Tracking Mechanistic Pathway of Photocatalytic CO2 Reaction at Ni Sites Using Operando, Time-Resolved Spectroscopy

Article abstract of DOI:10.1021/jacs.9b12443

Harvesting solar energy for catalytic conversion of CO₂ into valuable chemical fuels/feedstocks is an attractive yet challenging strategy to realize a sustainable carbon-cycle utilization. Homogeneous catalysts typically exhibit higher activity and selectivity as compared with heterogeneous counterparts, benefiting from their atomically dispersed catalytic sites and versatile coordination structures. However, it is still a black box how the coordination and electronic structures of catalysts dynamically evolve during the reaction, forming the bottleneck for understanding their reaction pathways. Herein, we demonstrate to track the mechanistic pathway of photocatalytic CO₂ reduction using a terpyridine nickel(II) complex as a catalyst model. Integrated with a typical homogeneous photosensitizer, the catalytic system offers a high selectivity of 99% for CO₂-to-CO conversion with turnover number and turnover frequency as high as 2.36 × 10⁷ and 385.6 s^-1, respectively. We employ operando and time-resolved X-ray absorption spectroscopy, in combination with other in situ spectroscopic techniques and theoretical computations, to track the intermediate species of Ni catalyst in the photocatalytic CO₂ reduction reaction for the first time. Taken together with the charge dynamics resolved by optical transient absorption spectroscopy, the investigation elucidates the full mechanistic reaction pathway including some key factors that have been often overlooked. This work opens the black box for CO₂ reduction in the system of homogeneous catalysts and provides key information for developing efficient catalysts toward artificial photosynthesis.

Full text of DOI:10.1021/jacs.9b12443

Subscriber access provided by University of Newcastle, Australia
Article
Tracking Mechanistic Pathway of Photocatalytic CO2 Reaction
at Ni Sites Using Operando, Time-Resolved Spectroscopy
Yangguang Hu, Fei Zhan, Qian Wang, Yujian Sun, Can Yu, Xuan Zhao, Hao Wang, Ran
Long, Guozhen Zhang, Chao Gao, Wenkai Zhang, Jun Jiang, Ye Tao, and Yujie Xiong
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b12443 • Publication Date (Web): 04 Mar 2020
Downloaded from pubs.acs.org on March 4, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Tracking Mechanistic Pathway of Photocatalytic CO2 Reaction
at Ni Sites Using Operando, Time-Resolved Spectroscopy
Yangguang Hu,†,¶ Fei Zhan,‡,§,¶ Qian Wang,†,¶ Yujian Sun,† Can Yu,‡,£ Xuan Zhao,‖ Hao Wang, ‡,§ Ran
Long,† Guozhen Zhang,† Chao Gao,†,* Wenkai Zhang,‖,* Jun Jiang,† Ye Tao,‡,* and Yujie Xiong†,,*
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
58
59
60
†Hefei National Laboratory for Physical Sciences at the Microscale, iChEM, School of Chemistry and Materials Science, and
National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei, Anhui 230026, China
‡Beijing Synchrotron Radiation Facility, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049,
China
§Chemistry and Chemical Engineering Guangdong Laboratory, Shantou, Guangdong 515031, China
£University of Chinese Academy of Sciences, Beijing 100049, China
‖Department of Physics, Beijing Normal University, Beijing 100875, China
Dalian National Laboratory for Clean Energy, Dalian, Liaoning 116023, China
KEYWORDS. Time-resolved spectroscopy; Operando spectroscopy; Photocatalysis; Mechanistic pathway; CO2 Reduction.
Supporting Information Placeholder
ABSTRACT: Harvesting solar energy for catalytic conversion of CO2 into valuable chemical fuels/feedstocks is an attractive yet
challenging strategy to realize a sustainable carbon-cycle utilization. Homogeneous catalysts typically exhibit higher activity and
selectivity as compared with heterogeneous counterparts, benefiting from their atomically dispersed catalytic sites and versatile
coordination structures. However, it is still a “black box” how the coordination and electronic structures of catalysts dynamically
evolve during the reaction, forming the bottleneck for understanding their reaction pathways. Herein, we demonstrate to track the
mechanistic pathway of photocatalytic CO2 reduction using terpyridine nickel(II) complex as a catalyst model. Integrated with a
typical homogeneous photosensitizer, the catalytic system offers a high selectivity of 99% for CO2-to-CO conversion with turnover
number and turnover frequency as high as 2.36107 and 385.6 s-1, respectively. We employ operando and time-resolved X-ray
absorption spectroscopy, in combination with other in situ spectroscopic techniques and theoretical computation, to track the
intermediate species of Ni catalyst in the photocatalytic CO2 reduction reaction for the first time. Taken together with the charge
dynamics resolved by optical transient absorption spectroscopy, the investigation elucidates the full mechanistic reaction pathway
including some key factors that have been often overlooked. This work opens the “black box” for CO2 reduction in the system of
homogeneous catalysts, and provides key information for developing efficient catalysts towards artificial photosynthesis.
coordinated to the metal center, the redox potential and
electronic structure of the catalyst can be readily modulated to
tune catalytic activity and selectivity. Moreover, the atomically
dispersed active sites in homogeneous catalysts are better
chemically defined, facilitating the determination of reaction
mechanisms.8 These advantages designate the homogeneous
system an ideal model to explore the reaction mechanisms of
photocatalytic CO2 reduction, which would provide valuable
information for rationally designing heterogeneous catalysts.
However, the dynamic evolution of coordination and electronic
structures for homogeneous catalysts during photocatalytic CO2
reduction closely associated with reaction pathways are still a
“black box”. Such a black box is rather complex than common
assumption in the sense that even the non-protonic solvent used
for homogeneous catalysts, typically CH3CN, may participate
in the reaction pathways and impact on the catalytic
performance of CO2 reduction. To gain key information on how
homogeneous catalysts work efficiently during the reaction, it
INTRODUCTION
Increasing concerns about energy crisis and excessive CO2
emission have urged scientists to exploit new clean energy to
address the irreplaceable demand for fossil fuels. The solar-
driven reduction of CO2 into chemical fuels/feedstocks is a
promising strategy to address the issues while highly efficient
and selective reduction of CO2 still remains a grand challenge
owing to the large bond energy of CO2 and the diversity of CO2
reduction products.1-3 In the past few decades, great research
efforts have been devoted to exploring efficient catalysts for
photocatalytic CO2 reduction, striving to achieve artificial
photosynthesis.4-7
Typically, homogeneous catalysts exhibit high activity and
selectivity towards artificial photosynthesis compared with
heterogeneous counterparts.8 The homogeneous catalysts are
commonly referred to as metal complexes, in which metal ions
are coordinated by organic ligands. Altering the ligands
ACS Paragon Plus Environment

Journal of the American Chemical Society
is imperative to track the mechanistic pathways using operando,
Page 2 of 8
Endstation (BL01B) in the National Synchrotron Radiation
Laboratory (NSRL) in Hefei, China. Each spectrum was
recorded by averaging 128 scans at a resolution of 2 cm-1. The
samples were mixed with KBr by a mortar and were held in a
custom-fabricated IR reaction chamber which was specifically
designed to examine highly scattered powder samples in diffuse
reflection mode. The chamber was sealed with two ZnSe
windows. Pure KBr powder was first measured to obtain a
background spectrum. During the in situ characterization, pure
CO2 (99.999%) or N2 (99.999%) saturated by CH3CN/H2O
vapor was continually introduced into the chamber. The spectra
were collected under dark condition or after a certain irradiation
time using a 5 W UV LED (UVEC-4II). The spectra were
obtained by subtracting the background from the spectra of
samples.
1
2
3
4
5
6
7
8
time-resolved and element-specific spectroscopic techniques.
9
Scheme 1. The photocatalytic system consisting of [Ni(tpy)2]2+
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
58
59
60
catalyst, [Ru(bpy)3]2+ photosensitizer, and BIH electron donor.
Herein, we demonstrate the elucidation of reaction pathways
using time-resolved X-ray absorption spectroscopy (TR-XAS)
in combination with other spectroscopic techniques and
theoretical computation. As a proof of concept, we employ
[Ni(tpy)2]2+ (tpy = 2,2':6',2"-terpyridine) complex as a
homogeneous catalysts model to open the “black box” in
photocatalytic CO2 reduction. Polypyridyl transition metal
complexes are one class of the most extensively studied
homogeneous catalysts for CO2 reduction, benefiting from the
redox active nature of polypyridine backbone. The reduced
polypyridyl rings can act as the reservoir of photo-induced
electrons, and in turn, the metal center mediates the transfer of
the electrons to CO2 molecules.9-11 Using the model catalyst of
[Ni(tpy)2]2+ complex, we achieve highly efficient and selective
reduction of CO2 to CO in a photocatalytic system with
[Ru(bpy)3]2+ (bpy = 2,2′-bipyridine) as a photosensitizer and
1,3-dimethyl-2-phenyl-2,3-dihydro-1Hbenzo[d]imidazole
(BIH) as an electron donor (Scheme 1). The high activity and
selectivity for CO2 reduction urge us to track the intermediate
species of [Ni(tpy)2]2+ catalyst during the reaction in terms of
dynamic coordination and electronic structures by employing
the TR-XAS and other in situ spectroscopic techniques. In
combination with charge dynamics information, this research
can elucidate the full mechanistic pathway involving the
catalyst, photosensitizer, electron donor and solvent.
Operando surface-enhanced Raman spectroscopy
(SERS). Operando SERS spectra were recorded using a WITec
alpha 300R confocal Raman microscope with
a 600
grooves/mm diffraction grating. As the excitation source, a 633
nm solid-state laser with an exact power of 20 mW was used
with attenuation by a filter. Calibration was performed using a
silicon wafer standard (520 cm−1). A CCD with 1650 × 200
pixels was used to collect the spectra. The CO2-saturated
sample solution and indium tin oxide (ITO, 2 × 10 mm) plated
with 50 nm Au nanocubes were sealed in the quartz cuvette (3
× 1 × 1 cm). During the in situ characterization, the spectra were
collected continuously under UV irradiation using a 5 W UV
LED (UVEC-4II).
Time-resolved optical transient absorption (OTA)
spectroscopy.
The nanosecond-to-microsecond transient
spectroscopy measurements were performed using
a
commercial nanosecond laser flash photolysis spectrometer
(LP980-KS, Edinburgh Instruments Ltd, Livingston, UK) at
room temperature. A pulsed Nd: YAG laser (Quanta-Ray
LAB190, Spectra-Physics) with an optical parametric oscillator
(PrimoScan ULD400, Spectra-Physics) for wavelength
conversion was used for excitation, and the probing light was
generated by a 150 W pulsed xenon arc-lamp at a repetition of
1 Hz and perpendicular to the pump light. Typical experiments
employed laser excitation at 440 nm (10 mJ/pulse, 3-6 ns fwhm).
The single-wavelength transient absorption was monitored at
the indicated wavelengths (0.3-2.0 nm band widths) with a
photomultiplier tube (PMT) detector (Hamamatsu R928),
which was connected to a Tektronix Model MDO3052 (100
MHz 1.25 GS/s) digital oscilloscope. The transient spectra were
recorded using an ICCD camera (DH320T, Andor) at chosen
indicated time delays. For all records, the probe background
was collected between two laser shots and subtracted from the
signal, and emission background was subtracted where relevant.
Transient absorption spectra and kinetics were acquired using
the L900 software (Edinburgh Instruments) and processed
using Origin 8.5 software. A quartz cell cuvette with a 1×1 cm
was used for measurements, and the samples were bubbled with
Ar before measurement.
EXPERIMENTAL SECTION
Time-resolved X-ray absorption spectroscopy (TR-XAS).
The TR-XAS measurements (the setup as shown in Figure S1)
were performed at the beamline 1W2B in the Beijing
Synchrotron Radiation Facility (BSRF), China, and the detail of
the setup has been described in a previous report.12 The high
repetition rate and ultra-short pulse fiber laser (Tangerine HP,
Amplitude systems) was utilized to excite sample. The laser
spot size is 150  160 µm. Laser repetition rate was set to 75
kHz with a power of 1.5 W to sufficiently excite the sample.
The sample was circulated by a peristaltic pump to avoid
radiation damage and showed no degradation after X-ray and
laser illumination. The sample chamber provided an air-tight
environment with CO2 gas purging during the measurement.
The K-edge spectra of the terpyridine nickel (II) complex were
collected by fluorescence mode using homemade avalanche
photodiode detectors (APDs).
Theoretical calculations. The hybrid functional B3LYP
was used for electronic structure calculations as implemented in
Gaussian 09 program. Geometry optimization was performed
with 6-31G(d) basis sets on all elements. The solvation energy
in the mixture of water and acetonitrile was calculated using the
SMD continuum solvation model. D3 dispersion correction,
proposed by Grimme, was also added. Analytic frequency
calculations were carried out on all optimized structures at the
In situ diffuse reflectance infrared Fourier-transform
spectroscopy (DRIFTS). In situ DRIFTS measurements were
performed using a Bruker IFS 66v Fourier transform
spectrometer equipped with a Harrick diffuse reflectance
accessory at the Infrared Spectroscopy and Microspectroscopy
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
same level of theory, to identify all stationary points as minima
(no imaginary frequency) and to obtain zero-point energy
corrections.
ligand corresponds to a distinct compression along the
molecular z axis of the NiN6 octahedron. The 3A2g ground state
of octahedral Ni(II) is Jahn-Teller stable while the NiN6
geometry solely reflects the strain effects imposed by the
ligands and/or steric packing effects in the unit cell (refer to the
Ni-N bond lengths, Table S2 and S3).18
1
2
3
4
RESULTS AND DISCUSSION
5
The [Ni(tpy)2]2+ (namely, Ni(II)) catalyst and BIH electron
donor are synthesized by following the protocols in SI and then
characterized by high-resolution mass spectroscopy (HRMS,
Figure S2), elemental analysis and UV-vis absorption
spectroscopy (Figure S3). We first evaluate the performance of
photocatalytic CO2 reduction in a quartz reactor containing
Ni(II) catalyst, [Ru(bpy)3]2+ (namely, Ru(II)) photosensitizer
and BIH electron donor in CH3CN/H2O solvent. After
optimizing the reaction conditions, the Ni(II) catalyst exhibits a
high turnover number (TON) of 2.36107 and turnover
frequency (TOF) of 385.6 s-1 with a CO selectivity up to 99%
(Figure S4), two orders of magnitude higher than the TON and
TOF records as ever reported in the state-of-the-art
homogeneous catalysts for photocatalytic CO2 reduction (Table
S1). The isotope-labeled 13CO2 experiments confirm that the
produced CO indeed originates from the CO2 reduction (Figure
S5). In addition to activity and selectivity, the Ni(II) catalyst
can undergo a complete catalytic cycle and maintain its valence
state after the reaction, offering excellent stability (Figure S6-
S8). The durability test shows a stagnation for the yield of CO
upon irradiation for 21 h only caused by the degradation of
Ru(II) photosensitizer, while the supplement of Ru(II) can
reactivate such nearly ceased photocatalytic CO2-to-CO
conversion.
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
58
59
60
Figure 1. The Ni transient X-ray absorption spectra obtained in a
system consisting of 1 mM Ni(II) catalyst, 1.5 mM Ru(II), and 10
mM BIH by 345 nm laser excitation in degassed CH3CN/H2O
solution: (a) Experimental XANES spectrum of Ni(II) complex at
ground state (black), the ground-state fitting XANES spectrum
based on the structure in Panel b, and the reconstructed transient
fitting spectrum based on the structure in Panel d. (b) The optimized
structure of ground-state Ni(II) complex. (c) Transient XANES
spectrum (red) at 0.4 µs after laser excitation and ground-state
spectrum (black dotted line). The difference spectra from three TR-
XAS scans (hollow circles) were obtained by subtracting the
ground-state spectrum from the transient one, respectively. The
corresponding XANES difference fitting spectrum (blue) was
presented with the fitted intermediate fraction 4.2%. (d) The
optimized structure of Ni intermediate. N1 is located at the benzene
ring in the middle of the terpyridine ligand, and N2 is located at the
benzene ring in the two sides of the terpyridine ligand.
Given that the Ni(II) complex has achieved high catalytic
activity and selectivity, we are now in a position to explore the
long-standing “black box” how exactly the homogeneous
catalysts work efficiently for the reaction. Such a puzzle refers
to the dynamic coordination and electronic structures of the
Ni(II) complex during the reaction, the role of the solvent on
catalytic performance, and the key steps for determining
product selectivity and conversion efficiency, constituting the
full mechanistic pathway. To decode the puzzle, we employ
operando TR-XAS to track the intermediate species of Ni(II)
complex during the photocatalytic CO2 reduction process. As a
powerful element-specific tool, TR-XAS has been applied to
resolve the information for the local coordination and electronic
structures of transient species in the system of homogeneous
hydrogen production with metal complexes.13-16 Nevertheless,
such a technique has not been employed to examine
photocatalytic CO2 reduction
–
a
substantially more
Upon light excitation, transient intermediate species is
formed within 0.4 µs. As the experimental transient spectrum
contains the signals from both the ground state and transient
state, the difference spectrum is obtained to highlight the
intermediate state by subtracting the ground signal from the
transient spectrum. Figure 1c shows the difference spectra from
three TR-XAS scans obtained for the given pump-probe delay
time (the corresponding XANES data are shown in Figure S9),
which provides the information on the local atomic structure
around the Ni atom and its oxidation state. The change in the Ni
coordination environment can be indicated by the combination
of XANES fitting with first-principles calculation. XANES
fitting reveals that the optimized intermediate structure
complicated system than hydrogen production. This
complicated system involves both CO2 and organic solvent so
that tracking the local coordination and electronic structures of
the catalyst exposed to the CO2 and solvent molecules is crucial
to the deep understanding of reaction mechanism. To track the
intermediates of Ni(II) complex during the photocatalytic CO2
reduction reaction, operando Ni K-edge X-ray absorption near-
edge structure (XANES) spectra are collected for the Ni(II)
complex before (laser-off) and after (laser-on) laser excitation.
The multiple scattering calculation is performed using FEFF9
software on a cluster with radius of 6.5 Å.17 Figure 1a shows the
Ni K-edge XANES spectra. The ground-state fitting XANES
spectrum indicates that the optimized ground-state geometry
structure of the Ni(II) complex is a six-coordinated Ni-N6
octahedral configuration as shown in Figure 1b. In the structure,
the predominant strain component of the rigid terpyridine
possesses
a
penta-coordinated distorted tetrahedron
configuration as shown in Figure 1d (the corresponding
difference XANES fitting procedure is shown in Figure S10).
Such penta-coordinated structure acquired from XANES fitting
ACS Paragon Plus Environment

Journal of the American Chemical Society
is consistent with the intermediate obtained from density
Page 4 of 8
1
2
3
4
5
6
7
8
function theory (DFT)-calculated free energy profiles, which
will be discussed in the following section. This indicates that
after photoexcitation, one terpyridine ligand is detached from
the Ni(II) complex when the central Ni ion is combined with
CO2 and CH3CN, resulting in a non-central geometric
symmetry. In such an intermediate structure, the Ni-N (Ni-N2)
bond on both sides of the terpyridine ligand is shortened while
the middle Ni-N (Ni-N1) bond is elongated (Table S2). The
positive feature at 8344 eV (peak A) originating from the
absorption edge shifts to lower energy, and the negative dip at
8354 eV (peak B) corresponding to the white-line peak for the
Ni intermediate is lower than that of the ground-state Ni(II)
complex. This observation indicates the formation of reduced
Ni(I) intermediate, induced by the transfer of photogenerated
electrons from the photosensitizer to Ni(II) complex. The
percentage of the Ni(I) intermediate obtained after laser
excitation is determined to about 4.2% by the XANES fitting.
The reconstructed transient fitting spectrum based on the
structure in Panel d is also shown in Figure 1a, convincing the
claim about the reduction of Ni(II) to Ni(I).
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
58
59
60
Figure 3. Optical transient absorption spectra of different systems
following 440 nm laser excitation (1 Hz, 10 mJ/pulse) in degassed
mixed solution: (a) Spectra of 0.15 mM Ru(II) after laser excitation
for 0-4 μs. (b) Spectra of 0.15 mM Ru(II) and 0.1 mM Ni(II) after
laser excitation for 0-4 μs. (c) Spectra of 0.15 mM Ru(II) and 1 mM
BIH after laser excitation for 0-200 μs. (d) Spectra of 0.15 mM
Ru(II), 0.1 mM Ni(II) and 1 mM BIH after laser excitation for 0-
200 μs. (e) Kinetic analysis of the Ru(I) signal growth probed at
510 nm in the absence (red trace) and presence (blue trace) of 0.1
mM Ni(II) catalyst. (f) Kinetic analysis of the Ru(I) signal decay
probed at 510 nm in the absence (red trace) and presence (blue
trace) of 0.1 mM Ni(II) catalyst.
Figure 2. (a) In situ DRIFTS spectra collected in N2 or CO2
atmosphere under irradiation in a reaction system containing Ru(II),
BIH and Ni(II) in CH3CN/H2O vapor. The system is irradiated with
a 365 nm LED lamp. (b) Operando SERS spectra collected during
photocatalytic CO2 reduction in a reaction system containing Ru(II),
BIH and Ni(II) in CH3CN/H2O solvent. The system is irradiated
with a 365 nm LED lamp, and 50 nm Au nanocubes are used as
SERS substrates.
To gain more information for chemical bonding, we further
employ in situ diffuse reflectance infrared Fourier-transform
spectroscopy (DRIFTS) and operando surface-enhanced
Raman spectroscopy (SERS) to detect the potential
intermediate species during the reaction process, in which the
background is subtracted to exclude environment factors as well
as the signals from Ni(II) complex and free molecular CO2. In
the in situ DRIFTS spectra, no additional vibration peaks can
be observed during irradiation up to 60 min in N2 atmosphere
(Figure 2a) or under dark condition (Figure S11). In
comparison, two bands at 1662 and 1622 cm-1 appear
significantly after irradiation for 5 min in CO2 atmosphere,
The information gleaned above has revealed the formation of
reduced Ni(I) intermediate which should be induced by the
transfer of photogenerated electrons from Ru(II)
photosensitizers to Ni(II) complex. Then a question naturally
arises how the photogenerated electrons behave during the
reaction. To investigate the dynamics of photogenerated
electrons involved in the system, electrochemical measurement
is first performed to examine the redox potentials of the
components  BIH and Ni(II) complex (Figure S12). As
indicated by the redox potentials (Figure S13), the transfer of
−
photo-induced
electrons from
the
excited
*Ru(II)
attributed to the formation of a Ni-CO2 or Ni-CO2H
photosensitizer (-0.81 V)24 to the Ni(II) catalyst (-1.27 V) is
thermodynamically unfavorable. As such, essentially the
catalytic process should occur through a reduction quenching
pathway. The excited *Ru(II) is quenched by BIH to form Ru(I)
which in turn delivers one electron to the Ni(II) catalyst to yield
a reduced Ni(I) intermediate for CO2 reduction, simultaneously
regenerating Ru(II). To further prove the dynamic process, we
collect photoluminescence (PL) spectra and analyze
complex.19,20 Moreover, complementary information for
intermediate species is provided by operando SERS
measurement. As shown in Figure 2b, no peaks are present for
CO2-saturated CH3CN/H2O solvent before irradiation. Upon
irradiation, three peaks are observed at 1604, 1553, and 1485
cm-1 that can be assigned to carboxylate COO-, while another
peak around 2040 cm-1 is attributed to C≡O stretching
vibration.21-23 These results from in situ DRIFT and operando
SERS further confirm the existence of intermediate species
tracked by TR-XAS.
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
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
58
59
60
Figure 4. (a) The free energy profiles for CO2 reduction by [Ni(tpy)]0 in both low-spin and high-spin states (ACN: acetonitrile) calculated
by DFT. (b) Schematic mechanistic pathway for the CO2 reduction reaction catalyzed by [Ni(tpy)]0.
luminescence lifetime. It turns out that BIH can quench the
luminescence of Ru(II) photosensitizer while Ni(II) cannot
(Figure S14), indicating that photogenerated electrons are not
directly transferred from the Ru(II) photosensitizer to the Ni(II)
catalyst. In the presence of BIH, however, the addition of Ni(II)
catalyst can further quench the luminescence of Ru(II)
photosensitizer. This phenomenon can be ascribed to the
mechanism that the formed Ru(I) delivers one electron to the
Ni(II) catalyst by regenerating Ru(II), shifting the quenching
equilibrium of excited *Ru(II) to Ru(I). This argument is also
supported by the luminescence lifetime comparison as shown in
Figure S15. The luminescence lifetime of Ru(II) decays from
256.9 ns to 170.6 ns with the addition of BIH while the sole
presence of Ni(II) catalyst does not significantly alter the
lifetime (245.4 ns). In the presence of BIH, the luminescence
lifetime of Ru(II) further decays from 170.6 ns to 127.6 ns by
shifting the quenching equilibrium of excited *Ru(II) to Ru(I)
with the addition of Ni(II) catalyst, further confirming the
efficient electron transfer from Ru(I) to Ni(II).
To elucidate the electron dynamics, optical transient
absorption (OTA) spectroscopy is further employed to examine
the photocatalytic system (Figure 3, S15-S18). As shown in
Figure 3a-d, upon laser excitation of the system at 440 nm, the
bleaching signal of ground-state Ru(II) and the absorption
attenuation signal of excited-state *Ru(II) appear at 446 nm and
365 nm, respectively. With the addition of BIH, an additional
absorption band located at 510 nm can be observed (Figure 3c,
d). This suggests the formation of Ru(I) by quenching excited-
state *Ru(II) with BIH, consistent with the findings from PL
spectra and lifetime characterizations (Figure S14 and S15).
Furthermore, the addition of Ni(II) catalyst does not alter the
kinetics of forming Ru(I) but remarkably facilitates the decay of
Ru(I) species. As shown in Figure 3e, there is no significant
difference for the signal growth of Ru(I) in the presence and
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 8
absence of Ni(II) (172 ns vs. 198 ns). However, the signal decay
photogenerated electrons to form [NiI(tpy)(tpy)•-]0 (b) that
subsequently loses a neutral terpyridine ligand, leaving a
coordinatively unsaturated and catalytically active Ni species
(c). As a result, the vacant site on the catalytically active Ni
species can be combined with CO2 and solvent CH3CN
1
2
3
4
5
6
7
8
of Ru(I) is accelerated significantly by the addition of Ni(II)
with a rate constant of 1.36×109 M-1 s-1 (Figure 3f, 7.89 μs
without Ni(II) vs. 3.8 μs with Ni(II)). Such a twice faster decay
is a solid indicator of efficient electron transfer from Ru(I) to
Ni(II) through BIH, manifesting the reduction quenching
mechanism of the Ru(II)/Ni(II)/BIH photocatalytic system.
molecules,
generating
a
five-coordinated
Ni(tpy)(CO2)(CH3CN) species (d). The next step is protonation
of the anionic O=C–O- in the intermediate complex to form the
intermediate adduct (e). Following closely, CO formation
requires an extra electron so as to trigger C–O bond cleavage,
and complexes (f) and (g) are formed successively via a coupled
electron-proton transfer-bond breaking step.29 Eventually, CO
and solvent CH3CN molecules are released to complete the
cycle.
Based on the information on the coordination and electronic
structures of intermediate species, we further depict the free
energy profiles of CO2-to-CO reduction reaction catalyzed by
[Ni(tpy)]2+ by DFT calculations. Since [Ni(tpy)2]2+ can accept
one electron by divalent Ni metal ions to form [Ni(tpy)2]+ and
further accept one electron by a terpyridine ligand to form
[Ni(tpy)2]0 before initiating the CO2 reduction reaction
(CO2RR),25 we compare the formation energies of [Ni(tpy)]+
(35.45 kcal/mol for low spin and 49.43 kcal/mol for high spin,
Figure S20) and [Ni(tpy)]0 (17.80 kcal/mol for low spin and
32.06 kcal/mol for high spin, Figure 4a). The comparison
reveals that the latter requires less energy input in the following
step of detaching a terpyridine ligand. In this case, we mainly
present the reaction profiles associated with [Ni(tpy)]0 here. In
addition, the calculations show that the simultaneous
coordination of acetonitrile and CO2 to the Ni center is stronger
than that of H2O and CO2 by 2.4 kcal/mol in binding energy (see
supporting information for details, Table S4), suggesting that
the Ni center prefers to simultaneously coordinate with
acetonitrile and CO2. Furthermore, owing to stereo repulsion,
H2O could not bind to Ni center when coordinated with
acetonitrile and CO2. Meanwhile, although H2O may provide
protons to facilitate the CO2 reduction to CO through the
dihydroxyl pathway in aqueous environment, both the
selectivity and yield of CO production decrease with the
increase of water content due to the competitive hydrogen
evolution (Figure S4c), indicating that H2O would hinder the
CO2 reduction. Taken together, the calculation results can reach
the conclusion that the acetonitrile and CO2 rather than H2O and
other ligands are simultaneously coordinated to the Ni center.
This conclusion is consistent with the result from the difference
XANES fitting for TR-XAS spectra, which suggests that the
intermediate is a five-coordinated Ni species with the most
likely structure Ni(tpy)(CO2)(CH3CN) as depicted in Figure 1d.
DFT calculation shows that the attached CO2 is bent with a ∠
OCO of 144.96° at low spin and 140.68° at high spin,
respectively. As revealed by Mulliken population analysis, the
low-spin and high-spin states carry a negative charge of 0.38 e-
and 0.49 e-, respectively. This clearly indicates the formation of
anionic O=C–O- in the intermediate complex, which is
generally assumed as the rate-limiting step of CO2RR.21 As the
reduction reaction proceeds to generate CO, it can go either the
dihydroxyl pathway involving protons in aqueous environment
or the carbonate pathway involving a second CO2 in non-
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
58
59
60
CONCLUSION
In conclusion, we have successfully identified the dynamic
coordination and electronic structures of a Ni complex catalyst
in a homogeneous photocatalytic system for CO2 reduction by
operando TR-XAS spectroscopy. It has revealed that the
reduced Ni(I) intermediate is coordinated with a terpyridine
ligand, CO2 and solvent CH3CN molecules to generate a five-
coordinated Ni(tpy)(CO2)(CH3CN) species. Such an
intermediate specie has also been confirmed by in situ DRIFTS
and operando SERS. The formation of reduced Ni(I)
intermediate is induced by the transfer of photogenerated
electrons from photosensitizer to terpyridine nickel(II) complex
whose dynamics has been resolved by OTA spectroscopy.
Combined with the detail DFT calculation, the experimental
evidence based on operando and time-resolved spectroscopy
provides a direct view for the full mechanistic pathway. The
findings in this work represent a step forward in opening the
“black box” of CO2 reduction in highly efficient homogeneous
photocatalytic systems. For instance, this work highlights the
key role of solvent molecules in reactions, which has been often
overlooked, and elucidates how homogeneous photosensitizer,
catalyst and electron donor work synergistically for
photocatalytic CO2 reduction. This work also provides insights
for rationally designing and developing advanced
photocatalytic systems for artificial photosynthesis from the
angle of catalytic sites.
ASSOCIATED CONTENT
Supporting Information
Detailed experimental section, characterization methods, and
additional material characterizations. The Supporting Information
is available free of charge on the ACS Publications website.
AUTHOR INFORMATION
Corresponding Author
+
aqueous environment.20, 26 By adopting the Zundel (H5O2 ) as
*yjxiong@ustc.edu.cn; taoy@ihep.ac.cn; wkzhang@bnu.edu.cn;
gaoc@ustc.edu.cn
the form of hydronium in our calculation,27,28 we reveal that the
dihydroxyl pathway is more favorable than the carbonate
pathway in both low spin (-21.33 kcal/mol vs. -17.39 kcal/mol)
and high spin (-12.55 kcal/mol vs. -6.80 kcal/mol) states,
regarding to their respective free energy changes. As a result,
we conclude that CO2RR in our catalytic system should undergo
the dehydroxyl pathway.
Author Contributions
¶These authors contributed equally.
The manuscript was written through contributions of all authors.
Notes
The authors declare no competing financial interests.
Taken together, the coordination structure, electronic
structure, charge dynamics and reaction free energy provide the
essential information for proposing the full reaction pathway.
As illustrated in Figure 4b, [NiII(tpy)2]2+ (a) accepts two
ACKNOWLEDGMENT
This work was supported by National Key R&D Program of China
(2017YFA0207301), NSFC (21725102, 91961106, 21703220,
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
resolved X-ray absorption spectroscopy. J. Phys. Chem. C 2016, 120,
20049.
U1332205, 21773012 and 91753118), CAS Key Research Program
1
2
3
4
5
6
7
8
of Frontier Sciences (QYZDB-SSW-SLH018), DNL Cooperation
Fund, CAS (DNL201922), Beijing Natural Science Foundation
(L172028) and CAS Interdisciplinary Innovation Team. TR-XAS
characterization was performed at the beamline 1W2B in the
Beijing Synchrotron Radiation Facility (BSRF), China. In situ
DRIFTS was collected at the Infrared Spectroscopy and
Microspectroscopy Endstation (BL01B) in the National
Synchrotron Radiation Laboratory (NSRL) in Hefei, China.
(14) Moonshiram, D.; Gimbert-Suriñach, C.; Guda, A.; Picon, A.;
Lehmann, C. S.; Zhang, X.; Doumy, G.; March, A. M.; Benet-
Buchholz, J.; Soldatov, A.; Llobet, A.; Southworth, S. H. Tracking the
structural and electronic configurations of a cobalt proton reduction
catalyst in water. J. Am. Chem. Soc. 2016, 138, 10586.
(15) Li, Z. J.; Zhan, F.; Xiao, H.; Zhang, X.; Kong, Q. Y.; Fan, X. B.;
Liu, W. Q.; Huang, M. Y.; Huang, C.; Gao, Y. J.; Li, X. B.; Meng, Q.
Y.; Feng, K.; Chen, B.; Tung, C. H.; Zhao, H. F.; Tao, Y.; Wu, L. Z.
Tracking Co(I) intermediate in operando in photocatalytic hydrogen
evolution by X-ray transient absorption spectroscopy and DFT
calculation. J. Phys. Chem. Lett. 2016, 7, 5253.
(16) Smolentsev, G.; Soldatov, M. A.; Probst, B.; Bachmann, C.;
Azzaroli, N.; Alberto, R.; Nachtegaal, M.; vanꢀBokhoven, J. A.
Structure of the CoI intermediate of a Cobalt pentapyridyl catalyst for
hydrogen evolution revealed by time-resolved X-ray spectroscopy.
ChemSusChem 2018, 11, 3087.
(17) Rehr, J. J.; Kas, J. J.; Vila, F. D.; Prange, M. P.; Jorissen, K.
Parameter-free calculations of X-ray spectra with FEFF9. Phys. Chem.
Chem. Phys. 2010, 12, 5503.
(18) Folgado, J. V.; Henke, W.; Allmann, R.; Stratemeier, H.; Beltran-
Porter, D.; Rojo, T.; Reinen, D. Fluxionality in hexacoordinated copper
(II) complexes with 2, 2': 6', 2''-terpyridine (Terpy) and related ligands:
structural and spectroscopic investigations. Inorg. Chem. 1990, 29,
2035.
(19) Sampson, M. D.; Froehlich, J. D.; Smieja, J. M.; Benson, E. E.;
Sharp, I. D.; Kubiak, C. P. Direct observation of the reduction of carbon
dioxide by rhenium bipyridine catalysts. Energy Environ. Sci. 2013, 6,
3748.
(20) Kuehnel, M. F.; Orchard, K. L.; Dalle, K. E.; Reisner, E. Selective
photocatalytic CO2 reduction in water through anchoring of a molecular
Ni catalyst on CdS nanocrystals. J. Am. Chem. Soc. 2017, 139, 7217.
(21) Chernyshova, I. V.; Somasundaran, P.; Ponnurangam, S. On the
origin of the elusive first intermediate of CO2 electroreduction. Proc.
Natl. Acad. Sci. U.S.A. 2018, 115, E9261.
(22) Jiang, S.; Klingan, K.; Pasquini, C.; Dau, H. New aspects of
operando Raman spectroscopy applied to electrochemical CO2
reduction on Cu foams. J. Chem. Phys. 2019, 150, 041718.
(23) Tammer, M. G. Sokrates: Infrared and Raman characteristic group
frequencies: tables and charts. Colloid Polym. Sci. 2004, 283, 235.
(24) Prier, C. K.; Rankic, D. A.; MacMillan, D. W. C. Visible light
photoredox catalysis with transition metal complexes: applications in
organic synthesis. Chem. Rev. 2013, 113, 5322.
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
58
59
60
REFERENCES
(1) Kuriki, R.; Sekizawa, K.; Ishitani, O.; Maeda, K. Visible-light-
driven CO2 reduction with carbon nitride: enhancing the activity of
ruthenium catalysts. Angew. Chem. Int. Ed. 2015, 54, 2406.
(2) Rao, H.; Schmidt, L. C.; Bonin, J.; Robert, M. Visible-light-driven
methane formation from CO2 with a molecular iron catalyst. Nature
2017, 548, 74.
(3) Guo, Z.; Chen, G.; Cometto, C.; Ma, B.; Zhao, H.; Groizard, T.;
Chen, L.; Fan, H.; Man, W. L.; Yiu, S. M.; Lau, K. C.; Lau, T. C.;
Robert, M. Selectivity control of CO versus HCOO− production in the
visible-light-driven catalytic reduction of CO2 with two cooperative
metal sites. Nat. Catal. 2019, 2, 801.
(4) Cui, X.; Wang, J.; Liu, B.; Ling, S.; Long, R.; Xiong, Y. Turning
Au nanoclusters catalytically active for visible-light-driven CO2
reduction through bridging ligands. J. Am. Chem. Soc. 2018, 140,
16514.
(5) Wang, J.; Xia, T.; Wang, L.; Zheng, X.; Qi, Z.; Gao, C.; Zhu, J.; Li,
Z.; Xu, H.; Xiong, Y. Enabling visible-light-driven selective CO2
reduction by doping quantum dots: trapping electrons and suppressing
H2 evolution. Angew. Chem. Int. Ed. 2018, 57, 16447.
(6) Choi, K. M.; Kim, D.; Rungtaweevoranit, B.; Trickett, C. A.;
Barmanbek, J. T. D.; Alshammari, A. S.; Yang, P.; Yaghi, O. M.
Plasmon-enhanced photocatalytic CO2 conversion within metal–
organic frameworks under visible light. J. Am. Chem. Soc. 2017, 139,
356.
(7) Gao, C.; Chen, S.; Wang, Y.; Wang, J.; Zheng, X.; Zhu, J.; Song,
L.; Zhang, W.; Xiong, Y. Heterogeneous single-atom catalyst for
visible-light-driven high-turnover CO2 reduction: the role of electron
transfer. Adv. Mater. 2018, 30, 1704624.
(8) Gao, C.; Wang, J.; Xu, H.; Xiong, Y. Coordination chemistry in the
design of heterogeneous photocatalysts. Chem. Soc. Rev. 2017, 46,
2799.
(9) Elgrishi, N.; Chambers, M. B.; Wang, X.; Fontecave, M. Molecular
polypyridine-based metal complexes as catalysts for the reduction of
CO2. Chem. Soc. Rev. 2017, 46, 761.
(25) Wang, M.; England, J.; Weyhermüller, T.; Wieghardt, K.
Electronic structures of “low-valent” neutral complexes [NiL2]0 (S = 0;
L = bpy, phen, tpy) – an experimental and DFT computational study.
Eur. J. Inorg. Chem. 2015, 2015, 1511.
(26) Barlow, J. M.; Yang, J. Y. Thermodynamic considerations for
optimizing selective CO2 reduction by molecular catalysts. ACS Cent.
Sci. 2019, 5, 580.
(10) Wang, J. W.; Liu, W. J.; Zhong, D. C.; Lu, T. B. Nickel complexes
as molecular catalysts for water splitting and CO2 reduction. Coordin.
Chem. Rev. 2019, 378, 237.
(11) Elgrishi, N.; Chambers, M. B.; Artero, V.; Fontecave, M.
Terpyridine complexes of first row transition metals and
electrochemical reduction of CO2 to CO. Phys. Chem. Chem. Phys.
2014, 16, 13635.
(12) Wang, H.; Yu, C.; Wei, X.; Gao, Z.; Xu, G. L.; Sun, D. R.; Li, Z.;
Zhou, Y.; Li, Q. J.; Zhang, B. B.; Xu, J. Q.; Wang, L.; Zhang, Y.; Tan,
Y. L.; Tao, Y. Development of picosecond time-resolved X-ray
absorption spectroscopy by high-repetition-rate laser pump/X-ray
probe at Beijing Synchrotron Radiation Facility. J. Synchrotron. Radiat
2017, 24, 667.
(13) Moonshiram, D.; Guda, A.; Kohler, L.; Picon, A.; Guda, S.;
Lehmann, C. S.; Zhang, X.; Southworth, S. H.; Mulfort, K. L.
Mechanistic evaluation of a nickel proton reduction catalyst using time-
(27) Dahms, F.; Costard, R.; Pines, E.; Fingerhut, B. P.; Nibbering, E.
T. J.; Elsaesser, T. The hydrated excess proton in the Zundel cation
+
H5O2 : the role of ultrafast solvent fluctuations. Angew. Chem. Int. Ed.
2016, 55, 10600.
(28) Park, M.; Shin, I.; Singh, N. J.; Kim, K. S. Eigen and Zundel forms
of small protonated water clusters:ꢁ structures and infrared spectra. J.
Phys. Chem. A 2007, 111, 10692.
(29) Bonin, J.; Maurin, A.; Robert, M. Molecular catalysis of the
electrochemical and photochemical reduction of CO2 with Fe and Co
metal based complexes. Recent advances. Coordin. Chem. Rev. 2017,
334, 184.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
1
2
3
SYNOPSIS TOC
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
58
59
60
ACS Paragon Plus Environment