Angewandte
Communications
Chemie
Table 1: Optimization studies and control experiments.[a]
Entry
Deviation
Yield [%][b]
ee [%]
1
2
3
4
5
6
7
8
9
none
65 (56)[c]
75
80
80
28
77
12
0
Figure 2. Proposed catalytic mechanism for the visible-light-driven
asymmetric nickel-catalyzed acyl cross-coupling process.
NiBr2 instead of NiCl2
NiCl2dme instead of NiCl2
acetonitrile instead of THF
dioxane instead of THF
L2 instead of L1
L3 instead of L1
no light
no catalyst NiCl2
53
48
43
38
31
34
0
the catalytically active nickel intermediate and to secure
turnover. In the first catalytic cycle, the excited-state inter-
mediate 1* would reduce, by two discrete SETevents, the NiII
precatalyst to afford the active Ni0 intermediate A (Ep(NiII/
Ni0) = À1.2 V versus SCE in DMF).[13] The resulting, highly
unstable radical cation 1C+ would then undergo homolytic
–
–
0
[a] Reaction performed in THF [0.167m] at 108C for 48 h on a 0.1 mmol
scale using 2 equiv of 1a and 1 equiv of lutidine as base under
illumination by a single high-power (HP) LED (lmax =405 nm) with an
irradiance of 75 mWcmÀ2. [b] Yield determined by 1H NMR analysis of
the crude mixture using mesitylene as the internal standard. [c] The
number in parentheses indicates the yield of the isolated 3a after
chromatography purification on silica gel.
cleavage to generate a secondary C(sp3)-centered radical B.
2
À
Oxidative addition into the C(sp ) O bond of the anhydride 2
would afford the NiII–acyl complex C, which would intercept
the stabilized secondary radical, leading to the NiIII inter-
mediate D. Reductive elimination would then provide the
cross-coupling chiral ketone product 3. We anticipated that an
appropriate chiral ligand would provide control of the
stereoselectivity.[14] Finally, the generated NiI complex would
undergo SET reduction by the excited alkyl-DHPs 1*,
completing the nickel catalytic cycle while regenerating the
C(sp3) radical intermediate B.
ligands, including representative examples that have been
useful in other nickel-catalyzed enantioconvergent cross-
couplings (entries 6 and 7).[3,4] Control experiments con-
firmed that the reaction could not proceed in the absence of
light or a nickel catalyst (entries 7 and 8).
To validate our plan, the commercially available and
stable butyric anhydride 2a was selected as the acyl precursor
(Table 1). For the radical precursor, we chose the indole-
containing racemic DHP 1a because this would form product
3a bearing a stereogenic center a to the indole nitrogen. This
structural motif is synthetically interesting because it is found
in many natural products and pharmaceutical drugs.[15] How-
ever, it is a difficult target as testified to by the paucity of
asymmetric catalytic protocols available for the preparation
of enantioenriched N-alkylated indoles.[16] We conducted our
experiments in THF under irradiation by a single high-power
visible-light-emitting diode (LED, lmax = 405 nm) with an
irradiance of 75 mWcmÀ2, as controlled by an external power
supply (full details of the illumination set-up are reported in
the Supporting Information). By examining a range of
reaction parameters, we determined that NiCl2 and the
chiral box ligand L1[17] can accomplish the enantioconvergent
photochemical cross-coupling in good yield and high ee (3a
formed in 65% yield and 75% ee; entry 1). Other nickel salts
provided slightly improved stereocontrol, but at the expense
of chemical yield (entries 2 and 3). Because of the good
compromise between reactivity and enantioselectivity, we
selected NiCl2 for further optimization. No improvement was
achieved with other solvents (entries 4 and 5) or chiral
Using the optimized conditions described in Table 1,
entry 1, we tested the generality of the photochemical cross-
coupling process (Figure 3). We first evaluated the scope of
the radical precursors 1. Several halogen substituents on the
indole scaffold were tolerated well, affording the correspond-
ing N-alkylated chiral indoles in good yields and stereoselec-
tivity (products 3b–d). A carbazole scaffold was also intro-
duced within the final products, albeit at the expense of
stereoselectivity (adducts 3e and 3 f). We then evaluated
different anhydrides as acyl coupling partners. Different
substitution patterns were tolerated, including an aryl moiety
(products 3i, 3k, 3l), a ketone (3h and 3j), and an alkyl
chloride (3g). We failed to synthesize DHP radical precursors
1 bearing a substitution pattern other than methyl. This
limitation is somehow mitigated by the ability to forge
a ketone moiety with a methyl a-stereogenic center. This is an
important synthetic achievement,[18] for which there are few
effective catalytic asymmetric protocols.
We then sought to extend the applicability of this photo-
chemical cross-coupling strategy to the stereocontrolled
preparation of acyclic a,a-aryl,alkyl ketones, which are
versatile synthetic intermediates for the synthesis of natural
products and pharmaceutical agents.[19] This required the
preparation of racemic DHP radical precursors bearing both
2
ꢀ 2019 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2019, 58, 1 – 6
These are not the final page numbers!