Nanoscale - Nanyang Technological University

Nanoscale
View Article Online
Published on 08 January 2013. Downloaded by Nanyang Technological University on 25/03/2015 05:07:10.
COMMUNICATION
Cite this: Nanoscale, 2013, 5, 1503
Received 17th November 2012
Accepted 5th January 2013
View Journal | View Issue
High performance carbon nanotube–Si core–shell wires
with a rationally structured core for lithium ion battery
anodes†
Yu Fan, Qing Zhang,* Congxiang Lu, Qizhen Xiao, Xinghui Wang and Beng kang Tay
DOI: 10.1039/c3nr33683b
www.rsc.org/nanoscale
Core–shell Si nanowires are very promising anode materials. Here, we
synthesize vertically aligned carbon nanotubes (CNTs) with relatively
large diameters and large inter-wire spacing as core wires and
demonstrate a CNT–Si core–shell wire composite as a lithium ion
battery (LIB) anode. Owing to the rationally engineered core structure, the composite shows good capacity retention and rate performance. The excellent performance is superior to most core–shell
nanowires previously reported.
To meet the energy demands of future portable electronics and
hybrid electric vehicles (HEVs), advanced electrode materials
with high capacities are desired to produce LIBs with not only
high energy density but also high power capability.1 Silicon is a
very promising candidate for anodes because it exhibits an
extremely high theoretical capacity of 4200 mA h g1 and some
other advantages such as abundance and non-toxicity.2,3
Unfortunately, traditional Si anodes based on bulk or micronsized Si exhibit very poor capacity retention for several reasons
such as huge volume change, unstable solid electrolyte interface
(SEI) and poor conductivity. Among them, the signicant
volume change induced fracture and pulverization have been
identied as the major cause of the poor cyclability. In recent
years, nanostructuring has proved to be an effective approach to
alleviate the volume change problem because nanostructures
could more efficiently release the internal stress, in comparison
with their bulk or micron-sized counterparts.
Among the various Si nanostructures/composites such as
nanotubes,4–6 nanospheres,7–10 and nanowires,11 etc.,12–15 core–
shell nanowires rooted on current collectors16–20 have attracted
much attention because their unique structure offers several
distinct benets such as enhanced energy density due to elimination of binders and effective charge transport owing to the
NOVITAS, Nanoelectronics Centre of Excellence, School of Electrical and Electronic
Engineering, Nanyang Technological University, Singapore 639798. E-mail:
eqzhang@ntu.edu.sg
† Electronic supplementary
10.1039/c3nr33683b
information
(ESI)
available.
This journal is ª The Royal Society of Chemistry 2013
See
DOI:
robust core nanowires. Actually, the essential benet of such a
structure is that the core nanowires greatly increase the virtual
surface area of the current collector such that a very thick Si lm
on a conventional at current collector could be accommodated
in such a structure with a very thin Si coating.
However, it should also be noted that a Si thin lm deposited
on conventional at current collectors shows much better electrochemical performance than a core–shell composite with the
same Si coating thickness. For example, Takamura et al.
demonstrated a 50 nm Si thin lm on a at nickel foil current
collector that exhibited a high capacity exceeding 2000 mA h g1
and superior cyclability of over 1000 cycles at a rate of 12 C.21 On
the other hand, none of the core–shell nanowire composites
demonstrated until now could achieve a comparable performance even if their Si coating thickness is smaller. Therefore, it is
reasonable to conclude that structures of the core nanowires
(alignment, inter-wire distance, wire diameter, etc.) also play very
important roles in such core–shell composites. For instance, a Si
thin lm on a at current collector is subjected to stresses in both
in-plane and perpendicular-to-plane directions, owing to the
expansion/contraction of the Si lm in the two corresponding
directions. In core–shell nanowires, the two stresses correspond
to radial and axial stresses. At the same time, curvature of the core
nanowires introduces another stress, hoop stress, which is
conrmed by recent calculations and in situ transmission electron microscopy (TEM) investigations.19,22,23 Wang et al. even
observed crack propagation in a Si shell driven by the hoop
stress.23 In this regard, core nanowires with large diameters would
be advantageous because of their reduced curvature effect.
Another structure parameter that denitely plays a role is the
inter-wire spacing. A Si thin lm has sufficient space above the
surface to accommodate the volume change perpendicular to the
plane and for access to the electrolyte. On the other hand, core–
shell nanowires with random orientation would inevitably suffer
from insufficient space at some locations such as the bottom.
Therefore, we believe that vertically aligned core–shell nanowires
with a relatively large inter-wire spacing would be advantageous
in both volume change accommodation and electrolyte access.
Nanoscale, 2013, 5, 1503–1506 | 1503
View Article Online
Published on 08 January 2013. Downloaded by Nanyang Technological University on 25/03/2015 05:07:10.
Nanoscale
Unfortunately, although many works on core–shell nanowire
composites have been demonstrated, most were dedicated to
synthesis of new core nanowire materials and very few focused
on engineering core nanowire structures. This is not surprising
because in most of the works, the core nanowires were fabricated through thermal chemical vapour deposition (CVD)
processes that usually produce randomly aligned thin nanowires and have almost no control over nanowire structures
other than the length.16–19 Here, for our proof of concept, we
synthesize vertically aligned carbon nanotubes (CNTs) with
large diameters and large inter-wire spacing as the core wires
and investigate the electrochemical performance of the CNT–Si
core–shell wire arrays as LIB anodes.
The CNTs are synthesized through a plasma enhanced CVD
(PECVD) process with a Ni thin lm deposited on stainless steel
(SS) current collectors as the catalyst. In this process, diameter
and inter-wire spacing of the synthesized CNTs could be
controlled through the thickness of the Ni catalyst lm. The
produced CNTs are also straight and vertically aligned due to
the existence of a strong electric eld. Fig. 1(a) shows a tilted
angle scanning electron microscopy (SEM) image of the fabricated CNT core array. It can be seen that the CNTs are vertically
aligned with a relatively large inter-wire spacing. Although some
CNTs with small diameters can be observed, most CNTs exhibit
diameters of 150 nm, larger than that of most core nanowires
previously reported.16–19 The structure of the CNT array presented here is thus distinctly different from that of previously
reported core nanowires. Finally, a smooth Si shell is coated on
the CNT core wires in a PECVD system to realize a CNT–Si core–
shell wire array. Fig. 1(b) shows a top view SEM image of the
CNT–Si wire array. Aer Si coating, the inter-wire spacing is still
quite large, which is benecial for volume change accommodation as well as efficient access of the electrolyte to the bottom
of the structure. Fig. 1(c) displays a TEM image of a randomly
selected CNT–Si wire, showing a rather smooth Si coating
of 60 nm thick. From the TEM image, the diameter ratio of the
CNT–Si wire to the CNT core is 2 : 1. The mass ratio of Si to
CNT is calculated to be 2 : 1 (Si : CNT) (refer to the ESI†).
Raman characterization (Fig. S1†) of the composite conrms
that the Si coating is mainly amorphous.
To evaluate the electrochemical performance of the CNT–Si
composite, button type half-cells are assembled with the CNT–
Si wire array directly rooted on SS as the working electrode and
lithium foil as the counter electrode. Fig. 2(a) presents the
capacity retention performance of the composite over the
potential range of 0.01–1.2 V at 0.2 C (1 C ¼ 4200 mA g1). In
this demonstration, only the mass of Si is considered to reveal
quality of the structure. Although the CNTs are also active
against Li+, their contribution is rather small, as demonstrated
in the ESI (Fig. S2†). Fig. 2(a) shows that in the rst several
cycles, charge–discharge capacities of the anode gradually
increase. The phenomenon can be attributed to an activation
process involving reconstruction and activation of Si and electrode/electrolyte interface, as supported by the cyclic voltammogram (CV) curves shown in Fig. S3.† In the 7th cycle, the
anode achieves the maximum de-lithiation capacity of 2755 mA
h g1, aer which the capacity remains quite stable. Fig. 2(b)
1504 | Nanoscale, 2013, 5, 1503–1506
Communication
Fig. 1 (a) Tilted angle SEM image of the synthesized CNT array, (b) top view
SEM image of the CNT–Si core–shell composite and, (c) TEM image of a single
CNT–Si wire.
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Published on 08 January 2013. Downloaded by Nanyang Technological University on 25/03/2015 05:07:10.
Communication
Fig. 2 (a) Variation of charge–discharge capacities of the CNT–Si anode versus
cycle number and (b) typical charge–discharge voltage profiles of the CNT–Si
anode.
displays typical voltage proles of the anode in a full charge–
discharge cycle (aer the 7th cycle), which shows the typical
sloping behaviour of Si anodes. At the end of the 80th cycle, a delithiation capacity of 2510 mA h g1 is still retained, corresponding to a capacity retention of 91% in comparison with its
maximum value. In addition, the anode maintains a very high
Coulombic efficiency of over 99% for all cycles (except for the
rst 3 activation cycles), indicating a very good electrode
structural stability.
To further conrm the structural stability of the CNT–Si
composite, a half-cell is disassembled in the fully de-lithiated
state aer 50 galvanostatic cycles for inspection. As shown in
Fig. 3, the electrode still maintains its original structure and no
wire detachment is found. Fig. S4† shows a TEM image of a
single CNT–Si wire aer cycling. Compared to the pristine wire
shown in Fig. 1(c), the Si shell appears brighter aer cycling,
indicating an increased porosity, which is as expected. In
addition, the Si shell surface is covered by a very rough lm,
which is most likely the SEI. It is well known that the SEI formed
on the pristine Si surface is not stable and de-lithiation induced
signicant volume contraction would break the SEI formed
during the previous lithiation process. However, it is apparent
that the Si shell remains adhered to the CNT core strongly,
without noticeable pulverization. Because of this, the composite
anode shows quite a good cycling performance.
Several factors could account for the good stability. First, the
thick CNTs are robust enough to be free of damage caused by
the stress at the CNT–Si interface. In contrast, a very thin
metallic backbone is found to get bent or distorted during
lithiation/de-lithiation.11,24 Second, the large CNT diameter
reduces the hoop stress in the Si shell, resulting in less stress
accumulation. Third, the relatively large inter-wire spacing in
the vertically aligned CNT–Si array can efficiently accommodate
the lithiation/de-lithiation induced volume changes. For
comparison, we also fabricated CNTs with smaller diameters
This journal is ª The Royal Society of Chemistry 2013
Nanoscale
Fig. 3 SEM image of the CNT–Si anode in the de-lithiated state after 50
galvanostatic cycles at 0.2 C.
and smaller inter-wire spacing as the core by using a thinner Ni
catalyst lm. Fig. S5a and b† present a SEM image of the
fabricated CNT array and a TEM image of a single CNT–Si wire,
respectively. The diameter ratio of the CNT–Si wire to the CNT
core for this control sample is also 2 : 1. Fig. S6† shows cycling
performance of the thin CNT–Si. Aer 50 cycles, 91% of the
maximum de-lithiation capacity is retained. On the other hand,
the retention rate of the thick CNT–Si is 96.4%, clearly
evidencing the superiority of employing core wires with large
diameters and large inter-wire spacing. In addition, in situ TEM
investigation on lithiation/de-lithiation of a single carbon ber–
Si core–shell wire showed that the adhesion between coated Si
and carbon core is very strong.22 Therefore, the Si coating could
strongly adhere to the CNT core even if cracks have developed in
the Si coating. We notice that previously vertically aligned Cu
rods with a relatively large diameter have been fabricated
through electro-deposition with anodic alumina membranes
(AAO) as the template and investigated as core wires,16 but
capacity retention of such Cu–Si core–shell composites is not so
good as our CNT–Si composite. The inferior adhesion between
Cu and Si, compared with that of carbon and Si, is probably an
important reason.
Actually, the cycling performance of our CNT–Si composite
could be further improved through either lithiation voltage
control or surface passivating coating. It has been demonstrated in many works that by setting a relatively high lithiation
voltage of above 50 mV to limit the amount of Li+ intercalated
and therefore, suppress the volume changes, signicantly
improved cycling performance could be achieved.19,20 But in that
case, the capacity of Si would be compromised a lot. Another
effective approach is to apply a surface coating such as carbon
to passivate the unstable SEI on the Si surface.17,25,26 The method
has been intensively investigated and therefore is not demonstrated here. Even though, owing to the good core structure, the
CNT–Si composite in this work still shows a very good cycling
performance that is better than most of core–shell composites
including those with surface coatings demonstrated before.
Nanoscale, 2013, 5, 1503–1506 | 1505
View Article Online
Nanoscale
Communication
Published on 08 January 2013. Downloaded by Nanyang Technological University on 25/03/2015 05:07:10.
Notes and references
Fig. 4
Rate performance of the CNT–Si anode.
Another output parameter that is of signicance, especially
for future LIBs in HEV applications, is rate performance. Fig. 4
presents the rate performance of the CNT–Si wire composite. In
the rst 10 cycles, a relatively low charge–discharge rate of 0.2 C
is applied to activate the materials. At the 10th cycle, a de-lithiation capacity of 2670 mA h g1 is retained, consistent with
the value in Fig. 2(a). Aer that, the rate is gradually increased to
0.5, 1, 2, 4, 8 and 12 C, and the de-lithiation capacities at
the corresponding rates are 2280, 2010, 1730, 1480, 1170 and
985 mA h g1, respectively. Finally, when the current is reverted
to 0.2 C, the de-lithiation capacity recovers to over 2500 mA h
g1. The remarkable rate performance suggests that the CNT–Si
composite is able to facilitate fast charge transport and maintain structural stability at very high rates.
The rate performance demonstrated here is superior to most
core–shell nanowires previously reported. We believe that the
superiority could be attributed to the relatively large inter-wire
spacing in our core–shell structure. Consequently, the whole Si
coating including that at the bottom of the array can gain
sufficient access to the electrolyte, effectively alleviating the
concentration polarization problem at high charge–discharge
rates. In addition, the previous in situ TEM investigation
showed that due to the fast diffusion of Li+ in carbonaceous
materials, lithiation in such carbon–Si core–shell nanowires
occurs from both the Si shell surface and the carbon–Si interface, which eventually depletes the Si in the middle of the
shell.23 In other words, the carbon–Si interface could serve as a
fast Li+ transport channel and could be another reason of the
good rate performance of our CNT–Si anode. This feature
indicates that a carbonaceous core is perhaps more advantageous in fast charge transport than electrochemically inert core
materials such as Cu.
In conclusion, we have demonstrated a CNT–Si core–shell
wire composite that employs vertically aligned CNTs with a
large diameter and large inter-wire spacing as the core to alleviate the shortcomings of previously reported core–shell nanowires. Owing to the rationally engineered core structure, the
CNT–Si wire demonstrates good capacity retention and rate
performance, much superior to most core–shell nanowires in
previous reports, strongly verifying the important role the core
structure plays in such core–shell composites.
Acknowledgements
This work is supported by MOE AcRF Tier2 Funding, Singapore
(MOE2011-T2-1-137).
1506 | Nanoscale, 2013, 5, 1503–1506
1 J. M. Tarascon and M. Armand, Nature, 2001, 414, 359–367.
2 U. Kasavajjula, C. Wang and A. J. Appleby, J. Power Sources,
2007, 163, 1003–1039.
3 D. Larcher, S. Beattie, M. Morcrette, K. Edstrom, J.-C. Jumas
and J.-M. Tarascon, J. Mater. Chem., 2007, 17, 3759–3772.
4 T. Song, J. Xia, J.-H. Lee, D. H. Lee, M.-S. Kwon, J.-M. Choi,
J. Wu, S. K. Doo, H. Chang, W. I. Park, D. S. Zang, H. Kim,
Y. Huang, K.-C. Hwang, J. A. Rogers and U. Paik, Nano
Lett., 2010, 10, 1710–1716.
5 M.-H. Park, M. G. Kim, J. Joo, K. Kim, J. Kim, S. Ahn, Y. Cui
and J. Cho, Nano Lett., 2009, 9, 3844–3847.
6 C. Zhao, Q. Li, W. Wan, J. Li, J. Li, H. Zhou and D. Xu,
J. Mater. Chem., 2012, 22, 12193–12197.
7 L. Su, Z. Zhou and M. Ren, Chem. Commun., 2010, 46, 2590–
2592.
8 A. Esmanski and G. A. Ozin, Adv. Funct. Mater., 2009, 19,
1999–2010.
9 J. Liu, W. Li and A. Manthiram, Chem. Commun., 2010, 46,
1437–1439.
10 L. Su, Y. Jing and Z. Zhou, Nanoscale, 2011, 3, 3967–3983.
11 C. K. Chan, H. Peng, G. Liu, K. McIlwrath, X. F. Zhang,
R. A. Huggins and Y. Cui, Nat. Nanotechnol., 2008, 3, 31–35.
12 Y. Fan, K. Huang, Q. Zhang, Q. Xiao, X. Wang and X. Chen,
J. Mater. Chem., 2012, 22, 20870–20873.
13 H. Kim, B. Han, J. Choo and J. Cho, Angew. Chem., Int. Ed.,
2008, 47, 10151–10154.
14 L. Ji and X. Zhang, Energy Environ. Sci., 2010, 3, 124–129.
15 X. Li, P. Meduri, X. Chen, W. Qi, M. H. Engelhard, W. Xu,
F. Ding, J. Xiao, W. Wang, C. Wang, J.-G. Zhang and J. Liu,
J. Mater. Chem., 2012, 22, 11014–11017.
16 H. Guan, X. Wang, S. Chen, Y. Bando and D. Golberg, Chem.
Commun., 2011, 47, 12098–12100.
17 F.-F. Cao, J.-W. Deng, S. Xin, H.-X. Ji, O. G. Schmidt, L.-J. Wan
and Y.-G. Guo, Adv. Mater., 2011, 23, 4415–4420.
18 X. Chen, K. Gerasopoulos, J. Guo, A. Brown, C. Wang,
R. Ghodssi and J. N. Culver, ACS Nano, 2010, 4, 5366–5372.
19 L. Hu, H. Wu, Y. Gao, A. Cao, H. Li, J. McDough, X. Xie,
M. Zhou and Y. Cui, Adv. Energy Mater., 2011, 1, 523–527.
20 H. T. Nguyen, M. R. Zamr, L. D. Duong, Y. H. Lee,
P. Bondavalli and D. Pribat, J. Mater. Chem., 2012, 22,
24618–24626.
21 T. Takamura, S. Ohara, M. Uehara, J. Suzuki and K. Sekine,
J. Power Sources, 2004, 129, 96–100.
22 C.-M. Wang, X. Li, Z. Wang, W. Xu, J. Liu, F. Gao, L. Kovarik,
J.-G. Zhang, J. Howe, D. J. Burton, Z. Liu, X. Xiao,
S. Thevuthasan and D. R. Baer, Nano Lett., 2012, 12, 1624–1632.
23 J. W. Wang, X. H. Liu, K. Zhao, A. Palmer, E. Patten,
D. Burton, S. X. Mao, Z. Suo and J. Y. Huang, ACS Nano,
2012, 6, 9158–9167.
24 M. T. McDowell, S. Woo Lee, C. Wang and Y. Cui, Nano
Energy, 2012, 1, 401–410.
25 H. Jia, P. Gao, J. Yang, J. Wang, Y. Nuli and Z. Yang, Adv.
Energy Mater., 2011, 1, 1036–1039.
26 Y. Yu, L. Gu, C. Zhu, S. Tsukimoto, P. A. van Aken and
J. Maier, Adv. Mater., 2010, 22, 2247–2250.
This journal is ª The Royal Society of Chemistry 2013