Visible light driven reduction of CO2 catalyzed by an abundant manganese catalyst with zinc porphyrin photosensitizer

Article abstract of DOI:10.1016/j.apcata.2016.04.035

Visible light driven photocatalytic CO₂ conversion has been investigated using abundant metal complexes catalyst fac-[Mn(phen)(CO)₃Br] and photosensitizer ZnTPP, which efficiently produced carbon monoxide and formic acid in an aqueous acetonitrile solution. The photochemical and electrochemical properties of the catalysis system proved the good ability of the light utilization and the reduction. During the 180-min irradiation experiments with the Mn/Zn ratio of 2:1, the TONs reached 64 for the CO formation and 16 for the formic acid formation, respectively. An integrated mechanism was proposed for the CO₂ reduction in this system. This photocatalyst system not only shows environmental friendly and sustainability, but also retains the product selectivity in the CO₂ photoreduction system.

Full text of DOI:10.1016/j.apcata.2016.04.035

Accepted Manuscript
Title: Visible light driven reduction of CO2 catalyzed by an
abundant manganese catalyst with zinc porphyrin
photosensitizer
Author: Jun-Xiao Zhang Chang-Ying Hu Wei Wang Hui
Wang Zhao-Yong Bian
PII:
S0926-860X(16)30238-1
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.04.035
APCATA 15868
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
23-1-2016
13-4-2016
28-4-2016
Please cite this article as: Jun-Xiao Zhang, Chang-Ying Hu, Wei Wang, Hui Wang,
Zhao-Yong Bian, Visible light driven reduction of CO2 catalyzed by an abundant
manganese catalyst with zinc porphyrin photosensitizer, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2016.04.035
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Visible light driven reduction of CO2 catalyzed by an abundant
manganese catalyst with zinc porphyrin photosensitizer
Jun‐Xiao Zhanga, Chang‐Ying Hua, Wei Wangb, Hui Wangb,**, Zhao‐Yong Biana,*
1College of Water Sciences, Beijing Normal University, Beijing 100875, PR China
2College of Environmental Science and Engineering, Beijing Forestry University, Beijing 100083, PR
China
*Corresponding author. Tel.: +86 10 58802736; Fax: +86 10 58802739
E‐mail address: bian@bnu.edu.cn
** Corresponding author. Tel.: +86 10 62336615; Fax: +86 10 62336596
E‐mail address: wanghui@bjfu.edu.cn

Graphical abstract

Highlights
(1) We investigated the visible light driven reduction of CO2 using an abundant
metal system containing tricarbonyl manganese catalyst with zinc porphyrin.
(2) Abundant and cheap metal manganese and zinc complexes were used in this
study to replace novel rare metal complexes.
(3) In the aqueous acetonitrile reaction solution, carbon monoxide and formic
acid were efficiently produced as the gas and liquid products.
Visible light driven photocatalytic CO2 conversion has been investigated using abundant metal complexes catalyst fac‐
[Mn(phen)(CO)3Br] and photosensitizer ZnTPP, which efficiently produced carbon monoxide and formic acid in an aqueous
acetonitrile solution. The photochemical and electrochemical properties of the catalysis system proved the good ability of the
light utilization and the reduction. During the 180‐min irradiation experiments with the Mn/Zn ratio of 2:1, the TONs reached 64
for the CO formation and 16 for the formic acid formation, respectively. An integrated mechanism was proposed for the CO2
reduction in this system. This photocatalyst system not only shows environmental friendly and sustainability, but also retains the
product selectivity in the CO2 photoreduction system.
Keywords: Visible light catalysis; CO2 reduction; abundant catalyst; Mn;
Zn

Introduction
In recent years, photocatalytic CO2 conversion has attracted significant attention because of a couple of
serious problems such as global warming and shortage of energy sources. Many attempts have been taken to
design a system to convert CO2 to useful fuel with the help of novel methods, such as electrochemistry as well as
photochemistry methods.1 In 1979, Inoue and his coworkers first announced that TiO2 has the ability to catalyze
the reduction of CO2, which induced the research of photocatalytic reduction of CO2.2 Since then, researchers
continued to looking for new efficient catalysts for the CO2 conversion, in which various kinds of macrocyclic
complexes have been designed and synthesized, which are capable for the absorption of visible light. For instance,
Hawecker and his coworkers firstly used [Re(bpy)(CO)3Cl] (bpy=2,2’‐bipyridine) as both photocatalyst and
photosensitizer for the CO2 reduction in a homogeneous system with the turnover number (TONCO)=48 in a four‐
hour photoreaction.3 Beyond that, more Ru and Re macrocycle complexes showed good photocatalytic ability and
were used as the catalyst for CO2 photo‐reduction.4 Recently, Kuramochi and his coworkers found that trans‐
[Ru(bpy)(CO)2Cl2] achieved an unexpected production of CO and formic acid in a photocatalysis, in which
[Ru(bpy)3]2+ served as a photosensitizer.5
Although these systems with such catalysts loaded with precious rare metals get efficiencies in the CO2
photocatalysis, their applications are limited by the low concentration in earth and high price of the noble metals.
Hence, abundant metals have more advantages for broader application because of their easier achievement and
the lower cost. Therefore, researchers have always been trying to look for suitable abundant metals for the load
for photo‐catalyst, while for decades rhenium has always been regarded as one of the best one.6 For example, Liu
and his co‐workers used a copper porphyrin based metal organic framework in chemical adsorption capture and
photoreduction of CO2.7 In recent years, researchers began to explore manganese catalysts, because manganese,
which shares the same family as rhenium, has similar properties in some aspects. As for the catalytic reduction of
CO2, manganese compounds could act like Rhenium complexes. For example, Mukhopadhyay and his coworkers
used a bis(imino)pyridine manganese electrocatalyst realizing the promotion of H+ reduction by the addition of
CO2.8 It was also found that manganese have the ability to substitute rhenium in the electrocatalyst system.9 In
order to make the extensive study of the manganese catalyst, researchers began to study the similarities and
differences between materials loaded with these two elements. Smieja and his coworkers found that the addition
of Brønsted acid to CO2‐sturated solutions of Mn complexes and subsequent reduction of the complexes could
lead to the stable and efficient production of CO from CO2.10 Riplinger and his coworkers have also studied the
characters of the two elements and found that with the weak Brønsted acid manganese compound can act well in
the electronic CO2 reduction.11 Furthermore, researchers also began to study the effect of manganese catalysts in
both electrocatalytic and photocatalytic reduction of CO2. Agarwal and his coworkers has found manganese
complexes electrocatalyzed the CO2 reduction with a turnover frequency (TOF) = 0.14 s‐1.12 Zeng and his
coworkers used a manganese(I) tricarbonyl complex containing a nonaromatic α‐diimine ligand in the
electrocatalytic reduction of CO2.13 Using [Ru(dmb)3]2+ as the photosensitizer, Takeda and his coworkers has
reported a carbonyl manganese complex acted as a photocatalyst for the CO2 reduction and formed CO with
TONCO = 13.14 However, many Mn compounds do not get the good ability in the using of visible light. Thus there is
a necessity to looking for a suitable photosensitizer helping to absorb the visible light in this research.
Researchers have used precious metal complexes, such as ruthenium complexes, as the photosensitizer for a
long time. There are enormous types of macrocycle photosensitizers applied loaded with precious and abundant
metals during past years. Among these compounds ruthenium complexes have been used in plenty of
photocatalysis systems both as photocatalyst and photosensitizer, for which it is also regarded as one of the most
efficient materials. For instance, Shakeri and his co‐workers used ruthenium(II) complex as a photosensitizer in a
rhenium photocatalytic system.3d However, despite the high absorbing efficiency, ruthenium complex costs too
much so that it makes an unsustainable system, which could not be widely used. It is necessary to replace the
precious metal with abundant metal as the coordinated metal for sustainable application, for which transitional
metals, such as zinc and copper, have come into the sight of researchers. In a research done by Li and Diau
showed that the power conversion efficiency of porphyrin dyes has grown a lot to 12% from the year of 1991 to
2012 in excess of Ru‐based complexes and organic dyes.15 This result illustrates the trend of the research of
porphyrins in recent years, because of the low cost and likely efficiency, especially for porphyrins loaded with
transition metals. For instance, Amao and his coworkers reported a photocatalytic system containing zinc
tetraphenyl porphyrin (ZnTPP), serving as a photosensitizer, MV2+, FDH (formate dehydrogenase), AldDH
(aldehyde dehydrogenase), ADH (aldehyde), TEOA, and NaHCO3 in potassium phosphate buffer solution (pH 8), in
which methanol as product was found after three hours of irradiation.16
In this research, we reported a homogeneous photochemical system using fac‐[Mn(phen)(CO)3Br] (phen

=1,10‐phenanthroline) as a precursor of the catalyst and [ZnTPP] as a photosensitizer. In this system, CO2 could be
reduced to CO as gas product and formic acid as liquid product. In this experiment, the use of ZnTPP prevented
the generation of hydrogen from the reduction of proton, which showed high product selectivity to the non‐H2
products of the CO2 reduction.
Experimental Section
Reactions were performed in oven‐dried round‐bottomed flasks unless otherwise noted. Reactions that
required an inert atmosphere were conducted under pressure of N2. Air and moisture sensitive reagents were
transferred using standard syringe or cannula techniques. Reagents and solvents were of reagent grade and used
as received.
Tetraphenyl porphyrin (TPP) was synthesized by a method modified from the classical Adler method.17 9.6 g
of aluminum chloride anhydrous was dissolved in 165 mL N,N‐dimethylformamide (DMF) in a three‐necked,
round‐bottomed flask containing a stirring bar. When the solution was heated to reflux, 6.4 g of pyrrole and 7.7 g
of benzaldehyde were added into the flask. The reaction was stopped after heating for additional 2 hours. A
certain volume of absolute ethyl alcohol was added into the device after cooled to room temperature. After
overnight standing, a large amount of solid precipitated and was filtrated out. The purple crystals were achieved
after washed by ethanol and hot water in sequence for at least two times. The crude product was then purified by
column chromatography (eluent: dichloromethane). The compound was characterized by NMR and elemental
1
analysis. H NMR (CDCl3) δ: ‐2.70 (N‐H); 7.71‐7.85 (othor‐ and para‐ H); 8.97 (meta‐H). Element analysis: TPP:
Found: C, 86.0; H, 5.1; N, 9.7. C44H30N4 requires C, 85.9; H, 4.9; N, 9.2%.
Zinc tetraphenyl porphyrin (ZnTPP) was synthesized following the reported method11 and purified by
recrystallization and centrifugation. The compound was characterized by ESI‐MS and elemental analysis. Through
the ESI‐MS analysis, it could be seen that peaks at 677.5 got the highest value. According to the molecular weight
of ZnTPP (676), the values in the picture could prove that the zinc element was loaded successfully to TPP. From
element analysis of ZnTPP (Found: C, 77.9; H, 4.1; N, 8.4. C44H30N4 requires C, 78.1; H, 4.1; N, 8.3%), it could be
easily found that the sum of the percentage of the three elements is 90.4%, which is much fewer than 100%,
while the difference stands for the existence of zinc.
fac‐[Mn(phen)(CO)3Br] was synthesized by a reflux‐condensation method and purified by recrystallization
and filtration. The material was characterized by ESI‐MS, which showed the molecular weight was 335, which is
the same as the theoretical value calculated by taking away the Br and adding the CH3CN radical, which leads to
the formation of [Mn(phen)(CO)3CH3CN] structure. Therefore, the result matched the exact calculating value well.
1H NMR spectra were recorded using BRUKER AVANCE 500Hz and Mass spectrometry was recorded using
Quattro Micro API from Waters Micromass. Electrochemistry was performed using a CHI‐660 electrochemical
workstation. Cyclic voltammetry (CV) was performed using a standard three‐electrode configuration. The working
electrode was a polished glassy carbon electrode and a piece of platinum wire was used as the counter electrode,
while a saturated calomel electrode (SCE) was used as the reference electrode. Prior to electrolysis the solution
was dissolved by MeCN bubbled with nitrogen and CO2 gas for 40 min, respectively. With all these steps, the cell
was sealed and electrolysis initiated. UV‐vis absorption spectra were acquired by UV‐3600 from Shimazu
Corporation. All absorption spectra were recorded at room temperature and all samples were prepared in
dichloromethane (DCM). Samples for fluorescence analysis were prepared in an analogous method to that
described above for the preparation of samples for UV‐vis spectroscopy. For ZnTPP, samples were excited at
λex=499 nm and emission was monitored from 509 nm to 800 nm with a step size of 1 nm. The experiment began
with the addition of 120 mL MeCN into a 250 mL flask, which was bubbled with nitrogen for 1 h. The ZnTPP was
resolved in bubbled acetonitrile solution and prepared as a 0.1 mM MeCN solution. The ZnTPP solution was
quenched using the 0.1M TEA‐MeCN solution bubbled by nitrogen for 40 min. The quenching experiment was
also done using the TEA solution.
fac‐[Mn(phen)(CO)3Br], ZnTPP and TEA of certain quantities are added into tube shaped reactor containing
25 mL solvent of the mixture of MeCN and deionized water(v/v=20:1). The solution is bubbled by nitrogen for 40
min to wipe off oxygen and then by CO2 for 40 min to make sure the solution is saturated with CO2. The tubes of
solution with magnetic rotors are sealed up to make sure that the gas inside would never come out and put into
an equipment installed with a lamp to do the experiment. After a certain period, samples are taken out and
placed in a dark container, followed by the detection of products, including gas chromatograph and ion
chromatograph. The turnover numbers are the average of at least three individual acquisitions.
Results and Discussion
Electrochemical properties

The catalysis material fac‐[Mn(phen)(CO)3Br] was analyzed by electrochemical cyclic voltammetry. It could be
easily seen that there are three reduction peaks in the CV spectrum under the nitrogen pressure (as shown in Fig
1). The first peak appeared at ‐1.2V vs. SCE, while the second occured at ‐1.5 V and the third at ‐1.75 V,
respectively. Under CO2 (red trace), the current enhancement of peaks, relative to the nitrogen scan at each peak,
were 1.98×, 1.88×, and 4.58×, respectively. These three peaks stood for the three reaction processes of fac‐
[Mn(phen)(CO)3Br] under different potentials. And the considerable current enhancement illustrated that there
were reactions between CO2 and fac‐[Mn(phen)(CO)3Br] in the system. The first peak stood for the protonation of
[Mn(phen)(CO)3(MeCN)], which could be observed that the current difference was not big. Under the potential of
‐1.5 V vs. SCE, [Mn(phen)(CO)3]2, the key compound in the whole reaction, was formed. However, it quickly
‐
received energy from the electrode and then transformed to [Mn(phen)(CO)3 ]. It was also declared in Bourrez’s
research that for this kind of manganese compounds, lower potential could lead to the decrease of selectivity
because of the production of a small amount of hydrogen in a process of electrocatalytic reduction.9 In this
research, the electrolysis producing CO was done under a potential of ‐1.70 V vs. Ag/Ag+, obtaining the TONCO=13.
However, it was also found that when the potential becomes more negative than ‐2.1 V vs. Ag/Ag+, there would
be hydrogen generated from the system. That is to say, in order to increase the selectivity of the catalytic
reduction, less negative reduction potential is needed.
In an electroreaction, the driving force and energy were given by the applied potential. However, in a
photoreaction, the irradiation energy absorbed by a photocatalyst will motivate the proposed reaction. If the
potential was no more negative than ‐2.1 V vs. Ag/Ag+ (‐1.8 V vs. SCE), there would be no hydrogen generated. As
shown in Fig 2, it could be observed that there were two obvious reduction peaks, one at ‐1.05 V vs. SCE and the
other at ‐1.65 V vs. SCE. The first peak (‐1.05 V vs. SCE) stood for the reduction of ZnTPP to an excited state of
ZnTPP (ZnTPP E1), which was a transition form of the ZnTPP complex. The second state (‐1.65 V vs. SCE) stood for
the reduction of ZnTPP E1 to ZnTPP‐. In a typical electroreaction, when the potential was between these two state
(not more negative ‐1.8 V vs SCE), there would be no hydrogen generated. However, in a typical photoreaction,
the energy was got from the irradiation and electrons were provided by the electron donor. The energy
transferred from excited ZnTPP should be between ‐1.65 V and ‐1.05 V vs. SCE which was positive than ‐1.8 V vs.
SCE. Therefore, the use of ZnTPP as a photosensitizer could prevent the production of hydrogen.
With the comparison of the current peaks under the two reduction potentials, it could be found that the
peak currents and the square roots of the scan rate match better the linear relationship (R2=0.983). It means that
the electron transfer process of ZnTPP was controlled by the diffusion process in the solution.
In addition, ZnTPP is one of excellent materials for the visible light utilization, due to the good ability of
absorbing visible light. As shown in Fig 3, there were three absorption peaks, appearing at 424 nm, 556 nm and
593 nm. According to the former research, the Soret‐band of ZnTPP appeared at 425 nm, while Q‐bands appeared
at 559 nm and 599 nm.17b This indicated the good absorption ability of ZnTPP sample in this research. In the
fluorescence spectrum of ZnTPP, it could be easily seen that there are two strong emission peaks, one at 600 nm
and another at 650 nm. It indicated that ZnTPP could transform the absorbed light energy into other kinds of light
and the excited state of ZnTPP had a lifetime long enough for the energy utilization.
Quenching Experiment
In order to use ZnTPP as a photosensitizer in this research, a quenching experiment was done to confirm the
sacrificial reductant. Triethylamine (TEA) is regarded as one of the most efficient reductant, also known as
electron donor. Fluorescence quenching measurements in the acetonitrile solutions containing ZnTPP with
increasing concentrations of TEA were analyzed by the Sterm‐Volmer equation:
In this equation, I0 and I stand for the fluorescence intensity without and with the quencher, respectively. kq
is the quenching rate constant and [Q] is the quencher concentration. τ0 is the excited‐state lifetime of ZnTPP
without the quencher which was 2ns in a acetonitrile solution.18
According to spectrums above, it could be figured out that the evolution of the fluorescence ratio with the
concentration of TEA was linear. The quenching rate constant kq could be calculated as 2.7*1010 L mol‐1 s‐1. The
conclusion could be made that the reductive electron transfers between ZnTPP and TEA was efficient.
Photoreaction
The molar ratio of the catalyst and photosensitizer in a photocatalysis system is important for the
photocatalysis operation. A group of experiments was set to determine the suitable ratio of the concentration of

the catalyst and the photosensitizer. In a typical photoreaction experiment, a 20:1 mixed solution of acetonitrile
and deionized water (v/v) containing the catalyst fac‐[Mn(phen)(CO)3Br], photosensitizer ZnTPP, and reductant
TEA was irradiated using xenon lamp under a CO2 atmosphere. The concentration of catalyst fac‐
[Mn(phen)(CO)3Br] was 0.5 mM, while different concentrations of the photosensitizer were set to investigate the
effect of ratio of catalyst and photosensitizer on the photocatalytic ability to the CO2 reduction.
Table 1 summarized catalytic reduction efficiencies under different operation conditions. In these
photoreaction experiments, the main reduction products were CO as a gas product and formic acid as a liquid
product. There was no hydrogen gas could be determined. From the control experiment (No. 6), the catalyst fac‐
[Mn(phen)(CO)3Br] formed little products without the photosensitizer ZnTPP under the same conditions. Because
the catalyst fac‐[Mn(phen)(CO)3Br] had weak absorption in the visible light region and could not efficiently utilize
the visible light energy. It indicated that the visible light was the main driving force for this CO2 conversion.
Additionally, in experiment 7, there was no products detected in the reaction system without catalyst fac‐
[Mn(phen)(CO)3Br]. It indicated that both the catalyst and photosensitizer played vital roles in this CO2 conversion.
The whole experiments could be divided into two parts with controlling variable method. On the one hand, the
concentration of ZnTPP was set as a constant (0.5mM), while the concentration of fac‐[Mn(phen)(CO)3Br] was
changed from 0.5mM to 2mM (Group 1, 2 and 4). In this series of experiment, it could be seen that with the
increase of the concentration of fac‐[Mn(phen)(CO)3Br], the production of both CO and formic acid was enhanced.
Specifically, when the concentration ratio of catalyst fac‐[Mn(phen)(CO)3Br] and the photosensitizer ZnTPP
increased from 1:1 to 3:1 and ultimately 4:1, the TONCO increased from 12 to 97 and finally 119, while TONHCOOH
rose from 10 to 18 and finally 19. On the other hand, in the other series of experiments, the concentration of fac‐
[Mn(phen)(CO)3Br] was the constant, while the concentration of ZnTPP changed from 0.25mM to 1mM (Group 3,
4 and 5). It could also be found that with the concentration ratio increased, the TON of the formed products
became bigger. For the gas product CO, TON increased from 8 to 64 when the Mn/Zn ratio changed from 1:2 to
2:1. Meanwhile, TON of formed liquid product formic acid increased from 6 to 16. It illustrated that the
combination of fac‐[Mn(phen)(CO)3Br] and ZnTPP in this system favored a high selectivity as well as a high
productivity in the catalytic reduction of CO2.
Group 3 was chosen as the research target to study the formation of the two products during the whole
reaction. Fig.6 shows the production of CO and formic acid with the irradiation time. It could be obviously seen
that formic acid generated quickly during the first one hour of the reaction (TONHCOOH quickly changed to 7 and
then 11), and then the production slowed down during the three hours irradiation (TONHCOOH ultimately reached
16). Additionally, a dramatic amount of CO was produced with TONCO=64, which was the main product in this
reaction. During this progress CO generated with a surprising speed in the first one hour and then slowed down
gradually until TONCO reached 64 during the experiment period.
According to the table above, there are two kinds of products detected in the reaction, in which the main
one is CO while the minor one is formic acid. Some mechanisms were reported to describe the solo product
formation process. For example, both manganese and rhenium compounds were reported to reduce CO2 into
formic acid when combined with protons in the first place, while leaded to the CO product if combined with CO2
firstly3d. In some other systems using manganese compounds, the mechanisms show that certain electron donors,
such as TEOA, could take part in reactions of Mn catalyst, to make a combination of proton and the catalyst.6 And
some Mn species existing in the system could control the product selectivity.14 Herein, an integrated
photocatalysis mechanism was proposed to discuss the CO2 conversion (shown in Scheme 1). As discussed in the
former part of this article, ZnTPP could successfully transfer the visible light energy into electron energy, which
transferred to the catalyst with the help of electron donor (TEA). Under the irradiation of visible light, the ground
state ZnTPP firstly turned into excited state (S1). This could be proved by the UV‐vis and fluorescence spectrum of
ZnTPP (Fig 3). And the long enough excited‐state life time of ZnTPP (τ=2 ns18) could allow the complex turn into a
lower‐energy but long‐lived T1 state (τ=1.5 ms).19 Thus [ZnTPP]‐ could be quenched more easily by TEA. [ZnTPP]‐
then transferred electrons to the catalyst and turns into the ground state to fulfill the reaction. (as shown in
Scheme 1)
The catalytic mechanism of fac‐[Mn(phen)(CO)3Br] could be summarized in two processes which were
illustrated in Scheme 1. For both processes, in the solvent, the Mn compound was firstly combined with the
solvent molecular to replace Br and got electrons from [ZnTPP‐] at the same time. This part of the reaction was

proved by the cyclic voltammograms of fac‐[Mn(phen)(CO)3Br]. According to Fig 1, it has been declared that there
are three reduction peaks in the fac‐[Mn(phen)(CO)3Br] cyclic voltammogram under CO2 condition, ‐1.2 V, ‐1.5 V
and ‐1.75 V vs. SCE. In a certain electrochemical reaction, fac‐[Mn(phen)(CO)3Br] firstly lost the Br‐ ion and then
reacted with MeCN, the solvent molecule. And then a Mn‐Mn dimer, [Mn(phen)(CO)3]2 was formed. This process
could be confirmed by the reduction peak at ‐1.5 V vs. SCE in the cyclic voltammetry of fac‐[Mn(phen)(CO)3Br].
The similar observation was reported and attributed to the formation of Mn‐Mn dimer. 9 The dimer quickly
received electrons from one‐electron reduced [ZnTPP‐] and then transformed to [Mn(phen)(CO)3]‐. According to
the comparison of the cyclic voltammetry, the peak current achieved a 1.88 times enhancement under CO2
compared with that in N2 atmosphere. It indicated that there were reactions between the manganese compound
and CO2. The existence of the metal dimer in the catalytic system was also confirmed by other reports.4d,9,14 And
for this dimer, there was only one reaction path, in which it reacted with protons to form a [Mn(phen)(CO)3H]
structure. Following that step, the acetonitrile molecules were then attached to the metal part (manganese) of
the catalyst, which also leaded to the addition of CO2 to replace the solvent molecule.
H2O, one kind of the weak Brønsted acid, also plays an important role to help with the catalytic reaction in
this process. As shown in many other metal carbonyl catalysts, water was a necessary reagent to acquire the
catalytic activity.20 In a manganese catalyst case, the addition of a weak Brønsted acid made the stabilization of
the manganese‐carbon dioxide intermediate through a protonation, which also promoted the break of one of the
C‐O bonds of CO2 to form CO.21 In Scheme 1, the proton from H2O was attached to the oxygen atom, which was
linked with the carbonyl in the molecule, to form a [Mn(phen)(CO)3(COOH)]+ structure (one of the C=C double
bond in the CO2 molecule already broken up in the previous step). A similar process was reported by Shakeri and
his co‐workers in the research of the photoreduction of CO2 with Re(I) complex.22 In the reported research, the
[C=O] fraction falls out to produce CO after the shedding of the [–O–H] fraction after receiving electrons in the
homogeneous system.
The formic acid formation could be attributed to the reaction between TEA and [Mn(phen)(CO)3]‐ which has
no linkage with Br atom as well as the acetonitrile molecule. The concentration of TEA reached 0.1M in the
homogeneous system, which was far more than that of ZnTPP (0.05 mM). Thus although TEA was added as the
electron donor and sacrificial reagent to quench the excited ZnTPP, its concentration was high enough for the its
reaction with [Mn(phen)(CO)3]‐ species even the reaction rate is slow. The electrons transferred from TEA in aid of
the bonding between [Mn(phen)(CO)3]‐ and protons.6 Then the CO2 molecules attached with the manganese
species to form a hybrid with a structure of [Mn(phen)(CO)3(COOH)]. Because this species was neutral, it was not
likely to react in the similar way as Process I. Finally, another proton combined with [Mn(phen)(CO)3(COOH)] to
fulfill the CO2 conversion and release formic acid.
Reusability
The reusability of the photocatalyst is also one of the most important factors for application. In order to
confirm the reusability of the photocatalyst system, Group 3 in Table 1 was chosen as the target to study the
reusability of the photocatalyst system. The catalyst solutions were irradiated for one hour and sampled for the
products determination. Then the solutions were refreshed with CO2 in dark condition. Four cycles of experiment
data were compared.
It could be seen from Fig 7 that after 1 hour of photoreduction, TONCO and TONHCOOH were detected as 53
and 11. And after four runs, TONCO and TONHCOOH were 41 and 7, respectively. Although the amount of the
products decreased slightly, the catalyst system kept the high product selection and yield for the CO2 conversion.
Conclusion
Abundant manganese phenanthroline complex was combined with the zinc porphyrin to perform the
efficient CO2 conversion under the visible light irradiation. The zinc porphyrin was acted as photosensitizer to
absorb the visible photon and induce the formation of reduced manganese complex in the presence of electron
donor. The ratio of manganese and zinc complexes in the solution affected the efficiency of CO2 reduction. The
TON of CO formation was 64 when the ratio of Mn/Zn was 2:1. The liquid product formic acid was achieved with
TON of 16 under the same conditions. In addition, the zinc porphyrin photosensitizer prevented the hydrogen
formation from protons in this system.
Acknowledgments
This work was supported by the Fundamental Research Funds for the Central Universities (No. 2015ZCQ‐HJ‐
02), the National Natural Science Foundation of China (No. 21373032 and 51278053), and grant‐in‐aid from Kochi
University of Technology.
Reference

1.
(a) Wang, W. H.; Himeda, Y.; Muckerman, J. T.; Manbeck, G. F.; Fujita, E., Chem. Rev. 2015, 115 (23), 12936‐
12973.; (b) Wu, B. H.; Guo, C. Y.; Zheng, N. F.; Xie, Z. X.; Stucky, G. D., J. Am. Chem. Soc. 2008, 130 (51), 17563‐
17567.; (c) Saveant, J. M.; Tard, C., J. Am. Chem. Soc. 2016, 138 (3), 1017‐1021.
2.
3.
4.
Inoue, T.; Fujishima, A.; Konishi, S.; Honda, K., Nature 1979, 277 (5698), 637‐638.
Hawecker, J.; Lehn, J. M.; Ziessel, R., J. Chem. Soc. Chem. Commun. 1983, 9, 536‐538.
(a) Bian, Z. Y.; Sumi, K.; Furue, M.; Sato, S.; Koike, K.; Ishitani, O., Inorg. Chem. 2008, 47 (23), 10801‐10803.;
(b) Takeda, H.; Koike, K.; Inoue, H.; Ishitani, O., J. Am. Chem. Soc. 2008, 130 (6), 2023‐2031.; (c) Andrade, G. A.;
Pistner, A. J.; Yap, G. P. A.; Lutterman, D. A.; Rosenthal, J., Acs Catal. 2013, 3 (8), 1685‐1692.; (d) Sampson, M. D.;
Froehlich, J. D.; Smieja, J. M.; Benson, E. E.; Sharp, I. D.; Kubiak, C. P., Energ. Environ. Sci. 2013, 6 (12), 3748‐3755.;
(e) Bian, Z. Y.; Wang, H.; Fu, W.; Li, L.; Ding, A. Z., Polyhedron 2012, 32, 78‐85.
5. Kuramochi, Y.; Itabashi, J.; Fukaya, K.; Enomoto, A.; Yoshida, M.; Ishida, H., Chem. Sci. 2015, 6 (5), 3063‐3074.
6.
7.
Fei, H. H.; Sampson, M. D.; Lee, Y.; Kubiak, C. P.; Cohen, S. M., Inorg. Chem. 2015, 54 (14), 6821‐6828.
Liu, Y.; Yang, Y.; Sun, Q.; Wang, Z.; Huang, B.; Dai, Y.; Qin, X.; Zhang, X., ACS Appl. Mater. Interfaces 2013, 5
(15), 7654‐7658.
8. Mukhopadhyay, T. K.; MacLean, N. L.; Gan, L.; Ashley, D. C.; Groy, T. L.; Baik, M. H.; Jones, A. K.; Troyitch, R. J.,
Inorg. Chem. 2015, 54 (9), 4475‐4482.
9.
Bourrez, M.; Molton, F.; Chardon‐Noblat, S.; Deronzier, A., Angew. Chem.‐Int. Edit. 2011, 50 (42), 9903‐9906.
10. Smieja, J. M.; Sampson, M. D.; Grice, K. A.; Benson, E. E.; Froehlich, J. D.; Kubiak, C. P., Inorg. Chem. 2013, 52
(5), 2484‐2491.
11. (a) Riplinger, C.; Sampson, M. D.; Ritzmann, A. M.; Kubiak, C. P.; Carter, E. A., J. Am. Chem.Soc. 2014, 136
(46), 16285‐16298.; (b) Riplinger, C.; Carter, E. A., Acs Catal. 2015, 5 (2), 900‐908.
12. Agarwal, J.; Shaw, T. W.; Stanton, C. J.,; Majetich, G. F.; Bocarsly, A. B.; Schaefer, H. F., Angew. Chem.‐Int. Edit.
2014, 53 (20), 5152‐5155.
13. Zeng, Q.; Tory, J.; Hartl, F., Organometallics 2014, 33 (18), 5002‐5008.
14. Takeda, H.; Koizumi, H.; Okamoto, K.; Ishitani, O., Chem. Commun. 2014, 50 (12), 1491‐1493.
15. Li, L. L.; Diau, E. W. G., Chem. Soc. Rev. 2013, 42 (1), 291‐304.
16. Amao, Y.; Watanabe, T., Appl. Catal. B Environ. 2009, 86 (3‐4), 109‐113.
17. (a) Pereira, C.; Ferreira, H. G.; Schultz, M. S., Inorg. Chimi. Acta 2005, 358 (13), 3735‐3744.; (b) Gamboa, M.;
Campos, M.; Torres, L. A., J Chem. Thermodyn. 2010, 42 (5), 666‐674.; (c) Adler, A. D.; Shergali, W.; Longo, F. R., J.
Am. Chem. Soc. 1964, 86 (15), 3145.
18. Ohno, O.; Kaizu, Y.; Kobayashi, H., J. Chem. Phys. 1985, 82 (4), 1779‐1787.
19. La Porte, N. T.; Moravec, D. B.; Hopkins, M. D., Proc. Nat. Acad. Sci. U.S.A. 2014, 111 (27), 9745‐9750.
20. (a) Arai, T.; Sato, S.; Uemura, K.; Morikawa, T.; Kajino, T.; Motohiro, T., Chem. Commun. 2010, 46 (37), 6944‐
6946.; (b) Christensen, P. A.; Hamnett, A.; Higgins, S. J.; Timney, J. A., J Electroanal. Chem. 1995, 395 (1‐2), 195‐
209.; (c) Koji Tanaka, D. O., Coord. Chem. Rev. 2002, 226 (1‐2), 221‐218.
21. Wong, K. Y.; Chung, W. H.; Lau, C. P., J Electroanal. Chem. 1998, 453 (1‐2), 161‐169.
22. Shakeri, J.; Farrokhpour, H.; Hadadzadeh, H.; Joshaghani, M., RSC Adv. 2015, 5 (51), 41125‐41134.

Fig 1 Cyclic voltammogram of fac‐[Mn(phen)(CO)3Br] in MeCN/H2O(v/v=20:1) bubbled by nitrogen (a) and
saturated with CO2 (b)
Fig 2 Cyclic Voltammetry of the ZnTPP under the condition of different current from 10mV/s to 2000mV/s
Fig 3 UV‐vis (a) and fluorescence (b) spectrum of ZnTPP

Fig 4 Fluorescence quenching data of ZnTPP with the concentration of TEA from 0 mM to 30 mM
Fig 5 The linear plot of the ZnTPP fluorescence ratio versus TEA concentration according to the Stern−Volmer
equation
Fig 6 The formation of CO (a) and formic acid (b) during the whole photoreaction of Group 3.

Fig 7 Reusability of the photocatalyst system under the solar irradiation.

Scheme 1 Proposed Reaction Mechanism

Table 1 Photocatalytic reactions
Group
Number
fac‐[Mn(phen)(CO)3Br]/ZnTPPa
TONb
CO
119
97
64
12
8
HCOOH
1
2
3
4
5
6
7
4:1
3:1
2:1
1:1
19
18
16
10
6
1:2
Only fac‐[Mn(phen)(CO)3Br]c
2
0
1
0
Only ZnTPPd
Irradiation with 500w xenon lamp conducted on a solution containing a Mn complex (0.5 mM), ZnTPP in a certain
concentration, TEA (0.1 M) in a MeCN‐Water (v/v=20:1) solution under a CO2 atmosphere. The reaction time in
this experiment was set to be 180 minutes.
a In this photoreaction, the concentration of fac‐[Mn(phen)(CO)3Br] and ZnTPP were set as the following part: 1.
fac‐[Mn(phen)(CO)3Br] 2mM, ZnTPP 0.5mM; 2. fac‐[Mn(phen)(CO)3Br] 1.5mM, ZnTPP 0.5mM; 3. fac‐
[Mn(phen)(CO)3Br] 0.5mM, ZnTPP 0.25mM; 4. fac‐[Mn(phen)(CO)3Br] 0.5mM, ZnTPP 0.5mM; 5. fac‐
[Mn(phen)(CO)3Br] 0.5mM, ZnTPP 1mM.
b According to the existing formula, TON = n(product)/n(catalyst).
c In this group, only Mn(phen)(CO)3Br was added into the system, with the concentration of 0.5mM.
d In this group, only ZnTPP was added into the system, with the concentration of 0.5mM.