.
Angewandte
Communications
elevated reaction temperature and prolonged reaction time
caused preferential formation of tetralone 3a over diol 2a
(entries 6 and 7). At 1 mol% catalyst loading, diol 2a was
formed cleanly in good yield (89%) and excellent enantio-
selectivity (e.r. > 99:1) (entry 8).[5] Further decreases in cata-
lyst loading led to incomplete conversion (entry 9). Other
ligands that gave diol 2a included (S,S)-DIOP (entry 10) and
DPPF (entry 11), albeit in lower yields, the latter required
heating at 758C. A lower e.r. of 87:13 was observed by using
(S,S)-DIOP as the chiral ligand. When (S)-BINAP (entry 12)
and BIPHEP (entry 13) were used, the dimeric product 4a
was obtained in good yields through a RhI-catalyzed cyclo-
dimerization process that has been previously described.[6]
However, our conditions employed a neutral RhI catalyst in
the presence of water whereas the literature reports used
cationic RhI sources under strictly anhydrous conditions. Both
protocols gave the product in excellent enantioselectivity
(e.r. > 99:1). Finally, reactions carried out in neat water using
a water-soluble ligand TPPDS did not yield any product
(entry 14).[7] We also screened other RhI sources such as
[Rh(cod)OH]2, [Rh(CO)2Cl]2, and Rh(cod)2OTf, all of which
showed no reactivity.[8]
which suggested the formation of a common achiral inter-
mediate during isomerization. Furthermore, the process was
catalyst-controlled since the (S,R)-ligand gave the opposite
enantiomer of 3a. These results supported the notion that
a RhI-catalyzed domino reaction[10] took place to generate 3a
from 1a (cf. Scheme 1): diol 2a was first formed from
oxabicycle 1a by RhI-catalyzed ARO with water as the
nucleophile, at a higher catalyst loading 2a was subsequently
converted to tetralone 3a by RhI-catalyzed enantioselective
isomerization. The presence of the allylic alcohol is crucial for
the isomerization since the analogous allylic methyl ether
(obtained from MeOH-induced ARO) did not isomerize after
17 h at 508C.
To investigate the generality of RhI-catalyzed ARO with
water, we next studied the scope of oxabicyclic alkenes
(Table 2). The effects of the remote substituents on the
Table 2: Scope of oxabicyclic alkenes in RhI-catalyzed ARO with water:
diol formation.[a]
We speculated that 2a is an intermediate en route to 3a
through an isomerization process. To test this hypothesis, we
subjected 2a (various enantiomeric compositions) to the
catalytic conditions (Scheme 2). In all cases, 2a isomerized to
3a in good yields.
Entry Substrate
Product
Catalyst
Yield[c]
loading[b] (e.r.)
83
(>99:1)
1[d]
2/4.4
82
(>99:1)
2
1/2.2
64
(>99:1)
3
0.5/1.1
Scheme 2. RhI-catalyzed enantioselective isomerization of diol 2a to
tetralone 3a.
68
(>99:1)
4[e]
5/10
5/10
An interesting feature of this isomerization is that a much
higher e.r. was observed than in the one-pot process using
oxabicycle 1a (cf. Scheme 1). At 258C, the enantio-enriched
(R)-3a was obtained in 15:85 e.r. whereas the one-pot process
only gave 33:67 e.r.[9] When repeating this reaction with
different concentrations, we found a significant impact on
conversion. At 0.1m concentration, only 12% product was
isolated (e.r. = 24:76) with 88% unreacted starting material.
At 0.4m concentration, the product was obtained in a high
yield of 89% (e.r. = 22:78). However, changing the concen-
tration did not lead to any improvement in enantioselectivity.
Preliminary studies showed that the conversion increased
linearly with catalyst loading and suggested that the reaction
is first order in the Rh catalyst (see Supporting Information).
The enantioselectivities (no significant differences) were not
dependent on the enantiopurity of the starting material,
79
(>99:1)
5[e]
[a] Reaction conditions: substrate 1b–f (0.2–1.2 mmol scale),
[Rh(cod)Cl]2 (0.5–5 mol%), (R,S)-PPF-PtBu2 (1.1–10 mol%), H2O
(56 equiv), THF (0.2m). [b] Mol% [Rh(cod)Cl]2/mol% (R,S)-PPF-PtBu2.
[c] Yield [%] of isolated product. [d] Reaction was run for 48 h. [e] No
formation of tetralone product at 258C or 508C.
oxabicyclic alkenes were studied by using difluoro-substituted
1b, dimethoxy-substituted 1c, and methylenedioxy-substi-
tuted 1d. Chiral trans-1,2-diol products 2b–d were obtained
with excellent enantioselectivities (e.r. > 99:1) (entries 1–3).
2
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!