Organic Letters
Letter
a
metric preparation of chiral boron reagents of 1 (M = B), only
recently have catalytic asymmetric variants emerged to
promote the catalytic generation of 1.2,10,11 For example,
Krische2a (Figure 1B) recently disclosed an enantioselective Ir-
catalyzed aminoallylation of aldehydes through hydrogen
transfer from alcohol 5 to allenamide 6 generating the
necessary α,γ-aminoanion nucleophile (1, M = Ir) and the
aldehyde electrophile. Additionally, our group10a recently
reported an orthogonal method for the aminoallylation of
ketone electrophiles enabled by Cu-catalyzed reductive
coupling12,13 (Figure 1C). Here, the readily available Evans-
auxiliary derived chiral allenamide 9 was employed for
stereochemical control affording high diastereoselectivies.10a,11
While this method is practical due to the low cost of the Evans
auxiliary, we appreciated the fact that absolute stereochemical
control by a chiral Cu-catalyst with an achiral allenamide
would increase atom efficiency (Figure 1D). Significantly,
enantioselective metal catalyzed aminoallylation of ketone
electrophiles is unknown and can be more challenging than
aldehydes due to the decreased reactivity and steric differ-
entiation of ketones versus aldehydes. Herein we report the
development of the first Cu-catalyzed enantioselective amino-
allylation of ketone electrophiles.
Initial studies focused on identifying an appropriate chiral
ligand scaffold to afford high diastereo- and enantioselectivity
in the reductive coupling of ketones and achiral allenamide 11
(Table 1, entries 1−7).14 Importantly, in all cases, variable
amounts of carbonate migration product 13a were formed
from internal trapping of the Cu-alkoxide11,14 intermediate
with only single diastereomers of 12a and 13a observed. Of the
ligands studied, W8 was identified as the best candidate
providing the desired branched reaction product 12a in good
enantioselectivity as a single diastereomer (entry 5).
Interestingly, (S,S)-Ph-BPE (entry 1) and J11 (entry 7) were
inferior to W8 despite their widespread use in Cu-catalyzed
reductive coupling reactions.12,13
Table 1. Chiral Ligand Survey
b
b
c
entry
ligand
% 12a
12a:13a
er 12a
1
(S,S)-Ph-BPE
64
82
51
64
77
58
60
71
58
61
89:11
83:17
86:14
90:10
81:19
91:9
78:22
87:13
>99:1
90:10
20:80
18:82
30:70
57:43
2
3
4
5
6
7
8
9
(R)-BINAP
(R)-Segphos
W3
W8
J6
J11
W8
W8
W8
d
93:7
15:85
28:72
87:13
88:12
91:9
85:15 ,
97:3
96:4
e
f
fg
10 ,
h
i
j
ik
11
W8
W8
W8
45
50
77
52:48
fh
i
j
12 ,
>99:1
fg h
,
i
j
13 ,
>99:1
a
1a (0.25 mmol), 11 (0.375 mmol), and 0.50 mmol Me(MeO)2SiH
in 0.5 mL of toluene. In all cases, a single diastereomer of 12a and 13a
was obtained (1H NMR spectroscopic analysis). See the Supporting
b
Information for additional details. Determined by 1H NMR
spectroscopy on the unpurified reaction mixture using dimethylfu-
marate as standard. Value determined by chiral HPLC analysis. Er
of 13a was 50:50. Using 10 equiv of silane. 2 equiv of t-BuOH
added. PhCF3 used as solvent. Propiophenone (8b) used in place of
c
d
e
f
g
h
i
j
k
8a. Value for 12b. Ratio of 12b:13b. Er fo 13b was 60:40.
Compounds 12 and 13 were produced in differing
enantiopurities (Table 1, entries 5 and 11); one explanation
consistent with this outcome is reversibility in the allylcupra-
tion.14 Reversible allylation15 in metal catalyzed reductive
coupling reactions has not been identified prior to our
work,11,16 and this issue would have significant ramifications
on catalyst stereocontrol. For instance, the enantiopurity of
product 12 would be dependent on the subsequent rate of
silylation vs carbonate migration of the intermediate Cu-
alkoxide formed after allylcupration leading to catalyst turnover
if the allylcupration step was reversible.14 Along these lines,
and in an effort to improve enantioselection, we reasoned that
enantioselectivities may be improved if the rate of trapping of
the Cu-alkoxide intermediate formed after allylcupration could
be increased relative to carbonate migration. This may be
achieved either through an increased silylation rate or by use of
protic additives capable of quenching the Cu-alkoxide by
protonolysis17 (e.g., t-BuOH). In this regard, examination of
alternate silane reducing agents or reaction solvents led to no
improvements.14 Use of excess silane (Table 1, entry 8)
reduced the amount of 13a formed but also reduced the
enantiopurity of 12a. Addition of t-BuOH as a proton donor
generally mitigated the formation of 13 but afforded reduced
yields presumably due to competitive protonation of the N-
allyl(Cu) nucleophile (entries 5 vs 9/10 and 11 vs 12/13).
This effect was more pronounced when a more sterically
demanding ketone was used (propiophenone (8b), entries
11−13), and use of PhCF3 as solvent led to improved yields
(entry 12 vs 13). The large amounts of 13 formed in the
absence of t-BuOH with 8b are consistent with an increased
rate of carbonate migration due to an enhanced Thorpe−
Ingold effect.
With optimized conditions in hand (Table 1, entries 5 and
13), the ketone scope was examined (Scheme 1). Notably, the
Me-group of 8 could be replaced with increased substitution
providing products in high enantioselectivities (12a−12c),
which can often be challenging due to the decreased steric bias
of the two ketone substituents. Para-substitution of the ketone
Ph-group generally led to a decrease in enantioselectivities
(12d−12j); however, enantiopurity could be improved
through the use of t-BuOH as an additive (Method B).
Here, electron-poor ketones (8d−e) afforded good yields
whereas electron-rich ketones (8f−i) provided poor yields with
Method B due to incomplete conversion from competitive
protonation of the allenamide, likely due to the reduced rate of
addition to these less electrophilic ketones. Interestingly,
ketones with meta-substitution (8k−o) returned to typical
enantioselection levels as was obtained with 8a; however,
addition of t-BuOH led to reduced er with the exception of the
bromo derivative 12m. For ketones containing both meta- and
B
Org. Lett. XXXX, XXX, XXX−XXX
Klake, Raphael K.
