Guest-encapsulation properties of a self-assembled capsule by dynamic boronic ester bonds

Article abstract of DOI:10.1021/ja9084918

Two molecules of tetrakis(dihydroxyboryl)-cavitand 1a as an aromatic cavity and four molecules of 1,2-bis(3,4-dihydroxyphenyl)ethane 2 as an equatorial linker self-assemble into capsule 3a via the formation of eight dynamic boronic ester bonds in CDCl

Full text of DOI:10.1021/ja9084918

Published on Web 12/15/2009
Guest-Encapsulation Properties of a Self-Assembled Capsule
by Dynamic Boronic Ester Bonds
Naoki Nishimura,† Kenji Yoza,‡ and Kenji Kobayashi*,†
Department of Chemistry, Faculty of Science, Shizuoka UniVersity, 836 Ohya, Suruga-ku,
Shizuoka 422-8529, Japan, and Bruker AXS, 3-9-B Moriya, Kanagawa-ku, Yokohama 221-0022, Japan
Received October 6, 2009; E-mail: skkobay@ipc.shizuoka.ac.jp
Abstract: Two molecules of tetrakis(dihydroxyboryl)-cavitand 1a as an aromatic cavity and four molecules
of 1,2-bis(3,4-dihydroxyphenyl)ethane 2 as an equatorial linker self-assemble into capsule 3a via the
formation of eight dynamic boronic ester bonds in CDCl3 or C6D6. Capsule 3a encapsulates one guest
molecule, such as 4,4′-disubstituted-biphenyl and 2,6-disubstituted-anthracene derivatives, in a highly
selective recognition event, wherein the guest substituents are oriented to both aromatic cavity ends of 3a,
1
as confirmed by a H NMR study and X-ray crystallographic analysis. Capsule 3a showed a significant
solvent effect on guest encapsulation. The association constant (Ka) of 3a with guests in C6D6 was much
greater than that in CDCl3 (450-48 000-fold). The encapsulation of guests within 3a in C6D6 was enthalpically
driven, whereas that in CDCl3 tended to be both enthalpically and entropically driven. Thermodynamic
studies suggest that the small Ka value in CDCl3 arises from the character of CDCl3 as a competitor guest
molecule for 3a, and not from the difference in stability of the boronic ester bonds of 3a in both solvents.
We propose a linker partial dissociation mechanism for the guest uptake and release into and out of 3a
based on the kinetic studies of guest@3a using 2D EXSY analysis, as well as structural analysis of a
guest@3b. The rotation behavior of 4,4′-diacetoxy-2,2′-disubstituted-biphenyls within 3a was also
investigated, where the elongation of 2,2′-disubstituents of guests put the brakes on guest rotation within
3a.
Introduction
namic control; namely, they combine both the strength of
covalent bonds and the reversibility of noncovalent interactions.6
Carcerands and hemicarcerands, in which two calix[4]resorcin-
arene cavitands are held together by four covalent linkages, have
been developed by Cram and others and have attracted
considerable attention from the viewpoint of stabilization of
reactive intermediates and as microvesicles for drug delivery,
given the confinement of guest molecules inside the capsules
away from a bulk phase.1 Error correction through thermody-
namic equilibration, minimization of synthetic effort by use of
modular subunits, and control of assembly processes through
subunit design are characteristics of supramolecular approaches
to self-assembly. Based on this concept, cavitand-based capsules
have been constructed under thermodynamic control using
noncovalent interactions, such as hydrogen bonds,2 metal-
coordination bonds,3 ionic interactions,4 and solvophobic in-
teractions.5 As an alternative strategy, dynamic covalent chem-
istry offers great advantages in supramolecular syntheses
because dynamic covalent bonds contain reversible covalent
bond-forming and bond-breaking processes under thermody-
The reversibility of the imine bond-forming reaction in the
presence of a catalytic amount of CF3CO2H has been applied
to cavitand-based capsule synthesis.7 The use of a redox buffer
has allowed a disulfide-linked cavitand capsule to form under
reversible conditions.8 Boronic ester formation is another reliable
synthon for dynamic covalent chemistry.9 Supramolecular
syntheses based on the reversible formation of boronic or
boronate esters, such as polymers,10 macrocycles,11 and cap-
sules,12 have been reported recently. Self-assembled architec-
tures using spiroborate formation have also been synthesized.13
However, the supramolecular synthesis of cavitand-based
capsules via the dynamic formation of boronic esters has not
been achieved so far. Recently, we have reported the self-
(2) (a) Chapman, R. G.; Sherman, J. C. J. Am. Chem. Soc. 1995, 117,
9081–9082. (b) MacGillivray, L. R.; Atwood, J. L. Nature 1997, 389,
469–472. (c) Heinz, T.; Rudkevich, D. M.; Rebek, J., Jr. Nature 1998,
394, 764–766. (d) Shivanyuk, A.; Paulus, E. F.; Bo¨hmer, V. Angew.
Chem., Int. Ed. 1999, 38, 2906–2909. (e) Kobayashi, K.; Shirasaka,
T.; Yamaguchi, K.; Sakamoto, S.; Horn, E.; Furukawa, N. Chem.
Commun. 2000, 41–42. (f) Ebbing, M. H. K.; Villa, M.-J.; Valpuesta,
J.-M.; Prados, P.; de Mendoza, J. Proc. Natl. Acad. Sci. U.S.A. 2002,
99, 4962–4966. (g) Kobayashi, K.; Ishii, K.; Sakamoto, S.; Shirasaka,
T.; Yamaguchi, K. J. Am. Chem. Soc. 2003, 125, 10615–10624. (h)
Sansone, F.; Baldini, L.; Casnati, A.; Chierici, E.; Faimani, G.;
Ugozzoli, F.; Ungaro, R. J. Am. Chem. Soc. 2004, 126, 6204–6205.
(i) Ajami, D.; Rebek, J., Jr. J. Am. Chem. Soc. 2006, 128, 5314–5315.
(j) Yamanaka, M.; Ishii, K.; Yamada, Y.; Kobayashi, K. J. Org. Chem.
2006, 71, 8800–8806. (k) Kitagawa, H.; Kobori, Y.; Yamanaka, M.;
Yoza, K.; Kobayashi, K. Proc. Natl. Acad. Sci. U.S.A. 2009, 106,
10444–10448.
† Shizuoka University.
‡ Bruker AXS.
(1) (a) Cram, D. J.; Cram, J. M. Container Molecules and Their Guests;
Royal Society of Chemistry: Cambridge, 1994. (b) Jasat, A.; Sherman,
J. C. Chem. ReV. 1999, 99, 931–967. (c) Warmuth, R.; Yoon, J. Acc.
Chem. Res. 2001, 34, 95–105. (d) Sherman, J. Chem. Commun. 2003,
1617–1623. (e) Barrett, E. S.; Irwin, J. L.; Edward, A. J.; Sherburn,
M. S. J. Am. Chem. Soc. 2004, 126, 16747–16749. (f) Makeiff, D. A.;
Sherman, J. C. J. Am. Chem. Soc. 2005, 127, 12363–12367. (g)
Warmuth, R.; Makowiec, S. J. Am. Chem. Soc. 2007, 129, 1233–
1241.
9
10.1021/ja9084918  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 777–790 777

A R T I C L E S
Nishimura et al.
Scheme 1. Formation of Capsule 3 from Cavitand Tetraboronic
assembly of tetrakis(dihydroxyboryl)-cavitand 1a as a polar
bowl-shaped aromatic cavity and 1,2-bis(3,4-dihydroxyphe-
nyl)ethane 2 as an equatorial bis(catechol)-linker14 into capsule
Acid 1 and Bis(catechol)-Linker 2a
(3) (a) Jacopozzi, P.; Dalcanale, E. Angew. Chem., Int. Ed. Engl. 1997,
36, 613–615. (b) Fox, O. D.; Dalley, N. K.; Harrison, R. G. J. Am.
Chem. Soc. 1998, 120, 7111–7112. (c) Ikeda, A.; Yoshimura, M.;
Udzu, H.; Fukuhara, C.; Shinkai, S. J. Am. Chem. Soc. 1999, 121,
4296–4297. (d) Fox, O. D.; Drew, M. G. B.; Beer, P. D. Angew. Chem.,
Int. Ed. 2000, 39, 136–140. (e) Pinalli, R.; Cristini, V.; Sottili, V.;
Geremia, S.; Campagnolo, M.; Caneschi, A.; Dalcanale, E. J. Am.
Chem. Soc. 2004, 126, 6516–6517. (f) Kobayashi, K.; Yamada, Y.;
Yamanaka, M.; Sei, Y.; Yamaguchi, K. J. Am. Chem. Soc. 2004, 126,
13896–13897. (g) Haino, T.; Kobayashi, M.; Chikaraishi, M.; Fuka-
zawa, Y. Chem. Commun. 2005, 2321–2323. (h) Park, S. J.; Shin,
D. M.; Sakamoto, S.; Yamaguchi, K.; Chung, Y. K.; Lah, M. S.; Hong,
J.-I. Chem.sEur. J. 2005, 11, 235–241. (i) Jude, H.; Sinclair, D. J.;
Das, N.; Sherburn, M. S.; Stang, P. J. J. Org. Chem. 2006, 71, 4155–
4163.
(4) (a) Corbellini, F.; Fiammengo, R.; Timmerman, P.; Crego-Calama,
M.; Versluis, K.; Heck, A. J. R.; Luyten, I.; Reinhoudt, D. N. J. Am.
Chem. Soc. 2002, 124, 6569–6575. (b) Oshovsky, G. V.; Reinhoudt,
D. N.; Verboom, W. J. Am. Chem. Soc. 2006, 128, 5270–5278.
(5) (a) Gibb, C. L.; Gibb, B. C. J. Am. Chem. Soc. 2004, 126, 11408–
11409. (b) Giles, M. D.; Liu, S.; Emanuel, R. L.; Gibb, B. C.; Grayson,
S. M. J. Am. Chem. Soc. 2008, 130, 14430–14431.
(6) (a) Lehn, J.-M. Chem.sEur. J. 1999, 5, 2455–2463. (b) Rowan, S. J.;
Cantrill, S. J.; Cousins, G. R. L.; Sanders, J. K. M.; Stoddart, J. F.
Angew. Chem., Int. Ed. 2002, 41, 898–952. (c) Severin, K.
Chem.sEur. J. 2004, 10, 2565–2580. (d) Corbett, P. T.; Leclaire, J.;
Vial, L.; West, K. R.; Wietor, J.-L.; Sanders, J. K. M.; Otto, S. Chem.
ReV. 2006, 106, 3652–3711. (e) Mal, P.; Schultz, D.; Beyer, K.;
Rissanen, K.; Nitschke, J. R. Angew. Chem., Int. Ed. 2008, 47, 8297–
8301.
a Molecular models of 3 and 4,4′-diacetoxybiphenyl encapsulating capsule
4@3 are calculated at the PM3 level of Spartan ′06,20 wherein the side
chains R of 3 are replaced by methyl groups.
3a via the dynamic formation of boronic esters (Scheme 1).15
The self-assembled cavitand-based capsule 3a encapsulates one
guest molecule, such as 4,4′-disubstituted-biphenyl. The merit
of 3a based on the dynamic boronic ester bond is the on/off
control of capsule formation with guest encapsulation by the
removal/addition of MeOH.15 Herein, we report on a compre-
hensive study on the guest-encapsulation properties of the self-
assembled capsule 3a, including X-ray crystal structure, guest-
encapsulation selectivity, remarkable solvent effect on guest
encapsulation, and thermodynamics and kinetics on guest en-
capsulation. We propose a linker partial dissociation mechanism
for guest encapsulation and release into and out of 3a. We also
describe guest-rotation behavior within 3a directed toward the
formation of a supramolecular gyroscope.
(7) (a) Ro, S.; Rowan, S. J.; Pease, A. R.; Cram, D. J.; Stoddart, J. F.
Org. Lett. 2000, 2, 2411–2414. (b) Liu, X.; Liu, Y.; Li, G.; Warmuth,
R. Angew. Chem., Int. Ed. 2006, 45, 901–904. (c) Liu, X.; Warmuth,
R. J. Am. Chem. Soc. 2006, 128, 14120–14127. (d) Liu, Y.; Liu, X.;
Warmuth, R. Chem.sEur. J. 2007, 13, 8953–8959. (e) Xu, D.;
Warmuth, R. J. Am. Chem. Soc. 2008, 130, 7520–7521. (f) Skowronek,
P.; Gawronski, J. Org. Lett. 2008, 10, 4755–4758.
(8) Sun, J.; Patrick, B. O.; Sherman, J. C. Tetrahedron 2009, 65, 7296–
7302.
(9) (a) Fujita, N.; Shinkai, S.; James, T. D. Chem. Asian J. 2008, 3, 1076–
1091. (b) Zhu, L.; Shabbir, S. H.; Gray, M.; Lynch, V. M.; Sorey, S.;
Anslyn, E. V. J. Am. Chem. Soc. 2006, 128, 1222–1232.
(10) For self-assembled polymers based on boronic ester formation, see:
(a) Coˆte´, A. P.; Benin, A. I.; Ockwig, N. W.; O’Keeffe, M.; Matzger,
A. J.; Yaghi, O. M. Science 2005, 310, 1166–1170. (b) Niu, W.;
O’Sullivan, C.; Rambo, B. M.; Smith, M. D.; Lavigne, J. J. Chem.
Commun. 2005, 4342–4344. (c) Tilford, R. W.; Gemmill, W. R.; zur
Loye, H.-C.; Lavigne, J. J. Chem. Mater. 2006, 18, 5296–5301. (d)
Niu, W.; Smith, M. D.; Lavigne, J. J. J. Am. Chem. Soc. 2006, 128,
16466–16467. (e) El-Kaderi, H. M.; Hunt, J. R.; Mendoza-Corte´s, J. L.;
Coˆte´, A. P.; Taylor, R. E.; O’Keeffe, M.; Yaghi, O. M. Science 2007,
316, 268–272. (f) Coˆte´, A. P.; El-Kaderi, H. M.; Furukawa, H.; Hunt,
J. R.; Yaghi, O. M. J. Am. Chem. Soc. 2007, 129, 12914–12915. (g)
Tilford, R. W.; Mugavero, S. J.; Pellechia, P. J.; Lavigne, J. J. AdV.
Mater. 2008, 20, 2741–2746. (h) Hunt, J. R.; Doonan, C. J.; LeVangie,
J. D.; Coˆte´, A. P.; Yaghi, O. M. J. Am. Chem. Soc. 2008, 130, 11872–
11873.
Results and Discussion
Synthesis of Self-Assembled Cavitand-Based Capsule 3a:
1H NMR Study. The cavitand tetraboronic acid 1a (side chain,
R ) (CH2)6CH3)15,16 and the bis(catechol)-linker 215 have low
solubility in CDCl3 by themselves. However, a 2:4 heteroge-
neous mixture of 1a and 2 in CDCl3 gives a homogeneous
solution on heating at 323 K for 3 h, and this quantitatively
(11) For self-assembled macrocycles based on boronic ester formation, see:
(a) Christinat, N.; Scopelliti, R.; Severin, K. Chem. Commun. 2004,
1158–1159. (b) Barba, V.; Villamil, R.; Luna, R.; Godoy-Alca´ntar,
C.; Ho¨pfl, H.; Beltram, H. I.; Zamudio-Rivera, L. S.; Santillan, R.;
Farfa´n, N. Inorg. Chem. 2006, 45, 2553–2561. (c) Christinat, N.;
Scopelliti, R.; Severin, K. J. Org. Chem. 2007, 72, 2192–2200. (d)
Christinat, N.; Scopelliti, R.; Severin, K. Angew. Chem., Int. Ed. 2008,
47, 1848–1852. (e) Hutin, M.; Bernardinelli, G.; Nitschke, J. R.
Chem.sEur. J. 2008, 14, 4585–4593.
produces capsule 3a (Scheme 1).15 The H NMR spectrum of
1
the reaction mixture shows a highly symmetrical single species
and the disappearance of the OH groups of the 1a and 2 units
(Figure 1c vs Figure 1a and 1b), indicating the quantitative
1
formation of 3a. The H NMR chemical shift change of 3a
relative to 1a and 2, ∆δ (δcomplex - δfree), is 0.14, -0.20, 0.10,
(12) For, self-assembled capsules based on boronic or boronate ester
formation, see: (a) Iwasawa, N.; Takahagi, H. J. Am. Chem. Soc. 2007,
129, 7754–7755. (b) Kataoka, K.; James, T. D.; Kubo, Y. J. Am. Chem.
Soc. 2007, 129, 15126–15127. (c) Ic¸li, B.; Christinat, N.; To¨nnemann,
J.; Schu¨ttler, C.; Scopelliti, R.; Severin, K. J. Am. Chem. Soc. 2009,
131, 3154–3155.
(14) For catechol-based metal ion assisted self-assembly of capsules, see:
(a) Schalley, C. A.; Lu¨tzen, A.; Albrecht, M. Chem.sEur. J. 2004,
10, 1072–1080. (b) Fiedler, D.; Leung, D. H.; Bergman, R. G.;
Raymond, K. N. Acc. Chem. Res. 2005, 38, 351–360.
(15) Nishimura, N.; Kobayashi, K. Angew. Chem., Int. Ed. 2008, 47, 6255–
6258.
(13) For self-assemblies based on spiroborate, see: (a) Katagiri, H.;
Miyagawa, T.; Furusho, Y.; Yashima, E. Angew. Chem., Int. Ed. 2006,
45, 1741–1744. (b) Danjo, H.; Hirata, K.; Yoshigai, S.; Azumaya, I.;
Yamaguchi, K. J. Am. Chem. Soc. 2009, 131, 1638–1639.
(16) Sherman, J. C.; Knobler, C. B.; Cram, D. J. J. Am. Chem. Soc. 1991,
113, 2194–2204.
9
778 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Encapsulation Properties of a Self-Assembled Capsule
A R T I C L E S
This guest orientation is consistent with that determined
previously by the 1H NMR analysis of 4@3a in solution.15 The
close C(dO)CH3 · · ·Cπ contact distance between the acetoxy
protons of 4 and the aromatic cavity of 3b is ca. 2.80 Å. The
carbonyl oxygen atoms of the acetoxy groups interact with inner
protons of the methylene-bridge rims (O-CHinHout-O) of 3b
with CdO· · ·HinC distances of 2.53 and 2.71 Å.2k
A molecular model of 3 (side chain of 1 unit, R ) CH3),
calculated using the Spartan ′06 software package at the PM3
level,20 suggests that four linker 2 units adopt an anti-
conformation, and the northern polar 1 unit is twisted by ca.
12.4° with respect to the southern polar 1 unit about a C4 polar
axis (Scheme 1).15 Namely, capsule 3 may exist as a 1:1 mixture
of the twisted (M)- and (P)-enantiomers (twistomers),21,22 when
four linkers with anti-conformation are arranged in the same
direction. However, this is not the case for the actual crystal
structure of 4@3b. All linker 2 units adopted an approximate
anti-conformation, but they were highly disordered, even at 100
K, and were not necessarily arranged in the same direction
(Figure 2c-2e).
In Figure 2c-2e, the occupancy factor (occ-f) of the B(1)
atom was occ-f ) 1.00, but the other three boron atoms were
observed at two positions, wherein both the B(2) and B(3) atoms
showed occ-f ) 0.796 (orange colored atoms) and occ-f ) 0.204
(blue colored atoms) and the B(4) atom showed occ-f ) 0.50
each (orange and green colored atoms). The two linkers
connecting the B(2) and B(3) atoms with the major occ-f, which
are placed diagonally across from each other, are arranged in
the opposite direction to produce a meso-form. The two linkers
connecting the B(1) and B(4) atoms can be arranged in the same
or opposite directions to give enantio- or meso-forms, respec-
tively. Thus, the four linker 2 units in 3b behave as semimobile
linkers. As a result, in the crystal structure of 4@3b, the northern
and southern polar 1b units were not twisted but were slipped
ca. 0.8 Å relative to each other.22b,23 In the crystal structure,
3b exists as an average form of a mixture of the (M)- and (P)-
twistomers and slipped-forms (Figure 2f) and possesses an inner
cavity with an approximate dimension of 8.5 × 19.9 Å2
(interatomic distances) and four equatorial portals with a
dimension of approximately 6.7 × 11.9 Å2. These dimensions
are almost the same as those in the calculated twisted form of
3.15 Interconversion between the (M)- and (P)-twistomers and
slipped forms in 3a is very fast in solution, and their confor-
mational isomers are observed as an average form.
Figure 1. 1H NMR spectra (400 MHz, 296 K). In CDCl3, (a) [2] ) 10
mM if soluble (heterogeneous), (b) [1a] ) 5 mM if soluble, and (c) capsule
3a (homogeneous after heating [1a] ) 5 mM and [2] ) 10 mM at 50 °C
for 3 h. In C6D6, (d) [2] ) 10 mM if soluble (heterogeneous), (e) [1a] )
5 mM if soluble, and (f) capsule 3a (heterogeneous after heating [1a] ) 5
mM and [2] ) 10 mM at 50 °C for 3 h). The signals marked ‘a-i’ are
assigned in Scheme 1. The signals marked ‘s’ are the spinning sidebands
of the residual solvent.
and 0.12 ppm for Ha-Hd of the 1a unit, respectively, and 0.42,
0.49, 0.45, 0.20 ppm for Hf-Hi of the 2 unit, respectively.
Heating a 2:4 mixture of 1a and 2 in C6D6 at 323 K for 3 h
does not give a homogeneous system, although 3a was formed
quantitatively (Figure 1f). The solubility of 3a in C6D6 at 296
1
K was 1.3 mM based on the H NMR integration ratio with
p-dimethoxybenzene as an internal standard. The ∆δ values
were 0.59, 0.02, 0.08, and 0.19 ppm for Ha-Hd of the 1a unit,
respectively, and 0.10, 0.46, 0.11, and -0.35 ppm for Hf-Hi
of the 2 unit, respectively. In contrast, heating this mixture in
the presence of an appropriate guest gives a homogeneous
solution and quantitatively produces the guest-encapsulating
capsule, guest@3a (vide infra).15
X-ray Crystal Structure of 4@3b. Single crystals of 4@3b
suitable for X-ray diffraction analysis were obtained by slow
evaporation of a benzene solution of a 2:4:2 mixture of cavitand
tetraboronic acid 1b (side chain, R ) (CH2)2Ph), bis(catechol)-
linker 2, and 4,4′-diacetoxybiphenyl 4 at room temperature after
heating this mixture at 323 K for 3 h. The single-crystal X-ray
diffraction measurement was performed at 100 K.17 As shown
in Figure 2a and 2b, two molecules of 1b at the polar positions
are indirectly held together by eight boronic ester bonds
involving four molecules of 2 at the equatorial positions to self-
assemble into capsule 3b. Exactly one molecule of 4 is
encapsulated in the cavity of 3b, and the acetoxy groups at the
4,4′-positions of 4 are oriented toward the two aromatic cavity
ends of 3b to maximize the guest-capsule CH-π interaction.18,19
IR Study of Guest@3a. The IR spectrum of 4@3a (KBr) is
characterized by the disappearance of the OH groups from 1a
and 2, showing the formation of the boronic ester.17 The CdO
stretching band of the acetoxy group of 4@3a appears at ν )
(19) (a) Kobayashi, K.; Kitagawa, R.; Yamada, Y.; Yamanaka, M.;
Suematsu, T.; Sei, Y.; Yamaguchi, K. J. Org. Chem. 2007, 72, 3242–
3246. (b) Kitagawa, H.; Kawahata, M.; Kitagawa, R.; Yamada, Y.;
Yamanaka, M.; Yamaguchi, K.; Kobayashi, K. Tetrahedron 2009, 65,
7234–7239.
(20) Spartan ′06; Wavefunction Inc.: Irvine, CA, USA, 2006.
(21) (a) Chapman, R. G.; Sherman, J. C. J. Am. Chem. Soc. 1999, 121,
1962–1963. (b) Paek, K.; Ihm, H.; Yun, S.; Lee, H. C. Tetrahedron
Lett. 1999, 40, 8905–8909. (c) Chapman, R. G.; Sherman, J. C. J.
Org. Chem. 2000, 65, 513–516.
(22) (a) Cram, D. J.; Blanda, M. T.; Paek, K.; Knobler, C. B. J. Am. Chem.
Soc. 1992, 114, 7765–7773. (b) Robbins, T. A.; Knobler, C. B.; Bellew,
D. R.; Cram, D. J. J. Am. Chem. Soc. 1994, 116, 111–122.
(23) (a) v. d. Bussche-Hu¨nnefeld, C.; Bu¨hring, D.; Knobler, C. B.; Cram,
D. J. J. Chem. Soc., Chem. Commun. 1995, 108, 5–1087. (b) Helgeson,
R. C.; Paek, K.; Knobler, C. B.; Maverick, E. F.; Cram, D. J. J. Am.
Chem. Soc. 1996, 118, 5590–5604. (c) Helgeson, R. C.; Knobler, C. B.;
Cram, D. J. J. Am. Chem. Soc. 1997, 119, 3229–3244.
(17) See the Supporting Information.
(18) Kobayashi, K.; Ishii, K.; Yamanaka, M. Chem.sEur. J. 2005, 11,
4725–4734.
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 779

A R T I C L E S
Nishimura et al.
Figure 2. X-ray crystal structure of 4@3b at 100 K. One of molecular structures of 4@3b: (a) front view and (b) its 45°-rotation view, wherein two linkers
with occupancy factor (occ-f) of 0.796 are placed at the diagonal position and two linkers (one side) with occ-f of 0.500 are placed at the other diagonal
position. Molecular structures of 3b containing all disordered linkers: (c) front view, (d) side view, and (e) top view, wherein the linkers with B(2) and B(3)
are shown in gray (occ-f ) 0.796) and blue (occ-f ) 0.204) and the half-linkers with B(4) are shown in gray and green (occ-f ) 0.50 each). (f) Schematic
representation of conformational isomers of 3b. In Figure 2c-f, the side chains of 3b and encapsulated 4 are omitted for clarity.
excess 8. This result indicates that the exchange of 8 into and
out of 3a is slow on the NMR time scale. The 1H NMR signals
of 8 encapsulated in 3a were shifted upfield by 3.99 ppm for
the acetoxy protons and 0.61 ppm for the methyl ester protons
relative to those of free 8; namely, in 8@3a the acetoxy protons
are shifted much more upfield than the methyl ester pro-
tons are. This result clearly shows that the acetoxy groups at
the 4,4′-positions are oriented toward both polar aromatic cavity
ends of 3a and the methyl ester groups at the 2,2′-positions are
directed toward the equatorial portals of 3a (Figure 4). This
guest orientation in solution is in good agreement with that
shown in the X-ray crystal structure of 4@3b. Based on the
integration ratio of the 1H NMR signals, capsule 3a encapsulates
one molecule of 8, and the association constant (Ka) of 3a with
8 was estimated to be Ka ) 44 M-1 in CDCl3 at 298 K.
Heating a 2:4:2 mixture of 1a, 2, and 8 or a 1:2 mixture of
Figure 3. Association behavior of capsule 3a with guest 8 monitored by
as-prepared 3a and 8 ([8] ) 1 mM) in C6D6 at 323 K for 3 h
gave a 1:1 mixture of 8@3a and 8 (Figure 3b), wherein the
guest exchange was slow on the NMR time scale. No guest-
free 3a was observed, indicating a very large value of Ka in
1H NMR (400 MHz): (a) [3a] ) 2.5 mM and [8] ) 12.5 mM in CDCl3 at
298 K and (b) [3a] ) 0.5 mM and [8] ) 1.0 mM in C6D6 at 313 K. The
solid circle, open circle, solid square, and open square indicate the
representative signals of 3a of 8@3a, free 3a, 8 of 8@3a, and free 8,
respectively.
1
C6D6 compared with the value of Ka in CDCl3. The H NMR
signals of 8@3a in C6D6 were shifted upfield by 2.81 ppm for
the acetoxy protons and 0.17 ppm for the methyl ester protons
relative to those of free 8. The relatively smaller change in
1761 cm-1, which is shifted to a higher wavenumber by 16 cm-1
relative to the spectrum of 4 alone. In 11@3a, where 11 is 2,6-
diacetoxy-9,10-anthraquinone, the CdO stretching bands of the
acetoxy group and quinone moiety are shifted to higher
wavenumbers by 13 and 4 cm-1, respectively, relative to the
spectrum of 11 alone.17 These results suggest that the acetoxy
groups are constricted by both aromatic cavity ends of 3a.
1H NMR Study of Guest@3a: Guest-Encapsulation Selectiv-
ity and Solvent Effect. A 1H NMR study for the encapsulation
of 4 in 3a has been reported previously.15 Figure 3 shows the
1H NMR spectra for the encapsulation of 4,4′-diacetoxy-2,2′-
bis(methoxycarbonyl)biphenyl 8 within 3a. When 5 equiv of 8
was added to 3a (2.5 mM) in CDCl3, three species were
independently observed after heating (Figure 3a), i.e., 34%
guest-encapsulating capsule 8@3a, 66% guest-free 3a, and
(24) The 1H NMR chemical shifts (δ) of the acetoxy and methyl ester
protons of free 8 and 8@3a were as follows: for free 8, 2.33 (acetoxy)
and 3.63 ppm (methyl ester) in CDCl3 and 1.70 and 3.22 ppm in C6D6;
for 8@3a,-1.66 and 3.02 ppm in CDCl3 and-1.11 and 3.05 ppm in
C6D6. Thus, ∆δG (δencapsulated-guest- δfree-guest) values of 8 in C6D6 showed
less upfield shifts by ca. 0.4-1.2 ppm relative to those in CDCl3. This
result can be interpreted as follows. (1) Free 8 is solvated by C6D6
and is subject to the ring current effect of C6D6-solvation. (2) Upon
encapsulation of 8 in 3a, desolvation of C6D6 from 8 occurs, and the
8 encapsulated in 3a is no longer subject to the ring current effect of
C6D6-solvation. (3) Instead, the 8 encapsulated in 3a is subject to the
ring current effect of the aromatic cavity of 3a. Consequently, the
∆δG values of 8 in C6D6 are cancelled and seemingly shifted less
upfield relative to those in CDCl3.
9
780 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Encapsulation Properties of a Self-Assembled Capsule
A R T I C L E S
Item 1. The values of ∆δG of 5-7, 9-14, and 17-19 were
almost the same as those of 4 and 8, mentioned above. This
result shows that all guests encapsulated in 3a are oriented with
the guest long axis aligned along the long axis of capsule 3a.
Item 2. The value of Ka of guest@3a in C6D6 increases in
the order 20 < 16 < 18 < 17 < 19 < 14 < 15 < 6 < 7 < 5 < 13
< 10 < 9 < 8 < 4 < 12 < 11, whereas the value of Ka of guest@3a
in CDCl3 increases in the order 6 < 10 < 13 e 15 < 9 < 5 < 12
< 8 < 7 < 4 < 11. In CDCl3, encapsulation of 14 and 16-20
was scarcely detected. Although the guest-binding order of 3a
in C6D6 is not necessarily consistent with that in CDCl3, it is
noteworthy that the value of Ka of guest@3a in C6D6 is 2 to 4
orders of magnitude greater than the value of Ka of guest@3a
in CDCl3 (Ka(in C6D6)/Ka(in CDCl3) ) 450-(4.8 × 104)). Thus,
3a shows a significant solvent effect on guest encapsulation.
Item 3. Capsule 3a strictly discriminates between functional
groups of a guest. In a series of 4,4′-disubstituted-biphenyl
derivatives with a similar molecular length, the value of Ka of
guest@3a in C6D6 increases in the order 17 (4,4′-C(dO)OCH3,
Ka ) 567 M-1) , 6 (4,4′-OCH2CH3, Ka ) 1.30 × 104 M-1) <
5 (4-OC(dO)CH3-4′-OCH2CH3, Ka ) 1.24 × 105 M-1) < 4
(4,4′-OC(dO)CH3, Ka ) 1.26 × 106 M-1). This selectivity arises
from a combination of CHsπ and CH · · · OdC interactions,
which were found in the X-ray crystal structure of 4@3b (Figure
2a and 2b). The electrostatic potential repulsion between the
lone pair electron of the carbonyl oxygen atom of the ester group
of 17, which is inwardly directed toward the aromatic cavity of
3a, and the electron-rich aromatic cavity of 3a would be one
of the reasons for the low encapsulation ability of 17 in 3a
compared with 4-6.19a
Figure 4. Molecular model of 8@3 calculated with Spartan ′06 at the PM3
level:20 (a) front, (b) 45°-rotation, and (c) top views. (d) Space-filling
representations of guest-free 3 and 8. The side chains of 3 are replaced by
methyl groups.
chemical shift in C6D6 than that in CDCl3 upon encapsulation
of 8 in 3a is the result of a solvent effect.24
Item 4. In a series of 4,4′-disubstituted-diphenylacetylene
derivatives, the value of Ka of guest@3a in C6D6 increases in
the order 16 (4,4′-Br, Ka ) 245 M-1) < 14 (4,4′-OCH3, Ka )
4.57 × 103 M-1) < 15 (4,4′-I, Ka ) 8.71 × 103 M-1). This
result indicates that the CH · · ·I interaction between the polarized
inner proton of the methylene-bridge rim (O-CHinHout-O) of
3a and the δ(-) equatorial region of the iodo atom in the C-I
bond of 15 and the I · · ·π interaction between the δ(+) polar
region of the iodo group and the aromatic cavity of 3a are
effective for guest encapsulation.18,19 The change in the 1H NMR
chemical shift (∆δ) of the inner proton of the methylene-bridge
rim of guest@3a relative to that of free 3a in C6D6 (CDCl3)
was ∆δ ) 0.40 ppm (0.15 ppm) for 15 based on a CH · · · I
interaction, ∆δ ) 0.26 (0.04) for 4 based on a CH · · · OdC
interaction, and ∆δ ) 0.17 (-0.08) for 6.
Item 5. Capsule 3a strictly discriminates between a difference
of one or two carbon atoms in guest size. In contrast to 5, 6, and
15 encapsulated in 3a, guests 25, 21-24, and 26 were scarcely
encapsulated, respectively. In a series of 4,4′-dialkynyl-biphenyl
derivatives, the value of Ka of guest@3a in C6D6 increases in the
order 20 < 18 < 19. Thus, a fit between the guest and the cavity of
3a in size is essential for encapsulation.3g,18,19,26
Guest molecules 4-20 encapsulated in 3a are shown in Chart
1. In all cases, except for 20, the exchange of guests into and
out of 3a was slow on the NMR time scale in both CDCl3 and
C6D6, as shown in Figure 3 for 8@3a.17 Thus, using the H
1
NMR integration change in signal of guest@3a as a function
of the guest concentration, the association constant (Ka) of 3a
with guests can be estimated.25 When the value of Ka is very
large, a comparison of the signal integration between guest@3a
and standard-guest@3a, namely, competitive encapsulation
experiments, can be used to evaluate the guest-encapsulation
ability of 3a.17 In all cases, guest exchanges between guest-
A@3a and guest-B were fast on the human time scale and
reached equilibration within a period of 5-10 min in both
CDCl3 and C6D6 at room temperature, but these are very slow
on the NMR time scale, and the 1H NMR signals of guest@3a
and standard-guest@3a were independently observed. Thus,
before the measurements of Ka, a mixture of as-prepared 3a,
guest-A, and guest-B (standard-guest) in CDCl3 and C6D6 was
heated at 323 K for 1 h in competitive encapsulation experi-
ments. The value of Ka of 3a with guests in CDCl3 at 298 K
1
and in C6D6 at 313 K and the change in H NMR chemical
shift of guest@3a relative to the free guests at the terminal
functional groups (∆δG ) δencapsulated-guest - δfree-guest) are
summarized in Table 1. For comparison, the previously reported
data for 4, 5, 6, and 12 are also shown in Table 1.15 The
following features (Items 1-5) are noteworthy concerning the
guest encapsulation in 3a.
Thermodynamics of Guest@3a and the Solvent Effect. The
thermodynamics of guest@3a in CDCl3 and C6D6 were studied
to elucidate the reasons for the difference in Ka of guest@3a in
CDCl3 and C6D6. Based on the value of Ka as a function of the
temperature, the van’t Hoff plots gave the thermodynamic
parameters of the enthalpic (∆H°) and entropic (∆S°) contribu-
tions for guest encapsulation in 3a.17 The results are summarized
in Table 2.27 The following features (Items 1-4) are noteworthy
regarding the thermodynamics of guest@3a.
(25) The exchange of 20 in and out of 3a was fast on the NMR time scale
in C6D6, and the signals were observed as averaged forms. Thus, the
Ka of 3a with 20 in C6D6 at 313 K was calculated from a nonlinear
curve fitting, wherein [3a] ) 2.5 mM (constant) and [20] ) 2.0-
10.0 mM (variable).
(26) Mecozzi, S.; Rebek, J., Jr. Chem.sEur. J. 1998, 4, 1016–1022.
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 781

A R T I C L E S
Nishimura et al.
Chart 1. Encapsulation and No Encapsulation Guests in 3a
Table 1. Association Constants (Ka) of 3a with Guests in CDCl3
Table 2. Thermodynamic Parameters for the Encapsulation of
a
and C6D6,a and the H NMR Chemical Shift Changes (∆δG) of
Guests within 3a in CDCl3 and C6D6
1
Guest@3a Relative to Free Guests at the Terminal Functional
Groupsb
∆H°
∆S°
∆G°(298K)
entry guest
solvent
(kcal mol-1
)
(cal mol-1 K-1
)
(kcal mol-1
)
Ka (M-1
)
Ka (M-1
)
Ka(in C6D6)/
1
2
3
4
5
6
7
8
4
8
11
11
7
7
6
CDCl3
CDCl3
CDCl3
sat. D2O-CDCl3
CDCl3
C6D6
C6D6
-1.87
1.31
2.25
11.92
4.70
-2.54
-2.24
-3.17
-3.16
-2.41
-6.99
-6.91
-5.66
guest in CDCl3 at 298 K ∆δG (ppm) in C6D6 at 313 K ∆δG (ppm) Ka(in CDCl3)
4c d
81.7
33.0
-4.04
1.26 × 106
-2.85
1.54 × 104
,
-1.77
-0.97
-2.69
-18.60
-26.33
-14.14
5c
-4.08e 1.24 × 105
-2.86e 3.76 × 103
7.36
-3.80f
-2.82f
-0.93
-38.97
-65.18
-28.47
6c
7
8.1
-3.84
-4.04
-3.99
-4.00
-3.99
-4.07
-4.11
-4.05
1.30 × 104
6.14 × 104
5.42 × 105
4.75 × 105
3.41 × 105
4.25 × 106
1.83 × 106
2.12 × 105
4.57 × 103
-2.87
-2.99
-2.81
-2.79
-2.80
-2.73
-2.88
-2.88
-2.20
1.61 × 103
1.16 × 103
1.23 × 104
2.14 × 104
2.54 × 104
1.90 × 104
4.78 × 104
1.16 × 104
53.1
44.2
22.2
13.4
224
38.3
18.2
ndg
8
9
14
C6D6
10
11
12c
13
14
15
16
17
18
19
a ∆H° and ∆S° were obtained by van’t Hoff plots: ln Ka ) -(∆H°/
R)(1/T) + (∆S°/R). ∆G°(298K) ) ∆H° - T∆S°, wherein T ) 298 K.
Item 2. In a series of 4,4′-diacetoxybiphenyl derivatives with
substituents at the 2,2′-positions in CDCl3, the encapsulation
of the relatively more polar guest 8 (2,2′-CO2CH3, entry 2)
showed a major entropic contribution, whereas the encapsulation
of the relatively less polar 7 (2,2′-CH3, entry 5) indicated nearly
an enthalpic contribution, and the encapsulation of the medium
polarity 4 (2,2′-H, entry 1) exhibited both enthalpic and entropic
contributions. Upon guest encapsulation, such as 8 in 3a, a
CDCl3-solvated guest would release CDCl3 molecules giving
rise to an enhancement of the entropic contribution.22a,28
19.3
-0.33h 8.71 × 103
0.83h 450
ndg
245
1.14h
-2.91
-3.18
-2.98i
-3.02j
ndg
567
ndg
507
ndg
2.71 × 103
k
20
ndg
174k
-
a Ka ) [G@3a]/{[3a]f[G]f} or Ka/Ka-Standard ) {[G@3a][GStandard]f}/
{[GStandard@3a][G]f}. b ∆δG ) δencapsulated-guest - δfree-guest c See ref 15.
.
d The Ka of 3a with 4 in CD2Cl2 at 298 K was 2.50 × 103 M-1 with
∆δG ) -4.00 ppm. e Acetoxy group. f Ethoxy group. g Nd is not
detected due to too small Ka. h Aromatic protons at 3(3′) and 5(5′)
positions. i Propynyl group. j Ethynyl group. k See ref 25.
(27) To minimize experimental errors, in CDCl3, guests 4, 7, 8, and 11
were chosen, which showed large Ka values (44-224 M-1 at 298 K)
among a series of guests@3a in CDCl3. On the other hand, in C6D6,
guests 6, 7, and 14 with moderate Ka values (4.6 × 103-6.1 × 104
M-1 at 313 K) were chosen among a series of guests@3a in C6D6. In
C6D6, guests 4, 8, and 11 showed too large Ka values, and their Ka
could not be determined directly but just by competitive encapsulation
experiments.
Item 1. The encapsulation of guests within 3a in C6D6 are
enthalpically driven with large values of Ka (entries 6-8),
whereas the encapsulation of guests in CDCl3 tended to be both
enthalpically and entropically driven with small values of Ka
(entries 1-5).
9
782 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Encapsulation Properties of a Self-Assembled Capsule
A R T I C L E S
Table 3. Molar Ratio of 28 and 30 at 298 K in the Reaction
Mixture after Heating a 1:1 Mixture of 28 and 29 To Form 30 in
CDCl3 and C6D6 at 323 K for 12 ha
entry
Ar
solvent
28:30
1
2
3
4
5
6
7
8
a
a
a
a
b
b
b
b
CDCl3
C6D6
2.0:98.0
5.9:94.1
9.5:90.5
15.1:84.9
72.7:27.3
83.9:16.1
98.9:1.1
97.3:2.7
CDCl3-D2Ob
C6D6-D2Ob
CDCl3
Figure 5. Plot of the enthalpy vs entropy values for the encapsulation of
guests within 3a in CDCl3 and C6D6 based on the data of Table 2: (A) 4 in
CDCl3, (B) 8 in CDCl3, (C) 11 in CDCl3, (C′) 11 in sat. D2O-CDCl3, (D)
7 in CDCl3, (D′) 7 in C6D6, (E) 6 in C6D6, and (F) 14 in C6D6.
C6D6
CDCl3-D2Ob
C6D6-D2Ob
a [28]i ) [29]i ) 17 mM in solvent (600 µL).31 b To a solution of the
resulting reaction mixture (entries 1, 2, 5, and 6) were added 5 µL of
D2O (27 equiv), and the mixture was heated again at 323 K for 12 h.
However, the value of Ka of 8 with 3a was not greater than
that of 7 with 3a in CDCl3. This result suggests that CDCl3
solvation of a guest would not contribute strongly to an increase
of Ka.
Item 3. Guest encapsulation in D2O-saturated CDCl3 enhances
the trend toward entropic contribution compared with that in
CDCl3 (entry 4 vs entry 3). However, under both conditions,
the value of Ka was almost the same.
Item 4. Based on the thermodynamic parameters for the
encapsulation of guests within 3a in CDCl3 and C6D6 shown in
Table 2, the encapsulation process shows an enthalpy-entropy
compensation effect (Figure 5).28
is larger than CD2Cl2 in size, and the CD · · ·π and Cl · · ·π
interactions18,19 between the solvent and the aromatic cavity of
3a would be more favorable for CDCl3 than for CD2Cl2 based
on the difference in the number of electronegative Cl atoms.
Solvent Effect on the Thermodynamic Stability of Arylbo-
ronic Acid Catechol Esters. As a model for studying the
thermodynamic stability of the boronic ester bond of 3a, the
reaction of arylboronic acids 28 with catechol 29 to produce
the arylboronic acid catechol esters 30 was investigated in CDCl3
and C6D6. The results are summarized in Table 3.
The small value of Ka and considerable entropic contribution
for guest@3a in CDCl3 suggest that CDCl3 acts as a guest
competitor for 3a. Several molecules of CDCl3 may be
encapsulated in 3a or may interact with the inner aromatic cavity
of 3a,29 and upon encapsulation of guest into 3a, the encapsu-
lated CDCl3 molecules would be released from the cavity of
3a to give a positive ∆S°.30 When pure 4@3a, prepared in
benzene and dried in vacuo, was dissolved in CDCl3, the release
of 4 from 4@3a into the bulk phase of CDCl3 occurred
immediately at 296 K. The ratio of 4@3a:free-3a was 60:40
after 3 min and 48:52 after 24 h, reaching an equilibrium.15 In
C6D6, which is less polar than CDCl3, noncovalent interactions
between the guests and the inner cavity of capsule 3a, such as
CHsπ and CH · · ·OdC interactions, would work more ef-
fectively, leading to a significant enthalpic contribution and a
large value of Ka in C6D6. The value of Ka of as-prepared 3a
with 4 in CD2Cl2 was compared with the value of Ka in CDCl3
and C6D6.17 The value of Ka of 3a with 4 in solvents increased
in the order CDCl3 (82 M-1 at 298 K) < CD2Cl2 (2.50 × 103
M-1 at 298 K) < C6D6 (1.26 × 106 M-1 at 313 K), whereas the
dielectric constant increases in the order C6H6 (2.28) < CHCl3
(4.81) < CH2Cl2 (8.93). This result also indirectly supports the
assertion that CDCl3 acts as a guest competitor for 3a. CDCl3
The 1H NMR spectra of the reaction mixture of 4-methylphe-
nylboronic acid 28a with 29 (17 mM each) in CDCl3 and C6D6
at 298 K after heating at 323 K for 12 h showed a mixture of
28a and its catechol ester 30a in ratios of 2.0:98.0 and 5.9:
94.1, respectively.17,31 Upon addition of 5 µL (27 equiv) of D2O
to these mixtures and then reheating at 323 K for 12 h, the
equilibrium ratio of 28a and 30a shifted to 9.5:90.5 in CDCl3
and 15.1:84.9 in C6D6 at 298 K. On the other hand, the reaction
of 2,6-dimethoxyphenylboronic acid 28b with 29 under the same
conditions gave a mixture of 28b and 30b in a ratio of 72.7:
27.3 in CDCl3 and 83.9:16.1 in C6D6.17.31 On addition of D2O
and reheating under the same conditions, 30b was almost
hydrolyzed back into 28b in both CDCl3 and C6D6. These results
clearly indicate that (1) the arylboronic acid catechol esters 30a
and 30b are thermodynamically more stable in CDCl3 than in
C6D6 and (2) 30b is less stable than 30a is, which is probably
the result of the near orthogonal conformation between the aryl
group and the catecholatoboryl plane of 30b with respect to
the coplanar conformation of 30a.
Thus, the fact that arylboronic acid catechol esters 30 are
thermodynamically more stable in CDCl3 than in C6D6 also
1
(31) The H NMR spectra of 28a (17 mM) in CDCl3 and C6D6 at 298 K
(28) Leung, D. H.; Bergman, R. G.; Raymond, K. N. J. Am. Chem. Soc.
2008, 130, 2798–2805.
showed a mixture of 28a and its boroxine (trimeric anhydride of 28a)
in ratios of 36.8:63.2 and 61.1:38.9, respectively, after 30 min of
dissolution.17 After reaching an equilibrium in 36 h at 298 K, the ratio
of 28a and its boroxine was 43.5:56.5 in CDCl3 and 60.0:40.0 in C6D6.
This result indicates that the boroxine is thermodynamically more
stable in CDCl3 than in C6D6. Upon addition of 29 to this mixture,
the boroxine almost disappeared to produce 30a concomitantly with
28a. In contrast, 28b (17 mM) remained intact in both CDCl3 and
C6D6 and its boroxine was not detected even after heating at 323 K
for 12 h.17
1
(29) In the H NMR spectra of 3a in CHCl3 even at 218 K with CDCl3
sealed in a capillary for NMR deuterium-locking or a mixture of 3a
and 100 equiv of CHCl3 in C6D6, no signal of CHCl3 encapsulated in
3a was observed. These results suggest a small Ka and very fast
exchange of CHCl3 in and out of 3a on the NMR time scale.
(30) (a) Kang, J.; Rebek, J., Jr. Nature 1996, 382, 239–241. (b) Yamanaka,
M.; Shivanyuk, A.; Rebek, J., Jr. J. Am. Chem. Soc. 2004, 126, 2939–
2943.
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 783

A R T I C L E S
Nishimura et al.
Figure 6. Schematic representations of possible mechanisms for the guest
encapsulation and release into and out of 3a: (a) an extended portal
mechanism, (b) a linker dissociation mechanism, and (c) a linker partial
dissociation mechanism. (d) Conformational change of the two linkers
connected with the two cavitands from anti to eclipsed-like conformation.
indirectly supports that the small Ka value for guest@3a in
CDCl3 compared with that in C6D6 does not arise from the
difference in stability of the boronic ester bonds of 3a in CDCl3
and C6D6, but probably from the character of CDCl3 as a guest
competitor for 3a.
Mechanism of Guest Encapsulation in 3a: A Kinetic Study.
Capsule 3a is composed of two molecules of tetrakis(dihy-
droxyboryl)-cavitand 1a as a polar aromatic cavity and four
molecules of 1,2-bis(3,4-dihydroxyphenyl)ethane 2 as an equa-
torial bis(catechol)-linker connected by eight dynamic boronic
ester bonds. Three mechanisms are possible for guest encapsula-
tion in capsule 3a (Figure 6).
Figure 7. Variable temperature 2D NOESY spectra of a mixture of as-
prepared 3a, 8, and free 2 in CDCl3 under the conditions of [3a] ) 2.5
mM, [8] ) 12.5 mM, and [2 (saturated)] ) 0.94 mM, pulse delay ) 3.0 s,
and mixing time ) 1.0 s: (a) at 298 K and (b) at 323 K. The representative
signals of 8 of 8@3a and free 8 are marked with solid and open squares,
respectively. The signals of free 2 are marked with asterisks.
The first mechanism is an extended portal mechanism based
on a conformational change of the linker 2 unit (Figure 6a).22a
However, the following facts rule out this mechanism. In the
crystal structure of 4@3b, all linkers adopt an approximate anti-
conformation, although they are highly disordered, even at 100
K, and are not necessarily arranged in the same direction (Figure
2). The anti-conformation of the linker is the most extended
form between two catechol moieties among all the possible
conformations. Therefore, it is no longer possible for 3a to
extend the equatorial portal. In fact, the dimension of the
equatorial portal of 3b in the crystal structure, in which 3b was
observed as an average form between the twistomers and slipped
forms, was almost the same as that in the calculated twisted
form of 3 (vide supra). The intact portal at the equatorial position
of 3a is not large enough to allow guest molecules as large as
4 and 8 to enter and egress the inner phase of capsule 3a, as
shown in Figures 2a, 2b, and 4d.
of a guest into 3a and its egression out of 3a. A catalytic amount
of adventitious water leading to the reversibility of boronic ester
bond is inevitably present in CDCl3 and C6D6.
However, a linker dissociation mechanism, wherein two of
the four linkers completely dissociate from 3a (Figure 6b) would
not be also plausible based on the following results. In the 2D
NOESY spectra of a mixture of as-prepared 3a (2.5 mM), 5
equiv of 8, and 0.38 equiv of free 2 in CDCl3,32 exchange cross-
peaks between 8@3a and free 2, between guest-free 3a and
free 2, and between 8@3a and free 8 were not observed below
308 K, even for a mixing time of 1.0 s (Figure 7a). At 323 K,
exchange cross-peaks between 8@3a and free 2 and between
guest-free 3a and free 2 were also not observed, whereas
exchange cross-peaks between 8@3a and free 8 were observed
in CDCl3 (Figure 7b). These results indicate that (1) the linker
exchange of 3a is much slower than the guest-exchange rate or
(2) the complete dissociation of the two linkers from 3a does
not occur on an NMR time scale.
Consideration of a CPK model suggests that a complete or
partial dissociation of the two linkers from 3a (Figure 6b or
6c), followed by the conformational change of the remaining
two linkers connected with the two caVitands from the anti to
eclipsed-like conformation (Figure 6d) is required for the
opening of the equatorial portal of 3a, followed by the ingression
(32) The saturated concentration of free 2 was 0.28 and 0.94 mM in CDCl3
at 298 and 323 K, respectively, and 0.19 and 0.90 mM in C6D6 at 298
and 323 K, respectively, based on the 1H NMR integration ratio with
p-dimethoxybenzene as an internal standard.
9
784 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Encapsulation Properties of a Self-Assembled Capsule
A R T I C L E S
The most plausible mechanism for guest encapsulation and
release into and out of 3a is a linker partial dissociation
mechanism (Figure 6c),7a wherein dissociation of each one of
the respective two boronic ester bonds of the two linkers among
the four linkers of 3a occurs. Two opposite or adjacent linkers
among the four linkers would open up the capsule (the partial
dissociation of two adjacent linkers is depicted in Figure 6c
and 6d).33 If this linker partial dissociation mechanism is correct,
then the linker exchange of 3a would not be observed in its 2D
NOESY spectrum. The reversible mechanism shown in Figure
6c and 6d is composed of three steps. The first step is a linker
partial dissociation to give a gate-opening guest-free 3a (the
following parentheses denote the gate-opening guest@3a step
as the reverse reaction). The second step is the guest uptake
into the gate-opening 3a (or guest release out of the gate-opening
guest@3a) that occurs simultaneously with the conformational
change of the two linkers connected with the two cavitands from
the anti to the eclipsed-like conformation. The final step is a
linker reassociation and gate-closing process to form guest@3a
(or guest-free 3a). To elucidate this mechanism, the pseudo-
first-order exchange rate constants for release (k-1) and encap-
sulation (k1) of the guest from and into 3a, respectively, and
second-order rate constant of guest encapsulation (k1* ) k1/
[guest]) in CDCl3 and in C6D6 were determined from a 2D
EXSY analysis based on the integration values of the exchange
cross-peaks and diagonal peaks of the encapsulated and free
guests in the 2D NOESY spectra (Figures 7 and 8).17,30b,34 The
results, together with the calculated free energy of activation
(∆G-1‡, ∆G1 , and ∆G1*‡), are summarized in Table 4.35 There
‡
are several features in the kinetic study shown in Table 4.
Item 1. For 8, the value of k1 increases with guest concentra-
tion, and the values of k1* and k-1 are roughly independent of
guest concentration for both CDCl3 and C6D6 (entries 1, 2, and
4, and 5, 7, and 8, respectively).
Item 2. The external addition of linker 2 to a solution of 3a
and 8 barely increases the value of k1 (k1*) and k-1 for both
CDCl3 and C6D6 (entries 2 vs 3, and 5 vs 6, respectively).
Item 3. For 8, 5, and 6 in C6D6, both the values of k1 (k1*)
and k-1 increase in the order 8 < 5 < 6, indicating a significant
dependency on the characteristics of the guest molecule (entries
5 vs 10, and 9 vs 11, respectively). This order could be related
to the difference in size of the guest functional groups or the
interaction ability of them with 3a (or partially dissociated 3a),
such as hydrogen bonding and CH-π interactions.
Item 4. For 8, the value of k1 (k1*) and k-1 in C6D6 is
apparently greater than the value of k1 (k1*) and k-1 in CDCl3
‡
(entries 1 vs 5, 2 vs 7, and 4 vs 8, respectively). Namely, ∆G1
‡
(∆G1*‡) and ∆G-1 in CDCl3 are higher than those in C6D6.
The results in Items 1-3 indicate that the guest uptake in 3a
follows second-order kinetics and the guest release out of 3a
(33) Craig, S. L.; Lin, S.; Chen, J.; Rebek, J., Jr. J. Am. Chem. Soc. 2002,
124, 8780–8781.
(34) (a) Abel, E. W.; Coston, T. P. J.; Orrell, K. G.; Sik, V.; Stephenson,
D. J. Magn. Reson. 1986, 70, 34–53. (b) Pons, M.; Millet, O. Prog.
Nucl. Magn. Reson. Spectrosc. 2001, 38, 267–324.
(35) Guests 5, 6, and 8 were chosen for the 2D EXSY analysis because
the diagonal peaks of encapsulated and free guests in the 2D NOESY
spectra did not overlap with each other or those of the capsule 3a.
However, when excess amounts of guests were used, those faintly
overlapped with each other in some cases, increasing experimental
errors for k1 and k-1. The use of excess amounts of guests or low
8@3a/free-8 would produce a large difference in integration values
between the diagonal peak of the free guest and the exchange cross-
peak, also increasing experimental errors for k1 and k-1. Note that the
values of k1 and k-1 contain an experimental error of 10%-15%.
Figure 8. 2D NOESY spectra of a mixture of 3a and guests in C6D6 for
2D EXSY analysis under the conditions of pulse delay ) 3.0 s, and mixing
time ) 1.0 s: (a) [3a] ) 0.5 mM, [8] ) 1.0 mM at 323 K, (b) [3a] ) 2.5
mM, [5] ) 5.0 mM at 313 K, and (c) [3a] ) 2.5 mM, [6] ) 5.0 mM at 313
K. The representative signals of guest of guest@3a and free guest are marked
with solid and open squares, respectively.
follows first-order kinetics,30b and suggest the linker partial
dissociation mechanism. Item 4 may reflect the solvation effect
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 785

A R T I C L E S
Nishimura et al.
Table 4. Pseudo-First-Order Exchange Rate Constants for Release (k-1) and Encapsulation (k1) of Guests from and into Capsule 3a,
Respectively, and Second-Order Rate Constant of Guest Encapsulation (k1* ) k1/[guest]) in CDCl3 and in C6D6,a and Free Energy of
Activation of Release (∆G-1‡) and Encapsulation (∆G1‡, ∆G1*‡) of Guest@3ab
∆G-1
∆G1‡
∆G1*‡
‡
entry
guest
solvent
[G]/[3a]c
T (K)
k-1 (s-1
)
k1 (s-1
)
k1* (s-1 M-1
)
(kcal mol-1
)
(kcal mol-1
)
(kcal mol-1
)
1
2
3
4
5
6
7
8
8
8
8
8
8
8
8
8
5
5
6
CDCl3
CDCl3
CDCl3
CDCl3
C6D6
C6D6
C6D6
C6D6
C6D6
2
323
323
323
323
323
323
323
323
313
323
313
0.002
0.002
0.005
0.007
0.013
0.016
0.017
0.018
0.120
0.342
0.836
0.034
0.076
0.099
0.157
0.044
0.035
0.097
0.238
0.207
0.537
2.46
6.80
6.08
7.92
6.28
44.0
35.0
38.8
47.6
41.4
107
22.9
22.9
22.3
22.1
21.7
21.6
21.6
21.5
19.7
19.6
18.5
21.1
20.6
20.4
20.1
20.9
21.1
20.4
19.9
19.3
19.3
17.8
17.7
17.8
17.6
17.8
16.5
16.7
16.6
16.5
16.0
15.9
14.5
5
5d
10
2
2d
5
10
2
2
9
10
11
C6D6
C6D6
2
493
a Pseudo-first-order exchange rate constants (k-1, k1) were estimated by 2D EXSY experiments. b ∆G-1 ) -RT ln(hk-1/kBT), ∆G1 ) -RT ln(hk1/
kBT), and ∆G1*‡ ) -RT ln(hk1*/kBT). c [G]/[3a] ) [guest]/[3a]; [3a] ) 2.5 mM for 8@3a in CDCl3, 5@3a in C6D6, and 6@3a in C6D6; and [3a] ) 0.5
mM for 8@3a in C6D6. d Linker 2 was added (saturated concentration of 2 at 323 K: 0.94 mM in CDCl3 and 0.90 mM in C6D6).
‡
‡
of the guest, 3a (guest@3a), or the partially linker-dissociated
3a (guest@3a).
It is known that the stability of hemicarceplexes formed
between covalently bonded hemicarcerands and their guests is
governed by two free energy terms: the intrinsic binding (free
energy of complexation, thermodynamic ∆G°) and the constric-
tive binding (free energy of activation for complex formation,
kinetic ∆G‡association - ∆G‡dissociation),1c,22,36 wherein the balance
of the free energy is (∆G‡
- ∆G‡dissociation) ≈ ∆G°.
association
However, this is not true in the case of guest@3a, wherein the
balance of the free energy for 8@3a in CDCl3 at 323 K (entry
2) was (∆G1*‡ - ∆G-1‡) ) -5.1 kcal/mol < ∆G° ) -2.5 kcal/
mol, and the balance of the free energy for 8@3a in C6D6 (entry
7) was (∆G1*‡ - ∆G-1‡) ) -5.0 kcal/mol at 323 K > ∆G° )
-8.2 kcal/mol at 313 K. The difference between (∆G1*‡ -
∆G-1‡) and ∆G° in guest@3a may reflect the linker partial
dissociation, as well as the solvation effect.
The linker partial dissociation mechanism shown in Figure
6c also facilitates the understanding of the mechanism of the
interconversion between the (M)- and (P)-twisted forms (racemic
twistomers) through the slipped forms in 3a (Figures 9 and 2f).
In one linker 2 unit in 3a, one boronic ester bond dissociation
and a subsequent rotation of the CH2-Ar bond and the
reformation of the boronic ester, followed by a progression of
the same process for the other boronic ester moiety, would
complete the interconversion of the (M)- to the (P)-form for
one linker (Figure 9). Application of the same process to the
other three linkers would complete the interconversion of the
(M)- to the (P)-form of 3a through an intramolecular mixture
of (M)- and (P)-forms of the four linkers in 3a (Figure 2f).
Guest Rotation within 3a. Molecular gyroscopes have at-
tracted considerable attention in the field of mechanical mo-
lecular devices.37,38 They consist of a rotator with a spinning
axis and a stator framework, wherein the rotator is encased and
protected by a stator. So far, molecular gyroscopes have been
synthesized using several types of covalent-bond strategies38,39
or by a metal-coordination strategy, such as a metal-centered
Figure 9. Schematic representation for the interconversion of one linker 2
unit in 3a.
gyroscope.40 Recently, it has been reported that a self-assembled
molecular capsule with an encapsulated guest behaves as a
supramolecular gyroscope, in which an encapsulated guest is
the rotator and a self-assembled capsule is the stator.2k,41
To explore the capability of guest@3a as a supramolecular
gyroscope, the rotation behavior of 4,4′-diacetoxy-2,2′-disub-
stituted-biphenyls 8-10 within 3a was investigated. Guests
8-10 were oriented with their long axis aligned along the long
axis of 3a, as mentioned above, and can rotate along this axis
within 3a. The inner and outer protons of the methylene-bridge
(39) (a) Bedard, T. C.; Moore, J. S. J. Am. Chem. Soc. 1995, 117, 10662–
10671. (b) Dominguez, Z.; Dang, H.; Strouse, M. J.; Garcia-Garibay,
M. A. J. Am. Chem. Soc. 2002, 124, 2398–2399. (c) Setaka, W.;
Ohmizu, S.; Kabuto, C.; Kira, M. Chem. Lett. 2007, 36, 1076–1077.
(d) Nun˜ez, J. E.; Natarajan, A.; Khan, S. I.; Garcia-Garibay, M. A.
Org. Lett. 2007, 9, 3559–3561.
(36) (a) Houk, K. N.; Nakamura, K.; Sheu, C.; Keating, A. E. Science 1996,
273, 627–629. (b) Yoon, J.; Cram, D. J. Chem. Commun. 1997, 1505–
1506.
(37) (a) Kottas, G. S.; Clarke, L. I.; Horinek, D.; Michl, J. Chem. ReV.
2005, 105, 1281–1376. (b) Kay, E. R.; Leigh, D. A.; Zerbetto, F.
Angew. Chem., Int. Ed. 2007, 46, 72–191.
(40) (a) Shima, T.; Hampel, F.; Gladysz, J. A. Angew. Chem., Int. Ed. 2004,
43, 5537–5540. (b) Narwara, A. J.; Shima, T.; Hampel, F.; Gladysz,
J. A. J. Am. Chem. Soc. 2006, 128, 4962–4963.
(38) Khuong, T.-A. V.; Nunez, J. E.; Godinez, C. E.; Garcia-Garibay, M. A.
Acc. Chem. Res. 2006, 39, 413–422.
(41) Scarso, A.; Onagi, H.; Rebek, J., Jr. J. Am. Chem. Soc. 2004, 126,
12728–12729.
9
786 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Encapsulation Properties of a Self-Assembled Capsule
A R T I C L E S
Figure 10. Top views of half-structure model of 3 (a) with and (b) without
guest 8. (c) Four-site model for the encapsulated-guest rotation along the
long axis of 3a, wherein schematic representation of the cross section of
guest@3a is shown. The centered rectangle and the Y represent the half-
biphenyl ring and the ester group of guests 8-10, respectively. The CdO
means the carbonyl moiety of the guest acetoxy group. The peripheral circle
and rectangles represent the aromatic cavity of the 1a unit and the catechol
rings of the 2 unit in capsule 3a, respectively. The Ain, Bin, Cin, and Din
(Aout, Bout, Cout, and Dout) represent the relative positions of the inner protons
Hin (outer protons Hout) of the methylene-bridge rim of the 1a unit of 3a
with respect to the positions of the carbonyl moiety and the substituent-Y
of the encapsulated-guest rotating in 3a.
Figure 11. Guest-rotation behavior within 3a, monitored by 1H NMR (400
MHz, C6D6, [3a] ) 2.5 mM and [guest] ) 5 mM) in the region of the
inner (Hin) and outer (Hout
) protons of the methylene-bridge rims
(O-CHinHout-O) of 3a: (a) 8@3a at 298 K, (b) 10@3a at 298 K, (c) 10@3a
at 343 K, (d) 9@3a at 298 K, (e) 9@3a at 323 K, and (f) 9@3a at 343 K.
The signals marked ‘Ain-Din’ and ‘Aout-Dout’ are assigned in Figure 10.
The signal marked ‘c’ is assigned in Scheme 1.
rims (O-CHinHout-O) of 3a act as molecular probes for the
encapsulated-guest rotation within 3a.2k Figure 10 depicts a four-
site model for the encapsulated-guest rotation along the long
axis of 3a. In this model, if the encapsulated-guest rotation along
the long axis of 3a is too fast on the NMR time scale, then the
1H NMR signals of the inner and outer protons (Hin and Hout)
would appear as a respective one doublet signal, i.e., as the
averaged form. In contrast, if the encapsulated-guest rotation
is slow on the NMR time scale or comes to a stop, then the
doublet signals of Hin and Hout would split into four sets of
doublet signals with an integration ratio of Ain:Bin:Cin:Din (Aout:
Bout:Cout:Dout) ) 1:1:1:1.
the elongation of the ester group at the 2,2′-positions of 4,4′-
diacetoxybiphenyl guests puts the brakes on guest rotation within
3a.
Figure 12 shows the 2D NOESY spectrum of a 1:1 mixture
of 9@3a and free 9 in C6D6 at 298 K (mixing time ) 0.5 s),
wherein the guest exchange into and out of 3a was not observed,
and the acetoxy signal of the encapsulated 9 showed NOE
correlations with four split inner protons (Ain-Din) of the
methylene-bridge rim of 3a (Figure 12a). Furthermore, it should
be noted that all the inner protons (Ain-Din) and all the outer
protons (Aout-Dout) of 3a showed exchange cross-peaks with
one another in addition to the NOE correlations (Figure 12b).
For example, signal Ain showed an NOE correlation with Aout
and exhibited exchange cross-peaks with Bin, Bout, Cin, Cout, Din,
and Dout. Similar correlations were also observed in C6D5CD3
at 298 K (Figure 13a). This result shows that 9 within 3a rotates
at 298 K, although the rotation is slower than the NMR time
scale. In marked contrast, at 243 K, only NOE correlations
between the inner and outer protons were observed, and
exchange cross-peaks were no longer observed (Figure 13c vs
13a), although the chemical shift of the inner and outer protons
of 9@3a changed depending on the solvent used and the
temperature (Figures 12 and 13). This result clearly indicates
that the rotation of 9 within 3a ceases at 243 K.
1
Figure 11 shows the H NMR spectra of guest@3a (Y )
CO2CH3 in 8, Y ) CO2(CH2)7CH3 in 9, and Y ) CO2(CH2)4Ph
in 10) in C6D6 at various temperatures.17 This gave significant
information on the encapsulated-guest rotation along the long
axis of capsule 3a, namely, of the rotational steric barrier effect
for guests within 3a. In the 1H NMR spectrum of 8@3a at 298
K (Figure 11a), the inner (Hin) and outer (Hout) protons of the
methylene-bridge rims of 3a appear as respective one sharp
doublet signal. This spectrum remained unchanged, even at 213
K in C6D5CD3.17 This result undoubtedly indicates that the
rotation of 8 within 3a is too fast on the NMR time scale, even
1
at 213 K. On the other hand, the H NMR signals of Hin and
Hout of 10@3a were extremely broad and nearly coalesced at
298 K (Figure 11b) and were still broadened, even at 343 K
1
(Figure 11c). In marked contrast, the H NMR signals of Hin
1
and Hout of 9@3a at 298 K were split into four sets of doublet
signals of Ain, Bin, Cin, and Din, and Aout, Bout, Cout, and Dout in
an integration ratio of 1:1:1:1, respectively (Figures 11d and
10c). These signals were nearly coalesced at 343 K (Figure 11f).
This result clearly indicates that the rotation of 9 within 3a is
very slow on the NMR time scale at 298 K. Thus, the guest
rotation within 3a increases in the order 9 < 10 , 8. Namely,
Variable low temperature H NMR spectra of 9@3a and
10@3a showed that four sets of doublet signals of the inner
(Ain-Din) and outer (Aout-Dout) protons of the methylene-bridge
rims of 3a were shifted, coalesced, and split at several points,
depending on the temperature.17 This result suggests that the
guest rotation within 3a, the interconversion of the twistomers
of 3a, and the internal rotation of the biphenyl moiety of the
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 787

A R T I C L E S
Nishimura et al.
of free 9 (Figure 14c vs 14b).42 This arises from the desym-
metrization of the two geminal R-protons given the emergence
of a chiral axis (racemization) in the encapsulated 9, because
the internal rotation of the biphenyl moiety of 9 is slow on the
NMR time scale on encapsulation in 3a. This phenomenon was
also observed in 10@3a.
Conclusions
We have described the self-assembly of two molecules of
tetrakis(dihydroxyboryl)-cavitand 1a and four molecules of 1,2-
bis(3,4-dihydroxyphenyl)ethane 2 into capsule 3a by the forma-
tion of eight dynamic boronic ester bonds in CDCl3 and C6D6.
Capsule 3a encapsulates one guest molecule, such as 4,4′-
disubstituted-biphenyl and 2,6-disubstituted-anthracene deriva-
tives, which is oriented with the long axis of the guest directed
1
along the long axis of 3a, as confirmed by a H NMR study
and X-ray crystallographic analysis. Capsule 3a strictly dis-
criminates the functional groups of a guest, as well as a
difference of one or two carbon atoms in guest size, leading to
a highly selective guest recognition on encapsulation. Capsule
3a shows a significant solvent effect on guest encapsulation.
The association constant of 3a with guests in C6D6 is much
greater than that in CDCl3 (450-48 000-fold). The encapsulation
of guests within 3a in C6D6 is enthalpically driven, whereas
the encapsulation of guests within 3a in CDCl3 tends to be both
enthalpically and entropically driven. Thermodynamic studies
suggest that the small value of Ka with a considerable entropic
contribution for guest@3a in CDCl3 arises from the character
of CDCl3 as a guest competitor for 3a, and not from the
difference in stability of the boronic ester bonds of 3a in both
solvents. We propose a linker partial dissociation as the
mechanism for guest uptake and release into and out of 3a based
on kinetic studies of guest@3a by 2D EXSY analysis, as well
as structural analysis of guest@3a. The 2D EXSY studies also
suggest that guest uptake in 3a follows second-order kinetics
and guest release out of 3a follows first-order kinetics. The
rotation behavior of 4,4′-diacetoxy-2,2′-disubstituted-biphenyls
within 3a was also investigated, where the elongation of 2,2′-
disubstituents of guests hindered on guest rotation within 3a.
Areas for further investigation are (1) asymmetric guest
recognition of a chiral capsule43 to be derived from 1a and
incorporating a C2 symmetric chiral bis(catechol)-linker in place
of 2 and (2) application of an asymmetric-guest@chiral-capsule
to a supramolecular gyroscope directed toward unidirectional
guest rotation.37a,44
Figure 12. 2D NOESY spectrum of 9@3a in C6D6 at 298 K (400 MHz,
[3a] ) 2.5 mM and [9] ) 5 mM, pulse delay ) 1.5 s, and mixing time )
0.5 s) in the regions (a) between -1.5 and 8.5 ppm and (b) between 4.6
and 6.2 ppm. The signals marked ‘Ain-Din’ and ‘Aout-Dout’ are assigned
in Figure 10. The signals marked ‘c’, ‘e’, and ‘i’ are assigned in Scheme
1. The solid and open squares indicate the representative signals of 9 of
9@3a and free 9, respectively.
guest coexist in cooperation with each other. Given this
complexity, we could not determine the exchange rate constant
for guest rotation within 3a at this stage.
Experimental Section
General. THF and CH2Cl2 were distilled from sodium-ben-
zophenone ketyl and CaH2, respectively, under an argon atmosphere.
The other solvents and all commercially available reagents were
used without any purification. 1H and 13C NMR spectra were
recorded at 400 and 100 MHz, respectively, on a JEOL JNM-AL400
spectrometer. IR spectra were recorded on a JASCO FT/IR-460Plus
spectrometer. The CDCl3 and CD2Cl2 employed in all the 1H NMR
experiments were stored over K2CO3 prior to use, and the C6D6
1
In 4@3a in C6D6 or C6D5CD3, the H NMR signal of the
ethylene moiety of the linker 2 unit of 3a appeared as a single
sharp singlet at 298 K, and this broadened without splitting,
even at 213 K.17 On the other hand, the linker signal of 8@3a
in C6D5CD3 appeared as a multiplet at 298 K and split into two
multiplets at 263 K.17 In marked contrast, the linker signal of
9@3a in C6D6 or C6D5CD3 split into five signals at 298 K (three
major broad signals and two minor broad signals, with one of
the minor signals overlapping one of the major signals, Figures
12a and 14c). Furthermore, at least one of the three major signals
split into two signals at 233 K.17 In 9@3a, the 1H NMR signal
of the R-protons of the octyl ester group (CO2-CH2(CH2)6CH3)
of the encapsulated 9 split into two sets of double triplet signals
with an upfield shift relative to the triplet signal of the R-protons
(42) The biphenyl rotational barrier of guest 9 in C6D5CD3 was estimated
to be ∆G‡ ) <14.9 kcal mol-1, which was determined by the VT 1H
NMR.17
(43) (a) Yoon, J.; Cram, D. J. J. Am. Chem. Soc. 1997, 119, 11796–11806.
(b) Castellano, R. K.; Nuckolls, C.; Rebek, J., Jr. J. Am. Chem. Soc.
1999, 121, 11156–11163. (c) Mateos-Timoneda, M. A.; Crego-Calama,
M.; Reinhoudt, D. N. Chem. Soc. ReV. 2004, 33, 363–372. (d) Amaya,
T.; Rebek, J., Jr. J. Am. Chem. Soc. 2004, 126, 6216–6217. (e)
Schramm, M. P.; Rebek, J., Jr. New J. Chem. 2008, 32, 794–796.
9
788 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010

Encapsulation Properties of a Self-Assembled Capsule
A R T I C L E S
Figure 13. 2D NOESY and 1H-1H COSY spectra of 9@3a in C6D5CD3 (400 MHz, [3a] ) 2.5 mM and [9] ) 5 mM, pulse delay ) 1.5 s, and mixing time
1
1
) 0.5 s): (a) 2D NOESY and (b) H-1H COSY at 298 K; (c) 2D NOESY and (d) H-1H COSY at 243 K. The signals marked ‘Ain-Din’ and ‘Aout-Dout
’
are assigned in Figure 10. The signal marked ‘c’ is assigned in Scheme 1.
dihydroxyphenyl)ethane 2, and guest 15 were synthesized according
to the literatures.15,45 The syntheses of other guests, X-ray data
collection and crystal structure determination of 4@3b, and 2D
EXSY experiments of guest@3a were described in the Supporting
Information.
Tetrakis(dihydroxyboryl)-Cavitand 1b (R ) (CH2)2Ph). To
a solution of tetrabromo-cavitand (R ) (CH2)2Ph)46 (1.00 g, 0.788
mmol) in dry THF (60 mL) at -78 °C under an argon atmosphere
was added a hexane solution of n-BuLi (1.59 M, 2.5 mL, 4.0 mmol).
After stirring for 1 h, to the resulting solution was added B(OMe)3
(1.6 mL, 14 mmol) at -78 °C. The resulting mixture was stirred
at -78 °C for 1 h and then allowed to warm to room temperature
overnight. The reaction mixture was quenched with 2 M HCl (10
mL) at 0 °C and stirred for 30 min. After evaporation of solvents,
the residue was partitioned between EtOAc (300 mL) and water
(100 mL), and the organic layer was washed with water and brine
and dried over Na2SO4. After evaporation of solvents, the residue
was recrystallized from EtOAc-hexane to give 1b as a bis-adduct
of EtOAc (off-white crystals) after vacuum drying at room
temperature for 12 h (617 mg, 60% yield). Mp 282 °C (decomp.);
Figure 14. 1H NMR spectra (400 MHz, C6D6, 298 K): (a) [3a] ) 2.5
mM, (b) [9] ) 5 mM, and (c) [3a] ) 2.5 mM and [9] ) 5 mM (9@3a:
free-9 ) 1:1). The signals marked ‘e’ and ‘i’ are assigned in Scheme 1.
The representative signals of 9 of 9@3a and free 9 are marked with solid
and open squares, respectively.
(44) (a) Zheng, X.; Mulcahy, M. E.; Horinek, D.; Galeotti, F.; Magnera,
T. F.; Michl, J. J. Am. Chem. Soc. 2004, 126, 4540–4542. (b) Horinek,
D.; Michl, J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 14175–14180.
(45) Orita, A.; Taniguchi, H.; Otera, J. Chem.sAsian J. 2006, 1, 430–
437.
(46) Sherman, J. C.; Knobler, C. B.; Cram, D. J. J. Am. Chem. Soc. 1991,
113, 2194–2204.
and C6D5CD3 were stored over molecular sieves 4A prior to use.
Tetrakis(dihydroxyboryl)-cavitand 1a (R ) (CH2)6CH3), 1,2-bis(3,4-
9
J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010 789

A R T I C L E S
Nishimura et al.
1H NMR (DMSO-d6) δ 8.08 (s, 8H), 7.53 (s, 4H), 7.16-7.25 (m,
20H), 5.53 (d, J ) 7.8 Hz, 4H), 4.63 (t, J ) 7.3 Hz, 4H), 4.43 (d,
J ) 7.8 Hz, 4H), 3.30-3.33 (m, 8H), 2.53-2.62 (m, 8H); 13C NMR
(DMSO-d6) δ 155.5, 141.8, 137.2, 128.6, 128.4, 128.3, 125.8, 121.5,
98.7, 36.5, 34.2, 31.9. Anal. Calcd for C64H60B4O16 ·2(EtOAc): C,
66.29; H, 5.87. Found: C, 66.08; H, 6.15.
6.55 (d, J ) 7.8 Hz, 8H), 6.49 (s, 8H), 5.72 (d, J ) 6.8 Hz, 8H),
5.29 (t, J ) 7.8 Hz, 8H), 4.93 (d, J ) 6.8 Hz, 8H), 2.34-2.42 (m,
16H), 2.28 (s, 16H), 1.20-1.44 (m, 80H), 0.95 (t, J ) 6.8 Hz,
24H).
Guest@3a. To as-prepared 3a in an NMR tube was added a
stock solution of a guest in CDCl3 or C6D6. The resulting mixture
was heated at 50 °C for 30 min to give a solution of guest@3a.17
Alternatively, to a 2:4 mixture of 1a·OEt2 and 2 in an NMR tube
was added a stock solution of a guest in CDCl3 or C6D6, and the
resulting mixture was heated at 50 °C for 3 h to give a solution of
guest@3a. A solution of 8@3a in C6D6 was prepared at 0.5 mM
because of its solubility problem. In all other cases, a solution of
guest@3a or a mixture of guest@3a and free 3a were prepared at
2.5 mM in total.
Capsule 3a. A suspension mixture of 1a·OEt2 (50.00 mg, 42.4
mmol) and 2 (20.89 mg, 84.8 mmol, 2 equiv) in CDCl3 (or CHCl3)
1
(8.5 mL) was stirred at 50 °C for 3 h. The H NMR spectrum of
the resulting homogeneous solution showed the quantitative forma-
tion of 3a (see Figure 1c). After evaporation of solvents, the residue
was dried in vacuo at room temperature for 5 h. The dried sample
of 3a was initially less soluble in CDCl3 at room temperature, but
after heating in CDCl3 at 50 °C for several minutes, 3a became
soluble again at room temperature. Heating a 2:4 mixture of 1·OEt2
and 2 in C6D6 (or C6H6) at 50 °C for 3 h with stirring did not give
a homogeneous system, although 3a was quantitatively formed
(Figure 1f). The solubility of 3a in C6D6 at 23 °C was 1.3 mM
Acknowledgment. We thank Prof. A. Orita (Okayama Univer-
sity of Science) for the gift of guest 15. We also thank Prof. T.
Haino (Hiroshima University) for valuable discussions. This work
was supported in part by PRESTO, JST.
1
based on the H NMR integration ratio with p-dimethoxybenzene
1
as an internal standard. H NMR (CDCl3, 296 K, 2.5 mM) δ 7.37
(s, 8H), 7.27 (d, J ) 7.8 Hz, 8H), 7.11 (d, J ) 1.9 Hz, 8H), 7.06
(dd, J ) 1.9 and 7.8 Hz, 8H), 5.67 (d, J ) 7.3 Hz, 8H), 4.89 (t, J
) 8.3 Hz, 8H), 4.63 (d, J ) 7.3 Hz, 8H), 2.97(s, 16H), 2.27-2.37
(m, 16H), 1.25-1.53 (m, 80H), 0.93 (t, J ) 6.3 Hz, 24H); 1H NMR
(C6D6, 296 K, 1.3 mM) δ 7.77 (s, 8H), 6.95 (d, J ) 7.8 Hz, 8H),
Supporting Information Available: Additional data and X-ray
crystallographic data (CIF). This material is available free of
charge via the Internet at http://pubs.acs.org.
JA9084918
9
790 J. AM. CHEM. SOC. VOL. 132, NO. 2, 2010