Table 1 Screening of ligands and conditions with n-butyllithium
(Scheme 2)a
Table 2 Cu-catalyzed ring opening of 1a and 1b (Scheme 3)a
Anti : syn Yieldb 3 (%) ee 3 (%)
Entry Substrate RLi
Conversion
BF3ÁOEt2 T/1C Solvent (%)
eeb
Anti : syn (%)
1
2
3
4
5
6
7
8
9
1a
1b
1a
1b
1a
1b
1a
1b
1a
n-BuLi (2a)
n-BuLi (2a)
EtLi (2b)
>99 : 1 84
>99 : 1 82
>99 : 1 80
>99 : 1 71
>99 : 1 81
>99 : 1 82
>99 : 1 86
>99 : 1 96
97
95
98
97
97
93
97
93
43
Entry
L
1
L1
L2
L2
L2
L2 Yes
L2 Yes
L3 Yes
L4 Yes
L5 Yes
L2 Yes
—
—
—
—
À80 CH2Cl2
—
—
nd
nd
nd
nd
nd
2c
3d
4d
5c
6
À30 C2H4Cl2 50
3 : 97
16 : 84
3 : 97
—
>99 : 1 97
98 : 2
93 : 7
EtLi (2b)
0
0
C2H4Cl2 45
n-HexLi (2c)
n-HexLi (2c)
i-BuLi (2d)
i-BuLi (2d)
MTBE Fulle
À40 C2H4Cl2 Fullf
À80 CH2Cl2 Full
À80 CH2Cl2 76
À80 CH2Cl2 53
À80 CH2Cl2 75
À80 CH2Cl2 Full
7
8
À69
TMSCH2Li (2e) >99 : 1 65c
À77
a
Reagents and conditions: 1 (0.2 mmol), n-butyllithium (1.1 eq.),
CuBrÁMe2S (5 mol%), L (6 mol%), BF3ÁOEt2 (1.1 eq.), CH2Cl2, À80 1C,
9
>99 : 1 À76
10g
>99 : 1 97
b
c
full conversion after 16 h. Isolated yield. Conversion E 70%.
a
Reagents and conditions: 1a (0.2 mmol), n-butyllithium (1.5 eq.),
CuBrÁMe2S (5 mol%), L (6 mol%), solvent, temperature, 2 h.
b
c
ee of the anti product. 16 h.
d
e
5 h. Formation of side
%
products. Formation of 5 and 6 (1 : 1). n-Butyllithium (1.1 eq.),
16 h.
f
g
Scheme 3 Copper-catalyzed ring opening of oxabicyclic alkenes 1a
and 1b with organolithium reagents.
formation of the anti diastereomers was observed with excellent
enantioselectivity (97% ee) (Table 1, entry 6). The use of other
phosphoramidite ligands resulted in lower enantioselectivity
compared to the values obtained with ligand (R,R,R)-L2
although excellent anti : syn ratios were achieved in all cases
(Table 1, entries 7–9). The amount of organolithium reagent
could be lowered to 1.1 eq. without affecting the selectivities
(Table 1, entry 10). In this case the reaction mixture was
stirred overnight to ensure full conversion while undesired side
reactions were prevented.
Scheme 2 Copper-catalyzed asymmetric ring opening of oxabicyclic
alkenes using n-butyllithium.
the optimized conditions using Taniaphos L1 as a ligand for
the allylic alkylation of allylic bromides were employed
(Table 1, entry 1).16 Unfortunately, no ring opening of the
oxabicyclic alkene occurred at À80 1C.
We anticipated that higher temperatures were needed but in
order to raise the temperature the solvent had to be changed to
1,2-dichloroethane in order to prevent carbene formation.19,20
This led to 50% conversion and syn selectivity for the ring-
opening reaction giving a racemic product due to the uncatalyzed
reaction (Table 1, entry 2). We also switched to phosphoramidite
ligands21 because they have been shown to perform very well
both with organolithium reagents in the allylic alkylation16 as
well as in the copper-catalyzed ring opening of oxabicyclic
alkenes using organozinc reagents developed by our group.10
Increasing the temperature even further did not lead to higher
conversion or selectivity (Table 1, entry 3). Switching solvents
from 1,2-dichloroethane to methyl tert-butyl ether (MTBE) led
to mainly syn product together with a considerable amount of
side products (Table 1, entry 4). We then added a Lewis acid
(BF3ÁOEt2, 1.1 eq.) to activate the substrate and performed the
reaction at À40 1C overnight (Table 1, entry 5). In this case
full conversion was reached, but a 1 : 1 mixture of 1-naphthol 5,
originating from Lewis acid-catalyzed ring opening, and
n-butyl naphthalene 6, originating from addition of the butyl
group followed by elimination/aromatization, was observed
(Table 1, entry 6).
After optimization of the reaction conditions the use of
several organolithium reagents was examined for the ring
opening of oxabicyclic alkenes 1a and 1b (Table 2). In all
cases full conversion was obtained after stirring overnight
using only a slight excess of organolithium reagent (1.1 eq.)
and the anti product was formed exclusively (Scheme 3).
The ring opening of 1b using n-butyllithium gave a slightly
lower enantioselectivity than 1a (Table 2, entry 2). Both
ethyllithium as well as n-hexyllithium gave excellent enantio-
selectivities and isolated yields (Table 2, entries 3–6). Further-
more, the use of the more bulky reagent i-butyllithium also
resulted in an excellent yield and enantioselectivity of the
desired product 3ad (Table 2, entries 7 and 8). The use of a
trimethylsilyl-substituted organolithium reagent led to a decrease
in reactivity and enantioselectivity (Table 2, entry 9).
Organolithium reagents commercially available as solutions
in ethereal solvents were not effective under the optimized
reaction conditions and no conversion was observed. Further-
more, substrates bearing electron donating groups were not
compatible under the reaction conditions, leading to elimina-
tion of the alcohol group and aromatization of the naphthyl
group. This substrate incompatibility has also been observed
in the copper-catalyzed ring opening of oxabicyclic alkenes
with aluminium and Grignard reagents.12,14
Upon lowering the temperature to À80 1C in dichloro-
methane full conversion was obtained after 2 h, after which
the reaction had to be stopped to prevent side-reactions.22
Using phosphoramidite ligand (R,R,R)-L2 (Fig. 1) exclusive
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 1748–1750 1749