Controllable Pt nanoparticle deposition on carbon nanotubes as an anode catalyst for direct methanol fuel cells

Article abstract of DOI:10.1021/jp0555448

We report a novel process to prepare well-dispersed Pt nanoparticles on CNTs. Pt nanoparticles, which were modified by the organic molecule triphenylphosphine, were deposited on multiwalled carbon nanotubes by the organic molecule, which acts as a cross linker. By manipulating the relative ratio of Pt nanoparticles and multiwalled carbon nanotubes in solution, Pt/CNT composites with different Pt content were achieved. The so-prepared Pt/CNT composite materials show higher electrocatalytic activity and better tolerance to poisoning species in methanol oxidation than the commercial E-TEK catalyst, which can be ascribed to the high dispersion of Pt nanoparticles on the multiwalled carbon nanotube surface.

Full text of DOI:10.1021/jp0555448

22212
J. Phys. Chem. B 2005, 109, 22212-22216
Controllable Pt Nanoparticle Deposition on Carbon Nanotubes as an Anode Catalyst for
Direct Methanol Fuel Cells
†
†
†
‡
Yongyan Mu, Hanpu Liang, Jinsong Hu, Li Jiang, and Lijun Wan*
Institute of Chemistry, Chinese Academy of Sciences (CAS), Beijing 100080, China
ReceiVed: September 29, 2005
We report a novel process to prepare well-dispersed Pt nanoparticles on CNTs. Pt nanoparticles, which were
modified by the organic molecule triphenylphosphine, were deposited on multiwalled carbon nanotubes by
the organic molecule, which acts as a cross linker. By manipulating the relative ratio of Pt nanoparticles and
multiwalled carbon nanotubes in solution, Pt/CNT composites with different Pt content were achieved. The
so-prepared Pt/CNT composite materials show higher electrocatalytic activity and better tolerance to poisoning
species in methanol oxidation than the commercial E-TEK catalyst, which can be ascribed to the high dispersion
of Pt nanoparticles on the multiwalled carbon nanotube surface.
Introduction
immobilization on CNTs includes generating functional groups
on the external walls, mostly by harsh oxidative treatments such
To satisfy mankind’s ever-increasing energy needs and tackle
the daunting environmental issues, one has to consider alterna-
tive energy sources to replace the currently dominant fuels,
petroleum and natural gas. Direct methanol fuel cells (DMFCs),
which provide a new way to store and convey energy, have
as refluxing in HNO3, and then deposition of metals on activated
1
8-23
CNT walls.
Such surface functionalization provides an
avenue for metal precursors to correlate with CNTs and prompts
the deposition of metal on the external walls. While these
strategies are effective, the controllability has not been fully
attracted much attention as green power sources for automobiles
2
4,25
and portable electronics.1-3 In this electrochemical cell, metha-
realized. The other approaches, such as physical evaporation,
2
6,27
28,29
electroless deposition,
and electrodeposition,
were also
nol is directly oxidized with air to carbon dioxide and water to
produce electricity. DMFC can use alkaline electrolyte. How-
ever, the carbonation of carbon dioxide is a serious problem of
DMFCs in an alkaline electrolyte, which decreases the fuel cell
reported recently. Close examination of the existing methods
reveals that they involved either a tedious pretreatment procedure
to modify CNT walls or complex equipment to achieve high
dispersion of Pt nanoparticles on CNTs. Therefore, it is desirable
to develop a simple and effective synthetic route that provides
well-dispersed Pt nanoparticles while maintaining some degree
of control of particle size and size distribution on CNTs.
In this paper, we report a novel process to prepare well-
dispersed Pt nanoparticles on multiwalled CNTs (MWCNTs).
Different from the methods reported, which produce Pt/CNT
composite by modifying CNTs to achieve high Pt dispersion,
the method we adopted was to modify Pt nanoparticles with
organic molecule triphenylphosphine (PPh3). In contrast to
CNTs, the functionalization of Pt nanoparticles was facile and
effective under much more benign conditions, such as no harsh
acid and room temperature. Pt nanoparticles prepared in ethylene
glycol kept their small particle size and narrow size distribution
after being modified with PPh3 and deposited on the surface of
CNTs. Transmission electron microscopy (TEM) and high-
resolution transmission electron microscopy (HRTEM) char-
acterization were carried out to determine the particle size and
distribution of the Pt catalyst. The so-prepared Pt/CNT com-
posite materials showed higher electrocatalytic activity in
methanol oxidation than the commercial catalyst.
efficiency by decreasing the electrolyte conductivity and
increasing the concentration polarization.2,4 Therefore, the
practical DMFCs now operate with acid electrolyte, which are
carbon dioxide rejecting. One of the challenges in developing
DMFCs with acid media is to enhance the exchange current
density due to the low reactivity of methanol oxidation in acid
4
electrolyte. To date, the most promising catalytic materials used
for the methanol oxidation at room temperature are supported
5
-11
Pt catalyst.
It is well-known that the specific activity of
catalysts is strongly related to their size, distribution, and the
support. Highly distributed catalyst nanoparticles with small size
and narrow size distribution are ideal for high electrocatalyst
activity owing to their large surface-to-volume ratio. Among
the possible supports, carbon black has been widely used as an
4-5,12,13
electrode, which disperses Pt nanoparticles,
but carbon
nanotubes (CNTs) are considered to be a more attractive
candidate owing to their outstanding mechanical characteristics
such as high tensile strength coupled with high surface area,
1
4-16
high electric conductivity, and thermal conductivity.
Pt
nanoparticles on the external walls are easier to make contact
with the reactant than those trapped in the pores of the carbon
black electrode. However, the effective attachment of Pt
nanoparticles uniformly dispersed onto CNTs remains a formi-
dable challenge because of the inertness of the CNT walls.
So far, the most established protocol for catalytic metal
Experimental Section
17-21
Nanoparticle Synthesis. Pt nanoparticles were synthesized
3
0
according to Wang’s method, details of which are as follows.
An ethylene glycol solution of H2PtCl6‚6H2O (50 mL, 1.93 mM)
was mixed with an ethylene glycol solution of NaOH (50 mL,
0.5 M) with magnetic stirring. This makes a transparent yellow
solution. Then the solution was refluxed at 160 °C for 3 h with
*
To whom correspondence should be addressed. E-mail:
wanlijun@iccas.ac.cn.
†
‡
Also at the Graduate School of CAS, Beijing 100064, China
Current address: Schlumberger-Doll research, Ridgefield, CT 06877.
1
0.1021/jp0555448 CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/02/2005

Preparation of Well-Dispersed Pt Nanoparticles on CNTs
J. Phys. Chem. B, Vol. 109, No. 47, 2005 22213
the protection of N2. The color of the solution began to change
at around 130 °C, implying that Pt precursors were being
reduced and Pt nanoparticles were synthesized. The obtained
Pt colloidal solution is dark brown and homogeneous.
Modification of Pt Nanoparticles. The obtained Pt colloidal
solution (10 mL) was added to 40 mL of ethanol. Then the
mixed solution was added into a toluene solution of PPh3 (50
mL, 3.81 mM) with stirring. A homogeneous, dark brown
solution was obtained. After 15 min, 50 mL of distilled water
was added to the solution. A liquid in two phases was obtained.
The upper dark-brown solution was PPh3-modified Pt/toluene
solution and the bottom colorless solution was the mixed
solution of water and ethanol. The PPh3-modified Pt/toluene
solution was washed with distilled water three times to remove
the ethylene glycol and transfer the Pt nanoparticles from glycol
solution to toluene, which is essential for the deposition of Pt
nanoparticles on CNT walls.
Deposition of Pt Nanoparticles on CNTs. MWCNTs were
used ultrasonically as received in toluene solution for 4 h to
break big CNT aggregates and disperse CNTs. Then PPh3-
modified Pt/toluene solution was added to the CNT/toluene
solution and the mixture was ultrasonicated for 5 days until all
Pt nanoparticles were deposited on CNTs. The toluene solution
became colorless, suggesting that the PPh3-modified Pt nano-
particles were completely deposited on the surface of CNTs.
The Pt/CNT composite was separated from the solution by
centrifugation and washed by toluene three times.
Figure 1. Schematic illustrations of the synthesis procedures of Pt/
CNT composite.
Characterization. The size of the Pt nanoparticles and
morphology of the Pt/CNT composite were observed by TEM
and HRTEM. TEMs were performed on a JEOL2010 equipped
with an energy-dispersive X-ray analyzer (Phoenix) (200 KV).
HRTEM was carried out on a Philips F30 with an acceleration
voltage of 300 kV. For TEM observation, the samples were
dispersed in ethanol by ultrasonic treatment and dropped on
copper grids with carbon films followed by solvent evaporation
in air at room temperature. X-ray powder diffractions (XRD)
were carried out on a Rigaku D/max-2500, using filtered Cu
KR radiation.
Figure 2. TEM images of Pt/CNT composite with different Pt
content: 40.6 (a) (inset: enlarged image of panel a), 24.0 (b), 19.1
(c), and 3.1 wt % (d).
CNTs surface. At last, it links Pt nanoparticles with CNTs and
leads to the uniform dispersion of Pt nanoparticles on CNT
surfaces.
Electroactivities of the catalyst were measured by cyclic
voltammetry with use of a three-electrode test cell at room
temperature after the Pt/CNT composite was calcined at 400
Pt/CNT composites with different Pt content having distinct
application potential were produced by simply manipulating the
ratio of Pt nanoparticles vs CNTs in solution. The as-prepared
Pt/CNT composite nanomaterials were investigated by TEM.
Figure 2a-d shows representative TEM images of as-prepared
Pt/CNT composite with varying content of Pt, from (a) 40.6,
(b) 24.0, (c) 19.1, to (d) 3.1 wt %, respectively. Figure 2a is a
typical TEM image showing the CNTs covered with a continu-
ous Pt nanoparticle adlayer, which extends the overall length
of CNTs. Owing to no pretreatment, the CNTs are flexible and
as long as several tens of micrometers. The insert in Figure 2a
is a high-magnification TEM image of the Pt/CNT composite
nanomaterials. It can be seen that there exists a well-dispersed
Pt nanoparticle adlayer on the wall of a CNT. The PPh3
molecules effectively isolated adjacent Pt nanoparticles. There-
fore, Pt nanoparticles are distributed on tube walls quite
uniformly and do not aggregate with each other to form larger
clusters. The average distance between Pt nanoparticles is less
than 1 nm, which is consistent with the stabilizer PPh3
molecules.
°
C for 1 h to remove PPh3. A thin film electrode technique was
used to make the measurements. A glassy carbon disk (3 mm
in diameter) held in a Teflon cylinder was used as the working
electrode, on which a thin layer of Nafion-impregnated catalyst
was cast. The Pt loading of both catalysts on electrode was 1
2
mg/cm . A Pt wire served as the counter electrode and a
saturated calomel electrode (SCE) was used as the reference
electrode. The electrolyte for electrochemical measurements was
a solution of 2 M methanol in 1 M H2SO4. The solution was
deaerated with ultrahigh-purity N2 before scanning.
Results and Discussion
In view of the inertness of CNT walls, we prepared Pt
colloidal solution in toluene, which is an excellent solvent to
wet CNTs. As illustrated in Figure 1, the organic molecule PPh3
acts as a cross linker for Pt nanoparticles and CNTs. At the
end of the molecule, the free electron pair of P interacts with
the space orbital of Pt to form a coordination compound. And
at the other end of the molecule, the phenyl rings correlate with
the backbone of CNTs. There are three functions of PPh3. At
first, it causes Pt nanoparticles transfer from glycol solution to
toluene solution. Second, it prevents Pt nanoparticles from
aggregating with each other either in toluene solution or on the
As the content decreases, Pt nanoparticles randomly deposit
on the wall of CNTs and no continuous layer can be found on
the CNT surface as shown in Figure 2, panels b to d. However,
no free Pt nanoparticles can be resolved in these four TEM
images. It is evident that Pt nanoparticles strongly adsorb on

22214 J. Phys. Chem. B, Vol. 109, No. 47, 2005
Mu et al.
Figure 3. (a) TEM image of Pt/CNT composite (24.0 wt %) after thermal treatment at 400 °C for 1 h. (b) EDS from the Pt/CNT composite shown
in panel a. (c) HRTEM image of the Pt/CNT composite shown in panel a.
Figure 4. Size distribution of Pt nanoparticles on CNTs (a) before and (b) after thermal treatment.
the CNT surface. By adjusting the relative ratio of CNTs and
Pt nanoparticles in toluene solution, different coverage of Pt
nanoparticles on the CNT surface can be accomplished, which
is crucial to their industrial applications as catalysts. The above
results have clearly demonstrated the effectiveness of a simple
route for the assembly of Pt nanoparticles on CNTs and that Pt
nanoparticles are loaded on the outside walls of CNTs, not on
the inside walls.
The Pt/CNT composite was investigated by annealing treat-
ment because it is necessary to remove PPh3 molecules
surrounding Pt nanoparticles before the Pt/CNT composite can
be used as a catalyst. Figure 3a is the TEM image of Pt/CNT
composite after thermal treatment at 400 °C for 1 h in ambient
environment. The sample is the same as that shown in Figure
that of Pt nanoparticles in toluene solution. Figure 4a is a
statistical result from more than 200 Pt nanoparticles showing
the distribution of the particle size. After thermal treatment, the
mean particle size shown in Figure 3a is measured to be 2.7 (
1.0 nm and most Pt nanoparticles are smaller than 4 nm. Figure
4b is a statistical result from more than 200 Pt nanoparticles
after heat treatment. From the results in Figure 4a,b, it is clear
that the size distribution of Pt nanoparticles is quite narrow
1
3,31
compared to previous reports.
After thermal treatment, Pt
nanoparticles are still well dispersed, which is critical to their
future application as catalysts. One of the advantages of thermal
treatment is to increase the stability of the composite because
the working temperature of DMFCs is much lower than the
thermal treatment temperature. Figure 3b demonstrates the
energy-dispersive X-ray spectroscopy (EDS) of the Pt/CNT
composite shown in Figure 3a. The existence of elemental Pt
can be clearly seen. Figure 3c is a high-resolution transmission
electron micrograph (HRTEM) of the composite shown in
Figure 3a, displaying a layer-by-layer morphology of the tube
wall of the MWCNT and the crystalline facets of Pt nanopar-
ticles, which demonstrate the highly crystalline character of
metallic Pt particles.
2
b. The PPh3 protection layer on Pt nanoparticles should have
been effectively removed because PPh3 decomposes at about
16 °C according to thermogravimetric analysis (not shown
2
here). With the removal of PPh3, Pt nanoparticles are fused and
aggregate into larger particles. By comparing the size of Pt
nanoparticles in Figures 2b and 3a, it is evident that Pt
nanoparticles in Figure 3a are larger due to the aggregation of
neighboring Pt nanoparticles after heat treatment. The average
size of the Pt nanoparticles on CNTs before thermal treatment
shown in Figure 2b is measured to be 1.9 ( 0.5 nm, close to
The crystalline nature of the Pt nanoparticles was further
confirmed by recording the X-ray power diffraction (XRD)

Preparation of Well-Dispersed Pt Nanoparticles on CNTs
J. Phys. Chem. B, Vol. 109, No. 47, 2005 22215
to the reverse current peak (Ib), If/Ib, is an index of the catalyst
tolerance to the poisoning species, PtdCdO. A higher ratio
indicates more effective removal of the poisoning species on
the catalyst surface. The If/Ib ratio of Pt/CNT composite is 0.88,
higher than that of the E-TEK catalyst (0.74), showing better
catalyst tolerance of the Pt/CNT composite.
Conclusions
In summary, a novel method has been successfully developed
to load Pt nanoparticles on CNTs without pretreating the CNTs.
It is unique in the system that the PPh3 offers strong adhesion
of Pt on CNT surface and sufficient protection for Pt nanopar-
ticles from being aggregated. Pt/CNT composite with different
Pt content was conveniently controlled by manipulating the
relative concentration of CNTs vs Pt nanoparticles. Pt nano-
particles highly dispersed on CNTs kept their initial small sizes
in solution. Even though thermal treatment led to some extent
of aggregation of Pt nanoprticles, the sizes of most Pt nano-
particles are still smaller than 4 nm. The produced Pt/CNT
composite shows higher electrocatalytic activities and catalyst
tolerance that can be applied in DMFCs in comparison with
the E-TEK catalyst. The present technique represents a highly
feasible approach to produce Pt/CNT composite, which should
have great potential in DMFC applications.
Figure 5. XRD pattern of Pt/CNT composite shown in Figure 4a.
Acknowledgment. Financial support from National Natural
Science Foundation of China (Nos. 20520140277, 20121301,
and 20575070), National Key Project on Basic Research (Grant
No. G2000077501), and Chinese Academy of Sciences is
gratefully acknowledged.
Figure 6. Cyclic voltammograms of methanol oxidation on Pt/CNT
composite (solid line) and commercial E-TEK Pt/C catalyst (dash line)
-
1
in 2 M CH
temperature.
3 2 4
OH/1 M H SO electrolyte at 20 mV S at room
References and Notes
(
(
(
1) Olah, G. A. Angew. Chem., Int. Ed. 2005, 44, 2.
2) Winter, M.; Brodd, R. J. Chem. ReV. 2004, 104, 4245.
3) Mcnicol, B. D.; Rand, D. A. J.; Williams, K. R. J. Power Sources
spectrum (Figure 5) of Pt/CNT composite after thermal treatment
shown in Figure 3a. The peak at 2θ ) 26.5° corresponds to the
(
4
002) planes of graphitized CNTs. And the peaks at 2θ ) 39.7°,
1999, 83, 15.
(4) Lamy, C.; L e´ ger, J. M.; Srinivasan, S. Mod. Aspects Electrochem.
6.2°, 67.4°, and 81.2° can be assigned to the (111), (200), (220),
2
001, 34, 73.
and (311) crystalline planes of fcc Pt, respectively, which
indicates that the Pt nanoparticles are composed of pure
crystalline Pt.
(5) Matsumoto, T.; Komatsu, T.; Arai, K.; Yamazaki, T.; Kijima, M.;
Shimizu, H.; Takasawa, Y.; Nakamura, J. Chem. Commun. 2004, 840.
(6) Yu, J. S.; Kanf, S.; Yoon, S. B.; Chai, G. J. Am. Chem. Soc. 2002,
24, 9382.
1
The electrochemical performance of the Pt/CNT composite
(
(
7) Chen, W. X.; Lee, J. Y.; Liu, Z. L. Chem. Commun. 2002, 2588
8) Liang, H. P.; Zhang, H. M.; Hu, J. S.; Guo, Y. G.; Wan, L. J.; Bai,
(24 wt %, shown in Figure 3a) was tested for methanol
oxidation, which is at the heart of DMFC application in the
anodic half-cell reaction. For comparison, the commercial Pt
catalyst (E-TEK, 20 wt %) was also employed. In Figure 6, the
cyclic voltammograms for methanol oxidation at Pt/CNT
composite prepared by our method (solid line) and the com-
mercial catalyst (dash line) are compared. The voltammetric
C. L. Angew. Chem., Int. Ed. 2004, 43, 1540
(9) Anderson, M. L.; Stroud, R. M.; Rolison, D. R. Nano Lett. 2002,
2, 235.
10) Vinodgopal, K.; Haria, M.; Meisel, D.; Kamat, P. Nano Lett. 2004,
, 415.
11) Schmidt, T. J.; Paulus, U. A.; Gasteiger, H. A.; Behm, R. J. J.
Electroanal. Chem. 2001, 508, 41.
(12) Bock, C.; Paquet, C.; Couillard, M.; Botton, G. A.; MacDougall,
B. R. J. Am. Chem. Soc. 2004, 126, 8028.
13) Liu, Z.; Ling, X. Y.; Su, X.; Lee, J. Y. J. Phys. Chem. B 2004,
08, 8234.
14) Ebbesen, T. W.; Lezec, H. J.; Hiura, H.; Bennett, J. W.; Ghaemi,
H. F.; Thio, T. Nature 1996, 382, 54.
(15) Baughman, R. H.; Zakhidov, A. A.; Heer, W. A. Science 2002,
(
4
(
6-8,18
features are in good agreement with literature,
in which
the typical methanol oxidation current peak on Pt catalyst is at
about 0.74 V vs SCE in the forward scan. Figure 6 clearly shows
that the oxidation current observed with Pt/CNT composite is
considerably higher than that of the E-TEK catalyst. This
significant improvement in the catalytic performance can be
attributed to the high level of dispersion of Pt nanoparticles on
CNTs. In the reverse scan, an oxidation peak is observed around
(
1
(
2
97, 787.
(
16) Lee, K. M.; Li, L. C.; Dai, L. M. J. Am. Chem. Soc. 2005, 127,
122.
(17) Rajesh, B.; Thampi, K. R.; Bonard, J. M.; Xanthopoulos, N.;
Mathieu, H. J.; Viswanathan, B. J. Phys. Chem. B 2003, 107, 2701.
18) Guo, D. J.; Li, H. L. J. Electroanal. Chem. 2004, 573, 197.
19) Tang, H.; Chen, J. H.; Huang, Z. P.; Wang, D. Z.; Rn, Z. F.; Nie,
L. H.; Kuang, Y. F.; Yao, S. Z. Carbon 2004, 42, 191.
20) Ang, L. M.; Andy Hor, T. S.; Xu, G. Q.; Tung, C. H.; Zhao, S.;
Wang, J. L. S. Chem. Mater. 1999, 11, 2115.
21) Yu, R. Q.; Chen, L. W.; Liu, Q. P.; Lin, J. Y.; Tan, K. L.; Ng, S.
4
0
.53 V, which is primarily associated with the removal of the
4,13,32,33
residual carbon species formed in the forward scan.
The
(
(
residual carbon species are oxidized according to the following
reaction:
(
+
-
PtOH + PtdCdO f CO + 2Pt + H + e
ad
2
(
C.; Chan, H. S. O.; Xu, G. Q.; Andy Hor, T. S. Chem. Mater. 1998, 10,
718.
Therefore, the ratio of the forward oxidation current peak (If)

22216 J. Phys. Chem. B, Vol. 109, No. 47, 2005
Mu et al.
(
22) Rajalakshmi, N.; Ryu, H.; Shaijumon, M. M.; Ramaprabhu, S. J.
Power Sources 2005, 140, 250.
23) Li, W. Z.; Liang, C. H.; Zhou, W. J.; Qiu, J. S.; Zhou, Z. H.; Sun,
G. Q.; Xin, Q. J. Phys. Chem. B 2003, 107, 6292.
(28) Nelson, T.; Vinodgopal, K.; Kumar, G. G.; Kamat, P. Electrochem.
Soc. Proc. 2004, 12, 152.
(
29) Quinn, B. M.; Dekker, C.; Lemay, S. G. J. Am. Chem. Soc. 2005,
(
1
1
27, 6146.
(30) Wang, Y.; Ren, J.; Deng, K.; Gui, L.; Tang, Y. Chem. Mater. 2000,
(24) Kong, J.; Chapline, M.; Dai, H. AdV. Mater. 2001, 13, 1384.
(25) Bezryadin, A.; Lau, C. N.; Tinkham, M. Nature 2000, 404, 971.
(26) Choi, H. C.; Shim, M.; Bangsaruntip, S.; Dai, H. J. Am. Chem.
2, 1622.
(
(
31) Xing, Y. J. Phys. Chem. B 2004, 108, 19255.
32) Yajima, T.; Uchida, H.; Watanabe, M. J. Phys. Chem. B 2004, 108,
Soc. 2002, 124, 9058.
27) Girishkumar, G.; Vinodgopal, K.; Kamat, P. V. J. Phys. Chem. B
004, 108, 19960.
2
654.
(33) Zhu, Y. M.; Uchida, H.; Yajima, T.; Watanabe, M. Langmuir 2001,
17, 146.
(
2