View Article Online / Journal Homepage / Table of Contents for this issue

ChemComm Dynamic Article Links

Cite this: Chem. Commun., 2012, 48, 853–855

www.rsc.org/chemcomm COMMUNICATION
Morphology-directed synthesis of Co3O4 nanotubes based on modified
Kirkendall effect and its application in CH4 combustionw
Published on 02 December 2011. Downloaded by University of York on 16/07/2013 11:02:22.

Zhaoyang Fei,a Shengchao He,a Lei Li,a Weijie Ji*a and Chak-Tong Au*b
Received 26th September 2011, Accepted 16th November 2011
DOI: 10.1039/c1cc15976c

We reported the morphology-directed synthesis of Co3O4 nano-

tubes via interfacial reaction of NaOH with pre-fabricated
CoC2O4
2H2O nanorods based on modified Kirkendall effect.
The as-obtained Co3O4 nanotubes showed excellent activity and
durability in catalytic combustion of CH4.

In recent years, nanoparticles (NPs) with hollow structure

have received great attention due to their unique properties
and potential application in different fields.1–3 Many synthetic
approaches have been developed for constructing various
kinds of hollow structures. By the use of different functional
building blocks and/or unique processes such as acid etching,
solid-state thermal decomposition and so forth, hollow nano-
structures of Cu2O,2 Cu7S4,4 Fe2O3,5 MnO2,6 and NiO7 were
synthesized. However, the processes were generally conducted
at high temperature and/or involve the use of corrosive acid
and costly template. To fabricate hollow nanostructures of
controlled morphology under milder and more environment-
friendly conditions is hence a challenge.
In principle, hollow nanostructures can be generated
through a mechanism called the Kirkendall effect in which
pores can be created inside solid NPs due to the difference in
diffusion rates between two components of a diffusion couple.
The related fabrication route has recently been applied to the
making of hollow structures.8 Compared to conventional methods,
the Kirkendall-based one is easy to conduct, template-free,
and suitable for large-scale production. To date, the examples
of preparation of nanotubes (NTs) based on the Kirkendall
effect are limited.9
Co3O4 is not only an important p-type semiconductor but also
shows unique magnetic property.10 There are reports on Co3O4
nanostructures of different morphologies.11 Co3O4 NTs can be
prepared through thermal decomposition of Co(NO3)2
6H2O in
the presence of AAO template.12 In the present study, we adopted
a
Key Laboratory of Mesoscopic Chemistry, MOE, School of
Chemistry and Chemical Engineering, Nanjing University,
Nanjing 210093, P. R. China. E-mail: jiwj@nju.edu.cn;
Fax: +86-25-83317761
b
Department of Chemistry, Hong Kong Baptist University,
Hong Kong, P. R. China. E-mail: pctau@hkbu.edu.hk;
Fax: +852-3411-7348
w Electronic supplementary information (ESI) available: Experimental
details; physicochemical properties of various Co3O4 samples; nitrogen
adsorption–desorption isotherms; TGA profiles; TEM images; SEM
images; H2-TPR; O2-TPD; durability test. DOI: 10.1039/c1cc15976c
Scheme 1 Illustration of the generation of Co3O4 NTs based on
modified Kirkendall effect.
a facile approach based on modified Kirkendall effect for the
generation of Co3O4 NTs with high yield, namely, a liquid
phase reaction between CoC2O4 NRs and NaOH (Scheme 1).
The process is operated under mild conditions without the need
for pressurized vessel and specific template. The approach
provides an alternative to synthesize other inorganic tubular
nanostructures. It was found that the as-synthesized Co3O4
NTs displayed superior catalytic activity and durability in CH4
combustion.
The details about sample synthesis, characterization and
reaction are provided in the ESI.w
It was found that the concentration of oxalic acid and
sodium hydroxide is the key factor for the production of final
Co3O4 NTs. Fig. S1 (ESIw) shows the impact of concentration
of oxalic acid solution on the morphology of CoC2O4 NRs.
The oxalic acid solution of lower concentration (0.01 M)
results in long and thin (not well developed) CoC2O4 NRs
(Fig. S1a, ESIw); while the oxalic acid solution of higher
concentration (0.5 M) brings about thick, stacked, and non-
uniform CoC2O4 NRs (Fig. S1c, ESIw). These CoC2O4 NRs
are not good templates for producing Co3O4 NTs through
Kirkendall effect. On the other hand, the oxalic acid solution
of medium concentration (0.1 M) leads to the well developed
yet uniform CoC2O4 NRs (Fig. S1b, ESIw) which are suitable
templates for generating uniform Co3O4 NTs via Kirkendall
effect. Fig. S2 (ESIw) demonstrates the effect of NaOH
concentration on the morphology of Co(OH)2 NTs: the
NaOH solutions of 0.25 and 0.5 M produce non-uniform
(Fig. S2b, ESIw) or highly collapsed Co(OH)2 NTs (Fig. S2c,
ESIw) while a NaOH solution of 0.1 M generates the most

This journal is c The Royal Society of Chemistry 2012 Chem. Commun., 2012, 48, 853–855 853

uniform Co(OH)2 NTs (Fig. S2a, ESIw) which are easily
transformed into uniform Co3O4 NTs upon calcination.
The BET surface areas of the Co3O4 NTs, NRs, and NPs are
in the range of 36–48 m2 g 1 (Table S1, ESIw). The pore size of
Co3O4 NTs increases with rise in calcination temperature for the
generation of NTs (Fig. S3, ESIw). Such a phenomenon can be
explained based on the results of TEM observation (see later).
Fig. S4 (ESIw) shows the XRD patterns of the as-prepared
Co3O4 NTs and NRs as well as the related precursors. The
CoC2O4
2H2O phase (JCPDS No. 25-0250) was generated
Published on 02 December 2011. Downloaded by University of York on 16/07/2013 11:02:22.
after reacting CoCl2 with oxalic acid. After reacting CoC2O4
with NaOH in an ethanol solution, there was the production of
b-Co(OH)2 phase (JCPDS No. 30-0443). Through calcination
of b-Co(OH)2 or CoC2O4
2H2O species, a phase of Co3O4 was
generated (JCPDS No. 74-1657).
The TGA measurement of CoC2O4
2H2O and b-Co(OH)2
precursors was conducted (Fig. S5, ESIw). The precursor of
CoC2O4
2H2O showed weight losses in the 100–200 1C and
270–300 1C ranges. The first weight loss of 19.6% is due to
dehydration of crystal water, in good agreement with the
expected theoretical value (19.7%). The second weight loss
of 36.5% is due to the decomposition of anhydrous cobalt
oxalate to Co3O4 species (corresponding to a calculated weight
loss of 36.6%). The b-Co(OH)2 precursor only shows a weight loss
of 14.1% in the 70–350 1C range, suggesting the transformation
of b-Co(OH)2 to Co3O4 upon thermal treatment at 350 1C
(theoretical weight loss being 14.0%).
The TEM images of CoC2O4
2H2O (Fig. 1a) display a
morphology of NRs, with diameters in the range of 100–200 nm
and lengths of several microns. When suspended in ethanol
solution in the presence of NaOH, the CoC2O4
2H2O NRs were
converted to b-Co(OH)2 NTs, and the thickness of the tube walls
was found to be ca. 30–40 nm. One can see that the wall of
Co3O4NTs (350 1C) is made up of fine Co3O4 NPs of ca. 10 nm
(Fig. 1b) whereas the wall of Co3O4 NPs (550 1C) is made up
of Co3O4 NPs of bigger size (Fig. 1c). Nonetheless, the Co3O4
NTs obtained after calcination at 350 and 550 1C are similar in
diameter and length, suggesting insignificant sintering effect on
Fig. 1 TEM and HRTEM images of (a) b-Co(OH)2 NTs, (b) Co3O4NTs
(350 1C), (c) Co3O4NTs (550 1C), and (d) Co3O4 NRs.
854 Chem. Commun., 2012, 48, 853–855
View Article Online
the Co3O4 NTs even calcined at 550 1C. However, the building
blocks of Co3O4 NTs, namely, the Co3O4 NPs in the wall of
Co3O4 NTs, seem to grow up somehow after calcination at
550 1C. Such increase in particle size of Co3O4 building blocks
brings about an increase in pore size distribution as shown in
Fig. S3 (ESIw). In both cases, the HRTEM images (insets of
Fig. 1b and c) suggest that the major exposed crystal plane of
the Co3O4 NPs is of (112) orientation, and is normal to the
(220) planes that have a lattice space of 0.286 nm.
The major exposed crystal plane of Co3O4 NRs is of (110)
orientation, and is normal to the (311) planes that have a
lattice spacing of 0.242 nm. In these three cases, (111) planes
with lattice spacing of 0.467 nm can be detected (insets of
Fig. 1b–d). The SEM images of Fig. S6 (ESIw) indicate the
transformation of CoC2O4
2H2O NRs to Co3O4 NTs upon
reaction with NaOH and calcination in air at 350 1C (Fig. S6a
and b, ESIw). After calcination at 550 1C, there is growth of
the building blocks of Co3O4 NTs (Fig. 1c). In the present
study, the strategy of morphology directed synthesis of uniform
Co3O4 NTs needs the uniform CoC2O4 NRs as an appropriate
template in the NaOH-containing medium. By tuning the concen-
tration of oxalic ions, uniform and well developed CoC2O4 NRs
can be obtained (Fig. S6a, ESIw), which lie on a solid base to
produce Co(OH)2 NTs through the Kirkendall mechanism and
eventually Co3O4 NTs upon following calcination. In these
consecutive procedures, the original morphology of CoC2O4
NRs can be essentially retained in the final Co3O4 NTs. In other
words, there is morphology-directed effect that originates from
the first precursor CoC2O4
2H2O NRs and can last to the final
Co3O4 NTs product. Fig. S6d (ESIw) also shows the SEM images
of Co3O4 NPs obtained through the route of urea combustion.
One can see that the NPs are uniform in size (ca. 10 nm).
Based on the reaction [CoC2O4 + 2OH - Co(OH)2 +
C2O42 ], the product Co(OH)2 can be obtained. In this reaction,
the Co2+ (in CoC2O4) and OH ions comprise a diffusion pair.
The coupled reaction/diffusion at the solid–liquid interface could
lead to quick formation of an inter-connected Co(OH)2 shell over
the external surface of CoC2O4 NRs (Fig. S7a and b, ESIw). The
outer shell of Co(OH)2 prevents direct reaction between CoC2O4
and OH species, and further reaction relies on the diffusion rates
of Co2+ and OH ions through the Co(OH)2 shell. Since the Co2+
ions diffuse faster than OH ones, a net outward flow of Co2+ ions
through the Co(OH)2 shell results in the creation of inner lattice
vacancies. These vacancies would condense to form the Kirkendall
voids, an evidence of the consequence of Kirkendall effect
(Fig. S7c, ESIw). The voids can be further developed to
produce the eventual Co(OH)2 NTs (Fig. S7d, ESIw).
H2-TPR investigation on the NTs, NRs, and NPs of Co3O4
reveals the low and high-temperature peaks (Fig. S8, ESIw)
which can be ascribed to the reduction of Co3+ to Co2+ and
Co2+ to Co0, respectively.13 The temperature for the reduction
of Co3O4NTs (550 1C) shifts slightly toward the higher end.
This trend is more obvious over Co3O4NRs (350 1C) and
Co3O4 NPs. The reducibility of oxide is associated with
the formation of oxygen vacancies:14 the more susceptibly
reducible an oxide is, the easier it can generate oxygen
vacancies. Therefore, one can deduce that as compared to
Co3O4 NRs and NPs, Co3O4 NTs have more oxygen vacancies
on the surface.
This journal is c The Royal Society of Chemistry 2012

Published on 02 December 2011. Downloaded by University of York on 16/07/2013 11:02:22.
Fig. 2 Temperature dependence of CH4 conversion over the catalysts.
Generally, the oxygen adspecies can undergo successive
transformation with electron gain: O2 (ad) - O2 (ad) -
O (ad) - O2 (ad). The molecularly adsorbed oxygen O2
(ad) and the oxygen adspecies of O2 (ad) and O (ad) are
weakly bonded to substrate surface, while the O2 (ad)species
is usually thought to be a kind of surface lattice oxygen, O2
(lattice), and is difficult to remove. Fig. S9 (ESIw) indicates
that there are surface O2 (ad) and O (ad) species desorbing
below 300 1C; while surface lattice oxygen species desorb
beyond 350 1C. A clear trend in desorption temperature of
the oxygen adspecies can be observed over the Co3O4 nano-
structures: Co3O4NTs (350 1C) E Co3O4NTs (550 1C) o
Co3O4 NRs o Co3O4 NPs. It is well known that the surface
oxygen species play an important role in catalytic oxidation
reactions.15,16 Based on the H2-TPR and O2-TPD results, one
can expect that there are more oxygen vacancies as well as
more reactive oxygen species on Co3O4 NTs.
In gas turbine industry, a detection limit of 5% lower
explosive limit for methane has been implemented for safety
purposes. While catalytic combustion is the method of choice
for such regulation, it is necessary to find efficient catalyst for
methane complete oxidation at relatively low concentration
and temperature.17 The data of Fig. 2 suggest that on the basis
of per unit mass of catalyst, the Co3O4 NTs (350 1C) are
notably more active than Co3O4 NRs and NPs, and slightly
better than Co3O4 NTs (550 1C). 100% CH4 conversion is
achieved at 325 1C over Co3O4 NTs (350 1C) while conversion
over Co3O4 NPs at 400 1C is only 55%. On the basis of per
unit area of catalyst, the Co3O4 NTs (350 1C) are comparable
to the Co3O4 NTs (550 1C), but still obviously more active
than Co3O4 NRs and NPs (Table S1, ESIw). The presence of
adsorbed and lattice oxygen of higher reactivity over Co3O4
NTs (as revealed by H2-TPR and O2-TPD studies) is responsible
in part for the supremacy of catalytic activity. The catalyst
stability of Co3O4 NTs (350 1C) and Co3O4 NTs (550 1C) at T90
was tested over 30 h, and the results suggest that there is no
activity loss in both catalysts (Fig. S10, ESIw). Moreover, the
TEM images of the used samples (Fig. S11, ESIw) do not show
apparent sintering effect on the Co3O4 NTs in both cases after
reaction.
In summary, we reported for the first time the fabrication
of Co3O4 NTs through an interfacial reaction of NaOH with
This journal is c The Royal Society of Chemistry 2012
View Article Online
pre-synthesized CoC2O4
2H2O NRs making use of the
Kirkendall effect. The adopted method may be applied for the
synthesis of hollow nanostructures of other inorganic compounds.
In this approach, the morphology of the original precursor has a
‘‘directing effect’’ on the morphology of the final product. Among
the Co3O4 nanostructures obtained, Co3O4 NTs showed excellent
activity and good stability towards the catalytic combustion of
CH4. The superior catalytic activity can be attributed to: (i) the
specifically and dominantly exposed (112) crystal plane; and (ii)
high reactivity of oxygen adspecies as well as surface lattice oxygen
on Co3O4 NTs.
Notes and references
1 X. Pan, Z. Fan, W. Chen, Y. Ding, H. Lou and X. Bao, Nat.
Mater., 2007, 6, 507.
2 Y. M. Sui, W. Y. Fu, Y. Zeng, H. B. Yang, Y. Y. Zhang, H. Chen,
Y. X. Li, M. H. Li and G. T. Zou, Angew. Chem., Int. Ed., 2010,
49, 4378.
3 C. W. Guo, Y. Cao, S. H. Xie, W. L. Dai and K. N. Fan, Chem.
Commun., 2003, 700.
4 H. L. Cao, X. F. Qian, C. Wang, X. D. Ma, J. Yin and Z. K. Zhu,
J. Am. Chem. Soc., 2005, 127, 16024.
5 K. An, S. G. Kwon, M. Park, H. B. Na, S. Baik, J. H. Yu, D. Kim,
J. S. Son, Y. W. Kim, I. C. Song, W. K. Moon, H. M. Park and
T. Hyeon, Nano Lett., 2008, 8, 4252.
6 J. B. Fei, Y. Cui, X. H. Yan, W. Qi, Y. Yang, K. W. Wang, Q. He
and J. B. Li, Adv. Mater., 2008, 20, 452; L. Z. Wang, F. Q. Tang,
K. Ozawa, Z. G. Chen, A. Mukher, Y. C. Zhu, J. Zou,
H. M. Cheng and G. Q. Lu, Angew. Chem., 2009, 121, 7182
(Angew. Chem., Int. Ed., 2009, 48, 7048).
7 H. Pang, Q. Y. Lu, Y. C. Li and F. Gao, Chem. Commun., 2009,
7542.
8 M. Ibanez, J. D. Fan, W. H. Li, D. Cadavid, R. Nafria, A. Carrete
and A. Cabot, Chem. Mater., 2011, 23, 3095; L. Wang, B. Peng,
X. F. Guo, W. P. Ding and Y. Chen, Chem. Commun., 2009, 1565;
H. J. Fan, M. Knez, R. Scholz, K. Nielsch, E. Pippel, D. Hesse,
M. Zacharias and U. Gosele, Nat. Mater., 2006, 5, 627;
J. G. Railsback, A. C. Johnston-Peck, J. W. Wang and
J. B. Tracy, ACS Nano., 2011, 4, 1913; J. Zhou, J. Liu,
X. D. Wang, J. H. Song, R. Tummala, N. S. Xu and
Z. L. Wang, Small, 2007, 3, 622.
9 H. J. Fan, U. Gosele and M. Zacharias, Small, 2007, 3, 1660;
J. J. Miao, L. P. Jiang, C. Liu, J. M. Zhu and J. J. Zhu, Inorg.
Chem., 2007, 46, 5673; T. Huang and L. Qi, Nanotechnology, 2009,
20, 025606; H. Colfen and S. Mann, Angew. Chem., Int. Ed., 2003,
42, 2350; Y. Yin, C. K. Erdonmez, A. Cabot, S. Hughes and
A. P. Alivisatos, Adv. Funct. Mater., 2006, 16, 1389.
10 J. B. Tracy, D. N. Weiss, D. P. Dinega and M. G. Bawendi, Phys.
Rev. B, 2005, 72, 064404.
11 E. Hosono, S. Fujihara, I. Honma and H. Zhou, J. Mater. Chem.,
2005, 15, 1938; P. A. George, S. Elias and D. D. Dionysios,
J. Phys. Chem. B, 2005, 109, 13052; X. S. Shi, S. Han,
R. J. Sanedrin, C. Galvez, F. Zhou and M. Selke, Nano Lett.,
2002, 2, 289.
12 W. Y. Li, L. N. Xu and J. Chen, Adv. Funct. Mater., 2005, 15, 851.
13 C. B. Wang, C. W. Tang, S. J. Gau and S. H. Chien, Catal. Lett.,
2005, 101, 59; P. Arnoldy and J. A. Moulijn, J. Catal., 1985, 93, 38;
M. Vob, D. Borgmann and G. Wedler, J. Catal., 2002, 212, 10.
14 P. Li, C. He, J. Cheng, C. Y. Ma, B. J. Dou and Z. P. Hao, Appl.
Catal., B, 2011, 101, 570.
15 B. Z. Tian, X. Liu, L. Solovyov, Z. Liu, H. Zang, S. Xie, F. Zhang,
B. Tu, C. Yu, O. Terasaki and D. Zhao, J. Am. Chem. Soc., 2004,
126, 865.
16 Q. Liu, L. C. Wang, M. Chen, Y. Cao, H. Y. He and K. N. Fan,
J. Catal., 2009, 263, 104.
17 G. Laugel, J. Arichi, M. Molie, A. Kiennemann, F. Garin and
B. Louis, Catal. Today, 2008, 138, 38.
Chem. Commun., 2012, 48, 853–855 855