Organic Letters
Letter
toxicity, and oxidizing activity.8 Furthermore, the byproducts
of carboxylic acid and iodobenzene can be recycled after the
decarboxylation coupling.9 Here, we achieved the ortho
alkylation of heteroarene N-oxides by employing hypervalent
iodine(III) carboxylates under copper photocatalysis (Figure
2c). Although the system using a copper complex as a
photocatalyst under visible light is still in its infancy, the
copper complex has great potential in photocatalysis because it
is cheap and abundant on the Earth. Compared to ruthenium
or iridium photocatalysts, copper photocatalysts not only have
economic advantages but also play a unique role in promoting
the transfer of electrons to organic substrates.10 In this work,
we report a highly site-selective C−H functionalization using a
copper complex in addition as a photocatalyst.
directly increased to 77% (entry 8). Finally, other 1,10-
phenanthroline ligands, such as 1,10-phenanthroline L2, 4,7-
dimethoxy-1,10-phenanthroline L3, and 2,9-dimethyl-4,7-
diphenyl-1,10-phenanthroline L4, were also explored for this
transformation, but no further increase in the yield of 3a was
achieved (entries 9−11, respectively). Control experiments in
the absence of ligand or visible light were also conducted, and
no coupling product could be obtained (entry 12 or 13,
respectively). It is noteworthy that this reaction requires only
16 W white light instead of blue light that may irritate eyes and
can also react in the air.
After optimizing the reaction conditions for our decarbox-
ylative C−H alkylation of heteroarene N-oxides, we next
expanded the scope of the hypervalent iodine(III) carbox-
ylates. As shown in Scheme 1, a variety of primary, secondary,
To optimize this reaction, we chose hypervalent iodine(III)
cyclohexanoate 1a and quinoline N-oxide 2a as the model
substrates. At the beginning of the study, we used metal-free
a
Scheme 1. Scope of Hypervalent Iodine(III) Carboxylates
photocatalyst (Acr-Mes)+ClO4 or Eosin Y (Table 1, entry 1
−
a
Table 1. Optimization of the Reaction Conditions
yield (%)
ligand
(equiv)
b
c
c
entry
catalyst (equiv)
3a
3a′
3a″
1
2
3
4
5
6
7
8
(Acr-Mes)+CIO4− (0.1)
Eosin Y (0.1)
Cu(OAc)2 (0.1)
CuBr2 (0.1)
CuBr (0.1)
CuCI (0.1)
CuI (0.1)
CuCI (0.2)
CuCI (0.2)
CuCI (0.2)
CuCI (0.2)
CuCI (0.2)
CuCI (0.2)
−
−
41
30
40
trace
45
12
6
0
0
0
0
0
0
0
0
0
0
0
5
trace
0
L1 (0.15)
L1 (0.15)
L1 (0.15)
L1 (0.15)
L1 (0.15)
L1 (0.3)
L2 (0.3)
L3 (0.3)
L4 (0.3)
−
0
0
0
0
0
0
0
0
55
0
d
77 (81)
a
9
45
40
64
trace
0
The reactions were carried out using 2a (0.2 mmol, 1.0 equiv), 1
10
11
12
(0.3 mmol, 1.5 equiv), CuCl (0.2 equiv), BPhen (0.3 equiv), and
DCM (2.0 mL) under white light-emitting diodes for 12 h at rt.
Isolated yields based on quinoline N-oxides 2a.
0
0
e
13
L1 (0.3)
a
and tertiary alkyl decarboxylative couplings with quinoline N-
oxide 2a were successful. Hypervalent iodine(III) 1a−c with
the decarboxylative centers in six- and five-membered rings can
furnish the desired products 3a−c in good yields of 77%, 75%,
and 63%, respectively. Even with more strained cyclobutyl and
cyclopropyl hypervalent iodine(III) carboxylates 1d and 1e,
the coupling products 3d and 3e, respectively, could be
obtained in good to moderate yields. Furthermore, general
secondary alkyl hypervalent iodine(III) carboxylate 1f gave
target product 3f in excellent yield. The primary alkyl
hypervalent iodine(III) carboxylates 1g−j were also found to
be suitable for the reaction, affording the corresponding
products 3g−j, respectively, in good to excellent yields, and
these mild reaction conditions displayed good compatibility
with sensitive groups, such as amino, alkenyl, and alkynyl
groups. Interestingly, this transformation can also provide an
enantiopure amino acid derivative 3k by employing aspartic
Reaction conditions: hypervalent iodine(III) cyclohexanoate 1a (0.3
mmol, 1.5 equiv), quinoline N-oxide 2a (0.2 mmol), catalyst, ligand in
DCM (2.0 mL) under white light-emitting diodes for 12 h at rt.
b
c
d
e
Isolated yields. GC yields. N2 atmosphere. No light.
or 2, respectively), and the product 2-cyclohexylquinoline 1-
oxide 3a can be obtained in a yield of 41% or 30%,
respectively. At the same time, two byproducts, 3a′ and 3a″,
were detected by GC-MS (entry 3). Then we examined copper
salts under the same conditions, and the results showed that
divalent copper CuBr2 provided only a trace of the product
(entry 4); in contrast, product 3a can be obtained in a yield of
45% by using monovalent copper CuBr (entry 5). When CuBr
was replaced by CuCl, the yield of 3a increased to 55% (entry
6). For CuI, no desired product was generated (entry 7). As
the catalytic amount of CuCl was increased to 20% and that of
the ligand BPhen was changed to 30%, the yield of 3a was
B
Org. Lett. XXXX, XXX, XXX−XXX