Structural characterization of micro- and mesoporous carbon materials using in situ high pressure 129Xe NMR spectroscopy

Article abstract of DOI:10.1021/cm501102y

In situ high pressure ¹²⁹Xe NMR spectroscopy in combination with volumetric adsorption measurements were used for the textural characterization of different carbon materials with well-defined porosity including microporous carbide-derived carbo

Full text of DOI:10.1021/cm501102y

Article
pubs.acs.org/cm
Structural Characterization of Micro- and Mesoporous Carbon
129
Materials Using In Situ High Pressure Xe NMR Spectroscopy
Martin Oschatz,† Herbert C. Hoffmann, Julia Pallmann, Jana Schaber, Lars Borchardt,
,#
‡,#
‡,#
‡,#
§,⊥
†
†
∥
∥
,†
́
Winfried Nickel, Irena Senkovska, Soledad Rico-Frances, Joaquín Silvestre-Albero, Stefan Kaskel,*
,‡
and Eike Brunner*
†Department of Chemistry and Food Chemistry, Institute of Inorganic Chemistry, TU Dresden, Bergstrasse 66, D-01062 Dresden,
Germany
‡
Department of Chemistry and Food Chemistry, Bioanalytical Chemistry, TU Dresden, Bergstrasse 66, D-01062 Dresden, Germany
§
Institute for Chemical and Bioengineering, ETH Zu
̈
rich, Wolfgang-Pauli-Strasse 10, CH-8093 Zu
Laboratorio de Materiales Avanzados, Departamento de Química Inorganica-Instituto Universitario de Materiales, Universidad de
Alicante, Ap. 99, E-03080 Alicante, Spain
S Supporting Information
̈
rich, Switzerland
∥
́
*
ABSTRACT: In situ high pressure 129Xe NMR spectroscopy in combination
with volumetric adsorption measurements were used for the textural
characterization of different carbon materials with well-defined porosity
including microporous carbide-derived carbons, ordered mesoporous carbide-
derived carbon, and ordered mesoporous CMK-3. Adsorption/desorption
isotherms were measured also by NMR up to relative pressures close to p/p = 1
0
at 237 K. The 1 Xe NMR chemical shift of xenon adsorbed in porous carbons is
found to be correlated with the pore size in analogy to other materials such as
zeolites. In addition, these measurements were performed loading the samples
with n-nonane. Nonane molecules preferentially block the micropores.
However, 1 Xe NMR spectroscopy proves that the nonane also influences
the mesopores, thus providing information about the pore system in
hierarchically structured materials.
29
29
1
5,16
INTRODUCTION
catalysis,
or as electrode materials in electrochemical
■
1
4,17,18
double-layer capacitors (EDLCs).
Porous carbon materials are crucial components in various
fields including adsorption/separation of gases,
1
−3
However, certain applications, such as drug delivery or the
removal of cytokines from blood plasma require the adsorption
of large organic molecules, which cannot be immobilized in
narrow micropores due to their dimensions exceeding the size
electro-
4
−6
7,8
chemical energy storage,
catalysis, capacitive desalination
9
10,11
of water, and biomedicine.
They stand out due to their
high thermal stability, chemical inertness, and electrical
conductivity. Appropriate materials need to exhibit a large
and well-defined contact area between the carbon surface and
the surrounding species because many applications include
adsorption processes. High specific surface areas are provided
by high contents of micropores (<2 nm according to the
IUPAC classification). Therefore, their controlled integration
into porous carbon materials is of fundamental interest in
present day’s materials science. The selective removal of metal
atoms from carbide precursors using a treatment in hot
halogens (most commonly chlorine gas) is one method for the
production of microporous carbon materials (carbide-derived
of the cavities. In such a case, mesopores (2−50 nm) are
10,11
required as adsorption sites.
In addition, they serve for
enhanced materials transport within the adsorptive system. This
14,17
makes their presence highly important.
The synthesis of
mesoporous carbon materials is mostly carried out using
methods based on structure-directing agents (denoted as soft-
1
9,20
templating)
or solid-state templates (denoted as hard-
21
templating or “nanocasting”). One of the most widely known
families of materials produced by nanocasting are the so-called
CMKs (Carbons Mesostructured by KAIST). They are
produced by infiltration of carbon sources (e.g., sucrose) into
ordered mesoporous silica (e.g., SBA-15 or KIT-6) followed by
carbonization and template removal. The resulting carbon
materials offer a system of ordered and uniformly sized
12
carbons, CDCs) with sub-Ångstrøm accuracy. The pore size
and carbon microstructure can be controlled by the
halogenation conditions and the distribution of carbon atoms
in the carbide precursor. The very narrow pore-size distribution
Received: March 28, 2014
Revised: April 29, 2014
Published: May 7, 2014
2
together with high specific surface areas of up to 3000 m /g
2
,13,14
make CDCs attractive candidates for gas storage,
©
2014 American Chemical Society
3280
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
mesopores and are therefore highly suitable for size-selective
applications.
high NMR sensitivity and a large electron cloud which is very
2
2
sensitive to different interactions with various environments,
1
29
Both, the CDC approach and the nanocasting technique can
be combined, and the resulting materials exhibit an ordered
mesopore system coupled with well-defined micropores. These
ordered mesoporous carbide-derived carbons (OM-CDCs) are
obtained by infiltration of polycarbosilane polymers into
ordered mesoporous silica followed by pyrolytic conversion
to silicon carbide (SiC) and wet chemical template removal.
The resulting ordered mesoporous SiC (OM-SiC) is then
transferred to OM-SiC-CDC by selective silicon extraction
associated with the insertion of large micropore volumes in the
carbon nanorods under the conservation of the mesopore
resulting in a wide chemical shift range. Consequently, Xe
NMR spectroscopy offers many interesting parameters, such as
the chemical shift, the line width, the chemical shift anisotropy,
and the longitudinal relaxation time T that can be correlated to
1
structural properties of the porous materials. This technique
2
9
was first described by Ito and Fraissard in 1980. Since then,
different porous solids, such as silica, zeolites, and carbons were
30−48
129
characterized.
Recently, Xe NMR spectroscopy was also
used to monitor the adsorption-induced breathing transitions in
4
9−51
metal−organic frameworks (MOFs).
In general, the overall chemical shift of adsorbed 129Xe
depends on xenon-surface interactions and density-dependent
xenon−xenon interactions:
δ(T, ρ) = δ (T) + δXe−Xe(T, ρ)
1
4,23
structure.
These materials offer outstanding performance in
29
various electrochemical energy storage applications and
capacitive deionization of water due to their impressive specific
2
S
(1)
surface areas of up to 3000 m /g and total nanopore volumes as
3
9,14,23
high as 2.0 cm /g.
Whereas the microporous carbon
where ρ denotes the xenon density and T the absolute
temperature. For low xenon densities, i.e., low relative
pressures, the density-dependent term can be written as
follows:
nanorods serve for high capacities, the mesopore channels in
between facilitate efficient materials transport and rapid access
of guest molecules to the adsorption sites. Consequently, OM-
CDCs offer advanced kinetics when used as electrode materials
2
3
14
δXe−Xe(T, ρ) = ΔXe−Xe(T)ρ
(2)
in EDLCs or in adsorption processes.
The use of such complex materials in applications related to
adsorption processes requires a detailed understanding of their
pore structure (i.e., their size distribution, curvature, accessi-
bility, volume, and surface properties) in order to optimize
them for a specific field of interest and to understand basic
mechanisms. While characterization based on spectroscopy, X-
ray scattering, electron microscopy, calorimetry, or mercury
intrusion as well as combinations of them are very suitable, the
physisorption of gases will likely remain the most frequently
used technique for textural imaging of porous solids because it
enables the calculation of important properties such as the
specific surface area, total pore volume and pore-size
with the temperature-dependent slope ΔXe−Xe(T). Extrapola-
tion of the measured 1 Xe chemical shift to zero density/
pressure therefore yields the xenon-surface interaction term
δS(T). Previous investigations on various amorphous carbon
materials revealed that this value depends on a number of
different materials properties. It could be shown to be
29
38
correlated with the concentration of acidic groups, strong
40
42
adsorption sites, heteroatoms, and the average pore size.
The virial coefficient ΔXe−Xe, i.e., the slope of the chemical shift
with increasing xenon density inside the pores can also deliver
information about the microporous carbon material under
4
1,43,44
129
study.
However, the previously published Xe NMR
2
4
distribution. Commonly, physisorption of nitrogen (77 K),
argon (87 K), and carbon dioxide (273 K/298 K) or
combinations of these techniques are useful for the character-
investigations on carbon materials were limited to rather low
30−47
relative pressures.
Within the present work, we investigated a series of
nanostructured carbon materials (CMK-3, OM-SiC-CDC-800,
and two purely microporous TiC-derived carbons) with well-
defined or even hierarchical porosity in the range from 0.6 to
4.5 nm using high-pressure in situ 1 Xe NMR spectroscopy up
to 18 bar at 237 K. This allowed the NMR study of adsorption/
desorption isotherms up to a relative pressure close to 1. The
NMR parameters can be well correlated with the pore
structures, rendering this spectroscopic method as highly useful
for the characterization of complex porous carbon structures.
The shift of the NMR signals after saturation of micropores
with n-nonane provides additional information about the
mesopore systems, pore connectivity, and the adsorption
state of the hydrocarbon molecules.
25
ization of various nanoporous structures. However, the most
complex and hierarchical structures are often difficult to
analyze. Therefore, additional adsorption techniques have
been developed.
29
For instance, nitrogen physisorption before and after
preadsorption of n-nonane is a useful tool to better distinguish
between the meso- and micropores because the latter are
selectively blocked by the hydrocarbon molecules. This method
was proven to deliver deep insights into the pore structure of
ordered mesoporous silica materials (OMS) as well as
hierarchical activated carbons with a cavitation-linked meso-
2
6,27
structure.
Very recently, this method was used for the
characterization of carbide- and carbon materials with ordered
28
mesopores and varying amounts of micropores. Preadsorp-
tion of n-nonane provides information about the pore
connectivity, and contributions of the different pore systems
to the total pore volume can be directly determined. However,
while some effects of the n-nonane adsorbed in the micropores
upon the mesopore system of the materials were already
observed, these phenomena are still neither fully understood
nor quantified.
EXPERIMENTAL SECTION
■
Synthesis of CMK-3. Hexagonally ordered mesoporous carbon
(CMK-3) was synthesized by mixing a 10 mL aqueous solution of 2.5
g sucrose and 0.28 g concentrated (96%) sulfuric acid with SBA-15
2
8
(
hydrothermally treated at 403 K) in a Petri dish. This mixture was
then heated to 373 K and kept there for 6 h followed by heating to 433
K and keeping for another 6 h ensuring complete polymerization of
the carbohydrate molecules. For complete infiltration of the SBA-15
pore system, the infiltration was repeated with a 10 mL solution of 1.6
g sucrose and 0.18 g sulfuric acid in water again followed by heating to
373 and 433 K. Carbonization was achieved by heating the composite
up to 1173 K for 2 h under flowing argon (150 K/h heating rate) in an
129Xe NMR spectroscopy of adsorbed xenon is another very
useful method for the characterization of porous solids because
it allows us to directly investigate the physical state of adsorbed
molecules. Xenon is an inert, nonpolar spin-1/2 atom with a
3
281
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
alumina tube in a horizontal tubular furnace. Template removal was
performed by a 3 h treatment in a 1:1:1 mixture (by volume) of water,
ethanol, and 35% aqueous hydrofluoric acid solution followed by
filtration, washing with large amounts of ethanol, and drying at room
temperature.
for our apparatus. After pressurizing the samples, a subsequent
equilibration phase of at least 15 min has been deferred. In the case of
cooled samples a further phase of at least 1 h was incorporated for
thermal equilibration. All experiments were performed using an
Avance 300 (Bruker, Karlsruhe, Germany) NMR spectrometer
coupled with a BIOSPIN SA BCU-Xtreme unit (Bruker, Karlsruhe,
Germany). The 1 Xe NMR spectra were recorded at a resonance
frequency of 83.02 MHz using a 10 mm HR probe, a pulse length of 6
μs, and relaxation delays of 5 s. Temperature calibration and
referencing of the 1 Xe NMR chemical shift have been performed
Synthesis of OM-SiC-CDC-800. For the synthesis of ordered
29
mesoporous carbide-derived carbons (OM-SiC-CDCs), 2.0 g of SBA-
28
1
5 (hydrothermally treated at 403 K) were infiltrated with a mixture
of 2.1 mL liquid allylhydridopolycarbosilane (SMP-10, Starfire
Systems) and 0.5 mL of the cross-linker para-divinylbenzene by the
incipient wetness method. The polymeric precursor was thermally
decomposed at 1073 K for 2 h under flowing argon atmosphere (60
K/h heating rate) in an alumina tube in a horizontal tubular furnace.
Silica removal was achieved under the same conditions as described for
the CMK-3. Approximately 1.0 g of the obtained OM-SiC was then
transferred to a quartz boat and placed in a quartz tube (inner
diameter: 25 mm) into the isothermal zone of a horizontal tubular
furnace. The precursor was heated to 1073 K under an argon flow of
29
4
9
as previously described.
RESULTS AND DISCUSSION
■
The structural properties of the four investigated carbon
materials derived from the nitrogen adsorption/desorption
measurements (Table 1 and Supporting Information (SI)
a
150 mL/min (450 K/h heating rate) and subsequently, the gas flow
Table 1. Porosity Data of the Porous Carbon Samples
was changed to a mixture of 70 mL/min argon and 80 mL/min
chlorine for 3 h. Cooling of the OM-SiC-CDC-800 to a temperature of
SSA
TPV
MPV
APS
2
3
3
sample
CMK-3
(m /g)
(cm /g)
(cm /g)
(nm)
873 K under an argon flow of 150 mL/min was followed by 1 h of
1396
1652
1412
2838
1.42
0.74
0.60
2.05
0.1
1.0/4.5
0.8
postreductive treatment under 80 mL/min of flowing hydrogen for the
removal of chlorine and chlorides from the porous carbon.
TiC-CDC-1000
TiC-CDC-600
0.59
0.54
0.55
Synthesis of Microporous TiC-CDCs. The microporous CDCs
were obtained by high-temperature chlorination followed by
postreductive hydrogen treatment of crystalline TiC powder (Sigma-
Aldrich, 95%; particle size ∼4 μm) under the same conditions as for
the above-described OM-SiC-CDC-800. The maximum chlorination
temperature was either 873 K/600 °C (TiC-CDC-600) or 1273 K/
0.6
OM-SiC-CDC-800
1.0/4.1
a
SSA: Specific Surface Area, TPV: Total Pore Volume, MPV:
Micropore Volume, and APS: Average Pore Size.
53
1
000 °C (TiC-CDC-1000).
Figure S1) have been previously described. Due to the purely
microporous character of the TiC-CDCs, their nitrogen
physisorption isotherms are of type I according to the
IUPAC classification. In the semilogarithmic plots of the
isotherms (SI Figure S2), higher nitrogen uptake at low relative
pressures can be observed for the TiC-CDC-600 due to the
presence of rather narrow micropores of 0.6 nm diameter in
this sample compared to TiC-CDC-1000 containing 0.8 nm
sized pores (SI Figure S3). The largest micropores of 1.0 nm
are present in OM-SiC-CDC-800. CMK-3 shows a relatively
n-Nonane Preadsorption. Preadsorption of n-nonane was
performed in a homemade manometric equipment developed at the
Advanced Materials Group (LMA, University of Alicante), now
commercialized as N2Gsorb-6 (Gas to Materials Technologies) using
the following procedure: after degassing the material at 523 K for 4 h,
the sample was exposed to n-nonane (Aldrich, 99%) for 30 min at 77
K and then left in contact with the liquid for 3 h at room temperature.
Characterization. After degassing for at least 5 h at 423 K,
nitrogen physisorption experiments were carried out at 77 K on a
Quadrasorb apparatus (Quantachrome Instruments, U.S.A.). The high
resolution nitrogen physisorption isotherms were collected at 77 K on
an Autosorb 1C instrument (Quantachrome Instruments, U.S.A.).
Specific surface areas were calculated according to the Multi-Point
low uptake of nitrogen below a relative pressure of p/p = 0.1
0
due to the small micropore volume (Table 1). In contrast to
the controlled insertion of the micropores into OM-SiC-CDC-
BET method (p/p = 0.05−0.2 for OM-SiC-CDC-800 and CMK-3
0
8
00, the presence of micropores in CMK-3 is due to the
and 0.01−0.1 for the TiC-CDCs). Total pore volumes (TPVs) were
formation of defects during decomposition of the carbon
precursor. In consequence, they are not as well connected to
the mesopore system as the micropores in the hierarchical
CDC material.
determined from the amount adsorbed at p/p = 0.97. Pore size
0
distributions (PSDs) were calculated using the Quenched Solid
Density Functional Theory (QSDFT) method for nitrogen on carbon
5
2
(
slit/cylindrical pores, adsorption branch kernel). Micropore
volumes (MPVs) are estimated from the cumulative pore volumes at
a diameter of 2 nm. The Autosorb 1C was also used for the carbon
dioxide physisorption measurements at 273 K. Volumetric xenon
adsorption/desorption measurements at 237 ± 3 K were performed
on a Belsorb HP apparatus (Bel, Japan). Raman spectroscopic
investigations were carried out on a Holospec f/1.8 Spectrograph
Carbon dioxide physisorption measurements at 273 K up to
1
bar (p/p ≈ 0.03) additionally show the large differences in
0
the micropore structure of the investigated samples (Figure 1).
Under these conditions, CO adsorption is very sensitive to the
2
1
,28
presence of narrow micropores.
As a consequence, the
highest amount of gas at 1 bar is adsorbed in the pore system of
the TiC-CDC-600 due to the presence of the narrowest
micropores followed by TiC-CDC-1000 and OM-SiC-CDC-
800 in agreement with their increasing micropore size. CMK-3
shows the lowest uptake due to the lower micropore volume
and relatively large micropore diameter of 1.0 nm.
(
Kaiser Optical systems, U.S.A.) working at 785 nm excitation
wavelength.
1
29Xe NMR Spectroscopy. All 129
Xe NMR experiments were
4
9
carried out using a homemade in situ high-pressure apparatus.
Samples were transferred into the single crystal sapphire tube under
argon atmosphere (Glovebox). Afterward, the samples were activated
overnight using high vacuum (10−8 to 10 bar) to prevent artifacts
due to surface adsorption of atmospheric molecules. The application of
−7
The presence of distinct hysteresis loops in the nitrogen
physisorption isotherms of the OM-SiC-CDC-800 and CMK-3
relative pressures p/p up to 1 would require an absolute pressure of
0
at relative pressures of 0.4 < p/p < 0.7 indicates the presence
0
ca. 60 bar at room temperature. Although the sapphire tube withstands
such pressures, the tubing connecting the tube inside the magnet with
the outside pump rack and xenon reservoir does not allow this
pressure. At the chosen lower temperature (237 ± 3 K), however, a
relative pressure of 1 corresponds to ca. 19 bar which is well accessible
of small and narrowly distributed mesopores in these carbon
materials. Their sizes were calculated to be 4.1 and 4.5 nm for
OM-SiC-CDC-800 and CMK-3, respectively (SI Figure S3).
The good agreement between the QSDFT method and the
3
282
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
Two of the samples under study, TiC-CDC-1000 and TiC-
00 exclusively show micropores of 0.8 and 0.6 nm average
diameter. Correspondingly, the Xe NMR spectra of these
samples only exhibit one signal due to adsorbed xenon (Figure
6
129
2
). The highest chemical shift is observed for xenon adsorbed at
TiC-CDC-600. It should be noted that an inverse correlation
129
between the pore size and the Xe chemical shift is well-
54
known, e.g. for zeolites. The observed behavior therefore
indicates that the micropore diameter in TiC-CDC-600 is
smaller than in TiC-CDC-1000in agreement with the
nitrogen and carbon dioxide adsorption data (Table 1, Figure
Figure 1. Carbon dioxide adsorption/desorption (filled symbols/
empty symbols) isotherms of OM-SiC-CDC-800 (circles), CMK-3
1
, and SI Figure S1). It is also remarkable that the line width in
TiC-CDC-600 strongly exceeds the line width in the other
samples (Figure 2). This can be explained in two ways: (i)
Xenon is more strongly adsorbed in TiC-CDC-600, probably
due to the narrow micropores of 0.6 nm in size that is close to
the kinetic diameter of xenon (0.44 nm). (ii) Surface
inhomogeneity (despite the well-defined pore size) due to
the rather amorphous carbon microstructure in this sample can
(
squares), TiC-CDC-1000 (triangles), and TiC-CDC-600 (diamonds)
measured at 273 K.
experimental data for all samples (SI Figure S4) indicates that
this method calculates reliable pore-size distributions for the
investigated samples.
The 1 Xe NMR spectra measured at a temperature of 237 K
29
38
129
be another reason for line broadening. Recently, the higher
concentration of hydrogen atoms on the surface of TiC-CDC-
and a xenon pressure of 11 bar show at least two resolved Xe
NMR signals (Figure 2). The signals at low chemical shifts of
6
00 compared to materials obtained at higher chlorination
1
temperature such as TiC-CDC-1000 has been shown by H
53
MAS NMR measurements. (iii) In addition, xenon clusters of
different size may be formed within the micropores which
would give rise to a chemical shift distribution.
To better characterize the degree of graphitization of the
materials under study, Raman spectroscopic investigations have
been performed (SI Figure S5). Two characteristic bands in the
Raman spectra of such materials denoted as D and G are
5
5,56
usually considered (SI Figure S5,A).
The disorder-induced
D-band arises from a double-resonance process (intervalley
scattering), whereas the graphitic G-band is related to in-plane
2
stretching vibrations of sp sites. The increasing intensity of
both bands in carbon materials prepared at increasing
temperature indicates an increasing graphitization (i.e., the
2
growth and enlargement of the sp -bonded sites in the pore
walls). Up to a sample treatment temperature of ca. 1000 °C,
the D-band intensity increases even stronger than the G-
55
band. That means, the initially increasing I /I band intensity
D
G
ratio in carbide-derived carbon materials prepared at increasing
temperature indicates the proceeding graphitization, i.e., the
2
growth and enlargement of the sp -bonded sites in the pore
walls. Finally, at very high temperatures, the D-band starts to
decrease. The highest I /I ratio of 1.44 is observed for TiC-
D
G
CDC-1000. The lowest value of 1.26 occurs for TiC-CDC-600,
while OM-SiC-CDC-800 (1.43) and CMK-3 (1.43) are
comparable in graphitization to TiC-CDC-1000 (SI Figure
S5,B). These observations are characteristic for CDCs prepared
Figure 2. 129Xe NMR spectra for isothermal xenon adsorption on
OM-SiC-CDC-800, CMK-3, TiC-CDC-1000, and TiC-CDC-600
9
,13,17,55,56
at comparable temperatures.
In addition, the
measured at 237 K at a xenon pressure of 11 bar (p/p = 0.64).
decreasing full width at half-maximum of the D band indicates
a lower degree of graphitization in the carbon materials
0
1
3,16
prepared at lower temperatures (SI Figure S5,C).
That
ca. 0−20 ppm are due to xenon in the gas phase. The signals of
adsorbed xenon occur at higher chemical shifts of 100−300
ppm. In addition to the intense signal of adsorbed xenon and
the gas phase signal, a third signal appears for TiC-CDC-600
means the surface of TiC-CDC-600 is less graphitized and
contains more hydrogen atoms than the other porous carbon
materials.
Interestingly, two distinct signals can be detected for
adsorbed xenon in CMK-3 (Figure 2): signal (1) at ca. 190
ppm and signal (2) at ca. 230 ppm with an intensity ratio of
approximately 90:10, in agreement with previous observations
(
denoted by an asterisk in Figure 2) at ca. 35 ppm. Its intensity
amounts to only 4% of the signal of adsorbed xenon at 264
ppm. Due to the low chemical shift and intensity, the signal can
be attributed to macropores, e.g., between the particles. Xenon
in these macropores interacting with the outer surface of the
TiC-CDC-600 particles would give rise to such a chemical shift.
46
for CMK-3. These signals were interpreted as being due to
mesopores (1) and micropores (2). In agreement with nitrogen
and carbon dioxide physisorption isotherms, (Table 1, Figure 1,
3
283
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
and SI Figure S1), the chemical shifts of the micropore signal
are lower than those observed for the micropores in the TiC-
CDCs due to the larger micropores in CMK-3. The mesopores
represent a more than ten times higher fraction of the overall
pore volume. At 11 bar, signal (1) exhibits an asymmetry which
becomes even more pronounced at higher pressure, finally
resulting in two closely neighbored signals (SI Figure S6). This
can be attributed to the presence of two slightly different
mesopores due to the partial presence of tubular structures
within the carbon nanorods. At lower pressures, rapid exchange
results in an averaged signal. However, this exchange obviously
becomes slow at increasing pore filling.
signal intensity of the two signals in CMK-3 (Figure 4). Both
the chemical shift as well as the intensity of the signal (1)
The pressure-dependence of the NMR chemical shift of the
xenon adsorbed in purely microporous samples TiC-CDC-600
(
Figure 3) and TiC-CDC-1000 (SI Figure S7) closely
Figure 3. 129Xe NMR chemical shift for TiC-CDC-600 measured at
37 K as a function of relative pressure.
2
Figure 4. Top: 129Xe NMR chemical shift of the signals due to
resembles a type I adsorption/desorption isotherm. Note that
the chemical shift steeply increases up to a relative pressure p/
p0 of ca. 0.2 for the micropores. Beyond a relative pressure of
ca. 0.2, the chemical shift levels off and reaches a constant value.
mesopores and micropores observed for CMK-3 at 237 K as a function
129
of relative pressure. Bottom: Xe NMR signal intensity of the signals
due to mesopores and micropores observed for CMK-3 at 237 K as a
function of relative pressure.
This indicates that the pore system is filled with xenon at
relative pressures of p/p = 0.2 for microporous samples. This
0
behavior could be expected for the signal intensity since the
Xe NMR signal intensity represents the amount of adsorbed
follow a dependence characteristic for adsorption isotherms of
mesoporous materials (cf. Figures 2 and 4). In contrast to TiC-
CDCs, the final chemical shift value for xenon inside the
mesopores is close to 200 ppm, since the xenon condensed
inside the pores is similar to the bulk liquid xenon. For signal
(2), the curves resemble type I adsorption isotherms which are
characteristic for microporous materials. Over the entire range
of relative pressure, the chemical shift of the signal (2) is lower
compared to the TiC-CDCs due to the larger size of the
micropores in CMK-3.
1
29
xenon in an isothermal adsorption/desorption measurement.
However, the 1 Xe NMR chemical shift is also correlated with
the xenon density ρ, i.e., the amount of adsorbed xenon (eq 1).
Therefore, the pressure-dependence of the chemical shift
should also resemble the shape of the adsorption/desorption
isothermas it is indeed observed. Over the entire relative
pressure range, TiC-CDC-600, which exhibits the smallest
average pore diameter, shows the highest chemical shift,
followed by TiC-CDC-1000. It is furthermore remarkable that
the “saturation value” of the chemical shift of xenon in
micropores clearly exceeds the chemical shift value of 203 ppm
29
The advantage of the chemical shift measurement is its
inherently high accuracy compared with the measurement of
the signal intensity. At the chosen measurement temperature of
237 K, neither the volumetrically measured adsorption/
desorption isotherm (SI Figure S8) nor the 1 Xe NMR signal
intensity measurements (Figure 4, bottom) show a hysteresis
which is observed at lower temperature in the nitrogen
adsorption isotherms (SI Figure S1). However, the very
sensitive 1 Xe NMR chemical shift clearly reflects the onset
of a corresponding hysteresis behavior at relative pressures p/p0
of ca. 0.5−0.6 (Figure 4, top and Figure 5).
49
observed for liquid xenon at 237 K. This is due to the fact
that xenon−wall interactions are predominating in microporous
materials even if the pore system is completely filled. In
contrast, the final chemical shift values for xenon inside
mesopores is close to 200 ppm since the xenon condensed
within the mesopores is similar to the bulk liquid (see below).
It is interesting to compare these observations with the
29
29
1
29
pressure-dependence of the Xe NMR chemical shift and the
3
284
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
2
03 ppm expected for liquid xenon at the experiment
49
temperature of 237 K. In contrast, the signal due to xenon
atoms in the mesopores of CMK-3 only exhibits a δmax-value of
2
05 ± 3 ppm close to the value expected for liquid xenon. This
observation shows that the δmax-value pronouncedly depends
on the presence of xenon−wall interactions and is obviously
correlated with the average pore diameter. Advantageously, δ
max
can be easily measured since the slope of the chemical shift in
the high-pressure regime is small and the signal intensity is
highin contrast to the chemical shift δ determined by
S
extrapolation to zero pressure (see above, eq 1). We, therefore,
suggest the use of δmax as a measure for the pore size which can
be determined precisely in an inherently sensitive high-pressure
129
Xe NMR experiment.
In addition to the mesoporous CMK-3 and the TiC-CDCs,
we have investigated the hierarchically ordered mesoporous
carbon sample, OM-SiC-CDC-800, which contains significant
amounts of both, mesopores and micropores (Table 1). In
contrast to CMK-3, however, only one single narrow signal of
adsorbed xenon is observed for OM-SiC-CDC-800 (see Figure
2
). Interestingly, the signal appears at 209 ppm, i.e., in between
the chemical shift values of micropores (230−260 ppm) and
the chemical shift of xenon inside the mesopores of CMK-3
(
190 ppm) at 11 bar xenon pressure. In contrast to CMK-3, the
well interconnected micro- and mesopores in OM-SiC-CDC-
00 obviously allow a fast exchange of adsorbed xenon between
Figure 5. 129Xe NMR spectra of CMK-3 measured after xenon
8
adsorption and during xenon desorption at 237 K for 8 bar (p/p =
0
micro- and mesopores. As a result of this fast exchange, a signal
at the averaged chemical shift is detected as previously reported
for the adsorption of electrolyte ions in this hierarchically
0
.46) and 10 bar (p/p = 0.58). Note that the chemical shift for xenon
0
inside the mesopores (signal (1)) is different for the adsorption and
desorption spectra at identical pressures (hysteresis) in contrast to the
micropore signal (2).
53
structured carbon material. This is observed over the full
relative pressure range up to p/p = 1 (Figure 6). It is also
0
The presence of two distinct signals for mesopores and
micropores in CMK-3 shows that the exchange between the
two pore systems must be slow at the NMR time scale, i.e.,
compared with the critical time constant which is defined by
−1
(
2πΔν) . Δν denotes the frequency difference between the
two signals which amounts to 3300 Hz (40 ppm) at 11 bar
Figure 2). Thus, the exchange of xenon between micro- and
(
mesopores must be considerably slower than the critical time
constant of 50 μs. Indeed, 2D Exchange Spectroscopy (EXSY)
was unable to detect any exchange between the different pore
systems even at a mixing time of 500 ms (SI Figure S9). It
should be noted that an exchange of xenon between micro- and
mesopores of CMK-3 could be detected at room temperature
and low xenon pressure p = 0.53 bar within a mixing time of
Xe
4
6
5
0 ms. At 237 K and high pressure, the mobility of xenon is
obviously very limited, which effectively suppresses xenon
exchange between the two pore systems in CMK-3. In addition,
the accessibility of the micropores in CMK-3 might be hindered
due to narrow entrances resulting from their uncontrolled
growth during carbonization further hindering the exchange of
xenon between both pore systems.
Figure 6. 129Xe NMR chemical shift for OM-SiC-CDC-800 measured
at 237 K as a function of relative pressure. For comparison, the
corresponding data for CMK-3 are given.
Note that the micropores in the three samples, TiC-CDC-
00, TiC-CDC-1000, and CMK-3 exhibit a similar general
remarkable that the chemical shift δmax extrapolated to a relative
pressure of 1 amounts to ca. 225 ± 3 ppm. This is in between
the value observed for mesopores (205 ppm) and micropores
(235 ppm) in CMK-3 (see above). This is in line with the
conclusion that the xenon atoms exchange rapidly between
mesopores and micropores in OM-SiC-CDC-800. Moreover,
the observed average signal exhibits a lower line width than the
other samples under study. This also indicates a high mobility
of the xenon and/or well-defined pore size of this material.
Finally, we have characterized the hierarchical samples after
6
dependence on the relative pressure resembling a type I
adsorption isothermbut with pronounced quantitative differ-
ences. In particular, the “plateau value” δ of the chemical
max
shift obtained by extrapolation of the chemical shift to p/p = 1
0
is obviously influenced by the pore size. It amounts to 260 ± 3
ppm for TiC-CDC-600 (0.6 nm average pore diameter), 250 ±
3
2
ppm for TiC-CDC-1000 (0.8 nm average pore diameter), and
35 ± 3 ppm for the micropores in CMK-3 (1 nm average pore
1
29
diameter). These values are all larger than the chemical shift of
n-nonane loading using
Xe NMR spectroscopy. This
3
285
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
treatment preferentially closes the micropores thus making
them practically inaccessible for other molecules, whereas the
mesopore system remains accessible. Note that the presence of
a low but measurable amount of micropores in CMK-3 as well
as a well-interconnected hierarchical pore structure in the OM-
SiC-CDC-800 with large contributions from both micro- and
mesopores were demonstrated by nitrogen physisorption
penetrate into the larger pores explaining the shift of the
hysteresis loop toward lower relative pressure.
2
6,28
A similar situation is observed in OM-SiC-CDC-800. An
even larger volume of micropores is blocked by the
hydrocarbon molecules significantly decreasing the overall
porosity of OM-SiC-CDC-800 + n-nonane compared with
28
129
CMK-3. This should result in a significantly decreasing Xe
NMR chemical shift provided the mesopores remain
uninfluenced by the nonane. However, the signal of the
adsorbed xenon atoms shows a minor change toward higher
chemical shift (Figure 7, bottom and SI Figure S11). That
means, loading with n-nonane also influences the mesopores as
observed for CMK-3, which gives rise to the increased chemical
shift. Two effects may contribute to this chemical shift increase,
namely (i) the decreasing pore size of the mesopores due to the
presence of n-nonane at the pore walls and (ii) a stronger
interaction between xenon and the nonane molecules
compared with the graphene-like pore walls in nonane-free
2
8
measurements before and after n-nonane-preadsorption.
The 1 Xe NMR spectra of the two mesoporous samples
CMK-3 and OM-SiC-CDC-800 before and after n-nonane
loading at a pressure of 11 bar and 237 K (Figure 7) prove the
29
53
samples.
CONCLUSIONS
■
In situ high pressure 129Xe NMR spectroscopy in combination
with volumetric adsorption measurements and n-nonane
preadsorption were used for the textural characterization of
different carbon materials with well-defined porosity. While
xenon adsorbed in the micro- and mesopores of CMK-3 gives
two different signals due to their spatial separation, only one
averaged signal can be observed in the strongly hierarchical
OM-SiC-CDC-800, due to the well-connected pore systems
allowing fast exchange of adsorbed xenon. This method can be
used to clearly distinguish between different micropore sizes as
shown by the investigations on two TiC-derived carbons with
different pore size. Full adsorption/desorption isotherms up to
relative pressures p/p close to 1 were recorded. The chemical
0
shift δmax extrapolated to a relative pressure of 1 is suggested as
a sensitive measure for the micropore diameter. It can be
measured with high accuracy and sensitivity. The agreement
with the volumetric adsorption data renders this method as
highly useful for the pore structure analysis of different carbon
materials. The changes in the spectra of the hierarchical
materials after the blocking of the micropores with n-nonane
provides information about the adsorption state of the
hydrocarbon molecules in the carbon structure leading to a
better understanding of the connectivity of micro- and
mesopores which is of interest for various applications.
Figure 7. 129Xe NMR spectra of the mesoporous samples CMK-3 and
OM-SiC-CDC-800 before and after loading with n-nonane measured
at 237 K at a xenon pressure of 11 bar (p/p = 0.64).
0
different pore structures present in both samples. The
quantitative blocking of the micropores by the hydrocarbon
molecules causes the disappearance of the signal due to
adsorbed xenon in the micropores of CMK-3 at ca. 230 ppm.
However, the signal of xenon adsorbed in the mesopores is
located at higher chemical shift after n-nonane loading. This
observation indicates that the mesopores are also modified by
hydrocarbon loading. The pressure-dependence of the chemical
shift (SI Figure S10) shows that the hysteresis loop shifts
toward a slightly lower relative pressure compared to the
nonane-free material. These observations indicate that the
mesopores become smaller during n-nonane loading. Moreover,
the chemical shift value δmax of the xenon adsorbed in the
mesopores extrapolated to relative pressures of 1 is also slightly
larger than before n-nonane loading indicating a stronger
influence of xenon−wall interactions caused by the presence of
n-nonane in the mesopores. It is known for hierarchically
structured materialsespecially for materials with an ordered
mesopore structurethat n-nonane molecules blocking the
micropores in the neighborhood of mesopores can also partly
ASSOCIATED CONTENT
Supporting Information
■
*
S
Figure S1: Nitrogen adsorption/desorption isotherms meas-
ured at 77 K. Figure S2: Semilogarithmic plot of the nitrogen
adsorption/desorption isotherms. Figure S3: QSDFT pore-size
distributions. Figure S4: QSDFT fitting comparisons. Figure
S5: Raman spectra, I /I ratio and fwhm of the D band. Figure
D
G
129
S6: Pressure-dependence of the Xe NMR spectrum of xenon
1
29
adsorbed on CMK-3 measured at 237 K. Figure S7: Xe NMR
chemical shift for TiC-CDC-1000 measured at 237 K as a
function of relative pressure. Figure S8: Volumetric xenon
adsorption/desorption isotherm of CMK-3 at 237 K. Figure S9:
2D EXSY spectrum of xenon adsorbed at 10 bar and 237 K on
CMK-3. Figure S10: 1 Xe NMR chemical shift for CMK-3
after n-nonane loading measured at 237 K as a function of
relative pressure. Figure S11: 1 Xe NMR chemical shift for
OM-SiC-CDC-800 after n-nonane loading measured at 237 K
29
29
3
286
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
as a function of relative pressure. This material is available free
of charge via the Internet at http://pubs.acs.org.
(16) Borchardt, L.; Hasche, F.; Lohe, M. R.; Oschatz, M.; Schmidt,
F.; Kockrick, E.; Ziegler, C.; Lescouet, T.; Bachmatiuk, A.; Buechner,
B.; Farrusseng, D.; Strasser, P.; Kaskel, S. Carbon 2012, 50, 1861.
(
17) Oschatz, M.; Borchardt, L.; Pinkert, K.; Thieme, S.; Lohe, M. R.;
AUTHOR INFORMATION
■
Hoffmann, C.; Benusch, M.; Wisser, F.; Ziegler, C.; Giebeler, L.;
Ruemmeli, M.; Eckert, J.; Eychmueller, A.; Kaskel, S. Adv. Energy
Mater. 2014, DOI: 10.1002/aenm.201300645.
Corresponding Authors
E-mail: stefan.kaskel@chemie.tu-dresden.de.
E-mail: eike.brunner@tu-dresden.de.
*
(
18) Chmiola, J.; Yushin, G.; Gogotsi, Y.; Portet, C.; Simon, P.;
Taberna, P. L. Science 2006, 313, 1760.
19) Deng, Y.; Liu, C.; Yu, T.; Liu, F.; Zhang, F.; Wan, Y.; Zhang, L.;
Wang, C.; Tu, B.; Webley, P. A.; Wang, H.; Zhao, D. Chem. Mater.
007, 19, 3271.
20) Chuenchom, L.; Kraehnert, R.; Smarsly, B. M. Soft Matter. 2012,
8, 10801.
*
Present Address
Institute for Chemical and Bioengineering, ETH Zu
Wolfgang-Pauli-Strasse 10, CH-8093 Zurich, Switzerland.
(
⊥
̈
rich,
2
(
Author Contributions
#
The manuscript was written through contributions of all
authors. M.O., H.C.H., J.P., and J.S. contributed equally to the
presented work. All authors have given approval to the final
version of the manuscript.
(
(
21) Lu, A.-H.; Schueth, F. Adv. Mater. 2006, 18, 1793.
22) Kruk, M.; Jaroniec, M.; Kim, T.-W.; Ryoo, R. Chem. Mater.
2
(
003, 15, 2815.
23) Rose, M.; Korenblit, Y.; Kockrick, E.; Borchardt, L.; Oschatz,
M.; Kaskel, S.; Yushin, G. Small 2011, 7, 1108.
Funding
(
(
24) Thommes, M. Annu. Rev. Nano Res. 2010, 3, 515.
25) Silvestre-Albero, J.; Silvestre-Albero, A.; Rodriguez-Reinoso, F.;
Financial support from the Deutsche Forschungsgemeinschaft
DFG Grant No. Br 1278/20-2) is gratefully acknowledged.
(
Thommes, M. Carbon 2012, 50, 3128.
26) Silvestre-Albero, A.; Jardim, E. O.; Bruijn, E.; Meynen, V.; Cool,
Notes
(
The authors declare no competing financial interest.
P.; Sepulveda-Escribano, A.; Silvestre-Albero, J.; Rodriguez-Reinoso, F.
Langmuir 2009, 25, 939.
ACKNOWLEDGMENTS
(27) Silvestre-Albero, A.; Goncalves, M.; Itoh, T.; Kaneko, K.; Endo,
M.; Thommes, M.; Rodriguez-Reinoso, F.; Silvestre-Albero, J. Carbon
■
Thanks are due to Dr. S. Paasch for excellent experimental
assistance.
2
(
012, 50, 66.
28) Oschatz, M.; Borchardt, L.; Rico-Frances, S.; Rodriguez-
Reinoso, F.; Kaskel, S.; Silvestre-Albero, J. Langmuir 2013, 29, 8133.
(29) Ito, T.; Fraissard, J. Proc. Int. Conf. Zeolites 5th 1980, 510.
REFERENCES
■
(
(
30) Clewett, C. F. M.; Pietra, T. J. Phys. Chem. B 2005, 109, 17907.
31) Bansal, N.; Foley, H. C.; Lafyatis, D. S.; Dybowski, C. Catal.
(
1) Silvestre-Albero, A.; Rico-Frances, S.; Rodriguez Reinoso, F.;
Kern, A. M.; Klumpp, M.; Etzold, B. J. M.; Silvestre-Albero, J. Carbon
Today 1992, 14, 305.
32) Ago, H.; Tanaka, K.; Yamabe, T.; Miyoshi, T.; Takegoshi, K.;
Terao, T.; Yata, S.; Hato, Y.; Nagura, S.; Ando, N. Carbon 1997, 35,
781.
2
(
013, 59, 221.
2) Gogotsi, Y.; Dash, R. K.; Yushin, G.; Yildirim, T.; Laudsio, G.;
Fischer, J. E. J. Am. Chem. Soc. 2005, 127, 16006.
3) Wang, J.; Senkovska, I.; Oschatz, M.; Lohe, M. R.; Borchardt, L.;
Heerwig, A.; Liu, Q.; Kaskel, S. ACS Appl. Mater. Interfaces 2013, 5,
160.
4) Galeano, C.; Baldizzone, C.; Bongard, H.; Spliethoff, B.;
Weidenthaler, C.; Meier, J. C.; Mayrhofer, K. J. J.; Schueth, F. Adv.
Funct. Mater. 2014, 24, 220.
5) Vu, A.; Li, X.; Phillips, J.; Han, A.; Smyrl, W. H.; Buhlmann, P.;
Stein, A. Chem. Mater. 2013, 25, 4137.
6) Zhai, Y.; Dou, Y.; Zhao, D.; Fulvio, P. F.; Mayes, R. T.; Dai, S.
Adv. Mater. 2011, 23, 4828.
(
1
(
(33) McGrath, K. J. Carbon 1999, 37, 1443.
(34) Raftery, D.; Long, H.; Meersmann, T.; Grandinetti, P. J.; Reven,
3
(
L.; Pines, A. Phys. Rev. Lett. 1991, 66, 584.
(
(
35) Tsiao, C.; Botto, R. E. Energy Fuels 1991, 5, 87.
36) Saito, K.; Kimura, A.; Fujiwara, H. Magn. Reson. Imaging 2003,
2
(
1, 401.
37) Romanenko, K. V.; d’Espinose, d. l. C. J.-B.; Fraissard, J.;
Reshetenko, T. V.; Lapina, O. B. Microporous Mesoporous Mater. 2005,
1, 41.
38) Suh, D. J.; Park, T. J.; Ihm, S. K.; Ryoo, R. J. Phys. Chem. 1991,
5, 3767.
39) Radovic, L. R.; Menon, V. C.; Leon, Y. L. C. A.; Kyotani, T.;
Danner, R. P.; Anderson, S.; Hatcher, P. G. Adsorption 1997, 3, 221.
40) Simonov, P. A.; Filimonova, S. V.; Kryukova, G. N.; Moroz, E.
(
(
8
(
9
(
(
(
7) Rodriguez-Reinoso, F. Carbon 1998, 36, 159.
8) Taguchi, A.; Schueth, F. Microporous Mesoporous Mater. 2004, 77,
1.
(
9) Porada, S.; Borchardt, L.; Oschatz, M.; Bryjak, M.; Atchison, J. S.;
Keesman, K. J.; Kaskel, S.; Biesheuvel, P. M.; Presser, V. Energy
Environ. Sci. 2013, 6, 3700.
10) Yushin, G.; Hoffman, E. N.; Barsoum, M. W.; Gogotsi, Y.;
Howell, C. A.; Sandeman, S. R.; Phillips, G. J.; Lloyd, A. W.;
(
M.; Likholobov, V. A.; Kuretzky, T.; Boehm, H. P. Carbon 1999, 37,
591.
(41) Chen, F.; Chen, C.-L.; Ding, S.; Yue, Y.; Ye, C.; Deng, F. Chem.
Phys. Lett. 2004, 383, 309.
(42) Garsuch, A.; Boehlmann, W.; Sattler, R. R.; Fraissard, J.; Klepel,
(
Mikhalovsky, S. V. Biomaterials 2006, 27, 5755.
11) Presser, V.; Yeon, S.-H.; Vakifahmetoglu, C.; Howell, C. A.;
(
Sandeman, S. R.; Colombo, P.; Mikhalovsky, S.; Gogotsi, Y. Adv.
Healthcare Mater. 2012, 1, 796.
12) Presser, V.; Heon, M.; Gogotsi, Y. Adv. Funct. Mater. 2011, 21,
10.
13) Oschatz, M.; Borchardt, L.; Thommes, M.; Cychosz, K. A.;
Senkovska, I.; Klein, N.; Frind, R.; Leistner, M.; Presser, V.; Gogotsi,
Y.; Kaskel, S. Angew. Chem., Int. Ed. 2012, 51, 7577.
14) Oschatz, M.; Kockrick, E.; Rose, M.; Borchardt, L.; Klein, N.;
O. Carbon 2006, 44, 1173.
(43) Romanenko, K. V.; Py, X.; D’Espinose, d. L. J.-B.; Lapina, O. B.;
Fraissard, J. J. Phys. Chem. B 2006, 110, 3055.
(44) Romanenko, K. V.; d’Espinose, d. L. J.-B.; Lapina, O.; Fraissard,
J. Microporous Mesoporous Mater. 2007, 105, 118.
(45) Gotoh, K.; Ueda, T.; Eguchi, T.; Kawabata, K.; Yamamoto, K.;
Murakami, Y.; Hayakawa, S.; Ishida, H. Bull. Chem. Soc. Jpn. 2009, 82,
1232.
(
8
(
(
Senkovska, I.; Freudenberg, T.; Korenblit, Y.; Yushin, G.; Kaskel, S.
(46) Onfroy, T.; Guenneau, F.; Springuel-Huet, M.-A.; Gedeon, A.
Carbon 2010, 48, 3987.
Carbon 2009, 47, 2352.
(
15) Kockrick, E.; Borchardt, L.; Schrage, C.; Gaudillere, C.; Ziegler,
(47) Romanenko, K. V.; Fonseca, A.; Dumonteil, S.; Nagy, J. B.;
d’Espinose, d. L. J. B.; Lapina, O. B.; Fraissard, J. Solid State Nucl.
Magn. Reson. 2005, 28, 135.
C.; Freudenberg, T.; Farrusseng, D.; Eychmueller, A.; Kaskel, S. Chem.
Mater. 2011, 23, 57.
3
287
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288

Chemistry of Materials
Article
(
48) Bonardet, J.-L.; Fraissard, J.; Gedeon, A.; Springuel-Huet, M.-A.
Catal. Rev. - Sci. Eng. 1999, 41, 115.
49) Hoffmann, H. C.; Assfour, B.; Epperlein, F.; Klein, N.; Paasch,
S.; Senkovska, I.; Kaskel, S.; Seifert, G.; Brunner, E. J. Am. Chem. Soc.
011, 133, 8681.
50) Klein, N.; Herzog, C.; Sabo, M.; Senkovska, I.; Getzschmann, J.;
(
2
(
Paasch, S.; Lohe, M. R.; Brunner, E.; Kaskel, S. Phys. Chem. Chem.
Phys. 2010, 12, 11778.
(
51) Springuel-Huet, M.-A.; Nossov, A.; Adem, Z.; Guenneau, F.;
Volkringer, C.; Loiseau, T.; Ferey, G.; Gedeon, A. J. Am. Chem. Soc.
010, 132, 11599.
52) Gor, Y. G.; Thommes, M.; Cychosz, K. A.; Neimark, A. V.
Carbon 2012, 50, 1583.
53) Borchardt, L.; Oschatz, M.; Paasch, S.; Kaskel, S.; Brunner, E.
Phys. Chem. Chem. Phys. 2013, 15, 15177.
2
(
(
(
54) Demarquay, J.; Fraissard, J. Chem. Phys. Lett. 1987, 136, 314.
(
55) Gogotsi, Y.; Nikitin, A.; Ye, H.; Fischer, J. E.; Yi, B.; Foley, H.
C.; Barsoum, M. W. Nat. Mater. 2003, 2, 591.
56) Osswald, S.; Chmiola, J.; Gogotsi, Y. Carbon 2012, 50, 4880.
(
3
288
dx.doi.org/10.1021/cm501102y | Chem. Mater. 2014, 26, 3280−3288