Study of the ion implanted Fe depth distribution in Si after pulsed ion beam treatment

Abstract

The behaviour of Fe implanted into Si(100) during subsequent pulsed ion - beam treatment (PIBT) has been studied. A two-step 56Fe+ ion implantation at energies of 60 and 20 keV and total doses of 1016 - 2×1017 cm-2 was used. As crystallization due to PIBT took place, Fe segregated towards the surface of the samples for the lower dose used (1016 cm-2) and diffusion into the bulk of the Si samples ccured for higher doses (1×1017 cm-2 and 2×1017 cm-2). The Fe concentration profile was shifted rigidly, without Fe losses. Both the movement of the Fe layer and the maximum concentration of Fe in the Si crystallized region were characterized by Rutherford backscattering spectroscopy (RBS) in combination with channelling (RBS/C).

Full text

JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS Vol. 9, No. 2, February 2007, p. 307 - 310
Study of the ion implanted Fe depth distribution in Si
after pulsed ion beam treatment
CH. ANGELOV, S. GEORGIEV*, B. AMOV, E. GORANOVAa, V. MIKLI b, I. DÉZSI c, E. KÓTAI c
Institute for Nuclear Research and Nuclear Energy, 72 Tzarigradsko Chaussee Blvd., 1784 Sofia, Bulgaria
a Central Laboratory for Solar Energy and New Energy Sources, Bulgarian Academy of Sciences, 72 Tzarigradsko
Chaussee Blvd., 1784 Sofia, Bulgaria
b Centre for Materials Research, Tallinn Technical University, Ehitajate 5, Tallinn 79086, Estonia
c KFKI Research, Institute for Particle and Nuclear Physics, H-1525 Budapest P.O. Box 49, Hungary
The behaviour of Fe implanted into Si(100) during subsequent pulsed ion - beam treatment (PIBT) has been studied. A two-
step 56Fe+ ion implantation at energies of 60 and 20 keV and total doses of 1016 – 2x1017 cm-2 was used. As crystallization
due to PIBT took place, Fe segregated towards the surface of the samples for the lower dose used (1016 cm-2) and diffusion
into the bulk of the Si samples occured for higher doses (1x1017 cm-2 and 2x1017 cm-2). The Fe concentration profile was
shifted rigidly, without Fe losses. Both the movement of the Fe layer and the maximum concentration of Fe in the Si
crystallized region were characterized by Rutherford backscattering spectroscopy (RBS) in combination with channelling
(RBS/C).
(Received November 1, 2006; accepted December 21, 2006)
Keywords: Fe+ implantation, Pulsed ion-beam treatment, Impurity distribution
1. Introduction
The conventional ion beam synthesis (IBS) technique
for the formation of semiconducting silicides consists of
two steps: ion implantation of metal ions into Si and
annealing. Several methods of annealing have been used
for the preparation of silicides: conventional furnace
annealing, rapid thermal annealing (RTA), laser annealing
and pulsed ion-beam treatment (PIBT) [1-4]. Among these
PIBT is very suitable because it affects only a local surface
layer of the material (~1μm) during a short time (< 1μs),
which eliminates the unwanted impurity diffusion into the
base material [4].
Ion implantation is a non - equilibrium process. An
excess of implanted atoms over the solid solubility in the
system results in segregation or precipitate formation after
annealing. This effect is considered to be important for the
formation of a stoichiometric silicide phase in Si by IBS.
Semiconducting β-FeSi2 is one of the most extensively
studied materials due to its potential applications for
optical sources and silicon-based optoelectronic
components.
In the present work the depth profiles of Fe ions of as-
implanted and PIBT-annealed Si wafers were studied by
Rutherford backscattering spectroscopy (RBS).
2. Experimental
Ion implantation was carried out in a type ILU-4 ion
accelerator. In order to form an unburied surface layer a
two-step 56Fe+ ion implantation in n-type (100) Si wafers
at energies of 60 and 20 keV was performed, using three
different doses: D1=5×1015, D2=5×1016 and
D3=1×1017 cm-2, resulting in total doses of 1×1016 cm-2,
1×1017 cm-2 and 2×1017 cm-2 , respectively.
The samples were subjected to PIBT from a TEMP
accelerator [4] using a high-power nanosecond ion beam
(80% of C+, 20% of H+, E = 300 keV, τ = 50 ns, and
W = 1.2 J cm-2; five pulses were used). The total dose of
C+ and H+ ions in the course of PIBT did not exceed
1×1014 cm -2 per pulse. The C and H ranges in Si are well
beyond the thicknesses of the initially formed amorphous
layers. The C range is about 703 nm with a straggle of 118
nm and the more penetrating H component has a range of
3.08 µm with a straggle of 157 nm in Si [5]. Computer
simulation of Si heating shows that during PIBT the melt
thickness amounts to 1 µm and the melt duration amounts
to 400 ns [4].
The depth distribution of ion-beam implanted Fe
atoms in Si wafers was studied by backscattering
spectrometry. RBS measurements were performed using a
3 MeV He+ beam from a Van de Graaff accelerator in
glancing angle geometry at 97°, under random and
channeling conditions.
3. Results and discussion
The initial 56Fe+ implantation profiles, shown in Fig. 1
were simulated by the SRIM computer code [5]. The
implanation conditions are given in Table 1. Fe peak
volume concentrations calculated by SRIM and
experimentally obtained from the RBS spectra of as-
implanted samples are in good agreement.

Ch. Angelov, S. Georgiev, B. Amov, E. Goranova, V. Mikli, I. Dézsi, E. Kótai
308
Fig. 1. Fe + Implantation profiles simulated by SRIM.
Table 1. Summary of the implantation conditions,
theoretical calculations and experimental results for the
peak Fe concentration of the as-implanted depth profiles.
Sample
Ion impl. 56Fe+
Energy,
Doses
SRIM
NFe
(%)
Exp.
NFe
(%)
D11
(as-impl.)
D12 (PIBT)
60 кеV,
5 × 1015 сm-2
plus
20 кеV,
5 × 1015 сm-2
5 4.6
D21
(as-impl.)
D22 (PIBT)
60 кеV,
5 × 1016 сm-2
plus
20 кеV,
5 × 1016 сm-2
32 23
D31
(as-impl.)
D32
(PIBT)
60 кеV,
1 × 1017 сm-2
plus
20 кеV,
1 × 1017 сm-2
52 49
RBS random and aligned (channeling (RBS/C))
spectra of a Fe implanted Si substrate with a D1 dose,
before and after annealing, are shown in Fig. 2 a,b,c. An
apparent difference was observed between the RBS
random and aligned spectra of Fe (Fig. 2(b) and Fig. 2(c),
respectively). The depth of the initially formed
amorphous layer (Fig. 2(b)) was estimated to be 108 nm.
The peak positions from a SRIM simulation, indicated in
the inset of Fig. 2(a) coincides with the measured
spectrum. During PIBT, the amorphous layer completely
recrystallizes. All implanted Fe atoms after PIBT are in
substitutional sites (Fig. 2(c)). Fig. 2(a) shows random
spectra of as-implanted and annealed samples. After
annealing, the Fe peak shifts to the surface. The
redistribution of Fe atoms to the Si surface (segregation),
characteristic of low-solubility impurities in
Si (N
Fe ~ 1015 cm-3 at 1000 oC [6]), is observed.
1200 1600 2000 2400 280
0
40
80
120
160
200
240
0
2
4
6
8
10
TRIM
60 keV Fe+
NFe (%)
Depth (nm)
10080604020 0
TRIM
20 keV Fe+
60 keV+20 keV Fe+ n-Si(100)
PIBT: W=1.2 J/cm2, N=5 pulses
Yield (counts)
Energy (keV)
as-implanted random
PIBT annealed random
Fe
Si
(a)
020406080100
0
10
20
30
40
50
60
D1
D2
D3
Fe CONCENTRATION ( at % )
DEPTH ( nm )
1200 1600 2000 2400 280
0
0
50
100
150
200
250 60 keV+20 keV Fe+ n-Si(100)
4He+ E=3 MeV
θ
=97o
(b)
Fe
impl
layer
Yield (counts)
Energy (keV)
as-implanted random
as-implanted aligned
Si
1200 1600 2000 2400 280
0
0
50
100
150
200
250
(c)
Si
60 keV+20 keV Fe+ n-Si(100)+PIBT
RBS: 4He+ E=3 MeV
θ
=97o
Yield (counts)
Energy (keV)
PIBT annealed random
PIBT annealed aligned
Fe
Fig. 2. RBS random and aligned spectra of implanted
Si with an ion dose D1.
In the case of the higher D2 dose, the Fe impurity
profile has a different behavior (Fig. 3(a)). Due to the high
sputtering yield for Fe in Si (2.39 at.ion-1 for 20 keV and
2.64 at.ion-2 for 60 keV [5]) the initial implantation profile
shifts to the surface. The maximum Fe concentration in the
initial depth profile is ~ 23 %. The higher Fe concentration
within the a-Si affects the redistribution during the
crystallization process. A large fraction of the implanted
Fe atoms diffuses toward the Si substrate. At the same
time, some of the Fe atoms diffused to the Si surface are
sputtered during the implantation process. Figs. 3(b) and
3(c) show RBS/C for the same samples. The depth of the
amorphous layer created after the implantation – 130 nm
(Fig. 3(b)) is close to that theoretically predicted by SRIM
– 132 nm. All of the deeper-diffused Fe atoms occupy
substitutional sites, while for the diffused atoms close to

Study of the ion implanted Fe depth distribution in Si after pulsed ion beam treatment
309
the surface, some of the Fe atoms (~21 %) are on
interstitial sites (Fig. 3(c)).
A similar behavior of the crystallization process after
PIBT of Fe implanted Si with the highest total dose of
D2 = 2×1017 cm-2 (Fig. 4(a)) is observed. The accumulation
of defects in the tail of the initial implantation profile leads
to a slight increase in the depth of the amorphous layer -
133 nm (Fig. 4(b)). The Fe concentration at the maximum
of the profile is close to the theoretical calculated value
(Table 1). Fig. 4(a) shows significant diffusion during
PIBT, toward the substrate. The RBS/C spectrum (Fig.
4(c)) reveals complete recrystallization of the Si matrix.
Similarly to the case of the D2 dose, the deeper diffused Fe
atoms occupy substitutional sites while the atoms close to
the surface are on interstitial sites. An indication of this is
the difference (~ 27%) between the aligned and random
spectra (Fig. 4(c)) estimated by Rutherford Universal
Manipulation Program (RUMP) [7]. A similar behavior of
Fe implanted atoms in Si after laser annealing was
observed by Bayasitov et al. [3].
1200 1600 2000 2400 280
0
0
50
100
150
200
250 60 keV+20 keV Fe+ n-Si(100)
PIBT: W=1.2 J/cm2, N=5 pulses
Fe
Si
Yield (counts)
Energy (keV)
as-implanted random
anneal random
(a)
1200 1600 2000 2400 280
0
0
50
100
150
200
250 60 keV+20 keV Fe+ n-Si(100)
4He+ E=3 MeV
θ
=97o
(b)
Fe
impl
layer
Yield (counts)
Energy (keV)
as-implanted random
as-implanted aligned
Si
1200 1600 2000 2400 280
0
0
50
100
150
200
250 60 keV+20 keV Fe+ n-Si(100)+PIBT
RBS: 4He+ E=3 MeV
θ
=97o
(c)
Si
Fe
Yield (counts)
Energy (keV)
PIBT annealed random
PIBT annealed aligned
Fig. 3. RBS spectra of the Fe-implanted samples with a total
dose of 1x1017 cm-2, before and after PIBT annealing.
1200 1600 2000 2400 280
0
0
50
100
150
200
250 60 keV+20 keV Fe+ n-Si(100)
PIBT: W=1.2 J/cm2, N=5 pulses
(a)
Fe
Si
Yield (counts)
Energy (keV)
as-implanted random
PIBT annealed random
1200 1600 2000 2400 280
0
0
50
100
150
200
250
60 keV+20 keV Fe+ n-Si(100)
4He+ E=3 MeV
θ
=97o
(b)
Fe
Si
impl.
layer
Yield (counts)
Energy (keV)
as-implanted random
as-implanted aligned
1200 1600 2000 2400 280
0
0
50
100
150
200
250 60 keV+20 keV Fe+ n-Si(100)+PIBT
RBS: 4He+ E=3 MeV
θ
=97o
Fe
Si
(c)
Yield (counts)
Energy (keV)
PIBT annealed random
PIBT annealed aligned
Fig. 4. RBS spectra of the Fe-implanted samples with a
total dose of 2
×
1017cm-2 before and after PIBT annealing.
4. Conclusions
During PIBT, significant redistribution of ion
implanted Fe atoms in Si takes place. Depending on the
initial Fe concentration, either segregation of the dopant to
the surface or diffusion into the Si substrate occurs.
For the lowest implanted dose (1×1016 cm-2) all
implanted Fe atoms are incorporated in substitutional sites
near the surface. For the intermediate dose (1×1017 cm-2)
implanted atoms diffuse into the Si and some of them
remain in interstitial sites (~ 23 %). For the highest
implanted dose (2×1017 cm-2), the fraction of Fe atoms in
the interstitial sites increases to 27 % after PIBT.

Ch. Angelov, S. Georgiev, B. Amov, E. Goranova, V. Mikli, I. Dézsi, E. Kótai
310
The results of a computer simulation agree well with
the experimental data on the iron depth distribution of the
as implanted samples.
The results of this investigation show the possibility
of forming structures with a high concentration of Fe
dopants by a non-equilibrium method – IBS.
Acknowledgements
This work was partly supported by OTKA Grant No.
T 34332.
References
[1] M. Sugiyama, Y. Maeda, Thin Solid Films, 381, 225
(2001).
[2] M. M. Mitan, D. P. Pivin, T. L. Alford, J. W. Mayer,
Appl. Phys. Lett. 78, 2727 (2001).
[3] R. M. Bayazitov, R. I. Batalov, I. B. Khaibullin, G. D.
Ivlev, I. Dézsi, E. Kotai, J. Phys. D: Appl. Phys. 37,
468 (2004).
[4] R. I. Batalov, R. M. Bayazitov, I. B. Khaibullin, E. I.
Terukov, V. Kh. Kudoyarova, Nanotechnology 12,
409 (2001).
[5] J. Ziegler, The Stopping and Range of Ions in Solids,
New York, Pergamon Press, 1986.
[6] E. R. Weber, in: Properties of Silicon, INSPEC, 1988,
p. 409.
[7] L. R. Doolitle, Nucl. Instr. and Meth. B 9, 344 (1985).
___________________
*Corresponding author: stefan@inrne.bas.bg