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: jegarcia@icb.csic.es

mailto:jegarcia@carbon.icb.csic.es

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.