of olefin can occur in tandem to construct the tetrahydropyran
unit (Scheme 1).
presence of DDQ as C-H bond activation reagent, and it
was discovered that the SnBr4/DDQ/CH2Cl2 system provided
the desired tetrahydropyran product 2a at ambient temper-
ature with an excellent yield and without detectable epimer-
ization at the C4-position. Addition of 4 Å molecular sieves
was believed to be of necessity since it could effectively
prevent the hydrolysis of oxonium ion.8 Meanwhile, we
found that InCl3 also induced the cyclization of homoallyl
ether to afford the corresponding 4-chlorotetrahydropyran,
but the reaction was much slower and a certain amount of
decomposed product (p-methoxybenzaldehyde) was obtained
even in the presence of 4 Å MS. Other Lewis acids (e.g.,
TiCl4, TiBr4, and AlBr3) resulted in only trace conversion
or decomposition of the starting material. Finally, no desired
product was observed with a lack of any Lewis acid.
Efforts toward investigating the scope of the electron-rich
benzyl substrate are outlined in Table 2. Under the optimized
conditions, a wide range of aromatic substrates underwent
reactions with high isolated yields and specified diastereo-
selectivities (Table 2, entries 1-11). When the unsubstituted
benzyl ether 1b was subjected to DDQ/SnBr4/4 Å MS, the
cyclized product 2b was formed in moderate yield, although
with a relatively slower rate (Table 2, entry 2). As revealed
in entries 3-7, a variety of methyl- or methoxyl-substituted
benzyl ethers (1c, 1d, 1e, 1f, and 1g) could be successfully
utilized in this transformation. These results clearly indicated
that the reaction activities were consistent with the electronic
effect of the benzyl ring: the lower oxidation potential and
the greater capacity for stabilizing the intermediate cation.
In the cases of the moderate electron-rich benzyl ether 1h
and the 1-naphthyl methyl ether 1i, the reactions provided
products 2h and 2i in satisfactory yields. Entry 10 demon-
strated a silyl ether 1j could be tolerated even when strong
Lewis acid was added to the reaction. Toward regiochemistry
of this transformation, 1k proved to be a suitable substrate
for the cyclization in that only alkoxybenzyl hydrogen prefer
to be activated and offer a similar skeleton precursor of (()-
centrolobine.9
Scheme 1
.
Strategy for Prins Cyclization through Benzylic/
Allylic C-H Activation
The discovery and development of methods to function-
alize selectively sp3 C-H bonds represent a long-standing
goal in organic synthesis.6 The single-electron oxidative
process has been a player in this field, and recently, several
reports have highlighted the utility of DDQ to promote
chemo- and regioselective C-H bond functionalization.7
Recently, DDQ-mediated oxidative annulation reaction, in
which the enol acetate serves as nucleophile, has been
reported by Floreancig’s group.7h,i Inspired by these studies,
we postulated that benzylic or allylic sp3 C-H bonds could
be directly activated via a relay-activated procedure and that
both single-electron oxidative reagent (DDQ) and Lewis acid
generate highly activity oxonium ion precursor, followed by
π-nucleophilic attack of an unactivated olefin, which leads
to the tetrahydropyran structure via a Prins cyclization
procedure.
Following this designed strategy, we chose the p-meth-
oxybenzyl (PMB)-protected homoallylic alcohol 1a as the
standard substrate in our efforts to find an effective relay-
activated condition for this type of Prins cyclization. As
shown in Table 1, several Lewis acids were screened in the
Next, we turned our attention to allylic ethers to expand
the scope of potential substrates (Table 3). As anticipated,
although allylic hydrogen is relatively inert toward single-
electron oxidative condition than benzylic hydrogen, suc-
cessful cyclization examples were observed with uniformly
exclusive diastereoselectivity. Trisubstituted alkene 1l and
phenylalkene 1p were found to result in an increase in both
Table 1. Optimization of the Prins Cyclization of Homoallylic
Ether 1a through C-H Activationa
(6) For recent reviews and commentary, see: (a) Godula, K.; Sames, D.
Science 2006, 312, 67. (b) Davies, H. M. L. Angew. Chem., Int. Ed. 2006,
45, 6422. (c) Murai, S. AdV. Synth. Catal. 2003, 345, 1033.
entry Lewis acid Nu- equiv
T (°C)
time (h) yieldb (%)
(7) (a) Xu, Y.-C.; Kohlman, D. T.; Liang, S. X.; Erikkson, C. Org. Lett.
1999, 1, 1599. (b) Ying, B.-P.; Trogden, B. G.; Kohlman, D. T.; Liang,
S. X.; Xu, Y.-C. Org. Lett. 2004, 6, 1523. (c) Zhang, Y.; Li, C.-J. Angew.
Chem., Int. Ed. 2006, 45, 1949. (d) Zhang, Y.; Li, C.-J. J. Am. Chem. Soc.
2006, 128, 4242. (e) Wang, L.; Seiders, J. R., II.; Floreancig, P. E. J. Am.
Chem. Soc. 2004, 126, 12596. (f) Seiders, J. R., II.; Wang, L.; Floreancig,
P. E. J. Am. Chem. Soc. 2003, 125, 2406. (g) Jung, H. H.; Seiders, J. R.,
II.; Floreancig, P. E. Angew. Chem., Int. Ed. 2007, 46, 8464. (h) Tu, W.;
Liu, L.; Floreancig, P. E. Angew. Chem., Int. Ed. 2008, 47, 4184. (i) Tu,
W.; Floreancig, P. E. Angew. Chem., Int. Ed. 2009, 48, 4567.
(8) When 4 Å MS was not used to remove water that might be present
in the reaction mixture, an amount of p-methoxylbenzaldehyde was observed
by TLC.
1
2
3
4
5c
6
7
SnBr4
SnBr4
TiBr4
TiCl4
InCl3
AlBr3
Br
Br
Br
Cl
Cl
Br
1.1 -20 to rt
1.1 rt
1.1 -20 to rt
1.1 -20 to rt
1.1 rt
1
70
0.5
0.5
0.5
6
89
dec
dec
55
1.1 rt
0.5
6
trace
0
no acid
rt
a All of the reactions were performed on a 1 mmol scale in DCM
(0.1 M). b Yield of isolated product. c Accompanied by 32% of p- me-
thoxybenzaldehyde byproduct. DCM ) dichloromethane.
(9) For a recent synthesis of centrolobine via the Prins cyclization, see:
Chan, K.-P.; Loh, T. -P. Org. Lett. 2005, 7, 4491, and ref 4d.
Org. Lett., Vol. 11, No. 15, 2009
3443
Yu, Binxun
