Highly selective and durable photochemical co2 reduction by molecular mn(i) catalyst fixed on a particular dye-sensitized TiO2 Platform

Article abstract of DOI:10.1021/acscatal.8b03816

A Mn(I)-based hybrid system (OrgD-|TiO₂|-MnP) for photocatalytic CO₂ reduction is designed to be a coassembly of Mn(4,4′-Y₂-bpy)(CO)₃Br (MnP; Y = CH₂PO(OH)₂) and (E)-3-[5-(4-(diphenylamino)phenyl)-2,2′-bithiophen-2′-yl]-2-cyanoacrylic acid (OrgD) on TiO₂ semiconductor particles. The OrgD-|TiO₂|-MnP hybrid reveals persistent photocatalytic behavior, giving high turnover numbers and good product selectivity (HCOO^- versus CO). As a typical run, visible-light irradiation of the hybrid catalyst in the presence of 0.1 M electron donor (ED) and 0.001 M LiClO₄ persistently produced HCOO^- with a >99% selectivity accompanied by a trace amount of CO; the turnover number (TON_formate) reached ?250 after 23 h of irradiation. The product selectivity (HCOO^-/CO) was found to be controlled by changing the loading amount of MnP on the TiO₂ surface. In situ FTIR analysis of the hybrid during photocatalysis revealed that, at low Mn concentration, the Mn-H monomeric mechanism associated with HCOO^- formation is dominant, whereas at high Mn concentration, CO is formed via a Mn-Mn dimer mechanism.

Full text of DOI:10.1021/acscatal.8b03816

Subscriber access provided by Iowa State University | Library
Article
Highly Selective and Durable Photochemical CO2 Reduction by
Molecular Mn(I) Catalyst Fixed on Particular Dye-Sensitized TiO2 Platform
Sung-Jun Woo, Sunghan Choi, So-Yoen Kim, Pil Soo Kim, Ju Hyoung
Jo, Chul Hoon Kim, Ho-Jin Son, Chyongjin Pac, and Sang Ook Kang
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b03816 • Publication Date (Web): 31 Jan 2019
Downloaded from http://pubs.acs.org on February 1, 2019
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 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
Highly Selective and Durable Photochemical CO2
Reduction by Molecular Mn(I) Catalyst Fixed on
Particular Dye-Sensitized TiO2 Platform
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
Sung-Jun Woo, Sunghan Choi, So-Yoen Kim, Pil Soo Kim, Ju Hyoung Jo, Chul Hoon Kim, Ho-
Jin Son,* Chyongjin Pac,* and Sang Ook Kang*
Department of Advanced Materials Chemistry, Korea University, Sejong 30019, Korea
KEYWORDS: CO2 to formate conversion, organic-inorganic hybrid systems, molecular catalyst
TiO2 immobilization, heterogeneous catalysis, photocatalysis.
ABSTRACT. A Mn(I)-based hybrid system (OrgD-|TiO2|-MnP) for photocatalytic CO2 reduction
is designed to be a co-assembly of Mn(4,4′-Y2-bpy)(CO)3Br (MnP; Y = CH2PO(OH)2) and (E)-3-
[5-(4-(diphenylamino)phenyl)-2,2′-bithiophen-2′-yl]-2-cyanoacrylic acid (OrgD) on TiO2
semiconductor particles. The OrgD-|TiO2|-MnP hybrid reveals persistent photocatalytic behavior,
giving high turnover numbers and good product selectivity (HCOO– versus CO). As a typical run,
visible-light irradiation of the hybrid catalyst in the presence of 0.1 M electron donor (ED) and
0.001 M LiClO4 persistently produced HCOO– with a >99% selectivity accompanied by a trace
amount of CO; the turnover number (TONformate) reached ~250 after 23 h irradiation. The product
selectivity (HCOO–/CO) was found to be controlled by changing the loading amount of MnP on
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 44
1
2
3
4
5
6
7
8
9
the TiO2 surface. In-situ FTIR analysis of the hybrid during photocatalysis revealed that at low Mn
concentration, the Mn‒H monomeric mechanism associated with HCOO– formation is dominant,
whereas at high Mn concentration, CO is formed via an Mn‒Mn dimer mechanism.
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
INTRODUCTION
The combustion of fossil fuels has resulted in the rapid rise of CO2 concentration (the most
abundant greenhouse gas) in the atmosphere, consequently causing global warming. Thus, the
photocatalytic reduction of CO2 to useful C1 feedstock (i.e., CO, formate, CH3OH, and CH4) is
considered to be the best option to resolve the on-going environmental hazard caused by carbon-
based fuels.[1-8] Strenuous efforts have been made to design an efficient catalyst for carbon dioxide
conversion. Among several strategies, synthetic homogeneous[9,10] and heterogeneous[10,11]
catalytic systems based on molecular transition metal complexes have received much attention as
a potential candidate for photo- and electro-catalytic CO2 reduction, because of the easy control of
their redox potentials through ligand modifications, facile coordination of CO2 to the metal center
in the reduction state(s), and oxidation-state jump of the metal center allowing multi-electron
reduction processes for CO2 reductions.[12-18] In many cases reported so far, Ru[19-21]- and Re[10,21-
24]-based complexes with 2,2′-bipyridine (bpy) or related ligands have been extensively studied for
the electrochemical and photocatalytic reductions of CO2, mainly to CO and HCOO–. However,
the major drawbacks of Ru- and Re-catalysts have very often been pointed out concerning the cost
of such noble metals, which should limit their practical usage at large scale. In this regard, first-
row transition metal complexes based on Mn(I), Fe(II), Co(II), Ni(II), and Cu(II) have been
proposed as potential candidates for the practical catalysis of CO2 reduction.[9,10,17,18,25-46] Among
them, a molecular complex incorporating Mn, (bpy-R)MnICO3Br (bpy-R = 2,2′-bipyridine-based
ACS Paragon Plus Environment
2

Page 3 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
ligand), has recently received much attention, because of the structural/mechanistic similarity to
related Re(I) complexes, which have been intensively studied due to the selective and efficient
formation of CO by the photo- and electro-reduction of CO2.[28,47-50] However, compared with the
Re(I) complex analogues, the Mn(I) complexes show lower product selectivity (formation of both
CO and HCOO–) and lower durability under uncontrolled conditions. It is suggested that HCOO–
is formed via an Mn–H intermediate after one-electron reduction of the Mn complex in
competition with CO formation via the so-called dimeric mechanism involving the formation of
(R-bpy)(CO)3Mn0‒Mn0(CO)3(bpy-R). It has been also known that the Mn–Mn dimer formation
causes a lowering of the catalytic activity of the Mn catalyst due to the participation of two
competitive catalytic pathways, as well as an increase of overpotential for the two electron
reduction to [(bpy-R)Mn(CO)3]–.[28,47,51,52] For this reason, possible strategies to avoid the dimeric
catalytic pathway should be required for the selective and efficient CO2 reduction by the Mn(I)-
based complexes. Cohen and Kubiak have demonstrated that the incorporation of Mn catalyst in
metal-organic frameworks (MOFs) prevents dimerization of the singly-reduced Mn complex to
result in the selective photocatalytic formation of formate.[53] In a similar strategy, Fontecave and
Mougel have also reported that immobilization of Mn catalyst on bipyridine functionalities of
periodic mesoporous organosilicas (PMO) greatly enhances the photoreduction activities of CO2
into CO and HCOO– with inhibited bimolecular processes by site isolation of Mn catalysts.[54] In
electrochemical CO2 reduction, Reisner and co-workers have reported that the adsorption of a Mn
catalyst via the pyrene anchoring group on a multiwall carbon nanotube (MWCNT) electrode
enhances the electrocatalytic activity and durability with variable product selectivity that is
sensitive to the concentration of the Mn complex: at low Mn catalyst loading, HCOO– was the
major product (efficiently alleviating the Mn‒Mn dimeric mechanism with a spatial isolation of
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
3

ACS Catalysis
Page 4 of 44
1
2
3
4
5
6
7
8
9
Mn catalyst), whereas at high Mn catalyst loadings, CO was the main CO2 reduction product.[55]
Kubiak and co-workers have also reported that the dimer formation is effectively suppressed
through incorporating a bulky moiety to the coordinating bipyridine ligand.[47] While recent
advances for tuning the catalytic pathway in Mn(I)-based CO2 reduction continue to emerge (in
the field of electrochemistry), limited reports appear on the photosensitizer-driven reaction of Mn(I)
complex with controlled product selectivity. Therefore, a new strategy that focuses on improving
product selectivity with efficient/durable activity is highly needed in photocatalytic CO2 reduction
based on Mn(I) complex.
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
Recently, the immobilization of a molecular catalyst on a semiconductor (typically TiO2) for
photocatalytic CO2 reduction has emerged as an effective tool; 1) to improve catalytic durability
by suppressing unwanted interactions between long-lived reactive species of the catalyst and 2) to
enhance the catalytic efficiency by smooth electron supply to the anchored catalyst through TiO2
after electron injection from an excited-state photosensitizer.[56-61] This strategy can be expected
to apply to spatial isolation of Mn catalyst on the surface of TiO2 that might allow the dominant
participation of the monomeric mechanism associated with high product selectivity in the
photocatalytic CO2 reduction. Scheme 1 illustrates the present reaction system with the key
components and overall photochemical reaction pathways. Herein, we present the selective and
sustainable photocatalytic behavior of the Mn-based hybrid catalyst prepared by the co-
immobilization of the Mn(I) catalyst (MnP) and the visible-light photosensitizer (OrgD) on TiO2
particles
(OrgD-|TiO2|-MnP);
OrgD
is
(E)-2-cyano-3-(5′-(5″-(p-
(diphenylamino)phenyl)thiophen-2″-yl)thiophen-2′-yl)-acrylic acid, MnP is fac-[Mn(4,4′-Y2-
bpy)(CO)3Br] (Y = CH2PO(OH)2), and the sacrificial electron donor (ED) is 1,3-dimethyl-2-
phenyl-1,3-dihydrobenzimidazole (BIH). This approach offers details of the photocatalytic CO2
ACS Paragon Plus Environment
4

Page 5 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
reduction activities of the Mn complex-based ternary hybrid system. It was found that different
transient intermediates can be probed by in-situ FTIR analysis of the hybrid system. This analysis
showed that low surface loading of MnP leads to high selectivity toward HCOO– production.
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
Scheme 1. Conceptual representation for the reaction system and components used in this study.
EXPERIMENTAL SECTION
Synthesis
and
characterization.
The
organic
sensitizer
(OrgD),[62]
4,4′-
Bis(diethoxyphosphorylmethyl)-2,2’-bipyridine,[59,60] and sacrificial electron donor (BIH)[63] were
prepared according to the method reported in previous work. The Mn complex (MnP) was
synthesized by modifying synthetic protocols described by previous reports, details of which are
described below (see Scheme 2 and the experimental section).
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 44
1
2
3
4
5
6
7
8
9
fac-[Mn(4,4′-bis-(diethoxyphosphorylmethyl)-2,2′-bipyridine)(CO)3Br] (MnPE). A Et2O
solution (50 mL) containing a mixture of 4,4′-Bis(diethoxyphosphorylmethyl)-2,2′-bipyridine
(0.524 g, 1.148 mmol) and Mn(CO)5Br (0.292 g, 1.043 mmol) was refluxed for 4 h and then cooled
to room temperature. The reaction mixture was further maintained at –20 °C using ice/NaCl bath
for 30 min to give a yellow solid powder. The residue was purified by silica gel chromatography
using dichloromethane/methanol (20:1 v/v) as eluent to give MnPE as a light yellow solid. Yield
45% (0.701 g). 1H NMR (300 MHz, DMF-d7) δ 9.27 (d, J = 5.4 Hz, 2H), 8.65 (s, 2H), 7.81 (d, J
= 5.4 Hz, 2H), 4.08 (m, 8H) 3.62 (d, J = 22.8 Hz, 4H), 1.22 (t, J = 6.9 Hz, 12H). ESI-MS (m/z):
calcd. for C23H30BrMnN2O9P2 [M]: 673.9990, found [M–H]–: 672.8930. Anal. Calcd (%) for
C23H30BrMnN2O9P2: C, 40.91; H, 4.48; N, 4.15. Found (%): C, 40.79; H, 4.45; N, 4.17.
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
fac-[Mn(4,4′-bis-(dihydroxyphosphorylmethyl)-2,2′-bipyridine)(CO)3Br]
(MnP).
Bromotrimethylsilane (TMSBr) (6.9 mL, 52.6 mmol) was added dropwise to a dry CH3Cl solution
(30 mL) containing a mixture of MnPE (0.500 g, 0.74 mmol). The reaction mixture was refluxed
for 24 h. To this solution were added methanol (17 mL) and stirred for 3 h, and then yellow solid
precipitated was collected by filtration. After evaporation under reduced pressure, the remaining
solid was recrystallized from CH2Cl2/MeOH/n-Hexane mixture solvent to give the product (MnP)
as a light yellow solid. Yield 67% (0.300 g). 1H NMR (300 MHz, DMSO-d6) δ 9.07 (d, J = 5.4 Hz,
2H), 8.42 (s, 2H), 7.60 (d, J = 5.4 Hz, 2H), 3.28 (d, J = 22.2 Hz, 4H). ESI-MS (m/z): calcd. for
C15H14BrMnN2O9P2 [M]: 561.8738, found [M–H]–: 560.9206. Anal. Calcd (%) for
C15H14BrMnN2O9P2: C, 32.00; H, 2.51; N, 4.98. Found (%): C, 31.85; H, 2.48; N, 5.01.
Crystal Structure Determination. Fine crystal of MnPE obtained from a CH2Cl2/n-hexane
solution was sealed in glass capillaries under argon and mounted on a diffractometer. The
preliminary examination and data collection were performed by Bruker SMART APEX CCD X-
ACS Paragon Plus Environment
6

Page 7 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
ray diffractometer at the Korea Basic Science Institute (KBSI Seoul center) equipped with a sealed-
tube X-ray source (50 kV × 1 mA) using graphite monochromated Mo Kα radiation (λ = 0.71073
Å). The preliminary unit-cell constants were determined using a set of 50 narrow frame (0.30° in
ω) scans. The double pass method of scanning was used to exclude noise. The collected frames
were integrated using an orientation matrix determined from the narrow-frame scans. The SMART
software package was used for data collection, and SAINT was used for frame integration.[64] The
final cell constants were determined through global refinement of the xyz centroids of the
reflections harvested from the entire data set. Structure solution and refinement were performed
using the SHELXTL-PLUS software package.[64] Crystallographic data was deposited with the
Cambridge Crystallographic Data Centre as supplementary publications (CCDC-1824333
(MnPE)). Additional crystallographic data are available in the Tables S1–S3 of the Supporting
Information (SI).
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
Preparation of Mn-immobilized OrgD-|TiO2|-MnP Hybrid Catalyst: Commercially available
TiO2 particles (Hombikat UV-100) were thoroughly washed with distilled water, ultrasonically
treated in water, separated by centrifugation, and then dried in an oven under N2. The specific
Brunauer-Emmett-Teller (BET) surface areas were determined to be >250 m2/g. The TiO2 particles
(0.1 g) were stirred overnight in an acetonitrile/tert-butanol solution of OrgD (15 μmol) in the dark
and then subjected to centrifugation. The collected solids were washed with the solvent and then
dried in an oven under N2. For the preparation of OrgD-|TiO2|-MnP sample (Figure S6 in the SI),
the Dye-deposited TiO2 powders (0.1 g) were dispersed into an acetonitrile/tert-butanol solution
of MnP (1 to 20 μmol) and allowed to stand overnight under stirring. The collected solids were
separated by centrifugation, washed with the solvent and then dried in an oven (70 °C) and stored
under N2 in the dark. The complete loading of MnP was confirmed by absorbance comparison
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 44
1
2
3
4
5
6
7
8
9
before and after each adsorption step of MnP: the supernatant, which is separated after
centrifugation of the MnP–treated suspensions, showed the negligible absorption intensity of
MnP.
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
Photocatalyzed CO2 Reduction. Suspensions of OrgD-|TiO2|-MnP particles (10 mg with
photosensitizer (OrgD) and reduction catalyst (MnP) in 3 mL distilled DMF containing BIH (0.1
M) were placed in a Pyrex cell (1 cm pass length; 6.0 mL total volume), bubbled with CO2 for 30
min, and sealed with a septum. A series of samples were set on a homemade merry-go-round
apparatus and then irradiated under magnetic stirring with a LED lamp (λ >400 nm, 60 W, model
Fc-6051, Cree Inc.). Homogeneous-solution photocatalysis was carried out in 3 mL of DMF
2+
solutions containing Ru(bpy)3 (0.05 mM), MnPE (0.05 mM), and BIH (0.1 M). The gaseous
products (CO and H2) accumulated in the overhead space of the cell were determined by gas
chromatography (HP6890A GC equipped with a TCD detector) using a SUPELCO CarboxenTM
1010 PLOT fused silica capillary column (5 Å). The liquid phase of the irradiated samples was
subjected to high-performance liquid chromatography (HPLC) analysis using a Waters 515 pump,
a Waters 486 UV detector operated at UV 210 nm, a RSpak KC-811 Column (Shodex), and 0.05
M H3PO4 aqueous solution eluent.
Apparent Quantum Yield Measurement. The apparent quantum yield Φ(CO) for HCOO–
production was determined for the OrgD-|TiO2|-MnP suspensions, a band-pass filter (420−450
nm) was used to isolate the 436 nm light from the emission light of a high-pressure mercury lamp
(1000 W, model 6171, Newport Corporation), and the incident light flux was determined using a
0.2 M ferrioxalate actinometer solution.[65] The apparent quantum yield (AQY) of HCOO–
formation for the hybrid system in the presence of 0.001 M LiClO4 was determined in a linear
time–conversion region. As defined in Eq. 1, the measured AQY is not the real quantum yield with
ACS Paragon Plus Environment
8

Page 9 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
scientific definition, but represents a relative estimate for the utilization efficiency of
photogenerated electrons into HCOO– formation with respect to the incident light intensity,
because the number of photons absorbed by the sensitizer cannot be exactly determined. The
relatively low AQY appears to arise, at least in part, from poor light harvesting by the dye, due to
extensive light scattering in the particle dispersion system.
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
amount of formate generated per unit time
=
x 100
AQY(%)
(Eq. 1)
number of incident photons per unit time
In-situ Infrared Spectroscopy. A homemade in-situ FTIR spectroscopy tool is designed and
made for analyzing a chemical intermediate of Mn(I) molecular catalyst immobilized on OrgD-
|TiO2 particles during photocatalysis. Figure S19 and S20 of the SI show the detailed description
of the accessory parts designed for this measurement. Three mirror reflectance accessories were
installed in the Nicolet 6700 FTIR spectrometer (from Thermo Electron Corporation). After PTFE
gasket type ring was placed in the cell base body, CaF2 window was inserted, and Teflon spacer
was placed on the CaF2 window. In order to conduct the IR study under an environment similar to
the homogeneous-solution photocatalytic reaction, a CO2-saturated DMF solution containing
2+
MnPE (0.05 mM), Ru(bpy)3 (0.05 mM), and BIH (0.1 M) was injected through a silicon tube
into the IR beam-penetrating active cell equipped with a CaF2 window. In the case of IR
measurements for heterogeneous ternary system (OrgD-|TiO2|-MnP), a mesoporous TiO2 film was
prepared on a CaF2 plate and then sequentially treated with solutions of OrgD and MnP for the
immobilization of the components on the TiO2 film. The composite plate was set at the center of
active cell filled with CO2-saturated DMF solution containing 0.1 M BIH. In this FTIR study, the
loading amount of MnP was set at (~0.67, 3.3, or ~6.6) μmol, ~(7, 33, or 66) times higher than
that (0.1 μmol per 10 mg TiO2 particles) employed in the photocatalyzed CO2 reduction experiment.
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 44
1
2
3
4
5
6
7
8
9
The loaded Mn concentrations are determined based on 1.5 mg mesoporous TiO2 deposited onto
CaF2 window. Note that the 0.1 μmol loading amount of MnP (similar to actual photocatalytic
condition) is too low to allow the detection of the relevant IR absorption bands. After purging the
DMF solution containing 0.1 M BIH with CO2 gas in a dark condition for 30 min, the 0.05 mL of
purged BIH solution was injected into the active spacer (0.1 mm pass length; 0.03 mL total volume)
of the in-situ FTIR tool, through which the IR-beam can pass with simultaneous LED irradiation
(see Figure S19 of the SI). The change of IR peaks was traced over time ((0 to 60) min), under
LED irradiation (>400 nm).
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
RESULTS AND DISCUSSION
Preparation and characterization of MnP
Scheme 2 describes the synthetic procedure of MnP. Refluxing of 4,4′-Y2-bipyridine (Y = -
CH2PO(OEt)2) with [MnBr(CO)5] in diethyl ether gave MnPE in 45% yield, which was
subsequently treated with bromotrimethylsilane (TMSBr) to produce the desired MnP in 67%
yield. The prepared Mn complexes (MnPE and MnP) were characterized by 1H NMR and high-
resolution ESI-MS spectroscopy (see the Experimental Section, and Figure S1–S4 of the SI). The
solid-state structure of MnPE was confirmed by X-ray crystallography (Figure S5 in the SI). Table
S1 of the SI summarizes the crystal data. Table S2 and S3 of the SI list the selected bond distance
and angles. MnPE showed an Mn1‒N1 bond length of 2.042(2) Å, an Mn1‒Br1 bond length of
2.515(6) Å, and a torsion angle for N1‒C5‒C6‒N2 in the 2,2′-bipyridine of ‒10.00(3)°. The values
are within the ranges of normal metal-carbon/nitrogen bond and torsion angle found in other
bipyridyl Re analogues.[66] In comparison, [ReBr(4,4′-diethylphosphonate-2,2′-bipyridine)(CO)3]
ACS Paragon Plus Environment
10

Page 11 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
displayed a Re1‒N1 bond length of 2.613(7) Å, Re1‒Br1 bond length of 2.629(10) Å, and a torsion
angle (N1‒C1‒C2‒N2) in the 2,2′-bipyridine of 6.3(11)°. Overall, MnPE has metal‒N and metal‒
C bond lengths that are shorter than those of bipyridyl Re complex due to the relatively small
radius of first row (3d) Mn metal ion compared to third row (5d) Re metal.
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
Scheme 2. Synthetic Route to fac-[Mn(4,4′-bis-(dihydroxyphosphorylmethyl)-2,2′-
bipyridine)(CO)3Br] (MnP)[a]
O
P
O
P
O
P
O
O
HO
HO
O
O
Cl
CO
Mn
CO
Mn
Br
CO
CO
CO
CO
N
N
N
N
(i)
(iii)
(ii)
O
O
O
P
N
N
N
N
O
HO
HO
O
P
P
Cl
Br
O
O
MnPE
MnP
[a]Conditions: (i) triethylphosphite (10 equiv), 24 h, 140 C
̊ , N2 (88% yield); (ii) [Mn(CO)5Br] (0.9 equiv), Et2O, 4 h,
reflux, N2 (45% yield); (iii) TMSBr, CHCl3, 24 h, reflux, N2 (67% yield).
Integration and characterization of Dye-|TiO2|-MnP ternary hybrid
The successful anchoring of the functional components (OrgD and MnP) on TiO2 nanoparticles
(Hombikat UV-100, Huntsman with Brunauer-Emmett-Teller (BET) surface areas of greater than
250 m2/g) was confirmed by the IR absorption peaks characteristic of the CO ligands of MnP
((2031, 1936, and 1919) cm‒1) (Figure 1) and by diffuse-reflectance spectra (DRS) showing the
absorption maximum of OrgD at 450 nm (Figure S8 of the SI). The supernatant obtained by
centrifugation after each adsorption step of OrgD and MnP was transparent, a result confirming a
high-efficiency loading of the components supplied.
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 44
1
2
3
4
5
21.5 µM MnP
Supernatant after adsorption of MnP on TiO2
(a)
(b)
Bare TiO2
TiO2|-MnP
6
7
8
0.1
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
+ TiO2
+ MnP
0.0
2400
2200
2000
1800
1600
1400
400
500
600
Wavenumber / cm−1
Wavelength / nm
Figure 1. IR spectra of bare TiO2 and TiO2|-MnP in KBr discs (sample: KBr ≈ 1:100) (a, left) and
comparable absorption spectra of DMF solution of MnP before (black line) and after (blue line)
adsorption process of MnP with TiO2 particles (b, right).
Photocatalytic activity
Dispersed solution of OrgD-|TiO2|-MnP (10 mg) in 3 mL of CO2-saturated DMF containing 0.1
M BIH was irradiated at ≥400 nm using an LED lamp (Cree Inc., 60 W). Table 1 summarizes the
compositions of HCOO–, CO, and H2 observed under various conditions, including turnover
numbers (TONs, mol of the products/mol of MnP) and net formation amounts in μmol. First, the
effects of loaded amounts of OrgD and MnP on the photocatalytic behavior were investigated to
identify the optimal conditions. First of all, it should be noted that HCOO– is the exclusive product
when the loaded amount of MnP is lower than 0.1 μmol. The photocatalytic activity for the HCOO–
formation increased with the increase in the loaded amount of MnP on 10 mg OrgD-|TiO2|
particles from (0.025 to 0.1) μmol (entries 1 to 3) to give a maximum at 0.1 µmol and then
decreased with further increase of loaded MnP amount (entry 4 to 8). This is reminiscent of similar
ACS Paragon Plus Environment
12

Page 13 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
observations in the case of a Re(I)-based ternary analogue (OrgD-|TiO2|-Re(I) catalyst).[59] On the
other hand, further increase in the amount of the MnP on TiO2 surface of (0.1 to 3.2) μmol results
in a substantial decrease of HCOO– production accompanied by significant CO production (Figure
2 and Figure S14 of the SI). In turn, the optimum loading amount of OrgD was determined to be
1.484 µmol on 10 mg TiO2|-MnP(0.1 µmol) (entries 9 to 11). Figure 3 shows typical plots of TON
versus irradiation time for the formation of HCOO–, CO, and H2 under optimum conditions.
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
Table 1. Results of Visible-Light-Driven CO2 Reduction with Mn Complex-Based Hybrid TiO2
Catalyst in Different Conditions[a]
HCOO–
μmol
CO
μmol
H2
μmol
<0.1
<0.1
tirr.
[h]
entry OrgD(μmol)-|TiO2|-MnP(μmol)
TON
TON
<1
<1
1
OrgD(1.48)-|TiO2|-MnP(0.025) 26 60.0 ± 4.0 1.5 ± 0.1
<0.1
<0.1
<0.1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
OrgD(1.48)-|TiO2|-MnP(0.05)
OrgD(1.48)-|TiO2|-MnP(0.10)
OrgD(1.48)-|TiO2|-MnP(0.20)
OrgD(1.48)-|TiO2|-MnP(0.40)
OrgD(1.48)-|TiO2|-MnP(0.80)
OrgD(1.48)-|TiO2|-MnP(1.60)
OrgD(1.48)-|TiO2|-MnP(3.20)
OrgD(2.97)-|TiO2|-MnP(0.1)
OrgD(1.48)-|TiO2|-MnP(0.1)
OrgD(0.74)-|TiO2|-MnP(0.1)
OrgD(1.48)-|TiO2
26 50.0 ± 2.0 2.5 ± 0.1
28 143 ± 11
14.3 ± 1.1 <1
<0.1
28 55.0 ± 6.0 11.0 ± 1.2 12.0 ± 2.5 2.4 ± 0.5
28 26.3 ± 3.0 10.5 ± 1.2 9.8 ± 1.0
28 10.4 ± 2.0 8.3 ± 1.6 6.5 ± 1.0
0.15 ± 0.01
0.27 ± 0.02
0.26 ± 0.02
0.17 ± 0.01
3.9 ± 0.4
5.2 ± 0.8
7.8 ± 0.9
28 6.3 ± 0.6
28 4.0 ± 0.3
31 92.5
10.0 ± 1.0 4.9 ± 0.6
12.8 ± 0.9 4.4 ± 0.4
14.2 ± 1.2 0.60 ± 0.05
9.25
15.1
4.2
<1
<1
<1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
31 151
30 42
33
10
10
10
[b]
[b]
[b]
[b]
-
-
-
-
-
-
-
[b]
[b]
[b]
[b]
[b]
[b]
[b]
OrgD(1.48)-|TiO2|-MnP(0.1)[c]
OrgD(1.48)-|TiO2|-MnP(0.1)[d]
OrgD(1.48)-|TiO2|-MnP(0.1)[e]
OrgD(1.48)-|ZrO2|-MnP(0.1)
OrgD(1.48)-|TiO2|-MnP(0.1)[f]
Ru(bpy)32+ + MnPE[g]
-
-
-
-
[b]
[b]
[b]
[b]
-
-
-
-
[b]
[b]
[b]
[b]
-
-
-
-
[b]
[b]
10 7.4
0.74
-
-
<0.1
[b]
[b]
[b]
20 165 ± 20
10 86
16.5 ± 2.0
12.9
-
-
-
52.6
7.9
1.17
[a]The following standard conditions were employed: 0.1 M BIH and 10 mg of TiO2 are commonly used as sacrificial electron
donor and immobilizer of functional components (MnP and OrgD); the hybrids were modified with various amounts of MnP of
(0.025 to 3.20) μmol and OrgD of (0.74 to 2.97) μmol. The total volume of the solution or suspension was 3 mL. All photocatalytic
reactions are performed in CO2-saturated DMF at greater than 400 nm emitted from a light-emitting diode (LED) lamp (60 W, Cree
Inc.). [b]Not detected. [c]Under dark condition. [d]Ar bubbling (no CO2 atmosphere). [e]No BIH. [f]DMA (N,N-dimethylacetamide) as
solvent. [g]Ru(bpy)32+ (0.05 mM) + MnPE (0.05 mM) in the presence of 0.1 M BIH.
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 44
1
2
3
4
5
6
7
8
20
15
10
5
H2
(a)
(b) 150
CO
CO
HCOO−
HCOO−
120
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
90
60
30
0
µ
0
0.1
0.2
0.8
1.6
3.2
[MnP] / µm0.4ol per 10 mg TiO2
0.1
0.2
0.4
0.8
1.6
3.2
[MnP] / µmol per 10 mg TiO2
Figure 2. Dependencies of the net formation amounts (μmol) (a) and TON (b) of the products
(H2/CO/HCOO–) on the loading amounts of MnP on 10 mg of OrgD(1.48 μmol)-|TiO2 particles;
irradiation of 10 mg of OrgD-|TiO2|-MnP in 3 mL DMF containing 0.1 M BIH for 28 h.
16
HCOO−
CO
H2
12
µ
8
4
0
0
5
10
15
20
25
30
tirr. / h
Figure 3. Temporal plots for photocatalyzed CO2 reduction by OrgD(1.48 μmol)-|TiO2|-MnP(0.1
μmol) in CO2-saturated DMF containing 0.1 M BIH under irradiation at ≥400 nm using LED (60
W, Cree Inc.).
ACS Paragon Plus Environment
14

Page 15 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
Some control experiments were undertaken (entries 12 to 17 in Table 1). The OrgD-|TiO2| or
|TiO2|-MnP half hybrid prepared without immobilization of either OrgD or MnP showed little
photocatalytic activity for the CO2 reduction, indicating that the co-assembly of both OrgD and
MnP on TiO2 is essential for the efficient photoreduction of CO2. In the dark or in the absence of
sacrificial electron donor (BIH), HCOO– was formed only in a trace amount or was not detected
at all. It was confirmed that an OrgD-|ZrO2|-MnP hybrid prepared by using ZrO2 particles in place
of TiO2 reveals very low catalytic activity, a result that demonstrates the essential role of TiO2 in
the effective electron transport from photoexcited OrgD to MnP; the high conduction band level
of ZrO2, 1.3 eV higher than TiO2,[67] causes a very low probability of electron injection from photo-
excited OrgD into ZrO2 which consequently results in the poor formation of reduced MnP species
(Figure S12 of the SI). Another control experiment was performed by using N,N-
dimethylacetamide (DMA) solvent in place of DMF in order to confirm whether or not the
hydrolysis of DMF would participate in the HCOO– formation. Irradiation of suspensions of OrgD-
|TiO2|-MnP in 3 mL DMA containing 0.1 M BIH gave (165 ± 20) µmol of HCOO– for 20 h, an
amount comparable to those observed for DMF solvent, while acetic acid (the hydrolysis product
of DMA) was not detected. Therefore, it is evident that the photocatalytic CO2 reduction is the
dominant consequence for the HCOO– formation with little participation of hydrolysis of DMF in
the present ternary system (Table 1). The linear plot of HCOO– formation in Figure 3 suggests that
both OrgD and MnP should persistently work without significant consumption within the
irradiation time investigated. This was verified by FTIR spectra of the hybrid catalyst taken before
or after 10 h photolysis (Figure S15 of the SI), which revealed no discernible intensity change in
the CO peaks of MnP. An isotope tracer experiment was carried out for this Mn-based ternary
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
15

ACS Catalysis
Page 16 of 44
1
2
3
4
5
6
7
8
9
13
13
system in CO2–saturated DMF-d7 solvent containing 0.1 M BIH. The C NMR spectrum of
irradiated ternary sample shows a sharp signal at 167.6 ppm corresponding to the resonance of
H13COO–, indicating that the produced formate originated from injected CO2 gas (Figure S16 of
the SI).
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
2+
For comparison, a conventional homogeneous system using Ru(bpy)3 as visible-light
photosensitizer and MnPE as CO2 reduction catalyst was investigated in CO2-saturated DMF
solution containing 0.1 M BIH under irradiation at >400 nm (entry 18 in Table 1). In contrast to
the ternary hybrid system, the photocatalytic activity showed a leveling-off tendency at an early
stage and no further increase after 8 h, producing 12.9 μmol of HCOO– and 7.9 μmol of CO after
10 h irradiation. This result is consistent with most reported studies on homogeneous Mn(I)-based
photochemical CO2 reduction (see Figure S13 of the SI).[49,50,68] The comparable formation of both
HCOO– and CO (HCOO–/CO = 1.63/1) implies that the catalytic reaction in homogeneous solution
proceeds via both of the monomeric and dimeric mechanistic pathways, as reported. The relatively
early leveling-off tendency is thought to derive from losses of the original components resulting
from unwanted interactions between various long-lived reactive species and intermediates, which
lead to as yet fully undefined deformation (i.e., CO2-coordinated Mn complexes and one-electron
reduced molecule).[69-71]
Additive Effects
Since Brønsted acids and metal cations are known to exert a positive effect on Re(I)- and Mn(I)-
based CO2 reduction,[56,60] the photocatalytic performance of this hybrid system was further tested
in the presence of such additives as lithium perchlorate (LiClO4), trifluoroethanol (TFE), and
ACS Paragon Plus Environment
16

Page 17 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
methanol. Figure 4 shows that variation of LiClO4 concentration in DMF solvent of (0 to 10‒1) M
led to a dramatic change of TONformate for OrgD-|TiO2|-MnP. The maximum effect was obtained
with 0.001 M LiClO4; TONformate reaches >250 for a prolonged irradiation time (23 h). The
apparent quantum yield (AQY) of HCOO– at 436 nm under the optimum condition was determined
to be (1.05 ± 0.05) × 10–2. In the case of TFE, the photocatalytic conversion efficiency increased
with increasing concentration of TFE up to 0.2 M, but no further increase of activity was observed
at higher TFE concentration (0.3 M) (see entries 4 to 6 in Table 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
with 0.001 M LiClO4
No additive
25
250
200
(a)
200
20
16
12
8
20
with 0.001 M LiClO
25
160
4
µ
µ
15
150
120
10
100
No additive
80
50
5
20
0
0
0
40
4
10-5
10-4
10-3
10-2
10-1
5
10
15
20
25
tirr. / h
log([LiClO4] / M)
Figure 4. Plots of HCOO– formation versus irradiation time for OrgD-|TiO2|-MnP in the presence
and absence of 0.001 M LiClO4 (a) and dependence of TONformate on concentration of LiClO4 for
10 mg of OrgD-|TiO2|-MnP irradiated for 23 h in CO2-saturated DMF containing 0.1 M BIH (b);
irradiation at λ ≥420 nm at 293 K.
Table 2. Additive Effects on the Photocatalytic CO2 Reduction by OrgD-|TiO2|-MnP[a]
b
[ ]
tirr.
[h]
HCOO–
CO
μmol
H2
entry
additives in DMF
TON
72.1
μmol
7.21
TON
<1
<1
μmol
1.44
2.24
<0.1
<0.1
1
2
MeOH
MeOH
2.5 vol%
5 vol%
21.6
21.6
102.7
10.27
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 44
1
2
3
4
5
6
7
8
9
<1
<1
<1
<1
<1
<1
<1
<1
<1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
3
4
5
6
7
8
9
10
11
MeOH
TFE
TFE
10 vol%
0.1 M
0.2 M
21.6
80.7
194.5
178
160.7
48.4
8.07
19.45
17.8
16.07
4.84
5.59
0.35
0.39
1.26
2.11
0.43
<0.1
<0.1
<0.1
21.6
21.6
21.6
23
23
23
TFE
0.3 M
LiClO4
LiClO4
LiClO4
LiClO4
LiClO4
0.1 M
10–2 M
10–3 M
10–4 M
10–5 M
41.4
4.14
218.8
128.8
114.4
21.88
12.88
11.44
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
23
23
[a]The following standard conditions were employed: 0.1 M BIH and 10 mg of TiO2 are commonly used as sacrificial electron
donor and immobilizer of 0.1 µmol MnP and 1.48 µmol OrgD, respectively. The total volume of the solution or suspension was 3
mL. All photocatalytic reactions are performed in CO2-saturated DMF at greater than 400 nm emitted from a light-emitting diode
(LED) lamp (60 W, Cree Inc.). [b]Irradiation time (hour).
In-situ FTIR characterization
Equations 2 and 3 in Scheme 3 illustrates a possible mechanism for homogeneous Mn(I)-based
photocatalysis; a key intermediate for the CO formation is the Mn–Mn dimer complex generated
after one-electron reduction of the Mn(I) complex (Eq. 3) and another one for the HCOO–
formation is the Mn–H complex generated by sequential two electron reduction of the Mn(I)
complex followed by protonation (Eq. 2).[28,47-50]
Scheme 3. Proposed Mechanism for HCOO– or CO Production in Homogeneous Photocatalysis
by L(CO)3MnBr complex (L: bpy-R); PS = photosensitizer.
−
+ CO2
Mn-H]0
HCOO
− from PS
- Br−
− from PS
e
+ H+
Mn]-
e
Mn]0
L(CO)3MnBr
(Eq. 2)
L(CO)3MnBr
[L(CO)3
[L(CO)3
[L(CO)3
−
+ Br-
2
1
Mn H
+ CO2
Monomeric catalytic pathway
- CO
2
hv
Dimerization
L(CO)3Mn(CO2
)
e−
+
, H
+ H+
+ H
2O
CO
+ Br−
− from PS
e
1/2 [L(CO)3Mn]2
Mn-COOH]0
[L(CO)3
L(CO)3MnBr
(Eq. 3)
−
Mn Mn dimer
Mn(COOH)
+ CO2
Dimeric catalytic pathway
ACS Paragon Plus Environment
18

Page 19 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
In order to elucidate the sharp differences from the selective formation of HCOO– in the present
ternary system from the concomitant formation of HCOO– and CO in homogeneous systems,
comparative in-situ FTIR measurements were performed with a homemade in-situ FTIR tool for
both homogeneous and ternary hybrid systems; in the latter case, a particular CaF2 plate coated
with a mesoporous TiO2 film was used (see experimental section and Figures S19 and S20 of the
SI for details of the in-situ FTIR tool). In the case of homogeneous sample, a CO2-saturated DMF
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
2+
solution containing MnPE (0.05 mM), Ru(bpy)3 (0.05 mM), and BIH (0.1 M) was injected
through a silicon tube into the IR beam-penetrating active cell equipped with a CaF2 window. FTIR
spectra were recorded under the successive LED illumination of (0 to 50) min and differential
spectra were extracted by comparing them with the reference spectrum taken at time zero. Figure
5 shows the differential FTIR spectra in the range of 1800 ~ 2100 cm–1, revealing the rapid
appearance of new absorptions accompanied by the negative changes at (1905, 1926, and 2018)
cm–1 characteristic of the MnPE CO ligand absorptions. Moreover, the three new peaks at (1844,
1870, and 1968) cm‒1 are almost identical to the reported absorption bands of the Mn–Mn
dimer.[52,73] The coincident growth of the absorption bands for the dimer was further confirmed by
the generalized 2D-correlation spectroscopy. In the synchronous spectra obtained by the 2D-
correlation spectral analysis, four cross-peaks at (1844, 1870, 1968, and 1990) cm–1 were resolved
with the same sign (+ or –), indicating those are developed together under the external perturbation.
The additional peak at around 1990 cm–1 can be attributed to the CO vibration peak of the Mn–H
intermediate, because the absorption bands of Mn–H have been reported to appear at (1880 and
1990) cm–1.[74] The disappearance of the other peak at 1880 cm–1 may be due to the interference
arising from the neighboring peak at 1905 cm–1. Please note that the unknown weak absorption at
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 44
1
2
3
4
5
6
7
8
9
1975 cm–1 was also resolved in the 2D-correlation analysis, which may be attributable to the dimer
because its lifetime is comparable to that of the dimer bands (non-zero amplitude at 50 min. in
Figure 5a). The present results are in line with those reported by other workers,[49,68] indicating
that the competitive pathways via the Mn–H and Mn–Mn dimer intermediates (Eqs. 2 and 3 in
Scheme 3) proceed in the present homogeneous system. This means that our in-situ FTIR method
might be effectively applicable to the ternary system.
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
CO
PO3H2
OC
N
Mn
H
N
OC
H2O3P
H2O3P
PO3H2
N
H2O3P
H2O3P
CO
Mn
CO
N
CO
Mn
CO
(a)
N
CO
N
CO
Mn–H
Mn–Mn dimer
0.004
0.000
-0.004
-0.008
-0.012
-0.016
tirr. / min
0 min
0.02 min
0.15 min
0.34 min
0.95 min
1 min
10 min
30 min
50 min
∆
1800
1850
1900
1950
2000
2050
2100
Wavenumber / cm−1
Figure 5. Differential IR spectral changes taken during the photocatalytic CO2 reduction for a
2+
CO2-saturated DMF solution of MnP (0.5 mM), Ru(bpy)3 (0.5 mM), and BIH (0.1 M); LED
irradiation at λ ≥400 nm at 293 K.
For FTIR measurements of the ternary system, a mesoporous TiO2 film was prepared on a CaF2
plate and then sequentially treated with solutions of OrgD and MnP for the immobilization of the
ACS Paragon Plus Environment
20

Page 21 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
components on the TiO2 film.[71] The composite plate was set at the center of active cell filled with
CO2-saturated DMF solution containing 0.1 M BIH. In this FTIR study, the loaded amount of
MnP was set to be 0.67 μmol (sample 1), 3.3 μmol (sample 2), or 6.6 μmol (sample 3) per 10 mg
TiO2 ~(7, 33, or 66) times higher than that (0.1 μmol) employed optimally in the photocatalyzed
CO2 reduction experiment; the 0.07 μmol loaded amount of MnP is too low to allow the detection
of the relevant IR absorption bands (see Figure S21 in the SI). Figure 6 shows the FTIR spectral
changes following irradiation time. The spectral features are relatively complex due to optical
effects of the mesoporous TiO2 film and presumably due to absorptions of injected electrons in the
conduction band and trap sites of TiO2.[75,76] However, the growth of both the Mn–H peaks at (1997
and 1885) cm‒1 and the Mn–Mn dimer bands at (1976, 1873, and 1831) cm‒1 can be clearly seen,[55]
suggesting that both the monomeric and dimeric mechanisms should proceed for the two samples
with the high loading amounts of MnP. A notable point is that the relative intensities of the Mn–
Mn absorptions are significantly different among samples 1–3; the increase of loaded Mn amount
rises the band intensities at 1976, 1873, and 1831 cm–1 representing Mn–Mn dimer species, which
is supported by the good match between previously reported[25,55] and observed band frequencies
(see Figure 6d and 7). This observation strongly suggests that the two processes should be
controlled by the loading amount of MnP on TiO2 surface. Under the low loading conditions
employed for the photocatalytic CO2 reduction experiment (0.1 μmol MnP), the Mn–H mechanism
(Eq. 2) should dominantly proceed, whereas the very high loading amount of MnP (≥1 μmol)
might allow competitive participation of the Mn–Mn dimer process (Eq. 3). These results are in
good agreement with the Mn concentration-sensitive product selectivity observed in the above
photocatalysis (Figure 2 and Table 1). These trends are also in line with the results reported by
Reisner et al., such that the Mn surface concentration modulates the mole fraction ratio of CO/
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
21

ACS Catalysis
Page 22 of 44
1
2
3
4
5
HCOO– in the total products (CO, HCOO–, and H2) electrochemically produced from Mn-
immobilized MWCNT electrode.[55]
6
7
8
9
It should be noted that our ternary system exhibits a relatively much higher product selectivity
with the retarded hydrogen production in aprotic solvent: the product selectivity for HCOO– is
>99% at less than 0.1 μmol of MnP surface concentration.
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
(a) 0.15
(b)
sample 1
sample 2
0.3
0.2
0.1
0.0
tirr. / min
t
irr. / min
2.2
2.0
0.10
2.3
2.0
1.8
1.6
1.9
1.7
1.6
1.4
1.2
0.05
1.4
1.2
1.0
1.0
0.0
0.0
2100 2050 2000 1950 1900 1850 1800
2100 2050 2000 1950 1900 1850 1800
Wavenumber / cm−1
Wavenumber / cm−1
0.5
(d)
(c)
sample 3
0.5
tirr. / min
0.4
0.3
0.2
0.1
0.4
0.3
0.2
0.1
tirr. / min
sample 3
2.2
2.0
1.9
1.7
1.6
2.2
2.2
sample 2
sample 1
1.5
1.3
1.2
2.3
1.1
0.0
2100 2050
2000 1950 1900
1850 1800
2100 2050 2000 1950 1900 1850 1800
Wavenumber / cm−1
Wavenumber / cm−1
Figure 6. FTIR spectral changes following irradiation time for mesoporous films of OrgD-|TiO2|-
MnP on a CaF2 plate with loaded amounts of MnP corresponding to 0.67 μmol per 10 mg TiO2
(a, sample 1), 3.3 μmol per 10 mg TiO2 (b, sample 2), and 6.6 μmol per 10 mg TiO2 (c, sample 3);
the sample plate dipped in 0.03 mL CO2-saturated DMF solution containing 0.1 M BIH was
irradiated at ≥400 nm using LED lamp at 293 K. Mn concentration-dependent FTIR spectra of
ACS Paragon Plus Environment
22

Page 23 of 44
ACS Catalysis
1
2
3
4
5
6
mesoporous Mn hybrid films (sample 1–3) after ~2 min LED irradiation (d). Orange dot lines
located at 1975 and 1830 cm–1 indicate the Mn–Mn dimer.
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
MnP concentration
per 10 mg TiO2
6.6 µmol (sample 3)
∆A 0.2
3.3 µmol (sample 2)
0.67 µmol (sample 1)
2100 2050 2000 1950 1900 1850 1800
Wavenumber / cm−1
Figure 7. Comparison of FTIR difference spectra for mesoporous MnP-bound dye-sensitized
TiO2 films (OrgD-|TiO2|-MnP) with variation of loaded MnP amount corresponding to 6.6 μmol
per 10 mg TiO2 (top, sample 3), 3.3 μmol per 10 mg TiO2 (middle, sample 2), and 0.67 μmol per
10 mg TiO2 (bottom, sample 1). Two bands at 1997 and 1885 cm–1 (green numbers) indicate the
peak positions that are assigned to Mn–H intermediate. The overall band intensities at 1976, 1873,
and 1831 cm–1 representing Mn–Mn dimer species rise with increase of Mn loading amount.
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 44
1
2
3
4
Discussions
5
6
7
8
9
Scheme 4 illustrates possible mechanistic pathways working in the present photocatalytic CO2
reduction. Upon visible-light absorption of anchored OrgD, fast electron injection from excited-
state OrgD (1OrgD*) into the conduction band (CB) of TiO2 semiconductor (~400 fs) occurs much
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
1
faster than the decay of OrgD* (~2 ns),[62,77] probably due to strong electronic coupling of the
anchored OrgD with TiO2. Injected electrons (TiO2(e–)) deposited in TiO2 nanoparticles should be
sequentially transferred to the MnP center through CB and by way of trap-to-trap hopping,[78]
generating such TiO2-bound key intermediates as [L(CO)3Mn]0 (TiO2-bound 1) formed after the
loss of Br– from MnP•– and [L(CO)3Mn–H] (TiO2-bound Mn–H) formed by the protonation of
[L(CO)3Mn]–1 (TiO2-bound 2); the reaction of TiO2-bound Mn–H with CO2 results in the major
CO2 reduction catalysis to produce HCOO–. In this photocatalysis, BIH serves as the sacrificial
reducing agent for the recovery of OrgD by reducing the oxidized organic dye (OrgD•+) formed
by the electron injection, completing the photocatalytic cycle for CO2 reduction.
The electron transfer rate from TiO2(e–) to MnP should be determined by the relationship
between the reduction potential of MnP and CB of TiO2. Based on the reported flat-band potentials
(Vfb) of TiO2 nanoparticle films (‒(1.5 ~ 1.97) V versus saturated calomel electrode (SCE) in
organic solvent depending on the contents of water or alkali metal ions added),[56,79-81] it can be
easily predicted that, under our experimental conditions, TiO2(e–) should be smoothly transferred
to MnP (‒1.14 V vs SCE) and to [L(CO)3Mn]0 (‒1.48 V vs SCE) (Figure S17 of the SI).[82] On
the other hand, the Mn–Mn dimer intermediate supposed to be responsible for the CO formation
has a reduction potential of –1.88 V versus SCE,[49] which is similar to or significantly more
negative than Vfb of TiO2 depending on reaction conditions. Therefore, in the present
photocatalytic CO2 reduction, TiO2(e–) should be preferentially directed toward the monomeric
ACS Paragon Plus Environment
24

Page 25 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
catalytic pathway, even under the assumption that the Mn–Mn dimer would be formed on the TiO2
surface. Such hypothesis could be supported by the early leveling off tendency of CO production
(<3 h), which is commonly observed in all photolysis with surface loading control of MnP (>0.2
μmol per 10 mg TiO2) (Figure S14 of the SI).
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
The in-situ FTIR analysis revealed the effectiveness in detection of catalytically relevant Mn‒
Mn dimer and Mn–H formed during photocatalysis. The observed intensity ratios of the IR
absorptions characteristic of the two key intermediates show an apparent parallelism with the
photocatalytic product selectivity (HCOO– via the Mn-hydride (Mn–H) and CO via the Mn‒Mn
dimer) depending on the loading amounts of MnP. At low surface loading of MnP, TiO2-bound 1
species should exclusively exist as the isolated form without significant dimer formation due to
the sparse occupation of MnP on the TiO2 surface. The proton-coupled transfer of TiO2(e–) to
isolated TiO2-bound 1 should yield TiO2-bound Mn–H. The insertion of CO2 into the Mn–H bond
generates Mn(I)–OC(O)H complex, the direct precursor for the catalytic formation of HCOO–.
On the other hand, at high surface loading of MnP, CO was substantially formed. Although the
mechanistic origin of the CO formation is still unclear, the first possible assumption is the
competitive participation of the Mn–Mn dimer route (Eq. 3 in Scheme 3) that would occur via the
local dimerization of TiO2-bound 1 to the TiO2-bound Mn–Mn dimer.[47-50] However, the
dimerization of TiO2-bound 1 would apparently be unlikely to occur, according to simple
consideration of the low averaged coverage of MnP even at the maximum loading (0.256 at 3.2
μmol of MnP on 10 mg TiO2, area of components/2.5 m2).[83] A possible speculation is that the
complex and rough surface morphology of TiO2 particle agglomerate would allow local
assembling of MnP on specific surface sites to extents enough for the dimerization of TiO2-bound
1 to occur under high loading conditions, as reported for the aggregation of dyes on mesoporous
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 44
1
2
3
4
5
6
7
8
9
TiO2 films.[55,84] Alternatively, TiO2(e–) should be mostly consumed for the generation of TiO2-
bound 1 at high loading amounts of MnP so that the available amount of TiO2(e–) would become
relatively short for the second electron transfer to TiO2-bound 1 associated with the HCOO–
formation. As a consequence, a considerable portion of TiO2-bound 1 might stay without further
reduction, being “long” lived enough to undergo CO2 binding to an Mn(COOH) intermediate,
which is thought to be responsible for the CO formation. Another possibility would be desorption
of MnP from the TiO2 surface into solution that should allow the participation of the solution
processes shown in Scheme 3. The higher the loading amount of MnP, the greater the net amount
of liberated MnP in solution would be. The contribution of the solution mechanism involving
liberated MnP in solution would become apparent at high loading amount of MnP, while the net
amount of liberated MnP should be negligible under low loading conditions. Besides, the CO
production at high Mn concentration might be also originated from Mn–Mn dimer formation that
would be generated from a sequence of events such as temporary desorption, re-orientation, and
then re-adsorption of weakly physisorbed Mn(I) complex, as recently suggested by Reisner and
Walsh.[85,86]
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
On the other hand, as BI•, radical species generated from one electron oxidation of BIH with its
proton transfer, has been shown to have a higher reduction potential (Eox = –1.66 V versus SCE in
CH3CN),[63] the presence of BI• might be another possible electron source for the reduction of Mn(I)
catalytic site (Ered = ‒1.14 V vs SCE) as suggested by Fontecave et al.[54] However, such route
should be minor because most of BI• is consumed for the regeneration of oxidized photosensitizer
(1OrgD•+) generated with fast oxidative quenching of OrgD on TiO2 semiconductor possessing its
low-lying shallow trap sites in this hybrid. The poor CO2 reduction yields observed when ZrO2 is
used as immobilizer clearly evidence that the TiO2(e–) is the major electron source for the chemical
ACS Paragon Plus Environment
26

Page 27 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
CO2 reduction process of Mn(I) catalyst (Scheme 4) rather than direct electron supply by radical
BI• species. Regarding the resource of protons needed for the formation of Mn–H intermediates in
photolysis with the addition of LiClO4 or absence of additives, the protons could be supplied from
both first oxidation of the added BIH and water pre-adsorbed on TiO2 surface.
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
Scheme 4. Schematic representation and proposed MnP concentration-/intermediate-dependent
mechanistic pathways in photochemical CO2 reduction of the ternary hybrid system (OrgD-|TiO2|-
MnP).
We now believe that the relatively low AQY of the present Mn hybrid system probably arises
from poor light harvesting by the loaded dye due to extensive light scattering in the particular TiO2
sample and the complicated chemical process involving the slow coordination of CO2 molecules
to the TiO2-bound 17-electron species (pentacoordinate [Mn(L)(CO)3]•) compared to other
homogeneous analogues.[13,87] A further performance enhancement can be achieved by introducing
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 44
1
2
3
4
5
6
7
8
9
a transparent sol solution of TiO2 immobilized with a photosensitizer and a reduction catalyst,[88-
90] which are defined as an efficient antenna function without a loss of light harvesting by scattering
effect. Alternatively, the introduction of long-wavelength sensitizer could be also a viable way of
overcoming such an optical drawback by dispersions of the hybrid particles in solution since the
penetration of red and near-IR light is generally superior than blue and green light having a
relatively high frequency wave. This work is more amenable to mechanistic study on standard dye-
sensitized TiO2 platform than full report involving the performance optimization.
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 these investigations, we employed a rational/directed co-immobilization of functional
components on the semiconductor as a method to ensure the product selectivity and durability in
photocatalytic CO2 reduction using earth-abundant Mn(I) complex as electrocatalyst. The
enhancement of conversion efficiency of the resulting Mn-immobilized TiO2 hybrid (OrgD-|TiO2|-
MnP) with almost absolute product selectivity (>99%) can be understood in terms of the
stabilization of molecular Mn catalyst by TiO2 immobilization and the economical catalytic
process directed by suppression of Mn–Mn dimer formation. Under optimum condition with
cationic additive (0.001 M LiClO4) and low surface loading of MnP (0.1 μmol per 10 mg TiO2),
TONformate reached ~250 over a prolonged irradiation time (23 h) with high product selectivity
(>99%). To the best of our knowledge, this surpasses not only the related homogeneous system,
but also the other heterogenized Mn photocatalytic systems. The in-situ FTIR
spectrophotochemical analysis to scrutinize the mechanistic pathway demonstrated that the
chemical process (via either monomeric or Mn‒Mn dimeric intermediates) can be tuned by
ACS Paragon Plus Environment
28

Page 29 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
changing the loading amount of TiO2-bound MnP. Overall, this study is of scientific significance
because many research groups have focused on the polypyridyl complexes of earth-abundant and
economical first-row transition metal such as manganese for catalytic CO2 reduction. Further
improvement of durability towards introducing other alternative first-row transition metal-based
reduction catalysts and photo-stable visible-light sensitizers for durable and selective CO2
reduction is currently underway.
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
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI:
xx.xxxx/acscatal.xxxxxx.
Characterization data of MnPE and MnP (1H NMR and ESI-Mass spectra), X-ray crystallographic
data of MnPE, results of photophysical (UV-Vis absorption data and DRS) and electrochemical
data (cyclic voltammogram) of components, 13C isotope labeled CO2 NMR experiment, and plots
of CO and formate formation versus time and in-situ FTIR spectra for hybrid catalysts
Corresponding Author
*E-mail: hjson@korea.ac.kr.
*E-mail: jjpac@korea.ac.kr.
*E-mail: sangok@korea.ac.kr.
Notes
The authors declare no competing financial interest.
ACS Paragon Plus Environment
29