An asymmetric ortholithiation approach to inherently chiral calix[4]arenes

Article abstract of DOI:10.1021/ol902238p

A general asymmetric synthesis of inherently chiral calix[4]arenes is described: using a chiral oxazoline derived from L-valine, an ortholithiation strategy Is employed to give inherently chiral calix[4]arenes with high (93%) enantiomeric excesses. A crys

Full text of DOI:10.1021/ol902238p

ORGANIC
LETTERS
2009
Vol. 11, No. 21
4986-4989
An Asymmetric Ortholithiation Approach
to Inherently Chiral Calix[4]arenes
Simon A. Herbert and Gareth E. Arnott*
Department of Chemistry and Polymer Science, UniVersity of Stellenbosch,
Matieland 7602, South Africa
arnott@sun.ac.za
Received September 28, 2009
ABSTRACT
A general asymmetric synthesis of inherently chiral calix[4]arenes is described: using a chiral oxazoline derived from L-valine, an ortholithiation
strategy is employed to give inherently chiral calix[4]arenes with high (93%) enantiomeric excesses. A crystal structure of a phosphine oxide intermediate
has been obtained, unambiguously assigning the major diastereomer in the reaction; a mechanism explaining this result is proposed.
Calix[4]arenes are large bowl-shaped molecules that have
been well studied for their properties in host-guest chem-
istry.1 Since calix[4]arenes are nonplanar, it is possible to
functionalize them in such a way as to produce an inherently
chiral framework; the first example of this was published
more than 25 years ago by Gutsche.2 Although there are
potentially many exciting applications of inherently chiral
calix[4]arenes as ligands for asymmetric reactions,3 their
study has been severely hampered by the difficulty in
obtaining them in an enantiomerically pure form.4 Recently,
there have been reports on methods for obtaining nonracemic
inherently chiral calix[4]arenes via resolution techniques,5
but none of these demonstrates the use of an asymmetric
synthetic methodology. Intrigued by this, we considered
whether the asymmetric ortholithiation approaches used to
introduce planar chirality in ferrocenes6 and chromium-
arenes7 would translate to the realm of calix[4]arenes. We
reasoned that the shape of the calixarene allows for facial
differentiation of the aromatic rings (i.e., “inside bowl” or
“outside-bowl”) which may be comparable to the metal-arene
complex’s “metal-side” and “open-side”. To test this hy-
pothesis, we envisaged employing a chiral oxazoline as an
ortholithiation directing group,8 which would allow for
asymmetric functionalization of the upper rim of the calix-
arene. Herein we report our findings.
(5) Lower-rim lipase transesterification: (a) Browne, J. K.; McKervey,
M. A.; Pitarch, M.; Russell, J. A.; Millership, J. S. Tetrahedron Lett. 1998,
39, 1787–1790. Lower-rim resolution: (b) Yakovenko, A. V.; Boyko, V. I.;
Danylyuk, O.; Suwinska, K.; Lipkowski, J.; Kalchenko, V. I. Org. Lett.
2007, 9, 1183–1185. (c) Boyko, V. I.; Yakovenko, A. V.; Matvieiev, Y. I.;
Kalchenko, O. I.; Shishkin, O. V.; Shishkina, S. V.; Kalchenko, V. I.
Tetrahedron 2008, 64, 7567–7573. (d) Kliachyna, M. A.; Yesypenko, O. A.;
Pirozhenko, V. V.; Shishkina, S. V.; Shishkin, O. V.; Boyko, V. I.;
Kalchenko, V. I. Tetrahedron 2009, 65, 7085–7091. Upper-rim resolution:
(e) Xu, Z. X.; Zhang, C.; Zheng, Q. Y.; Chen, C. F.; Huang, Z. T. Org.
Lett. 2007, 9, 5331–5331. (f) Xu, Z.-X.; Li, G.-K.; Chen, C.-F.; Huang,
Z.-T. Tetrahedron 2008, 64, 8668–8675. (g) Xu, Z.-X.; Zhang, C.; Yang,
Y.; Chen, C.-F.; Huang, Z.-T. Org. Lett. 2008, 10, 477–479.
(6) Atkinson, R. C. J.; Gibson, V. C.; Long, N. J. Chem. Soc. ReV. 2004,
33, 313–328.
(1) Gutsche, C. D., Ed. Calixarenes: An Introduction, 2nd ed.; RSC:
Cambridge, 2008; 276 pp.
(2) No, K. H.; Gutsche, C. D. J. Org. Chem. 1982, 47, 2713–2719.
(3) Examples of inherently chiral calix[4]arenes in asymmetric catalysis:
(a) Shirakawa, S.; Moriyama, A.; Shimizu, S. Org. Lett. 2007, 9, 3117–
3119. (b) Shirakawa, S.; Moriyama, A.; Shimizu, S. Eur. J. Org. Chem.
2008, 5957–5964. (c) Shirakawa, S.; Kimura, T.; Murata, S.; Shimizu, S.
J. Org. Chem. 2009, 74, 1288–1296. (d) Shirakawa, S.; Shimizu, S. Eur. J.
Org. Chem. 2009, 1916–1924.
(7) Rosillo, M.; Dom´ınguez, G.; Pe´rez-Castells, J. Chem. Soc. ReV. 2007,
36, 1589–1604.
(4) First asymmetric synthesis of a calix[4]resorcinarene:Page, P. C. B.;
Heaney, H.; Sampler, E. P. J. Am. Chem. Soc. 1999, 121, 6751–6752.
(8) For an overview of directed metalations of aromatic compounds,
see: Clayden, J. Chem. Organolithium Compd. 2004, 1, 495–646.
10.1021/ol902238p CCC: $40.75
Published on Web 10/08/2009
 2009 American Chemical Society

Scheme 1
.
Synthesis of Isopropyloxazoline Calix[4]arene 4
Table 1. Selected Results of Ortholithiation of 4 with Different
Alkyllithiums
entrya
RLib (equiv)
time (h)
yieldc (%)
ded (%)
1
2
3
4
5
6
7
8
n-BuLi (3)
t-BuLi (6)
s-BuLi (3)
c-HexLi (6)
i-PrLi (3)
i-PrLi (3)
c-PentLi(6)
c-PentLi(5)
4
4.5
4.5
4.5
4
18
4.5
30
0
<5
66
<5
17
65
30
78
17
75
82
86
84
93
93
Chiral oxazoline calix[4]arene 4 was targeted as a model
to study this reaction (Scheme 1); starting from known amino
calix[4]arene 1,9 the desired oxazoline calix[4]arene was
readily obtained in three steps. Knochel’s recent version of
the Sandmeyer reaction converted arylamine 1 to the aryl
iodide 2 in 79% yield.10 Introduction of the carboxyl group
proved to be capricious when using a lithium-halogen
exchange protocol (followed by CO2 quench), often returning
up to 70% of the protonated byproduct. Various methods
and conditions were explored with the greatest success
coming from Knochel’s modified Grignard reagent,11 which
afforded the desired carboxyl calix[4]arene 3 in at least 60%
yield. The oxazoline could then be introduced using standard
conditions with L-valinol in 94% yield.
With the desired oxazoline calix[4]arene 4 in hand, our
attention turned to the feasibility of ortholithiation. Compet-
ing benzylic lithiation of the calix[4]arene was considered
in light of a report from Bennet et al.12 which demonstrated
this on a calix[4]arene; however, the relatively harsh condi-
tions (10-fold excess of n-butyllithium at room temperature)
suggested that this would not present any problems in our
studies.
For the electrophile it was decided to use an F+ reagent
since the diastereomeric ratio of the products could then be
determined by 19F NMR spectroscopy. Selectfluor was
unsuitable due to its poor solubility under the reaction
conditions, but N-fluorobenzenesulfonimide (NFSI) proved
to give good yields.13
Our preliminary attempts to ortholithiate 4 using n-
butyllithium, with or without the addition of N,N,N′,N′-
tetramethylethylenediamine (TMEDA), failed to produce any
product on quenching with NFSI, even at temperatures up
a Reactions performed in Et2O at -78 °C with 2 equiv of TMEDA per
equivalent of alkyllithium. b Number of equivalents of alkyllithium used.
c Determined by 1H NMR. d Determined by 19F-NMR. c-Hex ) cyclohexyl;
c-Pent ) cyclopentyl.
to 0 °C (Table 1, entry 1). No benzylic lithiation was
observed, with only starting material being recovered.
However the use of tert-butyllithium with TMEDA at -78
°C (entry 2) resulted in trace quantities of product with a
modest de of 17% (diastereomer signals found at -113.29
ppm and -114.25 ppm in the 19F NMR spectrum, Figure
1). Attempts to improve the yield by performing the reaction
at higher temperatures resulted in the major product arising
from the addition of tert-butyllithium onto the oxazoline.14
The use of sec-butyllithium, however, returned up to 66%
of the product with a de of 75% (entry 3). At higher
temperatures, the diastereoselectivity dropped off, e.g., at
-45 °C the reaction returned a lower de of 68%. It was also
found that TMEDA was essential for product formation when
the reactions were performed in diethyl ether.
Although others have observed dramatic variations in de’s
obtained using different alkyllithiums with chiral oxazoline
ortholithiation groups, no explanation for this has been
forthcoming. We were curious as to whether other sec-
alkyllithiums would return results similar to those obtained
using sec-butyllithium. We found that cyclohexyllithium
(entry 4) and isopropyllithium15 (entry 5) gave slightly better
de’s, although the yields were significantly lower (no side
product from alkyllithium addition onto the oxazoline was
observed). The low yield was presumably due to these
alkyllithium-TMEDA complexes being less reactive than
the sec-butyllithium-TMEDA complex with calix[4]arene
4. By extending the reaction with isopropyllithium to 18 h
the yield could be improved to a reasonable 65% (entry 6).
(9) Alemi, A. A.; Shaabani, B.; Dilmaghani, K. A.; Ganjali, S. T.
Molecules 2001, 6, 417–423.
(10) Krasnokutskaya, E. A.; Semenischeva, N. I.; Filimonov, V. D.;
Knochel, P. Synthesis 2007, 81–84.
(11) Krasovskiy, A.; Knochel, P. Angew. Chem., Int. Ed. 2004, 43, 3333–
3336.
(14) Addition of the base to the oxazoline has been observed previously:
Beak, P.; Kerrick, S. T.; Gallagher, D. J. J. Am. Chem. Soc. 1993, 115,
10628–10636.
(12) Scully, P. A.; Hamilton, T. M.; Bennett, J. L. Org. Lett. 2001, 3,
2741–2744.
(15) Surry, D. S.; Fox, D. J.; Macdonald, S. J. F.; Spring, D. R. Chem.
Commun. 2005, 2589–2590.
(13) Differding, E.; Ofner, H. Synlett 1991, 187–189.
Org. Lett., Vol. 11, No. 21, 2009
4987

Table 2. Various Electrophilic Quenches and Oxazoline
Removal
entrya electrophile
E
5 (yield, %)b 7 (yield, %)c
1
2
3
4
5
6
7
NFSI
NFSI
TMS-Cl
Ph2P-Cl
Cl-CO2Et
Me2S2
F
F
TMS
P(O)Ph2
CO2Et
5a (73)
ent-5a (63)
5b (62)
5c (70)
5d (73)
5e (73)
5f (72)
7a (78)
ent-7a (88)
7b (0)d
7c (99)
7d (72)
7e (41)
7f (72)
e
Selected 19F NMR spectra: (a) representative from a
f
Figure 1
.
poorly selective s-BuLi reaction at -45 °C; (b) representative from
c-PentLi reaction.
SMe
Br(CH2)2Br Br
a Conditions: -78 °C, 6 equiv of c-PentLi, 12 equiv of TMEDA, 18-24
h. b Isolated yields; de >92% as judged by 1H NMR. c Conditions: (i) AcOH,
170 °C, 1 h; (ii) NaOH, ethanol, 140 °C, 1 h. d Starting material
decomposition. e Isolated as a single diastereomer. f Ester hydrolysis
concomitant with removal of oxazoline.
The use of cyclopentyllithium gratifyingly gave a much
improved de of 93% and conversion of 78% after 30 h
(entries 7 + 8).16
We then turned our attention to the removal of the
oxazoline chiral auxiliary, which is known to be a stable
group and resistant to hydrolysis. We examined a number
of procedures but found that a two-step microwave reaction
was the most efficient in our hands. Thus, heating the
oxazoline in aqueous acetic acid at 170 °C for 1 h afforded
the amide intermediate which was transferred to ethanolic
sodium hydroxide and heated to 140 °C for a further 1 h. In
this way, we obtained the inherently chiral calixarene acid
7a in 78% yield. Without microwave heating, the reaction
was incomplete after 3 days of reflux.
mide afforded the expected products, though hydrolysis of
the methyl thioether calix[4]arene 5e returned a very poor
yield with concomitant degradation products being observed.
We then turned our attention to elucidating the structure
of the major diastereomer in this reaction. It was found that
phosphine oxide calix[4]arene 5c gave diffraction-quality
crystals when grown by slow evaporation of dichloromethane
from ethanol. The X-ray crystal structure (Figure 2) unam-
biguously revealed calix[4]arene 5 to be the major diaste-
reomer in this reaction.
In order to demonstrate the generality of the reaction, we
performed a series of electrophilic quenches, followed by
hydrolysis of the chiral oxazoline (Table 2). It was found
that the enantiomer of oxazoline calix[4]arene 4 resulted in
the same de (93%) being observed, giving the enantiomer
of 5a. Using chlorotrimethylsilane as an electrophile afforded
the expected product 5b; however, the hydrolysis of the
oxazoline resulted in decomposition of the starting material
(entry 3). Quenching with diphenylphosphine chloride (entry
4) resulted in a product that rapidly oxidized to phosphine
oxide 5c on workup. This product was crystalline and was
We next considered the origin of this diastereoselectivity.
The mechanism of ortholithiation is generally accepted to
occur via two steps: a fast reversible coordination of the
alkyllithium-TMEDA complex to the nitrogen of the
oxazoline followed by a slower deprotonation step.17 It is
therefore reasonable to suggest that aryllithium 5 (E ) Li)
is the major reaction intermediate and that calix[4]arene 5
is the major diastereomer for all reactions. The formation of
calix[4]arene 5 allows us to propose a likely transition state
for this reaction (Figure 3). For this the conformation of the
oxazoline is such that the bulky isopropyl group faces toward
the calix[4]arene cavity, allowing for an unobstructed ap-
proach of the alkyllithium from the calix[4]arene outer face.18
Current work is focused on gaining a better understanding
of the mechanism of this reaction, i.e., whether it is governed
by thermodynamic (i.e., conformational control of the
oxazoline or reactive intermediate) or kinetic factors (reaction
faster when isopropyl group facing toward cavity). For
1
obtained as a single diastereomer as judged by the H, 13C,
and 31P NMR spectra. Using ethyl chloroformate as an
electrophile afforded the diacid 7d after hydrolysis. This
compound proved to be sparingly soluable in most organic
solvents (DMSO, CHCl3, EtOH, MeCN, acetone, and
benzene), and its NMR spectra exhibited broad signals even
at elevated temperatures. It was found to be soluble in acetic
acid, and its NMR spectra (recorded in d4-acetic acid)
displayed sharp signals consistent with the proposed struc-
ture. Quenches with dimethyl disulfide and ethylene dibro-
(17) For a recent investigation into the mechanism, see: Chadwick, S. T.;
Ramirez, A.; Gupta, L.; Collum, D. B. J. Am. Chem. Soc. 2007, 129, 2259–
2268.
(18) This transition state is an analogous to that proposed for the work
on ferrocenes; see: Sammakia, T.; Latham, H. A. J. Org. Chem. 1996, 61,
1629–1635.
(16) A 93% de represents the lowest value we have measured over a
number of reactions.
4988
Org. Lett., Vol. 11, No. 21, 2009

Figure 3. Proposed transition state.
excesses,19 opening the door for further study into their
applications as chiral ligands for asymmetric synthesis. The
use of sec-butyllithium, which is more readily available from
commercial sources, still allows for a diastereoselectivity
much higher than that previously shown. This is the first
time that cyclopentyllithium (and cyclohexyllithium) have
been used as reagents for aromatic lithiation; the results
obtained using cyclopentyllithium suggest that this reagent
may have been overlooked in the literature and may have
an important role to play in other stereoselective lithiations.
Acknowledgment. We thank Dr. T. Jacobs for CIF
preparation and the University of Stellenbosch for financial
support; S.A.H. thanks the NRF, HB Thom, and Harry
Crossley Foundations for funding.
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds,
including spectra; crystallographic data for 5c (CIF).This
material is available free of charge via the Internet at
http://pubs.acs.org.
Figure 2. X-ray crystal stucture of phosphine oxide calix[4]arene
5c.
example, synthesis of a calix[4]arene analogue of 4 without
upper rim tert-butyl groups would help determine whether
they contribute any cooperative effects in determining the
diastereoselectivity.
In conclusion, we have demonstrated a practical synthesis
of inherently chiral calix[4]arenes with high enantiomeric
OL902238P
(19) ee’s of 7 are inferred to be the same as de’s of precursor 5
since racemization would involve the breaking and formation of aryl-X
bonds.
Org. Lett., Vol. 11, No. 21, 2009
4989