Communications
Table 1: Scope of the tandem allylic oxidation/oxa-Michael reaction.
The stereochemical outcome of this tandem allylic
oxidation/oxa-Michael reaction can be rationalized on the
basis that the unfavorable 1,3-diaxial interaction of the C6
a,b-unsaturated carbonyl group and the C4 dithiane group in
conformation 5b is larger than that of the C6 hydrogen atom
and the dithiane group in conformation 5a; thus, 6a is formed
preferentially. It is known that a 2,6-cis-THP is thermody-
namically more favorable than a 2,6-trans-THP in equilibri-
um.[11] To identify the origin of the stereoselective formation
of 6a in the tandem reaction, the trans isomer 6b was
subjected to the reaction conditions for the cyclization
(MnO2, CH2Cl2, 258C, 24 h). As no formation of 6a was
observed in this case, we could conclude that the formation of
6a was kinetically controlled (see the Supporting Information
for details).
Encouraged by the preliminary results, we set out to
explore the scope of the tandem allylic oxidation/oxa-Michael
reaction. First, we examined a range of oxidation conditions
for the tandem reaction. The Parikh–Doering oxidation of 4
(SO3·pyridine, Et3N/dimethyl sulfoxide (DMSO)/CH2Cl2
(1:1:4), 258C, 24 h) provided the desired 2,6-cis-THP 6a
with excellent stereoselectivity (d.r. > 20:1, 93%). Other
oxidation conditions (pyridinium chlorochromate, tetrapro-
pylammonium perruthenate, or Dess–Martin periodinane)
also afforded 6a (d.r. > 20:1), but in low to moderate yields
(21–51%) owing to competitive oxidation of the secondary
hydroxy group, decomposition of the aldehyde intermediate,
or possible oxidation of the 1,3-dithiane group. Swern
oxidation of 4 did not provide 6a, but significant decom-
position of the aldehyde intermediate was observed (see the
Supporting Information for details).
Entry
Substrate
Conditions[b]
Product
(yield [%])[a]
(yield [%], d.r.[c])
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
12a (88, >20:1)
12a (87, >20:1)
13a (86, >20:1)
12b (75, >20:1)
12b (87, >20:1)
13b (81, >20:1)
12c (81, >20:1)
12c (75, >20:1)
13c (76, >20:1)
12d (96, >20:1)
12d (94, >20:1)
13d (76, >20:1)
12e (86, NA[d])
12e (87, NA[d])
13e (75, NA[d])
1
2
3
4
5
11a (74)
11b (72)
11c (76)
11d (83)
11e (67)
[a] Yield of isolated 11 from the coupling reaction (tBuLi, HMPA/THF
(1:10), ꢁ788C, 5 min, then 10, ꢁ788C, 0.5–2 h). [b] A) MnO2, CH2Cl2,
258C, 24 h; B) SO3·pyridine, Et3N/DMSO/CH2Cl2 (1:1:4), 258C, 24 h;
C) I MnO2, CH2Cl2, 258C, 1.5 h; II dimethyltriazolium iodide, MnO2,
DBU, MeOH, 4 ꢀ MS, 258C, 23 h. [c] The diastereomeric ratio (2,6-cis-
THP/2,6-trans-THP) was determined by integration of the 1H NMR
spectrum of the crude product. [d] Not applicable.
When 4 was subjected to the conditions for the tandem
reaction in the presence of dimethyltriazolium iodide,[12] the
THP methyl ester 7 was obtained in a single step from 4 in a
one-pot allylic oxidation/oxa-Michael/oxidation reaction
(Scheme 2). This MnO2 oxidation catalyzed by an N-hetero-
conditions examined proceeded smoothly to provide the
corresponding 2,6-cis-THP aldehydes 12a–d or esters 13a–d
with excellent diastereoselectivity (d.r. > 20:1; Table 1,
entries 1–4). Even the sterically hindered tertiary alcohol
11e was converted into the THP aldehyde 12e and ester 13e
in good yield (75–87%) under the mild reaction conditions
(Table 1, entry 5).
We hypothesized that the 1,3-dithiane group would be
critical to overcoming the low reactivity of oxygen nucleo-
philes and the reversibility of the reaction by promoting an
ideal conformation for cyclization through the gem-disub-
stituent effect.[14] To prove this hypothesis, we prepared
substrates 14 and 16 with no or a diminished gem-disubstitu-
ent effect and subjected them to the reaction conditions for
the tandem reaction (Table 2). Although the MnO2 oxidation
of 14 provided the desired 2,6-cis-THP 15a (d.r. 7:1; Table 2,
entry 1), the reaction also produced ketone 15b (15a/15b 3:1)
as a result of a slow oxa-Michael addition step owing to the
absence of the gem-disubstituent effect and the competing
oxidation of the benzylic hydroxy group. Parikh–Doering
oxidation of 14 failed to provide 15a; instead, the exclusive
formation of 15b was observed (Table 2, entry 2). When gem-
dimethyl substitution was introduced in the substrate 14, the
tandem reaction was accelerated (Table 2, entries 3 and 4).
Thus, the reaction of 16 with MnO2 or under Parikh–Doering
conditions provided 17a with good stereoselectivity
Scheme 2. One-pot allylic oxidation/oxa-Michael/oxidation reaction to
give the THP ester 7: I) MnO2, CH2Cl2, 258C, 1.5 h; II) dimethyltriazo-
lium iodide, MnO2, DBU, MeOH, 4 ꢀ MS, 258C, 23 h, 61%.
DBU=1,8-diazabicyclo[5.4.0]undec-7-ene.
cyclic carbene proved to be a reliable method for the
oxidation of aldehydes containing sensitive electron-rich
sulfur atoms to the corresponding carboxylic acids or esters.[13]
To investigate the scope and stereochemical outcome of
the tandem reaction with respect to substituents at the C2
position, we prepared allylic alcohols 11a–e by coupling 3
with the commercially or readily available epoxides 10a–e
and subjected them to MnO2 or Parikh–Doering oxidation, or
MnO2 oxidation in the presence of dimethyltriazolium iodide
(Table 1). The tandem reaction of 11a–d under the reaction
7578
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7577 –7581