A deep, water-soluble cavitand acts as a phase-transfer catalyst for hydrophobic species

Article abstract of DOI:10.1002/anie.200600405

Overcoming inhibition: A deep, water-soluble cavitand has been shown to extract various N-substituted maleimides into aqueous solution through the hydrophobic effect. These complexes react with a water-soluble thiol to confer water solubility to the produ

Full text of DOI:10.1002/anie.200600405

Angewandte
Chemie
Host–Guest Chemistry
guest to be replaced by the starting material, thus leading to
turnover and the use of the cavitand as a phase-transfer
catalyst (Figure 1). For example, sodium decyl sulfate (SDS)
is a good guest for 1 below its critical micelle concentration,[8]
but no binding is observed at higher concentrations; the guest
only binds if dissolution in water is unfavorable.[9]
DOI: 10.1002/anie.200600405
A Deep, Water-Soluble Cavitand Acts as a Phase-
Transfer Catalyst for Hydrophobic Species**
Richard J. Hooley, Shannon M. Biros, and
Julius Rebek, Jr.*
Product inhibition is generally a problem for catalysis by
synthetic receptors.[1] Supramolecular catalysts do not show
the exquisite recognition of transition structures versus
products required for high turnover and, with few excep-
tions,[2,3] their reactions slow down, then stop. We have now
encountered cases of phase-transfer catalysis[4] with the water-
soluble tetracarboxylate cavitand 1 (Scheme 1).[5] The reac-
Figure 1. The phase-transfer process. a) The water-insoluble guest is
extracted into aqueous solution by cavitand 1. b) The bound guest,
with its reactive group protruding fromthe cavity, reacts with an
aqueous partner. c) The product of this union gains water solubility
and leaves the cavitand, thus allowing for turnover and catalysis.
Org=organic phase, Aq=aqueous phase.
We used a “two-part” reactant to effect the process: a
hydrophobic anchor (to provide water insolubility and shape
complementarity for the cavitand) and a reactive function-
ality, N-substituted maleimides. Species such as N-adaman-
tylmaleimide 3 show no water solubility by NMR spectro-
scopic analysis, yet maleimides 3–9 are all extracted into a
solution of 1 in D2O. Their NMR spectra show that the
hydrocarbon portions of substrates 3–9 are bound inside the
cavity, whereas the maleimide function is exposed to the
solvent.
A water-soluble coupling partner that reacts well with the
maleimide and confers water solubility on the product was
found in the commercially available thiol 2. Reaction of 2 with
maleimide 3 proceeds smoothly to completion in less than
10 minutes in a homogeneous THF/H2O solution to give
product 10. Water-soluble succinimide 10 shows no affinity for
1, thus promising that the product would be displaced from
the cavitand by the starting material.
Table 1 shows the results of the addition of thiol 2 to
various N-substituted maleimides 3–9 in a two-phase (CD2Cl2/
D2O) system in the presence and absence of 1 as the catalyst.
The reactions in the presence of 2mol% 1 showed a
significant rate enhancement relative to the uncatalyzed
process. The lack of affinity of the water-soluble products for
1 allows the use of only 2mol% catalyst, thus suggesting that
very little (if any) product inhibition occurs.
The larger maleimides required longer times for comple-
tion of the reaction, and the rate acceleration was corre-
spondingly greater. Only N-n-octylmaleimide 7 was a poor
substrate for the reaction, with the starting material still
present after 2days. Maleimide 7 was almost completely
insoluble in water, and even the anionic product 14 was only
sparingly soluble. The background rate for this reaction was
very slow; less than 1% completion was observed after 51 h,
and only 3% conversion occurred after 220 h. Both N-
Scheme 1. Water-soluble synthetic receptor 1 and a representation of
the complex of 1 with THF (Maestro v7.0.2; AMBER forcefield;[6] some
groups are omitted for clarity).
tions produce anionic products that are replaced with a
neutral reactant, thus forcing high turnover. The results
suggest that moving beyond shape complementarity in the
recognition event can lead to efficient catalysts.
Normal alkanes, which are essentially insoluble in water,
can be extracted into aqueous solutions containing 1 and form
stoichiometric 1:1 complexes.[7] Hydrophobic stabilization
drives this binding and leads to very high association
constants. The open-ended nature of the receptor exposes a
portion of the guest to the aqueous environment. When the
guest gains a water-soluble functionality by chemical trans-
formation, the hydrophobic stabilization of the guest is
decreased along with the binding affinity. This allows the
[*] Dr. R. J. Hooley, S. M. Biros, Prof. J. Rebek, Jr.
The Skaggs Institute for Chemical Biology and
Department of Chemistry
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax: (+1)858-784-2876
E-mail: jrebek@scripps.edu
[**] We are grateful to the Skaggs Institute and the National Institutes of
Health (GM 27932) for financial support and to Dr. L.B. Pasternak
and Dr. D-H. Huang for NMR assistance. R.J.H. is a Skaggs
Postdoctoral fellow.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or fromthe author.
Angew. Chem. Int. Ed. 2006, 45, 3517 –3519
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3517

Communications
Table 1: Scope of the phase-transfer process.[a]
of the nucleophilic addition step was determined by treating a
preformed complex of 1 and 3 (1 mm) with an equimolar
amount of 2. Complete conversion into the product was
observed in less than 1 minute, thus indicating that the
second-order nucleophilic addition step is much faster than
the overall process. The rates of addition for each of the
complexed maleimides should be approximately the same. A
similarly rapid rate was observed when a 1 mm aqueous
solution of N-methylmaleimide was treated with 2.
The rate of the overall process is determined by the
transfer rate of the maleimide reactant into the aqueous layer.
The innate water solubility of the maleimide and its affinity
for the cavitand are responsible for the differences in the
transfer rates. Certain fast-reacting species, namely, 8 and 9,
are slightly water soluble, and so the effectiveness of their
phase transfer is not determined solely by binding affinity
(they have poor shape complementarity for 1). However, as
the size of the alkyl substituent increases, the water solubility
drops to a level at which it no longer dominates the phase-
transfer process. These larger species 3–7 are transferred in
accordance with their binding properties.
Maleimide
t [h]
24
Yield [%][b]
(2 mol% 1)
Yield [%]
(0 mol% 1)[b]
3
4
5
6
81
82
76
80
13
1
4.5
9
7
4.5
15
The rate difference between 3 and 4 is surprising; the
addition of only one methylene unit causes a significant
increase in reaction rate. The 1H NMR spectra of the
complexes give an illustration of the relative positions of
the maleimide group when bound in 1 (see the Supporting
Information for the 1H NMR spectra). The chemical shift for
the olefinic protons in N-methylmaleimide in D2O (d =
6.88 ppm) can be used as a reference to approximate the
extent to which the maleimide is unaffected by the magnetic
anisotropy of the cavitand. The further upfield the olefinic
protons of the guest are shifted, the “deeper” the maleimide
resides in the cavity. These signals of the longer guests 4, 6,
and 7 were essentially unshifted (d = 6.83, 6.80, and 6.96 ppm,
respectively), thus showing that the large alkyl substituent fills
the cavity and leaves the maleimide near the rim of the
cavitand. Maleimides 5 and 8 are bound more deeply because
of their small size, with olefinic CH shifts of d = 6.74 and
6.73 ppm, respectively, but the olefinic protons of 3 were
much more highly shifted (d = 6.58 ppm), thus indicating that
the guest is located even deeper than the small cyclohex-
ylmaleimide 8.[10] These Dd values are not large and indicate
that the maleimide function is situated at the cavitand
opening, well positioned for reaction.
Table 1 shows that 4 is far more reactive in the phase-
transfer process than 3, even though their sizes (and water
solubilities) are similar. The major difference between the
two is that 4 is a far better guest for 1 than 3. When an
equimolar mixture of solids 4 and 3 is exposed to a solution of
1, 4 is preferentially encapsulated by a ratio of 6:1 (Figure 2).
As discussed above, the maleimide moiety is positioned
deeper in the cavity in 3 as opposed to 4. The adamantyl group
has good shape complementarity for 1, but the binding of N-
adamantylmaleimide (3) evidently forces the carbonyl groups
of the maleimide too close to the benzimidazole walls, thereby
lowering the binding affinity. In the case of 4, the maleimide
group is positioned further from the walls, and so binding
affinity (and consequently the rate of phase transfer) is
increased. Figure 2shows molecular models of the complexes
7
8
9
51
2
25
98
78
1
12
60
1
[a] Maleimide (70 mm in CD2Cl2) was combined with 1 (1.4 mm) and 2
(100 mm) in D2O and stirred at 238C. [b] Yields determined by dilution
with a stock solution (6.2 mm) of dipotassium1,2-benzenedisulfonate in
D2O and comparison of integrals in the 1H NMR spectrum.
cyclohexylmaleimide (8) and N-phenylmaleimide (9) reacted
quickly, but showed relatively high background reaction rates
because of their partial solubility in water. A competition
experiment between 3 and 8 showed that selectivity was
possible. When reacted in a two-phase system in the presence
of 2mol% 1, significant selectivity (> 8:1) for the reaction of
8 over 3 was observed. When the reaction was performed in
the absence of catalyst under homogeneous conditions (1:1
THF/H2O), the ratio of products was 1:1.
A good turnover was achieved with 2mol% catalyst, but
lower concentrations of catalyst were also effective. The
reaction of 6 and 2 for 2h in the presence of 1 and 0.5 mol% 1
gave yields of 31 and 15%, respectively. The conversion
observed at 0.2mol% catalyst (8%) was not significantly
greater than that of the background reaction.
Why do the rates vary so much for processes with very
similar reactants? There are two steps to the phase-transfer
process which could contribute to the observed differences:
the transfer from the organic phase to the aqueous phase
(Figure 1, step a) and the reaction (step b). The relative rate
3518
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3517 –3519

Angewandte
Chemie
observed product inhibition. The efficiency of these reactions
is governed by both the intrinsic solubility of the reactants and
the binding affinity of the substrates for the cavitand.
Accordingly, selective reactions can be conceived for suitably
sized molecules bearing a hydrophobic “handle”. Further
studies on the applications of 1 to promote reactions of such
species are underway.
Received: January 30, 2006
Published online: April 20, 2006
Keywords: aqueous reactions · hydrophobic effect · molecular
.
recognition · NMR spectroscopy · phase-transfer catalysis
[1] F. Hof, S. L. Craig, C. Nuckolls, J. Rebek, Angew. Chem. 2002,
114, 1556 – 1578; Angew. Chem. Int. Ed. 2002, 41, 1488 – 1508.
[2] a) J. M. Kang, G. Hilmersson, J. Santamaria, J. Rebek, J. Am.
Chem. Soc. 1998, 120, 3650 – 3656; b) J. Chen, J. Rebek, Org.
Lett. 2002, 4, 327 – 329; c) B. W. Purse, A. Gissot, J. Rebek, J.
Am. Chem. Soc. 2005, 127, 11222 – 11223.
[3] Some successful examples of supramolecular catalysis include:
a) W. L. Mock, T. A. Irra, J. P. Wepsiec, T. L. Manimaran, J. Org.
Chem. 1983, 48, 3619 – 3620; b) A. McCurdy, L. Jiminez, D. A.
Stauffer, D. A. Dougherty, J. Am. Chem. Soc. 1992, 114, 10314 –
10321; c) J. M. Kang, J. Santamaria, G. Hilmersson, J. Rebek, J.
Am. Chem. Soc. 1998, 120, 7389 – 7390; d) D. Fiedler, R. G.
Bergman, K. N. Raymond, Angew. Chem. 2004, 116, 6916 – 6919;
Angew. Chem. Int. Ed. 2004, 43, 6748 – 6751; e) H. Ito, T.
Kusukawa, M. Fujita, Chem. Lett. 2000, 598 – 599.
[4] a) R. Breslow, S. D. Dong, Chem. Rev. 1998, 98, 1997 – 2011; b) S.
Shimizu, T. Suzuki, S. Shirakawa, Y. Sasaki, C. Hirai, Adv. Synth.
Catal. 2002, 344, 370 – 378; c) M. Baur, M. Frank, J. Schatz, F.
Schildbach, Tetrahedron 2001, 57, 6985 – 6991.
Figure 2. 1H NMR spectra of the complexes of cavitand 1 (1 mm, D2O)
and a) N-adamantylmaleimide (3), b) N-adamantylmethylmaleimide
*
(4), and c) an equimolar mixture of 3 and 4 ( , peak from“free”
cavitand 1 binding THF). Representations of the complexes of 1 with
d) 4 and e) 3 (Maestro v7.0.2; AMBER forcefield;[6] some groups are
omitted for clarity).
[5] S. M. Biros, E. C. Ullrich, F. Hof, L. Trembleau, J. Rebek, J. Am.
Chem. Soc. 2004, 126, 2870 – 2876.
[6] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M.
Lipton, C. Caufield, G. Chang, T. Hendrickson, W. C. Still, J.
Comput. Chem. 1990, 11, 440 – 467.
[7] R. J. Hooley, S. M. Biros, J. Rebek, Chem. Commun. 2006, 5,
509 – 510.
[8] L. Trembleau, J. Rebek, Science 2003, 301, 1219 – 1220.
[9] L. Trembleau, J. Rebek, Chem. Commun. 2004, 58 – 59.
[10] Under the conditions used to form the complexes, N-phenyl-
maleimide (9) was hydrolyzed to N-phenylmaleamic acid as a
3·1 and 4·1, from which the relative positions of the maleimide
oxygen atoms may be seen.
The cavitand is also useful to effect transfer between the
solid and liquid phases. When solid 3 was added to a solution
of 2 and 2mol% 1, no dissolution was observed; the
maleimide merely floated on top of the solution. After
2hours of vigorous stirring, quantitative consumption of
starting material had occurred, thus yielding product 10, as
observed in the solid–liquid cases described above. In the
absence of the catalyst, the maleimide was present even after
4 days of stirring. Quantitative analysis of this process proved
difficult, however, as the rate of the reaction was highly
dependent not only upon the rate of stirring, but also on the
physical nature of the solid maleimide: recrystallized mal-
eimide and maleimide purified by chromatography on silica
À
result of the extra lability of the C N amide bond. In the phase-
transfer process, no products from hydrolysis were detected. The
nucleophilic addition of 2 is evidently much faster than the
competing hydrolysis.
[11] S. Narayan, J. Muldoon, M. G. Finn, V. G. Fokin, H. C. Kolb,
K. B. Sharpless, Angew. Chem. 2005, 117, 3339 – 3343; Angew.
Chem. Int. Ed. 2005, 44, 3275 – 3279.
1
gel, identical by H NMR spectroscopic analysis, gave large
variances in rate (the large acceleration in the presence of the
catalyst was still observed in each case). The slow rate of
reaction for the uncatalyzed process (relative to the rapid
conversion observed in THF/H2O) suggests that this reaction
is not an example of acceleration “on water”, as described
recently.[11]
In conclusion, water-soluble tetracarboxylate cavitand 1
extracts insoluble maleimides into the aqueous phase, where-
upon reaction with thiol 2 occurs. The anionic products are
released from the cavitand and replaced by the reactant, thus
allowing catalytic phase-transfer processes with little-to-no
Angew. Chem. Int. Ed. 2006, 45, 3517 –3519
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3519