On the nature of the adsorbed hydrogen phase in microporous metal-organic frameworks at supercritical temperatures.
ABSTRACT Hydrogen adsorption measurements on different metal-organic frameworks (MOFs) over the 0-60 bar range at 50 and 77 K are presented. The results are discussed with respect to the materials' surface area and thermodynamic properties of the adsorbed phase. A nearly linear correlation between the maximum hydrogen excess amount adsorbed and the Brunauer-Emmett-Teller (BET) surface area was evidenced at both temperatures. Such a trend suggests that the adsorbed phase on the different materials is similar in nature. This interpretation is supported by measurements of the adsorbed hydrogen phase properties near saturation at 50 K. In particular it was found that the adsorbed hydrogen consistently exhibits liquid state properties despite significant structural and chemical differences between the tested adsorbents. This behavior is viewed as a consequence of molecular confinement in nanoscale pores. The variability in the trend relating the surface area and the amount of hydrogen adsorbed could be explained by differences in the adsorbed phase densities. Importantly, the latter were found to lie in the known range of bulk liquid hydrogen densities. The chemical composition and structure (e.g., pore size) were found to influence mainly how adsorption isotherms increase as a function of pressure. Finally, the absolute isotherms were calculated on the basis of measured adsorbed phase volumes, allowing for an estimation of the total amounts of hydrogen that can be stored in the microporous volumes at 50 K. These amounts were found to reach values up to 25% higher than their excess counterparts, and to correlate with the BET surface areas. The measurements and analysis in this study provide new insights on supercritical adsorption, as well as on possible limitations and optimization paths for MOFs as hydrogen storage materials.
- SourceAvailable from: Jorge Roque de la Puente[Show abstract] [Hide abstract]
ABSTRACT: High pressure H2 adsorption isotherms at N2 liquid temperature were recorded for the series of cubic nitroprussides, Ni1−xCox[Fe(CN)5NO] with x = 0, 0.5, 0.7, 1. The obtained data were interpreted according to the effective polarizing power for the metal found at the surface of the cavity. The cavity volume where the hydrogen molecules are accumulated was estimated from the amount of water molecules that are occupying that available space in the as-synthesized solids considering a water density of 1 g/cm3. The calculated cavity volume was then used to obtain the density of H2 storage in the cavity. For the Ni-containing material the highest storage density was obtained, in a cavity volume of 448.5 Å3 up to 10.4 hydrogen molecules are accumulated, for a local density of 77.6 g/L, above the density value corresponding to liquid hydrogen (71 g/L). Such high value of local density was interpreted as related to the electrostatic contribution to the adsorption potential for the hydrogen molecule within the cavity.Research highlights►High density hydrogen storage. ►Hydrogen adsorption forces. ►Hydrogen storage in nanocavities. ►Hydrogen storage in nanoporous solids.International Journal of Hydrogen Energy 12/2010; 35(23):12864-12869. · 2.93 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: D2 loaded to saturation at 25 K into the porous coordination framework Cu3(1,3,5-benzenetricarboxylate)2 was studied using neutron powder diffraction. A saturation loading equivalent to 6.1 wt % H2 is reached at 25 K. D2 is located at up to nine distinct sites with six stable at saturation and three “metastable” sites at doses intermediate to saturation. Filling of the trimodal pore system is complex, and a concentration dependency on the optimal D2 arrangement is noted. A dynamic rearrangement of the D2 in the pores to a cubic-closed packed equilibrium structure occurs at doses close to saturation, with 4 D2 molecules arranged tetrahedrally in the smallest pore, 32 D2 arranged in a truncated octahedron with capped hexagonal faces in the intermediate sized pore, and 48 D2 arranged in a face-capped rhombic dodecahedron in the largest pore. The equilibrium structure of D2 in the largest pore at doses close to and at saturation was found to be the same arrangement that is optimal for 48 circles on a sphere. The structural response of the framework to D2 adsorption is dependent on the amount of D2, where the host lattice expands with increasing amount of adsorbed D2 at low and high D2 doses, and contracts upon adsorption of intermediate amounts of D2, commensurate with the known flexibility exhibited by this framework.The Journal of Physical Chemistry C 04/2011; 115(17). · 4.84 Impact Factor
- Chemical Reviews 11/2012; · 45.66 Impact Factor
NRC Publications Archive (NPArC)
Archives des publications du CNRC (NPArC)
Publisher’s version / la version de l'éditeur:
Langmuir, 25, 20, pp. 12169-12176, 2009-09-01
On the Nature of the Adsorbed Hydrogen Phase in Microporous
Metal – Organic Frameworks at Supercritical Temperatures
Poirier, Eric; Dailly, Anne
Contact us / Contactez nous: firstname.lastname@example.org.
LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site
Web page / page Web
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Langmuir 2009, 25(20), 12169–12176 Published on Web 09/23/2009
©2009 American Chemical Society
On the Nature of the Adsorbed Hydrogen Phase in Microporous
Metal-Organic Frameworks at Supercritical Temperatures
Eric Poirier*,†,‡,§and Anne Dailly†
‡College of Engineering, Purdue University, West Lafayette, Indiana 47907.
Research Council Canada, Canadian Neutron Beam Centre, Chalk River Laboratories, Building 459, Station
143C, Chalk River ON, K0J 1J0 Canada.
§Current address: National
Received May 12, 2009. Revised Manuscript Received August 24, 2009
Hydrogen adsorption measurements on different metal-organic frameworks (MOFs) over the 0-60 bar range at 50
and 77 K are presented. The results are discussed with respect to the materials’ surface area and thermodynamic
properties of the adsorbed phase. A nearly linear correlation between the maximum hydrogen excess amount adsorbed
and the Brunauer-Emmett-Teller (BET) surface area was evidenced at both temperatures. Such a trend suggests that
the adsorbed phase on the different materials is similar in nature. This interpretation is supported by measurements of
the adsorbed hydrogen phase properties near saturation at 50 K. In particular it was found that the adsorbed hydrogen
consistently exhibits liquid state properties despite significant structural and chemical differences between the tested
trend relating the surface area and the amount of hydrogen adsorbed could be explained by differences in the adsorbed
phase densities. Importantly, the latter were found to lie in the known range of bulk liquid hydrogen densities. The
chemical composition and structure (e.g., pore size) were found to influence mainly how adsorption isotherms increase
as a function of pressure. Finally, the absolute isotherms were calculated on the basis of measured adsorbed phase
volumes, allowing for an estimation of the total amounts of hydrogen that can be stored in the microporous volumes at
50 K. These amounts were found to reach values up to 25% higher than their excess counterparts, and to correlate with
theBETsurfaceareas.Themeasurements andanalysisinthisstudyprovide newinsights onsupercriticaladsorption, as
well as on possible limitations and optimization paths for MOFs as hydrogen storage materials.
The environmental and economical concerns associated with
the current use of fossil fuels strongly stimulate research and
development on alternate fuels such as hydrogen. In the perspec-
development of a safe, compact, and efficient on-board fuel
storage system. Many hydrogen storage strategies are currently
under investigation in order to improve on the traditional
compression and liquefactionapproaches, which have significant
efficiency penalties.1The adsorption of molecular hydrogen in
microporous materials attracts significant interest because it can
allow the storage of a larger amount of gas than single compres-
sion under given pressure and temperature (P, T). Conversely, it
can provide for a lower operating pressure facilitating the use of
conformable tanks.1In addition, because of weak physical
solid-gas interactions, the adsorption mechanism is generally
fast and reversible, facilitating the handling of the fuel. The latter
characteristics are also critical for on-board application. The
hydrogen storage capacity of an adsorbent is determined by its
most interesting materials in that respect are the metal-organic
frameworks (MOFs). The latter have specific surface areas and
microporevolumes that canexceedthat oftraditional adsorbents
such as zeolites and activated carbons. MOFs are hybrid inorga-
or clusters, through rigid organic ligands. The variety of cations
and molecular bridges that can be combined in the framework
yields an extended range of materials with diverse pore sizes and
functionalities. Efforts are now being made to tune these adsor-
that regard, the extent to which pore geometry and functionality
affect the nature of the adsorbed hydrogen phase is a crucial
question in determining how these materials can be improved.
Because of their crystalline nature, MOFs offer a well-defined
structuralplatformfrom which, in principle, thesorptionproper-
ties and mechanism can be studied. This is illustrated in the
Figure 1 and 2 which show how the surface area of IRMOF-1
(Zn(BDC)) is generated by the framework. In this case, the
Connolly surface was plotted using the Material Studio software.
Clearly both the extent of surface area and accessible porous
volume appear as determinant factors for the hydrogen storage
capacity of a MOF. An important step for the tuning of
ship between the adsorption enthalpy, the surface area and the
porous volume. At low pressures the hydrogen uptake correlates
with the adsorption enthalpy.2This quantity is a measure of the
strength of the solid-gas interactions and is determined by
the nature of the chemical constituents and the structure of the
adsorbent. At higher pressures, i.e., when the excess maximum is
reached, hydrogen uptakes can be related to the specific surface
area and the porous volume. The Brunauer-Emmett-Teller
(BET) specific surface area, along with the micropore volume
*Corresponding author. E-mail: Eric.M.Poirier@nrc.gc.ca. Tel: 613-584-
8811 ext. 44179. Fax: 613-584-4040.
(1) Poirier, E.; Chahine, R.; B? enard, P.; Cossement, D.; Lafi, L.; M? elanc -on, E.;
Bose, T. K.; D? esilets, S. Appl. Phys. A: Mater. Sci. Process 2004, 78, 961–967.
(2) Frost, H.; D€ uren, T.; Snurr, R. S. J. Phys. Chem. B 2006, 110, 9565–9570.
Langmuir 2009, 25(20), 12169–12176
Article Poirier and Dailly
as it is demonstrated for activated carbons and zeolites.3,4Good
correlations between these quantities were also obtained on a set
of MOFs.5Microporous MOFs provide strong confinement of
the guest species, with pores of dimensions inferior to 2 nm.
Whether the BET method determines actual geometric areas is
questionable and, in such case, the surface area may be qualified
as “apparent”.6In fact, owing to confinement in micropores, the
potential wells from opposite walls can overlap which results in
micropore filling by a liquid.7This phenomenon has been
observed in several MOFs in the 50-100 K range.8-10More
specifically, the adsorbed hydrogen phase near pore saturation
was found to behave like an incompressible fluid and to reach
bulk liquid densities.8-10The isotherms revealing these typical
liquid state properties could be modeled successfully using a
Dubinin-Astakhov micropore filling equation. Hence, as as-
sumed in earlier works, such mechanism is not restricted to
subcritical temperatures.11These investigations have shown
how the determination of some thermodynamic properties of
the adsorbed phase near saturation can reveal both the physical
limits of the adsorbent and its influence on the guest molecules.
on the nature and on the behavior of the adsorbed hydrogen
phase in MOFs in connection with their specific surface area,
structure and chemical functionality. The influence of the struc-
ture and composition of the adsorbents on the nature of the
adsorbed phase is investigated in that perspective. The approach
involves the synthesis of an extended set of MOFs along with the
measurement and analysis of their excess hydrogen adsorption
isotherms, and their porosity. Traditional adsorbents such as
activated carbons and molecular sieves were also tested for
comparison. Hydrogen adsorption measurements were per-
adsorbed phase densities and volumes, and the corresponding
absolute isotherms. These quantities were used to investigate the
state of the adsorbed supercritical hydrogen. The results are
discussed in the perspective of on-board hydrogen storage.
2. Theoretical Background
Solid-gas adsorption consists essentially of the enrichment of
guest molecules(i.e., adsorbate), in the vicinity ofthe surface ofa
solid over the concentration of the bulk gas phase. The total
amount of adsorbate molecules present in the space Vain which
absolute amount adsorbed Nadefined as12
where Fais the density of the adsorbed phase. The amount Na
the adsorbed phase is not measurable under usual (P, T) condi-
with reference to a nonadsorbing system and the corresponding
measured quantity is referred to as the excess amount adsorbed
Nex.13The latter corresponds to the fraction of all the adsorbate
and Nexare formally related as
Nexis characterized by a maximum that occurs when Fa
approaches monotonically a constant value while the gas density
Nexexhibits near saturation a characteristic decrease from which
adsorbed phase properties can be studied. Given that sufficient
data are available near saturation (e.g., at very low temperature
and relatively high pressure), the volume Vaoccupied by the
adsorbed phase can be extracted as8-10,14
= Vafor Naf constant
If eq 3 behaves linearly, that is Vadoes not change with
Figure 1. Unit cell structure of the Zn(BDC) (IRMOF-1) as
represented with the Material Studio software platform.
Figure 2. Surface area of the Zn(BDC) as illustrated using the
(3) Chahine,R.;Bose,T.K.HydrogenEnergyProgressXI;Veziroglu,T.N., etal.,
Eds.; International Association of Hydrogen Energy: Coral Gables, FL,1996; p 1259.
(4) Sing, K. Colloids Surf., A 2001, 187-188, 3–9.
(5) Hirsher, M.; Panella, B. Scripta Mater. 2006, 56, 809–812.
(6) Fletcher, A.J.; Thomas, K. M.;Rosseinsky, M.J. J. Solid State Chem. 2005,
(7) Dubinin,M.M.;Astakhov, V.A. Izv.Akad.NaukSSSRSer.Khim.1971,1,
5-11 (translated from Russian).
(8) Poirier, E.; Dailly, A. J. Phys. Chem. C 2008, 112, 13047–13052.
(9) Poirier, E.; Dailly, A. Nanotechnology 2009, 20, 204006.
(10) Poirier, E.; Dailly, A. Energy Environ. Sci. 2009, 2(4), 420–425.
(11) Amankwah, K. A. G.; Schwarz, J. A. Carbon 1995, 33, 1313–1319.
(12) Rouquerol, F.; Rouquerol, J.; Sing, K. Adsorption by Powders and Porous
Solids; Academic Press: London, 1999.
(13) B? enard, P.; Chahine, R. Langmuir 2001, 17, 1950–1955.
(14) Menon, P. G. Chem. Rev. 1968, 68, 277–294.
Langmuir 2009, 25(20), 12169–12176
Poirier and Dailly Article
isotherm near saturation can be directly associated with the
displacement of gas by the adsorbed phase. The density of the
adsorbed phase can be approximated from the expression
Equation4 reduces toFa= Fgwhen Nex= 0,reflecting thatno
“gain” over compression is made when the gas density is high
enough. In the absence of phase transition, neither Vanor Fa
should vary inthe saturationregime. This will be verified later by
performing linear analysis of the isotherms near saturation, and
extrapolating down to Nex= 0. The quantities Vaand Faare
closely related to the underlying physics of the system and its
ability to adsorb hydrogen.15They can reveal of the effects of
chemical functionality, pore size and gas-gas interactions. More-
over, Vacan be used to calculate Naby means of eqs 3 and 4 as
Na= Nexþ FgVa
approaches, without independent assumptions about the porous
volume of the adsorbent (e.g., crystallographic volume). The
quantity Faprovides a basis for the calculation of the mean
spacing between adsorbed molecules, and quantum effects. The
classical or quantum nature of the adsorbed hydrogen can be
identified using the de Broglie thermal wavelength, which is
where β = (kbT)-1and p = h/2π, m is the mass of a particle, kb
and p are respectively the Boltzmann and Planck constants. Λ is
usually compared with the mean spacing between hydrogen
molecules, a0, which can be deduced from Faas a0= Fa-1/3. A
classical treatment is valid if Λ is much smaller than a0, that is
Λ/a0,1; otherwise quantum effects can be considered. Because
the quantum behavior of molecules tends to increase their
3. Experimental Section
3.1. Materials. In this study we examine the hydrogen
adsorption properties of a large number of MOFs composed of
different inorganic clusters and linking organic units. All labora-
tory scale samples were prepared via a one-pot reaction by
solvothermal methods from a solution or slurry of the metal ion
salt with the corresponding linking organic unit. MOFs for
hydrogen storage purposes have extended three-dimensional
structures with structural stability incorporating uniform pores
and a network of channels. The activation of the porosity was
achieved by the removal of noncoordinating guests’ species like
The first group of materials we studied is first-generation MOFs
based on aromatic carboxylate ligands such as BDC (1,4-
tribenzoate),21,22TTDC (thieno[3,2-b]thiophene -2,5- dicarbo-
xylate),23TCN (3,5,30,50-tetracarboxylate naphthalene),24TCP
(3,5,30,50-tetracarboxylate phenanthrene),24TPTC (terphenyl
DOBDC (2,5-dihydroxy-1,4-benzenedicarboxylate).28We also
have examined microporous coordination solids analogous to
the MOFs where carboxylate ligands are replaced by structurally
analogous functional nitrogen-based bridging ligands while
essentially keeping the same specific surface area. The linking
organic units are BDP (1,4-benzenedipyrazolate), BTT (1,3,5-
methane).31Finally we explored two MOFs based on metal
imidazolates: PhIM (benzimidazolate) and MeIM (2-methyl-
imidazolate).32The molecular structures of the mentioned ex-
tended ligands are shown in Figure 3. Considering that some
MOFs are known to be air/moisture sensitive, they were system-
known coconut shell KOH-activated carbon obtained from AK
Research Inc., as well as a type Y molecular sieve, obtained from
Alfa Aesar, were also tested for comparison purposes.
3.2. Hydrogen Sorption and Specific Surface Area Mea-
vacuum up to 373 K prior to each adsorption experiment. The
hydrogen adsorption isotherms were measured using an auto-
to control precisely the temperature in the pressurized cell within
adsorbed in excess was calculated from a set of gas expansion by
comparing the amount of hydrogen sent to the sample cell with
the amount of residual gaseous hydrogen in the same cell after
summed as a function of pressure to build the excess adsorption
isotherms. The equilibrium conditions were carefully identified
using graphical plotters that monitor the system pressure and the
temperature as a function of time. The dead space volumes were
determined at room temperature using helium as a negligibly
adsorbing gas. Ultrahigh purity hydrogen and helium (99.999%
purity) obtained from Airgas Inc. were used. The porosity of the
(15) Frost, H.; Snurr, R. Q. J. Phys. Chem. C 2007, 111, 18794–18803.
(17) Li, H.; Eddaoudi, M.; O’Keeffe, M.; Yaghi, O. M. Nature 2008, 402, 276–
(18) Barthelet, K.; Marrot, J.; Riou, D.; Ferey, G. Angew. Chem. Int. Ed. 2002,
S.; Margiolaki, I. Science 2005, 309, 2040–2042.
(20) Loiseau, T.; Serre, C.; Huguenard, C.; Fink, G.; Taulelle, F.; Henry, M.;
Bataille, T.; Ferey, G. Chem.;Eur. J. 2004, 10, 1373–1382.
(21) Furukawa, H.; Miller, M. A.; Yaghi, O. M. J. Mater. Chem. 2007, 17(30),
2001, 291(5506), 1021–1023.
(23) Rowsell, J. L. C.; Yaghi, O. M. J. Am. Chem. Soc. 2006, 128, 1304–1315.
(24) Yang, S. H.; Lin, X.; Dailly, A.; Blake, A. J.; Hubberstey, P.; Champness,
N. R.; Schroder, M. Chem.;Eur. J. 2009, 15(19), 4829–4835.
Zoppi, M.; Walker, G. S.; Thomas, K. M.; Mays, T. J.; Hubberstey, P.;
Champness, N. R.; Schroder, M. J. Am. Chem. Soc. 2009, 131(6), 2159–2171.
(26) Mueller, U.; Schubert, M.; Teich, F.; Puetter, H.; Schierle-Arndt, K.;
Pastre, J. J. Mater. Chem. 2006, 16, 626–636.
(27) Loiseau, T.; Lecroq, L.; Volkringer, C.; Marrot, J.; Ferey, G.; haouas, M.;
(28) Caskey, S. R.; Wong-Foy, A. G.; Matzger, A. J. J. Am. Chem. Soc. 2008,
(29) Dinca, M.; Dailly, A.; Liu, Y.; Brown, C. M.; Neumann, D. A.; Long, J. R.
J. Am. Chem. Soc. 2006, 128(51), 16876–16883.
(31) Dinca, M.; Dailly, A.; Long, J. R. Chem.;Eur. J. 2008, 14(33), 10280–
(32) Park, K. S.; Ni, Z.; Cote, A. P.; Choi, J. Y.; Huang, R.; Uribe-Romo, F. J.;
Chae, H. K.; O’Keeffe, M.; Yaghi, O. M. Proc. Natl. Acad. Sci. 2006, 103(27),
(33) Kaye, S. S.; Dailly, A.; Yaghi, O. M.; Long, J. R. J. Am. Chem. Soc. 2007,
Langmuir 2009, 25(20), 12169–12176
Article Poirier and Dailly
activated materials was measured using a Quantachrome Auto-
sorb-1 automated gas-sorption apparatus. The specific surface
areas of the samples were evaluated from nitrogen sorption
isotherms at 77 K using the BET model. The total pore volumes
were extracted from argon adsorption isotherms measured at
4. Results and Discussion
of data, including excess isotherms of hydrogen adsorption and
desorption, maximum hydrogen uptakes as well as Bru-
nauer-Emmett-Teller (BET) specific surface areas (N2), have
been measured. The 77 K excess hydrogen adsorption isotherms
MOFs materials. Note the hydrogen excess adsorption uptakes
are reported in specific units, mass of hydrogen per gram of
basis than the surface area units. As expected from a physisorp-
tionprocess, the hydrogen adsorptionisotherms were completely
reversible for all the samples except for Al(BDC) and Co(BDP);
which showed hysteretic behaviors, as discussed later. In most
5 min. The samples could also be regenerated by degassing at
room temperature, showing weak solid-gas interactions. The
materials showed reproducible hydrogen uptakes following sev-
eral of such adsorption/desorption cycles, revealing their struc-
tures were stable in that respect. Figure 4 presents a survey of the
function of the BET specific surface areas. The metal ion of the
SBU and the linking organic unit are indicated for each studied
MOF. The higher maximum excess hydrogen uptakes measured
at 77 K range between 6 and 7.5 wt % for Zn(BTB) (MOF-177),
Zn(BDC)(BTB) (UMCM-1), Zn(BDC) (IRMOF-1), and Zn-
(TTDC) (IRMOF-20). These MOFs, with very high surface
is limited to about 5.5 wt % hydrogen in excess at 77 K. This
illustrates that MOFs, as an emerging technology, have signifi-
cantcapabilities formolecular hydrogenstorageonagravimetric
surface areas as evidenced by the trend curve shown in Figure 4.
Figure 3. Organic linkers used in the coordination frameworks presented in this study.
Figure 4. Maximum excess hydrogen uptakes measured at 77 K
versus BET specific surface areas for different types of MOFs.
Langmuir 2009, 25(20), 12169–12176
Poirier and DaillyArticle
per 500 m2g-1. This value corresponds essentially to Chahine’s
rule established for activated carbons and zeolites.3This correla-
data were indeed obtained on qualitatively different materials,
including porous frameworks built from the linking of different
building blocks and organic units with different functionalities.
The error on the presented trend is, in part, inherent to the BET
method.34It can also be associated to the presence of a relatively
high number of very small pores in some materials. Because of
their larger molecular size, the nitrogen molecules used for the
also occur because of micropore filling and differences in adsor-
bate packing densities. As revealed in a previous report, the
adsorbed phase density of hydrogen may vary up to 40% on
different materials near saturation at 50 K.9This aspect will now
be further investigated.
4.2. ExcessHydrogenAdsorption Measurementsat 50K.
In order to get further insights into the nature of the adsorbed
phase and its influence on the correlation with the surface area,
the excess adsorption isotherms of some specific microporous
materials were analyzed close to saturation at 50 K. It can be
assumed that the adsorption mechanism and the state of the
as already observed on Zn(BDC), Zn(BTB), Cu(TPTC), and Cu
materials were, in fact, successfully modeled with a same set of
Dubinin-Astakhov model parameters over the 50-77 K range.
The applicability of the model is owed to the small pore sizes in
these materials, which are about 1.7, 1.5, 0.73, and 0.65 nm
respectively for the Zn(BDC), Zn(BTB), Cu(TPTC), and Cu-
(BPTC).21,37,38In this section, these materials are studied side by
sites which make them interesting because of their enhanced
affinity for guest species.39Consistent with eqs 3 and 4, the
measured excess adsorption isotherms are expressed in Figure 5
as a function of the gas phase density.40The behaviors of the
isotherms at low filling (pressures) vary for the different samples,
differences inthe organicunits.41For instance,the higheraffinity
for molecular hydrogen in the open metal sites materials such as
the Cu-based MOFs is a likely cause of a steeper increase of
adsorption as a function of pressure.41In the near saturation
region, observed past the excess maxima, the excess isotherms on
a function of Fg. In accordance with eq 3, this remarkable feature
indicates that an incompressible adsorbed phase is formed on all
of the present samples. From that standpoint, it is interesting to
note that the adsorbed hydrogen phase on the MOFs does not
differ from that on the traditional adsorbents (i.e., AX-21 and
MS). The adsorbed phase volumes and densities were calculated
using eqs 3 and 4 in the saturation regions of the 50 K excess
isotherms; the results are presented in Table 1. These adsorbed
phase volumes are, in many cases, consistent with the volumes
measured using argon porosimetry, which are also presented in
the table for comparison. It was observed that the values of
adsorbed hydrogen phase volumes exceed to some extent that of
reported porous volumes calculated from crystallographic data.
However, the volume ratios between samples are comparable
using either quantity, e.g., in the case of the Cu-MOFs one finds
Va(BPTC)/Va(TPTC)= 0.63; the same ratio can be obtained using
the porousvolumes calculatedby Lin etal. fromcrystallographic
data (i.e., using Vp(BPTC)= 0.68 and Vp(TPTC)= 1.08 mL g-1).38
The difference between these volumes and those measured using
eq 3 might be due to defects and structural transformations such
asstretching, rotational,“breathing” and scissoringmechanisms,
exacerbated at the lowest temperatures. The adsorbed phase
Table 1. Summary of Hydrogen Adsorption Measurements at 50 K and Structural Propertiesa
aNote that Nex
measured with Argon; Fa
max: maximum excess hydrogen uptake; ABET: BET specific surface area; Va: adsorbed hydrogen phase volume; Vp
max: maximum adsorbed hydrogen phase density; Na
Ar: porous volume
max: maximum absolute hydrogen uptake.
Figure 5. Plots of the measured excess hydrogen adsorption at
50 K expressed as a function of the gas density. The saturation
regimes past the excess maxima were linearly fitted and extrapo-
lated to Nex= 0 .
(34) Gregg S. J.; Sing K. S. W. Adsorption, surface area and porosity, 2nd ed.;
Academic Press: New York, 1982.
(35) Poirier, E.; Chahine, R.; Bose, T. K. Int.J. Hydrogen Energy2001, 26, 831–
(36) Ahn, C. C.; Ye, Y.; Ratnakumar, B. V.; Witham, C.; Bowman, R. C., Jr.;
Fultz, B. Appl. Phys. Lett. 1998, 73, 3378–3380.
(38) Lin, X.; Jia, J.; Zhao, X.; Thomas, K. M.; Blake, A. J.; Walker, G. S.;
Champness, N. R.; Hubberstey, P.; Schr€ oder, M. Angew. Chem., Int. Ed. 2006, 45,
(39) Panella, B.; Hirscher, M.; Putter, H.; Muller, U. Adv. Funct. Mater. 2006,
(40) Lemmon E. W.; Peskin A. P.; McLinden M. O.; Friend D. G. NIST12
Database 23, Version 8.0; U.S. Secretary of Commerce: Washington, DC, 2007.
(41) Rowsell, J. L. C.; Millward, A. R.; Park, K. S.; Yaghi, O. M. J. Am. Chem.
Soc. 2004, 126, 5666–5667.
(42) Fletcher, A. J.; Thomas, K. M.; Rosseinsky, M. J. J. Solid State Chem.
2005, 178, 2491–2510.
Langmuir 2009, 25(20), 12169–12176
ArticlePoirier and Dailly
Notably, this range of densities corresponds to that of liquid
In this view, differences in the affinities of the adsorbents for
hydrogen seem to mimic changes in the equilibrium (P, T)
conditions of the bulk liquid. An inspection of figure 5 also
clearly shows that linear extrapolations down to Nex= 0, where
Fa= Fg, yield adsorbed phase densities consistent with those
constant a little past the excess maxima, meaning the latter
indicates well thebeginningofthesaturationregime.Thisfeature
bulk supercritical gas does not have a well-defined saturation
Interestingly, it is apparent from Figure 5 that the amount
adsorbed rises steeply with increasing pressure on both Cu-
(BPTC) and the MS. On the other hand, these samples also have
reduction of the pore size may not lead to any gain in the
maximum hydrogen packing densities despite enhancing so-
lid-gas interactions. For instance, this can be evidenced
by considering the adsorbed phase density of Cu(BPTC)
(51 g L-1), which is smaller than that of Zn(BTB) (55 g L-1),
despite smaller pores. The behavior of the curves in Figure 5 is
qualitatively comparable to the excess hydrogen adsorption
isotherms for different slit pore widths calculated by B? enard
et al. using a Grand Canonical Simulations approach.43In the
present view, one advantage of the higher hydrogen affinity in
small pores would be potentially milder (P, T) operating condi-
tions. This aspect will be discussed in section 4.4 in relation to
adsorption enthalpy calculations.
At the microscopic level, some limiting factors for the packing
of molecules in small pores are associated with electronic repul-
sion between adsorbate molecules, and quantum effects. The
latter become particularly important at low temperatures. The
magnitude of such effect can be assessed using the de Broglie
wavelength. For hydrogen at 50 K, eq 6 yields Λ = 0.17 nm.
Considering the mean spacing between hydrogen molecules is a0
= Fa-1/3, then for the observed densities varying between 50 and
70 g L-1one obtains a0= 0.41-0.36. Hence the ratio Λ/a0lies
between about 0.4-0.5 showing the adsorbed hydrogen has a
significant wave-like character. This phenomenon is likely to
oppose the densification of the adsorbed phase and limit the
hydrogen uptake in the present conditions.9,44
4.3. Absolute Adsorption Isotherms at 50 K. The pre-
viously calculated Vaand Favalues can be used to evaluate the
absolute amounts adsorbed such as Na= FaVa. The latter are
presented for the specific set of samples in the last column of
Table 1. On this basis, considerable amounts can be stored at
50 K, up to 25% more than the maximum excess values. In
particular, a substantial capacity of about 12 wt % is reached on
the Zn(BTB). The extent of the absolute uptake is mainly
determined by the adsorbed phase volume. Note that Faappears
as a relatively less important factor to reach high hydrogen
storage capacities as it does not vary as much as Vawith respect
using eq 5, along with their excess counterparts are plotted, for
example, for the Zn(BTB) and Cu(BPTC) in Figure 6. Both
absolute isotherms are of type I and clearly exhibit the typical
monotonic behavior associated with a saturation regime at high
pressure. The comparatively lower performance of Cu(BPTC) is,
as exposed in the previous section, a consequence of reduced Va
and Favalues on this material.
Figure 7 expresses the maximum absolute and excess amounts
adsorbents including those presented in Table 1.A linearcorrela-
tion obtained with the absolute amount is slightly better than the
the slopes. Also, in the case of excess adsorption, the slope, m0=
0.0255 mg m-2, is about 25% higher than that obtained at 77 K
(i.e., 0.02 mg m-2). This is naturally a result of the higher excess
adsorption maxima at 50 K which, in turn, might be due to an
increasing contribution of weakly attractive sites as the tempera-
The variability in these curves can possibly be explained by
differences in the hydrogen packing densities and by the occur-
rence of micropore filling. The later may be quantitatively
equivalent to the superposition of adsorbed layers in the pores.
In order to investigate this matter one may consider the mean
lateral spacing between adsorbed molecules deduced from the
apparent “surface” coverage. In essence, this quantity can be
estimated as ~ a0= (1/m0)1/2. It is found from the slope of the
absolute amount as a function of the BET surface area that ~ a0=
0.32 nm. This spacing is smaller than the previous spacing
calculated from the adsorbed phase densities and, in fact, it is
even smaller than under a solid hydrogen density. This unlikely
result suggests the microporous space above the surface might be
filled, leading to an overestimation of apparent surface coverage.
Figure 6. Plots of excess and absolute isotherms on Zn(BTB) and
Cu(BPTC) at 50 K.
Figure 7. Absolute and excess amounts adsorbed at 50 K as a
function of the BET specific surface area.
(43) B? enard, P.; Chahine, R.; Chandonia, P. A.; Cossement, D.; Dorval-
Douville, G.; Lafi, L.; Lachance, P.; Paggiaro, R.; Poirier, E. J. Alloys Compd.
2007, 446-447, 380–384.
(44) Garberoglio, G.; Skoulidas, A. I.; Johnson, J. K. J. Phys. Chem. B 2005,
Langmuir 2009, 25(20), 12169–12176
Poirier and DaillyArticle
the surface area (ABET) and the pore volume: Va(ABET)-1= na0,
where the quantity n would represent the “apparent” number of
be interpreted as various “stacking” arrangements of adsorbed
molecules in the pores. These different arrangements may also be
related to the variability in trends relating the amounts adsorbed
and the BET surface areas.
4.4. Adsorption at Different Temperatures. In order to
assess the magnitude of the solid-gas interactions and their
influence on hydrogen uptakes, adsorption at different tempera-
tures has been examined. Excess hydrogen adsorption on Zn-
(BDC), Zn(BTB), Cu(BPTC), and Cu(TPTC) materialshas been
previously measured between 50 and 87 K and up to 40 bar.8-10
The isotherms measured on this set of MOFs materials were also
successfully modeled using the Dubinin-Astakhov micropore
filling equationin a formadapted for excessadsorption. Figure 8
the adsorption enthalpy as a function of fractional filling and
temperature using the Clausius-Clapeyron equation. Note the
complete modeling approach was described in details in previous
articles.8-10The magnitude of the adsorption enthalpy (ΔH) at
with the smaller pore size, Cu(BPTC), shows the highest ΔH, in
pressure noted in section 4.2. All the enthalpies lie in a range of
values consistent with physical types ofinteractions. The diminu-
tion of ΔH with fractional filling, particularly rapid for the
material with smaller pores Cu(BPTC), may be due to more
variability in the energetic heterogeneity. This observation sug-
gests that adsorption in these conditions departs from a mechan-
the amounts adsorbed is revealed by comparing how the excess
maxima vary as a function of temperature for the different
materials, as shown in Figure 10.
This figure shows that amounts adsorbed by the Zn-based
MOFs surpass considerably that of other adsorbents over the
whole 50-87K range.The Zn(BTB), for instance, has maximum
hydrogen excess uptakes 30-40% larger than AX-21 activated
carbon. The decrease ofthe curves with increasing temperature is
shortest linker and the highest ΔH, i.e. Cu(BPTC), exhibits
steadier maximum excess amounts with increasing temperature.
Therefore, although reducing the pore size may not enhance the
maximum adsorbed phase density, it favors the retention of the
hydrogen as the temperature increases as a result of the stronger
solid-gas interaction. This observation clearly exposes an inher-
ent contradiction between the large pore volumes required to
enhance hydrogen storage capacity and the resulting decrease in
seen from the reverse order of the curves between Figures 9 and
10. Hence, the potential (P, T) operating conditions may be
ΔH, but atthecostofareductionofporevolumewhilenogain is
made in term of maximum adsorbed phase density. These con-
clusions are consistent with some recent independent theoretical
calculations.2More experiments will be needed to determine if
similar observations can be made beyond this particular set of
Hydrogen adsorption measurements performed over an ex-
maximum excess amounts adsorbed at 77 and 50 K and the BET
phase near saturation is essentially independent of the structure
is supported by measurements at 50 K which showed that the
adsorbed phase behaves like an incompressible fluid reaching
Figure 8. Measured (data points) and modeled (dashed) excess
adsorption isotherms on Zn(BDC) over the 50-87 K range
(adapted from ref 8). A form of the Dubinin-Astakhov equation
adapted for excess adsorption was used for modeling.
Figure 9. Magnitudeofthehydrogenadsorptionenthalpyat50K
as a function of filling on Zn- and Cu-based MOFs.
Figure 10. Profile of the excess adsorption maxima of various
MOFs as function of temperature. The activated carbon AX-21
is also shown for comparison.
Langmuir 2009, 25(20), 12169–12176
ArticlePoirier and Dailly
a set of qualitatively different samples, shows the adsorbed phase
near saturation is not altered much by differences in the sample’s
affinity for molecular hydrogen. The variability (∼10%) in the
correlation between the maximum hydrogen uptake and the
surface area can be explained by variations in the adsorbate
packing consistent with the above-cited range of bulk liquid
hydrogen densities. The present conclusions on the state of the
adsorbed hydrogen are consistent, in essence, with simulation
results.45,46The usual assumption of monolayer coverage of
hydrogen near saturation does not seem to hold in the present
conditions. The stronger hydrogen affinity on the materials with
smaller pores leads to a displacement of the excess maximum to
of temperature. However, the enhancement of the solid-gas
interactions does not necessarily lead to higher packing of
molecular hydrogen, as it is sometimes supposed. Hence, (P, T)
adsorption conditions can be somehow milder on the materials
volumes and storage capacities. The determination of the ad-
sorbed phase volumes allowed for the calculation of the absolute
isotherms which were found to reach values up to 25% higher
than their excess counterparts. This leads to substantial storage
capacities, reaching 10-12 wt % at 50 K, on materials with large
porevolumes already capable ofadsorbing largeexcessamounts.
The absolute amounts were also found to correlate well with the
surface areas, showing the adsorbed phase properties are compa-
tible with surface area considerations. Owing to the rapid evolu-
tion of this class of material, it seems likely that new MOFs could
have higher storage capacities, mitigating the efficiency penalties
associated with low operating temperatures. In particular, MOFs
with a lower framework mass could be particularly interesting for
on-board applications. From the present standpoint, such feature
could compensate for potential limits to the densification of the
adsorbed phase. On the other hand, strategies to retain hydrogen
at higher temperatures are also being explored actively.47,48
Flexible MOFs also have shown some interesting features for
hydrogen storage. For instance, the adsorption/desorption curve
pore to a close-pore structure as a function of temperature.49The
material Co(BDP) also shows a significant hydrogen uptake with
of the channel pores.50Such unusual behaviors offer interesting
paths to tailor adsorbent for specific (P, T) operating conditions.
Acknowledgment. The authors are grateful to Dr. Xiang Lin
and Prof. Martin Schr€ oder of the University of Nottingham
(U.K.) and to Dr. Mircea Dinca and Prof. Jeffrey Long of
(45) Belof, J. L.; Stern, A. C.; Eddaoudi, M.; Space, B. J. Am. Chem. Soc. 2007,
(46) Belof, J. L.; Stern, A. C.; Space, B. J. Phys. Chem. C 2009, 113(21), 9316–
(47) Li, Y.; Yang, R. T. J. Am. Chem. Soc. 2006, 128, 8136–8137.
(48) Han, S. S.; Goddard, W. A., III. J. Am. Chem. Soc. 2007, 129, 8422–8423.
(49) Liu, Y.; Her, J.-H.; Dailly, A.; Ramirez-Cuesta, A. J.; Neumann, D. A.;
Brown, C. M. J. Am. Chem. Soc. 2008, 130, 11813–11818.