We report a record-high SO2 adsorption capacity of 12.3 mmol g-1 in a robust porous material, MFM-601, at 298 K and 1.0 bar. SO2 adsorption in MFM-601 is fully reversible and highly selective over CO2 and N2. The binding domains for adsorbed SO2 and CO2 molecules in MFM-601 have been determined by in situ synchrotron X-ray diffraction experiments, giving insights at the molecular level to the basis of the observed high selectivity.
We report a record-high SO2 adsorption capacity of 12.3 mmol g-1 in a robust porous material, MFM-601, at 298 K and 1.0 bar. SO2 adsorption in MFM-601 is fully reversible and highly selective over CO2 and N2. The binding domains for adsorbed SO2 and CO2 molecules in MFM-601 have been determined by in situ synchrotron X-ray diffraction experiments, giving insights at the molecular level to the basis of the observed high selectivity.
The International Energy Outlook
2017 report produced by the U.S. Energy Information Administration
predicts that fossil fuels will account for ∼77% of global
energy production in 2050, with coal comprising a major component
of the overall energy supply.[1] The burning
of bituminous, sub-bituminous and lignite coals produces SO2 at a concentration of between 500 and 3000 ppm, around 95% of which
is removed via limestone-scrubbing or the wet-sulfuric-acid processes.[2] Although these processes are effective, as much
as 400 ppm of SO2 can remain in the exhaust gas, which,
when vented to the atmosphere, poses significant health risks and
environmental impact.[3] Small amounts of
SO2 in flue gas can also react with organo-amines used
for CO2-scrubbing causing permanent loss of activity.[4] Physisorption of SO2 by porous materials
such as zeolites,[5] mesoporous silica,[6] ionic microgels[7] and
activated carbons[8] using supramolecular
host–guest interactions is a promising approach that can give
high selectivity, reversibility and low energy penalty for system
regeneration. However, these materials generally suffer from low adsorption
capacities and exposure to SO2 often leads to irreversible
structural degradation.Metal–organic frameworks (MOFs)
are emerging porous materials
constructed from metal clusters with organic linkers,[9] and their tuneability makes them interesting candidates
for many applications.[10] Although much
work has focused on the study of gas separations in MOFs, very little
effort has been devoted to the sequestration of SO2,[11] since it often leads to severe structural degradation
of the material and/or irreversible uptake. Here, we report the exceptional
adsorption and reversible binding of SO2 in two Zr-based
MOFs, MFM-600 and MFM-601. MFM-601 is obtained by postsynthetic removal
of a linker in an ordered, predictable pattern from MFM-600. This
is accompanied by significant enhancement of gas adsorption in MFM-601,
notably simultaneous increases in SO2 uptake (146%) and
selectivity over CO2 and N2. Significantly,
MFM-601 shows a SO2 adsorption capacity of 12.3 mmol g–1 at 298 K and 1.0 bar, representing the highest value
observed to date in porous materials under the same conditions. In
addition, the binding domains for adsorbed CO2 and SO2 in MFM-610 have been determined by in situ synchrotron X-ray diffraction.4,4′,4″,4′′′-(1,4-Phenylenebis(pyridine-4,2,6-triyl))tetrabenzoic acid (H4L, Scheme ) was prepared via a three-step synthesis (Scheme S1). [Zn6(μ3-O)4(μ3-OH)4(OH)4(L)2(H4L)0.7]
(MFM-600) was synthesized by heating a mixture of ZrCl4, H4L and benzoic acid as modulator in DMF at 120 °C
for 24 h and was isolated as plate-shaped colorless single crystals.
The single crystal X-ray structure shows that MFM-600 contains [Zr6(μ3-O)4(μ3-OH)4(OH)4]8+ clusters, first observed in
UiO-66.[12] In this cluster the six ZrIV ions form an octahedron with the 8 faces each capped by
a μ3-O or μ3-OH. Eight of the edges
of the octahedron are bridged by L4– linkers via
bidentate carboxylate groups with the remaining four equatorial edges
each bridged by a H4L linker (occ = 0.35) that is monodentate
to each equatorial Zr(IV) center and bound through the C=O
moiety of the carboxylate groups (Figure S4). The coordination sphere at the equatorial Zr(IV) centers is completed
by a terminal OH– ligand.
Scheme 1
Synthesis of MFM-601
from 4,4′,4″,4′′′-(1,4-Phenylenebis(pyridine-4,2,6-triyl))tetrabenzoic
Acid (H4L)
MFM-600 was treated with an 8 M solution of HCl, which
led to a
structural phase transition to MFM-601, [Zn6(μ3-O)4(μ3-OH)4(OH)4(H2O)4(L)2], the structure
of which was determined by high resolution synchrotron powder X-ray
diffraction with the {Zr6} node assigned to the MIX-staggered
proton topology.[13] In MFM-601, the H4L monodentate linkers from the equatorial positions of the
{Zr6} cluster have been removed (Figure ) and replaced with terminal H2O molecules. As a result, the alternating linkage between the [Zr6(μ3-O)4(μ3-OH)4(OH)4(H2O)4]8+ cluster and the L4– linker yields two types of
channels of ∼9 Å (cylinder-shaped) and ∼12 Å
(waterdrop-shaped) diameter running through the structure of MFM-601
along the c-axis. This leads to an increase in the
BET surface area from 2281 m2g–1 for
MFM-600 to 3644 m2g–1 for MFM-601, a
rise of ∼60%. We note that MFM-601 has recently been reported
(as BUT-15) as an adsorbent for FeIII ions in water.[14] Acid treatment has been previously reported
as an efficient approach to remove free ligands such as surplus benzoic
acid modulator in Zr-based MOFs[15] and as
a method of improving the N2 uptake in UiO-66.[16]
Figure 1
Views of (a) disordered H4L linker present
in MFM-600;
(b) “fully bound” L4− linker present
in MFM-600 and MFM-601; (c) {Zr6} node present in MFM-600
and MFM-601; (d) structure of MFM-600; (e) structure of MFM-601.
Views of (a) disorderedH4L linker present
in MFM-600;
(b) “fully bound” L4− linker present
in MFM-600 and MFM-601; (c) {Zr6} node present in MFM-600
and MFM-601; (d) structure of MFM-600; (e) structure of MFM-601.Sorption isotherms of SO2, CO2 and N2 were measured on desolvated samples.
As expected, the higher
BET surface area of MFM-601 allows a greater total uptake of all three
gases. MFM-600 displays a two-step Type-IV isotherm for SO2 with an excess uptake of 5.0 mmol g–1 at 298 K
and 1.0 bar. Interestingly, in comparison, MFM-601 exhibits a type-I
isotherm for SO2 leading to an exceptionally high uptake
of 12.3 mmol g–1 (146% enhancement) at 298 K and
1.0 bar, representing the highest value observed in porous solids
and notably higher than a range of best-behaving SO2 sorbents
(Table S1). Importantly, no loss of adsorption
capacity of SO2 was observed in MFM-601 over five cycles
of adsorption/desorption (Figure S17b).
The SO2 uptake in MFM-601 is further increased to 16.9
mmol g–1 at 273 K and 1.0 bar (Figure S17a). MFM-601 shows fully reversible adsorption of
SO2 at both 273 and 298 K, allowing the total evacuation
of the framework upon desorption under pressure-swing conditions (Figure ) with no structural
phase change or framework collapse, as evidenced by the postsorption
PXRD patterns (Figure S8). This is in contrast
to the current leading SO2 sorbent, MFM-202a, which undergoes
irreversible phase change to MFM-202b upon SO2 adsorption,[17] and Ni(bdc)(ted)0.5 and MOF-74(Mg),
which release up to 40% of the captured SO2 upon desorption.[18] The desorption of SO2 from SIFSIX-1-Cu
has not been reported.[20a] Interestingly,
the high uptake of SO2 in MFM-601 does not correlate with
equivalent increases in uptake of CO2 or N2.
The uptake of CO2 in MFM-601 exhibits a type V isotherm
with slow initial uptake at low pressures before the uptake increases
more rapidly at intermediate pressures reaching a plateau at ∼24
mmol g–1 at 30 bar and 298 K (Figure S15). Adsorption of N2 in MFM-601 follows
a type-I isotherm with a total uptake of ∼12 mmol g–1 at 80 bar 298 K. Strikingly, despite the high uptake of SO2 at 298 K and 1.0 bar, under the same conditions the CO2 uptake in MFM-601 compares poorly to other MOFs in the literature,[19] with a total CO2 uptake of only 1.3
mmol g–1 and a negligible N2 uptake.
The isosteric heat of adsorption (Q) for SO2, CO2 and N2 in MFM-600
and MFM-601 are compared in Figure c. The same trend is followed for both MOFs with N2 having the lowest Qst, followed
by CO2 and then SO2 with the highest Qst. Interestingly the Qst values for SO2 in MFM-600 are generally higher
than in MFM-601, consistent with the smaller cages of MFM-600 affording
greater overlap potentials between SO2 molecules and the
pore interior. In MFM-600 at low loadings SO2 interacts
with binding sites at the MOF surface. As these sites become less
available Qst reduces, but once SO2 fills the pore, SO2–SO2dipole
interactions lead to an increase in Q.
Figure 2
(a) SO2 sorption isotherms in MFM-600 and MFM-601.
(b)
Comparison of SO2, CO2 and N2 isotherms
at 298 K in MFM-601. (c) Qst values for
SO2, CO2 and N2 in MFM-600 and MFM-601.
(d) IAST selectivities of SO2/CO2 and SO2/N2 at 298 K in MFM-600 and MFM-601 as a function
of substrate ratios.
(a) SO2 sorption isotherms in MFM-600 and MFM-601.
(b)
Comparison of SO2, CO2 and N2 isotherms
at 298 K in MFM-601. (c) Qst values for
SO2, CO2 and N2 in MFM-600 and MFM-601.
(d) IAST selectivities of SO2/CO2 and SO2/N2 at 298 K in MFM-600 and MFM-601 as a function
of substrate ratios.Analysis of adsorption selectivities at 298 K using ideal
adsorbed
solution theory (IAST) shows that the transformation of MFM-600 to MFM-601
is accompanied by an increase in selectivity
of SO2 over both CO2 and N2 (Figure ). The selectivity
for equimolar mixtures of SO2/CO2 and SO2/N2 in MFM-601 at 298 K are 32 and 255, respectively,
comparable with the leading MOFs in the literature.[20] The capability of selective separation of SO2 from gas mixtures using fixed-bed packed with MFM-601 has been confirmed
by dynamic breakthrough experiments (Figure and SI).
Figure 3
Breakthrough
curves of 2500 ppm of SO2 diluted in (a)
1:1 He:N2 and (b) 1:1 He:CO2 through a fixed-bed
packed with MFM-601 at 298 K and 1 bar.
Breakthrough
curves of 2500 ppm of SO2 diluted in (a)
1:1 He:N2 and (b) 1:1 He:CO2 through a fixed-bed
packed with MFM-601 at 298 K and 1 bar.In order to examine the origins of the preferential adsorption
of SO2 over CO2 in MFM-601, in situ synchrotron X-ray powder diffraction studies were carried out (Figures and 5). Rietveld refinement of PXRD data for CO2-loaded
MFM-601 at 200 K revealed six binding sites within the pores, giving
a formula of [Zr6(μ3-O)4(μ3-OH)4(OH)4(L)2][CO2]5.52. The primary binding site of CO2I in MFM-601 (occupancy = 0.50) overlaps with the terminal
oxygen of the zirconium cluster due to the removal of four H2O molecules from each {Zr6} cluster upon activation, leaving
a terminal hydroxyl and an open zirconium site on each of the four
equatorial edges of the [Zr6(μ3-O)4(μ3-OH)4(OH)4]8+ cluster, thus accounting for the 0.5 occupancy of the terminal
oxygen and CO2I. This positioning of CO2I puts it within close proximity to the open zirconium
site [OCO2···Zr = 2.40(10) Å] as well
as being within binding distance of the terminal hydroxyl [CCO2···O= 2.00(24) Å] and the μ3-O of the Zr6O8 cluster [CCO2···μ3-O= 2.70(11) Å], similar to that observed in UiO-66.[21] CO2II (occupancy = 0.38)
and CO2VI (occupancy = 0.16) occupy similar
environments either side of the zirconium cluster in that they are
both in close proximity to the terminal hydroxyl [(OCO2···O = 2.58(23) Å) and [OCO2···O
= 3.31(44) Å], respectively, as well as the μ3-O [(OCO2···μ3-O = 2.20(10)
Å) and [OCO2···μ3-O
= 3.37(14) Å], respectively. The remaining sites, CO2III (occupancy = 0.33), CO2IV (occupancy
= 0.30) and CO2V (occupancy = 0.17), are not
within binding distance of the {Zr6} cluster but are closer
to the L4– linker. CO2III is
near the phenyl ring of the linker [OCO2···Ph
= 2.79(8) Å] whereas CO2IV and CO2V are close to the pyridyl rings [OCO2···NPyr = 4.38(5) Å)] and [OCO2···HPyr = 3.91(25) Å], respectively. All CO2 sites
are within intermolecular binding distance of one another with C···O
distances between 2.18(14) and 4.38(12) Å. However, no ordered
CO2 molecule was present at the center of the large pore
of MFM-601, thus confirming that CO2 positioning is dominated
by host–guest interactions.
Figure 4
View of SO2 (top) and CO2 (bottom) in MFM-601;
the size of the colored balls represent the occupancy at each site.
Positions refined by in situ synchrotron PXRD.
Figure 5
Binding sites of SO2 (top) and CO2 (bottom)
in MFM-601.
View of SO2 (top) and CO2 (bottom) in MFM-601;
the size of the colored balls represent the occupancy at each site.
Positions refined by in situ synchrotron PXRD.Binding sites of SO2 (top) and CO2 (bottom)
in MFM-601.The crystal structure
of SO2-loaded MFM-601 at 298 K
also shows six binding sites within the pore giving a formula of [Zr6(μ3-O)4(μ3-OH)4(OH)4(H2O)4(L)2][SO2]10.71. The three SO2 positions
with the highest occupancies, SO2I (occupancy
= 1.00), SO2II (occupancy = 0.87) and SO2III (occupancy = 0.63) form an intermolecular dipole
bonding network with S···O bond distances between 3.59(7)
and 3.24(6) Å, thus exhibiting efficient packing as exhibited
by the similarity to the intermolecular distances observed in solid
SO2.[22] The highest occupied
SO2 position (SO2I) is located near
the terminal hydroxyl of the {Zr6} cluster [OSO2···O = 2.53(6) Å] with a similar SO2-hydroxyl interaction to that observed in MFM-300(In).[20b] SO2II sits away from
the zirconium cluster, close to the pyridyl group of the L4– linker [OSO2···Pyr = 3.06(6) Å].Conversely, SO2III is not in close proximity
to the pore wall and is only within interaction distance of other
SO2 molecules. SO2IV (occupancy =
0.46) occupies a position between two phenyl rings of two opposite
linkers related by a mirror plane. It is in close proximity to both
the terminal hydroxyl [OSO2···O = 1.99(10)
Å] and the μ3-O [SSO2···μ3-O = 3.42(5) Å] of the [Zr6(μ3-O)4(μ3-OH)4(OH)4(H2O)4]8+ cluster. As for SO2IV, SO2V (occupancy = 0.31)
resides in between the phenyl rings of two opposite L4– linkers and also is in close proximity to the terminal hydroxyl
[OSO2···O = 3.48(15) Å] and the μ3-O [OSO2···μ3-O
= 3.06(15) Å] of the {Zr6} cluster. The least occupied
SO2 site (SO2VI; occupancy = 0.31)
lies near to a pyridyl ring of the organic linker [OSO2···HPyr = 2.75(19)Å] and also exhibits
the potential for dipole–dipole interaction with SO2V [OVI···SV = 3.53(18)Å]
(Figures and 5).In conclusion, exceptional adsorption of
SO2 and selective
sorption of SO2 over CO2 and N2 have
been demonstrated in a robust Zr-based MOF, MFM-601. By locating the
positions of CO2 and SO2 in MFM-601, key understanding
of the observed uptake and selectivity has been gained. It has been
shown that the dipole moment of SO2 can be utilized by
MFM-601 not only to provide stable binding within the pores but also
to drive intermolecular interactions between SO2 molecules.
Authors: Omar M Yaghi; Michael O'Keeffe; Nathan W Ockwig; Hee K Chae; Mohamed Eddaoudi; Jaheon Kim Journal: Nature Date: 2003-06-12 Impact factor: 49.962
Authors: Nora Planas; Joseph E Mondloch; Samat Tussupbayev; Joshua Borycz; Laura Gagliardi; Joseph T Hupp; Omar K Farha; Christopher J Cramer Journal: J Phys Chem Lett Date: 2014-10-15 Impact factor: 6.475
Authors: Mathew Savage; Yongqiang Cheng; Timothy L Easun; Jennifer E Eyley; Stephen P Argent; Mark R Warren; William Lewis; Claire Murray; Chiu C Tang; Mark D Frogley; Gianfelice Cinque; Junliang Sun; Svemir Rudić; Richard T Murden; Michael J Benham; Andrew N Fitch; Alexander J Blake; Anibal J Ramirez-Cuesta; Sihai Yang; Martin Schröder Journal: Adv Mater Date: 2016-08-16 Impact factor: 30.849
Authors: Sihai Yang; Leifeng Liu; Junliang Sun; K Mark Thomas; Andrew J Davies; Michael W George; Alexander J Blake; Adrian H Hill; Andrew N Fitch; Chiu C Tang; Martin Schröder Journal: J Am Chem Soc Date: 2013-03-25 Impact factor: 15.419
Authors: Timothy C Wang; Nicolaas A Vermeulen; In Soo Kim; Alex B F Martinson; J Fraser Stoddart; Joseph T Hupp; Omar K Farha Journal: Nat Protoc Date: 2015-12-17 Impact factor: 13.491
Authors: I Matito-Martos; A Martin-Calvo; J J Gutiérrez-Sevillano; M Haranczyk; M Doblare; J B Parra; C O Ania; S Calero Journal: Phys Chem Chem Phys Date: 2014-10-07 Impact factor: 3.676
Authors: Weiyao Li; Jiangnan Li; Thien D Duong; Sergei A Sapchenko; Xue Han; Jack D Humby; George F S Whitehead; Iñigo J Victórica-Yrezábal; Ivan da Silva; Pascal Manuel; Mark D Frogley; Gianfelice Cinque; Martin Schröder; Sihai Yang Journal: J Am Chem Soc Date: 2022-07-18 Impact factor: 16.383
Authors: Jiangnan Li; Gemma L Smith; Yinlin Chen; Yujie Ma; Meredydd Kippax-Jones; Mengtian Fan; Wanpeng Lu; Mark D Frogley; Gianfelice Cinque; Sarah J Day; Stephen P Thompson; Yongqiang Cheng; Luke L Daemen; Anibal J Ramirez-Cuesta; Martin Schröder; Sihai Yang Journal: Angew Chem Int Ed Engl Date: 2022-07-25 Impact factor: 16.823
Authors: Erik Svensson Grape; J Gabriel Flores; Tania Hidalgo; Eva Martínez-Ahumada; Aída Gutiérrez-Alejandre; Audrey Hautier; Daryl R Williams; Michael O'Keeffe; Lars Öhrström; Tom Willhammar; Patricia Horcajada; Ilich A Ibarra; A Ken Inge Journal: J Am Chem Soc Date: 2020-09-17 Impact factor: 15.419