Tuning hydrophilic properties of carbon nanotubes: A challenge for enhancing selectivity in Pd catalyzed alcohol oxidation

Article abstract of DOI:10.1016/j.cattod.2011.09.041

CNTs were prepared by CCVD and functionalized with inorganic acid, with the aim to study the effect of the surface properties on the catalytic performance in the selective alcohol oxidation using polar (water) and apolar (cyclohexane) solvent. The match b

Full text of DOI:10.1016/j.cattod.2011.09.041

Catalysis Today 186 (2012) 76–82
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Tuning hydrophilic properties of carbon nanotubes: A challenge for enhancing
selectivity in Pd catalyzed alcohol oxidation
a
a
a
b
b
b
a,∗
A. Villa , M. Plebani , M. Schiavoni , C. Milone , E. Piperopoulos , S. Galvagno , L. Prati
a
Dipartimento di Chimica Inorganica Metallorganica e Analitica L.Malatesta, Università degli Studi di Milano, via Venezian 21, 20133 Milano, Italy
Dipartimento di Chimica Industriale e Ingegneria dei Materiali, Università degli Studi di Messina, Contrada di Dio, 98166 Messina, Italy
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 May 2011
Received in revised form
CNTs were prepared by CCVD and functionalized with inorganic acid, with the aim to study the effect of
the surface properties on the catalytic performance in the selective alcohol oxidation using polar (water)
and apolar (cyclohexane) solvent. The match between properties of the reactant and the solvent allows
tuning the selectivity of the reaction.
2
9 September 2011
Accepted 30 September 2011
Available online 1 November 2011
©
2011 Elsevier B.V. All rights reserved.
Keywords:
Carbon nanotubes
Surface functionalization
Alcohol oxidation
1
. Introduction
oxidation of alcohol carried out in water. The main problems that
can be envisaged to perform the reaction of hydrophobic alcohol
in an aqueous media concern with the intimate contact of reac-
tants (alcohol and oxygen with the catalyst) and the lower oxygen
solubility in water with respect of organic solvent [5].
Active carbons have been extensively studied as support in
the liquid phase oxidation of alcohols [6–9]. However, more
recently, it has been demonstrated that the more crystalline
(graphene/graphite) nature of carbon nanotubes could be beneficial
in terms of activity and durability of catalytic systems [10]. Surfaces
of active carbons are normally hydrophobic indeed be wetting in a
better way in apolar solvent than in polar ones [11].
Alcohol oxidation has been always foreseen as an important task
for chemical transformation especially in the case of fine chem-
ical production. However in this field a lot of drawbacks have
up-to-now limited the development of oxidative processes on an
industrial scale. In fact when O2 is use as the oxidant and reactions
are carried out in condensed phase, deactivation of catalyst often
occurred and high selectivity are not reached [1].
Lot of efforts has been made in order to increase both selectivity
and durability of the catalyst by changing the metallic active sites,
the solvent, the temperature and O2 pressure. Generally speaking
catalyst life is increased by using higher temperature, lower oxygen
pressure and a metal less prone to be over-oxidize like gold [2–4].
Solventless conditions have been also investigated showing good
results in terms of both activity and selectivity. However, as well as
gas phase conditions, also carrying out the reaction without solvent
can be not applicable to all the substrate. In fact high boiling or ther-
mally unstable alcohol cannot be vaporized to be oxidized in gas
phase like solid alcohols cannot be directly oxidized in condensed
phase without the addition of a solvent.
In the present paper we investigated the impact of using apo-
lar or polar solvents in the liquid phase oxidation of hydrophilic
or hydrophobic alcohol by modifying the surface of CNTs by
means of different HNO /H SO treatment in order to change the
3
2
4
hydrophilic/hydrophobic properties of their surface. The final aim
of the study is to find out the optimized conditions (support char-
acteristics and solvent) for each class (water-soluble or insoluble)
of alcohols in terms of activity and selectivity.
Indeed CVD prepared CNTs have been differently treated
for introducing functionalities and defects thus varying surface
hydrophilicity. Pd nanoparticles (PdNP) were deposited onto the
CNTs producing 1%Pd/CNTs materials which have been used as
catalyst for the selective benzyl alcohol oxidation in cyclohex-
ane and in water. In addition the same materials were used
in the glycerol selective oxidation as a test of very hydrophilic
substrate.
From an environmental and safety point of view, water should
be considered the solvent of choice but apart from water-soluble
substrate, only a few reports in the literature report selective
∗ Corresponding author. Fax: +39 02503 14405.
E-mail address: Laura.Prati@unimi.it (L. Prati).
0
920-5861/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2011.09.041

A. Villa et al. / Catalysis Today 186 (2012) 76–82
77
2
. Experimental
(0.01 M) or NaOH (0.10 M) were used as titrant, starting from the
initial pH of the CNTs suspension.
2
.1. Materials
The volatile matter content was estimated by Thermogravime-
try (TA Instruments SDTQ 600) in the temperature range
−
1
Carbon nanotubes (CNTs) were prepared by Catalytic Chemical
473–1273 K, heating rate 20 K min . Each sample, placed in a plat-
inum pan, was kept at 473 K in argon until balance stabilisation and
subsequently temperature was raised until the desired value.
Vapor Deposition (CCVD) of isobutane (i-C H10) over Fe supported
on Al O catalyst (iron load 29 wt.%) [12].
The synthesis was carried out placing the catalyst (0.5 g) in a
quartz boat inside a quartz tube located in a horizontal electric fur-
nace. The catalyst was reduced at 773 K under a H /He flow (volume
ratio 1:1, total flow 120 sccm/min) for 1 h then the temperature
was raised to 873 K and helium was replaced by isobutane. The
H /i-C H volume ratio was 1:1 and the total flow was kept at
4
2
3
2
Specific surface area (m /g), were determined by
adsorption–desorption of dinitrogen at −77 K, after outgassing
−
4
(10 mbar) the samples at 353 K for 2 h using Surface Area
Analyzers – Qsurf Series.
2
Morphology and dimensions of the CNTs were investigated by
using a TEM (JEOL JEM 2010), operating at 200 kV and equipped
with a Gatan 794 Multi-Scan CCD camera. The nanotubes were dis-
persed in isopropanol and deposited on a grid covered with holey
carbon film.
2
4
10
1
20 sccm/min for 2 h. After synthesis, the product was cooled down
to room temperature under helium atmosphere.
Support and iron particles are subsequently removed by reflux-
ing the solid discharged from the reaction with a solution of NaOH
A Hitachi S-5200 in lens UHR FE-SEM was used for SEM obser-
vation.
(
1 M) at 353 K and then with a solution of HCl (1 M).
CNTs were finally washed thoroughly with distilled water and
dried at 353 K for 12 h.
Oxidation of CNTs in nitric–sulphuric acid mixture was under-
taken in an ultrasonic bath at 333 K for 6 h. 3 g of CNTs were weighed
Temperature programmed desorption profiles were obtained
with a flow reactor equipped with a quadrupole mass spectrom-
eter (HPR 20 HIDEN ANALYTICAL instrument). Samples (30 mg)
were placed in a U-shaped quartz tube inside an electrical furnace
−
1
and added to 300 mL of acid mixture (67% HNO3 and 98% H SO )
and heated at 10 K min up to 1373 K using a constant flow rate
of helium (30 sccm/min). The mass signals m/z = 28 (CO), 30 (NO),
44 (CO ), and 64 (SO ) were monitored during the analysis. The
2
4
having a proper volume ratio. Three different nitric–sulphuric
volume–ratio mixtures were used, 3:1, 1:1 and 1:3, respec-
tively. After treatment, carbon material was separated by filtration
through 0.2 ␮m filter paper, washed with water to neutral pH and
dried at 353 K for 12 h.
2
2
amounts of CO, CO , NO and SO were calibrated at the end of each
2
2
analysis with pure gases. After the treatment of the acquired data,
−
1
−1
s ) were
the TPD spectra of CO, CO , NO and SO2 (in ␮mol g
2
Table 1 reports the catalysts code which summarizes infor-
mation relative their treatment (CNTs stands for purified and
CNTsN:S stands for oxidized). The subscript N:S indicates the
nitric–sulphuric volume ratio used for the oxidation.
obtained.
Water vapor uptake was evaluated by a thermo-gravimetric
technique based on the use of a laboratory McBain. Samples were
◦
heated at T = 150 C in vacuum (P = 0.01 mbar) for 12 h, cooled down
◦
to T = 30 C (in vacuum) and then connected to the evaporator at
a (absolute) vapor pressure of P = 10 mbar. Measurements ended
when samples reached a stable weight.
2.2. CNTs characterization
The purity degree, PD% of the carbonaceous materials, evalu-
2
.3. Catalyst preparation
ated by measuring the residual weight upon combustion of samples
at 1073 K, a temperature largely exceeding the range of tempera-
ture the more oxidation-resistant CNTs burns [12], was estimated
as 100-wI, wI representing the relative weight of un-combusted
impurities.
Pd sol: solid Na PdCl4 (0.043 mmol) and 0.22 mL PVA solution
2
(2%, w/w) (Pd/PVA 1:1, w/w) were added to 130 mL of H O. After
2
3
min, 0.860 mL of 0.1 M NaBH4 solution was added to the yellow-
brown solution under vigorous magnetic stirring. The brown Pd(0)
sol was immediately formed. An UV–vis spectrum of the palladium
sol was recorded for ensuring the complete reduction of Pd(II).
Within few minutes from their generation, the colloids (acidified
at pH 2, by sulphuric acid) were immobilized by adding the sup-
port under vigorous stirring. The amount of support was calculated
in order to obtain a final metal loading of 1 wt% (on the basis of
quantitative loading of the metal on the support). The catalyst was
Water vapor uptake was calculated as:
(
mt − m )
0
Water vapor uptake (wt%) = 100
m
0
◦
where m represents the CNTs weight after degassing at 150 C and
0
mt represents the final total weight (CNTs + H O).
2
Potentiometric titration of the basic sites was carried out using
a Mettler Toledo Titrator. Approximately 0.2 g of sample was sus-
◦
thoroughly washed with deionized water at 50 C. TGA performed
◦
−3
in the range 100–200 C revealed the presence of residual PVA in
pended in 50 mL of KCl 10 M and then sonicated and equilibrated
for several hours. Prior to each measurement, the suspension
was continuously saturated with argon to eliminate the influence
a very low amount (less than 1 wt% respect to the added amount).
The metal loading was also confirmed by burning off the CNS and
performing ICP (Jobin Yvon JV24) analyses of the solution.
of CO , until the pH was constant. Volumetric standards of HCl
2
Table 1
Sample code, nitric–sulphuric acid mixture volume ratio used for CNTs oxidation, purity degree (PD%), volatile content (wt.%) and surface area (SA) of purified and oxidized
carbon nanotubes.
Codea
Nitric–sulphuric (v:v)
PD (%)
Volatile (wt%)
SA (m /g)
2
CO/CO2b (␮mol g−1
)
NOb
SO2b (␮mol g−1
)
Water vapor uptake (%)
CNTs (6.4)
–
94.1
95.4
95.2
96.6
2.0
4.0
12.0
36.0
196
195
194
172
4.4
2.7
1.7
1.3
n.d.c
3
4
2
n.d.c
6
182
448
2
4
8
CNTs3:1 (4.4)
CNTs1:1 (3.7)
CNTs1:3 (2.8)
3:1
1:1
1:3
11
a
In parenthesis the value of auto-generated pH.
CO/CO2 ratio and concentration (␮mol g ) of NO and SO2 as monitored by TPD analysis of carbon materials are also reported.
n.d.: not detected.
b
c
−1

7
8
A. Villa et al. / Catalysis Today 186 (2012) 76–82
2.4. Catalyst characterization
which could enhance the hydrophilicity of the surface, the CNTs
have been dispersed in a mixture of HNO /H SO varying the ratio
3
2
4
Metal particle distribution was investigated by using a TEM
of HNO3 to H SO4 from 3:1 to 1:3.
2
(
7
(
JEOL JEM 2010), operating at 200 kV and equipped with a Gatan
94 Multi-Scan CCD camera. Transmission Electron Microscopy
TEM) was performed with a Philips model CM12 electron micro-
Table 1 summarizes the results of textural and proximate anal-
ysis of purified and oxidized CNTs. Specific surface area (SA) of the
samples appears to be scarcely influenced by the oxidation con-
ditions. Indeed, as shown in Table 1, close SA values are found
for CNTs, CNTs3:1 and CNTs1:1, while it only slightly reduces for
the sample treated with the most sulphuric acid concentrated acid
mixture (CNTs1:3).
The liquid phase treatment of CNTs leads to a progressive
increase of PD as nitric–sulphuric acid mixture volume ratio low-
ers. The residue obtained upon CNTs oxidation is mainly due to the
presence of iron catalyst encapsulated by the graphitic structures
during CCVD synthesis [13] that is rather difficult to be dissolved
under the mild post synthesis purification with HCl 1 M.
The volatile matter content (wt%) of the investigated samples,
reported in Table 1, also progressively increases with the sulphuric
acid concentration, i.e. lowering the nitric–sulphuric acid mixture
volume ratio. Since these volatiles include products released from
the decompositions of the surface groups the above results clearly
indicates that sulphuric rich acid mixture is more efficient towards
the oxidation of the CNTs surface.
scope operating at 120 kV and directly interfaced with a computer
for real-time image processing. The powdered samples were ultra-
sonically dispersed in isopropyl alcohol, and then a few droplets
of the suspension were deposited on a holey carbon support film
copper grid. After the solvent evaporated, the specimens were
introduced into the microscope column. At least 300 palladium par-
ticles were measured from each catalyst to obtain a good statistical
particle size distribution.
The average size diameter (d) of the palladium particles was
calculated using the following formula:
ꢀ
nidi
d =
n
i
where n is the number of the metal particles of diameter d . The
standard deviation (ꢀ) was calculated from the formula:
i
i
ꢁ
ꢀ
ꢂ
1/2
2
(
d − d)
i
ꢀ
=
ꢀ
Moreover it was revealed that the water vapor uptake
n
i
(
%) increase increasing the H SO /HNO molar ratio, mean-
2 4 3
ing that the surface hydrophilicity increased in the order
CNTs1:3 > CNTs1:1 > CNTs3:1 > CNTs, in agreement with the increase
of the volatile matter content, i.e. with amount of functional groups
introduced onto the CNTs surface. This fact resulted of great impor-
tance by considering the different wetness of surfaces during liquid
phase reactions.
Transmission electron microscopy (TEM) analysis of CNTs and
CNTsN:S, displayed in Fig. 1, evidences that purified CNTs present
as highly entangled long carbon filaments, with external diame-
ter comprised between 5 and 20 nm (Fig. 1a); regardless the liquid
phase treatment conditions the tubular structure of carbonaceous
material is preserved and that no other carbon structures, such as
amorphous carbon and/or graphite platelets, form.
2
.5. Catalytic reactions
The reactions were carried out in a thermostatted glass reac-
tor (30 mL) provided with an electronically controlled magnetic
stirrer connected to a large reservoir (5000 mL) containing oxy-
gen at 2 atm. The oxygen uptake was followed by a mass flow
controller connected to a PC through an A/D board, plotting a
flow/time diagram. Benzyl alcohol oxidation was carried out in
the presence of cyclohexane or water as solvent (0.0125 mol sub-
strate, substrate/metal = 10,000 (mol/mol), Benzyl alcohol/solvent:
5
◦
0/50 (vol%), 100 C, pO = 2 atm). In the case of cyclohexane as sol-
2
vent, periodic removal of samples from the reactor was performed,
whereas in the case of water solvent, after the end of the reaction
the catalyst was filtered off and the product mixture was extracted
with CH Cl . Recoveries were always 98% ± 3 with this procedure.
The chemical modification of the carbonaceous materials sub-
jected to the oxidative treatment is evaluated by means of TPD
monitoring the CO (m/e = 28) and CO2 (m/e = 44) molecules result-
ing from the decomposition of various oxygen containing groups
such as carboxyl, carbonyl, phenol, lactones present and/or intro-
duced upon oxidation onto the carbon surfaces. In addition to
CO and CO , the evolution of NO (m/e = 30) and SO (m/e = 64)
2
2
For the identification and analysis of the products a GC-MS and GC
(
Agilent Technologies 7820A Gas Chromatograph equipped with
a capillary column, HP-5 30 m × 0.32 mm, 0.25 ␮m Film, made by
Agilent Technologies), were used by comparison of the authen-
tic samples. For the quantification of the reactant–products the
external calibration method was used. Glycerol oxidation was
carried out in the presence of water as solvent: 0.3 M Glycerol
solution, NaOH (glycerol/NaOH = 4 mol/mol) and the Pd catalyst
2
2
molecules, for all the samples is monitored.
The TPD results also allow the identification and quantifica-
tion of the functional groups present on the materials surface by
peak assignment, as described elsewhere [14–16]. TPD spectra are
(
glycerol/metal = 1000 mol/mol) were added (total volume 10 mL).
shown in Fig. 2. CNTs release a very low amount of CO in the range
2
The reactor was pressurised at 3 atm of O2 and thermostatted at
of temperature 400–650 K (Fig. 2a), typical of the decomposition of
the carboxylic groups [14]. A very narrow CO peak is also observed,
at high temperature, likely arising from decomposition of carbonyl
groups (Fig. 2b) [14].
◦
5
0 C. After an equilibration time of 5 min, the reaction was initi-
ated by stirring and samples were taken every 15 min and analysed
by HPLC on a Varian 9010 HPLC equipped with a Varian 9050 UV
(
210 nm) and a Waters R.I. detector in series. An Alltech OA-1000
SEM analysis of catalysts evidences the presence of discrete
bundles with a broad size distribution ((10–100 ␮m). Upon func-
tionalization CNTs is known to agglomerate in a major extent [17].
Upon oxidation with nitric–sulphuric acid mixture CO2 and CO
evolution augments with the increasing of sulphuric acid concen-
tration, indicating an enhancement of functional groups introduced
onto the CNTs surface as the oxidative condition become more pow-
erful, in agreement with the increases of the volatile matter content
(Table 1). Fig. 2a shows that CO2 released from CNTsN:S samples
occurs in a broad range of temperature (400–1000 K) suggesting
that different functionalities such as carboxylic acids, anhydrides
and lactones are introduced [14]. CO desorption profiles also
column (300 mm × 6.5 mm) was used with aqueous H PO 0.1%
3
4
wt/wt M (0.5 mL/min) as the eluent. Samples of the reaction mix-
ture (0.5 mL) were diluted (5 mL) using the eluent. Products were
assigned by comparison with authentic samples.
3
. Results and discussion
3
.1. CNTs characterization
CNTs have been prepared by CVD of isobutane according to the
procedure reported in [12]. In order to introduce functional groups,

A. Villa et al. / Catalysis Today 186 (2012) 76–82
79
Fig. 1. Morphology of purified and acid treated CNTs as monitored by TEM. Photos refer to CNTs (a), CNTs3:1 (b), CNTs1:1 (c) and CNTs1:3 (d), respectively.
dramatically enlarge due to the overlap of different contribution
such as carboxylic anhydride groups, occurring at T < 900 K where
CO and CO2 partly covers, phenols, ethers and carbonyls (Fig. 2b).
The sharp CO peak present in the CNTs profile instead progres-
sively decreases for CNTs3:1 and CNTs1:1 and disappears on the
most oxidized CNTs1:3 likely due to the over-oxidation of carbonyl
functionalities.
NO and SO2 evolution (not shown) is only observed for the
acid treated CNTs. The concentration of the monitored species is
reported in Table 1.
Fig. 2. TPD profiles of CNTs and CNTsN:S samples; CO2 (a), CO (b).

8
0
A. Villa et al. / Catalysis Today 186 (2012) 76–82
Table 2
Supported PdNPs: mean size and distribution.
Code
Nitric–sulphuric (v:v)
TEM
CO-chemisorption
2
Statistical median (nm)
Standard deviation ꢀ
Metallic surface area (m /g)
CNTs
–
5.4
5.6
5.1
4.8
4.9
0.9
1.3
1.0
0.92
1.03
1.04
1.03
CNTs3:1
CNTs1:1
CNTs1:3
3:1
1:1
1:3
NO release occurs at low extent in the range of temperature
00–600 K and is due to the decomposition of the nitro-groups
3.3. Alcohol oxidation
4
introduced by nitration of aromatic rings [22] while SO2 release,
which occurs within 450–800 K, is likely due to the decompo-
sition of thiol and sulfonic acid groups [23,24]. The amount of
SO2 released increases in the order CNTs1:3 > CNTs1:1 > CNTs3:1 in
agreement with the increase of sulphuric acid concentration in the
oxidant mixture (Table 1).
CO/CO2 ratio, usually taken as an indicator of the acid/basic
properties of the carbonaceous materials [18–21] and reported in
Table 1, evidences that CNTs show the higher base character, i.e.
higher CO/CO2 ratio; as result of oxidation by acid mixture the
enhancement of the acidic character of CNTs with the increase of
sulphuric acid concentration occurs. These evidences agree with
the auto-generated pH values (Table 1) that decreases from 6.4 to
Pd on CNTs has been tested in alcohol oxidation using the well
studied benzylic alcohol oxidation evaluating the different impact
of CNT surface properties on the catalytic performances. The activ-
ity of the catalyst has been evaluated in terms of conversion per
2
m
exposed metal per hour and the selectivity as single product
over the alcohol converted has been produced. We choose two
different systems: the first where the alcohol is completely sol-
uble (cyclohexane) (Table 3), the second where benzyl alcohol is
only partly soluble (water) (Table 4). We used a very low amount
of catalyst (10 mg,: substrate/metal = 10,000 (mol/mol)) to ensure
a kinetic regime. In fact by varying the amount of catalyst in the
substrate/metal range 3000–20,000 we observed almost constant
TOF. Moreover maintaining the substrate/metal ratio constant, we
2
2
.8 over the most oxidized CNTs1:3
.
observed a constant “Mol converted/Pd m h” by using different
However it should be also considered that apart the acid func-
loaded catalysts (0.5–2%). According to the latter Madon–Boudart
test [25] we can conclude that internal and external diffusion lim-
itation were avoided.
tionalities introduced upon oxidation on CNTs surface, the presence
of sulphur group can also contribute to the lowering of pH being
thyol and sulphonic acid fairly and strong acid respectively.
When cyclohexane is used as the solvent the activities of the
catalysts follows their hydrophobic properties, being the most
hydrophilic support (CNTs1:3) the less active (Table 3). All the cat-
alysts behaved similarly in terms of selectivity, producing almost
the same amount of benzaldehyde (60–70%) and toluene (25–35%).
As reported by Sumathi et al. [26] and more recently by Hutch-
ings’ group [27,28] toluene constitutes the main by-product. A
disproportionation mechanism of two mols of benzyl alcohol in
a helium atmosphere gives an equimolar amount of benzaldehyde
and toluene [27–29]. The toluene amount decrease as the oxygen
pressure increase and the oxidation pathway prevails producing a
higher amount of benzaldehyde. The ratio between benzaldehyde
and toluene was used to roughly quantify the two different mech-
anisms in which a relevant role was played by the reaction media
or the catalytic surface [27]
3
.2. Catalyst preparation and characterization
We prepared Pd catalyst using the sol immobilization technique
with the aim of obtaining as much as possible similar metal disper-
sion. The sol was prepared in a unique batch and then divided for
four samples: CNTs, CNTs3:1, CNTs1:1, CNTs1:3. Table 2 reported the
mean size and standard deviation (ꢀ) of the PdNPs supported on the
differently treated support. However, as clearly appeared from TEM
images (Fig. 3), deposition of colloidal Pd over pristine CNTs result
in the main formation of agglomerate PdNPs while over CNTsN:S
a
metal dispersion is highly improved. The particle size distribution
did not vary appreciably over oxidized CNTs
Indeed the particle distribution on the support become progres-
sively better as the acidic treatment of the support has been more
Looking at our results (Table 3) we could argued that in the case
of cyclohexane solvent, the role of the support is limited in deter-
mining the preferential reaction route even if its role is important
aggressive, i.e. as the H SO4 content increased.
2
Table 3
Benzyl alcohol oxidation in cyclohexane.
Catalysta
Mol converted/Pd (m2 h)
benzaldehyde (%)
Toluene (%)
Benzylbenzoate (%)
Pd/CNTs
32.4
34.8
21.7
8.4
75
63
65
64
25
37
35
36
0
0
0
0
Pd/CNTs3:1
Pd/CNTs1:1
Pd/CNTs1:3
a
◦
Reaction conditions: substrate/metal = 10,000 (mol/mol), Benzyl alcohol/solvent: 50/50 (vol%), 100 C, pO2 = 2 atm.
Table 4
Benzyl alcohol oxidation in water.
Catalysta
Mol converted/Pd (m2 h)
benzaldehyde (%)
Toluene (%)
Benzoic acid (%)
Pd/CNTs
32.2
34.5
21.7
24.4
74
60
61
>99
25
40
38
0
1
Pd/CNTs3:1
Pd/CNTs1:1
Pd/CNTs1:3
0.5
0.2
0
a
◦
Reaction conditions: substrate/metal = 10,000 (mol/mol), Benzyl alcohol/solvent: 50/50 (vol%), 100 C, pO2 = 2 atm.

A. Villa et al. / Catalysis Today 186 (2012) 76–82
81
Fig. 3. TEM images of Pd/CNTs, Pd/CNTs3:1, Pd/CNTs1:1, Pd/CNTs1:3.
in terms of activity. Indeed, it appeared that the more the sur-
face is hydrophobic the more the catalyst becomes active, i.e. the
hydrophobic reactant (benzyl alcohol) has an easier access to the
active sites.
Using water as the solvent a real biphasic system is obtained
dispersing the alcohol in water. Considering the lower solubility
of oxygen in water than in cyclohexane, it was quite surprising
to obtain almost the same activities in water as in cyclohexane
most hydrophilic one and therefore the one that should be taken
the highest advantage in the contact surface-solvent.
The selectivities did not change considerably for the differ-
ent catalysts (Table 4) remaining almost unchanged for pristine,
Pd/CNTs3:1 and Pd/CNTs1:1 with respect to the values obtained
using cyclohexane as the solvent. However, we noted a progressive
decrease of toluene production with an increase of the hydrophilic-
ity of the support. In the case of Pd/CNTs1:3 an almost total
selectivity to aldehyde was obtained. We could thus conclude that
the beneficial effect of using water as the solvent, in terms of both
activity and selectivity, become effective only when the support
is highly hydrophilic, i.e. showed a high affinity with the solvent.
(
Table 4). This could be plausible only admitting that in water the
elementary steps of the reaction proceeded faster than in cyclohex-
ane. The activity of Pd/CNTs1:3 has been the only one considerably
increased. This is not surprising as the surface of this catalyst is the

8
2
A. Villa et al. / Catalysis Today 186 (2012) 76–82
Table 5
Glycerol oxidation in water.
Catalysta
Mol converted/Pd (m2 h)
Glycerate and tartronate (%)
Oxalate and glycolate (%)
Formate (%)
Pd/CNTs
2.9
1.9
1.2
1.0
85
80
78
49
10
12
10
25
5
8
12
23
Pd/CNTs3:1
Pd/CNTs1:1
Pd/CNTs1:3
a
◦
Reaction conditions: substrate/metal = 1000 (mol/mol), 0.3 M glycerol, 50 C, 4 eq NaOH, pO2 = 3 atm.
Moreover in this latter case the dispropornation mechanism lead-
ing to toluene formation is suppressed.
References
[
[
1] T. Mallat, A. Baiker, Chem. Rev. 104 (2004) 3037.
2] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C.J. Kiely, G.J. Hutchings, Phys.
Chem. Chem. Phys. 5 (2003) 1329.
Following these results we investigated what happened when
also
a hydrophilic reactant is used. We carried out some
experiments using glycerol, a highly hydrophilic reactant, as the
substrate to oxidize. The results are reported in Table 5.
[3] F. Porta, L. Prati, J. Catal. 224 (2004) 397.
[
[
4] S. Demirel-Gulen, M. Lucas, P. Claus, Catal. Today 102–103 (2005) 166.
5] R.W. Cargill, C.L. Young, R. Battino, P.L. Long, H.L. Clever, IUPAC Commission on
Solubility Data, Solubility Data Series 7 (1981) 190–213.
Contrarily to what expected, also in this case the catalyst activi-
ties followed the same trend as before, i.e. the activity decreased by
increasing the hydrophilicity of the support surface. The increased
affinity of the surface for the reactant seems lowering the desorp-
tion of the products thus decreasing the overall rate of the reaction.
We can find a confirmation of this in the product distribution: to
a highly hydrophilic surface (CNTs1:3) corresponded the highest
amount of products derived by C–C bond cleavage (oxalate, gly-
colate, formate) which raised 48% (Table 5). The adsorbed specie
probably due to the high affinity with the surface, desorbed slowly
thus lowering the reaction rate but on the same time having time
to undergo oxidative cleavage.
[6] A.B. Crozon, M. Besson, P. Gallezot, New J. Chem. 22 (1998) 269–273.
[
[
7] T. Mallat, A. Baiker, Catal. Today 19 (1994) 247–284.
8] P. Vinke, D. deWit, A.T.J.W. de Goede, H. van Bekkum, New Developments
in Selective Oxidation by Heterogeneous Catalysis Studies in Surface Sci-
ence and Catalysis, vol. 72, (Elsevier) Amsterdam, The Netherlands, 1992,
pp. 1–20.
[
9] P. Gallezot, Catal. Today 37 (1997) 405–418.
[
10] A. Villa, D. Wang, N. Dimitratos, D. Su, V. Trevisan, L. Prati, Catal. Today 150
2010) 8–15.
(
[11] P. Serp, J.L. Figueiredo (Eds.), Carbon Materials for Catalysis, Wiley-VCH, Wein-
heim, 2009, pp. 45–92.
12] M.G. Donato, S. Galvagno, M. Lanza, G. Messina, C. Milone, E. Piperopoulos, A.
Pistone, S. Santangelo, J. Nanosci. Nanotechnol. 9 (4) (2009) 3815–3823.
[13] S. Santangelo, G. Messina, M.G. Donato, M. Lanza, C. Milone, A. Pistone, J. Appl.
Phys. 100 (10) (2006) 104311–104315.
[
[
[
14] J.L. Figueiredo, M.F.R. Pereira, Catal. Today 150 (1–2) (2005) 2–7.
15] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J. Orfao, Carbon 37 (9) (1999)
4
. Conclusions
1
379–1389.
16] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M. Orfao, Ind. Eng. Chem. Res.
6 (12) (2007) 4110–4115.
17] W. Xia, C. Jin, S. Kundu, M. Muhler, Carbon 47 (2009) 919–922.
[
We prepared Pd catalysts which differ by support surface prop-
erties (in particular hydrophilic/hydrophobic property) and tested
them in liquid phase oxidation reactions. We changed the solvent
4
[
[18] Y.C. Chiang, W.H. Lin, Y.C. Chang, Appl. Surf. Sci. 257 (2011) 2401–2410.
[
[
[
[
19] A. Silvestre-Albero, J. Silvestre-Albero, A. Sepúlveda-Escribano, F. Rodríguez-
Reinoso, Microporous Mesoporous Mater. 120 (2009) 62–66.
20] C. Moreno-Castilla, M.V. Lopez-Ramon, F. Carrasco-Marın, Carbon 38 (2000)
1995–2001.
21] A. Erhan Aksoylu, M. Madalena, A. Freitas, M. Fernando, R. Pereira, J.L.
Figueiredo, Carbon 39 (2001) 175–185.
(
cyclohexane and water) and the reactant (benzyl alcohol and glyc-
erol). The activities of the catalysts depended on the match among
reactant, solvent and support surface.
When the reactant is hydrophobic (benzylic alcohol) as the
solvent (cyclohexane), the support surface hydrophobicity can
improve the activity of the catalyst (but not the selectivity). How-
ever, when the solvent is water a complex triphasic system is
formed and the hydrophilicity of the surface can mediate the con-
tact between reactant. In this case the selectivity is also affected.
Conversely when the reactant is hydrophilic (glycerol) as well
as the solvent (water) it has been showed that the hydrophilicity
of the support can limit the activity of the catalyst by lowering the
desorption of the reaction products.
22] C. Moreno-Castillia, M.N. Lopez-Ramon, F. Carrasco-Marin, Carbon 37 (1999)
1379–1389.
[23] H.T. Gomes, S.M. Miranda, M.J. Sampaio, A.M.T. Silva, J.L. Faria, Catal. Today 151
1–2) (2010) 153–158.
(
[
[
24] A.P. Terzyk, J. Colloid Interface Sci. 268 (2) (2003) 301–329.
25] R.J. Madon, M. Boudart, Ind. Eng. Chem. Fundam. 21 (1982) 430.
[26] R. Sumathi, K. Johnson, B. Viswanathan, T.K. Varadarajan, Appl. Catal. A 172
1998) 15.
(
[
27] M. Sankar, E. Nowicka, R. Tiruvalam, Q. He, S.H. Taylor, C.J. Kiely, D. Bethell,
D.W. Knight, G.J. Hutchings, Chem. Eur J., doi:10:1002/chem.201003484.
[28] M. Sankar, E. Nowicka, P.J. Miedziak, G.L. Brett, R.L. Jenkins, N. Dimitratos, S.H.
Taylor, D.W. Knight, D. Bethell, G.J. Hutchings, Faraday Discuss. 145 (2010)
Concluding we have shown that it is possible to tune the
activity/selectivity of a reaction by accurately matching the
hydrophilicity of the surface depending on the solvent and the
alcohol to be oxidized.
341.
[
29] G. Kovtun, T. Kameneva, S. Hladyi, M. Starchevsky, Y. Pazdersky, I. Stolarov, M.
Vargaftik, I. Moiseev, Adv. Synth. Catal. 344 (2002) 957.