J. P. M. Nunes et al. / Tetrahedron Letters 50 (2009) 3706–3708
3707
100
90
80
70
60
50
40
30
20
10
0
O
O
O
DMF
a
OH
OAc
+
O
DMSO
Ot-Bu
Ot-Bu
Ot-Bu
1
(+)-2
(-)-4
HMPA
( )-2%
Scheme 2. Direct synthesis of enantiomerically pure cyclopentenone (À)À4 from
pyranone 1: (a) (0.59 mmol) t-BuOH (1 mL), DABCO (0.3 equiv), Lipase AK
1
Butyl acetate
n-heptane
‘Amano’ 20 (0.5 mg/mg of 1), vinyl acetate (5 equiv) 50 °C, 10 d, (+)À2 and 3 (35%,
91:9 ratio, ee (+)À2 11%); (À)À4: 55%, ee 80%.
As can be seen from Table 1, the use of a protic solvent to speed
up the reaction such as tert-butanol was highly beneficial and al-
lowed us to significantly reduce the reaction time (entries 4 and
5). It is also noteworthy that prolonged reaction times led to an
undesirable ratio of 2 to 3 with the higher temperature (entry 4).
Further studies on the effect of the base also revealed some useful
observations. The use of the base DABCO was found to be far more
effective than that of triethylamine (80% vs. 30% GC yield of ( )-2
after 20 min) and we were able to reduce the quantity of base from
five equivalents in our original procedure, to sub-stoichiometric
quantities (10%-15%) (entries 6 and 7). Thus we have developed a
new organocatalysed protocol for the isomerisation of ( )-2 under
milder conditions. It is noteworthy that, under these conditions,
the quantity of 3 produced is very modest (3%).10
tert-butanol
0
500
1000
1500
2000
time/min
Figure 1. Selected solvent effect on the rearrangement of pyranone 1 to ( )À2
(70 °C, 0.29 M, Et3N (5 equiv), determined by GLC using n-decane as internal
standard).
( )-2 followed by enzymatic kinetic resolution. In the context of
the present work it is noteworthy that Ramström and co-workers9
recently described a very elegant combined nitro aldol and enzy-
matic kinetic resolution protocol.
In order to develop an effective one-pot protocol it was essen-
tial to improve the conditions for the isomerisation of the pyra-
none to the cyclopentenone. The existing conditions required the
use of triethylamine (5 equiv) and a prolonged reaction time. Our
first objective was to improve these conditions with a view to ren-
dering them compatible with a resolution step. In order to opti-
mize the conditions for the rearrangement of pyranone 1, a range
of solvents were screened at 70 °C in the presence of triethylamine
(5 equiv) and each reaction was followed by gas chromatography
(GC). A considerable solvent effect can be seen for some represen-
tative examples presented in Figure 1.
From Figure 1 it can be seen that the rearrangement proceeds
slower for apolar or weakly polar solvents. For more polar aprotic
and protic solvents the reaction is faster which is consistent with a
proposed mechanism involving an intermediate enol derived from
electrocyclic ring opening of 1, since such an enol would be stabi-
lized in polar protic solvents.4 It is noteworthy that although the
use of these solvents is beneficial, the quantity of ( )-2 can eventu-
ally decrease as a function of reaction time. This can be attributed
to decomposition or conversion to the undesired enone 3. A more
detailed study of the effect of solvent, base and temperature was
carried out and led us to identify conditions that minimised the
formation of 3 and also optimised the formation of ( )-2. Selected
examples of these studies are presented in Table 1.
These milder conditions for the rearrangement step prompted
us to explore the possibility of effecting a combined one-pot se-
quence of reactions involving organocatalysed rearrangement of
pyranone 1 to ( )-2 followed by in situ enzymatic resolution. After
screening several enzymes and reaction conditions it was possible
to isolate (À)-4 in 55% yield, ee 80% and a mixture of (+)-2 and 3 in
35% yield (91:9 ratio, ee (+)-2 11%) (Scheme 2).11 Interestingly, the
remaining alcohol (+)-2 was isolated in low ee which can only be
explained by racemisation of the remaining enantiomerically en-
riched (+)-2 under the reaction conditions. This may offer the
opportunity to further develop this methodology to incorporate a
dynamic kinetic resolution (DKR) step with the possibility of
improving the conversion further.
In conclusion, we have shown that there is a remarkable solvent
effect on the rearrangement of pyranone to cyclopentenone and
this has led to a new organocatalysed version of this transforma-
tion. The improvement in the conditions resulting from this opti-
misation study has provided us with the opportunity of
developing a combined one-pot sequence involving organocataly-
sed rearrangement followed by kinetic resolution leading to a con-
venient asymmetric synthesis of (À)-4 directly from readily
available pyranone 1. Further work will focus on extending this
methodology to other classes of pyranones and towards the possi-
bility of developing a dynamic kinetic resolution approach.
Table 1
Studies on the rearrangement of pyranone 1 to ( )À2 and 3
Entry
Solvent
Temp (°C)
Base (equiv)
Reaction time (h)
Yield ( )-2, 3a
Ratio ( )À2:3b
1c
2c
3c
4c
5c
6d
7e
DMF
80
80
80
80
80
50
50
Et3N (5)
Et3N (5)
Et3N (5)
Et3N (5)
Et3N (5)
DABCO (0.1)
DABCO (0.15)
24
24
24
48
5
74
22
22
59
83
77
85
91:9
DMSO
MeOH
t-BuOH
t-BuOH
t-BuOH
t-BuOH
49:51
0:100
21:79
78:22
96:4
48
24
97:3
a
Isolated yield by flash chromatography.
b
c
Ratio of ( )À2:3 determined by 1H NMR.
Reactions were performed using 1 (50 mg/mL).
Reaction was performed using 1 (100 mg/mL).
Reaction was performed using 1 (200 mg/mL).
d
e