COMMUNICATION
Table 2. Optimization of the reaction conditions.[a]
derivatives. In these studies, the reaction was catalyzed by
gold, palladium, or silver.[14] This situation prompted us to
develop a new catalyst system. Initial experiments were per-
formed using phenols 7a and 8a as model substrates, which
were treated with 5 mol% CuI and 10 mol% Cs2CO3 in dry
benzene in a sealed tube at 1108C for 26 h. Gratifyingly, the
desired spiroketal product 11a could be isolated in 42%
yield (Table 1, entry 1). Remarkably, the addition of PPh3
(5 mol%) improved the yield to 52%, which probably indi-
Entry
Solvent
T [8C]
t [h]
Product
Yield [%][b]
1
2
3
4
5
6
7
8
THF
110
110
110
110
110
100
90
100
100
100
26
26
26
26
26
26
26
26
22
31
11a
11a
11a
9a, 12a
9a, 12b
11a
11a
11a
29
31
64
dioxane
CHCl3
MeOH
EtOH
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
85, 92[c]
82, 99[c]
79
Table 1. Screening of catalysts.[a]
52
74[d]
77
9
10
11a
11a
70
[a] Reaction conditions: 7a (0.1 mmol), 8a (0.12 mmol), CuI (5 mol%),
PPh3 (5 mol%), solvent (0.5 mL) in a sealed tube, reacted at the temper-
ature indicated. [b] Yields were calculated after column chromatography.
[c] Based on 7a and 8a, respectively. [d] CuI (2.5 mol%), PPh3
(2.5 mol%) were used.
Entry Catalyst [mol%]
8a/7a [mol/mol] Yield [%][b]
1
CuI/Cs2CO3 (5/10)
1
1
1
1
1
1
1
1
1
1
1
1
42
52
36
48
38
41
52
45
38
26
51
0
2
3
4
5
CuI/PPh3/Cs2CO3 (5/5/10)
CuCl/PPh3/Cs2CO3 (5/5/10)
CuBr/PPh3/Cs2CO3 (5/5/10)
CuCl2/PPh3/Cs2CO3 (5/5/10)
6
CuN
yields (Table 2, entries 4 and 5). Scheme 3 accounts for the
formation of 12.[15] The collective results (Table 1, entry 12
and Table 2, entries 4 and 5) provide evidence for the mech-
anism of the domino process: compounds 8 are easily con-
7
8
9
10
11
12
13
CuI/PPh3/K3PO4 (5/5/10)
CuI/PPh3/LiOH·H2O (5/5/10)
CuI/PPh3/piperidine (5/5/10)
CuI/PPh3/TMG[c] (5/5/10)
CuI/PPh3 (5/5)
none
CuI/PPh3 (5/5)
1.2
59
[a] Reaction conditions: 7a (0.1 mmol), 8a, catalysts, dry benzene
(0.5 mL) in a sealed tube, 1108C for 26 h. [b] Yields were calculated after
column chromatography. [c] TMG=1,1,3,3-tetramethylguanidine.
Scheme 3. The formation of 12.
cates CuI activation by PPh3 (Table 1, entry 2). We then ex-
amined the efficiency of other copper salts under the same
experimental conditions. Compared to CuI, other copper
salts were less active and afforded lower yields (Table 1, en-
tries 3–6). The effect of added base was also examined
(Table 1, entries 7–11). Surprisingly, a good yield of spiroke-
tal can be maintained in the absence of added base (Table 1,
entry 11), which is quite different from the results of other
research groups.[14] On the other hand, after heating the re-
action mixture for 26 h without a catalyst, 8a was decom-
posed and 7a was recovered (entry 12), indicating that 7a
was neither converted to 9a (R1 =H) nor did it undergo
hetero-Diels–Alder cycloaddition with the o-QMs in the ab-
sence of the catalyst. Further research showed that the yield
of spiroketal can be promoted simply by increasing the
amount of 8a to 1.2 equivalents (entry 13).
An efficient method was identified for construction of ar-
omatic spiroketals in a one-pot manner. To improve yields,
further optimization studies were made. First, the influence
of solvents was examined. The results show that the same
reaction can also be performed in THF, 1,4-dioxane
(Table 2, entries 1 and 2), or CHCl3, which afforded the
highest yield (Table 2, entry 3). When alcohols were used as
the solvents, however, undesired 1,4-addition products 12
and the enol ether 9a (R1 =H) were obtained in excellent
verted to o-QMs 10 at high temperature and 9 are formed
very quickly from 7 in the presence of the catalyst. As enol
ethers 9 are good dienophiles and o-QMs 10 are very reac-
tive diene species, once 9 and 10 are formed, they immedi-
ately react with each other to produce the spiroketals 11. In
addition, o-QMs 10 react more readily with nucleophiles
(such as alcohols) to form 1,4-addition products than they
do with dienophiles. Next, the effects of temperature and re-
action time were examined. An obvious increase in the yield
was obtained when the reaction was performed at 1008C for
26 h (Table 2, entry 6). Further research demonstrated that
lower temperature (Table 2, entry 7) or reduced catalyst
loading (Table 2, entry 8) resulted in decreased yields, while
attempts to improve the yield by altering the reaction time
were unsuccessful (Table 2, entries 9 and 10).
As a high diastereomeric ratio of the product was detect-
ed by NMR spectroscopy (>20:1), our attention next
moved to assignment of the stereochemistry of the spiroke-
tal 11a. The relative configuration of 11a was finally estab-
lished to be cis by NOE analysis,[16] which was exactly oppo-
site to that obtained with Reissigꢁs protocol.[8 h] Because in-
termolecular hydrogen bonding in asymmetric Diels–Alder
cycloadditions has been demonstrated by both theoretical
904
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2012, 7, 903 – 906