Japanese Journal of Applied

Physics

Related content
- Long Length, High-Density Carbon
Growth of Dense, Vertical and Horizontal Nanotube Film Grown by Slope Control of
Temperature Profile for Applications in
Graphene and Its Thermal Properties Heat Dissipation
Akio Kawabata, Tomo Murakami, Mizuhisa
Nihei et al.
To cite this article: Akio Kawabata et al 2013 Jpn. J. Appl. Phys. 52 04CB06 - Fabrication of Graphene Directly on SiO2
without Transfer Processes by Annealing
Sputtered Amorphous Carbon
Motonobu Sato, Manabu Inukai, Eiji
Ikenaga et al.
View the article online for updates and enhancements. - Wafer-Level Electrical Evaluation of
Vertical Carbon Nanotube Bundles as a
Function of Growth Temperature
Bart Vereecke, Marleen H. van der Veen,
Masahito Sugiura et al.

Recent citations
- Polymeric composites for powder-based
additive manufacturing: Materials and
applications
Shangqin Yuan et al

- Xiaoxing Lu

- Low temperature growth of ultra-high mass

density carbon nanotube forests on
conductive supports
Hisashi Sugime et al

This content was downloaded from IP address 193.231.133.66 on 11/03/2019 at 12:37

Japanese Journal of Applied Physics 52 (2013) 04CB06 REGULAR PAPER
http://dx.doi.org/10.7567/JJAP.52.04CB06

Growth of Dense, Vertical and Horizontal Graphene and Its Thermal Properties
Akio Kawabata
, Tomo Murakami, Mizuhisa Nihei, and Naoki Yokoyama
Collaborative Research Team Green Nanoelectronics Center (GNC), National Institute of Advanced Industrial Science and Technology (AIST),
Tsukuba, Ibaraki 305-8569, Japan
E-mail: kawabata.akio@aist.go.jp
Received September 21, 2012; accepted January 22, 2013; published online April 22, 2013

We have grown an extremely dense nanocarbon structure on a Si substrate at a temperature of 450  C, in which vertical graphene layers were
formed densely below horizontal graphene layers. We named this carbon structure ‘‘dense, vertical, and horizontal graphene (DVHG)’’. The
density of this structure was an extremely high at 1.4 g/cm3 , which is 63% of that of graphite (2.2 g/cm3 ). Although DVHG showed poor thermal
properties, we confirmed that vertical thermal conductivity increased by a factor of 10 by removing the horizontal graphene layers from the top of
DVHG. This result indicates that the thermal conductivity parallel to the graphene plane is several orders of magnitude higher than that
perpendicular to the graphene plane. # 2013 The Japan Society of Applied Physics

1. Introduction
Carbon nanotubes (CNTs), grapheme, and other nanocarbon
materials hold promise for applications in electronics
because of their unique electrical1,2) and thermal character-
istics.3) Their known electrical characteristics include the
ballistic transport of carrier conduction along the tube and
tolerance to a high-density current. The application of those
electrical characteristics to LSI via interconnects was
reported previously.4–6) There are also a number of reports
on their thermal characteristics; however, large differences
remain between their ideal properties and experimental data,
particularly those due to issues related to CNT density
factors.7–11)
There are some reports on the high-density growth of
CNTs. The MIRAI-Selete group has reported CNT densities
of more than 1012 /cm2 .12,13) The Cambridge group has also
reported CNT densities of more than 1013 /cm2 .14) In these
reports, the growth temperature is lower than that under
conventional conditions. In this study, we focused on the
low-temperature growth of a previously unseen dense,
vertical, and horizontal graphene (DVHG) structure, in
which vertical graphene layers were formed densely below
horizontal graphene layers, perpendicularly to the substrate.
The density of this structure was measured by X-ray
reflectivity (XRR) measurement. We also attempted to
measure the thermal characteristics of DVHG structures by
picosecond thermoreflectance measurement.15) Fig. 1. DVHG structures observed by (a) SEM and (b) parallel-sectional
TEM. (c) Schematics of DVHG structure and magnified (d) upper (e) lower
2. Experimental Procedure parts.
Experiments were conducted with a newly developed
multichamber process that enables substrate preprocessing,
catalyst metal deposition, and nanocarbon thermal chemical of 1 kPa. The growth time was 30 min. The CVD growth
vapor deposition (CVD) growth in a vacuum batch. The base temperature was as low as 450  C.
vacuum pressure was 3:0  105 Pa.
We deposited a TaN layer (15 nm thick) on SiOC 3. Results and Discussion
dielectric/Si substrate having via holes with a diameter of A high-resolution scanning electron microscopy (SEM)
160 nm in SiOC dielectric layer. The substrate was subjected observation revealed that a columnar carbon structure was
to Ar plasma etching in a preprocessing chamber to etch the obtained after the growth in via holes, as shown in Fig. 1(a).
TaN surface very slightly. Next, layers of Ti (1.0 nm thick) The surface of the structure is flat and mirror like. A cross-
and Co (4.0 nm thick) were deposited in a vacuum vapor sectional transmission electron microscopy (TEM) observa-
deposition chamber. The substrate was then moved to a tion confirmed the contrast in the columnar carbon structure.
thermal CVD chamber, where it was heated to a growth What is notable here is that no empty structures are seen in
temperature of 450  C in a vacuum. Subsequently, a mixture the CNT [Fig. 1(b)], and nearly all vertical graphene layer
of acetylene and argon gas was introduced at a gas pressure are formed. Looking at the entire cross-sectional image, a
04CB06-1 # 2013 The Japan Society of Applied Physics
Jpn. J. Appl. Phys. 52 (2013) 04CB06
(c)
(d)
(e)
(f)
(g)
Experimental
Calculation
Critical
Intensity (arb. Unit)
angle
0 0.5 1.0 1.5 2.0
2θ (degree)
Fig. 2. (Color online) Schematics of DVHG growth mechanism. Cross-
sectional TEM images of DVHG in initial a 5 min growth and (b) 10 min
growth. (c) The horizontal graphene layers grow on the Co catalyst. (d) The
Co catalyst aggregates to a round shape, (e) resulting in round graphene
layers along the Co particles. (f ) Graphene layers are formed vertically to
the substrate. (g) Critical angle from XRR measurement.
contrast is confirmed in the lower and middle parts of the
structure, which may be the catalyst metal, but no contrast
is observed in the upper part. Moreover, a magnified view
of the upper part [Fig. 1(d)] confirms the formation of a
horizontal graphene layer about 2–4 nm in thickness. A
magnified view of the lower part [Fig. 1(e)] clearly confirms
the contrast of graphene formed in a direction perpendicular
to the substrate. From this finding, we know that there exists
a horizontal graphene layer in the upper part of the structure
and a vertical graphene layer in the lower part, indicating a
previously unseen structure. Figure 1(c) shows an illustra-
tion of the DVHG structure obtained in this study. A CNT/
graphene structure in which the uppermost part is graphene
was reported by Kondo et al.;16) however, the structure
reported here features the formation of a vertical graphene
layer with a high density of CNTa.
Figure 2(a) shows a cross-sectional TEM image of the
DVHG structure in the initial 5 min growth. There are
horizontal grapheme layers on the Co catalyst. Figure 2(b)
shows that of the DVHG structure in the initial 10 min
growth. The Co catalyst gradually aggregates to a round
04CB06-2
A. Kawabata et al.
shape, resulting in round graphene layers along the Co
particles. From this observation, we can suggest the growth
mechanism of DVHG. First, the graphene layers grow
horizontally on the Co catalyst, because the Co catalyst is
flat at the early stage of graphene growth [Fig. 2(c)]. Next,
the Co catalyst gradually aggregates to a round shape,
resulting in round graphene layers along the Co particles
[Fig. 2(d)]. Finally, graphene layers are formed vertically to
the catalyst metal and substrate [Figs. 2(e) and 2(f )]. The
density of DVHG was calculated, by fitting the experimental
data at the critical angle [Fig. 2(g)]. It shows an extremely
high value of 1.4 g/cm3 , which is higher than that of high-
density packed SWNTs (0.57 g/cm3 ) by a factor of 2.4 as
reported by Futaba et al.17) and is 63% of that of graphite
(2.2 g/cm3 ). The dense structure of DVHG is due to its
growth mechanism. Conventional CNTs were grown from
catalyst particles. Therefore, the density of CNTs will
depend on the density of catalyst particles, which is as low
as 1010 /cm2 at 540  C.7) On the other hand, DVHG is grown
from the catalyst film. This is due to the low-temperature
growth, in which the catalyst film does not aggregate. Thus,
a growth temperature as low as 450  C is important for
DVHG growth. This result suggests that it is possible to
grow high-density nanocarbon materials by low-temperature
CVD.
We measured the thermal characteristics with a picose-
cond thermoreflectance measurement system (PicoTherm).
The measurement method, which is referred to as pulse
optical heating thermoreflectance, involves the formation of
a Mo thin film (reference) on the object to be measured and
the measurement of the thermal effusivity of the material
beneath the Mo thin film from the reflectance of the heated
Mo surface [Fig. 3(a)].
The thermal effusivity and conductivity are related as
expressed by
pffiffiffiffiffiffiffiffi
b ¼ c;
where b is the thermal effusivity,  is the density,  is the
thermal conductivity, and c is the specific heat capacity.
In this measurement, it is important to have the laser light
reflected by the Mo surface; thus, the surface of the object
to be measured must be flat. Because the surface of the
structure obtained in this work consists of horizontal
grapheme layers at the upper part, the surface remains flat
after the formation of the Mo layer and measurement is
possible. Because horizontal graphene layers the surface of
this structure, we do not expect the thermal effusion to
proceed directly to the vertical structure. For this reason, we
introduced the process of Ar plasma etching, which is
performed in the preprocessing chamber to remove the
horizontal grapheme layers. The Ar gas pressure was 0.5 Pa
and the plasma power was 400 W. The etching times were
120, 300, and 1200 s. As an example, the signals measured
when the etching times were 300 and 1200 s are shown in
Fig. 3(b).
As shown in Fig. 3(c), without etching, the thermal
effusivity was 975 J S0:5 m2 k1 , the lowest value for any
of the measurement conditions. An improvement was seen in
the thermal effusivity for all etching conditions. The most
distinct improvement was obtained with an etching time
of 300 s, for which the thermal effusivity was 3162
# 2013 The Japan Society of Applied Physics
Jpn. J. Appl. Phys. 52 (2013) 04CB06
(a)
(b)
300 s Ar etching
1200 s Ar etching
(c)
Fig. 3. (Color online) (a) Picosecond thermoreflectance measurement
system. (b) Thermoreflectance signal. (c) Ar etching time vs Thermal
effusivity.
J S0:5 m2 k1 . Using a graphite density of 1.4 g/cm3 and
a graphite specific heat value of 710 kJ kg1 K1 ,18,19) we
obtained thermal conductivities of 1.0 and 10 W m1 K1
in the cases without and with etching 300 s, respectively. In
the case of a longer Ar etching time of 1200 s, etching
damage was seen. As expected, we could improve the
vertical thermal conductivity by removing the horizontal
graphene layers from the DVHG structure. This result
indicates that the thermal conductivity parallel to the
graphene plane is several orders of magnitude higher than
that perpendicular to the graphene plane. However, this
thermal conductivity is lower than that of graphite. We
considered that the graphene sheets are not continuous
because of their low growth temperature. In general, highly
crystalline graphene can be obtained by high-temperature
growth. Thus, we expect that the thermal conductivity could
be improved by combining high-temperature growth with
our DVHG growth conditions.
04CB06-3
A. Kawabata et al.
4. Conclusions
We demonstrated a DVHG structure fabricated by a vacuum
batch multichamber process. The density of DVHG was an
extremely high value of 1.4 g/cm3 . We determined the
thermal characteristics of DVHG by picosecond thermore-
flectance measurement. The thermal conductivity of DVHG
was measured to be 1.0 W m1 K1 when horizontal
grapheme layers were formed, and 10 W m1 K1 when
they were removed. These results indicate that, as expected,
the conductivity parallel to the graphene plane is several
orders of magnitude higher than that perpendicular to the
graphene plane. To improve the thermal conductivity of
DVHG further, we plan to obtain a higher-quality vertical
graphene structure by applying higher temperature growth
conditions.
Acknowledgements
This research was supported by the Japan Society for the
Promotion of Science (JSPS) through its Funding Program
for World-Leading Innovative R&D on Science and
Technology (FIRST), and was partly conducted at the
Nano-Processing Facility and Nano-Instrumentation Facil-
ity, supported by the IBEC Innovation Platform, AIST.
1) Z. Yao, C. L. Kane, and C. Dekker: Phys. Rev. Lett. 84 (2000) 2941.
2) J. Kong, E. Yenilmez, T. W. Tombler, W. Kim, H. Dai, R. B. Laughlin, L.
Liu, C. S. Jayanthi, and S. Y. Wu: Phys. Rev. Lett. 87 (2001) 106801.
3) P. Kim, L. Shi, A. Majumdar, and P. L. McEuen: Phys. Rev. Lett. 87
(2001) 215502.
4) S. Sato, M. Nihei, A. Mimura, A. Kawabata, D. Kondo, H. Shioya, T. Iwai,
M. Mishima, M. Ohfuti, and Y. Awano: IEEE IITC, 2006, p. 230.
5) M. Nihei, A. Kawabata, T. Hyakushima, S. Sato, T. Nozue, D. Kondo, H.
Shioya, T. Iwai, M. Ohfuti, and Y. Awano: Ext. Abstr. Solid State Devices
and Materials (SSDM), 2006, p. 140.
6) A. Kawabata, S. Sato, T. Nozue, T. Hyakushima, M. Norimatsu, M.
Mishima, T. Murakami, D. Kondo, K. Asano, M. Ohfuti, H. Kawarada, T.
Sakai, M. Nihei, and Y. Awano: IEEE IITC, 2008, p. 230.
7) M. Horibe, M. Nihei, D. Kondo, A. Kawabata, and Y. Awano: Jpn. J. Appl.
Phys. 43 (2004) 7337.
8) D. J. Yang, Q. Zhang, G. Chen, S. F. Yoon, J. Ahn, S. G. Wang, Q. Zhou,
Q. Wang, and J. Q. Li: Phys. Rev. B 66 (2002) 165440.
9) H. Wang, J. Y. Feng, X. J. Hu, and K. M. Ng: Chem. Eng. Sci. 65 (2010)
1101.
10) J. Hone, M. C. Llaguno, N. M. Nemes, A. T. Johnson, J. E. Fischer, D. A.
Walters, M. J. Casavant, J. Schmidt, and R. E. Smalley: Appl. Phys. Lett.
77 (2000) 666.
11) H. Xie: J. Mater. Sci. 42 (2007) 3695.
12) A. Kawabata, T. Nozue, T. Murakami, T. Daidou, M. Sato, T.
Hyakushima, D. Kondo, S. Sato, M. Nihei, and Y. Awano: NT09 2009.
13) Y. Yamazaki, M. Katagiri, N. Sakuma, M. Suzuki, S. Sato, M. Nihei, M.
Wada, N. Matsunaga, T. Sakai, and Y. Awano: Appl. Phys. Express 3
(2010) 055002.
14) G. Zhong, J. H. Warner, M. Fouquet, A. W. Robertson, B. Chen, and J.
Robertson: ACS Nano 6 (2012) 2893.
15) T. Baba: 1st Int. Symp. Thermal Design and Thermophysical Property for
Electronics, 2008.
16) D. Kondo, S. Sato, and Y. Awano: Appl. Phys. Express 1 (2008) 074003.
17) D. N. Futaba, K. Hata, T. Yamada, T. Hiraoka, Y. Hayamizu, Y. Kakudate,
O. Tanaike, H. Hatori, M. Yumura, and S. Iijima: Nat. Mater. 5 (2006) 987.
18) A. Mizel, L. X. Benedict, M. L. Cohen, S. G. Louie, A. Zettl, N. K. Budraa,
and W. P. Beyermann: Phys. Rev. B 60 (1999) 3264.
19) C. Masarapu, L. L. Henry, and B. Wei: Nanotechnology 16 (2005) 1490.
# 2013 The Japan Society of Applied Physics