Kinetic gate-opening process in a flexible porous coordination polymer

Article abstract of DOI:10.1002/anie.200705822

(Graph Presented) Open the gate! A porous coordination polymer shows abrupt changes in its adsorption isotherm with guest-dependent gate-opening pressure (see picture). Kinetic analysis reveals that adsorption with structure transformation follows a gate-opening model, in which adsorption proceeds through the formation of an intermediate by a gate-opening process. This process provides a significant difference in onset pressure between similar gas molecules.

Full text of DOI:10.1002/anie.200705822

Communications
DOI: 10.1002/anie.200705822
Coordination Polymers
Kinetic Gate-Opening Process in a Flexible Porous Coordination
Polymer**
Daisuke Tanaka, Keiji Nakagawa, Masakazu Higuchi, Satoshi Horike, Yoshiki Kubota,
Tatsuo C. Kobayashi, Masaki Takata, and Susumu Kitagawa*
Recently, there has been growing interest in the nature of
flexible and dynamic porous coordination polymers (PCPs)
that reversibly change their structures and properties in
response to guest adsorption.[1] After one of the present
authors predicted their importance,[2] many flexible PCPs
have been prepared in only a decade. So-called structural
dynamism has been identified as a key principle for high
selectivity,[3] accommodation,[4] and molecular sensing.[5] One
of the most interesting phenomena in flexible PCPs is
stepwise adsorption caused by the guest-induced framework
transition.[3,4,6] In particular, a gate effect occurs when the
framework structure changes during the adsorption process
from a closed structure to an open one at a specific pressure,
which generates an S-shaped or sigmoidal adsorption profile.
The onset pressure at which the gates of the closed-structure
grooves become open is referred to as the gate-opening
pressure (Pgo) and is related to the structural transformation
of the host framework.[7] Interestingly, the value of Pgo shows
a guest dependency.[8] This difference between adsorbates
suggests that flexible PCPs have potential applications in
many fields, including separation, sensors, and switching
materials. In particular, small gaseous adsorbates, such as O2,
Ar, and N2, are attractive targets for research, first because
the differences in sorption behavior between such similar gas
molecules have attracted wide commercial interest for
processes such as air separation, and also because of scientific
interest as a result of their simple structures and small
differences in physical properties.[9]
However, several questions remain unanswered. Why
does adsorption not occur below Pgo? What determines Pgo?
How can the difference in Pgo for different guests be
enhanced? Generally, a sigmoid isotherm with small hyste-
resis has been understood to be the result of cooperative
phenomena.[10] Several flexible PCPs actually show isotherm
discontinuities at high relative pressure, with large hysteresis,
where kinetics would play an essential role.[11,12] In spite of
their importance, few attempts have been made to determine
the kinetics of the gate effect.[11,13] Herein, we present the
synthesis, crystal structure, and gas sorption properties of a
flexible PCP, {[Cd(bpndc)(bpy)]}n (1; bpndc = benzophenone-
4,4’-dicarboxylate, bpy = 4,4’-bipyridyl), which shows a large
difference in Pgo between O2, Ar, and N2 (Figure 1). To
[*] D. Tanaka,K. Nakagawa,Dr. M. Higuchi,Dr. S. Horike,
Prof. Dr. S. Kitagawa
Department of Synthetic Chemistry and Biological Chemistry
Graduate School of Engineering,Kyoto University
Katsura,Nishikyo-ku,Kyoto 615-8510 (Japan)
Fax: (+81)75-383-2732
E-mail: kitagawa@sbchem.kyoto-u.ac.jp
Figure 1. Adsorption at low pressure.
Dr. Y. Kubota
Department of Physical Science,Graduate School of Science
Osaka Prefecture University,Sakai,Osaka 599-8531 (Japan)
understand the mechanism behind the differences between
similar gases, we treat this phenomenon with the aid of a new
diffusion model in which adsorption proceeds through the
formation of an intermediate. Kinetic analysis revealed that
the formation of the intermediate, which can be described as a
gate-opening process, governs Pgo and enhances the differ-
ence between gases.
Prof. Dr. T. C. Kobayashi
Department of Physics,Okayama University
Okayama 700-8530 (Japan)
Dr. M. Takata
Structural Materials Science Laboratory,Harima Institute
RIKEN SPring-8 Center and CREST,JST,Sayo-gun,Hyogo 679-5148
and
The solvothermal reaction of Cd(NO3)2·4H2O, H2bpndc,
and bpy in dimethylformamide (DMF) produced the solvated
framework compounds {[Cd(bpndc)(bpy)](dmf)(H2O)}n
(1ꢀSolvents). Single-crystal X-ray analysis demonstrated
that cadmium ions are connected by bpndc to produce 1D
double-chain structures of {[Cd(bpndc)]}n along the c axis,
and are linked by bpy along the b axis to give a 2D sheet motif
(Figure 2a and b). The 2D layers are mutually interdigitated
to create a 3D assembled framework (Figure 2c). This is
similar to previously reported PCPs that show a flexible
nature.[14] The 3D structure of 1ꢀSolvents consists of a
Department of Advanced Materials Science
The University of Tokyo (Japan)
[**] We thank Prof. Dr. M. Miyahara,Dr. H. Tanaka,Dr. S. Watanabe,H.
Sugiyama,and S. Shimomura of Kyoto University and Dr. R.
Matsuda of Kyushu University for encouragement of this research
and valuable information. We thank Dr. K. Kato and Dr. K. Osaka in
JASRI for their experimental help at the SPring-8. This work was
supported by ERATO,JST,and a Grant-in-Aid for Scientific Research
in a Priority Area “Chemistry of Coordination Space” (434) from the
MEXT (Japan).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
3914
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 3914 –3918

Angewandte
Chemie
cavities are created, isolated from each other, with a size of
about 6 ” 5 ” 3 Š3. The solvent-accessible volume of this
3
cavity is 132.6 Š and Vvoid is 11.3% of the total crystal
volume, as estimated by Platon.[15]
We measured the adsorption isotherms of O2, Ar, and N2
by volumetric adsorption (Belsorp-18, BEL Japan) at various
temperatures (77, 90, and 100 K) to study the effect of
structural transformation. The isotherms at 90 K showed a
sudden increase at gate-opening pressures Pgo of O2, Ar, and
N2 at 3.9, 40.1, and 55.3 kPa, respectively, and attained
saturation (Figure 3a).[16] It is surprising that the small
Figure 2. Crystal structures of 1ꢀSolvents and 1 showing a) 1D double
chain of {[Cd(bpndc)]}n,b) 2D layer structure of 1ꢀSolvents (gray,
purple,red,and yellow are C,N,O,and Cd,respectively),c) 3D
assembled structure of 1ꢀSolvents,and d) 3D assembled structure of
1. Solvents are omitted for clarity. Cavities and windows are high-
lighted.
unidirectional set of nonintersecting linear arrays of cavities
(12 ” 7 ” 4 Š3) with narrow connecting windows (with cross
sections of about 2 ” 6 Š2) along the [110]or [1 ꢁ10]
directions, as shown in Figure 2c and Figure S3a in the
Supporting Information. The void volume Vvoid is 29.4% of
the total crystal volume, as calculated using the Platon
program.[15] The solvent-accessible volume of the cavity is
403.4 Š3, which includes two water and two DMF molecules.
Thermogravimetric (TG) analysis of 1ꢀSolvents showed
the release of solvent molecules with increasing temperature
up to 1408C, to produce the guest-free phase 1 (Figure S1 in
the Supporting Information). No further weight loss was
observed between 140 and 2308C, which indicates that the
guest-free phase was stable. As the X-ray powder diffraction
(XRPD) pattern of 1 shows sharp diffraction peaks (Figure S2
in the Supporting Information), the porous framework was
largely maintained, even without guest molecules, and 1
retains its single-crystallinity. Single-crystal diffraction data
were then collected. The crystal structure of 1 was solved in
space group P21/a, which is different from the space group for
1ꢀSolvents (C2/c). The c axes in 1 correspond to half of the
c axes in 1ꢀSolvents. The cell volume per Z value decreases
by 14.3%, which indicates a dramatic contraction of the
framework. Crystal structure analysis revealed that the 2D
layer motif is maintained. On the other hand, shrinking of the
layer gap along the a axis by about 0.9 Š occurred, which is
attributable to a gliding motion of the 1D double-chain
moieties (Figure 2d).
Figure 3. a) Sorption isotherms for adsorbates in 1 at 90 K. 1) O2
adsorption (red open circles) and desorption (red filled circles); 2) N2
adsorption (blue open triangles) and desorption (blue filled triangles);
3) Ar adsorption (purple open squares) and desorption (purple filled
squares). Inset: O2 adsorption isotherm at low pressure. b,c) Compar-
ison of calculated profiles for the pressure versus time graphs from
*
the DE model (red lines) with the experimental profiles ( ) for b)
point x to point y of the inset in (a),and c) point y to point z.
difference in physical properties between these gas molecules
should have such a large effect on Pgo. On the other hand,
desorption isotherms showed adsorption volumes decreasing
at lower pressures and large hysteresis loops in the N2 and Ar
isotherms. Adsorption measurements at 77 and 100 K showed
that Pgo increased as the measurement temperature increased
(see the Supporting Information). The adsorption volumes of
O2, Ar, and N2 at 90 K were 158.8, 129.5, and 150.7 mLgꢁ1,
respectively, when the pressure approached 100 kPa. These
isotherms reveal that the gas molecules cannot diffuse into the
isolated cavities of 1 at pressures below Pgo. At Pgo, the gates
of the grooves of 1 open and the structure is transformed from
closed phase to open phase.
The interesting aspect of the desolvated structure is the
presence of interlayer noncovalent interactions. There is close
contact between hydrogen atoms of bpndc and the ketone
oxygen or aromatic carbon atoms of neighboring bpndc (d(H–
O) = 2.89 Š, d(H–C) = 3.44 and 3.53 Š). The void spaces are
separated by interlayer interactions, and zero-dimensional
To observe the structure transition of 1 under gas pressure,
we performed in situ XRPD experiments under various gas
atmospheres (Figure S4 in the Supporting Information). Cell
Angew. Chem. Int. Ed. 2008, 47, 3914 –3918
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3915

Communications
parameters of the O2-adsorbed form were determined by
introduce the GO equation where P is the pressure at time t,
Pe is the pressure at equilibrium, and Pi is the initial pressure
[see the Supporting Information for its derivation; Eq. (2)]:[21]
Le Bail analysis to be a = 15.1588(4), b = 11.7324(1), and c =
30.9112(6) Š, b = 104.430(3), and V= 5324.10 Š3, with space
group P21/c. This is close to the cell parameters of 1ꢀSolvents
(a = 15.52(3), b = 11.79(2), c = 30.92(6) Š; b = 103.87(2), V=
5492 Š3), which indicates that the O2-adsorbed form is open
and expanded. Powder patterns for N2- and Ar-adsorbed
forms showed structures similar to those of the O2-adsorbed
form.
ꢀ
kgo expðꢁkd1 tÞꢁkd1 expðꢁkgo tÞ
P ¼
2 ðkgoꢁkd1
Þ
ð2Þ
ꢁ
kgo expðꢁkd2 tÞꢁkd2 expðꢁkgo tÞ
þ
ðPiꢁPeÞ þ Pe
2 ðkgoꢁkd2
Þ
To clarify the origin of the shape of the adsorption
isotherms and the large difference in Pgo between similar gas
molecules, we analyzed the adsorption phenomena from a
kinetic point of view. Figure 3b and c show how pressure
decays, during O2 adsorption, from point x to point y and
from point y to point z, as labeled in the inset of Figure 3a. In
these volumetric measurements, pressure change is propor-
tional to the adsorption volume. Within these interdigitated
structures, there are barriers as a result of diffusion both
through the windows and along the pore cavities.[14,17] This can
be described by a double-exponential (DE) equation.[18] The
pressure decay from point y to point z follows the DE
equation with good accuracy, thus indicating that the kinetic
behavior of this process is dominated by diffusion (Figure 3c).
On the other hand, the pressure decay curve at Pgo from
point x to point y has an inflection point and cannot be
described by the DE model (Figure 3b). This type of decay
curve has been reported in flexible PCPs[19,20] and the pseudo-
plateau at the beginning of this curve suggests the existence of
a certain “growing process”.
We collected kinetic data by one-point volumetric
adsorption measurements. The pressure Pi of an adsorbate
was introduced into the evacuated sample cell and allowed to
equilibrate at Pe. We monitored the pressure change with time
for various values of Pi at different temperatures. A compar-
ison of the profile calculated according to the GO equation
with the experimental profile for O2 adsorption at 90 K with
Pi = 11.8 kPa is shown in Figure 5a. It is apparent that the GO
model provides a good description of the kinetics for
adsorption with structure transformation.
To describe the decay curve that includes the pseudo-
plateau, another diffusion process could be introduced. In this
model, adsorption proceeds through the formation of an
intermediate [see Figure 4; Eq. (1)]:
Figure 5. a) Variation of pressure with time for adsorption of O2 at
*
90 K with Pi =11.8 kPa ( ) and the calculated profile for the GO model
(gray line). b) Plot of the logarithm of kgo (O2 adsorption at 90 K)
against the logarithm of Pi.
kgo
kd1,kd2
C
I
A
ð1Þ
ƒ! ƒƒƒ!
The kinetic data obtained from one-point measurements
of O2 adsorption on 1 at 90 K with various initial pressures
(Pi = 11.3, 11.8, 12.5, 12.9, and 13.6 kPa) also followed the GO
model, thus affording an experimental estimate of kgo. The
kinetic data are given in the Supporting Information. A plot
of the logarithms of the experimental kgo values against the
logarithms of Pi is a straight line with slope 14.5 (Figure 5b),
which indicates that we could write kgo = k’go(Pi)a in the
measurement region (where a is the slope of the line). The
abruptly changing adsorption isotherms of 1 are explained by
the large pressure dependency of kgo, as follows. When Pi is
much lower than Pgo, the gate-opening process does not
proceed (kgo is nearly zero) and the diffusion process does not
start. When the pressure is close to Pgo, however, kgo increases
very rapidly and adsorption then occurs. Adsorption does not
occur below Pgo because gate opening is required before the
diffusion process can begin. In other words, the kinetic
behavior of the gate-opening process determines Pgo.
Figure 4. The gate-opening (GO) model.
Here, a closed phase C is transformed to an intermedia-
te I, which can adsorb gas molecules. The formation of I can
be referred to as the gate-opening process. We will suppose
that gas molecules adsorb only on the crystal surface during
the gate-opening process, and therefore that the adsorption
volume during this process is quite small and does not affect
the pressure. The gate-opening process (C to I) can be
described by a first-order rate law (with rate constant kgo).
The intermediate I, formed from C, adsorbs gas and decays to
the adsorbed form A. This diffusion process can be described
by the DE model (with rate constants kd1 and kd2). We call this
model the gate-opening (GO) model. To describe pressure
decay in volumetric measurements using the GO model, we
We also measured the adsorption kinetics of 1 with
different gases (O2, N2, and Ar) at 90 K. The adsorption of Ar
on 1 with Pi = 45.3 kPa followed the GO model well (kgo
=
4.7 ” 10ꢁ4, d1 = 1.3 ” 10ꢁ5, d2 = 3.4 ” 10ꢁ6). The pressure
k
k
decay of N2 for various values of Pi (59.0, 72.5, and
90.4 kPa) followed the GO equation, but with large errors
3916
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 3914 –3918

Angewandte
Chemie
technology Support Project of MEXT (Proposal No. 2006A1677/
because the gate-opening process is slow and pressure decay
affects kgo.[21] The small kgo values for N2 and Ar adsorption
indicate that the Pgo of N2 and Ar are dominated by the gate-
opening process. If this process did not exist, adsorption
would occur under lower pressure conditions. The gate-
opening process dramatically increases the difference in Pgo
between similar gases. Comparison of the kinetic data shows
that the order of kgo for the same pressure is O2 > Ar> N2,
which is consistent with the guest dependency of Pgo. This
order is also consistent with that of the boiling points (90.2,
87.3, and 77.4 K for O2, Ar, and N2, respectively), which
suggests that the intermolecular interaction force of the guest
molecules governs this process.
BL02B2). The adsorption isotherm for N2, Ar, and O2 was measured
with BELSORP-18 volumetric adsorption equipment attached to a
closed-cycle helium cryostat (BEL Japan, Inc.).
X-ray structure determination: Measurements were made on a
Rigaku Mercury CCD system with MoKa radiation. In all cases, the
structure was solved by direct methods and refined by full-matrix
least-squares techniques on F2 (SHELXL-97).[23] All non-hydrogen
atoms were anisotropically refined, while all hydrogen atoms were
placed geometrically and refined with a riding model with Uiso
constrained to be 1.2 times Ueq of the carrier atom.
Crystal data for 1ꢀSolvents: C28H23CdN3O7, Mr = 625.89, mono-
clinic, space group C2/c, (#15), a = 15.52(3), b = 11.79(2), c =
30.92(6) Š, b = 103.87(2), V= 5492(18) Š3, Z = 8, T= 173 K, 1calcd
=
1.514 gcmꢁ3
,
m(MoKa) = 0.845 cmꢁ1
,
2qmax = 50.08,
l(MoKa) =
0.71070 Š, 11161 reflections measured, 4726 unique, 4227 > 2s(I)
were used to refine 347 parameters, no restraints, R(Rw) = 0.0450
(0.0908), GOF = 1.124. Crystal data for 1: C25H16CdN2O5, Mr =
536.80, monoclinic, space group P21/a, (#14), a = 13.758(7), b =
11.753(5), c = 15.648(7) Š, V= 2352.8(18) Š3, Z = 4, T= 223 K,
1calcd = 1.515 gcmꢁ3, m(MoKa) = 0.966 cmꢁ1, 2qmax = 50.08, l(MoKa) =
0.71070 Š, 12226 reflections measured, 3978 unique, 3366 > 2s(I)
were used to refine 298 parameters, no restraints, R(Rw) = 0.0997
(0.2175), GOF = 1.202. CCDC-671180 and CCDC-671181 contain the
supplementary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
We also carried out one-point O2 adsorption measure-
ments at various temperatures (88, 90, 92, and 95 K with Pi in
the range 11.80–11.93 kPa). As the temperature increased,
the value of kgo decreased, which is consistent with the
temperature dependency of Pgo (see the Supporting Informa-
tion). Such anti-Arrhenius behavior was observed in clathra-
tion reactions of an organic host.[13] Anti-Arrhenius behavior
is usually indicative of a multistep mechanism (see the
Supporting Information).[22] The features of the gate-opening
process can be summarized as follows: 1) the adsorption
volume is quite small; 2) kgo shows a strong dependency on
pressure (or concentration of adsorbate); 3) an adsorbate
with a higher boiling point provides a larger kgo; and 4) kgo
decreases with increasing temperature. These features should
be attributable to the condensation of adsorbate on a crystal
surface, and some structural transformation might then be
induced.
In conclusion, we have prepared a flexible PCP that shows
abrupt changes in its adsorption isotherms with guest-
dependent gate-opening pressure. We also demonstrated
that the kinetics of the gate-opening process provides for
large differences in onset pressure between similar gas
molecules. These kinetic analyses indicate that the gate-
opening process could be associated with the condensation of
adsorbate on a crystal surface. Modification of a crystal
surface could provide new mechanisms for adsorption phe-
nomena. A detailed understanding of the gate-opening
process would be the key to further fine-tuning of the gas-
adsorption performance of PCPs.
Received: December 19, 2007
Published online: April 11, 2008
Keywords: adsorption · coordination polymers · kinetics ·
.
microporous materials · phase transitions
[1]a) A. J. Fletcher, K. M. Thomas, M. J. Rosseinsky, J. Solid State
Chem. 2005, 178, 2491; b) K. Uemura, R. Matsuda, S. Kitagawa,
J. Solid State Chem. 2005, 178, 2420; c) C. J. Kepert, Chem.
Commun. 2006, 695; d) M. P. Suh, Y. E. Cheon, Aust. J. Chem.
2006, 59, 605; e) J. P. Zhang, X. M. Chen, Chem. Commun. 2006,
1689; f) D. Bradshaw, J. E. Warren, M. J. Rosseinsky, Science
2007, 315, 977; g) C. Serre, C. Mellot-Draznieks, S. SurblØ, N.
Audebrand, Y. Filinchuk, G. FØrey, Science 2007, 315, 1828; h) S.
Kitagawa, R. Kitaura, S. Noro, Angew. Chem. 2004, 116, 2388;
Angew. Chem. Int. Ed. 2004, 43, 2334.
[2]S. Kitagawa, M. Kondo, Bull. Chem. Soc. Jpn. 1998, 71, 1739.
[3]P. L. Llewellyn, S. Bourrelly, C. Serre, Y. Filinchuk, G. FØrey,
Angew. Chem. 2006, 118, 7915; Angew. Chem. Int. Ed. 2006, 45,
7751.
[4]Y. Kubota, M. Takata, R. Matsuda, R. Kitaura, S. Kitagawa, T. C.
Kobayashi, Angew. Chem. 2006, 118, 5054; Angew. Chem. Int.
Ed. 2006, 45, 4932.
[5]G. J. Halder, C. J. Kepert, B. Moubaraki, K. S. Murray, J. D.
Cashion, Science 2002, 298, 1762.
[6]A. Kondo, H. Noguchi, L. Carlucci, D. M. Proserpio, G. Ciani, H.
Kajiro, T. Ohba, H. Kanoh, K. Kaneko, J. Am. Chem. Soc. 2007,
129, 12362.
[7]D. Li, K. Kaneko, Chem. Phys. Lett. 2001, 335, 50.
[8]R. Kitaura, K. Seki, G. Akiyama, S. Kitagawa, Angew. Chem.
2003, 115, 444; Angew. Chem. Int. Ed. 2003, 42, 428.
[9]C. R. Reid, I. P. Oꢀkoye, K. M. Thomas, Langmuir 1998, 14, 2415.
[10]T. Dewa, K. Endo, Y. Aoyama, J. Am. Chem. Soc. 1998, 120,
8933.
Experimental Section
Synthesis of 1ꢀSolvents and 1: A mixture of Cd(NO3)2·4H2O (1.0 g,
3.24 mmol), H2bpndc (0.88 g, 3.26 mmol), and bpy (0.50 g, 3.20 mmol)
was suspended in DMF (40 mL) and heated in a round-bottomed
flask at 1208C for 1 day. A colorless crystalline precipitate 1ꢀSolvents
was formed. These crystals were collected, washed with ethanol, and
dried in a vacuum at 1208C, which resulted in guest-free solid 1.
Elemental analysis of 1 calcd (%) for C25H16CdN2O5: C 55.93, H 3.00,
N 5.22; found: C 55.68, H 3.40, N 5.30.
Physical measurements: TG analyses were performed using a
Rigaku Thermo plus TG 8120 apparatus in the temperature range
between 303 and 773 K in a N2 atmosphere and at a heating rate of
10 Kminꢁ1
. Elemental analyses were performed on a Thermo
Finnigan EA1112 apparatus. XRPD data of 1ꢀSolvents and 1 were
collected on a Rigaku RINT-2200HF (Ultima) diffractometer with
CuKa radiation. In situ synchrotron radiation experiments were
performed at SPring-8 with the approval of JASRI as a Nano-
[11]K. Uemura, S. Kitagawa, K. Fukui, K. Saito, J. Am. Chem. Soc.
2004, 126, 3817.
[12]B. L. Chen, S. Q. Ma, F. Zapata, F. R. Fronczek, E. B. Lobkov-
sky, H. C. Zhou, Inorg. Chem. 2007, 46, 1233.
Angew. Chem. Int. Ed. 2008, 47, 3914 –3918
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3917

Communications
[13]L. J. Barbour, M. R. Caira, T. le Roex, L. R. Nassimbeni, J.
[21]Generally, volumetric adsorption measurement is not appropri-
ate to kinetic analysis because pressure decay affects the value of
the rate constant. Therefore, we do not discuss the absolute
values of kd1 and kd2, although the DE equation describes
pressure decay in the diffusion process well. However, when the
Chem. Soc. Perkin Trans. 2 2002, 1973.
[14]S. Horike, D. Tanaka, K. Nakagawa, S. Kitagawa, Chem.
Commun. 2007, 3395.
[15]A. L. Spek, J. Appl. Crystallogr. 2003, 36, 7.
[16]An abrupt increase in adsorption volume and decrease in
pressure was observed at Pgo. This phenomenon can be explained
according to the method of measuring volumetric adsorption. At
Pgo, 1 transforms from a closed to an open structure and suddenly
adsorbs some of the gas in the measurement cell, which results in
a loss of pressure.
gate-opening process is much faster than diffusion (that is, kgo
@
kd1, kd2), the gate-opening process has finished before the
pressure changes significantly and pressure decay does not affect
the gate-opening process. For instance, when the extent of gate
opening was 70% for Pi = 11.8 kPa (kgo = 3.5 ” 10ꢁ3, kd1 = 6.4 ”
10ꢁ4, kd2 = 2.3 ” 10ꢁ4), the pressure decreased by only 0.45 kPa.
On the other hand, when kgo is relatively small, the error
becomes larger and the GO model cannot describe the pressure
decay. Herein, we mainly discuss cases where kgo is more than
twice the value of kd1 and kd2.
[17]E. J. Cussen, J. B. Claridge, M. J. Rosseinsky, C. J. Kepert, J. Am.
Chem. Soc. 2002, 124, 9574.
[18]A. J. Fletcher, E. J. Cussen, D. Bradshaw, M. J. Rosseinsky, K. M.
Thomas, J. Am. Chem. Soc. 2004, 126, 9750.
[19]S. Shimomura, S. Horike, R. Matsuda, S. Kitagawa, J. Am. Chem.
Soc. 2007, 129, 10990.
[22]J. W. Moore, R. G. Pearson, Kinetics and Mechanism, Wiley,
New York, 1981.
[20]M. Eddaoudi, H. L. Li, O. M. Yaghi, J. Am. Chem. Soc. 2000,
122, 1391.
[23]G. M. Sheldrick, SHELXL-97, University of Goettingen (Ger-
many), 1997.
3918
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 3914 –3918