Hideshi Muto, Center of General Education, Tokyo University of Science, Suwa, 5000-1, Toyohira, Chino Nagano 391-0292, JapanY. Ohshiro, Y. Kotaka, S. Yamaka, S. Watanabe, H. Yamaguchi, S. Shimoura, Center for Nuclear Study, University of Tokyo, 2-1 Hirosawa, Wako Saitama 351-0198, Japan

M. Kase, S. Kubono, K. Kobayashi, M. Nishimura, Nishina Center for Accelerator-Based Science, RIKEN, 2-1 Hirosawa, Wako Saitama 351-0198, JapanM.Oyaizu, Institute of Particle and Nuclear Studies, High Energy Accelerator Research Organization, 1-1 Oho, Tsukuba Ibaraki 305-0801, Japan

T. Hattori, Heavy Ion Cancer Therapy Center, National Institute of Radiological Sciences, 49-1 Anagawa, Inage Chiba 263-8555, Japan

INTRODUCTION

A grating monochromator with a photomultiplier has been used for beam tuning at the Center for Nuclear

Study Hyper-ECR ion source [1,2]. Hyper-ECR ion source has been successfully used as an injector of the

multi-charged ion beams of high intensity for RIKEN Azimuthal Varying Field (AVF) cyclotron [3]. Light

intensity observation is an especially useful technique for an identification of the ions of the same charge to

mass ratio in the plasma [4]. These ions are difficult to separate by an analyzer magnet. Before the

operation of multi-charged metal ion beams chamber baking (degassing from the plasma wall) must be

done to obtain a required vacuum condition. At the beginning low RF power (~100W) is fed to the residual

gas in the plasma chamber, and a degassing process is conducted with increasing RF power gradually until

the vacuum gauge reading is settled (1~5 x 10-5 Pa order) to start a metal rod insertion into the plasma

chamber. In this paper we describe the sublimation pump effect of Mg and Ti ions of ECR ion source during

chamber baking and beam tuning.

EXPERIMENTAL

RESULTS and DISCUSSIONS

Figure 1: Light intensity spectrum of the

residual gas ions after baking for three

hours. The peaks of the spectrum are

mostly Fe I and Fe II. The pressure and

microwave power were 5.7 x 10-5 Pa and

100 W, respectively.

24Mg8+ ion beam tuningFigure 1 shows the optical line spectrum of the

Hyper-ECR ion source under plasma chamber

baking after three hours from the start. A vacuum

gauge reading was 5.7 x 10-5 Pa. A drain current

(an extraction current) was 12 mA. RF power was

100 W. In this figure most of all peaks were Fe I

and Fe II. There were some C, N and O optical

lines in the spectrum. However, those lines were

all disturbed by Fe I and Fe II strong lights, and

therefore it was difficult to separate those.

Relative intensities of those Fe I and Fe II are

quite strong. Figure 2 shows the optical line

spectrum of the ECR plasma during 24Mg8+ ion

beam tuning. A vacuum gauge reading was 1.7 x

10-5 Pa. A drain current was 1. 8 mA. RF power

was 611 W. Line intensities of Fe I and Fe II

almost disappeared, and Mg light intensities

appeared. Especially, Mg VIII line spectrum

(l=279.64 nm) was clearly obtained to identify the

existence of 24Mg8+ ions in the ECR plasma. In

this figure new Fe I (l=559.2 nm) and Fe II

(l=570.3 nm) light intensities were observed. A

non-magnetic stainless steel cover was used for a

smooth heat transfer to the tip of the MgO rod

from the plasma. The edge of the cover was

melted by plasma as shown in fig.3. Therefore,

these two strong Fe lines are thought to be

present because of the melted stainless cover.

Figure 2: Optical line spectrum during24Mg8+ ion beam tuning. The shape of the

spectrum drastically changed from that of

residual gas plasmas. Fe light intensities

of residual gas disappeared, and Mg I, III,

IV and VIII lines were clearly observed.

Figure 4: Charge state distribution of Mg, Helium,

Oxygen and residual gas ions. The ion source was

tuned for production of 24Mg8+ ions. The beam

intensity of 24Mg8+ was 22 emA. Small peaks of

near multi-charged 24Mg ions were those of 25Mg

and 26Mg.

[1] H. Muto et al., Rev. Sci. Instrum. 84, 073304 (2013).

[2] H. Muto et al. , Rev. Sci. Instrum. 85, 02A905 (2014).

[3] Y. Ohshiro et al., Rev. Sci. Instrum. 85, 02A912 (2014).

[4] H. Muto et al., Rev. Sci. Instrum. 84, 073304 (2013).

[5] www.nist.gov/pml/data/asd.cfm for NIST Atomic Spectra Data base.

Figure 7: Optical line spectrum during 48Ti13+ ion beam tuning

with He supporting gas. Ti XIII line was clearly observed.Figure 8: Charge state distribution of multi-charged Ti

beams. The beam intensity of 48Ti 13+ was 1.0emA.

During plasma chamber baking observed light intensities were mostly Fe I and Fe II. Fe ions were relatively

heavy and not easy to extract from the plasma chamber. Those atoms were present for a long time in the

vacuum chamber. Therefore, stainless steel is thought to be an unsuitable material for a plasma chamber to

extract multi-charged ions. Aluminum or Magnesium based light alloy is better for plasma chamber materials

for degassing and extraction.

CONCLUSIONS

REFERENCES

Observation of sublimation effect of Mg and Ti ions at the Hyper-Electron Cyclotron Resonance ion source

24Mg8+ and 48Ti13+ ions have been produced in the 14.2 GHz Hyper-ECR ion source. The structure and

present operation condition of the ion source are described in Ref. 3. At the beginning of the chamber

baking RF power of ~100 W was fed to the residual gas of the plasma chamber. Extraction voltage was set

to 10 kV. Then a vacuum gauge reading rapidly dropped down to less than 10-4 Pa from 10-5 Pa order, and a

brake-down of the high voltage power supply happened because of a huge extraction current. Several hours

later the extraction voltage was recovered, and vacuum gauge reading also reached 10-5 Pa order. RF

power gradually increased to ~ 600 W until obtaining a required vacuum condition (1~5 x 10-5 Pa), and a

row extraction current of less than 2 mA. After baking of the plasma chamber, a pure metal or an oxidized

metal rod was gradually inserted into the chamber without an excessive heat. An excessive heat causes a

brake-down of the power supply because of a huge extraction current. The RF power was ranging between

500 and 600W for a highly multi-charged ion production. Argon, Neon, Oxygen and Helium gases were

used as supporting gases to keep the plasma condition stable. A grating monochromator (JASCO CT-25C)

and a photomultiplier (Photosensor module H11462-031, Hamamatsu Photonics) were used for a light

intensity observation during chamber baking and beam operation. Beam resolution of the grating is 0.1 nm

(FWHM). L-37 and R-64 filters are used for preventing both second and third order light signals.

Wavelengths of the observed lines were determined in accordance with the NIST Atomic Spectra Database

[5].

m

Mg

VII

I

Mg I

Mg

III

O I

IM

g I

Mg

IV

Mg

VM

g I

II

O I

I (F

e II

)

He

II

m

Fe

II

Fe

I

Figure 3: MgO rod with a stainless steel cover.

Oxygen and Helium lines also appeared in the

spectrum because Helium and Oxygen gases were

fed during the beam tuning as supporting gases to

keep a required plasma condition. Figure 4 shows the

charge distribution of ion beams extracted from the

Hyper-ECR ion source. The ion source was tuned for

the production of the 24Mg8+ ions. Beam intensity of

the 24Mg8+ ions was set at 22 emA.

48Ti13+ ion beam tuningFigure 5 shows light intensity spectrum of residual

gas plasma just after inserting TiO2 rod. Vacuum

condition became stable after insertion of TiO2 rod.

However, a sublimation effect is too strong, and a

production of highly multi-charged Ti ions was

complex. Then, a pure Ti rod instead of TiO2 rod was

used to make 48Ti13+ beam. Figure 6 shows the optical

line spectrum during 48Ti13+ beam tuning with Ne

supporting gas. Natural Neon gas has 9.22 % of 22Ne

and 22Ne6+ is the same charge to mass ratio of 48Ti13+

(q/m=0.27). Beam intensity of these beams was 3.0

emA. Ti XIII light intensity was observed clearly.

However, it was difficult to separate these ions by a

magnetic analyzer. Then, Ar was selected as a

supporting gas instead of Ne to separate 48Ti13+ ions.

However, the plasma condition became unstable, and

the desired 48Ti13+ ions were not obtained. As a result,

He gas was selected as a supporting gas in order to

produce 48Ti13+ ion beam. Figure 7 shows light

intensity spectrum during 48Ti13+ ion beam tuning with

He supporting gas. Ti XIII line was clearly observed

and 48Ti13+ beam intensity was 1.0 emA. Beam

intensity of 48Ti11+ was also set at 9.6 emA. Figure 8

shows the charge distribution of these multi-charged

Ti beams.

m

Mg

VII

I

Mg I

Mg

III

O I

IM

g I

Mg

IV

Mg

VM

g I

II

O I

I (F

e II

)

He

II

m

Fe

II

Fe

I

Figure 5: Optical line spectrum

of residual gas plasma just after

inserting TiO2 rod. Vacuum

gauge reading was 1.9 x 10-5

Pa and RF power was 100 W.

5

10

15

200 300 400

0

500 600 700 800

Wave length [nm]

Lig

ht

inte

nsit

y [m

V]

65

6.1

3 T

i I

48

5.5

Ti

I, I

II

46

4.3

N I

I 45

3.2

Ti

I 4

33

.87 T

i II

I

41

0.3

Ti

I

37

5.7

6 T

i II

I

27

9.3

7 T

i II

55

5.4

2 A

r II

Figure 6: Optical line spectrum

during 48Ti13+ ion beam tuning

with Ne supporting gas. Ti XIII

line was clearly observed.

However, Ne lines also were

present.

5

10

15

200 300 400

0

500 600 700 800

Wave length [nm]

Lig

ht

inte

nsit

y [m

V]

70

3.1

4 T

i II

I

66

8.0

1 T

i II

65

4.4

7 T

i I

64

0.1

Ne I

61

4.1

5 T

i II

58

5.2

5 N

e I

, T

i I

53

5.4

N I

, N

e I

50

3.6

4 T

i I

47

2.5

6 N

e I

V

44

5.5

6 N

e I

41

3.1

Ti

XI

39

9.9

Ti

I, N

e I

36

9.9

Ti

I 36

4.2

Ti

I 35

6.7

2 B

III

3

48

.1 N

e I

I, T

i I

334

.5 T

i I

27

7.4

Ti

XII

I 26

2.7

7 T

i N

e I

I

Ti

I (7

28.0

7 n

m)

Ti

I (6

56.1

3 n

m)

Ti

I (6

67.7

nm

)

Ti

II (

64

0.6

2 n

m)

He I

(5

87

.56 n

m)

Ti

I +

Mo I

(5

46

.56 n

m)

Ar

II (

55

5.4

2 n

m)

Ti

II (

54

1.1

nm

)

Ti

I (5

01.4

2 n

m)

Ti

I (5

04.8

2 n

m)

Ne I

(48

5.9

6 n

m)

Ti

III

(49

1.4

3 n

m)

He I

I (4

68

.58

nm

)

Ti

III

(43

3.8

7 n

m)

Ti

III

(41

0.0

5 n

m)

Ti

I (3

96.4

3 n

m)

Ti

I (3

88.8

nm

)

Ti

II (

36

1.2

7 n

m)

Ti

II (

32

0.2

5 n

m)

Ti

XII

I (2

77

.4 n

m)

5

10

15

200 300 400

0

500 600 700 800

Wave length [nm]

Lig

ht

inte

nsit

y [m

V]

0 20 30 40 50 60 70 80 90

Analyzer current [A]

0

2

4

6

8

10

12

14

16

18

20

Beam

cu

rren

t [e

mA

]

N6+

N5+ N4+

⑬

(1.0)

⑪

(9.6)

( )emA

q = 48Tiq+