Content uploaded by Valdek Mikli
Author content
All content in this area was uploaded by Valdek Mikli on Oct 13, 2016
Content may be subject to copyright.
JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS Vol. 9, No. 2, February 2007, p. 311 - 314
Solid phase and ion beam epitaxial crystallization of Si
implanted with Zn and Pb ions
CH. ANGELOV*, S. GEORGIEV, B. AMOV, V. MIKLIa, T. LOHNERb
Institute for Nuclear Research and Nuclear Energy, Bulgarian Academy of Sciences, 72 Tzarigradsko Chaussee Blvd.,
1784 Sofia, Bulgaria
aCentre for Materials Research, Tallin Technical University, Ehitajate 5, Tallin 79086, Estonia
bResearch Institute for Technical Physics and Materials Science, Konkoly Thege Miklos ut 29-33, H-1121 Budapest,
Hungary
Rutherford backscattering spectroscopy in combination with channelling have been applied to investigate the surface
structural changes of silicon implanted with low-soluble species (Zn and Pb) and subsequently thermally and ion beam
annealed. For the fast diffusing Zn complete recrystallization during annealing of layers amorphized by ion implantation was
observed, while for the slow diffusing Pb incomplete recrystallization took place. The results are discussed with respect to
the corresponding experimental conditions.
(Received November 1, 2006; accepted December 21, 2006)
Keywords: Ion implantation, Solid phase epitaxial crystallization, Impurity distribution
1. Introduction
The unique feature of ion implantation is that impurity
atoms can be introduced into silicon in concentrations well
above the thermal equilibrium solubility value. Precise
control and reproducibility of the depth distribution of
dopants is required for the fabrication of semiconductor
devices. The activation energy for conventional solid
phase epitaxial growth (SPEG) in Si has been measured to
be around 2.7 eV [1,2]. At the same time, ion beam
induced epitaxial crystallization (IBIEC) in Si occurs at
temperatures as low as 200-400 oC, well below those of
the conventional furnace or rapid thermal annealing
processes (500-1000 oC). The activation energy of such
ion induced epitaxy has been measured to be around
0.3 eV, which is one order of magnitude lower than SPEG.
It is thus interesting to compare other physical properties
between IBIEC and SPEG. Zn is characterized by a fast
diffusion coefficient in Si (D
Czochralski-grown Si (100) wafers, which were p-
type, with a resistivity of 17 Ω cm, were amorphized with
100 keV Zn
Zn=10-7-10-6 cm-2 s-1), while
Pb is a very slow diffuser, even in amorphous
Si (DPb~10-25 cm-2 s-1). The impurity elements used in the
present study have very low solid solubilities in
Si (6×1016 cm-3 for Zn [3]) and, for the dose used of
1×1015 cm-2, the Zn and Pb concentrations greatly exceed
the solid solubility limits in Si. Thus it possible to produce
non-equilibrium structures with impurities trapped in
crystalline Si (c-Si) at such concentrations. Ion
implantation of insoluble elements seems to be important
in the future fabrication of integrated circuits, and it is also
promising for the formation of shallow p-n junctions [4].
In the present work, Rutherford backscattering
spectroscopy (RBS) in combination with channeling
(RBS/C) has been used to investigate the evolution of the
depth profiles of the implanted ions in silicon during
IBIEC and SPEG. Spectroscopic ellipsometry (SE) was
used for precise measurement of the values of the near-
surface amorphization. Reflection high energy electron
diffraction (RHEED) and scanning electron microscopy
(SEM) were used for structure investigations.
2. Experimental
+ and Pb+ ions at liquid nitrogen temperature
(LNT). In both cases, the implantation dose was
1×1015 cm-2, which resulted in peak concentrations of 0.3
at.% for Zn and 0.8 at.% for Pb. Before the SPEG and
IBIEC annealing, pre-annealing was performed at 450 oC
for 1 h in vacuum, to obtain sharp crystalline-amorphous
(c-a) interfaces. A thin layer of SiO2, 4.3 nm for Zn
implantation and 5.4 nm for Pb implantation, was formed
on the Si surface during this pre-annealing treatment, as
measured by a SOPRA ES4G rotating polarizer
spectroscopic ellipsometer. These processes resulted in an
amorphous layer at the surface with a thickness of 145.1
nm for Zn or 78.6 nm for Pb. A standard tube furnace was
used for isothermally annealing in a dry Ar atmosphere for
20, 40 and 60 min at 525 oC. Another set of the as-
implanted samples was irradiated with 3 MeV Si+ ions
from a 3 MV tandetron in order to stimulate IBIEC. The
beam current density was 0.075 μA cm-2 and the doses
were 5×1015, 1×1016 and 2×1016 cm-2. The total energy
deposition at the depth of the initial c-a interface was 120
eV/atom for the highest dose used, and was nearly
constant within the measured depth range. The
temperature rise of the irradiated area was negligible and
the sample holder temperature was stabilized at
400 oC (±2 oC).
Ch. Angelov, S. Georgiev, B. Amov, V. Mikli, T. Lohner
312
RBS measurements were performed using a
1.8 MeV He
+ beam from a Van de Graaff accelerator in
glancing angle geometry at 100o of the detector to obtain
better depth resolution. Some of the Zn implanted and
IBIEC annealed samples were investigated in a backward
geometry at 165o, to avoid overlapping of the Si and Zn
RBS spectra (see Fig. 1). A silicon surface-barrier detector
was used with an overall resolution of about 15 keV
FWHM. Energy calibrations were performed using 200 Å
Au on Si and bulk SiO2 samples. Complementary
information on the nature of the implantation induced
structures and surface morphology of the samples was
obtained by RHEED and SEM. The measurements were
carried out using a TEM-EMV-100B transmission electron
microscope working at an accelerating voltage of 100 kV
and a Jeol JSM-840A scanning electron microscope. SE
measurements were performed using a M-88 rotating
analyzer ellipsometer in the wavelength range 280-760
nm, at an angle of incidence of 75.1o for Zn and 70.2o for
Pb implanted samples. The depth distributions of both
vacancies and implanted atoms were calculated by SRIM-
2003 [5].
3. Result and discussion
The implantation and annealing conditions used in
this study, and the resulting re-growth thickness of the
amorphous Si (a-Si) and SiO2, due to the thermal
annealing, are summarized in Table 1.
Table 1. Summary of the implantation conditions and the
SE results for the thickness of a-Si and SiO2 layers of
SPEG samples, obtained in the present work.
Projectile
ion
100 keV,
1×1015 cm-2
Thickness of a-Si
and SiO2 (nm)
as-
impla
nted
20
min
40
min
60
min
Zn SiO2
a-Si 4.3
145.1 4.2
73.1 4.2
1.7 4.2
1.2
Pb SiO2
a-Si 5.4
78.6 5.4
41.0 5.4
26.0 4.2
20.2
After ion implantation with a dose of 1×1015 cm-2 and
energy of 100 keV, no reflections were present in the
RHEED pattern for Zn and for Pb, indicating that
complete surface amorphization of the Si wafer had taken
place. The surface morphology of the as-implanted
samples was revealed by SEM. No structural features can
be seen for as-implanted with Zn and Pb samples, at a dose
of 1×1015 cm-2. The SPEG and IBIEC annealing did not
lead to any apparent changes in the surface morphology.
Fig. 1 shows random (a) and channeling (b) spectra of
samples implanted with Zn, before and after SPEG
annealing for various times. The amorphous layer initially
formed was 145.1 nm with 4.3 nm SiO2 on the Si surface
(Table 1). The regrowth thickness increased with
increasing annealing time, and almost complete
recrystallization took place for the annealing time of 40
min. The mean re-growth rate was 3.6 nm/min. During
SPEG annealing, the zink impurity peak shifted to the
surface. This shift and the decrease of the width of the
peak are attributed to outdiffusion of the Zn atoms. The
height of Zn peak reached a maximum in the case of 40
min annealing (Fig. 1 (a)). Most of the Zn atoms (~65%)
are in a substitutional sites. Increasing the annealing time
to 60 min led to small decrease of a-Si layer (Table 1). A
backward diffusion of the Zn atoms occurred, occupying
~82 % substitutional sites (Fig. 1, aligned yield/random
yield). The crystallization process after SPEG of Pb-
implanted Si (Fig. 2) was different. The initially formed
amorphous layer was 78.6 nm with 5.4 nm SiO2 on the Si
surface (Table 1). The regrowth thickness increased with
increasing the annealing time. Even for the highest
annealing time (60 min) incomplete recrystallization was
observed (Table 1, Fig. 2 (b)). The mean re-growth rate
decreased from 1.8 nm/min for the first part of the
annealing to 0.3 nm/min for highest annealing time.
Redistribution of the initially formed Pb profile in a-Si
occurred when the c-a interface passed through it under
SPEG. A double Pb peak in the near-surface region was
detected for the final annealing step. The substitutional Pb
in the recrystallized part of the Si layer was about 86 %.
325 350 375 400 450
0
1000
2000
3000
4000
YIELD (counts)
CHANNEL
as-implanted
SPEG, 525 oC, 20 min
SPEG, 525 oC, 40 min
SPEG, 525 oC, 60 min
x15
(a)
Si
Zn
250 300 350 400 450
0
1000
2000
3000
4000
YIELD (counts)
CHANNEL
as-implanted
SPEG, 525 oC, 20 m
i
SPEG, 525 oC, 40 m
i
SPEG, 525 oC, 60 m
i
x15
Si
Zn
(b)
Fig. 1. RBS random (a) and aligned spectra (b) of Zn+
implanted Si with a dose of 1
×
1015 cm-2 at a glancing
angle (100o).
Solid phase and ion beam epitaxial crystallization of Si implanted with Zn and Pb ions
313
The implantation and annealing conditions used in our
study and the resulting re-growth thickness of a-Si and
SiO2 resulting from the ion beam annealing are
summarized in Table 2.
Fig. 3 (a) shows the channeling spectra of Si samples
implanted with Zn before and after irradiation at 3 MeV
Si+ beam at various annealing doses. The re-growth
thickness increased with increasing annealing dose from
5×1015 cm-2 to 2×1016 cm-2. The highest annealing dose
was sufficient for almost complete recrystallization of the
amorphised Si layer. As in the case of SPEG, the out-
diffusion of Zn took place during IBIEC.
300 350 450 500
0
500
1000
1500
2000
2500
YIELD (counts)
CHANNEL
as-implanted
SPEG, 525 oC, 20 min
SPEG, 525 oC, 40 min
SPEG, 525 oC, 60 min
x2
Si
Pb
(a)
300 350 450 500
0
500
1000
1500
2000
YIELD (counts)
CHANNEL
as-implanted
SPEG, 525 oC, 20 min
SPEG, 525 oC, 40 min
SPEG, 525 oC, 60 min
x2
Si
Pb
(b)
Fig. 2. RBS random (a) and aligned spectra (b) of Pb+
implanted Si with a dose of 1x1015 cm-2 at a glancing
angle (100o).
Table 2. Summary of the implantation conditions and the
SE results for the thickness of a-Si and SiO2 layers of
IBIEC samples, (1 – 5
×
1015 cm-2, 2 – 1
×
1016 cm-2,
3 - 2
×
1016 cm-2).
Projectile
ion
100 keV,
1×1015 cm-2
Thickness of a-Si
and SiO2 (nm)
as-
impl.
1
2
3
Zn SiO2
a-Si 4.3
145.1 4.6
75.6 5.8
56.6 6.4
0.9
Pb SiO2
a-Si 5.4
78.6 6.1
35.4 5.1
10.7 6.4
1.7
For the highest annealing dose of 2×1016 cm-2, the
retained amount of Zn atoms in the Si substrate was only a
few percent, compared to the respective implanted dose of
Zn which was measured to be 1×1015 cm-2. It suggests that
massive diffusion, accompanied by strong liberation of Zn
from the near-surface region occurred during the IBIEC.
The crystallization process after IBIEC of Pb-
implanted Si (Fig. 3 (b)) was different. For the highest
dose of annealing (2 × 1016 cm-2) incomplete
recrystallization was observed (Table 2). The out-diffusion
of Pb atoms in a-Si occurred when the c-a interface passed
through it under IBIEC.
225 250 275 350 375 40
0
2000
4000
6000
8000
YIELD (counts)
CHANNEL
as-implanted
IBIEC, 5x1015
c
IBIEC, 1x1016
c
IBIEC, 2x1016
c
x20
Si
Zn
(a)
300 350 400 450 500
0
500
1000
1500
2000
2500
Pb
Si
YIELD (counts)
CHANNEL
as implanted
IBIEC, 5x1015
c
IBIEC, 1x1016
c
IBIEC, 2x1016
c
(b)
Fig. 3. RBS spectra of Zn+ at a backscattering angle of
165o in a channeling condition (a) and Pb+ (b) implanted
Si at a glancing angle 100o, for a random orientation
after IBIEC.
During the time of the second annealing dose
(1×1016 cm-2) of the IBIEC processing (20000 s), the
substitutional part of the implanted Pb species reached
saturation (~62%). After this some of the substitutional Pb
escaped to interstitial positions. With the help of Si-
interstitials, these species diffused during the remaining
the time (more than 20000 s) for the highest dose of IBIEC
(2×1016 cm-2). This result is consistent with the model
proposed by Cho et al. [6].
In attempting to explain these observations, it should
be noted that SPEG was performed at 525 oC and IBIEC
was performed at 400 oC during irradiation. The melting
points of these species are Tm,Zn = 419.5 oC and
Tm,Pb = 327.6 oC.
Ch. Angelov, S. Georgiev, B. Amov, V. Mikli, T. Lohner
314
The most plausible explanation of the Zn and Pb
push-out effects under SPEG and IBIEC is a melt-
mediated crystallization process occurring during
annealing. Zn atoms diffuse rapidly Si, and are able to
escape from the c-a boundary as it moves to the surface.
The presence of slow diffusing Pb in a-Si inhibits the
crystallization processes at temperature 525 oC.
According to Fig. 3(b), it appears that in the case of
Pb implantation, the phase transformation in the a-Si layer
during IBIEC is controlled by transient enhanced diffusion
[7] as well.
4. Conclusion
The evolution of Zn and Pb species, implanted in Si
during the pure thermal and ion beam induced
crystallization of the amorphized substrates has been
analyzed using RBS, SE, RHEED and SEM. Massive
redistribution of the initially formed profiles of Zn and Pb
was observed during the annealing. In the case of SPEG,
the impurity atoms with low solubility and high diffusivity
(Zn) could escape to the surface with the c-a interface, as
shown in Fig. 1.
The impurities with low diffusivity (Pb) could not
follow the moving c-a interface. As a result, they were
“frozen” in the crystalline phase after the c-a interface
passed through, resulting in a higher impurity
concentration than the normal solid solubility (i.e.
supersaturation). The present results indicated that IBIEC
also achieves the supersaturation of impurity atoms, but
the impurity substitution is less than that for SPEG. The
results of this study clearly show the possibility of the
formation of structures with high volume concentrations of
insoluble dopants in a non-equilibrium way.
References
[1] R. G. Elliman, J. S. Williams, R. V. Knoel, Nucl. Instr.
and Meth. B 19-20, 435 (1987).
[2] F. Priolo, Mat. Sci. Rep. 5, 319 (1990).
[3] A. J. Milnes, Deep Impurities in Semiconductors
(Wiley, New York; Russian edition, Mir, M., 1977),
p. 38.
[4] Y. Ishikawa, I. Kobayashi, I. Nakamichi, Jpn.
J. Appl. Phys. 29, 1929 (1990).
[5] J. Ziegler, The Stopping and Range of Ions in Solids,
New York, Pergamon Press, (1986).
[6] K. Cho, M. Numan, T. G. Finstad, W. K. Chu, J. Liu,
Appl. Phys. Lett. 47, 1321 (1985).
[7] S. C. Jain W. Schoenmaker, Decoutere, M. Willander,
J. Appl. Phys. 91, 8919 (2002).
______________________
*Corresponding author: changelov@inrne.bas.bg
