ACS Catalysis
Research Article
a
Table 1. Screening for Optimal Reaction Conditions
photocatalytic activity in this hydrodecarboxylation trans-
2
BCN1000 than that of BCN (1337 vs 534 m /g) (Figure S4),
40
thus expanding more active sites to the reaction.
With the optimal reaction conditions in hand, the generality
of the hydrodecarboxylation protocol was investigated (Table
2
). Delightfully, a wide variety of substrates, including aromatic
acids, arylacetic acids, arylpropionic acids, and heterocyclic
acid, went through the reaction successfully with good to
excellent yields. Further investigation revealed that moderately
electron-rich aromatic acids were tolerated under the standard
reaction conditions (2a−2d). To arylacetic acids, the more
electron rich arylacetic acids gave a higher yield than those of
less electron-rich ones (2e−2h). 1-Naphthoic acid was
successfully converted into the corresponding 2-methylnaph-
thalene (2i) in 64% yield with 24 h irradiation. Secondary and
tertiary substituted carboxylic acids undergo the process
smoothly to the corresponding products in excellent yields
(2j and 2k). With 2,3-diphenylpropanoic acid and 3-(p-
tolyl)propionic acid as the substrates, a 63% yield of 2l was
isolated and a 50% yield of 2m was tested by gas
chromatography−mass spectrometry (GC-MS), respectively.
Furthermore, heterocyclic acids also underwent the trans-
formation smoothly in moderate yields (2n−2p).
entry
substrate
photocatalyst
BCN40
solvent
yield (%)
1
2
3
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1h
1h
1h
MeOH
MeOH
MeOH
EtOH
60 (55)
n.d.
n.d.
53
35
10
n.d.
n.d.
trace
n.d.
n.d.
85
b
BCN40
BCN40
BCN40
BCN40
BCN40
BCN40
BCN100
BCN500
BCN1000
BCN40
BCN1000
BCN1000
4
5
6
7
8
9
iPrOH
H O
2
DMSO
MeCN
MeOH
MeOH
MeOH
MeOH
10
11
12
13
14
c
c
c
H O
20
95 (93)
2
MeOH
a
Standard reaction conditions: 0.2 mmol of 1 and 30 mg of catalyst in
solvent (6 mL) for 48 h under a 15 W 420 nm LED irradiation at 40
C. Yields were determined by gas chromatography, and yields are
referred to isolated yield in parentheses. “n.d.” means “not detected”.
The system is able to produce deuterium products, which is
important in pharmaceutical chemicals by integrating a
distinguishing signal into a molecule without changing its
functionalization. The aromatic acid (1a), arylacetic acid (1e),
and arylpropionic acid (1l) were transferred into deuterium
products in moderate to good yields with excellent selectivity
°
b
c
No light. Twenty-four hours.
Other heterogeneous catalysts, such as hexagonal boron nitride
29a
29b
(
1
2a-D, 2e-D, and 2l-D). In particular, heteroaromatic substrate
o could be converted into the heavy drug 2o-D smoothly
(
h-BN, band gap 5.7 eV), CdS (band gap 2.4 eV), BiVO
4
29c 29d
(band gap 2.4 eV), TiO (rutile) (band gap 3.0 eV), and
2
29e
were examined, and no 2a was
with 86% deuterium incorporation in the final product. As
mentioned above, protic solvent (methanol) was essential for
the effectiveness of the photocatalytic hydrodecarboxylation.
Deuterium-labeling experiments led to deep insight into the
role of methanol as well as the reaction mechanism. When
replacing 3,5-dimethoxybenzoic acid with 3,5-dimethoxyben-
zoic acid-d in the standard conditions, 8% deuterium
indicating a proton exchange process (Figure S11). While
observed (Table S3, entries 1−5), demonstrating the photo-
catalytic effectiveness of BCN for decarboxylation reactions.
The catalyst contributes a lot in the process, not only to the
formation of carbon-centered radicals in the photooxidation
process but also to the oxidative quenching of the resulting
radical intermediate to close the catalytic cycle. On the basis of
26
previous reports, doping h-BN with different amounts of
carbon can modify the redox potential window and improve
the specific surface area, thus turning the insulator into a
visible-light-responsive semiconductor and providing open
reactive sites to improve reaction rates (see synthesis method
using CH OH-d , 92% deuterium incorporation was produced
3 4
in the reaction, which unambiguously confirmed that methanol
serves as the hydrogen atom source (Scheme 2, 1). Next, the
3 3
(
of carbon in the h-BN matrix. Also, elemental analysis by XPS
Table S1) reveals that the carbon content is gradually
S13, Scheme 2, 2). The results reveal that the hydrogen atom
comes from the hydroxyl group of the methanol. The KIE
(kinetic isotope effect) was determined to be 1.25, indicating
that the O−H bond-breaking event was not involved in the
rate-determining step (Scheme 2, 3).
increasing with increasing amounts of glucose. Therefore, the
performance of BCN40 and BCN (where x represents the
x
percentage weight content of glucose to boric acid precursor)
was compared, and results indicate that the photodecarbox-
ylation activity of BCN is better than that of other BCN in a
Motivated by the efficiency of BCN1000 to the photocatalytic
hydrodecarboxylation reaction, the system then was applied to
a decarboxylative C−C cross-coupling with cyanopyridines,
40
x
3
0
which were well-known electron acceptors. The C−C
coupling began using 4-methoxyphenylacetic acid (1q) and
4-cyanopyridine (3a) as the model substrates (Table S4). A
higher yield and selectivity were achieved using K CO as the
level of BCN is 2.04 V vs NHE (Figure S6), thermodynami-
4
0
•
+
while BCN1000 (VB level at 1.62 V vs NHE, Figure S6) are
2
3
exogenous base (yield of 3a is 23% without K CO and 78%
2
3
2
3
•
+
19a
for 1h (E(1h/1h ) = 1.55 V vs NHE, Figure S7c), BCN ,
carboxylate species (Table S4, entries 1 and 6).
optimized reaction conditions are 0.6 mmol of 1q, 0.2 mmol of
3a, 0.6 mmol of K CO , and 10 mg of BCN in 2 mL of
The
4
0
BCN1000, BCN100, and BCN500 are adequate to afford the
and 14; Table S3, entries 6 and 7). BCN1000 shows prior
2
3
1000
DMSO (Table S4, entry 6). The substrate scope was then
3
042
ACS Catal. 2021, 11, 3040−3047
Shi, Jiale
