A new approach to enantiocontrol and enantioselectivity amplification: Chiral relay in Diels-Alder reactions

Full text of DOI:10.1021/ja016396b

8444
J. Am. Chem. Soc. 2001, 123, 8444-8445
A New Approach to Enantiocontrol and
Enantioselectivity Amplification: Chiral Relay in
Diels-Alder Reactions
Mukund P. Sibi,* Lakshmanan Venkatraman, Mei Liu, and
Craig P. Jasperse
Department of Chemistry
North Dakota State UniVersity
Fargo, North Dakota 58105-5516
Figure 1.
is effective, the chiral ligand is only required to bias the
conformation of the N(1) substituent, which in turn controls face
selectivity; the ligand itself need not shield the reaction center.
But any inherent bias of the chiral Lewis acid could be either
consonant or dissonant with the relay group. Consonant chiral
relay would be an attractive approach to amplify selectivity.
To examine if our chiral relay design is operative, Diels-Alder
cycloaddition of cyclopentadiene to crotonates was undertaken
2
using Cu(OTf) /bisoxazolines as the chiral Lewis acid (eq 1).
The reaction choice was based on two key factors: (1) the well-
established square planar-like geometry for Cu(II) complexes with
bisoxazoline ligands and bidentate substrates and (2) literature
ReceiVed June 12, 2001
ReVised Manuscript ReceiVed July 5, 2001
Development of new strategies to access enantiomerically pure
1
compounds is at the forefront of synthetic organic chemistry. In
this context, chiral Lewis acid catalysis has emerged as one of
the premiere methods to control stereochemistry. Much effort has
8
gone into the design of superior ligands with increased steric
extension, to shield distant reactive sites.2 We have instead
considered a “chiral relay” approach, focusing on the improved
design of achiral templates which may indirectly relay and amplify
stereoselectivity from ligands, including ligands with minimal
steric bias.
9
data for DA reactions with nonrelay substrates for easy compari-
son.
Davies has recently introduced the relay concept, using a chiral
auxiliary to install asymmetry.3 Our approach differs from Davies
in that in our relay network, asymmetry originates with a chiral
Lewis acid but is then relayed/amplified via an achiral template.5
There are no reports in the literature in which an achiral template
has been applied for stereocontrol in a systematic way. The design
of our achiral template took into consideration several require-
ments: (1) easy variation of the relay group R, (2) ready
accessibility, (3) easy attachment of the appropriate reaction
fragment, and (4) the presence of donor sites suitable for rotamer
and Lewis acid coordination control.
,4
The DA reactions were carried out at room temperature with
substrates 4a-e and excess cyclopentadiene using 15 mol % of
the catalyst (equation 1, Table 1). On the average, the DA
reactions took 24 h for completion, and yields ranged from 85 to
90%. The endo selectivity was generally good. For ease of ee
determination and absolute stereochemistry analysis, the DA
adducts 6 were converted to the known benzyl ester 7. For our
initial experiments, we chose bisoxazoline 8 whose C-4 methyl
substituent is too small to provide high selectivity in the absence
of chiral relay. As the effective size of the relay group increases
(H < Et < Bn < CH -2-naphthyl, CH -1-naphthyl), so does the
enantioselectivity (entries 1-5). Use of the bulky 1-naphthyl-
methyl relay group (compound 4e) gave 86% ee (entry 5).
Decreasing the reaction temperature to -23 °C further increased
the ee to 96% (entry 6). The data presented in the table suggests
that enantioselectiVity for the major endo adduct correlates
directly with the size of the relay group. The high ee for reaction
with 4e clearly indicates that chiral relay is operatiVe. This high
selectivity with ligand 8 is very impressive since one requires
A novel class of achiral templates which met the above criteria
6
were the pyrazolidinones 1 (Figure 1). The parent (1 R ) H) is
readily available by conjugate addition of hydrazine to 3,3-
7
dimethylacrylate. The relay group R was installed either by
N
simple S 2 alkylation or by reductive amination of aldehydes,
taking advantage of the higher reactivity of the N-1 nitrogen. We
reasoned that in the presence of a chiral Lewis acid, the tetrahedral
N(1)-nitrogen would undergo inversion (2f3) and preferentially
equilibrate to an asymmetric conformation 2 or 3. With the N(1)
conformation determined by the Lewis acid, the substituent R
2
2
(shown as a circle) should in turn provide shielding. In essence,
the chiral Lewis acid would effectively convert an achiral auxiliary
into a chiral auxiliary. It should be emphasized that if chiral relay
(
1) Stinson, S. C. Chem. Eng. News 2001, May 14, 45-56.
(
2) For recent monographs see: Catalytic Asymmetric Synthesis, Ojima, I.
Ed.; Wiley-VCH: New York; 2000. ComprehensiVe Asymmetric Catalysis;
Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer: New York; 1999.
2
the bulky chiral Lewis acid tert-butylbisoxazoline/Cu(OTf) to
(3) (a) Bull, S. D.; Davies, S. G.; Fox, D. J.; Garner, A. C.; Sellers, T. G.
obtain high selectivity in DA reactions with nonrelay oxazolidi-
R. Pure Appl. Chem. 1998, 70, 1501. (b) Bull, S. D.; Davies, S. G.; Epstein,
S. W.; Ouzman, J. V. A. Chem. Commun. 1998, 659. (c) Bull, S. D.; Davies,
S. G.; Epstein, S. W.; Leech, M. A.; Ouzman, J. V. A. J. Chem. Soc., Perkin
Trans. 1 1998, 2321.
9
none crotonate 11. The absolute stereochemistry of the major
9
adduct of 6e was determined to be (2S,3R). That the stereo-
chemical outcome with substrates 4a-e is identical to that
obtained using oxazolidinone crotonate 11 suggests similar
coordination geometries in the two series. The DA reactions with
ligands 9 and 10, which have medium-sized substituents (i-Pr, 9
and Bn, 10), follow the same trend as with ligand 8. Once again,
substrate 4e gave the highest selectivity: 92 and 85% ee at room
temperature and 99 and 96% ee at -23 °C respectiVely. The
results with 4e are in stark contrast to the low selectiVities
(
4) Clayden, J.; Pink, J. H.; Yasin, S. A. Tetrahedron Lett. 1998, 39, 105.
(
5) Only a few examples have been reported to date that apply this concept
in enantioselective reactions. (a) Quaranta, L.; Renaud, P. Chimia 1999, 53,
3
(
64. (b) Quaranta, L.; Ph. D thesis, University of Fribourg, Switzerland, 2000.
c) Balsells, J.; Walsh, P. J. J. Am. Chem. Soc. 2000, 122, 1802. (d) Hiroi,
K.; Ishii, M. Tetrahedron Lett. 2000, 41, 7071. (e) Wada, E.; Pei, W.;
Kanemasa, S. Chem. Lett. 1994, 2345. (f) Wada, E.; Yasuoka, H.; Kanemasa,
S. Chem. Lett. 1994, 1637. (g) Evans, D. A.; Campos, K. R.; Tedrow, J. S.;
Michael, F. E.; Gagn e´ , M. R. J. Am. Chem. Soc. 2000, 122, 7905.
(
6) For detailed procedures of synthesis and characterization data see
Supporting Information.
7) For the effect of gem alkyl substitution on conformation control see:
a) Onimura, K.; Kanemasa, S. Tetrahedron 1992, 48, 8631. (b) Bull, S. D.;
Davies, S. G.; Key, M.-S.; Nicholson, R. L.; Savory, E. D. Chem. Commun.
000, 1721.
(
(8) (a) Dias, L. C. J. Braz. Chem. Soc. 1997, 8, 289. (b) Johnson, J. S.;
Evans, D. A. Acc. Chem. Res. 2000, 33, 325.
(9) Evans, D. A.; Miller, S. J.; Lectka, T.; von Matt, P. J. Am. Chem. Soc.
1999, 121, 7559.
(
2
1
0.1021/ja016396b CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/07/2001

Communications to the Editor
J. Am. Chem. Soc., Vol. 123, No. 34, 2001 8445
Table 1. Diels-Alder Reactions with Relay Templates Using C
2
Chiral Bisoxazolinesa
a
For reaction details see Supporting Information. Endo/exo ratios were determined by NMR, and ee determination was carried out using chiral
b
c
HPLC. Yields for isolated column purified material averaged around 85-90%. Reaction at -23 °C using 50 mol % chiral Lewis acid. 8% ee
was reported in ref 10 for this adduct.
Table 2. Reactions with Relay Templates Using C
Ligands
1
Chiral
with ligand 12 clearly establishes that the relay group primarily
provides the face shielding, not the ligand, and that chiral relay
is indeed operative. Cycloadditions using chiral Lewis acids
derived from ligands 13 and 14 showed the same trend as with
a
12.
The results detailed in Tables 1 and 2 can be explained using
9
a square planar geometry model relative to Cu, with the
bisoxazoline occupying two sites, and with bidentate coordination
by the template in an s-cis conformation (complex with 12 is
shown). The ligand isopropyl group in the lower left front
quadrant orients the relay group to the upper right rear quadrant,
where the relay group blocks the back face of the crotonate. Thus
the 1-naphthylmethyl group “relays” chirality from the ligand’s
isopropyl group to the reactive centers in the substrate. In reactions
2
with the C chiral bisoxazoline 9, the additional ligand isopropyl
group would occupy the upper left rear quadrant and reinforce
shielding of the back face of the crotonate. Thus in reactions with
a
Endo/exo ratios were determined by NMR and ee determination
was carried out using chiral HPLC. Yields for isolated column purified
b
material averaged around 90%. Reaction at -23 °C using 50 mol %
chiral Lewis acid.
obserVed with oxazolidinone crotonate 11, a nonrelay substrate
(
entry 7).10 These results exemplify chiral amplification between
the chiral ligand and the relay group.
A more stringent test for demonstrating chiral relay was devised
using non-C symmetric ligands 12-14, which afford almost no
2
11
enantioselectivity in the absence of chiral relay (Table 2). We
surmised that ligand 12, like 9, should provide a square planar
complex, but should now force the relay group to occupy an
opposite quadrant relative to the ligand’s isopropyl group.
Reaction using 12 showed the same trend as was observed with
2
the C ligands both the relay group and the C-4 substituent act in
concert to mutually amplify the selectivity. These observations
2
the C symmetric ligands: larger relay groups gave higher
are similar to double diastereoselection experiments described by
Evans involving both chiral Lewis acids and chiral auxiliaries.
9
12
selectivity (entries 1-5) with 4e again being the best (entry 5).
Lowering the temperature (compare entry 5 with 6) gave
selectivity as high as 88% at -23 °C. The stereochemical outcome
for reaction using ligand 12 was the same as with ligand 9 (2S,3R).
DA reactions with oxazolidinone crotonate 11 using ligands 12-
In conclusion, we have demonstrated that chiral relay is a novel
and promising strategy for obtaining high selectivity in enantio-
selective transformations using simple ligands. This approach
should also find utility in situations where one requires enhance-
ment of enantioselectivity for synthetic applications. Experiments
are underway to evaluate chiral relay in other enantioselective
transformations of general interest.
14 were nearly racemic (entry 7). The high selectivity observed
(
10) Takacs, J. M.; Lawson, E. C.; Reno, M. J.; Youngman, M. A.; Quincy,
D. A. Tetrahedron: Asymmetry 1997, 8, 3073.
11) There are reports on the use of non-C
(
2
symmetric ligands in DA
Acknowledgment. Our research programs are supported by grants
from NSF and NIH.
reactions. For selected examples see: (a) Ichiyanagi, T.; Shimizu, M.; Fujisawa,
T. J. Org. Chem. 1997, 62, 7937. (b) Sagasser, I.; Helmchen, G. Tetrahedron
Lett. 1998, 39, 261.
Supporting Information Available: Characterization data for com-
pounds 1-14 and experimental procedures (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
(12) With both symmetric and nonsymmetric ligands, minor enantiomers
may result from imperfect shielding within the indicated model or if the relay
group partially populates the upper right front quadrant. In reactions with
nonsymmetric ligands 12-14, alternate ligand coordination such that the
isopropyl group is in the upper left rear quadrant could further erode selectivity.
JA016396B