mirror-image related molecules is one of the best methods
for assessing affinity-independent specificity.5
Scheme 2. Facial Selectivity at the Dipole Ester
Both the regio- and diastereoselective applications of this
method are accomplished by exploiting primary6 and second-
ary7 orbital effects, furnishing exclusively bicyclic constructs
of syn/endo topology.3 Therefore, manipulation of absolute
stereochemistry remains a challenge for the synthesis of these
molecules in a stereoselectiVe manner.
The catalytic enantioselective cycloaddition of some 1,3-
dipoles, such as nitrones, has been reported.8 The application
of a chiral rhodium(II) catalyst for the asymmetric induction
of a carbonyl ylide cycloaddition has also been reported with
moderate selectivity (ca. 50% ee).9 This low enantioselec-
tivity, in conjunction with the evidence of metal-free
dipoles,10 indicates the transient nature of metal coordination
with the isomu¨nchnone dipole. While several diastereo-
selective cycloadditions with chiral isomu¨nchnones have been
reported,11 the stereogenic center was an integral part of the
molecule, not allowing use as a removable auxiliary.
To fulfill the criteria for the stereospecific construction
of scaffold 1, we required a removable enantioselective 1,3-
dipolar cycloaddition auxiliary. By incorporating a chiral
substituent at the ester position of the dipole (OR3, Scheme
1), we hoped to create an inductive effect that would be
proximal to the dipole to induce facial selectivity, yet
sensitive enough for subsequent removal. A variety of
potential auxiliaries were evaluated by coupling the corre-
sponding chiral malonic acid ester with methylacetamide.12
Diazotransfer subsequently furnished the 1,3-dipole precur-
sors 4, 6, and 8, derivatized with (-)-menthol, (-)-borneol,
and (R)-(-)-pantalactone respectively (Scheme 2). Treatment
of R-diazoimides 4, 6, and 8 with a catalytic amount of Rh2-
(pfbm)4 in the presence of tert-butyl vinyl ether led to
formation of the corresponding cycloadducts 5, 7, and 9 in
good to excellent yield. While each chiral dipole underwent
efficient cycloaddition, the terpene-based chiral auxiliary
failed to exert the required π facial bias and provided only
a modest level of selectivity.13 Despite low diastereoselec-
tivity, these preliminary results were encouraging enough
to warrant further exploration of this approach. An amino
acid based chiral auxiliary would allow the introduction of
stereoelectronic bias,14 while providing a modular design for
systematic optimization. In addition, amino acids afford a
diverse array of functionality and are available in enantio-
meric pairs.
R-Diazoimide 10 was synthesized with a phenylalanine-
based chiral auxiliary. Interestingly, upon exposure to Rh2-
(pfbm)4 in the presence of tert-butyl vinyl ether, L-phenyl-
alanine derivative 10 led to a 50:50 mixture of bicyclic
adducts (2R,5S)-11 and (2S,5R)-11. It was postulated that
the lack of selectivity was due to the flexibility of the
methylene spacer between the dipole and the chiral center.
R-Diazoimide 12 contains a more sterically demanding
â-branched isopropyl group, thereby limiting flexibility
adjacent to the dipole. Treatment of R-diazoimide 12 with
Rh2(pfbm)4 in the presence of tert-butyl vinyl ether provided
cycloadduct 13 in a 84% de in excellent yield. The absolute
stereochemistry of the major isomer was determined via
single-crystal X-ray analysis, which identified the (2Re-5Re)
face of the dipole as the preferred approach of dipolarophile.
Fortunately, both the syn-regiocontrol and endo-diastereo-
control were maintained, despite the added steric congestion
from the auxiliary.
(4) The importance of chirality in drug design has been recently
reviewed: Beroza, P.; Suto, M. J. Drug DiscoVery Today 2000, 5, 364.
(5) Liu, F.; Johnson, E. F.; Austin, D. J.; Anderson, K. F. Submitted.
(6) Osterhout, M. H.; Nadler, W. R.; Padwa, A. Synthesis 1994, 123.
(7) Savinov, S. N. Ph.D. Dissertation, Yale University, New Haven, CT,
2000.
(8) (a) Gothelf, K. V.; Jorgensen, K. A. Chem. Commun. 2000, 16, 1449-
58. (b) Jensen, K. B.; Roberson, M.; Jorgensen, Karl A. J. Org. Chem.
2000, 65, 9080-4.
(9) Hodgson, D. M.; Stupple, P. A.; Johnstone, C. Tetrahedron Lett. 1997,
38, 6471.
(10) Padwa, A.; Snyder, J. P.; Curtis, E. A.; Sheehan, S. M.; Worsencroft,
K. J.; Kappe, C. O. J. Am. Chem. Soc. 2000, 122, 8155-67.
(11) (a) Padwa, A.; Prein, M. J. Org. Chem. 1997, 62, 6842. (b) Padwa,
A.; Prein, M. Tetrahedron 1998, 54, 6957. (c) Angell, R.; Fengler-Veith,
M.; Finch, H.; Harwood, L. M.; Tucker, T. T. Tetrahedron Lett. 1997, 38,
4517.
To better understand how amino acid substitution affects
the level of diastereofacial induction, a series of R-hydroxy
acids, readily accessible from their corresponding R-amino
acids,15 were used to construct R-diazoimides with incre-
mental substitution. In analogy with the parent system 12,
four diazo substrates, 14, 16, 18, and 21, were prepared from
(14) A π-complex is a feature common to many diastereofacially robust
chiral auxiliaries: (a) Aoyagi, S.; Tanaka, R.; Naruse, M.; Kibayashi, C. J.
Org. Chem. 1998, 63, 8397. (b) Mezrhab, B.; Dumas, F.; d'Angelo, J.; Riche,
C. J. Org. Chem. 1994, 59, 500. (c) Maddaluno, J. F.; Gresh, N.; Giessner-
Prettre, C. J. Org. Chem. 1994, 59, 793.
(12) Rigo, B.; Fasseur, D.; Cauliez, P.; Couturier, D. Tetrahedron Lett.
1989, 30, 3073.
(13) The percent diastereoselectivity (% de) was determined by 1H NMR
integration of the diastereomeric pairs.
(15) Johnson, R. L. J. Med. Chem. 1980, 23, 666.
1416
Org. Lett., Vol. 4, No. 9, 2002