.
Angewandte
Communications
Table 1: Selected optimizations.
Table 2: Scope of the cascade olefin isomerization/intramolecular
Diels–Alder reaction.[a]
Entry
Catalyst[a]
Solvent
T [8C]
t [h]
Yield [%][b]
1
2
3
4
5
6
7
8
–
o-Cl2C6H4
1,4-dioxane
THF
xylene
toluene
o-Cl2C6H4
o-Cl2C6H4
o-Cl2C6H4
o-Cl2C6H4
o-Cl2C6H4
o-Cl2C6H4
181
6
48
48
72
72
6
6
6
6
6
–
A1
A1
A1
A1
A1
A2
A1[d]
A1[e]
A1
B[g]
101
trace
3
100[c]
138
22
20
64
57
32
51
41
–
150[c]
181
181
181
9
10
11
181
181[f]
181
6
[a] NHCs were generated using equimolar tBuOK except as noted.
[b] Yields of isolated products following flash column chromatography
on silica gel. [c] Sealed tube. [d] A1 generated and isolated from salt by-
products. [e] A1 generated with LiHMDS (10 mol%). [f] Heated using
microwave irradiation. [g] B generated with KHMDS (10 mol%). Entry in
bold marks optimized reaction conditions. Bn=benzyl, Dip=2,6-
diisopropylphenyl, Mes=2,4,6-trimethylphenyl.
[a] Yields of isolated products following flash column chromatography
on silica gel. [b] Diastereoselectivity determined by 1H NMR analysis.
and in low yield. Similar results were obtained in THF heated
within a sealed tube (Table 1, entry 3), while xylene at reflux
or toluene in a sealed tube provided 8a in 22 and 20% yield
respectively (Table 1, entries 4 and 5). When the reaction was
heated at reflux in ortho-dichlorobenzene, bicyclooctane 8a
was isolated in an acceptable yield of 64% after 6 hours
(Table 1, entry 6). Changing the catalyst to the more hindered
IPr (A2; Table 1, entry 7) decreased the yield, as did the use
of IMes (A1) in the absence of salt by-products,[14] or
generated using LiHMDS, or with microwave heating
(Table 1, entries 8–10). Finally, using the electron-rich triazol-
ylidene B, designed for substrates in the ester oxidation
state,[6i] only decomposition of the starting materials was
observed (Table 1, entry 11).
The generality of the reaction was examined with a,b,g,d-
unsaturated enol esters 6a–q (Table 2). Using para-substi-
tuted acetophenone derivatives, the sensitivity of the reaction
to electronic effects was investigated. When either electron-
rich (6b and d) or electron-poor substrates (6c) were
subjected to the standard reaction conditions, bicyclo-
[2.2.2]octanes 8b–d were isolated in comparable yields.
Bicyclic aromatic moieties could be incorporated (i.e. 8e),
while the furan-containing product 8 f could only be accessed
in modest yield. Unfortunately, the use of alkyl-substituted
enol esters failed to provide the expected product (i.e. 8g).
Regarding the modest yield of furan 8 f, we postulated that
olefin isomerization should be favored by non-hydrogen
substituents at C5 (R2 = CH3), and hence the modest yield of
bicyclo[2.2.2]octanes should improve. This proved to be the
case with product 8h, which was prepared in 44% yield
(compare with 8 f, 34% yield). Similarly, naphthyl-containing
substrate 6i and acetophenone derivatives 6j–m provided the
expected bicyclo[2.2.2]octanes 8i–m with yields that were
increased by around 10%. Changing R2 to ethyl (i.e. 6n)
unfortunately hampered the reaction, with 8n isolated in
11% yield. Next, aldehyde-derived enols were examined
using phenylacetaldehyde-derived substrate 6o, which was
converted to 8o in a low yield. Unfortunately, the introduc-
tion of a methyl substituent at C5 with this type of enol
decreased the yield (8o versus 8p), presumably as a conse-
quence of steric crowding effects. Finally, the diastereoselec-
tivity of the reaction was examined with 6q, a substrate that
bears a stereogenic methyl group at C6 (R2 = CH3). While this
reaction was successful, bicyclo[2.2.2]octane 8q was formed
with little diastereoselectivity.
While nucleophilic catalysis of the olefin isomerization
provides a reasonable mechanistic explanation for this trans-
formation (Figure 1), an alternate mechanism might involve
the deprotonation of 6a by the NHC to generate enolate 9,
which is protonated to give an unconjugated diene en route to
8a (Scheme 1). To test this scenario, the reaction was
attempted using a range of non-nucleophilic Brønsted bases.
In all cases, unreacted starting materials were isolated
[Eq. (5)].[15] While this result is consistent with a process
mediated by a Lewis base, more compelling support was
obtained using DMAP as the catalyst. While DMAP is a good
Lewis base, it has low Brønsted basicity, and would be unable
to generate enolate 9.[16] In the event, the use of DMAP
provided the expected lactone 8a in 23% yield of isolated
product, presumably via hemiacetal pyridinium intermediate
10 [Eq. (6)].
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!