Kubas-type hydrogen storage in V(III) polymers using tri- and tetradentate bridging ligands

Article abstract of DOI:10.1021/ja110243r

Oxalic acid, oxamide, glycolic acid, and glycolamide were employed as 2-carbon linkers to synthesize a series of one-dimensional V(III) polymers from trismesityl vanadium(III)·THF containing a high concentration of low-valent metal sites that can be exploited for Kubas binding in hydrogen storage. Synthesized materials were characterized by powder X-ray diffraction (PXRD), nitrogen adsorption (BET), X-ray photoelectron spectroscopy (XPS), infrared spectroscopy (IR), Raman spectroscopy, thermogravimetric analysis, and elemental analysis. Because each of these organic linkers possesses a different number of protons and coordinating atoms, the products in each case were expected to have different stoichiometries with respect to the number of mesityl groups eliminated and also a different geometry about the V(III) centers. For example, the oxalate and glycolate polymers contained residual mesityl groups; however, these could be exchanged with hydride via hydrogenolysis. The highest adsorption capacity was recorded on the product of trismesityl vanadium(III)·THF with oxamide (3.49 wt % at 77 K and 85 bar). As suggested by the high enthalpy of adsorption (17.9 kJ/mol H₂), a substantial degree of performance of the vanadium metal centers was retained at room temperature (25%), corresponding to a gravimetric adsorption of 0.87 wt % at 85 bar, close to the performance of MOF-177 at this temperature and pressure. This is remarkable given the BET surface area of this material is only 9 m ²/g. A calculation on the basis of thermogravimetric results provides 0.88 hydrogen molecule per vanadium center under these conditions. Raman studies with H₂ and D₂ showed the first unequivocal evidence for Kubas binding on a framework metal in an extended solid, and IR studies demonstrated H(D) exchange of the vanadium hydride with coordinated D₂. These spectroscopic observations are sufficient to assign the rising trends in isosteric heats of hydrogen adsorption observed previously by our group in several classes of materials containing low-valent transition metals to the Kubas interaction.

Full text of DOI:10.1021/ja110243r

ARTICLE
pubs.acs.org/JACS
Kubas-Type Hydrogen Storage in V(III) Polymers Using Tri- and
Tetradentate Bridging Ligands
Tuan K. A. Hoang,† Ahmad Hamaed,† Golam Moula,† Ricardo Aroca,† Michel Trudeau,‡ and
David M. Antonelli*,†,§
†Department of Chemistry and Biochemistry, University of Windsor, 401 Sunset Avenue, Windsor, Ontario N9B 3P4, Canada
‡Emerging Technologies, Hydro-Quꢀebec Institute, 1800 Boul. Lionel-Boulet, Varennes, Quebec J3X 1S1, Canada
§Sustainable Environment Research Center, University of Glamorgan, Pontypridd CF37-1DL, United Kingdom
S Supporting Information
b
ABSTRACT: Oxalic acid, oxamide, glycolic acid, and glycola-
mide were employed as 2-carbon linkers to synthesize a series of
one-dimensional V(III) polymers from trismesityl vanadium-
(III) THF containing a high concentration of low-valent metal
3
sites that can be exploited for Kubas binding in hydrogen
storage. Synthesized materials were characterized by powder
X-ray diffraction (PXRD), nitrogen adsorption (BET), X-ray
photoelectron spectroscopy (XPS), infrared spectroscopy (IR),
Raman spectroscopy, thermogravimetric analysis, and elemen-
tal analysis. Because each of these organic linkers possesses a
different number of protons and coordinating atoms, the products in each case were expected to have different stoichiometries with
respect to the number of mesityl groups eliminated and also a different geometry about the V(III) centers. For example, the oxalate
and glycolate polymers contained residual mesityl groups; however, these could be exchanged with hydride via hydrogenolysis. The
highest adsorption capacity was recorded on the product of trismesityl vanadium(III) THF with oxamide (3.49 wt % at 77 K and
3
85 bar). As suggested by the high enthalpy of adsorption (17.9 kJ/mol H2), a substantial degree of performance of the vanadium
metal centers was retained at room temperature (25%), corresponding to a gravimetric adsorption of 0.87 wt % at 85 bar, close to the
performance of MOF-177 at this temperature and pressure. This is remarkable given the BET surface area of this material is only
9 m2/g. A calculation on the basis of thermogravimetric results provides 0.88 hydrogen molecule per vanadium center under these
conditions. Raman studies with H2 and D2 showed the first unequivocal evidence for Kubas binding on a framework metal in an
extended solid, and IR studies demonstrated H(D) exchange of the vanadium hydride with coordinated D2. These spectroscopic
observations are sufficient to assign the rising trends in isosteric heats of hydrogen adsorption observed previously by our group in
several classes of materials containing low-valent transition metals to the Kubas interaction.
’ INTRODUCTION
yet to meet the stringent DOE goals. Physisorption materials,
such as carbon8 and metalÀorganic frameworks (MOFs),9 possess
much lower binding enthalpies (5À13 kJ/mol) and demonstrate
excellent hydrogen uptake with no kinetic barrier and little heat
management issues at 77 K and moderate pressure. For example,
the total hydrogen uptake of MOF-177 and NOTT-122 sur-
passes the DOE 2015 system target under these conditions;10
however the cost of cooling still makes this technology prohibi-
tive. To improve performance at higher temperatures, some
MOF materials have been synthesized in hopes to exploit the role
of open metal sites11 or the Kubas interaction,12 a type of
hydrogen binding to transition metals with enthalpies in the
20À50 kJ/mol range.13À15 While there has yet to be conclusive
spectroscopic evidence of the Kubas interaction in MOFs, it has
been observed by vibrational spectroscopy in Cu-ZSM-5.16
Hydrogen is one of the best candidates for an alternative
energy carrier and in its pure form can be stored in compressed
gas cylinders at pressures as high as 700 bar. Since these pressures
are considered unsafe and the ranges of vehicles using such tanks
are still not optimal, new methods of storing hydrogen are
currently being investigated.1 Extensive research has been carried
out using a variety of systems as chemical carriers based around
physisorption2 and metal hydrides;3 however, these approaches
have so far fallen short of the U.S. Department of Energy (DOE)
2015 system target4 of 5.5 wt % and 40 kg/m2 H2 at 298 K. Some
chemical hydrides, which generally possess high adsorption
enthalpies in the 70 kJ/mol range, contain enough hydrogen to
satisfy the DOE 2015 target;5 however, the kinetics of charging
and discharging hydrogen and the heat management issues
prevent them from practical usage.6 Catalyzed hydrogen spillover
on high surface area substrates,7 is another promising method
employing dissociative hydrogen adsorption; however, it too has
Received: November 23, 2010
Published: March 10, 2011
r
2011 American Chemical Society
4955
dx.doi.org/10.1021/ja110243r J. Am. Chem. Soc. 2011, 133, 4955–4964
|

Journal of the American Chemical Society
ARTICLE
Extensive studies of metal binding sites have also been carried out
by our group on Ti, V, Cr supported silica systems using different
oxidation states, ligand types, preparation methods, and doping
levels.17,18 In these studies, Ti(III), V(III), Cr(II), and Cr(III)
were identified as the best metal hosts for what was believed to be
Kubas interaction with hydrogen at room temperature and
moderate pressure on the basis of rising enthalpies with surface
covereage.19 Subsequent computational studies have supported
this assignment;20 however, no spectroscopic evidence has yet
been obtained due to the small amounts of hydrogen absorbed by
the ca. 5% metal in these systems. To maximize the number of
potential Kubas sites per unit volume, cyclopentadienyl chro-
mium hydrazide gels21 and vanadium hydrazide gels22 were
synthesized by the reaction of low valent organometallic pre-
cursors with anhydrous hydrazine. Hydrazine was used as a linker
because of its low molecular weight and ability to bridge two
metal centers, providing a diffusion pathway for hydrogen with-
out creating excess void space, while also maintaining the low
ligation number of the metal coordination sphere. These X-ray
amorphous materials possess linear isotherms at both 77 and 298
K that do not saturate at 80 bar, ideal adsorption enthalpies in the
20À40 kJ/mol range, and maintain up to 49% of their perfor-
mance at 298 K as compared to 77 K. The volumetric adsorption
of the vanadium material at 77 K and 80 bar (60 kg/m3)
surpasses all MOFs as well as the 2010 DOE target. At room
temperature, this material demonstrates a volumetric adsorption
of 23.2 kg/m3 H2, approaching the 2010 DOE target. While
transition metal hydrazides show enormous promise, hydrazine
has many safety issues and can also bind in multiple coordination
modes, leading to amorphous systems that are difficult to
characterize. Because of this, we are continuing to explore other
low-molecular weight ligands that support low-coordinate metal
sites useful for Kubas binding. In this work, trismesitylvanadium-
collected by filtration and washed several times with a solution of THF
and ether (THF:ether = 1:3 by volume) before drying in a vacuum.24
Preparation of V-Oxamide100 and V-Oxamide150. V-Ox-
amide100 was synthesized as follows: V(Mes)3 THF (3 g, 6.24 mmol)
3
was dissolved in 75 mL of dry toluene at room temperature in an
Erlenmeyer flask. 0.5608 g (6.24 mmol) of oxamide was then added with
vigorous stirring. The solution was capped and stirring was continued for
12 h. The solution was then heated to 100 °C for 3 h with stirring. After
this, the system was filtered, and a black solid was obtained. The filtrate
was colorless, indicating complete reaction of the V precursor. This solid
was transferred to an air-free tube and was then dried under a vacuum
(10À2 mmHg) for 12 h, followed by heating at 60 °C under a vacuum
(10À2 mmHg) for a period of 6 h and another 6 h at 100 °C under a
vacuum (10À2 mmHg). V-Oxamide150 was obtained by continuing to
heat V-Oxamide100 at 150 °C for 6 h in a vacuum (10À2 mmHg).
Preparation of V-Oxalate150, V-Glycolate150, and V-Gly-
colamide150. V-Oxalate150 was synthesized by the same method as
V-Oxamide150, using 0.5618 g (6.24 mmol) of oxalic acid instead of
oxamide. V-Glycolate150 and V-Glycolamide150 were also synthesized
in this same way, using glycolic acid (0.4973 g; 6.24 mmol) and glyco-
lamide (0.4780 g; 6.24 mmol), respectively. The filtrate was colorless in
each case, indicating complete reaction of the V precursor.
Preparation of H2-V-Oxalate and H2-V-Glycolate. V-Oxa-
late150 and V-Glycolate150 were treated with hydrogen at 150 °C and
85 bar for 3 h, and the resulting materials were then collected and stored
under Ar.
Analysis. Powder X-ray diffraction (PXRD) was performed on a
Siemens diffractometer D-500 with Cu KR radiation (40 KV, 40 mA)
source. The step size was 0.02°, and the counting time was 0.3 s for each
step. Diffraction patterns were recorded in the 2θ range of 2.3À52°.
Samples for PXRD analysis were put in a sealed capillary glass tube to
protect the sample from air and moisture during experiment.
Nitrogen adsorption and desorption data were collected on a Micro-
meritics ASAP 2010. All XPS studies were conducted using a Physical
Electronics PHI-5500 spectrometer using charge neutralization, and
emissions were referenced to the carbon CÀ(C, H) peak at 284.8 eV.
Elemental analysis was performed by Galbraith Laboratories, Knoxville,
TN. Thermogravimetric analysis was conducted on a Mettler Toledo
TGA SDTA 851e, using helium (99.99%) as purging gas with the rate of
30 mL/min. Samples were held at 25 °C for 20 min before being heated
to 550 °C at a rate of 5 °C/min.
FT-IR was conducted on all synthesized materials using a Bruker
Vector 22 instrument. In a typical experiment, 2 mg of sample was mixed
with 500 mg of dry KBr, and a compressed disk was made in an inert
atmosphere. The discs were then transferred outside, and the measure-
ments were conducted immediately. Because this technique was pri-
marily used to identify the hydrocarbon in the sample, which would not
oxidize on exposure to air, rigorously air-free conditions were not
employed.
HydrogenÀDeuterium Exchange Experiment. 100 mg of
V-Oxamide150 was placed in a Schlenk tube connected to a bubbler and
a deuterium cylinder. A deuterium stream was passed over the sample for
7 days at a rate of 1 L per hour. After this, the deuterium-treated sample
was left under a vacuum for 6 h. The original and deuterium treated
materials were then studied by infrared spectroscopy using a Bruker
Tensor 27 instrument. Three milligrams of material was mixed with
500 mg of KBr, and a compressed disk was made in inert atmosphere.
The discs were transferred outside, and the measurements were con-
ducted immediately.
Raman Experimental Details. Samples were loaded in NMR
tubes fitted with Young’s valves, and spectra were recorded under 1 atm of
H2, D2, or Ar. For the samples recorded after vacuum, the samples were
left for 1 h under H2 or D2 and then placed under vacuum for 10 min
before backfilling with Ar. All micro-Raman scattering experiments were
(III) THF was treated with various commercially available
3
multiproton, multidentate chelating ligands capable of bridging
two metal centers in a controlled stoichiometric fashion. The
one-to-one metal-to-ligand stoichiometry used in this study
should also favor a 1-D system with much more flexibility around
the tetrahedral metal center than more rigid 3-D systems such as
zeolites,23 which do not possess a high number of active and
strong adsorption centers per unit volume, to optimize the
orbital geometries for H2 coordination.
’ EXPERIMENTAL SECTION
All chemicals, unless otherwise stated, were obtained from Aldrich.
H2 grade 6.0, N2, Ar, and He were obtained from Praxair Canada. All
solvents were dried by distillation in argon atmosphere, using sodium/
benzophenone to remove moisture and oxygen. Distilled solvent was
transferred to a Schlenk flask using a canula. Dry argon gas was bubbled
through the flask for 30 min, followed by degassing before transferring
into a drybox.
Preparation of V(Mes)3 THF. 33.33 mL of dry tetrahydrofuran
3
(THF) was added to 50 mL of mesityl magnesiumbromide 1 M
(MesMgBr) solution in THF, followed by the portion-by-portion
addition of 6.22 g of VCl3 3THF 97%. The resulting solution was
3
stirred vigorously at room temperature for 2 h, and a clear blue solution
was obtained. 21.66 mL of dioxane was then added to the solution with
stirring. After 2 more hours, stirring was ceased, and the solution was left
to settle before filtration. The filtrate was collected and concentrated in a
vacuum until crystals formed. 16.67 mL of diethylether was then added,
and the remaining product precipitated out. The solid product was then
4956
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
Scheme 1. Proposed Reaction between V(Mes)3 THF and Various Ligands: (a) Oxamide, (b) Oxalic Acid, (c) Glycolic Acid, and
3
(d) Glycolamide
conducted using a Renishaw InVia system, with laser excitation at
514.5 nm, and powers of 10À20 μW at the sample. Each scan was
duplicated to ensure that any features observed were reproducible. All
measurements were made in a backscattering geometry, using a 20Â
microscope objective with a numerical aperture value of 0.40, providing
scattering areas of ∼20 μm2. Single-point spectra were recorded with
4 cmÀ1 resolution and 50 s accumulation times. Data acquisition and
analysis were carried out using the WIRE software for windows and
Galactic Industries GRAMS C software.
Hydrogen Adsorption Measurements. Hydrogen adsorption
isotherms were obtained by using an Advanced Materials PCI and the
same method described in our previous publications on metal
hydrazides.21,22 Adsorption enthalpies were calculated from adsorption
isotherms at 77 and 87 K, using a method described elsewhere.25 True
volumetric adsorption of all materials were calculated on the basis of
gravimetric adsorption and skeletal densities of materials, while abso-
lute volumetric adsorption was calculated on the basis of gravimetric
adsorption using bulk densities.25,26
reaction of V(Mes)3 THF with the various ligands is shown in
Scheme 1.
3
The four protons present in oxamide are enough to eliminate
three mesityl groups with one proton remaining, putatively
resulting in a tetrahedral V(III) center bound to 2 O and 2 N
atoms from oxamide, capable of binding 3 H2 according to the
18-electron rule.27 Because the expected weight of the V-oxamide
monomer is 136 g/mol, this corresponds to 4.41 wt % H2, close
to the 2010 DOE goal. Oxalic acid has only two protons per
ligand so one remaining mesityl is expected in V-oxalate, result-
ing in a five-coordinate environment with the 4 oxalate oxygens,
while in the glycolate material there are only two available
protons and three ligating atoms, so retention of one aryl group
to create a 4-coordinate center V(III) is expected. In the glyco-
lamide material, there are three protons available and three
ligating atoms per ligand; however, elimination of all three
mesityls would result in a three-coordinate V center, which is
not likely. Therefore, it is possible that there is a remaining aryl
functionality on the V-center with a proton still remaining on the
ligand, or conversely that all mesityls have been eliminated and
the extra coordination site is taken up by THF, which could be
clarified by XPS and IR spectroscopy.
’ RESULTS AND DISCUSSION
In this work, tris(mesityl) vanadium(III) THF was treated
3
with one molar equivalent of either oxamide, oxalic acid, glycolic
acid, and glycolamide to create polymers with VÀO and VÀN
linkages through protonolysis of the mesityl groups. The samples
were then dried at either 100 or 150 °C (to create, for example,
V-Oxamide100 or V-oxamide150, respectively), and their hydro-
gen storage properties were tested. Because the materials dried at
150 °C possessed superior hydrogen storage performance, the
majority of this Article focuses on these samples. The proposed
The PXRD pattern (Figure S1aÀf; Supporting Information)
of V-Oxamide150 showed a wide reflection at 2θ < 5°, suggesting
possible mesoscopic periodicity with no long-range order in the
material. Similar XRD patterns were obtained with vanadium
hydrazide gels, which were synthesized by employing the same
V(Mes)3 THF precursor, but with anhydrous hydrazine as
3
linking agent.22 The materials synthesized from oxalic acid,
4957
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
due to incomplete reaction stoichiometries or error in the
elemental analysis process due to incomplete digestion of the
V species in HF, so these values may not reflect the true V
content in the materials. Repeat analysis yielded similar results,
so this may reflect the limitations of this technique on these
materials.
Infrared spectroscopy (IR) was also used to characterize the
materials. All samples show strong stretches for metal-coordi-
nated carbonyl groups at ca. 1600 cmÀ1. The elimination of
hydrocarbon was supported by the decrease in intensity of the
CÀH stretches at 2850À2950 cmÀ1 in the spectrum of the
products as compared to V(Mes)3 THF (Figure S2; Supporting
3
Information). Because V-glycolate150 and V-glycolamide150
contain methylene protons in the R-carbon to the carbonyl, it
is not possible to distinguish the CÀH stretches arising from
these functionalities from those of residual mesitylene. The CÀC
aromatic stretches expected at 1500 cmÀ1 in all materials are
obscured by other bands and are therefore not useful in this
study. The IR spectrum of the V-Oxamide150 shows an unusual
feature at 2200 cmÀ1 that can be assigned to a VÀH.29 The presence
of this species must arise from migration of the additional
hydrogen atom from the organic ligand to the V center through
an R-hydrogen migration.30 Because only three of the four
oxamide protons are used in eliminating the aryl groups, the
remaining proton, not present in the other three ligands, is free to
migrate to the V(III) center.
Figure 1. Nitrogen adsorptionÀdesorption isotherms of V-Oxa-
mide100, V-Oxamide150. Samples were measured on an ASAP-2010
instrument at 77 K.
XPS studies of the synthetic materials were conducted, and the
results are shown in Figures S3ÀS6 (Supporting Information).
In general, these spectra support the formulations suggested by
the IR spectra, but expand on our knowledge of the oxidation
states and binding modes present in each material. Multiple
oxidation states of vanadium were detected in the vanadium 2p
1/2, 3/2 regions of all samples (Figure S3; Supporting In-
formation). For V-Oxamide150, a small amount of a V(I) species
is detected at 513.8 and 519.7 eV, assigned by comparison with
literature values.31 Emissions at 515.1 and 521.9 eV can be
assigned to V(III).32 V(IV) species are also present at 516.2
and 523.4 eV.33 V(V) is detected with emissions at 517.6 and
524.9 eV.34 Similar disproportionation was observed for vana-
dium hydrazides studied by our group and is common in the
chemistry of lower oxidation sates of this element. Four oxidation
states of vanadium were also detected in V-Oxalate150 and
V-Glycolate150, but the emissions are slightly shifted to higher
binding energies with ∼0.3 eV difference. The V-glycolamide150
material also has multiple oxidation states, but there appear to be
no V(I) species present. The reason for this is not understood but
may be related to the relative disproportionation pathways open
to the metals in the different ligand environments.
The O 1s region of the XPS spectra of all materials is shown in
Figure S4. The emission at 530.4 eV, which is very strong in the
case of V-Oxamide150 and moderately strong in the case of
V-Glycolamide150, can be attributed to coordinated THF on the
basis of our previous work on vanadium hydrazides, which show a
substantial amount of THF retention.22 This is consistent with
the EA of these materials, which exhibits a higher amount of
carbon than expected. The fact that the THF is not observed in
the glycolate or oxalate samples is consistent with the presence of
the one remaining sterically demanding mesityl group expected
by proton stoichiometry. The emission at 532 eV can be
attributed to V-bound oxygen in the amide carbonyl or carbox-
ylate group in all four species on the basis of previous XPS spectra
of metal coordinated carboxylate species,35,36 while the small
Figure 2. Nitrogen adsorptionÀdesorption isotherms of V-Oxalate150,
V-Glycolate150, H2-V-Oxalate150, H2-V-Glycolate150, and V-Glyco-
lamide150.
glycolic acid, or glycolamide do not possess any distinct XRD
feature and appear to be completely amorphous. Nitrogen
adsorption and desorption isotherms for these materials displayed
BrunauerÀEmmettÀTeller (BET) surface areas in the range of
9À220 m2/g (Figures 1 and 2). The isotherms were all type II,
representing nonporous materials with most of the surface area
arising from textural porosity.28 Similar isotherms and surface
areas were observed for V- and cyclopentadienyl Cr-hydrazide
gels studied previously by our group.21,22 The isotherms for the
V-Oxamide materials show anomalies in the desorption branch
passing under the adsorption branch, which can be attributed to
the very low surface areas of these materials (9 and 28 m2/g)
approaching the detection limit of the instrument on the sample
sizes used in the measurement.
The C, H, N, and V elemental analysis data of these materials
are shown in Table S1, and the expected molecular formulas and
compositions of the stoichiometric products are shown in Table
S2 (Supporting Information). The high carbon values for the
oxamide and glycolamide samples indicate that the reaction did
not go to completion, either due to retention of THF or because
of incomplete protolysis of the mesityl groups. In contrast, the
carbon values for the oxalic acid and glycolic acid samples were
lower than expected, perhaps due to incomplete combustion of
the material during the elemental analysis process leading to
carbide formation. The V values are systematically low, potentially
4958
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
emission at 533.8 eV corresponds to an oxygen with two single
bonds in either CÀOÀV or CÀOÀH species.37 The latter
would be expected in the oxamide and glycolamide materials due
to tautomerization of the amide functionality. The N 1S XPS
spectrum of V-Oxamide150 shows a large peak, which can be
simulated as three emissions at 396.8, 398.8, and 400.3 eV
(Figure S5; Supporting Information). As compared to the values
reported for monomeric compounds, the emission at 400.3 eV
could be attributed to a bound nitrogen from a À(CdO)ÀNH2
species.38 The emission at 398.8 eV can be assigned to the
nitrogen in À(CdO)ÀNHÀ species on the basis of previous
XPS on metal coordinated amides.35,39 The small emission at
396.7 eV can be attributed to a free amino N bound directly to
two vanadium centers,40 possibly formed through a small amount
of reductive CÀN cleavage due to decomposition. The high
intensity of the central emission being assigned to a proton-free
N species is consistent with three of the four oxamide protons
being used to eliminate the aryl group. The XPS of the N 1s
region of V-Glycolamide150 can be simulated as two peaks at
400.2 and 398.8 eV and can thus be assigned to analogous
À(CdO)ÀNH2 and À(CdO)ÀNHÀ species by comparison
to the oxamide species. The presence of NÀH is consistent with
incomplete elimination of the aryl group. No emission is detected
at 396.8 eV, indicating that there are no free amino groups in the
materials. This can be further confirmed by IR, as there are NÀH
stretches present from 3400À3500 cmÀ1 in V-Oxamide150 but
not in the IR of the other materials. XPS of all samples in the
valence region were also obtained (Figure S6; Supporting
Information). All materials with the exception of V-Oxamide
150 display a high density of states at the Fermi level consistent
with metallic behavior in these samples. This was confirmed by
the fact that charge neutralization was only required on the
oxamide sample.
To gain further insight into the composition of these materials,
we performed thermogravimetric analysis (TGA), which is often
used as a compliment to elemental analysis. The TGA curves for
all four samples are shown in the Supporting Information (Figure
S7) and show a steady drop in mass with heating at a rate of
5 °C/min from 200 to 350 °C, followed by a more gradual decrease
to 550 °C. For V-oxamide150, 45.09% mass was remaining at the
end oftheexperiment. Assuming completeconversiontoV2O5, the
%V in the original sample would be 25.26%. If the molecular
formula of this material was C2HN2O2V, then 37.46% of the
sample would be V, with 66.87% mass remaining after stoichio-
metric conversion to V2O5. If 1 THF per V center is included, as
suggested by the IR and XPS, then the material would have a
formula of C6H9N2O3V, with 24.48% V and 43.7% mass remaining
after conversion to V2O5. The TGA data are thus more consistent
with the later formulation of this material as a THF adduct as given
in Table S2. Likewise, V-Glycolamide150 can be formulated more
accurately as a THF adduct on the basis of 50.95% mass remain-
ing after combustion, while the remaining glycolic and oxalic acid
materials appear to be correctly formulated as described in
Scheme 1, with 49.13% and 42.01% mass remaining, respec-
tively, after combustion in the TGA experiment. The %V as
derived from TGA is shown for all materials in Table S1 and in
all cases more closely matches the expected values listed in
Table S2 than do the EA results.
Figure 3. Excess hydrogen adsorption isotherms at 77 K, 298 K of
V-Oxamide100, V-Oxamide150. Desorption at 298 K is left out for
clarity.
Table 1. At 77 K, there is an initial sharp rise to ∼0.5 wt % at the
beginning of adsorption process, followed by linear region up to
85 bar. The cyclopentadienyl chromium hydrazide materials
synthesized by our group possess similar adsorption isotherms
with a sharp rise at low pressure.21 This can be attributed to a
small amount of physisorption arising from the relatively low
surface area of the material and the adsorption of hydrogen
molecules on the outermost adsorption centers, which are coor-
dinatively unsaturated and exposed. At 85 bar, V-Oxamide100
adsorbs 2.25 wt % hydrogen at 77 K and 0.51 wt % at 298 K
(Table 1). Increasing the activation temperature to 150 °C
affords a material with 3.49 wt % adsorption at 77 K and 85
bar. From 0 to 30 bar, this material adsorbs up to 0.55 wt % at 298
K. After that, the isotherm is linear up to 85 bar, and an
adsorption capacity of 0.87 wt % is achieved. These results corr-
espond to a true volumetric density of 51 kg/m3 at 77 K and
13 kg/m3 at room temperature. The true volumetric density was
quoted previously by our group for vanadium hydrazide gels and
is defined as the product of the skeletal density and the excess
storage.22 While it is not typical in the physorption of carbons
and MOFs to use the skeletal densities in calculating the volum-
etric density, this is because of the large amount of void space in
these materials and the correspondingly large discrepancy be-
tween the bulk density and skeletal density. In our polymeric
materials, the void space has been minimized to optimize the
amount of adsorption arising from Kubas sites, so there is only a
small difference between skeletal density and bulk density. The
absolute volumetric adsorption of V-Oxamide150 material was
obtained by using a preweighed portion of material, which was
compressed at 500 psi pressure to remove void spaces between
particles. The compressed material possesses a bulk density of
0.9739 g/cm3. At 85 bar, it adsorbs 2.25 wt % of hydrogen at 77 K
and 0.99 wt % at 298 K. This corresponds to an absolute
volumetric adsorption of 21.91 and 9.6 kg/m3, respectively, at
77 and 298 K. This compares to an absolute volumetric adsorp-
tion of 30 kg/m3 for MOF 177 and a volumetric storage of MOF-
177, including the compressed gas, of 50.3 kg/m3 at 77 K and 8
kg/m3 at 298 K.26 Because V-Oxamide150 adsorbs hydrogen in a
linear fashion without saturation at 80 bar and 298 K, this
indicates that higher adsorption results could be expected at
higher pressures.
Figure 3 shows the excess hydrogen adsorption and desorp-
tion isotherms at 77 K and excess hydrogen adsorption at 298 K
of the V-Oxamide100 and V-Oxamide150 materials. The hydro-
gen storage results for this and other samples are summarized in
The excess storage isotherms of V-Oxalate150, V-Glyco-
late150, and V-Glycolamide150 are shown in Figures 4, 5, and
6, respectively. At 77 K, a sharp initial rise up to 0.65 wt % from
1 to 10 bar is observed in V-Oxalate150. This is attributed to
4959
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
Table 1. Summary of Excess Storage Data on Various Vanadium Materials and Carbon AX-21a
material
BET surface area (m2/g) skeletal density (g/cm3) gravimetric adsorption (wt %) volumetric adsorption (kg/m3) retention (%)
V-Oxamide100
28
9
1.2308
1.4636
1.6452
1.4581
1.4407
1.3110
1.2390
2.103
2.25 (at 77 K)
0.51 (at 298 K)
3.49 (at 77 K)
0.87 (at 298 K)
1.10 (at 77 K)
0.23 (at 298 K)
1.40 (at 77 K)
0.44 (at 298 K)
1.60 (at 77 K)
0.28 (at 298 K)
1.70 (at 77 K)
0.35 (at 298 K)
1.40 (at 77 K)
0.37 (at 298 K)
4.2 (at 77 K, 65 bar)
0.55 (at 298 K)
5.1 (at 77 K)
28 (at 77 K)
23
25
21
32
17
22
27
13
6.2 (at 298 K)
51 (at 77 K)
V-Oxamide150
V-Oxalate150
H2-V-Oxalate
V-Glycolate150
H2-V-Glycolate
V-Glycolamide150
AX-21
13 (at 298 K)
18 (at 77 K)
168
169
220
163
15
3.8 (at 298 K)
20 (at 77 K)
6.4 (at 298 K)
23 (at 77 K)
4.0 (at 298 K)
22 (at 77 K)
4.9 (at 298 K)
17 (at 77 K)
4.6 (at 298 K)
14 (at 77 K, 65 bar)
3225
3534
MOF-5(*)
5.5
0.28 (at 298 K)
a Data are taken at 85 bar. MOF adsorption data obtained from ref 41.
Figure 4. Hydrogen adsorption properties at 77 K, 298 K of V-Ox-
alate150. Desorption at 298 K is left out for clarity.
Figure 6. Hydrogen adsorption properties at 77 K, 298 K of V-Glyco-
lamide150. Desorption at 298 K is left out for clarity.
V-Glycolate150, followed by linear adsorption up to 1.6 wt %, is
also observed (Figure 3).
At room temperature and 85 bar, V-Oxalate150 and V-Gly-
colate150 adsorb 0.23 and 0.28 wt % of hydrogen, respectively.
The V-Glycolamide150 material does not exhibit a sharp physi-
sorption increase at low pressure and 77 K, and the adsorption
isotherm is linear in the pressure range up to 85 bar, at which it
absorbs 1.4 wt % hydrogen. A sharp increase is observed in the
desorption, however, as it plateaus at 0 bar and 0.5 wt % hydro-
gen. Placing the sample under a vacuum after the desorption
cycle removes this hysteresis as the second adsorption cycle is
then identical to the first. The reason this behavior is not obser-
ved in V-Oxamide150 is not understood, but may be related to
different pore structures in the material. This would introduce a
slight kinetic barrier to initial adsorption and desorption at low
pressure in this material, due to the low porosity and energy
required for hydrogen diffusion through the void space within
the materials.42 Moreover, at 298 K and 85 bar, V-Glycolamide
150 adsorbs 0.37 wt % hydrogen, which is higher than that of H2-V-
Glycolate materials, even though the V-Glycolamide150 possesses
Figure 5. Hydrogen adsorption properties at 77 K, 298 K of V-Glyco-
late150. Desorption at 298 K is left out for clarity.
physisorption and is followed by a linear region increasing up to
1.1 wt % at 85 bar (Figure 2). A similar rise up to 0.5 wt % for
4960
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
Table 2. V Concentration of Various Vanadium Materials
and Average Number of Hydrogen Molecules Adsorbed on
Each Vanadium Site at 85 bara
number of H2/V
at 77 K
number of H2/V
at 298 K
material
V-Oxamide150
V-Oxalate150
3.50
1.19
1.42
1.81
1.41
1.25
0.88
0.25
0.45
0.31
0.29
0.33
H2-V-Oxalate
V-Glycolate150
H2-V-Glycolate
V-Glycolamide150
a Numbers calculated from experimental TGA results.
Figure 7. Hydrogen binding enthalpies of V-Oxamide100, V-Oxa-
mide150, V-Oxalate150, V-Glycolate150, V-Glycolamide150, and car-
bon AX-21.
a much smaller nitrogen surface area. According to XPS results,
V-Glycolamide150 possesses the second highest concentration
of V(III) in the sample series, and these active species account
for the hydrogen adsorption capacity, physisorption in this case
representing minor adsorption capacity contribution.
Hysteretic adsorptionÀdesorption was also observed in
V-Oxalate150 and V-Glycolate150, however, to a much lower
degree. At room temperature, V-Glycolamide150 adsorbs up to
0.37 wt % at 85 bar, which is higher than that of V-Oxalate150
and V-Glycolate150, but lower than observed in the V-Oxamide
series. By comparing the gravimetric adsorption capacities at
298 and 77 K, the retention of excess adsorption capacities is
calculated and ranges from 17% to 32%. These values are higher
than those of MOF-5, which retains 5.5%,37 and carbon AX-21,
which retains 13%.21
To summarize these results, V-oxamide150 possesses both the
lowest surface area and the highest excess storage capacities. The
combination of low surface area and a high concentration of Kubas
binding sites in this material minimizes physisorption, allowing
the majority of adsorption to arise via the Kubas interaction. This
is clearly seen in Table 2, which shows the comparison of H2/V in
the materials studied. On the basis of the TGA results, the
oxamide material absorbs 3.5 H2/V at 77 K and 0.88 H2/V at 298
K, while the other materials range from 0.25 to 0.45 H2/V
(Table 2). By comparison, the vanadium hydrazides range in
1.13À1.96 H2/V center;22 the value of 3.5 H2/V at 77 K reflects
the presence of a component of physisorption in the total amount
absorbed, introducing a systematic error in the H2/V values. This
accounts for the fact that the maximum allowable by the 18 elec-
tron rule with THF coordination is 2 H2/V. Thus, a value of
2 H2/V is more reasonable and compares to a value of 2.18 H2/V
at 77 K for a compressed pellet of this material (Figure S8). So,
while not entirely accurate, these values are still useful bench-
marks in comparing the relative efficiency of H2 adsorption in
these materials. Thus, while these materials possess lower gravi-
metric and volumetric capacities of the best vanadium hydrazide
materials,22 which exhibit 1.17 wt % and 23.2 kg H2/m3, the more
controlled coordination environment around the V centers afford-
ed by the oxamide ligand as compared to hydrazine, which can
bond and bridge in several ways, may make these materials easier
to study and characterize.
adsorption enthalpy (Figure 7).13 This rising enthalpy trend has
been observed in all materials studied in our group involving
reduced early transition metal species, where the Kubas interac-
tion is believed to be operative as an adsorption mechanism.15
This phenomenon been predicted in calculations on Sc-coated
buckyballs,14 and evidence for the Kubas interaction in these
and related materials was provided by recent ESR work on
V-hydrazide gels.22 V-Oxalate150 and V-Glycolamide150 exhibit
adsorption heat in the range of 6.6À9.5 kJ/mol H2 and 2.1À
11.9 kJ/mol H2, respectively. The lower enthalpies correspond
with the lower hydrogen adsorption of these samples at 77 and
298 K. V-Glycolate150 possesses the lowest adsorption enthalpy at
∼2 kJ/mol with traditional physisorption behavior that decreases
with hydrogen concentration. This indicates that there is little or no
Kubas binding in the mechanism of adsorption of this material.
A 20-cycle adsorption run with pressure up to 85 bar was
carried out on the V-Oxamide150 sample. The result shows no
significant loss of excess adsorption capacity through cycling
(Figure S9, Supporting Information). The adsorption results
range randomly between 0.82 and 0.91 wt %, falling within the
instrument error ((0.05%, instrument’s user manual), and the
average adsorption capacity of this 20-cycle run is 0.87 wt %.
Hydrogenation was carried out on V-Oxalate150 and V-Gly-
colate150 according to Scheme 2 to replace residual mesityl
groups with hydride ligands, using experimental conditions deve-
loped previously by our group.19 Theoretical studies on the
interaction of hydrogen with different ligands showed that transi-
tion metal species with hydride ligands possess higher heat of
adsorption than those with alkyl, allyl, aryl, or benzyl groups.20
This is because alkyl, allyl, aryl, or benzyl groups are stronger π-
acceptors and exhibit more steric hindrance than hydride. More-
over, the removal of carbon will lead to a decrease in the sample
weight and a concomitant increase in gravimetric hydrogen
storage capacity. Thus, hydrogenation at 85 bar and 150 °C of
V-Oxalate150 and V-Glycolate150 decreases the carbon concen-
tration from 28.14% to 23.99% for V-Glycolate, while V-oxa-
late150 showed very little change. There was also an increase in
vanadium concentration from 19.64 to 21.60 wt % in the
V-Glycolate150 material, while again the V-Oxalate150 sample
showed little change. The TGA results (Figure S7) showed
44.87% remaining mass after combustion for the hydrogenated
V-Oxalate material as compared to 42.01% in the parent material,
indicating that only a small amount of hydrocarbon had been
removed. Likewise, 54.83% mass remained for the hydrogenated
Using a variant of ClausiusÀClapeyron I equation, the hydro-
gen adsorption enthalpies were calculated by employing two
sets of adsorption data at 77 and 87 K. V-Oxamide100 and
V-Oxamide150 possess adsorption enthalpies that rise versus
surface coverage and from 2.3 to 19.8 kJ/mol H2, approaching
the 20À30 kJ/mol H2 range, believed to be the ideal hydrogen
4961
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
Scheme 2. Proposed Hydrogenolysis Reactions of (a) V-Oxalate150 and (b) V-Glycolate150
Figure 8. Hydrogen adsorption properties at 77 K, 298 K of H2-V-
Oxalate. Desorption at 298 K is left out for clarity.
Figure 9. Hydrogen adsorption properties at 77 K, 298 K of H2-V-
Glycolate. Desorption at 298 K is left out for clarity.
V-Glycolate as compared to 49.13% in the starting material. If
complete hydrogenation had occurred, TGA would show 65.39%
and 72.18% mass remaining for the V-Oxalate and V-Glycolate
materials, respectively. Hydrogenation of V-Glycolate150 led to a
detectable decrease in surface area (from 220 to 163 m2/g), but
no significant change was observed in the surface area of the
hydrogenated V-Oxalate150 material. The IR spectra of these
materials both show a very minor decrease in the CÀH stretch,
indicating that only a small change has occurred in both samples.
There was also little detectable change in the XPS spectra of
either material. The difficulty in hydrogenating off the mesityl
group from V was observed previously in silica supported vana-
dium mesityl fragments.19
Hydrogenation has a more pronounced effect in the hydrogen
adsorption performance of V-Glycolate150 than that of V-Ox-
alate150 on the basis of elemental analysis and thermogravi-
metric analysis. H2-V-Oxalate and H2-V-Glycolate possess linear
hydrogen adsorption isotherms from 0 to 85 bar at room tem-
perature with better linearity than those of V-Oxalate150 and
V-Glycolate150. The excess storage isotherms of H2-V-Oxalate
and H2-V-Glycolate are shown in Figures 8 and 9, respectively.
H2-V-Oxalate demonstrated a rise to 0.7 wt % at 77 K and low
pressure, after which the isotherm became linear and rose to
1.4 wt % at 85 bar, which is 0.3 wt % higher than the adsorption
capacity of the non-hydrogenated V-Oxalate material. A 0.1 wt %
adsorption increment with respect to the non-hydrogenated
sample was observed at 77 K and 85 bar in H2-V-Glycolate.
The two hydrogenated materials possess linear adsorption iso-
therms at room temperature, H2-V-Oxalate and H2-V-Glycolate
Figure 10. Hydrogen binding enthalpies of H2-V-Oxalate150, H2-V-
Glycolate150.
adsorbing up to 0.44 and 0.35 wt %, respectively. Increasing the
hydrogenation reaction temperatures led to further decrease of
carbon content for the V-Glycolate150 sample, but no enhance-
ment of hydrogen adsorption capacities.
Heats of hydrogen adsorption of these two samples are
calculated using adsorption isotherms at 77 and 87 K and are
shown in Figure 10. In line with the negligible change in H2-V-
Oxalate on hydrogenation, there was little change in the enthalpy
behavior or magnitude in this sample. In the case of H2-V-
Glycolate, however, hydrogenation leads to a dramatic increase
in enthalpy, now rising from 2.2 to 17.3 kJ/mol H instead
4962
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
consistent with Raman studies on previous Kubas compounds43
and demonstrate for the first time that the rising enthalpies are
indeed related to the Kubas interaction, as the ESR evidence in a
previous report on V-hydrazides shows only that H2 is interacting
with the V(III) center, but does not give any information on the
intimate nature of this interaction. The position and lower
intensity of the υ(DÀD) is expected on the basis of previous
studies in the literature.43 This is thus the first Raman evidence of
the Kubas interaction in an extended solid in which the metals are
a part of the framework structure. Previous work on MOFs with
open metal sites suggested Kubas interactions but only Coulombic
interactions were observed;44À46 however, all Raman spectra
showed little or no bathochromic shift expected for this bonding
mode, all of the υ(HÀH) appearing above 4000 cm
.
À1 44 Raman
spectroscopy has been performed on Cu-exchanged ZSM-516 and
has shown a υ(HÀH) of 3079 and 3130 cmÀ1, indicative of a weak
Kubas interaction;however, the Cu is not a part of the framework in
this material, and υ(HÀH) values are relatively high as compared
to the majority of Kubas compounds in the literature.43 Further-
more, in the design of materials exploiting the Kubas interaction for
hydrogen storage, it is important that the metal comprises as large a
fraction as possible of the overall weight of the system and is not just
an ion or metal complex exchanged onto the surface of a much
heavier structure that does not absorb H2 in a Kubas fashion.
Figure 11. Raman spectra of V-Oxamide150 under H2 and D2.
of falling from 2.0 kJ/mol. The change in enthalpy behavior
corresponds to a noticeable change in hydrogen adsorption capacity
at 77 K and 85 bar (1.6À1.8 wt %) and the change in this material’s
composition. The enhancement of adsorption performance of H2-V-
Glycolate at 298 K and 85 bar (0.28À0.35 wt %) is smaller but still
noticeable. The most likely reason for the increase in enthalpy and
absorption is the opening of new more active V centers on loss of
hydrocarbon ligand as evidenced by the EA results. However, this
adsorption performance is much less than that of V-Oxamide150
because there are a limited number of Kubas hydrogen coordina-
tion sites available in V-Glycolate150 to begin with, so opening a
small amount of sites of higher enthalpy only leads to a moderate
change in gravimetric adsorption.
Vibrational spectroscopy studies were carried out to clarify the
interaction between hydrogen and the adsorption centers. In
previous work on vanadium hydrazides, we used ESR to confirm
that V(III) was interacting with H2 at room temperature.22 Vib-
rational spectroscopy is a more traditional method of studying
Kubas binding to metals because it potentially provides more
direct information on the mode of hydrogen interaction with the
metal, and the length of the HÀH bond can be used to gauge the
strength of the binding interaction.43 A free HÀH stretch is
Raman active, but IR inactive, while coordinated M(HÀH),
which becomes formally IR allowed due to polarization along the
M(HÀH) axis, is weak or invisible unless coupled to other bands
such as M(CO), and typically appears in the region from 2100
’ CONCLUSIONS
In summary, a series of new polymeric solids were synthesized
by the reaction between V(Mes)3 THF with various polyprotic
3
multidentate ligands. These polymers were designed with the
aim of creating well-defined Kubas sites for hydrogen storage that
could be studied by spectroscopy and gas sorption techniques.
The V-Oxamide150 sample showed the highest performance
with adsorption value of 3.49 wt % with adsorption enthalpy of
17.9 kJ/mol H2 at 77 K. At 298 K, the materials adsorb 0.87 wt %
at 85 bar with linear isotherms without reaching saturation, which
increases the value of these materials at higher pressure. While
these values fall short of the DOE goals, they are comparable to
the best MOFs at 298 K and 80 bar. Raman spectroscopy and
deuterium labeling confirmed the presence of Kubas binding in
these solids, for the first time unequivocally tying the rising
enthalpy trends observed previously in our systems based on low-
valent metals to the Kubas interaction. The DOE guidelines
specify a maximum pressure of 100 bar; however, commercial
pressure vessels for hydrogen transport are normally rated at 200
bar or higher, and our materials possess linear isotherms that do
not saturate at 298 K and 80 bar. We thus anticipate further gains
not only on applying this strategy to different metal systems, but
also recording isotherms at higher pressures.
to 2700 cm
.
À1 43 This band loses intensity and shifts to 1700À
1900 cmÀ1 in the case of the D2 analogue, demonstrating a
smaller net isotope effect than observed for MÀH(D).43 In
accordance with these expectations, V-Oxamide 150 shows no
change in the IR on treatment with H2 or D2 at 1 bar for 1 h
before sample preparation. After treatment with D2 for several
days, however, V-Oxamide150 exhibits a weaker intensity of the
VÀH band at 2200 cmÀ1 (Figure S9; Supporting Information),
suggesting that a small amount of deuterium substitution has
occurred.29 This phenomenon could be attributed to the forma-
tion of a metastable VH(D2) complex, which could yield either
VÀD or the original VÀH after the sample is evacuated.
’ ASSOCIATED CONTENT
S
Supporting Information. XRD, BET surface areas, IR
b
spectroscopy, and XPS of all samples dried at 150 °C, elemental
analysis, TGA, and adsorption of V-Oxamide150 compressed
pellet, and the 20-cycle hydrogen adsorption run of V-Oxa-
mide150. This material is available free of charge via the Internet
at http://pubs.acs.org.
Figure 11 shows the Raman spectrum of V-Oxamide150 under
H2 followed by the same sample after evacuation, this evacuated
sample under D2, and finally this D2 sample after vacuum. The
sample prior to any treatment with H2 or D2 is also shown for
reference. The strong band at 2553 cmÀ1 disappears under
vacuum, and after treatment with D2 a new band appears at
1927 cmÀ1, which also disappears after vacuum. These results are
’ AUTHOR INFORMATION
Corresponding Author
dantonel@glam.ac.uk
4963
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964

Journal of the American Chemical Society
ARTICLE
’ ACKNOWLEDGMENT
(20) Skipper, C.; Hamaed, A.; Antonelli, D.; Kaltsoyannis, N. J. Am.
Chem. Soc. 2010, 132, 17296.
(21) Mai, H. V.; Hoang, T. K. A.; Hamaed, A.; Trudeau, M.;
Antonelli, D. M. Chem. Commun. 2010, 46, 3206.
NSERC is acknowledged for funding.
(22) Hoang, T. K. A.; Webb, I. M.; Mai, H. V.; Hamaed, A.; Walsby,
C. J.; Trudeau, M.; Antonelli, D. M. J. Am. Chem. Soc. 2010, 132, 11792.
(23) Langmi, H. W.; Walton, A.; Ah-Mamouri, M. M.; Johnson,
S. R.; Book, D.; Speight, J. D.; Edwards, P. P.; Gameson, I.; Anderson,
P. A.; Harris, I. R. J. Alloys Compd. 2003, 356À357, 710.
(24) Seidel, V. W.; Kreisel, G. Z. Anorg. Allg. Chem. 1977, 435, 146.
(25) Hu, X.; Skadtchenko, B. O.; Trudeau, M.; Antonelli, D. M.
J. Am. Chem. Soc. 2006, 128, 11740.
’ REFERENCES
(1) (a) Schlapbach, L.; Zuettel, A. Nature 2001, 414, 353. (b)
Seayad, A. M.; Antonelli, D. M. Adv. Mater. 2004, 16, 765. (c) Yaghi,
O. M.; O’Keeffe, M.; Ockwig, N. W.; Chae, H. K.; Eddaoudi, M.; Kim, J.
Nature 2003, 423, 705. (d) Mark Thomas, K. Dalton Trans. 2009, 1487.
(e) Wang, L.; Yang, R. T. Energy Environ. Sci. 2008, 1, 268. (f) Lochan,
R. C.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2006, 8, 1357. (g)
Eberle, U.; Felderhoff, M.; Schueth, F. Angew. Chem., Int. Ed. 2009, 48, 2.
(h) Satyapal, S.; Petrovic, J.; Read, C.; Thomas, G.; Ordaz, G. Catal.
Today 2007, 120, 246. (i) Liu, C.; Li, F.; Ma, L.-P.; Cheng, H.-M. Adv.
Mater. 2010, 22, E28.
(2) (a) Langmi, H. W.; Book, D.; Walton, A.; Johnson, S. R.; Al-
Mamouri, M. M.; Speight, J. D.; Edwards, P. P.; Harris, I. R.; Anderson,
P. A. J. Alloys Compd. 2005, 404À406, 637. (b) Hirscher, M.; Panella, B.
Scr. Mater. 2007, 56, 809. (c) Bꢀenard, P.; Chahine, R. Scr. Mater. 2007,
56, 803. (d) Hu, Y. H.; Zhang, L. Adv. Mater. 2010, 22, E1. (e) Zhao, J.;
Shi, J.; Zhang, X.; Cheng, F.; Liang, J.; Tao, Z.; Chen, J. Adv. Mater. 2010,
22, 394.
(3) (a) Orimo, S.; Nakamori, Y.; Eliseo, J. R.; Zuettel, A.; Jensen,
C. M. Chem. Rev. 2007, 107, 4111. (b) Sandrock, G. J. Alloys Compd.
1999, 293, 877. (c) Bogdanovic, B.; Eberle, U.; Felderhoff, M.; Schueth,
F. Scr. Mater. 2007, 56, 809.
(4) Hydrogen, Fuel Cells and Infrastructure Technologies Program:
Multi-year Research, Development and Demonstration Plan: Planned
Program Activities for 2005À2015. http://www1.eere.energy.gov/hydro-
genandfuelcells/mypp/pdfs/storage.pdf; retrieved September 16th, 2009.
(5) Dornheim, M.; Doppiu, S.; Barkhordarian, G.; Boesenberg, U.;
Klassen, T.; Gutfleisch, O.; Bormann, R. Scr. Mater. 2007, 56, 841.
(6) Graetz, J.; Reilly, J. J. Scr. Mater. 2007, 56, 835.
(26) Furukawa, H.; Miller, M. A.; Yaghi, O. J. Mater. Chem. 2007,
17, 3197.
(27) Seidel, W.; Kreisel, G. Z. Chem. 1974, 14, 25.
(28) Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area and Porosity,
2nd ed.; Academic Press: London, 1982; pp 41À42.
(29) Heinekey, D. M.; Oldham, W. J., Jr. Chem. Rev. 1993, 93, 913.
(30) Antonelli, D. M.; Schaefer, W. P.; Parkin, G.; Bercaw, J. E.
J. Organomet. Chem. 1993, 462, 213.
(31) Groenenboom, C. J.; Sawatzky, G.; Meijer, H. J. D.; Jellinek, F.
J. Organomet. Chem. 1974, 76, C4–C6.
(32) Horvath, B.; Strutz, J.; Geyer-Lippmann, J.; Horvath, E. G. Z.
Anorg. Allg. Chem. 1981, 483, 181.
(33) Kasperkiewicz, J.; Kovacich, J. A.; Lichtman, D. J. Electron
Spectrosc. Relat. Phenom. 1983, 32, 123.
(34) Wagner, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F.;
Muilenberg, G. E. Handbook of X-Ray Photoelectron Spectroscopy: Physical
Electronics Division; Perkin-Elmer Corp.: Eden Prairie, MN, 1979.
(35) Whelan, C. M.; Cecchet, F.; Baxter, R.; Zerbetto, F.; Clarkson,
G. J.; Leigh, D. A.; Rudolf, P. J. Phys. Chem. B 2002, 106, 8739.
(36) Sivastava, S.; Badrinarayanan, S.; Mukhedkar, A. J. Polyhedron
1985, 4, 409.
(37) Batich, C. D.; Donald, D. S. J. Am. Chem. Soc. 1984, 106, 2758.
(38) Gervais, M.; Douy, A.; Gallot, B.; Erre, R. J. Colloid Interface Sci.
1988, 125, 146.
(39) Larkins, F. P.; Lubenfeld, A. J. Electron Spectrosc. Relat. Phenom.
1979, 15, 137.
(40) Kafizas, A.; Hyett, G.; Parkin, I. P. J. Mater. Chem. 2009, 19
1399.
(41) Panella, B.; Hirscher, M.; Puetter, H.; Mueller, U. Adv. Funct.
(7) Yang, R. T.; Wang, Y. J. Am. Chem. Soc. 2009, 131, 4224.
(8) (a) Strobel, R.; Garche, J.; Moseley, P. T.; Jorissen, L.; Wolf, G. J.
Power Sources 2006, 159, 781.
(9) (a) Wong-Foy, A. G.; Matzger, A. J.; Yaghi, O. M. J. Am. Chem.
Soc. 2006, 128, 3494. (b) Dailly, A.; Vajo, J. J.; Ahn, C. C. J. Phys. Chem. B
2006, 110, 1099. (c) Kaye, S. S.; Long, J. R. J. Am. Chem. Soc. 2005,
127, 6506. (d) Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim,
J.; O’Keeffe, M.; Yaghi, O. M. Science 2003, 300, 1127.
Mater. 2006, 16, 520.
(42) (a) Kang, J.; Wei, S.-H.; Kim, Y.-H. J. Am. Chem. Soc. 2010,
132, 1510. (b) Yamada, K.; Tanaka, H.; Yagashita, S.; Adachi, K.;
Uemura, T.; Kitagawa, S.; Kawata, S. Inorg. Chem. 2006, 45, 4322.
(43) Bender, B. R.; Kubas, G. J.; Jones, L. H.; Swanson, B. I.; Eckert,
J.; Capps, K. B.; Hoff, C. D. J. Am. Chem. Soc. 1997, 119, 9179.
(44) Murray, L. J.; Dincꢁa, M.; Long, J. R. Chem. Soc. Rev. 2009,
38, 1294.
(45) Kaye, S. S.; Long, J. R. J. Am. Chem. Soc. 2008, 130, 806.
(46) Centrone, A.; Siberio-Pꢀerez, D. Y.; Millward, A. R.; Yaghi,
O. M.; Matzger, A. J.; Zerbi, G. Chem. Phys. Lett. 2005, 411, 516.
(10) Yan, Y.; Lin, X.; Yang, S.; Blake, A. J.; Dailly, A.; Champness,
N. R.; Hubberstey, P.; Schroeder, M. Chem. Commun. 2009, 1025.
(11) (a) Zhou, W.; Yildirim, T. J. Phys. Chem. C 2008, 112, 8132. (b)
Dincꢁa, M.; Dailly, A.; Liu, Y.; Brown, C. M.; Neumann, D. A.; Long, J. R.
J. Am. Chem. Soc. 2006, 128, 12876.
(12) (a) Kubas, G. J.; Ryan, R. R.; Swanson, B. I.; Vergamini, P. J.;
Wasserman, H. J. J. Am. Chem. Soc. 1984, 106, 451. (b) Kubas, G. J.
Chem. Rev. 2007, 107, 4152. (c) Heinekey, D. M.; Lledꢀos, A.; Lluch, J. M.
Chem. Soc. Rev. 2004, 33, 175.
(13) (a) Vitillo, J. G.; Regli, L.; Chavan, S.; Ricchiardi, G.; Spoto, G.;
Dietzel, P. D. C.; Bordiga, S.; Zecchina, A. J. Am. Chem. Soc. 2008,
130, 8386. (b) Dincꢁa, M.; Long, J. R. Angew. Chem., Int. Ed. 2008,
47, 6766.
(14) Zhao, Y.; Kim, Y.; Dillon, A. C.; Heben, M. J.; Zhang, S. B. Phys.
Rev. Lett. 2005, 94, 155504.
(15) Hoang, T. K. A.; Antonelli, D. M. Adv. Mater. 2009, 21, 1787.
(16) Georgiev, P. A.; Albinati, A.; Mojet, B. L.; Ollivier, J.; Eckert, J.
J. Am. Chem. Soc. 2007, 129, 8086.
(17) Hamaed, A.; Trudeau, M.; Antonelli, D. M. J. Am. Chem. Soc.
2008, 130, 6992.
(18) (a) Hoang, T. K. A.; Hamaed, A.; Trudeau, M.; Antonelli, D. M.
J. Phys. Chem. C 2009, 113, 17240. (b) Hu, X.; Trudeau, M.; Antonelli,
D. M. Chem. Mater. 2007, 19, 1388. (c) Hamaed, A.; Hoang, T. K. A.;
Trudeau, M.; Antonelli, D. M. J. Organomet. Chem. 2009, 694, 2793.
(19) Hamaed, A.; Mai, H. V.; Hoang, T. K. A.; Trudeau, M.;
Antonelli, D. M. J. Phys. Chem. C 2010, 114, 8651.
4964
dx.doi.org/10.1021/ja110243r |J. Am. Chem. Soc. 2011, 133, 4955–4964