Table 2. Scope of asymmetric hydroamination with chiral aldehydes[a]
enantioenriched products with moderate enantiocontrol (3
examples, 45–78% ee; Table 1, entry 1). However, this alde-
hyde is notorious for its ready epimerization, and prelimina-
ry efforts thus focused on the identification of a more robust
aldehyde. Increased enantioselectivity was observed with
the more easily handled diphenyl analogue 1b (88% ee;
Table 1, entry 2), but an early screening of the substrate
scope showed that this procedure was not general and was
plagued by some catalyst epimerization. Unfortunately, the
Ley aldehyde[9] (1c), the Garner aldehyde[10] (1d), and an al-
dehyde lacking an a proton (1e) led to poor reactivity and
enantiocontrol (Table 1, entries 3–5). The low yields ob-
tained with these aldehyde catalysts suggest that their steric
bulk has a negative impact on the preassociation step, which
involves the formation of the mixed aminal I. However, the
acyclic aldehyde 1 f also showed poor reactivity (Table 1,
entry 6), which highlights the benefits associated with the
cyclic structures present in 1a and 1b. It is also worth noting
that catalysts 1a–d and 1 f had the common problem that
epimerization (via enamine formation, for example) would
lead to catalyst racemization. In contrast, the bicyclic alde-
hydes 1g and 1h possess several stereocenters embedded in
a rigid bicyclic structure that could help prevent epimeriza-
tion and help ensure retention of the original chirality.[11]
Encouragingly, high enantioselectivities were observed with
both the 6- and 5-membered bicyclic aldehydes 1g (85% ee;
Table 1, entry 7) and 1h (94% ee; Table 1, entry 9). Catalyst
1h showed remarkable stability as the reaction proceeds in
72 h with high selectivity. In addition, these catalysts led to
the R enantiomer of the diamine, which complements the
ability of catalyst 1b to provide access to the S enantiomer.
Fortunately, two additional observations were made during
this screening process that led to higher enantioselectivities
with catalyst 1b. We found that addition of the catalyst last
proved optimal, and higher enantioselectivities were ob-
tained with C6F6 as solvent. By following this revised proce-
dure, catalyst 1b afforded the S enantiomer in 97% ee
(Table 1, entry 11)!
Entry
Product
Catalyst
Yield
[%][b]
ee
[%][c]
1
2
3
4
R3 =H (2a)
R3 =H (2a)
R3 =Cl (2b)
R3 =OMe (2c)
1b
1h
1b
1b
91
79
81
86
97
À88
94
92
5
6
7
8
9
R1 =3,5-(CF3)2Bn (2d)
R1 =3,5-(CF3)2Bn (2d)
R1 =iPr (2e)
1b
1h
1b
1b
1h
82
74
60
63
60
82
À92
60
R1 =(CH2)3Ph (2 f)
R1 =(CH2)3Ph (2 f)
71
À77
10
11
12
13
14
15
16
17
18
19
20
R2 =Me (2g)
1b
1b
1h
1b
1b
1b
1h
1b
1h
1b
1h
91
85
76
83
81
62
51
75
71
73
66
56
82
R2 =allyl (2h)
R2 =allyl (2h)
R2 =p-NO2Bn (2i)
R2 =BrBn (2j)
À88
95
90
60
R2 =CH2CH
R2 =CH2CH
U
À88
R2 =CH2CO2Et (2l)
R2 =CH2CO2Et (2l)
R2 =(CH2)CO2Me (2m)
R2 =(CH2)CO2Me (2m)
72
À91
82
À90
[a] Performed with hydroxylamine (1 equiv), allylamine (1.5 equiv), and
catalyst (0.2 equiv) in solvent (1m) under argon, for 24 h with catalyst 1b
(and 72 h with 1h) at room temperature. [b] Yields of the isolated prod-
uct. [c] Determined by chiral HPLC analysis of derivatized products (see
the Supporting Information).
With efficient conditions giving access to either enantiom-
er of the diamine motifs with catalysts 1b and 1h,[12] the ap-
plicability of this enantioselective reaction was evaluated
(Table 2). Electron-rich and electron-poor benzylic hydrox-
ylamines displayed excellent enantioselectivity with benzyl-
allylamine as substrate (Table 2, entries 1–6). In contrast, re-
duced enantioselectivity was seen with two aliphatic hydrox-
ylamines (Table 2, entries 7–9). To determine if this effect
was steric or electronic in nature, the reaction of benzylhy-
droxylamine and methylallylamine was also performed. This
reaction led to reduced enantioselectivity (56% ee; Table 2,
entry 10), which suggests that the size of the allylic amine is
important for high enantioselectivity. Additionally, the lower
enantioselectivity could result from reduced diastereoselec-
tivity in the formation of the mixed aminal I or from an
amine-catalyzed aldehyde epimerization process (see
below). To probe this, many secondary allylic amines were
reacted with benzylhydroxylamine and catalysts 1b and 1h
(Table 2, entries 11–20). Collectively, these results reveal
that the enantioselectivities obtained with catalyst 1b are
quite sensitive to the structure of the allylic amine (Table 2,
entries 10, 11, 13, 15, 17, and 19), and that the lowest enan-
tioselectivities are observed with the least nucleophilic
amines (60–95% ee; Table 2, entries 15 and 17 vs. 19). In
contrast, 1h afforded the enantiomeric products reliably in
high enantioselectivities (88–91% ee; Table 2, entries 12, 16,
18, and 20). Overall, the enantioselectivities obtained are, in
many cases, the highest obtained for intermolecular hydroa-
minations of unactivated alkenes by any method[7i,j] (includ-
ing metal-catalyzed reactions) and clearly validate the teth-
ering strategy as an effective method in asymmetric organo-
catalysis.
The enantioselectivity trend highlighted above for alde-
hyde 1b suggests that catalyst epimerization occurs under
the reaction conditions. To probe this, the aldehyde was sep-
arately exposed to catalytic amounts of each reagent and
the enantiopurity of the catalyst was determined after 24 h
(Figure 1). Both experiments showed significant epimeriza-
&
2
&
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!