ARTICLE IN PRESS
DIAMAT-05191; No of Pages 6
Diamond & Related Materials xxx (2009) xxx–xxx
Contents lists available at ScienceDirect
Diamond & Related Materials
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / d i a m o n d
State of the catalyst during carbon nanotube growth
C.T. Wirth, S. Hofmann, J. Robertson ⁎
Engineering Dept, Cambridge University, Cambridge CB2 1PZ, UK
a r t i c l e
i n f o
Available online xxxx
Keywords:
carbon nanotubes
growth mechanism
chemical vapour deposition
catalyst
a b s t r a c t
We review our in-situ X-ray photoemission (XPS) and in-situ transmission electron microscopy studies
which determined that the catalyst is in the metallic state for Fe, Co and Ni catalysts. We show that the
existence of surface carbide phases in related catalytic reactions could account for the observation of carbide
peaks in XPS. The observed catalytic activity of gold is discussed in terms of carbon solubility, reaction rates,
and surface coordination numbers.
© 2009 Elsevier B.V. All rights reserved.
Carbon nanotubes can be grown by three methods, laser ablation,
carbon arc and chemical vapour deposition (CVD). CVD is the most
interesting for applications, because it can be used for bulk growth or
for applications requiring the growth of nanotubes bound to a surface
[1–10]. CVD uses a catalyst to promote growth. In order to optimise
the CVD process, it is important to understand how the catalyst works,
and what is the nature of the catalyst. Although it is supplied in the
experiment as a metal, is the active form of the catalyst a metal, oxide
or a carbide? Is a solid or liquid form necessary for single wall carbon
nanotubes? Carbon nanotubes have been grown for 50 years, and it is
also useful to recognise the early catalyst literature in which it was
often desired to stop the growth of ‘filamentous carbon’ [11,12]. There
is also a desire to know which elements are capable of catalysing CNT
growth in the wider sense. The most active catalysts for carbon
nanotube growth are the iron group elements Fe, Co and Ni. A wide
range of other transition metals such as Mo, Ru, Rh, Pd, etc are also
active [13–15], to varying degrees. Recently even the noble metal gold
was found to be able to catalyst growth, contradicting some
predictions. We study here the catalytic process by a combination of
in-situ photoemission and previous in-situ electron microscopy,
together with a range of processing studies. Mechanisms are discussed
in terms to the results and the standard catalyst literature.
The question of whether a catalyst is in the solid or liquid form
arises as follows. In the laser or arc methods, a high temperature gas of
carbon and metal is created which is then condensed, which is
followed by nanotube growth. In-situ optical measurements have
followed the condensation process and found that single wall
nanotubes (SWNTs) grow on condensed clusters at temperatures
above the catalyst solidification temperature [16].
Growth on solid surfaces usually involves much lower temperatures. However, the melting temperature of the catalyst is lowered by
⁎ Corresponding author.
E-mail address: [email protected] (J. Robertson).
two effects. First, the melting temperature is depressed by the Gibbs–
Thomson effect for small diameter particles [17],
ΔTm =
2Tm γ
Lρr
where Tm is the melting temperature in K, γ is the surface tension, L is
the latent heat, ρ is the density and r is the particle radius. Second, the
melting point is depressed by forming a eutectic with carbon. In the
laser or arc method, more than one nanotube grows from each particle
[18], so the catalyst nanoparticles tend to be larger. On the other hand,
for CVD, a single nanotube tends to grow from each catalyst
nanoparticle [19] so the diameter of the catalyst tends to equal that
of the resulting nanotube. Thus the typical nanoparticle diameters are
smaller, of order 1–2 nm. This can reduce melting temperatures by up
to 700–800 °C, down to 550 °C (Fig. 1), a significant reduction. This
trend has been confirmed by molecular dynamics simulations [20,21].
In addition some thermal measurements have shown the presence of
a liquid phase for SWNT growth [22].
Fig. 1. Calculated melting points of Ni as a function of particle size, data from Bolton [20].
0925-9635/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.diamond.2009.01.030
Please cite this article as: C.T. Wirth, et al., Diamond Relat. Mater. (2009), doi:10.1016/j.diamond.2009.01.030
ARTICLE IN PRESS
2
C.T. Wirth et al. / Diamond & Related Materials xxx (2009) xxx–xxx
Fig. 2. TEM image of Ni catalyst nanoparticle at 480 °C and 360 °C showing lattice planes of a crystal.
For bulk CVD such as in the injection methods for growing SWNTs
which use temperatures of order 1000 °C [23,24], the catalyst is likely
to be in the liquid state. For low temperature CVD on surfaces, recent
in-situ transmission electron microscopy studies of growth [25] at
480 °C as in Fig. 2 shows the presence of lattice planes in the catalyst
nanoparticles, which suggests that the catalyst has remained in the
solid state. These same experiments also show that the catalyst
particle undergoes severe mechanical re-shaping during the tip
growth of multi-wall nanotubes [25–27]. This can give the impression
that the catalyst is liquid. However, this re-shaping occurs by creep in
the solid state because of the large forces exerted by the surrounding
carbon tube, and it is compatible with a solid form. The creep rate by
power law creep increases for nano-meter sized particles [28]. The
catalyst is solid in the remaining data.
The chemical state of the catalyst is important for optimising the
catalyst. The nature of the catalyst is often derived by TEM or diffraction measurements of the lattice constant of catalyst after growth.
However, during growth the catalyst contains excess carbon, which,
on cooling, can precipitate as a carbide phase [29] as the carbon
solubility limit drops. A second problem is that metal and carbide
phases can have rather similar lattice constants, so they are not so easy
to distinguish by diffraction especially if the particles are small, and
the lattice constant could be distorted by their small size. If TEM is
used, care must be taken that the projection is correctly oriented
against the lattice planes. For these reasons, the most reliable method
Fig. 3. X-ray photoemission spectra of the Ni 2p edge of a nominal 0.13 nm thick Ni
film on a SiO2 support as it is thermally annealed from 200 °C to 500 °C in a vacuum of
10− 6 mbar. The 3+ peak is due to Ni2O3, the metal peak is due to Ni metal from [31].
is to use a combination of in-situ photoemission and in-situ TEM
measurements [25,30–32]. In this way, the measurements are carried
out at realistic temperatures and carbon activities.
We have previously carried out a series of in-situ photoemission
measurements of the carbon 1 s core level and the metal core levels
before and during nanotube growth, with growth carried out at low
(2 × 10− 7 mbar) and moderate (2 × 10− 3 mbar) acetylene pressures
[25,30,31]. Fig. 3 shows the Ni 2p core level spectra of a nominally
0.13 nm thick Ni film on a SiO2 support as the film is heated up in a
vacuum of 10− 6 mbar [31]. Initially, the spectrum is dominated by the
Ni3+ peak at 856 eV. Above 200 °C, this peak reduces, and the Ni metal
peak at 852.6 eV appears and dominates. Thus, the nominal Ni film has
Fig. 4. Evolution of the C 1 s spectrum for growth of nanotubes on Ni on SiO2 at
10− 6 mbar of acetylene at 580 °C, after [25].
Please cite this article as: C.T. Wirth, et al., Diamond Relat. Mater. (2009), doi:10.1016/j.diamond.2009.01.030
ARTICLE IN PRESS
C.T. Wirth et al. / Diamond & Related Materials xxx (2009) xxx–xxx
3
Fig. 5. Evolution of the C 1 s and Ni 2p spectra before, during an after growth on Ni on a SiO2 support at 10− 3 mbar acetylene at 500 °C, from [31].
changed under vacuum annealing from that of a surface oxide to that
of a metal.
Fig. 4 shows the evolution of the carbon 1 s core level of growth on
a Ni catalyst on SiO2 support with exposure to undiluted acetylene at a
pressure of 2 × 10− 7 mbar at 580 °C [25]. (The catalyst had previously
been exposed to 1 mbar ammonia at 480 °C to reduce any oxide.) We
see that initially there is a peak at 282.6 eV which shifts to 283.2 eV,
and after about 100 s, a new peak at 284.5 eV appears and it begins to
dominate. The 282.6 eV peak was assigned to surface carbon, and the
283.2 eV peak to a surface carbide phase, following the literature. The
284.5 eV is due to sp2 carbon and corresponds to the over-growth of
carbon nanotubes, after an induction period.
Fig. 6. Evolution of the C 1 s and Fe 2p spectra before, during an after growth on Fe on a SiO2 support at 10− 3 mbar acetylene at 500 °C, from [31].
Please cite this article as: C.T. Wirth, et al., Diamond Relat. Mater. (2009), doi:10.1016/j.diamond.2009.01.030
ARTICLE IN PRESS
4
C.T. Wirth et al. / Diamond & Related Materials xxx (2009) xxx–xxx
Fig. 9. Heats of formation of transition metal carbides, per C atom across the transition
metal series.
Fig. 7. Schematic of reactions and processes on a catalytic nano-particle during
nanotube growth (tip growth illustrated).
Fig. 5 shows the Ni and C core levels of a Ni catalyst on SiO2 support
after exposure to acetylene at 2 × 10− 3 mbar pressure at 500 °C [31].
The Ni spectrum is dominated by the peak at 852.6 eV which
corresponds to Ni metal, and the carbon spectrum is dominated by a
284.3 eV peak which corresponds to the sp2 carbon of nanotube walls.
The carbide peak of Fig. 4 is not seen. This is because the growth rate is
higher. These spectra (Figs. 4 and 5) suggest that the Ni catalyst is in
the form of metal during growth, not as an oxide or a carbide.
Fig. 6(a) shows the spectrum of a nominal 0.3 nm thick Fe film on
SiO2 support after annealing at 580 °C in vacuum of 2 × 10− 10 mbar
[31]. The Fe is in metallic form here. Note that this pressure is much
lower than that used to reduce the Ni oxide.
Fig. 6(b) shows the Fe and C core levels during growth of
nanotubes on Fe at 500 °C at an acetylene pressure of 2 × 10− 3 mbar
[31]. The Fe peak is dominated during and after growth by the peak at
707.3 eV due to Fe metal. There are weak features in the 711–709 eV
range which would be due to Fe oxides, but only before growth starts.
The carbon spectrum is dominated by the peak at 284.6–284.5 eV due
to sp2 carbon. This suggests that Fe may initially be as a (surface)
oxide, but during growth it is in the metallic state. It is not as a carbide.
Our data suggest that for the main nanotube growth catalysts, the
active state is the metal. This is consistent with most recent work. On
the other hand, we should note that in-situ photoemission measurements of Arcos [33,34] favoured the oxide form. Yoshida et al. [52] also
show TEM evidence in favour of a carbide catalyst.
The results help to explain the behaviour of Fe, Co and Ni catalysts.
Ni is a frequently used catalyst, which is active over a wide tempera-
Fig. 8. Schematic of partial reactions during a catalysed reaction.
ture range. On the other hand, Fe is favoured because it gives rise to
the highest density nanotube forests, that is it has the highest yield.
However, Fe is less active at low temperatures than Ni and, in surface
growth, it is less active at highest temperatures. We believe that the
decline in activity of Fe below 450 °C is due to it being still as an oxide.
FeOx can be formed during sample transfer or preparation and it is
more difficult to reduce to the metallic form, as its heat of formation is
higher than that of NiO. Thus, it can be necessary to use ammonia to
reduce FeOx. On the other hand, Fe also has a metastable carbide,
while Ni has none. Thus, Fe catalyst could become slightly deactivated by the side reaction of carbide formation at higher temperatures. We have previously discussed the unusually high activity of Fe
on Al2O3 in forming vertically aligned nanotube mats [30].
What about other transition metal elements as catalysts? Other
transition metals have been found to work as catalysts, such as Ru, Re,
Rh, Pd, etc. Fig. 7 shows schematically the intermediate reactions in
nanotube growth. The precursor molecule absorbs on the catalyst
surface, dissociates, and sticks to the surface. This releases atomic C,
which dissolves in the bulk or surface of the catalyst. The C diffuses in
the bulk or over the surface. The C then precipitates on the opposite
surface to form the nanotube walls. This illustrates that the catalyst
must perform two key functions – hydrocarbon dissociation and
carbon solubility.
Catalytic activity is often linked to carbon solubility. However, the
highest solubilities occur for those metals, which form carbides. A
catalyst works by lowering the energy barrier to the overall reaction
(Fig. 8). This often occurs by forming intermediate species. On the
other hand, any intermediate species such as a carbide should not be
too stable, or it will not dissociate to give the product.
Fig. 10. TEM image of carbon nanotubes grown at 850 °C on Au catalyst.
Please cite this article as: C.T. Wirth, et al., Diamond Relat. Mater. (2009), doi:10.1016/j.diamond.2009.01.030
ARTICLE IN PRESS
C.T. Wirth et al. / Diamond & Related Materials xxx (2009) xxx–xxx
5
on subsurface C in Pd [36] suggests that the mechanism might involve
subsurface diffusion of C in gold.
Fig. 11 shows how we should think of the reaction kinetics. Gold
has a low solubility for carbon. But more importantly, it is less good at
dissociating the hydrocarbon precursors. This would limit reaction
rates at lower temperatures, but it can still give growth a higher
temperatures.
Acknowledgements
The authors are very grateful to the help of R. Sharma, C. Ducati,
R. E. Dunin-Borkowski, C Mattevi, C Cepek, R Blume, C T Wirth, M
Cantoro, R Sharma, C Ducati, M Havecker, S Zafeiratos, P Schnoerch, A
Oestereich, D Teschner, M Albrecht, A Knop-Gericke, and R Schlogl.
Fig. 11. Schematic on temperature limits for reactions on a Au catalyst.
Fig. 9 plots the formation energy of the carbides of the transition
metals, across the series. The metals on the left form stable carbides,
and this lowers their suitability to act as growth catalysts. The
transition metals Fe, Co and Ni form metastable or unstable carbides
and have a moderate C solubility, which is one reason that they act as
catalysts.
Hofmann et al. [35] studied the diffusion mechanism of carbon on a
Ni particle. They showed that carbon atoms diffuse on top of the closepacked (111) surface, whereas they diffuse sub-surface for the less
close packed (100) surface, especially across the open ridges. It turns
out that subsurface incorporation and diffusion of C in transition
metals is more general. The reason is that the Ni–C bond is quite strong.
However, C enters an octahedral interstitial site in the Ni lattice, which
requires a volume expansion. A surface site has fewer bonds than a
bulk interstitial site, but it requires the volume expansion. These leads
to the optimum site being a subsurface interstitial, which is fully
bonded, but which allows some relaxation by pushing up the surface
layer above it. This is consistent with recent depth-resolved photoemission spectra of hydrocarbon reactions on Pd surfaces, which
showed that the carbon actually lay 1–2 layers below the surface [36–
38]. Thus, the presence of a carbide signal in XPS does not necessary
mean the presence of a bulk carbide during growth, just a subsurface
carbide as in these reactions.
Hofmann [35] argued that surface (and subsurface) diffusion of C
was the most probable diffusion path for thermal CVD of SWNTs, as
well as in plasma enhanced CVD [39]. A similar conclusion was reached
by Raty et al. [40].
Gold is a noble metal. It was expected to be a poor catalyst for
nanotube growth because carbon is rather insoluble in bulk gold
(5 × 10− 4 wt.% at 730 °C). this is consistent with the calculations of Raty
et al. [39] who found that C atoms are unstable in the centre of a gold
cluster. Also C–Au bonds are weaker than on most transition metals, so
the nucleation cap would have difficulty to form [41,42].
Recently, it has been found that gold nano-particles can be active,
they are catalytic active for some oxidation reactions. There have been
two basic explanations. Firstly, while bulk gold is inactive, nanoparticles have a much larger surface area and much more of the atoms
are active surface atoms with low coordination [43]. This might also
alter their band structure. Second, electronegative metals can interact
with defects in the oxide support and lead to a charge transfer and the
Au− states catalyse reactions [44–46].
Experimentally, a number of groups have been able to grow carbon
nanotubes or carbon nanostructures using gold as a catalyst but with
low yield [47–51]. Radial breathing modes show that SWNTs are
formed in some cases. Fig. 10 shows a TEM image of CNTs grown by us
at 850 °C in acetylene. This requires that gold both catalyses
hydrocarbon dissociation and that C has a finite solubility. The work
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
H. Dai, Acc Chem. Res. 35 (2002) 1035.
B.Q. Wei, R. Vajtai, P.M. Ajayan, Appl. Phys. Lett. 79 (2001) 1172.
M. Nihei, A. Kawabata, D. Kondo, Y. Awano, Jpn. J. Appl. Phys. 43 (2004) 1856.
J. Robertson, G. Zhong, H. Telg, C. Thomsen, J.H. Warner, G. Briggs, U. Detlaff, S.
Roth, Appl. Phys. Lett. 93 (2008) 163111.
A. Javey, J. Guo, Q. Wang, M. Lundstrom, H.J. Dai, Nature 424 (2003) 654.
D.N. Futaba, K. Hata, T. Yamada, T. Hiraoka, Y. Hayamizu, Y. Kakudate, O. Tanake, H.
Hatori, M. Yumura, S. Iijima, Nat. Matters 5 (2006) 987.
T. Yamada, T. Namai, K. Hata, D.N. Futaba, K. Mizuno, J. Fan, M. Yudasaka, M.
Yumura, S. Iijima, Nat. Nanotechnol. 1 (2007) 131.
S. Hofmann, C. Ducati, B. Kleinsorge, J. Robertson, Appl. Phys. Lett. 83 (2003) 4661.
W.I. Milne, K. Teo, G. Amaratunga, P. Legagneux, L. Gangloff, J.P. Schnell, V. Semet,
V.T. Binh, P. Groning, J. Mater. Chem. 14 (2004) 933.
M. Cantoro, S. Hofmann, S. Pisana, V. Scardaci, A. Parvez, C. Ducati, A.C. Ferrari, A.M.
Blackburn, K.Y. Wang, J. Robertson, Nano Lett. 6 (2006) 1107.
R.T.K. Baker, J. Catal. 26 (1972) 51.
J.R. Rostrup-Nielsen, J. Catal. 33 (1974) 184.
H. Katuara, Y. Kumazawa, Y. Maniwa, Y. Ohtsuka, R. Sen, S. Suzuki, Y. Achiba, Carbon
38 (2000) 1691.
X. Li, X. Tu, S. Zaric, K. Weischer, H. Dai, J. Am. Chem. Soc. 129 (2007) 15770.
S.M. Bachilo, L. Balzano, J.E. Herrera, F. Pompeo, D.E. Resasco, R.B. Weisman, J. Am.
Chem. Soc. 125 (2003) 11186.
A.A. Puretzky, D.B. Geohegan, X. Fan, S.J. Pennycook, Appl. Phys. Lett. 76 (2000)
182.
P. Buffat, J.P. Borel, Phys. Rev. A 13 (1976) 2287.
J. Gavillet, A. Loiseau, C. Journet, F. Willaime, F. Ducastelle, J.C. Charlier, Phys. Rev.
Lett. 87 (2001) 275504.
Y. Li, W. Kim, Y. Zhang, M. Rolandi, D. Wang, H. Dai, J. Phys. Chem. B 106 (2002)
12361.
F. Ding, K. Bolton, A. Rosen, J. Vac. Sci. Technol. A 22 (2004) 1471.
Y. Qi, T. Cagin, W.L. Johnson, W.A. Goddard, J. Chem. Phys. 115 (2001) 385.
A. Harutyunyan, T. Tokuna, E. Mora, Appl. Phys. Lett. 87 (2005) 051919.
(a) H. Ago, S. Ohshima, K. Uchida, M. Yumura, J. Phys. Chem. B 105 (2001) 10453;
(b) T. Saito, S. Ohshima, W.C. Xu, H. Ago, M. Yumura, S. Iijima, J. Phys. Chem. B 109
(2005) 10647.
Y.L. Li, I.A. Kinloch, A.H. Windle, Science 304 (2006) 276.
S. Hofmann, R. Sharma, C. Ducati, G. Du, C. Mattevi, C. Cepek, M. Cantoro, S. Pisana,
A. Parvez, F. Cervantes-Sodi, A.C. Ferrari, R.E. Dunin-Borkowski, J. Robertson, Nano
Lett. 7 (2007) 602.
S. Helveg, C. Lopez-Cartes, J. Sehested, P.L. Hansen, B.S. Clausen, J.R. RostrupNielsen, F. Abild-Pedersen, J.K. Norskov, Nature 427 (2004) 426.
M. Lin, J.P.Y. Tan, C. Boothroyd, K.P. Loh, E.S. Tok, Y.L. Foo, Nano Lett. 6 (2006) 449.
(a) H.J. Frost, M.F. Ashby, Deformation mechanism maps, Pergamon Press, Oxford,
1982;
(b) M.F. Ashby, Acta Met. 20 (1972) 887.
C. Ducati, I. Alexandrou, M. Chhowalla, J. Robertson, G.A.J. Amaratunga, J. Appl.
Phys. 95 (2004) 6387.
C. Mattevi, C.T. Wirth, S. Hofmann, R. Blume, M. Cantoro, C. Ducati, C. Cepek, A.
Knop-Gericke, S. Milne, C. Castellarin-Cudia, S. Dolafi, A. Goldoni, R. Schlogl, J.
Robertson, J. Phys. Chem. C 112 (2008) 12207.
S. Hofmann, R. Blume, C.T. Wirth, M. Cantoro, R. Sharma, C. Ducati, M. Havecker, S.
Zafeiratos, P. Schnoerch, A. Oestereich, D. Teschner, M. Albrecht, A. Knop-Gericke,
R. Schlogl, J. Robertson, J. Phys. Chem. C 113 (2009) 1648.
M. Cantoro et al., accepted J App Phys (2008).
T. de los Arcos, M.G. Garnier, J.W. Seo, P. Oelhafen, V. Thommen, D. Mathys, J. Phys.
Chem. B 108 (2004) 7728.
T. de los Arcos, P. Oelhafen, V. Thommen, D. Mathys, J. Phys. Chem. C 111 (2007)
16392.
S. Hofmann, G. Csanyi, A.C. Ferrari, M.C. Payne, J. Robertson, Phys. Rev. Lett. 95
(2005).
D. Teschner, A. Vass, M. Havecker, S. Zafeiratos, P. Schnorch, H. Sauer, A. KnopGericke, R. Schlogl, M. Chamann, A. Wootsch, A.S. Canning, J.J. Gamman, S.D.
Jackson, J. McGregor, L.F. Gladden, J. Catal. 242 (2006) 26.
Please cite this article as: C.T. Wirth, et al., Diamond Relat. Mater. (2009), doi:10.1016/j.diamond.2009.01.030
ARTICLE IN PRESS
6
C.T. Wirth et al. / Diamond & Related Materials xxx (2009) xxx–xxx
[37] E.M. Vass, M. Havecker, Z. Zafeiratos, D. Teschner, A. Knop-Gericke, R. Schlogl,
J. Phys. Condens. Matter 20 (2008) 1.
[38] L. Gracia, M. Catatayud, J.A.C. Minot, M. Salmeron, Phys. Rev. B 71 (2005) 033407.
[39] S. Hofmann, C. Ducati, B. Kleinsorge, J. Robertson, Appl. Phys. Lett. 83 (2003) 135.
[40] J.V. Raty, F. Gygi, G. Galli, Phys. Rev. Lett. 95 (2005) 096103.
[41] F. Ding, P. Larsson, A. Larsson, R. Ahuja, H. Duan, A. Rosen, K. Bolton, Nano Lett.
8 (2008) 463.
[42] W. Deng, X. Xu, W.A. Goddard, Nano Lett. 4 (2004) 2331.
[43] N. Lopez, T.V.W. Janssens, B.S. Clausen, Y. Xu, M. Mavrikas, T. Bligaard, J.K. Norskov,
J. Catal. 223 (2004) 232.
[44] A. Sanchez, S. Abbet, U. Heiz, W.D. Schneider, H. Hakkinen, R.N. Barnett, U.
Landman, J. Phys. Chem. A 103 (1999) 9573.
[45] G. Pacchioni, S. Sicolo, C. DiValentin, M. Chiesa, E. Giamello, J. Amer. Chem. Soc.
130 (2008) 8690.
[46] J. Robertson, O. Sharia, A. Demkov, Appl. Phys. Lett. 91 (2007) 132912.
[47] D. Takagi, Y. Homma, H. Hibino, S. Suzuki, Y. Kobayashi, Nano Lett. 6 (2006) 2642.
[48] D. Takagi, Y. Kobayashi, H. Hibino, S. Suzuki, Y. Homma, Nano Lett. 8 (2008) 832.
[49] S. Bhaviripudi, E. Mile, S.A. Steiner, A.T. Zare, M.S. Dresselhaus, A.M. Belcher, J.
Kong, J. Amer. Chem. Soc. 129 (2007) 1516.
[50] S.Y. Lee, M. Yamada, M. Miyake, Carbon (2005) 2654.
[51] N. Yoshihara, H. Ago, M. Tsuji, Jpn. J. Appl. Phys. 47 (2008) 1944.
[52] H. Yoshida, S. Takeda, T. Uchiyama, H. Kohno, Y. Homma, Nanoletts 8 (2008) 2086.
Please cite this article as: C.T. Wirth, et al., Diamond Relat. Mater. (2009), doi:10.1016/j.diamond.2009.01.030