9885
allylstannane (1 equiv.) on the reaction of acetophenone (1 equiv.) in the presence of 10 mol%
Zn(OTf) and 10 mol% pyridine in CH Cl at room temperature for 1 day: 94% of 3 was
2
2
2
obtained with 1, 59% of 3 with (CH ꢀCHCH ) SnBu , and 19% of 3 with (CH ꢀCHCH )SnBu ;
2
2 2
2
2
2
3
acetophenone was recovered in the last two cases. It may be said that the use of 1 is not
necessarily desirable since only one of the four allyl groups can be utilized and the remaining
three allyl groups have to be discarded. Fortunately, the use of 2,6-lutidine, instead of pyridine,
enhanced the yield of 3 even in the reaction of (CH ꢀCHCH )SnBu : 79% of 3 was obtained
2
2
3
along with 20% of the recovered acetophenone. This result should be compared with that of the
Sc(OTf) catalyzed allylation with (CH ꢀCHCH )SnBu : only 54% of 3 was obtained even using
3
2
2
3
the much stronger and expensive Sc(OTf) . Then we searched for the optimized reaction
3
conditions in the Zn(OTf) –2,6-lutidine promoted reaction of acetophenone (1 equiv.) with 1 (1
2
equiv.). The following three conditions (amount of Zn(OTf) , amount of 2,6-lutidine, reaction
2
time) gave 3 in quantitative yield; (10 mol%, 1 mol%, 1 day), (1 mol%, 10 mol%, 1 day), (1
mol%, 1 mol%, 1 day). If we used 10 mol% Zn(OTf) and 10 mol% 2,6-lutidine, 94% of 3 was
2
obtained even after 1 h.
The allylation of several different kinds of ketone with 1 was examined in the presence of 10
mol% Zn(OTf) –10 mol% 2,6-lutidine in CH Cl at room temperature for 1 day, and the results
2
2
2
are summarized in Table 2. Arylmethyl ketones bearing an EWG at the para-position of the
aromatic ring gave the corresponding allylation products in quantitative yield (entries 1–3).
Arylmethyl ketones bearing an EDG such as p-Me or p-MeO gave quantitative or high yields
of the allylation products (entries 4 and 5). However, a p-Me N substituent halted the allylation
2
completely (entry 6), the reason for which is not clear at present. In general, aliphatic ketones
and a,b-unsaturated ketones underwent the allylation without problem (entries 7–11). Not only
arylmethyl ketones but also phenylethyl ketone and benzophenone afforded the corresponding
allylation products in quantitative yield (entries 12 and 13).
It is clear that the Zn(OTf) –2,6-lutidine catalyst is very effective for the allylation of ketones
2
with 1. Two remaining questions at this point were (1) whether other metal triflates can be
6
applied or not, and more importantly (2) whether allyltributylstannane is applicable or not. The
reaction of acetophenone with allyltributylstannane in 10 mol% catalyst in CH Cl at room
2
2
temperature for 1 day was examined. (Catalyst, yield of 3, recovery of acetophenone) is shown
below. (Sc(OTf) , 54, 15), (Sc(OTf) –2,6-lutidine, 93, 7), (Hf(OTf) –2,6-lutidine, 52, 46),
3
3
4
(
Cp Ti(OTf) –2,6-lutidine, 23, 49). Accordingly, the Sc(OTf) –2,6-lutidine catalyst is more
2
2
3
effective than the Zn(OTf) –2,6-lutidine system in the reaction with allyltributylstannane.
2
A representative procedure is shown for the reaction of acetophenone with tetraallylstannane
using a Zn(OTf) –2,6-lutidine catalyst. To a mixture of Zn(OTf) (14.6 mg, 0.04 mmol) and
2
2
2
,6-lutidine (4.7 ml, 0.04 mmol) in dry CH Cl were added acetophenone (48 ml, 0.4 mmol) and
2
2
tetraallylstannane (110 ml, 0.44 mmol) under an Ar atmosphere. The reaction was carried out at
room temperature for 1 day and quenched with sat. aqueous NaHCO solution (4 ml). The
3
product was extracted with Et O and dried with Na SO . The solvent was evaporated, and the
2
2
4
product was purified by silica gel column chromatography, giving 3 (64.8 mg) in a quantitative
yield.
7
In conclusion, we have found that the Lewis acid–Lewis base catalysts, such as Zn(OTf) –2,6-
2
lutidine and Sc(OTf) –2,6-lutidine, are quite effective for converting ketones into the correspond-
3
ing homoallyl alcohols. Further investigation on diastereo- and enantio-selective allylation using
the new catalyst system is currently in progress in our laboratories.