Template control over dimerization and guest selectivity of interpenetrated coordination cages

Article abstract of DOI:10.1021/ja403184a

We have previously shown that the self-assembly of dibenzosuberone-based bis-monodentate pyridyl ligands L¹ with Pd^II cations leads to the quantitative formation of interpenetrated coordination cages [BF ₄@Pd₄L¹₈]. The BF₄ ^- anion inside the central cavity serves as a template, causing the outer two pockets to show a tremendous affinity for allosteric binding of two small chloride anions. Here we show that derivatization of the ligand backbone with a bulky aryl substituent allows us to control the dimerization and hence the guest-binding ability of the cage by the choice of the templating anion. Steric constraints imposed by L² prevent the large BF ₄^- anion from serving as a template for the formation of interpenetrated double cages. Instead, a single isomer of the monomeric cage [Pd₂L²₄] is formed. Addition of the small anionic template Cl^- permits dimerization, yielding the interpenetrated double cage [Cl@Pd₄L²₈], whose enlarged outer pockets show a preference for the binding of large anions such as ReO₄^-.

Full text of DOI:10.1021/ja403184a

Communication
pubs.acs.org/JACS
Template Control over Dimerization and Guest Selectivity of
Interpenetrated Coordination Cages
Sabrina Freye,† Reent Michel,† Dietmar Stalke,† Martin Pawliczek,‡ Holm Frauendorf,‡
and Guido H. Clever*,†
†Institute for Inorganic Chemistry, Georg-August-Universitat Gottingen, Tammannstraße 4, 37077 Gottingen, Germany
̈
̈
̈
‡Institute for Organic and Biomolecular Chemistry, Georg-August-Universitat Gottingen, Tammannstraße 2, 37077 Gottingen,
Germany
̈
̈
̈
S
* Supporting Information
template control of the pocket sizes in interpenetrated10
coordination cages. Toward the realization of the latter
approach, we recently reported the anion-binding capabilities
ABSTRACT: We have previously shown that the self-
assembly of dibenzosuberone-based bis-monodentate
pyridyl ligands L1 with PdII cations leads to the
of a dimeric intercatenated coordination cage, [BF4@Pd4L1 ],
8
−
quantitative formation of interpenetrated coordination
containing one templating BF4 anion tightly bound in its
−
cages [BF4@Pd4L1 ]. The BF4 anion inside the central
−
central cavity and two loosely bound BF4 anions in the
8
symmetry-equivalent outer pockets.11 The outer two anions can
be replaced by two halide anions in a positive cooperative-
binding process (Figure 1a).12 Chloride is bound with
tremendous affinity (Knet ≈ 1020 M−2), as exemplified by the
cage’s ability to dissolve AgCl in acetonitrile.13 We further
showed that the allosteric binding mechanism induces a
shrinking of the double cage along the Pd4 axis by 3.3%,
accompanied by a relative torsion of the two intercatenated
cavity serves as a template, causing the outer two pockets
to show a tremendous affinity for allosteric binding of two
small chloride anions. Here we show that derivatization of
the ligand backbone with a bulky aryl substituent allows us
to control the dimerization and hence the guest-binding
ability of the cage by the choice of the templating anion.
−
Steric constraints imposed by L2 prevent the large BF4
anion from serving as a template for the formation of
interpenetrated double cages. Instead, a single isomer of
the monomeric cage [Pd2L2 ] is formed. Addition of the
4
small anionic template Cl− permits dimerization, yielding
the interpenetrated double cage [Cl@Pd4L2 ], whose
8
enlarged outer pockets show a preference for the binding
−
of large anions such as ReO4 .
anoscopic rings and cages based on hydrogen-bonded1 or
N
metal-mediated2 self-assembly are widely used as hosts
for the encapsulation of guest species. Recent work by Fujita,
Stang, Nitschke, and others has spurred the understanding of
the assembly principles with respect to the size, shape, and
guest-binding capabilities of such structures.3 Anion binding by
positively charged coordination cages has been applied in
various sensing, transport, and separation tasks4 and in the
construction of functional supramolecular systems such as
stacked metal arrays5 and assemblies capable of light-triggered
crystallization.6 One prerequisite for the formation of such non-
covalent assemblies is a large value of the host−guest
association constant (K). K is a function of the guest’s charge
(i.e., its charge density and distribution) and size.7 The
dependence on the latter parameter can be optimized by
screening a series of homologous cage derivatives with the aim
of finding the ideal match between the size of the anionic guest
and the cavity.
Figure 1. (a) The addition of [Pd(CH3CN)4](BF4)2 to ligand L1 leads
to the formation of the thermodynamically unstable monomeric cage
−
[Pd2L1 ](BF4)4, which rearranges quantitatively to give the BF4 -
4
templated, interpenetrated double cage [BF4@Pd4L1 ](BF4)7. The
8
latter is able to bind two chloride anions in its small outer pockets. (b)
Addition of [Pd(CH3CN)4](BF4)2 to ligands L2a−c (Ra = CF3; Rb =
OCH3; Rc = CH3) leads to the formation of the thermodynamically
stable monomeric cages [Pd2L2a−c4](BF4)4. Addition of 0.5 equiv of
chloride then leads to the templated formation of the dimers [Cl@
Pd4L2a−c8](BF4)7. The latter can bind two perrhenate anions in their
−
large outer pockets. All external and all loosely bound BF4 counter
anions (inside the monomeric cages and the outer pockets of the
dimeric cages) have been omitted for clarity.
To minimize the synthetic effort associated with the
generation of such cage libraries, we envisioned two alternative
strategies for the preparation of adjustable anion-binding
systems:8 one approach uses light-switchable cage structures
to modulate the cavity size,9 and the other method is based on
Received: April 2, 2013
Published: May 22, 2013
© 2013 American Chemical Society
8476
dx.doi.org/10.1021/ja403184a | J. Am. Chem. Soc. 2013, 135, 8476−8479

Journal of the American Chemical Society
Communication
{Pd2L1 } units by 8°.13,14 This mechanism poses the following
question: Can the choice of the anion inside the central pocket
be used to control the anion-binding selectivity in the outer two
pockets via a structural relay effect?
parent ligand L1, which is just a kinetic intermediate on the way
4
to the thermodynamic end product [BF4@Pd4L1 ].12 We
8
attribute this behavior to the steric influence of the attached
side arms of L2, which would clash with the other inter-
penetrating unit in a hypothetical [BF4@Pd4L2 ] species.
To investigate this question, we synthesized the modified bis-
monodentate pyridyl ligands L2a−c (Ra = CF3; Rb = OCH3; Rc =
CH3) carrying a bulky aryl substituent attached to the ligand’s
central backbone carbonyl atom (Figure 1b) via Grignard
additions to L1. When ligand L2a was treated with a
stoichiometric amount of [Pd(CH3CN)4](BF4)2 in acetonitrile,
8
Interpenetration of two {Pd2L2 } units could, however, be
4
enforced when 0.5 equiv of chloride was added to the solution
containing monomeric cage [Pd2L2 ] (Figure 1b). In this case,
4
warming of the mixture led to a gradual transition into a
solution in which the interpenetrated double cage [Cl@Pd4L2 ]
8
a single isomer of the monomeric coordination cage [Pd2L2a
]
4
was the most abundant species, although it was obviously in
equilibrium with minor amounts of [Pd2L2 ] and free ligand L2
was formed quantitatively, as indicated by the shift (but not
4
1
(when F− or Br− was added as the template instead of Cl−,
double cages were also formed, but in lower relative amounts).
splitting) of the signal observed in the H NMR spectrum
[Figure 2a(i,ii)] and the occurrence of peaks for the species
In full agreement with the findings for the double cage [BF4@
1
Pd4L1 ], the H NMR spectrum of [Cl@Pd4L2a8] showed the
8
typical pattern for an interpenetrated dimer with a twofold
splitting of each signal and a pronounced downfield shift of the
signals for the protons next to the pyridine N atoms, especially
for the g′ protons pointing into the central cavity [Figure
2a(iii)]. Furthermore, the most intense signals in the HR-ESI-
FTICR mass spectrum can be assigned to the species
5+
[2BF4+Cl@Pd4L2a
]
and [2BF4+Cl@Pd4L2a8+BF4]4+, in
8
which the central pocket contains one chloride anion and
each of the two outer pockets is filled with a BF4− anion (Figure
2c). In addition, we were able to obtain single crystals of [Cl@
Pd4L2b8] suitable for X-ray analysis by slow evaporation of a
solution in acetonitrile. Figure 3 compares the structures of the
previously reported BF4 -templated double cage [BF4@Pd4L1 ]
−
8
(Figure 3a, left)12 and the Cl−-templated double cage [Cl@
Pd4L2b8] (Figure 3a, right) with respect to the Pd−Pd distance,
the size of the templating anion, and the volume of the outer
two pockets available for guest binding. Whereas the distance
between the inner Pd centers decreases from 8.25 to 6.26 Å in
−
going from the BF4 -templated to the Cl−-templated double
cage, the distance between an outer Pd and the nearer inner Pd
increases from 8.09 to 8.79 Å. This change is also reflected in
the volumes observed for the inner and outer pockets (Table
1).
We can explain the structural differences between [BF4@
Pd4L1 ] and [Cl@Pd4L2 ] in terms of the cooperation of two
8
8
factors: in L2, the introduction of steric bulk prevents the
−
formation of a BF4 -templated double cage. Hence, only a small
anion such as chloride can template the dimerization, even in
−
the presence of excess BF4 , since it allows the aryl substituents
of L2 to steer clear of the ligand arms of the interpenetrating
cage fragment (Figure 3b,c). This results in the formation of a
small central cavity and two large outer pockets (visualized in
Figure 3a by depicting the results of VOIDOO16 cavity
calculations).
1
Figure 2. (a) H NMR spectra of (i) ligand L2a, (ii) monomeric cage
[Pd2L2a4](BF4)4, and (iii) dimeric cage [Cl@Pd4L2a8](BF4)7 (400
MHz, 298 K, CD3CN). (b−d) ESI-FTICR mass spectra of (b)
monomeric cage [Pd2L2a4](BF4)4, (c) double cage [2BF4+Cl@
Pd4L2a8](BF4)5, and (d) the host−guest complex [2ReO4+Cl@
Pd4L2a8](BF4)5.
Next, we examined the anion-binding capabilities of the
double cage [Cl@Pd4L2 ]. As expected, the selectivity of the
8
[Pd2L2a4]4+, [Pd2L2a4+BF4]3+, and [Pd2L2a4+2BF4]2+ in the
high-resolution electrospray ionization Fourier transform ion
cyclotron resonance (HR-ESI-FTICR) mass spectrum (Figure
2b).15 Comparison of the calculated relative energies of the four
possible stereoisomers further supported the formation of a
single isomer in which all of the aryl substituents exhibit the
same rotational sense [see the Supporting Information (SI)].
Interestingly, this monomeric cage species is thermodynami-
cally stable without showing any signs of interpenetration in the
outer pockets shifted toward larger guests. Whereas [BF4@
Pd4L1 ] was found to be a strong binder of halide anions
8
(especially chloride), [Cl@Pd4L2 ] showed a preference for the
8
binding of larger anions such as perchlorate, hexafluorophos-
phate, and in particular perrhenate (see the SI). Figure 4a
shows the results of the 1H NMR titration of a CD3CN
solution of [Cl@Pd4L2a8](BF4)7 with NBu4ReO4. In particular,
protons a, g, and f′, all of which point into the outer pockets,
−
display downfield shifts upon ReO4 binding. In contrast,
−
presence of BF4 anions in acetonitrile solution, in stark
protons g′ and a′, which are close to the encapsulated chloride
anion, are shifted upfield, most likely because the inner cavity
contrast to the monomeric [Pd2L1 ] species formed by the
4
8477
dx.doi.org/10.1021/ja403184a | J. Am. Chem. Soc. 2013, 135, 8476−8479

Journal of the American Chemical Society
Communication
second guests, respectively. The fact that K2 is 3.4 times larger
than the statistically expected value of K1/4 indicates positive
cooperativity of the system. Furthermore, the 2:1 guest:host
ratio was supported by Job plot analysis (Figure 4b) and the
HR-ESI-FTICR mass spectrum of [2ReO4+Cl@Pd4L2a8](BF4)5
(Figure 2d).
The observation that the strongly bound guest Cl− in [2Cl
+BF4@Pd4L1 ] showed slow exchange in the NMR experi-
8
−
ments while the weakly bound guest ReO4 in [2ReO4+Cl@
Pd4L2 ] showed fast exchange kinetics is worth discussing. A
8
plausible explanation can be given on the basis that the energy
difference between the starting materials in the two processes is
much smaller than the energy difference between the products
of the guest encapsulation. This leads to a large difference in
the ΔG° values for the two reactions and hence a large
difference in the K values. If it is assumed that the rate
constants for binding (the on rates) are comparable for the two
processes, the difference in exchange kinetics is dominated by
the rate constant for anion release from the pockets (the off
rates), which is much smaller for Cl− than for ReO4 .
−
To compare the anion selectivities, we performed cross
experiments: when Cl− was titrated into a solution of [Cl@
1
Pd4L2a8], a release of free ligand L2a was observed in the H
NMR spectrum instead of a shift in the proton signals of the
double cage. This behavior was also observed when excess Cl−
was added to the chloride-saturated host−guest complex [2Cl
Figure 3. (a) Comparison of the X-ray structures of [BF4@
Pd4L1 ](BF4)7 (left)12 and [Cl@Pd4L2b8](BF4)7 (right). Color
+BF4@Pd4L1 ].12
8
8
scheme: C, light/dark gray; N, blue; O, red; Cl, yellow; F, green; B,
brown; Pd, beige. For clarity, H atoms, solvent molecules, and anions
not in the central pockets have been omitted. (b) Space-filling top
views and (c) stick side views of the inner Pd(py)4 planes (H atoms
omitted) illustrating the differences in the relative positions of the
The outcome of the addition of ReO4− to [BF4@Pd4L1 ] was
8
unexpected. Although shifts in some of the proton signals were
observed in the 1H NMR titration, a binding constant could not
be calculated in this case because mass spectrometric
monitoring of the titration showed the partial formation of
interpenetrating cage fragments in [BF4@Pd4L1 ](BF4)7 (left) and
8
the species [3ReO4@Pd4L1 ] encapsulating a ReO4 anion
−
[Cl@Pd4L2b8](BF4)7 (right). In (c), the distances between the N4 and
8
instead of BF4− in the central pocket. We attribute this process
to the better match of the size of a ReO4 anion with the
O4 planes are given.
−
a
Table 1. Pocket Volumes and Packing Coefficients
volume of the inner cavity (packing coefficient of 58% for
−
ReO4− vs 47% for BF4 ). Interestingly, the equilibrium between
b
pocket
[BF4@Pd4L1 ]
[Cl@Pd4L2b
]
8
8
−
−
the BF4 - and ReO4 -templated double cages shifted toward
the latter compound upon addition of chloride. This
observation demonstrates the possibility of further extending
the range of anionic templates in the central cavity.
before guest
addition
outer
inner
outer
inner
49.9 (BF4−: 103%)
108.9 (BF4−: 47%)
23.5 (Cl−: 96%)
183.7 (BF4−: 28%)
16.2 (Cl−: 139%)
183.7 (ReO4−: 35%)
c
after guest
addition
c
155.4 (BF4−: 33%)
16.2 (Cl−: 139%)
Herein we have shown that the interpenetration and guest-
binding abilities of dimeric coordination cages can be
controlled by ligand derivatization and the choice of the
templating anion. In contrast to our previous findings,12 the
introduction of bulky substituents into the ligand backbone
allowed us to isolate monomeric cage species and induce
dimerization by addition of the template at a later time. We
have further shown that the size of the template inside the
central pocket controls the size selectivity of allosteric anion
binding in the two outer pockets. We think that our
observations add to the understanding of anion -binding
processes in dynamic supramolecular and biological systems.
Furthermore, this strategy might find application in adaptive
sensors and selective anion cotransporter systems.
a
Pocket volumes in Å3 and (guests: packing coefficients) of [BF4@
Pd4L1 ] and [Cl@Pd4L2b8] before and after guest addition are shown.
8
b
c
Data from ref 13. Based on the host structure before guest addition.
Figure 4. (a) Plot of 1H NMR signal shifts observed in the titration of
[Cl@Pd4L2a8](BF4)7 with NBu4ReO4 (CD3CN, 293 K). (b) Job plot
ASSOCIATED CONTENT
■
−
analysis showing a ReO4 :[Cl@Pd4L2a8] ratio of 2:1.
S
* Supporting Information
Synthetic procedures; analytical data for L2a−c, [Pd2L2a−c
]
4
-
(BF4)4, and [Cl@Pd4L2a−c8](BF4)7; and NMR titration, ESI-
MS, molecular modeling, and X-ray data. This material is
available free of charge via the Internet at http://pubs.acs.org.
slightly relaxes upon binding of ReO4− inside the outer pockets.
Nonlinear data fitting17 delivered the values K1 = 2158 61
M−1 and K2 = 1848
30 M−1 for binding of the first and
8478
dx.doi.org/10.1021/ja403184a | J. Am. Chem. Soc. 2013, 135, 8476−8479

Journal of the American Chemical Society
Communication
(15) Examples of other [M2L4] cages: (a) McMorran, D.; Steel, P.
Angew. Chem., Int. Ed. 1998, 37, 3295. (b) Chand, D. K.; Biradha, K.;
Fujita, M. Chem. Commun. 2001, 1652. (c) Su, C.; Cai, Y. P.; Chen, C.;
Smith, M. D.; Kaim, W.; zur Loye, H. C. J. Am. Chem. Soc. 2003, 125,
8595. (d) Yue, N. L. S.; Eisler, D. J.; Jennings, M. C.; Puddephatt, R. J.
Inorg. Chem. 2004, 43, 7671. (e) Clever, G. H.; Tashiro, S.; Shionoya,
M. Angew. Chem., Int. Ed. 2009, 48, 7010. (f) Liao, P.; Langloss, B. W.;
Johnson, A. M.; Knudsen, E. R.; Tham, F. S.; Julian, R. R.; Hooley, R.
J. Chem. Commun. 2010, 46, 4932. (g) Kishi, N.; Li, Z.; Yoza, K.; Akita,
M.; Yoshizawa, M. J. Am. Chem. Soc. 2011, 133, 11438. (h) Lewis, J. E.
M.; Gavey, E. L.; Cameron, S. A.; Crowley, J. D. Chem. Sci. 2012, 3,
778.
AUTHOR INFORMATION
Corresponding Author
gclever@gwdg.de
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the DFG through Grant CL 489/
2-1 and the Fonds der Chemischen Industrie. We thank David
M. Engelhard for the VOIDOO calculations.
(16) Kleywegt, G. J.; Jones, T. A. Acta Crystallogr., Sect. D 1994, 50,
178.
(17) Thordarson, P. Chem. Soc. Rev. 2011, 40, 1305.
REFERENCES
■
(1) Palmer, L.; Rebek, J., Jr. Org. Biomol. Chem. 2004, 2, 3051.
(2) Reviews: (a) Pluth, M. D.; Raymond, K. N. Chem. Soc. Rev. 2007,
36, 161. (b) Dalgarno, S. J.; Power, N. P.; Atwood, J. L. Coord. Chem.
Rev. 2008, 252, 825. (c) Tranchemontagne, D. J.; Ni, Z.; O’Keeffe, M.;
Yaghi, O. M. Angew. Chem., Int. Ed. 2008, 47, 5136. (d) Chakrabarty,
R.; Mukherjee, P. S.; Stang, P. J. Chem. Rev. 2011, 111, 6810.
(e) Ronson, T. K.; Zarra, S.; Black, S. P.; Nitschke, J. R. Chem.
Commun. 2013, 49, 2476. (f) Cook, T. R.; Zheng, Y.-R.; Stang, P. J.
Chem. Rev. 2013, 113, 734.
(3) Further recent examples of self-assembled coordination cages:
(a) Sun, Q. F.; Iwasa, J.; Ogawa, D.; Ishido, Y.; Sato, S.; Ozeki, T.; Sei,
Y.; Yamaguchi, K.; Fujita, M. Science 2010, 328, 1144. (b) Sun, Q.-F.;
Murase, T.; Sato, S.; Fujita, M. Angew. Chem., Int. Ed. 2011, 50, 10318.
(c) Zheng, Y.-R.; Lan, W.-J.; Wang, M.; Cook, T. R.; Stang, P. J. J. Am.
Chem. Soc. 2011, 133, 17045. (d) Turega, S.; Whitehead, M.; Hall, B.
R.; Haddow, M. F.; Hunter, C. A.; Ward, M. D. Chem. Commun. 2012,
48, 2752. (e) Chepelin, O.; Ujma, J.; Barran, P. E.; Lusby, P. J. Angew.
Chem., Int. Ed. 2012, 124, 4270. (f) Bivaud, S.; Balandier, J.-Y.; Chas,
M.; Allain, M.; Goeb, S.; Salle, M. J. Am. Chem. Soc. 2012, 134, 11968.
́
(g) Riddell, I. A.; Smulders, M. M. J.; Clegg, J. K.; Hristova, Y. R.;
Breiner, B.; Thoburn, J. D.; Nitschke, J. R. Nat. Chem. 2012, 4, 751.
́
(h) Pasquale, S.; Sattin, S.; Escudero-Adan, E. C.; Martínez-Belmonte,
M.; de Mendoza, J. Nat. Commun. 2012, 3, 785.
(4) (a) Sessler, J. L.; Gale, P. A.; Cho, W.-S. Anion Receptor
Chemistry; RSC Publishing: Cambridge, U.K., 2006. (b) Custelcean,
R.; Bosano, J.; Bonnesen, P. V.; Kertesz, V.; Hay, B. P. Angew. Chem.,
Int. Ed. 2009, 48, 4025. (c) Thomas, J. A. Dalton Trans. 2011, 40,
12005. (d) Ma, Z.; Moulton, B. Coord. Chem. Rev. 2011, 255, 1623.
(5) Clever, G. H.; Kawamura, W.; Tashiro, S.; Shiro, M.; Shionoya,
M. Angew. Chem., Int. Ed. 2012, 51, 2606.
(6) Clever, G. H.; Tashiro, S.; Shionoya, M. J. Am. Chem. Soc. 2010,
132, 9973.
(7) Clever, G. H.; Kawamura, W.; Shionoya, M. Inorg. Chem. 2011,
50, 4689.
(8) Clever, G. H. In Molecules at Work; Pignataro, B., Ed.; Wiley-
VCH: Weinheim, Germany, 2012.
(9) Han, M.; Michel, R.; He, B.; Chen, Y.-S.; Stalke, D.; John, M.;
Clever, G. H. Angew. Chem., Int. Ed. 2013, 52, 1319.
(10) Other examples of interpenetrated coordination cages: (a) Fujita,
M.; Fujita, N.; Ogura, K.; Yamaguchi, K. Nature 1999, 400, 52.
(b) Yamauchi, Y.; Yoshizawa, M.; Fujita, M. J. Am. Chem. Soc. 2008,
130, 5832. (c) Fukuda, M.; Sekiya, R.; Kuroda, R. Angew. Chem., Int.
Ed. 2008, 47, 706. (d) Westcott, A.; Fisher, J.; Harding, L. P.;
Rizkallah, P.; Hardie, M. J. J. Am. Chem. Soc. 2008, 130, 2950.
(e) Heine, J.; Schmedt auf der Gunne, J.; Dehnen, S. J. Am. Chem. Soc.
̈
2011, 133, 10018. (f) Sekiya, R.; Fukuda, M.; Kuroda, R. J. Am. Chem.
Soc. 2012, 134, 10987.
(11) Charges and counteranions have been partially omitted in
formulas for simplicity.
́
(12) Freye, S.; Hey, J.; Torras-Galan, A.; Stalke, D.; Herbst-Irmer, R.;
John, M.; Clever, G. H. Angew. Chem., Int. Ed. 2012, 51, 2191.
(13) Freye, S.; Engelhard, D. M.; John, M.; Clever, G. H. Chem.
Eur. J. 2013, 19, 2114.
(14) Dieterich, J. M.; Clever, G. H.; Mata, R. A. Phys. Chem. Chem.
Phys. 2012, 14, 12746.
8479
dx.doi.org/10.1021/ja403184a | J. Am. Chem. Soc. 2013, 135, 8476−8479