In situ X-ray structural studies of a flexible host responding to incremental gas loading

Article abstract of DOI:10.1002/anie.201201281

Crystallographic pressure-lapse snapshots of a porous material responding to gas loading were used to investigate the stepwise uptake of carbon dioxide and acetylene molecules into discrete confined spaces. Based on the data, a qualitative statistical mec

Full text of DOI:10.1002/anie.201201281

Angewandte
Chemie
DOI: 10.1002/anie.201201281
Flexible Pores
In situ X-ray Structural Studies of a Flexible Host Responding to
Incremental Gas Loading**
Tia Jacobs, Gareth O. Lloyd, Jan-Andrꢀ Gertenbach, Kristian K. Mꢁller-Nedebock,
Catharine Esterhuysen, and Leonard J. Barbour*
Synthetic materials with sub-nanometer scale pores and
cavities are of considerable interest for catalysis, as well as
the separation, storage and sensing of gases.[1] Much effort is
therefore being devoted to thwarting close-packing tenden-
cies with a view to engineering and even tuning empty space
in crystals.[2] One of our approaches to the formation of
porous materials is to utilize metallocyclic host molecules that
exhibit poor self-complementarity thus precluding the assem-
bly of close-packed structures.[3]
Porous materials with large voids and/or internal surface
areas are highly desirable for applications where guest storage
capacity is of primary importance.[4] However, materials that
possess more restricted guest-accessible spaces are also of
interest[5]—despite their often low storage capacities they
have potential for discriminating between different guests[6]
based on criteria such as size or electrostatics, and such
selectivity may be useful for applications involving sensing
and separation. Indeed, we are particularly interested in
carrying out in situ structural studies to explore host–guest, as
well as guest–guest interactions within constrained environ-
ments with a view to gaining a better molecular-level under-
standing of important processes.
Scheme 1. Formation of the solvated metallocycle [Cd2Cl4L2]·2CH3OH
(1·2MeOH).
We have now extended our previous work on metal-
locyclic hosts to engineer larger discrete cavities that are each
capable of accepting exactly two volatile guest molecules.
This incremental increase in complexity offers an excellent
opportunity to study gas–solid interactions involving tightly
confined guest-accessible spaces, as well as to construct
a statistical mechanics model that appears to qualitatively
explain subtle features in the experimental sorption isotherms
that arise due to host flexibility.
The ligand 4,4’-bis(2-methylimidazol-1-ylmethyl)biphenyl
(L) was combined with CdCl2·2.5H2O to afford crystals
suitable for X-ray diffraction studies (SCD) (see Supporting
Information). Structure elucidation reveals a [Cd2Cl4L2]
·2CH3OH metallomacrocycle (1·2MeOH, Scheme 1) that
stacks to form columns in the solid state as shown in
Figure 1a, with each complex encircling two methanol
molecules. Analysis of the solvent-accessible volume[7] reveals
that paired methanol guests are situated in discrete voids of
123 ꢀ3. Although permanent channels linking the guest
pockets are not apparent, thermogravimetric analysis (Sup-
porting Information) indicated spontaneous loss of the
solvent molecules from the crystals, even under ambient
conditions. This was confirmed by subsequent SCD as guest
desolvation occurs in a single-crystal to single-crystal fashion
(Figure 1b, Supporting Information).
Since the molecular volumes of CO2 and C2H2 are
approximately the same as that of methanol (Supporting
Information), we surmised that 1 might absorb either of these
two gases to achieve a maximum capacity of two guest
molecules per void. A comparison of the gas sorption
isotherms (Figure 2 and Supporting Information) indicates
that 1 has marginally greater affinity for C2H2 than for CO2,
and both isotherms display Type I behavior with a noticeable
inflection occurring at ca 2 bar for the sorption of C2H2. Since
CO2 and C2H2 have similar dimensions and boiling points, but
inverse electrostatic profiles, it is reasonable to presume that
subtle dissimilarities in the shapes of the isotherms may be
due to differences in how the two gases interact with the host
framework. We therefore consider this system to be an ideal
test case for structural studies involving incremental gas
[*] Dr. T. Jacobs, Dr. G. O. Lloyd, Dr. J. Gertenbach,
Prof. C. Esterhuysen, Prof. L. J. Barbour
Department of Chemistry, University of Stellenbosch
7600 Stellenbosch (South Africa)
E-mail: ljb@sun.ac.za
Prof. K. K. Mꢀller-Nedebock
Institute of Theoretical Physics, Department of Physics
University of Stellenbosch, 7600 Stellenbosch (South Africa)
[**] The authors thank the National Research Foundation of South
Africa for financial support.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201201281.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
We first determined the crystal structure of desolvated
1 under vacuum at room temperature (RT= 228C) (hereafter
designated as 1x, where x = gas pressure in bar). X-ray data
measured at RT for 1 under CO2 pressures of 0.5, 1, 2, 4, 6, 8,
10, 12, and 14 bar (20.5 to 214) did not yield reliable atomic
coordinates for the CO2 molecules. Nevertheless, it was still
possible to determine the CO2 occupancies by summing
electron densities within the guest-accessible volume.[10] An
X-ray diffraction measurement under 10 bar of CO2 pressure
and at À408C (LT210) yielded a fully occupied and ordered
structure in which each CO2 molecule (Supporting Informa-
tion) is positioned approximately perpendicular to the Cl···Cl
vector across the long axis of the cavity (Figure 3). The
position and orientation of the CO2 molecule are consistent
Figure 1. Columnar stacking of 1 along [001] (horizontal direction) as
viewed down [010]. Each guest-accessible cavity (semi-transparent
surface) is defined by four metallocycles and has an approximately
rectangular cross-section terminated by a chloride ion (semi-trans-
parent sphere) at each of the short ends. a) Each cavity (123 ꢁ3) in the
structure of 1·2MeOH is occupied by a pair of methanol molecules.
b) Upon desolvation at room temperature the crystals remain intact
and the cavities maintain their shape (117 ꢁ3).
Figure 3. Crystal structure of 2 determined at À408C under a pressure
of 10 bar of CO2. The CO2 molecules are positioned at approximately
right angles relative to the Cl···Cl vector across the cavity (137 ꢁ3), and
form a weak C(d+)···ClÀ interaction with the host complex.
with the existence of
a weak electrostatic interaction
(10.48 kJmolÀ1 from DFT calculations; Supporting Informa-
tion) between its electropositive carbon atom and the
negatively charged chloride ion of the host.
Successive X-ray intensity data sets were collected at RT
with 1 exposed to C2H2 pressures of 0.5, 1, 2, 4, 6, 8, and 16 bar
(30.5 to 316). The C2H2 molecules were located in difference
electron density maps and their positions and orientations
contrast sharply with those of CO2 in LT210. At 2 bar the linear
C2H2 molecule is aligned approximately along the Cl···Cl axis
across the cavity (Figure 4a) and does not allow room for
a second molecule within the confined space of the void. At
16 bar the cavities are almost all doubly occupied (i.e. the host
is 95% occupied according to sorption analysis) by C2H2
molecules (Figure 4b). By comparing 32 and 316, it is possible
to rationalize the structural consequences of inserting
a second molecule into the cavity—the first molecule to
occupy the cavity is required to change its orientation in order
to accommodate the second within the confined space. The
CH···Cl hydrogen bond is maintained (decreasing in strength
from 16.1 to 9.2 kJmolÀ1, according to DFT calculations;
Supporting Information), but the two molecules align them-
selves across the cavity such that the CH···Cl angle changes by
approximately 258 relative to that in 32.
Figure 2. Sorption isotherms (solid circles) measured at 228C for
a) CO2 and b) C2H2. Both gases exhibit Type I sorption behavior.
Superimposed lines are the result of the statistical physical model with
elastic interactions (using structural data, reasonable energy estimates
and two fitting parameters as described in the Supporting Informa-
tion). Solid lines represent the total occupancy expression, which can
be decomposed into the sum due to the singly (dashed lines) and
doubly (dotted lines) occupied cavities.
loading,[8] and a series of single-crystal X-ray data collections
was carried out using an environmental gas cell (Supporting
Information).[9]
Tracking the pressure-dependent changes in the lattice
parameters as well as in the host conformation and packing
(Table 1 and Supporting Information) proves very useful with
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
rich triple bonds and the host chloride anions. Instead the
CH···Cl hydrogen bonds force the two C2H2 molecules into
orientations that bring them into van der Waals contact with
each other. In order to accommodate this crowded arrange-
ment of the two guest molecules, it is necessary for the host
cavity to deform.
In rationalizing a sorption process involving essentially
identical cavities that can each accommodate a maximum of
two guest molecules, it is useful to consider that any individual
cavity can only exist in one of three possible occupancy states
at any given moment (Figure 5, top): the void can be a) empty
Figure 4. Crystal structures of 3 determined at pressures of 2 and
16 bar of C2H2. a) In 32 each void is, on average, occupied by
approximately one molecule of C2H2. The molecule is situated at either
end of the void and is partially stabilized in that position by the
À
formation of a CH··· Cl hydrogen bond (C···Cl 3.54(3) ꢁ, ]C H···Cl
169.48). The C2H2 molecule is disordered over two equivalent positions
(ii and iii) and the model therefore shows overlapping molecules (i
and iv) indicating that either, but not both, of the positions can be
occupied at any given moment. b) In 316 each cavity accommodates
two molecules of C2H2. The CH···Cl hydrogen bonds are maintained
À
(C···Cl 3.41(3) ꢁ, ]C H···Cl 142.18).
regard to correlating structural data with the sorption
isotherms. It is apparent that deformation of the host
structure with increasing pressure is far more significant for
C2H2 than for CO2. During C2H2 uptake, the size and shape of
the cavity change with the void volume increasing smoothly
from 113 ꢀ3 for 10 to 139 ꢀ3 for 316. As the pressure increases
from vacuum to 16 bar, the void stretches and narrows: the
Cl···Cl distance across the cavity expands by 11% overall
(from 10.61 ꢀ to 11.85 ꢀ) while the Cd···Cd distance across
the cavity contracts by 9% (from 9.44 to 8.59 ꢀ). The
analogous changes due to CO2 loading are far less striking due
to the electrostatically controlled orientation of the CO2. In
contrast, C2H2 molecules cannot be accommodated in the
same manner because of repulsion between their electron-
Figure 5. The three different occupancy states of the C2H2-loaded
structures. Top: crystallographic models showing an individual cavity
in its a) zero-, b) single-, and c) double-occupancy states. The length of
the cavity is related to the Cl···Cl vector, d. Bottom: the statistical
mechanical model presumes different elastic properties of the cavities
depending on the state of occupancy, and is shown schematically for
the three occupancy states. Different equilibrium deformations (y=0,
y1, y2) are associated with energetic minima (e=0, e1, e2) in each
type of occupancy state. The presence of guest molecules changes the
local elastic constant (given by the curvature of the potential at the
minimum).
or it can be occupied by either b) one or c) two gas molecules.
For a representative cavity, these three states most likely
interchange rapidly with one another as gas molecules travel
through the material, and the probability of the existence of
each state should be a constant value at equilibrium.
Although the sorption isotherm provides information regard-
ing the time-averaged occupancy of each cavity at equilibri-
um, it does not yield information about the absolute or even
relative distributions of the three possible states. To this end
we endeavored to combine the experimental observations
with a statistical mechanical treatment in order to derive the
individual isotherms of the three possible occupancy states.
The energy of the system decreases when guest molecules
are accommodated in the framework, but when deformation
of the host is required there is also an energetic penalty
(Figure 5, bottom). Owing to cooperative effects acting
through the lattice, deformation of an individual cavity
could also cause deformation of neighboring cavities, thus
affecting the ability of those voids to accommodate guest
molecules. In the context of three possible occupancy states as
Table 1: Selected parameters of the vacuum structure and the carbon
dioxide (2) and acetylene (3) loaded structures.
Label
Void
[ꢁ3]
a [ꢁ]
b [ꢁ]
c [ꢁ]
B [8]
V
Cl···Cl
[ꢁ]
[ꢁÀ3
]
10
20.5
30.5
21
31
22
32
24
34
26
36
28
38
113
122
126
121
125
123
131
119
131
122
133
127
132
18.208
18.046
18.046
18.033
17.901
18.030
17.788
18.009
17.501
18.028
17.313
18.025
17.269
15.527
15.525
15.525
15.535
15.618
15.535
15.640
15.546
15.637
15.537
15.610
15.544
15.608
9.353
9.394
9.394
9.398
9.447
9.400
9.506
9.403
9.693
9.409
9.832
9.410
9.855
114.9
114.5
114.5
114.5
114.3
114.5
114.1
114.4
113.8
114.4
113.6
114.3
113.5
2398
2394
2394
2396
2407
2397
2413
2398
2427
2401
2435
2403
2435
10.61
10.78
10.83
10.79
10.94
10.81
10.95
10.84
11.51
10.85
11.73
10.86
11.81
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
described above, we formulated a simple statistical mechan-
ical approach (see Supporting Information). The model
assumes host–guest and guest–guest interactions within
a cavity and incorporates cavity deformation as well as elastic
coupling between neighboring cavities in a one-dimensional
crystal. A reduced version of this model, intended to
represent the CO2 sorbing system, assumes that host–guest
and guest–guest interactions occur, but that host deformation
is negligible (Supporting Information). The calculations for
both systems are analogous to that of a standard lattice gas
derivation of the adsorption isotherm,[11] with host deforma-
tion determining the interaction between nearest sites (much
related work[12] has been applied in different contexts to
elastic lattice gases). They yield the total occupancy as
a function of pressure, as well as the contribution to total
occupancy by singly and doubly filled voids (Figure 2). The
full model represents the C2H2 sorbing system, but is also
applicable to CO2. Assuming reasonable elastic constants and
using calculated interaction energies (Supporting Informa-
tion), the inflection in the overall sorption isotherm (as
observed experimentally for acetylene) becomes apparent
(Figure 2b). The qualitative origin of this feature represents
the excess energy required to add a second molecule to
a singly occupied cavity as deformation of the surrounding
voids is taken into account. As filling of the voids continues,
further occupation is eased because much of the required
deformation has already taken place through elastic coupling
to already filled voids. The model also enables the derivation
of expressions for the contributions due to singly and doubly
occupied cavities to the total occupancy that can then be
superimposed on the experimental data. Upon applying two
fitting parameters, the model agrees almost exactly with the
experimental data. We note, however, that steps have also
been observed for open frameworks and host distortion is
therefore not exclusively responsible for inflections[13] present
in sorption isotherms.
We have developed a relatively simple procedure suitable
for routine single-crystal structure analysis with the sample
exposed to controlled gas pressures. As a proof of concept,
a series of “pressure-lapse” crystal structure snapshots was
recorded to smoothly track subtle changes resulting from gas
loading by a porous host. These experiments are analogous to
the measurement of a sorption isotherm, but provide detailed
structural information regarding the underlying mechanism
of the gas uptake process. We observed that carbon dioxide
and acetylene are accommodated in different orientations
within the available space and were able to rationalize this on
the basis of intermolecular host–guest interactions and space
filling considerations.[8h] Furthermore, we have utilized our
observations both quantitatively and qualitatively to formu-
late a statistical mechanical model that accounts for the
shapes of the experimental sorption isotherms. These models
allow deconvolution of the total occupancy into components
representing the statistical distributions of the three possible
occupancy states for a representative cavity with a maximum
possible occupancy of two guest molecules. Future routine
application of our systematic method in conventional SCD
laboratories promises to yield significant advances in the field
of gas sorption studies by providing an improved under-
standing of the processes that occur at the molecular level.
Received: February 15, 2012
Published online: && &&, &&&&
Keywords: acetylene · carbon dioxide · flexible pores ·
.
gas sorption · inclusion compounds
[1] a) J.-R. Li, R. J. Kuppler, H.-C. Zhou, Chem. Soc. Rev. 2009, 38,
1477 – 1504; b) S. Ma, Pure Appl. Chem. 2009, 81, 2235 – 2251;
c) J. Liu, P. K. Thallapally, B. P. McGrail, D. R. Brown, J. Liu,
Chem. Soc. Rev. 2012, 41, 2308 – 2322.
[2] a) O. M. Yaghi, M. OꢁKeeffe, N. W. Ockwig, H. K. Chae, M.
Eddaoudi, J. Kim, Nature 2003, 423, 705 – 714; b) Y. L. Liu, V. C.
Kravtsov, R. Larsen, M. Eddaoudi, Chem. Commun. 2006,
1488 – 1490; c) S. J. Dalgarno, P. K. Thallapally, L. J. Barbour,
J. L. Atwood, Chem. Soc. Rev. 2007, 36, 236 – 245.
´
[3] a) L. Dobrzanska, G. O. Lloyd, H. G. Raubenheimer, L. J.
Barbour, J. Am. Chem. Soc. 2005, 127, 13134 – 13135; b) L.
´
Dobrzanska, G. O. Lloyd, H. G. Raubenheimer, L. J. Barbour, J.
Am. Chem. Soc. 2006, 128, 698 – 699.
[4] a) K. Koh, A. G. Wong-Foy, A. J. Matzger, J. Am. Chem. Soc.
2009, 131, 4184 – 4185; b) G. Fꢂrey, C. Mellot-Draznieks, C.
Serre, F. Millange, J. Dutour„ S. Surble, I. Margiolaki, Science
2005, 309, 2040 – 2042; c) H. K. Chae, D. Y. Siberio-Perez, J.
Kim, Y. Go, M. Eddaoudi, A. J. Matzger, M. OꢁKeeffe, O. M.
Yaghi, Nature 2004, 427, 523 – 527.
[5] a) J. L. Atwood, L. J. Barbour, A. Jerga, Science 2002, 296, 2367 –
2369; b) J. L. Atwood, L. J. Barbour, A. Jerga, B. L. Schottel,
Science 2002, 298, 1000 – 1002; c) T. Tozawa, J. T. A. Jones, S. I.
Swamy, S. Jiang, D. J. Adams, S. Shakespeare, R. Clowes, D.
Bradshaw, T. Hasell, S. Y. Chong, C. Tang, S. Thompson, J.
Parker, A. Trewin, J. Bacsa, A. M. Z. Slawin, A. Steiner, A. I.
Cooper, Nat. Mater. 2009, 8, 973 – 978.
[6] J. L. Atwood, L. J. Barbour, A. Jerga, Angew. Chem. 2004, 116,
3008 – 3010; Angew. Chem. Int. Ed. 2004, 43, 2948 – 2950.
[7] M. L. Connolly, Science 1983, 221, 709 – 713.
[8] a) S. Takamizawa, E. Nakata, T. Saito, Angew. Chem. 2004, 116,
1392 – 1395; Angew. Chem. Int. Ed. 2004, 43, 1368 – 1371; b) S.
Takamizawa, E. Nakata, T. Saito, T. Akatsuka, Inorg. Chem.
2005, 44, 1362 – 1366; c) S. Takamizawa, E. Nakata, T. Saito, K.
Kojima, CrystEngComm 2003, 5, 411 – 413; d) S. Takamizawa, E.
Nataka, T. Akatsuka, R. Miyake, Y. Kakizaki, H. Takeuchi, G.
Maruta, S. Takeda, J. Am. Chem. Soc. 2010, 132, 3783 – 3792;
e) G. Fꢂrey, C. Mellot-Draznieks, C. Serre, F. Millange, J.
Dutour, S. Surblꢂ, I. Margiolaki, Science 2005, 309, 2040 – 2042;
f) E. C. Spencer, J. A. K. Howard, G. J. McIntyre, J. L. C.
Rowsell, O. M. Yaghi, Chem. Commun. 2006, 278 – 280;
g) D. G. Samsonenko, H. Kim, Y. Y. Sun, G. H. Kim, H. S. Lee,
K. Kim, Chem. Asian J. 2007, 2, 484 – 488; h) J. P. Zhang, X. M.
Chen, J. Am. Chem. Soc. 2009, 131, 5516 – 5521; i) R. Vaidhya-
nathan, S. S. Iremonger, G. K. H. Shimizu, P. G. Boyd, S. Alavi,
T. K. Woo, Science 2010, 330, 650 – 653.
[9] D. S. Yufit, J. A. K. Howard, J. Appl. Crystallogr. 2005, 38, 583 –
586.
[10] A. L. Spek, J. Appl. Crystallogr. 2003, 36, 7 – 13.
[11] L. D. Gelb, K. E. Gubbins, R. Radhakrishnan, M. Sliwinska-
Bartkowiak, Rep. Prog. Phys. 1999, 62, 1573 – 1659.
[12] a) D. J. Bergman, B. I. Halperin, Phys. Rev. B 1976, 13, 2145 –
2175; b) M. Presber, B. Dꢃnweg, D. P. Landau, Phys. Rev. E
1998, 58, 2616 – 2623.
[13] H. Li, M. Eddaoudi, A. Laine, M. OꢁKeeffe, O. M. Yaghi, J. Am.
Chem. Soc. 1999, 121, 6096 – 6097.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Flexible Pores
Crystallographic pressure-lapse snap-
shots of a porous material responding to
gas loading were used to investigate the
stepwise uptake of carbon dioxide and
acetylene molecules into discrete con-
fined spaces. Based on the data a qual-
itative statistical mechanical model was
devised that reproduces even subtle
features in the experimental gas sorption
isotherms.
T. Jacobs, G. O. Lloyd, J. Gertenbach,
K. K. Mꢁller-Nedebock, C. Esterhuysen,
L. J. Barbour*
&&&& — &&&&
In situ X-ray Structural Studies of
a Flexible Host Responding to
Incremental Gas Loading
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!