CARBON 4 8 ( 2 0 1 0 ) 3 1 5 7 –3 1 6 8

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/carbon

A comparison of physical activation of carbon xerogels with

carbon dioxide with chemical activation using hydroxides

Marı́a S. Contreras a,b, Carlos A. Páez a, Leire Zubizarreta c, Angélique Léonard a,

Silvia Blacher a, Claudio G. Olivera-Fuentes b, Ana Arenillas c,
Jean-Paul Pirard a, Nathalie Job a,*
a
Laboratory of Chemical Engineering, Department of Applied Chemistry, University of Liège, B6a, Sart-Tilman, B-4000 Liège, Belgium
b
TADiP Group, Department of Thermodynamics and Transport Phenomena, Simón Bolı́var University, Caracas 1080, Venezuela
c
Instituto Nacional del Carbon, CSIC, Apartado 73, E-33080 Oviedo, Spain

A R T I C L E I N F O A B S T R A C T

Article history:
Received 10 November 2009
Accepted 30 April 2010
Available online 8 May 2010
Carbon xerogels synthesized with a fixed resorcinol/sodium carbonate molar ratio (R/C)
were physically activated using CO2. The effect of activation temperature and activation
time on the final properties of the activated carbon xerogels was evaluated. The specific
surface area increases from 600 m2 g1 to 2000 m2 g1 and more by increasing the temper-
ature and duration of the activation step. A comparison between physical activation with
CO2 and chemical activation with hydroxides was also performed: it was found that both
processes produce an increase of the micropore volume and specific surface area without
altering the mesoporosity developed during the synthesis. However, chemical activation
can lead to the development of the narrow microporosity mainly whereas, in physical acti-
vation, the widening of the narrow micropores takes place whatever the process
conditions.
 2010 Elsevier Ltd. All rights reserved.

1. Introduction diffusional limitations in adsorption process, like in the case

Porous carbon materials are required in many different appli-
cations, like adsorption in liquid and gaseous phases, cataly-
sis or electrocatalysis. The extent of porosity development
and the characteristics of this porosity are linked to the
source and mode of preparation of the carbon [1]. In particu-
lar, it is possible to modify the porosity, i.e. the inner surface,
pore size and pore volume, via activation procedures. In dy-
namic adsorption processes for instance, both the surface
and the pore size distribution are key parameters to perfor-
mance: the total adsorption capacity of the material is linked
to the surface developed, while kinetics of adsorption de-
pends on the pore size distribution, via diffusion of molecules
within the material pore structure. Microporous materials
with low macropore or mesopore volume may thus induce
of supported catalysis. The presence of large amounts of
meso- and/or macropores allows minimizing diffusional lim-
itations inside carbon supports [2]; in this ambit, porous car-
bon materials with high specific surface area and large
volume of meso-macropore with well-controlled size are
interesting.
Carbon gels produced from resorcinol–formaldehyde (RF)
aqueous gels are texture-controlled porous materials that
can be obtained following different procedures [3,4]. The
preparation mainly consists of three steps: (i) gel synthesis,
involving the formation of a three-dimensional polymer in a
solvent (gelation), followed by a curing period, (ii) gel drying,
where the solvent is removed to obtain an organic gel; and fi-
nally (iii) pyrolysis under an inert atmosphere to form the por-
ous carbon material (carbon gel). Different drying methods

* Corresponding author: Fax: +32 4366 3545.

E-mail address: Nathalie.Job@ulg.ac.be (N. Job).
0008-6223/$ - see front matter  2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbon.2010.04.054
3158 CARBON 4 8 ( 2 0 1 0 ) 3 1 5 7 –3 1 6 8

have been used: supercritical drying [3–6], freeze-drying [4,7– such treatment on the surface chemistry of carbon gels was
9] and evaporative drying [4,10–14]. The materials obtained by not analyzed. So, the present work deals with the study of
these drying methods are called, after pyrolysis, carbon aero- physical activation of highly mesoporous carbon xerogels
gels, cryogels and xerogels, respectively. with CO2 and the influence of the activation conditions on
Published work on RF carbon gels agree that the synthesis the pore texture and surface chemistry of the activated mate-
and drying steps determine the pore size distribution at the rials. The obtained pore textures were then compared to
mesopore and macropore levels while the microporosity is those obtained by chemical activation of carbon xerogels with
mostly developed during the pyrolysis [2–4,10–14]. These car- hydroxides.
bon materials are composed of microporous interconnected
sphere-like nodules, formed during the gel synthesis via a
microphase separation mechanism induced by polymer
2. Experimental
growth [15,16], the size of which is mainly regulated by the
2.1. Preparation of carbon xerogels
synthesis pH. The development of the micropores takes place
within the nodules during the pyrolysis step, but the mass
Organic aqueous gels were synthesized by polycondensation
loss, and thus the final specific surface area, remains rela-
of resorcinol and formaldehyde in water. Sodium carbonate
tively low and constant [11]. For specific applications, and de-
was used as basification agent. Mechanical and textural prop-
spite their higher cost, carbon xerogels proved interesting as
erties of organic xerogels depend on the synthesis variables:
alternative to classical activated carbons prepared from natu-
(i) the resorcinol/formaldehyde molar ratio (R/F), (ii) the resor-
ral sources: the accurate tailoring of the pore texture led to
cinol/sodium carbonate molar ratio (R/C), and (iii) the dilution
increasing significantly the performance of catalytic [2] and
ratio (D = total solvent/reactants molar ratio) [3,4]. In the pres-
electrocatalytic [17] processes. The same could apply to dy-
ent work, D, R/F and R/C were fixed at 5.7, 0.5 (stoichiometric
namic adsorption. However, to increase the adsorption capac-
ratio), and 1000, respectively, in order to obtain micro-meso-
ity, the relatively low inner surface (600 m2 g1) of carbon
porous carbon xerogel with large mesopores (32–45 nm) [4].
gels should be developed via activation.
160.0 g resorcinol (Vel, 99%) and 0.1540 g sodium carbonate
Basically, one can distinguish two different types of activa-
were first mixed with 303.5 mL deionised water under mag-
tion processes [1,18,19]: chemical activation and physical
netic stirring. After dissolution, 263.1 mL of formaldehyde
activation. In chemical activation, the carbon precursor is
solution (Aldrich, 37 wt.% in water; stabilized by 10–15 wt.%
impregnated with a chemical agent and carbonized. The
methanol) were added to the mixture. So as to avoid too large
chemical is typically a Brönsted acid (H3PO4, by far the most
temperature gradients during gelation and ageing, the solution
used in industry), a Lewis acid (ZnCl2) or a strong base (KOH);
was divided into three equal parts and poured into sealable
note that many other activation agents of these types are
flasks of 250 mL. Gelation and ageing were then performed at
found in the literature, but not used industrially (e.g. NaOH,
70 C in an oven during 72 h. Organic xerogels were obtained
AlCl3, MgCl2, K2CO3, etc.). Then, the raw material is carbonized
by vacuum drying without any sample pre-treatment, i.e.
at temperature ranging from 450 to 900 C). In physical activa-
without solvent exchange prior to drying. The unsealed flasks
tion, the precursor carbonaceous material is usually contacted
were simply kept at 60 C and the pressure was progressively
with CO2, steam or a combination of both during thermal treat-
decreased in two days from 105 Pa to 103 Pa. The samples were
ment. Oxidation by the activating gas significantly increases
then heated to 150 C at 103 Pa during 12 h.
the microporosity of the final product. Although residues
After drying, the gels were pyrolysed at 900 C in
may remain in the final product, chemical activation is
4.1 · 103 mol min1 flowing nitrogen in a horizontal tubular
preferred over physical activation owing to the lower tempera-
furnace. The heating program included the following sequen-
tures and shorter time needed for activating the material.
tial steps [11]: (i) ramp at 1.7 C/min to 150 C and hold for
In previous work [20,21], carbon xerogels with high surface
15 min; (ii) ramp at 5 C/min to 400 C and hold for 60 min;
area (up to 1800 m2 g1) were obtained by chemical activation
(iii) ramp at 5 C/min to 900 C and hold for 120 min; and (iv)
with alkali hydroxides. The method proved efficient for
cool slowly to room temperature. Note that the pyrolysis pro-
increasing the specific surface area without altering the
cedure can be simplified: tests have shown that the process
meso-macropore texture, and it was concluded that surface
can be reduced to one heating ramp and one temperature
development occurs through the appearance of new microp-
step without damaging the final pore texture.
ores, formed within the carbon nodules by the alkali. How-
ever, the pore texture development may differ from one
activation process to another [18] and, to our knowledge, no 2.2. Activation of carbon xerogels
study comparing the activation of carbon xerogels using var-
ious processes was previously published. In fact, few data The pyrolyzed samples were crushed and sieved between 0.5
concerning the activation of carbon gels with CO2 are avail- and 1.0 mm. Note that crushing carbon xerogels does not
able in the literature. In 2000, Lin and Ritter [10] activated a modify their pore texture. Indeed, gas adsorption performed
mainly microporous carbon xerogel with CO2 under very spe- on both monoliths and powders with granulometry down to
cific conditions. In 2006, Fairén-Jiménez et al. [22] reported the 30–50 lm lead to the same result [2,17]: after crushing, one ob-
activation of a carbon aerogel prepared with paratoluene sul- tains the so-called ‘‘micromonoliths’’ that retain the pore tex-
fonic acid as pH modifier. However, the effect of CO2 activa- ture of the pristine non-crushed material. The samples were
tion strongly depends on the pristine carbon properties, crushed and sieved for two reasons: (i) to avoid CO2 concen-
such as its structure and composition [23], and the effect of tration gradients within the particles that could lead to
 CARBON 4 8 ( 20 1 0 ) 3 1 5 7–31 6 8
inhomogeneous activation (the size of the carbon particles
influences the development and the characteristics of the
porosity [23]) and (ii) because their final use (adsorption in li-
quid media) required small particles of homogeneous size.
In order to study the influence of the activation tempera-
ture, Tact, and activation time, tact, i.e. the duration of the acti-
vation step, the samples were then activated in the horizontal
tubular furnace used for pyrolysis. The unactivated carbon
xerogel was heated at 5 C/min from room temperature to
the chosen activation temperature under a constant nitrogen
flow (4.1 · 103 mol min1). Once the temperature was
reached, the nitrogen flow was replaced by a constant flow
of pure CO2 (4.1 · 103 mol min1) for activation. Two series
of activated carbon xerogels were obtained by using two dif-
ferent activation temperatures, Tact, equal to 800 C and
900 C. The temperature and gas flow were left constant
and, within a series, the activation time, tact, was modified.
After activation with CO2, the nitrogen flow was restored,
helping cooling down the samples in inert atmosphere. For
the samples activated at 800 C, tact was chosen equal to 4,
16 and 24 h. In the case of samples activated at 900 C, tact
of 4, 8 and 16 h were selected.
2.3. Sample designation
The samples obtained in this work are labelled as follows. The
first two letters refer to the type of precursor: CX for carbon
xerogels. These letters are followed by the activation temper-
ature. The last number refers to the activation time. For
example, sample CX-900-4 was activated with CO2 at 900 C
during 4 h.
2.4. Sample characterization
2.4.1. Pore texture characterization
The pore texture of both pyrolyzed and activated carbon xero-
gels was characterized by physical adsorption of nitrogen;
adsorption–desorption isotherms were obtained at 196 C
using a Sorptomatic Carlo Erba 1900 device. Before nitrogen
adsorption, the samples were outgassed for 24 h (100 C,
103 Pa). The analysis of the isotherms provided the BET spe-
cific surface area, SBET, the micropore volume calculated by
the Dubinin–Radushkevich equation, VDUB-N2 , and the pore
volume calculated from the adsorbed volume at saturation,
Vp. The minimum and maximum mesopore diameters, dp,min
and dp,max, i.e. the limit under which smaller pores represent
5% and 95% of the mesopore volume, respectively, were de-
duced from mesopore size distribution calculated by the
Broekhoff–de-Boer method using the adsorption branch of
the isotherm [11].
For samples containing micropores and mesopores only,
the isotherms of which display a plateau at saturation, the
nitrogen adsorption technique is sufficient for determining
the total void volume Vv; consequently, Vv and Vp can be con-
sidered equal. Since a plateau was observed at saturation for
all the samples, N2 adsorption should provide an accurate va-
lue of the total pore volume. However, measurements of
micropore volume by nitrogen adsorption might be not accu-
rate enough for samples containing pores with a width lower
than 0.7 nm: adsorption of nitrogen in such small pores is
3159
hampered by very slow diffusion, and several weeks may be
required to obtain accurate adsorption isotherms [24]. So,
the use of carbon dioxide adsorption at 0 C has been estab-
lished as an effective procedure for appropriate textural char-
acterization of the narrow micropores [24,25]. In this case,
the higher adsorption temperature (0 C instead of 196 C)
results in a larger kinetic energy of the CO2 molecules,
which are thus able to reach the narrow microporosity de-
spite slightly bigger in size than the molecules of N2. The
combination of N2 and CO2 adsorption provides complemen-
tary and relevant information about the whole micropore
range.
Carbon dioxide adsorption isotherms of samples were
thus determined at 0 C in a TriStar 3000 from Micromeritics.
The Dubinin–Radushkevich method [26] was applied to the
carbon dioxide adsorption isotherms in order to obtain the
narrow micropore volume, VDUB-CO2 , and the carbon-CO2 inter-
action energy, E0-CO2 . The corresponding micropore specific
surface area, SCO2 , was evaluated from:
VDUB-CO2
SCO2 ¼ 2 ð1Þ
L
where SCO2 and VDUB-CO2 are expressed in m2 g1 and m3 g1,
respectively, and L (m) is the micropore size calculated either
from Stoeckli and Ballerini [27] for E0-CO2 values between 20
and 42 kJ/mol, or from the Dubinin equation for lower values
of E0-CO2 [25]. Finally, the skeletal density of the carbons, qs,
was obtained by helium pycnometry using a pycnometer
Micromeritics Accupyc 1330. The bulk density, qbulk, was cal-
culated by:
1
qbulk ¼ ð2Þ
VV þ q1s
Following previous studies [2,4], the bulk density values
obtained by this method combining nitrogen adsorption and
helium pycnometry are similar to those obtained by direct
bulk density measurement using mercury pycnometry.
2.4.2. Surface chemistry characterization
Various oxygen surface groups have been identified on car-
bon surfaces [28–30]. Experimental techniques can be used to
characterize these functional groups, such as Temperature-
Programmed Desorption (TPD) of CO and CO2 [29,30] or Point
of Zero Charge (PZC) measurements [31]. TPD was carried out
with an automatic analyzer Micromeritics AutoChem II mon-
itoring the amount of desorbed CO, nCO, and CO2, nCO2 , with an
Onmistar Pfeiffer mass spectrometer. The temperature was
increased at 5 C/min from room temperature up to 1000 C;
the argon flow rate was fixed at 2.05 · 103 mol min1. The
PZC of non-activated and activated carbon xerogels, i.e. the
pH value at which the electrical charge density on the support
surface is zero, was determined by the method of Park and
Regalbuto (equilibrium pH at high loading, EpHL) [31]: the por-
ous solid was soaked in water solutions of various starting pH
and, after stabilization, the pH was measured again. The PZC
of the solid corresponds to a plateau in the final pH vs. initial
pH plot. The equilibrium pH of the carbon xerogels was mea-
sured over a wide range of initial pH values (from pH = 2.0 to
3160 CARBON 4 8 ( 2 0 1 0 ) 3 1 5 7 –3 1 6 8
12.4), adjusted using HNO3 or NaOH solutions, in 20 mL solu-
tion. A critical operating variable when comparing samples
with different surface areas is the total material surface area
in solution, i.e. the surface loading, SL (m2 L1). SL was kept at
104 m2 L1, in line with previous studies [31,32].
3. Results
3.1. Textural properties of carbon materials
Table 1 provides the activation conditions (activation tempera-
ture, Tact, and activation time, tact), the burn-off degree, Dm/m,
and the average specific gasification rate, rs , calculated by:
mi  mf
Dm=m ¼ 100 ð3Þ
mi
rs ¼
Dm=m 1 p /p 0
 ð4Þ
100 tact
where mi denotes the initial sample mass, and mf the sample
mass at the end of tact.
For the samples activated at 800 C with activation time
ranging from 4 to 24 h, Table 1 shows that the burn-off degree,
Dm/m, increases from 4% (CX-800-4) to 14% (CX-800-24), and
the average specific gasification rate, rs , decreases from
0.010 h1 (CX-800-4) to 0.006 h1 (CX-800-24). When Tact in-
creases up to 900 C, Dm/m increases from 18% (CX-900-4) to
60% (CX-900-16); in parallel, rs decreases from 0.045 h1
(CX-900-4) to 0.038 h1 (CX-900-16). As expected, for the same
activation temperature, the burn-off degree increases with
time; also, for the same activation time, the burn-off degree
increases with the temperature.
Fig. 1 shows the nitrogen adsorption–desorption isotherms
of carbon xerogels synthesised using R/C = 1000, and physi-
cally activated at 900 C during various activation times. Each
isotherm is a combination of type I and type IV isotherms
according to the BDDT classification [33], which corresponds
to microporous and mesoporous materials. Physical activation
with CO2 induces an increase of nitrogen adsorption capacity at
low relative pressure, p/p0: the isotherms are simply shifted
upwards, while the hysteresis shape and position do not
change much, indicating that the mesopores are not altered
by the activation while the microporosity develops. Both the
specific surface area, SBET, and the micropore volume, VDUB-N2 ,
increase (Table 1). Indeed, for the samples activated at 800 C
with activation time ranging from 4 to 24 h, SBET increases from
630 m2 g1 (CX) to 910 m2 g1 (CX-800-24), while VDUB-N2 in-
creases from 0.26 to 0.43 cm3 g1, respectively. When Tact
increases up to 900 C, SBET increases from 630 m2 g1 (CX) to
2180 m2 g1 (CX-900-16); in accordance, VDUB-N2 increases up
to 0.98 cm3 g1. The microporosity development increases with
increasing Tact and tact. The results obtained show that a good
development of microporosity is obtained by physical activa-
tion of carbon xerogels with CO2.
As stated above, the hysteresis shape of the N2 adsorption–
desorption isotherms is almost the same for all the samples
activated physically (Fig. 1), indicating that physical activation
does not modify the size and volume of mesopores. In Table 1
are reported the mesopore size ranges, i.e. the minimum mes-
opore diameter (dp,min = 28–30 nm), and the maximum meso-
2000
V ads (cm3 g-1, STP)
1000
0
0 0.5 1
Fig. 1 – N2 adsorption–desorption isotherms of carbon
xerogels synthesised using R/C = 1000, non-activated and
physically activated at 900 C with carbon dioxide during 4,
8 and 16 h: (·) CX, (j) CX-900-4, () CX-900-8, (d) CX-900-16.
pore diameter (dp,max = 40–43 nm): the differences are within
the measurement error. The mesopore volume can be roughly
estimated from the difference between Vv and VDUB-N2 ; this
volume remains almost constant at relatively low burn-off
degree (1.0–1.3 cm3 g1 for Dm/m = 4–18%) and increases
when Dm/m increases further (1.8 cm3 g1 for sample
CX-900-16). This could indicate a deterioration of the 3D
structure of the carbon xerogel at high burn-off degree.
Fig. 2 collects the CO2 adsorption isotherms of samples
activated at 800 C or 900 C, with various activation times,
and the non-activated reference (CX). As already detected by
nitrogen adsorption, CO2 adsorption confirms the develop-
ment of the microporosity by physical activation. Indeed,
the global adsorption capacity of the samples, i.e. the volume
adsorbed at p/p0  0.03 increases with tact. However, as tact in-
creases, some differences in the curves appear at low relative
pressure. On the one hand, the adsorption capacity at low p/p0
increases as tact increases from 0 to 8 h, with very few differ-
ences between CX-900-4 and CX-900-8; in addition, the curva-
ture of the three isotherms (CX, CX-900-4 and CX-900-8) is
similar and they simply seem displaced upwards. This corre-
sponds to the development of narrow micropores. On the
other hand, increasing tact up to 16 h leads to decreasing the
adsorption capacity at low p/p0, the isotherm of CX-900-16
being less incurved than the other three. This reflects the
development of larger micropores from the narrow ones.
The micropore enlargement is confirmed by the decrease of
the CO2–carbon interaction energy, E0-CO2 : indeed, when
activation is performed at 900 C, E0-CO2 decreases from
25.0 kJ/mol (CX) to 17.1 kJ/mol (CX-900-16). In addition, calcu-
lated values of the average micropore size, L, show that the
micropores tend to get enlarged by activation: indeed, L in-
creases regularly from 0.79 nm (CX) to 1.40 nm (CX-900-16)
as the burn-off percentage increases.
Results gathered in Table 1 also indicate that the volume
corresponding to narrow micropores, VDUB-CO2 , increases while
 Table 1 – Activation conditions and texture characterization.
Sample Activation N2 adsorption CO2 adsorption He pycnometry
Tact tact Dm/m rs SBET VDUB-N2 Vv dp,min dp,max VDUB-CO2 E0-CO2 L SCO2 qs qbulk
(C) (h) (%) (h1) (m2 g1) (cm3 g1) (cm3 g1) (nm) (nm) (cm3 g1) (kJ/mol) (nm) (m2 g1) (g cm3) (g cm3)
±2 ±0.005 ±10 ±0.01 ±0.1 ±2 ±2 ±0.01 ±0.5 ±0.05 ±10 ±0.02 ±0.05

CX –a –a –a –a 630 0.28 1.3 28 40 0.23 25.0 0.79 580 1.89 0.55

CX-800-4 800 4 4 0.010 670 0.28 1.2 29 43 –b –b –b –b 1.94 0.58
CX-800-16 800 16 12 0.008 810 0.33 1.3 28 42 –b –b –b –b 2.00 0.56
CX-800-24 800 24 14 0.006 910 0.37 1.5 30 41 0.30 23.2 0.92 660 2.04 0.50
CX-900-4 900 4 18 0.045 1015 0.42 1.7 29 40 0.32 22.7 0.96 670 2.13 0.46
CX-900-8 900 8 32 0.040 1365 0.57 2.0 28 43 0.38 21.0 1.13 680 2.15 0.41
CX-900-16 900 16 60 0.038 2180 0.98 2.8 29 41 0.40 17.1 1.40 570 2.18 0.31
Tact: activation temperature; tact: activation time; Dm/m: relative mass loss during activation (i.e. percentage of burn-off); rs : average specific gasification rate; SBET: specific surface area calculated by
the BET equation from N2 adsorption data; VDUB-N2 : micropore volume determined from nitrogen adsorption by the Dubinin–Radushkevich equation; Vv: total pore volume determined from nitrogen
adsorption at saturation; dp,min and dp,max: minimum and maximum mesopore diameters, deduced from the mesopore size distributions provided by the Broekhoff–de–Boer method applied to N2

CARBON
adsorption data; VDUB-CO2 : micropore volume determined from CO2 adsorption by the Dubinin–Radushkevich equation; E0-CO2 : CO2–carbon interaction energy calculated by the Dubinin–Radushkevich
equation; L: average micropore size determined from Stoeckli or Dubinin equation; SCO2 : specific surface area determined from CO2 adsorption, calculated from VDUB-CO2 and L; qs: skeletal density
measured by helium pycnometry; qbulk: bulk density obtained from Vv and qs.
a
Not pertinent.

4 8 ( 20 1 0 ) 3 1 5 7–31 6 8
b
Not measured.

CX-800-24, (j) CX-900-4, () CX-900-8, (d) CX-900-16.

900 C with carbon dioxide during 4, 8 and 16 h: (·) CX, (D)
activated at 800 C with carbon dioxide during 24 h; and at
synthesised using R/C = 1000, non-activated and physically
Fig. 2 – CO2 adsorption isotherms of carbon xerogels
CO2 adsorption.
percentage. Data obtained from (.) N2 adsorption and (5)
Radushkevich equation as a function of the burn-off
Fig. 3 – Micropore volume obtained by the Dubinin–

VDUB-CO2 reaches a maximum value of 0.38–0.40 cm3 g1

VDUB-CO2 reaches a plateau for Dm/m higher than 30%:
VDUB-N2 increases linearly with the burn-off degree, while
series, and shows that for samples activated at 900 C,
the burn-off degree. Fig. 3 regroups the data of the CX-900
both adsorption techniques can be plotted as a function of
increase continuously. The micropore volume obtained by
pore volume measured by N2 adsorption, VDUB-N2 , both
trast, the specific surface area, SBET, and the Dubinin micro-
reflects the increase of the average micropore size. By con-
imum when increasing the burn-off percentage. Again, this
the corresponding surface area, SCO2 passes through a max-
V DUB (cm3 g-1) V ads (cm3 g-1, STP)

100
0.5

50
0
0

0
0
20
Δm/m (%)

0.015
p /p 0
40 60

0.03
80

3161
3162 CARBON 4 8 ( 2 0 1 0 ) 3 1 5 7 –3 1 6 8

(samples CX-900-8 and CX-900-16) while VDUB-N2 increases
continuously up to 0.98 cm3 g1 (CX-900-16). This behavior is
typical, and was observed for other activated carbons derived
from various precursors [25,34]: (i) for low burn-off degree, the
activation process develops the narrow microporosity inac-
cessible to nitrogen; (ii) from a sufficiently high burn-off de-
gree, the narrow micropore volume does not increase
anymore (and eventually decreases at high Dm/m) while the
micropore volume detected by nitrogen adsorption still in-
creases. This indicates that the increase of total micropore
volume occurs through (i) the creation of narrow micropores
15
and (ii) the widening of these narrow micropores, both phe-
nomena occurring during the whole activation process [25].
Helium pycnometry was used to measure the skeletal den-
sity, qs, of the carbons, and the bulk density, qbulk, was calcu-
lated according to Eq. (2). Globally, one can observe a slight
increase of the skeletal density with activation (Table 1): from
1.89 g cm3 (CX) to 2.04 g cm3 (CX-800-24) for the CX-800 ser-
ies, and up to 2.18 g cm3 for the CX-900 series. This could re-
flect the opening of closed microporosity, the skeletal density
of graphite being close to 2.25 g cm3. The bulk density, qbulk,
decreases when the burn-off degree increases: for example, in
the CX-900 series, qbulk decreases from 0.46 g cm3 (CX-900-4:
Dm/m = 18%) to 0.31 g cm3 (CX-900-16: Dm/m = 60%). This re-
flects both the development of microporosity for all the acti-
vated samples, and the increase of mesopore volume
observed at very high burn-off degree (CX-900-16).

10 3.2. Surface chemistry characterization

Final pH

The initial pH vs. final pH curves leading to the determination

5 to 8.6, whatever the activation treatment. The observed differ-
0
0 5
Initial pH
Fig. 4 – Determination of the PZC of carbon xerogels
synthesised using R/C = 1000, non-activated and physically
activated with carbon dioxide at 800 C during 24 h; and at
900 C during 4, 8 and 16 h: (·) CX, (D) CX-800-24, (j)
CX-900-4, () CX-900-8, (d) CX-900-16.
of the PZC of the various materials are presented in Fig. 4. All
the curves clearly show a plateau at final pH ranging from 8.2
ences are not significant. In the case of samples activated at
800 C, only CX-800-24 was analyzed; since no difference is
observed between this sample and the non-activated carbon
(CX), one can reasonably conclude that intermediate activa-
tion time does not lead to PZC change.
10 15
Fig. 5 shows the TPD spectra of carbon xerogels activated
with CO2. The quantities of CO and CO2 desorbed, nCO and
nCO2 , are quite low. Indeed, nCO ranges between 405 and
927 lmol g1, and nCO2 between 114 and 163 lmol g1. For com-
parison, carbons activated by oxidation with nitric acid or oxy-
gen can display nCO2 values reaching 1000–1500 lmol g1 and
nCO amounts up to 6000–6500 lmol g1 [29].

a 0.60 b 0.08
n CO2 (µmol g-1 s-1) , relative units
n CO (µmol g-1 s-1) , relative units

0.30 0.04

0.00 0.00
0 500 1000 0 500 1000
T (ºC) T (ºC)

Fig. 5 – CO (a) and CO2 (b) desorption profiles for carbon xerogels synthesised using R/C = 1000, non-activated and physically
activated with carbon dioxide at 800 C during 24 h; and at 900 C during 4, 8 and 16 h: (·) CX, (D) CX-800-24, (j) CX-900-4, ()
CX-900-8, (d) CX-900-16. For clarity, the curves are shifted upwards, and the corresponding ground levels are indicated by
dotted horizontals.
 CARBON 4 8 ( 20 1 0 ) 3 1 5 7–31 6 8
a 0.06
rs (h-1)
0.03
0 0
0 12 24
t act (h)
Fig. 6 – (a) Average specific gasification rate for activated xerogels as a function of activation time, (b) specific surface area for
activated xerogels as a function of activation time (.) CX-900 series, (5) CX-800 series.
The presence of peaks also allows determining the nature
of the oxygen groups present on the carbon surface. The oxy-
gen groups decompose with the evolution of CO (ethers,
anhydrides, phenols, carbonyls, quinones) or CO2 (carboxylic
acids, anhydrides, lactones) [29]. However, since the amounts
desorbed are very small, it is not easy to attribute the peaks
observed to distinct oxygen surface groups and quantify their
relative proportions. A qualitative analysis remains possible.
The CO desorption observed below 500 C for CX and
CX-800-24 (Fig. 5a) could be attributed to some carboxylic
anhydrides groups. Phenol and ether groups lead to CO
desorption around 600–700 C while detection of CO at tem-
perature higher than 700 C could reflect the elimination of
carbonyl and quinone groups. The peak observed near
800 C for all samples can be attributed to the presence of car-
bonyl and quinone groups; the intensity of this peak is almost
twice higher for CX-800-24 than for the CX-900 series. Con-
cerning CO2 desorption, peaks located between 200 and
400 C indicate the presence of carboxylic acids, and maybe
carboxylic anhydrides, which would match the desorption
of CO in the same temperature range. The small shoulder ob-
served for all the samples at 500 C could also indicate the
presence of carboxylic anhydrides. For all samples activated,
CO2 desorption is also observed between 500 and 800 C, indi-
cating maybe the presence of lactone groups. As a general
observation, the nature and amounts of oxygen surface
groups do not change drastically with the activation condi-
tions: all the obtained CO and CO2 desorption spectra are sim-
ilar, and the amounts desorbed are quite low.
4. Discussion
In physical activation of carbon xerogels, the effect of the
activation time, at constant activation temperature, on the
porosity development can be studied through a comparison
of the sample series activated either at 800 C or at 900 C.
In the same way, the effect of temperature can be determined
through a comparison between CX-800 and CX-900 samples
3163
b 2500
2000
S BET (m2 g-1)
1500
1000
500
0 12 24
t act (h)
prepared at fixed activation time. One can observe (Fig. 6a)
that, at constant activation temperature, the average specific
gasification rate, rs , decreases slightly with activation time:
indeed, rs decreases from 0.010 to 0.008 h1 for the series set
at 800 C when increasing tact from 4 to 16 h while, for the
CX-900 series, rs decreases from 0.045 to 0.038 h1 for the
same tact range. This could be due to the fact that, at the
beginning of the gasification, some carbon surface atoms
are more prone to elimination by gasification with CO2; with
time, these preferential sites disappear, and gasification pro-
ceeds with the elimination of more stable carbon atoms. In
addition, comparison of CX-800 and CX-900 series shows that
rs globally increases with the temperature (Fig. 6a), which is in
agreement with previous studies [35].
The specific surface area, SBET, develops with activation
time (Fig. 6b and Table 1). The development of microporosity,
which can be interpreted in terms of increase of specific sur-
face area, SBET, is directly related to the activation tempera-
ture (which increases the average specific gasification rate)
and activation time. As shown in Fig. 6b, SBET increases quite
linearly with tact. However, the average specific gasification
rate is about three times higher at 900 C than at 800 C, as
also shown by the slope difference in the SBET vs. tact plot of
CX-800 and CX-900 series.
Many studies report specific surface area development
using chemical activation or physical activation [1,18,19,36–
43]. Since carbon xerogels were previously activated with
hydroxides (NaOH and KOH) [20,21], the present study allows
for comparing the two procedures. Fig. 7 shows the specific
surface area, SBET, for xerogels chemically activated with
NaOH and KOH as a function of the activation agent/carbon
precursor mass ratio (A/P), and Fig. 8 shows the CO2 adsorp-
tion isotherms of two carbon xerogels, activated with KOH
at various A/P values (A/P = 2, sample KOH-2 and A/P = 4,
sample KOH-4). It was observed from nitrogen adsorption iso-
therms that chemical activation produces an increase in
micropore volume, VDUB-N2 , and thus specific surface area,
SBET; however, the activation modifies neither the 3D structure
3164 CARBON
2000
* ing agent/carbon precursor ratio (A/P) increases, the micro-
1500
* *
S BET (m2 g-1)
1000
* *
500
0 plots show that, in the case of chemical activation, the iso-
0 1 2 3 4
A/P
Fig. 7 – Specific surface area for chemically activated
xerogels produced at various activating agent/carbon
precursor mass ratios, data from Ref. [20]: (*) NaOH, (+) KOH.
140
V ads (cm3 g-1, STP)
70
0 chemical activation delays this phenomenon. So, if narrow
0 0.015 0.03
p /p 0
Fig. 8 – CO2 adsorption isotherms of carbon xerogels
synthesised using R/C = 1000, non-activated, physically
activated with carbon dioxide at 900 C during 4 and 16 h;
and chemically activated with KOH: (·) CX, (j) CX-900-4, (d)
CX-900-16, ( ) sample activated with KOH, A/P = 2, (+)
sample activated with KOH, A/P = 4 (data from Ref. [20]).
of the carbon xerogel, except at very high activating agent/
carbon precursor mass ratio, A/P, nor the meso- or macropo-
rosity formed during the gel synthesis step, which corre-
sponds to the results obtained in the present study with
CO2 activation. Looking only at nitrogen adsorption, one
could then conclude that both chemical and physical pro-
cesses lead to the same porosity development. However, dif-
ferences are observed in CO2 adsorption. In the case of
physical activation, one observes that both VDUB-N2 and
VDUB-CO2 increase, even at low burn-off degree, which means
that enlargement of the narrow pores takes place immedi-
4 8 ( 2 0 1 0 ) 3 1 5 7 –3 1 6 8
ately. On the contrary, in the case of chemical activation of
carbon xerogels with KOH [20] and NaOH [21], as the activat-
pore volume obtained by CO2 adsorption, VDUB-CO2 , increases
first, while the micropore volume deduced from N2 adsorp-
tion, VDUB-N2 , remains quite constant. This shows that chemi-
cal activation first develops narrow micropores only, and that
the enlargement of the micropores only takes place later.
Fig. 8 compares the CO2 adsorption isotherms of several sam-
ples: the non-activated reference CX, two carbon xerogels
activated with CO2 (CX-900-4 and CX-900-16) and two carbon
xerogels, prepared in the same conditions as the present
study, but activated with KOH at various A/P values (A/P = 2,
sample KOH-2 and A/P = 4, KOH-4; samples from [20]). The
therms superimpose to that of the non-activated reference
CX at very low p/p0; then the adsorbed volume increases as
A/P increases and the isotherms are outstretched upwards.
This hints at the development of the micropore volume
through the appearance of narrow micropores. The same
behavior is observed for physical activation with low burn-
off degree (CX-900-4). However, at high burn-off degree
(CX-900-16), the isotherm curvature at low p/p0 changes dras-
tically: even at very low p/p0 (<0.01), the adsorbed volume is
lower than in the case of CX, and the isotherm is out-
stretched. This reflects again the increase of micropore size
in the case of physical activation. The effect can also be
emphasized from VDUB-CO2 and VDUB-N2 values. In the case of
physical activation, VDUB-CO2 barely reaches 0.38–0.40 cm3 g1
for CX-900-8 and CX-900-16; in parallel, VDUB-N2 increases up
to 0.57 and 0.98 cm3 g1, respectively (Table 1). On the con-
trary, chemical activation with low amounts of KOH allows
obtaining VDUB-CO2 = 0.50 cm3 g1 for lower corresponding
VDUB-N2 values (0.26 cm3 g1 for A/P = 1 and 0.65 cm3 g1 for
A/P = 4, R/C = 1000 [20]).
This clearly shows that, while micropore enlargement
takes place immediately in the case of physical activation,
micropores are required, chemical activation with KOH or
NaOH with low A/P ratio should be used. On the contrary, if
the micropores must be enlarged, physical activation with
CO2 gives more appropriate materials. In addition, one can
conclude that chemical activation of carbon xerogels leads
to microporosity more homogeneous in size, while physical
activation widens the micropore size distribution. This goes
in line with results obtained with other carbon carbon precur-
sors like, for instance, carbon fibers [43], Saran char [34], bitu-
minous coal [44] or almond shell char [45], and should be
taken into account in the ambit of future applications. The
difference probably comes from the activation mechanism.
Indeed, in chemical activation, reactions involve the transfor-
mation of hydroxide and carbon into carbonate, hydrogen
and metallic Na or K [42]. So, the chemical activation of the
carbon precursor presumably takes place at privileged spots,
where K- or Na-containing clusters are formed by deposition
of hydroxides after impregnation and solvent evaporation.
Chemical activation thus probably occurs via the ‘drilling’ of
new micropores; finally, further activation leads to the re-
moval of walls between two micropores, and global enlarge-
ment at high A/P ratio. On the contrary, CO2 can access the
 CARBON 4 8 ( 20 1 0 ) 3 1 5 7–31 6 8 3165

difficult to eliminate during the washing step [21]. However,

Table 2 – Surface chemistry characterization.
salts should be completely eliminated by deeper washing.
Sample PZC nCO nCO2 For physically activated carbon xerogels, the only salt present
(–) (lmol g1) (lmol g1) is Na2CO3, which was used in the synthesis step and found
±0.1 ±20 ±20
back under the same form after thermal treatments [11]. Cal-
CX 8.2 500 148 culation leads to a maximum percentage of 0.20% in sample
CX-800-4 –a –a –a CX-900-16, i.e. the sample with maximum burn-off, which ex-
CX-800-16 –a –a –a plains the difference in surface accessibility.
CX-800-24 8.2 927 163
Finally, one can compare the surface chemistry of carbon
CX-900-4 8.6 457 114
xerogels activated either chemically with KOH or physically
CX-900-8 8.6 405 133
CX-900-16 8.6 427 123 with CO2. Note that some samples, prepared by both methods,
KOH-4b –a 1927 277 were re-analyzed after a longer storage in air (closed vessels),
a
Not measured.
and no difference was found: the samples were considered as
b
Data from Ref. [20]. stable with regard to oxidation. From the PZC measurement
(Table 2) and the TPD spectra (Fig. 5a and b), one can conclude
that physical activation does not lead to the formation of oxy-
whole carbon inner surface by diffusion during the activation gen surface groups. Indeed, the PZC of the activated samples is
process, and the micropore enlargement starts immediately. identical to that of the non-activated reference; the CO and
The comparison of Fig. 6b with Fig. 7 shows one additional CO2 amounts evolved during TPD are quite low and in the
difference between physical and chemical activations. In the same order of magnitude as for the non-activated carbon
case of physical activation, the specific surface area obtained xerogel. The surface chemistry remains almost constant.
from N2 adsorption, SBET, increases quite linearly with the Since CO2 activation takes place at a temperature higher than
activation time and with the burn-off degree (Fig. 3), tact and that of the oxygen surface group elimination, these groups can
Dm/m being directly linked. On the contrary, for chemical acti- only be formed by re-oxidation when the activated sample is
vation, the SBET vs. A/P relationship is not linear. As an exam- removed from the oven. On the contrary, for a carbon chemi-
ple, the specific surface area, SBET, decreases slightly at low cally activated with KOH, the TPD spectra show higher quan-
A/P value (Fig. 7), which was concluded to be due to micropore tities of CO and CO2 desorbed. Fig. 9 compares the TPD
blocking by salt clusters coming from the activating agent spectra of samples CX, CX-900-16 and KOH-4 (sample acti-
[20,21]. Indeed, KOH or NaOH removal by washing may in vated with KOH, A/P = 4 [20]). Chemical activation clearly in-
some cases be difficult [18,20,21,42,43], especially when the creases the amount of CO and CO2 desorbed, which reflects
microporosity is not developed enough. Results obtained for the presence of higher amounts of oxygen surface groups.
carbon xerogels indicate that the access to the material sur- The total amount of CO an CO2 desorbed for KOH-4 are
face is more difficult in the case of chemical activation. In 1927 lmol g1and 277 lmol g1, respectively, i.e. almost four
addition, the final ash content of chemically activated carbon and two times the total quantity observed for physically acti-
xerogels can remain high after a standard washing procedure: vated carbons (nCO  400–500 lmol g1 and nCO2  110–
around 2 wt.% in the case of activation with KOH, and up to 160 lmol g1). In the case of chemical activation, two CO peaks
10 wt.% for samples activated with NaOH, sodium being more can be observed at 700 and 800 C, showing that the formation

a 1.00 b 0.16
n CO2 (µmol g-1 s-1) , relative units
n CO (µmol g-1 s-1) , relative units

0.50 0.08

0.00 0.00
0 500 1000 0 500 1000
T (ºC) T (ºC)

Fig. 9 – CO (a) and CO2 (b) desorption profiles for carbon xerogels synthesised using R/C = 1000, non-activated, physically
activated with carbon dioxide at 800 C during 24 h, physically activated with carbon dioxide at 900 C during 16 h; and
chemically activated with KOH: (·) CX, (D) CX-800-24, (d) CX-900-16, (+) sample activated with KOH, A/P = 4 (data from Ref.
[20]). For clarity, the curves are shifted upwards, and the corresponding ground levels are indicated by dotted horizontals.
3166 CARBON 4 8 ( 2 0 1 0 ) 3 1 5 7 –3 1 6 8
of high quantities of phenol, ether, quinone and/or carbonyls
groups. The large CO2 peak observed for KOH-4 at 250 C is
probably due to carboxylic groups, the development of which
is clearly higher in the case of chemical activation. Again, dur-
ing TPD experiments, most of the oxygenated surface groups
are eliminated at temperatures lower than that of the activa-
tion treatment (750 C [20,21]). The difference observed be-
tween these two activation procedures could be due to the
difference of re-oxidation mechanisms. Indeed, potassium
or sodium introduced as chemical reagents remain in the car-
bonaceous material until re-exposure to air, and could there-
fore induce consequent surface re-oxidation. On the
contrary, in physical activation, the carbon surface is treated
thermally and the oxygen-containing surface groups are re-
moved. So, re-oxidation can only be induced by very reactive
carbon sites left at the surface that can react with the oxygen
when they are exposed to the room-temperature air. Much
likely, the amount of these active sites is rather low, which ex-
plains the low amounts of oxygen surface groups detected,
while oxidation induced by potassium or sodium is easier,
leading to increasing the surface oxidation.
5. Conclusions
The physical activation of carbon xerogels using carbon diox-
ide allows the production of materials with a well developed
microporosity without any alteration of the mesoporosity
formed during the synthesis. Increasing the activation temper-
ature from 800 C to 900 C leads to increasing the gasification
rate from 0.01 to 0.04 h1; moreover, the developed surface
depends linearly on the burn-off degree. Thus, it is possible
to control the surface development easily. However, the activa-
tion time remains very long to reach very high specific surface
areas, even at 900 C (10 h for SBET  1500 m2 g1); physical
activation should then be conducted at higher temperature,
which is a major drawback of the process from the point of view
of costs.
Physical activation of carbon xerogels was compared to
chemical activation using hydroxides. Globally, both pro-
cesses lead to the development of the microporosity without,
to some extent, damaging the meso-macroporosity obtained
via sol–gel chemistry. Chemical activation has several advan-
tages: (i) as in the case of many other precursors, the temper-
ature required is lower, which is why this process is generally
preferred, (ii) the time required to produce equivalent surface
development is shorter and (iii) the micropore development of
carbon xerogels is more uniform, the enlargement of narrow
micropores taking place at high chemical agent concentration
only. However, physical activation (i) does not require any cor-
rosive agents, (ii) leads to activated carbon xerogels with very
low ash content without requiring a washing treatment and
(iii) leaves a surface containing much lower concentration of
oxygen surface groups than in the case of the xerogels ob-
tained by chemical activation with hydroxides. Since con-
trolled oxidation remains possible afterwards [29,30],
physical activation produces carbon xerogels with more flex-
ible surface chemistry characteristics.
Concerning microporosity development, it was observed
that the activation mechanism is not the same for both pro-
cesses. Comparison of micropore volumes obtained either by
N2 or CO2 adsorption shows that, during chemical activation
with hydroxides, the narrow microporosity first develops via
the appearance of new narrow micropores; at high A/P ratio,
enlargement of micropores, which could occur via the breaking
of walls separating individual micropores, is detected. In the
case of physical activation, the enlargement of narrow microp-
ores starts immediately, whatever the activation conditions,
which leads to the development of microporosity with a wider
size range. So, should narrow micropores be required in the fi-
nal activated carbon xerogels, the use of chemical activation
with relatively low A/P ratio is recommended. If extremely pure
materials are necessary, physical activation would better be
used. Concerning the mesopores and macropores, the texture
of carbon xerogels is preserved as long as the degree of carbon
removal is not too high (i.e. relatively low burn-off degree in the
case of CO2 activation, and A/P ratio < 4 in the case of chemical
activation with KOH). However, (i) in the case of activation with
CO2, one observes an increase of the mesopore volume at high
burn-off degree and (ii) chemical activation with KOH may lead
to meso/macropore enlargement at high A/P ratio [20]. From re-
sults presented in this study and in previous work [20,21], it is
not possible to conclude about the mechanisms of meso/mac-
ropore development during the activation of carbon xerogels;
however, the mechanism is much probably quite different from
one activation technique to another. Indeed, whereas the for-
mation of micro, meso, and macropores in a physical activation
process is usually the consequence of carbon burn-off and pro-
gressive pore widening, this is not necessarily the case for
chemical activation processes, in which pore development
can be the result of very complex reaction schemes [46].
Both activation types could be directly subsequent to pyro-
lysis in a continuous process. However, it was demonstrated
previously [20] that coupling with pyrolysis is not possible in
the case of chemical activation with hydroxides, because
the direct activation of organic gels leads to the destruction
of the initial meso-macropore texture. The effect of direct
CO2 activation of organic resorcinol–formaldehyde gels on
the textural properties of the final carbon should be the scope
of further works aiming at the optimization of the physical
activation process.
Acknowledgements
N.J. and C.P. are postdoctoral researchers of the F.R.S.-FNRS
(Belgium). A.L. is research associate of the F.R.S.-FNRS (Bel-
gium). The Belgian authors thank the Fonds de Recherche
Fondamentale Collective, the Ministère de la Région Wallonne
and the Interuniversity Attraction Pole (IAP-P6/17) for their
financial support.
R E F E R E N C E S
[1] Derbyshire F, Jagtoyen M, Thwaites M. Activated carbons –
production and applications. In: Patrick JW, editor. Porosity in
carbons. London: Wiley; 1995. p. 227–52.
 CARBON 4 8 ( 20 1 0 ) 3 1 5 7–31 6 8
[2] Job N, Heinrichs B, Lambert S, Pirard J-P, Colomer J-F,
Vertruyen B, et al. Carbon xerogels as catalyst supports:
study of mass transfer. AIChE J 2006;52(8):2663–76.
[3] Al-Muhtaseb SA, Ritter A. Preparation and properties of
resorcinol formaldehyde organic and carbon gels. Adv Mater
2003;15(2):101–14.
[4] Job N, Théry A, Pirard R, Marien J, Kocon L, Rouzaud J-N, et al.
Carbon aerogels, cryogels and xerogels: Influence of the
drying method on the textural properties of porous carbon
materials. Carbon 2005;43(12):2481–94.
[5] Pekala RW, Kong FM. A synthetic route to organic aerogels –
mechanism, structures and properties. Rev Phys Appl
1989;24(C4):33–40.
[6] Moreno-Castilla C, Maldonado-Hódar FJ. Carbon aerogels for
catalysis applications: an overview. Carbon 2005;43(3):455–65.
[7] Mathieu B, Michaux B, Van Cantfort O, Noville F, Pirard R,
Pirard J-P. Synthesis of resorcinol–formaldehyde aerogels by
the freeze-drying method. Ann Chim – Sci Mat
1997;22(1):19–29.
[8] Mathieu B, Blacher S, Pirard R, Pirard J-P, Sahouli B, Brouers S.
Freeze-dried resorcinol–formaldehyde gels. J Non-Cryst
Solids 1997;212(2–3):250–61.
[9] Tamon H, Ishizaka H, Yamamoto T, Suzuki T. Influence of
freeze-drying conditions on the mesoporosity of organic gels
as carbon precursors. Carbon 2000;38(7):1099–105.
[10] Lin C, Ritter JA. Carbonization and activation of sol–gel
derived carbon xerogels. Carbon 2000;38(6):849–61.
[11] Job N, Pirard R, Marien J, Pirard JP. Porous carbon xerogels
with texture tailored by pH control during sol–gel process.
Carbon 2004;42(3):619–28.
[12] Léonard A, Job N, Blacher S, Pirard J-P, Crine M, Jomaa W.
Suitability of convective air drying for the production of
porous resorcinol–formaldehyde and carbon xerogels.
Carbon 2005;43(8):1808–11.
[13] Job N, Sabatier F, Pirard J-P, Crine M, Léonard A. Towards the
production of carbon xerogel monoliths by optimizing
convective drying conditions. Carbon 2006;44(12):2534–42.
[14] Reichenauer G, Pfrang T, Hofmann M. Drying of meso- and
macroporous gels – length change and drying dynamics.
Colloid Surf A 2007;300(1–2):211–5.
[15] Pekala RW, Schaefer DW. Structure of organic aerogels. I.
Morphology and Scaling. Macromolecules
1993;26(20):5487–93.
[16] Gommes CJ, Job N, Pirard J-P, Blacher S, Goderis B. Critical
opalescence points to thermodynamic instability: relevance
to small-angle X-ray scattering of resorcinol–formaldehyde
gel formation at low pH. J Appl Crystallogr 2008;41(4):663–8.
[17] Job N, Lambert S, Chatenet M, Gommes CJ, Maillard F,
Berthon-Fabry S, et al. Preparation of highly loaded Pt/carbon
xerogel catalysts for proton exchange membrane fuel cells by
the strong electrostatic adsorption method. Catal Today
2010;150(1–2):119–27.
[18] Linares-Solano A, Lozano-Castelló D, Lillo-Ródenas MA,
Cazorla-Amorós D. Carbon activation by alkaline hydroxides.
Preparation and reactions, porosity and performance. In:
Radovic LR, editor. Chemistry and physics of carbons, vol.
30. Boca Raton: CRC Press, Taylor & Francis Group; 2008. p.
1–62.
[19] Ahmadpour A, Do DD. The preparation of active carbons
from coal by chemical and physical activation. Carbon
1996;34(12):471–9.
[20] Zubizarreta L, Arenillas A, Pirard J-P, Pis JJ, Job N. Tailoring the
textural properties of activated carbon xerogels by chemical
activation with KOH. Micropor Mesopor Mater
2008;115(3):480–90.
[21] Zubizarreta L, Arenillas A, Pis JJ, Pirard J-P, Job N. Studying
chemical activation in carbon xerogels. J Mater Sci
2009;44(24):6583–90.
3167
[22] Fairén-Jiménez D, Carrasco-Marı́n F, Moreno-Castilla C.
Porosity and surface area of monolithic carbon aerogels
prepared using alkaline carbonates and organic acids as
polymerization catalysts. Carbon 2006;44:2301–7.
[23] Byrne JF, Marsh H. Introductory overview. In: Patrick JW,
editor. Porosity in carbons. London: Wiley; 1995. p. 15–8.
[24] Marsh H, Wynne-Jones WFK. Surface properties of carbon – I.
The effect of activated diffusion in the determination of
surface area. Carbon 1964;1(3):269–79.
[25] Lozano-Castelló D, Cazorla-Amorós D, Linares-Solano A.
Usefulness of CO2 adsorption at 273 K for the
characterization of porous carbons. Carbon
2004;42(7):1233–42.
[26] Dubinin MM, Zaverina ED, Radushkevich LV. Sorption and
structure of active carbons. I. Adsorption of organic vapors.
Zh Fiz Khim 1947;21:1351–62.
[27] Stoeckli F, Ballerini L. Evolution of microporosity during
activation of carbon. Fuel 1991;70(4):557–9.
[28] Rodrı́guez-Reinoso F. The role of carbon materials in
heterogeneous catalysis. Carbon 1998;36(3):159–75.
[29] Figueiredo JL, Pereira MFR, Freitas MMA, Órfão JJM.
Modification of the surface chemistry of activated carbons.
Carbon 1999;37(9):1379–89.
[30] Mahata N, Pereira MFR, Suárez-Garcı́a F, Martı́nez-Alonso A,
Tascón JMD, Figueiredo JL. Tuning of textural and surface
chemistry of carbon xerogels. J Colloid Interf Sci 2008;324(1–
2):150–5.
[31] Regalbuto JR. Strong electrostatic adsorption of metals on
catalyst supports. In: Regalbuto JR, editor. Catalyst
preparation: science and engineering. Boca Raton: CRC
Press, Taylor & Francis Group; 2007. p. 297–318.
[32] Lambert S, Job N, D’Souza L, Ribeiro MFR, Pirard R, Heinrichs
B, et al. Synthesis of very highly dispersed platinum
catalysts supported on carbon xerogels by the strong
electrostatic adsorption method. J Catal 2009;261(1):23–33.
[33] Brunauer S, Emmett T, Teller E. Adsorption of gases in
multimolecular layers. J Am Chem Soc 1938;60:309–20.
[34] Lozano-Castelló D, Cazorla-Amorós D, Linares-Solano A. Can
highly activated carbons be prepared with a homogeneous
micropore size distribution? Fuel Process Technol 2002;77–
78:325–30.
[35] Byrne JF, Marsh H. Introductory overview. In: Patrick JW,
editor. Porosity in carbons. London: Wiley; 1995. p. 1–48.
[36] Rodrı́guez-Reinoso F, Pastor A, Marsh H, Martı́nez M.
Preparation of activated carbon cloths from viscous rayon.
Part II. Physical activation processes. Carbon
2000;38(3):379–95.
[37] Pastor AC, Rodrı́guez-Reinoso F, Marsh H, Martı́nez MA.
Preparation of activated carbon cloths from viscous rayon.
Part I. Carbonization procedures. Carbon 1999;37(8):1275–83.
[38] Rodrı́guez-Reinoso F, Pastor A, Marsh H, Martı́nez M.
Preparation of activated carbon cloths from viscous rayon.
Part III. Effect of carbonization on CO2 activation. Carbon
2000;38(3):397–406.
[39] Berger J, Siemieniewska T, Tomkov K. Development of
porosity in brown-coal chars on activation with carbon
dioxide. Fuel 1976;55(1):9–15.
[40] Rodrı́guez-Reinoso F, Martı́n-Martı́nez JM, Molina-Sabio M,
Pérez-Lledó I, Prado Burguete C. A comparison of the porous
texture of two CO2 activated botanic materials. Carbon
1985;23(1):19–24.
[41] Lozano-Castelló D, Lillo-Ródenas MA, Cazorla-Amorós D,
Linares-Solano A. Preparation of activated carbons from
Spanish anthracite I. Activation with KOH. Carbon
2001;39(5):741–9.
[42] Lillo-Ródenas MA, Cazorla-Amorós D, Linares-Solano A.
Understanding chemical reactions between carbons and
NaOH and KOH. Carbon 2003;41(2):267–75.
3168 CARBON 4 8 ( 2 0 1 0 ) 3 1 5 7 –3 1 6 8

[43] Maciá-Agulló JA, Moore BC, Cazorla-Amorós D, Linares-
Solano A. Activation of coal tar pitch carbon fibres: physical
activation vs. chemical activation. Carbon 2004;42(7):1367–70.
[44] Linares-Solano A, Martin-Gullon I, Salinas-Martinez de Lecea
C, Serrano-Talavera B. Activated carbons from bituminous.
Coal: effect of mineral matter content. Fuel 2000;79(6):
635–43.
[45] Cazorla-Amorós D, Ribes Pérez DE, Román-Martı́nez MC,
Linares-Solano A. Selective porosity development by
calcium-catalyzed carbon gasification. Carbon
1996;34(7):869–78.
[46] Jagtoyen M, Derbyshire F. Activated carbons from yellow
poplar and white oak by H3PO4 activation. Carbon 1998;36
(7–8):1085–97.