Journal of the American Chemical Society
Article
distilled water and filtered again. To the filtrate, neat HCl was slowly
added until a pH between 1 and 2 was reached. The resulting crude
product was collected via filtration. Recrystallization using 50 mL of
acetone and 50 mL of water per gram of crude material afforded 0.68 g
(40%) of pure product as a white powder. Anal. Calcd for C14H10O6:
C, 61.32; H, 3.68. Found: C, 61.34; H, 3.60. IR (KBr) 1659 (vs), 1612
(s), 1481 (s), 1450 (vs), 1319 (s), 1290 (s), 1240 (s), 1099 (w), 1047
(w), 894 (w), 825 (m), 792 (m), 723 (w), 686 (m), 573 (w), 530 (w)
cm−1. 1H NMR (300 MHz, dmso-d6): δ = 10.75−11.95 (br, 2H), 7.93
(s, 2H), 7.74 (d, 2H, J = 11.2 Hz), 7.00 (d, 2H, J = 8.6 Hz).
Zn2(dobpdc)(DEF)2·DEF·H2O (DEF-1). To a 2-mL Pyrex tube,
H4dobpdc (4.0 mg, 0.015 mmol), ZnBr2·2H2O (8.9 mg, 0.034 mmol),
and 0.5 mL of mixed solvent (1:1 DEF:EtOH; DEF = N,N-
diethylformamide) were added. The tube was sealed and placed in a
pre-heated oven at 100 °C. After 72 h, needle-shaped, colorless crystals
had formed. The crystals were isolated by filtration and washed with
hot DEF to afford 3.7 mg (35%) of product. Anal. Calcd for
C29H41N3O10Zn2: C, 47.84; H, 5.64; N, 5.79. Found: C, 48.21; H,
5.72; N, 5.82. IR (KBr): 1655 (vs), 1610 (s), 1544 (s) 1462 (s), 1412
(vs), 1286 (s), 1234 (s), 1149 (m), 1103 (m), 1047 (w), 881 (m), 825
(m), 760 (w), 690 (m), 586 (m), 505 (w) cm−1.
Mg2(dobpdc)(DEF)2·DEF1.5·H2O (DEF-2). Into a 10-mL Pyrex cell,
H4dobpdc (24 mg, 0.088 mmol), MgBr2·6H2O (60 mg, 0.21 mmol),
and 3 mL of solvent (1:1 DEF:EtOH) were loaded and sealed with a
PTFE cap. The mixture was irradiated in a microwave reactor (CEM
Discover) for 30 min at 120 °C. After 30 min, the solution was cooled,
and the resulting solid was collected via filtration and washed with hot
DEF. The solid was dried under vacuum to yield 57.5 mg (95%) of
product as a white powder. Anal. Calcd for C31.5H46.5Mg2N3.5O10.5: C,
54.77; H, 6.78; N, 7.10. Found: C, 54.85; H, 7.07; N, 6.86. IR (KBr):
1661 (vs), 1612 (s), 1570 (s), 1468 (vs), 1419 (s), 1298 (s), 1242 (s),
1149 (m), 1111 (m), 1047 (w), 945 (w), 885 (m), 842 (m), 825 (m),
725 (w), 692 (m), 660 (w), 590 (m), 501 (w), 447 (w) cm−1. Heating
at 420 °C for 65 min in vacuo yielded the fully activated adsorbent
Mg2(dobpdc) (2).
diffractometer using Cu Kα radiation (λ = 1.5406 Å). The unit cell
dimensions of DEF-2 and mmen-2 were determined by performing a
full-pattern decomposition using the Le Bail method, as implemented
in TOPAS-Academic.35 Owing to the isomorphism with Zn2(dobpdc),
the trigonal space group P3221 was used for the refinements. Crystal
data for DEF-2: a = 21.761(2) Å, c = 6.9721(7) Å, V = 2859.1(5) Å3
(Rwp = 0.093, Rp = 0.067). Crystal data for mmen-2: a = 21.500(2) Å, c
= 6.8275(9) Å, V = 2733.2(6) Å3 (Rwp = 0.042, Rp = 0.033).
Gas Adsorption Measurements. Gas adsorption data for
pressures in the range 0−1.1 bar were obtained by volumetric
methods using a Micromeritics ASAP2020 instrument. All gases were
99.998% purity or higher. Isotherms at 77 K were measured in liquid
nitrogen baths. Isotherms at 25, 35, 45, 50, and 75 °C were measured
using liquid circulators to maintain a constant temperature. Isotherms
at 100 and 120 °C were measured using a heated sand bath controlled
by a programmable temperature controller. BET surface areas were
calculated from N2 adsorption at 77 K. DFT pore size distributions
and pore sizes were calculated from N2 adsorption at 77 K with the
Micromeritics DFT Plus Models Kit (Ver. 2.02) software suite with
cylinder pore geometries for an oxide surface. The compound mmen-2
was regenerated at 100 °C under dynamic vacuum for 4 h after
measurement of each isotherm.
Isosteric Heats of Adsorption Calculations. A dual-site
Langmuir−Freundlich equation (eq 1) was employed to model the
CO2 adsorption at 25, 50, and 75 °C for mmen-2 in the region before
the step in the isotherms.
qsat,AbApα
qsat,BbBpα
A
B
q =
+
1 + bApα
1 + bBpα
A
B
(1)
Here, q is the amount of CO2 adsorbed (mmol/g), p is the pressure
(bar), qsat is the saturation capacity (mmol/g), b is the Langmuir−
Freundlich parameter (bar−α), and α is the Langmuir−Freundlich
exponent (dimensionless) for two adsorption sites A and B. In order
to model the CO2 adsorption in the region after the step, a modified
Langmuir−Freundlich equation (eq 2) was employed.
Mg2(dobpdc)(mmen)1.6(H2O)0.4 (mmen-Mg2(dobpdc) or mmen-
2). A sample of fully activated 2 (77 mg, 0.24 mmol) was immersed in
anhydrous hexane, and 20 equiv of N,N′-dimethylethylenediamine
(mmen, 0.53 mL, 4.8 mmol) was added. The suspension was stirred
for one day, filtered, and rinsed copiously with hexanes. The solid was
then evacuated of residual solvents at 100 °C for 24 h to afford 87 mg
(77%) of product as a gray-white powder. Anal. Calcd for
C20.4H26Mg2N3.2O9.6: C, 52.46; H, 5.62; N, 9.60. Found: C, 52.15,
H, 5.41; N, 9.52. IR (ATR, neat): 3320 (w), 2952 (w), 2910 (w), 2862
(w), 2806 (w), 1616 (s), 1575 (s), 1538 (w), 1468 (vs), 1421 (vs),
1295 (m), 1244 (s), 1152 (m), 1104 (m), 1053 (m), 1000 (w), 887
(m), 844 (m), 828 (m), 727 (w), 692 (s), 618 (m), 589 (s) cm−1.
X-ray Structure Determination. A crystal of DEF-1 was
mounted on a cryoloop under a cooling stream of dinitrogen.
Diffraction data were collected with synchrotron radiation using a 6B
MX-I ADSC Quantum-210 detector with a silicon (111) double-
crystal monochromator at the Pohang Accelerator Laboratory. The
ADSC Quantum-210 ADX program (Ver. 1.92) was used for data
collection and HKL2000 (Ver. 0.98.699) was used for cell refinement,
data reduction, and absorption corrections. The structure was solved
by direct methods and refined by full-matrix least-squares analysis
using anisotropic thermal parameters for non-hydrogen atoms with the
SHELXTL program.33 The C2 and C5 atoms were isotropically
refined due to poor thermal behavior. Guest molecules in the pores
were highly disordered and unable to be modeled. To account for this
electron density, the program SQUEEZE34 was employed. All
hydrogen atoms were calculated at idealized positions and refined
using a riding model. Crystal data for DEF-1: empirical formula =
C12H14NO4Zn, Mr = 301.61, T = 100(2) K, space group = P3221, a =
21.698(3) Å, c = 6.8690(14) Å, α = 90°, β = 90°, γ = 120°, V =
2800.7(8) Å3, Z = 6, Dcalc = 1.073 g/cm3, μ = 1.319 mm−1, 13927
reflections collected, 2941 unique (Rint = 0.0418), R1 = 0.0466, wR2 =
0.1204 (I > 2σ(I)).
qsat,AbA(p − p*)α
qsat,BbB(p − p*)α
A
B
q =
+
1 + bA(p − p*)αA
1 + bB(p − p*)α
B
qsat,CbC(p − p*)α
C
+
1 + bC(p − p*)α
C
(2)
Here, adsorption is considered at three sites, A, B, and C, and the extra
parameter p* is used to account for the pressure at which the step in
the isotherm occurs and the strongest adsorption sites are first
populated. After carefully refining the parameters in eqs 1 and 2,
excellent agreement was achieved between the experimental isotherm
data and the corresponding isotherm fits (see Figures S10−S13).
Using the appropriate isotherm fits, Mathematica software was used to
solve for the exact pressures, p, corresponding to constant amounts of
CO2 adsorbed, q, at 25, 50, and 75 °C. The Clausius−Clapeyron
equation (eq 3) was then used to calculate the isosteric heats of
adsorption, Qst, by determining the slope of the best-fit line for ln p
versus 1/T at each loading.
⎛Q st
⎞
⎛
⎜
⎞
⎟
1
T
(ln p) =
+ C
⎜
⎟
⎠
q
⎝
⎠
⎝
R
(3)
As indicated by the residual sum of square values, R2, of close to 1
(see Figure S14), the isotherm data were consistent with the
Clausius−Clapeyron equation across the entire loading range
considered, even with the changes in the location of the step in the
isotherms.
The isosteric heats of adsorption for CO2 in the unmodified
Mg2(dobpdc) were determined by fitting the adsorption isotherms at
25, 35, and 45 °C with a dual-site Langmuir−Freundlich equation (eq
1). Each temperature was fit independently, and the Clausius−
Clapeyron equation was used to determine Qst as a function of loading.
All fit parameters for Mg2(dobpdc) and mmen-Mg2(dobpdc) are
specified in Tables S2−S4.
Powder X-ray Diffraction. Powder X-ray diffraction data were
collected with either a Rigaku Ultima III or a Bruker D8 Advance
7058
dx.doi.org/10.1021/ja300034j | J. Am. Chem. Soc. 2012, 134, 7056−7065