Organic Letters
Letter
aromatic C−H oxygenation enabled by iron photocatalysis. The
advantages of our current method should include room
temperature reaction conditions, much lower price of iron
than palladium, copper, and ruthenium, and high tolerance to
halogen substitutes which would be left untouched.
spectra illustrate that the solution of 1:2 iron(III) and L3 has a
as FeCl , FeCl ·4H O, and Fe(OAc) were compatible with the
reaction although with lower yields (entries 8−10). Not
surprisingly, a decrease in yield was observed with 440 and
456 nm irradiation (entries 11 and 12) because the iron(III)
critical roles of the oxidant, iron catalyst, and visible light in the
protocol were demonstrated through the control experiments
3
2
2
2
We started our investigation into this intramolecular oxygen-
ation by exposing the water/alcohol solution of 2-biphenylcar-
boxylic acid, Fe(NO ) ·9H O, and NaBrO to visible light
3
3
2
3
(
Kessil 40 W 427 nm LED) for 16 h (Table 1). Whereas water,
a
Table 1. Optimization of the Reaction Conditions
(
entries 13−15). Without NaBrO , only a slight amount of
3
cyclization product 1 was observed as this is a net oxidative
transformation and the oxidant should be responsible to bring
iron(II) back to iron(III). No desired product 1 was isolated in
the absence of either iron catalyst or light. The intramolecular
oxygenation was severely inhibited by the addition of free radical
scavengers, such as TEMPO (2,2,6,6-tetramethylpiperidine-1-
oxyl) and BHT (3,5-di-tert-butyl-4-hydroxytoluene). These
radical quenching experiments suggest a radical mechanism
where the aroyloxy radicals are likely involved (entries 16 and
1
7). Finally, it is worth mentioning that the reaction at 10 mmol
gram scale could afford 1.53 g of the product benzo-3,4-
coumarin without further optimization (entry 18).
entry
iron(III)
ligand
none
solvent
wavelength yield (%)
Having determined the optimal conditions, we sought to
evaluate the scope of this aromatic C−H oxygenation method.
Substituents on the 2-biphenylcarboxylic acid substrates played
an important role in the reaction possibly through the electronic
and steric effects. Thus, when ligand L1 failed to afford a satisfied
yield for a specific substrate, L2, L3, or L4 would be used for a
better result. As illustrated in Scheme 1, a wide range of 2-
biphenylcarboxylic acids could furnish the corresponding
cyclization products in good to excellent yields. The reaction
proceeded smoothly with electron-donating/withdrawing sub-
stituents at both rings. Notably, a variety of functional groups,
such as halides (6, 7, 8, 12, 16, 17, 18, 21, 22), ethers (5, 10, 13,
14, 19), alkyl groups (2, 3, 4, 11, 15, 20), and trifluoromethyl
groups (9, 23) were well-tolerated. However, we noticed that
substituents on the 6, 2′, and 6′ positions of 2-biphenylcarbox-
ylic acids would heavily retard the reaction perhaps because of
the steric effect. Finally, the substrate with a naphthalene moiety
was also transformed into the lactone product in a diminished
but synthetically useful yield (24).
A possible mechanism for the intramolecular aromatic C−H
oxygenation is illustrated in Scheme 2. We postulated that
coordination of 2-biphenylcarboxylic acid to iron(III) 25
followed by deprotonation would readily form the aryl
carboxylate−iron(III) complex 26. Photoexcitation of the iron
complex 26 should render an intramolecular ligand-to-metal
charge-transfer event to afford the reduced iron(II) complex 27
and an aroyloxy radical 28, which rapidly underwent an
intramolecular electrophilic addition to the other phenyl ring.
The resulting radical intermediate 29 could then be oxidized by
iron(III) after deprotonation to furnish the rearomatization
1
2
3
4
5
6
7
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
5 mol %
none
HFIP
HFIP
HFIP
MeCN
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
427 nm
427 nm
427 nm
427 nm
427 nm
427 nm
427 nm
427 nm
427 nm
427 nm
440 nm
456 nm
427 nm
427 nm
dark
17
92
60
90
42
77
75
70
77
58
82
72
9
5 mol % Ll
5 mol % L2
5 mol % L2
5 mol % L3
10 mol % L3
5 mol % L4
5 mol % Ll
5 mol % Ll
5 mol % Ll
5 mol % Ll
5 mol % Ll
5 mol % Ll
5 mol % Ll
5 mol % Ll
5 mol % Ll
5 mol % Ll
b
8
9
c
d
1
1
1
1
1
1
1
1
1
0
1
2
3
4
5
6
7
8
e
0
0
5 mol %
5 mol %
5 mol %
5 mol %
f
427 nm
427 nm
12
6
g
h
5 mol % Ll
427 nm
78
a
b
c
d
Isolated yields. FeCl used. FeCl ·4H O used. Fe(OAc) used.
3
2
2
2
e
f
g
Without NaBrO . 5 equiv of TEMPO added. 3 equiv of BHT
3
h
added. At 10 mmol scale.
methanol, ethanol, and isopropyl alcohol are amenable to the
reaction, hexafluoroisopropanol (HFIP) proved to be the
superior solvent. It afforded the intramolecular oxygenation
product 1 in 17% yield (entry 1). To our delight, the yield of 1
increased dramatically up to 92% in the presence of 5 mol % of
ligand L1 (2,2′-bipyridine-6,6′-dicarboxylic acid, entry 2). The
UV−vis spectra show a significant optical absorption in the
Information). The reaction with ligand L2 in acetonitrile also
provided a high yield of 90% (di(2-picolyl)amine, entries 3 and
product 1. The terminal oxidant NaBrO would serve to bring
3
iron(II) back to iron(III) to close the catalytic cycles.
In summary, we have developed a mild and efficient protocol
for the intramolecular aromatic C−H oxygenation of 2-
biphenylcarboxylic acids via iron photocatalysis. This new
reaction exhibits a broad functionality tolerance at both phenyl
rings. Based on the radical quenching experiments and related
4
), and ligands L3 and L4 promoted the transformation
moderately (picolinamide and 2,2′-bipyridyl N,N′-dioxide,
entries 5−7). During exploration of the substrate scope, we
found that, in general, HFIP was superior than MeCN as a
solvent. It is of note that with the bidentate ligand L3, the 1:2
ratio of iron(III) to L3 led to a higher yield. Also, the UV−vis
25,26
mechanism studies in the literature,
we speculate that the
aryl carboxylate−iron(III) complexes should generate the
B
Org. Lett. XXXX, XXX, XXX−XXX