Appl. Phys. Lett., Vol. 72, No. 21, 25 May 1998
Hausmann et al.
2721
rable to broad beam ion implantation where the thermally
induced defect annihilation prevents an amorphization of the
silicon wafer. It is also helpful to compare the present results
to ion beam induced crystallization and amorphization ex-
periments performed by Linnros et al.17 with pulsed ion
beams. As explained by Jackson18 at a low pulse frequency
the results are comparable to experiments with a constant
beam with the same current. For high pulse frequencies the
results correspond to experiments with a constant beam but
with the averaged current of the pulsed beam. If this knowl-
edge is transferred to our results, although the conditions are
rather different, this means that for short dwell-times the
effective ion arrival rate per pixel is reduced. In comparison
with long dwell-times, when the steady-state case is as-
sumed, the effective current density for the 40ϫ40 m2 im-
planted areas with short dwell-times is a factor of 2.5ϫ105
smaller ͑for 1 s͒. This means that the effective current den-
sity is now on the order of A cmϪ2 which is comparable to
the current density of broad beam ion implantation.
In summary 70 keV Co2ϩ was implanted with a focused
ion beam system into a heated silicon target to form continu-
ous cobalt disilicide layers. A strong influence of the dwell-
time on the CoSi2 layer formation was found. For sufficiently
short dwell-times it was possible to form continuous CoSi2
layers while for long dwell-times it was shown that the in-
creased damage accumulation of the silicon target prevented
the formation of a continuous CoSi2 layer.
The authors would like to thank the Deutsche For-
schungsgemeinschaft for their financial support under Con-
tract No. Te 250/1-1. They gratefully acknowledge the SEM
FIG. 3. ͑a͒ Schematic view of the sample preparation with small implanted
areas using a conventional RBS/channeling setup. ͑b͒ RBS/channeling spec-
tra of an as-implanted sample. Implantation with a low dose 35 keV Coϩ
FIB into Si͑111͒ at 430 °C and a dwell-time of 1 s. The aligned spectra is
multiplied with the ratio of the total area of the opened hole in the resist to
the implanted area (ϭ25/9). ͑c͒ Same as in ͑b͒ except a dwell-time of 250
s. The damage level for 1 s dwell-time implantation is reduced in con-
trast to that of 250 s.
¨
investigations by E. Christalle and Dr. R. Muller and the
technical assistance by I. Beatus.
1 A. E. White, K. T. Short, R. C. Dynes, J. P. Garno, and J. M. Gibson,
Appl. Phys. Lett. 50, 95 ͑1987͒.
2 S. Mantl and H. L. Bay, Appl. Phys. Lett. 61, 267 ͑1992͒.
3 S. Mantl, M. Dolle, St. Mesters, P. F. P. Fichtner, and H. L. Bay, Appl.
Phys. Lett. 67, 3459 ͑1995͒.
balt is possible even at those high doses of about 1017 cmϪ2
.
4 S. Mantl, Mater. Sci. Rep. 8, 1 ͑1991͒.
5 T. Aoki, K. Gamo, S. Namba, T. Shiokawa, K. Toyoda, H. Okabayashi,
H. Mori, and H. Fuijita, Nucl. Instrum. Methods Phys. Res. B 39, 291
͑1989͒.
In addition, broad beam implantation of cobalt in random
direction ͑7° tilt angle͒ was successfully simulated12 by high
dose TRIM ͑HDTRIM͒,13 indicating that the implantation pro-
files are governed by collisional effects rather than diffusion.
A similar dwell-time effect was also found for GaAs. Low-
dose Ga FIB implantation with different dwell-times has
been investigated14 on GaAs wafers at room temperature and
the corresponding samples were analyzed by RBS channel-
ing. The samples implanted with longer dwell-times showed
a higher degree of damage than those with short dwell-times.
Tamura et al.15 showed for FIB implantation of boron into
silicon that the electrical activity after annealing is higher for
samples implanted with a slow scanning speed. This is an
indication of a higher damage for longer dwell-times. The
same was found by Madokoro et al.16 for FIB implantation
of phosphorus into silicon.
6 L. Bischoff, J. Teichert, E. Hesse, D. Panknin, and W. Skorupa, J. Vac.
Sci. Technol. B 12, 3523 ͑1994͒.
L. Bischoff, E. Hesse, D. Janssen, K. F. Naehring, F. Notzold, G. Schmidt,
7
¨
and J. Teichert, Microelectron. Eng. 13, 367 ͑1991͒.
8 F. Herrmann and D. Grambole, Nucl. Instrum. Methods Phys. Res. B 104,
26 ͑1995͒.
9 N. M. Zimmerman, J. A. Liddle, A. E. White, and K. T. Short, Appl. Phys.
Lett. 62, 387 ͑1993͒.
10 G. Bai and M.-A. Nicolet, J. Appl. Phys. 70, 649 ͑1991͒.
11
¨
¨
¨
R. Jebasinski, PhD thesis, Universitat Koln/Forschungszentrum Julich,
1993.
12 R. Jebasinski, S. Mantl, L. Vescan, and Ch. Dieken, Appl. Surf. Sci. 53,
264 ͑1991͒.
13
¨
¨
¨
H. Kohlhof, PhD thesis, Universitat Koln/Forschungszentrum Julich,
1989.
14 C. R. Musil, J. Melngailis, S. Etchin, and T. W. Haynes, J. Appl. Phys. 80,
3727 ͑1996͒.
15 M. Tamura, S. Shukuri, S. Tachi, T. Ishitani, and H. Tamura, Jpn. J. Appl.
Phys., Part 2 22, L700 ͑1983͒.
Thus, the damage of the silicon lattice has a strong in-
fluence on the CoSi2 layer formation. For long dwell-times
the damage accumulation of the silicon lattice during the
cobalt implantation prevents the formation of continuous
16 Y. Madokoro, S. Shukuri, K. Umemura, and M. Tamura, Nucl. Instrum.
Methods Phys. Res. B 39, 511 ͑1989͒.
17 J. Linnros, W. L. Brown, and R. F. Elliman, Mater. Res. Soc. Symp. Proc.
100, 369 ͑1988͒.
CoSi2 layers. For short dwell-times the results are compa-
18 K. A. Jackson, J. Mater. Res. 3, 1218 ͑1988͒.
138.251.14.35 On: Tue, 23 Dec 2014 15:29:31