6126
S. L. Schaefer et al. / Tetrahedron Letters 54 (2013) 6125–6128
OBz OBz
( )4
SO2Ph
OH
( )4
( )4
SmI2
n-BuLi, PhCHO
OBz
5
SmI2, DMPU
9
+
DMPU
then BzCl
( )4
SO2Ph
6
or
1a
14
5
( )4
SmI2, H2O
OBz
(65-75%)
expected
( )4
7
OBz
( )4
SO2Ph
D
8
SmI2, D2O
( )n
12 or 14
15 (n = 3, 4)
( )4
observed
9
LnSm
( )n
Scheme 3. Unexpected elimination of a bis-benzoyloxysulfone.
16
ent elimination/isomerization process with the goal of developing
a useful method for the preparation of synthetically challenging
non-conjugated olefinic systems.
Upon closer inspection it is perhaps not surprising that 8 was
not obtained from the reaction of 5. Following chemoselective
elimination generating 8, further samarium-mediated elimination
of the benzoyl group would now give a resonance stabilized radical
intermediate 10 (Scheme 4). Selectivity for the formation of 9 from
this intermediate as opposed to the corresponding conjugated iso-
mer 11 was however not immediately clear.
Scheme 6. Samarium/water elimination and deuterium labeling.
15, suggestive that both substrates converge to the same organosa-
marium intermediate of type 16.10
Regarding the selectivity for formation of the non-conjugated
olefin product, previous work from Yoshida et al. using a palladium
tetrakis triphenylphosphine (Pd(PPh3)4)/SmI2 system suggested
that this is primarily due to steric reasons.11 This group reported
that the reduction of allylic phosphonate 17 with Pd(Ph3)4/SmI2
and H2O gave primarily the c-isomer via internal delivery of a pro-
To test the intermediacy of 8 in the production of 9 from 5,
analogue 12 was prepared by the addition of n-butyllithium
(n-BuLi) to cinnamaldehyde followed by trapping with benzoyl-
chloride (Scheme 5). Treatment of 12 to elimination conditions
using SmI2 and DMPU gave 137 along with minor amounts of the
conjugated isomer in essentially the same ratio as 9 to 11 obtained
from compound 5 (Scheme 4). This is consistent with a reaction
pathway for 5 proceeding by first chemoselective elimination to
generate 8 and subsequent elimination/isomerization affording 9.
Placement of the benzoyl group at the benzylic position as in 14
or the reaction using SmI2 and H2O8 gave comparable results in
terms of both yield and ratio of non- to conjugated products
(Scheme 6).9 When the reaction was performed with 12 or 14 in
D2O the major product for each was the monodeuterated adduct
ton from a sterically preferred organosamarium–H2O complex 18
(Scheme 7). Switching to tert-butanol (tBuOH) as the proton source
leads to the
a-product as the major isomer proceeding through a
presumably open protonation mechanism.
In analogy, the results in Schemes 4–6 can be explained as a ste-
ric preference for intermediate 16 to exist as the
D
1,2-structure
(Scheme 8). Internal protonation with H2O (upon quench when
using DMPU under anhydrous conditions) would then give com-
pounds 9 or 13.
The use of tBuOH in place of H2O with compound 12 resulted in
an erosion of this selectivity consistent with the observations of
Yoshida et al. (Scheme 9). The ratio of products could also be af-
fected by manipulating the substrate structure. For instance reduc-
tion of the tert-butyl substrate 19 using SmI2 with DMPU, H2O or
tBuOH gave little product selectivity perhaps due to minimal pref-
erence for either the
D D
1,2- or 2,3-organosamarium intermediate.
OBz
For each, the product was obtained as a mixture with what have
been tentatively assigned as radical dimerization adducts conceiv-
ably formed as the rate of protonation is slowed for this sterically
encumbered substrate.12
The present method thus appears to be in line with the results
described by Yoshida et al. In order to directly compare the two
reactions, compounds 2013 and 2114 were prepared and subjected
to samarium elimination conditions15 (Scheme 10). The reactions
with SmI2/H2O gave a complex mixture of products. However with
SmI2/tBuOH both 20 and 21 were converted primarily to allylben-
zene with identical yield and selectivity, again consistent that each
SmI2, DMPU
SmI2
( )4
5
chemoselective
elimination
8
( )4
9 (major)
6
( )4
10
+
1a
( )4
resonance stabilized
11
(minor)
aRatios determined by 1H NMR.
Scheme 4. Proposed formation of 9 from 8.
Pd(PPh3)4
X
+
OBz
α
γ
SmI2
n-BuLi then
CHO
SmI2, ROH
( )3
17
H2O
tBuOH
(γ/α = 32:68)
BzCl
(97%)
DMPU
(73%)
12
X = OP(O)OEt2
(γ/α = 93:7)
H
O
via
H+
SmLn
SmLn
H
( )3
(minor)
( )3
+
13 (major)
Ph
Ph
18
preferred
Ph
SmLn
5 : 1a
Scheme 5. Samarium allylic benzoate elimination/isomerization.
Scheme 7. Yoshida et al. palladium-catalyzed allylic samarium reduction.