Organic Letters
Letter
subsequent stereospecific oxetane fragmentation, which
accounts for the high selectivity in products. Insights gained
from this work are expected to guide future reaction
development and catalyst design to expand and improve the
synthetic utility of available protocols.
The initial studies toward the development of an
intermolecular carbonyl−olefin metathesis reaction focused
on benzaldehyde 12 and 2-methyl-2-butene 13 as substrates
(Table 1). Early efforts identified benzaldehyde 12 and 2-
pairs,17 resulting upon halide abstraction from neutral metal
salts (MXn) with silver salts (AgX),18 can function as Lewis
acidic superelectrophiles for catalytic carbonyl−olefin meta-
thesis reactions. The addition of catalytic amounts of silver
salts in combination with FeCl3 as the Lewis acid resulted in
increased yields of the desired carbonyl−olefin metathesis
product 14 (entries 9−13, Table 1). Specifically, AgBF4 was
identified as the superior silver salt together with FeCl3,
providing (E)-olefin in 51% yield (entry 13, Table 1).
Stoichiometric amounts of AgBF4 under otherwise identical
conditions resulted in diminished yields of 14 in 28% (entry
14, Table 1). Decreasing loadings of AgBF4 of 10 and 20 mol
% also proved inferior and formed 14 in 27 and 20%,
respectively (entries 15 and 16, Table 1). GaCl3 (which had
similar yields to FeCl3) was evaluated in combination with 30
mol % of AgBF4; however, the transformation provided a
diminished yield of 35% (entry 17, Table 1). Importantly, (E)-
olefin 14 was observed as the exclusive carbonyl−olefin
metathesis product for all Lewis acids and reaction conditions
evaluated (Table 1).
Subsequent efforts focused on obtaining experimental
support for heterobimetallic ion pairs as the active catalytic
species under the optimal reaction conditions. Several distinct
Lewis acidic species could be operative as the active catalyst:
FeCl3 (A), AgBF4(B), heterobimetallic ion pairs
[FeCl2]+[BF4]− (C), and [Fe]3+3[BF4]− (D), resulting from
chloride abstraction, or FeCl2F (E), FeF3 (F), and BF3 (G),
formed via fluoride transfer or the decomposition of C and D
(Table 2). As previously demonstrated, substoichiometric
amounts of FeCl3 formed metathesis product 14 in 19% yield,
whereas the sole use of AgBF4 failed to promote the desired
carbonyl−olefin metathesis reaction (entries 1 and 2, Table 2).
Equimolar loadings of 10 mol % FeCl3 and AgBF4 were also
able to catalyze the reaction, although in a low yield of 20%
Table 1. Reaction Optimization for Intermolecular
a
Carbonyl−Olefin Metathesis
entry
Lewis acid
additive
mol %
solvent
yield 14 (%)
b
1
BF3·Et2O
FeCl3
FeCl3
FeCl3
GeCl3
AlCl3
Fe(OTf)3
Sc(OTf)3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
30
30
30
30
30
30
30
30
30
30
30
DCM
DCM
DCE
28
19
16
2
17
0
30
26
0
31
24
36
51
28
27
20
35
27
19
b
2
b
3
b
4
toluene
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
b
5
b
6
b
7
b
8
c
9
AgOTS
AgAsF6
AgSbF6
AgPF6
AgBF4
AgBF4
AgBF4
AgBF4
AgBF4
AgBF4
AgBF4
c
10
11
12
13
14
15
16
17
c
c
d
c
c
c
c
e
a
18
19
Table 2. Determination of Active Catalytic Species
f
a
Conditions: All reactions were performed using 5.0 equiv of the
substrate 12 and 1 equiv of 13 in DCM (0.3 M) at 25 °C for 24 h.
b
c
Yields are reported based on NMR analysis. Yields are reported
d
based on GC analysis. Yields are reported based on isolated yield.
e
f
2.0 equiv of 12 was used. 1.0 equiv of 12 was used.
methyl-2-butene 13 in a 5:1 ratio as optimal for the
́
details.) Franzen and coworkers were able to achieve 60% yield
of 14 with 20 mol % of TrBF4 as the cationic catalyst and
otherwise identical conditions as entry 1, Table 1.15 Catalytic
amounts of BF3·OEt2 and FeCl3 promoted the desired
intermolecular carbonyl−olefin metathesis reaction in 28 and
19% yield, respectively (entries 1 and 2, Table 1). In
comparison, decreased yields of 14 were observed with other
solvents, including dichloroethane and toluene, under
otherwise identical reaction conditions (entries 3 and 4,
Table 1). GaCl3 was analogous to FeCl3 and resulted in the
desired metathesis product in 17% yield (entry 5, Table 1),
whereas stronger Lewis acids, such as AlCl3, proved ineffective
in promoting the desired carbonyl−olefin metathesis reaction
(entry 6, Table 1). Promising results were also obtained with
catalytic amounts of metal triflates, Fe(OTf)3 and Sc(OTf)3,
resulting in the formation of (E)-olefin 14 in increased yields
of 30 and 26%, respectively (entries 7 and 8, Table 1).
Recently, we were able to show that heterobimetallic ion
entry
species
Lewis acid
additive
X mol %
yield 14 (%)
b
1
A
B
C
D
E
F
FeCl3
−
FeCl3
FeCl3
FeCl3
FeCl3
FeF3
−
AgBF4
AgBF4
AgBF4
AgF
AgF
−
−
30
10
30
10
30
−
19
0
20
51
4
9
0
28
c
2
de
,
3
4
5
6
7
8
c e
,
c
c
b
b
F
G
BF3·Et2O
−
−
a
Conditions: All reactions were performed using 5.0 equiv of the
substrate 12 and 1.0 equiv of 13 in DCM (0.3 M) at 25 °C for 3 h.
Yields were determined via NMR with PhMe3Si as an internal
standard. Yields were based on isolated yield. Yields were reported
based on GC analysis. Formation of AgCl was observed.
b
c
d
e
B
Org. Lett. XXXX, XXX, XXX−XXX