Chemical vapor deposition of nickel-group metals on multiwall carbon nanotubes

Article abstract of DOI:10.1139/V07-014

Chemical vapor deposition (CVD) of Ni, Pd, and Pt films and of Ni/Pd and Pd/Pt bimetallic films on multiwall carbon nanotubes (MWCNTs) can be effected at low temperature if the nanotubes are pretreated by CVD of titanium carbide. In the absence of the pretreatment, the CVD leads to formation of isolated nanoparticles of the nickel-group metals. The metallized MWCNTs are curved or kinked, as a result of the interaction with the metal. Preliminary oxidation of the carbon nanotubes allows easier metallization, and the bending of the metallated nanotubes is not observed in this case.

Full text of DOI:10.1139/V07-014

645
Chemical vapor deposition of nickel-group metals
on multiwall carbon nanotubes1
Maoqi Feng and Richard J. Puddephatt
Abstract: Chemical vapor deposition (CVD) of Ni, Pd, and Pt films and of Ni/Pd and Pd/Pt bimetallic films on
multiwall carbon nanotubes (MWCNTs) can be effected at low temperature if the nanotubes are pretreated by CVD of
titanium carbide. In the absence of the pretreatment, the CVD leads to formation of isolated nanoparticles of the
nickel-group metals. The metallized MWCNTs are curved or kinked, as a result of the interaction with the metal. Pre-
liminary oxidation of the carbon nanotubes allows easier metallization, and the bending of the metallated nanotubes is
not observed in this case.
Key words: Chemical vapor deposition, platinum, palladium, nickel, carbon, nanotube.
Résumé : Si on opère avec des nanotubes qui ont été prétraités par un dépôt de vapeur chimique (DVC) de carbure de
titane, on peut effectuer à basse température des dépôts de vapeurs chimiques de films de Ni, Pd et Pd et de films bi-
métalliques de Ni/Pd et Pd/Pt sur des nanotubes de carbone à parois multiples (NCPM). En l’absence d’un prétraite-
ment, le DVC conduit à la formation de nanoparticules isolées du métal du groupe du nickel. Les NCPM métallisées
sont recourbées ou déviées en raison de l’interaction avec le métal. Une oxydation préliminaire des nanotubes de car-
bone facilite la métallisation et, dans ce cas, on n’observe pas de courbure dans les nanotubes métallisés.
Mots-clés : dépôt de vapeur chimique, platine, palladium, nickel, carbone, nanotube.
[Traduit par la Rédaction] Feng and Puddephatt 650
Introduction
isolated nanoparticles, which are often located at defect sites
of the nanotube. One method that has been successful in
coating nanotubes with metals by electron-beam evaporation
is to apply a preliminary coating of titanium, which forms a
stable carbide at the interface, so it gives a continuous and
uniform coating (9). Subsequent evaporation of other metals
onto the titanium-coated carbon nanotubes can then give
uniform metal coatings (9). This article reports the
metallization of multiwall carbon nanotubes (MWCNT’s)
with the metals nickel, palladium, and platinum or with the
bimetallics Ni/Pd and Pd/Pt by CVD techniques. Because di-
rect deposition of the metals led to the formation of isolated
metal nanoparticles, a new method was developed in which
preliminary CVD of titanium carbide was carried out to give
a continuous coating, followed by CVD of the nickel-group
metal. An unexpected discovery was that the CVD on
the MWCNT can lead to the bending of the carbon nano-
tube. We have previously reported the low-temperature
CVD of palladium and platinum on organic materials, in-
cluding deposition on polystyrene beads to give
macroporous films of the metal (10).
The surface metallization of inorganic and organic materi-
als has relevance in areas such as geochemistry and nano-
electronics, so it is an active area of research (1, 2). For ex-
ample, there has been much interest in the use of chemical
vapor deposition (CVD) to form thin films of transition met-
als (3), including nickel (4), palladium (5), and platinum (6),
on metals, semiconductors, and insulating materials. The
metallization of carbon nanotubes to give a hybrid material
for potential applications in electronic devices or catalysis is
rapidly developing; however, although transition-metal nano-
particles are often used as catalysts for forming carbon
nanotubes (7), the use of CVD as a method of metallizing
carbon nanotubes has been neglected (8, 9). Instead, carbon
nanotubes are usually coated by solution methods, such as
electroless deposition, or by metal evaporation–sputtering
methods (8, 9). In all coating methods, it is important to rec-
ognize that most late transition metals bind only weakly to
the nanotube surface; hence, metal deposition tends to give
Received 17 November 2006. Accepted 26 January 2007.
Published on the NRC Research Press Web site at
canjchem.nrc.ca on 11 April 2007.
Results and discussion
Dedicated to Mike Bancroft, friend, colleague, and
distinguished scientist.
Direct deposition experiments
The atmospheric-pressure CVD experiments were con-
ducted by using a hot-wall CVD apparatus (10) with the
multiwall carbon nanotubes supported on the surface of a
silicon wafer and using hydrogen as the carrier gas. Initial
experiments were carried out to deposit nickel, using the
precursor [Ni(Cp)2]; palladium, using the precursor [Pd(hfac)2];
and platinum, using the precursor [Pt(COD)Me2], with the
M. Feng and R.J. Puddephatt.2 Department of Chemistry,
University of Western Ontario, Chemistry Building, Room
100, London, ON N6A 5B7, Canada.
1This article is part of a Special Issue dedicated to Professor
G. Michael Bancroft.
2Corresponding author (e-mail: pudd@uwo.ca).
Can. J. Chem. 85: 645–650 (2007)
doi:10.1139/V07-014
© 2007 NRC Canada

646
Can. J. Chem. Vol. 85, 2007
Table 1. Surface compositions (atom%) of the metallized
precursor warmed to 60 °C and the substrate heated to 200–
250 °C. CVD of bimetallic alloys Ni/Pd and Pd/Pt was
carried out by using equimolar amounts of the appropriate
precursors. The metallized MWCNT’s were then structurally
characterized by imaging, SEM or TEM, and the composi-
tion of the film was studied by XPS and XRD.
MWCNTs determined by XPS.
Metal
C
O
Nia
69
66
85
87
79
80
17
18
15
13
21
20
19
25
18
16
14
16
0
0
0
0
8
8
0
Nib
Pda
The metallic phases were characterized by the binding en-
ergies of Pt 4f7/2 at ~71.2 eV, Pd 3d5/2 at 335 eV, and Ni
2p3/2 at 853 eV, with only minor variations as a function of
film thickness and alloy formation. It has previously been
shown that the metals are formed in high purity by CVD on
silicon or glass by using hydrogen as a carrier gas (6). In the
present experiments, CVD was carried out on carbon nano-
tubes, and XPS analysis always showed the presence of car-
bon. For example, a sample of carbon nanotubes coated with
palladium was analyzed as (atom%): Pd, 35; C, 65 atom%,
but the composition varied across the sample. It is shown
later by SEM or TEM imaging that the nanotubes are not
evenly coated by the direct CVD route, so it is clear that
much of the detected carbon is the nanotube substrate (and
graphitic impurity in the commercial samples of MWCNT).
The composition of the Ni/Pd and Pd/Pt alloys, formed by
CVD, was determined by XPS analysis to be Ni:Pd ≈ 1:1
and Pd:Pt ≈ 3:1, which reflects the relative ease of decompo-
sition of the two precursors in an individual CVD experi-
ment. The nickel-coated MWCNT samples gave significant
oxygen analysis (Table 1), although NiO was not detected by
XRD analysis. It is likely that the nickel surface is oxidized
during storage and transfer, prior to XPS analysis.
Pdb
Pta
Ptb
Ni/Pda
Ni/Pdb
Pd/Pta
Pd/Ptb
Ni,38; Pd,35
Ni,33; Pd,34
Pd,60; Pt,22
Pd,61; Pt,23
0
a,bValues for films on unoxidized and oxidized carbon nanotubes,
respectively.
Fig. 1. SEM image of a MWCNT coated with nickel
nanoparticles.
SEM examination of the metallized carbon nanotubes
showed the presence of isolated metal nanoparticles in all
cases. For example, Fig. 1 shows an SEM image of a carbon
nanotube coated with nickel, and Fig. 2 shows a TEM image
of a carbon nanotube coated with palladium. They are repre-
sentative of similar images obtained for nanotubes coated
with nickel, palladium, or platinum or with Ni/Pd or Pd/Pt
alloys. Clearly, there is no continuous coating, but isolated
nanoparticles of nickel or palladium are formed, with aver-
age diameter of ~20 nm. This structure arises because the
nucleation rate is low, and growth at the forming metal sur-
face is fast, while the metal – carbon nanotube interactions
are relatively weak. The autocatalysis of the CVD of nickel,
palladium, and platinum by the reactive metal surface is well
documented (5, 6).
Pre-coating with titanium carbide
Titanium carbide can be formed at high temperature
(>800 °C) by the reaction of titanium atoms with carbon
nanotubes, and it has been used to connect individual
nanotubes (11). In the present work, titanium carbide was
deposited by CVD at 400 °C on carbon nanotubes, using
[Cp2TiCl2] as a precursor (12). The film formed this way
was analyzed by XPS and XRD. The Ti 2p3/2 peak for TiC
was broad with components from 455–459 eV, whereas crys-
talline TiC is reported at 455.1 eV and TiO2 (anatase or
rutile) at 459 eV (13). The data are similar to those reported
for partially oxidized TiC(O) films, for which components
with Ti 2p3/2 = 456 and 458 eV were observed (13). In con-
firmation, an O 1s peak at 530 eV was observed, close to the
literature values for TiO2 of 529.7–529.9 eV. The C 1s peak
was dominated by the peak for the carbon nanotubes at
284.5 eV, but there was a lower-energy peak at ~283.5 eV,
close to the value reported for carbide in Ti–O–C films at
283.45 eV (13). Finally, XRD analysis confirmed the pres-
ence of a crystalline TiC phase (13). XPS analysis gave
(atom%) Ti:C:O (49:40:11). Thus, the thin films are identi-
fied as partially oxidized TiC.
An unexpected feature of the images obtained by SEM or
TEM was that many of the TiC-coated carbon nanotubes
were curved or kinked, whereas all of the uncoated
nanotubes were linear. Figure 3 shows a TEM image of one
such curved nanotube. The deposition of titanium carbide on
fullerenes has been shown to cause surface reconstruction
(9, 14), while the bending of carbon nanotubes may be
© 2007 NRC Canada

Feng and Puddephatt
647
Fig. 3. A TEM image of a MWCNT coated with TiC, showing
Fig. 2. TEM image of a MWCNT coated with palladium
nanoparticles.
the curvature induced by the CVD process.
Fig. 4. An SEM image of a carbon nanotube coated first with
TiC and then with Pd/Pt alloy. The background material arises
from a graphitic impurity in the carbon nanotubes.
caused by replacement of carbon hexagons by pentagons
(14), so it is possible that the bending observed in Fig. 3
arises as a result of surface reconstruction of the carbon
nanotube, caused by the deposition of titanium carbide.
Nickel-group elements were deposited on the titanium
carbide-coated carbon nanotubes by CVD, as previously de-
scribed; however, under these conditions, the coatings were
continuous and more even. Qualitatively, the deposition of
nickel-group metals was faster on the TiC-coated nanotubes
because the induction period for nucleation on the carbon
surface of the pure nanotube was no longer present. Thus, it
is shown that TiC is an effective intermediate layer because
it can interact strongly with both the carbon nanotube sub-
strate and the nickel-group metal. Figure 4 shows an SEM
image of a carbon nanotube coated first with TiC and then
with Pd/Pt alloy. Comparing with Fig. 1, it is clear that the
metal coating is more continuous and even, with no isolated
metal nanoparticles.
Some typical analytical data by EDX for the continuous
nickel-group coatings obtained are listed in Table 1. In sam-
ples with thicker coatings, the TiC buffer layer was not de-
tected, but the samples always contained residual carbon
impurity in the range of 15–21 atom%. It is not obvious
whether the carbon impurity arises from the substrate, from
carbon in the ligands used, or from a combination of both.
The samples containing Ni or Ni/Pd coatings also gave sig-
nificant oxygen content, indicating partial oxidation of the
surface nickel. XPS binding energies showed that Ni, Pd,
and Pt coatings contained the elements in the metallic state.
Figure 5 shows TEM images of nanotubes coated with
TiC and then with a nickel-group metal. Figure 5(a) shows a
nanotube coated with TiC and then with a thin coat of
nickel. The kink in the nanotube and the even coating of the
nanotube with nickel are obvious features. Less obvious is
the presence of a helical appearance along the nanotube.
Figure 5b shows CVD-coated nanotubes first with TiC and
then with a thicker coating of Ni/Pd alloy. The apparent heli-
cal structure is now very pronounced and appears to arise
© 2007 NRC Canada

648
Can. J. Chem. Vol. 85, 2007
Fig. 5. TEM images of MWCNTs coated by CVD: (a) TiC, fol-
lowed by Ni; (b) TiC, followed by Pd/Ni alloy.
Fig. 6. A TEM image of a bundle of oxidized carbon nanotubes
with buffer layer of TiC/TiO2 and final coating of palladium.
from multiple bends of the coated nanotubes. The metallic
coating is again uniform and continuous along the length of
the nanotube. These images further show that the problem of
forming disconnected particles of Ni, Pd, Pt, and Ni/Pd or
Pd/Pt bimetallic alloys by direct CVD onto the MWCNTs
(Fig. 2) is overcome by precoating the nanotubes with TiC
(Figs. 4, 5), which forms a strong interaction with
the MWCNTs and gives a high nucleation rate for subse-
quent metal deposition.
unoxidized carbon nanotubes (Table 1). However, no bend-
ing or kinking of the coated nanotubes was observed by
SEM or TEM imaging, in contrast to the observations with
unoxidized carbon nanotubes described earlier. Figure 6
shows a TEM image of a bundle of oxidized carbon
nanotubes coated first with TiC/TiO2 and then with palla-
dium, showing that the nanotubes remained linear.
The uniform coating of the oxidized carbon nanotubes
with TiC/TiO2 suggested that it might be possible to coat the
nanotubes with a pure metal oxide (15, 17). This hypothesis
was tested by CVD of zirconium dioxide, using [Zr(DPM)4]
as a precursor (DPM = dipivaloylmethanate). Zirconia formed
a continuous coating on the oxidized carbon nanotubes
(Fig. 7a), and it was identified by XRD (Fig. 8). Further
coating with nickel gave a continuous film (though with
some visible imperfections) rather than isolated nano-
particles (Fig. 7b). A thick film gave the composition
(atom%) of Ni:C:O (62:19:19), similar to that obtained for
direct deposition on the oxidized carbon nanotubes. Again,
no bending or kinking of the oxidized carbon nanotubes was
observed as a result of the CVD-coating processes.
Pre-oxidation of the carbon nanotubes
Oxidation of carbon nanotubes by nitric acid is known to
introduce carboxylic acid, carbonyl, and hydroxyl groups at
the surface, and these functional groups can act as nucle-
ation centres in the deposition of metals, such as silver, gold,
and platinum from solution (15).
CVD onto the oxidized carbon nanotubes, using
[Cp2TiCl2] as a precursor, gave a mixture of TiC and TiO2 in
a ratio of roughly 1:3, as determined by XRD and XPS anal-
yses. XRD peaks for TiO2 are broadened, indicating the for-
mation of small particles, and the average particle size is
estimated to be about 6 nm, as calculated from the Scherrer
equation (16). The much greater degree of oxidation of tita-
nium to form TiO2 on the oxidized compared with the un-
treated carbon nanotubes indicates that the oxophilic
titanium abstracts oxygen atoms from the oxygen-containing
functional groups on the nanotube walls. Nevertheless, it is
possible to deposit nickel-group metals by CVD to give con-
tinuous coatings on these treated nanotubes, and the compo-
sitions for thick films are very similar to those for films on
Conclusions
The deposition of the transition metals Ni, Pd, and Pt, and
of the alloys Ni/Pd and Pd/Pt, on MWCNTs by CVD, using
a buffer layer of TiC, provides a viable method for the
metallization of carbon nanotubes, with potential uses as
nanowires etc. The unusual bending of the MWCNTs, ob-
served during CVD, is caused by the interaction of
© 2007 NRC Canada

Feng and Puddephatt
649
Fig. 7. TEM images of carbon nanotubes coated with (a) ZrO2
and (b) ZrO2, followed by Ni.
section of silicon wafer (1 cm in diameter), followed by
evaporation of the solvent at 60 °C.
CVD was conducted at atmospheric pressure by using a
horizontal glass hot-wall CVD apparatus (10). The carrier
gas was ultra-high purity H2 at a flow rate of 100 mL min–1.
The precursor was heated to 60 °C to give a suitable rate of
evaporation in the carrier-gas stream. The substrate was
heated as follows: deposition of TiC at 400 °C, Ni at 200 °C,
Pd or Pt (or alloys of these elements) at 250 °C. Under these
conditions, a typical CVD experiment was complete in 2 h.
After the deposition, the silicon wafer was removed from the
reactor, and the metallized MWCNTs were removed from
the silicon surface and stored in vacuum for further analysis.
For CVD of alloy materials, a mixture of [Ni(Cp)2]–
[Pd(hfac)2] or [Pd(hfac)2]–[Pt(COD)Me2] was used for the
deposition of Ni/Pd or Pd/Pt, respectively.
XPS spectra were recorded by using a SSL SSX-100
small-spot XPS surface-analysis instrument with a mono-
chromatized Mg Kα X-ray source (1253.6 eV); binding en-
ergies were reproducible to within 0.1 eV. X-ray diffraction
(XRD) profiles were measured using a Rigaku diffracto-
meter with Co Kα radiation (wavelength = 1.790210 Å) at
40 kV and 40 mA, operating over a 2° < 2θ < 82° angular
range. TiC was identified by comparison to standard data
from the Joint Committee for Powder Diffraction Studies
(JCPDF) card 32–1383. SEM images were obtained using a
Hitachi S-4500 scanning electron microscope. TEM images
were obtained using a Philips EM300 transmission electron
microscope system, operating at 80 keV.
Acknowledgments
Fig. 8. XRD of carbon nanotubes coated with ZrO2. The broad-
ening of the ZrO2 peaks is attributed to the nanocrystalline na-
ture of the zirconia.
We thank Surface Science Western for SEM measure-
ments and EMK (Ontario) and the Natural Sciences and En-
gineering Research Council of Canada (NSERC) for
financial support. RJP thanks the Government of Canada for
a Canada Research Chair.
References
1. (a) G.E. Jean and G.M. Bancroft. Geochim. Cosmochim. Acta,
50, 1455 (1986); (b) Z. Zhang, E.S. Yamaguchi, M. Kasrai,
and G.M. Bancroft. Tribology Lett. 19, 211 (2005).
the MWCNTs with the adsorbate. Thus, although a detailed
explanation is not possible at this stage, it appears likely that
adsorbate-induced reconstruction of the nanotube surface oc-
curs, and that the distortions are amplified by the subsequent
epitaxial growth of the metal films.
2. Y. Zhou. Curr. Nanosci. 2, 123 (2006).
3. (a) T.T. Kodas and M.J. Hampden-Smith. The chemistry of
metal CVD. VCH, New York. 1994; (b) R.J. Puddephatt. Poly-
hedron, 13, 1233 (1994); (c) J.C. Hierso, R. Feurer, and P.
Kalck. Chem. Mater. 12, 390 (2000); (d) J.R.V. Garcia and T.
Goto. Mater. Trans. 44, 1717 (2003).
Experimental
4. (a) L. Mond, C. Langer, and F. Quincke. J. Chem. Soc., Trans.
749 (1890); (b) L. Brissoneau and C. Vahlas. Chem. Vap. De-
position, 5, 135 (1999); (c) V.V. Bakovets, V.N. Mitkin, and
N.V. Gelfond. Chem. Vap. Deposition, 11, 112 (2005).
5. (a) W. Lin, S.R. Wilson, and G.S. Girolami. Inorg. Chem. 33,
2265 (1994); (b) Z. Yuan, D. Jiang, S.J. Naftel, T.-K. Sham,
and R.J. Puddephatt. Chem. Mater. 6, 2151 (1994); (c) Y.
Zhang and R.J. Puddephatt. Chem. Vap. Deposition, 3, 81
(1997); (d) Y. Zhang, Z. Yuan, and R.J. Puddephatt. Chem.
Mater. 10, 2293 (1998); (e) Y. Zhang, S.W.-K. Choi, and R.J.
Puddephatt. J. Am. Chem. Soc. 119, 9295 (1997); ( f ) Y.H.
Liu, Y.C. Cheng, Y.L. Tung, Y. Chi, Y.L. Chen, C.S. Liu, S.M.
Peng, and G.H. Lee. J. Mater. Chem. 13, 135 (2003); (g) U.
The precursors used for CVD of TiC, Ni, Pd, and Pt were
[Cp2TiCl2], [Ni(Cp)2], [Pd(hfac)2] and [Pt(COD)Me2].
[Cp2TiCl2] and [Ni(Cp)2] (Cp = η-C5H5) were commercial
samples (Aldrich, Strem), and [Pd(hfac)2] (hfac
=
hexafluoroacetylacetonate) and [Pt(COD)Me2] (COD = 1,5-
cyclooctadiene) were prepared as described previously
{[Pd(hfac)2], [Pt(COD)Me2]} (5, 6). Multiwall carbon
nanotubes (MWCNT, 7–12 nm outer diameter, 0.5–10 µm
long) were purchased from Alfa Aesar in the form of a sus-
pension in toluene. The substrates for CVD were prepared
by adding a drop of the MWCNT suspension in toluene to a
© 2007 NRC Canada

650
Can. J. Chem. Vol. 85, 2007
Weckenmann, S. Mittler, S. Kramer, A.K.A. Aliganga, and
R.A. Fischer. Chem. Mater. 16, 621 (2004); (h) Q.-H. Wu, M.
Gunia, T. Strunskus, G. Witte, M. Muhler, and C. Woell.
Chem. Vap. Deposition, 11, 355 (2005).
11. (a) Y. Zhang, T. Ichihashi, E. Landree, F. Nihey, and S. Ijima.
Science (Washington, D.C.), 285, 1719 (1999); (b) J.-O. Lee,
C. Park, J.-J. Kim, J. Kim, J.W. Park, and K.-H. Yoo. J. Phys.
D: Appl. Phys. 33, 1953 (2000).
6. (a) R. Kumar, S. Roy, M. Rashidi, and R.J. Puddephatt. Poly-
hedron, 8, 551 (1989); (b) Z. Xue, M.J. Strouse, D.K. Shuh,
C.B. Knobler, H.D. Kaesz, R.F. Hicks, and R.S. Williams. J.
Am. Chem. Soc. 111, 8779 (1989); (c) N.H. Dryden, R.
Kumar, E. Ou, M. Rashidi, S. Roy, P.R. Norton, and R.J.
Puddephatt. Chem. Mater. 3, 677 (1991); (d) B. Nixon, P.R.
Norton, E.C. Ou, R.J. Puddephatt, S. Roy, and P.A. Young.
Chem. Mater. 3, 222 (1991); (e) Z. Xue, H. Thridandam, H.D.
Kaesz, and R.F. Hicks. Chem. Mater. 4, 162 (1992); ( f ) T.P.
Martin, C.P. Tripp, and W.J. DeSisto. Chem. Vap. Deposition,
11, 170 (2005); (g) G.A. Battiston, R. Gerbasi, and A. Rodri-
guez. Chem. Vap. Deposition, 11, 130 (2005).
7. (a) J.F. Geng, D. Jefferson, and B.F.G. Johnson. J. Mater.
Chem. 15, 844 (2005); (b) S.Y. Lee, M. Yamada, and M.
Miyake. Sci. Tech. Adv. Mater. 6, 420 (2005).
8. (a) L.M. Ang, T.S.A. Hor, G.Q. Xu, C.H. Tung, S.P. Zhao, and
J.L.S. Wang. Chem. Mater. 11, 2115 (1999); (b) X.R. Ye, Y.H.
Lin, and C.M. Wai. Chem. Commun. 642 (2003); (c) C.H.
Yen, X.L. Cui, H.B. Pan, S.F. Wang, Y.H. Lin, and C.M. Wai.
J. Nanosci. Nanotechnol. 5, 1852 (2005); (d) Y. Wang, X. Xu,
Z.Q. Tian, Y. Zong, H.M. Cheng, and C.J. Lin. Chem. Eur. J.
12, 2542 (2006).
12. J. Slifirski, G. Huchet, A. Reynes, A. Marty, and F.
Teyssandier. Chem. Mater. 7, 622 (1995).
13. (a) Handbook of X-ray Photoelectron spectroscopy.
PerkinElmer Corp., 1992; (b) V. Schier, H.-J. Michel, and J.
Halbritter. Fres. J. Anal. Chem. 346, 227 (1993); (c) L.
Ramqvist, K. Hamrin, G. Johansson, A. Fahlman, and C.
Nordling. J. Phys. Chem. Solids, 30, 1835 (1969); (d) H. Ihara,
Y. Kumashiro, A. Itoh, and K. Maeda. Jpn. J. Appl. Phys. 12,
1462 (1973); (e) U. Muller and R Hauert. Thin Solid Films,
290–292, 323 (1996); ( f ) G.A. Swift and R. Koc. J. Mater.
Sci. 34, 3083 (1999).
14. (a) A.J. Maxwell, P.A. Bruhwiler, S. Andersson, N.
Martensson, and P. Rudolf. Chem. Phys. Lett. 247, 257 (1995);
(b) M. Terrones. Annu. Rev. Mater. Res. 33, 419 (2003); (c) L.
Chico, V.H. Crespi, L.H. Benedict, S.G. Louie, and M.L. Co-
hen. Phys. Rev. Lett. 76, 971 (1996).
15. (a) T.W. Ebbesen, H. Hiura, M. E. Bisher, M.M.J. Treacy, J.L.
Shreeve-Keyer, and R.C. Haushalter. Adv. Mater. 8, 155
(1996); (b) M. Kosaka, T.W. Ebbesen, H. Hiura, and K.
Tanigaki. Chem. Phys. Lett. 233, 47 (1995).
16. A.D. Krawitz. Introduction to diffraction in materials science
and engineering. Wiley, New York. 2001.
9. (a) Y. Zhang and H.J. Dai. Appl. Phys. Lett. 7, 3015 (2000);
(b) Y. Zhang, N.W. Franklin, R.J. Chen, and H.J. Dai. Chem.
Phys. Lett. 33, 35 (2000).
17. (a) W.-Q. Han and A. Zettl. Nano Lett. 3, 681 (2003); (b) D.F.
Johnson, B.J. Craft, and S.M. Jaffe. J. Nanopart. Res. 3, 63
(2001).
10. M. Feng and R.J. Puddephatt. Chem. Mater. 15, 2696 (2003).
© 2007 NRC Canada