Full Papers
doi.org/10.1002/cssc.202002161
ChemSusChem
ure 1b). This fairly large KIE value (>6.4) strongly suggests the
HAT to be the rate-determining step, and the process might be
considerably assisted by hydrogen atom tunneling.[11] The
resulting ketyl radical is a very short-lived species and a strong
reductant, which readily performs electron transfer to generate
more stable carbonyls.[12] Accordingly, the intermediate IV can
reduce another molecule of PA at its excited state, thus making
the catalytic loop complete and converting it into the reduced
form of the catalyst, PAH2. As a distinct proof of the PAH2
generation, it was isolated from a stoichiometric reaction under
anaerobic condition and examined by NMR spectroscopy and
high-resolution MS (Supporting Information, Figures S4–S6).
Importantly, PAH2 can be easily converted back to the catalyti-
cally active state PA by simple aerial oxidation, forming H2O2 as
the byproduct. The formation of H2O2 was unambiguously
authenticated by standard iodometric detection (Figure S14).
The use of molecular oxygen or even simpler aerial oxygen as a
terminal oxidant makes the process environmentally benign
and appealing. As proposed through the catalytic cycle in
Scheme 4d, the role of the terminal oxidant is only to
regenerate the catalyst by oxidizing its reduced form PAH2. To
prove that oxygen is only essential for catalyst regeneration, we
performed a stoichiometric alcohol oxidation under inert
conditions, keeping all other reaction parameters intact, and
found a 59% yield of the aldehyde formation. Furthermore, the
H/D KIE measurement under this stoichiometric condition was
observed to be 11.1�0.2 (Scheme 4a).
Establishing the efficient alcohol oxidation by PA and KOtBu
combination under visible-light-induced condition, we planned
to survey the scope of this method in value-added heterocycle
formation. From the perspective of sustainability, visible-light-
induced and metal-free systems are becoming very popular,
and such a system has been described recently to phosphor-
ylate thioflavones.[16] As a testament to the broad scope for this
bio-inspired aerobic oxidation methodology, we demonstrate
the synthesis of two model heterocycles, quinoline and
pyrimidine, under very mild conditions. The ease of both
primary and secondary alcohol oxidations under aerobic
atmosphere prompted us to choose a 2-aminobenzyl alcohol
and a secondary alcohol 1-phenylethanol, so that dehydrogen-
ation reactions followed by aldol condensation and ring
cyclization may lead to the formation of quinoline. Gratifyingly,
2-phenyl quinoline was isolated in 84% yield when the reaction
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
°
mixture was heated at 70 C under visible light for 8 h. Although
major experiments were done under blue light, we note that a
household light bulb or even sunlight can also easily promote
the reaction. Furthermore, a diverse range of substituted
secondary alcohols were tested to examine the wide scope of
the method, and the respective quinolines (2a–2f) were
isolated in good to excellent (64–84%) yields (Scheme 5).
Interestingly, secondary alcohols with heteroaromatic rings as
well as aliphatic secondary alcohols were well tolerated under
the mild reaction conditions and offered quinolines 2g–2n in
good yields. Notably, such an iminoquinone-based redox
catalyst circumvents the requirements of transition metal, high
temperature, and extended time of many prior synthetic reports
The almost identical KIE with catalytic conditions unequiv-
ocally establishes that the main mechanistic cycle of the
reaction remains same in the presence or absence of oxygen.
To unambiguously prove the intermediacy of a radical during
hydrogen transfer, we further performed a Hammett analysis
(Figure S22). For the alcohol oxidation, six different benzyl
alcohols were chosen, and the substituents were varied at its
para-position ranging from strongly electron-donating À OMe to
electron-withdrawing À NO2. Interestingly, plotting the Hammett
constants[13] σp with respect to the relative rate of product
formation shows a straight line (R2 =0.96) with minimal slope.
The 1 value (slope of the staright line) obtained from the plot is
À 0.36 (Figure S22), which can be attributed to the lack of
strong electronic influence of the para-substituents on the HAT
rate. Such a small 1 value has been repeatedly alluded to the
radical-mediated processes in
a diverse set of reactions
including metal oxo-based CÀ H activation reactions.[14] Further-
more, to gather compelling evidence for a one-electron radical
process, a radical clock experiment with phenyl-(2-phenyl-
cyclopropyl) methanol was performed, which offered a large
amount of ring-opened product (42%, Scheme 4c). Moreover,
as an additional evidence of the carbon-based radical gener-
ation upon HAT, the transiently generated ketyl radical was
intercepted by forming its galvinoxyl adduct. The isolated
adduct was authenticated by high-resolution MS at 542.3791
amu (Figure S11). Significantly, while TPQ operates in the 2eÀ
pathway via addition–elimination or transamination, this PA-
catalyzed dehydrogenation is distinctly regulated by 1eÀ
processes.[3a,15]
Scheme 5. PA-catalyzed oxidative synthesis of substituted quinolines. Reac-
tion conditions: PA (5 mol%, with respect to alcohol), 2-aminobenzyl alcohol
(1 mmol), secondary alcohols (1 mmol ), KOtBu (40 mol%), toluene (2 mL),
°
blue light, 70 C, 8 h.
ChemSusChem 2020, 13, 1–7
4
© 2020 Wiley-VCH GmbH
��
These are not the final page numbers!