Strong CO2 binding in a water-stable, triazolate-bridged metal-organic framework functionalized with ethylenediamine

Article abstract of DOI:10.1021/ja903411w

(Figure Presented) Reaction of CuCl₂·2H₂O with 1,3,5-tris(1H-1,2,3-triazol-5-yl)benzene (H₃BTTri) in DMF at 100°C generates the metal-organic framework H₃[(Cu ₄Cl)₃(BTTri)₈(DMF)<

Full text of DOI:10.1021/ja903411w

Published on Web 06/08/2009
Strong CO2 Binding in a Water-Stable, Triazolate-Bridged Metal-Organic
Framework Functionalized with Ethylenediamine
Aude Demessence, Deanna M. D’Alessandro, Maw Lin Foo, and Jeffrey R. Long*
Department of Chemistry, UniVersity of California, Berkeley, California 94720
Received April 27, 2009; E-mail: jrlong@berkeley.edu
The development of more efficient processes for capturing CO2
from power plant flue streams is critical for the reduction of
greenhouse gas emissions implicated in global warming. A typical
flue gas has an overall pressure of ca. 1 bar and contains mainly
N2, H2O, and CO2 in a ca. 6.5:1:1 ratio by weight.1 Conventional
CO2 capture processes involving the chemisorption of CO2 by
alkylamine-containing liquids present several disadvantages, includ-
ing the considerable heat required to regenerate the liquid and the
necessary use of inhibitors for corrosion control. Attempts to address
these limitations have focused on the adsorption of CO2 in porous
solids such as zeolites2 and amine-modified silicas.3 In the latter,
surface modification facilitates the selective adsorption of CO2
(typically at 40-60 °C) via the formation of carbamate or
bicarbonate species and permits the stripping of CO2 at temperatures
of 100-120 °C. Owing to their exceptionally high surface areas
and chemical tunability, much recent attention has turned toward
Figure 1. A portion of the structure of the sodalite-type framework of
the use of metal-organic frameworks as CO2 capture materials.4
Cu-BTTri (1) showing surface functionalization of a coordinatively unsatur-
ated CuII site with ethylenediamine, followed by attack of an amino group
In terms of capacity, it has been shown that certain frameworks
on CO2. Purple, green, gray, and blue spheres represent Cu, Cl, C, and N
are capable of storing more than 150 wt % CO2 at 298 K and 40
atoms, respectively; framework H atoms are omitted for clarity.
bar.4a,h Most of the compounds studied, however, show little uptake
in the low pressure regime of 0.1-0.2 bar relevant to capture from
protons, which are presumably bound to either the N3 atoms of
flue streams. In addition, the viability of the materials under realistic
the triazolate rings or guest water molecules.
flue stream conditions requires air and water stability, corrosion
resistance, high thermal stability, and high selectivity for CO2 over
other components in flue gas.
The incorporation of pendant alkylamine functionalities within
the pores of metal-organic frameworks offers a significant op-
portunity to produce highly efficient capture materials by virtue of
The chemical stability of 1-DMF was tested in water and acidic
media. After soaking the solid for 3 days in boiling water or 1 day
in a solution of HCl (0.001 M pH ) 3), the powder X-ray diffraction
pattern of the compound remained unchanged. Thermogravimetric
analysis (TGA) and X-ray diffraction experiments further demon-
strated the remarkable thermal stability of the framework, with no
the affinity of alkylamines for CO2. Two major strategies are
loss of framework composition or structure occurring up to at least
envisaged to achieve this goal: (i) the direct use of an amine-based
bridging ligand to generate a three-dimensional network5 and (ii)
postsynthetic approaches that covalently modify a bridging ligand4h,6
or graft an alkylamine functionality onto a coordinatively unsatur-
ated metal center.7 Herein, we utilize the latter strategy in a new
air- and water-stable metal-organic framework, H3[(Cu4Cl)3-
(BTTri)8] (1; H3BTTri ) 1,3,5-tris(1H-1,2,3-triazol-5-yl)benzene),
featuring open CuII coordination sites.
Triazolate-bridged frameworks were sought as analogues of
known tetrazolate-bridged structures8 with enhanced chemical and
thermal stability as a result of the increased metal-ligand bond
strength. Reaction of CuCl2 ·2H2O with H3BTTri in a mixture of
DMF and aqueous HCl at 100 °C afforded H3[(Cu4Cl)3(BTTri)8-
(DMF)12] · 7DMF · 76H2O (1-DMF) as a purple powder. Powder
X-ray diffraction data show the compound to be isotypic with the
cubic sodalite-type structure of H[Cu(DMF)6][(Cu4Cl)3(BTT)8-
(H2O)12]·3.5HCl·12H2O·16CH3OH (Cu-BTT),8c displaying a simi-
lar unit cell parameter of a ) 18.647(6) Å. Here, the framework
structure consists of chloride-centered [Cu4Cl]7+ squares linked via
triangular BTTri3- ligands (see Figure 1), and each CuII center has
a terminal DMF ligand directed toward the interior of the larger
pores. The anionic charge of the framework is balanced solely by
270 °C. Soaking 1-DMF in boiling water or methanol leads to
exchange of the guest solvent and coordinated DMF molecules to
give 1-H2O and 1-MeOH, respectively. The abrupt weight-loss step
in the TGA traces for 1-H2O and 1-MeOH in the range 25-40 °C
correspond to the release of guest solvent, with the former indicating
that the hydrophilic pores can take up 33 wt % water.
Prior to gas adsorption measurements, 1-MeOH was activated
by heating at 180 °C under a reduced pressure to obtain the fully
desolvated framework 1. A color change from purple to red occurred
upon removal of the coordinated solvent molecules to yield
unsaturated CuII centers. Diffuse reflectance UV-vis spectra
revealed a shift in the d-d transitions of CuII from 18 020 cm-1 in
1-MeOH to 19 070 cm-1 in 1. The sensitivity of these metal-based
transitions to the geometry change, which occurs upon conversion
of the CuII centers from six- to five-coordination, is reminiscent of
the color variations observed for other frameworks in which open
CuII sites can be generated.9
The N2 and H2 adsorption isotherms for 1 at 77 K are depicted in
Figure 2, together with the CO2 isotherm at 195 K. The data for N2
correspond to a type I isotherm characteristic of a microporous material,
with BET and Langmuir fits affording surface areas of 1770 and 1900
9
8784 J. AM. CHEM. SOC. 2009, 131, 8784–8786
10.1021/ja903411w CCC: $40.75  2009 American Chemical Society

C O M M U N I C A T I O N S
Figure 3. Adsorption isotherms for the uptake of CO2 at 298 K in 1 (black
squares) and 1-en (red circles) and for the uptake of N2 in 1 (black diamonds)
and 1-en (red triangles). Filled and open symbols represent adsorption and
desorption, respectively. The inset shows the higher uptake of CO2 for 1-en
compared with 1 at low pressures.
Figure 2. Adsorption isotherms in 1 for the uptake of N2 (blue circles)
and H2 (green triangles) at 77 K and for CO2 (black squares) at 195 K.
Filled and open symbols represent adsorption and desorption, respectively.
m2/g, respectively. The H2 adsorption data indicate fully reversible
uptake of 1.7 wt % at 1.2 bar, which is considerably less than the 2.4
wt % uptake observed for Cu-BTT under the same conditions.8c This
result is consistent with a reduced affinity of the CuII centers for H2 as
a consequence of the greater basicity of the triazolate ligands. Indeed,
the H2 adsorption isotherms for 1 at 77 and 87 K indicate an isosteric
heat of adsorption at zero coverage of just 6.1 kJ/mol, considerably
less than the 9.5 kJ/mol observed for Cu-BTT.8c The CO2 adsorption
isotherm for 1 at 195 K reveals a maximum uptake of 90 wt % at 1
bar, placing it among the highest capacity CO2 adsorbents under such
conditions.4j,k The S-shape character of the isotherm in the low pressure
region (0-0.1 bar) has been observed previously for inflexible
frameworks and can be attributed to the attractive electrostatic
interactions between CO2 molecules.10
Alkylamine functionalization of the coordinatively unsaturated
CuII sites was accomplished using a procedure analogous to that
reported recently for MIL-101.7 An excess amount (2.3 equiv per
Cu) of ethylenediamine (en) was added to a suspension of 1 in
toluene, prompting an immediate color change from red to blue.
The suspension was heated at reflux for 12 h to optimize the extent
of grafting, and the resulting solid was washed copiously with
hexane to ensure complete removal of nonappended en. The diffuse
reflectance UV-vis spectrum of the product (1-en) displays a shift
and broadening in the CuII d-d bands compared with 1. Aside from
minor changes in the peak intensities, the powder X-ray diffraction
pattern is unchanged from that of 1, indicating preservation of the
framework structure. The infrared spectrum of 1-en provides further
support for en grafting through (i) observation of en-based νNH and
however, the CO2 uptake for 1-en is less than that for 1, leading to
a significant difference in their capacities at 1 bar: 1.27 mmol/g
(5.5 wt %) for 1-en versus 3.24 mmol/g (14.3 wt %) for 1. This
behavior can be explained by the greater attraction of CO2 to the
more basic amine sites at low pressures, followed by a filling of
the available pore volume at higher pressures. While the CO2
desorption isotherms for 1 and 1-en indicate reversible sorption
processes, both exhibit a slight hysteresis which is more pronounced
in 1-en, reflecting its greater affinity for CO2. Note that the binding
of one CO2 molecule per en ligand in 1-en would lead to an uptake
of 6.8 wt %. Thus, while en grafting does lead to a steeper rise in
the CO2 adsorption isotherm at low pressures, the original pore
size of 1 appears to be too small to permit complete exposure of
the resulting terminal alkylamines.
The selectivity for adsorption of CO2 over N2 is a prerequisite
for the application of a framework as a capture material. At 298
K, the CO2/N2 selectivity for 1 is 10:1 at 0.09 bar and 21:1 at 1
bar. Owing to a reduction in the N2 uptake, these ratios improve
slightly for 1-en to 13:1 at 0.1 bar and 25:1 at 1 bar. Importantly,
despite the reduction in CO2 capacity upon en grafting, the
alkylamine functionalities endow the material with a higher CO2/
N2 selectivity over the entire pressure range measured.
The isosteric heat of CO2 adsorption, Qst, for 1 and 1-en were
calculated from the Clausius-Clapeyron equation using adsorption
data collected at 303, 308, and 313 K (see Figure 4). The
corresponding adsorption curves for N2 over the same temperature
range are provided in the Supporting Information. For 1, Qst remains
constant within error at a mean value of 21 kJ/mol as the degree
of CO2 loading is varied. Note that this is only slightly higher than
the enthalpy of liquefaction of CO2 (17 kJ/mol), indicating that
the metal-organic framework surface provides only a minor
improvement over the affinity of the gas for itself. Indeed, the Qst
value is significantly lower than has been observed for other
frameworks featuring exposed CuII centers, such as the 35 kJ/mol
in Cu3(BTC)2 (BTC3- ) 1,3,5-benzenetricarboxylate).4h It is well
established, however, that the enthalpies of adsorption for gases in
frameworks of this type are dependent on both the metal ion and
the nature of the ligand donor atoms that contribute to its electronic
structure.4j Here, the strong bonds between CuII and the triazolate
ligands reduce the Lewis acidity of the exposed metal sites relative
to those within carboxylate-based frameworks.
ν
CH stretching vibrations and (ii) a shift in the νCH bands to higher
energies in 1-en (2887 and 2947 cm-1) compared with free en (2848
and 2924 cm-1),7 as expected when en binds a Lewis acid.
Thermogravimetric analysis of 1-en showed a 12% weight loss
at 160 °C, which, coupled with the elemental analysis data, is
consistent with incorporation of 1.25 en molecules per [Cu4Cl]7+
unit. The incomplete grafting of significantly less than one en
molecule per CuII center is likely due to steric hindrance within
the pores. Indeed, N2 adsorption data collected at 77 K on a sample
of 1-en that had been activated at 100 °C reveal a much reduced
gas uptake capacity, with BET and Langmuir fits affording surface
areas of just 345 and 376 m2/g, respectively.
The CO2 and N2 adsorption isotherms for 1 and 1-en at 298 K
are compared in Figure 3. At low pressures (up to 0.06 bar), 1-en
takes up a greater amount of CO2: 0.366 mmol/g (1.6 wt %) versus
0.277 mmol/g (0.92 wt %) in 1. At pressures above 0.1 bar,
Significant increases in Qst are apparent for the adsorption of
CO2 in 1-en, particularly at very low loadings. At zero coverage,
9
J. AM. CHEM. SOC. VOL. 131, NO. 25, 2009 8785

C O M M U N I C A T I O N S
as the ability of Cu-BTTri (1) to incorporate en molecules that afford
a highly selective affinity for CO2 binding. Significantly, the en-
functionalized framework 1-en exhibits a higher uptake of CO2 at
very low pressures compared with the nongrafted material and
displays a record isosteric heat of adsorption of 90 kJ/mol. Future
efforts will attempt to increase the CO2 capture capacity by
extending this approach to frameworks containing larger pores and
to adjust the heat of adsorption by varying the amine substituents11
and investigating the effects of the presence of water.12
Acknowledgment. This research was funded by the Sustainable
Products & Solutions Program at the University of California,
Berkeley and General Motors Corporation. We thank the 1851
Royal Commission and the American-Australian Association for
Research Fellowships for support of D.M.D. and Dr. S. Horike and
Dr. H. J. Choi for helpful discussions.
Figure 4. Isosteric heat of adsorption for CO2 in 1 (black) and 1-en (red).
Supporting Information Available: Complete experimental details,
including characterization data and sorption analysis data. This material
is available free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Powell, C. E.; Qiao, G. G. J. Membr. Sci. 2006, 279, 1. (b) Figueroa,
J. D.; Fout, T.; Plasynski, S.; McIlvried, H.; Srivastava, R. D. Int. J.
Greenhouse Gas Control 2008, 2, 9.
(2) (a) Harlick, P. J. E.; Tezel, F. H. Microporous Mesoporous Mater. 2004, 76,
71. (b) Jia, W.; Murad, S. J. Chem. Phys. 2005, 122, 234708. (c) Lee, K. B.;
Beaver, M. G.; Caram, H. S.; Sircar, S. Ind. Eng. Chem. Res. 2008, 47, 8048.
(d) Xu, X.; Zhao, X.; Sun, L.; Liu, X. J. Nat. Gas Chem. 2008, 17, 391.
(3) (a) Franchi, R. S.; Harlick, P. J. E.; Sayari, A. Ind. Eng. Chem. Res. 2005,
44, 8007. (b) Hicks, J. C.; Drese, J. H.; Fauth, D. J.; Gray, M. L.; Qi, G.;
Jones, C. W. J. Am. Chem. Soc. 2008, 130, 2902. (c) Zelenak, V.; Halamova,
D.; Gaberova, L.; Bloch, E.; Llewellyn, P. Microporous Mesoporous Mater.
2008, 116, 358.
(4) (a) Millward, A. R.; Yaghi, O. M. J. Am. Chem. Soc. 2005, 127, 17998.
(b) Hayashi, H.; Coˆte, A. P.; Furukawa, H.; O’Keeffe, M.; Yaghi, O. M.
Nat. Mater. 2007, 6, 501. (c) Ma, S.; Sun, D.; Wang, X.-S.; Zhou, H.-C.
Angew. Chem., Int. Ed. 2007, 46, 2458. (d) Chen, B.; Ma, S.; Hurtado,
E. J.; Lobkovsky, E. B.; Zhou, H.-C. Inorg. Chem. 2007, 46, 8490. (e)
Dietzel, P. D. C.; Johnsen, R. E.; Fjellvåg, H.; Bordiga, S.; Groppo, E.;
Chavan, S.; Blom, R. Chem. Commun. 2008, 5125. (f) Bae, Y.-S.; Farha,
O. K.; Spokoyny, A. M.; Mirkin, C. A.; Hupp, J. T.; Snurr, R. Q. Chem.
Commun. 2008, 4135. (g) Bastin, L.; Ba´rcia., P. S.; Hurtado, E. J.; Silva,
J. A. C.; Rodrigues, A. E.; Chen, B. J. Phys. Chem. C 2008, 112, 1575.
(h) Llewellyn, P. L.; Bourrelly, S.; Serre, C.; Vimont, A.; Daturi, M.;
Hamon, L.; De Weireld, G.; Chang, J.-S.; Hong, D.-Y.; Hwang, Y. K.;
Jhung, S. H.; Fe´rey, G. Langmuir 2008, 24, 7245. (i) Bae, Y.-S.; Mulfort,
K. L.; Frost, Houston; Ryan, P.; Punnathanam, S.; Broadbelt, L. J.; Hupp,
J. T.; Snurr, R. Q. Langmuir 2008, 24, 8592. (j) Caskey, S. R.; Wong-Foy,
A. G.; Matzger, A. J. J. Am. Chem. Soc. 2008, 130, 10870. (k) Park, H. J.;
Suh, M. P. Chem.sEur. J. 2008, 14, 8812. (l) Thallapally, P. K.; Tian, J.;
Kishan, M. R.; Fernandez, C. A.; Dalgarno, S. J.; McGrail, P. B.; Warren,
J. E.; Atwood, J. L. J. Am. Chem. Soc. 2008, 130, 16842. (m) Bae, Y.-S.;
Farha, O. K.; Hupp, J. T.; Snurr, R. Q. J. Mater. Chem. 2009, 19, 2131.
(n) Keskin, S.; Sholl, D. S. Ind. Eng. Chem. Res. 2009, 48, 914. (o)
Banerjee, R.; Furukawa, H.; Britt, D.; Knobler, C.; O’Keeffe, M.; Yaghi,
O. M. J. Am. Chem. Soc. 2009, 131, 3875.
(5) (a) Arstad, B.; Fjellvåg, H.; Kongshaug, K. O.; Swang, O.; Blom, R.
Adsorption 2008, 14, 755. (b) Costa, J. S.; Gamez, P.; Black, C. A.;
Roubeau, O.; Teat, S. T.; Reedijk, J. Eur. J. Inorg. Chem. 2008, 1551.
(6) (a) Wang, Z.; Cohen, S. M. J. Am. Chem. Soc. 2007, 129, 12368. (b) Goto,
Y.; Sato, H.; Shinkai, S.; Sada, K. J. Am. Chem. Soc. 2008, 130, 14354.
(7) Hwang, Y. K.; Hong, D.-Y.; Chang, J.-S.; Jhung, S. H.; Seo, Y.-K.; Kim,
J.; Vimont, V.; Daturi, M.; Serre, C.; Fe´rey, G. Angew. Chem., Int. Ed.
2008, 47, 4144.
(8) (a) Dinca, M.; Yu, A. F.; Long, J. R. J. Am. Chem. Soc. 2006, 128, 8904.
(b) Dinca, M.; Dailly, A.; Liu, Y.; Brown, C. M.; Neumann, D. A.; Long,
J. R. J. Am. Chem. Soc. 2006, 128, 16876. (c) Dinca, M.; Han, W. S.; Liu,
Y.; Dailly, A.; Brown, C. M.; Long, J. R. Angew. Chem., Int. Ed. 2007,
46, 1419.
Figure 5. Gas cycling experiment for 1-en at 30 °C, employing a flow of
15% CO2 in N2 followed by a flow of pure N2.
a maximum value of 90 kJ/mol occurs, indicating a strong and
selective interaction of CO2 with the alkylamine functionalities. This
value compares well with previously reported data obtained for
alkylamine-functionalized silicas (100 kJ/mol).10a To our knowl-
edge, this is the highest binding energy reported to date for CO2
adsorption in a metal-organic framework, surpassing the previous
record of 63 kJ/mol in MIL-1004h and the values for amine-
functionalized frameworks in which the NH2 groups are directly
bound to aromatic ligands (50 kJ/mol).6 The high binding energy
for CO2 in 1-en at low coverage is indicative of a chemisorption
interaction, akin to the chemisorption mechanisms that are operative
for CO2 separation in industry using alkylamine solvents. As the
available alkylamine groups bind CO2 with increasing pressure, Qst
decreases, approaching the values observed for 1 as CO2 is adsorbed
onto less reactive surface sites.
The reversibility of the CO2 adsorption process in 1-en during
gas cycling at 30 °C was measured using a thermogravimetric
analysis apparatus, with use of a 15% CO2 in N2 mixture to simulate
the major components of flue gas (see Figure 5). A 0.75 wt %
weight change was observed over repeated cycles, indicating that
the material is able to withstand cyclic exposure to the mixed gas
stream. Significantly, the CO2 adsorption ability of 1-en is
maintained over repeated cycling, and the material can be regener-
ated by switching the gas stream to N2. This may render the material
suitable for use as an adsorbent in a pressure swing absorption type
process for CO2 capture.
(9) Prestipino, C.; Regli, L.; Vitillo, J. G.; Bonino, F.; Damin, A.; Lamberti,
C.; Zecchina, A.; Solari, P. L.; Kongshaug, K. O.; Bordiga, S. Chem. Mater.
2006, 18, 1337.
(10) (a) Kno¨fel, C.; Descarpentries, J.; Benzaouia, A.; Zelea´k, V.; Mornet, S.;
Llewellyn, P. L.; Hornebecq, V. Microporous Mesoporous Mater. 2007,
99, 79. (b) Walton, K. S.; Millward, A. R.; Dubbeldam, D.; Frost, H.; Low,
J. J.; Yaghi, O. M.; Snurr, R. Q. J. Am. Chem. Soc. 2008, 130, 406.
(11) Dibenedetto, A.; Aresta, M.; Fragale, C.; Narracci, M. Green Chem. 2002, 4,
439.
(12) Yazaydin, A. O.; Benin, A. I.; Faheem, S. A.; Jakubczak, P.; Low, J. J.;
Willis, R. R.; Snurr, R. Q. Chem. Mater. 2009, 21, 1425.
The foregoing results demonstrate the use of BTTri3- in
generating an air- and water-stable analogue of Cu-BTT, as well
JA903411W
9
8786 J. AM. CHEM. SOC. VOL. 131, NO. 25, 2009