Independent verification of the saturation hydrogen uptake in MOF-177 and establishment of a benchmark for hydrogen adsorption in metal-organic frameworks

Article abstract of DOI:10.1039/b703608f

Hydrogen isotherms for MOF-177, Zn₄O(1,3,5-benzenetribenzoate) ₂, crystals were independently measured by volumetric and gravimetric methods at 77 K to confirm its hydrogen uptake capacity and to establish the importance of calibrating gas adsorption instrumentation prior to evaluating H₂ storage capacities. Reproducibility of hydrogen adsorption experiments is important because non-systematic errors in measurements can easily occur leading to erroneous reports of capacities. The surface excess weight percentage of hydrogen uptake in MOF-177 samples is 7.5 wt% at 70 bar, which corresponds to an absolute adsorbed amount of 11 wt%. These values are in agreement with our previous report and with those found independently by Southwest Research Institute. Considering its well-known structure and its significant H₂ uptake properties, we believe MOF-177 is an excellent material to serve as a benchmark adsorber. The Royal Society of Chemistry.

Full text of DOI:10.1039/b703608f

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Independent verification of the saturation hydrogen uptake in MOF-177
and establishment of a benchmark for hydrogen adsorption in
metal–organic frameworks{{
Hiroyasu Furukawa,*a Michael A. Millerb and Omar M. Yaghi*a
Received 9th March 2007, Accepted 23rd April 2007
First published as an Advance Article on the web 10th May 2007
DOI: 10.1039/b703608f
Hydrogen isotherms for MOF-177, Zn4O(1,3,5-benzenetribenzoate)2, crystals were independently
measured by volumetric and gravimetric methods at 77 K to confirm its hydrogen uptake capacity
and to establish the importance of calibrating gas adsorption instrumentation prior to evaluating
H2 storage capacities. Reproducibility of hydrogen adsorption experiments is important because
non-systematic errors in measurements can easily occur leading to erroneous reports of capacities.
The surface excess weight percentage of hydrogen uptake in MOF-177 samples is 7.5 wt% at
70 bar, which corresponds to an absolute adsorbed amount of 11 wt%. These values are in
agreement with our previous report and with those found independently by Southwest Research
Institute. Considering its well-known structure and its significant H2 uptake properties, we believe
MOF-177 is an excellent material to serve as a benchmark adsorber.
any porous materials and clearly show that in principle the
Introduction
DOE targets can be achieved at 77 K. In light of these
The discovery of hydrogen adsorption in porous metal–
developments and the extensive work being done on MOFs
we propose the establishment of MOF-17712 as a benchmark
organic frameworks (MOFs) and subsequent related studies
has firmly established these materials as interesting candidates
material for researchers in the field. This is important as the
for hydrogen storage applications.1–8 This is due to the
field of hydrogen storage has often suffered from reports
availability of a large number of well-characterized MOFs
of high hydrogen uptake which were later found to be
erroneous.13 Accordingly, our DOE program officers have
and to the flexibility with which their organic and inorganic
components can be varied.9 In a recent review10 we delineated
advised that independent measurements of hydrogen uptake
numerous strategies that can be used in MOF chemistry for
should be performed at a DOE-approved facility (Southwest
Research Institute1; SwRI) to independently verify hydrogen
achieving the targets for on-board hydrogen storage systems
set by the US Department of Energy (DOE) for use of
uptake measurements reported for MOF-177. Here we report
hydrogen as a fuel. The gravimetric (6.0 wt%) and volumetric
jointly on the results of the verification tests, point out some of
(45 g L21) densities to be achieved by 201011 are important in
the intricacies of hydrogen measurements, and suggest a
the context of the present report as many research groups are
standard for reporting results of hydrogen adsorption.
aiming to achieve these targets.
We remark that MOF-177 is an ideal material to use as a
Recently, under a DOE-funded program, we showed that
benchmark for hydrogen adsorption measurements on MOF
MOF-177 (Fig. 1) can store 7.5 wt% hydrogen with a
materials due to the following facts: (1) it has high gravimetric
volumetric capacity of 32 g L21 at 77 K and 70 bar.4
Another MOF was found to store 5.1 wt% with a volumetric
capacity of 43 g L21 at 77 K and 45 bar.5 These results are
exciting as these MOFs exhibit the highest hydrogen uptake of
and volumetric uptake capacity4 that is 3–5 times higher
than in well-studied commercial materials, such as activated
carbons or zeolites,14 leading to ease of high accuracy mea-
surements; (2) its synthesis is simple and highly reproducible;15
aDepartment of Chemistry and Biochemistry, Center for Reticular
Chemistry at the California NanoSystems Institute, University of
California, Los Angeles, 607 Charles E. Young Drive East, Los Angeles,
California 90095-1569, USA. E-mail: furukawa@chem.ucla.edu;
yaghi@chem.ucla.edu; Tel: +1-310-206-0398
bDepartment of Materials Engineering, National Testing Laboratory for
Solid-State Hydrogen Storage Technologies, Southwest Research
Institute, 6220 Culebra Road, San Antonio, Texas 78238-5166, and
Department of Chemistry, University of Texas at San Antonio, One
UTSA Circle, San Antonio, Texas 78249-1644, USA.
Tel: +1-210-522-2189
{ This paper is part of a Journal of Materials Chemistry theme issue on
New Energy Materials. Guest editor: M. Saiful Islam.
{ Electronic supplementary information (ESI) available: synthesis of
H3BTB link, powder X-ray diffraction pattern and TGA analysis for
MOF-177, and adsorption data. See DOI: 10.1039/b703608f
Fig. 1 (A) Molecular structure of the BTB link and (B) crystal
structure of MOF-177. An approximately spherical pore is shown as
˚
the large sphere with a 17 A diameter.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3197–3204 | 3197

View Online
and (3) it has a crystalline structure that is well-characterized
in terms of atomic connectivity and chemical composition.
In this study, independent investigations were undertaken
to assess the H2 saturation uptake on MOF-177 samples
using volumetric and gravimetric instruments located at the
University of Michigan, UCLA and SwRI. We demonstrate
the validity of MOF-177 as a benchmark material based on
the agreement between volumetric (7.5 wt% at SwRI) and
gravimetric (7.3 wt% at UCLA) H2 adsorption measurements
at 77 K, which are in agreement with our original report4
(volumetric data at UM, 7.5 wt%).
volumetric analyzer at 77 K using a liquid nitrogen (LN2)
bath.16 H2 isotherms were collected at 87 K using a liquid
argon bath. The analysis station of the volumetric adsorption
apparatus was equipped, in addition to the standard pressure
transducers in the dosing volume (manifold) of the apparatus,
with high-precision pressure transducers (Baratron MKS)
dedicated to read the pressure in the sample cell alone.
Hence, isolation of the sample cell during equilibration
ensured a very small effective dead volume and, therefore, a
highly accurate determination of the amount of gas adsorbed.
To provide high accuracy and precision in the determination
of P/P0, the saturation pressure P0 was measured throughout
the entire analysis by means of a dedicated saturation pressure
transducer, which allowed us to monitor the vapor pressure for
each data point. It should be noted that H2 adsorption data
should not be displayed as P/P0 because there is no saturation
pressure defined for supercritical H2. An ultra-high purity
(UHP, 99.999% purity) grade of N2, H2 and He gases was used
throughout the adsorption experiments.
Experimental
Synthesis of MOF-177
N,N-Diethylformamide (DEF, BASF) was treated by addition
of charcoal carbon (ca. 0.5 g, Acros, CAS No.: 7440-44-0) into
1 L and stirred for 12 h at room temperature. The black
suspension solution was filtered through a silica-gel column
(ca. 350 g, Silica gel 60, 230–400 mesh, EMD Chemicals) to
remove carbon powder. Pure MOF-177 was prepared by
dissolving Zn(NO3)2?6H2O (0.187 g, 0.63 mmol, Aldrich) and
4,49,40-benzene-1,3,5-triyl-tribenzoic acid{ (H3BTB, 0.0395 g,
0.090 mmol) in DEF (10 mL) in a 20 mL vial. The vial was
capped tightly and placed in an oven at 85 uC for two days to
yield clear block crystals. After cooling, the yellow solution
was decanted, and the crystals were washed with 3 6 20 mL
N,N-dimethylformamide (reagent-grade, Fisher). The product
was then immersed in chloroform (20 mL, pentene stabilized,
reagent-grade, Fisher) for three days, during which time the
activation solvent was decanted and freshly replenished three
times. The crystals were transferred to a glass tube (7 mm
i.d. 6 300 mm length) as a chloroform suspension by using a
glass pipette to prevent exposure to air. The material was
evacuated to 1023 Torr at room temperature for 2 h, heated at
a constant rate (1 uC min21) to 120 uC for 6 h, then cooled at a
constant rate of 1 uC min21 to room temperature. Powder
X-ray diffraction patterns for as-synthesized and activated
materials were coincident and in agreement with the simulated
pattern, indicating the bulk purity of MOF-177 samples.
Although MOF-177 will not be decomposed even if the
samples are quickly transferred to a gas sorption cell in the air,
activated MOF-177 samples were handled using standard
techniques to avoid exposure to air or moisture.
Volumetric high-pressure gas adsorption measurements at
Southwest Research Institute
Low-temperature hydrogen adsorption isotherms were inde-
pendently measured (SwRI) from the same synthetic batch of
MOF-177 as in ref. 4 (see Table 1). Approximately 210 mg of
MOF-177 sample was transferred to a high-pressure sample
vessel constructed from 316L stainless-steel tubing (4.6 mm
i.d. 6 45 mm length). A polished stainless-steel dowel was
used to lightly compact the sample in the vessel, thereby
minimizing the intergranular free volume. The remaining free
volume above the sample was minimized by inserting segments
of polished stainless-steel rod, precisely milled to be slightly
smaller than the inner diameter of the vessel, and accommodat-
ing a center bore for the transfer of gas. A high-pressure Sieverts-
type volumetric analyzer (Hy-Energy Scientific Instruments,
Model PCTPro2000) was used to quantify the isothermal Gibbs
excess for H2 adsorption in the MOF-177 sample at 77 K.16
Prior to analysis, the total free volume of the sample and vessel
was determined from the expansion of low-pressure helium
(,5 bar) while maintaining isothermal conditions. A typical free
volume determined for this system was 9.50 ¡ 0.03 cm3. The
assumption that helium does not interact adsorptively with
MOF-177 was independently confirmed by measuring the
helium-derived free volume of pure silicon granules (99.999%,
Aesar) at low pressures (,5 bar), using a mass of silicon
equivalent in volume to the helium-derived volume previously
determined for the sample. The difference between the two free
volume measurements was in fact not measurable.
Low-pressure gas adsorption measurements
Low-pressure N2 and H2 adsorption measurements (up to
1 bar) were performed on an Autosorb-1 (Quantachrome)
Table 1 Summary of high-pressure H2 adsorption measurements for MOF-177
Surface excess amount
Absolute adsorbed amount
Investigator
Surface areaa/m2 g21
mg g21
g L21
(pressure/bar)
mg g21
g L21
(pressure/bar)
Method
SwRI #1
SwRI #2
UCLA
UMb
5640 (4750)
5640 (4750)
5250 (4630)
5640 (4750)
b
76
74
73
75
32
32
31
32
(66)
(57)
(52)
(69)
112
115
111
114
48
49
48
49
(72)
(72)
(75)
(78)
Volumetric
Volumetric
Gravimetric
Volumetric
a
Langmuir model (BET model). Data from ref. 4.
3198 | J. Mater. Chem., 2007, 17, 3197–3204
This journal is ß The Royal Society of Chemistry 2007

View Online
After re-evacuating the sample vessel, low-temperature
isotherms were achieved by immersing the sample vessel in a
LN2 bath, then allowing the vessel and sample to equilibrate at
77 K for 2 h before commencing the experiments. The liquid
level of LN2 was carefully adjusted to a predefined mark on the
sample vessel above the sample position so as to mitigate
density fluctuations of gas in the column above the vessel,
which are reflected as pressure fluctuations by the pressure
transducers of the instrument. H2 adsorption isotherms were
constructed in replicate by dosing the initially evacuated
sample in a step-wise fashion, allowing sufficient time for
equilibration between each pressure increment in the profile
such that the variance in the time course profile for H2 uptake
was less than 0.001 wt%.
To obtain the excess adsorption isotherm, all data points
were corrected for buoyancy and the thermal gradient that
arises between the balance (313 K) and the sample bucket
(77 K). Buoyancy and thermal-gradient effects exhibited by the
bucket and the components associated with the magnetic
suspension balance were corrected on the basis of the change
in mass of the empty bucket within the analyte gas at 77 K.
The weight loss due to the buoyancy of the adsorbent was
determined by multiplying the volume of the MOF-177
20
framework skeleton by the density of H2 (i.e., corrected
mass for buoyancy is Vskeleton 6 rbulk). The volume of the
MOF-177 framework skeleton was determined from the
helium (,15 bar) buoyancy curve at 298 K using the same
gravimetric system.19 The absolute amount of gas adsorbed
was calculated from the density of the framework skeleton
(inverse of the skeleton volume) and the crystallographic
density (d = 0.427 g cm23), leading to an accessible pore
volume (Vpore = 1.69 cm3 g21) for MOF-177.18
To account for the temperature gradient that arises between
the calibrated dosing volume (304 K) and the sample vessel
(77 K), a linear correction factor (Cf) was derived by
calibrating the effect of this gradient on the gas density in
this segment of the instrument. Experimental determination of
Cf was achieved by replacing the sample with high purity
silicon granules and measuring the adsorption isotherm (77 K)
over the requisite pressure regime.
Estimation of absolute adsorbed amount
There are two components of gas adsorption: namely, the
surface excess and the absolute adsorption.21 Fig. 2 illustrates
these components based on the definition of the Gibbs dividing
surface.22 When an adsorbent is exposed to gas molecules,
some of the gas molecules are physisorbed on the surface.
However, since this process is an equilibrium between
adsorption and desorption, the bulk density (i.e., the amount
not adsorbed but present in the free volume) of the adsorbate
is usually not zero. Therefore, the observed mass change in the
sample is represented by the difference between the total
adsorbed amount and the bulk density of the adsorbate.
Experimental measurements are reported as a surface excess
amount.21 To estimate the absolute amount of gas adsorbed,
the thickness of the adsorbed layer must be known. However,
this variable, and its spatial profile, cannot be measured
experimentally. Instead, the absolute adsorption can only be
estimated theoretically using, for example, Monte Carlo
simulations.23
Hydrogen gas concentrations were corrected for non-ideal
behavior by computing the pressure dependence of the
compressibility factor derived semi-empirically from PVT
tables.17 The adsorbed concentrations of H2 (excess mass)
were then computed using the previously measured free-
volume and thermal-gradient correction factor (Cf).
The accessible volume was calculated in the Void Space
2
routine in Cerius using the default settings (1.3 A probe radius
˚
which is half of the kinetic diameter of helium), which was used
for the estimation of absolute adsorbed amount (i.e., Vpore
=
1.56 cm3 g21).18 In the unit conversion (mg g21 A g L21), the
bulk density of MOF-177 (d = 0.427 g cm23) was applied to
the amount of H2 uptake (i.e., 10 mg g21 = 4.27 g L21).
Gravimetric high-pressure gas adsorption measurements at
UCLA
Gravimetric H2 sorption isotherms were measured by use of a
GHP-300 (Gravimetric High Pressure analyzer) from the VTI
Corporation.16 A Rubotherm magnetic suspension balance
MC-5 was used to measure the change in mass of samples
suspended within a tube (22 mm i.d.) constructed from Inconel
625 under a chosen atmosphere. Prior to admittance of the
analyte gas, the entire chamber and manifold were evacuated
at room temperature, and the weight of the Al sample bucket
(12 mm i.d. 6 21 mm length) was measured. After loading the
MOF-177 sample (ca. 200 mg) the system was purged at room
temperature with hydrogen and helium, and the sample was
outgassed, using a turbomolecular drag pump (Pfeiffer, TSH
071 E), until a constant mass was attained. When H2 gas was
used, water and other condensable impurities were removed
with a LN2 trap. Pressure was measured with a MKS Baratron
transducer 120AA (0 to 1000 Torr) and an electronic Bourdon
gauge-type transducer (Mensor, up to 1500 psi). The adsorbate
was added incrementally, and data points were recorded when
no further change in mass was observed. The temperature in
the Inconel tube was also monitored with a platinum resistance
thermometer.
Although the surface excess mass is a useful concept, from
the viewpoint of hydrogen storage, the total amount that a
material can store is more relevant to use for hydrogen as a
fuel. However, at high temperatures and pressures (i.e., above
the critical temperature and pressure of hydrogen), the density
profile of the adsorbed phase becomes more diffused and,
therefore, it is not possible to distinguish between the adsorbed
and bulk phases with the present techniques. In this situation
the surface excess is the only experimentally accessible
quantity, and there is not a reliable method to estimate the
absolute adsorbed amount with high accuracy, although many
efforts have been devoted to resolve this issue.24 Therefore, we
estimate the absolute amount of hydrogen adsorbed using a
simple equation:25,26
Nabs = Nex + rbulkVpore
(1)
where Nabs is the absolute adsorbed amount, Nex is the surface
excess amount, rbulk is the bulk density of H2,19 and Vpore is
the pore volume for MOF-177.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3197–3204 | 3199

View Online
Fig. 2 (A–C) Schematic representation of the relation between the surface excess mass and absolute adsorbed amount. Dark gray regions in the
top diagrams represent the surface excess mass. Since absolute adsorbed amounts should include all gas molecules (dark gray and white circles in
the bottom boxes) in the adsorbed layer, summation of the dark gray and light gray areas in the diagrams corresponds to the absolute adsorbed
amount. The density of excess mass is not necessarily higher than the bulk fluid density at all locations, and negative excess mass can be obtained
under large bulk density. (D) The surface excess amount (solid curve) possesses a maximum value because the excess mass can be decreased in the
high-pressure region, while total gas uptake (dashed curve) should increase monotonically with an increase in the pressure because of the
contribution from the large bulk density (dotted curve, rbulk) of the adsorbate.
In general, when gas molecules are adsorbed on the pore
Results and discussion
surface, heat is released, indicating that the adsorbed amount
Low-pressure gas adsorption
should be influenced by the measurement temperature.
MOF-177 samples showed significant N2 uptake at low pres-
sures, whose profile can be classified as a Type I isotherm
(Fig. 3). The N2 uptake (1200 cm3 g21 at P/P0 = 0.25)
corresponds to an apparent surface area of 5250 m2 g21
using the Langmuir model (4630 m2 g21 for the BET model27
using the adsorption branch of the N2 isotherm over 0.03 ,
P/P0 , 0.07).
Low-pressure H2 isotherms of MOF-177 were measured at
77 and 87 K, respectively, and are shown in Fig. 4. Under these
conditions, H2 adsorption isotherms are not saturated, because
these temperatures are above the critical temperature (33 K for
hydrogen). The rapid pressure equilibration (ca. 5 min per step)
and absence of hysteresis indicate that H2 molecules are
reversibly physisorbed by MOF-177. The maximum uptake at
800 Torr for 77 and 87 K was 142 and 82 cm3 g21, respectively.
Fig. 3 N2 isotherm for MOF-177 measured at 77 K. Filled and open
The isotherm measured at 77 K shows good agreement with
previously reported gravimetric data (12.45 mg g21 at 751 Torr,
see secondary axis for adsorbed amount of H2 in Fig. 4) in Ref. 2.
circles represent adsorption and desorption branches respectively.
(Inset) Logarithmic relative pressure expression of N2 isotherm for
MOF-177.
3200 | J. Mater. Chem., 2007, 17, 3197–3204
This journal is ß The Royal Society of Chemistry 2007

View Online
required to adequately describe the isotherms. From these
results, the isosteric heat of adsorption is calculated from
m
X
Qst~{R
aiNi
(3)
i~0
where R is the universal gas constant.
The initial value of the isosteric heat of adsorption for
MOF-177 is estimated to be 4.4 kJ mol21, which decreases
slightly with an increase in the surface coverage of hydrogen
(Fig. 5). The profile is quite similar to that of IRMOF-1,3 aside
from the slightly lower adsorption enthalpy compared to
IRMOF-1. It should be noted that the nearly constant Qst
value is attributed to the linear shape of the H2 isotherms in
the low-pressure region. Although the results indicate that
MOF-177 does not have a strong hydrogen binding site, as
indicated by the low Qst value, the constant Qst value for the
isotherms in accordance with Henry’s law is a desirable
property for H2 storage materials.32 In such a scenario, H2
storage capacity is not influenced by surface coverage, leading
to a large amount of available H2 as the low-pressure limit of
the delivery system is approached.
Fig. 4 H2 isotherms measured at 77 K (circles) and 87 K (squares) for
MOF-177. Filled and open symbols represent adsorption and
desorption branches respectively. These symbols are traced by fitting
curves from eqn (2) with the presented parameters.
Therefore, the temperature dependence of H2 uptake provides
useful information regarding H2–adsorbent interactions. In
particular, estimation of the coverage-dependent isosteric heat
of adsorption (Qst), which may allow estimation of the number
of strong binding sites per unit cell, is a useful diagnosis. For
the calculation of the isosteric heat of adsorption, an empirical
equation (for instance, the Langmuir–Freundlich model) is
applied,28 so, in most cases, the differential enthalpy derived at
low coverage is suspect because the model fits were poor (see
Fig. 5), even though the low-coverage region is mechanistically
more interesting than at high coverage. To estimate reliable
isosteric heats of adsorption, a virial-type expansion compris-
ing the temperature-independent parameters ai and bi should
be used and was applied:3,5,29
High-pressure H2 adsorption
High-pressure hydrogen uptake properties for MOF-177 were
evaluated by use of volumetric and gravimetric techniques.
Isotherms acquired volumetrically are illustrated in Fig. 6A.
These isotherms saturate at near 60 bar with maximum surface
excess amounts of 75 mg g21. The profiles are similar to
the data which we have reported previously (Fig. 6C).4
Gravimetric measurement provides a similar profile as the
volumetric isotherm measured at 77 K (Fig. 6B). The reversible
isotherm indicates that H2 is physisorbed even in the high-
pressure region. After buoyancy correction the uptake (surface
excess amount) is 73 mg g21 at a saturation pressure of
approximately 50 bar. We presume that the small difference
between volumetric and gravimetric uptake is attributed to the
difference in the apparent surface area of MOF-177 samples
(see inset of Fig. 6). It should be noted that the saturation
pressure for gravimetric measurements is slightly lower than
that determined by volumetric measurements. This may be
attributed to small inaccuracies in the He buoyancy correction,
because the total volume of adsorbents (i.e., MOF-177 sample
and adsorbed H2 molecules) can be slightly increased when gas
molecules are adsorbed on the surface of adsorbents, while we
assume that no He molecule is adsorbed on the MOF-177
surface. Consequently buoyancy attributed to the volume
of adsorbed H2 cannot be corrected, leading to an under-
estimation of H2 uptake in the high-pressure region. In
principle, it is possible to compensate for these differences,
however, we will address this issue in future work.33
m
n
X
X
1
ln P~ ln Nz
aiNiz
biNi
(2)
T
i~0
i~0
where P is pressure, N is the adsorbed amount,30 T is
temperature, and m and n represent the number of coefficients
Absolute adsorbed amounts at 70 bar are estimated to be
113 mg g21 from the measurements, which corresponds to
48.3 g L21. Since the absolute adsorbed amount cannot be
measured experimentally, it is not possible to confirm whether
or not this value is reasonable. Therefore, we compared the
absolute H2 uptake with N2 adsorption data which were taken
at 77 K (Table 2).34 In contrast to the difference in gravimetric
uptake, the volumes of adsorbed gases between N2 and H2
Fig. 5 Coverage dependence of the isosteric heat of adsorption for H2
in MOF-177 calculated from fits of its 77 and 87 K isotherms (solid
curve). For comparison, the isosteric heat of adsorption was also
estimated based on the Langmuir–Freundlich model (broken curve).
The obtained Qst value is apparently overestimated at low coverage
region due to poor fitting. Although the virial-type expansion fits
better than the Langmuir–Freundlich model, it is reasonable to apply
three or more temperatures for the Qst calculation with high accuracy.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3197–3204 | 3201

View Online
Fig. 6 High-pressure H2 isotherms for MOF-177 taken by volumetric (A) and gravimetric (B) methods (circles and triangles). As a reference,
volumetric data in ref. 4 are shown in (C) and all plots are overlaid in (D). Filled and open symbols represent adsorption and desorption branches,
respectively. Square and diamond symbols which are absolute adsorbed amounts are obtained from eqn (1). Connecting traces are guides for eyes.
Inset: surface area dependence of H2 uptake.
(i.e., cm3 g21) are quite similar. Consequently, the number of
adsorbed N2 and H2 molecules in the formula unit is almost
the same. This clearly demonstrates that the micropores in
MOF-177 are in large part occupied by H2 molecules in the
high-pressure region. This observation can be supported by
the ratio between the absolute adsorbed amount for H2 and
the liquid density of H2 (ca. 68%, see Table 2). Thus it is
presumed that H2 adsorption under these conditions can be
explained by a micropore filling model, in contrast to a
monolayer coverage model which might otherwise be widely
assumed on the basis of weak adsorbent–dihydrogen interac-
tions. Indeed, the idea that the H2 storage capacity strongly
depends on the pore volume of the adsorbent is consistent with
Monte Carlo simulations for various MOFs.23 In light of the
fact that the molecular diameter of H2 is small and that the
isotherm for the absolute adsorbed amount is not saturated,
the DOE 2010 volumetric target of 45 g L21 could probably be
met if a high-pressure H2 container is engineered for operation
at cryogenic temperatures. However, as predicted,32 higher
binding energies of adsorption (Qst) than exhibited by
unmodified MOFs are required for room temperature H2
storage in addition to the high surface-areas afforded by these
materials.
Under more practical conditions, such as actual gas
cylinders, it is important to realize that the discharge pressure
of a cylinder needs to be at least 1.5 bar.32 In other words, a
large isosteric heat of adsorption may reduce the amount of H2
available for use. In order to consider this issue, absolute H2
amounts for MOF-177 were plotted. As shown in Fig. 7, the
amount of H2 delivered for MOF-177 is 40 g L21 when the
pressure ranges from 1.5 to 70 bar. Therefore, we propose that
an alternative metric for achieving the DOE target should be
defined on the basis of the total amount of delivered H2 (e.g.,
an absolute adsorbed amount ranging from 1.5 to 100 bar),32
instead of a simple surface excess mass of H2.
Conclusions
The utility of MOF-177 as a benchmark material for high-
pressure and low temperature H2 adsorption was demon-
strated by verification of its uptake capacity [surface excess
Table 2 H2 and N2 adsorption data for MOF-177
Adsorbed amounta
No. of guests
per f.u.b
Ratioc
(%)
Guest
cm3 g21
mg g21
g L21
H2
N2
a
1270
1200
113
1500
48.3
641
65
61
68
79
The absolute adsorbed amount at 77 K and 70 bar for H2; low-
b
pressure data at 77 K for N2. Formula unit (f.u.) is Zn4O(BTB)2.
c
Comparison with corresponding liquid densities (71 g L21 for H2,
Fig. 7 Absolute adsorbed amount (triangles) and surface excess
806 g L21 for N2).
amount (circles) for MOF-177 expressed on a log scale.
3202 | J. Mater. Chem., 2007, 17, 3197–3204
This journal is ß The Royal Society of Chemistry 2007

View Online
weight uptake of 75 mg g21and total (absolute) H2 uptake of
110 mg g21 (47 g L21) at 70 bar, 77 K]. The theoretical H2
uptake capacity depends on the pore volume of the adsorbent
rather than the enthalpy of adsorption, especially at high
pressures and low temperature. However large binding energy
(15–20 kJ mol21) is required for viable room temperature
storage.32 Adsorbents possessing large binding energy and
high density of binding sites need to be developed to confine
large amounts of H2 molecules for use as a fuel in automotive
applications.
12 H. K. Chae, D. Y. Siberio-Pe´rez, J. Kim, Y.-B. Go, M. Eddaoudi,
A. J. Matzger, M. O’Keeffe and O. M. Yaghi, Nature, 2004, 427,
523.
13 For instance, (1) non-ideality of H2 gas is emphasized under low
temperature and high-pressure conditions, and (2) temperature
gradient effect between a sample cell and a pressure transducer
could bring a huge experimental error since the deviation increases
with an increase in the inverse of measurement temperature. In
other words, even if the experimental value at room temperature
looks reasonable, there is no guarantee that the sorption system
will also work properly at cryogenic conditions.
14 R. Chahine and T. K. Bose, Int. J. Hydrogen Energy, 1994, 19, 161;
H. W. Langmi, A. Walton, M. M. Al-Mamouri, S. R. Johnson,
D. Book, J. D. Speight, P. P. Edwards, I. Gameson, P. A. Anderson
and I. R. Harris, J. Alloys Compd., 2003, 356–357, 710.
15 X. Li, F. Cheng, S. Zhang and J. Chen, J. Power Sources, 2006,
160, 542.
Acknowledgements
We are grateful to Dr Matthias Thommes (Quantachrome)
and Dr Augie F. Venero (VTI Corporation) for valuable
advice and discussions, and to Ms Jun Yang and Mr Qiaowei
Li for samples of MOF-177. We also thank Drs George
Thomas, Jesse Adams, Sunita Satyapal and Carole Read
(DOE) for their valuable input. This work was supported by
the U.S. Department of Energy (DE-FG36-05GO15001 to
OMY and DE-FC36-02AL67619 to MAM).
16 S. Lowell, J. E. Shields, M. A. Thomas and M. Thommes,
Characterization of Porous Solids and Powders: Surface Area, Pore
Size and Density, Kluwer Academic Publishers, Dordrecht, 2004;
J. Keller and R. Staudt, Gas Adsorption Equilibria: Experimental
Methods and Adsorption Isotherms, Springer Science + Business
Media, New York, 2005.
17 R. D. McCarty, J. Hord and H. M. Roder, Selected Properties of
Hydrogen (Engineering Design Data), National Bureau of
Standards, 1981, Monograph 168.
18 Ideal MOF-177 samples do not have any external surface but only
an internal pore system, while actual crystalline samples (1) have
finite size (i.e., sub-millimeter scale), (2) may be partly decom-
posed, (3) contain small amount of guest (mainly water) molecules.
Therefore, it is difficult to conclude which value reflects an actual
system. Considering that estimation of the true density of porous
materials is not easy,19 further consideration is beyond the scope of
this article. However, we cannot fully exclude the possibility
of helium adsorption on the MOF-177 surface. In case the slope
of helium isotherm for the buoyancy correction can be under-
estimated, the Vp at UCLA may be overestimated.
References
1 N. L. Rosi, J. Eckert, M. Eddaoudi, D. T. Vodak, J. Kim,
M. O’Keeffe and O. M. Yaghi, Science, 2003, 300, 1127;
J. L. C. Rowsell, J. Eckert and O. M. Yaghi, J. Am. Chem. Soc.,
2005, 127, 14904.
2 J. L. C. Rowsell, A. R. Millward, K. S. Park and O. M. Yaghi,
J. Am. Chem. Soc., 2004, 126, 5666.
3 J. L. C. Rowsell and O. M. Yaghi, J. Am. Chem. Soc., 2006, 128,
1304.
4 A. G. Wong-Foy, A. J. Matzger and O. M. Yaghi, J. Am. Chem.
Soc., 2006, 128, 3494.
19 K. Murata, K. Kaneko, F. Kokai, K. Takahashi, M. Yudasaka
and S. Iijima, Chem. Phys. Lett., 2000, 331, 14.
20 NIST chemistry webbook (thermophysical properties of fluid
systems); http://webbook.nist.gov/chemistry/fluid/.
21 S. Sircar, Ind. Eng. Chem. Res., 1999, 38, 3670; A. L. Myers,
J. A. Calles and G. Calleja, Adsorption, 1997, 3, 107.
5 M. Dinca˘, A. Dailly, Y. Liu, C. M. Brown, D. A. Neumann and
J. R. Long, J. Am. Chem. Soc., 2006, 128, 16876.
6 D. N. Dybtsev, H. Chun, S. H. Yoon, D. Kim and K. Kim, J. Am.
Chem. Soc., 2004, 126, 32; S. S. Kaye and J. R. Long, J. Am. Chem.
Soc., 2005, 127, 6506; X. Lin, J. Jia, X. Zhao, K. M. Thomas,
A. J. Blake, G. S. Walker, N. R. Champness, P. Hubberstey and
M. Schro¨der, Angew. Chem., Int. Ed., 2006, 45, 7358.
7 L. Pan, M. B. Sander, X. Huang, J. Li, M. Smith, E. Bittner,
B. Bockrath and J. K. Johnson, J. Am. Chem. Soc., 2004, 126,
1308; B. Kesanli, Y. Cui, M. R. Smith, E. W. Bittner,
B. C. Bockrath and W. Lin, Angew. Chem., Int. Ed., 2005, 44,
72; B. Chen, N. W. Ockwig, A. R. Millward, D. S. Contreras and
O. M. Yaghi, Angew. Chem., Int. Ed., 2005, 44, 4745; H. Kabbour,
T. F. Baumann, J. H. Satcher, Jr., A. Saulnier and C. C. Ahn,
Chem. Mater., 2006, 18, 6085; P. M. Forster, J. Eckert,
B. D. Heiken, J. B. Parise, J. W. Yoon, S. H. Jhung, J.-S. Chang
and A. K. Cheetham, J. Am. Chem. Soc., 2006, 128, 16846; D. Sun,
S. Ma, Y. Ke, D. J. Collins and H.-C. Zhou, J. Am. Chem. Soc.,
2006, 128, 3896; S. Ma, D. Sun, M. Ambrogio, J. A. Fillinger,
S. Parkin and H.-C. Zhou, J. Am. Chem. Soc., 2007, 129, 1858.
8 M. Latroche, S. Surble´, C. Serre, C. Mellot-Draznieks,
P. L. Llewellyn, J.-H. Lee, J.-S. Chang, S. H. Jhung and
G. Fe´rey, Angew. Chem., Int. Ed., 2006, 45, 8227; A. Dailly,
J. J. Vajo and C. C. Ahn, J. Phys. Chem. B, 2006, 110, 1099; J. Jia,
X. Lin, C. Wilson, A. J. Blake, N. R. Champness, P. Hubberstey,
G. Walker, E. J. Cussena and M. Schro¨der, Chem. Commun., 2007,
840.
22 F. Rouquerol, J. Rouquerol and K. Sing, Adsorption by Powders
and Porous Solids, Academic Press, London, 1999.
23 H. Frost, T. Du¨ren and R. Snurr, J. Phys. Chem. B, 2006, 110, 9565.
24 K. Murata, M. El-Merraoui and K. Kaneko, J. Chem. Phys., 2001,
114, 4196.
25 A. V. Neimark and P. I. Ravikovitch, Langmuir, 1997, 13, 5148.
26 Actual pore volume after H2 adsorption may be smaller than the
accessible pore volume, Vpore. However, in the Gibbs definition of
dividing surface, the adsorbed phase is a surface therefore it does
not have a volume. Considering this, in terms of H2 storage, the
important factor is not the thickness of the adsorbed layer but the
number of H2 molecules in the pores (which includes the bulk
region), we believe that rough estimation of absolute adsorbed
amounts is an important issue. However, this is beyond the scope
of this paper.
27 The BET method is still widely used for determining the surface
area of micro–mesoporous materials. It is also recognized that the
BET area should not be accepted as the true internal area of
microporous solids having pores of molecular dimensions, such as
in MOF materials. In this case, the value obtained should be
regarded as an apparent or equivalent surface area.
28 A. Malek and S. Farooq, AIChE J., 1996, 42, 3191.
29 L. Czepirski and J. Jagiełło, Chem. Eng. Sci., 1989, 44, 797;
J. Jagiello, T. J. Bandosz, K. Putyera and J. A. Schwarz, J. Chem.
Eng. Data, 1995, 40, 1288.
30 Strictly speaking, the heat of adsorption needs to be estimated
based on the absolute adsorbed amount.31 However, since the bulk
density of H2 under experimental conditions is small, we assumed
9 H. Li, M. Eddaoudi, M. O’Keeffe and O. M. Yaghi, Nature, 1999,
402, 276; O. M. Yaghi, M. O’Keeffe, N. W. Ockwig, H. K. Chae,
M. Eddaoudi and J. Kim, Nature, 2003, 423, 705.
10 J. L. C. Rowsell and O. M. Yaghi, Angew. Chem., Int. Ed., 2005,
44, 4670.
11 Energy Efficiency and Renewable Energy Hydrogen, Fuel Cells
and Infrastructure Technologies Program http://www1.eere.energy.
gov/hydrogenandfuelcells/.
Nex # Nabs
.
31 M. M. K. Salem, P. Braeuer, M. v. Szombathely, M. Heuchel,
P. Harting, K. Quitzsch and M. Jaroniec, Langmuir, 1998, 14,
3376.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 3197–3204 | 3203

View Online
32 S. K. Bhatia and A. L. Myers, Langmuir, 2006, 22, 1688.
33 This does not mean that volumetric measurements are better
than gravimetric ones. Even in volumetric measurements
either estimation of dead volume or unidealistic behavior of
gas molecules under supercritical conditions can contain
uncertainties.
thus the surface excess Nex corresponds to the adsorbed amount
Nads, i.e. Nex # Nabs
13 This is the typical situation encountered for
.
N2 and Ar adsorption at their boiling temperatures (77 K and 87 K,
respectively). In contrast, adsorption in significant amounts occurs
above the critical temperature only at higher pressures. The bulk
gas density here is so high that it cannot be neglected any more,
and as indicated in eqn (1) adsorption data are therefore usually
given in terms of the surface excess.
34 At sufficiently low temperatures and pressures, the bulk gas density
(rbulk) is negligibly small against the density near the surface and
3204 | J. Mater. Chem., 2007, 17, 3197–3204
This journal is ß The Royal Society of Chemistry 2007