Exo Selective Enantioselective Nitrone Cycloadditions

Article abstract of DOI:10.1021/ja039087p

We have developed a novel method for accessing exo adducts with high enantioselectivity in nitrone cycloadditions to enoates. Pyrazolidinones proved to be effective achiral templates in the cycloadditions, providing exo adducts typically in >15:1 selectivity and 90-98% ee. The use of Lewis acids that form square planar complexes, such as copper triflate, was important for obtaining high exo selectivity. Copyright

Full text of DOI:10.1021/ja039087p

Published on Web 12/30/2003
Exo Selective Enantioselective Nitrone Cycloadditions
Mukund P. Sibi,* Zhihua Ma, and Craig P. Jasperse
Department of Chemistry, North Dakota State UniVersity, Fargo, North Dakota 58105
Received October 16, 2003; E-mail: mukund.sibi@ndsu.nodak.edu
Scheme 1
Dipolar cycloadditions of nitrones to alkenes have received much
synthetic interest1 because they generate three contiguous stereo-
centers, and the products can be converted by reductive cleavage
to â-amino acids. A number of different chiral Lewis acids have
been used to catalyze nitrone cycloadditions that are regioselective,
diastereoselective, and enantioselective.1,2 Highly enantioselective
examples involving electrophilic olefins 1 have consistently pro-
duced endo diastereomers 4.2,3 At present, there are no reported
methods which provide access to the exo isomers 3 with both high
enantioselectivity and high chemical efficiency (Scheme 1).4 In this
paper, we describe examples of highly exo- and enantioselective
nitrone cycloadditions using novel templates5 developed in our
laboratory.
Table 1. Exo Selective Nitrone Additions: Effect of Chiral Lewis
Acidsa
When Lewis acids coordinate to substrates 1, the pathway leading
to exo product 3 is frequently destabilized because the R2 group
interacts with the metal complex. In keeping with literature
precedents,4 we reasoned that a square planar metal complex would
reduce these steric interactions and be more amenable to an exo
mode of nitrone attack.
Given our interest in pyrazolidinone templates,5 we evaluated
nitrone additions to substrate 5, using catalytic amounts (30 mol
%) of chiral Lewis acids derived from metal salts and bisoxazolines
9-12 (eq 1).6 These results are tabulated in Table 1. Reaction using
Cu(OTf)2 and ligand 9 as a Lewis acid (entries 1-4) was effective
and gave a high yield of the exo isomer in good ee (entry 1).7
Increasing the bulk of the chiral ligand (10, 11) did not lead to
enhancement in ee (entries 1-3). The modest ee using tert-butyl
ligand 11 is interesting because this ligand in combination with
cupric salts often gives excellent selectivity in a variety of
transformations.8 Much better results were observed using the
aminoindanol-derived ligand 12 (entry 4), which resulted in 98%
ee for the exo isomer 7 in high yield and with 96:4 exo-endo
diastereoselectivity. These results establish for the first time that
the exo adduct can be obtained in high yield and enantioselectivity.
Molecular sieves (MS) have been reported to play an influential
role in a variety of nitrone cycloadditions.9 In our reactions, addition
of MS led to a dramatic reversal (entry 5 versus 1) or reduction
(entry 6 versus 4) in exo/endo selectivity, but without compromising
the enantioselectivity of the exo isomer.10 A similar impact on exo/
endo selectivity by MS was observed when Cu(ClO4)2 was used
(compare entry 7 with 8). A series of other Lewis acids were also
screened using ligand 9 (compare entries 9-12 with entry 1). Zinc
triflate was modestly exo-selective, but with low enantioselectivity
(entry 9). In contrast, magnesium and iron Lewis acids gave nearly
equal amounts of the exo/endo adducts with low ee’s (entries 10
and 11), while scandium triflate gave only the endo adduct with
no selectivity (entry 12). These studies suggest that a square planar
metal complex is best for providing high exo selectivity.11
Having identified a viable chiral Lewis acid for providing the
exo adduct, we evaluated the role of the template with respect to
enantioselectivity (eq 2, Table 2). The bisoxazoline 9 with the
smallest shielding group was used in these experiments to better
% yieldd
(SM)
exo/
endo
exo
endo
c
ent
Lewis acidb
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(ClO4)2‚6H2O
Cu(ClO4)2‚6H2O
Zn(OTf)2
Mg(OTf)2
Fe(NTf2)2
Sc(OTf)3
lig
%(ee)e
%(ee)e
1
2
3
9
96
90/10
79/21
83/17
96/4
19/81
63/37
83/17
32/68
86/14
57/43
45/55
2/98
75
71
73
98
72
98
71
67
47
10
11
12
9
12
9
9
9
9
9
91 (5)
24 (71)
94 (4)
88 (9)
92 (6)
93
50
4
5f
6f
7
47
80
8f
9
96
52
90 (7)
82 (15)
44 (53)
98
10
11
12
27
-17
-12
9
2
a For details of the reaction conditions, see Supporting Information. b 30
mol % Lewis acid. c Diastereomer ratio determined by 1H NMR (500 MHz).
d Isolated yield for the product after chromatography. The amount of starting
material recovered is after column chromatography. e Determined by chiral
HPLC. f MS 4 A was added.
assess the importance of the substituents on the pyrazolidinone ring.
In the presence of a chiral Lewis acid, the fluxional nitrogen in the
template can adopt a preferred configuration such that the pyra-
zolidinone ring effectively becomes a chiral auxiliary which can
at times amplify stereoselectivity. Initially, we investigated the effect
of the N1-substituent (the “chiral relay” group) in substrate 5, with
R1 fixed. The enantioselectivity for the major exo adduct correlated
with the steric bulk of the N1-substituent: increasing the bulk of
the relay group from ethyl to benzyl to naphthylmethyl led to an
increase in ee from 78% to 86% (entries 1-3). Decreasing the
reaction temperature to 3 °C led to a small improvement in ee from
86% to 91% (compare entry 3 with 4). The nature of R1 also
impacted the ee of the major exo adduct. With benzyl as the N1-
substituent, varying R1 from methyls to cyclohexyl to benzyls also
led to a systematic increase in the ee for the exo adduct (compare
9
718
J. AM. CHEM. SOC. 2004, 126, 718-719
10.1021/ja039087p CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
Table 2. Evaluation of Relay Templates
ent
R
R
1
no.
yielda
exo/endob
exo (ee)c
Figure 1. Stereochemical model.
1
2
3
4d
5
6
7
CH3
phenyl
1-naphthyl
1-naphthyl
phenyl
CH3
CH3
CH3
CH3
cyclohexyl
benzyl
5b
5a
5c
5c
5d
5e
5f
96
96
93
94
93
94
97
92/8
78
75
86
91
79
93
85
90/10
89/11
87/13
95/5
91/9
90/10
gave high diastereo- and enantioselectivity (entry 6). Both electron-
deficient (6c) and electron-rich (6d) nitrones gave addition products
in modest yields, although enantioselectivity remained high in both
cases (entries 7 and 8). Reaction with N-phenyl substituted nitrone
6e was very efficient and highly enantioselective, but the exo/endo
selectivity was low (entry 9).
Our results suggest that the key to exo selectivity is the use of
a Lewis acid that forms square planar complexes (Figure 1 for
substrate 5c, nitrone 6a, and ligand 9, and copper triflate).11 The
exo attack is not sterically encumbered in this complex. As was
demonstrated in our previous work,5a the pyrazolidinone relay
template in a square planar arrangement works in concert with the
ligand to amplify enantioselectivity.
phenyl
1-naphthyl
cyclohexyl
a Isolated yield. b Diastereomer ratio determined by 1H NMR (400 or
500 MHz). c Determined by chiral HPLC. d Reaction at 3 °C.
Table 3. Reactions with Various Dipolarophiles and Nitrones
Acknowledgment. This work was supported by NSF-CHE-
0316203 and NSF-EPS-0132289.
yielda
(SM)
exo/
endob
ee exo
(endo)c
ent
subs. (R)
R
1
R
2
Supporting Information Available: Characterization data for
compounds 5-12 and experimental procedures (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
1
2
3
H 5g
CH3 5a
Et 5h
CH3
CH3
CH3
CH3
CH3
Bn
Bn
Bn
Ph
Ph 6a
85
94
66/34
96/4
94/6
84/16
67/33
93/7
85/15
95/5
99 (12)
98
99
75
85 (45)
99
98
98
99 (90)
Ph 6a
Ph 6a
Ph 6a
Ph 6a
Ph 6b
4-ClPh 6c
4-MeOPh 6d
Ph 6e
88 (10)
23 (76)
44 (37)
92
52 (41)
45
4d,e
5f
6
Ph 5i
CO2Et 5j
CH3 5a
CH3 5a
CH3 5a
CH3 5a
References
(1) (a) Karlssson, S.; Hogberg, H.-E. Org. Prep. Proced. Int. 2001, 33, 103.
(b) Gothelf, K. V.; Jørgensen, K. A. Chem. Commun. 2000, 1449. (c)
Kanemasa, S. Synlett 2002, 1371. (d) Confalone, P. N.; Huie, E. M. Org.
React. 1988, 1.
7
8g
9
85 (13)
52/48
(2) (a) Kanemasa, S.; Oderaotoshi, Y.; Tanaka, J.; Wada, E. J. Am. Chem.
Soc. 1998, 120, 12355. (b) Karlsson, S.; Hogberg, H.-E. Tetrahedron:
Asymmetry 2002, 13, 923. (c) Jen, W. S.; Wiener, J. J. M.; MacMillan,
D. W. C. J. Am. Chem. Soc. 2000, 122, 9874. (d) Kobayashi, S.;
Kawamura, M. J. Am. Chem. Soc. 1998, 120, 5840. (e) Iwasa, S.;
Tsushima, S.; Shimada, T.; Nishiyama, H. Tetrahedron 2002, 58, 227.
(f) Desimoni, G.; Faita, G.; Mortoni, A.; Righetti, P. Tetrahedron Lett.
1999, 40, 2001.
a Isolated yield. b Diastereomer ratio determined by 1H NMR (400 or
500 MHz). c Determined by chiral HPLC. d Ligand 9 was used in place of
12. e MS 4 A was added. f 50 mol % catalyst was used. g O2 balloon was
used.
entries 2, 5, and 6). On the other hand, when both N1- and C5-
substituents are large, the exo adduct enantioselectivity does not
change (compare entry 3 with 7). These results suggest that there
is a fine balance between the N1 and C5 substituents in providing
optimal levels of selectivity. When the pyrazolidinone ring was
replaced by a simple oxazolidinone, the reaction proceeded in high
yield (98%) and exo selectivity (98:2), but with modest enantiose-
lectivity (40% ee). This result suggests that the chiral relay templates
amplify enantioselectivity but are not the source of exo selectivity.
We have carried out a small breadth and scope study by varying
the enoyl substrate (entries 1-5) as well as the nitrone (entries
6-10), and these results are tabulated in Table 3 (eq 3). We chose
ligand 12 which had provided high selectivity in Table 1. The parent
acrylate 5g gave the exo isomer in good yield and selectivity (entry
1). As illustrated earlier, the crotonate (5a) gave high selectivity
(entry 2). The â-ethyl (5h) and cinnamate (5i) substrates were less
reactive, but the ee’s for the major exo adducts were high (entries
3 and 4). The fumarate (5j) was less efficient and somewhat less
selective (entry 5). In terms of nitrone structure, compound 6a was
very reactive and also gave the highest exo:endo selectivity (entry
2). A benzyl group on the nitrogen as in 6b was well tolerated and
(3) For a discussion, see: Gothelf, K. V.; Hazell, R. G.; Jørgensen, K. A. J.
Org. Chem. 1996, 61, 346.
(4) For reactions proceeding with good exo diastereoselectivity or enanti-
oselectivity, see: (a) Jensen, K. B.; Gothelf, K. V.; Hazell, R. G.;
Jørgensen, K. A. J. Org. Chem. 1997, 62, 2471. (b) Gothelf, K. V.;
Jørgensen, K. A. J. Org. Chem. 1994, 59, 5687. (c) Crosignani, S.;
Desimoni, G.; Fiata, G.; Filippone, S.; Mortoni, A.; Righeti, P. P.; Zema,
M. Tetrahedron Lett. 1999, 40, 7007. (d) Hori, K.; Kodama, H.; Ohta,
T.; Furukawa, I. J. Org. Chem. 1999, 64, 5017. (e) Gothelf, K. V.; Hazell,
R. G.; Jørgensen, K. A. J. Org. Chem. 1998, 63, 5483.
(5) Achiral templates with fluxional groups in synthesis, see: (a) Sibi, M.
P.; Venkatraman, L.; Liu, M.; Jasperse, C. P. J. Am. Chem. Soc. 2001,
123, 8444. (b) Corminboeuf, O.; Quaranta, L.; Renaud, P.; Liu, M.;
Jasperse, C. P.; Sibi, M. P. Chem.-Eur. J. 2003, 9, 28. (c) Sibi, M. P.;
Liu, M. Org. Lett. 2001, 3, 4181.
(6) See Supporting Information for details on the synthesis of the starting
materials, nitrones, stereochemistry analysis, and experimental conditions.
(7) The identities of the exo and endo adducts were established by converting
them to known compounds. See Supporting Information for details.
(8) Johnson, J. S.; Evans, D. A. Acc. Chem. Res. 2000, 35, 325.
(9) (a) References 2a, 4a, and 4c. (b) Kawamura, M.; Kobayashi, S.
Tetrahedron Lett. 1999, 40, 3213.
(10) These results suggest that there may be competing pathways when MS
are added. The high enantioselectivity for the exo adduct with or without
MS suggests that it may be formed without the involvement of the sieves.
(11) This conclusion is consistent with the high exo selectivity with copper as
a Lewis acid (square planar complex) rather than magnesium or iron Lewis
acids (typically tetrahedral or octahedral geometry).
JA039087P
9
J. AM. CHEM. SOC. VOL. 126, NO. 3, 2004 719