Publised in Microporous and Mesoporous MaterialsVolume 222, 1 March 2016, Pages 5562
Bio-sourced mesoporous carbon doped with heteroatoms (N,S) synthesised using
one-step hydrothermal process for water remediation.
Laura Roldán, Yanila Marco, Enrique García-Bordejé
Instituto de Carboquímica (ICB-CSIC), Miguel Luesma Castán 4, E-50018 Zaragoza,
Spain, jegarcia@icb.csic.es
Abstract
Mesoporous carbon monoliths doped with nitrogen, sulphur or both have been prepared
in one step hydrothermal carbonization from bio-based precursors. Glucose and pyrrole
carboxaldehyde or glucose and thiophene carboxaldehyde aqueous solutions have been
used as precursors for nitrogen and sulphur doped carbon materials, respectively. In the
same hydrothermal process, a salt-templating approach was used to endow the material
with mesoporosity. Additionally, the materials were also pyrolyzed in N2 at 973 K. The
average mesopore size of resulting xerogels is tuned by the used dopant. Both the
doping and mesoporosity enhanced substantially the adsorption of large dye molecules.
Accordingly, these metal-free, cost-effective and sustainable materials are excellent
candidates for liquid phase environmental and energy applications where the dopant
may play a role as catalytic active phase or electronic modulator.
* Corresponding author: Tel.:+ 34 976733977; fax.: +34 976733318 E-mail address: jegarcia@icb.csic.es
1
Keywords
hydrothermal carbon; mesoporosity; heteroatom-doped carbon, nitrogen, sulfur
1. Introduction
The rational design of high performance and cheap nanomaterials for multiple
sustainable energy and environmental applications is extremely urgent but remains
challenging. Carbons doped with heteroatoms (N, S) with tailored chemical
composition are required for some applications in which the incorporation of
heteroatoms showed some benefits such as metal-free catalysis or electrocatalysis. The
doping element affects conductivity, active sites and wettability of the carbon material
which is worth to explore. Top-down processes typically used for the synthesis of
porous carbons (such as physical or chemical activation) does not allow a well-defined
doping with heteroatoms. Hydrothermal carbonization (HTC) of carbohydrates is a
bottom-up process that allows the preparation of carbon materials with tailored
heteroatom content [1]. Nitrogen doped HTC [2] and Nitrogen and Sulphur co-doped
HTC [3,4] were prepared in one step using glucose plus a building block containing the
heteroatom. These materials have been used in several applications such as
electrocatalyst for oxygen reduction reaction [2-4] or for capacitive deionization [5].
Other important aspect is the development of porosity, especially mesoporosity, which
is crucial for the adsorption or catalytic application in liquid phase involving large
molecules (high weight compounds, dyes or biomolecules) [6-8]. Sometimes a
sacrificial template such silica nanoparticles has been used to create mesoporosity, but
this approach requires multiple steps and the use of hazardous reagents [9-16]. It is
beneficial if mesoporosity can be generated in one-pot during hydrothermal
2
carbonization. The use of a special precursor (ovoalbumina) led to a mesoporous
material [3] but the surface area is not very high (below 300 m2 g-1) even for the
pyrolized material. Borax was used to prepare N-doped HTC with mesoporosity[4] and
after pyrolysis at 1173 K the maximum surface area was 427 m2 g-1 formed by pores
between 4-24 nm. A similar approach is the so-called “salt templating” method using
ionic liquids or eutectic salt mixtures as porogen [17]. Using this approach, Fechler et
al. prepared aerogel-like mesoporous bodies by using hypersaline conditions [18]. The
basic concept behind is that the hypersaline conditions stabilize the surface of the asformed primary small nanoparticles (<50 nm) to avoid Ostwald ripening or excessive
particle growth. These primary particles at sufficiently high concentration then turn
collectively unstable, undergoing spinodal phase separation and cross-linking toward
the final porous carbon gels. The more salt is added, the smaller the primary particles
are and hence the higher the surface area is. Thus, this method allows structural control
by varying the salt concentration and salt type. Recycling of the reaction medium in all
these cases is very simple: the salt is washed away with water, filtered, and can be
reused after evaporation of the water. This procedure creates a mesoporous monolithic
bodies in one single step. A variant of this approach is when ZnCl2 is not washed after
hydrothermal carbonization and the material is pyrolised. During pyrolysis, the
remaining ZnCl2 has a second role as an activating agent to develop futher
microporosity [19]. Following this later approach, other authors prepared mesoporous
carbon materials introducing ZnCl2 during hydrothermal carbonization of coconut shell
in a 2:1 weight ratio [20,21]. In that case, the solid was not washed after hydrothermal
synthesis but pyrolyzed it at 1073 K.
As far as we know, salt templating approach has not been used to generate mesoporosity
in heteroatom doped carbons. Herein, we explored how this approach allows tuning the
3
dopant content and mesopore size in one pot. The prepared materials have showed
enhanced performance in the adsorption of dyes even without the need of an activation
step. It is foreseen that these materials will exhibit enhanced performance in liquid
phase reactions of energy and environmental interest.
2. Experimental
For the preparations, anhydrous Glucose (panreac), Pyrrole-2-carboxaldehyde (SigmaAldrich), 2-Thiophenecarboxaldehyde (Sigma-Aldrich) were used.
For the synthesis of N-doped carbon materials, 3 g of anhydrous Glucose, 4,5 g of
ZnCl2, 0,50 g (5.3 mmol) of Pyrrole-2-carboxaldehyde and 1,5 mL of H2O were
thoroughly mixed. The mixture was transferred a glass vessel that was introduced in a
Teflon-lined autoclave and kept at 463 K under autogenous pressure for 19 h.
For the synthesis of S-doped carbon materials, 3 g of anhydrous Glucose, 4,5 g of
ZnCl2, 0,59 g (5.3 mmol) of 2-Thiophenecarboxaldehyde and 1,5 mL of H2O were
thoroughly mixed and treated in the autoclave in the same conditions mentioned above.
For the synthesis of dual N,S-doped carbon materials, 3 g of anhydrous Glucose, 4,5 g
of ZnCl2, 0,50 g of Pyrrole-2-carboxaldehyde, 0,59 g of 2-Thiophenecarboxaldehyde
and 1,5 mL of H2O were thoroughly mixed and treated in the autoclave as mentioned
above.
After hydrothermal synthesis, the solid was introduced in 500 ml of water and kept
overnight. Subsequently, it was filtered, washed with abundant water and dried at 383 K
in an oven.
Surface areas were determined by N2 adsorption at 77 K (BET) using a Micromeritics
ASAP 2020 apparatus, after outgassing for 4 h at 423 K. From the physisorption
measurements with N2, the specific surface area has been calculated by the BET
4
(Brunauer, Emmet, and Teller) theory in the relative pressure range 0.01–0.10 following
standard ASTM-4365, which is applicable to microporous materials. Total pore volume
(VT) was calculated from the amount of N2 adsorbed at a relative pressure of 0.99.
Pore-size distribution was obtained from the desorption branch of the N2 isotherm
according to the BJH method (Barrett-Joyner-Halenda) using the DataMaster V4.0
software and assuming slit pore geometry. BJH model developed in 1951 which is
based on the Kelvin equation and corrected for multilayer adsorption, is most widely
used for calculations of the pore size distribution in the mesoporous and part of the
macroporous range [22]. Microporous volume (Vµ) was estimated by Density
Functional Theory (DFT method), which is applicable to materials with both micro and
mesopores [23]. The model fitted quite well to the isotherm with a standard deviation of
~0.05 cm³ g-1 STP. Alternatively, it was also determined by CO2 adsorption (DubininRadushkevich) at 273 K in the same apparatus, after outgassing under the same
conditions.
CHS analyses were carried out by combustion in a ThermoFlash 1112 elemental
analyzer equipped with a TCD detector. Oxygen analysis was done by direct assay
which involves pyrolysis of the sample at 1343 K in a nickel/carbon bed under a known
He flow. The outlet flow, after passing through a separation column, ends in a TCD
detector. The oxygen content of sample is quantified on the basis of CO analyzed.
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
samples were sputtered with gold previously to measurements. The images were
obtained from the secondary electron signal.
HRTEM of was carried out using a FEI TECNAI F30 electron microscope equipped
with Gatan Energy Filter and cold field emission gun (FEG) operated at 300 kV with
5
1.5 Å lattice resolution. TEM specimens were prepared by ultrasonic dispersion in
ethanol and a drop of the suspension was applied to a holey carbon support grid.
Ex-situ XPS spectra were recorded with an ESCA+ (Omicron) system equipped with
Al/Mg 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. The program used to do curve fitting
of the spectra was CasaXPS after baseline Shirley method.
For the adsorption tests, two dyes with different molecular sizes were employed,
namely, methylene blue and Rhodamine B. To this end, 2 mg of the porous xerogels
was immersed into 12 mL of a 0.05 mg mL−1 aqueous solution of the dye, which was
then gently shaken in an incubator at 20 °C for 72 h to ensure that the adsorption
equilibrium is reached. The adsorbed amounts were then calculated from the difference
between the initial and equilibrium concentration of the dye, as determined with UV–vis
absorption spectroscopy.
3. Results and discussion
The hydrothermal carbonisation produced xerogel bodies which adopt the form of the
mould in which they are synthetized (Figure S1 supplementary material). The
microscopy inspection of the materials (Figure 1) revealed that the prepared materials
have similar morphology, irrespective if they are undoped, N-doped or S-doped. They
are formed by the aggregation of primary particles of size smaller than 20 nm (Figure
2a,b) and the high magnification shows that they have an onion-like morphology. When
the material is pyrolized at 973 K, it is possible to see small nanodomains leaving pores
6
between them (Figure 2 c, d) although the morphology of the particles transforms
slightly from rounded to flake-like.
b
a
4 µm
50 µm
d
c
4 µm
4 µm
Figure 1
7
a
b
20 nm
10 nm
c
d
50 nm
50 nm
Figure 2.
8
Table 1. Textural parameters obtained from N2 (77 K)
sample
BET Area
SN2
Total pore
Micropore
volume (VT)
volume
a
(Vµ)
Mesopore
Percentage
Average
volume
mesopores
pore
(Vm)
(Vm/VT*100) diameter
(Dp)
m2 g-1
cc g-1
cc g-1
cc g-1
%
nm
7
0.014
0.002
0.012
-
<2
HTC-973
541
0.22
0.18
0.04
18
<2
HTC-Zn
373
0.26
0.12
0.14
53
3.5
HTC-Zn-973
520
0.27
0.20
0.07
26
3.5
HTC-N-Zn
503
0.60
0.15
0.44
73
12.0
HTC-N-Zn-973
450
0.48
0.14
0.33
68
8.2
HTC-N-Zn-
1510
0.77
0.54
0.22
28
34
HTC-S-Zn
288
0.56
0.06
0.50
89
32
HTC-S-Zn-973
559
0.66
0.15
0.50
75
2
HTC-N-S-Zn
174
0.27
0.04
0.23
85
32
HTC-N-S-Zn-973 477
0.37
0.14
0.23
62
33
HTC
unwashed-973
a
obtained by DFT model applied to N2 adsorption isotherm
9
Table 2. Textural parameters obtained from CO2 physisorption (273 K) for selected
samples.
Surface area
ultramicropore volume
(SCO2)
(VCO2)
m2g−1
cm3g−1
HTC
142
0.06
HTC-973
608
0.24
HTC-Zn
252
0.1
HTC-Zn-973
511
0.22
sample
Table 1 compiles the textural parameters of different samples. HTC sample, which was
prepared without salt templating, showed negligible surface area measured by N2
physisorption (Figure 3). For HTC calcined at 973 K (HTC-973), the surface area is not
negligible and the isotherm is of type I indicative of microporous material. Thus, the
pyrolysis of HTC develops microporosity accessible to N2. To get further insight about
microporosity, HTC and other selected samples were also characterized by CO2
adsorption (Figure 4 and Table 2) showing a non-negligible surface area of 142 m2g-1.
In samples HTC and HTC-973, SCO2>SN2 and VCO2>VT, indicating the presence of
ultramicorporosity. This is due to kinetic restrictions to N2 diffusion in micropores of
dimensions similar to those of the adsorbate [24,25]. The Pore size distribution
estimated by application of DFT method to CO2 adsorption isotherm determine a mean
pore size of 0.55 nm. On the other hand, in HTC-Zn and HTC-Zn-973, SCO2<SN2 and
VCO2<VT because they are predominantly mesoporous as explained next.
Figure 5, 6, 7 and 8 display the N2 adsorption isotherms and pore size distribution for
salt-templated undoped, N-doped, S-doped and dual (N,S) doped HTC materials,
10
respectively. Unlike to HTC, the adsorption isotherms are of Type IV according to
IUPAC with hysteresis loops, indicative of mesoporous materials. The hysteresis loop is
not completed for the as-synthesised hydrothermal carbons but it is closed for the
pyrolized carbons. This indicates that pyrolized samples have a “permanent porosity”
similar to that of conventional activated carbons while as-synthesised carbons have a
“non-permanent” porosity which suffers some irreversible swelling upon gas
adsorption. The mesoporosity accounts for 50%, 73%, 89% and 85% of total pore
volume and the average pore sizes are 3.5 nm, 12 nm, 34 and 32 nm for un-doped, Ndoped, S-doped, dual N,S-doped salt-templated samples, respectively (Table 1 and
Figures 5,6,7 and 8). Thus, the dopant molecule tunes the average mesopore size. The
pores are widely distributed for doped samples, between 5-30 nm for N-doped and
between 10-100 for S-doped and N,S-doped samples. Upon pyrolizing at 973 K, the
microporosity increases for all the samples, except for N-doped samples. The increase
of microporosity may be rationalized by the release of H2O and COx gases as
temperature rises, leaving pores behind. What is more important is that mesopore
volume remains after pyrolysis and it is only reduced 46%, 25%, 0% and 0% after
pyrolysis of un-doped, N-doped, S-doped, dual N,S-doped samples, respectively.
As a standard procedure, ZnCl2 was thoroughly washed with water after hydrothermal
synthesis. Only one sample was pyrolized without washing and the remaining ZnCl2
acted as a chemical activating agent leading to the sample named as HTC-N-Znunwashed-973. Its textural properties were very different from washed samples,
exhibiting a very high surface (1500 m2g) but with 70% of total pore volume in the
microporous range (dotted line in Figure 4). Accordingly, the pyrolysis without washing
the Zn leads to a reconstruction of the structure, destroying mesoporosity and creating
micropores by activation.
11
Adsorbed volume (cm3g-1 STP)
160
140
120
100
80
60
40
20
0
0
0.2
0.4
0.6
0.8
1
Relative Pressure (P/Po)
Figure 3
Adsorbed volume (cm3g-1 STP)
80
70
60
50
40
30
20
10
0
0
0.01
0.02
Relative Pressure (P/Po)
Figure 4
12
0.03
Figure 5.
Figure 6.
13
Figure 7
Figure 8
In summary, a set of samples with tuneable textural properties have been prepared using
a salt templating approach and different heteroatoms dopants, ranging from
predominantly microporous to mesoporous with different pore sizes. The surface
14
chemistry of the samples is also different and the results of chemical characterization
are shown next.
The main results of chemical characterisation by elemental analysis (e.a) and XPS are
compiled in Table 3. The materials prepared by one-step hydrothermal carbonisation
incorporated the heteroatoms that are initially present in the precursors. In HTC-N-Zn,
the nitrogen content is 5.7 wt% and 4.21 wt% determined by elemental analysis and
XPS, respectively. The similar values determined by both techniques suggest that the N
is distributed quite uniformly throughout the xerogel. For HTC-S-Zn materials, the
amount of sulphur determined by XPS doubles that measured by elemental analysis
suggesting that the surface is enriched in sulphur. For dual doped sample (HTC-N,SZn), the amount of N and S incorporated are in the same range as for the single doped
materials although slightly lower. After pyrolysis, the O content decreases drastically
due to its release as H2O or COx. This decomposition did not affect the mesoporous
structure significantly (vide supra). The N/C ratio remains at similar values after
pyrolysis indicating that N is not removed. In contrast, S/C ratio decreases about 50%
after pyrolysis which may be explained by its release as SO2.
The N 1s peak of of N-doped materials (Figure 9a) was deconvoluted in four
components denoted as pyridine at 398.3±0.1 eV, pyrrolic at 400 ±0.1 eV and
quaternary at 401 ±0.1 eV as described in the literature [26-28] . A transformation of the
type of nitrogen occurs after pyrolysis at 973 K. The as-synthetized HTC-N-Zn contains
pyrrolic nitrogen in agreement with the used precursor. In the pyrolised sample (HTCN-Zn-973), the pyrrolic nitrogen has been transformed to pyridinic and quaternary, i.e
the structure has changed from 5-membered to 6-membered ring. This confirms the
graphitization of the material when it undergoes the thermal treatment in N2 at 973 K, in
15
agreement with XRD measurements of HTC materials (Figure S3 supplementary
information) and with results of other authors [29].
Table 3. Chemical characterisation by elemental analysis and XPS
Elemental analysis
Weight ratios
wt%
C
O
N
S
e.a.
HTC
66.3
27
0.0
0.0
0.0
HTC-Zn
67.6
27
0.0
0.0
HTC-Zn-973
95.3
3.5
0.16
HTC-N-Zn
69.1
22
HTC-N-Zn-973
88.5
N/C
S/C
Zn/C
wt%
wt%
at%
XPS
e.a.
XPS
0.0
0.0
0.0
0.00
0.0
0.0
0.0
0.0
0.01
0.0
0.0
0.0
0.0
0.0
0.09
3.9
0.00
5.7
4.2
0.0
0.0
0.02
5.7
4.7
0.00
5.3
4.0
0.0
0.0
0.11
77
17
4.2
0.00
5.5
3.9
0.0
0.0
0.15
HTC-S-Zn
64.8
23
0.08
8.3
0.0
0.0
12.8
25.6
1
HTC-S-Zn-973
88.7
3.3
0.10
6.7
0.0
0.0
7.6
15.0
2.3
HTC-N-S-Zn
63.7
24
2.8
5.7
4.4
4.2
9.0
17.1
4.0
HTC-N-S-Zn-973
84.9
6.3
3.7
3.6
4.3
4.0
4.2
10.3
5.7
HTC-N-Znunwashed-973
The 2p XPS peak of S-doped materials (Figure 9b) shows a doublet peak at 163.7 and
165 eV which is attributed the the C-S-C bonding derived from tiophenic group [30].
16
There is another peak which intensifies after pyrolysis at 161.7 eV which is ascribable
to ZnS sphalerite [30]. As can be observed in last column of table 3, the Zn content
determined by XPS is negligible for most of the samples, indicating that the washing
step after hydrothermal synthesis is very efficient removing ZnCl2 salt, except for S-
N1
containing-samples
in whichpyridinic
the washing of Zn is not so effective and 1-2 at% of Zn is
detected by XPS. The peak at 161.7 eV confirms that Zn remains bound to S, hindering
pyrrolic
the washing out
N2 of Zn with water, and eventually leading to ZnS upon pyrolysis.
quaternary
N3
Intensity (a.u.)
N4
HTC-N-Zn
Oxidized pyridinic
HTC-Zn-N-973
403
a
400
Binding Energy (eV)
17
398
395
3/2
2p
HTC-S-Zn
1/2
2p
Intensity (a.u)
ZnS
HTC-Zn-S-973
168
b
167
166
165
164
163
162
161
160
Binding Energy (eV)
Figure 9
The materials prepared with different pore texture and dopants were tested in the
adsorption of two dye molecules, namely Methylene blue and Rhodamine B. The
dimensions of these molecules are 1.43 nm × 0.61 nm × 0.4 nm and 1.59 nm × 1.18
nm× 0.56 nm for Methylene blue and Rhodamine B, respectively. The latter one is
slightly larger than the former. The bars in Figure 10 are the dye adsorption capacity
normalized per surface area. The numbers inside the bars are the absolute adsorption
values. In general, the adsorption capacity of Methylene blue is larger than that of
Rhodamine B, which is may be attributed to the slightly smaller size because both are
cationic molecules and the interaction with support should not vary much[31]. The
doped materials adsorb significantly larger amounts of dye before pyrolysis than after
pyrolysis. This can be explained by the higher oxygenated groups content (Table 3) and
lower graphitization degree for the un-pyrolised samples, which render a more
hydrophilic surface for the penetration of aqueous phase and for enhancing interactions
(electrostatic, H-bond) with the dye. The surface area-normalized adsorption follows the
18
same trend for the two dyes HTC-Zn<HTC-N-Zn<HTC-S-Zn. Apparently, the presence
of the dopant is beneficial for the adsorption compared to the undoped sample. It is
probable that this trend is not consequence of the dopant itself but of the different
average pore sizes which is an indirect effect of the dopant heteroatom. It is worth
noting that the sample HTC-N-Zn-unwashed-973 adsorbed comparable dye quantity as
HTC-N-Zn despite having three-fold higher surface area. In addition, the adsorption
kinetics are faster on HTC-N-Zn (Figure S4 supplementary material). This difference
could be attributed either to the higher mesoporosity or richer surface chemistry of
HTC-N-Zn. Studies of ionic strength would help to separate the effect of mesoporosity
and electrostatic interactions but this is beyond the present research. The enhanced dye
adsorption for the mesoporous doped materials demonstrates that they are excellent
candidates to be used in liquid phase applications beyond water environmental
remediation, such as in the field of metal-free catalysis [32-34], electrocatalysis [3,3537] or capacitive deionization [5,38]. Mesoporosity will favour the penetration of the
aqueous phase and the diffusion of voluminous molecules in the pores while doping and
rich surface chemistry will enhance the interactions with the reactant molecules. In the
case of electrocatalytic oxygen reduction reaction, it has been proposed that the reaction
mainly occurs in the mesopores, whereas micropores can be detrimental for the
diffusion of O2 and electrolyte [39]. Moreover, the carbon materials are prepared using
straightforward method and a bio-based precursors, in contrast to the preparation of
other mesoporous carbons which involves multiple steps and hazardous reagents. The
next step is testing the doped mesoporous carbon in oxygen reduction reaction as metalfree electrocatalyst to replace noble metal catalysts.
19
0.6
0.5
HTC-S-Zn
92 mg g-1
0.4
HTC-N-Zn
38 mg g-1
51 mg g-1
HTC-S-Zn-973
42 mg g-1
166 mg g-1
HTC-N-Zn-973
107 mg g-1
HTC-973
0
HTC-N-Znunwashed-973
46 mg g-1
38 mg g-1
0.1
HTC-Zn-973
43 mg g-1
88 mg g-1
0.2
105 mg g-1
159 mg g-1
HTC-Zn
106 mg g-1
123 mg g-1
0.3
13 mg g-1
Dye adsorption capacity per Surface Area (mg m-2)
HTC-N-S-Zn
HTC-N-S-Zn-973
Figure 10
4. Conclusions
Mesoporous carbon materials doped with nitrogen, sulfur or dual doped have been
prepared via hydrothermal carbonisation of glucose and building blocks containing the
heteroatom. In the same single step, mesoporosity has been developed in the materials
by introducing ZnCl2 as easily removable template. The average mesopore size varies
with the type of doping, providing materials with average mesopore sizes of 3.5 nm, 12
nm, 34 and 32 nm for un-doped, N-doped, S-doped, and dual N,S-doped samples,
respectively. After thermal treatment at 973 K in N2 gas, the 5-membered rings
transforms into 6-membered rings due to an increased graphitization but mesoporosity
and N, S content are preserved. The as-prepared hydrothermal carbon has demonstrated
substantially higher surface area-normalized adsorption capacity of bulky dye molecules
than pyrolised materials and carbon with even three-time higher surface area but mainly
20
microporous. These findings emphasizes that this material can be an excellent candidate
for other applications such as electrocatalyst or catalyst in energy and environmental
related applications due to the enhanced diffusion of large molecules in aqueous
environment. The sustainable and simple nature of the preparation technique guarantees
economic and environmental feasibility, a property which graphene and CNT based
materials are lacking yet. Moreover, the bottom-up doping approach allows a high level
of control over the dopant, which is an elusive goal for graphitic carbon materials that
are usually doped following a top-down approach.
5. Aknowledgements.
The financial support of European Commission (FREECATS project, FP7 Grant
agreement nº 280658) from Spanish Ministry MINECO and the European Regional
Development Fund (project ENE2013-48816-C5-5-R), and Regional Government of
Aragon (DGA-ESF-T66 Grupo Consolidado) are gratefully acknowledged.
References
[1] M. M. Titirici, M. Antonietti, Chem.Soc.Rev., 39 (2010) 103-116.
[2] N. Brun, S. A. Wohlgemuth, P. Osiceanu, M. M. Titirici, Green Chem., 15
(2013) 2514-2524.
[3] S. A. Wohlgemuth, R. J. White, M. G. Willinger, M. M. Titirici, M. Antonietti,
Green Chem., 14 (2012) 1515-1523.
[4] S. A. Wohlgemuth, T. P. Fellinger, P. Jaker, M. Antonietti, J.Mater.Chem.A, 1
(2013) 4002-4009.
[5] S. Porada, F. Schipper, M. Aslan, M. Antonietti, V. Presser, T. P. Fellinger,
ChemSusChem, (2015) n/a.
21
[6] K. M. de Lathouder, J. Bakker, M. T. Kreutzer, F. Kapteijn, J. A. Moulijn, S. A.
Wallin, Chem.Eng.Sci., 59 (2004) 5027-5033.
[7] M. Sevilla, P. Valle-Vigón, P. Tartaj, A. B. Fuertes, Carbon, 47 (2009) 25192527.
[8] S. van Donk, A. H. Janssen, J. H. Bitter, K. P. de Jong, Catal.Rev., 45 (2003)
297-319.
[9] N. Brun, K. Sakaushi, L. Yu, L. Giebeler, J. Eckert, M. M. Titirici,
Phys.Chem.Chem.Phys., 15 (2013) 6080-6087.
[10] S. Ikeda, K. Tachi, Y. H. Ng, Y. Ikoma, T. Sakata, H. Mori, T. Harada, M.
Matsumura, Chem.Mater., 19 (2007) 4335-4340.
[11] J. B. Joo, Y. J. Kim, W. Kim, P. Kim, J. Yi, Catal.Comm., 10 (2008) 267-271.
[12] S. Kubo, I. Tan, R. J. White, M. Antonietti, M. M. Titirici, Chem.Mater., 22
(2010) 6590-6597.
[13] M. M. Titirici, A. Thomas, M. Antonietti, Adv.Funct.Mater., 17 (2007) 10101018.
[14] M. M. Titirici, A. Thomas, M. Antonietti, J.Mater.Chem., 17 (2007) 3412-3418.
[15] L. Yu, N. Brun, K. Sakaushi, J. +. Eckert, M. M. Titirici, Carbon, 61 (2013)
245-253.
[16] R. mir-Cakan, P. Makowski, M. Antonietti, F. Goettmann, M. M. Titirici,
Catalysis Today, 150 (2010) 115-118.
[17] N. Fechler, T. P. Fellinger, M. Antonietti, Adv.Mater., 25 (2013) 75-79.
[18] N. Fechler, S. A. Wohlgemuth, P. Jaker, M. Antonietti, J.Mater.Chem.A, 1
(2013) 9418-9421.
[19] F. Cesano, M. M. Rahman, S. Bertarione, J. G. Vitillo, D. Scarano, A. Zecchina,
Carbon, 50 (2012) 2047-2051.
[20] A. Jain, V. Aravindan, S. Jayaraman, P. S. Kumar, R. Balasubramanian, S.
Ramakrishna, S. Madhavi, M. P. Srinivasan, Sci.Rep., 3 (2013) .
[21] A. Jain, S. Jayaraman, R. Balasubramanian, M. P. Srinivasan, J.Mater.Chem.A,
2 (2014) 520-528.
[22] E. P. Barrett, L. G. Joyner, P. P. Halenda, J.Am.Chem.Soc., 73 (1951) 373-380.
[23] R. Evans, P. Tarazona, Physical review letters, 52 (1984) 557-560.
[24] D. Lozano-Castelló, D. Cazorla-Amorós, A. Linares-Solano, Carbon, 42 (2004)
1233-1242.
[25] H. Marsh, W. F. K. Wynne-Jones, Carbon, 1 (1964) 269-279.
22
[26] I. Florea, O. Ersen, R. Arenal, D. Ihiawakrim, C. Messaoudi, K. Chizari, I.
Janowska, C. Pham-Huu, J.Amer.Chem.Soc., 134 (2012) 9672-9680.
[27] T. C. Nagaiah, A. Bordoloi, M. D. Sánchez, M. Muhler, W. Schuhmann,
ChemSusChem, 5 (2012) 637-641.
[28] J. R. Pels, F. Kapteijn, J. A. Moulijn, Q. Zhu, K. M. Thomas, Carbon, 33 (1995)
1641-1653.
[29] L. Yu, C. Falco, J. Weber, R. J. White, J. Y. Howe, M. M. Titirici, Langmuir, 28
(2012) 12373-12383.
[30] R. S. Smart, W. M. Skinner, A. R. Gerson, Surf.Interf.Anal., 28 (1999) 101-105.
[31] C. C. Wang, J. R. Li, X. L. Lv, Y. Q. Zhang, G. Guo, Energy Environ.Sci., 7
(2014) 2831-2867.
[32] Y. Gao, G. Hu, J. Zhong, Z. Shi, Y. Zhu, D. S. Su, J. Wang, X. Bao, D. Ma,
Angew.Chem.Int.Ed., 52 (2013) 2109-2113.
[33] S. Indrawirawan, H. Sun, X. Duan, S. Wang, J.Mater.Chem.A, 3 (2015) 34323440.
[34] H. Watanabe, S. Asano, S. i. Fujita, H. Yoshida, M. Arai, ACS Catal., 5 (2015)
2886-2894.
[35] R. Liu, D. Wu, X. Feng, K. Müllen, Angew.Chem.Int.Ed., 49 (2010) 2565-2569.
[36] M. Sevilla, L. Yu, T. P. Fellinger, A. B. Fuertes, M. M. Titirici, RSC Adv., 3
(2013) 9904-9910.
[37] X. Sheng, N. Daems, B. Geboes, M. Kurttepeli, S. Bals, T. Breugelmans, A.
Hubin, I. F. J. Vankelecom, P. P. Pescarmona, Appl.Catal.B: Environ., 176177
(2015) 212-224.
[38] L. Liu, L. Liao, Q. Meng, B. Cao, Carbon, 90 (2015) 75-84.
[39] X. Wang, J. S. Lee, Q. Zhu, J. Liu, Y. Wang, S. Dai, Chem.Mater., 22 (2010)
2178-2180.
Captions to figures
Figure 1. Representative SEM images of the hydrothermal carbon aerogels: (a) particle
of crushed HTC-Zn xerogel body; (b) HTC-Zn; (c) HTC-N-Zn; (d)HTC-S-Zn
23
Figure 2. Representative TEM image of HTC-N-Zn (a, b), HTC-N-Zn-973 (c) and
HTC-S-Zn-973 (d).
Figure 3. N2 adsorption isotherms and pores size distribution estimated using the BJH
method (inset) for HTC (solid line) and HTC-973 (dashed line).
Figure 4. CO2 adsorption isotherms and pores size distribution estimated using the
density functional theory method (inset) for HTC (solid line) and HTC-973 (dashed
line).
Figure 5. N2 adsorption isotherms (a) and pores size distribution estimated using the
BJH method (b) for HTC-Zn (solid line) and HTC-Zn-973 (dashed line).
Figure 6. N2 adsorption isotherms and pores size distribution estimated using the BJH
method for HTC-N-Zn (solid line), HTC-N-Zn-973 (dashed line) and HTC-N-Znunwashed-973 (dotted line).
Figure 7. N2 adsorption isotherms and pores size distribution estimated using the BJH
method for HTC-S-Zn (solid line) and HTC-S-Zn-973 (dashed line).
Figure 8. N2 adsorption isotherms and pores size distribution estimated using the BJH
method for HTC-N,S-Zn (solid line) and HTC-N,S-Zn-973 (dashed line).
Figure 9. N 1s (a) and S 2p (b) XPS peaks of HTC-N-Zn and HTC-S-Zn, respectively,
before and after pyrolysis
24
Figure 10. Surface area normalized adsorption capacity of dyes at equilibrium:
methylene blue (grey bars), Rhodamine B (white bars). The absolute absorption values
are inside the bars
25