Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://danp.sinp.msu.ru/Articles_GSIB/nimb_GCIBSystem_Yamada.pdf Äàòà èçìåíåíèÿ: Fri Oct 7 22:12:40 2005 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 22:33:32 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: magnetic field

ARTICLE IN PRESS

Nuclear Instruments and Methods in Physics Research B xxx (2005) xxx­xxx www.elsevier.com/locate/nimb

Development of a size-selected gas cluster ion beam system for low-damage processing
Noriaki Toyoda *, Shingo Houzumi, Isao Yamada
Laboratory of Advanced Science and Technology for Industry, University of Hyogo, 3-1-2 Kouto, Kamigori, Hyogo 678-1205, Japan Available online

Abstract A high-current gaseous cluster ion beam (GCIB) system with a magnetic size separation function has been developed. This system was equipped with a permanent magnet with a magnetic flux density of 1.2 T. The system consists of a magnetic spectrometer with sliding detector and sample holder on a guiding rail perpendicular to the incoming cluster beam axis. Depending on the momentum of the cluster ions, positioning of the sample on the guiding rail allows bombardment of the sample with the desired cluster size. In this study, preliminary results demonstrating cluster mass spectra and mass resolution are presented. Finally, scanning tunneling microscopy is used to characterize the morphology of individual cluster impacts after irradiation. ñ 2005 Elsevier B.V. All rights reserved.
PACS: 36.40.Wa; 79.20.Rf; 41.85.þp Keywords: Cluster ion; Size-selection; Damage; Crater formation

1. Introduction Molecular dynamics simulations have demonstrated that cluster size plays an important role in damage formation induced by energetic cluster impacts [1]. Even though the total acceleration energy is the same, the damage formation, crater size, secondary electron and secondary ion yields
Corresponding author. Tel.: +81 791 58 0428; fax: +81 791 58 2666. E-mail address: ntoyoda@lasti.u-hyogo.ac.jp (N. Toyoda).
*

in solid materials are all different depending on the size of the impinging cluster ion [2]. There are cluster ion beam systems that can select cluster size, however, the ion currents are very low or the systems are operated in a pulsed mode. Also, the cluster size has typically been relatively small. The difficulties of mass separation drastically increase with cluster size. As the average cluster size of the current gas cluster ion beam (GCIB) system is typically several thousands, it is difficult to perform experiments that require high ion-dose by size-selected cluster ion beams.

0168-583X/$ - see front matter ñ 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2005.07.079


ARTICLE IN PRESS

2

N. Toyoda et al. / Nucl. Instr. and Meth. in Phys. Res. B xxx (2005) xxx­xxx

Recently, we have developed a high-current GCIB system with size separation capabilities for the first time. This system is equipped with a permanent magnet having a magnetic flux density of 1.2 T. There is a sliding detector and sample holder on a guiding rail perpendicular to the incoming cluster beam axis. When the sample was positioned at a certain point on the rail, the desired size of cluster ion can be used to continuously irradiate the sample with high-current densities. In this study, preliminary results of mass spectra and performance of magnetic spectrometer are discussed. Then, size-selected Ar-GCIB was used to irradiate a graphite substrate to observe craters created by individual cluster ion impacts with a scanning tunneling microscope.

2. Results and discussion The permanent magnetic spectrometer was designed and built by Epion corporation to study the cluster size effects on GCIB processing. Fig. 1 shows schematic diagram of the size-selected GCIB irradiation system. The cluster ion beam source is a commercial Epion 200 lA GCIB source and evacuated by cryogenic vacuum pump. A permanent magnet is installed after the ionizer and a scanning detector is mounted on a linear motion rail to select a cluster size. The spectrometer was designed to analyze clusters with p/q = 0.6 GeV/ c, where p and q is a momentum and a velocity of light, respectively. The designed mass resolution is M/DM = 5 for a ±20 mm and ±20 mr input

beam. The magnet uses a NdFeB permanent magnet and achieves 1.2 T on the center axis. The magnet has a 50 mm gap with a 6.6° wedge angle to give an vertical focusing. The effective field length is 450 mm. Fig. 2 shows a comparison of measured and computed magnetic field values for one straight line path through the center of the magnet. Magnetic fields were measured by Hall probe measurements using a computer driven XY stage. The plot shows that the deviation between the computed and measured values is less than 1.3%. By using these calculated magnetic fields, the trajectories of cluster ions for various energies or momentums can be calculated. The momentum (p) of the particle is expressed as a function of the slit position (X), p = 208.39 · Xþ1.0226. Therefore, the atomic mass or cluster size of ions can be obtained from this momentum and the acceleration energy. Mass spectra of the Ar cluster ion beam were measured with this magnetic spectrometer and it is shown in the Fig. 3. The flow rate of the Ar gas into the nozzle was changed from 200 to 600 sccm. With increasing the flow rate, the cluster beam intensity and the peak cluster size position increased. This tendency is the same as the previous data obtained with time of flight mass spectrometer [3]. To evaluate the performance of the magnetic spectrometer, a time of flight (TOF) mass spectrometer was mounted on the scanning rail instead of the existed Faraday cup as shown in Fig. 1. The Ar cluster beam was chopped by a deflector and the flight length was 600 mm from the chopper. The ion current was amplified by a current ampli-

Cluster size selection Linear motion

Neutral cluster source

Ionizer
Permanent Magnet Sample holder Faraday Cup

Fig. 1. Schematic diagram of size-selected gas cluster ion beam irradiation system.


ARTICLE IN PRESS

N. Toyoda et al. / Nucl. Instr. and Meth. in Phys. Res. B xxx (2005) xxx­xxx
14000 12000 10000
Ampheres Calculation

3

B(z) (Gauss)

Hall Probe Measurement

8000 6000 4000 2000 0 -2000 -60

y
X
Origin is the center of the magnet gap

-40

-20

0

20

40

60

Magnet X axis (cm)
Fig. 2. Comparison of measured and computed magnetic field at the center of the magnet.

0.5 Va = 5.5 keV Ve = 400 V 0.4
1.0 0 1x10
5

Mass [a.m.u]
2x10
5

3x10

5

4x10

5

5x10

5

Ion current [µA]

0.3 500 0.2 300 0.1 250 0.0 0 200 5000 10000 15000 20000 350 600sccm

0.8

Before mass separation

Intensity [a.u.]

0.6

0.4

After mass separation

0.2

Cluster size [atoms/cluster]
Fig. 3. Mass spectra of Ar cluster ion beam measured with the magnetic spectrometer.
0.0 0 2000 4000 6000 8000 10000 12000

Cluster size [atoms/cluster]
Fig. 4. TOF spectrum of Ar-GCIB after mass separation with the permanent magnet.

fier and TOF spectra were stored in a digital oscilloscope. As the total ion currents varied depending on the slit position (X), the TOF spectra were normalized to the peak ion current. Fig. 4 shows TOF mass spectra of the 5 keV Ar cluster ion beam at the position (X) of 250 mm from the beam center. From Fig. 4, the peak cluster size was 1250 atoms/

cluster and the cluster size distributed ±250 atoms from the peak position. Compared to the original cluster size distribution, the cluster size distribution was quite narrowed and the cluster size distri-


ARTICLE IN PRESS

4

N. Toyoda et al. / Nucl. Instr. and Meth. in Phys. Res. B xxx (2005) xxx­xxx

bution was reduced by 1/20 from the original distribution. In Fig. 5, the mass resolution of the magnetic spectrometer obtained from TOF measurements is shown for both experimental and designed values. In the TOF measurement, M/DM was 3.5 at X = 250 mm (Ar cluster size of 1250), while that of designed valued was 4.5. These values are adequate for the irradiation experiments. In addition, the cluster ion current after mass filtering was several hundreds of nA at an Ar cluster size of 1250.

6
Design

5

4
Experiment

3

2

1

0

50

100

150

200

250

300

350

400

Position [mm]
Fig. 5. Designed and measured mass resolution (M/DM) of the magnetic spectrometer.

It was almost 100 times higher than that obtained by mass filtering with time of flight method. Thus, it was proved that the high-current and size-selected cluster beam was successfully obtained by using the magnetic spectrometer. Subsequently, irradiation experiments using size-selected GCIB were performed. To observe individual traces by GCIB impacts, graphite substrates were irradiated and subsequently observed by a scanning tunneling microscope (STM). Fig. 6 shows STM images of a highly oriented pyrolytic graphite (HOPG) surface after Ar cluster irradiation at 30 keV. Fig. 6(a)­(c) corresponds to the irradiation of Ar cluster ions of sizes 500, 1000 and 5000, respectively. As the total acceleration energy was 30 keV, the energies per atom were 60, 30 and 6 eV/atom, respectively. As reported in a previous study, large crater like structures were formed by Ar cluster ion impacts [4­6], which were observed in Fig. 6(a) and (b). However, with decreasing the energy per atom of cluster ions, these craters were not observed as shown in Fig. 6(c), which corresponds to the cluster size of 5000 (6 eV/atom). From molecular dynamics simulations of Ar cluster ion impact onto Si, it is known that the number of displaced Si atoms from their lattice cites suddenly decreased with increasing cluster sizes around 2000 atoms/cluster [7]. This effect is due to the decrease of energy per atoms in Ar cluster ions. The MD simulation and experimental results showed very good agreement for the damage formation. MD simulations also indicated that Ar

Fig. 6. STM images of HOPG surfaces irradiated with (a) Ar100, (b) Ar1000 and (c) Ar 30 keV.

Mass resolution [M/M]

5000

cluster ions at total acceleration energy of


ARTICLE IN PRESS

N. Toyoda et al. / Nucl. Instr. and Meth. in Phys. Res. B xxx (2005) xxx­xxx

5

cluster ion deposits 30% of its acceleration energy even though there is no damage formation. This indicates that there is a possibility of surface processing for solid surface without causing damage formation.

formation due to the decrease of energy per atoms of Ar cluster ions.

References
[1] T. Aoki, T. Seki, J. Matsuo, Z. Insepov, I. Yamada, Nucl. Instr. and Meth. B 153 (1999) 264. [2] I. Yamada, J. Matsuo, N. Toyoda, A. Kirkpatrick, Mater. Sci. Eng. R 34 (2001) 231. [3] N. Toyoda, M. Saito, N. Hagiwara, J. Matsuo, I. Yamada, in: Proceedings of 1998 International Conference on Ion Implantation Technology, 1999, p. 1234. [4] M.W. Matthew, R.J. Beuhler, M. Ledbetter, L. Friedman, Nucl. Instr. and Meth. B 14 (1986) 448. [5] T. Seki, T. Kaneko, D. Takeuchi, T. Aoki, J. Matsuo, Z. Insepov, I. Yamada, Nucl. Instr. and Meth. B 121 (1997) 498. [6] D. Takeuchi, T. Seki, T. Aoki, J. Matsuo, I. Yamada, Mater. Chem. Phys. 54 (1998) 76. [7] T. Aoki, J. Matsuo, G. Takaoka, Nucl. Instr. and Meth. B 202 (2003) 278.

3. Summary A size-selected cluster ion beam irradiation system was developed to study the cluster size effect for surface processing and damage formation. The cluster size distribution was narrowed by 1/ 20 from original the distribution by using the magnetic spectrometer. The mass resolution M/DM of the system was 3.5 at Ar cluster size of 1250, which was close to the designed values (4.5). The STM observation of irradiated graphite surfaces showed that there was a threshold cluster size for crater