Asymmetric intramolecular radical vinylation using enantiopure sulfoxides as temporary chiral auxiliaries

Article abstract of DOI:10.1021/ja992818u

A very diastereoselective addition-elimination sequence affords cyclopentane derivatives in high enantiomeric purities. The enantiopure sulfoxide unit serves as a very efficient temporary chiral auxiliary in this tandem reaction. Interestingly, the presen

Full text of DOI:10.1021/ja992818u

J. Am. Chem. Soc. 1999, 121, 11395-11401
11395
Asymmetric Intramolecular Radical Vinylation Using Enantiopure
Sulfoxides as Temporary Chiral Auxiliaries
Be´ne´dicte Delouvrie´, Louis Fensterbank, Emmanuel Lacoˆte, and Max Malacria*
Contribution from the Laboratoire de Chimie Organique de Synthe`se associe´ au CNRS,
UniVersite´ Pierre et Marie Curie, 4, place Jussieu-Tour 44-54, Case 229, 75252 Paris Cedex 05, France
ReceiVed August 6, 1999
Abstract: A very diastereoselective addition-elimination sequence affords cyclopentane derivatives in high
enantiomeric purities. The enantiopure sulfoxide unit serves as a very efficient temporary chiral auxiliary in
this tandem reaction. Interestingly, the presence of the MAD Lewis acid totally reverses the stereochemical
outcome of this reaction. Several determining parameters of this sequence have been investigated: the substitution
of the vinyl sulfoxide moiety, the nature of the prochiral radical, the aromatic substituent of the sulfoxide
group, the tether, and the role played by different Lewis acids.
For the past decade, radical reactions have occupied a
prominent position in the realm of asymmetric synthesis.1 This
is due in part to the high compatibility of radical reactions with
a large number of interesting functionalities, notably present
on chiral auxiliaries, and to the possibility of optimizing the
stereoselectivities on using Lewis acids.2 For instance, the
Figure 1.
addition of a carbon-centered (alkyl or vinyl) radical to an alkene
Scheme 1
moiety bearing a chiral auxiliary has been well studied (Figure
1). Generally higher diastereoselectivities are obtained when the
addition occurs in the position R to the chiral auxiliary.3-6 Good
to excellent â-diastereoselectivities could also be observed,7-10
the use of Lewis acids being critical in the case of chiral
acrylates11,12 and N-enoyloxazolidinones.13
We have focused for some years on the use of chiral sulfur-
based auxiliaries, mainly sulfoxides, because of their easy
introduction, their low cost, and their versatile final function-
alization. Our initial approach, based on the Michael addition
of a vinyl radical onto vinyl sulfoxides, gave mixed results. Very
high diastereoselectivities were obtained for â-alkoxy vinyl
sulfoxides,9 while the pure carbon systems have so far led to
(1) Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of Radical
Reactions; VCH: Weinheim, 1996.
(2) Renaud, P.; Gerster, M. Angew. Chem., Int. Ed. 1998, 37, 2562-
2579. It is worthy of note that this chemistry has been nicely extended to
the use of chiral Lewis acids in a catalytic manner: Sibi, M. P.; Ji, J. J.
Org. Chem. 1997, 62, 3800-3801.
(3) Porter, N. A.; Lacher, B.; Chang, V. H.-T.; Magnin, D. R. J. Am.
Chem. Soc. 1989, 111, 8309-8310.
poor results, even in the case of N-sulfinimines.14 This failure
is illustrated by the reaction of (E)-vinyl sulfoxide 1 (Scheme
1). When submitted to low-temperature radical reaction condi-
tions (Bu3SnH, Et3B/O2), two diastereomers of the methylene
cyclopentane 2 were formed in an equimolar ratio. Running the
same reaction at a lower temperature (-78 °C) does not yield
any improvement. A rationale for this is the likely absence of
control of the s-cis or s-trans vinyl sulfoxide conformation
(Figure 2).15-19 Since sulfoxides are known to be compatible
with aluminum-based Lewis acids,20,21 we ran this reaction in
(4) Porter, N. A.; Scott, D. M.; Lacher, B.; Giese, B.; Zeitz, H. G.;
Lindner, H. J. J. Am. Chem. Soc. 1989, 111, 1679-1682.
(5) Porter, N. A.; Scott, D. M.; Rosenstein, I. J.; Giese, B.; Veit, A.;
Zeitz, H. C. J. Am. Chem. Soc. 1991, 113, 1791-1799.
(6) Scott, D. M.; McPhail, A. T.; Porter, N. A. Tetrahedron Lett. 1990,
31, 1679-1682.
(7) Stack, J. G.; Curran, D. P.; Geib, S. V.; Rebek, J. J.; Ballester, P. J.
Am. Chem. Soc. 1992, 114, 7007-7018.
(8) Ru¨ck, K.; Kunz, H. Synlett 1992, 343-344.
(9) Zahouily, M.; Journet, M.; Malacria, M. Synlett 1994, 366-368.
(10) Mase, N.; Watanabe, Y.; Ueno, Y.; Toru, T. J. Org. Chem. 1997,
62, 7794-7800.
(14) Lacoˆte, E.; Malacria, M. C. R. Acad. Sci. Fr., Ser. IIc 1998, 191-
194.
(11) Nishida, M.; Ueyama, E.; Hayashi, H.; Ohtake, Y.; Yamaura, Y.;
Yanaginuma, E.; Yonemitsu, O.; Nishida, A.; Kawahara, N. J. Am. Chem.
Soc. 1994, 116, 6455-6456.
(15) Pyne, S. G.; Griffith, R.; Edwards, M. Tetrahedron Lett. 1988, 29,
2089-2092.
(16) Hoffmann, R. W. Chem. ReV. 1989, 89, 1841-1860.
(12) Nishida, M.; Nishida, A.; Kawahara, N. J. Org. Chem. 1996, 61,
3574-3575.
(17) Kahn, S. D.; Hehre, W. J. J. Am. Chem. Soc. 1986, 108, 7399-
7400.
(13) Sibi, M. P.; Jasperse, C. P.; Ji, J. J. Am. Chem. Soc. 1995, 117,
10779-10780.
(18) Koizumi, T.; Arai, Y.; Takayama, H. Tetrahedron Lett. 1987, 28,
3689-3692.
10.1021/ja992818u CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/30/1999

11396 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
DelouVrie´ et al.
selenylated malonates 6-14 (Scheme 3). Malonate 6 was
obtained from the methoxyselenylation of the prenyldimethyl-
malonate.31 The other selenylated malonates 7-14 were pre-
pared through radical phenylselenyl group transfer according
to the method of Byers.32 The synthesis of the mesityl and the
trisyl (R)-vinylsulfoxides was based on Kagan’s procedure using
chiral sulfate derivatives.33 With these compounds in hand,
precursors 31-33 were prepared by similar chemistry, albeit
in lower yields (see Supporting Information). Finally, alkylation
of the lithium anion of (R)-cyclohexylidene-p-tolyl sulfoxide
3434 with the iodide 35 gave a high yield of precursor 36.
Racemic precursors were synthesized from the commercially
available phenylvinyl sulfoxide.
2. Radical Cyclizations. The radical cyclizations were
generally conducted at low temperature from -78 to 0 °C in
the presence of triethylborane and O2 as an initiating system.35
The racemic products were obtained by running the reaction in
classical conditions: AIBN in refluxing benzene. In both cases
an excess of initiator and tin hydride had to be used, since the
process is not a chain reaction. Indeed, the resulting sulfinyl
radical disproportionates rapidly and does not carry on the
radical chain properly.36
3. Determination of Enantiomeric Excesses. The cyclization
products 37-48 were reduced with LiAlH4 to provide diols 49-
58. The ee determination was performed on these compounds
using the chiral phosphorus derivatives NMR method of
Alexakis and Mangeney,37which can be applied to 1,3-diols
bearing a proximate stereogenic center. However, in our case,
we sometimes faced some depreciated values and some prob-
lems of reproducibility. This may be attributed to the fact that
the stereogenic center on diols 49-58 is too far away from the
phosphorus atom. Fortunately, the enantiomeric diols could also
be separated by chiral GC analysis, using a Chirasil-Dex CB
column.
4. The Vinyl Sulfoxides Substituents (R2 and R3 Substit-
uents). Our approach was initially tested with E precursor 19,
bearing an isopropyl group on the vinyl sulfoxide moiety
(Scheme 6). Under low-temperature radical cyclization condi-
tions, 19 underwent an exclusive 5-exo-trig radical cyclization
to afford cyclopentyl derivative 37 in 60% yield (Table 1, entry
1). One substituent on the vinyl sulfoxide at the â-position is
sufficient here to preclude the 6-endo-trig mode of cyclization.
Moreover, no cyclopentyl derivative incorporating the sulfoxide
moiety was observed, which confirmed the efficiency of the
â-elimination of the sulfoxide auxiliary. The promising stereo-
selectivity (73% ee) of this sequence was equally interesting.
A similar result in terms of yield and stereoselectivity was
obtained with cyclopropyl precursor 20 (entry 3). No sulfoxide
adduct showing the opening of the cyclopropyl ring was isolated
in this reaction, presumably suggesting that the â-elimination
of the sulfoxide moiety is even faster than the rearrangement
of the traditional radical clock, the R-cyclopropyl radical. This
Figure 2.
Scheme 2
*
*
the presence of triisobutyl aluminum and MAD (methylalumi-
num bis(2,6-di-tert-butyl-4-methylphenoxide)), anticipating that
we would freeze the reactive conformations and boost the
stereoselectivity. The best result (22% de), though still quite
mediocre, was obtained with the bulkier MAD. Obviously, no
synthetic application could ensue from these findings. One could
figure out that an acceptable level of diastereoselectivity could
be attained on a more sophisticated reacting system, for instance
by implementing a second Lewis base which would set the stage
for a chelation in the presence of a Lewis acid. This was indeed
achieved in an intermolecular version by Toru10 who solved
the diastereoselective intermolecular addition of alkyl radicals
on 2-arylsulfinyl-2-cyclopentenones. In the presence of alumi-
num- or titanium-based Lewis acids, and by adjusting the size
of the aromatic moiety on the sulfoxide (2,4,6-trimethylphenyl
or 2,4,6-triisopropylphenyl), very high diastereomeric excesses
were observed.
We preferred modifying our strategy according to the tandem
reaction depicted in Scheme 2. The sequence would consist of
a 5-exo-trig cyclization of a prochiral radical in an anti-Michael
orientation, followed by the previously reported elimination of
â-sulfinyl radicals22-28 to furnish alkylidene cyclopentane 5.29
Implying an a priori quite favorable R-selectivity, this radical
addition should be highly diastereoselective. To test this reaction,
we synthesized precursors 19-33 (Scheme 3). The scope and
the limitations are given in this article and we varied the
following parameters: the substituents of the double bond, the
nature of the prochiral radical, the role of Lewis acids, the
sulfoxide moiety, and the tether.
Results and Discussions
1. Synthesis of Precursors. To prepare precursors 19-30,
we coupled allyl bromides 15-18, easily prepared from the
chemistry developed by Maignan,30 with the sodium anion of
(19) Tietze, L. F.; Schuffenhauer, A.; Schreiner, P. R. J. Am. Chem. Soc.
1998, 120, 7952-7958.
(20) Renaud, P.; Bourquard, T.; Gerster, M.; Moufid, N. Angew. Chem.,
Int. Ed. 1994, 33, 1601-1603.
(21) Renaud, P.; Moufid, N.; Kuo, L. H.; Curran, D. P. J. Org. Chem.
1994, 59, 3547-3552.
(22) Miyamoto, N.; Fukuoka, D.; Utimoto, K.; Nozaki, H. Bull. Chem.
Soc. Jpn. 1974, 47, 503-504.
(30) Bonfand, E.; Gosselin, P.; Maignan, C. Tetrahedron: Asym. 1993,
4, 1667-1676.
(23) Boothe, T. E.; Greene, J. L. J.; Shevlin, P. B.; Wilcott, M. R., III;
Inners, R. R.; Cornelis, A. J. Am. Chem. Soc. 1978, 100, 3874-3879.
(24) Russell, G. A.; Tashtoush, H.; Ngoviwatchai, P. J. Am. Chem. Soc.
1984, 106, 4622-4623.
(31) Garratt, D. G.; Kabo, A. Can. J. Chem. 1980, 58, 1030-1041.
(32) Byers, J. H.; Lane, G. C. J. Org. Chem. 1993, 58, 3355-3360.
(33) Rebiere, F.; Samuel, O.; Ricard, L.; Kagan, H. B. J. Org. Chem.
1991, 56, 5991-5999.
(25) McCarthy, J. R.; Huber, E. W.; Le, T. B.; Laskovics, F. M.;
Matthews, D. P. Tetrahedron 1996, 52, 45-58.
(26) Aldabbagh, F.; Russell Bowman, W. Tetrahedron Lett. 1997, 38,
3793-3794.
(27) Clark, A. J.; Rooke, S.; Sparey, T. J. Tetrahedron Lett. 1996, 37,
909-912.
(28) Xiang, J.; Jiang, W.; Fuchs, P. L. Tetrahedron Lett. 1997, 38, 6635-
6638.
(34) Craig, D.; Daniels, K.; Roderick MacKenzie, A. Tetrahedron 1993,
49, 11263-11304.
(35) Nozaki, K.; Oshima, K.; Utimoto, K. J. Am. Chem. Soc. 1987, 100,
2579-2580.
(36) Chatgilialoglu, C. In The Chemistry of Sulphones and Sulphoxides;
Patai, S., Rappoport, Z., Stirling, C. J. M., Eds.; John Wiley & Sons Ltd.:
New York, 1988; pp 1081-1087.
(29) Lacoˆte, E.; Delouvrie´, B.; Fensterbank, L.; Malacria, M. Angew.
Chem., Int. Ed. 1998, 37, 2116-2218.
(37) Alexakis, A.; Frutos, J. C.; Mutti, S.; Mangeney, P. J. Org. Chem.
1994, 59, 3326-3334.

Asymmetric Intramolecular Radical Vinylation
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11397
Scheme 3
Scheme 4
Scheme 6
Scheme 5
Table 1. Results of the Tandem Reaction (yields and
stereoselectivity) with Variation of the R2 and the R3 Groups
product;
ee, %b
entry precursor
R2, R3
T, °C yield, %a (abs config)
1
1
3
4
5
6
7
8
9
19
19
20
20
21
21
22
22
22
22
R2 ) i-Pr, R3 ) H
R2 ) i-Pr, R3 ) H
R2 ) c-Pr, R3 ) H
R2 ) c-Pr, R3 ) H
-78 37; 60
37; 74
-78 38; 52c
38; 90
73 (S)
65 (S)
0
73 (S)
0
42d (S)
88 (S)
R2 ) R3 ) -(CH2)5- -40 39; 62
R2 ) R3 ) -(CH2)5-
R2 ) R3 ) Me
R2 ) R3 ) Me
R2 ) R3 ) Me
R2 ) R3 ) Me
0
39; 77
86 (S)
-78 40; 54
>98 (S)
>98 (S)
98 (S)
-40 40; 67
0
40; 93
10
10e 40; 89
92 (S)
a See Experimental Section. b Determined by chiral GC analysis.
c Starting material (38%) was recovered. d Determined by 31P NMR,
see ref 37. e AIBN, sun lamp initiation.
was run at 0 °C, the yield of cyclization adduct 38 dramatically
increased (90%, entry 4). The given ee (42%) does not really
mean a loss of stereoselectivity. More likely, it just illustrates
the depreciation of ee that we sometimes observed with the
phosphorus method (see above). To probe this, at -78 °C (entry
3), the phosphorus method gave a ee of 48% instead of 73%
by GC. It should also be pointed out that in both cases, as
determined by NOE measurements, the E configuration of the
alkene moiety of 37 and 38 was maintained.
result is not so surprising in view of the estimated rate of about
109 s-1 for the â-elimination of the arylsulfinyl radical,38
compared to the generally admitted rate of 5 × 107 s-1 for the
opening of related R-cyclopropyl radicals.39 When the reaction
(38) Wagner, P. J.; Sedon, J. H.; Lindstrom, M. J. J. Am. Chem. Soc.
1978, 100, 2579-2580.
(39) Newcomb, M. Tetrahedron 1993, 49, 1151-1176.

11398 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
DelouVrie´ et al.
Scheme 7
Scheme 8
We thus decided to examine the behavior of terminally
disubstituted vinyl sulfoxides, anticipating that the addition of
a substituent cis to the sulfoxide moiety should create some
additional allylic strain and thus should freeze the reactive
conformations and enhance the stereoselectivity. This proved
correct since terminally disubstituted vinyl sulfoxides 21 and
22 afforded much higher stereoselectivities, up to 98% ee (Table
1, entries 5 to 9). In both cases, no significant decrease of
stereoselectivity was observed when running the reaction at 0
°C and the chemical yield was greatly improved. With a sun-
lamp and AIBN initiation, the stereoselectivity remained very
high (Table 1, entry 10), thus discarding any role played by the
triethylborane in the stereoselectivity of the reaction. The high
enantiomeric purity of 40 led us to ozonolyze this compound,
to provide ketone 59, on which a CD measurement was
performed to determine the absolute configuration of the
stereogenic center (Scheme 7). An intense positive Cotton effect
was observed, and the configuration was deduced to be (S) on
40.40-42 Ozonolysis of 37-39 also afforded ketone 59 with
specific rotations of positive sign as in the case of 40, thus
suggesting a similar major (S) absolute configuration for 37-
39.
Scheme 9
The results obtained with the cyclopropyl precursor 20
intimate that the â-elimination of the sulfinyl radical is faster
than the opening of the R-cyclopropyl radical. We wanted to
gain a better insight into this â-elimination, namely: does it occur
with inversion or retention of chemistry of the alkene moiety?
The cyclizations of 19 and 20 brought the first elements of
information, since, as previously mentioned, they proceeded with
a complete retention of configuration. We may wonder whether
this â-elimination is not under kinetic control, implying that it
is the more favorable pathway at any temperature. This would
corroborate the findings of Boothe,23 who reported that the
radical â-elimination of diastereomeric 2-bromo-3-phenylsulfi-
nylbutanes to give but-2-ene “occurs before rotation about the
C-C bond and consequent equilibration of radical conforma-
tions can occur”. However, we could not synthesize the Z
isomers of 19 and 20, because of the E configuration of the
obtained allyl bromides,30 and this forbids any firm conclusion.
We thus decided to prepare precursor 60, isolated as a E:Z 1:1
mixture of diastereomers. The stereochemical outcome of the
radical cyclization of 60 proved to be temperature dependent
(Scheme 8). At low temperature, the reaction is highly diaste-
reoselective in favor of the E isomer 62, while raising the
temperature significantly increases the amount of the Z isomer
63. This finding suggested to us that in this case, the â-elimina-
tion of the sulfinyl radical is under thermodynamic control. The
formation of a tetrasubstituted double bond is probably less
favorable than the formation of a trisubstituted one and at low
temperature the relative rates (rotation vs â-elimination) must
change, mainly because of a less contributing -T∆S# term for
the â-elimination process. Consequently, the bond rotation
would be more favorable from 61 in a cold medium, resulting
in the formation of the more stable diastereomer 62. At higher
temperature, the â-elimination would augment, thus preventing
the thermodynamic equilibrium to be completely reached and
giving the 62-63 mixture.
5. The Prochiral Radical (R1 group). To have an insight
into the synthetic potential associated with this reaction, we
altered the nature of the prochiral radical, by varying the R1
substituent (Scheme 9). The obtained results call for several
comments. First, as previously observed, there is no significant
loss of stereoselectivity on changing the reaction temperature
from -78 to 0 °C or from -20 to 0 °C (entry 3 vs 4, 7 vs 8,
and 10 vs 11); however, a substantial gain in yield is generally
recorded. Second, on a general basis, the bulkier the R1
substituent, the higher the stereoselectivity. Notably, high ee
values were obtained with bulky secondary or tertiary substit-
uents (R1 ) cyclohexyl, t-Bu).
All these data can be rationalized by the following model
(Scheme 10). We presumed that the preferred conformation for
the sulfoxide when R2 and R3 are alkyl groups is the one that
eclipses the lone pair and the C-R3 bond.16,19 The radical then
cyclizes in the pseudochair 65, anti to the bulky aromatic group,
and placing the R1 group in pseudoequatorial position, to avoid
1,3-diaxial interactions with the sulfoxide moiety and an ester
group. This produces the major enantiomer 67, consistent with
our experimental results of Tables 1 and 2. However, for R3 )
H (precursors 19 and 20), the s-cis conformation becomes more
accessible17 and radical cyclization through transition state 64,
anti to the p-tolyl group, is now possible. This would explain
why we obtained cyclization products 37 and 38 with a lower
enantioselectivity.
(40) Djerassi, C.; Records, R.; Ouannes, C.; Jacques, J. Bull. Soc. Chim.
Fr. 1966, 2378-2381.
(41) Sundaraman, P.; Barth, G.; Djerassi, C. J. Am. Chem. Soc. 1981,
103, 5004-5007.
(42) Stien, D.; Samy, R.; Nouguier, D.; Crich, D.; Bertrand, M. P. J.
Org. Chem. 1997, 62, 275-286.

Asymmetric Intramolecular Radical Vinylation
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11399
Scheme 11
Scheme 10
Table 3. Effect of Lewis Acids
product;
yield, %a
ee, %b
(abs config)
entry
precursor
additive
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
22
22
20
23
24
24
25
26
29
27
28
28
28
30
30
Et2AlCl
MAD
MAD
MAD
MAD
ATPH
MAD
MAD
MAD
MAD
MAD
MAPH
ATPH
MAD
MAPH
40; 63
40; 52
38; 46
41; 35
42; 65
42; 23
43; 66
44; 84
47; 89
45; 70
46; 88
46; 75
46; 56
60 (S)
92 (R)
82 (R)
52 (R)
76 (R)
38 (R)
84 (R)
84 (R)
92 (R)
95 (S)
93 (R)d
34 (R)
86 (R)
Table 2. Effect of the Prochiral Radical (R1 Group)
product;
ee, %b
entry precursor
R1
T, °C yield, %a (abs config)
1
2
3
4
5
6
7
8
9
23
24
25
25
26
27
28
28
29
30
30
OEt
0
0
-20
41; 95
42; 100
43, 42
43; 77
44; 89
45; 88
46; 43
46; 72
47; 100
48; 95
48; 94
66c (S)
4d (R)
26e (R)
26e (R)
22 (R)
47 (S)
84 (R)
85 (R)
31f (S)
94 (R)
89 (R)
c
CH2OH
CH2OTBS
CH2OTBS
CH2OEt
n-Bu
cyclohexyl
cyclohexyl
CH(OMe)2
t-Bu
-
48; 55
57 (R)
0
0
0
a All reactions were run at 0 °C, see Experimental Section.
b Determined by chiral GC analysis. c Reduced starting material was
isolated (80%). d A CD measurement on the corresponding ketone 70
showed an intense negative Cotton effect, suggesting a (R) configuration
for 46.
-78
0
0
10
11
-20
Scheme 12
t-Bu
0
a See Experimental Section. b Determined by chiral GC analysis.
c Determined by 31P NMR. d Determined directly on 42. e Determined
on 42 after desilylation of 43. f Determined by transacetalisation with
(R,R)-(+)-diphenyl-1,2-ethanediol.
plexation of the sulfoxide moiety should modify the steric
environment, and possibly reverse the group priorities (Scheme
11). We initially focused on precursor 22 which offered the
best results. The addition of 4 equiv of diethylaluminum chloride
proved to be encouraging, since a notable decrease of selectivity
was obtained, suggesting (Table 3, entry 1) that the two
antagonist attacks 68 and 69 (Scheme 12) took place. Presum-
ably, by using a very bulky Lewis acid like MAD, we should
be able to restrict the cyclization through approach 68, and
indeed, we obtained the (R) antipode of 40 with a ee of 92%
(Table 3, entry 2). Furthermore, on a general basis, the addition
of MAD allowed a reversal of the stereochemical outcome of
the reaction, with higher enantioselectivities than the reactions
without MAD (up to 62% for precursor 26).
We also examined the behavior of other Lewis acids
developped by Yamamoto and Maruoka, such as MAPH and
ATPH44,45 (Figure 4). In most cases, we also witnessed an
inversion of the stereoselectivity (Table 3, entries 6, 12, and
13); however, the outcome was not as satisfactory as the one
obtained with MAD. The ATPH Lewis acid is an extremely
bulky one; moreover, it possesses three oxygenated ligands so
that the Lewis acidity is decreased. This presumably could result
in a less tight complexation of the oxygen of the sulfinyl moiety,
Figure 3.
Two examples of Table 2 are worthy of further discussion.
We studied the behavior of alcohol 24, anticipating that we could
lock the transition state through hydrogen bonding between the
hydroxy group and the sulfoxide moiety.43 In comparison to
the ether precursors 25 and 26, a notable drop of selectivity
was recorded (Table 2, entry 2 vs. 4 and 5), as if to a slight
extent some hydrogen bonding setting the CH2OH group in
pseudoaxial position (see Figure 3) indeed intervened and would
balance out the usual pseudoequatorial attack. This too weak
effect for a reversal of stereoselectivity could be rationalized
by a relatively loose ten-membered-ring chelate.
Another striking piece of data is the cyclization of 23. A fair
selectivity is obtained (66% ee) with this compound, especially
when compared to R1 groups of similar size (precursors 26 and
27). Complexation of the ethoxy group with surrounding
triethylborane molecules, resulting overall in a larger group,
probably implies a more pronounced pseudoequatorial orienta-
tion.
6. Effect of Lewis Acids. Reexamining Scheme 10 drove us
to the utilization of aluminum-based Lewis acids. The com-
(44) Maruoka, K.; Ooi, T.; Hokke, Y. Angew. Chem., Int. Ed. 1997, 36,
1181-1183.
(43) Mase, N.; Wake, S.; Watanabe, Y.; Toru, T. Tetrahedron Lett. 1998,
39, 5553-5556.
(45) Saito, S.; Yamamoto, H. J. Chem. Soc., Chem. Commun. 1997,
1585-1592.

11400 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
DelouVrie´ et al.
Scheme 13
Table 5. Reactivity and Stereoselectivity of a Precursor with No
gem-Dicarboxyl Groups
Figure 4.
yield of ee, yield of
Table 4. Effect of the Arysulfinyl Group
a
entry precursor
conditions
Et3B/O2, 0 °C
73, %
%
74, %
aromatic
substituent
product;
yield, %a
ee, %b
(abs config)
1
2
36
36
0
65
-
0
90
35
entry
precursor
AIBN, toluene,
1
2
3
31
32
33
trisyl
mesityl
trisyl
44; 52
46; 80
46; 69
49 (S)
68 (R)
66 (R)
∆, 4 × 10-4 mol‚h-1
AIBN, toluene, 10 °C,
hν, 13 × 10-4 mol‚h-1
Et3B/O2, 0 °C, MAD
3
36
15
0
30
80
90
4
36
a All reactions were run at 0 °C, see Experimental Section.
b Determined by chiral GC analysis.
a Determined by chiral GC analysis.
tion. Moreover, this orientation could be enhanced by some
dipolar repulsions between the ethoxy and the sulfinyl groups.
8. The Tether. The radical cyclization of 36 is a nice
illustration of the gem-dicarboalkoxy effect evidenced by Jung
in cycloadditions (Scheme 13).46 No cyclization adduct 73 can
be obtained at low temperature (Table 5, entry 1), the reduction
product 74 being isolated in high yield. Only in refluxing toluene
and with a slow addition of tin hydride could we isolate 73,
albeit in average yield. As expected, in these conditions, no
stereoselectivity was detected. At 10 °C, the radical cyclization
proceeded slugglishly (15% of 73) and with a quite mediocre
stereoselectivity (Table 5, entry 3), compared to the product
40. Therefore, the malonate moiety, presumably because of
possible 1,3-diaxial interactions, not only favors the cyclization
by restricting the number of reacting conformations, but also
appears to play a key role in the stereoselectivity of the
sequence. The presence of MAD in the reaction medium did
not bring any amelioration in terms of reactivity.
9. Conclusion. An efficient 5-exo-trig radical cyclization in
an anti-Michael orientation followed by the â-elimination of a
homochiral sulfinyl moiety opens a new access to enantiopure
alkylidenecyclopentanes. This sequence relies on the use of a
temporary chiral auxiliary. Very high enantioselectivities can
be reached, up to 98% ee. The â-elimination process was shown
to be faster than the opening of a cyclopropyl ring in the case
of a trisubstituted alkene and competitive with bond rotation in
the case of a tetrasubstituted alkene. Gratifyingly, the enantio-
selectivity can be reversed and improved by introducing the
bulky MAD Lewis acid in the reaction medium, giving a new
illustration of the versatile use of Lewis acids for stereoselective
radical reactions. Applications of this methodology in the
synthesis of enantiopure heterocyclic compounds are underway.
Figure 5.
and consequently in a drop of the selectivity. The MAPH is
less sterically demanding than ATPH and MAD and this would
be directly reflected by the weaker enantiomeric excesses we
obtained (entries 12 and 15). The t-Bu precursor 30 displayed
a rather peculiar behavior with Lewis acids certainly because
of steric effects. No cyclization in the presence of MAD could
be achieved, as if the approach syn to the p-tolyl group as of
the bulky radical 68 was completely impeded. No inversion of
the enantioselectivity was observed with MAPH, instead a
decrease of selectivity (Table 3, entry 15), which suggests a
poor complexation and recalls the result of precursor 22 in the
presence of Et2AlCl (entry 1) occurred.
7. The Arylsulfinyl Groups. The recent work of Toru
illustrates the important role played by the aromatic substituents
of the sulfoxide auxiliary. In our case, we could anticipate that
a bulky substituent would lock the reacting conformations and
notably force the pseudoequatorial orientation of the R1 group.
The results obtained with the OEt precursor 31 were encouraging
in this sense, since an increase of ee of 27% was observed (Table
4, entry 1 vs Table 2, entry 5). We next wanted to check if we
could render the total stereoselectivity of the reaction with R1
) cyclohexyl (Table 2, entries 7 and 8). However, with the
cyclohexyl precursors 32 and 33, the stereoselectivity confus-
ingly dropped by 20%. Reinspecting the models showed us that
conformation 71, which eclipses the lone pair, is no longer the
best one (Figure 5). Interactions involving the bulky aromatic
moieties (trisyl or mesityl) with the axial ester group and the
nearby methylene take place. In contrast, the s-trans conforma-
tion 7215,18 appears more plausible. The R1 group would still
be placed in pseudoequatorial position. However, in the case
of the bulky cyclohexyl group, the pseudoequatorial approach
is also hampered by the substituent of the aromatic ring at the
2- or 6-position. This additional constraint would result in a
larger amount of pseudoaxial attack and thus justify the erosion
of the ee (Table 4, entry 3). In comparison, the smaller CH2-
OEt group would still fit for a major pseudoequatorial orienta-
Acknowledgment. B.D. and E.L. acknowledge the Ministe`re
de l’Education Nationale et de l’Enseignement Supe´rieur et de
la Recherche for grants. The authors thank Professor Dennis P.
Curran (University of Pittsburgh), Professor Alexandre Alexakis
(Universite´ de Gene`ve), and Dr. Pierre Mangeney (Universite´
Pierre et Marie Curie) for helpful discussions, Mr. Buisine
(Universite´ Pierre et Marie Curie) for his high skills and patience
on chiral GC analysis, and Professor Jacques Bolard (Universite´
Pierre et Marie Curie) for CD measurements.
(46) Jung, M. E.; Vu, B. T. Tetrahedron Lett. 1996, 37, 451-454. See
also: Jung, M. E. Synlett 1999, 843-846.

Asymmetric Intramolecular Radical Vinylation
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11401
Supporting Information Available: All experimental pro-
cedures and spectral data for new compounds, as well as H
52 showing all experiments with precursor 22 (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
1
NMR spectra for 7, 10, 13, 19, 20, 21, 24, 25, 26, 27, 28, 30,
A, B, 31, 32, 33, 36, 41, 45, 53, 57, 60, 70, 73, and 74 and 13
C
NMR spectra for 23, 29, 39, and 59; chiral GC spectra of diols
JA992818U