Growth kinetics of silicon nanowires by platinum assisted vapour–liquid–solid
mechanism
H. Jeong, T.E. Park, H.K. Seong, M. Kim, U. Kim, H.J. Choi
*
Department of Materials Science and Engineering, Yonsei University, Seoul 120-749, Republic of Korea
article info
Article history:
Received 9 April 2008
In final form 5 November 2008
Available online 12 November 2008
abstract
The growth kinetics of Si nanowires produced by a vapour–liquid–solid (VLS) mechanism in conjunction
with Pt and Au catalysts, respectively, was investigated and compared. Pt was employed as a VLS catalyst
for single-crystal Si nanowires in a SiCl
4
-based chemical vapour deposition process. The growth rates
were higher with Pt than with Au under all processing conditions. The activation energy was measured
as 80 and 130 kJ/mol with the Pt and Au catalysts, respectively. The present results suggest that the rate-
determining step is the incorporation of Si atoms in the lattice at the liquid/solid interfaces and, further-
more, the metal catalysts affect this step, resulting in different activation energy.
Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction
Silicon (Si) nanowires have novel properties as well as comple-
mentary metal oxide semiconductor (CMOS) compatibility. As
such, they are among the most promising materials in terms of
serving as building blocks for the next generation of nano devices.
Among the many available methods to grow Si nanowires, chem-
ical vapour deposition (CVD) processes via a vapour–liquid–solid
(VLS) mechanism have been widely used since Wagner and Ellis
contrived this process in 1964 [1] for Si whiskers. The bulk of re-
search in this area has focused on the use of an Au catalyst to

luted H
2
(100 standard cubic centimeter per minute, sccm) and Ar
(100 sccm) gas was initiated. SiCl
4
was carried into the reactor
quartz tube by H
2
that had been passed through a bubbler that
maintained SiCl
4
in a liquid state at 0 °C. To investigate the growth
kinetics, processing time (1, 5, 10, and 20 min), temperature (900–
1100 °C), and SiCl
4
/H
2
concentration (5, 10, 15, and 20 sccm) were
manipulated.
3. Results and discussions
Both Pt and Au successfully catalyzed the growth of Si nano-
wires. Fig. 1 shows scanning electron microscopy (SEM) images
and transmission electron microscopy (TEM) observations. The
SEM images shown in Fig. 1a and b reveal Si nanowires with diam-
eters of 100 nm and lengths of several
l
m. The Si nanowires in
Fig. 1a and b were grown under identical conditions with the
exception of the metal catalyst that was used. The figures show
that the nanowires grew vertically on the substrate in both cases.

ubility of Si in Au is 18.6 at% [9], whereas in Pt it is as high as 67 at%
[10] at their eutectic points. Therefore, it is possible that the Pt cat-
alyst under a liquid state has a Si-rich composition. The composi-
tional differences may also be attributable to the formation of
platinum silicide upon cooling. We characterized the catalyst glob-
ule/nanowires interfaces by HRTEM and observed platinum silicide
layers showing different lattice fringes from Si in the nanowires
grown with Pt (Fig. 1g). However, no such layers were observed
at the interfaces with Au (Fig. 1h).
Fig. 2 shows the length of the Si nanowires according to the
holding time for the Pt and Au catalysts, respectively, at 1000 °C
with a SiCl
4
flow rate of 20 sccm. The growth rate of the Si nano-
wires was constant with time in both cases. This indicates that Pt
maintains a catalytic role throughout the growth of the Si nano-
wires, as does Au, without loss of components by chemical reaction
or evaporation. Thus, it has been demonstrated that Pt is a stable
catalyst for the VLS mechanism. The growth rate of Si nanowires
with Au was 5.20
l
m/min while Pt showed a 2.28 times faster
growth rate (11.86
l
m/min), with a 20 sccm SiCl
4
flow rate in both
cases.
Fig. 3 shows the growth rate of the Si nanowires with the flow
rate of SiCl

n
(P/P
0
), n = 1.5–2
[11,12], and thus the growth rate should follow the power law.
However, this is not the case in the present study as well as in pre-
vious studies [11–14].
The primary evidence for regarding step (2) as the rate-limiting
step is that the growth rate is proportional to the partial pressure
of reactant gas. However, this does not fully support the argument
since the growth process consists of two activated steps in series
[11]. The dependence of the growth rate on the reactant vapour
concentration is not in itself evidence that any of them is the
rate-determining step. Rather, it simply reflects the dependence
of the growth rate on supersaturation. Furthermore, as shown in
Fig. 3, the growth rate is saturated at a high SiCl
4
flow rate of
20 sccm; i.e., in the present growth conditions, the growth rate is
not proportional to the partial pressure of reactant gas. Therefore,
the rate-determining step would be step (4), incorporation of Si
atoms in a crystal lattice. Accordingly, Pt would affect this step,
resulting in a different growth rate.
We determined the activation energy for the growth of the Si
nanowires, as shown in Fig. 4, by plotting the growth rate with
0 5 10 15
0
50
100
150

flow rate of 20 sccm. Under these conditions,
the growth rate is saturated and thus supersaturation of the cata-
lyst is nearly maximized. Therefore, step (2) can be excluded as the
rate-determining step. Fig. 4 shows that the growth rate does not
follow the power law of D = D
0
(T/T
0
)
n
(P/P
0
) for gas diffusion coeffi-
cient [11,12] but strongly depends on temperature. This supports
the previous argument that step (1) is not the rate-determining
step. From Fig. 4, the calculated activation energy values were 80
and 130 kJ/mol for the Pt and Au catalysts, respectively. These val-
ues are much higher than the activation energy for the diffusion of
Si atoms throughout most liquid metals ranges, i.e., between 10.5
and 36 kJ/mol, as well as the activation energy for Si diffusion in
Au–Si liquid (about 25 kJ/mol) and that in Pt–Si liquid (54 kJ/
mol) [15,16]. This supports the argument that step (3) is not the
rate-determining step. Since steps (1)–(3) can be excluded again
as described above, it is concluded that step (4), incorporation of
Si atoms in a crystal lattice at the liquid/solid interfaces, is the
rate-limiting step under the present conditions, for both the Pt
and Au catalyst cases.
The lower activation energy with Pt implies that step (4) is
dependent on the type of catalyst. This could be explained by
the effect of the catalyst on the barrier energy to the nucleation

of the metal catalyst on the rate-determining step.
Acknowledgments
This work was supported by the Program of the National Re-
search Laboratory (Grant R0A-2007-000-20075-0) of the Korean
Ministry of Science and Technology (MOST), and the Korean Re-
search Foundation (MOEHRD, KRF-2005-042-D00203).
References
[1] R. Wagner, W. Ellis, Appl. Phys. Lett. 4 (1964) 89.
[2] J. Yoshino, Y. Okamoto, J. Morimoto, T. Miyakawa, Appl. Phys. A 66 (1998)
323.
[3] E. Garnett, W. Liang, P. Yang, Adv. Mater. 19 (2007) 2946.
[4] R. He, P. Yang, Nature Nanotechnol. 1 (2006) 42.
[5] T. Baron, M. Gordon, F. Dhalluin, C. Ternon, Appl. Phys. Lett. 89 (2006) 233111.
[6] J. Weyher, J. Cryst. Growth 42 (1977) 235.
[7] Y. Wang, V. Schmidt, S. Senz, U. Gosele, Nature Nanotechnol. 1 (2006) 186.
[8] T. Stelzner, G. Andra, E. Wendler, W. Wesch, R. Scholz, U. Gosele, S.
Christiansen, Nanotechnology 17 (2006) 2895.
[9] ASM International Handbook Committee, ASM Handbook, ASM International,
Ohio, 1990.
[10] T. Massalski, J. Murray, L. Bennett, H. Baker, Binary Alloy Phase Diagrams,
American Society for Materials, Florida, 1986.
[11] E.I. Givargizov, J. Cryst. Growth 31 (1975) 20.
[12] T.I. Kamins, R.S. Williams, D.P. Basile, T. Hesjedal, J.S. Harris, J. Appl. Phys. 89
(2001) 2.
[13] K. Lew, J. Redwing, J. Cryst. Growth 254 (2003) 14.
[14] J. Kikkawa, Y. Ohno, S. Takeda, Appl. Phys. Lett. 86 (2005) 123109.
[15] W.C. Yang, H. Ade, R.J. Nemanich, Phys. Rev. B 69 (2004) 045421.
[16] T.R. Anthony, H.E. Cline, J. Appl. Phys. 43 (1972) 2473.
[17] I.V. Markov, Crystal Growth for Beginners, World Scientific, 1995, p. 41
(Chapter 1).