Attaching Cobalt Corroles onto Carbon Nanotubes: Verification of Four-Electron Oxygen Reduction by Mononuclear Cobalt Complexes with Significantly Improved Efficiency

Article abstract of DOI:10.1021/acscatal.9b00213

Cobalt complexes have been extensively explored in catalyzing the oxygen reduction reaction (ORR), which is the cathode reaction in fuel cells. Although they show high activities, mononuclear Co complexes typically mediate the 2e reduction of O₂ to H₂O₂, and two Co sites are generally required to catalyze the 4e reduction of O₂ to H₂O. Herein we report the significantly improved efficiency of covalently grafted Co corroles on carbon nanotubes (CNTs) for the 4e ORR. Azide-containing Co corroles can be attached to alkyne-modified CNTs via azide-alkyne cycloaddition. This attachment can avoid the formation of dimeric face-to-face Co corroles. The resulted hybrid catalyzes the 4e ORR in 0.5 M H₂SO₄ aqueous solutions with a half-wave potential at 0.78 V versus reversible hydrogen electrode (RHE). This performance makes this hybrid one of the most efficient Co-based molecular ORR electrocatalysts. Control studies using Co corrole analogues loaded on CNTs via noncovalent interactions give ORR half-wave potentials at 0.68-0.61 V versus RHE. This work is significant in demonstrating that with proper covalent bond interactions to CNTs mononuclear Co corroles can become intrinsically active for the 4e ORR with significantly improved efficiency.

Full text of DOI:10.1021/acscatal.9b00213

Subscriber access provided by BUFFALO STATE
Article
Attaching Cobalt Corroles onto Carbon Nanotubes: Verification
of Four-Electron Oxygen Reduction by Mononuclear
Cobalt Complexes with Significantly Improved Efficiency
Jia Meng, Haitao Lei, Xialiang Li, Jing Qi, Wei Zhang, and Rui Cao
ACS Catal., Just Accepted Manuscript • Publication Date (Web): 10 Apr 2019
Downloaded from http://pubs.acs.org on April 10, 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
Attaching Cobalt Corroles onto Carbon Nanotubes:
Verification of Four-Electron Oxygen Reduction by
Mononuclear Cobalt Complexes with Significantly
Improved Efficiency
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
Jia Meng,‡ Haitao Lei,‡ Xialiang Li, Jing Qi, Wei Zhang, and Rui Cao*
Key Laboratory of Applied Surface and Colloid Chemistry, Ministry of Education,
School of Chemistry and Chemical Engineering, Shaanxi Normal University, Xi’an
710119, China
‡These authors contributed equally to this work.
*Correspondence E-mail: ruicao@ruc.edu.cn
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 44
1
2
3
4
Abstract
5
6
7
8
Cobalt complexes have been extensively explored in catalyzing the oxygen
reduction reaction (ORR), which is the cathode reaction in fuel cells. Although they
show high activities, mononuclear Co complexes typically mediate the 2e reduction of
O2 to H2O2, and two Co sites are generally required to catalyze the 4e reduction of O2
to H2O. Herein we report the significantly improved efficiency of covalently grafted
Co corroles on carbon nanotubes (CNTs) for the 4e ORR. Azide-containing Co
corroles can be attached to alkyne-modified CNTs via azide-alkyne cycloaddition.
This attachment can avoid the formation of dimeric face-to-face Co corroles. The
resulted hybrid catalyzes the 4e ORR in 0.5 M H2SO4 aqueous solutions with a
half-wave potential at 0.78 V vs reversible hydrogen electrode (RHE). This
performance makes this hybrid one of the most efficient Co-based molecular ORR
electrocatalysts. Control studies using Co corrole analogues loaded on CNTs via
noncovalent interactions give ORR half-wave potentials at 0.68-0.61 V vs RHE. This
work is significant to demonstrate that with proper covalent bond interactions to
CNTs, mononuclear Co corroles can become intrinsically active for the 4e ORR with
significantly improved efficiency.
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
Keywords: oxygen reduction, cobalt corrole, covalent immobilization, mononuclear,
electrocatalysis.
2
ACS Paragon Plus Environment

Page 3 of 44
ACS Catalysis
1
2
3
4
5
Introduction
6
7
8
9
Developing efficient and robust electrocatalysts for O2 reduction has attracted
considerable interests because this process is sluggish in kinetics but is important as a
cathode reaction in fuel cells.1-3 Platinum and its alloys are very active for ORR,4 but
their uses are limited due to the very low natural abundance and high cost of this
noble metal. As an alternative, extensive efforts have been made in the last decade to
make ORR electrocatalysts derived from earth-abundant and cheap transition metal
elements.5-17 For mononuclear transition metal catalysts, O2 can be reduced by four
electrons to form H2O or by two electrons to form H2O2 (Scheme 1).3 In general, late
transition metal complexes typically catalyze ORR via the 2e pathway because it is
challenging for the peroxo intermediates M−OOH to undergo heterolytic O−O bond
cleavage, leading to the generation of terminal metal-oxo species that are high in
energy. However, the 4e ORR can still be accomplished by late transition metal
complexes through the formation of dinuclear peroxo intermediates.
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
Cobalt complexes have been widely studied and are shown to be highly active
for ORR.18-30 As inspired by cytochrome c oxidases (CcOs) with heme active sites for
the 4e ORR,31 a variety of Co porphyrins and derivatives have been investigated as
ORR catalysts. Mononuclear and cofacial dinuclear Co porphyrin catalysts are
presented by Anson, Fukuzumi, Kadish and others.32-45 Mononuclear Co corroles and
Co complexes of porphyrin-corrole dyads and biscorroles are studied by Kadish,
Gross, Cao and others.46-55 Although porphyrin and corrole ligands are negatively
charged and thus are effective in stabilizing metal ions in their high valent states to
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 44
1
2
3
4
5
6
7
8
9
benefit the O−O bond cleavage, dimeric Co-based forms are still usually required for
the 4e ORR.56-59 Xia and co-workers performed theoretical studies and suggested that
the highest-occupied 3d orbital of Co porphyrins has a relatively low energy level,
leading to the weak activity for the 4e ORR.60 It is necessary to note that mononuclear
Co complexes can form face-to-face dimeric Co species in both solution and solid
states to catalyze the 4e ORR. This phenomenon has been demonstrated by Girault
and co-workers, who studied ORR using Co porphyrins at liquid-liquid interfaces.61-63
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. The reduction of O2 catalyzed by mononuclear transition metal complexes,
showing the possible reaction pathways for the 4e and 2e ORR.
Synthetic modeling studies of CcOs provide fundamental knowledge for the
rational design of efficient ORR catalysts.64,65 Collman and co-workers revealed that
rapid electron transfer to heme was essential for CcOs to catalyze the 4e reduction of
O2.66 In nature, this rapid electron transfer is realized by involving extra redox centers
(i.e., a tyrosine unit and a Cu ion) located in the vicinity of the heme unit. For
biomimetic systems, Dey and co-workers reported that Fe porphyrins with ferrocene
units attached at the second coordination sphere were able to catalyze the 4e ORR.67,68
4
ACS Paragon Plus Environment

Page 5 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
The Dey’s group also showed the switch between 2e vs 4e ORR by controlling the
electron transfer rate from the electrode to the Fe porphyrin unit through different
linkers.69 On the other hand, Collman and co-workers revealed that hydrogen bonding
interactions were able to stabilize the O2-adduct and to assist the O−O bond cleavage,
leading to much improved 4e ORR selectivity.70 Similar effects of hydrogen bonding
interactions were reported by Mayer, Dey and others with the use of Fe porphyrin
model complexes.71-76 Nocera and co-workers demonstrated that intramolecular
acid/base groups of hangman Co porphyrins and corroles could lead to increased 4e
ORR activity.77,78 These results not only shed light on understanding the reaction
mechanism of CcOs for O2 reduction but also provide valuable information for
designing new efficient ORR catalysts.
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
Herein we report the high efficiency of covalently grafted mononuclear Co
corroles on CNTs for the 4e reduction of O2. Recently, we reported the covalent
immobilization of azide-containing Co corroles onto alkyne-modified CNTs through
azide-alkyne cycloaddition.79 CNTs are chosen as the support for electrocatalysis due
to their good electrical conductivities and large surface areas.80 With the use of this
method, we can graft Co corrole 1 onto CNTs (Figure 1). The resulted hybrid CNT-1
is highly efficient in catalyzing the 4e ORR in 0.5 M H2SO4 aqueous solutions with a
half-wave potential at 0.78 V vs RHE. This performance is significantly improved in
comparison with Co corrole analogues immobilized on CNTs through noncovalent
interactions and is also outstanding as compared with previously reported Co-based
molecular ORR catalysts. Because such covalently attached Co corroles cannot form
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 44
1
2
3
4
5
6
7
8
9
dimeric face-to-face Co species, this work provides direct experimental evidence that
mononuclear Co corroles can be intrinsically active for the 4e ORR. As demonstrated
above, extensive efforts have been made to prove and improve the 4e ORR activity of
Fe porphyrins. However, the 4e ORR activity of mononuclear Co complexes has been
poorly understood, although a large number of Co-based catalysts have been reported
to be able to catalyze ORR with high efficiency. This work is therefore significant to
show the significantly improved 4e ORR activity of mononuclear Co corroles and will
bring new insights into the ORR with complexes of late transition metal elements.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. (a) Molecular structures of Co corroles 1, 2, and 3. (b) Illustration showing
the preparation of CNT-1, CNT-2, and CNT-3.
6
ACS Paragon Plus Environment

Page 7 of 44
ACS Catalysis
1
2
3
4
5
Experimental Section
6
7
General methods and materials
8
9
Materials that are sensitive to air and moisture were manipulated under N2
using standard Schlenk line techniques. All reagents were commercially available and
were used directly unless otherwise noted. Dry solvents, including dichloromethane,
tetrahydrofuran, acetonitrile, and dimethylformamide were used in this work. Co
corroles were synthesized according to the method we reported previously.79 All
aqueous solutions were prepared freshly using the Milli-Q water. Multi-walled CNTs
(>95% purity, < 8 nm outside diameter, 3 nm intside diameter) were commercially
available. MALDI-TOF mass spectra were obtained using Brüker MAXIS. UV-vis
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
spectra were measured using a Hitachi U-3310 spectrophotometer. H NMR spectra
were acquired on a Brüker spectrometer. Transmission electron microscopy (TEM)
images were obtained using JEOL JEM-2100 with a 200-kV accelerating voltage.
Hitachi SU8020 cold-emission field emission scanning electron microscope (FE-SEM)
was used to study the sample morphologies with an accelerating voltage of 1 kV.
Energy dispersive X-ray (EDX) spectra were measured from three areas of each
sample to ensure the chemical homogeneity. X-ray photoelectron spectroscopy (XPS)
data were collected by ESCALab 220i-XL electron spectrometer from VG Scientific
using 300 W Al Kα radiation. The C 1s peak at 284.6 eV arising from adventitious
hydrocarbon was used for the correction of binding energies. All samples used were
carefully prepared via centrifugation and re-dispersion.
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 44
1
2
3
4
5
Synthesis of Co corrole 1
6
7
8
9
To the solution of 5,15-bis-(pentafluorophenyl)-10-(4)-(1-azido)phenylcorrole
(19 mg, 0.025 mmol) and anhydrous NaOAc (30 mg, 0.36 mmol) in 10 mL methanol,
were added Co(OAc)24H2O (30 mg, 0.12 mmol) and triphenylphosphine (PPh3, 50
mg, 0.19 mmol). The resulted mixture was stirred for 1.5 h at room temperature, and
the solvent was evaporated. The residue was then subjected to silica chromatography
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
(petroleum ether : dichloromethane = 5 : 1) to afford purple solids (22 mg, 85%). H
NMR (400 MHz, CDCl3): δ = 8.69 (d, J = 4.4 Hz, 2H), 8.34 (brs, 2H), 8.21 (d, J = 4.4
Hz, 2H), 8.01 (d, J = 4.4 Hz, 2H), 7.76-7.69 (m, 2H), 7.57-7.49 (m, 2H), 7.03 (t, J =
7.3 Hz, 3H), 6.69 (t, J = 6.6 Hz, 6H), 4.68 (s, 6H) (Figure 2a). HRMS: calcd for
C55H28CoF10N5P [1-N2+H+]+, 1038.1249; found, 1038.1245 (Figure 2b).
Preparation of CNTs modified with 4-ethynylaniline
Sonication was performed to disperse CNTs to avoid any agglomeration and
sedimentation. CNTs (30 mg) were mixed with 4-ethynylaniline (307 mg, 2.6 mmol)
and sonicated together in 70 mL of hydrochloric acid solution (0.5 M). Sodium nitrite
(1.5 equivalent to 4-ethynylaniline) dissolved in 15 mL of deionized water was then
added to generate the diazonium salt from 4-ethynylaniline. The addition of iron
powder (500 mg) could reduce the diazonium salt to generate the corresponding aryl
radical. Generated aryl radicals can be grafted onto available surfaces, and successive
radical substitutions finally yielded dehydrobenzene coatings on CNTs. The crude
products were stirred in excessive hydrochloric acid to remove the residue iron, and
8
ACS Paragon Plus Environment

Page 9 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
were washed with ethanol, acetone, and water for several times until no Fe content
was detected in both the washing solution and the hybrid. The modified CNTs can be
obtained after drying at room temperature.
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
Caution! The addition of iron powder to the hydrochloric acid solution is potentially
explosive and should be handled with care and in small amounts.
Preparation of CNT-1 through the azide-alkyne cycloaddition
CNTs (10 mg) modified with 4-ethynylaniline were mixed with 5 mg of Co
corrole 1 in 5 mL of dimethylformamide, and the mixture was stirred at 85 °C
overnight. The suspension was kept in a sonication bath for 1 h, and the crude CNT-1
was collected by centrifugation. The solids were re-suspended in dichloromethane to
remove simply adsorbed complex 1 on CNTs, and then collected by centrifugation
again, and were dried in the dark at room temperature. The Co content in the hybrid
CNT-1 was determined by using inductively coupled plasma mass spectrometry
(ICP-MS, Bruker aurora). CNT-1 was immersed in a concentrated high purity HNO3
solution under ultrasonication to dissolve the Co species. The HNO3 solution was
filtered and was then diluted into 20 mL for ICP-MS measurements. The Co ICP-MS
standards (AccuStandard) with a concentration of 50, 100, 200, 300, 400, and 500 ppb
were used for the calibration.
Preparation of catalyst-loaded electrodes
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 44
1
2
3
4
5
6
7
8
9
In order to prepare CNT-1-loaded electrode, 1.0 mg of CNT-1 was added to
1.0 mL of dimethylformamide (DMF), and the resulted mixture was treated with an
ultrasonic cleaner to give a homogeneous CNT-1 suspension (1 mg mL−1). The glassy
carbon (GC) electrode was rinsed with ethanol and then dried. The CNT-1 DMF
suspension was added onto the GC electrode (5 µL) or GC disk electrode (16 µL)
using a pipet, and the electrode was dried slowly in the dark. In order to prepare
electrodes loaded with non-covalently immobilized Co corroles, 5 mg of CNTs were
added to 5 mL of DMF, and the resulted mixture was treated with an ultrasonic
cleaner to give a homogeneous CNT suspension (1 mg mL−1), which was then added
onto the GC electrode (5 µL) or GC disk electrode (16 µL) using a pipet. The
electrode was then dried slowly. The acetonitrile solution of Co corrole 2 or 3 (1 mg
mL−1) was dropwisely added to the CNT-coated GC electrode (5 µL) or GC disk
electrode (16 µL), and the electrode was dried in the dark at room temperature. The
electrode was then gently rinsed with acetonitrile to remove weakly adsorbed corroles,
and was dried in the dark at room temperature. The Co contents in these samples of
noncovalent immobilization were also determined by ICP-MS.
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
Electrochemistry
Electrochemical measurements were carried out using CH Instruments (model
CHI660D Electrochemical Analyzer). Cyclic voltammogram was collected in 0.1 M
Bu4NPF6 acetonitrile solution using a three-compartment cell with a 0.07 cm2 GC
working electrode, graphite rod counter electrode, and Ag/AgNO3 reference electrode.
10
ACS Paragon Plus Environment

Page 11 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
At the end of each measurement, ferrocene was added as an internal standard to
calibrate the potential. In aqueous solutions, Ag/AgCl reference electrode was used.
Bipotentiostat (model DY2300 Electrochemical Analyzer) was used for rotating
ring-disk electrode measurements. The rotating ring-disk electrode has a GC disk
electrode (0.125 cm2) and a platinum ring electrode (0.188 cm2, ALS RRDE-2). The
collection efficiency of 0.425 was evaluated for this ring-disk electrode using the
[Fe(CN)6]3−/4− redox couple.
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
Synthesis and characterization of Co corroles
Co corrole 1 (Figure 1a) was synthesized by the reaction of Co acetate and
5,15-bis-(pentafluorophenyl)-10-(4)-(1-azido)phenylcorrole. Triphenylphosphine was
added as an axial ligand during the reaction. Purple crystals of 1 were obtained with a
yield of 85% and were used for subsequent reaction with the alkyne-modified CNTs.
1
Complex 1 was characterized by H NMR (Figure 2a) and high-resolution mass
1
spectrometry (HRMS, Figure 2b). In the H NMR spectrum, the eight β-hydrogen
atoms of the corrole ring and the four hydrogen atoms of the 10-(4)-(1-azido)phenyl
substituent are found as sharp peaks in the range of 8.80-7.50 ppm. The rest three
hydrogen peaks at 7.03 (3 H), 6.69 (6 H), and 4.68 (6H) ppm are due to the axial PPh3
unit. The large up-field movement of the PPh3 hydrogen peaks is caused by the
aromatic effect of the corrole ring, which is strong evidence to support the binding of
PPh3 at the Co axial site. The HRMS of 1 shows an ion with the mass-to-charge ratio
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 44
1
2
3
4
5
6
7
8
9
of 1038.1245. This value matches the calculated value of 1038.1249 for the
monocation of [1-N2+H+]+ with the expected isotopic distribution. This result
suggested that during the HRMS measurement, the axial PPh3 ligand was retained but
the azide group decomposed to lose a N2 molecule. Additionally, the UV-vis spectrum
of 1 shows characteristic corrole-based Soret bands at 375 and 411 nm and Q bands at
556 and 587 nm (Figure 2c).81 On the basis of these results, we can confirm the
identity and purity of Co corrole 1.
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 complexes of 5,15-bis(pentafluorophenyl)-10-(4)-(1-pyrenyl)phenylcorrole
(2) and 5,10,15-tris(pentafluorophenyl)corrole (3) were synthesized according to our
previously reported methods.81 The redox properties of Co corroles 1-3 were then
studied by using cyclic voltammetry in acetonitrile. As shown in Figure 2d, the cyclic
voltammogram (CV) of 1 on a GC electrode displays two quasi-reversible reduction
couples at E1/2 = −0.86 and −1.87 V vs ferrocene in acetonitrile, which are ascribed to
the formal CoIII/CoII and CoII/CoI couples, respectively. These values are very similar
to those of Co corroles 2 and 3. For complex 2, the two reduction couples have E1/2
=
−0.76 and −1.88 V vs ferrocene, and for complex 3, they are E1/2 = −0.72 and −1.83 V
vs ferrocene. These results demonstrate that all three Co corroles have very similar
redox behaviors, which are important for direct comparison of their ORR activities
when loaded on CNTs.
12
ACS Paragon Plus Environment

Page 13 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
Figure 2. (a) H NMR spectrum of Co corrole 1 in CDCl3. (b) HRMS of 1 in
methanol. The observed isotopic distribution pattern is identical to that calculated
using the formula C55H28CoF10N5P, [1-N2+H+]+. (c) UV-vis absorption spectrum of 1
in DMF at room temperature. (d) CVs of Co corroles 1, 2, and 3 in 0.1 M Bu4NPF6
acetonitrile. Conditions: GC electrode, 100 mV s−1 scan rate, 20 °C.
Covalent immobilization of 1 on CNTs
To graft the azide-containing Co corrole 1 onto CNTs through covalent bonds,
CNTs were first functionalized with 4-ethynylaniline to introduce surface alkyne
groups.79,82 The as-prepared alkyne-functionalized CNTs were denoted as
CNT-alkyne hereafter. Notably, iron powder used in this synthesis was carefully
removed by stirring the products in hydrochloric acid and then washing with ethanol,
acetone, and water. This process was repeated for several times until no Fe content
was detected in the washing solution by using ICP-MS. The as-prepared CNT-alkyne
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 44
1
2
3
4
5
6
7
8
9
was further analyzed by XPS to exclude the presence of residual Fe. This washing
step is essential because Fe-doped CNTs are known to have some activities for ORR.
The alkyne-functionalized CNTs were analyzed by Raman, showing the decrease of
the ID/IG ratio (Figure 3a, ID and IG represents the intensity of the D and G bands at
1341 and 1583 cm−1, respectively). Although diazonium treatment will usually cause
the increase of the ID/IG ratio, but the surface-modification of CNTs with alkyne
groups, which contain abundant graphite-like C atoms, will lead to the decrease of the
ID/IG ratio.83-85 Subsequently, Co corrole 1 was grafted onto the alkyne-modified
CNTs through the azide-alkyne cycloaddition to make hybrid CNT-1 (Figure 1b).
Notably, during this cycloaddition, no metal catalysts were added to avoid the
contamination of the CNT hybrid with other metal species.
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 successful attaching of molecules of 1 onto CNTs was confirmed. The
Raman spectrum of CNT-1 showed further decrease of the ID/IG ratio to 0.64 (Figure
3a), which is consistent with the surface-modification with corrole moieties that
contain abundant sp2 C atoms. When a GC electrode was loaded with blank CNTs and
was immersed in an acetonitrile solution of 1, the CV showed two quasi-reversible
reduction couples at E1/2 = −0.89 and −1.91 V vs ferrocene (Figure 3b, blue line).
These values match very well with the reduction couples of 1 measured by using a
freshly cleaned GC electrode (Figure 3b, black line).
14
ACS Paragon Plus Environment

Page 15 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. (a) Raman spectra of CNT-1, CNT-alkyne, and unmodified CNT. (b) CV of
GC electrode loaded with CNT-1 in acetonitrile, CVs of GC electrode loaded with or
without blank CNTs in the acetonitrile solution of 1. (c) FTIR spectra of CNT-1, 1,
and CNT. (d) XPS survey scan spectra of CNT-1 and CNT. (e) SEM image of CNT-1.
(f) TEM image of CNT-1.
Importantly, the CV of GC electrode loaded with CNT-1 showed two small
but significant reduction processes at similar potentials in acetonitrile (Figure 3b, red
line). This result suggests the covalent attachment of 1 on CNTs, although the small
CV current implies the very low loading of Co corrole molecules. Notably, if GC
electrodes loaded with CNT-2 or CNT-3 were immersed in acetonitrile, no signals
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 44
1
2
3
4
5
6
7
8
9
from Co corroles could be identified in CV measurements. We rationalize that Co
corrole molecules in CNT-2 and CNT-3 have noncovalent interactions with CNTs and
will re-dissolve into acetonitrile, making its detection by CV very challenging. This
difference is noteworthy and strongly supports the covalent immobilization of 1 on
CNTs in the hybrid CNT-1.
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
Additional evidence for the covalent attachment of molecules of 1 in CNT-1
was provided by infrared spectroscopy. As shown in Figure 3c, complex 1 presents a
characteristic azido group resonance band at 2113 cm−1. This band disappears in the
infrared spectrum of CNT-1, indicating the reaction of this azido group with alkyne
groups on the surface of CNTs. The resonance bands of 1 are well reserved in the
infrared spectrum of CNT-1, confirming the presence of intact molecules of 1 in the
hybrid. Notably, in the infrared spectrum of unmodified CNTs, no such resonance
bands are found in the same range.
In addition, CNT-1 was characterized by XPS (Figure 3d) and EDX (Figure
S4). Comparison of the XPS spectra of CNTs before and after cycloaddition clearly
showed the appearance of Co, N, and F signals from 1. The corresponding Co 2p
narrow scan spectrum was depicted in Figure S5, showing Co 2p3/2 and Co 2p1/2
signals at 778 and 790 eV. This result implies the presence of CoIII ions in the hybrid,
which is consistent with the Co oxidation state of Co corrole 1. Moreover, CNT-1 was
analyzed by FE-SEM (Figure 3e) and TEM (Figure 3f), excluding the presence of
aggregated particles on CNT surfaces. We also loaded Co corroles 2 and 3 onto CNTs
through noncovalent interactions to give CNT-2 and CNT-3, respectively (Figure 1b).
16
ACS Paragon Plus Environment

Page 17 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
As we mentioned above, the coordination structures of Co corroles 1-3 and their
redox behaviors are almost identical. These factors are essential to analyze the effect
of covalent and noncovalent immobilization methods on ORR electrocatalysis.
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
Electrocatalytic oxygen reduction studies
The electrocatalytic ORR features of the three hybrids were then examined in
0.5 M H2SO4 aqueous solutions. The CV of a GC electrode loaded with CNT-1
showed no catalytic current under N2, but showed a large catalytic current under O2
with the half-wave potential at 0.78 V vs RHE (Figure 4a, all potentials reported in
0.5 M H2SO4 aqueous solution in this work are referenced to RHE unless otherwise
noted). This result indicates electrocatalytic reduction of O2 by CNT-1.
The linear sweep voltammetry (LSV) of CNT-1 loaded on a GC electrode was
also measured, giving the same result as obtained by CV (Figure S7). This ORR
performance of CNT-1 is close to that obtained with the commercial Pt/C material,
which shows the half-wave potential at 0.84 V in 0.5 M H2SO4 solution. The LSV of
GC electrodes loaded with CNT-2 and CNT-3 showed catalytic ORR currents with
the half-wave potential at 0.68 and 0.61 V, respectively. In comparison with CNT-2
and CNT-3, the catalytic ORR wave with CNT-1 shifts to the anodic direction by
more than 100 mV under the same conditions. Moreover, the loading of Co corrole
molecules on CNTs can be determined by ICP-MS, giving 2.81, 11.93, and 11.17 µg
Co per mg of CNT-1, CNT-2, and CNT-3, respectively (Table S1). By normalizing
the LSV data per Co content, we can see that the catalytic performance of CNT-1 is
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 44
1
2
3
4
5
6
7
8
9
much improved in comparison with that of CNT-2 and CNT-3 regarding both the
overpotentials and the catalytic currents (Figure 4b). Because complexes 1-3 contain
the same Co corrole core structures and they have almost identical redox behaviors,
this improved ORR performance of CNT-1 suggested the synergistic effect between
molecules of 1 and carbon supports for ORR. More importantly, after ultrasonication
in dichloromethane, CNT-1 retained the initial ORR activity, but both CNT-2 and
CNT-3 almost completely lost their activities due to the re-dissolution of Co corrole
molecules into organic solvents (Figure S8). This result provides strong support for
the covalent immobilization of Co corrole 1 on CNTs in CNT-1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. (a) CVs of GC electrode loaded with CNT-1 and CNT under O2 in 0.5 M
H2SO4 solution. (b) Normalized LSVs of GC electrode loaded with CNT-1, CNT-2,
and CNT-3 under O2 in 0.5 M H2SO4 solution. Conditions: 50 mV s−1 scan rate, 20
°C.
The selectivity of O2 reduction with CNT-1 was then evaluated by using
rotating ring-disk electrode (RRDE) measurements. As shown in Figure 5a and 5b,
CNT-1 displayed a pronounced catalytic ORR activity in RRDE. The GC disk
electrode was scanned by LSV from 1.0 to 0.20 V, while the Pt ring electrode was
18
ACS Paragon Plus Environment

Page 19 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
applied with a constant potential at 1.20 V to detect H2O2 that might be produced by
partially reduced O2 at the disk electrode (Figure S9). The H2O2 yield can be
calculated according to current densities at the disk (Id) and the ring (Ir) electrodes
using the following equation.2,50,51,86,87
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
Ir /N
%H2O2 = 200
Id + Ir /N
in which N = 0.425 is the current collection efficiency of the Pt ring. For CNT-1, this
result gives the number of electrons (n) transferred per O2 molecule to be 3.73 (Figure
5f). Moreover, the catalytic feature of CNT-1 for ORR was studied in 0.1-0.5 M
H2SO4 solutions. The ionic strength of these H2SO4 solutions is maintained the same
with addition of Na2SO4. As shown in Figure S10, the half-wave potential shows a
linear relationship with the solution pH, giving a slope of −63 mV per pH. This is
consistent with the proton-coupled electron transfer process of ORR. For selectivity,
similar n values were obtained in 0.1-0.5 M H2SO4 solutions (Figure S11). In addition,
CNT-2 (Figure 5c), CNT-3 (Figure 5d), and CNT-alkyne (Figure 5e) were subjected
to RRDE measurements. Similar to LSV studies mentioned above, the catalytic ORR
wave with CNT-1 shifts to the positive direction by more than 100 mV in comparison
with that with CNT-2 and CNT-3 as observed in RRDE. The n values for CNT-2 and
CNT-3 are 3.68 and 3.61, respectively. Importantly, CNT-alkyne displayed much
smaller ORR catalytic current under much more negative potentials, confirming that
the ORR activity of CNT-1 is due to covalently attached Co corroles instead of the
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 44
1
2
3
4
5
6
7
8
9
alkyne groups on the surface of CNTs. In addition, we examined the catalytic ORR
feature of complex 1 loaded directly on the GC electrode. As shown in Figure S12,
molecules of 1 on GC electrode showed very poor ORR activity regarding both the
overpotentials and catalytic currents. Moreover, the n value of 1 on GC electrode was
determined to be 2.93. These results confirmed the significantly improved ORR
performance of 1 by attaching to CNTs through covalent bonds. As shown in Table 1,
CNT-1 is one of the most efficient Co-based ORR electrocatalysts regarding both the
half-wave potentials and the n value.
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
Additionally, the n value was further determined using the Koutecky-Levich
(K-L) and the Levich analysis in rotating disk electrode (RDE) experiments. The n
value for CNT-1 calculated from the K-L and Levich analysis was very similar to that
obtained from RRDE measurements under the applied potential range of 0.20-0.40 V
(Figure S13 and S14). The linearity and the near-parallelism as observed in the K-L
plots in the potential range of 0.20 to 0.40 V suggested a first-order reaction kinetics
on dissolved O2 concentration and similar ORR n values at different potentials (Figure
S13b). It is necessary to note that the Levich analysis is in general only applicable
under the totally mass-transfer-limited ORR conditions, which could be regarded at
the applied potential range of 0.20-0.40 V for CNT-1. The n values for CNT-2 (Figure
S15) and CNT-3 (Figure S16) were also examined from RDE measurements, giving
similar values to those from RRDE.
20
ACS Paragon Plus Environment

Page 21 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. (a) O2 reduction in RDE at the GC disk electrode loaded with CNT-1,
CNT-2, and CNT-3 in O2-saturated 0.5 M H2SO4 solutions at 2500 rpm. O2 reduction
in RRDE at the GC disk electrode loaded with CNT-1 (b), CNT-2 (c), CNT-3 (d), and
CNT-alkyne (e) in O2-saturated 0.5 M H2SO4 solutions at 2500 rpm. The Pt ring
electrode was polarized at 1.20 V. (f) The ORR n value with CNT-1, CNT-2, CNT-3
and CNT-alkyne determined from RRDE. Conditions: GC disk electrode (area 0.125
cm2), Pt ring electrode (area 0.188 cm2), 10 mV s−1 scan rate, 20 °C.
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 44
1
2
3
4
5
Table 1. Comparison of electrocatalytic ORR performances for Co-based complexes.
6
7
8
9
Half-wave potentials
(vs RHE)
n values
Catalysts
Conditions
References
This work
from RRDE
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
CNT-1
CNT-2
CNT-3
0.78 V
0.68 V
0.61 V
0.58 V
0.70 V
0.63 V
0.85 V
0.25 V
0.74 V
0.80 V
0.65 V
0.82 V
0.56 V
0.73 V
0.75 V
0.45 V
0.62 V
0.62 V
0.76 V
0.51 V
3.73
3.68
3.61
2.14
3.10
3.96
3.10
3.84
3.85
3.90
3.93
3.65
3.61
2.00
3.60
3.40
3.60
3.80
3.80
3.70
0.5 M H2SO4
0.5 M H2SO4
0.1 M KOH
0.5 M H2SO4
0.1 M KOH
0.5 M H2SO4
0.1 M KOH
0.1 M NaOH
0.5 M H2SO4
0.1 M KOH
0.1 M HClO4
0.1 M KOH
0.5 M H2SO4
0.5 M H2SO4
0.5 M H2SO4
0.5 M H2SO4
0.1 M KOH
1.0 M H2SO4
PloyCoTAC
47
Oxa-Co
MWCNT/Co(TCPP)pyr4
Vc/Nf/Co2Pc2
59
45
28
21
29
42
rGo/(Co-THPP)7
CoTNPc/PGr
MWCNT-CoP-1
C-COP-P-Co
30
GO-CorCo
[Co(tpfcBr8)]/BP2000
CoHPX-1
46
49
78
50
51
43
41
CoTPFC/MWCNT
Co-1/MWCNT
CoCOF-py-0.05rGO
MWCNT-Co-porphyrin
22
ACS Paragon Plus Environment

Page 23 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
The durability of CNT-1 for ORR was examined by chronoamperometric tests.
During 10-h measurements under potential 0.60 V, the current with CNT-1 decreased
by only 10% in an air-saturated solution (Figure 6a), while the current with Pt/C
decreased by 24%. It is worth noting that although Pt/C shows high durability for
electrocatalytic ORR in alkaline media, it is much less stable in acidic solutions.30,88,89
Moreover, CNT-1 exhibited a very small current response in the methanol crossover
test (Figure 6b), but Pt/C showed an apparent increase in the electrocatalytic current
upon the addition of methanol, which corresponds to the methanol oxidation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. (a) Controlled-potential electrolysis of GC electrode loaded with CNT-1
and Pt/C at 0.60 V in air-saturated 0.5 M H2SO4 solutions. (b) Chronoamperometric
responses of CNT-1 and Pt/C upon adding 2.0 M methanol. (c) CV of GC electrode
loaded with CNT-1 in N2-saturated 0.5 M H2SO4 after adding H2O2. (d) Histogram
illustration of currents of GC electrodes loaded with CNT-1 at different electrolysis
potentials under O2-saturated or N2-saturated 0.5 M H2SO4 solutions. In N2-saturated
0.5 M H2SO4 solution, H2O2 was added to get the final concentration of 2.0 mM.
23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 44
1
2
3
4
5
6
7
8
9
In order to distinguish the two possible pathways of the 4e reduction of O2,
namely the direct 4e ORR (i.e., O2 → 2H2O) and the stepwise 2e + 2e ORR (i.e., O2
→ H2O2 → 2H2O),90 we examined the electrocatalytic activity of CNT-1 for the
reduction of H2O2. As shown in Figure 6c, CNT-1 is virtually inactive in catalyzing
the H2O2 reduction. This result excludes the possibility that H2O2 is first generated
from O2 reduction and is subsequently reduced to H2O by CNT-1. In other words,
CNT-1 is able to catalyze the direct 4e reduction of O2 to H2O. Importantly, in the
reverse scan of this test, we can see a pronounced catalytic current, which is due to the
oxidation of H2O2 to evolve O2 (Figure 6c). In the subsequent cathodic scan, the
generated O2 is reduced by CNT-1, confirming the generation of O2 from the H2O2
oxidation. Moreover, we did controlled potential electrolysis with CNT-1 at 0.40-0.80
V in either an O2-saturated solution or an N2-protected solution with addition of 2.0
mM H2O2 (Figures S17-S21). The concentration of H2O2 used here is more than the
1.30 mM concentration of dissolved O2 in an aqueous solution under 1.0 atm O2 at 23
°C. As shown in Figure 6d, the currents observed in O2-saturated solutions in the
potential range of 0.40-0.80 V are much larger than those observed in N2-protected
solutions with addition of H2O2. These results provide strong evidence that CNT-1 is
ineffective in catalyzing the reduction of H2O2. Consequently, CNT-1 is an efficient
electrocatalyst for the direct 4e reduction of O2 to H2O. In addition, chemical O2
reduction by ferrocene (Fc) with CNT-1 in acetonitrile gives n value of 3.2 based on
generated Fc+ and H2O2 (Figure S22). This relatively smaller n value in the chemical
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
24
ACS Paragon Plus Environment

Page 25 of 44
ACS Catalysis
1
2
3
4
5
6
7
reduction studies is likely due to the less efficient electron transfer between Fc and
CNTs in comparison with that between the electrode and CNTs.
8
9
By further analyzing the results mentioned above, we can also exclude the
possibility that small ring currents during the RRDE measurements of CNT-1 are due
to the trapping of H2O2 by CNTs or the disproportionation of H2O2 as catalyzed by
CNT-1. First, as shown in Figure 5e, the RRDE measurement of CNT-alkyne shows
large ring currents. This result confirms the reduction of O2 to form H2O2 at the disk
electrode by CNT-alkyne and suggests that CNTs are not likely to trap H2O2, as large
ring currents are observed. It is worthy noting that the disk current of CNT-alkyne is
much smaller than that of CNT-1 but the ring current of CNT-alkyne is similar with
that of CNT-1 under the same conditions. Second, data from Figure 5e and Figure S12
suggest that the disproportionation of H2O2 by CNTs and complex 1 is at least not
significant. Even with these small disk currents as observed for CNT-alkyne and
complex 1, large currents can be observed at the ring electrode. In addition, as shown
in Figure 6c, if the disproportionation of H2O2 by CNT-1 was quick and significant,
we should observe large catalytic ORR currents during the first run of CV (O2 is
generated in situ from the disproportionation of H2O2). However, no such current is
observed in the first CV run in Figure 6c. This result strongly argues against the
possibility that the formed H2O2 at the disk electrode undergoes disproportionation as
catalyzed by CNT-1, making its detection at the ring electrode unsuccessful.
Moreover, as shown in Figure S17-S21, during the electrolysis with H2O2 under N2,
we did not observe the increase of currents, indicating that the disproportionation 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
25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 44
1
2
3
4
5
6
7
8
9
H2O2 by CNT-1 was also not likely during the long time electrolysis studies.
Mononuclear late transition metal complexes, such as Co complexes, are in
general thought to catalyze the 2e ORR because the heterolytic O−O bond cleavage of
their M−OOH units towards the 4e ORR activity leads to the generation of terminal
late transition metal oxo species, which are high in energy. However, mononuclear Co
porphyrins and corroles have still shown 4e ORR activities.48-51 In these studies,
self-assembled dimers or oligomers of Co porphyrins and corroles are suggested to
form on the surface of electrode materials, which play crucial roles in improving the
4e ORR selectivity. For example, Girault and co-workers studied and established that
the formation of self-assembled face-to-face dimeric Co porphyrins would improve
the 4e selectivity of ORR.61-63 Consequently, dimeric face-to-face Co complexes of
bisporphyrins, biscorroles, and porphyrin-corrole dyads have been designed and
extensively studied as 4e ORR catalysts.33-36,54,56-59
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
In our studies, by using the ICP-MS result, we can calculate that each Co atom
of CNT-1 is supported by about 400 C atoms of the outmost wall of CNTs (please see
Table S1 and Figure S24 and their corresponding discussion for details). Because the
diazonium of CNT surfaces is a nonselective process, it is unlikely to form dimeric
face-to-face Co corrole species in subsequent azide-alkyne cycloaddition, which gives
very low loading density of Co corrole molecules on the surface of CNTs. This is
further supported by the absence of blue shift for the Soret band of Co corroles in
CNT-1 (Figure S23), which is characteristic for the formation of cofacial
“face-to-face” structures.91 Although we cannot completely rule out the presence of
26
ACS Paragon Plus Environment

Page 27 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
possible trace amount of dimeric face-to-face Co corrole species, the majority of
covalently attached Co corrole molecules have to be presented as monomers.
Therefore, our results demonstrate that the mononuclear Co corrole molecules grafted
on CNTs can indeed catalyze the 4e ORR. We rationalized that the short conjugated
linkers enable rapid electron transfer between Co corroles and electrode materials. As
the reduction of peroxo species and the dissociation of peroxo from Co sites are two
competing reactions, the increased electron transfer rates will eventually improve the
4e ORR activity.
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
Conclusions
In summary, we report the electrocatalytic ORR features of Co corrole 1
covalently immobilized on CNTs. The resulted CNT-1 is highly efficient to catalyze
the 4e ORR in 0.5 M H2SO4 aqueous solutions with a half-wave potential at 0.78 V.
As compared with noncovalent immobilization of Co corrole analogues on CNTs, this
covalent immobilization causes more than 100 mV anodic shift for ORR with
significantly improved activity and stability. More importantly, the covalent
attachment with short linkers and very low loading density prevents the formation of
dimeric face-to-face Co corrole species. This result demonstrates that with rapid
electron transfer rates, mononuclear Co corroles can become intrinsically active for
the direct 4e ORR. Although it is shown previously that rapid electron transfer
abilities can considerably improve the 4e ORR activity of Fe porphyrins, it is very
rare to demonstrate similar effect for Co corroles because of the challenging in
27
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 44
1
2
3
4
5
6
7
8
9
molecular design and synthesis. This work is therefore significant to demonstrate that
with sufficient electron transfer abilities, mononuclear Co complexes can display
significantly improved activity for the 4e ORR.
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
Notes: The authors declare no competing financial interest.
Supporting Information: The Supporting Information is available free of charge on
the ACS Publications website.
Figures S1-S24 and Table S1 as described in the text (PDF).
Acknowledgments: We are grateful for support from the “Thousand Talents Program”
of China, the Fok Ying-Tong Education Foundation for Outstanding Young Teachers
in University, the National Natural Science Foundation of China (Grants 21101170,
21573139, and 21773146), the Fundamental Research Funds for the Central
Universities, and the Research Funds of Shaanxi Normal University.
References
(1)
Dey, S.; Mondal, B.; Chatterjee, S.; Rana, A.; Amanullah, S.; Dey, A.
Molecular Electrocatalysts for the Oxygen Reduction Reaction. Nat. Rev.
Chem. 2017, 1, 0098.
(2)
Pegis, M. L.; Wise, C. F.; Martin, D. J.; Mayer, J. M. Oxygen Reduction by
28
ACS Paragon Plus Environment

Page 29 of 44
ACS Catalysis
1
2
3
4
5
6
7
Homogeneous Molecular Catalysts and Electrocatalysts. Chem. Rev. 2018,
118, 2340-2391.
8
9
(3)
Zhang, W.; Lai, W. Z.; Cao, R. Energy-Related Small Molecule Activation
Reactions: Oxygen Reduction and Hydrogen and Oxygen Evolution Reactions
Catalyzed by Porphyrin- and Corrole-Based Systems. Chem. Rev. 2017, 117,
3717-3797.
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
(4)
(5)
Nie, Y.; Li, L.; Wei, Z. D. Recent Advancements in Pt and Pt-Free Catalysts
for Oxygen Reduction Reaction. Chem. Soc. Rev. 2015, 44, 2168-2201.
Miner, E. M.; Gul, S.; Ricke, N. D.; Pastor, E.; Yano, J.; Yachandra, V. K.;
Voorhis, T. V.; Dincă, M. Mechanistic Evidence for Ligand-Centered
Electrocatalytic Oxygen Reduction with the Conductive MOF
Ni3(hexaiminotriphenylene)2. ACS Catal. 2017, 7, 7726-7731.
Iwase, K.; Yoshioka, T.; Nakanishi, S.; Hashimoto, K.; Kamiya, K.
Copper-Modified Covalent Triazine Frameworks as Non-Noble-Metal
Electrocatalysts for Oxygen Reduction. Angew. Chem. Int. Ed. 2015, 54,
11068-11072.
(6)
(7)
(8)
Wei, P. J.; Yu, G. Q.; Naruta, Y.; Liu, J. G. Covalent Grafting of Carbon
Nanotubes with a Biomimetic Heme Model Compound to Enhance Oxygen
Reduction Reactions. Angew. Chem. Int. Ed. 2014, 53, 6659-6663.
Guo, J. N.; Lin, C.-Y.; Xia, Z. H.; Xiang, Z. H. A Pyrolysis-Free Covalent
Organic Polymer for Oxygen Reduction. Angew. Chem. Int. Ed. 2018, 57,
12567-12572.
29
ACS Paragon Plus Environment