Chemical treatment of CNTs in acidic KMnO4 solution and promoting effects on the corresponding Pd-Pt/CNTs catalyst

Article abstract of DOI:10.1016/j.molcata.2011.12.032

Carbon nanotubes (CNTs) were chemically treated in acidic KMnO₄ solution. XRD, nitrogen physisorption, FTIR and chemical titration were used to investigate the influence of the treatment conditions on the surface structure, the bulk structure,

Full text of DOI:10.1016/j.molcata.2011.12.032

Journal of Molecular Catalysis A: Chemical 356 (2012) 114–120
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Chemical treatment of CNTs in acidic KMnO4 solution and promoting effects on
the corresponding Pd–Pt/CNTs catalyst
Jiuling Chena,b,∗, Qinghai Chena, Qing Mab, Yongdan Lia, Zhonghua Zhub
a Department of Catalysis Science & Technology, School of Chemical Engineering, Tianjin University, Tianjin 300072, China
b School of Chemical Engineering, the University of Queensland, St Lucia, Brisbane, QLD 4072, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 16 August 2011
Received in revised form 24 October 2011
Accepted 31 December 2011
Available online 9 January 2012
Carbon nanotubes (CNTs) were chemically treated in acidic KMnO4 solution. XRD, nitrogen physisorption,
FTIR and chemical titration were used to investigate the influence of the treatment conditions on the
surface structure, the bulk structure, the types and the amounts of the formed surface oxygen-containing
functional groups of CNTs. The results show that the structure is not changed clearly after treatment
and the formation of oxygen-containing groups, e.g., carboxyls, on the CNT surface is dependent on the
treatment conditions. Under the present conditions, the concentration of the employed sulphuric acid and
the weight ratio of the solid KMnO4 and CNTs have clear influence on the formation of carboxyls, however,
by comparison, the influence is slight when the treatment temperature is above 95 ◦C. With the increasing
of the amount of carboxyls, the dispersion of the corresponding Pd–Pt metal particles supported on CNTs
is increased and the catalytic activity is promoted in the profile reaction of naphthalene hydrogenation
to tetralin.
Keywords:
Carbon nanotubes
Chemical treatment
Acidic KnMO4 solution
Carboxyls
Pd–Pt catalyst
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
published literature have shown that compared with the conven-
tional catalyst supports, like Al2O3, SiO2 and activated carbon, the
Since the discovery of carbon nanotubes (CNTs) by Iijima in
1991, they have become one of the most active fields of nanoscience
and nanotechnology due to their unique properties, which make
them to qualify for many potential applications such as poly-
mer reinforcements for composites or breakthrough materials
for energy storage, electronics and catalysis [1–4]. In heteroge-
neous catalysis, compared with conventional carbon materials, like
activated carbon, CNTs attract a growing interest as a catalyst sup-
port because of their specific characteristics, e.g., high mechanical
strength, macroporous and mesoporous structure, possibility to
control the surface chemistry and easy recovery of precious metals
without introducing other metallic impurities by support burning
[3,4]. Additionally, a theoretical investigation about the interac-
tion of the supported nickel metal atoms with CNTs shows that
the curvature of the surface graphite layers of CNTs significantly
affects the values of magnetic moments of the nickel atoms and
the charge transfer direction between nickel and carbon can be
inverted [5]. Therefore, a certain peculiar metal–support interac-
tion may exist in this catalyst system. The research results in the
catalytic activity or the selectivity of the supported metal catalysts
like Ni, Rh and Pt on CNTs can be promoted pronouncedly due
to this peculiar interaction [3,4,6–9]. However, the surface of the
primitive CNTs, which are generally produced from decomposition
of hydrocarbons, is mostly covered with the inert C H bonds and
the corresponding interaction is very weak between the supported
metal and the CNT surface [3,4,10–13]. Therefore, the surface of
primitive CNTs has to be chemically pre-treated in the oxidants
to generate some oxygen-containing functional groups, like car-
boxyls and phenolic hydroxyls, for anchoring the supported metal
particles to improve the interaction between them [3,4,11–17].
The results in the published literature and in our lab show
that the acidic potassium permanganate (KMnO4) solution is an
efficient method to generate oxygen-containing groups, especially
strongly acidic groups, carboxyls [3,16–20]. Although there have
been many publications about this topic, to our best knowledge, up
to now there is no report concerning the influence of the treat-
ment conditions on the formation of surface oxygen-containing
functional groups, e.g., the ratio of the employed KMnO4 and CNTs,
the concentration of the employed acid and the treatment temper-
ature. Additionally, the further relationship between the treatment
conditions and the types and the amounts of the formed oxygen-
containing groups has not been reported, either.
∗
Corresponding author at: School of Chemical Engineering, the University of
Queensland, St Lucia, Brisbane, QLD 4072, Australia. Tel.: +61 7 33653687;
fax: +61 7 33654199.
In this work, the influence of the chemical treatment condi-
tions on the formation of the oxygen-containing functional groups
E-mail address: cjlchen@yahoo.com (J. Chen).
1381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2011.12.032

J. Chen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 114–120
115
was investigated in the details. Hydrogenation of naphthalene to
tetralin was employed as a profile reaction to investigate the pro-
moting effects of the formed functional groups on the activity of
the supported Pd–Pt/CNTs catalyst.
For the measurement of the total content of the oxygen-
containing groups, 50 ml of 0.01 M NaOH solution was firstly used
for soaking 100 mg of CNTs in a container sealed under nitrogen
for 24 h at RT in an ultrasonic bath and then the suspension was fil-
tered to obtain the filtrate. The solution of 0.0098 M HCl was used to
titrate 20 ml of the above filtrate to obtain the measurement titra-
tion curve. Additionally, by considering the disturbing of CO2, 20 ml
of the same NaOH solution without soaking CNTs was titrated to
obtain the reference titration curve. The consumed volumes of HCl
to neutralize NaOH in the reference titration and in the measure-
ment titration were referred as V1 and V2, respectively. The total
content of the oxygen-containing groups on CNT surface can be cal-
culated out by the expression, 0.244 × (V1 − V2) mmol g−1. The total
content of carboxyls, lactones and lactols and that of carboxyls can
be obtained using HCl to back titrate Na2CO3 and NaHCO3, respec-
tively, which have been used to soak CNTs. A PHS-2 pH meter was
used for monitoring the pH change during the chemical titration.
2. Experimental
2.1. Treatment of CNTs
CNTs were obtained from Tsinghua University. They were pro-
duced from methane catalytic decomposition in a continuous
commercial fluidized-bed reactor. The purity of the CNTs in the
primitive samples is above 95%. The external diameter of CNTs is in
the range of 10–30 nm with the majority around 20 nm. The inner
diameters are typically about one-third of the corresponding exter-
nal diameters. All the particle size of CNTs employed in this work
is between 60 and 80 mesh.
For the chemical treatment of CNTs in the acidic KMnO4 solu-
tion, solid KMnO4 and CNTs were put in sulphuric acid solution in
a flask with three necks and stirred for 100 min, then the suspen-
sion cooled down to room temperature (RT) and was filtered. The
obtained solid mixture was washed with water, filtered, washed
with concentrated HCl solution to remove the produced MnO2 and
then was re-filtered. Following this step, the obtained CNTs were
washed with water for several times until the filtrate became neu-
tral. Finally, CNTs were dried in air at 110 ◦C for 16 h. The obtained
CNT samples under different treatment conditions are listed in
Table 1.
2.3. Preparation and activity of Pd–Pt/CNTs catalyst
To investigate the promoting effects of the formed surface
functional groups on the dispersion of the supported metal cata-
lyst, Pd–Pt/CNTs were prepared by incipient wetness impregnation
method. The mixture solution of [Pd(NH3)4]Cl2 and [Pt(NH3)4]Cl2,
in which the atomic ratio of Pd and Pt was 4:1, was used as
the precursors of Pd–Pt. The results in our lab showed that the
ionic exchange capacities of [Pd(NH3)4]2+ and [Pt(NH3)4]2+ on the
treated CNT samples, were all above 1 wt% of the total content of
Pd–Pt metal. Therefore, to ensure the loading [Pd(NH3)4]2+ and
[Pt(NH3)4]2+ can fully interact with the functional groups, 1 wt.%
Pd–Pt/CNTs catalyst were prepared in this work. During the prepa-
ration, the mixed solution of [Pd(NH3)4]Cl2 and [Pt(NH3)4]Cl2 was
dropped on CNTs at RT and then dried in the ambient for 24 h. After-
wards, the samples were dried in air at 110 ◦C for 16 h and then
calcinated in a tube furnace in an ultra-high purity nitrogen flow
at 380 ◦C for 2 h. The obtained catalysts supported on different CNT
samples are referred as Pd–Pt-0# to Pd–Pt-7#, respectively. The
actual loading content of Pd–Pt metal was measured using a Hitachi
180-80 atomic absorption spectroscopy (AAS).
Hydrogen chemisorption was employed to measure the disper-
sion of the metal particles at RT in a pulse system equipped with
Thermo Finnigan TPDRO1100. The samples of 100 mg were reduced
in a hydrogen flow at 300 ◦C for 2 h, followed by flushing for 2 h
in an ultrahigh purity helium stream of 30 ml min−1 at 300 ◦C and
then cooled in helium to ambient. Hydrogen pulses of 100 ␮l each
were injected into an ultrahigh purity nitrogen flow and detected
in the outlet gas by a thermal conductivity detector. By compar-
ing the amount of hydrogen reaching the detector and the amount
of hydrogen injected, the quantity of hydrogen adsorbed could be
determined. Blank experiments on CNTs showed that there was no
measurable uptake of hydrogen on the support itself. The metal dis-
persion, in terms of hydrogen to metal molar ratio, was calculated
based on the moles of gas adsorbed on the catalyst and expressed
into H/(Pd + Pt), the atomic ratio of the adsorbed hydrogen atoms
and the total metal atoms of Pd and Pt in the catalyst system [24].
The hydrogenation of naphthalene was used as the profile
reaction to investigate the promoting effects on the reactivity of
Pd–Pt/CNTs catalysts. The reactions were carried out in a Parr 4842
high pressure autoclave. For each run, the catalyst sample of 300 mg
was firstly reduced at 300 ◦C for 2 h in hydrogen with a flow rate
of 20 ml.min−1 and then cooled to RT and moved to the reactor
pre-filled with 120 ml solvent of tridecane under the protection
of ultrahigh-purity nitrogen. Then 9 g of naphthalene were put in
the reactor. When the temperature was heated to 220 ◦C, the reac-
tion solution began to be stirred. During the reaction, the hydrogen
pressure was kept at 6.0 MPa, the reaction temperature was held
2.2. Characterization of CNTs
The XRD measurements of the CNT samples before and after
the treatment were conducted with a Bruker D8 advance research
X-ray diffractor (XRD) with Cu K␣ radiation at a scanning rate of
2◦ min−1
.
The texture structures of the samples were measured using
nitrogen physisorption in a NOVA 1200 adsorption analyser (Quan-
tachrome). The corresponding specific surface areas (Sg) were
calculated by the Brunauer–Emmett–Teller (BET) equation at rela-
tive pressure (P/P0) between 0.05 and 0.35. The total pore volumes
were calculated from the amount of vapor adsorbed at the relative
nitrogen pressure of 0.98, by assuming that all the pores were filled
with liquid N2.
The surface functional groups of the different CNT samples
were observed in a Nicolet NEXUS Fourier transform infrared spec-
troscopy (FTIR). CNT samples and KBr particles were first ground
into particles smaller than 300 mesh, dried in air at 110 ◦C for 5 h,
cooled to ambient, mixed physically with each other to uniformity
and finally pressured into wafers. Since CNTs can strongly absorb
IR light, the content of CNTs in the wafers was only 0.2–0.3 wt.%.
During the investigation, the wafers were located on the sample
shelf, which was exposed to the ambient in the instrument. IR
absorptions in the range of 3800–800 cm−1 were recorded.
According to the method suggested by Boehm [21–23], on car-
bon materials, e.g., activated carbon and carbon black, the surface
oxygen-containing groups, which differ in their acidities, can be
quantitatively measured in the process of chemical titration by neu-
tralization with NaHCO3, Na2CO3 and NaOH, respectively. NaHCO3
can only react with carboxyls (including carboxylic anhydride)
and Na2CO3 can react with lactones and lactols besides carboxyls,
and NaOH can react with carboxyls, lactone, lactols and phenolic
hydroxyls. Therefore, the differences in acidity of the various types
of functional groups allow differentiation by the simple titration
method. For example, the difference between NaOH and Na2CO3
consumption corresponds to the phenolic groups.

116
J. Chen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 114–120
Table 1
Employed CNT samples before and after treatment in the acidic KMnO4 solution under different conditions.
CNT samples
Treatment conditions
Temperature (◦C)
Concentration of H2SO4 (M)
Weight ratio of KMnO4 and CNTs
0#
1#
2#
3#
4#
5#
6#
7#
–
75
95
115
95
95
–
–
4.5
4.5
4.5
3
6
4.5
4.5
3:1
3:1
3:1
3:1
3:1
1:1
5:1
95
95
at 250 ◦C, the stirring speed was 660 rad min−1. The samples were
taken in the interval of 20 min and were analyzed using a gas chro-
matography equipped with a hydrogen flame ionization detector
and a HP-PONA methyl silicone capillary column.
A
1250
1400
1580
1630
1730
3460
3180
2930
1190
1130
1050
2860
3. Results and discussion
3#
2#
3.1. Influence on structure of CNTs
1#
0#
Fig. 1 shows that the XRD profiles of the CNTs before and after
chemical treatment under the different conditions. It can be found
that all the eight CNT samples possess the crystal structure similar
to that of graphite. The structures of CNTs after treatment under
the different conditions remain unchanged, which indicates that
the treatment does not change the bulk structure of the CNTs.
The specific surface areas, Sg, and the total pore volumes, Vp,
of CNTs obtained using N2 physisorption, are shown in Table 2. It
can be seen that there are only some small changes after treat-
ment, which may be attributed to the surface modification from
the oxidization of some active carbon layers on the surface, e.g.,
the unsaturated carbon atoms in the defects or the carbon layers
like pentagon and heptagon located on the ends of CNTs [3,18].
The above results of XRD and N2 physisorption indicate that the
treatment in acidic KMnO4 solution can not clearly alter the bulk
structure and the texture, which is advantageous to the use of CNTs
as a support.
3500
3000
3180
2500
2000
cm-1
1500
1000
B
1400
3460
1580
1630
1190
1130
1050
2930
1730
2860
5#
2#
4#
0#
3.2. Formation of surface functional groups
Fig. 2A–C shows FTIR spectra of CNT samples from 0# to
7#. To allow comparison, transmission levels of all spectra were
kept approximately same. It was ascertained that, within the
3500
3000
3180
2500
2000
1500
1000
1130
(cm-1)
C
1400
3460
1190
1580
1630
1730
1050
7#
6#
5#
2930
2860
7#
2#
4#
3#
6#
0#
2#
1#
0#
3500
3000
2500
2000
1500
1000
(cm-1)
20
30
40
50
60
70
2 (degree)
Fig. 2. IR spectra of the CNT samples before and after treatment in the acidic KMnO4
solution.
Fig. 1. XRD profiles of the CNT samples before and after treatment in the acidic
KMnO4 solution; Cu K␣; ꢀ, graphite.

J. Chen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 114–120
117
Table 2
Specific surface area (Sg) and pore volume (Vp) of the CNT samples.
CNT samples
0#
1#
2#
3#
4#
5#
6#
7#
Sg (m2 g−1
Vp (ml g−1
)
)
158
0.37
149
0.36
162
0.38
150
0.35
153
0.37
139
0.34
154
0.37
138
0.34
transmission window used, the intensity of the bands did not
depend upon the transmission level of the spectra. The absorptions
in the spectra can be assigned according to Refs. [25–31].
12
10
8
A
The peaks at 1050 and 1120 cm−1 are because of the stretching
vibration of C O in ethers [26,27]; the peaks at 1190 and 1250 cm−1
originate from the stretching vibration of C O in phenols [26,27];
the peak at 1400 cm−1 may be associated with H O bending vibra-
tion in water, phenols and carboxyls [25,28,29] and the peak at
3164 cm−1 may be attributed to H O stretching vibration in phe-
nols and carboxyls [26,27]; the peak at 1580 cm−1can be ascribed to
the stretching vibration of aromatic rings in the surface of CNTs. The
absorptions at 1630 cm−1 are due to water adsorbed on the KBr and
CNTs, possibly during the wafer preparation and the observation
[25,28]; the peak at 1715 cm−1 can be assigned to the C O stretch-
ing vibration in carboxyls or carbonyls [25,26,28]; the peaks of 2846
and 2930 cm−1 are in the range of CH3/CH2 stretching vibration
[25,26]. The absorptions at 3440 cm−1 are due to the H O stretch-
ing vibration of the adsorbed water on the KBr and CNTs [30,31].
The absorptions at 1400, 1630 and 3440 cm−1 are also observed to
be present in the KBr backgrounds.
reference
1#
2#
6
4
2
0
5
10
15
20
25
VHCl (ml)
12
10
8
B
Based on the above analysis, before treatment, there are only
small amounts of C O on the surface of CNTs, mainly in ethers
and a few in phenols. Oxygen in air may be chemisorbed at low
temperatures on unsaturated carbon atoms at some defects or dis-
locations that occur on the surface of the CNTs and may react further
to generate some surface oxygen-containing groups [10]. By com-
parison, some bands are formed apparently after treatment: one is
at 1580 cm−1 originating from an aromatic ring vibration [25,26];
reference
6
1#
2#
and the other is at 1715 cm−1 which can be assigned to the C
O
stretching vibration in carboxyls or carbonyls [25,26,28]. The peak
at 1580 cm−1 indicates aromatic skeleton rings sticking out of the
surface of CNTs after the oxidization of the surface and the for-
mation of C O shows the generation of carbonyls or carboxyls
4
2
in the surface of CNTs. The absorptions at 1190, 1250 cm−1 (C
O
0
5
10
15
20
25
stretching vibrations in phenols) are enhanced indicating that more
phenols were formed [26,27]. In Fig. 2A, it can be seen that the
amount of the formed C O or the phenols is higher at 95 and 115 ◦C
than at 75 ◦C. In Fig. 2B and C, they are both higher using 4.5 M and
6.0 M sulphuric acid than using 3.0 M or higher with the propor-
tion of KMnO4 and CNTs of 3:1 and 5:1 than with that of 1:1. Since
the surface functional groups are very complex and their vibration
ranges are adjacent to one another, it is difficult to make further
quantitative assignments. However, it can be found that after treat-
ment, phenols and carboxyls are formed on CNT surface and the
amount of them has a relationship with the treatment conditions.
According to the method suggested by Boehm [21–23], the
amounts of the different oxygen-containing surface groups can
be measured by using HCl to back titrate the solution of NaOH,
NaHCO3 and Na2CO3, respectively, which have been used to soak
CNT samples. As an example, the reference titration curves and the
measurement titration curves of two CNT samples 1# and 2# are
shown in Fig. 3A–C. The types and amounts of the surface func-
tional groups on CNTs were calculated out and listed in Table 3.
The relationship of them with treatment conditions is further illus-
trated in Fig. 4. It can be found that before treatment the amount
of the total oxygen-containing groups is very low and only a few
weakly acidic phenolic groups are existed, which is consistent with
VHCl (ml)
10
8
C
6
reference
1#
2#
4
2
0
5
10
VHCl (ml)
15
20
25
Fig. 3. Chemical titration curves using HCl to back titrate NaOH (A), Na2CO3 (B) and
NaHCO3 (C) for the quantitative measurement of the oxygen-containing groups on
CNT samples 1# and 2#.

118
J. Chen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 114–120
Table 3
Amounts of the surface oxygen-containing groups on the CNT samples.
CNT samples
0#
1#
2#
3#
4#
5#
6#
7#
Carboxyl (mmol g−1
)
–
–
–
0.38
0.35
0.52
1.25
0.54
0.27
0.5
0.55
0.26
0.53
1.34
0.47
0.14
0.45
1.06
0.71
0.21
0.52
1.44
0.28
0.29
0.41
0.98
0.65
0.3
0.53
1.48
Lactone and lactol (mmol g−1
)
)
Phenolic hydroxyl (mmol g−1
Total (mmol g−1
)
0.013
1.31
1.6
0.50
1.4
A
total groups
0.40
0.30
0.20
0.10
0.00
1.2
1.0
0.8
0.6
0.4
0.2
1.2
0.8
0.4
0
carboxyls
k
lactones & lactols
phenols
H/(Pd+Pt)
total
carboxyls
0#
1#
2#
3#
4#
5#
6#
7#
CNT sample
70
80
90
100
110
120
Fig. 5. Amount of total oxygen-containing functional groups and that of carboxyls
on CNTs, the atomic ratio of H/(Pd + Pt) during the measurement of metal dispersion
of Pd–Pt/CNTs and the pseudo-first-order reaction rate constant, k, on Pd–Pt/CNTs
in naphthalene hydrogenation to tetralin.
Treatment temperature (OC)
1.6
1.2
0.8
0.4
0.0
B
the FTIR observation. After treatment, the amount of the functional
groups can be increased pronouncedly. From Fig. 4A, it can be seen
that the amount of carboxyls is increased clearly when the treat-
ment temperature is increased from 75 to 95 ◦C (1# → 2#) and is
almost not changed from 95 to 115 ◦C (2# → 3#). The amount of
phenols is unchanged with the increase of the treatment temper-
ature. The amount of latones and lactols is decreased from 75 to
95 ◦C, then keeps flat from 95 to 115 ◦C. This indicates that the oxi-
dation ability of acidic KMnO4 solution can be enhanced from 75
to 95 ◦C and cannot be promoted further above 95 ◦C. In Fig. 4B, the
amount of carboxyls is flatly increased when the concentration of
the employed sulphuric acid is increased from 3 to 4.5 M (4# → 2#)
and is clearly increased when it is from 4.5 to 6 M (2# → 5#). The
amount of phenols keeps flat in the full range. The amount of lac-
tones and lactols is increased when it is from 3 to 4.5 M and then
decreased when it is from 4.5 to 6 M. In Fig. 4C, when the weight
ratio of KMnO4 and CNTs is increased from 1 to 5, the amount of
carboxyls is increased and that of phenols is flatly increased and
that of latones and lactols remains unchanged. In the acidic KMnO4
carboxyls
lactones & lactols
phenols
total
3
4
5
6
H2SO4 concentration (M)
1.6
1.2
0.8
0.4
0.0
C
carboxyls
lactones & lactols
phenols
total
Table 4
Actual loading content of Pd–Pt/CNTs, the atomic ratio of H/(Pd + Pt) during the
measurement of metal dispersion of Pd–Pt/CNTs and the corresponding pseudo-
first-order reaction rate constant, k, in naphthalene hydrogenation to tetralin.
Catalyst samples
Actual loading
content (%)
H/(Pd + Pt)
k (min−1
)
1
2
3
4
5
Pd–Pt-0#
Pd–Pt-1#
Pd–Pt-2#
Pd–Pt-3#
Pd–Pt-4#
Pd–Pt-5#
Pd–Pt-6#
Pd–Pt-7#
1.13
0.027
0.193
0.256
0.262
0.218
0.321
0.202
0.296
0.00033
0.0017
0.0033
0.0035
0.0020
0.0038
0.0020
0.0037
Weight ratio of KMnO4 and CNTs
0.985
0.937
0.956
1.01
0.936
0.991
0.946
Fig. 4. Effect of treatment condition on the formation of surface oxygen-containing
groups on CNTs: (A) treatment temperature; (B) concentration of H2SO4; (C) weight
ratio of KMnO4 and CNTs.

J. Chen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 114–120
119
40
35
30
25
20
15
10
5
0.35
A
B
Ptd-Pt-3#
Pd-Pt-2#
Pd-Pt-1#
Pd-Pt-0#
Pd-Pt-3#
0.30
Pd-Pt-2#
Pd-Pt-1#
0.25
0.20
0.15
0.10
0.05
0.00
Pd-Pt-0#
0
0
50
100
150
t (min)
200
250
20
40
60
80
100
t (min)
45
C
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
D
40
35
30
25
20
15
10
5
Pd-Pt-5#
Pd-Pt-2#
Pd-Pt-4#
Pd-Pt-0#
Pd-Pt-5#
Pd-Pt-2#
Pd-Pt-4#
Pd-Pt-0#
0
0
50
100
150
200
250
20
40
60
80
100
t (min)
t (min)
40
35
30
25
20
15
10
5
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
E
F
Pd-Pt-7#
Pd-Pt-2#
Pd-Pt-6#
Pd-Pt-0#
Pd-Pt-7#
Pd-Pt-2#
Pd-Pt-6#
Pd-Pt-0#
0
0
50
100
150
t (min)
200
250
20
40
60
80
100
t (min)
Fig. 6. Influence of the treatment of CNTs on conversion of naphthalene to tetralin on Pd–Pt/CNTs: (A, C and E) activity curves; (B, D and F) pseudo-first-order kinetic plots.
generated. Under the present treatment conditions, the con-
centration of the employed sulphuric acid and the weight ratio of
solid KMnO4 and CNTs have clear influence on the formation of
carboxyls, however, by comparison, the influence is slight when
the treatment temperature is above 95 ◦C.
solution, the oxidization process of solid carbon materials could
occur according to the reaction as below [32]:
3C + 4KMnO4 + 4H+ → 4MnO2 + 3CO2 + 4 K+ + 2H2O
Therefore, with the increase of the concentration of sulphuric
acid or the amount of the employed KMnO4, the reactivity of the
acidic KMnO4 solution is enhanced. Therefore, the unstable groups,
phenols, lactones or lactols could be further oxidized into car-
boxyls. In Fig. 5, it can be seen that except 3#, the tendency of
the amount of the total oxygen-containing groups is similar to that
of carboxyls, which indicates that the contribution of the amount
increase of the total oxygen-containing groups is mainly from car-
boxyls. The above results show that the types and the amounts of
the functional groups are dependent on the treatment conditions.
When the treatment becomes more severe, more carboxyls are
3.3. Dispersion and activity of Pd–Pt/CNTs
The actual loading content of Pd–Pt metal in the eight
Pd–Pt/CNTs catalysts was measured by AAS and is listed in Table 4.
It can be seen that they are all around 1 wt.% as expected.
Hydrogen chemisorption was employed to measure the disper-
sion of the metal particles and the results are listed in Table 4.
Fig. 5 illustrates the relationship between the metal dispersion of
the Pd–Pt/CNTs and the amount of the carboxyls. It can be found

120
J. Chen et al. / Journal of Molecular Catalysis A: Chemical 356 (2012) 114–120
that the order of the former is the same as that of latter. More
carboxyls are formed on the surface, is higher dispersion of the
corresponding catalyst obtained here. Carboxyls on the surface of
solid carbon support can have strong interaction with the loaded
metal precursors [Pd(NH3)4]2+ and [Pt(NH3)4]2+ through the for-
mation of chemical bonds, like COO−[Pt(NH3)4]2+−OOC , which
serve to hinder agglomeration and surface diffusion of metal ions,
and minimize the sintering propensity of the formed Pd–Pt par-
ticles across the CNT surface to obtain the high metal dispersions
in the calcination and reduction process [10]. The results in Fig. 5
indicate that the amount of the carboxyls is vital to the dispersion
of the Pd–Pt metal particles.
temperature is above 95 ◦C. The presence of carboxyls on CNT sur-
face can improve the dispersion of the supported Pd–Pt metal
particles and promote their activity in hydrogenation of naphtha-
lene to tetralin. It seems that acidic KMnO4 solution is an efficient
treatment method for CNTs before being employed as the catalyst
support.
Acknowledgment
This work was supported by the Natural Science Foundation of
China under contract number 20006012.
Hydrogenation of naphthalene to tetralin was employed as the
profile reaction to further show the promoting effect of the surface
groups on the activity of the supported metal catalyst because the
activity of the Pd–Pt metal particles in this reaction is increased
with the decrease of their size [33]. During the hydrogenation pro-
cess, tetralin is the only product. From the curve of conversion
with the reaction time extension in Fig. 6, it can be found that the
activity of Pd–Pt-0# is very low and after treatment, the activity
of Pd–Pt/CNTs is promoted obviously. Additionally, in an ideal stir-
ring batch reactor, the hydrogenation of naphthalene to tetralin can
be regarded as a pseudo-first-order reaction [34]. The relationship
between the conversion, x, and reaction time, t, can be expressed
by the equation, −ln(1 − x) = kt, in which k is the reaction rate con-
stant. Therefore, the first five conversions are used to show the
plot of the relationship, −ln(1 − x) ∼ t. The obtained k are listed in
Table 4. The tendency of both k and dispersion of Pd–Pt metal par-
ticles of the corresponding catalysts is illustrated in Fig. 5. It can
be found that the order of k is the same as that of the Pd–Pt dis-
persion. Since the catalyst samples have the similar loading content
and are obtained from the same preparation method, the difference
in activity should be attributed to the dispersion of the metal parti-
cles. Therefore, the formation of carboxyls during treatment in the
acidic KMnO4 solution can improve the dispersion of Pd–Pt metal
particles, which can promote the activity of the corresponding cat-
alyst. With the increasing of the amount of carboxyls, the Pd–Pt
dispersion is increased and the activity of the catalyst is enhanced
as well.
References
[1] S. Bose, R.A. Khare, P. Moldenaers, Polymer 51 (2010) 975–993.
[2] A. Reyhani, S.Z. Mortazavi, S. Mirershadi, A.Z. Moshfegh, P. Parvin, A.N. Golikand,
J. Phys. Chem. C 115 (2011) 6994–7001.
[3] P. Serp, M. Corrias, P. Kalck, Appl. Catal. A 253 (2003) 337–358.
[4] K.P. de Jong, J.W. Geus, Catal. Rev.-Sci. Eng. 42 (2000) 481–510.
[5] M. Menon, A.N. Andriotis, G.E. Froudakis, Chem. Phys. Lett. 320 (2000) 425–434.
[6] C. Park, R.T.K. Baker, J. Phys. Chem. B 103 (1999) 2453–2459.
[7] C.A. Bessel, K. Laubernds, N.M. Rodriguez, R.T.K. Baker, J. Phys. Chem. B 105
(2001) 1115–1118.
[8] R. Giordano, P. Serp, P. Kalck, Y. Kihn, J. Schreiber, C. Marhic, J.L. Duvail, Eur. J.
Inorg. Chem. 4 (2003) 610–617.
[9] J.Z. Luo, L.Z. Gao, Y.L. Leung, C.T. Au, Catal. Lett. 66 (2000) 91–97.
[10] F. Rodriguez-Reinoso, Carbon 36 (1998) 159–175.
[11] C. Lu, H. Chiu, Chem. Eng. J. 139 (2008) 462–468.
[12] J.L. Chen, Z.H. Zhu, Q. Ma, L. Li, V. Rudolph, G.Q. Lu, Catal. Today 148 (2009)
97–102.
[13] J.L. Chen, Z.H. Zhu, S.B. Wang, Q. Ma, V. Rudolph, G.Q. Lu, Chem. Eng. J. 156
(2010) 404–410.
[14] R.I. Jafri, T. Arockiados, N. Rajalakshmi, S. Ramaprabhu, J. Electrochem. Soc. 157
(2010) B874–B879.
[15] M.A.M. Motchelaho, H.F. Xiong, M. Moyo, J. Mol. Catal.
189–198.
[16] T.J. Aitchison, M. Ginic-Markovic, J.G. Matisons, G.P. Simon, P.M. Fredericks, J.
Phys. Chem. C 111 (2007) 2440–2446.
[17] Q.H. Chen, H. Liu, J.L. Chen, Y.D. Li, Tansu 3 (2006) 8–14.
[18] K. Hernadi, A. Siska, L. Thien-Nga, L. Forro, I. Kiricsi, Solid State Ionics 141–142
(2001) 203–209.
[19] N.Y. Zhang, J.N. Xie, V.K. Varadan, Smart Mater. Struct. 11 (2002) 962–965.
[20] M.O. Danilov, A.V. Melezhyk, J. Power Sources 163 (2006) 376–381.
[21] H.P. Boehm, E. Diehl, W. Heck, R. Sappok, Angew. Chem. Int. Ed. Engl. 3 (1964)
669–678.
A 335 (2011)
[22] H.P. Boehm, Carbon 32 (1994) 759–769.
[23] H.P. Boehm, Carbon 40 (2002) 145–149.
[24] S. Jongpatiwut, N. Rattanapuchapong, T. Rirksomboon, S. Osuwan, D.E. Resasco,
Catal. Lett. 122 (2008) 214–222.
[25] T.G. Ros, A.J. van Dillen, J.W. Geus, D.C. Koningsberger, Chem. Eur. J. 8 (2002)
1151–1162.
4. Conclusions
The treatment in the acidic KMnO4 solution cannot clearly
change the bulk structure and the surface texture of CNTs under
the present conditions, however, different oxygen-containing func-
tional groups can be formed on the surface. The types and the
amounts of the formed functional groups, e.g., carboxyls, are
dependent on the treatment conditions. The concentration of the
employed sulphuric acid and the weight ratio of the solid KMnO4
and CNTs have clear influence on the formation of carboxyls, how-
ever, by comparison, this influence is slight when the treatment
[26] P.E. Fanning, M.A. Vannice, Carbon 31 (1993) 721–730.
[27] E. Papirer, E. Guyon, N. Perol, Carbon 16 (1978) 133–140.
[28] M.S.P. Shaffer, X. Fan, A.H. Windle, Carbon 36 (1998) 1603–1612.
[29] C. Moreno-Castilla, M.V. Lopez-Ramon, F. Carrasco-Marin, Carbon 38 (2000)
1995–2001.
[30] H. Kandori, Y. Shichida, J. Am. Chem. Soc. 122 (2000) 11745–11746.
[31] W. Neagle, C.H. Rochester, J. Chem. Soc. Faraday Trans. 86 (1990) 181–183.
[32] J.F. Colomer, P. Piedigrosso, A. Fonseca, J.B. Nagy, Synth. Met. 103 (1999)
2482–2483.
[33] B.H. Cooper, B.B.L. Donnis, Appl. Catal. A 137 (1996) 203–223.
[34] A. Corma, A. Martinez, V. Martinez-Soria, J. Catal. 169 (1997) 480–489.