-
The formation of a hydrothermal carbon coating on graphite microfiber
felts for using as structured acid catalyst
Laura Roldan1,2
, Iván Santos1, Sabino Armenise
1, José María Fraile
2, Enrique García-
Bordejé1
1 Instituto de Carboquimica (ICB-CSIC), Miguel Luesma Castán 4, 50018 Zaragoza, Spain
2 Instituto de Síntesis Química y Catálisis Homogénea, Facultad de Ciencias, Universidad
de Zaragoza - C.S.I.C., Pedro Cerbuna 12, 50009 Zaragoza, Spain
Abstract
A hydrothermal carbon layer has been coated on graphite felt using a carbohydrate solution
(glucose, sucrose or starch) as precursor. The coating is uniform in thickness and coverage,
and exhibits excellent adhesion to the substrate. It is highly hydrophilic, very accessible to
reactants and the thickness can be tuned by the synthesis conditions. The coating was
functionalized with sulfonic acid groups and its morphology and surface chemistry
characterized by SEM, TPO-MS, IR-spectroscopy and XPS. The sulfonated material has
been tested in the esterification of palmitic acid with methanol exhibiting significant
activity and stability.
* Corresponding author: Tel.:+ 34 976733977; fax.: +34 976733318 E-mail address: [email protected]
mailto:[email protected]
-
1. Introduction
Material science devoted to valuable carbon materials has gained increasing interest
because of the potential applications as catalyst support, adsorbent, electrodes or gas
storage media for fuel cells or biological applications. Carbon materials have a long history
as catalyst support. Carbon has been used as catalyst support for noble metals in liquid
phase reactions such hydrogenations [1-3] and oxidations [4], or as electrocatalyst [5;6].
Recently, carbon materials have been also applied as metal-free catalyst for oxidative
dehydrogenations of alkanes [7;8] and for different reactions catalysed by bases [9;10] or
acids [11;12].
Many synthetic approaches to prepare carbon materials have been used. To name a few,
carbonisation, laser ablation, chemical vapour deposition, or hydrothermal carbonisation.
Each particular method allows the preparation of carbon materials with different shapes,
texture, crystallinity and chemical composition. An appealing method to prepare carbon
materials is hydrothermal carbonisation (HTC) [13]. This versatile method allows preparing
carbon nanomaterials and composites with different morphologies. Furthermore, it is
simple and sustainable because it requires a low specific energy input and uses biomass as
feedstock instead of fossil fuels.
When HTC method is carried out at low temperature (
-
hence very accessible to reactants. Moreover, these functional groups are amenable to
anchorage of different active phases for catalysis or adsorption. Several approaches have
been used to modulate the morphology of hydrothermal carbon or to increase its specific
surface area. For instance, the use of a sacrificial hard template such as anodized alumina
[15], porous silica beads [16], silica nanospheres [17;18], SBA-15 [19] or latex
nanoparticles [20]. Soft templates such as anionic[21] or nonionic surfactant [22] have been
also used. The approaches mentioned above usually lead to macro and mesopores.
Recently, micropores have also been created by chemical activation [23;24].
Due to the small size of nanospheres, some drawbacks are encountered for application in
adsorption or catalysis. For slurry operation, the separation of the carbon nanospheres can
be troublesome and cost-intensive. On the other hand, if they are used in a packed bed, the
high pressure drop, non-uniform flow distribution, plugging and release to the effluent
preclude its practical implementation. These drawbacks would be overcome by
immobilizing the hydrothermal carbon on macroscopic supports. Furthermore, supporting
the catalyst on structured reactors is highly desired because it has been put forward as a key
enabling factor to implement flow chemistry in organic reactions [25]. As far as we know,
there is no precedent on the macroshaping of hydrothermal carbon into monolithic
macroscopic structures.
Recently, sulfonated carbon materials have been used in esterification and
transesterification reactions with remarkable activity [26-32]. They were prepared via
incomplete carbonization of sugars (glucose), starch or cellulose. These materials are
claimed to be more resistant to desulfonation than sulfonic commercial resins due to
stabilizing effect exerted by the electronegative polycyclic aromatic hydrocarbons.
-
Furthermore, they have shown higher activity in the trans/esterification than many other
solid acid catalysts used in this reaction such as silica supported Nafion, sulfonated zirconia
or niobic acid. Toda and cols. [30] prepared a sulfonated carbon derived from D-glucose
which showed higher activity than conventional solid acid catalyst. This catalyst did not
lose activity, even for repeated reactions at temperatures between 80 and 180 ºC. Goodwin
and cols. [26] prepared a sulfonated carbon that was tested in esterification (acetic acid and
caprilic acid) and transesterification (triacetine, tricapriline and soy oil). It exhibited a
catalytic activity during 1-hour cycle comparable to H2SO4 and superior than the catalytic
activity using Nafion SAC-13. As far as we know, there is only one publication employing
sulfonated carbon microspheres prepared via hydrothermal carbonization as acid catalyst
[33]. The target reaction was the esterification of oleic acid with ethanol. This catalyst
exhibited slightly larger turnover frequencies than H2SO4 and p-toluene sulfonic acid and
much larger than Amberlyst-15.
Herein, the coating of a hydrothermal carbon (HC) layer on a macrostructured support such
as carbon felt (CF) using one-pot synthesis method is described. Several synthesis
conditions have been employed such as pretreatment or not with nitric acid, different
carbohydrate precursor (glucose, sucrose and starch) and different concentrations. The
morphology and surface chemistry of the hydrothermal carbon coated carbon felt has been
characterised by SEM, IR, XPS and TPO-MS. The adhesion strength of the coating layer
has been assessed by ultrasonic tests. Finally, the coated felt has been functionalised with
sulfonic acid groups and tested in esterification of palmitic acid.
-
2. Experimental
Graphite felt was purchased from Carbon Lorraine. It consists of an entangled network of
carbon microfilaments of a few centimetres length and around ten micrometres of diameter.
A 1 cm diameter circular piece was cut from 5 mm-thick graphite felt sheet. The piece of
felt was washed with acetone. In some preparations, it was also treated with nitric acid
(65%) at 80 ºC during 16 hours and washed with distilled water until neutrality. For the
hydrothermal carbon synthesis, a solution of the carbohydrate (glucose, sucrose or starch)
was prepared at different concentrations in 25 ml distilled water. The solution was poured
into the autoclave vessel. Subsequently, the piece of felt was immersed into the
carbohydrate solution. The HNO3-treated felt kept immersed without forcing it. However,
the as-received felt floated on the solution and it was necessary to force the felt inside the
solution by fixing its borders to the autoclave wall.
The autoclave vessel was closed and the hydrothermal synthesis was carried out under
autogenous pressure at 195 ºC for 18 h. After cooling down the autoclave, the felt was
introduced in ethanol and sonicated for 15 min (frequency 40 KHz) to remove loosely
adhere hydrothermal carbon. After drying, the ultrasonic treatment was repeated again.
Finally, the sulfonated materials was washed thoroughly with distilled water and
subsequently with ethanol and dried in an oven at 105 ºC. The carbon felt coated with a
hydrothermal carbon layer was denoted as HC-CF.
The sulfonation of the hydrothermally treated felt was carried out with concentrated
(>96%) sulfuric acid (1 g felt/20 mL H2SO4) at 150ºC under argon atmosphere for 16 h.
-
The resulting graphite felt was then washed thoroughly with hot distilled water and dried
overnight at 105ºC. The sample of HC-CF after sulfonation was denoted as HO3S-HC-CF.
Esterification reaction of palmitic acid with methanol was carried out in a round flask under
reflux using 7.5 mmol of palmitic acid (1.92 g), 75.0 mmol of methanol (3.04 ml), 210 mg
of graphite felt and 2.1 mmol of 1-methylnaphthalene (0.30 g) as internal standard. Small
aliquots were withdrawn at different reaction times and analyzed by means of a gas
chromatograph (HP-5890-II).
SEM analysis was carried out with a microscope SEM EDX Hitachi S-3400 N with
variable pressure up to 270 Pa and with an analyzer EDX Röntec XFlash of Si(Li). The
images were obtained both from the secondary and backscattered signal.
Temperature-programmed oxidation experiments (TPO) were performed in a SETARAM
Setsys Evolution. About 10 mg of sample was heated at 10 ºC/min rate until 700 ºC under
air while monitoring the weight. In parallel, TPO-MS experiments were carried out in the
same conditions in a quartz reactor and the gases were analysed using a quadrupole mass
spectrometer (Balzers) connected in line with the reactor.
For infrared analysis, monolith samples were first ground in an agate mortar to produce fine
powder that was mixed with pure KBr (Merck for spectroscopy) by a weight ratio of 1:100.
About 250 mg of the mixture was used to prepare each pellet by compressing at 10 tonnes
cm-2
for 10 min using a hydraulic press. Infrared spectra were recorded with a Vertex 70
Bruker spectrometer over the wavenumber range of 4000-400 cm-1
with 30 scans being
taken at 2 cm-1
spectral resolution. The spectrum of a pure KBr pellet was used as
background.
-
XPS spectra were recorded with an ESCAPlus Omnicrom system equipped with a Al Kα
radiation source to excite the sample. Calibration of the instrument was done with Ag 3d5/2
line at 368.27eV. All measurements were performed under UHV, better than 10-10
Torr.
Internal referencing of spectrometer energies was made using the dominating C 1s peak of
the support at 284.6 eV and Al 2p, at 74.3eV. The program used to do curve fitting of the
spectra was XPS Peak4.1 and XPS Graph routines after baseline Shirley method.
3. Results and discussion
When hydrothermal synthesis from a glucose solution is carried out in absence of graphite
felt, carbon nanospheres of diameters ranging between 500 and 1000 nm are produced
(figure 1a) in agreement with other works [13]. When hydrothermal synthesis is carried out
after introducing a piece of graphite felt (figure 1b) in the carbohydrate solution, the carbon
felt becomes coated by a uniform layer of hydrothermal carbon with some carbon
nanospheres attached to it (figures 2-4). The variation of synthesis parameters, i.e.
carbohydrate type and concentration and pre and post treatments, led to different
morphologies and loadings of hydrothermal carbon coating, which is explained in detail
below.
-
Figure 1. SEM images of different materials used and synthesised in this work: (a)
hydrothermal carbon nanospheres synthesised in the absence of graphite felt (HC); (b) as-
received graphite felt (CF).
Figure 2. SEM images of carbon felt alter hydrothermal carbon synthesis from three
different concentrations of glucose solution: (a, b) 0.3 M glucose solution ; (c, d) 0.5 M
glucose solution; (e, f) 1M glucose solution.
10 ma 20 mb
10 m 5 m
5 m10 m
10 m20 m
a b
c d
e f
10 m 5 m5 m
5 m5 m10 m10 m
10 m10 m20 m
a b
c d
e f
-
3.1. Effect of the carbohydrate concentration
Figure 2 shows SEM images prepared from glucose parent solutions of three different
concentrations (0.3, 0.5 and 1 M). It is apparent that the amount of produced carbon
nanospheres and layer thickness increase as the carbohydrate concentration increases. From
some SEM images in which the coating is abruptly interrupted, it is possible to measure the
hydrothermal carbon layer thickness, which is 200 nm and ca. 4000 nm for hydrothermal
carbon coating derived from the 0.3 M (figure 2b) and 1 M glucose solutions (figure 2f),
respectively. Thus, a large range of thicknesses can be prepared depending on the
concentration of carbohydrate solution. Nevertheless, for the thinnest coatings, SEM
observations cannot unambiguously prove that the coverage is complete.
500 m500 m 5 m5 m
1000 m 100 m
1000 m 100 ma b
c d
e f500 m500 m 5 m5 m
1000 m 100 m100 m
1000 m1000 m 100 m100 ma b
c d
e f
-
Figure 3. SEM images illustrating the influence of pre-treatment of carbon felt with boiling
HNO3 and post-treatment with ultrasound. (a,b) sample after hydrothermal synthesis using
as-received carbon felt; (c,d) sample after hydrothermal synthesis using HNO3-pretreated
carbon felt; (e,f) sample after post-treatment in ultrasound bath of samples corresponding to
figures (a,b). The synthesis was carried out in all cases with 0.5 M glucose solution
3.2. Effect of the pre-treatment with nitric acid and post-treatment in ultrasound
Figure 3 shows samples after synthesis of hydrothermal carbon using the same 0.5 M
glucose solution but two different carbon felts, namely as-received carbon felt (figure 3a, b)
and carbon felt pre-treated with 65% HNO3 at 80º C during 16 h (figure 3c, d). Comparison
of the two figures shows that HNO3-treated carbon felt yielded higher amount of
hydrothermal carbon spheres than as-received carbon felt. Agglomerates of nanospheres
even clog the macropores between the felt microfibers for HNO3-treated carbon felt (figure
3 c and d). Hydrothermal carbon loading determined by thermogravimetric analysis was 16
wt% and 26 wt% for the as-received and HNO3-treated hydrothermal carbon, respectively.
The higher amount of hydrothermal carbon on the HNO3-treated carbon felt can be
explained because graphite felt becomes more hydrophilic after HNO3 treatment. Therefore,
it is more easily wetted by aqueous carbohydrate solution, which can access readily to the
whole porosity of carbon felt.
Some produced carbon nanospheres are weakly attached to the carbon felt surface or just
entrapped within the macropores of the felt. In order to remove the hydrothermal carbon
that is not well attached, ultrasound treatment was carried out previously to further
-
characterisation. The hydrothermal carbon content determined by thermogravimetric
analysis was 16.1 wt% and 7.3 wt% before and after ultrasound treatment, respectively.
Figure 3(e) and (f) shows SEM images of the hydrothermal carbon coated carbon felt after
ultrasound treatment. Most nanospheres were removed and only the hydrothermal carbon
coating layer remains after ultrasound treatment. Ultrasound treatment is very aggressive
and commonly employed to check coating adhesion [34-36]. After this treatment, only the
well-attached coating remains with some semi-spheres embedded within the coating (figure
3f). Therefore, the adhesion of the remaining hydrochar layer to the graphite felt is
excellent, which underscores the mechanical resistance of coating for a real catalytic
application under flow conditions.
Figure 4. SEM images of hydrothermal carbon coated carbon felt prepared using 0.3 M
solutions of two different carbohydrates: (a) sucrose or (b) starch
3.3. Influence of the carbohydrate precursor type
Hitherto, the shown results correspond only to hydrothermal carbon prepared using glucose
as precursor. Besides glucose, other carbohydrate types have been tested such as sucrose
5 m 5 ma b
-
and starch (figure 4). The morphology of coating surface varies depending on the precursor.
The surface of sucrose-derived carbon is rough, with no embedded nanospheres but sharp
elongated features (figure 4a). In contrast, the surface of starch-derived carbon is smoother
and with rounded features (figure 4b), resembling the morphology of samples prepared
using glucose (figure 3f).
The loading of hydrothermal carbon on graphite felt can be accurately quantified by
gasification under air atmosphere in temperature programmed thermogravimetric analysis
as explained in next section. For the same carbohydrate molar concentration, the
hydrothermal carbon loading follows this increasing order, glucose < starch < sucrose
(figure S1 supplementary information).
3.4. Characterisation of HTC coating by TPO-MS and proposed formation mechanism
Figure 5 shows TPO gravimetric experiments for pure graphite felt, for felt coated with
hydrothermal carbon derived from glucose, starch and sucrose, respectively. All samples
were previously treated by ultrasound to remove looser nanospheres. Graphite felt does not
lose weight until temperatures above 930 K. On the other hand, hydrothermal carbon
coating is less stable to oxidation. It starts to be burnt out already at 500 K and it is
completely burn-out at 850 K. Therefore, the weight loss until 850 K gives an accurate
measure of the hydrothermal carbon loading.
-
Figure 5. Normalized weight loss in temperature programmed oxidation experiments under
air atmosphere for original carbon felt (CF) and carbon felt coated with hydrothermal
carbon (HC-CF) prepared from 0.3 M solutions of the three different types of saccharides
(glucose, starch and sucrose). Inset shows the derivative of weight loss.
Before hydrothermal synthesis the graphite felt is superhydrophobic and it floats on water
surface. After hydrothermal carbon synthesis, the felt remains immersed within water
indicating the functionalisation of the carbon coating (figure S2, supplementary
information). Some preliminary information about surface chemistry can be obtained using
temperature programmed oxidation coupled to mass spectrometry (TPO-MS). The
derivative of weight loss (inset in figure 5) shows two different burn-out rates, which take
place at similar temperatures for all the samples irrespective of the synthesis conditions
(saccharide type and concentration). The slower burn out is centred at ca. 585 K and the
faster one occurs at ca. 750 K. These two rates seem to indicate a heterogeneous
functionalisation along the coating thickness. This is confirmed by the analysis of MS
300 400 500 600 700 800 900
50
60
70
80
90
100
HC(starch)-CF
HC(sucrose)-CF
HC(glucose)-CF
No
rma
lize
d w
eig
h lo
ss (
%)
Temperature (K)
CF
300 400 500 600 700 800 900
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
Derivative o
f w
eig
ht
loss
Temperature (K)
-
signal (figure S3 supplementary information). The slower rate is mainly associated with the
formation of H2O (trace d in figure S3) by reaction of O2 gas with hydrogen on the highly
functionalised external surface. In the period of faster burning-out at ca. 750 K the
formation of CO2 prevails, which is produced by reaction of O2 gas with the less
functionalised inner part in contact with the felt surface. Thus, the innermost part of
hydrochar layer is hydrophobic, which creates a good anchorage with the graphite surface
while the external surface of the layer, which is in contact with the aqueous solution, is
hydrophilic. This agrees with the structure reported for carbon nanospheres, which have a
hydrophobic core and a hydrophilic surface [14;34]. Elsewhere, it was confirmed that the
hydrocarbon produced by condensation of saccharide has a composition based on furan
groups [37].
Based on SEM observations and TPO-MS, we propose the mechanism depicted in figure 6
for the synthesis of HT-carbon coating on graphite felt. Hydrothermal carbon nanospheres
are generated both in the carbohydrate solution and on graphite microfiber surface. As
hydrothermal process proceeds further, HT-carbon that is not in contact with aqueous
carbohydrate solution becomes enriched in carbon by dehydration as in the mechanism
reported for HT-carbon nanospheres [14;38]. When the carbon felt surface is densely
covered by nanospheres, they coalesce to form a HT-carbon layer. This layer has increasing
carbon content, i.e. decreasing functionalisation, towards the direction of the graphite
microfiber. This hydrofobicity gradient enhances the interaction of HT-carbon and graphite
felt surface leading to a good anchorage. As synthesis time increases, coalescence of new
spheres with the HT-carbon layer contributes to its thickening. This mechanism explains
-
the observations of increasing HTC layer thickness as time and carbohydrate concentration
rises.
Figure 6. Scheme proposed for the synthesis of HTC coating on graphite felt. The darkest
contrast of HTC indicates higher carbon content.
3.5. Surface chemistry characterisation by spectroscopic techniques
HTC coatingformation
HTC coatingthickening
Graphite felt
Hydrothermal carbon
4000 3000 2000 1000
HO3S-HC-CF
HC-CF
Wavenumber (cm-1)
CF
Abso
rba
nce
(a
.u.)
s(C=O)s(O=S=O)
C-S
b(C-H)
s(C-H)
s(O-H)
s(C=O)
s(C=C) s(C-OH)
-
Figure 7. Infrared spectra of original carbon felt (CF), carbon felt coated with hydrothermal
carbon (HC-CF) and after subsequent sulfonation (HO3S-HC-CF).
The surface chemistry of the original carbon felt (CF), the carbon felt composite coated
with hydrothermal carbon (HC-CF) and after subsequent sulfonation (HO3S-HC-CF) was
characterised exhaustively by IR and XPS spectroscopy. Figure 7 shows the
characterisation by IR. The original graphite felt is almost featureless indicating the
absence of functionalisation. After hydrothermal synthesis many bands emerge which
indicate a high degree of functionalisation in agreement with the increased hydrophilicity.
The bands are similar for all the used saccharides types and for all the tested synthesis
conditions. Among the most prominent bands, that at 1710 cm-1
can be attributed to C=O
(carbonyl, quinone, ester, or carboxyl) whereas the bands at 1620, 1505 and 1428 cm-1
are
attributed to skeletal C=C vibrations, and the bands at approximately 2900 and 3000–3700
cm-1
correspond to stretching vibrations of aliphatic C-H (methyl and methylene groups)
and O-H (hydroxyl or carboxyl), respectively [39-42]. In HO3S-HC-CF sample, the –
SO2OH groups could be identified with the intense band at 1221 cm-1
, representing the
O=S=O asymmetric and symmetric stretching modes [27;43]. In the low frequency range,
the band at 590 cm-1
is assigned to C–S stretching mode, suggesting the existence of
covalent sulfonic acid groups [27]. It is evident that the H2SO4 treatment at high
temperature enables the intensive covalent sulfonation of hydrothermal carbon.
-
Figure 8. X-ray photoelectron spectra of C1s (a) and O1s (b) peaks for graphite felt (CF),
graphite felt coated with glucose(0,5 M)-derived hydrothermal carbon (HC-CF) and this
latter after sulfonation (HO3S-HC-CF). (c) S2p peak of HO3S-HC-CF and sulfonated
carbon felt (HO3S-CF).
Additionally, X-ray photoelectron spectroscopy (XPS) was used to characterise the surface
functionalities. The detailed XPS C1s and O1s spectra are shown in figure 8. The relative
abundance of each peak after fitting of C1s and O1s region are displayed in tables S1 and
S2 (supplementary information). The C1s core-level spectrum of the carbon felt (CF)
exhibits a dominant peak at 284.6 eV due to carbon atoms of the graphene sheets (C=C, C-
H, C-C). However, the spectrum is remarkably different after coating with hydrothermal
carbon (HC-CF) and after subsequent sulfonation (HO3S-HC-CF). In these later samples,
282 284 286 288 290 292
-C-O
CF
Binding energy (eV)
HC-CF
Inte
sity (
au
)
HO3S-HC-CF
>C=O
-COO
C=C/C-C/C-H
528 530 532 534 536 538
CF
HO3S-HC-CF
Inte
sity (
au)
Binding Energy (eV)
-O-CR=O
>C=O -COOR
C-O-R
-O-RC=O
HC-CF
H2O
C1s O1s
282 284 286 288 290 292
-C-O
CF
Binding energy (eV)
HC-CF
Inte
sity (
au
)
HO3S-HC-CF
>C=O
-COO
C=C/C-C/C-H
528 530 532 534 536 538
CF
HO3S-HC-CF
Inte
sity (
au)
Binding Energy (eV)
-O-CR=O
>C=O -COOR
C-O-R
-O-RC=O
HC-CF
H2O
C1s O1s
174 172 170 168 166 164 162
Binding energy (eV)
HO3S-CF
HO3S-HC-CF
Inte
nsity (
a.u
.)
S2p
a b c
-
the peak-fitting allowed the identification of three additional signals at 285.7, 287.2, and
289.0 eV. They are attributed, respectively, to carbon coordinated to a single oxygen in
hydroxyl groups or ethers (C-O-R), to carbonyl or quinone groups (>C=O), and carbon
having a total of three bonds to oxygen in carboxylic groups (COOH/R) [44-46]. C-S bonds
cannot be distinguished from C-C bonds because of the similarity in electronegativity
between sulfur and carbon [47]. Comparing the intensity of the C1s peaks before and after
sulfonation (table S1), the decrease of the peak at 285.7 eV and the increase of peak at 287
eV could be attributed to the strong oxidising ability of H2SO4 treatment which generates
more carbonyl and carboxylate groups from hydroxyl groups.
The O/C ratio determined by XPS is 0.03, 0.17 and 0.22 for CF, HC-CF and HO3S-HC-CF,
respectively. Therefore, the functionalisation increases progressively in each preparation
step. For curve fitting of the O1s peak, the functional groups shown in figure 9 were
considered, namely, C=O groups at 531.1 eV, carbonyl oxygen atoms in esters, amides,
anhydrides and oxygen atoms in hydroxyls or ethers at 532.2 eV, the ether oxygen atoms in
esters and anhydrides at 533.3 eV, and the oxygen atoms in carboxyl groups at 534.2 eV
[40;42]. Additionally, the band at 535.5 eV is associated to chemisorbed water [42]. This
later band was especially characteristic of the highly hydrophilic sulfonated samples.
The bands in the sulfur 2p region were also analysed. No sulfur band was observed for CF
and HC-CF as expected. Figure 9 shows the S 2p bands of sulfonated as-received carbon
felt (HO3S-CF) and HO3S-HC-CF. The peak at 169 eV is assigned to a higher oxidation
state (+6) of sulfur in SO3H groups, while the peak at 163.5 eV is not present, which is
assigned in the literature to reduced forms of sulfur [39;48]. This shows that all the sulfur
fixed in the hydrothermal carbon coating is associated to sulfonic groups. The S/C ratio
-
determined by XPS was 0.34 wt% and 1.44 wt% for HO3S-CF and HO3S-HC-CF,
respectively. Therefore, the original carbon felt is more difficult to functionalise than the
hydrothermal carbon coated one.
3.6. Sulfonation and testing in esterification reaction
Previously to testing in esterification reaction, the carbon samples were sulfonated to create
the catalytic acid sites. The HT carbon coating was not removed after sulfonation when
glucose and starch were used as precursors. However, sulfonation of the sucrose-derived
composite led to the complete removal of carbon coating. Therefore, using these
sulfonation conditions, a suitable acid catalyst can be prepared from glucose or starch
precursors but not from sucrose precursor. SEM image of glucose-derived HT carbon
(figure 9a) after sulfonation shows that the morphology of the coating and the spheres
deposited on it are apparently unaltered by sulfonation process. In order to explain the
different chemical reactivity upon sulfonation and the different morphology of HT carbon
coating for the several carbohydrate precursors, it is necessary to look into the mechanism
of HT carbon formation. It is reported that the mechanism is a complex cascade of
coexisting reaction paths of furfural polymerization and aldol condensation reactions
involving both furan motifs and decomposition products, such as levulinic acid [37].
Depending on carbohydrates type, different types and proportions of furans
(hydroxyfurfural, furfural) and decomposition products are produced. These different
monomers will have various emulsification behaviours and connecting functions.
Therefore, the polymerization process will differ from one carbohydrate to another. In
-
reference [38], it is shown that hydrothermal carbon spheres prepared from glucose and
starch are smaller and more cross-linked than those prepared from sucrose. The lower
degree of cross-linking of carbon moieties for sucrose-derived coating could account for its
removal upon sulfonation.
Figure 9. SEM images of HT carbon coating (a) after sulfonation and (b) after esterification
reaction of sample prepared from 0,5 M glucose. Image (a) shows clearly that the HT
coating has uniform thickness of about 350 nm.
As a proof of concept, the catalytic activity of HO3S-HC-CF prepared using 0.5 M glucose
solution was tested in the esterification reaction of palmitic acid with methanol. The HO3S-
HC-CF catalyst led to 87% yield after 26 h (Figure 10). The hydrothermal carbon loading
on the felt was very similar before and after reaction (TPO, S4 supplementary information).
The loading is only slightly smaller after reaction due to the removal of some species
desorbing at the highest temperatures in TPO. The coating morphology after reaction
(figure 9b) is also comparable to that before reaction, corroborating the stability of the
10 m 3 ma b
-
hydrothermal carbon coating. Therefore, the HO3S-HC-CF composite was not damaged
even though it underwent intensive stirring during reaction. The direct sulfonation of the
graphite felt without hydrothermal carbon (HO3S-CF) gave no activity in the acid catalysed
reaction. Although the lower functionalisation of this support, i.e. 0.33 wt% sulfur
determined by XPS, may account for a lower activity, the total lack of activity might be
related to the different nature of the acidic sites, aryl-SO3H in the case of felt and alkyl-
SO3H in HC [37], and the corresponding lower pKa of the aromatic sulfonic groups.
A catalyst consisting of sulfonated hydrothermal carbon nanospheres (HO3S-HC) was also
prepared at the same synthesis conditions as HO3S-HC-CF, i.e from 0.5 M glucose
solution, but without introducing the carbon felt in the synthesis autoclave. The sulfur
content determined by XPS was 1.50 wt% and 1.44 wt% for HO3S-HC and HO3S-HC-CF,
respectively. Therefore, the density of sulfonic acid groups in both samples is similar. The
determination of sulphur content in HO3S-HC-CF sample by elemental analysis was not
possible due to difficulty of sample grinding for analysis that entailed irreproducible
measures. HO3S-HC sample could be grinded and a sulphur content of 1.7 wt% was
determined by elemental analysis, which is very coincident with that determined by XPS.
Therefore, sulphur contents determined by XPS is adopted as an approximation of sulfonic
acid groups density for comparison between samples. In the catalytic tests, the amount of
HO3S-HC catalyst was set equivalent to the amount of hydrothermal carbon coating on
HO3S-HC-CF. The conversion to palmitate using HO3S-HC was slightly smaller than that
of HO3S-HC-CF. The initial reaction rate was 38 and 81 mmol g-1
h-1
for HO3S-HC and
HO3S-HC-CF, respectively. Thus, catalytic sites in the sulfonated hydrothermal carbon
-
coating the felt are more efficiently used than the sulfonated carbon nanospheres. The
preparation and catalytic tests were repeated twice, exhibiting similar performance.
HO3S-HC-CF catalyst was used as macroscopic square pieces of the felts (0.5 cm side) in
contrast to the nanospheres (500 nm average size). The macrostructure of the felt has the
advantage of being easier to handle and separate from reaction medium than carbon
nanospheres. In addition, the higher ester yield using the HO3S-HC-CF catalyst
demonstrates that the liquid diffuses easily within the macropores of the felt. Work is
underway studying the stability and reuse of the catalyst.
Figure 10. Esterification of palmitic acid with methanol catalyzed by sulfonated catalyst.
Reaction conditions: 7.5 mmol palmitic acid; 75 mmol methanol; 0.210 g catalyst; under
reflux. HO3S-HC-CF and HO3S-HC were prepared from 0.5 M glucose precursor.
0 5 10 15 20 25 30
0
20
40
60
80
100
HO3S-HC-CF
HO3S-CF
HO3S-HC
Yie
ld o
f m
eth
yl palm
itate
(%
)
Time (h)
-
For continuous operation, the macroporous structure of the felt will enable that the liquid
reactant can flow through the monolithic felt with low pressure drop. Monolithic pieces of
graphite felt can be readily implemented inside microreactor plates [49], thus paving the
way for its use in flow chemistry. The carbon felt scaffolding imparts to the composite
physical reinforcement, versatility to mould in different macroscopic shapes and additional
heat and electric conductive pathways while the hydrothermal carbon layer provides
surface to anchor functionalities as demonstrated for sulfonic groups, showing high
performance in esterification reactions. The high thermal conductivity inherited from the
graphitic framework guarantees isothermal operation and rapid heating/cooling down in
chemical reactions performed at moderate or high temperatures. In addition, the high
electrical conductivity of the graphitic felt broadens the potential application field of
hydrothermal carbon for electrochemical devices.
4. Conclusions
The nanocoating of hydrothermal carbon layer on graphite microfibers felts is reported for
the first time. This macroporous structure is rarely achieved by standard chemistry and
processing of hydrothermal carbon. The method described here is one-pot, energy efficient
and uses biomass as feedstock. The coating is mechanically stable and uniform in thickness
and coverage. Furthermore, the thickness can be tailored depending on the synthesis
conditions down to a thickness of a few hundreds of nanometers. The sulfonation of the
hydrothermal carbon coated felt gave rise to an active and stable catalyst in esterification
reaction.
-
Acknowledgements
This work has been possible thanks to the financial support of Aragón regional government
(PI050/09), Spanish national government (MAT 2008-02365, CTQ2008-05138, Consolider
Ingenio 2010 INTECAT CSD2006-0003) and European Commission (Grant agreement No.
226347).
References
[1] Castillejos E, Debouttiere PJ, Roiban L, Solhy A, Martinez V, Kihn Y, et al. An
Efficient Strategy to Drive Nanoparticles into Carbon Nanotubes and the Remarkable
Effect of Confinement on Their Catalytic Performance. Angew Chem -Int Ed
2009;48(14):2529-33.
[2] Hoffer BW, Crezee E, Mooijman PRM, van Lagneveld AD, Kapteijn F, Moulijn JA.
Carbon supported Ru catalysts as promising alternative for Raney-type Ni in the
selective hydrogenation of D-glucose. Catal Tod 2003;79(1-4):35-41.
[3] Solhy A, Machado BF, Beausoleil J, Kihn Y, Gonçalves F, Pereira MFR, et al.
MWCNT activation and its influence on the catalytic performance of Pt/MWCNT
catalysts for selective hydrogenation. Carbon 2008;46(9):1194-207.
[4] Wang D, Villa A, Spontoni P, Su DS, Prati L. In Situ Formation of Au-Pd Bimetallic
Active Sites Promoting the Physically Mixed Monometallic Catalysts in the Liquid-
Phase Oxidation of Alcohols. Chem - A Eur J 2010;16(33):10007-13.
[5] Centi G, Perathoner S. Problems and perspectives in nanostructured carbon-based
electrodes for clean and sustainable energy. Catal Tod 2010;150(1-2):151-62.
[6] Wu BH, Hu D, Kuang YJ, Liu B, Zhang XH, Chen JH. Functionalization of Carbon
Nanotubes by an Ionic-Liquid Polymer: Dispersion of Pt and PtRu Nanoparticles on
Carbon Nanotubes and Their Electrocatalytic Oxidation of Methanol. Angew Chem -
Int Ed 2009;48(26):4751-4.
[7] Frank B, Zhang J, Blume R, Schlogl R, Su DS. Heteroatoms Increase the Selectivity
in Oxidative Dehydrogenation Reactions on Nanocarbons. Angew Chem -Int Ed
2009;48(37):6913-7.
[8] Zhang JA, Su DS, Blume R, Schlogl R, Wang R, Yang XG, et al. Surface Chemistry
and Catalytic Reactivity of a Nanodiamond in the Steam-Free Dehydrogenation of
Ethylbenzene. Angew Chem -Int Ed 2010;49(46):8640-4.
-
[9] Tessonnier JP, Villa A, Majoulet O, Su DS, Schlogl R. Defect-Mediated
Functionalization of Carbon Nanotubes as a Route to Design Single-Site Basic
Heterogeneous Catalysts for Biomass Conversion. Angew Chem -Int Ed
2009;48(35):6543-6.
[10] Villa A, Tessonnier JP, Majoulet O, Su DS, Schlogl R. Transesterification of
Triglycerides Using Nitrogen-Functionalized Carbon Nanotubes. Chemsuschem
2010;3(2):241-5.
[11] Budarin VL, Clark JH, Luque R, Macquarrie DJ. Versatile mesoporous carbonaceous
materials for acid catalysis. Chem Commun 2007;(6):634-6.
[12] Hara M, Yoshida T, Takagaki A, Takata T, Kondo JN, Hayashi S, et al. A carbon
material as a strong protonic acid. Angew Chem -Int Ed 2004;43(22):2955-8.
[13] Hu B, Wang K, Wu LH, Yu SH, Antonietti M, Titirici MM. Engineering Carbon
Materials from the Hydrothermal Carbonization Process of Biomass. Adv Mater
2010;22(7):813-28.
[14] Sevilla M, Fuertes AB. Chemical and Structural Properties of Carbonaceous Products
Obtained by Hydrothermal Carbonization of Saccharides. Chem - A Eur J
2009;15(16):4195-203.
[15] Kubo S, Tan I, White RJ, Antonietti M, Titirici MM. Template Synthesis of
Carbonaceous Tubular Nanostructures with Tunable Surface Properties. Chem Mater
2010;22(24):6590-7.
[16] Titirici MM, Thomas A, Antonietti M. Replication and coating of silica templates by
hydrothermal carbonization. Adv Funct Mater 2007;17(6):1010-8.
[17] Ikeda S, Tachi K, Ng YH, Ikoma Y, Sakata T, Mori H, et al. Selective Adsorption of
Glucose-Derived Carbon Precursor on Amino-Functionalized Porous Silica for
Fabrication of Hollow Carbon Spheres with Porous Walls. Chem Mater
2007;19(17):4335-40.
[18] Joo JB, Kim YJ, Kim W, Kim P, Yi J. Simple synthesis of graphitic porous carbon by
hydrothermal method for use as a catalyst support in methanol electro-oxidation.
Catal Comm 2008;10(3):267-71.
[19] Titirici MM, Thomas A, Antonietti M. Aminated hydrophilic ordered mesoporous
carbons. J Mater Chem 2007;17(32):3412-8.
[20] White RJ, Tauer K, Antonietti M, Titirici MM. Functional Hollow Carbon
Nanospheres by Latex Templating. J Amer Chem Soc 2010;132(49):17360-3.
[21] Sun X, Li Y. Hollow carbonaceous capsules from glucose solution. J Colloid Inter Sci
2005;291(1):7-12.
-
[22] Fang Y, Gu D, Zou Y, Wu ZX, Li FY, Che RC, et al. A Low-Concentration
Hydrothermal Synthesis of Biocompatible Ordered Mesoporous Carbon Nanospheres
with Tunable and Uniform Size. Angew Chem -Int Ed 2010;49(43):7987-91.
[23] Li M, Li W, Liu S. Hydrothermal synthesis, characterization and KOH activation of
carbon spheres from glucose. Carbohydrate Research 2011;346(8):999-1004.
[24] Sevilla M, Fuertes AB, Mokaya R. High density hydrogen storage in superactivated
carbons from hydrothermally carbonized renewable organic materials. Ener Environ
Sci 2011;4(4):1400-10.
[25] Wegner J, Ceylan S, Kirschning A. Ten key issues in modern flow chemistry. Chem
Commun 2011;47(16):4583-92.
[26] Mo X, Lopez DE, Suwannakarn K, Liu Y, Lotero E, Goodwin J, et al. Activation and
deactivation characteristics of sulfonated carbon catalysts. J Catal 2008;254(2):332-8.
[27] Mo X, Lotero E, Lu C, Liu Y, Goodwin J. A Novel Sulfonated Carbon Composite
Solid Acid Catalyst for Biodiesel Synthesis. Catal Lett 2008;123(1):1-6.
[28] Okamura M, Takagaki A, Toda M, Kondo JN, Domen K, Tatsumi T, et al. Acid-
catalyzed reactions on flexible polycyclic aromatic carbon in amorphous carbon.
Chem Mater 2006;18(13):3039-45.
[29] Takagaki A, Toda M, Okamura M, Kondo JN, Hayashi S, Domen K, et al.
Esterification of higher fatty acids by a novel strong solid acid. Catal Tod
2006;116(2):157-61.
[30] Toda M, Takagaki A, Okamura M, Kondo JN, Hayashi S, Domen K, et al. Green
chemistry - Biodiesel made with sugar catalyst. Nature 2005;438(7065):178.
[31] Wang X, Liu R, Waje MM, Chen Z, Yan Y, Bozhilov KN, et al. Sulfonated Ordered
Mesoporous Carbon as a Stable and Highly Active Protonic Acid Catalyst. Chem
Mater 2007;19(10):2395-7.
[32] Zong MH, Duan ZQ, Lou WY, Smith TJ, Wu H. Preparation of a sugar catalyst and
its use for highly efficient production of biodiesel. Green Chem 2007;9(5):434-7.
[33] Maciá-Agulló JA, Sevilla M, Diez MA, Fuertes AB. Synthesis of Carbon-based Solid
Acid Microspheres and Their Application to the Production of Biodiesel.
Chemsuschem 2010;3(12):1352-4.
[34] Jarrah N, van Ommen JG, Lefferts L. Development of monolith with a carbon-
nanofiber-washcoat as a structured catalyst support in liquid phase. Catal Tod
2003;79-80:29-33.
[35] Stefanescu A, van Veen AC, Mirodatos C, Beziat JC, Duval-Brunel E. Wall coating
optimization for microchannel reactors. Catal Tod 2007;125(1-2):16-23.
-
[36] Valentini M, Groppi G, Cristiani C, Levi M, Tronconi E, Forzatti P. The deposition of
[gamma]-Al2O3 layers on ceramic and metallic supports for the preparation of
structured catalysts. Catal Tod 2001;69(1-4):307-14.
[37] Baccile N, Laurent G, Babonneau F, Fayon F, Titirici MM, Antonietti M. Structural
Characterization of Hydrothermal Carbon Spheres by Advanced Solid-State MAS
13C NMR Investigations. J of Phys Chem C 2009;113(22):9644-54.
[38] Titirici MM, Antonietti M, Baccile N. Hydrothermal carbon from biomass: a
comparison of the local structure from poly- to monosaccharides and
pentoses/hexoses. Green Chem 2008;10(11):1204-12.
[39] Cano-Serrano E, Blanco-Brieva G, Campos-Martin JM, Fierro JLG. Acid-
Functionalized Amorphous Silica by Chemical Grafting-Quantitative Oxidation of
Thiol Groups. Langmuir 2003;19(18):7621-7.
[40] Figueiredo JL, Pereira MFR, Freitas MMA, Orfao JJM. Modification of the surface
chemistry of activated carbons. Carbon 1999;37(9):1379-89.
[41] Zawadzki J, Wisniewski M, Weber J, Heintz O, Azambre B. IR study of adsorption
and decomposition of propan-2-ol on carbon and carbon-supported catalysts. Carbon
2001;39(2):187-92.
[42] Zielke U, Huttinger KJ, Hoffman WP. Surface-oxidized carbon fibers: I. Surface
structure and chemistry. Carbon 1996;34(8):983-98.
[43] Zhou W, Xiao J, Chen Y, Zeng R, Xiao S, Nie H, et al. Sulfonated carbon
nanotubes/sulfonated poly(ether sulfone ether ketone ketone) composites for polymer
electrolyte membranes. John Wiley & Sons, Ltd.; 2010. p. DOI: 10.1002/pat.1666.
[44] Adams L, Oki A, Grady T, McWhinney H, Luo Z. Preparation and characterization of
sulfonic acid-functionalized single-walled carbon nanotubes. Physica E: Low-
dimensional Systems and Nanostructures 2009;41(4):723-8.
[45] Okpalugo TIT, Papakonstantinou P, Murphy H, McLaughlin J, Brown NMD. High
resolution XPS characterization of chemical functionalised MWCNTs and SWCNTs.
Carbon 2005;43(1):153-61.
[46] Xia W, Wang Y, Bergstraber R, Kundu S, Muhler M. Surface characterization of
oxygen-functionalized multi-walled carbon nanotubes by high-resolution X-ray
photoelectron spectroscopy and temperature-programmed desorption. Appl Surf Sci
2007;254(1):247-50.
[47] Chin HC. Preparation and characterization of carbon-sulfur surface compounds.
Carbon 1981;19(3):175-86.
-
[48] Konya Z, Vesselenyi I, Niesz K, Kukovecz A, Demortier A, Fonseca A, et al. Large
scale production of short functionalized carbon nanotubes. Chem Phys Lett
2002;360(5-6):429-35.
[49] Bromley B, Hessel V, Renken A, Kiwi-Minsker L. "Sandwich Reactor" for
heterogeneous catalytic processes: N2O decomposition as a case study. Chem Eng &
Tech 2008;31(8):1162-9.