M. Li, J. Li, B. Guo et al.
Journal of Catalysis 399 (2021) 111–120
the probability of electron transfer between photocatalyst and sub-
strate, which potentially drives the reaction easier to occur [20].
As the redox potentials of ground state and excited state of a
photocatalyst determine whether a photocatalytic reaction is actu-
ally able to be carried out or not, and which photocatalytic route
the reaction will take place, through either the oxidative or reduc-
tive quenching cycles [20], we next did cyclic voltammetry analy-
3. Results and discussion
3.1. Structural modifications of CP
To develop a novel organic photocatalyst with great photophys-
ical properties, natural product CP was structurally modified with
diverse substitutions. CP-methoxy and CP-acetate were synthe-
sized through the etherification [40] and esterification of hydroxyls
[41,42]. The seven-membered ring opening product pre-CP was
obtained by microbial fermentation [43]. Although there were a
few changes in their photophysical properties (Fig. S1), no signifi-
cant change in the photocatalytic activity was observed (see Sup-
porting Information). To our delight, more than 250-folds
increase of the fluorescence intensity was observed along with
the structural modification from perylenequinonoid core to pery-
lene phenol (Fig. S1), which needed further acetylation with acetic
anhydride because of its instability. We reasoned that this phe-
nomenon was probably induced by changes of photophysical and
electrochemical properties of this CP derivative, thus, prompting
us to unravel its molecular mechanism and then investigate
whether this version of CP derivative can function as a novel pho-
tocatalyst with higher photocatalytic activity than CP. In this study,
other modifications in the same manner using different groups
including donor group (methoxy, TMERCP) and acceptor group
(isobutyrate ester, HiBRCP) were further synthesized. However,
we did not observe obvious increase of fluorescence intensity no
matter what substituent was modified compared with HARCP
(Fig. S2). Considering the low yields and the separation difficulties
of HiBRCP and TMERCP, therefore, we here only focus on the most
easily realized acetylation product HARCP, which was obtained in
96% yield, to systematically study its photophysical properties
and photocatalytic activities.
sis to obtain the redox potentials of HARCP (Fig. 2c and S7).
red
Compared with CP (E (CP/CP ꢀꢁ) = -0.46 V vs SCE) [33], the reduc-
1=2
tion potential of HARCP was dramatically improved through this
red
structural modification (E (HARCP/HARCP ꢀꢁ) = -1.43 V vs SCE).
1=2
The reason for the above improvement are the simultaneous
changes of HOMO energy level (from ꢁ5.77 eV to ꢁ5.49 eV) and
LUMO energy level (from ꢁ3.24 eV to ꢁ2.51 eV) based on the
DFT theoretical calculations of frontier molecular orbitals for CP
and HARCP, respectively (Fig. 2d, Table S2). Especially, the signifi-
cant change of LUMO energy level reduced the electron affinity of
the molecule, which made electrons at infinite distance difficult to
fill in the LUMO orbital, and then results in the reduction potential
of HARCP toward a more negative direction [53,54]. Based on the
calculated energy levels of the first singlet excited S1 state of CP
and HARCP by TD-DFT (see Supporting Information), their redox
potentials of the excited states could be calculated according to
‘‘Gibbs Energy of Photoinduced Electron Transfer” [55] as described
in Fig. 2d (The calculation process was described at Supporting
Information). We then speculated that the remarkable improve-
ment of the ground redox potentials of HARCP made it superior
in more challenging photoredox reactions as a novel and versatile
photocatalyst than CP. Therefore, we next selected mechanistically
distinct and challenging reactions with aryl halides as substrates to
evaluate the photocatalytic activity of HARCP.
3.3. Applications of HARCP on photoreduction of aryl halides
Previously, we have shown that unmodified CP has the photo-
catalytic ability to perform C–C bond-forming arylation through
the consecutive PET process to overcome the photocatalytic limita-
tion of CP for photoreductive activation of aryl halides (Fig. 3a)
[29], but the stability of CP radical anion was severely affected
by solvent polarity and hydrogen-bond donor, which greatly lim-
ited its wide applications of this consecutive PET in photocatalysis
(Fig. 3a). As a fact, the photoreduction of aryl halides was of low
3.2. Photophysical and electrochemical properties of HARCP
Generally, to design or customize a novel photocatalyst, a set of
photophysical properties including the lowest-energy absorption
maximum kmax, the lifetime of fluorescence sf, the fluorescence
quantum yield Uf, and electrochemical properties (redox poten-
tials) should be carefully considered as they determine its photo-
catalytic reactivity in photoinduced electron transfer (PET)
process [20]. As a visible-light photocatalyst, the prerequisite is
to have a strong absorption capacity in the UV–Vis region, thus
we first did the measurement of kmax of HARCP. It showed that kmax
of HARCP was centered at 455 nm in CH3CN (Fig. 2a, red). Accord-
ing to the UV–Vis absorption spectra, its excellent photon absorp-
tion ability of HARCP in the UV–Vis range and its photostability
(Fig. S3 and S4) provide the prerequisite as a good organic photo-
catalyst even if a certain blue shift was observed, compared with
CP (kmax = 476 nm, Fig. 2a, black).
Since we observed a more than 250-folds increase of the fluo-
rescence intensity of the emission maxima (kem) of HARCP, cen-
tered at 503 nm in CH3CN (Fig. 2b, red) with a blue shift
compared with CP (kem = 605 nm, Fig. 2b, black), it prompted us
to study its photophysical properties to reveal its molecular mech-
anism, which could give an explanation of its great improvements
in photocatalytic activity [20]. We then measured sf and Uf for CP
and HARCP, respectively. It shows that the lifetime of fluorescence
yield when CP was used as the photocatalyst (Table 1).
red
1=2
Since the reduction potential of HARCP was measured as E
(HARCP/HARCP ꢀꢁ) = ꢁ1.43 V vs SCE, which reached or exceeded
the reduction potentials of aromatic halides [56,57], it means that
HARCP could activate reductive cleavage of ArꢁX of aromatic
halides to generate the key aryl radical intermediates (Fig. 3b)
without the requirement of the consecutive PET process like CP.
Therefore, photoreduction of aryl halides was selected to test the
photocatalytic activity of HARCP with methyl 4-bromobenzoate
(1a) as a model reaction under visible light conditions (Fig. S8).
After a comprehensive optimization of all reaction parameters
including photocatalysts (Table S3), electron donors (Table S4), sol-
vents (Table S5) and light sources (Table S6), aryl bromide 1a was
almost completely converted to reductive product 2a in the pres-
ence of 10 mol% HARCP, 8 equiv DIPEA, upon irradiation with
23 W CFL for 40 h in acetone (Table 1, entry 1). HARCP shows out-
standing photocatalytic activity compared with CP which has low
photocatalytic effect (Table 1, entry 2). CP-methoxy, CP-acetate
and pre-CP were also tried, but with low photocatalytic activities
similar to CP (Table S3, entries 1–4). HiBRCP and TMERCP, prepared
through the same strategy as HARCP, had the similar photocat-
alytic effect (Table S3, entries 5–6), which confirmed that this type
of structural modification for CP did bring great improvements of
for HARCP is sf = 4 ns (Fig. S5) and the fluorescence quantum yield
Uf is up to 58% (Table S1, Fig. S6), while the corresponding sf and
Uf for CP were 1 ns [33] and 0.33% (Table S1, Fig. S6), respectively.
With the prolongation of fluorescence lifetime from 1 ns of CP to
4 ns of HARCP and the fluorescence quantum yield Uf of HACRP
with 175-folds increase, it indicates that HARCP greatly enhances
114