Communications
mepy}4]BF4 provides a substantial rate enhancement. With
One of the significant challenges in asymmetric Lewis acid
catalysis is a 1,3-dipolar cycloaddition between nitrones and
enals[10–13] to form isoxazolidines. A variety of chiral catalysts
have been used for the transformation with methacrolein,
which occurs in variable yields, usually below room temper-
ature, with the use of 5–10 mol% of catalyst and excess
methacrolein (Table 2). Ruthenium and iron catalysts
(Table 2, entries 1–2) appear to favor the formation of
2,[10,11] and the only example of a nickel catalyst (Table 2,
entry 5, with chiral ligand 3, which is shown in Scheme 2)
shows complete selectivity for the formation of 2.[12] Chiral
dirhodium(II) carboxamidates have been unsuitable because
they lack catalytic activity (e.g., Table 2, entry 6). In contrast,
chiral Rh5+ carboxamidates show high catalytic activity
(Table 2, entry 7). [Rh2{(5S)-mepy}4]BF4 exhibited high regio-
selectivity for 2 with modest enantioselectivity. Increasing the
steric bulk of the ligand ester group from methyl (4+) to
isopropyl (5+)[9] and, reported for the first time, to (R)-
menthyl (6+) enhanced the enantioselectivity for 1, but not for
2, for which the ee value remained at nearly the same level
throughout the selection of dirhodium catalysts. Other chiral
cationic dirhodium carboxamidate catalysts, [Rh2{(4S)-
meox}4]BF4 and [Rh2{(4S)-meaz}4]BF4, gave regio-and enan-
tioselectivities that were lower than those from [Rh2{(5S)-
mepy}4]BF4. Results do not vary when the molar ratio of
methacrolein over the nitrone is varied from 1.4 to 10, and the
enantioselectivity is the same when the catalyst loading is
increased from 5 to 10 mol%. Furthermore, there is no
detectable variation of regioselectivity or enantioselectivity
ethyl glyoxylate, enantioselectivity rose from 20 to 74% ee
with the enantiomer of the same catalyst under the same
conditions, and the rate of reaction with [Rh2{(5S)-
mepy}4]BF4 is significantly faster than that with [Rh2{(5S)-
Table 1: Influence of the cationic dirhodium (5S)-mepy catalysts on
reactivity and selectivity in hetero-Diels–Alder reactions of the Dani-
shefsky diene with representative aldehydes.[a]
Aldehyde
Catalyst
Conv. [%][b]
ee [%][c]
p-NO2C6H4CHO
p-NO2C6H4CHO
C6H5CHO
[Rh2{(5S)-mepy}4]
[Rh2{(5S)-mepy}4]BF4
[Rh2{(5S)-mepy}4]
[Rh2{(5S)-mepy}4]BF4
[Rh2{(5S)-mepy}4]
[Rh2{(5S)-mepy}4]BF4
[Rh2{(5S)-mepy}4]PF6
[Rh2{(5S)-mepy}4]SbF6
53
70
<5
40
<10
100
100
100
73
93
–
88
20
74
76
76
C6H5CHO
EtOOCCHO
EtOOCCHO
EtOOCCHO
EtOOCCHO
[a] Reactions were performed with1.0 mol% catalyst at room temper-
ature in anhydrous dichloromethane with a reaction time of 24 h using
1
1.1 equiv diene. [b] Determined by H NMR analysis. [c] Determined by
HPLC on an OD-H or AD-H column.
mepy}4]. The estimated increase in rate by [Rh2{(5S)-
mepy}4]BF4 is at least a factor of ten. As can be seen from
5+
this data, the anion of the Rh2 complex has no measurable
effect on the enantioselectivity. In its use as a Lewis acid
5+
catalyst, coordination of the Rh2 complex with water was
expected to produce a protonic acid; to circumvent this
problem these reactions were performed in the presence of a
noncoordinating base.
À
À
with the anion of the cationic catalyst (BF4 vs. PF6 or
À
SbF6 ). Use of 2,6-di-tert-butylpyridine (DTBP) to remove
The oxidation of organic com-
pounds by dirhodium(II/III) car-
Table 2: Comparative influence of cationic dirhodium carboxamidate catalysts on reactivity and
selectivity in 1,3-dipolar cycloaddition reactions between C,N-diphenylnitrone and methacrolein.[a]
boxamidates was anticipated.
Indeed, we have known for many
5+
years that Rh2 complexes are
4+
reduced to Rh2 by diazoacetates,
and this is one of the factors that
allows dirhodium(II) catalysts to
be used with high TON.[9] How-
ever, we anticipated that there are
Entry
Catalyst
Mol%
T
[C8]
Yield
[%]
1/2[b]
1/2
ee [%][c]
other oxidizable substrates in
1
2
3
4
5
6
7
8
[CpRu{Ar2POCH*(Ph)C*H(Ph)OPAr2}][d]
[CpFe{Ar2POCH*(Ph)C*H(Ph)OPAr2}][d]
[Cp*Rh{Ph2PC*H(Me)CH2PPh2}][e]
[Cp*Rh{Ph2PC*H(Me)CH2PPh2}][e]
Ni(ClO4)2·6H2O+(R,R)-DBFOX/Ph( 3)[f]
[Rh2{(5S)-mepy}4] (4)
[Rh2{(5S)-mepy}4]BF4 (4-BF4)
[Rh2{(5S)-ippy}4]BF4 (5-BF4)
[Rh2{(5S,R)-menpy}4]SbF6 (6-SbF6)
[Rh2{(5S,R)-menpy}4]SbF6 (6-SbF6)[g]
[Rh2{(5S,R)-menpy}4]SbF6 (6-SbF6)[h]
5
5
5
5
10
5
5
5
5
5
À20
À20
À25
0
92
85
100
100
73
<5
80
64
40:60
20:80
63:37
53:47
0:100
3:97
13:87
12:88
33:67
37:63
24:76
94:76
91:87
90:75
85:68
–:96
22:20
30:64
63:71
88:67
94:71
95:53
5+
whose presence Rh2
will be
reduced to Rh24+; thus, one proce-
dural requirement of our investi-
gation has been to test the redox
stability of the reacting partners
RT
À20
À20
RT
RT
RT
5+
with the Rh2 catalyst. We have
found, for example, that [Rh2{(4S)-
meox}4]BF4 is reduced by the Dan-
ishefsky diene, and the rate of this
oxidation is competitive with ca-
talysis at temperatures above 408C.
In contrast, [Rh2{(5S)-mepy}4]BF4,
which has a much lower oxidation
potential (358 mV vs. 742 mV),[5] is
stable to reduction by the Dan-
ishefsky diene over long reaction
times.
9
10
11
88
90
95
1
RT
[a] Reactions were performed in anhydrous dichloromethane with a reaction time of 24 h using a slight
excess of methacrolein (1.4 equiv), 10 mol% 2,6-di-tert-butylpyridine, and 4- molecular sieves (0.5 g
per mmol of nitrone). [b] Yield determined by product mass and 1/2 ratio determined by 1H NMR
analysis of reaction mixture using easily distinguished aldehydic protons. [c] Determined by 1H NMR of
the diastereomeric imine protons formed from reaction of 1 and 2 with( S)-(À)-a-methylbenzylamine
(reference [11a]); racemic product was obtained by reaction withdirohdium caprolactamate.
[d] Reference [10a]. [e] Reference [11a]. [f] Reference [12]. [g] In situ generated catalyst. [h] Catalyst
generated in situ witha reaction time of 48 h.
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1439 –1442