.3×109 pps) with a spread out Bragg peak
(SOBP) of 140 mm at the maximum and a uniform
irradiation field of 150×150 mm. The accelerator consists of
an ECR ion source, an RFQ linac of 600 keV/u, an IH-DTL
of 4 MeV/u, a synchrotron of 63.3 m circumference, and
four irradiation rooms (3 for treatment and 1 for
development of scanning irradiation). The schematic view
of KeiGM is shown in Fig. 1. It is optimized for the highest
production rate of C4+, with an extraction voltage of 30 kV.
Both ECR and mirror field are generated by permanent
magnets; the maximum field is 0.8 T. Typically 300 W of
10 GHz microwave is applied through a travelling wave
tube (TWT) amplifier. Produced C4+ beams are converted
into C6+ by a charge-stripper foil placed at the downstream
of the IH-DTL.
B. OPERATION AND MAINTENANCE
Ring Permanent
Magnet
Iron Yoke
Ceramic Insulator
Microwave
feed port
A. Specification of KeiGM
The compact medical accelerator in GHMC provides
400 MeV/u (Maximum) C6+ beams with a fluence of
Gas Inlet
Ground
Electrode
a)
Contributed paper published as part of the Proceedings of the 15th
International Conference on Ion Source, Chiba, Japan, September, 2013.
b)
souda@gunma-u.ac.jp.
Sextupole
FIG. 1. (Color online).Schematic view of KeiGM.
The ion source was manufactured by Sumitomo Heavy
Industries and was installed in the accelerator room of
GHMC in February 2009. Commissioning of the facility
started in July 2009, and the therapeutic irradiation started
on 16 March 2010. In the current operation procedure, the
ion source is turned on around 7:45 in the following order:
the cooling water, the gas flow, the extraction voltage, and
the microwave. The output current is stabilized within 3
minutes after the microwave is applied. After it was turned
on, the extracted beam is utilized for daily measurements,
treatments, and experiments. The ion source continues to
provide beams until the daily shutdown around midnight,
therefore the operation time every day is approximately 16
hours.
carbon. The interval of the maintenance is determined by
the rate of contamination progress, which depends on the
gas flow and extracted current. With the current operation
condition and the maintenance interval of half year, the ion
source in GHMC has not experienced severe increase of
discharges.
There were only three major failures by which the
accelerator operation was stopped as follows:
● On 11 August 2010, the TWT amplifier was broken
after operating for 14100 hours. Treatment operation had
been continued by the same type amplifier borrowed from
NIRS during the repair.
● On 6 December 2010, the extraction high voltage was
tripped due to the insufficient aging after a planned outage.
The reflection power of the TWT amplifier was increased
after the trip and was repaired by the initialization of the
signal generator.
● On 13 October 2011, the high voltage of the Einzel lens
could not be kept higher than 17 kV due to frequent
discharges. It was caused by the surface contamination of
the ceramic insulator, and was recovered by cleaning of the
insulator and the replace of the crimping terminal of the high
voltage.
III. RESEARCH IN THE TEST STAND
A. Operation of the test stand
FIG. 2. (Color online). A long term trend of the beam current, vacuum
pressure, RF frequency, RF Power, and Biased disk voltage. Arrows
indicate maintenances including the cleaning of the electrode.
Figure 2 shows the long-term trend of the beam current
from February 2009. In 2010, the operation parameters of
RF frequency and biased disk voltage were not optimized
yet, there were several times of decrease of the beam current.
After 2011, the beam current had been kept between 200
and 250 A with slight tuning of the RF frequency, the RF
power, and the biased disk voltage. The CH4 gas flow is
controlled to keep the vacuum pressure between 1.2×10-4
and 1.7×10-4 Pa for the stable plasma density. The
fluctuation of the beam current is reduced within 10% by
such optimizations of the operation parameters.
The maintenance of the ion source is performed every
half year. In the maintenance, the plasma chamber is opened
and the electrode is changed to a cleaned one. This
maintenance is important to avoid the increase of discharge
due to the electrode surface contamination by the deposited
For research of the stability in a long-term operation, a
test stand of the ion source was constructed in the hospital
building of Gunma University. The test stand has the same
design as KeiGM. In the test stand, the ion source was
operated with more gas flow to increase the carbon
deposition on the electrode surface. The typical CH4 gas
flow and C4+ beam current on the Faraday cup is 0.1 sccm
and 350 A, around 4 and 1.5 times larger than those in the
treatment facility. With such a high gas flow and beam
current, the base current of the extraction electrode grows
from 1 mA to 10 mA and the frequency of the discharge also
increase within around 3 months.
B. Oxygen aging for electrode surface cleaning
In order to recover the base current and the withstand
voltage, aging operation with the use of oxygen gases has
been tested according to the preceding studies with oxygenincluding gas operations[5][6]. Two oxygen aging
experiments were carried out in August 2012 and June 2013.
In 2012, the oxygen gas of 0.2 sccm was introduced into the
chamber and an extraction voltage of 35 kV was applied,
but the microwave was not applied. After 40 hours of aging,
The base current was reduced from 4.4 mA to 2.5 mA, but
extracted beam current did not significantly change. In 2013,
the microwave of 30 W was applied with an extraction
voltage of 30 kV to generate the oxygen beam to clean the
electrode. The gas flow was 0.04 sccm for 33 hours at first
and later 0.2 sccm for 15 hours. The result aging with
microwave is rather clear with a reduction of the base
current from 8 mA to 1 mA with an output current of
220 A and the withstand voltage was recovered from 28 to
30 kV. The surface contamination of the electrode was
removed to some extent as shown in Fig. 3. Further
optimization of the aging condition and a systematic
investigation of the stability of the ion source will be carried
out in this test stand.
accelerators.
TABLE I: Current operation parameters of Kei type ion sources: Kei2 in
NIRS, KeiGM in Gunma University, and SAGA-HIMAT. Kei2 is used for
an experimental operation with large current and KeiGM in Gunma and
Saga is used for therapeutic operations. The base current represents the
extraction high voltage current without microwave.
Beam Current [A]
Pulse Width [ms]
Gas flow [sccm]
RF Frequency [GHz]
RF Fwd. Power [W]
RF Ref. Power [W]
Biased Disk Voltage[V]
Base Current [mA]
Einzel Voltage [kV]
Vacuum Pressure
Ion Source [×10-4 Pa]
Einzel lens [×10-4 Pa]
LEBT [×10-4 Pa]
IV. Individual Difference of Kei-type source
The original Kei2 source has been operated in NIRS,
and a new heavy ion therapy facility, SAGA-HIMAT [7]
employs the same type ECR ion source. The current
parameters of these sources are listed in Table I. Fig. 4
shows the trend graph of the C4+ beam current in SAGAHIMAT from December 2013. The output current of the
source is stable after the commissioning. There was only
Beam Current [A]
FIG. 3. (Color online) Extraction ground electrode before (left) and
after (right) the oxygen aging with microwave. Red arrows indicate
the positions with the minimum gaps.
350
300
250
200
150
100
50
0
1
51
101
151
201
251
day
FIG. 4. (Color online) Trend of extracted beam current in SAGA-HIMAT.
one major trouble: an increase of discharge due to the
carbon deposition mentioned above occurred in June 2013.
It was repaired by changing the electrode and the ceramic
insulator. These stable operations in three same type sources
with few troubles indicate the sufficient performance as a
standard production model of the ion source of compact
medical accelerators.
Summary
KeiGM has been operated for 4 years in GHMC. The
beam current is almost stabilized around 230A with
optimized parameters of microwave, biased disk voltage
and vacuum conditions. Although the maintenance interval
of half year is enough to prevent the increase of discharges,
cleaning the electrode by oxygen aging is under
investigation for further stable operations. The same type
sources in NIRS, Gunma, and Saga shows the similar and
stable performances, which are suitable for medical
NIRS
317
40
0.059
10.180
477
171
59.7
11.2
22.75
Gunma
238
30
0.023
9.956
236
16
50.0
0.17
21.63
Saga
260
50
0.0255
10.031
261
4
24.9
0.45
27.71
4.7
1.5
0.10
1.40
0.67
0.09
1.80
0.93
0.18
Acknowledgement
This work is partially supported by The Program for
Cultivating Global Leaders in Heavy Ion Therapeutics and
Engineering by Ministry of Education, Culture, Sports,
Science and Technology (MEXT) of Japan and Gunma
Heavy Ion Joint Research Center for Industry-academicgovernment Collaboration by Japanese Science and
Technology Agency (JST). The authors would greatly thank
to the operators of the Mitsubishi Electric Corporation for
the everyday operation and continuous recording of the
operation data.
1
M. Muramatsu, A. Kitagawa, Y. Sakamoto, S. Sato, Y. Sato, H. Ogawa,
S. Yamada, H. Ogawa, Y. Yoshida, and A. G. Drentje, Rev. Sci. Instrum.
76, 113304 (2005).
2
M. Muramatsu, A. Kitagawa, Hirotsugu Ogawa, Y. Iwata, K.
Yamamoto, S. Yamada, Hiroyuki Ogawa, T. Fujimoto, Y. Yoshida, and
A. G. Drentje, Rev. Sci. Instrum. 77, 03A307
3
S. Yamada, K. Yusa, M. Tashiro, K. Torikai, in Proceedings of NIRSEtoile Joint Symp. 2009 on Carbon Ion Radiotherapy (2009).
4
M. Muramatsu, A. Kitagawa, S. Hojo, H. Miyazaki, T. Ueno, K. Sawada,
M. Tsuchiyama, S. Ueda, Y. Kijima, K. Torikai, S. Yamada, in
Proceedings of ECRIS2010 (2010).
5
K. Tinschert, R. Iannucci, and R. Lang, Rev. Sci. Instrum. 79, 02C505
(2008)
6
G. Ciavola, S. Gammino, L. Celona, F. Maimone, A. Galatà, M. Pullia, R.
Monferrato, C. Bieth, W. Bougy, G. Gaubert, O. Tasset, Proc. of EPAC08,
415 (2008)
7
M. Endo et al., in Proceedings of PTCOG49, (2010).