Communications
Table 2: Annulation of 2-alkynylbenzaldehydes with allylsilanes.
Table 3: Annulation of 2-alkynyl-1-ketonylbenzenes with allylsilanes.
Entry Ketones[a]
Silanes t [h] Products (Yields [%])[c]
R=Ph 12 5a (66)
Entry Enynal[a]
Silane
t [h]
Products
(Yields [%])[c]
1
2
3
4
5
R1 =R2 =Me (4a)
R1 =Me, R2 =nBu (4b) R=Ph 10 5b (83)
X, Y=H
R1 =Me, R2 =Ph (4c) R=Ph[b] 11 5c (63), 6 (22, d.r.=1:1)
1
2
R1 =Ph (1b)
R2 =Ph
R1 =4-MeOC6H4- R2 =Ph
(1c)
20
24
3b (75)
3c (78)
R1 =R2 =nBu (4d)
R=Ph 11 5d (82)
R=Me 5e (57), 7 (24)
4b
6
3
4
5
R1 =Cy (1d)
R1 =Me (1e)
R1 =H (1 f)
R2 =Ph
R2 =Ph
R2 =Ph
10
15
14
3d (88)
[a] [substrate]=0.26m, 5% PtCl2/11% AgOTf, silane (1.5 equiv). [b] 2.0
equiv of silane used. [c] Yields are for products isolated after column
chromatography.
3e (72)
3 f (12, 31[b])
(10[b])
12
6
7
R1 =nBu (1a)
R1 =nBu (1a)
R2 =4-
3g (68)
3h (60)
CF3C6H4
R2 =4-
12
allyltrimethylsilane gave desired product 5e and chrysene 7 in
57% and 24% yields, respectively.[12]
To clarify the role of the Brønsted acid (Scheme 1), we
added p-toluenesulfonic acid (p-TSA, 5 mol%) to a mixture
of aldehyde 1a, 2-(phenylallyl)trimethylsilane, and PtCl2/CO
MeOC6H4
8
9
10
R1 =nBu (1a)
R1 =nBu (1a)
R1 =Me (1e)
R1 =nBu, Y=H
X=CF3 (1g)
R2 =2-thienyl 13
3i (81)
3j (63[b])
3k (80)
R2 =Me
R2 =Me
2
1
11
12
R2 =Ph
R2 =Ph
10
3l (68)
X=OMe (1h)
14
3m (34, 78[b])
(12[b])
R1 =nBu, X=H
Y=CF3 (1i)
Y=OMe (1j)
13
14
R2 =Ph
R2 =Ph
15
20
3n (70)
3o (82)
[a] 808C, [substrate]=0.26m, [CO]=1 atm, catalyst (5 mol%), silane
(1.5 equiv). [b] These values were obtained with 5% PtCl2/11% AgOTf.
[c] Yields are for products isolated after column chromatography. Cy=
cyclohexyl.
with various allylsilanes and water; the reactions were run
with 5 mol% PtCl2/CO in wet toluene. In three reactions
(Table 2, entries 5, 9, and 12) we used PtCl2/2AgOTf (5/
11 mol%) to improve the product yields. The reactions of
substrates 1b–1e, in which the R1 substituent of the aldehyde
is varied (Table 2, entries 1–4), gave oxatricyclic products 3b–
3e in good yields (72–88%). However, the unsubstituted
analogue gave 3 f in 31% yield (Table 2, entry 5). The
annulations of aldehydes 1a and 1e with various 2-substituted
allylsilanes proceeded smoothly in wet toluene to give
expected products 3g–3k in 60–81% yields (Table 2,
entries 6–10). The reaction is also compatible with different
substituents (X and Y) on the phenyl rings of aldehydes 1g–
1j, and satisfactory yields (> 68%) were obtained for their
annulated products 3l–3o.
Although ketones are generally inactive in the Lewis acid
promoted annulation with allylsilanes, ketones 4a–4d were
compatible with this new method. The results in Table 3 show
that the PtCl2/2AgOTf mixture (5/11 mol%) efficiently
catalyzed the annulation of these ketones with 2-(phenyl-
allyl)trimethylsilane to give desired oxacyclic compounds 5a–
5d (63–83% yields) in wet toluene (1008C, 10–12 h). We
obtained 1,2-dihydronaphthalene species 6 as a side product
(22% yield, d.r. = 1:1) with 4c as the substrate (Table 3,
entry 3).[11] The annulation of compound 4b with 2-(methyl)-
Scheme 1. Control experiments to clarify the role of the Brønsted acid.
(5 mol%) in wet toluene (808C, 12 h), and obtained oxatri-
cyclic product 3a and 1,2-dihydronaphthalene 8 (d.r. = 1.7:1)
in 45% and 33% yields, respectively. The use of stronger
Brønsted acids, such as HCl or HOTf, led to complete
decomposition of starting aldehyde 1a. The formation of
species 8 has been rationalized by work in an early report by
Yamamoto and co-workers.[13,14b,c] As proposed, the Brønsted
acid catalyzes the hydrative decomposition of 2-(phenyl-
allyl)trimethylsilane to 1-methylstyrene.[13] The results
(Table 1, entry 1–6) show that of 1,2-dihydronaphthalene 8
is not isolated. In a separate experiment, we found that
treatment of alcohol 2b with PtCl2/CO in dry toluene (808C,
12 h) predominantly gave 1H-isochromene 9 in 83–86%
yields in the presence of either 4 molecular sieves or a
proton-scavenger such as MgO. Additional treatment of 1H-
isochromene 9 with p-TSA (1 mol%) in hot toluene (808C,
0.5 h) gave desired oxatricyclic product 3a in a 96% yield. On
the basis of these observations, we concluded that the
concentration of the free Brønsted acid was rather small in
the PtCl2/CO/H2O system, but it is necessary for the final
annulation step.
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 5063 –5066