Effect of the surface oxidation of carbon nanotubes on the selective cyclization of citronellal

Article abstract of DOI:10.1016/j.apcata.2016.05.021

Commercial carbon nanotubes have been oxidized with HNO₃, H₂O₂ and air and they have been tested on the cyclization of citronellal. The materials have been characterized by N₂ adsorption/desorption isotherms, transmission electron microscopy, X-ray diffraction, temperature programmed desorption, potentiometry titration and X-ray photoelectron spectroscopy. The HNO₃-treated samples displayed the largest concentration of carboxylic acids which increased by increasing the oxidation time and by a previous ball-milling treatment. The carboxylic acid groups were found to be the active groups on the reaction and the activity increased by increasing its content. The selectivity to isopulegols was higher (98%) over the most active samples (activity 0.20-0.68 mmol/g min), which has been explained by their smaller pore size caused by the oxidation treatment. The stability of the most active CNT sample was also proved by recycling the catalyst in 3 cycles.

Full text of DOI:10.1016/j.apcata.2016.05.021

Accepted Manuscript
Title: Effect of the surface oxidation of carbon nanotubes on
the selective cyclization of citronellal
´
Author: A.B. Dongil L. Pastor-Perez J.L.G. Fierro N.
´
Escalona A. Sepulveda-Escribano
PII:
S0926-860X(16)30274-5
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.05.021
APCATA 15885
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
28-1-2016
13-4-2016
22-5-2016
´
Please cite this article as: A.B.Dongil, L.Pastor-Perez, J.L.G.Fierro, N.Escalona,
´
A.Sepulveda-Escribano, Effect of the surface oxidation of carbon nanotubes
on the selective cyclization of citronellal, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2016.05.021
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Effect of the surface oxidation of carbon nanotubes on the selective
cyclization of citronellal
A.B.Dongil a*, L.Pastor-Pérezb, J.L.G. Fierroc , N.Escalonad,e, A.Sepúlveda-Escribanob,
a
Universidad de Concepción, departamento de físicoquímica, laboratorio de catálisis
por metales, Edmundo Larenas 129, Concepción, Chile.
b
Laboratorio de Materiales Avanzados, Instituto Universitario de Materiales de
Alicante, Departamento de Química Inorgánica, Universidad de Alicante,
Apartado 99, E-03080 Alicante, Spain
c Instituto de Catálisis y Petroleoquímica CSIC, Grupo de Energía y Quimica Sostenible,
C/Marie Curie2, Cantoblanco, Madrid, Spain
dDepartamento de Ingeniería Química y Bioprocesos, Escuela de Ingeniería, Pontificia
Universidad Católica de Chile, Avenida Vicuña Mackenna 4860, Macul, Santiago, Chile.
eFacultad de Químicas, Pontificia Universidad Católica de Chile.
To whom correspondence should be addressed
*Dr. A.B.Dongil., e-mail: adongil@udec.cl, Tel: (56-41) 2207236.

Graphical abstract
O
O
A
OH
OH
363 K, 10 bar
Toluene
98% selectivity
0.68 mmol/g∙ min
Highlights
 Oxidation of commercial carbon nanotubes in liquid media and air was studied.
 The samples were tested on the cyclization of citronellal.
 The previous ball-milling treatment increases the carboxylic acid concentration.
 The activity is ascribed to the carboxylic groups and the selectivity to isopulegol seems
to be related to the pore size.

Abstract
Commercial carbon nanotubes have been oxidized with HNO3, H2O2 and air and they
have been tested on the cyclization of citronellal. The materials have been characterized
by N2 adsorption/desorption isotherms, transmission electron microscopy, X-ray
diffraction, temperature programmed desorption, potentiometry titration and X-ray
photoelectron spectroscopy. The HNO3-treated samples displayed the largest
concentration of carboxylic acids which increased by increasing the oxidation time and
by a previous ball-milling treatment. The carboxylic acid groups were found to be the
active groups on the reaction and the activity increased by increasing its content. The
selectivity to isopulegols was higher (98%) over the most active samples (activity 0.20-
0.68 mmol/g∙min), which has been explained by their smaller pore size caused by the
oxidation treatment. The stability of the most active CNT sample was also proved by
recycling the catalyst in 3 cycles.
Keywords: carbon nanotubes; oxidation; ball-milling; citronellal; isomerization;
isopulegol.

1.
Introduction
Citronellal is a renewable feedstock that can be naturally obtained from a great amount
of natural essential oils such as Cymbopogon nardus or Java citronella, and it can be used
in the chemical industry to replace petroleum in a wide variety of transformations. Among
them, the cyclization of citronellal is a relevant reaction in fine chemistry industry as its
product, isopulegol, is widely employed in flavor and perfume industry for the production
of fragrances and manufacturing menthol, which is an important ingredient in perfumery
and pharmaceuticals products [1].
Several heterogeneous catalysts have been studied for the cyclization of citronellal to
substitute the environmentally unfriendly Lewis acids (ZnCl2, AlCl3, SbCl3, etc.). Among
them, several oxides (SiO2–Al2O3, SiO2) [2,3], zeolites containing Lewis acid centers,
such as Sn and Zr atoms [4] and zirconia-based catalysts [1,5] have been found to be
active and selective to certain extent for the reaction. Both Lewis and Brönsted acids
have been identified as active sites and, although no precise correlation between acidity
and catalytic performance has been established, it has been reported that a proper
modulation of the acidity is needed to obtain an optimal activity while avoiding secondary
reactions such as condensation, etherification or cracking [6]. Besides acidity, the
catalytic performance depends largely on the textural properties of the catalyst and, for
example, a comparison between different zeolites and mesoporous materials found that
H-MCM-41 offered the best results [6].
In this context, carbon materials arise as excellent candidates given the possibility to
control the surface acidity with relatively easy procedures. Moreover, they are chemically
stable favouring the deposition of metal nanoparticles in acid and basic media. Their role
in catalysis has extensively been reported both as a support and as catalyst its own [7,8].

They can react with multiple molecules to incorporate different heteroatoms (O, N, H, Cl,
S, etc.), and the importance of their surface chemistry on a wide variety of reactions has
been reviewed [9,10,11]. Besides the surface chemistry it has been proved that the
morphology may affect the catalytic performance. Though most of the literature reports
the use of activated carbon, mainly due to its high adsorption capacity, for liquid phase
reactions a mesoporous material would offer additional advantages. In this sense, still few
examples can be found in literature where carbon nanotubes have been used as catalyst
its own [12,13,14].
The oxidation of carbon materials and CNT, in particular, has been performed employing
several liquid and gas reagents such as H2SO4, H2O2, HNO3, NaClO, HCl, ozone, air, etc
[15,16]. The type and concentration of surface groups depends on the oxidizer and
functionalization conditions [17,18,19,20]. Among the oxidizers, HNO3 has proved to
incorporate a high concentration of oxygen groups while preserving the CNT structure
[21]. In general, the oxidation with HNO3 and H2O2 produces acidic materials holding
carboxylic acids and, to a smaller extent, lactones, anhydrides and phenol groups.
Nonetheless, the obtained amount of oxygen functionalities is larger when HNO3 is
employed [22]. On the other hand, the treatment under air generates a larger proportion
of carbonyl and ethers which are neutral or may form basic structures, such as quinone
[23].
Therefore, we have explored the effect of different oxidation treatments (HNO3, H2O2
and air) on the surface chemistry of CNT and studied their catalytic performance on the
cyclization of racemic citronellal. The effect of the oxidation time with HNO3 and the
previous ball-milling treatment was also evaluated.

2.
Experimental
2.1 Oxidation of the CNT
High purity multi-wall carbon nanotubes (i.d. 2–6 nm) were supplied by Nanocyl
(Nanocyl 3100, > 95% purity) and submitted to different oxidation treatments. a) CNT
were treated with 50 ml/g HNO3 (65% wt) at 403 K, then they were washed with distilled
water until pH ∼7 and dried at 333 K overnight. The oxidation was performed for 24, 16
and 8 h obtaining the samples labeled as CNToxn24, CNToxn16 and CNToxn8
respectively; b) CNToxa: oxidation of CNT under air flow (30 ml/min) at 673 K for 2 h;
c) CNToxp: oxidation of CNT with H2O2 100 ml/g (30%), at 353 K for 24 h; d ) The CNT
were submitted to ball milling in a planetary stainless steel Spex 800 rpm for 1 h followed
by oxidation as applied for CNToxn8 obtaining the sample CNTmox8; e) Two portions
of the sample CNToxn24 were heat-treated in He (50 ml/min) with a heating ramp of 5
K/min at 373 and 473 K for 1 hour and they were labelled as CNToxn24-373 and
CNToxn24-473.
2.2 Characterization
Conventional TEM analysis was carried out with a JOEL model JEM-210 electron
microscope working at 200 kV and equipped with a INCA Energy TEM 100 analytical
system and a SIS MegaView II camera. Samples for analysis were suspended in methanol
and placed on copper grids with a holey-carbon film support.
X-ray powder diffraction patterns were recorded on a Rigaku diffractometer equipped
with a nickel-filtered CuK1 radiation (= 1.5418˚A), using a 2º·min−1 scanning rate. N2
adsorption-desorption analysis was performed at 77 K on a Micromeritics ASAP 2010
apparatus. Specific surface areas were determined via the BET (Brunauer–Emmett–
Teller) equation, using adsorption data in the relative pressure range of 0.05–0.3, and pore
size distributions were estimated using the BJH model.

The thermogravimetric analyses were performed with a Mettler Toledo
Thermogravimetric TGA/SDTA 851 system. The sample was treated under flowing He
for 2 h and then heated with a heating rate of 10 K·min-1 rate up to 1123 K.
Temperature-programmed desorption (TPD) experiments were carried out in a vertical
tube furnace with He as the carrier gas to thermally decompose the stable surface oxygen
complexes. 100 mg sample was placed in a quartz tube and heated to 383 K for 1 h, and
then the temperature was raised to 1000 K at 10 K min-1.
Acid site concentrations and acid strength measurements of the materials were
determined using a potentiometric method [24], whereby a suspension in acetonitrile
(Merck, 99.9%) was titrated with n-butylamine (Merck, 99%). The variation in electric
potential was registered on a Denver Instrument UltraBasic pH/mV meter.
Photoelectron spectra (XPS) were recorded using an Escalab 200 R spectrometer
equipped with a hemispherical analyser and using non-monochromatic Mg K X-ray
radiation (h = 1253.6 eV). The surface O/C atomic ratios were estimated from the
integrated intensities of C 1s and O 1s lines after background subtraction and correction
by the atomic sensitivity factors. The spectra were fitted to a combination of Gaussian–
Lorentzian lines of variable proportions.
2.3 Reaction
The catalytic performance was tested in a 100 mL stainless steel Parr- type reactor lined
with glass at 363 K and 10 bar of He with a stirring speed of 700 r.p.m. For each reaction
50 mg of catalyst, 30 ml of toluene and 0.16 M of racemic citronellal (Sigma Aldrich ≥
98.0%) was employed. For the recycling studies, 75 mg were initially employed keeping
the same reactant concentration and substrate/catalyst ratio. The catalysts were recovered
by filtration, extensively washed with acetone and dried under vacuum at 323 K for 6

hours and the amounts of reactant and solvent recalculated according to the catalyst mass
recovery. Analyses of reactants and products were followed by gas chromatography mass
spectrometry using a GC–MS instrument (Shimadzu GCMS-QP5050) with chiral-Dex
225, a 30-m column (Supelco), and helium as the carrier gas. Small aliquots of sample
were taken for analysis by gas chromatography (GC). The activity was calculated as mol
converted per minute at 50% conversion and the selectivity was estimated as the sum of
all isopulegol isomers divided by the sum of all products formed.
3. Results
3.1 Characterization
In Fig.1 representative TEM images of CNT, CNToxn24 and CNTmox8 are shown. It
can be observed that CNT presented quite long tubes with a spatial arrangement that
leaves an interstitial space between the nanotubes and that most of the tubes are closed.
The images of CNToxn24 suggested a larger proportion of agglomerated tubes compared
to CNT and, in general, few open-ends could be detected. On the other hand, for the
sample CNTmox8, the images showed that the ball-milling treatment was effective on
cutting the nanotubes, though the material also presented a marked proportion of
agglomerated tubes compared to CNT. Moreover, barely amorphous carbon was found
on the surface of the samples, though the presence of debris formed after nitric acid
treatment cannot be disregard as it has been reported previously [25]. These observations
were extracted from several TEM images and a selection is included in the supplementary
information.
The nitrogen adsorption/desorption isotherms at 77 K of the parent CNT and the modified
samples and are shown in Fig.2. The CNT showed the characteristic shape of mesoporous
materials with a well-defined hysteresis loop, a noticeable increase of nitrogen adsorption
at high relative pressure (0.8–0.9) and barely any nitrogen uptake at low relative pressure.

The isotherms of CNToxa and CNToxp presented similar shape compared to that of the
parent CNT but an increase of nitrogen uptake was observed at low relative pressure and
a small additional hysteresis loop was developed at medium relative pressures (0.3–0.6).
On the other hand, CNTmox8 and CNToxn24 displayed a different profile with much less
pronounced hysteresis loop at higher relative pressures. They have also developed a new
hysteresis loop at medium relative pressure, more evident than for CNToxa and CNToxp
samples, which can be originated by the inner hollow cavity of open-ended CNT created
by the treatments [26]. The estimated BET surface area and pore volumes are shown in
Table 1. The porosity of the CNT originates from the area at the surface of the tubes, the
interstitial space between nanotubes and the inner cavities of the open CNT. As expected,
CNT are non-microporous materials and they possess a high specific surface area and a
total pore volume above 1 cm3 g-1, which is primarily due to adsorption on the external
surface of the tubes. The oxidation treatment for CNToxa and CNToxp seems to have
created micropores, this increasing the surface area. On the other hand, the samples
treated with nitric acid displayed similar values of SBET, despite the clear changes on the
pore volume which can be explained along with the TEM observations. The chemical and
physical treatment may have opened some of the tubes which would increase the surface
area and the mesopores volume. However, as the TEM images showed the tubes on
CNToxn24 and CNTmox8 were more agglomerated than on the CNT sample and this
would in turn decrease the mesopores volume and the SBET. Similar findings have been
reported for other wet oxidation treatments [27]. Moreover, the increase in the V micro
observed for the HNO3-treated samples upon increasing the oxidation time can be
ascribed to the creation of micropores as suggested for the air and H2O2-treated samples.
The graphitic order of the materials was evaluated by XRD, and selected patterns are
shown in Fig. 3. The parent CNT displayed the characteristic reflections at 2= 26.1 º

and at 2= 43.1 º of the (002) and (100) planes of graphite respectively [28]. The XRD
patterns of the oxidized samples are similar to that of the parent CNT, this indicating that
the structure was preserved after the oxidizing and mechanical treatments.
In Figure 4.a-c the TPD profiles of CO2 and CO of selected modified samples are shown.
The thermogravimetry analyses, in the supplementary information Fig. S2, showed that
the as received CNT and CNToxa presented barely any weight loss up to 873 K.
Therefore, the H2O2 and HNO3-treated CNT profiles are shown. In Fig. 4a it can be
observed the evolution of CO2 along the temperature range 373-1073 K. All the samples
displayed three main peaks with maxima at around 523-593 K, 673-743 K and 823-953
K. These results indicate the presence on the surface of chemical species with different
thermal stability, which according to the general criteria proposed in literature would
correspond to the desorption of carboxylic acid, carboxylic anhydrides and lactone groups
respectively [9,13,29]. The desorption peaks appear at lower temperature for CNToxn24,
this indicating a lower thermal stability of those groups, in line with previous works which
also reported that by increasing the oxidation time, the decomposition temperature was
lower [30,31]. On the other hand, the CO profile in Fig. 4b showed a small peak in the
range 673-743 K that matches with the temperature for the second peak observed in the
CO2 profile and corresponds to the decomposition of anhydride groups which are known
to originate both CO and CO2; and a broad peak at higher temperatures with maximum at
953-993 K, this revealing the existence of phenolic and carbonyl/quinone groups
[9,13,30]. The intensity of this latter peak was higher for CNToxn16 than CNToxn24
which suggest a larger proportion of C-OH and C=O groups on the former. Furthermore,
Fig.4a also shows that by increasing the oxidation time on the HNO3-treated samples, the
intensity of the peak at low temperature corresponding to the decomposition of carboxylic
acids increases, as confirmed from the integration, and included in brackets. Moreover,

the desorption at the highest temperature is larger for CNToxn24 than for the other
samples, and the sample CNTmox8 displayed the largest peak of evolved CO2 at low
temperature although the oxidation time was 8 hours. The evolved amounts of CO2 and
CO were estimated from the integration of the area under the curves and from them, the
oxygen content was also calculated. The results, in Table 2, show that CNTmox8 owns
the largest amount of CO2, CO and O, while CNToxp had the lowest concentration of
evolved gases and oxygen. In addition, the oxygen concentration increases upon
increasing the oxidation time in agreement with literature [32] except for the sample
CNTmox8 which displayed the largest concentration among the samples. This is not
surprising considering that the development of oxygen-containing surface groups occurs
predominantly at defect sites on the surface and that the ball-milling treatment had created
a larger proportion of these reactive sites as the microscopy showed [33]. The estimated
amounts of CO2 and CO for the CNT samples are in quite good agreement with the
proposed oxidation mechanism which suggests that the oxidation of carbon materials
proceeds via introduction of carbonyls followed by their conversion into phenols,
carboxylic acids, carboxylic anhydrides and lactones [16]. Although the CO2/CO ratio
has been taken as an indication of the surface acidity, the possible presence of anhydrides
on the samples limits its applicability. Therefore, the proportion of carboxylic acid respect
to the total oxygen amount was calculated and the values, in Table 2, show that they
follow the same trend as those of the total amount of carboxylic acids except for the
sample CNToxp which displayed a larger value.
The surface acidity of the materials was estimated from the potentiometric titration curves
with n-butylamine as the probe molecule. In Table 3 are included the maximum acid
strength of the surface sites (derived from the initial electrode potential, E0) and the total
number of acid sites. The acid strength can be determined according to the criterion

proposed in literature [34]: E0 > 100 mV, very strong sites; 0 < E0 < 100 mV, strong
sites; −100 < E0 < 0 mV, weak sites; E0 < −100 mV, very weak sites. The results
confirmed that the parent CNT holds a basic surface and showed that the H2O2 and HNO3-
treatment produced very strong acid sites on the CNT, except for the sample CNToxn8
which displayed strong acid sites, whereas CNToxa displayed weak acid sites. As long as
the acid density site is concerned, the estimated amount showed that the HNO3-treated
samples displayed similar densities except for the sample previously submitted to milling
treatment which showed the largest value. The lowest density of acid sites was measured
for the CNT, CNToxa and CNToxp. These results are in line with those obtained by TPD
and thermogravimetry, where it is possible to confirm that among the oxidized samples,
CNToxa and CNToxp are the materials with the lowest concentration of oxygen–
containing functionalities decomposing as CO and CO2, while in the opposite, CNTmox8
is that with the highest concentration. Moreover, the acid strength correlates quite well
with the COOH/O ratio estimated by TPD. These results seem to confirm that the creation
of defects sites on CNTmox8, favored the oxidation to carboxylic acid, this increasing
the acidity of this sample.
XPS analyses were carried out to assess for the type of functional groups on the prepared
samples. The relative amounts of the different species, C and O, were calculated from the
corresponding peak areas divided by the sensitivity factors and are shown in Table 4. The
O/C ratios obtained showed a moderate increase in the oxygen content for the CNToxa
and CNToxp samples compared to the nitric-treated samples and that the samples with
the largest O/C ratio were CNTmox8 and CNToxn24. Furthermore, the oxygen
concentration on the samples treated with nitric acid increases upon increasing the
oxidation time, being more pronounced when the oxidation time increases from 8 to 16
hours in quite good agreement with the data obtained by TPD. In addition, the large O/C

ratio for the CNTmox8 sample, despite the treatment time, showed that the previous
milling of the CNT favoured the surface oxidation in line with the TPD results and
previous literature [33].
The O 1s region can be deconvoluted in three components to obtain information on the
nature of the surface oxygen groups: oxygen in carbonyl and quinone groups (531.4 eV);
C-O oxygen in hydroxyl and ether groups (532.6 eV); oxygen in acidic carboxyl groups
(534.6 eV) [35]. According to these contributions the resulting proportion of the oxygen
components are shown in Table 4. The proportion of the different oxygen species also
changes among the samples and it was found that the contribution ascribed to O=C-O
follows the same trend as that obtained from the TPD analyses, with CNTmox8 having
the largest concentration, and followed by the other HNO3-treated samples in decreasing
order of oxidation time.
3.2 Reaction results
The catalysts were tested in the cyclization of citronellal and the evolution of conversion
with time until 240 min. is shown Fig. 5. It was found that only the CNT treated with
nitric acid were active. Among them, the best catalyst in terms of activity and selectivity
was CNTmox8 which reached 100% conversion at 90 min. of reaction. The estimated
activity at 50% conversion is shown in Table 5 and it follows the trend: CNTmox8 >
CNToxn24 > CNToxn16 > CNToxn8.
The selectivity to isopulegol achieved at 100% conversion is shown in Table 5, where it
is shown that the catalysts CNTmox8, and CNToxn24 were 97-98% selective to
isopulegol. The less active catalysts were also the less selective and CNToxn16,
CNToxn8 afforded 78 and 69% selectivity respectively. These selectivity values being
nearly constant with conversion as shown in Fig. 6. The secondary products were
identified as isopulegol ethers, which are generated by condensation of two isopulegol

molecules. In order to assess for the stability of the most active catalyst, CNTmox8 was
employed in successive cycles and it was observed that it preserved relatively well the
activity as the values in Table 5 show.
The activity obtained with CNTmox8 was lower than that reported for the cyclization of
racemic citronellal over H-MCM-41 under similar conditions, though the surface area of
the CNT sample is also significantly lower [6]; and the conversion profile was similar to
that for sulfatated-ZrO2/carbon molecular sieve, despite the milder conditions employed
in the present study [36]. However, the selectivity to isopulegols over CNTmox8 was
higher than the reported in the references which was 64-65%.
The activity results showed that the most active sample was CNTmox8 and that the
activity decreased with the oxidation time, this suggesting that the carboxylic groups play
an important role in the reaction. Thus, to verify that the reaction was catalyzed by the
carboxylic groups, additional experiments with heat-treated CNToxn24 were performed.
Accordingly, two portions of the sample were heated to 373 and 473 K (see Experimental
Section for details) and then they were tested on the reaction. The conversion profile, in
Fig. 5, shows that while the sample treated at 373 K preserved relatively well the activity
compared to CNToxn24, the sample CNToxn24-473 nearly completely lost the activity.
These experiments seem to confirm that the oxygen groups responsible of the activity are
the carboxylic acids. The importance of Brönsted acid sites on the cyclization of
citronellal to isopulegol have already been reported and, for example, over zeolites it was
dependent of the number of Brönsted acids [37,38]. The reaction takes place according to
the following steps as shown in Fig.7: firstly, the carbonyl group is activated by hydrogen
bonding through the interaction with the proton of the surface acid groups. When the
carbonyl group is protonated it increases its electrophilicity, this favoring the nucleophilic
attack. Then, by intramolecular rearrangement a stable carbocation is formed followed by

deprotonation to obtain isopulegol [36, 38]. Our results show a clear dependence of the
activity of the catalysts with the carboxylic acid amount. The catalytic role of carboxylic
acid has already been reported in other reactions such as the alcohol dehydration, which
was found to be dependent of the amount and accessibility of these oxygen groups
[39,40]. Although the role of the carboxylic acids on the catalytic performance of the
cyclization of citronellal is clear, it does not follow a linear trend and the activity obtained
with CNTmox8 increases by a factor of three compared to CNToxn24. The
characterization performed does not suggest any parameter responsible for this higher
value. However, some authors suggested that if a marked proportion of oxygen groups
are adjacent on the surface, this may negatively affect the adsorption of the reactant by
steric hindrance [40].
As long as the selectivity is concerned, the less selective catalysts were CNToxn16 and
CNToxn8 which resulted also in the formation of isopulegol ethers. It has been suggested
that larger pores favored the formation of dehydration products inside the pores as well
as condensation products as diisipulegol ethers [6, 36]. The characterization showed that
the oxidation treatment changed the porosity of the samples, this decreasing the pore
volume for CNTmox8 and CNToxn24 as a consequence of the agglomeration of the
nanotubes. Therefore, a tentative explanation for the higher selectivity obtained with
CNTmox8 and CNToxn24 could be that the smaller pore size limited the formation of
bulkier molecules.
Conclusions
We have oxidized commercial carbon nanotubes by liquid treatment with H2O2 and HNO3
employing different times and oxidation under air. In addition, one of the samples was
ball-milled and then submitted to HNO3 oxidation. The samples were tested in the

cyclization of citronellal, a reaction which is known to be catalyzed by acids, and their
catalytic performances were compared. The characterization studies showed that the
samples treated with HNO3 displayed the largest amount of oxygen on the CNT and that
the proportion of carboxylic acids increased with the oxidation time. The sample
previously milled had the largest amount of carboxylic acid among the CNT samples
which has been ascribed to the creation of more edges on this sample. However, this
sample lost surface area during the milling process due to the agglomeration of the
nanotubes. The samples treated with air and H2O2 presented a low proportion of oxygen
and the CNT sample previously milled displayed the highest amount of oxygen among
the samples. It was found that only the samples treated in HNO3 were active on the
reaction and the activity was related to the presence of carboxylic acids. The selectivity
to isopulegols was high, 97-98%, except for the samples oxidized 8 and 16 hours. This
has been ascribed to the different porosity of the HNO3-treated samples. According to the
characterization, for the samples that presented the highest selectivity, the oxidation
treatment changed their porosity, reducing the pore volume and this may hinder the
formation of secondary products which are bulkier than the isopulegol molecule.
Acknowledgments
Financial support for the present study was received from CONICYT Chile, project
3130483 (Postdoctoral), Generalitat Valenciana, Spain (PROMETEOII/2014/004) is also
gratefully acknowledged.

References
[1] E. J. Lenardao, G.V. Botteselle, F. Azambuja, G.Perin, R. G. Jacob,
Tetrahedron 63 (2007) 6671–6712.
[2] A.F. Trasarti, A.J. Marchi, C.R. Apesteguía, J. Catal. 247 (2007) 1555-165.
[3] A.F. Trasarti, A.J. Marchi, C.R. Apesteguía Catal. Comm. 32 (2013) 62–66.
[4] Z. Yongzhong, N. Yuntong, S. Jaenicke, G.-K. Chuah, J. Catal. 229 (2005)
404–413.
[5] G.K Chuah, S.H Liu, S Jaenicke, L.J Harrison, J. Catal. 200 (2001) 352–359.
[6] P. Mäki-Arvela, N. Kumar, V.Nieminen, R.Sjöholm, T.Salmi, D.Y. Murzin,
J.Catal 225 (2004) 155-169.
[7] F. Rodríguez-Reinoso, Carbon 36 (1998) 159-175.
[8] L.R. Radovic, F. Rodríguez-Reinoso, Carbon materials in catalysis. In:
Thrower PA, editor, Chemistry and physics of carbon, (New York: Marcel
Dekker, 1997) Vol. 25, pp. 243– 363.
[9] J. L.Figueiredo, M.F.R. Pereira, Catal.Today 150 (2010) 2–7.
[10] J.L.Figueiredo, M.F.R. Pereira in: P.Serp J.L. Figueiredo (Eds.) Carbon
Materials for catalysis, John Wiley & Sons, INC, New Jersey, 2009, pp. 177-
206.
[11] P. Serp, B. Machado, Nanostructured Carbon Materials for Catalysis, Royal
Society of Chemistry, Cambridge, 2015.
[12] R. P. Rocha, J.P.S. Sousa, A.M.T. Silva, M.F.R. Pereira, J.L. Figueiredo,
Appl. Catal.B: Environmental 104 (2011) 330–336.
[13] A.G. Gonçalves, J.L. Figueiredo, J. J.M. Órfão, M.F.R. Pereira, Carbon 48
(2010) 4369–4381.

[14] B.Frank, J.Zhang, R.Blume, R.Schlögl, D.Sheng Su, Angew. Chem. Int. Ed. 48
(2009) 6913 –6917.
[15] J.L. Figueiredo, M.F.R. Pereira, M.M.A. Freitas, J.J.M.Orfao, Carbon 37 (1999)
1379-1389.
[16] P.Serp, M.Corrias, P.Kalck, Appl. Catal.A: General 253 (2003) 337–358.
[17] M.L. Toebes, J.M.P. van Heeswijk, J.H. Bitter, A.J. van Dillen, K.P. de Jong
Carbon 42 (2004) 307–315.
[18] I.D. Rosca, F.Watari, M.Uo, T.Akasaka, Carbon 43 (2005) 3124–3131
[19] R.R.N. Marques, B.F. Machado, J.L. Faria, A.M.T. Silva, Carbon 48 (2010) 1515 –
1523.
[20] I.Mazov, V. L. Kuznetsov, I.A. Simonova, A. I. Stadnichenko, A.V. Ishchenko, A.
I. Romanenko, E. N. Tkachev, O.B. Anikeeva, Appl. Surf. Sci. 258 (2012) 6272– 6280
[21] J. Zhang, H. Zou, Q.Qing, Y. Yang, Q. Li, Z. Liu, X.Guo, Z.Du , J Phys Chem B
107 (2003), 3712–3718.
[22] Moreno-Castilla C, Ferro-Garcia MA, Joly JP, Bautista-Toledo I, Carrasco-Marin
F, Rivera-Utrilla J, Langmuir 11 (1995), 4386–4392.
[23] E. Fuente, J.A. Menéndez, D. Suárez, M.A. Montes-Morán, Langmuir, 19 (2003),
p. 3505-3511.
[24] I. Tyrone Ghampson, C. Sepúlveda, R.Garcia, L.R. Radovic, J.L. García Fierro,
W.J. DeSisto, N.Escalona, Appl.Catal. A: Gen. 439–440 (2012) 111–124.
[25] Z.Wang, M.D. Shirley, S.T. Meikle, R. L.D. Whitby, S.V. Mikhalovsky, Carbon,
47, (2009) 73–79.
[26] Y.Q. Hong, H.P.Xiang, B.Shuo, W. Mao-Zhang, C. Hui-Ming. J. Chem
Phys Lett. 345 (2001) 18–24.

[27] B.Xionga, Y.Zhoua, Y.Zhaoa, J. Wanga, X.Chena, R. O’Hayrec, Z. Shao,
Carbon 50 (2012) 192–200.
[28] J.P. Tessonnier, D. Rosenthal, T.W. Hansen, C. Hess, M. E. Schuster, R.
Blume, F. Girgsdies, N. Pfänder, O. Timpe, D. Sheng Su, R. Schlögl, Carbon 47
(2009) 1779–1798.
[29] A.E. Aksoylu, M. Madalena, A. Freitas, M. Fernando, R. Pereira, J.L.
Figueiredo, Carbon 39 (2001) 175–185.
[30] C. Moreno-Castilla, F. Carrasco-Marín, F.J. Maldonado-Hódar, J. Rivera-
Utrilla, Carbon 36 (1998) 145-151.
[31] J. L. Figueiredo, M.F. R. Pereira , M.M. A. Freitas , J.J. M. Órfão, Ind
ChEng Chem Res, 46 (2007) 4110-4115.
[32] I.Mazov, V. L. Kuznetsov, I.A. Simonova, A. I. Stadnichenko, A.V.
Ishchenko, A.I.Romanenko, E.N. Tkachev, O.B. Anikeeva, Appl. Surf. Sci. 258
(2012) 6272– 6280.
[33] F.Liu, X.Zhanga, J.Cheng, J.Tu, F.Kong, W.Huang, C.Chen, Carbon 41
(2003) 2527–2532.
[34] R. Cid, G. Pecchi, Appl. Catal. 14 (1985) 15-21.
[35] A.B. Dongil, B. Bachiller-Baeza, A. Guerrero-Ruiz, I. Rodríguez-Ramos,
A. Martínez-Alonso and J. M. D. Tascón. J. Coll. Interf. Sci., 355 (2011) 179-
189.
[36] G. D. Yadav, J.J. Nair, Langmuir, 16 (2000) 4072–4079
[37] PJ . Kropp, G.W. Breton, S.L. Craig, S.D. Crawford, J. Org. Chem. 60
(1995) 4146.
[38] M. Fuentes, J. Magraner, C. de las Pozas, R. Rogque-Malherbe, Appl. Catal. 47
(1989) 367

[39] F. Carrasco-Marin, A. Mueden, C. Moreno-Castilla, J. Phys. Chem. B 102 (1998)
9239-9244.
[40] G.S. Szymanski, Z. Karpinski, S. Biniak, A. Swiatkowski, Carbon 40 (2002) 2627-
2639.

Figure 1. TEM micrographs of a) CNT, b) CNToxn24, c) CNTmox8.
b)
a)
c)

Figure 2. Nitrogen adsorption/desorption isotherms at 77 K of the materials.
50
40
30
20
10
0
300
250
200
150
100
50
CNT
CNToxp
CNToxa
CNTmox8
CNToxn24
CNToxn16
0
0.0
0.0
0.2
0.2
0.4
0.4
0.6
0.6
0.8
0.8
1.0
1.0
Relative pressure (P/P0)

Figure 3. X-ray diffraction patterns of CNT, CNTox24 and CNTmox8.
CNT
CNToxn24
CNTmox8
10
20
30
40
2
50
60
70
80

Figure 4. TPD analyses a) CO2 and b) CO profiles of CNToxn24 ( ) CNToxn16 (
)
CNToxn8 ( ) CNTmox8 ( ) CNToxp ( x ).
-3
2.0x10-9
1.8x10-9
1.6x10-9
1.4x10-9
1.2x10-9
1.0x10-9
8.0x10-10
6.0x10-10
4.0x10-10
2.0x10-10
0.0
.
2.0x10
a)
1.6x10-3
.
1.2x10-3
.
8.0x10-4
.
-4
.
4.0x10
0.0
373
473
573
673
773
873
973 1073 1173
Temperature (K)
4.0x10-9
b)
-3
.
1 6x10
3.0x10-9
.
-3
-4
-4
.
1 2x10
2.0x10-9
1.0x10-9
0.0
.
8 0x10
.
4.0x10
.
0.0
373
473
573
673
773
873
973 1073 1173
Temperature (K)

Figure 5. Evolution of conversion of citronellal with time.
100
80
60
40
20
0
CNTmox8
CNToxn24
CNToxn16
CNToxn8
CNTox24-373
CNTox24-473
0
50
100
150
200
250
time (min)

Figure 6. Selectivity to isopulegol vs conversion.
100
90
80
70
60
50
CNTmox8
CNToxn24
CNToxn16
CNToxn8
CNToxn24-373
0
20
40
60
80
100
Conversion (%)

Figure 7. Reaction scheme for cyclization of citronellal [38]
CH3
CH3
H3C
OH
OH
O
O
+
H3C
CH3
H3C
CH3
H2C
CH3
Isopulegol
Citronellal
Isopulegol ether

Table 1. Textural properties of the materials
S BET
V micro V meso V total
(cm3/g) (cm3/g) (cm3/g)
Sample
(m2/g)
CNT
300
493
463
295
310
289
301
0.01
0.18
0.16
0.10
0.05
0.03
0.01
1.76
1.46
1.06
0.49
1.04
1.17
0.41
1.77
1.64
1.22
0.59
1.09
1.20
0.42
CNToxa
CNToxp
CNToxn24
CNToxn16
CNToxn8
CNTmox8

Table 2. Density of functional groups decomposing as CO2 and CO estimated by TPD.
CO2
CO
(mmol/g)
O
(mmol/g)
1.5
Sample
COOH/O
(mmol/g)a
0.11
CNToxp
0.52 (0.13)
2.77 (0.72)
1.34 (0.49)
1.31 (0.29)
2.89 (1.13)
0.43
1.37
2.51
1.13
2.15
6.9
5.2
3.7
7.9
0.10
0.09
0.08
0.14
CNToxn24
CNToxn16
CNToxn8
CNTmox8
a
In brackets, estimated amount of carboxylic acid from the deconvolution of the first peak.