Angewandte
Chemie
entry 1). Another pyrrolidine-type catalyst (S)-3[13] gave a
lower yield and similar enantioselectivity (Table 1, entry 2).
In terms of enantioselectivity, no improvement was observed
with the binaphthyl-based amino alcohol catalyst (S)-4[14]
(Table 1, entry 3). We assumed that the poor enantioselectiv-
ity might arise from the sterically less-hindered oxygen atom
of 1. Thus, a binaphthyl-based secondary amine catalyst (S)-5,
containing bulky substituents at the 3,3’-positions, was
synthesized by the introduction of trimethylsilyl groups into
(S)-4. Gratifyingly, using the sterically more-congested cata-
lyst (S)-5, the desired aminoxylation product was obtained in
excellent enantioselectivity albeit with low yield (Table 1,
entry 4).
Encouraged by this promising result, the reaction con-
ditions were then optimized. Under the reaction conditions at
08C, 3-phenylpropanal was found to be oxidized into 3-
phenylpropanoic acid, and the catalyst could also be deacti-
vated via oxidation by 1 and/or BPO. These undesired side-
reactions could be suppressed somewhat by lowering the
reaction temperature and decreasing the amount of BPO;
higher concentration also resulted in an improved yield
(Table 1, entry 5). Switching solvent from dichloromethane to
tetrahydrofuran and toluene did not improve the yield
(Table 1, entries 6 and 7).
purity. In this transformation, the 2,2,6,6-tetramethylpiper-
idinyl group was not oxidized and acted as a protecting group.
The reaction of a-aminoxy aldehyde 7 with PhMgBr in
tetrahydrofuran proceeded smoothly to give the correspond-
ing half-protected 1,2-diol 9 in excellent diastereoselectivity
without loss of optical purity [Eq. (2)]. The observed
diastereoselectivity can be explained by non-chelation
control, which might be attributable to the bulky and non-
protic aminoxyl group of 7 (Figure 1, left), and contrasted
sharply with that observed in the chelate-controlled reactions
between Grignard reagents and a-aminoxy aldehydes gen-
erated in situ from nitroso compounds (Figure 1, right).[16]
This reaction system was then applied to various alde-
hydes (Table 2). Under the optimized conditions, the corre-
sponding a-aminoxylated products were obtained with good
to excellent enantioselectivity in all cases examined.
Table 2: Aminoxylation of various aldehydes.[a]
Entry
R
Yield [%][b] ee [%][c] Entry
R
Yield [%][b] ee [%][c]
1[d]
2
3
Me
Et
Bu
Bn
75
88
77
99
91
92
93
95
5
6
7
allyl
iPr
Cy
89
94
95
95
99
98
Figure 1. Possible transition-state models for diastereoselective nucle-
ophilic addition to a-aminoxy aldehydes.
4
[a] The reaction of an aldehyde (0.1 mmol), TEMPO (0.13 mmol), and
BPO (0.06 mmol) was carried out in CH2Cl2 (0.2 mL) in the presence of
(S)-5 (0.005 mmol). [b] Yield of isolated product. [c] The ee value of the
product was determined by HPLC analysis using a chiral column. [d] The
reaction time was 12 h.
For this aminoxylation reaction, two radical reaction
pathways and an ionic reaction pathway could be suggested:
1) The enamine radical cation 11, which is generated by
oxidation of the enamine intermediate 10 with BPO, reacts
with a TEMPO radical to give the iminium intermediate 12
(Scheme 2, path a). 2) The enamine 10 is oxidized by oxoam-
monium salt 1, which is generated from TEMPO and BPO, to
give the enamine radical cation 11 (path b). 3) Enamine 10
reacts directly with 1 in an ionic (nucleophilic addition)
pathway, giving 12 (path c).
Generation of oxoammonium salt 1 from TEMPO and
BPO was confirmed by an experiment in which treatment of
3-phenylpropanol with TEMPO (1 equiv) and BPO
(0.5 equiv) in dichloromethane led to the formation of 3-
phenylpropanal in 85% yield [Eq. (3)]. In addition, when the
aminoxylation of butanal was performed in the presence of 3-
phenylpropanol, the formation of a-aminoxy butanal 13 was
accompanied by oxidation of 3-phenylpropanol and the
aminoxylation of the resulting 3-phenylpropanal [Eq. (4)].
It should be noted that an a-aminoxyl aldehyde could be
isolated by column chromatography without reduction of the
carbonyl group,[15] and neither decomposition nor racemiza-
tion was observed. For instance, the isolated a-aminoxy
aldehyde 7 (89% yield, 96% ee) was stored in [D]chloroform
1
for 60 hours without any change observed by H NMR and
HPLC analyses (see the Supporting Information). To exam-
ine the synthetic utility of this aminoxylation reaction, an
optically enriched a-aminoxy aldehyde 7 was converted into
its corresponding a-hydroxy acid derivative [Eq. (1)]. Thus,
treatment of the a-aminoxy aldehyde 7 with NaClO2 in the
presence of NaH2PO4 and 2-methyl-2-butene resulted in
clean formation of a-aminoxy acid 8 without loss of optical
Angew. Chem. Int. Ed. 2010, 49, 6638 –6641
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim