Chemistry Reference
In-Depth Information
TABLE 8B.55. Allylic Substitution of Cyclic Sulfoximines
NMe
O
R
S
Ph
Cu reagent/LiI
THF or Et
2
O,
Me
2
S,
-
78°C
+
R
n
n
n
α
γ
-Product
-Product
Entry
n
Cu Reagent
R
Additive
Yield (%)
γ
:
α
ee (%)
1
1
RCu
RO(CH
2
)
4
B F
3
• OEt
90
98:2
71
2
1
R
2
CuLi
RO(CH
2
)
4
—
90
2:98
—
3
1
RCu
BF
3
• OEt
89
99:1
72
n
- Bu
4
1
R
2
CuLi
—
87
2:98
—
n
- Bu
5
2
RCu
BF
3
• OEt
83
99:1
60
n
- Bu
6
3
RCu
n
- Bu
BF
3
• OEt
90
98:2
60
7
0
RCu
n
- Bu
BF
3
• OEt
40
98:2
27
reacted with a comparable selectivity at the α-position (entries 2 and 4). Therefore, it is
relatively easy to get either the one or the other product by simply changing the reaction
conditions. Best enantioselectivities were obtained with the fi ve - membered - ring sub-
strate (
n
= 1; entries 1, 3, and 5), while the four-membered ring especially showed
a signifi cant drop in the yield and the enantiomeric excess (entry 7) but not the
regioselectivity.
Recently, Breit and Breuninger described highly selective allylations controlled by a
syn
-directing diphenylphosphinobenzoate (DPPB) leaving group [345]. They investi-
gated the reaction of Grignard reagents in the presence of different copper salts, and
observed that the regioselectivity depended on the Grignard concentration and the addi-
tion time. The lower the concentration and the slower the addition, the higher the S
N
2 ′
selectivity. This clearly indicates that the reaction seems to proceed via an organocopper
reagent rather than a cuprate. This is in good agreement with the results obtained by
Bäckvall et al. [338] and Gais et al. [344]. If enantiomerically pure diphenylphosphino-
ferrocene carboxylates were used, ees up to 88% could be obtained at 0°C in Et
2
O
(Table 8B.56, entry 1). Switching to CH
2
Cl
2
increased the ee to 95% in a comparable
yield (entry 2). Methyl and isopropyl substituents could also be introduced with good
results (entries 3 and 4), while the enantiomeric excess dropped dramatically in the
reaction of PhMgBr (entry 5). Aryl substituents at an (
E
)-allylic substrate are well toler-
ated (entries 6), while the corresponding (
Z
) - substrates gave signifi cantly worse ees and
the opposite enantiomer was formed preferentially.
The stereochemical outcome of the reaction can be explained by coordination of the
copper reagent toward the phosphino group (Fig. 8B.35). One can assume a conforma-
tion with minimized A
1,3
-strain in which the σ*-orbital of the leaving group is parallel
to the p-orbitals of the double bond. Coordination of the copper reagent and internal
alkyl transfer generates the (
S
) - confi guration as obtained.
