Ketones as Molecular Co-catalysts for Boosting Exciton-Based Photocatalytic Molecular Oxygen Activation

Article abstract of DOI:10.1002/anie.202003042

Excitonic processes in semiconductors open up the possibility for pursuing photocatalytic organic synthesis. However, the insufficient spin relaxation and robust nonradiative decays in semiconductors place restrictions on both quantum yield and selectivit

Full text of DOI:10.1002/anie.202003042

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Ketones as Molecular Co-catalysts for Boosting Exciton-Involved
Photocatalytic Molecular Oxygen Activation
Authors: Hui Wang, Shenlong Jiang, Wenxiu Liu, Xiaodong Zhang,
Qun Zhang, Yi Luo, and Yi Xie
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202003042
Angew. Chem. 10.1002/ange.202003042
Link to VoR: http://dx.doi.org/10.1002/anie.202003042
http://dx.doi.org/10.1002/ange.202003042

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
Ketones as Molecular Co-catalysts for Boosting Exciton-Involved
Photocatalytic Molecular Oxygen Activation
Hui Wang,+[a,b] Shenlong Jiang,+[a] Wenxiu Liu,[a] Xiaodong Zhang,* [a,b] Qun Zhang,*[a] Yi Luo, [a] and Yi
Xie*[a,b]
[a]
[b]
Dr. H. Wang,+ Dr. S. Jiang,+ W. Liu, Prof. X. Zhang, Prof. Q. Zhang, Prof. Y. Luo, Prof. Y. Xie
Hefei National Laboratory for Physical Sciences at the Microscale, CAS Center for Excellence in Nanoscience, Synergetic Innovation Center of Quantum
Information and Quantum Physics
University of Science and Technology of China
Hefei, Anhui 230026, P. R. China
Dr. H. Wang, Prof. X. Zhang, Prof. Y. Xie
Institute of Energy
Hefei Comprehensive National Science Center
Hefei, Anhui 230031, P. R. China
E-mail: zhxid@ustc.edu.cn
qunzh@ustc.edu.cn
yxie@ustc.edu.cn
These authors contributed equally to this work.
[+]
Supporting information for this article is given via a link at the end of the document.
Abstract: Excitonic processes of semiconductors open up the
possibility for pursuing photocatalytic organic syntheses. However,
the insufficient spin relaxation and robust nonradiative decays in
semiconductors place restrictions on both quantum yield and
selectivity of these reactions. Herein, by taking polymeric carbon
nitride (PCN)/acetone as a prototypical system, we propose that
extrinsic aliphatic ketones can serve as molecular co-catalysts for
synergistically promoting spin-flip transition and suppressing non-
radiative energy losses. Spectroscopic investigations indicate that
hot excitons in PCN can be transferred to ketones, while triplet
excitons in ketones can be transferred to PCN reversely. As such,
exciton–exciton annihilation (EEA) arising from the correlations
between excitons represents a crucial process that is adverse to
photocatalytic performance. Traditionally, EEA gives rise to the
formation of high-lying excitons (i.e., hot excitons) via the
inelastic collisions of low-lying excitons. The resulting hot
excitons then cool rapidly to low-lying excitonic states through
internal conversion, which leads to undesired energy losses in
the systems.[10,11] Therefore, searching effective strategies for
regulating these excitonic processes in the systems is urgently
needed.
To date tremendous efforts have been devoted to excitonic
regulation in semiconductors.[12−19] As for the regulation of spin-
state relaxation, much attention has been paid to the
manipulation of spin–orbit coupling and singlet–triplet energy
gap, which results in optimized strategies including structural
modification and heavy-atom incorporation for promoting ISC
efficiency.[12−15] As for the regulation of nonradiative decays, the
factors like dielectric environment and exciton density/diffusion
are considered to be critical to EEA rate, which leaves optimized
strategies such as phase engineering, substrate control, and
excitation regulation.[16−19] However, in terms of photocatalysis,
the applicability of the above strategies is quite limited:
regardless of their feasibility, these strategies interfere with the
investigation of the intrinsic properties of catalysts, and the
accompanying changes in such aspects as charge-carrier
mobility, surface area, stability, and band gap would complicate
the result interpretation. Most importantly, to the best of our
knowledge, the synergistically controllable optimization of spin
relaxation and nonradiative decays has never been achieved.
Recently the coupling between semiconductor and molecule
has been demonstrated to be favorable for achieving high-
efficiency exciton-involved energy utilization.[20−23] Considering
that electronic structures of organic molecules vary substantially
according to types of functional groups, it is rational to regulate
excitonic aspect of semiconductor-based photocatalysts by
introducing molecular co-catalysts with appropriate energy levels
and configurations. Here we focused our attention on a common
category of compounds, aliphatic ketones. Due to the presence
of carbonyl group, ketones tend to possess excited states
PCN/ketone
systems
exhibit
considerable
triplet-exciton
accumulation and extended visible-light response, leading to
excellent performance in exciton-involved photocatalysis like singlet
oxygen generation. This work provides a fundamental understanding
of energy harvesting in semiconductor/molecule systems, and paves
the way for optimizing exciton-involved photocatalysis via molecular
co-catalyst design.
Introduction
Semiconductor-based photocatalytic organic synthesis has
recently attracted tremendous interest, owing to its great
potential in converting solar energy to chemical energy.[1−7] For
the activation of molecular intermediates or substrates, excitonic
aspect in photoexcited semiconductors exhibits incomparable
advantage, by virtue of the feasible spin-involved energy
transfer.[7−9] However, the insufficient spin relaxation and the
robust nonradiative decays in semiconductors usually lead to
deficient excitonic spin-state control and fast photoinduced
species depopulation. Such features set limitations to not only
quantum efficiency but also selectivity of photocatalytic reactions.
For instance, owing to the unique spin configuration, triplet
excitons are usually involved in selective activation processes.[8,9]
Unfortunately, the generally low intersystem crossing (ISC)
efficiencies lead to intrinsically faint triplet exciton generation,
which clearly exerts
a negative effect on the relevant
photocatalytic activities. In respect of the nonradiative decays,
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
exhibiting mixed ππ* and nπ* configurations, which can result in
photocatalytic molecular oxygen activation of PCN and
high ISC yields and hence efficient triplet exciton generation.[24,25] PCN/acetone systems, using methyl orange (MO) as the probe
Moreover, such mixed configurations could establish
considerable coupling between high-lying excitonic states of
semiconductors and discrete excited states of ketones. Such a
coupling enables potential extraction of hot excitons competing
with energy loss through internal conversion. Besides, the high
stabilities of aliphatic ketones further promise their photocatalytic
applications. All these advantages hint that aliphatic ketones can
serve as molecular co-catalysts for synergistically optimizing the
spin relaxation and nonradiative decay processes in
semiconductor-based photocatalysts, as depicted in Scheme 1.
molecule. As a typical azo dye, MO can be oxidized by reactive
oxygen species (ROS) or photoinduced holes, which can result
in absorbance reduction in 350−520 nm. As shown in Figure 1a,
the absorbance reduction clearly indicated the consumption of
MO molecules by both PCN and PCN/acetone, while the latter
exhibited an obvious promotion in the consumption rate. The
absorbance evolution at 417 nm was used to estimate the MO
consumption. The negligible absorbance changes under blank-
and acetone-condition (Figure 1b) ruled out the existence of
direct interactions between acetone and MO molecules. The
distinctly different absorbance evolution observed in a set of
control tests under Ar- and O2-atmosphere (Figure 1c)
demonstrated that ROS is mainly responsible for the MO
consumption. As compared with the situations under oxygen-
containing conditions (that is, air and O2), the MO degradation
under Ar atmosphere showed
a
notable suppression.
Considering the non-existence of 1O2 or superoxide radical (O2•−),
we attribute the consumption of probe molecule under Ar
atmosphere to the photoinduced-hole-involved MO degradation
mechanism (see details in Supporting Note 2 and Figure S2).
The addition of carotene (a well-known scavenger of 1O2)
apparently slowed down the degradation of MO molecules,
revealing that the dominant reactive species in the PCN/acetone
case should be 1O2.
Scheme 1. Schematic illustration of acetone as co-catalyst for excitonic
regulation in PCN, where HET and TET denote hot exciton transfer and triplet
exciton transfer, respectively.
Herein, we proposed that ketones can serve as molecular
co-catalysts to optimize exciton-involved photocatalytic reactions.
By taking polymeric carbon nitride (PCN)/acetone as
a
prototypical system, we demonstrated that extrinsic acetone can
extract exciton–exciton annihilation (EEA)-induced hot excitons
from PCN and donate triplet excitons reversely. Such processes
lead to synergistic optimization of spin relaxation and
nonradiative decays in the system, thereby leaving highly-
efficiency triplet exciton harvesting and promoted visible light
response. Not only that, acetone also executes quite limited
impact on the intrinsic charge-carrier-involved photoexcitation
processes in PCN, endowing the system with selective excitonic
regulation. Benefiting from these, the PCN/ketone systems
exhibited
excellent
performance
in
exciton-involved
Figure 1. Photocatalytic molecular oxygen activation. (a) Time-dependent
absorption spectra of MO oxidation with PCN and PCN/acetone in air. (b)
Evolutions of the 417-nm absorbance of MO solution with different catalysts.
(c) Evolutions of the 417-nm absorbance of MO solution with PCN/acetone
under different atmospheres or in the presence of carotene. (d) ESR-trapping
tests of different samples in the presence of TEMP.
photocatalytic singlet oxygen (1O2) generation. The results not
only establish an in-depth understanding of exciton-involved
photocatalysis, but also present
strategy for gaining efficient solar energy utilization via molecular
co-catalyst design.
a synergistic optimization
Furthermore, electron spin resonance (ESR)-trapping tests
provided evidence for the photocatalytic ROS generation. In
detail, 2,2,6,6-tetramethylpiperidine (TEMP) was adopted as the
Results and Discussion
1
trapping agent for detecting O2. As displayed in Figure 1d, the
typical 1:1:1 triplet signals originating from 2,2,6,6-
PCN samples used in this work were obtained through the
typical thermal condensation of melamine (see details in
Supporting Information). The structural properties and chemical
compositions of the samples were confirmed according to the
XRD, FT-IR, and XPS characterizations (Supporting Note 1 and
Figure S1). To examine our proposal, we first investigated the
tetramethylpiperidine-N-oxyl
verified
photocatalytic
1O2
generation in PCN. Notably, PCN/acetone exhibited significantly
1
promoted O2 production, in accordance with the above results.
Negligible 1O2 was detected in the acetone case because of the
ill-matched excitonic energy levels between acetone and
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
molecular oxygen. This phenomenon confirms that the PCN
component is the main photocatalyst for 1O2 generation in the
PCN/acetone system. In addition, 5,5-dimethyl-1-pyrroline-N-
oxide (DMPO) was used as the trapping agent for detecting
excitation spectra of PCN and PCN/acetone monitored at 445
nm. As shown in Figure S6, the emission intensity of
PCN/acetone exhibited obvious reduction in the region below
325 nm. This threshold value coincided with the optical
absorption edge of acetone that corresponds to its lowest singlet
excited state (S1).[26] The slight reduction in the region above
325 nm is most likely due to the processes associated with EEA
in the PCN/acetone system, which endows the system with
potential visible light response (further discussion on this issue
will be given later). Besides, for both the emission and excitation
spectra, PCN and PCN/acetone systems exhibit similar spectral
profiles, which clearly demonstrate that the introduction of
acetone does not affect the intrinsic excitonic bands of the
polymeric semiconductor. To further understand the involved
excitonic processes, we carried out the time-resolved
fluorescence measurements by monitoring emission at 410 nm
(Figure 2b and Table S1). Here, a bi-exponential function was
employed to fit the decay in the acetone case (given the
potential fluorophore–fluorophore interactions), yielding an
average lifetime of ~2.25 ns in accordance with the previous
reports.[26,30] As for bare PCN, the bi-exponential fitting yielded
an average lifetime of ~0.66 ns. Notably, the PCN/acetone case
exhibited a very similar decay (~0.68 ns) to the PCN case (i.e.,
in terms of both time constants and statistical weights, see Table
S1). The absence of acetone’s kinetic feature suggested the
occurrence of an effective energy transfer from excited
molecular acetone to PCN.
•−
photoinduced-electron-related O2 generation. As shown in
Figure S3, PCN/acetone exhibited a slight increase (~10%) in
O2•− generation as compared to PCN. This can be related to the
increased yield of free carriers due to the increased long-lived
exciton concentration (via a route of exciton dissociation).
Obviously, such a slight increase cannot be responsible for the
distinct difference in MO degradation. Moreover, their similar
•−
O2 production indicates that acetone executes quite limited
impact on the intrinsic charge-carrier-involved photoexcitation
processes in PCN. The above results verified the promoted
exciton-involved photocatalysis in the PCN/acetone system as
expected.
Considering that the quantum yield of ISC in acetone is
nearly unity,[26] we carried out phosphorescence measurements
to further investigate the triplet exciton-related excitonic aspects
in the system. The phosphorescence spectrum of PCN (Figure
2c) obtained at a delay time of 10 μs exhibited a similar spectral
profile to that of the prompt fluorescence (Figure S7). This
phenomenon indicates the EEA-induced delayed fluorescence
feature of the long-lived emission, in accordance with the
previous report.[11] Strikingly, PCN/acetone exhibited a similar
emission to PCN. The quenching of the phosphorescence
emission from the acetone component clearly indicates the
efficient transfer of triplet excitons from excited acetone to PCN.
Besides, the nearly identical profiles of the phosphorescence
spectra of PCN and PCN/acetone (Figure 2c) suggested that
acetone possesses a negligible impact on the distribution of
triplet excitonic energy levels of PCN. Time-resolved
phosphorescence spectra (monitored at 485 nm, see Figure 2d
and Figure S8) also revealed a similar decay tendency between
the PCN and PCN/acetone cases. That is, they both exhibited
typical multi-stage traces, arising as a result of triplet−triplet
annihilation in the polymeric semiconductor.[11] The mono-
molecular exponential decays after 60 μs can be ascribed to the
suppression of triplet−triplet annihilation at low concentrations of
triplet excitons, with triplet lifetimes of ~125.2 and 126.6 μs for
the PCN and PCN/acetone cases, respectively. Such an
observation further confirms that the addition of acetone does
not influence the intrinsic radiative relaxation of triplet excitons in
PCN. Nevertheless, a subtle difference between the two decay
traces was found in the initial stage. As magnified in the inset of
Figure 2d, PCN/acetone exhibited a faster decay than PCN: (8.2
± 0.1) × 104 s−1 vs. (7.4 ± 0.1) × 104 s−1, suggesting the
promoted triplet–triplet annihilation rate (with a roughly 11%
acceleration) in the former case. Given that molecular acetone
does not influence both kinetics and thermodynamics of excitons
Figure 2. Photoluminescence measurements of different samples in
acetonitrile at room temperature (excitation at 300 nm). (a) Prompt
fluorescence spectra. (b) Time-resolved fluorescence spectra monitored at
410 nm. (c) Phosphorescence spectra (delay time: 10 μs). (d) Time-resolved
phosphorescence spectra monitored at 485 nm; Inset: magnified view of the
initial-stage decay traces.
To gain insights into the promoted photocatalytic 1O2, we
interrogated the photoexcitation processes of PCN and
PCN/acetone
systems
by
using
photoluminescence
spectroscopy. According to the UV−vis spectra of PCN and
acetone (Figure S4), the excitation wavelength was selected at
300 nm, ensuring excitation of both samples. Figure 2a displays
representative prompt fluorescence spectra of PCN, acetone,
and PCN/acetone. As for PCN and acetone, the fluorescence
spectra peaking at 445 and 410 nm are related to the emissions
from the lowest singlet excitonic states and excimer,[26,27]
respectively. Strikingly, PCN/acetone exhibited an obvious
suppression in the 445-nm emission, rather than the
enhancement observed in other donor–acceptor systems.[28,29]
Given the low absorbance of the system, the suppression would
not originate from the tiny difference in their UV−vis spectra
(Figure S5). Moreover, the contribution from the acetone
component turned out to exhibit no significant variation, which
might be linked to its low quantum yield.[26] To gain insights into
such changes, we further measured the photoluminescence
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
in PCN, such a difference clearly confirms the increased
concentration of triplet excitons in PCN/acetone. Moreover, the
accelerated decay in the initial stage was found to be closely
related to the acetone concentration (Figure S9), which further
confirms the positive role of acetone in optimizing triplet exciton
generation in the system. According to the above discussion, it
is safe to conclude that the extrinsic addition of acetone can
significantly enhance the triplet exciton accumulation without
affecting the intrinsic excitonic states in PCN.
picosecond (~0.3 ps) internal conversion in PCN was revealed in
the initial build-up kinetics (Figure S10), on
a timescale
comparable to that of energy transfer processes.[37–39] Note that
the accurate timescale of hot-exciton transfer is hardly to be
distinguished, owing to the resolution limitation of our pump–
probe system (~100 fs). However, the PCN/acetone case
exhibited a less negative TA signal than the bare PCN case at
the initial 1-ps stage (Figure S10a), suggesting that
a
considerable amount of hot excitons transfer from PCN to
acetone at a sub-ps timescale. These hot excitons are most
likely to originate from EEA, owing to the insufficient photon
energy (360 nm) for directly generating hot excitons. These
observations suggested that the EEA-induced hot excitons can
be transferred from PCN to acetone, thereby opening a pathway
for triplet exciton harvesting beyond the direct photoexcitation of
acetone. Strikingly, 315-nm excitation turned out to bring on a
spectral evolution to the contrary. As shown in Figure 3c, the
addition of a tiny amount of acetone led to a pronounced
spectral variation below ~515 nm, i.e., more probe bleach
emerged for PCN/acetone than for bare PCN. This observation
demonstrated that there must be a process of energy transfer
from excited acetone to PCN, which results in remarkably
promoted accumulation of photoinduced species in the
PCN/acetone system. The more negative TA signal for
PCN/acetone over the initial 1-ps range under 315-nm excitation
(Figure S10b) also confirms the ultrafast energy transfer from
acetone to PCN. It is worth pointing out that under 315-nm
excitation the EEA-induced energy transfer from PCN to acetone
exists concomitantly with the energy transfer from acetone to
PCN, whereas the latter should prevail over the former. In line
with this argument, the observed kinetics (Figure 3d) revealed
prolonged relaxation in PCN/acetone relative to PCN, verifying
the effective injection of photoinduced species to the PCN
component of PCN/acetone. Therefore, the above TA results
and analyses provided confirmative evidences for the exciton-
involved energy transfer in the photoexcited PCN/acetone
system.
Figure 3. Ultrafast transient absorption (TA) spectra of PCN and PCN/acetone.
(a) Representative TA spectra taken at the 5-ps probe delay; (b) the
corresponding kinetic traces taken at 450 nm under 315-nm excitation. (c)
Representative TA spectra taken at the 5-ps probe delay; (d) the
corresponding kinetic traces taken at 450 nm under 360-nm excitation. The TA
signal (i.e., the absorbance change, ∆Abs.) is given in the unit of mOD (OD:
optical density).
By virtue of its high sensitivity in tracking the exciton kinetics
in semiconductors, femtosecond time-resolved transient
absorption (TA) spectroscopy was further employed to
interrogate relaxation kinetics of the systems.[11,31–36] In the
pump−probe configuration, the pump wavelengths were chosen
at 315 or 360 nm, suitable for exciting the singlet excited state of
acetone or not, respectively.[34] The WLC (410–550 nm) probe
monitored the TA spectral evolution, where the probe bleach
signals (Figure 3) are known to correlate with the neutral singlet
excitons of PCN.[35,36] Note that the TA signal of PCN/acetone
should originate from the PCN component, since no signal was
discernible in the whole probing region for acetone. Figure 3a
presents the TA spectra taken at a probe delay of 5 ps (pump at
360 nm), where similar spectral profiles confirms that the
additional acetone (1 vol%) does not influence the band
distributions in PCN. Given that 360-nm locates out of the
spectral range for effective excitation of acetone, the reduction in
probe bleach of PCN/acetone (relative to bare PCN) would
correlate with the EEA-induced energy transfer from PCN to
acetone, echoing to the variation in photoluminescence
excitation spectra above 325 nm (Figure S6). Such an energy
transfer was further evidenced by the subtle difference in
relaxation kinetics (Figure 3b), from which one can see a slight
acceleration of photoinduced species consumption in
PCN/acetone as compared to bare PCN. Moreover, a sub-
The finding of the positive role of acetone in optimizing
exciton-involved photocatalysis inspired us to interrogate the
feasibility of using various aliphatic ketones as co-catalysts to
PCN for promoting 1O2-involved reactions. Here, oxidative
coupling of amines to imines, an extensively studied reaction by
virtue of its chemical and biological significances, was selected
to assess the relevant photocatalytic performance.[40,41] It has
been reported that 1O2 can selectively convert amines into
imines via an aldimine-mediated mechanism, where
benzaldehyde formed by the hydrolysis of aldimine can react
with benzylamine to gives rise to the formation of imine. (Figure
S11).[41] Table 1 lists the conversion rates and selectivities of
oxidative coupling reaction under different conditions. In
particular, PCN/acetone showed a rate of 95% for benzylamine
conversion (entry 1), ~4.5-fold higher than that of PCN (entry 2),
and both cases exhibited high imine selectivity (≥ 90%). The
formation of imine for bare PCN is ascribed to feasible charge-
transfer-mediated pathway, where reactive species including
•−
photoinduced hole and O2 are crucial to the reaction (see
details in Supporting Note 3 and Figure S12). As for the
PCN/acetone case, the conversion rate and selectivity of the
reaction are closely related to the concentration of acetone
(Figure S13), where higher concentrations led to side or
peroxidation reactions. To elucidate the role of 1O2 in the system,
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
we further performed atmosphere- and scavenger-control tests.
As shown in entry 3, the absence of oxygen can greatly
suppress both selectivities and conversion rates for both PCN
and PCN/acetone, confirming that the oxidative coupling
reaction mainly undergoes a ROS-mediated route. Significant
reduction of conversion rate in the scavenger-control test (with
additional NaN3, an efficient scavenger of 1O2; entry 4) further
case coincided with the longer decay time of triplet acetone-d6
originating from the reduced stretching vibrations of C–D bonds
(as compared to C–H bonds).[25] The above results confirmed
the universality of using ketones as co-catalysts to PCN for
optimizing the triplet exciton-involved photocatalysis. .
1
Table 1. Aerobic oxidative coupling of benzylamine under different
verified the crucial role of O2 in this oxidative coupling process.
conditions.[a]
ESR-trapping tests also provide direct evidence to the crucial
role of 1O2 in the benzylamine conversion (Figure S14). Note
that benzylamine serving as a scavenger of photoinduced hole
can lead to exciton dissociation to some extent, which, however,
exhibits quite limited impact on the overall excitonic processes in
PCN under catalytic condition (see details in Supporting Note 4,
Table S2 and Figure S15). Besides, the potential impact of
bound benzylamine on the photoexcitation processes of acetone
has also been excluded (Supporting Note 5 and Figure S16).
The incomplete suppression of the reaction might be ascribed to
the concomitant O2•−-involved coupling route. Moreover, it is
quite interesting that the imine yield in the presence of NaN3 for
PCN/acetone (23%) is quite close to the imine yield for bare
ketone
Conv.[b] Select.[c] Conv.
PCN/ketone
Entry
ketone
Select.
1
2[d]
3[e]
4[f]
5
acetone
‐
3
89
‐
95
21
92
93
50
70
87
96
85
90
89
85
87
‐
acetone
trace
trace
82
85
27
33
69
67
77
75
51
acetone
4
5
4
3
3
5
6
6
2‐butanone
2‐pentanone
2‐hexanone
2‐heptanone
3‐heptanone
4‐heptanone
3‐pentanone
6
82
85
•−
7
PCN (20%), echoing to their similar O2 generation. Such a
8
23
feature further verifies that the extrinsic acetone does not
influence the charge-carrier process involved in photoexcited
PCN, but selectively optimizes the excitonic processes therein.
To go further, a set of control tests with different amines as
substrates were conducted, of which the conversion-rate
promotion in PCN/acetone relative to PCN confirmed the
generality of acetone as co-catalyst for optimizing the oxidative
coupling of amines (see details in Supporting Note 6 and Table
S3)
9
88
97
80
10
11
29
34
4‐Methyl‐2‐
pentanone
12
5
99
35
76
5‐Methyl‐2‐
hexanone
13
14
5
5
80
63
44
97
80
90
acetone‐d6
To step further, we selected a wide range of aliphatic
ketones to evaluate the scope of ketone-based co-catalysts.
Given that the energy levels and ISC rates are variable in
different ketones, it can be expected to achieve different co-
catalytic activities by using ketones with certain structural factors.
Firstly, the impact of carbon chain length was investigated. As
shown in entry 5–8, the addition of straight-chain aliphatic
ketones including 2-butanone, 2-pentanone, 2-hexanone, and 2-
heptanone turned out to promote the conversion rates
substantially, proving the feasibility of using ketones as
molecular co-catalysts. As compared to the acetone case, the
lower promotions in these cases can be linked to the lower ISC
rates intrinsic to these ketones. Besides, the addition amount of
ketones for aerobic oxidative coupling of benzylamine was
controlled at 10 μL, which means acetone with lowest molecular
weight but similar density lead to the highest molar ratio. For
ketones with the same carbon chain length, the position of
carbonyl group can bring forth great impact on the relevant
photophysical properties, thereby determining the efficiency of
triplet exciton-involved energy transfer between ketones and
PCN. Taking 2-, 3-, and 4-heptanones as examples (entry 8–10),
the corresponding promotions in conversion rate exhibited an
obvious progressive decrease. Similar comparison results were
also found between 2- and 3-pentanones (entry 6,11). This
phenomenon can be ascribed to the reduced ISC rates caused
by the restricted vibrational motions in ketones.[42] Analogously,
relatively small promotions were found in the cases with methyl
substitution (entry 12,13). Moreover, the isotope effect was
examined by comparing the acetone and acetone-d6 cases
(entry 14). The slightly higher conversion rate in the acetone-d6
[a] Reaction conditions: benzylamine (0.5 mmol), catalyst (ketone, 10 μL; or
PCN/ketone, 2 mg/10 μL), acetonitrile (2 mL), O2 (1 atm), 20 °C, xenon lamp,
3 h. [b] Determined by NMR analyses, using 1,1,2,2-tetrachloroethane as the
internal standard, conversion
= 1 – [amine]terminal/[amine]initial, mol%. [c]
Selectivity = yieldimine × 2/conversionamine, mol%. [d] Without the addition of any
ketone. [e] Ar atmosphere. [f] Additional 50 mg of NaN3.
Last but not least, we investigated the wavelength-dependent
photocatalytic oxidative coupling of benzylamine in order to look
deeper into the relevant photoexcitation processes. As shown in
Figure 4a, the wavelength-dependent conversion rates of the
PCN case were in good agreement with the spectral distribution
of the sample absorbance (dashed black line), which exhibited
an obvious cut-off wavelength at around 450 nm. Such a feature
indicated the low photocatalytic reactivity under visible light
illumination, which would be detrimental to the potential
applications. As compared to bare PCN, PCN/acetone exhibited
remarkable enhancement in conversion rates over a broad
region spanning from UV to visible, where the sharp change in
the UV could be caused by the photoabsorption of acetone and
PCN. Dramatically, the PCN/acetone system exhibited
a
considerable reactivity under visible light illumination,
attributable to the EEA-induced energy transfer from
photoexcited PCN to acetone (Figure 4b). Such processes can
be understood as follow: under visible light illumination, the
selective photoexcitation of the PCN component induces the
accumulation of excitons; the strong correlations between these
excitons lead to the generation of hot excitons (refer to En in the
scheme) via an EEA mechanism; then the generated hot
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
excitons are transferred to acetone, indirectly exciting the
acetone component in the system. This understanding is based
on the following two facts: (i) the robust EEA in PCN yields a
considerable amount of hot excitons with suitable energy level
matching that of acetone, and (ii) the sub-picosecond build-up
time (~0.3 ps, linked to the internal conversion of hot excitons)
enables feasible hot exciton-involved energy transfer from PCN
to acetone prior to the relaxation to the lowest excitonic states
(refer to E1 in the scheme). This finding confirmed the feasibility
of using ketones as co-catalysts to PCN in optimizing the visible-
light-driven, exciton-involved photocatalysis. It should be pointed
out that the hot-exciton (generated via direct excitation or EEA)
transfer from PCN to acetone also exists under UV-light (below
325 nm) excitation. However, the simultaneous but dominant
excitation of acetone enables energy transfer from acetone to
PCN, prevailing over the hot-exciton transfer from PCN to
acetone.
This work was supported by the National Key R&D Program of
China (2017YFA0207301, 2017YFA0303500, 2019YFA0210004,
2018YFA0208702), the National Natural Science Foundation of
China (21922509, 21905262, 21890754, 61705133, 21633007,
21803067, 91950207), the Youth Innovation Promotion
Association of CAS (2017493), the Young Elite Scientist
Sponsorship Program by CAST, the Key Research Program of
Frontier Sciences (QYZDY-SSW-SLH011), and the Anhui
Initiative in Quantum Information Technologies (AHY090200).
The authors thank Dr. Kuai Yu (College of Electronic Science
and Technology, Shenzhen University) for his helpful suggestion
on spectroscopic analyses.
Keywords: co-catalysts • exciton • energy transfer • singlet
oxygen
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
I. Ghosh, J. Khamrai, A. Savateev, N. Shlapakov, M. Antonietti, B.
König, Science 2019, 365, 360–366.
D. Friedmann, A. Hakki, H. Kim, W. Choic, D. Bahnemann, Green
Chem. 2016, 18, 5391–5411.
N. J. Hestand, R. V. Kazantsev, A. S. Weingarten, L. C. Palmer, S. I.
Stupp, F. C. Spano, J. Am. Chem. Soc. 2016, 138, 11762–11774.
M. Z. Rahman, C. W. Kwong, K. Davey, S. Z. Qiao, Energy Environ. Sci.
2016, 9, 709–728.
M. Z. Rahman, K. Davey, S. Qiao, J. Mater. Chem. A 2018, 6, 1305–
1322.
L. Zhang, J. Ran, S.-Z. Qiao, M. Jaroniec, Chem. Soc. Rev. 2019, 48,
5184–5206.
C. Mongin, S. Garakyaraghi, N. Razgoniaeva, M. Zamkov, F. N.
Castellano, Science 2016, 351, 369–372.
Figure 4. Photocatalytic mechanism investigations. (a) Wavelength-dependent
photocatalytic oxidative coupling of benzylamine, where the dashed black line
is the UV-vis spectrum of PCN. (b) Schematic illustration of hot-exciton-
involved energy transfer in PCN/acetone (mediated by EEA) under visible light
illumination, where GS and IC denote ground state and internal conversion,
respectively.
T. R. Blum, Z. D. Miller, D. M. Bates, I. A. Guzei, T. P. Yoon, Science
2016, 354, 1391–1395.
M. C. DeRosa, R. J. Crutchley, Coord. Chem. Rev. 2002, 233–234,
351–371.
[10] Y. Z. Ma, L. Valkunas, S. L. Dexheimer, S. M. Bachilo, G. R. Fleming,
Phys. Rev. Lett. 2005, 94, 157402.
[11] H. Wang, S. Jiang, S. Chen, X. Zhang, W. Shao, X. Sun, Z. Zhao, Q.
Zhang, Y. Luo, Y. Xie, Chem. Sci. 2017, 8, 4087–4092.
[12] H. Wang, S. Jiang, S. Chen, D. Li, X. Zhang, W. Shao, X. Sun, J. Xie, Z.
Zhao, Q. Zhang, Y. Tian, Y. Xie, Adv. Mater. 2016, 28, 6940–6945.
[13] O. Bolton, K. Lee, H. J. Kim, K. Y. Lin, J. Kim, Nat. Chem. 2011, 3,
205–210.
Conclusion
In conclusion, we here propose that ketones can serve as
molecular co-catalysts for synergistically promoting spin
relaxation and suppressing nonradiative energy losses in
semiconductor-based photocatalysts. By taking polymeric
carbon nitride (PCN)/acetone as a prototypical system, we
demonstrate that the extrinsic introduction of ketones can lead to
promoted triplet exciton accumulation and extended visible light
response. According to spectroscopic investigations, we
attribute such optimizations to two types of exciton-involved
energy transfers between PCN and ketones. That is, the energy
transfer of exciton–exciton annihilation-induced hot excitons
from PCN to acetone and the energy transfer of molecular
ketone triplets (whose formation is due to the high intersystem
crossing efficiency) to PCN. Owing to these features, the
PCN/ketone systems exhibit significantly optimized performance
in exciton-involved photocatalytic reactions like singlet oxygen
generation. This work brings an in-depth understanding of
exciton-involved photocatalysis, and presents a prototypical
model for pursing optimized solar energy utilization through
introducing extrinsic molecular co-catalysts.
[14] M. S. Kwon, D. Lee, S. Seo, J. Jung, J. Kim, Angew. Chem. Int. Ed.
2014, 53, 11177–11181.
[15] X. Ma, C. Xu, J. Wang, H. Tian, Angew. Chem. Int. Ed. 2018, 57,
10854–10858.
[16] Y. Hoshi, T. Kuroda, M. Okada, R. Moriya, S. Masubuchi, K. Watanabe,
T. Taniguchi, R. Kitaura, T. Machida, Phys. Rev. B: Condens. Matter
Mater. Phys. 2017, 95, 241403.
[17] Y. Lee, G. Ghimire, S. Roy, Y. Kim, C. Seo, A. K. Sood, J. I. Jang, J.
Kim, ACS Photonics 2018, 5, 2904–2911.
[18] R. Tempelaar, T. L. C. Jansen, J. Knoester, J. Phys. Chem. Lett. 2017,
8, 6113–6117.
[19] L. Yuan, L. Huang, Nanoscale 2015, 7, 7402–7408.
[20] M. Tabachnyk, B. Ehrler, S. Gelinas, M. L. Bohm, B. J. Walker, K. P.
Musselman, N. C. Greenham, R. H. Friend, A. Rao, Nat. Mater. 2014,
13, 1033–1038.
[21] R. Younts, H. S. Duan, B. Gautam, B. Saparov, J. Liu, C. Mongin, F. N.
Castellano, D. B. Mitzi, K. Gundogdu, Adv. Mater. 2017, 29, 1604278.
[22] W. L. Chan, J. R. Tritsch, X. Y. Zhu, J. Am. Chem. Soc. 2012, 134,
18295–18302.
[23] X. Li, Z. Huang, R. Zavala, M. L. Tang, J. Phys. Chem. Lett. 2016, 7,
1955–1959.
[24] M. P. O'Sullivan, A. C. Testa, J. Am. Chem. Soc. 1970, 92, 5842–5844.
[25] Haas, Y. Photobiol. Sci. 2004, 3, 6–16.
Acknowledgements
[26] R. F. Borkman, D. R. Kearns, J. Chem. Phys. 1966, 44, 945–949.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
[27] G. Porter, S. K. Dogra, R. O. Loutfy, S. E. Sugamori, R. W. Yip, J.
Chem. Soc., Faraday Trans. 1 1973, 69, 1462–1474.
[28] C. Zhao, X. Zhang, K. Li, S. Zhu, Z. Guo, L. Zhang, F. Wang, Q. Fei, S.
Luo, P. Shi, H. Tian, W. H. Zhu, J. Am. Chem. Soc. 2015, 137, 8490–
8498.
[29] T. Zhu, L. Yuan, Y. Zhao, M. Zhou, Y. Wan, J. Mei, L. Huang, Sci. Adv.
2018, 4, eaao3104.
[30] A. Lozano, B. Yip, R. K. Hanson, Exp. Fluids 1992, 13, 369–376.
[31] N. A. Anderson, T. Lian, Annu. Rev. Phys. Chem. 2005, 56, 491–519.
[32] X. Luo, R. Lai, Y. Li, Y. Han, G. Liang, X. Liu, T. Ding, J. Wang, K. Wu,
J. Am. Chem. Soc. 2019, 141, 4186–4190.
[33] S. Nah, B. Spokoyny, C. Stoumpos, C. M. M. Soe, M. Kanatzidis, E.
Harel, Nat. Photonics 2017, 11, 285–288.
[34] E. W.-G. Diau, C. Kötting, A. H. Zewail, ChemPhysChem 2001, 2, 273–
293.
[35] K. L. Corp, C. W. Schlenker, J. Am. Chem. Soc. 2017, 139, 7904–7912.
[36] Z. Chen, Q. Zhang, Y. Luo, Angew. Chem. Int. Ed. 2018, 57, 5320–
5324.
[37] C. E. Rowland, I. Fedin, H. Zhang, S. K. Gray, A. O. Govorov, D. V.
Talapin, R. D. Schaller, Nat. Mater. 2015, 14, 484–489.
[38] S. van Bezouw, D. H. Arias, R. Ihly, S. Cambre, A. J. Ferguson, J.
Campo, J. C. Johnson, J. Defillet, W. Wenseleers, J. L. Blackburn, ACS
Nano 2018, 12, 6881–6894.
[39] E. J. O'Reilly, A. Olaya-Castro, Nat. Commun. 2014, 5, 3012.
[40] G. Jiang, J. Chen, J. Huang, C. Che, Org. Lett. 2009, 11, 4568–4571.
[41] D. B. Ushakov, K. Gilmore, D. Kopetzki, D. T. McQuade, P. H.
Seeberger, Angew. Chem. Int. Ed. 2014, 53, 557–561.
[42] D. A. Hansen, J. Chem. Phys. 1975, 62, 183–189.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202003042
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Ketones can serve as molecular co-catalysts for synergistically suppressing nonradiative energy losses and promoting spin transition.
The energy transfer of exciton–exciton annihilation-induced hot excitons from PCN to acetone and the reverse energy transfer of
triplet excitons from ketones to PCN are responsible to the optimizations, thereby endowing PCN/ketone systems with excellent
performance in exciton-involved photocatalytic reactions.
8
This article is protected by copyright. All rights reserved.