H. Zhang, T. Guo, M. Wu et al.
Tetrahedron Letters 67 (2021) 152878
time was extended, or performed under an oxygen atmosphere, the
yield was still not significantly improved, we speculated that there
should be a steric hindrance during the HAT process. To broaden
the substrates spectrum, we then investigate the oxidation of 1-
naphthalenemethanol 1j and cyclic benzyl alcohol. To our delight,
the rigid benzyl alcohol analogues with the steric hindrance of
benzylic position did not hamper the reaction (1k-n), and the
desired ketone products were isolated in good yields. Compared
with benzhydryl, a rigid structure was conducive to the reaction.
Subsequently, benzyl alcohols with electron-withdrawing groups
were tested. 4-Bromobenzyl alcohol 1o and 2-chlorobenzyl alcohol
1
p were converted into products with 70% and 86% yields respec-
tively, and the latter showed good reactivity [13]. Other substrates
with an electron withdrawing substitute (1q-t) were found to be
over-oxidized under standard conditions. If we removed Bu
4 2
NH -
PO , meanwhile extended the reaction time, desired aldehydes
4
could be obtained with moderate yields. Though this result was
unexpected, a clear mechanism was still under investigation.
Scheme 1. Oxidation of alcohols to carbonyl compounds.
Furthermore, heteroaromatic alcohol analogue 1u and
a-hydroxy
solvent of MeCN/DMSO was chose in our reaction. Replacing 4-
ClOBzBu N with tetrabutylammonium phosphate (Bu NH PO ),
ester analogue 1v could also be converted to the desired
aldehyde smoothly. What is striking about the results in this table
were 2w and 2x. The excellent selectivity was achieved on the
benzylic position only and the aliphatic alcohols were not
disturbed [17].
4
4
2
4
no significant improvement was observed (Table 1, Entry 6). Fortu-
nately, when 4 Å molecular sieves were added to the system, the
yield increased sharply from 50% to 88% (Table 1, Entry 7), we con-
sidered it improved the reaction efficiency by removing water and
boosting the decomposition of hydrogen peroxide described by
After substrate spectrum evaluation, control experiments were
carried out to investigate the actual rule of the reaction factors and
investigate the reaction mechanism (Table 3). Taking benzyl alco-
hol 1a as substrate, there was no formation of the product in the
absence of light (Table 3, Entry 2), photocatalyst (Table 3, Entry
4), or under Ar atmosphere (Table 3, Entry 3). Otherwise, benzalde-
hyde could be detected without quinuclidine, which indicated the
redox cycle was accelerated by quinuclidine (Table 3, Entry 5).
Then the effect of different quenchers on the reaction was exam-
ined. When butylated hydroxytoluene (BHT) (3 eq) was added into
the mixture of reactants as a radical scavenger, the yield dropped
sharply to 10% (Table 3, Entry 6). Using sodium azide as a singlet
oxygen scavenger, the monitored yield has also dropped dramati-
cally (Table 3, Entry 7). Meanwhile, benzoquinone as a superoxide
radical anion scavenger also inhibited the reaction (Table 3, Entry
2
Cibulka [10b]. Meanwhile, O (balloon) atmosphere also revealed
favorable yield (Table 1, Entry 8), but the air atmosphere was
eligible for reaction convenience in subsequent experiments.
With these optimized conditions in hand, we then explored the
substrate scope (Table 2). Various methoxy substituted benzyl
alcohols (1b-e) were firstly tested under the optimized reaction
conditions and could be oxidized with high yields up to 90%. Inter-
estingly, ortho-substitution in the aromatic ring did not hamper
the reaction process, which displayed distinct results compared
with previous work [13]. The benzyl alcohol substituted with an
amino group (1f, 1g) was then tested which has never been tried
under photocatalyzed oxidation. A time prolong was needed for
4
-(Boc-amino) benzyl alcohol 1f to get the product with 82% yield,
while 2-(Boc-amino) benzyl alcohol 1g bearing steric hindrance
2
8). Furthermore, the addition of CuCl hindered the process of reac-
reacted moderately under standard conditions with the yield of
tion (Table 3, Entry 9). The above results indicated there was a rad-
ical process involving single electron transfer and hydrogen atom
transfer in this photoredox system.
7
5%. When phenyl substituted benzyl alcohol was tested, 4-
biphenylmethanol 1h also showed excellent reactivity. On the con-
trary, benzhydryl 1i had a slightly decreased reactivity, the desired
benzophenone 2i was obtained with a yield of 66%. Even if reaction
To further investigate the reaction mechanism, we carried out
0
the oxidation of deuterium-labelled 1-indanol 1k under standard
Table 1
a
Optimization of the reaction conditions.
Entry
Solventb
Additivesc
Atmosphere
Yieldd (%)
1
2
3
4
5
6
7
8
MeCN/DMSO
MeCN/DMSO
MeCN
DMF
DMSO
MeCN/DMSO
MeCN/DMSO
MeCN/DMSO
w/o
Air
Air
Air
Air
Air
Air
Air
44
31
30
39
50
51
88
95
4-ClOBzBu
4-ClOBzBu
4-ClOBzBu
4-ClOBzBu
4
4
4
4
N
N
N
N
Bu
4
Bu
4
Bu
4
NH
NH
NH
2
2
2
PO
PO
PO
4
4
4
+ MS
+ MS
O
2
a
Reaction condition: benzyl alcohol (0.2 mmol), 4CzlPN (3 mol%), quinuclidine (10 mol%), solvent (1 mL), Air (balloon) or O
2
(balloon), room temperature (rt), under 8 W
blue LEDs irradiation. For the detailed optimization, see Table S1 and Table S2.
b
MeCN/DMSO: MeCN (1 mL) + DMSO (0.1 mL).
c
Bu
4
NH
2 4 4
PO (25 mol%), 4-ClOBzBu N (25 mol%), 200 mg 4 Å Molecular sieves were activated under vacuum at 200 °C for 30 min prior to use.
d
Yield determined by GC using n-dodecane as an internal standard.
2