Chelation-controlled 1,3-asymmetric induction in radical addition to γ- hydroxy- and γ-alkoxy-α-methylenecarboxylic esters

Article abstract of DOI:10.1055/s-1999-2554

The radical-mediated reactions of γ-hydroxy- and γ-alkoxy-α- methylenecarboxylic esters 3 (R¹ = Ph, i-Bu, and t-Bu, R² = H, Me, MOM, and MEM) with isopropyl iodide or cyclohexyl iodide performed in the presence of Lewis acids gave the syn-adducts 4 predominantly, whereas the anti-adduct 5 was the major product in the reaction of 3 (R¹ = Ph, R² = Me) with t-butyl iodide.

Full text of DOI:10.1055/s-1999-2554

LETTER
53
Chelation-Controlled 1,3-Asymmetric Induction in Radical Addition to
g-Hydroxy- and g-Alkoxy-a-methylenecarboxylic Esters
Hajime Nagano,* Satoko Toi, and Tomoko Yajima
Department of Chemistry, Faculty of Science, Ochanomizu University, Otsuka, Bunkyo-ku, Tokyo 112-8610, Japan
Fax: +81-3-5978-5348; e-mail: nagano@hososipc.chem.ocha.ac.jp
Received 29 September 1998
Abstract: The radical-mediated reactions of g-hydroxy- and g-
alkoxy-a-methylenecarboxylic esters 3 (R1 = Ph, i-Bu, and t-Bu, R2
= H, Me, MOM, and MEM) with isopropyl iodide or cyclohexyl io-
dide performed in the presence of Lewis acids gave the syn-adducts
4 predominantly, whereas the anti-adduct 5 was the major product
in the reaction of 3 (R1 = Ph, R2 = Me) with t-butyl iodide.
Key words: radical, 1,3-asymmetric induction, chelation, Lewis
acid, γ-alkoxy-α-methylenecarboxylic esters
The chelate ring formation of radical intermediates with
Lewis acid plays an important role in the stereochemical
control of acyclic radical reactions.1,2 We have recently
shown that the allylation of a-bromo-b-siloxy esters 1
conducted in the presence of Lewis acid proceeded
through the transition state model A involving a seven-
membered chelate ring and yielded the syn-product 2 pre-
dominantly (Scheme 1).3 We now report the chelation
controlled 1,3-asymmetric induction in radical addition to
g-hydroxy- and g-alkoxy-a-methylenecarboxylic esters 3.
Little is known about 1,3-asymmetric induction in radical
reactions.4
The Reformatsky reaction of aldehydes R1-CH=O (R1 =
Ph, t-Bu, and i-Bu) with ethyl a-bromomethylacrylate
gave racemic g-hydroxy-a-methylenecarboxylic esters 3
Scheme 1
in high yields (≥86%).5,6 Methylation of the alcohol 3 (R1
= Ph, R2 = H) with methyl iodide and silver(I) oxide gave
methyl ether 3 (R1 = Ph, R2 = Me) in 56% yield together
the comparison of their 1H NMR spectra with those of au-
thentic g-lactones prepared from a-methylene-g-lactone 6
following the reported procedures.8 Methylation of 4 and
5 (R1 = Ph, R2 = H) with methyl iodide and silver(I) oxide
gave the corresponding methyl ethers 4 and 5 (R1 = Ph, R2
= Me), respectively. The stereochemistry of 4 and 5 (R1 =
t-Bu and i-Bu, R2 = H, MOM, and MEM) was assigned by
comparing their 1H NMR spectral data with those of 4 and
5 (R1 = Ph, R2 = H and Me).9
with g-lactone 6. However, methyl ethers of the alcohols
3 (R1 = t-Bu, and i-Bu, R2 = H) were not obtained due to
the formation of the corresponding g-lactones. Meth-
oxymethyl (MOM) and methoxyethoxymethyl (MEM)
ethers 3 (R1 = Ph, t-Bu, and i-Bu, R2 = MOM and MEM)
were prepared from the corresponding alcohols 3 (R1 =
Ph, t-Bu, and i-Bu, R2 = H) following the standard proce-
dures.
After a 10 min complexation time, the alkylation of acry-
lates 3 was conducted with alkyl iodide R3I (3 equiv.), n-
Bu3SnH (2 equiv.), and Et3B (0.3 equiv.) as a radical
initiator7 in CH2Cl2 at 0 ∞C. The concentration of 3 was
0.07–0.13 mol dm-3 in all the reactions. The diastereomer
ratios of the products were determined by 1H NMR anal-
ysis. The stereochemistry of 4 and 5 was determined as
follows. Treatment of the mixture of hydroxy esters 4 and
5 (R1 = Ph, R2 = H, R3 = i-Pr; 4 : 5 = 2 : 1) with p-toluene-
sulfonic acid in benzene gave g-lactones 7 and 8 (7 : 8 = 2
: 1). The assignment of the g-lactones was performed by
A summary of the addition reactions is given in Table 1.
In the absence of Lewis acid, the reactions of 3 showed
poor stereoselectivity (4 : 5 = 1 : 1.4–1.6, entry 1) except
Synlett 1999, No. 1, 53–54 ISSN 0936-5214 © Thieme Stuttgart · New Yorkrk

54
H. Nagano et al.
LETTER
for that of 3 (R1 = i-Bu, R2 = MEM; entry 16). The diaste- groups. The high syn selectivity of 3 (R1 = t-Bu, R2 =
reoselectivity was remarkably affected when the reaction MOM; entry 14) reflects the very large interaction be-
was conducted in the presence of Lewis acid. Use of 3 tween the bulky t-butyl and i-butyl groups in model C.
equiv of MgBr2·OEt2 reversed the diastereoselectivity of The anti selectivity in the reaction of 3 (R1 = Ph, R2 = Me)
the reaction of alcohol 3 (R1 = Ph, R2 = H; entry 2) with with t-BuI (entries 9 and 10) may be ascribable to the
isopropyl iodide,2k but low selectivity (entry 2). The reac- shielding of the upper face of model B by the bulky neo-
tion of the methyl ether 3 (R1 = Ph, R2 = Me) with isopro- pentyl group. We have shown that the shielding of the up-
pyl iodide or cyclohexyl iodide performed in the presence per face of model A by the bulky t-BuPh2SiO group
of MgBr2·OEt2, MgBr2, ZnCl2, or Eu(fod)3 [= lowered the syn selectivity in the allylation of 1 (R =
tris(6,6,7,7,8,8,8-heptafluoro-2,2-dimethyl-3,5-octanedi-
onato)-europium] gave higher selectivities (entries 3–7).
As expected from our previous results in the allylation of
1,3 La(fod)3 was highly efficient (entry 8). MgI2 was less
effective, and tris(2,4-pentadionato)lanthanum and
tris(1,3-diphenyl-1,3-propanedionato)-lanthanum] had no
effect on the stereocontrol. The reaction of the methyl
ether 3 with t-butyl iodide performed in the presence of
Lewis acid gave the anti-product 5 predominately (entries
9 and 10).
SiPh2t-Bu).3
The MOM and MEM ethers 3 (R1 = Ph, R2 = MOM and
MEM; entries 11, 12, and 15) gave a poorer result than the
methyl ether 3. In the reactions of 3 (R1 = t-Bu and i-Bu,
R2 = MOM and MEM), use of Lewis acid reversed the
diastereoselectivity, but the selectivities were low (entries
13 and 16–18) except for 3 (R1 = t-Bu, R2 = MOM; entry
14 ).
References and Notes
(1) For reviews of stereoselective acyclic radical reactions, see:
Porter, N. A.; Giese, B.; Curran, D. P. Acc. Chem. Res. 1991,
24, 296. Smadja, W. Synlett, 1994, 1. Curran, D. P.; Porter, N.
A.; Giese, B. “Stereochemistry of Radical Reactions; Con-
cepts, Guidelines, and Synthetic Applications,” VCH, Wein-
heim (1996).
(2) (a) Wu, J. H.; Zhang, G.; Porter, N. A. Tetrahedron Lett. 1997,
38, 2067. (b) Miyabe, H.; Ushiro, C.; Naito, T. J. Chem. Soc.,
Chem. Commun. 1997, 1789. (c) Sibi, M. P.; Ji, J. Angew.
Chem., Int. Ed. Engl. 1997, 36, 274. (d) Sibi, M. P.; Ji, J. J.
Org. Chem. 1996, 61, 6090. (e) Gerster, M.; Schenk, K.;
Renaud, P. Angew. Chem., Int. Ed. Engl. 1996, 35, 2396.
(f) Gerster, M.; Audergon, L.; Moufid, N.; Renaud, P. Tetra-
hedron Lett. 1996, 37, 6335. (g) Guindon, Y.; Guérin, B.; Cha-
bot, C.; Ogilvie, W. J. Am. Chem. Soc. 1996, 118, 12528. (h)
Nishida, M.; Nishida, A.; Kawahara, N. J. Org. Chem. 1996,
61, 3574. (i) Nishida, A.; Hayashi, H.; Yonemitsu, O.; Kawa-
hara, N. Synlett 1995, 1045. (j) Urabe, H.; Yamashita, K.; Su-
zuki, K.; Kobayashi, K.; Sato, F. J. Org. Chem. 1995, 60,
3576. (k) Nagano, H.; Azuma, Y. Chem. Lett. 1996, 845, and
references cited therein.
(3) Nagano, H.; Kuno, Y.; Omori, Y.; Iguchi, M. J. Chem. Soc.,
Perkin Trans. 1 1996, 389. Nagano, H.; Kuno, Y. J. Chem.
Soc., Chem. Commun. 1994, 987.
(4) Mase, N.; Watanabe, Y.; Ueno, Y.; Toru, T. J. Chem. Soc.,
Perkin Trans. 1 1998, 1613. Radinov, R.; Mero, C. L.; Mc-
Phail, A. T.; Porter, N. A. Tetrahedron Lett. 1995, 36, 8183.
Zhu, Y.-H.; Vogel, P. Tetrahedron Lett. 1998, 39, 31.
(5) Hanessian, S.; Park, H.; Yang, R.-Y. Synlett 1997, 351.
(6) One enantiomeric form is shown arbitrarily.
(7) Nozaki, K.; Oshima, K.; Utimoto, K. Bull. Chem. Soc. Jpn.
1991, 64, 403.
In the absence of Lewis acid, n-Bu3SnH would approach
equally from the both faces of the radical center in an
open-chain transition state model to yield 4 and 5. In the
presence of the Lewis acids, the reaction of 3 with isopro-
pyl iodide or cyclohexyl iodide proceeds probably
through the transition state model B involving a seven-
membered chelate ring. n-Bu3SnH should attack from the
less hindered face of the model B to yield syn-adduct 4.
The transition model C yielding anti-adduct 5 is less pref-
erable due to the steric repulsion between R1 and CH2R3
(8) Urabe and Sato have reported that the addition of alkyl radical
to the g-substituted a-methylene-g-lactones gave syn-lactones
with high diastereoselectivity and the complexation of the ra-
dical intermediates with methylaluminium bis(2,6-di-t-butyl-
4-methylphenoxide) (MAD) reversed the diastereoselectivity.
Urabe, H.; Kobayashi, K.; Sato, F. J. Chem. Soc., Chem. Com-
mun. 1995, 1043.
(9) The signal of b-protons in the 1H NMR spectra of 4 were ob-
served consistently in lower field than those of 5.
Synlett 1999, No. 1, 53–54 ISSN 0936-5214 © Thieme Stuttgart · New York