the role of defect sites in nanomaterials for ...the role of defect sites in nanomaterials for...

Please cite this article in press as: Jia et al., The Role of Defect Sites in Nanomaterials for Electrocatalytic Energy Conversion, Chem (2019),https://doi.org/10.1016/j.chempr.2019.02.008

Review

The Role of Defect Sites in Nanomaterialsfor Electrocatalytic Energy ConversionYi Jia,1,5 Kun Jiang,2,5 Haotian Wang,3,* and Xiangdong Yao1,4,*

The Bigger Picture

Electrocatalytic energy

conversion technologies have

been widely considered a clean

and sustainable way to alleviate

the reliance on fossil fuels. The

development of efficient and

affordable electrocatalysts plays a

key aspect in energy conversion

processes by lowering the

reaction kinetic barriers and thus

boosting the efficiency and

selectivity of diverse

electrochemical reactions (e.g.,

oxygen and hydrogen evolution

reactions and oxygen and carbon

dioxide reduction reactions).

Recently, defect engineering has

emerged as a new strategy for

tailoring the electronic structures

and interface coordination;

however, the role of ‘‘defect’’-

related sites in as-designed

electrocatalysts has not yet been

fully understood. In this review, we

will shed light on the recent

advances in tailoring

nanomaterials from the aspects of

constructing defect-based motifs

as active sites for versatile

electrochemical energy

conversions as well as their

underlying mechanism on

structure-property correlations.

The development of advanced catalysts for efficient electrochemical energy

conversion technologies to alleviate the reliance on fossil fuels has attracted

considerable interest in the last decades. Insight into the roles of reactive sites

in nanomaterials is significant for understanding and implementing the design

principles of nanocatalysts. Recently, the essential role of defects, including va-

cancies, reconstructed defects, and doped non-metal (or metal)-defect-based

motifs, have been widely demonstrated to promote the diverse electrochemical

processes (e.g., O2 [or CO2] reduction reactions and H2 [or O2] evolution

reactions). Nevertheless, the in-depth exploration of the underlying defect elec-

trocatalytic mechanism is still in its infancy. This review summarizes the state-of-

the-art defect engineering strategies for designing highly efficient electrochem-

ical nanocatalysts with special emphasis on the correlation between defect

structures and electrocatalytic properties. Finally, some perspectives on the

challenges and future research directions in this promising area are presented.

INTRODUCTION

With the increasing depletion of resources and environmental concerns, electro-

chemical energy conversion technologies are regarded as promising pathways for

developing clean energy devices.1,2 The core of these clean energy devices involves

a series of electrochemical redox couples, for instance, oxygen reduction reaction

(ORR) and oxygen evolution reaction (OER) at the air electrodes of regenerative

fuel cells or rechargeable metal-air batteries, hydrogen evolution reaction (HER)

and OER at the cathode and anode of a water-splitting electrolyzer, and electroca-

talytic CO2 reduction reaction (CO2RR) for producing industrial chemicals and fuels.3

Electrocatalysts play a vital role in the above energy conversion processes because

they can lower the reaction kinetic barriers and thus boost the efficiency and selec-

tivity of the chemical transformations involved.4,5 However, the current scope of cat-

alysts utilized for these fundamental electrochemical reactions is dominated by

precious-metal-based materials, such as platinum (Pt)-based catalysts for ORR and

HER, iridium (Ir)- and ruthenium (Ru)-based catalysts for OER, and gold (Au)- and sil-

ver (Ag)-based catalysts for CO2RR, significantly limiting their practical and large-

scale deployment as a result of cost and scarcity.3,6,7

On the basis of the electrocatalytic reaction principles, gas-involving electrochemi-

cal reactions occur at the triple-phase boundary regions of solid electrocatalysts,

liquid electrolyte, and gaseous reactants and intermediates, where electrons, ions,

and molecules can transfer at catalytic active sites.8 In this regard, in the last decade,

the development of a new generation of electrocatalysts has mainly focused on

manipulating the components, sizes, porosity, and interface structures of the mate-

rials to promote multiple-proton-coupled electron transfer and mass transportation

in electrocatalysis.9–11 Recently, defect electrocatalysis has emerged as a new

perspective and attracted extensive research attention, which is based on the

Chem 5, 1–27, June 13, 2019 ª 2019 Elsevier Inc. 1


experimental and theoretical evidences that solid electrocatalysts (e.g., metal ox-

ides, carbides, transition-metal dichalcogenides [TMDs], zeolites, and carbons) con-

taining defects can exert a positive impact on reactivity and selectivity in ORR, OER,

HER, and CO2RR.12–16 Different from a perfect lattice surface, defective sites usually

present unique electronic properties that are particularly important to electrocata-

lytic reactions, where they can bind some reaction intermediates just properly to

deliver an optimized activity and selectivity. Remarkably, the defects can not only

alter the surrounding electronic structure but also serve as the ‘‘docking’’ site to

trap the atomic metal species and form a new synergetic coordination structure as

the active sites. Therefore, it is essential to get in-depth insight into the role of defec-

tive sites in nanomaterials for electrocatalytic energy conversion processes, which

would be particularly important to guide the design of ideal catalysts.

In this review, we start with a brief introduction of defect theory for the design prin-

ciples of electrocatalysts by encompassing defect types, synthetic strategies, and

electronic structure regulations. Afterward, we summarize recent advances in

tailoring nanomaterials from the aspects of constructing defect-based motifs as

active sites for versatile electrochemical energy conversions. In particular, we discuss

the structure-performance relationship in the field of electrocatalysis, which might

provide guidance for the rational design of defective nanomaterials with a superior

catalytic performance. Finally, we propose the current issues and opportunities in

defect engineering at the atomic level to suggest future research directions on

defect electrocatalysis.

1School of Environment and Science,Queensland Micro- and Nanotechnology Centre,Griffith University, Nathan Campus, Queensland4111, Australia

2Rowland Institute, Harvard University,Cambridge, MA 02142, USA

3Department of Chemical and BiomolecularEngineering, Rice University, Houston, TX 77005,

DEFECT THEORY FOR DESIGN PRINCIPLES OF ELECTROCATALYSTS

The key issue in the field of electrocatalysis is the ability to recognize and understand

the relationships between reaction intermediate adsorbed states on actives sites

(microscopic level) and reaction kinetics (macroscopic property).3,8 Defects present

widely in nanomaterials are commonly recognized as the active sites for electrocata-

lytic processes because of the tuned electronic and surface properties in the local

region.13,17–19 For instance, the electronic state is significantly enhanced at the to-

pological defect region (e.g., the combination of carbon rings with pentagon-

octagon-pentagon, denoted as D585) as compared with normal hexagon sites in

defective graphene (DG).20,21 This has been proved by a combination of experi-

mental results and theoretical calculations, which have indicated that active sites

for different electrochemical reactions (ORR, OER, and HER) are induced by specific

defect types as a result of particularly desired reactant binding energy for each elec-

trochemical reaction.22 In addition, the band gaps or electronic states near the Fermi

level of most two-dimensional (2D) TMDs can be tuned through crystal strain, which

can be induced by defects at the edges or basal plane.23 Therefore, defect engineer-

ing has been widely adopted for fabricating nanomaterials to boost their electroca-

talytic reactions.13,17–19 Nevertheless, controllable defect engineering methods are

essential for the precise study of defects and their roles in reaction mechanisms. Do-

ing so is practically expedient for gaining more insight into electrocatalyst design

principles for a specific catalytic process.

USA

4State Key Laboratory of Inorganic Synthesis andPreparative Chemistry, College of Chemistry, JilinUniversity, Changchun 130012, P.R. China

5These authors contributed equally

*Correspondence: [email protected] (H.W.),[email protected] (X.Y.)

https://doi.org/10.1016/j.chempr.2019.02.008

Categories of Defects and Preparation Strategies

According to the dimensions, defects in solid nanomaterials can be classified into

four main categories: zero-dimensional (0D) point defects (e.g., vacancy, reconstruc-

tion, and doping), one-dimensional (1D) line defects (e.g., dislocation), 2D planar

defects (e.g., grain boundary), and three-dimensional (3D) volume defects (e.g.,

spatial lattice disorder). In this review, we mainly concentrate on the point defects

2 Chem 5, 1–27, June 13, 2019

mailto:[email protected]

mailto:[email protected]

https://doi.org/10.1016/j.chempr.2019.02.008

Figure 1. Schematic Illustration of Categories of Defect-BasedMotifs in Nanomaterials and Their

Preparation Approaches


because they are the elementary motifs for endowing the functions of defect-based

electrocatalysts from the multi-dimensional scope. Point defects can be further sub-

divided into three representative configurations in metal compounds or non-metal

materials: (1) reconstructed or vacancy defects, (2) non-metallic-atom-doping

(such as N, S, B, and P)-induced defects, and (3) metal-defect coordination structures

(Figure 1).

Correspondingly, diverse approaches have been experimentally developed to fabri-

cate such defect-based motifs in nanomaterials, which can be mainly classified into

post-functionalized methods and in situ synthesis. With respect to post-functional-

ized methods, ball milling is an efficient and cost-effective approach to creating

and exposing a large amount of reconstructed and vacancy edge defects. For

example, Deng and co-workers demonstrated that the size reduction of graphite

nanosheets by simple ball milling can significantly enhance the electrocatalytic acti-

vation of oxygen, which originates from the reconstructed oxygen-defect motifs on

zigzag edges.24 It is also noteworthy that the plasma technology with good control-

lability (e.g., Ar, O2, and/or N2 plasma source type, power density, and processing

time) is another promising approach to obtaining uniformly vacancy defects in car-

bon- or metal-compound-based nanomaterials.25,26 Furthermore, the obtained

defect structures in carbons, transition-metal oxides, and TMDs (e.g., MoS2 and

CoSe2) can be further processed by non-metal- and metal-atom doping to form

various dopant-defect coordination motifs to not only stabilize the topological struc-

tures of defects but also further manipulate the local electronic distribution through

the diverse coordinated configurations.27,28 Analogous to plasma technology, a

high-pressure hydrogenation process has been recently reported to treat a 2D

iron-cobalt oxide (Fe1Co1Ox) via controllable hydrogen pressure and annealing tem-

perature with the purpose of tuning its oxygen vacancy density.29 Moreover, the

acidic etching of metallic nanoparticles anchored at the defect sites is another

post-treatment approach to obtaining metal-defect motifs at the atomic level.30,31

Chem 5, 1–27, June 13, 2019 3


Importantly, in the post-functionalized strategy, the construction conditions of

defect engineering (e.g., defect type and density) should be considered, which is

not only beneficial for generating expected defective centers with optimal amount

but also important for selecting suitable non-metal and metal species as guest dop-

ants for defect modification.

Alternatively, the in situ synthetic strategy is a bottom-up pathway for tuning the

type and density of intrinsic defects during the construction of defect-based motifs

in various nanomaterials, including template-based carbonization,32 dopant-

removal annealing,22 pyrolysis of Zn-contained metal-organic frameworks,33,34

and electrochemical activation.35 Specifically, Zhang and co-workers synthesized a

series of nitrogen-doped graphene (NG) meshes (NGMs) with different abundances

of topological defects byMgO-templated carbonization of carbon and nitrogen pre-

cursors.31 Yao and co-workers introduced different types of topological defects into

the graphene lattice via the removal of intentionally doped heteroatoms at a high

temperature; afterward, the aberration-corrected transmission electron microscopy

directly demonstrated the various defects (pentagons, heptagons, and octagons) on

the edge of the DG sheet at the atomic scale.22 Yao and co-workers carbonized a Zn-

enriched metal-organic framework (IRMOF-8) and obtained a highly porous defec-

tive carbon because the containing Zn could be removed completely at 950�C.33

Following this evaporated Zn-induced defect scenario, doping Zn cations in other

metal-contained MOFs (e.g., ZnCo-MOF) can exert the ‘‘fence’’ function to spatially

expand the distance of adjacent Co atoms. Thus, after pyrolysis, the induced va-

cancy defects due to the evaporated Zn nodes can in situ preferentially coordinate

with Co nodes to form a stable Co-N-C defect motif.34 In addition, electrochemical

activation can also be utilized to create defect sites in carbon-based materials given

that partial amorphous carbon can be oxidized and removed, resulting in simulta-

neous formation of atomic metals fixed by surrounding N-C defect sites in the pro-

cess of wet chemistry.35

Defect Sites for the Regulation of Electronic Structures

Unraveling diverse descriptors of catalytic reactivity is essential for quickly screening

catalysts for a given reaction. It has been demonstrated that the apparent rate of an

electrocatalytic reaction is strongly related to the intrinsic electronic structure (e.g.,

d-band position for metal compounds, valence orbital level for non-metals, andwork

function) of an electrocatalyst, which acts in determining the intermediate adsorp-

tion energetics, activation energies, and reaction energy barriers in specific electro-

chemical reactions.4,36,37 For instance, both theoretical calculations and experi-

mental data have revealed that topological point defects in the basal plane of

graphene can lead to opening a band gap or generating more electronic states

near the Fermi level in comparison with the perfect graphene (the p and p* bands

near the Fermi level are doubly degenerate) (Figure 2A).17,18,22 Analogous to carbon

materials, in defective metal compounds, such as MoS2 with S vacancies, band struc-

ture can be stepwisely tuned by varying concentrations of S vacancies, leading to the

reduction of the hydrogen adsorption free energy (DGH) (Figure 2B).38 With regard

to non-metallic-dopant-induced defect motifs, Zheng and co-workers selected six

non-metallic heteroatoms with various electron electronegativities (N and O act as

electron acceptors for the adjacent C, whereas F, S, B, and P act as electron donors)

to investigate their effects on the adjacent C atoms, which act as the active sites for

HER. The authors proposed that carbon-defect-based motifs via diverse coordina-

tion with multiple dopants can finely tailor the local charge populations in graphene

matrix (Figure 2C).39 Furthermore, Xia and co-workers proposed a novel activity

descriptor by considering the synergistic effect of electronegativity and electron

4 Chem 5, 1–27, June 13, 2019

Figure 2. Various Defect-Based Motifs for Electronic Structure Regulation

(A) Three types of topological carbon defects and their electronic density of states. Reprinted with permission from Kotakoski et al.17 Copyright 2011

American Physical Society.

(B) Schematic of the top (top) and side (bottom) views of MoS2 with strained S vacancies on the basal plane, and the surface energy in relation to the

densities of S vacancy and strain is shown in a colored contour plot. Reprinted with permission from Li et al.38 Copyright 2015 Springer Nature.

(C) Natural bond orbital (NBO) population analysis of six different non-metallic heteroatoms in a graphene matrix. pN and gN represent pyridinic and

graphitic N, respectively. The inset shows the proposed doping sites for different elements. Reprinted with permission from Zheng et al.39 Copyright

2014 American Chemical Society.

(D) Volcano plots obtained to correlate the ORR and OER overpotential and the intrinsic descriptor for doped graphene materials. Reprinted with

permission from Zhao et al.40 Copyright 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

(E) Different coordination motifs of dual dopants (B and N) in carbon nanotubes show distinct ORR performances as a result of tuned electronic

structures. Reprinted with permission from Zhao et al.41 Copyright 2013 American Chemical Society.

(F) The C bridge plays an essential role among various B-C-N configurations with B active sites as a function of the distance to a pyridinic N atom (sites I–

VI marked in the inset). Reprinted with permission from Zheng et al.42 Copyright 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

(G–J) Atomic-resolution STM image (G) and the corresponding simulation image (H) of single Cu atoms embedded in NG. Bader charge (left ordinate)

and corresponding chemical valence of Cu atoms (right ordinate) for different structures Cu-N-(C) coordination motifs (I) as well as corresponding

volcano plot of the relationship betweenORR activity andDEO (J). Reprinted with permission fromWu et al.43 Copyright 2016 Royal Society of Chemistry.

Chem 5, 1–27, June 13, 2019 5



affinity on charge redistribution to correlate the intrinsic material characteristics to

the catalytic performance: F = (EX/EC) 3 (AX/AC).40 The authors found that such

non-metallic-heteroatom-modulated carbon-defect-based motifs can also strongly

influence the electron transfer and reaction energy in ORR and OER with specific

electronic structures (Figure 2D). On the other hand, the location of the multiple het-

eroatom dopants in a carbon matrix is also essential. Zhao and co-workers reported

that two kinds of B- and N-co-doped carbon nanotubes (CNTs) dominated by the

bonded or separated B and N have been intentionally prepared by chemical-va-

por-deposition growth or after treatment, respectively, demonstrating distinct

ORR performances (Figure 2E).41 The authors revealed that inertness of the bonded

B- and N-co-doped CNT can be ascribed to the fact that the lone-pair electrons from

the N dopant are largely neutralized by the vacant orbital of B dopant, and few elec-

trons or vacant orbitals are left to conjugate with the carbonp system.41With respect

to this viewpoint, Zheng and co-workers found that the C bridge site plays a syner-

getic role among various B-C-N configurations in a graphene matrix with B active

sites as a function of the distance to a pyridinic N atom, and the strength of the syn-

ergistic effect decreases gradually as the distance between the B and pyridinic N

dopants increases (Figure 2F).42 Besides non-metallic dopants, the presence of de-

fects on supports can also act as the ‘‘docking’’ sites to capture atomic metal species

for the formation of unique metal-defect coordination motifs to enhance catalytic

performance. Particularly, the vacancies on the supporting matrix not only can

trap the diverse metal species at atomic scale depending on the geometric struc-

tures of defects but also can finely tune the electronic structures of the centered

metal atoms via different surrounding configurations. Wu and co-workers synthe-

sized atomically dispersed Cu-N-C coordinationmotifs with high density in a NGma-

trix by pyrolysis of copper phthalocyanine and dicyandiamide.43 The atomic-resolu-

tion scanning tunneling microscopy (STM) images show that embedded Cu atoms

render the surrounding N and C atoms electronically rich (Figures 2G and 2H). Bader

charge distribution of different Cu-N-(C) coordination motifs shows that the variable

chemical valence states of the Cu atom coordinated with different numbers of N

atoms (Figure 2I), and the corresponding volcano plot of the relationship between

ORR activity and DEO indicates that the optimal Cu-atom-modulated NC defect

motif for ORR is Cu-N2 (Figure 2J). In brief, the electronic structures of defect sites

can be artificially tuned by vacancy, non-metallic, or metallic dopants to achieve

the optimal binding energy of various adsorbed intermediate states for specific elec-

trochemical reactions.

CONSTRUCTING DEFECT-BASED MOTIFS IN NANOMATERIALS ASACTIVE SITES FOR ELECTROCHEMICAL ENERGY CONVERSION

Oxygen Reduction Reaction

The ORR on the cathode is a key process that limits the energy conversion efficiency

of fuel cells and metal-air batteries, in which commercial Pt is known as the current

benchmark catalyst. In the last decade, metal-free carbon nanomaterials (CNMs)

have been extensively investigated because of their low cost, high conductivity,

tunable microstructures, and strong tolerance to alkaline and acidic environments

for wide applications.12,15,37,44

Since 2009, N-doped CNT arrays have been discovered to be superior to Pt for the

electrocatalysis of the ORR,45 and extensive efforts have been made to adjust the

intrinsic electronic states and extrinsic physical characteristics of the CNMs, such

as tuning the types and location of doped atoms40–42,46,47 and hierarchically porous

microstructures48 to enhance ORR performance. Remarkably, it is inevitable to

6 Chem 5, 1–27, June 13, 2019

Figure 3. Edge Defects in Carbon Nanomaterials for ORR

(A and B) HAADF image of graphene possessing defects, where the hexagons, pentagons, heptagons, and octagons are labeled in orange, green, blue,

and red, respectively (A). ORR performance comparison of pristine graphene, NG, DG, and Pt/C in alkaline (B). Reprinted with permission from Jia

et al.22 Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

(C and D) Micro apparatus for the ORR electrochemical experiment (C). LSV curves of the ORR tested for a droplet located either on the edge or on the

basal plane of the HOPG (D). Reprinted with permission from Shen et al.49 Copyright 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

(E–H) HRTEM image of defective CNCs (E) and their ORR performance comparison with N-CNTs in 0.10 M KOH (F). Raman spectra (G) and LSV curves (H)

of the defective CNCs prepared at different temperatures. Reprinted with permission from Jiang et al.50 Copyright 2015 American Chemical Society.

(I–L) SEM image of NGM (I) and ORR and OER volcano plots of overpotential versus adsorption energy of OH* (J); the inset represents a graphene

nanoribbon with different kinds of N doping or topological defects used for DFT calculations. LSV curves of NGM, GM, NG, and commercial Pt/C and Ir/

C catalysts toward both ORR and OER (K). Open-circuit plots of the homemade primary Zn-air batteries with NGM and Pt/C (L). Reprinted with

permission from Tang et al.51 Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.


produce the vacancy (or reconstructed) defects in CNMs during the doping or micro-

structure tailoring process; however, the functionality of defects in electrocatalysis

has just been realized after long-standing neglect because of a lack of clear and

convincing ways to identify defects.8,13,14

Defect engineering insight for guiding the design of efficient and stable electroca-

talysts has recently attracted wide research attention because the existence of de-

fects can not only alter the electronic structures of defect sites for regulating binding

energy of various adsorbed intermediates but also serve as ‘‘docking’’ sites to cap-

ture and stabilize the non-metal and metal heteroatoms to form heteroatom-defect-

based motifs with diverse coordination environments and robust stability for effi-

cient and durable ORR performance. Yao and co-workers introduced different types

of topological defects into the graphene lattice by removing intentionally doped

heteroatoms. The aberration-corrected HRTEM image (Figure 3A) directly shows

various defects (pentagons, heptagons, and octagons) on the edge of the graphene

Chem 5, 1–27, June 13, 2019 7


at the atomic scale. Compared with the NG, the DG exhibited enhancedORR activity

(Figure 3B) due to a change in the charge and spin distributions in the DG and

adsorption energies of the intermediates on active sites.22 Moreover, Yao and co-

workers found that diverse configurations of topological carbon defects are distinc-

tively appealing to different electrochemical reactions (ORR, OER, and HER) as a

result of specific modulated electronic structures. To clarify whether the edge de-

fects serve as the active sites for ORR, Wang and co-workers developed a micro-

electrochemical testing apparatus to investigate the ORR activity at the edge and

the basal-plane of highly oriented pyrolytic graphite (HOPG) (Figure 3C).49 Linear

sweep voltammetry (LSV) was recorded at the specified location where the air-satu-

rated droplet was deposited and indicated an appreciably higher ORR activity at the

edges of HOPG with respect to the basal plane (Figure 3D). To further increase the

edge defect density, Hu and co-workers tailored the extrinsic microstructure of the

defective CNMs by employing carbon nanocages (CNCs) with abundant holes and

edges (Figure 3E).50 As a result of the existence of substantial defect sites, CNCs

delivered superior ORR activity compared with that of N-doped CNTs (Figure 3F).

Remarkably, CNCs with the highest defect density from the Raman spectra (Fig-

ure 3G) showed the best electrochemical ORR activity (Figure 3H). At the level of

intrinsic electron structure, all the factors (defects, edges, and dopants) that induced

the modification of charge and spin distribution in the carbon matrix could effec-

tively tune the adsorption energies of the intermediates, leading to enhanced

ORR performance. With these multiscale fundamentals of catalyst design, Zhang

and co-workers developed a NGM by synergistically combining the merits of

doping, defects, and morphology regulation.51 Scanning electron microscopy

(SEM) images clearly showed that the hierarchically porous NGMs were assembled

by ultrathin graphene flakes with a lateral size of z250 nm (Figure 3I). The authors

conducted density functional theory (DFT) investigations to systematically study

the roles of dopants, edges, and defects for both ORR and OER. The calculations re-

vealed that a nitrogen-free configuration with adjacent pentagon and heptagon car-

bon rings exhibits the lowest overpotential for both ORR and OER at the peak of vol-

cano plot, indicating that the reconstructed defects (compared with N-doped-based

motifs) could serve as the optimal active sites for ORR and OER (Figure 3J). Experi-

mentally, such defect-rich NGMs exhibited excellent bifunctional ORR and OER ac-

tivities, including a potential gap (DE) as low as 0.90 V (Figure 3K) and stable dura-

bility as tested in a primary Zn-air battery (Figure 3L).

High-performance ORR catalysts in acidic electrolytes are more significant than

those in alkaline environments because commercial proton-exchange membrane

FCs (PEMFCs) prefer to operate in acidic media. Although modified CNMs show ac-

tivities comparable to those of Pt in alkaline electrolytes, they are still not desirable in

acidic media because the Gibbs free energy in the ORR process under acidic condi-

tions is higher than that in alkaline conditions.46,47,52 To this end, Li and co-workers

developed a defective N- and S-co-doped carbon aerogel (NSCA) fabricated from

carrageenan (seaweed extract) through a two-step annealing method, which suc-

cessfully introduced structural defects into 3D hierarchical porous carbon aerogels

to generate atomic N-S-C-defect-based motifs (Figures 4A and 4B).53 The NSCA ex-

hibited superior ORR activity in 0.1 M HClO4 solution, in which the half-wave poten-

tial was only 26 mV lower than that of Pt/C commercial catalyst with higher current

density (Figure 4C). First-principle calculations based on DFT revealed activity ori-

gins of the atomic N-S-C-defect-based motifs in NSCA for acidic ORR (Figures 4D

and 4E). Particularly, the N-S-D-G-6 and N-S-D-G-4 sites (Figure 4D) presented

lower overpotentials of 0.15 and 0.40 eV, respectively, in acidic condition (pH = 0)

than Pt/C (0.45 eV), whereas the S-G site without defect modification still required

8 Chem 5, 1–27, June 13, 2019

Figure 4. Defective NSCA for Superior ORR in Acidic Media

(A) Schematic synthetic route of defective NSCA from carrageenan (seaweed extract).

(B) Color filtered STEM image of defect structures in NSCA.

(C) LSV curves of NSCA and Pt/C catalysts in 0.1 M HClO4 solution.

(D) Optimized structures of the N-S-D-G model. Gray, white, yellow, and claret balls represent C, H, S, and N atoms, respectively, whereas blue atoms

represent S- or N-adjacent C atoms.

(E) Free-energy diagram for N-S-D-G, S-D-G, and S-G at equilibrium potential.

Reprinted with permission from Li et al.53 Copyright 2018 Elsevier.


a high overpotential of 0.85 eV. This work indicates themultiscale synergy of doping,

microstructure, and defect engineering, which might open a new route toward

designing and developing low-cost, large-scale, metal-free carbon ORR catalysts

in acidic electrolytes.

Besides non-metallic-dopant-induced defect motifs in CNMs for efficient ORR, tran-

sition-metal-N-C-defect-based coordination sites (M-Nx-C, M = Fe, Co, Mn, etc.) in

CNMs have also been regarded as the most promising substitutes for Pt-based cat-

alysts.34,35,54,55 Li and co-workers reported stable Co single atoms on nitrogen-

doped porous carbon (Co SAs/N-C) with precise Co-N coordination in carbon de-

fects. The dominant active centers for Co SAs/N-C formed at 800�C and 900�Ccan be defined as planar Co-N4 and Co-N2, respectively.

34 The Co SAs/N-C with a

Co-N2 center exhibited higher ORR activity than commercial Pt/C and Co SAs/N-

C with Co-N4 in an alkaline solution as a result of a stronger interaction with peroxide

on Co-N2 sites to promote 4e� for ORR (Figure 5A). Zelenay and co-workers re-

ported an iron-nitrogen-carbon ((CM+PANI)-Fe-C) derived from cyanamide (CM)

and polyaniline (PANI) with atomically dispersed Fe-N4-C-defect-based motifs and

hierarchical porosity.54 This catalyst exhibited an excellent performance in a rotating

ring-disk electrode (RDE) test in 0.5 M H2SO4 (Figure 5B). The authors revealed that

the catalytic active sites were carbon-embedded Fe-N4 sites, which were directly

visualized by AC-STEM characterization (Figure 5B). Noticeably, albeit Fe (or Co)-

Nx-C-based motifs as the ORR active sites have been theoretically and experimen-

tally proved, they still suffer from insufficient durability, especially at the desirable

high voltages (>0.6 V), which limits their practical applications in PEMFCs.56 The rea-

sons of ORR activity degradation could originate from (1) the dissolution of active

metal sites due to surrounding corrosion of amorphous carbon, (2) the protonation

of the dopedN neighboring the active metal sites followed by anion adsorption, and

Chem 5, 1–27, June 13, 2019 9

Figure 5. Atomic-Metal-Defect Coordination Motifs for ORR

(A) LSV comparison curves of Co SAs/N-C (900) with Co-N2, Co SAs/N-C (800) with Co-N4, Co NPs/N-C, and Pt/C. Reprinted with permission from Yin


(B) ORR performance of (CM + PANI)-Fe-C catalyst and direct AC-HRTEM observation of the Fe-N-C-defect-based motif in this catalyst. Reprinted with

permission from Chung et al.54 Copyright 2017 AAAS.

(C–E) Schematic of the atomically dispersed Mn-N4-C site catalyst through a two-step doping and adsorption strategy (C). The effect of two-step

approach (first doping and second adsorption) on Mn-N-C catalyst ORR activity in 0.5 M H2SO4 (D). The stability tests of the 20Mn-NC-second catalyst

and 20Fe-NC-second catalyst at a constant potential of 0.8 V for 100 h in O2-saturated 0.5 M H2SO4 (E). Reprinted with permission from Wu et al.55

Copyright 2018 Springer Nature.

(F–K) CV (F) curves after different activation cycles in 0.5 M H2SO4 electrolyte and the evolution of the ECSA during activation (G). The morphology

evolution observation of A-CoPt-NC from Co-NC and the presence of Co(Pt)-Pt(Co)-N-C motifs in A-CoPt-NC by direct AC-HRTEM identification (H).

10 Chem 5, 1–27, June 13, 2019


Figure 5. Continued

LSV curves of Co-NC, A-CoPt-NC, and Pt/C in 0.1 M KOH electrolyte (I). LSV curves of A-CoPt-NC, A-Pt-NC, A-Co-NC, and Pt/C in 0.1 M KOH electrolyte

(J). Charge densities of a(Co-Pt)@N8V4 and a(Pt-Pt)@N8V4 motifs (K). Reprinted with permission from Jia and co-workers.35 Copyright 2018 American

Chemical Society.


(3) the Fenton side reaction (e.g., atomic Fe sites in Fe-N-Cmotifs are easily attacked

by H2O2 in the presence of the 2e� pathway in ORR, leading to significant loss of

active sites).56 To address these issues, Wu and co-workers recently reported an effi-

cient and stable ORR catalyst consisting of atomically dispersed nitrogen-coordi-

nated single Mn sites on partially graphitic carbon (Mn-N-C), which was synthesized

through a two-step strategy involving doping and adsorption processes (Fig-

ure 5C).55 The 20Mn-NC-second catalyst sample exhibited the most positive E1/2

in acidic ORR (0.5 M H2SO4) up to 0.8 V versus reversible hydrogen electrode

(RHE) because a much higher density of Mn-N-C active motifs benefitted from the

second adsorption step than from the first doping step (Figure 5D). More impor-

tantly, the 20Mn-NC-second catalyst exhibited more durable stability than the

20Fe-NC-second catalyst at a constant potential of 0.8 V for 100 h in O2-saturated

0.5 M H2SO4, suggesting that Fenton reactions involving Mn ions are insignificant

because of weak reactivity between Mn and H2O2 (Figure 5E). Furthermore, the

20Mn-NC-second catalyst was fabricated as a cathode in membrane electrode as-

semblies (MEAs) for H2-O2 practical fuel cell applications, which can achieve the

optimal power density of 0.46 W cm�2 at 1.0 bar partial pressure.55

On the one hand, the specific metallic atoms in M-N-C motifs possess distinct selec-

tivity in ORR (2e� or 4e� pathways). On the other hand, the surround coordination

structures via defect engineering can also play crucial roles for impacting the ORR

reaction pathways. For instance, the isolated atomic Pt species (denoted as IA-Pt)

were reported to possess a high selectivity for the production of H2O2 via a 2e�

pathway rather than H2O via a 4e� pathway, which is totally different from Pt bulk

because of a lack of synergy of the Pt-Pt bond.57 To get insight into the synergy of

adjacent dual-sites of atomic M1-M2 species in defect-based motifs, Yao and co-

workers developed an atomic Co-Pt nitrogen-carbon-based catalyst (denoted as

A-CoPt-NC) that directly utilized the induced defects in the shell of carbon capsules

to form atomic Co-Pt-N-C coordination structures as active sites through electro-

chemical activation (Figure 5F).35 Electrochemical active surface area (ECSA) analysis

of the Pt revealed that there was little Pt loading on the carbon shell (ECSA was

nearly zero) in the first 2,000 cyclic voltammetry (CV) cycles (zone A in Figure 5G)

because of the rarity of adsorption sites in the carbon. Thereafter, with the contin-

uous electrochemical activation (2,000–7,000 CV cycles), the ECSA increased

(zone B in Figure 5G), suggesting that more atomic Pt species are immobilized in

N-doped carbon defects to form Pt-N-C motifs, whereas partial amorphous carbon

in graphitic carbon shells is corroded during CV cycles. Simultaneously, creating

channels also allows ingress of acidic solvent and the gradual removal Co cores,

generating adjacent metallic dual sites of Co(Pt)-Pt(Co)-N-C motifs, as evidenced

by direct AC-HRTEM observations (Figure 5H). This A-CoPt-NC catalyst exhibited

a half-wave potential of 0.96 V versus RHE in 0.1 M KOH electrolyte, demonstrating

a much better ORR performance than that of Co-NC and Pt/C (the latter of which was

90 mV lower) (Figure 5I). The control samples containing atomic Co (A-Co-NC) and

atomic Pt (A-Co-NC) were prepared individually and showed worse performance

than A-CoPt-NC (Figure 5J), indicating the synergistic effect between atomic Co

and Pt for ORR. Further DFT calculations revealed that great ORR activity and high

selectivity for a 4e� ORR pathway was attributed to asymmetry in the electron distri-

bution around the Pt/Co metal centers, as well as the metal-atom coordination with

Chem 5, 1–27, June 13, 2019 11


specific defects (e.g., N8V4 defect, in which N8 represents the number of nitrogen

atoms and V4 indicates the number of vacant carbon atoms) (Figure 5K).

Oxygen Evolution Reaction

The OER represents the rate-determining step of electrocatalytic water splitting into

hydrogen and oxygen. Creating anion or cation vacancies and adjusting their den-

sity has proven to be an effective strategy for designing high-performance OER cat-

alysts because of the low formation energy.13,14 For instance, Zheng and co-workers

reported a facile method of preparing mesoporous Co3O4 nanowires (NWs) with

abundant oxygen vacancies (Figure 6A),58 where the pristine mesoporous Co3O4

NWs were treated with NaBH4 at room temperature to reduce Co3O4 NWs for intro-

ducing the oxygen vacancies. Compared with the pristine Co3O4 NWs, the reduced

Co3O4 NWs showed a 7-fold enhancement of water oxidation catalytic activity (at a

potential of 1.65 V versus RHE) as an efficient OER catalyst (Figure 6B). Moreover, the

chronoamperometric measurements showed that the current density of the reduced

Co3O4 remained at 69% of its initial current density (1.6 mA cm�2) after 6,000 s,

higher than that of the pristine Co3O4, which retained only 35% (Figure 6C), suggest-

ing the stability of oxygen vacancies serving as the active sites for OER. However, the

oxygen vacancies created during NaBH4 reduction are difficult to control. To

address this challenge, Yao and co-workers demonstrated a controllable hydroge-

nation treatment by tuning annealing temperatures and hydrogen pressures to

further precisely control the density of oxygen vacancies in oxide materials (Fig-

ure 6D).29 The X-ray diffraction (XRD) patterns of Fe1Co1Ox-origin and Fe1Co1Ox-

200�C-z samples showed characteristic peaks corresponding to the cubic Co3O4

phase, whereas the peaks shifted after being hydrogenated at 300�C or higher tem-

peratures, suggesting the occurrence of a phase transformation (Figure 6E). SEM

and transmission electron microscopy (TEM) images showed that compared with

Fe1Co1Ox-origin, the Fe1Co1Ox-200�C-2.0 MPa sample exhibited a much rougher

surface and retained the nanosheet morphology, implying the formation of more de-

fects (Figure 6F).

Furthermore, the particle agglomeration appeared to accompany the phase trans-

formation at the high temperature of 400�C (Figure 6F). The photoluminescence

(PL) spectra (Figure 6G) revealed that the Fe1Co1Ox-200�C-z samples with increased

hydrogenation pressures possessed increasingly higher amounts of oxygen va-

cancies than Fe1Co1Ox-origin. The OER LSV curves showed that Fe1Co1Ox-200�C-2.0 MPa had a much better OER performance than RuO2, Fe1Co1Ox-origin and Fe1-Co1Ox-200�C-0.5 MP, indicating that the OER performance can be significantly

enhanced by increased oxygen vacancies (Figure 6H). More importantly, such oxy-

gen vacancies can be well retained after a long-term stability test for 10,000 s (Fig-

ure 6I). Through DFT calculations, the authors revealed that an optimal oxygen va-

cancy density is the key to improving the OER activity (Figure 6J). Alternatively,

Wang and co-workers developed a plasma-engraving strategy to generate oxygen

vacancies on the surface of Co3O4 nanosheets to simultaneously adjust the elec-

tronic states and increase the surface area, leading to exposing more active sites

(Figure 6K).59 As expected, the Co3O4 nanosheets treated with Ar plasma for

120 s exhibited much higher OER performance than did pristine Co3O4 nanosheets

(Figure 6L) and a superior stability after 2,000 cycles (Figure 6M). Since then, such a

plasma-engraving strategy has also been employed in exfoliation of bulk 2D mate-

rials, such as CoSe2, to introduce defects and exposure more active sites.26 Besides

anion defects, cation defects have also been proved to promote OER properties.

Shao and co-workers found that the A-site cation defect of a perovskite type

(ABX3) LaFeO3 can greatly enhance both the OER and ORR activities because of

12 Chem 5, 1–27, June 13, 2019

Figure 6. The Roles of Vacancy Deficiencies in OER Catalysts

(A–C) Schematic of the NaBH4 reduction for in situ creation of oxygen vacancies in Co3O4 NWs for enhanced OER (A). OER LSV curves of reduced Co3O4

NWs, pristine Co3O4 NWs, IrOx, and Pt/C at 5 mV s�1 (B). Stability tests of reduced Co3O4 NWs and pristine Co3O4 NWs at 1.65 V versus RHE (C).

Reprinted with permission from Wang et al.58 Copyright 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

(D–J) Schematic of the hydrogenation process for the creation of oxygen vacancies in Fe1Co1Ox-origin (D). XRD patterns of Fe1Co1Ox-origin and

hydrogenated Fe1Co1Ox-y-z samples, in which y represents annealing temperature and z represents applied hydrogen pressure (E). Selected SEM and

TEM images (F). PL spectra of Fe1Co1Ox-origin and Fe1Co1Ox-200�C-z samples (G). OER LSV curves of Fe1Co1Ox-origin and Fe1Co1Ox-200

�C-z samples

in 1.0 M KOH (H). (I) Chronoamperometry curves of Fe1Co1Ox-200�C-2.0 MPa and RuO2 at a potential of 0.6 V versus Ag/AgCl. Free-energy diagrams for

OER pathway in alkaline media show the superior catalytic performance at the equilibrium potential (UNHE = 0.404 V) of Fe1Co1Ox�1 with one oxygen

vacancy (J). Reprinted with permission from Zhuang et al.29 Copyright 2018 Tsinghua University Press & Springer.

(K–M) Schematic of fabricating the Ar-plasma-engraved Co3O4 nanosheets with oxygen vacancies and high surface area (K). OER LSV curves of pristine

Co3O4 (0 s) and the plasma-engraved Co3O4 (120 s) (L). Stability of the plasma-engraved Co3O4 after 2,000 cycles (M). Reprinted with permission from Xu


(N–P) Schematic illustration of the formation of oxygen vacancy and Fe4+ in A-site-deficient La1-xFeO3-d perovskites (N). OER LSV curves of the RDE

(1,600 rpm) composed of LF, L0.98F, L0.95F, and L0.9F catalysts in 0.1 M KOH solution as well as the comparison in mass activity (MA) and specific activity

(SA) (O). Reprinted with permission from Zhu et al.60 Copyright 2016 American Chemical Society.

Chem 5, 1–27, June 13, 2019 13


Figure 7. The Roles of Non-metal (or Metal)-Defect Coordination Structures in OER Catalysts

(A–D) Schematic preparation of P-Co3O4 and VO-Co3O4 by a plasma-assisted approach (A). Co K-edge XANES spectra of pristine Co3O4, VO-Co3O4,

and P-Co3O4 (B); top inset magnifies the main peak region, and bottom insets magnify the pre-peak region. OER LSV cures of pristine Co3O4, VO-Co3O4,

and P-Co3O4 (C). Density of states for pristine Co3O4, VO-Co3O4, and P-Co3O4 systems (D). Reprinted with permission from Xiao et al.65 Copyright 2017

Royal Society of Chemistry.

(E–J) TEM image of CoSe2-x-Pt nanosheets (E). AFM image of the exfoliated CoSe2-x (F). k2-weighted Fourier-transformed EXAFS spectra of the Pt for Pt

foil, PtO2, and CoSe2-x-Pt (G). The processed and colored HAADF-STEM image of CoSe2-x-Pt (H). OER LSV cures of CoSe2-origin, CoSe2-origin-Pt,

CoSe2-x, and CoSe2-x-Pt performed in 1.0 M KOH (I); inset shows the effects of different loading amounts and dopants on OER performance. Energy

profiles of the CoSe2-origin, CoSe2-x, CoSe2-x-Pt, CoSe2-x-Ni, and CoSe2-x-Ru for OER (J). Reprinted with permission from Zhang and co-workers.28

Copyright 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.


the simultaneous creation of surface oxygen vacancies and Fe4+ species with an

optimal eg orbital filling (t2g3eg

1) (Figures 6N–6P).60

However, it is still a grand challenge to retain the stability of these defect sites, espe-

cially under the harsh redox process of electrocatalysis. To this end, coordination of

the defect sites with non-metal (or metal) atomic species in defect-rich carbons, g-

C3N4, TMDs, and other metal compounds could be a potential strategy for stabiliz-

ing the topological structures of defects, preventing atomic metal species from ag-

gregation, and even further manipulating the local electronic distribution through

the diverse coordinated configurations.30,61–64 Recently, Wang and co-workers suc-

cessfully filled the in-situ-generated oxygen vacancies of Co3O4 with P by a facile

plasma strategy as an electrocatalyst for the overall water splitting (Figure 7A).65

In Co K-edge XANES spectra, with P filled in the O vacancy site, the arising edge

of pre-peak of P-Co3O4 has the lowest energy intensity (Figure 7B), indicating that

the electron can be transferred more easily to the Co 3d orbital of P-Co3O4 than pris-

tine Co3O4 and VO-Co3O4. The P-Co3O4 exhibited higher OER performance than

pristine Co3O4 and VO-Co3O4 and only required an overpotential of 280 mV to

deliver the current density of 10 mA cm�2 (Figure 7C), which is attributed to the sig-

nificant shrink in band gap from pristine Co3O4 to P-Co3O4 (Figure 7D). Yao and co-

14 Chem 5, 1–27, June 13, 2019


workers successfully used the Ar plasma to create Se vacancies on the CoSe2 nano-

sheets with the thickness of 1.8 nm and then load the Pt single atoms onto the nano-

sheet surface through ultraviolet irradiation (Figures 7E and 7F).28 X-ray absorption

spectroscopy and high-angle annular dark-field scanning transmission electron mi-

croscopy (HAADF-STEM) tests both verified the existence of Pt single atoms in the

form of atomically coordinated Pt-Co-Se moieties (Figures 7G and 7H). The filling

of single Pt could remarkably enhance the OER activity of CoSe2 (CoSe2-x-Pt), which

wasmuch better than that of single Ni and even Ru atomic species filled CoSe2-x (Co-

Se2-x-Ni and CoSe2-x-Ru) (Figure 7I). Mechanism studies have unraveled that the sin-

gle Pt can induce much higher electronic distribution asymmetry degree than either

single Ni or Ru and synergistically benefit the interaction between the Co sites and

adsorbates (OH*, O*, and OOH*) duringOER process, leading to a superior OER ac-

tivity (Figure 7J).

Hydrogen Evolution Reaction

Electrochemical water splitting to produce H2 is recognized as a sustainable solution

for renewable energy storage and transportation, in which bulk Pt is known as the

best HER catalyst. In pursuit of a higher cost efficiency, different strategies such as

improving Pt atom efficiency and developing non-Pt catalysts have been widely

explored to date.11,66,67

By electrodepositing Pt onto a bismuth ultramicroelectrode, Bard and co-workers

illustrated a positive dependence of intrinsic Pt-catalyzed HER activity with its aggre-

gated atom number.68 Therefore, how to manipulate different Pt coordination struc-

ture toward a maximum Pt usage while maintaining the high activity is a prerequisite.

Wu and co-workers reported an iced-photochemical reduction method of preparing

atomically dispersed Pt coordinating with different carbon defects (single vacancy,

double vacancy, and edge sites) over mesoporous carbon (MC) substrate.69 As

shown in Figure 8A, this Pt1/MC catalyst delivered a current density of 100 mA

cm�2 at an overpotential of 65 mV, superior to that of the state-of-the-art Pt/C.

Calculated Gibbs free energies for atomic H adsorption, DGH*, on different MC de-

fects and edges were found to be smaller than those on Pt(111), confirming the

enhancement of HER activities in Pt-Cdefects motif. Similar to Pt-C,70 Pt-N-C71 and

Pt1-O2-Fe1-N4-C72 coordination have also been reported to boost HER performance

on atomic Pt sites.

In addition to this carbon-defect-based structure engineering, other HER active mo-

tifs (such as isolated Pt atoms anchored on CoP-based nanotube arrays73 or few-

layer MoS2 nanosheets,27 Pt-O coordination dispersed over TiO2,

74 and atomically

dispersed Cu-Pt dual sites on Pd nanorings75) have been successfully demonstrated.

Bao and co-workers came up with a strategy for triggering the HER activity of an inert

MoS2 surface by replacing Mo atoms in MoS2 with various single-atom metals (Fig-

ure 8B).27 In this system, the dopedmetal atoms can tune the adsorption behavior of

H atoms on the neighboring S sites, resulting in a significantly enhanced HER activity

on theMoS2 surface. More importantly, the reactants and harsh environments do not

directly contact the metal dopants because of protection from the trilayered sand-

wich structure of MoS2, which thereby increases the catalytic stability and anti-poi-

son ability. Atomic Pd-doped MoS2 to introduce sulfur vacancy and to trigger H

adsorption has been explored as well.76 Figure 8C shows the highly oxidized state

and localized Pt-O coordination structure of a Pt/TiO2 catalyst;74 the 5d orbital of

the oxidized Pt atoms appears to hybridize with the H 1s orbital to form weak Pt-H

valence bonds, leading to a DGH* approximately 0 eV and experimentally an excel-

lent electrocatalytic HER activity.

Chem 5, 1–27, June 13, 2019 15

Figure 8. Atomic Pt Embedded in Defective Substrates as HER Catalysts

(A) HAADF-STEM image of Pt/MC catalyst prepared via iced-photochemical reduction, as well as its HER performance and energetics analysis.

Reprinted with permission from Wei et al.69 Copyright 2017 Springer Nature.

(B) Volcano curve of HER currents (log(i0)) versus DGH over different metal-doped MoS2 motifs. Reprinted with permission from Deng et al.27 Copyright

2015 Royal Society of Chemistry.

(C) HER free-energy diagram and projected density of states for different H* absorption sites on a PtOx cluster and the TiO2(101) surface. Reprinted with

permission from Cheng et al.74 Copyright 2017 Royal Society of Chemistry.


Besides Pt coordination structure manipulation, developing other earth-abundant

metal coordination motifs is another strategy for upgrading HER cost efficiency.

Nanoscale MoS2 as a representative nonprecious TMD material has attracted

ever-growing interest as a HER catalyst. It is known that the active sites of pristine

MoS2 are located at the edges,77–79 leaving a large area of basal planes useless.

To better address this concern, Wang and co-workers theoretically evaluated the ca-

pabilities of 16 kinds of point defects and grain boundaries to activate the basal

plane of MoS2 monolayer (Figure 9A).80 Six types of defects have been identified

to greatly improve the HER performance of the in-plane domains of MoS2, among

which Vs on MoS2 point defects and S bridge and 4j8a grain boundaries exhibit

outstanding activity in both Heyrovsky and Tafel reactions. Experimentally, Xie

and colleagues put forward defect engineering on the basal (002) planes to increase

the exposure of active edge sites by forming cracks on the surfaces of the nano-

sheets (Figure 9B).81 Ajayan and co-workers demonstrated that either oxygen

plasma exposure or hydrogen treatment on pristine monolayer MoS2 could intro-

duce more active sites via the formation of defects within the monolayer, leading

to a high density of exposed edges and a significant improvement of the hydrogen

evolution activity.82 These as-fabricated defects are characterized at the scale from

macroscopic continuum to discrete atoms, as shown in Figure 9C. Besides MoS2,

carbon matrices are also widely explored because of their abundance of defect

16 Chem 5, 1–27, June 13, 2019


configurations for potential metal-support coordination engineering.83 Typically,

Mo-N-C,84 Co-N-C moieties dispersed over graphene (or graphitized carbon),85,86

and atomic Ni- or Fe-trapped in graphdiyne87 are recognized as highly active sites

for HER. Several representative atomic structure characterizations are adapted in

Figures 9D–9H. Taking Co-N-C for example, Figure 9I shows the HER activity com-

parison between N-C, Co-C, and Co-N-C moieties. The CoNx/C catalyst showed an

overpotential of only 133 mV at a current density of 10 mA cm-2, which is only

�100 mV lower than that of the benchmark Pt/C catalyst. Furthermore, rotating

ring-disk electrode measurement provides a more accurate evaluation of the onset

potential of HER, i.e., ca. 20 mV overpotential on Co-N-C, 60 mV on Co-C, and

220 mV on N-C motifs. Detailed Tafel slope analysis sheds light on the different

rate-limiting step (RLS) between these three coordination structures: N-C and Co-

C show a Tafel slope of �98 and �106 mV per decade, respectively, suggesting a

RLS of initial proton adsorption on these two catalysts. Co-N-C reveals a Tafel slope

of �57 mV per decade, which could suggest hydrogen production via the Volmer-

Heyrovsky mechanism, and that the electrochemical desorption step is rate limiting.

Another representative HER case is Mo1-N1-C2 versus Mo-N and Mo-C catalytic mo-

tifs,84 in which an onset overpotential of 13 mV has been achieved on Mo1-N1-C2,

much lower than that of the Mo-C (107 mV) and Mo-N (121 mV) counterparts. Theo-

retical calculations confirm that DGH* is lower on Mo-N-C than on Mo-C, Mo-N, and

N-C, highlighting the vital role of coordination-structure engineering to finely tune

HER performance. Besides metal-N-C-based coordination structures, bare carbon

defects with diverse topological motifs can also finely tune the electronic structure

of anchored atomic metal species.30,88,89 For instance, Yao and co-workers em-

ployed an impregnation method with acid leaching to prepare defective-gra-

phene-trapped atomic Ni catalysts (denoted as A-Ni@DG) with Ni loading up to

1.24%, which exhibited superb HER performance in 0.5 M H2SO4 electrolyte

(hHER = 70 mV at a current density of 10 mA cm�2) superior to that of Ni@DG and

DG (Figure 10A).30 Remarkably, the A-Ni@DG catalyst has exhibited much higher

turnover frequency (TOF) values (more than ten times) than those of non-precious-

metal-based catalysts at various overpotentials, suggesting the high utilizing effi-

ciency of Ni in the A-Ni@DG catalyst (Figure 10B). DFT calculations revealed that

the optimal hydrogen desorption energy of A-Ni@DG can be achieved by atomic

Ni-D5775 defect coordinated motifs, approaching that of Pt (Figure 10C).

Beyond point-defect motifs, Yao and co-workers further demonstrated a 2D hetero-

structured NiFe LDH-NS@DG hybrid by coupling exfoliated Ni-Fe layered double

hydroxide (LDH) nanosheet (NS) and DG as a bifunctional electrocatalyst for overall

water splitting.90 Remarkably, this catalyst has exhibited excellent electrocatalytic

performance in overall alkaline water splitting (20 mA cm�2 at 1.5 V), ranking better

than other non-noble-metal bifunctional catalysts (Figures 10D and 10E). For prac-

tical applications, this catalyst can be activated by only a 1.5 V solar panel (Fig-

ure 10F), which has high potential for the distributed energy storage and conversion

systems. The simulation of electronic structures in NiFe LDH-NS@DG catalyst has re-

vealed that the formed heterostructure can efficiently enhance the charge separa-

tion and redistribution (electron-rich zones on diverse carbon defect sites and elec-

tron-deficient zones on NiFe LDH-NS) (Figure 10G), thus facilitating the HER and

OER conducted on DG and NiFe LDH-NS, respectively (Figure 10H). More impor-

tantly, to meet the industrial application by mass production of efficient electrocata-

lysts, Yao and co-workers applied the developed defect tailoring strategies into

cheap and earth-abundant resources, such as activated carbons and seaweed- or

macadamia-nut-shell-derived carbon-based materials, which have been summa-

rized in the recent review on defect electrocatalysis.13

Chem 5, 1–27, June 13, 2019 17

Figure 9. Defective Structure Engineering of Non-Pt Catalysts toward HER

(A) Catalytic reaction pathways and energies of the HER on MoS2 basal with structural defects of grain boundaries and point defects. Reprinted with

permission from Ouyang et al.80 Copyright 2016 American Chemical Society.

(B) Cross-sectional HRTEM images for defective-rich MoS2 ultrathin nanosheets. Reprinted with permission from Xie et al.81 Copyright 2013 Wiley-VCH

Verlag GmbH & Co. KGaA, Weinheim.

(C) Morphology of monolayer MoS2 with either oxygen plasma exposure (left) or hydrogen annealing treatment (right), where the cracks and nanometer-

scale holes appear in the basal plane. Reprinted with permission from Ye et al.82 Copyright 2016 American Chemical Society.

(D) Representative STEM structure characterizations of Ni-doped graphene (D). Reprinted with permission from Qiu et al.88 Copyright 2015 Wiley-VCH

Verlag GmbH & Co. KGaA, Weinheim.

(E) Isolated Mo1N1C2 over N-doped carbon. Reprinted with permission from Chen et al.84 Copyright 2017 Wiley-VCH Verlag GmbH & Co. KGaA,

Weinheim.

(F) Atomic dispersed Co over N-doped graphene. Reprinted with permission from Fei et al.85 Copyright 2015 Springer Nature.

18 Chem 5, 1–27, June 13, 2019


Figure 9. Continued

(G and H) Atomic dispersed Ni (G) and Fe (H) over N-doped graphdiyne. Reprinted with permission from Xue et al.87 Copyright 2018 Springer Nature.

(I) HER polarization plots of the CoNx/C, N/C, Co/N, and Pt/C catalysts, together with the RRDE results and Tafel analysis in 0.5 M H2SO4. Reprinted with

permission from Liang et al.86 Copyright 2015 Springer Nature.


CO2 Reduction Reaction

Electrocatalytic CO2RR has attracted considerable interest in the past years by

serving as a promising solution for closing anthropogenic carbon cycle and delocal-

ized energy storage by converting CO2 into renewable fuels and chemical feed-

stocks.91–93 In addition to the well-known Cu-, Au-, and Ag-based catalysts, sin-

gle-metal or metalloid-heteroatom-doped carbon catalysts consisting of earth-

abundant and low-cost elements are an emerging material for selective CO2RR.94,95

From a theoretical view, Siahrostami and co-workers reported an energetically favor-

able N doping near the carbon defects arisen from the defect-induced strain to sta-

bilize N dopant and examined the CO2RR activity of a wide range of different

possible active sites in both pristine and defective carbon structures with or without

N doping (Figure 11A).19 Of special interest, the examined defective carbon active

sites do not fall on the known scaling line between *COOH and *CO adsorption en-

ergies for the transition metals because they do not bind CO. This is the reason that

carbon-based materials represent especially interesting alternatives for the 2e�

reduction of CO2 to CO. On the basis of a simple thermodynamic limiting potential

analysis, pyridinic N is suggested to be the most selective site for the CO2RR over

HER. Experimentally, Ajayan and co-workers reported a similar active assignment

of the pyridinic N, which retains a lone pair of electrons that are capable of binding

CO2 and lowering its activation barrier to *COOH, in nitrogen-doped CNTs (N-

CNTs; Figure 11B) for the CO2-to-CO conversion.96,97 Other carbon motifs with pyr-

idinic defects, such as 3D graphene foam,98 nanoporous carbon,99 quaternary and

graphitic nitrogen-doped CNTs,100,101 and steam-etched carbon frameworks,102

have also been explored to date. In addition to these sp2 hybridization carbon

matrices of graphene and CNTs, another kind of sp3 hybridization nanodiamond car-

bon with metalloid doping has also been investigated toward CO2 reduction prod-

ucts beyond CO. For example, a boron-doped diamond with a p-type doped surface

can produce formaldehyde at a maximum faradic efficiency (FE) of 74%,103 and a N-

doped nanodiamond supported on a Si rod array could generate acetate and

formate with 91.8% FE at �1.0 V with a proposed reaction pathway of CO2 /

CO2$– / (COO)2

$ / CH3COO– (Figure 11C).104

Bulk Cu electrode is known to generate multiple C2+ products from CO2RR; never-

theless, once atomic Cu sites are dispersed over Ni(OH)2105 or CeO2

106 substrates,

these isolated Cu centers can no longer catalyze the C-C coupling during CO2RR but

instead effectively convert CO2 into CO and CH4. Moreover, by delicate control of

the catalytic metal-C (or -N) motif at the atomic scale, other earth-abundant transi-

tion metals such as Ni, Co, and Fe can be employed to deliver a comparable CO2-

to-CO performance with a benchmark Au electrode.

Similar to molecular metal-porphyrin107 and metal-cyclam108 catalysts, M-N4 moi-

eties dispersed over different carbon supports with four neighboring nitrogen atoms

have been widely reported as active sites for reducing CO2 to CO.109–114 Wu and co-

workers prepared Co-Nx-C catalysts with fine-tuned Co-N coordination numbers by

controlling the volatile C-N fragments at different pyrolysis temperatures.115 Co-N2

sites exhibited the best CO2RR performance with a CO TOF of 18,200 h�1 at an over-

potential of 0.52 V. Fontecave and co-workers demonstrated a CO selectivity

Chem 5, 1–27, June 13, 2019 19

Figure 10. Defective Carbon Structures for Coupling Atomic Metal Species or 2D Nanomaterials toward Water Splitting

(A–C) HER polarization curves of DG, Ni@DG, A-Ni@DG, and Pt/C performed in 0.5 M H2SO4 electrolyte (A). TOF curve of A-Ni@DG and other reported

catalysts for HER (B). Energy profiles of three Ni-C coordination motifs and Pt for HER (C). Reprinted with permission from Zhang et al.30 Copyright 2018

Elsevier.

(D–H) LSV curve of NiFe LDH-NS@DG as an OER and HER bifunctional catalyst in 1 M KOH for overall water splitting (D); inset TEM image shows its

morphology. Comparison of the NiFe LDH-NS@DG catalyst with other state-of-the-art noble-metal-free bifunctional catalysts for overall water splitting

in alkaline (E). Demonstration of a solar-power-assisted water-splitting device with a voltage of 1.5 V (F). DFT calculation studies of Ni-Fe LDH-NS@DG

(DG-5, DG-585, or DG-7557)-based heterostructures (G). Schematic of the probable electrocatalytic mechanism of NiFe LDH-NS@DG as a bifunctional

catalyst for overall water splitting (H). Reprinted with permission from Jia et al.90 Copyright 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.


dependence on Fe loading over a series of Fe-N4-C catalysts derived from zeolitic

imidazolate frameworks,116 in which the electrocatalytic selectivity between

CO2RR and HER was governed by the ratio of isolated Fe-N4 sites versus Fe-based

nanoparticles. In practice, a simple acid-leaching pre-treatment could be used to re-

move residual metal clusters prior to CO2RR measurement to maximum CO2RR ef-

ficiency from the isolated coordinative metal motif.113,117

20 Chem 5, 1–27, June 13, 2019

Figure 11. Defective-Carbon-Motif-Catalyzed CO2RR

(A) Demonstration of optimized carbon defective structures, together with calculated chemisorption energies of *COOH versus *CO intermediates and

UL(CO2)�UL(H2). Reprinted with permission from Siahrostami et al.19 Copyright 2017 American Chemical Society.

(B) The onset potential and maximum FE for CO formation as a function of N content in the synthesized N-CNTs. Reprinted with permission from Sharma

et al.96 Copyright 2015, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

(C) N-doped nanodiamond supported on Si rod array with N-sp3 C species catalyzes CO2 reduction to carboxylic acid. Reprinted with permission from

Liu et al.104 Copyright 2015 American Chemical Society.


Figure 12A shows the design of single Ni atomic sites coordinated within graphene

defects.118 DFT calculations suggest that the graphene single-vacancy (SV) and/or

double-vacancy (DV) coordinated atomic Ni site is responsible for weakening CO

adsorption and suppressing HER side reactions. Three-dimensional atomic probe

tomography of over �10,000 individual Ni single atoms clearly reveals a direct

and predominant Ni-C coordination. As a result, these Ni-C active sites deliver a

maximum CO FE of 93.2% with a TOF of 28,800 h�1 at an overpotential of 0.7 V.

Furthermore, by replacing one or two neighboring carbon atoms with nitrogen,

Wang and co-workers comparatively investigated the CO2RR and HER pathways

over M@SV/DV with or without neighboring N.119 Whereas Co-N@SV/DV catalyzed

HER exclusively, the Ni-N@DV moieties maintained a CO FE above 90% for more

than 20 h of CO2RR electrolysis. In our latest work, by creating abundant O and N

defects over carbon spheres (Figure 12B), we developed a facile and scalable syn-

thesis of Ni-single-atom catalysts on commercial carbon black, which we further em-

ployed in a gas-phase electrocatalytic reactor under ambient conditions.120 As a

result, those single atomic sites in carbon vacancies yielded a CO2-to-CO reduction

current density above 100 mA cm-2 with nearly 100% selectivity for CO and around

Chem 5, 1–27, June 13, 2019 21

Figure 12. CO2RR over Heteroatom-Doped Carbon Matrix

(A) Thermodynamic analysis of CO2RR over different atomic sites of Ni-doped carbon vacancies with or without neighboring N, together with direct

coordination nature determined by 3D atomic probe tomography. Reprinted with permission from Jiang et al.118 Copyright 2017 Elsevier.

(B) Schematic and characterization of gram-scale synthetical procedure of defective carbon black supported atomic Ni catalysts, which delivers a

CO2RR current above 8 A and maintains a CO selectivity above 90%. Reprinted with permission from Zheng et al.120 Copyright 2019 Elsevier.

(C) STEM image and operando X-ray spectroscopy for CO2RR over Ni-NG. Reprinted with permission from Yang et al.121 Copyright 2018 Springer

Nature.

(D) Atomic characterization and Wavelet transforms for Co-N5 moieties supported on N-doped carbon. Reprinted with permission from Pan et al.122

Copyright 2018 American Chemical Society.


1% toward the HER side reaction. Moreover, the overall current in one unit cell could

easily ramp up to more than 8 A while maintaining an exclusive CO evolution with a

generation rate of 3.34 L h�1 per unit cell. This demonstration could pave the way

toward practical CO2 reduction in the future.

22 Chem 5, 1–27, June 13, 2019


Also noteworthy is the development of advanced operando spectroscopy, which

sheds light on the local coordination structure assignment and therefore provides

a better understanding of reaction mechanisms. By employing operando X-ray ab-

sorption and photoelectron spectroscopy measurements, Liu and co-workers iden-

tified the monovalent Ni(i) atomic center with a d9 electronic configuration in Ni-NG

as the catalytically active site for CO2RR (Figure 12C).121 A charge transfer was first

observed from the unpaired electron in the 3dx2-y2 orbital of Ni(i) to the carbon 2p

orbital in CO2 to form a CO2d� species at open circuit potential. Thereafter, the Ni

K-edge of Ni-NG shifted back to lower energy in CO2RR working potential, suggest-

ing the recovery of low-oxidation-state Ni after one cycle of CO2 reduction (Ni(i)* +

CO2 + 2H+ + 2e� / Ni(i)* + CO + H2O). In another recent report, Li and co-workers

anchored cobalt phthalocyanine (CoPC) on a polymer-derived hollow N-doped

porous carbon sphere support,122 which showed a 15.5-fold enhancement of

CO2-to-CO activity over the CoPc precursor (Figure 12D). By comparing the wavelet

transform contour plots between supported and unsupported CoPc, as well as the

fitting of Co K-edge extended absorption spectra, the authors demonstrated the

coordinative interaction between single Co atom in CoPc and N atom in N-doped

carbon substrate, leading to the active site of Co-N5 for CO2RR.

CONCLUSIONS AND PERSPECTIVES

In summary, from the viewpoint of crystallography, defects can be regarded as any

inhomogeneous composition in a material, which can include atom and ion va-

cancy or non-metal (or metal)-atom substitution in zero dimensions, irregular sur-

face or multiphase interface in high dimensions, and so on. The role of defects

can not only serve as the anchoring site to stabilize single non-metal (or metal) het-

eroatoms but also help to tune the electronic and/or geometric structures of the

defect-based coordination motifs, thus facilitating specific electrochemical reac-

tions (ORR, OER, HER, and CO2RR in this review). In addition, through manipula-

tion of material microstructures, spatial defects induced from in situ and ex situ

synthetic strategies can further promote the mass transport of intermediate reac-

tants on the solid-liquid-gas triple-phase boundary regions to enhance the kinetics

of electrocatalysis. However, despite the numerous research progress that has

been made in investigating the defects of electrocatalysts, several significant chal-

lenges still exist, and these deserve continuing attention for exploration of the ef-

fect of defects on electrocatalysts:

(1) Control of defect types: because different electrochemical reactions possess

distinct energy barriers, it is desirable to fabricate ideal electrocatalysts with a

single type of defect as the active site to maximize the catalytic process

without side effects. Thus, new control methods on defect engineering

need to be urgently explored to better elucidate the fundamental correlation

between the defects and the activity.

(2) Stability of defects: although various catalysts with defects show desirable

stability in diverse electrochemical reactions, it should be determined

whether the defects are changed or reconstructed under harsh working con-

ditions. Therefore, tracking the possible evolution of defect-based motifs is

very essential for researchers to establish the correct proof of concept for

catalyst design.

(3) Advanced operando characterizations: to address the above two challenges,

more accurate and advanced in situ measurements should be coupled with

theoretical insights, which could help us dynamically observe the role of de-

fects during the real-time process of electrocatalysis.

Chem 5, 1–27, June 13, 2019 23


(4) Last but not the least, further scaling up the energy conversion performance

of defective materials calls for continuous efforts in both robust catalyst

design and interface optimization with well-defined mass and charge trans-

port.

All in all, we believe that the fundamental understanding of the role of defects in cat-

alysts will promote the development of the whole field in electrocatalysis and push

them into industrial applications.

ACKNOWLEDGMENTS

The authors are grateful for the financial support of the Australian Research Council

(ARC; DP170103317). The work was supported by the Rowland Fellows Program at

Rowland Institute of Harvard University. H.W. acknowledges support from Rice Uni-

versity. Y.J. also thanks the ARC for the Discovery Early Career Researcher Award

(DE180101030).

AUTHOR CONTRIBUTIONS

X.Y. and H.W. proposed the topic of the review. Y.J. and K.J. contributed to inves-

tigating the literature and wrote the manuscript. All the authors discussed and

revised the manuscript.

REFERENCES AND NOTES

1. Chu, S., and Majumdar, A. (2012).Opportunities and challenges for asustainable energy future. Nature 488,294–303.

2. Obama, B. (2017). The irreversiblemomentum of clean energy. Science 355,126–129.

3. Seh, Z.W., Kibsgaard, J., Dickens, C.F.,Chorkendorff, I.B., Nørskov, J.K., andJaramillo, T.F. (2017). Combining theoryand experiment in electrocatalysis: insightsinto materials design. Science 355,eaad4998.

4. Jiao, Y., Zheng, Y., Jaroniec, M.T., and Qiao,S.Z. (2015). Design of electrocatalysts foroxygen- and hydrogen-involving energyconversion reactions. Chem. Soc. Rev. 44,2060–2086.

5. Stamenkovic, V.R., Strmcnik, D., Lopes, P.P.,and Markovic, N.M. (2016). Energy and fuelsfrom electrochemical interfaces. Nat. Mater.16, 57–69.

6. Liu, M., Pang, Y.J., Zhang, B., De Luna, P.,Voznyy, O., Xu, J.X., Zheng, X.L., Dinh, C.T.,Fan, F.J., Cao, C.H., et al. (2016). Enhancedelectrocatalytic CO2 reduction via field-induced reagent concentration. Nature 537,382–386.

7. Gasteiger, H.A., and Markovi�c, N.M. (2009).Chemistry. Just a dream-or future reality?Science 324, 48–49.

8. Tang, C., Wang, H.F., and Zhang, Q. (2018).Multiscale principles to boost reactivity ingas-involving energy electrocatalysis. Acc.Chem. Res. 51, 881–889.

9. Zhu, Y.P., Guo, C.X., Zheng, Y., and Qiao, S.Z.(2017). Surface and interface engineering ofnoble-metal-free electrocatalysts for efficient

24 Chem 5, 1–27, June 13, 2019

energy conversion processes. Acc. Chem.Res. 50, 915–923.

10. Asefa, T. (2016). Metal-free and noble metal-free heteroatom-doped nanostructuredcarbons as prospective sustainableelectrocatalysts. Acc. Chem. Res. 49, 1873–1883.

11. Chen, Y.J., Ji, S.F., Chen, C., Peng, Q., Wang,D.S., and Li, Y.D. (2018). Single-atom catalysts:Synthetic strategies and electrochemicalapplications. Joule 2, 1242–1264.

12. Yan, X.C., Jia, Y., and Yao, X.D. (2018). Defectson carbons for electrocatalytic oxygenreduction. Chem. Soc. Rev. 47, 7628–7658.

13. Jia, Y., Chen, J., and Yao, X.D. (2018). Defectelectrocatalytic mechanism: concept,topological structure and perspective. Mater.Chem. Front. 2, 1250–1268.

14. Yan, D.F., Li, Y.X., Huo, J., Chen, R., Dai, L.M.,and Wang, S.Y. (2017). Defect chemistry ofnonprecious-metal electrocatalysts foroxygen reactions. Adv. Mater. 29, 20.

15. Tang, C., and Zhang, Q. (2017).Nanocarbon for oxygen reductionelectrocatalysis: dopants, edges, anddefects. Adv. Mater. 29, 9.

16. Wang, Y.F., Han, P., Lv, X.M., Zhang, L.J., andZheng, G.F. (2018). Defect and interfaceengineering for aqueous electrocatalytic CO2

reduction. Joule 2, 2551–2582.

17. Kotakoski, J., Krasheninnikov, A.V., Kaiser, U.,and Meyer, J.C. (2011). From point defects ingraphene to two-dimensional amorphouscarbon. Phys. Rev. Lett. 106, 105505.

18. Hou, Z.F., Wang, X.L., Ikeda, T., Terakura, K.,Oshima, M., and Kakimoto, M. (2013).

Electronic structure of N-doped graphenewith native point defects. Phys. Rev. B 87, 16.

19. Siahrostami, S., Jiang, K., Karamad, M., Chan,K.R., Wang, H.T., and Nørskov, J. (2017).Theoretical investigations into defectedgraphene for electrochemical reduction ofCO2. ACS Sustainable Chem. Eng. 5, 11080–11085.

20. Zhao, H.Y., Sun, C.H., Jin, Z., Wang, D.W.,Yan, X.C., Chen, Z.G., Zhu, G.S., and Yao, X.D.(2015). Carbon for the oxygen reductionreaction: a defect mechanism. J. Mater.Chem. A 3, 11736–11739.

21. Zhang, L.P., Xu, Q., Niu, J.B., and Xia, Z.H.(2015). Role of lattice defects in catalyticactivities of graphene clusters for fuel cells.Phys. Chem. Chem. Phys. 17, 16733–16743.

22. Jia, Y., Zhang, L.Z., Du, A.J., Gao, G.P., Chen,J., Yan, X.C., Brown, C.L., and Yao, X.D. (2016).Defect graphene as a trifunctional catalyst forelectrochemical reactions. Adv. Mater. 28,9532–9538.

23. Liu, Y.W., Xiao, C., Huang, P.C., Cheng, M.,and Xie, Y. (2018). Regulating the charge andspin ordering of two-dimensional ultrathinsolids for electrocatalytic water splitting.Chem 4, 1263–1283.

24. Deng, D.H., Yu, L., Pan, X.L., Wang, S., Chen,X.Q., Hu, P., Sun, L.X., and Bao, X.H. (2011).Size effect of graphene on electrocatalyticactivation of oxygen. Chem. Commun.(Camb.) 47, 10016–10018.

25. Dou, S., Tao, L., Wang, R.L., El Hankari, S.,Chen, R., and Wang, S.Y. (2018). Plasma-assisted synthesis and surface modification ofelectrode materials for renewable energy.Adv. Mater. 30, 24.

http://refhub.elsevier.com/S2451-9294(19)30065-8/sref1



















































































































26. Wang, X., Zhuang, L.Z., Jia, Y., Liu, H.L., Yan,X.C., Zhang, L.Z., Yang, D.J., Zhu, Z.H., andYao, X.D. (2018). Plasma-triggered synergy ofexfoliation, phase transformation and surfaceengineering in cobalt diselenide forenhanced water oxidation. Angew. Chem. Int.Ed. 57, 16421–16425.

27. Deng, J., Li, H.B., Xiao, J.P., Tu, Y.C., Deng,D.H., Yang, H.X., Tian, H.F., Li, J.Q., Ren, P.J.,and Bao, X.H. (2015). Triggering theelectrocatalytic hydrogen evolution activity ofthe inert two-dimensional MoS2 surface viasingle-atom metal doping. Energy Environ.Sci. 8, 1594–1601.

28. Zhuang, L.Z., Jia, Y., Liu, H.L., Wang, X.,Hocking, R.K., Liu, H.W., Chen, J., Ge, L.,Zhang, L.Z., Li, M.R., et al. (2019). Defect-induced Pt–Co–Se coordinated sites withhighly asymmetrical electronic distribution forboosting oxygen-involving electrocatalysis.Adv. Mater. 30, e1805581.

29. Zhuang, L.Z., Jia, Y., He, T.W., Du, A.J., Yan,X.C., Ge, L., Zhu, Z.H., and Yao, X.D. (2018).Tuning oxygen vacancies in two-dimensionaliron-cobalt oxide nanosheets throughhydrogenation for enhanced oxygenevolution activity. Nano Res. 11, 3509–3518.

30. Zhang, L.Z., Jia, Y., Gao, G.P., Yan, X.C., Chen,N., Chen, J., Soo, M.T., Wood, B., Yang, D.J.,Du, A.J., et al. (2018). Graphene defects trapatomic Ni species for hydrogen and oxygenevolution reactions. Chem 4, 285–297.

31. Tang, C., Wang, B., Wang, H.F., and Zhang,Q. (2017). Defect engineering toward atomicCo-Nx-C in hierarchical graphene forrechargeable flexible solid Zn-Air batteries.Adv. Mater. 29, 7.

32. Wang, H.P., Li, X.B., Gao, L., Wu, H.L., Yang,J., Cai, L., Ma, T.B., Tung, C.H., Wu, L.Z., andYu, G. (2018). Three-dimensional graphenenetworks with abundant sharp edge sites forefficient electrocatalytic hydrogen evolution.Angew. Chem. Int. Ed. 57, 192–197.

33. Zhao, X.J., Zou, X.Q., Yan, X.C., Brown, C.L.,Chen, Z.G., Zhu, G.S., and Yao, X.D. (2016).Defect-driven oxygen reduction reaction(ORR) of carbon without any element doping.Inorg. Chem. Front. 3, 417–421.

34. Yin, P.Q., Yao, T., Wu, Y., Zheng, L.R., Lin, Y.,Liu, W., Ju, H.X., Zhu, J.F., Hong, X., Deng,Z.X., et al. (2016). Single cobalt atoms withprecise N-coordination as superior oxygenreduction reaction catalysts. Angew. Chem.Int. Ed. 55, 10800–10805.

35. Zhang, L.Z., Fischer, J.M.T.A., Jia, Y., Yan,X.C., Xu, W., Wang, X.Y., Chen, J., Yang, D.J.,Liu, H.W., Zhuang, L.Z., et al. (2018).Coordination of atomic Co-Pt couplingspecies at carbon defects as active sites foroxygen reduction reaction. J. Am. Chem. Soc.140, 10757–10763.

36. Armel, V., Hindocha, S., Salles, F., Bennett, S.,Jones, D., and Jaouen, F. (2017). Structuraldescriptors of zeolitic-imidazolateframeworks are keys to the activity of Fe-N-Ccatalysts. J. Am. Chem. Soc. 139, 453–464.

37. Liu, D.B., Ni, K., Ye, J.L., Xie, J., Zhu, Y.W., andSong, L. (2018). Tailoring the structure ofcarbon nanomaterials toward high-endenergy applications. Adv. Mater. 30.

38. Li, H., Tsai, C., Koh, A.L., Cai, L.L., Contryman,A.W., Fragapane, A.H., Zhao, J.H., Han, H.S.,Manoharan, H.C., Abild-Pedersen, F., et al.(2016). Activating and optimizing MoS2 basalplanes for hydrogen evolution through theformation of strained sulphur vacancies. Nat.Mater. 15, 48–53.

39. Zheng, Y., Jiao, Y., Li, L.H., Xing, T., Chen, Y.,Jaroniec, M., and Qiao, S.Z. (2014). Towarddesign of synergistically active carbon-basedcatalysts for electrocatalytic hydrogenevolution. ACS Nano 8, 5290–5296.

40. Zhao, Z.H., Li, M.T., Zhang, L.P., Dai, L.M., andXia, Z.H. (2015). Design principles forheteroatom-doped carbon nanomaterials ashighly efficient catalysts for fuel cells andmetal-air batteries. Adv. Mater. 27, 6834–6840.

41. Zhao, Y., Yang, L.J., Chen, S., Wang, X.Z., Ma,Y.W., Wu, Q., Jiang, Y.F., Qian, W.J., and Hu,Z. (2013). Can boron and nitrogen co-dopingimprove oxygen reduction reaction activity ofcarbon nanotubes? J. Am. Chem. Soc. 135,1201–1204.

42. Zheng, Y., Jiao, Y., Ge, L., Jaroniec, M., andQiao, S.Z. (2013). Two-step boron andnitrogen doping in graphene for enhancedsynergistic catalysis. Angew. Chem. Int. Ed.52, 3110–3116.

43. Wu, H.H., Li, H.B., Zhao, X.F., Liu, Q.F., Wang,J., Xiao, J.P., Xie, S.H., Si, R., Yang, F., Miao, S.,et al. (2016). Highly doped and exposedCu(I)-N active sites within graphene towardsefficient oxygen reduction for zinc-airbatteries. Energy Environ. Sci. 9, 3736–3745.

44. Liu, X., and Dai, L.M. (2016). Carbon-basedmetal-free catalysts. Nat. Rev. Mater. 1, 12.

45. Gong, K.P., Du, F., Xia, Z.H., Durstock, M., andDai, L.M. (2009). Nitrogen-doped carbonnanotube arrays with high electrocatalyticactivity for oxygen reduction. Science 323,760–764.

46. Guo, D.H., Shibuya, R., Akiba, C., Saji, S.,Kondo, T., and Nakamura, J. (2016). Activesites of nitrogen-doped carbon materials foroxygen reduction reaction clarified usingmodel catalysts. Science 351, 361–365.

47. Yang, H.B., Miao, J.W., Hung, S.F., Chen, J.Z.,Tao, H.B., Wang, X.Z., Zhang, L.P., Chen, R.,Gao, J.J., Chen, H.M., et al. (2016).Identification of catalytic sites for oxygenreduction and oxygen evolution in N-dopedgraphene materials: development of highlyefficient metal-free bifunctionalelectrocatalyst. Sci. Adv. 2, e1501122.

48. Liang, H.W., Zhuang, X.D., Bruller, S., Feng,X.L., and Mullen, K. (2014). Hierarchicallyporous carbons with optimized nitrogendoping as highly active electrocatalysts foroxygen reduction. Nat. Commun. 5, 4973.

49. Shen, A.L., Zou, Y.Q., Wang, Q., Dryfe,R.A.W., Huang, X.B., Dou, S., Dai, L.M., andWang, S.Y. (2014). Oxygen reduction reactionin a droplet on graphite: direct evidence thatthe edge is more active than the basal plane.Angew. Chem. Int. Ed. 53, 10804–10808.

50. Jiang, Y.F., Yang, L.J., Sun, T., Zhao, J., Lyu,Z.Y., Zhuo, O., Wang, X.Z., Wu, Q., Ma, J., andHu, Z. (2015). Significant contribution of

intrinsic carbon defects to oxygen reductionactivity. ACS Catal. 5, 6707–6712.

51. Tang, C., Wang, H.F., Chen, X., Li, B.Q., Hou,T.Z., Zhang, B.S., Zhang, Q., Titirici, M.M., andWei, F. (2016). Topological defects in metal-free nanocarbon for oxygen electrocatalysis.Adv. Mater. 28, 6845–6851.

52. Xue, L.F., Li, Y.C., Liu, X.F., Liu, Q.T., Shang,J.X., Duan, H.P., Dai, L.M., and Shui, J.L.(2018). Zigzag carbon as efficient and stableoxygen reduction electrocatalyst for protonexchange membrane fuel cells. Nat.Commun. 9, 3819.

53. Li, D.H., Jia, Y., Chang, G.J., Chen, J., Liu,H.W., Wang, J.C., Hu, Y.F., Xia, Y.Z., Yang,D.J., and Yao, X.D. (2018). A defect-drivenmetal-free electrocatalyst for oxygenreduction in acidic electrolyte. Chem 4, 2345–2356.

54. Chung, H.T., Cullen, D.A., Higgins, D., Sneed,B.T., Holby, E.F., More, K.L., and Zelenay, P.(2017). Direct atomic-level insight into theactive sites of a high-performance PGM-freeORR catalyst. Science 357, 479–484.

55. Li, J.Z., Chen, M.J., Cullen, D.A., Hwang, S.,Wang, M.Y., Li, B.Y., Liu, K.X., Karakalos, S.,Lucero, M., Zhang, H.G., et al. (2018).Atomically dispersedmanganese catalysts foroxygen reduction in proton-exchangemembrane fuel cells. Nat. Catal. 1, 935–945.

56. Banham, D., Ye, S., Pei, K., Ozaki, J.,Kishimoto, T., and Imashiro, Y. (2015). Areview of the stability and durability of non-precious metal catalysts for the oxygenreduction reaction in proton exchangemembrane fuel cells. J. Power Sources 285,334–348.

57. Choi, C.H., Kim, M., Kwon, H.C., Cho, S.J.,Yun, S., Kim, H.T., Mayrhofer, K.J.J., Kim, H.,and Choi, M. (2016). Tuning selectivity ofelectrochemical reactions by atomicallydispersed platinum catalyst. Nat. Commun. 7,10922.

58. Wang, Y.C., Zhou, T., Jiang, K., Da, P.M.,Peng, Z., Tang, J., Kong, B.A., Cai,W.B., Yang,Z.Q., and Zheng, G.F. (2014). Reducedmesoporous Co3O4 nanowires as efficientwater oxidation electrocatalysts andsupercapacitor electrodes. Adv. EnergyMater. 4, 7.

59. Xu, L., Jiang, Q.Q., Xiao, Z.H., Li, X.Y., Huo, J.,Wang, S.Y., and Dai, L.M. (2016). Plasma-engraved Co3O4 nanosheets with oxygenvacancies and high surface area for theoxygen evolution reaction. Angew. Chem. Int.Ed. 55, 5277–5281.

60. Zhu, Y.L., Zhou, W., Yu, J., Chen, Y.B., Liu,M.L., and Shao, Z.P. (2016). Enhancingelectrocatalytic activity of perovskite oxidesby tuning cation deficiency for oxygenreduction and evolution reactions. Chem.Mater. 28, 1691–1697.

61. Zheng, Y., Jiao, Y., Zhu, Y.H., Cai, Q.R.,Vasileff, A., Li, L.H., Han, Y., Chen, Y., andQiao, S.Z. (2017). Molecule-level g-C3N4

coordinated transition metals as a new classof electrocatalysts for oxygen electrodereactions. J. Am. Chem. Soc. 139, 3336–3339.

62. Dvorak, F., Camellone, M.F., Tovt, A., Tran,N.D., Negreiros, F.R., Vorokhta, M., Skala, T.,

Chem 5, 1–27, June 13, 2019 25


























































































































































































































Matolinova, I., Myslivecek, J., Matolin, V., et al.(2016). Creating single-atom Pt-ceria catalystsby surface step decoration. Nat. Commun. 7,10801.

63. Jones, J., Xiong, H.F., Delariva, A.T., Peterson,E.J., Pham, H., Challa, S.R., Qi, G.S., Oh, S.,Wiebenga, M.H., Pereira Hernandez, X.I.P.,et al. (2016). Thermally stable single-atomplatinum-on-ceria catalysts via atom trapping.Science 353, 150–154.

64. Liu, P.X., Zhao, Y., Qin, R.X., Mo, S.G., Chen,G.X., Gu, L., Chevrier, D.M., Zhang, P., Guo,Q., Zang, D.D., et al. (2016). Photochemicalroute for synthesizing atomically dispersedpalladium catalysts. Science 352, 797–801.

65. Xiao, Z.H., Wang, Y., Huang, Y.C., Wei, Z.X.,Dong, C.L., Ma, J.M., Shen, S.H., Li, Y.F., andWang, S.Y. (2017). Filling the oxygenvacancies in Co3O4 with phosphorus: an ultra-efficient electrocatalyst for overall watersplitting. Energy Environ. Sci. 10, 2563–2569.

66. Vesborg, P.C.K., Seger, B., and Chorkendorff,I. (2015). Recent development in hydrogenevolution reaction catalysts and their practicalimplementation. J. Phys. Chem. Lett. 6,951–957.

67. Duan, J.J., Chen, S., Jaroniec, M., and Qiao,S.Z. (2015). Heteroatom-doped graphene-based materials for energy-relevantelectrocatalytic processes. ACS Catal. 5,5207–5234.

68. Zhou, M., Dick, J.E., and Bard, A.J. (2017).Electrodeposition of isolated platinum atomsand clusters on bismuth-characterization andelectrocatalysis. J. Am. Chem. Soc. 139,17677–17682.

69. Wei, H.H., Huang, K., Wang, D., Zhang, R.Y.,Ge, B.H., Ma, J.Y., Wen, B., Zhang, S., Li, Q.Y.,Lei, M., et al. (2017). Iced photochemicalreduction to synthesize atomically dispersedmetals by suppressing nanocrystal growth.Nat. Commun. 8, 8.

70. Tavakkoli, M., Holmberg, N., Kronberg, R.,Jiang, H., Sainio, J., Kauppinen, E.I., Kallio, T.,and Laasonen, K. (2017). Electrochemicalactivation of single-walled carbon nanotubeswith pseudo-atomic-scale platinum for thehydrogen evolution reaction. ACS Catal. 7,3121–3130.

71. Cheng, N.C., Stambula, S., Wang, D., Banis,M.N., Liu, J., Riese, A., Xiao, B.W., Li, R.Y.,Sham, T.K., Liu, L.M., et al. (2016). Platinumsingle-atom and cluster catalysis of thehydrogen evolution reaction. Nat. Commun.7, 13638.

72. Zeng, X.J., Shui, J.L., Liu, X.F., Liu, Q.T., Li,Y.C., Shang, J.X., Zheng, L.R., and Yu, R.H.(2018). Single-atom to single-atom grafting ofPt-1 onto Fe-N-4 center: Pt-1@Fe-N-Cmultifunctional electrocatalyst withsignificantly enhanced properties. Adv.Energy Mater. 8, 8.

73. Zhang, L.H., Han, L.L., Liu, H.X., Liu, X.J., andLuo, J. (2017). Potential-cycling synthesis ofsingle platinum atoms for efficient hydrogenevolution in neutral media. Angew. Chem. Int.Ed. 56, 13694–13698.

74. Cheng, X., Li, Y.H., Zheng, L.R., Yan, Y., Zhang,Y.F., Chen, G., Sun, S.R., and Zhang, J.J.(2017). Highly active, stable oxidized platinum

26 Chem 5, 1–27, June 13, 2019

clusters as electrocatalysts for the hydrogenevolution reaction. Energy Environ. Sci. 10,2450–2458.

75. Chao, T.T., Luo, X., Chen, W.X., Jiang, B., Ge,J.J., Lin, Y., Wu, G., Wang, X.Q., Hu, Y.M.,Zhuang, Z.B., et al. (2017). Atomicallydispersed copper-platinum dual sites alloyedwith palladium nanorings catalyze thehydrogen evolution reaction. Angew. Chem.Int. Ed. 56, 16047–16051.

76. Luo, Z.Y., Ouyang, Y.X., Zhang, H., Xiao, M.L.,Ge, J.J., Jiang, Z., Wang, J.L., Tang, D.M.,Cao, X.Z., Liu, C.P., et al. (2018). Chemicallyactivating MoS2 via spontaneous atomicpalladium interfacial doping towards efficienthydrogen evolution. Nat. Commun. 9, 2120.

77. Jaramillo, T.F., Jørgensen, K.P., Bonde, J.,Nielsen, J.H., Horch, S., and Chorkendorff, I.(2007). Identification of active edge sites forelectrochemical H2 evolution from MoS2nanocatalysts. Science 317, 100–102.

78. Kibsgaard, J., Chen, Z.B., Reinecke, B.N., andJaramillo, T.F. (2012). Engineering the surfacestructure of MoS2 to preferentially exposeactive edge sites for electrocatalysis. Nat.Mater. 11, 963–969.

79. Kong, D.S., Wang, H.T., Cha, J.J., Pasta, M.,Koski, K.J., Yao, J., and Cui, Y. (2013).Synthesis of MoS2 and MoSe2 films withvertically aligned layers. Nano Lett. 13, 1341–1347.

80. Ouyang, Y.X., Ling, C.Y., Chen, Q., Wang,Z.L., Shi, L., and Wang, J.L. (2016). Activatinginert basal planes of MoS2 for hydrogenevolution reaction through the formation ofdifferent intrinsic defects. Chem. Mater. 28,4390–4396.

81. Xie, J.F., Zhang, H., Li, S., Wang, R.X., Sun, X.,Zhou, M., Zhou, J.F., Lou, X.W., and Xie, Y.(2013). Defect-rich MoS2 ultrathin nanosheetswith additional active edge sites for enhancedelectrocatalytic hydrogen evolution. Adv.Mater. 25, 5807–5813.

82. Ye, G.L., Gong, Y.J., Lin, J.H., Li, B., He, Y.M.,Pantelides, S.T., Zhou, W., Vajtai, R., andAjayan, P.M. (2016). Defects engineeredmonolayer MoS2 for improved hydrogenevolution reaction. Nano Lett. 16, 1097–1103.

83. Jiang, K., and Wang, H.T. (2018).Electrocatalysis over graphene-defect-coordinated transition-metal single-atomcatalysts. Chem 4, 194–195.

84. Chen,W.X., Pei, J.J., He, C.T., Wan, J.W., Ren,H.L., Zhu, Y.Q., Wang, Y., Dong, J.C., Tian,S.B., Cheong, W.C., et al. (2017). Rationaldesign of single molybdenum atomsanchored on N-doped carbon for effectivehydrogen evolution reaction. Angew. Chem.Int. Ed. 56, 16086–16090.

85. Fei, H.L., Dong, J.C., Arellano-Jimenez, M.J.,Ye, G.L., Kim, N.D., Samuel, E.L.G., Peng,Z.W., Zhu, Z., Qin, F., Bao, J.M., et al. (2015).Atomic cobalt on nitrogen-doped graphenefor hydrogen generation. Nat. Commun. 6,8668.

86. Liang, H.W., Bruller, S., Dong, R.H., Zhang, J.,Feng, X.L., and Mullen, K. (2015). Molecularmetal-N-x centres in porous carbon forelectrocatalytic hydrogen evolution. Nat.Commun. 6, 7992.

87. Xue, Y.R., Huang, B.L., Yi, Y.P., Guo, Y., Zuo,Z.C., Li, Y.J., Jia, Z.Y., Liu, H.B., and Li, Y.L.(2018). Anchoring zero valence single atomsof nickel and iron on graphdiyne for hydrogenevolution. Nat. Commun. 9, 1460.

88. Qiu, H.J., Ito, Y., Cong,W.T., Tan, Y.W., Liu, P.,Hirata, A., Fujita, T., Tang, Z., and Chen, M.W.(2015). Nanoporous graphene with single-atom nickel dopants: an efficient and stablecatalyst for electrochemical hydrogenproduction. Angew. Chem. Int. Ed. 54, 14031–14035.

89. Fan, L.L., Liu, P.F., Yan, X.C., Gu, L., Yang, Z.Z.,Yang, H.G., Qiu, S.L., and Yao, X.D. (2016).Atomically isolated nickel species anchoredon graphitized carbon for efficient hydrogenevolution electrocatalysis. Nat. Commun. 7, 7.

90. Jia, Y., Zhang, L.Z., Gao, G.P., Chen, H.,Wang, B., Zhou, J.Z., Soo, M.T., Hong, M.,Yan, X.C., Qian, G.R., et al. (2017). Aheterostructure coupling of exfoliated Ni-Fehydroxide nanosheet and defective grapheneas a bifunctional electrocatalyst for overallwater splitting. Adv. Mater. 29, 1700017.

91. Hori, Y. (2008). Electrochemical CO2

reduction on metal electrodes. In ModernAspects of Electrochemistry, 42, C.G.Vayenas, R.E. White, and M.E. Gamboa-Aldeco, eds. (Springer), pp. 89–189.

92. Jhong, H.-R., Ma, S., and Kenis, P.J.A. (2013).Electrochemical conversion of CO2 to usefulchemicals: current status, remainingchallenges, and future opportunities. Curr.Opin. Chem. Eng. 2, 191–199.

93. Editorial. (2018). Change is in the air. Nat.Catal. 1, 93.

94. Zheng, T.T., Jiang, K., and Wang, H.T. (2018).Recent advances in electrochemical CO2-to-CO conversion on heterogeneous catalysts.Adv. Mater. 30, 15.

95. Varela, A.S., Ju, W., and Strasser, P. (2018).Molecular nitrogen-carbon catalysts, solidmetal organic framework catalysts, and solidmetal/nitrogen-doped carbon (MNC)catalysts for the electrochemical CO2

reduction. Adv. Energy Mater. 8, 35.

96. Sharma, P.P., Wu, J.J., Yadav, R.M., Liu, M.J.,Wright, C.J., Tiwary, C.S., Yakobson, B.I., Lou,J., Ajayan, P.M., and Zhou, X.D. (2015).Nitrogen-doped carbon nanotube arrays forhigh-efficiency electrochemical reduction ofCO2: on the understanding of defects, defectdensity, and selectivity. Angew. Chem. Int. Ed.54, 13701–13705.

97. Wu, J.J., Yadav, R.M., Liu, M.J., Sharma, P.P.,Tiwary, C.S., Ma, L.L., Zou, X.L., Zhou, X.D.,Yakobson, B.I., Lou, J., et al. (2015). Achievinghighly efficient, selective, and stable CO2

reduction on nitrogen-doped carbonnanotubes. ACS Nano 9, 5364–5371.

98. Wu, J.J., Liu, M.J., Sharma, P.P., Yadav, R.M.,Ma, L.L., Yang, Y.C., Zou, X.L., Zhou, X.D.,Vajtai, R., Yakobson, B.I., et al. (2016).Incorporation of nitrogen defects for efficientreduction of CO2 via two-electron pathway onthree-dimensional graphene foam. NanoLett. 16, 466–470.

99. Li, W.L., Seredych, M., Rodrıguez-Castellon,E., and Bandosz, T.J. (2016). Metal-freenanoporous carbon as a catalyst for


































































































































































































































electrochemical reduction of CO2 to CO andCH4. ChemSusChem 9, 606–616.

100. Xu, J.Y., Kan, Y.H., Huang, R., Zhang, B.S.,Wang, B.L., Wu, K.H., Lin, Y.M., Sun, X.Y., Li,Q.F., Centi, G., et al. (2016). Revealing theorigin of activity in nitrogen-dopednanocarbons towards electrocatalyticreduction of carbon dioxide. ChemSusChem9, 1085–1089.

101. Liu, T.F., Ali, S., Lian, Z., Li, B., and Su, D.S.(2017). CO2 electoreduction reaction onheteroatom-doped carbon cathodematerials. J. Mater. Chem. A 5, 21596–21603.

102. Cui, X.Q., Pan, Z.Y., Zhang, L.J., Peng, H.S.,and Zheng, G.F. (2017). Selective etching ofnitrogen-doped carbon by steam forenhanced electrochemical CO2 reduction.Adv. Energy Mater. 7, 6.

103. Nakata, K., Ozaki, T., Terashima, C., Fujishima,A., and Einaga, Y. (2014). High-yieldelectrochemical production of formaldehydefrom CO2 and seawater. Angew. Chem. Int.Ed. 53, 871–874.

104. Liu, Y.M., Chen, S., Quan, X., and Yu, H.T.(2015). Efficient electrochemical reduction ofcarbon dioxide to acetate on nitrogen-dopednanodiamond. J. Am. Chem. Soc. 137, 11631–11636.

105. Dai, L., Qin, Q., Wang, P., Zhao, X.J., Hu, C.Y.,Liu, P.X., Qin, R.X., Chen, M., Ou, D.H., Xu,C.F., et al. (2017). Ultrastable atomic coppernanosheets for selective electrochemicalreduction of carbon dioxide. Sci. Adv. 3,e1701069.

106. Wang, Y.F., Chen, Z., Han, P., Du, Y.H., Gu,Z.X., Xu, X., and Zheng, G.F. (2018). Single-atomic Cu with multiple oxygen vacancies onceria for electrocatalytic CO2 reduction toCH4. ACS Catal. 8, 7113–7119.

107. Costentin, C., Robert, M., and Saveant, J.M.(2013). Catalysis of the electrochemicalreduction of carbon dioxide. Chem. Soc. Rev.42, 2423–2436.

108. Beley, M., Collin, J.P., Ruppert, R., andSauvage, J.P. (1986). Electrocatalytic

reduction of carbon dioxide by nickelCYCLAM2+ in water: study of the factorsaffecting the efficiency and the selectivity ofthe process. J. Am. Chem. Soc. 108, 7461–7467.

109. Nishihara, H., Hirota, T., Matsuura, K.,Ohwada, M., Hoshino, N., Akutagawa, T.,Higuchi, T., Jinnai, H., Koseki, Y., Kasai, H.,et al. (2017). Synthesis of orderedcarbonaceous frameworks from organiccrystals. Nat. Commun. 8, 109.

110. Su, P., Iwase, K., Nakanishi, S., Hashimoto, K.,and Kamiya, K. (2016). Nickel-nitrogen-modified graphene: an efficientelectrocatalyst for the reduction of carbondioxide to carbon monoxide. Small 12, 6083–6089.

111. Zhao, C.M., Dai, X.Y., Yao, T., Chen, W.X.,Wang, X.Q., Wang, J., Yang, J.,Wei, S.Q., Wu,Y.E., and Li, Y.D. (2017). Ionic exchange ofmetal-organic frameworks to access singlenickel sites for efficient electroreduction ofCO2. J. Am. Chem. Soc. 139, 8078–8081.

112. Li, X.G., Bi, W.T., Chen, M.L., Sun, Y.X., Ju,H.X., Yan, W.S., Zhu, J.F., Wu, X.J., Chu, W.S.,Wu, C.Z., et al. (2017). Exclusive Ni-N4 sitesrealize near-Unity CO selectivity forelectrochemical CO2 reduction. J. Am. Chem.Soc. 139, 14889–14892.

113. Ju, W., Bagger, A., Hao, G.P., Varela, A.S.,Sinev, I., Bon, V., Roldan Cuenya, B., Kaskel,S., Rossmeisl, J., and Strasser, P. (2017).Understanding activity and selectivity ofmetal-nitrogen-doped carbon catalysts forelectrochemical reduction of CO2. Nat.Commun. 8, 944.

114. Cheng, Y., Zhao, S.Y., Johannessen, B., Veder,J.P., Saunders, M., Rowles, M.R., Cheng, M.,Liu, C., Chisholm, M.F., De Marco, R., et al.(2018). Atomically dispersed transition metalson carbon nanotubes with ultrahigh loadingfor selective electrochemical carbon dioxidereduction. Adv. Mater. 30, e1706287.

115. Wang, X.Q., Chen, Z., Zhao, X.Y., Yao, T.,Chen, W.X., You, R., Zhao, C.M., Wu, G.,Wang, J., Huang, W.X., et al. (2018).

Regulation of coordination number oversingle Co sites: triggering the efficientelectroreduction of CO2. Angew. Chem. Int.Ed. 57, 1944–1948.

116. Huan, T.N., Ranjbar, N., Rousse, G., Sougrati,M., Zitolo, A., Mougel, V., Jaouen, F., andFontecave, M. (2017). Electrochemicalreduction of CO2 catalyzed by Fe-N-Cmaterials: a structure-selectivity study. ACSCatal. 7, 1520–1525.

117. Varela, A.S., Sahraie, N.R., Steinberg, J., Ju,W., Oh, H.S., and Strasser, P. (2015). Metal-doped nitrogenated carbon as an efficientcatalyst for direct CO2 electroreduction to COand hydrocarbons. Angew. Chem. Int. Ed. 54,10758–10762.

118. Jiang, K., Siahrostami, S., Akey, A.J., Li, Y.B.,Lu, Z.Y., Lattimer, J., Hu, Y.F., Stokes, C.,Gangishetty, M., Chen, G.X., et al. (2017).Transition-metal single atoms in a grapheneshell as active centers for highly efficientartificial photosynthesis. Chem 3, 950–960.

119. Jiang, K., Siahrostami, S., Zheng, T.T., Hu,Y.F., Hwang, S., Stavitski, E., Peng, Y.D.,Dynes, J., Gangisetty, M., Su, D., et al. (2018).Isolated Ni single atoms in graphenenanosheets for high-performance CO2

reduction. Energy Environ. Sci. 11, 893–903.

120. Zheng, T.T., Jiang, K., Ta, N., Hu, Y.F., Zeng,J., Liu, J.Y., andWang, H.T. (2018). Large-scaleand highly selective CO2 electrocatalyticreduction on nickel single-atom catalyst.Joule 3, 265–278.

121. Yang, H.B., Hung, S.F., Liu, S., Yuan, K.D.,Miao, S., Zhang, L.P., Huang, X., Wang, H.Y.,Cai, W.Z., Chen, R., et al. (2018). Atomicallydispersed Ni(I) as the active site forelectrochemical CO2 reduction. Nat. Energy3, 140–147.

122. Pan, Y., Lin, R., Chen, Y.J., Liu, S.J., Zhu, W.,Cao, X., Chen, W.X., Wu, K.L., Cheong, W.C.,Wang, Y., et al. (2018). Design of single-atomCo-N5 catalytic site: a robust electrocatalystfor CO2 reduction with nearly 100% COselectivity and remarkable stability. J. Am.Chem. Soc. 140, 4218–4221.

Chem 5, 1–27, June 13, 2019 27


























































































































































the role of defect sites in nanomaterials for ...the role of defect sites in nanomaterials for...

Documents