You are on page 1of 29

Review

pubs.acs.org/acscatalysis

Recent Trends and Perspectives in Electrochemical Water Splitting


with an Emphasis on Sulfide, Selenide, and Phosphide Catalysts of
Fe, Co, and Ni: A Review
Sengeni Anantharaj,† Sivasankara Rao Ede,† Kuppan Sakthikumar,† Kannimuthu Karthick,†
Soumyaranjan Mishra,†,‡ and Subrata Kundu*,†,§

Electrochemical Materials Science (ECMS) Division, CSIR-Central Electrochemical Research Institute (CECRI), Karaikudi-630006,
Tamil Nadu, India

Centre for Education (CFE), CSIR-Central Electrochemical Research Institute (CECRI), Karaikudi-630006, Tamil Nadu, India
Downloaded via NATL UNIV OF SCIENCES & TECHNOLOGY on May 22, 2023 at 09:06:11 (UTC).
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

§
Department of Materials Science and Mechanical Engineering, Texas A&M University, College Station, Texas 77843, United States

ABSTRACT: Increasing demand for finding eco-friendly and


everlasting energy sources is now totally depending on fuel cell
technology. Though it is an eco-friendly way of producing
energy for the urgent requirements, it needs to be improved to
make it cheaper and more eco-friendly. Although there are
several types of fuel cells, the hydrogen (H2) and oxygen (O2)
fuel cell is the one with zero carbon emission and water as the
only byproduct. However, supplying fuels in the purest form
(at least the H2) is essential to ensure higher life cycles and less
decay in cell efficiency. The current large-scale H2 production
is largely dependent on steam reforming of fossil fuels, which
generates CO2 along with H2 and the source of which is going
to be depleted. As an alternate, electrolysis of water has been
given greater attention than the steam reforming. The reasons are as follows: the very high purity of the H2 produced, the
abundant source, no need for high-temperature, high-pressure reactors, and so on. In earlier days, noble metals such as Pt
(cathode) and Ir and Ru (anode) were used for this purpose. However, there are problems in employing these metals, as they are
noble and expensive. In this review, we elaborate how the group VIII 3d metal sulfide, selenide, and phosphide nanomaterials
have arisen as abundant and cheaper electrode materials (catalysts) beyond the oxides and hydroxides of the same. We also
highlight the evaluation perspective of such electrocatalysts toward water electrolysis in detail.
KEYWORDS: water splitting, hydrogen evolution, oxygen evolution, overpotential, Tafel analysis, metal chalcogenides,
metal phosphides, electrolysis

1. INTRODUCTION electrolysis proceeds via the following two half-cell reactions:


An unusually increased rate of depletion of conventional fossil reduction of H+ ions at the cathode (2H+(aq) + 2e− → H2(g)),
fuels and the environmental hazards associated with the use of i.e., the hydrogen evolution reaction (HER),8−10 and oxidation
these carbon-emitting fuels have triggered the research of water (2H2O(l) → O2(g) + 4H+(aq) + 4e−), i.e., the oxygen
community to find an abundant, everlasting, zero-emitting, evolution reaction (OER).11−16 At first look, water electrolysis
and eco-friendly combined fuel-combustion technology as an may pretend to be an easy and problemless way to produce H2,
alternate energy server very urgently.1 Fuel cell technology is but it is not. Electrolysis of water is an area where the science
among the ways of producing energy in a more eco-friendly and technology need to be improved to overcome the issues
manner than the other existing ones. The fuel cell operating associated with it. One such issue is that the noble metals and
with H2 and O2 will be the choice among the available fuel cells their compounds catalyzing the HER (Pt)17−25 and OER (Ir,
as it emits water as the combustion product.2−7 Currently, Ru)26−43 need to be replaced by the available non-noble metals
steam reforming of fossil fuels and water electrolysis (H2O(l) and their compounds. To do so, recently attention has been
→ H2(g) + 1/2O2(g): ΔG° = +237 200 J/mol, ΔE° = 1.23 V vs diverted toward OER catalysts based on non-noble metals such
reversible hydrogen electrode (RHE)) are the ways of as Mn,44−47 Fe,48−53 Co,44,49,54−76 and Ni16,47,53,76−101 and
producing H2 to afford fuel for these fuel cells. Between HER catalysts based on Mo58,83,102−110 and W.111,112 Though
steam reforming and water electrolysis, the first one is not
environmentally friendly because it produces CO2 along with Received: August 29, 2016
H2, thereby reducing the purity of H2, which in turn affects the Revised: October 14, 2016
life cycle and the cell efficiency of fuel cells.2−7 Water Published: October 19, 2016

© 2016 American Chemical Society 8069 DOI: 10.1021/acscatal.6b02479


ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

the production of purest H2 is the main objective of water including the Krasil’shchkov, Bockris,46,116 Yeager, and Wade
electrolysis, we cannot neglect the counter-reaction (the OER), and Hackerman52 pathways in addition to the most recognized
as it is the sluggish one between them and affects the Faradaic electrochemical oxide pathway and the oxide pathway. Under
efficiency of the electrolytic cell to a greater extent. alkaline conditions, all of the proposed mechanisms begin with
The HER under acidic conditions is facile, as plenty of an essential elementary step of hydroxide coordination to the
protons are available, and it proceeds by a multistep reaction active site and proceed via different proposed other elementary
with two possible mechanisms.113,114 The mechanism by which steps.46,116 The kinetic barriers associated with each elementary
the HER proceeds is usually revealed by the experimentally step raise the overall overpotential required. An elementary step
obtained Tafel slope value.8−10,115 The first step in the with the most sluggish kinetics is the rate-determining step
multistep HER is discharging protons on the electrode surface (RDS).11,14
to form adsorbed hydrogen (Hads), which is called as the In general, the electrochemical reactions that occur at the
Volmer reaction (H+(aq) + e− → Hads). The Tafel slope for the anode (OER domain) and at the cathode (HER domain) in
Volmer reaction, b1,V, can be expressed as acidic and alkaline media are given as follows:
b1,V = 2.303RT /βF (1) in acid:

where R is the ideal gas constant, T is the absolute temperature, 4e− + 4H+(aq) → 2H 2(g) (5)
F is the Faraday constant, and β is the symmetry factor (equal
to 0.5). Depending on the coverage of Hads, the second step is 2H 2O(l) → 4H+(aq) + 4e− + O2 (g) (6)
either electrochemical desorption of H2, known as the
Heyrovsky reaction, or chemical H2 desorption, known as the in alkali:
Tafel reaction. If the coverage of Hads is low and the electrode 4e− + 4H 2O(l) → 4OH−(aq) + 2H 2(g) (4)
surface has sufficient active sites near the Hads sites, the
adsorbed H atoms will preferably join with a proton and an 4OH−(aq) → 2H 2O(l) + 4e− + O2 (g) (7)
electron simultaneously to evolve a molecule of H2 (Hads +
H+(aq) + e− → H2(g)). This is called the Heyrovsky reaction, In practice, there is no electrolytic cell that performs with
and its Tafel slope can be expressed as 100% Faradaic efficiency because of various thermodynamic
and kinetic hindrances. As stated above, the best catalyst that
b2,H = 2.303RT /(1 + β)F (2) catalyzes the HER is Pt, and the catalysts that catalyze the OER
In the case of high Hads coverage, two adjacent Hads will join are Ir and Ru and their compounds. However, use of such
together chemically and evolve a molecule of H2. This the Tafel relatively less abundant precious metals for a corrosive and
reaction, and its Tafel slope can be expressed as destructive water electrolysis is not advisible as it increases the
cost of H2 production, hinders the magnification of productivity
b2 ′ ,T = 2.303RT /2F (3) on a large scale, and more. Later, people have found that the
oxides and hydroxides of group VIII 3d metals, viz., Fe, Co, and
The calculated Tafel slopes of the above reactions under Ni alone and with a combination of other metals in the row,
standard conditions are 0.118 V/dec, 0.039 V/dec, and 0.029 can catalyze the OER as efficiently as Ir and Ru under alkaline
V/dec for the Volmer, Heyrovsky, and Tafel reactions, conditions. Nevertheless, the discovery is not fruitful as they did
respectively. An electrocatalytic HER with a Tafel slope of not consider the kinetics of the counter-reaction (the HER) in
0.029 V/dec is termed to follow the Tafel−Heyrovsky alkali. Moreover, the HER is sluggish even with Pt because of
mechanism, and the rate-limiting step would be the electro- the reasons stated above.
chemical desorption step. Under alkaline conditions, the HER To overcome this issue, there are two possibilities. The first
is comparatively more sluggish because it directly depends on one is to find efficient, durable, non-noble OER catalysts in
the anodic OER, which supplies the protons to the cathode by acidic media as alternates to Ir- and Ru-based catalysts.
the deprotonation of hydroxide ions and affects the HER However, this could not even eliminate the expensive Pt
kinetics. Under such conditions, the following reaction takes from the job of the HER. One way to get rid of the Pt would be
place at the cathode: to replace it with other non-noble metal-based catalysts such as
4e− + 4H 2O(l) → 4OH−(aq) + 2H 2(g) (4) sulfides, selenides, and phosphides, but the resulting electrolytic
cell will be the kind of asymmetric one, leading to other
The protons formed by the deprotonation of hydroxides at technical problems. The second way to overcome this pitfall is
anode get combined with the abundant OH− ions in alkaline to design a bifunctional catalyst out of non-noble metals that
solutions, making the HER struggle more to move forward can catalyze both the OER and HER at the same time without
further. the need for any separator. Such bifunctional catalysts made
The OER on the other hand has a different story. The from Fe, Co, and Ni are being reported, frequently with parallel
kinetics of the OER in acidic and alkaline media vary depending or in some cases better activities for both the HER and OER
on the material by which it is being catalyzed. Noble metal compared with Pt, Ir, and Ru. Such reported bifunctional
catalysts such as Ir and Ru and their compounds catalyze the catalysts are almost always phosphides, sulfides, and selenides
OER more easily in acidic media than under alkaline of Fe, Co, and Ni as nanostructures in various forms.
conditions. On the other hand, the catalysts derived from As an emerging field in the energy sector, there are few
VIII group 3d metals (Fe, Co, and Ni) catalyze the OER more reviews of many of these materials and their subsequent
favorably in alkaline media than in acidic media. This is mainly applications to water splitting. However, all of the available
correlated to the mechanism by which they catalyze it. In 1986, reviews to date have either been focused on one type of catalyst
Matsumoto and Sato13 published a detailed review on various formed out of these three metals (viz., Fe, Co, and Ni) or all
reported OER mechanisms in acidic and alkaline media, three types of catalysts formed out of a single metal. Examples
8070 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

for the first kind are the reviews of transition metal phosphides in acidic solution or hydroxide ion coordination in alkaline
and chalcogenides. Meanwhile, examples of the second kind are solution, which is then followed by a series of other elementary
the reviews available on heterogeneous electrocatalysis by steps. The kinetic barriers associated with each of these steps
catalysts derived out of Co, Ni, and Fe, in which for a single will contribute to the overall activation overpotential of an OER
metal the catalytic activity trend of its oxides, sulfides, selenides, catalyst. Man et al.14 and others13,116 have studied the
phosphides, and borides would have coherently been thermodynamics of the OER mechanism and proposed the
summarized. Hence, to have a better experience in under- following expression for calculating the theoretical OER
standing the basics to frontiers of electrochemical water overpotential (ηOER) under ideal conditions with U = 0 vs
splitting, a complete review on sulfide, selenide, and phosphide standard hydrogen electrode (SHE):
catalysts formulated out of Fe, Co, and Ni would highly be
desired. In this review, we present a detailed survey of the ηOER = (ΔGmax /e) − 1.23 V (8)
following topics: the parameters involved in the evaluation of
catalysts for both the HER and OER, the effects of the However, the theoretical and experimental values had very large
heteroatoms (S, Se, and P) on the material properties of the difference in the standard free energy change associated with
metal and the subsequent catalytic activities, the basic the elementary step involving the conversion of oxide to
requirements for an efficient bifunctional water electrolytic peroxide. This clearly says that the kinetic hindrances are not
catalyst, a detailed overview of the synthetic methodologies of considered in the thermodynamic prediction of the over-
metal (Fe, Co, Ni, and their mixed versions) phosphides, potential. Because of the varying kinetics of these elementary
sulfides, and selenides, and the applications of these materials in steps from material to material, instead of the onset
the electrochemical HER and OER. Finally, the challenges overpotential, the overpotential at a fixed current density (j)
ahead in developing new, efficient bifunctional catalysts and such as 10 mA/cm2 (η10) has now widely been accepted as an
opportunities are stressed. essential quantitative activity parameter to evaluate an electro-
catalyst.118−120 The same is also used for the HER. For
2. PARAMETERS USED TO EVALUATE THE materials with a strong redox peak (giving a current density
CATALYTIC ACTIVITY greater than 10 mA/cm2) within the potential window of gas
evolution and for high-performance catalysts (giving current
The following parameters are the widely recognized ones fpr densities greater than 500 mA/cm2) such as layered double
evaluating and comparing the catalytic activities of catalysts. In hydroxides,100,121−124 the overpotentials at higher current
this part of the review, the merits and demerits associated with densities such as 50 and 100 mA/cm2 are also used as alternate
each of these parameters are discussed and justified. activity parameters.
2.1. Overpotential (η). There is no electrochemical 2.2. Tafel Slope and Exchange Current Density (j0).
reaction that occurs at the potential predicted only by The Tafel plot of an electrocatalytic process is generally
thermodynamic considerations excluding the kinetic hindrances obtained by replotting the polarization curve (e.g., linear sweep
experienced in a real system.11,14,113,114 As a consequence of voltammogram (LSV)) as a plot of log(j) versus η. The slope of
these hindrances, an additional driving force in terms of an the linear portion of the Tafel plot is defined11 as the
extra potential is needed to effect such electrochemical dependence between the iR-compensated overpotential and the
reactions, which is called the overpotential (denoted by the current density, which is expressed as follows:
symbol η). For both the OER and HER, there are three sources
of overpotential, viz., the activation overpotential, the d log(j)
concentration overpotential, and the overpotential due to the = 2.303RT /αnF
dη (9)
uncompensated resistance (Ru), which is the resistance exerted
by the electrochemical interfaces. The activation overpotential The Tafel slope is inversely proportional to the charge transfer
is an intrinsic property of the material that catalyzes the coefficient (α), as the remaining other parameters are constants
electrode reaction and varies from one material to another. (viz., the ideal gas constant (R), temperature (T), Faraday
Hence, it can be minimized by choosing an efficient catalyst. constant (F), and the number of electrons transferred (n),
The concentration overpotential occurs as soon as the which is equal to 4 for the OER and 2 for the HER). This
electrode reaction begins as a result of the sudden drop in indicates that a catalyst with a high charge transfer ability
concentration near the interfaces. This can be minimized by should possess a small Tafel slope. This is the reason why it is
stirring the solution. The resistance overpotential can be often used as a primary activity parameter in determining the
removed by carrying out Ohmic drop compensation, which is catalytic activity.
now available in many electrochemical workstations.8 Other- However, with the above-mentioned path (i.e., converting
wise, it can also be done manually just by multiplying the the LSV into the Tafel plot), there are several issues that can
resultant current density by Ru, the result of which is a potential lead to a misinterpretation of the catalytic activity. In general,
(E). This drop in potential, known as the iR drop, needs to be the LSV obtained with lowest possible scan rate gives the Tafel
subtracted from the experimental potential. In case of the HER, slope with the least experimental inaccuracy. The scan rate at
the activation overpotential, which can be termed as the onset which the polarization curve used for Tafel plot was determined
overpotential, is more important than the others, as the kinetics becomes a serious problem when the catalyst is highly
of the HER is faster than that of the OER.8,9,115,117 This can be capacitive. This leads to large error while determining the
calculated from the polarization curve obtained by plotting the exchange current density because the exchange current density
overpotential versus the current density. is usually obtained by extrapolating the linear fit toward the
In contrast, the OER case differs and needs more attention corresponding current density on the logarithmic scale at the
on other parameters too in order to calculate the overpotential. equilibrium potential (i.e., at zero overpotential). In such a case
As we said above, all of the proposed mechanisms of the OER the catalyst with a high overpotential will have a large exchange
proceed through the first elementary step of water coordination current density value, which is not possible because a high
8071 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

exchange current density means that transferring electrons the slope of the plot of log(Rct) vs η (where Rct is the charge
across the catalytic interface will be facile and require a very low transfer resistance) is obtained. This is the exact Tafel slope of
activation energy. Hence, the result would be a low the catalyst and depends only on the charge transfer ability of
overpotential. the catalyst as Ru is excluded from the calculation. Depending
Besides the problem with capacitive currents, the current on the requirements and conditions, any one of these methods
observed in the Faradaic region is also not the true one, as it is can be chosen for finding the Tafel slope of an electrocatalyst.
not the steady-state current, and this leads to a considerable Apart from the methods of obtaining Tafel slope, there has
experimental inaccuracy in the determined Tafel slope value. been a serious discussion of considering the exchange current
This can be explained as follows. In a chronoamperometric density (j0) as an activity parameter in addition to the Tafel
analysis (at constant potential), it is common to observe the slope.11 The exchange current density is more often used as an
steady-state current only after few seconds. In the case of linear activity parameter for the HER than for the OER. As we stated
sweep voltammetry, even at a scan rate of 1 mV/s the current earlier, the kinetics of the HER is facile beyond the defined
obtained cannot be the real steady-state current. To overcome overpotential of the catalyst, so almost all reported catalysts
this issue, a traditional practice can be employed in which the have similar kinetics in the HER. This means that the exchange
steady-state current of the catalyst is obtained from the current density is directly correlated to the onset overpotential
chronoamperometric i−t curves obtained at various over- in the HER. Hence, it can be used there as an activity
potentials with regular small intervals (say 5 mV). A similar parameter. In contrast, the OER case is different, as the kinetics
experimental comparison of Tafel slopes obtained from cyclic differs from one catalyst to another one and is strongly pH-
voltammetry (CV) (run at 5 mV/s) and from the steady-state dependent. A classic example of this kind can be the
currents obtained by chronoamperometric curves for a comparison of the RuO2/IrO226,38,41,43,126−128 and Ni−
commercial IrTiO2 electrode is given in Figure 1.11 The linear Fe98,129−132 systems in alkaline media. In this case the onset
overpotential of IrO2/RuO2 is lower than that of Ni−Fe but the
Tafel slope is lower for the latter one. This implies that though
the driving force to start the reaction on IrO2/RuO2 is lower,
the rate of charge transfer/electron transfer is lower than for
the Ni−Fe system in alkaline media. This is one of the reasons
why people have adopted the overpotential at fixed current
density (e.g., 10 mA/cm2) as an additional activity parameter
rather than the exchange current density (j0) or the onset
overpotential for the OER.11 Hence, hereby we can conclude
that for the HER the Tafel slope and the onset overpotential
and/or exchange current density can be taken as activity
parameters. For the OER, the Tafel slope and the overpotential
at a defined current density can be taken as activity parameters.
2.3. Stability. The stability of an electrocatalyst is usually
tested by subjecting it to CV cycling at a higher scan rate, which
is otherwise known as the accelerated degradation test, and to
Figure 1. Tafel curves of a commercial IrTiO2 electrode (Umicore AG chronoamperometric or chronopotentiometric analyses. In the
& Co.) obtained by a cyclic voltammetry measurement at 5 mV/s and case of the HER, the accelerated degradation test is carried out
by a chronoamperometry experiment by holding each potential for 30 for several thousands of cycles as the polarization starts from 0
s in 20% O2-saturated 0.1 M HClO4 at room temperature and a
V vs NHE. In the case of the OER, the number of cycles
rotation speed of 900 rpm. Reproduced with permission from ref 11.
Copyright 2014 Royal Society of Chemistry. reported in accelerated degradation test ranges from 250 to
1000. Beyond 1000 cycles, it is rare to see such a report on an
portions of the Tafel plots obtained by the two methods are in OER catalyst with extreme stability. After the accelerated
close agreement with one another. However, in the lower- degradation test, the shift in onset overpotential (η0) (cathodic
overpotential region the story is different. It can be seen from for HER and anodic for OER) and the overpotential at a
Figure 1 that the Tafel plot obtained from CV shows a larger defined current density of 10 mA/cm2 (η10) are measured as
exchange current density than the one obtained from the indicative parameters of stability, with a smaller increase in
steady-state current, and the exchange current density from overpotential indicating a higher stability. Stability under
chronoamperometry will be more precise as it eliminates the constant exposure to a fixed potential (chronoamperometry)
capacitive current. Though the latter method seems to be better or a fixed current density (chronopotentiometry) is examined
than the method of replotting the polarization curves, it also for durations of several minutes to hours. It has now been
does not reflect the intrinsic and exact catalytic activity of a widely accepted that a stable current density (e.g., 10 mA/cm2)
catalyst because the potential drop due to the uncompensated for more than 12 h by chronoamperometry or a negligible
resistance cannot be easily excluded in this method as it is increase in overpotential at a current density of 10 mA/cm2 for
usually done in the former method of replotting polarization more than 12 h by chronopotentiometry is enough to recognize
curves by iR compensation. an efficient electrocatalyst for both the OER and HER. As
To avoid this issue, Hu and co-workers very recently highlighted in the overpotential case, for high-performance
proposed a new method for obtaining a more accurate Tafel materials and catalysts with strong redox peaks, other current
slope for a catalyst from electrochemical impedance spectros- densities such as 50 or 100 mA/cm2 and their corresponding
copy (EIS).125 In this method, the Nyquist plots of the catalyst potentials can also be used to run chronopotentiometric and
at various overpotentials at regular intervals are acquired, and chronoamperometric analyses for long times.17,27
8072 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

2.4. Faradaic Efficiency. Faradaic efficiency is another INA


TOF =
quantitative parameter used for both the HER and OER. It can AFn Γ (11)
be defined as the efficiency of an electrocatalyst to transfer
electrons provided by the external circuit across the interface to where I is the current, NA is the Avogadro constant, A is the
the electroactive species to effect the electrode reaction (in our geometrical surface area, n is the number of electrons
case, either the HER or OER). There are two methods to find transferred, and Γ is the surface or total concentration of
the Faradaic efficiency.8 The first one is the electrochemical catalyst in terms of number of atoms. There are several
method using a rotating ring disc electrode (RRDE), which is methods available to determine the surface or total
applicable to the OER only. The catalytically active material is concentration catalyst in terms of number of atoms. The
coated on the disc of the RRDE without disturbing the ring. redox peak in the cyclic voltammogram can be used to find the
The commonly used RRDE is the one with a glassy carbon surface concentration after activation of the catalyst by CV
(GC) disc and Pt ring. Prior to that, the collection efficiency of cycling.133−135 By the Avogadro’s number method, the total
the RRDE must be known or to be determined experimentally concentration of atoms can be calculated using the average
particle diameter of the catalyst.17 Another method is the
by studying the conventional ferrous/ferric redox system’s
assumption of a monolayer.27 If the catalyst surface is flat and
response at various rotation rates. The potential at the disk is
smooth or the catalyst has a sheet morphology, this assumption
then swept over the same experimental potential window of the
can be made. However, we are aware that each of these
OER, while at the same time the potential at the Pt ring is set at
methods has its own drawbacks. The first method may cause
a constant potential for the oxygen reduction reaction (ORR). potential error when there is more than one element in the
The ORR potential set at the Pt ring is varied depending on the catalyst or if the catalyst is not fully activated. The second
pH of the electrolyte solution. Equation 10 is used to calculate method does not reflect the exact catalytic property of the
the Faradaic efficiency (FE) of an OER catalyst:133−135 catalyst as it also includes the atoms in the core of the particle,
IR nD which actually do not participate in the catalytic cycle. The
FE = third method may lead to potential error when the material is
IDnR NCL (10) not completely flat, is prone to destruction under harsh
electrochemical conditions, or does not have a sheet
where IR and ID are the current at the ring and disc, morphology. Hence, it is advisible to adopt a method
respectively, nR and nD are the numbers of electrons transferred appropriate for the catalyst and its nature.
at the ring and disc, respectively, and NCL is the collection 2.6. Mass and Specific Activities. The mass and specific
efficiency of the RRDE used. This is an extremely useful activities of an electrocatalyst are two other quantitative active
technique to find the true activity of an OER catalyst that has parameters used to define the catalytic activity of an
the following possibilities of losing its Faradaic efficiency: an electrocatalyst. The current normalized by the catalyst loading
OER catalyst with one or more strong redox peaks within the is the mass activity, which is expressed in amperes per gram (A/
potential window of the OER (almost all Ni-, Co-, and Fe- g).136 On the other hand, the current normalized by the
based catalysts), an electrocatalyst that can facilitate other electrochemical surface area (ECSA) or the Brunauer−
unwanted side reactions, and an electrocatalyst that heats up Emmett−Teller (BET) surface area is the specific activity. As
during the electrocatalysis process. in the case of the TOF, the mass and specific activities of an
The second method of determining the Faradaic efficiency is electrocatalyst should be given at a defined overpotential. More
common for both the HER and OER. In this case, the quantity insights about mass and specific activities are described in detail
of gas (H2/O2) evolved is calculated by integration from the in section 3.1. Among the six parameters discussed to describe
chronoamperometric or chronopotentiometric analysis. Then the catalytic activity of an electrocatalyst, the overpotential,
the practically obtained gas (H2/O2) quantity is calculated, for Tafel slope, and stability parameters are the mandatory ones.
which any one of the following three methods can be
employed.8 The first method is the conventional water gas 3. METHODS OF NORMALIZING THE
displacement method, and the second is gas chromatography. EXPERIMENTALLY OBSERVED CURRENT AND THE
The third method is a spectroscopic technique applicable only ASSOCIATED MERITS AND DEMERITS
to the OER in which the evolved oxygen is excited from the Though the geometrical-surface-normalized current density has
triplet state to the singlet state and allowed to relax by now been widely used, it has its own merits and demerits in
fluorescence. The intensity of the fluorescence is a direct quantifying the activity of an electrocatalyst. A good survey of
measure of the quantity of oxygen evolved. However, the such normalization methods was briefly discussed by Fabbri et
method for determining the practically evolved gas quantity can al. in their recent review article on oxide-based OER catalysts.11
be chosen depending on the nature of the catalyst and the 3.1. Normalization of the Current by the Geometrical
availability of resources to do such studies. The ratio of the Surface Area. It is not good to have the current density
quantities of gas determined by the practical method and the normalized by the geometrical area of the substrate electrode
theoretical method is the Faradaic efficiency of the catalyst (in units of A/cm2geo), as this does not reflect the intrinsic
under study.8 The selection from among these methods is also catalytic property of the material. Moreover, it does not care
determined mainly by the demands of the catalyst used. about the catalyst loading, giving different potentials for the
2.5. Turnover Frequency (TOF). The TOF is another same current density with the varying catalyst loading. In other
quantitative parameter used to evaluate an electrocatalyst at a words, if the coverage of the loaded catalyst is 100% and
defined overpotential. The TOF of the catalyst is defined as the sufficient to form a monolayer, then the geometrical area of the
number of moles of O2/H2 evolved per unit time. Equation 11 substrate electrode can be used to normalize the current. In
is used to calculate the TOF for an electrocatalytic gas cases where the loading is low (in which there will be many
evolution reaction: catalytically inactive substrate electrode sites) or very high (in
8073 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Table 1. Merits and Demerits of Various Current Normalization Methods Used in Electrocatalysis of Water Splitting
normalization
method merits demerits
geometrical widely accepted and used does not reflect the intrinsic catalytic property of the catalyst
surface area method
fair comparison with existing may vary depending on catalyst loading and its optimization
literature reports is easy
good for planar electrodes ceometrical area of the substrate electrode is not equal to the actual surface area of the catalyst participating in catalysis
such as foils and deposited
thin films
ECSA can reflect the intrinsic difficulties in determining the ECSA
catalytic property of the
catalyst
loading-sensitive large experimental inaccuracies between one method and other, such as CV and EIS studies
comparison with existing reports would be tedious
BET surface ease of determining the BET does not reflect the intrinsic catalytic property
area surface area
most suitable for porous leads to large errors because not all gas adsorption sites are electrochemically active
materials and catalysts
loading-sensitive not suitable for planar and thin film electrodes
catalyst directly cares about the direct comparison with theory and experiment is not feasible
loading loading regardless of type
of catalyst
suitable when same material does not reflect the intrinsic catalytic property of the material
is used at different loadings
comparison of catalysts of varying particle size (e.g., mono- and polydispersed catalysts), density (e.g., metals and metal
aerogels), morphology (e.g., catalysts with spherical, rod, wire, and sheet structures), and topography (e.g., catalysts with
smooth and rough surfaces) is not possible
comparison with existing reports is also not possible

which only the surface layer participates in the electrode cm2ECSA) will reflect the intrinsic catalytic property of the
reaction, thereby excluding the catalyst present under the catalyst, unlike the one normalized by the geometrical surface
surface layer), the geometrical area of the substrate electrode is area. The current density obtained by this method is often
not equal to the actual surface area of the catalyst participating called the specific activity.136 However, the ECSAs determined
in catalysis. In such cases, normalization by the geometrical by cyclic voltammetry and impedance analysis differ from one
surface area will lead to large errors while quantifying the another significantly, which may lead to potential error.
catalytic activity of an electrocatalyst. However, this is the Recently, people have been using the BET surface area of the
conventional and widely accepted method for current normal- catalyst. However, this is also lacking in experimental accuracy.
ization. Hence, to get information on the optimum catalyst This is mainly due to the fact that all of the sites in the BET
loading, knowledge about the catalyst morphology (if a surface area determined by gas adsorption and desorption need
nanomaterial) and the wettability of the catalyst are needed. not be electrocatalytically active. The experimental incongruity
Beyond everything, an experimental study of the effect of becomes large when the catalyst is composed of more than one
loading can also reveal information about the minimum catalyst element.
loading to form a near-monolayer catalytically active surface. 3.3. Normalization of the Current by the Catalyst
The increase in catalyst loading will not affect the potential at a Loading. Fabbri et al.11 recently justified why the current
fixed current density beyond a certain amount of loading that is normalized by the catalyst loading is a more reliable activity
actually the minimum amount of catalyst required to form the parameter than the current normalized in other ways. The
monolayer. However, care must be taken while doing such a loading-normalized current density (in units of A/g) is
study because if the increase in catalyst loading from one trial to otherwise known as the mass activity. Though it has lesser
another is large, then the actual optimum loading may slip in experimental inaccuracy than the others, it has more
between somewhere. Moreover, if the loading difference is disadvantages than the other two methods explained above:
large, the catalyst resistance will drastically increase, which will direct comparative evaluation of the activity with the theoretical
in turn lead to an increase in the overpotential beyond certain activity is not possible, the intrinsic catalytic property of the
loading limits instead of a steady potential at a fixed current material is not soundly represented by this parameter, and it
density. This may lead to potential error in determining the does not allow a fair comparison of catalysts of varying particle
optimum catalyst loading to form a monolayer. Having size (e.g., mono- and polydispersed catalysts), density (e.g.,
discussed all such difficulties, we can conclude here that the metals and metal aerogels), morphology (e.g., catalysts with
geometrical surface area can be used normalize the spherical, rod, wire, and sheet structures), and topography (e.g.,
experimentally observed current when a planar electrode is catalysts with smooth and rough surfaces). To have a better
used to catalyze the electrode reaction. clarification of the merits and demerits of these methods, we
3.2. Normalization of the Current by the Electro- have summarized them in Table 1.
chemical Surface Area (ECSA) and the Brunauer− Having reviewed the advantages and disadvantages of each of
Emmett−Teller (BET) Surface Area. It is better to use the these normalization methods, we can hereby come to a
ECSA (or roughness factor)-normalized current density, as it is conclusion that depending on the nature of the catalyst, any
more sensitive to the catalyst loading and will vary with the one of these methods can be chosed for normalization of the
same. The current normalized by the ECSA (in units of A/ current. However, to enable fair comparison with earlier
8074 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Figure 2. (A, B) LSVs for the HER on various transition metal phosphide catalysts with current normalization by geometrical surface area and ECSA,
respectively. (C) Plots of average TOF vs E for various transition metal phoshides. (D−F) Volcano plots for various transition metal phosphides
using different activity parameters such as geometrical-area-normalized current, ECSA-normalized current, and average TOF per surface site.
Reproduced with permission from ref 143. Copyright 2015 Royal Society of Chemistry.

reports, it is essential to have data on the current density experimental and theoretical approach (Figure 2). In this
normalized to the geometrical surface area. Otherwise, it is particular report, the HER activities of these catalysts were
better to provide all of them together, and the same will help us benchmarked by plotting the free energy of H adsorption
to come to semiquantitative conclusion on the reliability of against the current normalized by both geometrical surface area
these methods. and ECSA and also against the TOF.
According to this study, an optimal free energy of H
4. BOND ENERGETICS OF INTERMEDIATES: adsorption is predicted for Fe0.5Co0.5P when the activity
THEORETICAL INSIGHT FROM DENSITY parameters such as the current normalized by ECSA and the
FUNCTIONAL THEORY (DFT) CALCULATIONS TOF were plotted against it. However, the Volcano plot
obtained by plotting the current normalized by the geometrical
For an electrocatalyst to be efficient, the bond strengths of the surface area against the H adsorption free energy showed a
intermediates should be neither too high nor too low. In case of slightly controversial activity trend and placed Fe0.5Co0.5P just
the HER, the intermediate is the H-adsorbed active site after below MoP/S and Fe0.25Co0.75P. This is mainly believed to be
electrochemical discharge of a proton. Although all of the Pt- an effect of the contribution of non-Faradaic capacitive current
group metals have been predicted to be highly active toward to the catalytic current. The same group also revealed the
the HER, the actual activity is in the order Pt > Pd > Ni.137−142 reason behind the exceptional HER activity trends observed
This is mainly because of the increased reluctance shown with these FeP-based catalysts from their dependence of the
toward desorption of Hads from the catalytic site by Pd and Ni. free energy of H adsorption on the increasing fraction of
The high metal−Hads bond strength ultimately results in monolayer H coverage, as shown in Figure 3. From Figure 3,
catalytic poisoning.137−142 As the standard redox potential for one can notice that the free energy of H adsorption decreases
hydrogen evolution is zero, the associated standard free energy for FeP and Fe0.5Co0.5P beyond a certain fraction of H coverage
of hydrogen adsorption is should also be zero. on these catalytic surfaces and approaches a negative free
The efficiencies of a set of HER catalysts are usually energy of H adsorption at high H coverages. This indicates that
determined by calculating the standard free energies of H beyond a certain amount of H adsorbed via the electrochemical
adsorption through DFT calculations and plotting them against discharge of protons, these catalytic surfaces turn into H2
a parameter such as the exchange current density (j0), the donors simultaneously. The exceptional activity observed with
current density at a defined overpotential, or the TOF (Sabatier these catalysts is attributed to this particular behavior.
volcano plot). The catalysts that are placed at or near the Later, a similar combined experimental−theoretical work on
summit of the plot are said to be highly active for HER revealing the activity trends of CoP catalysts for the HER
electrocatalysis. Such an analysis of the activity trends of various depending on the Fe dopant content was reported by Tang and
transition metal phosphides (viz., Ni2P, FeP, Fe2P, CoP, Co2P, co-workers.144 Interestingly, they too ended up with the same
MoP, MoP/S, Fe0.25C0.75P, Fe0.75Co0.25P, and Fe0.5Co0.5P) in conclusion that the Fe0.5Co0.5P is the catalyst with Pt-like HER
comparison with Pt nanoparticles (NPs) and Ti foil was activity in acid with a low overpotential of 37 mV at a current
performed by Kibsgaard and co-workers143 using a combined density of 10 mA/cm2. The scheme of the free energy of H
8075 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

theoretical and experimental studies on these catalysts such as


the above-said reports are highly desired.143,144
In the case of the OER, the reaction proceeds through
adsorption of water/hydroxide on the catalytic site (S−OH/S−
OH 2) followed by oxidation of the same along with
deprotonation to form the metal oxide (SO). Formation of
the hydroperoxide intermediate (S−O−OH) then occurs by
coordination of one more water molecule/hydroxide ion at the
same site, and this is followed by deprotonation and reductive
elimination of O2 from the active site, making it ready for the
next catalytic cycle. These are the common elementary steps in
all of the proposed pathways for the OER mechanism in acidic
and alkaline media.11,13,14,116 We can see from this that the
Figure 3. Plots of free energy of H adsorption on various metal OER is associated with a series of elementary steps and that
phosphide catalysts vs increasing fraction of H monolayer coverage. during each step the chemical environment around the active
Reproduced with permission from ref 143. Copyright 2015 Royal site is changing. If any one of these intermediates is not favored
Society of Chemistry. by its standard free energy of bonding, then the catalyst will not
be able to catalyze the OER. On the other hand, if the bond
adsorption with respect to reaction coordinate (Figure 4) strength is too high for any one of these intermediates, then
shows that Fe0.5Co0.5P has an optimal free energy of H catalyst poisoning occurs, resulting in a very high overpotential
adsorption very close to that of Pt. for the OER. In 2007, Rossmeisl et al.145 presented a good
theoretical survey of the efficiency of the OER activities of
oxides of Ir, Ru, and Ti in which they defined the theoretical
OER activity of the same by plotting their oxygen binding
energies against −ΔG (the negative of the change in the
standard free energy) associated with each of the above-said
elementary steps of the OER. That plot also contained a
horizontal dashed line indicating the thermodynamic over-
potential for the OER (1.23 eV). A material with high activity
has to stay on the horizontal line, but it will not be seen becaue
of the kinetic hindrances experienced by the catalysts in the real
system, which will fetch them an additional potential (the
overpotential). Moreover, the plot also says that regardless of
the nature of the binding of oxygen to the active site, the
activity is always limited by the oxide (SO*) and peroxide
Figure 4. Scheme of free energy of H adsorption with corresponding (S−O−O*) formation steps. However, catalysts with lower
reaction coordinate. Reproduced from ref 144. Copyright 2016 oxygen binding energies will stay closer to the theoretical
American Chemical Society. horizontal line of the OER (where RuOx is placed just above
IrOx as it binds more strongly to oxygen than RuOx does). In
the report of Man et al.,14 similar interpretations were made for
Although both of these reports say that Fe0.5Co0.5P is a better
nonprecious catalysts with Pt-like HER activity, it is not the various other binary metal oxides, including the three studied
bestthere is still room to improve the catalytic behavior. by Rossmeisl and co-workers along with some perovskites as
From Figure 3 we can see that only FeP and Fe0.5Co0.5P show OER catalysts. Here the difference in the standard free energies
such a reduction in the free energy of H adsorption, whereas of two subsequent steps, namely, hydroxide coordination and
the other catalysts do not show such a trend on the same. This oxidation of the same with deprotonation (ΔGO* − ΔGHO*)
could lead to the question of what makes them almost equally was plotted against the theoretical oxygen overpotential. It
active as FeP and Fe0.5Co0.5P. The answer lies in the higher should be noted that the standard free energy associated with
negative free energy of H adsorption at lower H coverage and the formation of hydroperoxide from the oxide intermediate
lower positive free energy of H adsorption at higher H coverage was not included here. This could lead to potential errors in
on these surfaces. This higher negative free energy of H predicting the catalytic activities of some catalysts with slow
adsorption at low H coverage shows its readiness to adsorb kinetics of hydroperoxide formation. As far as this work is being
hydrogen, and the lower positive free energy of H adsorption considered, the catalysts placed at the top of the plots, such as
the ease of H2 delivery from these catalytic surfaces and at the Co3O4, RuO2, IrO2, NiO, SrCoO3, LaNiO3, and SrNiO3, are
same time, its ability to achieve full Hads coverage as seen with the ones with the optimum oxygen bonding energies to catalyze
Pt. This is the reason why these FeP based systems have been the OER efficiently. From the above discussion, we can now
showing parallel activity to that of Pt in catalyzing HER. conclude here that the catalytic surface with high oxygen
Examples of this kind of catalyst are MoP and MoP/S, as binding energy would have the peroxide formation step as the
shown in Figure 3. This inference leads us to believe that the limiting step and the surface that binds oxygen weakly will have
better transition metal phosphide catalysts are still hidden and the hydroxide coordination and oxide formation steps as the
that changing the stoichiometric compositions among all of limiting steps. Therefore, an OER catalyst should neither bond
these known metal phosphides could lead to better HER too strongly nor too weakly to oxygen, as we have seen in the
catalysts. To carry out such worthy studies, combined case of the HER.
8076 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

5. EFFECT OF HETEROATOMS (S, SE, AND P) IN co-workers148 on P-terminated MoP surfaces focusing the
WATER SPLITTING Gibbs free energy of hydrogen adsorption (ΔG°H) revealed that
at lower Hads coverage ΔGH° is sufficiently negative (−0.34 eV)
A brief and nice highlight on the effect of P doping in metal
to accelerate the adsorption of hydrogen. However, when the
lattices on the electrocatalytic HER has recently been given by
surface becomes fully covered by Hads, ΔG°H becomes positive
Shi and Zhang.8 In this section of the review, we will also and accelerates the desorption of hydrogen, as observed earlier
elaborate the effect of other heteroatoms such as S and Se not in the cases of MoS2 with more S on the edges. This is in
only in the HER but also in the OER. agreement with the earlier reports of Kibsgaard et al.143 and
As highlighted by Shi and Zhang,8 the polarization-induced Tang et al.144 In the cases of S- and Se-doped metal lattices, a
partial negative charges localized on P centers in a metal similar mechanism should be responsible for the enhanced
phosphide structure with a P-terminated surface attracts HER catalytic property. However, the increased electro-
protons as a base and make their discharge easier, promoting negativity of S and the relatively higher strength of the S−H
the HER easily. As a consequence of this, it is obvious to expect bond (∼360 kJmol−1) do not allow many of the sulfides of Fe,
that metal phosphides with higher P content should show Co, and Ni to act as efficient HER catalysts. However, a few
higher HER catalytic activities. Interestingly, among the selected sulfides such as Ni3S2, CoS, CoS2, FeS2, and Co9S8 are
reported phosphides of Co, Ni, and Fe, the polymorphs with reported to be good in catalyzing the HER in acid. When the
higher P content are reported to be the highly active ones. bond strength is considered as an influencing factor in
Between CoP and Co2P, CoP is the more active one (Figure enhancing the hydrogen desorption from the catalytic site,
5a).146 Similarly, among Ni2P, Ni5P4, and Ni12P5, Ni5P4 with the selenides stand a step ahead of both sulfides and
the highest P content is the most active one (Figure 5b).147 phosphides, as the strengths of E−H bonds (E = S, P, Se)
Besides trapping the protons by acting as a base, these P centers are in the order S−H (363 kJ/mol) > P−H (322 kJ/mol) >
have recently been found to enhance the hydrogen desorption Se−H (276 kJ/mol).149 Moreover, like phosphides, both
at high Hads coverage. A DFT study carried out by Wang and sulfides and selenides can trap protons by acting as bases to
promote the discharge step faster.
The story on the influences of these heteroatoms in the OER
is different from the one we described above for the HER.
Unlike the HER, these heteroatoms do not have any direct
influence in enhancing the OER kinetics of these materials. In a
report by Subbaraman et al.,150 the trends in the electrocatalytic
OER activities of these 3d M2+ ions with oxide environments
are described, and it is stated that Ni2+ is more active than other
divalent cations that appear just before in the series. This is
mainly attributed to the increased 3d−2p repulsion between
the metal d-band center and the coordinated oxygen’s p-band
centers. However, when we have electronegative ligands (S2−,
Se2−, and P3−) in the vicinity of catalytically active metal sites,
there could be two effects. One of these is that the localized
negative charges on these heteroatoms (S and P only) will
actually deactivate the catalyst from coordinating with the
hydroxide ligand as a result of the increased 3p−2p repulsion.
This will increase the overpotential considerably. However,
after a tiresome coordination with hydroxide ligands, the
oxidation coupled with deprotonation followed by peroxide
formation will be supported because the deprotonation will
increase the electron population on the metal center and
thereby feed these electronegative heteroatoms. Similarly, the
formation of peroxide by coordination of an additional
hydroxide ligand happens just above the surface layers of
these heteroatoms and therefore is not affected during the
coordination processes. After formation of the peroxide
intermediate, the delivery of the dioxygen molecule is actually
accelerated by the enhanced 3p−2p repulsion between the
heteroatoms and the peroxide entity. In the case of Se in place
of S and P, we think that the coordination step with hydroxide
will be less affected because both the metal centers and the
ligand centers have 3d subshells as the outermost orbitals.
However, because of the negative charge localized on Se sites
and the cumulative 3d−2p repulsion between the metal d-band
Figure 5. (a) P-content-dependent HER activities in NixPy. center and Se d-band center, the delivery of the dioxygen
Reproduced with permission from ref 147. Copyright 2015 Royal molecule may become faster.
Society of Chemistry. (b) P-content-dependent HER activities in The facts discussed above on the OER performance of group
CoxP. Reproduced from ref 146. Copyright 2015 American Chemical VIII metal (Fe, Co, and Ni) sulfides, selenides, and phosphides
Society. are in resonance with the experimental reports. These are the
8077 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

reasons why the selenides are better OER catalysts than the of the procedures with a few necessary modification from one
phosphides and sulfides. In the HER, the selenides and material to another. The simple method of synthesizing metal
phosphides are better than the sulfides as a result of the sulfides is coprecipitation, in which the metal precursor salt is
strengths of the E−H bonds (E = S, Se, P) described above. placed in a solvent that dissolves both the metal ion and the
Hence, we conclude here by saying that an efficient water sulfide precursors. In this method, the reaction temperature and
splitting catalyst should be designed by considering the bond the pH should be carefully monitored to get reasonable yield.
energies of the intermediates formed and the electronic The main disadvantage of this method is that it requires
structure requirements of each element in the catalyst. additives such as surfactants, scaffolds, ligands, and chelating
agents to control the size and shape of the synthesized metal
6. OTHER REQUIREMENTS FOR A GOOD WATER sulfide nanostructure. Moreover, these additive are hard to
SPLITTING CATALYST remove after the synthesis. There are no reports on the
preparation of MxSy (M = Fe, Co, Ni) nanostructures by this
In addition to the key thermodynamic and kinetic requirements coprecipitation method for electrocatalytic water splitting
of an electrocatalyst discussed above, it is important to highlight applications.
some other essential requirements for an electrocatalyst to be The most common method of all time used to prepare metal
economically affordable in large-scale water electrolysis. sulfide nanostructures alone or with some three-dimensional
The first of these is the availability of resources of the (3D) matrix support is the solvothermal method. The common
catalytic material from which the catalyst can be formulated, as
sulfide sources used in this method are thiourea (TU,
we can no longer rely on noble metals (Pt, Ir, and Ru) to do
(NH2)2CS) and thioacetamide (TAA, CH3(CS)NH2).160,161
this simple water electrolysis. The second one concerns health
Among the sulfides of Fe, Co, and Ni, the sulfide
and environmental hazards. The catalyst should at least be less
nanostructures of Co have been relatively reported more than
harmful to persons working with it and to the environment in
other two metal sulfides. Di Giovanni et al.162 prepared FeS
which it is applied. Except for oxides and hydroxides of Fe, Co,
nanoparticles by solvothermal decomposition of Fe2S2(CO)6 at
and Ni, other compounds are potentially harmful. The third
230 °C for the HER under neutral conditions. Other available
one is the conductivity of the catalyst. Although heteroatom
reports describe Fe-based sulfides with Co as a dopant. Shen
doping of the metal lattice is fruitful in enhancing the catalytic
and co-workers reported the formation of Fe0.5Co0.5S@
activity, doping beyond certain limits will turn the metal
mesoporous N-doped graphitic carbon as an efficient ORR
surfaces into semiconducting and even insulating ones, which is
and OER catalyst under alkaline conditions by a solvothermal
not acceptable in electrocatalytic water splitting because it will
route where TU was used as the S source.163 Apart from these,
drastically increase the Ohmic drop. The fourth one is the need
Faber et al.164 recently reported pyrite-type FeS2 for the HER
for a high surface area of the catalyst, which would allow the
obtained by a two-step procedure that includes the formation of
amount of catalyst used to be reduced. The wettability is
a metallic thin film on a desired substrate by electron beam
another parameter to be considered while formulating a water
evaporation followed by thermal sulfidization taking elemental
splitting catalyst, as higher wettability lowers the Ohmic drop
S as the source directly. Co-based sulfides with varying S
caused by the formation of gas bubbles on the catalyst surface.
content such as CoS, CoS2, Co3S4, and Co9S8 have been
Another serious problem in dealing with these Fe-, Co-, and Ni-
prepared by solvothermal methods and reported with different
based catalysts is the current selectivity, as they tend to have
strong redox reactions within the potential window of the OER. morphologies. You et al.165 compared the HER electrocatalytic
In such cases we should be aware of where the applied current activities of hollow nanoprisms of CoS prepared by microwave
is being spent. A good electrocatalyst for water oxidation should (MW) and solvothermal methods taking TU as the S source.
have more current selectivity toward the OER rather than its Kornienko et al.166 showed the HER electrocatalytic activity of
own redox reactions. To know this, the Faradaic efficiency amorphous CoS with some O content in CoS2-like small
determination by the RRDE experiment can be used. The final amorphous clusters of CoS. Aslan and co-workers prepared
thing we should worry about is the corrosion resistance. Since CoS nanoparticles and anchored them on carbon nanotubes
the medium of water electrolysis is going to be either highly (CNTs) by a solvothermal method for the HER under alkaline
acidic or highly basic, the material that is supposed to catalyze conditions.167 Wang’s group recently published an interesting
the HER and OER should have very high corrosion resistance. report on the design of CoS@CTs@CP, an efficient 3D
With the sulfides, selenides, and phosphides of Fe, Co, and Ni, electrocatalyst for overall water splitting that comprises carbon
this problem is well documented in the literature and patents, tubes (CTs), carbon paper (CP), and CoS sheets (Figure 6).168
as the oxides of these heteroatoms (S, Se, and P) are well- Besides CoS, pyrite-type CoS2 is the next best documented
known corrosion inhibitors and resistors.151−159 Co-based sulfide nanocatalyst for water splitting. Faber et al.169
The above discussion certainly implies the advantageous use synthesized pyrite-type CoS2 by electron beam evaporation of
of the sulfides, selenides, and phosphides of group VIII 3d the metal target followed by thermal sulfidization and applied
metals in water electrolysis. In the following sections, the them to the HER. These pyrite CoS2 began to draw attention
synthetic strategies applied in designing such catalysts and their in mid-2014 and have been reported frequently with modified
subsequent applications in water electrolysis are highlighted in morphologies and support materials such as graphene oxide
detail. (GO), CNTs, N-doped CNTs (NCNTs), etc. The following
are some significant reports about them. It was Faber and co-
workers who reported the HER activity of CoS2 for the first
7. SYNTHETIC STRATEGIES EMPLOYED IN GROUP VIII time.169 Soon after that, Zhang’s group reported pyrite CoS2
3D METAL (FE, CO, AND NI) SULFIDE, SELENIDE, prepared by a solvothermal route as a versatile catalyst for the
AND PHOSPHIDE NANOSTRUCTURES HER over a wide pH range of 0−14.170 In the same year, Peng
7.1. Synthesis of Metal (Fe, Co, and Ni) Sulfides. The et al.171 fabricated CoS2 nanosheets on graphene@CNT as a
synthesis of metal sulfides has been achieved by following one flexible electrode for the HER. Then a similar report by
8078 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

graphene by a solvothermal approach using TU as the S source


was published by Liu and co-workers in 2013, and they applied
this material to the ORR and OER under alkaline conditions.184
Recently Liu et al.185 reported the electrodeposition of Ni−
Co−S and applied it to total water splitting. In a couple of
recent reports by Sun and co-workers, the sulfides of Ni and Co
have been fabricated into nanowire arrays via a solvothermal
route for application in total water splitting electrocataly-
sis.186,187 Ansovini et al.188 recently reported the formulation of
a 3D Co9S8−NixSy/Ni foam electrode for the HER by a
solvothermal method. The sulfide of Co−Fe alloy on N-doped
mesoporous carbon for the ORR and OER was recently
Figure 6. (a, b) SEM images of CP/CTs/Co−S. (c, d) TEM images, reported by Shen et al.163 Wang et al.189 reported the
(e) SAED pattern, and (f) elemental mapping of CT/Co−S. advantages of Co-doped FeS2 nanosheet/CNT arrays for
Reproduced from ref 168. Copyright 2016 American Chemical HER application. Similarly, Long et al.190 reported the
Society.
fabrication of ultrathin Fe−Ni−S nanosheets by a topotactic
sulfidization of Fe−Ni layered double hydroxides and examined
Ganesan et al.172 with S- and N-doped graphene as a support their HER properties under acidic conditions. Apart from these,
for CoS2 for use in the ORR and OER was published later in the sulfides of these three metals (Fe, Co, Ni) are also often
2015. Other than these two sulfides of Co, Co3S4 on N-doped alloyed with other metals to improve their electrocatalytic
CNTs was prepared by Wang and co-workers by a simple anion properties. One such catalyst, Ni−Mo−S on carbon fiber cloth
exchange method taking Co(OH)2/NCNTs as the starting synthesized by a solvothermal method for the HER under
material, and this material was applied to the OER.173 Feng et neutral conditions, was reported by Miao et al.191
al.174 proposed method for the formulation of Co9S8 armored For a better comparative overview of latest synthetic
with C that utilizes thiocyanuric acid as the S source and methodologies of these metal sulfides, we have tabulated the
thermal decomposition at 700 °C under an atmosphere of N2. materials, precursors, and methods in Table 2.
Very recently, Liang et al. published an interesting report on 7.2. Synthesis of Metal (Fe, Co, and Ni) Selenides. The
CoS and Zn-doped CoS for total water splitting where these synthesis of metal selenides resembles the synthesis of metal
catalysts were grown on a Ti mesh.175 sulfides in many ways. The most common method of forming
Like the sulfides of Fe and Co, Ni-based sulfides with good metal selenide nanostructures is the solvothermal route. The Se
electrocatalytic properties toward water splitting have been sources used for metal selenide synthesis are selenourea (SU),
reported. NiS and Ni3S2 are the most common Ni-based selenium metal powder, sodium hydrogen selenide (NaHSe)
sulfides applied for water splitting. The exceptional OER derived from the reaction of NaBH4 and crude Se powder, and
performance of Ni3S2 grown on Ni foam with low overpotential sodium selenide (Na2Se) in some cases, particularly when
was first reported by Zhou et al.176 in 2013. They synthesized aprotic solvents are used as the medium of synthesis.192 Unlike
this material by a simple solvothermal sulfurization of Ni foam the sulfides, the numbers of reports on the selenides of these
taking S directly as the sulfur source. In 2015, Tang et al.177 three metals are uneven. The most reported one is CoSe2 for
showed the HER performance of Ni3S2 nanosheets grown on both the OER and HER. The lone report on NiSe nanowires
Ni foam by a hydrothermal method for a pH range of 0−14. (NWs) was published by Tang et al.193 in 2015, where the NiSe
Then the same Ni3S2 in the form of a nanorod (NR) array on NWs were solvothermally grown on Ni foam and applied for
the same Ni foam for use in total water splitting was reported water splitting under strongly alkaline conditions (1 M KOH).
by Ouyang et al. later in 2015.178 In the last two reports of NaHSe derived from the reaction of crude Se powder and
Ni3S2 preparation, TU was used as the S source, and the sodium borohydride was used as the Se source during the
solvothermal treatment was kept constant in all three methods. solvothermal treatment of surface-etched-clean Ni foam. In the
In another report by Tang et al.,179 NiS hierarchical arrays case of Fe, there has been no report for either the HER or
grown on carbon cloth (CC) for efficient HER electrocatalysis OER. However, very recently Wang and co-workers reported
were obtained adapting a solvothermal route of synthesis. that the OER activity of NiSe was drastically increased when Fe
Another interesting report by Yang and co-workers achieved was introduced into the system, with a decrease in the
highly efficient and stable water splitting catalysis with Fe- overpotential of >100 mV.194 This indicates that although the
doped NiS nanaoarrays prepared by a solvothermal route.180 catalysts with Fe alone have not been shown to date to possess
Similarly, the electrocatalytic properties of NiS for the HER and any significant OER activity, their incorporation into the lattices
OER individually and also for bifunctional catalytic activity of active catalysts composed of Ni and Co may reduce the
were reported. Unlike Ni3S2, the bifunctional activity of NiS in overpotential and increase the activity. In that report, the Ni to
1 M KOH was first reported in 2015 by Zhu et al.181 following Fe ratio was maintained at 3:1 during the Ni−Fe ultrathin
similar hydrothermal route of sulfidizing Ni foam. Later in precursor sheet formation and also while it was selenized with
2015, Li et al. reported the atomic layer deposition of NiS on a NaHSe by a hydrothermal route.
silica substrate for the OER under alkaline conditions.182 Very Apart from the above two reports, the electrocatalysis by
recently, Mabayoje et al.183 comparatively reported the role of selenides of group VIII 3d transition metals is mainly occupied
the anion in nickel sulfide and selenide toward the OER. by CoSe2. Cobalt diselenide has been reported in both
Apart from the individual binary metal sulfides, it is common amorphous and crystalline forms with various morphologies
to see reports on sulfides of alloys of Fe, Co, and Ni. The and supports. Some of the important findings on the synthesis
sulfide of Ni−Co alloy is the most often reported alloy sulfide of this CoSe2 are briefly summarized below. Other than these
of this kind. The first report of the preparation of NiCo2S4@ there are two recent interesting reports on Ni3Se2 films and
8079 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Table 2. Methodologies and Precursors Used To Form the Sulfides of Fe, Co, and Ni
method and reaction
s. no. material and morphology metal precursor source of sulfur conditions ref
1 iron sulfide nanoparticles Fe2S2(CO)6 Fe2S2(CO)6 solvothermal heating at 230 °C 162
2 Co0.5Fe0.5S@N-MC cobalt(II) acetate and iron(III) nitrate thiourea self-assembly by F127 163
surfactant and annealing at
900 °C in Ar atmosphere
3 FeS2, CoS2, NiS2 films high-purity Fe, Co, and Ni metals sulfur powder electron beam evaporation and 164
thermal sulfidation
4 hollow cobalt sulfide nanoprisms cobalt acetate hydroxide thioacetamide microwave heating for 15 min 165
5 CoSx film CoCl2·6H2O thiourea electrodeposition at pH 7 166
6 CoS@CNT CoCl2·6H2O thioacetamide solvothermal method at 140 °C 167
for 24 h, N2 atmosphere
7 CoS@CNT@CFP sheets Co(NO3)2·6H2O thiourea electrodeposition for 8 min 168
8 CoS2 micro- and nanowires Co(NO3)2·6H2O and (Co(OH)(CO3)0.5·xH2O sulfur powder thermal sulfidation at 500 °C 169
for 1 h
9 CoS2 pyramids@Ti foil CoCl2·6H2O thiourea hydrothermal method at 180 170
°C for 15 h
10 CoS nanosheets@graphene@CNT Co(Ac)2·4H2O CS2 and thiourea hydrothermal method 171
combined with vacuum
filtration
11 CoS nanoparticles@N- and S-doped graphene Co(TU)4(NO3)2 complex Co(TU)4(NO3)2 solid-state thermolysis at 400, 172
oxide complex 500, and 600 °C
12 Co3S4@NCNTs CoCl2·6H2O Na2S and solvothermal method at 160 °C 173
thioacetamide for 15 h
13 Co9S8 nanoparticles@C Co(NO3)2·6H2O trithiocyanuric acid annealing at 700 °C in N2 174
atmosphere for 3 h
14 Zn0.76Co0.24 S/CoS2 nanowires@Ti mesh ZnCo2O4 nanowires sulfur powder annealing at 400 °C for 2 h 175
15 Ni3S2 nanorods@Ni foam nickel foam thioacetamide hydrothermal method at 180 176
°C 4 h
16 Ni3S2 nanosheets@Ni foam Ni(NO3)2·6H2O thiourea hydrothermal method at 120 177
°C 10 h
17 Ni3S2 nanorod arrays@nickel foam nickel foam thiourea hydrothermal method at 160 178
°C for 6 h
18 NiS2 nanosheets@CC Ni(NO3)2·6H2O sulfur powder hydrothermal method at 100 179
°C for 10 h
19 iron-doped nickel disulfide nanoarray@Ti NiFe LDH sulfur powder annealing at 400 °C for 60 min 180
under Ar atmosphere
20 nickel sulfide microsphere film@Ni foam nickel foam sulfur powder annealing at 300 °C with a rate 181
of 8 °C min−1 under Ar
atmosphere
21 NiSx films bis(N,N′-di-tert- butylacetamidinato)nickel(II) hydrogen sulfide vapor-phase atomic layer 182
deposition
22 NiS film NiCl2 thiourea electrodeposition 183
23 NiCo2S4 nanoparticles@graphene Ni(CH3COO)2·4H2O and Co(CH3COO)2·4H2O thiourea solvothermal method at 200 °C 184
for 6 h
24 NiCoS nanosheet films Ni(CH3COO)2·4H2O and Co(CH3COO)2·4H2O thiourea electrodeposition 185
25 CoS nanosheets@Ti mesh CoCl2·6H2O thiourea electrodeposition 186
26 NiCo2S4 nanowires@CC NiCl2·6H2O and CoCl2·6H2O sulfur powder annealing at 300 °C for 2 h 187
27 Co9S8−NixSy@Ni foam Co(NO3)2·6H2O and nickel foam thiourea annealing of cobalt thiourea 188
complex on Ni foam at 300
°C for 5 min
28 Fe1−xCoxS2/CNT Fe(NO3)3 and Co(Ac)2 thioacetamide 90 °C in an oil bath for 24 h. 189
29 iron−nickel sulfide nanosheets FeNi LDH thioacetamide hydrothermal method at 120 190
°C for 6 h
30 Ni−Mo−S nanosheets@CC Na2MoO4·2H2O and NiSO4·6H2O L-cysteine hydrothermal method at 200 191
°C for 24 h

NiSe2 nanowires fabricated via electrodeposition and solvo- performance shown by this CoSe2 NPs/CFP catalyst was better
thermal routes by Sun and co-workers for application in total than the previous report. In 2015, more attention was paid to
water splitting.195,196 Like the other two selenides (NiSe NWs synthesizing CoSe2 in various forms using various methods with
and Ni−Fe−Se nanosheets), the synthesis of CoSe2 is also the ultimate aim of total water splitting. In this regard, for the
achieved by the solvothermal method. It was Xu and co- first time Liu et al.199 reported the bifunctional catalytic activity
workers who in 2013 reported the exceptional HER activity of of CoSe2 as electrodeposited amorphous thin films on Ti foils
CoSe2 nanobelts to which Ni/NiO was anchored by electro- for full water splitting. Before this, Carim and co-workers had
deposition on a glassy carbon electrode (GCE).197 Soon after already reported the HER activity of electrodeposited CoSe
in 2014, the same was reported as CoSe2 NPs grown on carbon amorphous thin films in 2014.200 Then the attention was
fiber paper (CFP) by Kong et al.,198 who utilized thermal- deflected toward crystalline CoSe2. Zhang et al.201 reported the
pyrolysis-assisted cobalt oxide NP formation on CFP followed HER activity of polymorphic CoSe2 with mixed orthorhombic
by selenization under a Se vapor atmosphere. The HER and cubic phases. Recently, Liao et al.202 reported the OER
8080 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

activity of coral-like CoSe2 nanostructures prepared by a source, trioctylphosphine (TOP), is used along with
solvothermal route. As expected, the CoSe2 was then trioctylphosphine oxide (TOPO) as a structure-directing
composited with various other active catalysts such as MoS2 agent in various molar ratios. In the high-temperature solid-
and Ni−Fe LDH and synergistically enhancing compounds and state synthesis, the desired metal precursor is reacted with
substrates such as CeO2, TiO2, and graphene to improve the sodium hypophosphite (Na2HPO2), commercially known as
catalytic activity of CoSe2. A clever thing done by Hou and co- hypo, as the source of P under highly oxygen-free conditions.
workers was the synthesis of a CoSe2/Ni−Fe LDH on Other than hypo, elemental phosphorus (preferably red
graphene composite, which resulted in better bifunctional phosphorus) and ammonium hydrogen phosphate
water splitting catalytic activity than any other report.121 ((NH4)2HPO4) are also used as P sources.8,10 However, with
Similarly, Gao et al.105 synergistically improved the HER the latter two P sources, the reaction temperature needs to be
performance of CoSe2/MoS2 by making a composite of them. 2-fold higher than that required for the solution-phase synthesis
With a different synthetic route involving microwave using TOP. The inert atmospheres used to date for the
irradiation, Ullah et al.203 reported the same CoSe2 but as a synthesis of these metal phosphide nanostructures are Ar, H2,
composite with graphene and TiO2 with improved HER N2, and mixtures of H2 and N2 under either static or flowing
performance. Zheng and his group members showed the effect conditions. Similarly, in all of the liquid-phase syntheses the
of CeO2 on the OER catalysis led by CoSe2 with a nanobelt common metal precursors used are the acetate or acetylacet-
morphology.204 As a step ahead in the HER catalysis by CoSe2, onate complexes of the respective metal ions or the nitrates and
Liu et al.205 formed a 3D HER catalyst by simple hydrothermal chlorides of the metals with a phase-transfer catalyst (PTC)
growth of CoSe2 on carbon cloth (Figure 7). such as tetraoctylammonium bromide (TOAB).206,207 In the
case of the high-temperature solid−gas-phase reaction, the
nitrates are the most common metal precursors used. Unlike
the sulfides and selenides, the phosphides of Fe, Co, and Ni
have been equally explored in terms of synthesis and their
subsequent applications to water splitting. In the case of Fe, the
phosphides reported to date are FeP, FeP2, and Fe2P. Among
these, FeP is the more frequently reported one. As mentioned
in the earlier discussion on the effects of heteroatoms in water
splitting, the higher P content in FeP helps it to be more active
than Fe2P. Moreover, where bifunctionality is concerned,
among the available reports on the phosphides of Fe there is
only one report by Yan et al.,208 who grew ZnO NWs on CFP,
which were then doped with Fe3+ by hydrolysis followed by
phosphidization at 300 °C with hypo in an Ar atmosphere.
Similarly, there is a lone report by Xiong et al.209 on the OER
catalytic activity of FeP NRs supported on CFP obtained
through a set of hydrothermal and gas-phase phosphidization
reactions. Almost all of the other reported FeP materials in
various morphologies and crystalline forms are devoted only to
HER catalysis. Some of their synthesis sequences are briefly
summarized below. FeP has been reported as nanowires,
nanotubes, nanosheets, thin films, and nanocrystals and as a
Figure 7. (a) XRD patterns of the precursor and selenized product composite with N-doped CNTs for HER applications. The first
scratched down from CC. (b, c) SEM images of Co(OH)F NW/CC. report on HER electrocatalysis on FeP was published by Xu
(d) TEM image of a Co(OH)F NW. (e, f) SEM images of CoSe2 and co-workers, who synthesized FeP nanoporous nanosheets
NW/CC. (g) TEM image of a CoSe2 NW. (h) HRTEM image and (i) by an anion exchange method.210 Du and co-workers proposed
SAED pattern of a CoSe2 NW. (j) STEM image and EDX elemental an interesting method of preparing FeP NRs by a simple hard-
mapping of cobalt and selenium for a CoSe2 NW. (k) SEM image of a
CoSe2 MP. Reproduced from ref 205. Copyright 2015 American
template method using anodized aluminum oxide (AAO) as the
Chemical Society. desired hard template, onto which the Fe3+ precursor was
loaded through a sequence of soaking and drying processes
followed by phosphidization with hypo in an alumina boat at
We have summarized the methods and materials used for the 350 °C.211 Later, Liang et al.212 obtained nanoarrays (NAs) of
synthesis of these metal selenides in Table 3. FeP NRs on CC by a solvothermal treatment to form the
7.3. Synthesis of Metal (Fe, Co, and Ni) Phosphides. Fe2O3/CC precursor, which was then phosphidized using hypo
Although there are two good surveys on the synthesis of at relatively lower temperature than other methods (Figure 8).
transition metal phosphides and their applications to the HER, Using a similar synthetic route, Tian and co-workers obtained
it would be better to have a consolidated review of the OER 3D FeP NP thin films on CC that showed better HER
activities of Fe, Co, and Ni phosphides in addition to the performance than other phosphides in both acidic and neutral
sulfides and selenides of the same. The synthesis of metal electrolytes.213 In a similar report by Yang et al.,214 the same
phosphide nanostructures is much more difficult than that of CC was used once again for the formation of rugae-like FeP
simple sulfides and selenides and is almost always done at high and FeP2 nanocrystal (NC) arrays, which were used later for
temperature in an inert atmosphere. The metal phosphide HER electrocatalysis in acid. Jiang and co-workers chose a Ti
syntheses mainly fall into three categories. In the high- plate instead of CC as the substrate for the FeP NW array.215
temperature liquid-phase synthesis, the most common P The synthesis was a two-step process involving initial
8081 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Table 3. Methodologies and Precursors Used for the Synthesis of Selenides of Fe, Co, and Ni
source of
s. no. material and morphology metal precursor selenium method and reaction conditions ref
1 MoS2/CoSe2 Co(OAc)2·H2O and (NH4)2MoS4 Na2SeO3 hydrothermal method at 200 °C 105
for 10 h
2 Co0.85Se/NiFe LDH Co(NO3)2·6H2O, Ni(NO3)2·6H2O, and Fe(NO3)3·9H2O Na2SeO3 hydrothermal method at 150 °C 121
nanosheets for 48 h
3 NiSe nanowire film@Ni foam nickel foam Se powder hydrothermal method at 140 °C 193
for 12 h
4 (Ni0.75Fe0.25)Se2@CFC Ni(NO3)2·6H2O and Fe(NO3)3·9H2O Se powder hydrothermal method at 180 °C 194
for 24 h
5 Ni3Se2 film@Cu foam Ni(CH3COO)2·4H2O SeO2 electrodeposition 195
6 NiSe2 nanoparticle film NiCl2·6H2O SeO2 electrodeposition 196
7 Ni/NiO/CoSe2 nanocomposite Co(Ac)2·H2O and nickel(II) 2,4- pentanedionate Na2SeO3 hydrothermal method at 180 °C 197
for 16 h
8 CoSe2 Nanoparticles@CFP Co(NO3)2·6H2O Se powder thermal selenization in Ar 198
atmosphere
9 CoSe film@Ti mesh Co(Ac)2·4H2O SeO2 electrodeposition 199
10 amorphous cobalt selenide Co(Ac)2·4H2O SeO2 electrodeposition 200
films@Ti foil
11 polymorphic CoSe2 CoCl2 SeO2 electrodeposition 201
12 3D coral-like CoSe Co(NO3)2·6H2O Se powder hydrothermal method at 180 °C 202
for 15 h
13 CoSe2/graphene−TiO2 CoCl2 Se powder microwave heating for 15 min 203
heterostructure
14 CeO2/CoSe2 nanobelts Co(Ac)2·H2O and Ce(CH3COO)2 Na2SeO3 polyol reduction method at 278 204
°C for 1 h
15 CoSe2 nanowires Co(NO3)2·6H2O Se powder hydrothermal method at 140 °C 205
for 10 h

phosphidization under an atmosphere of N2 with hypo at 350


°C and applied these composites to the HER.216
The phosphides of Co have been studied more than those of
Fe in both the HER and OER. The phosphides of Co known to
date are CoP and Co2P. As seen in the case of FexP, CoP with
the higher P content is reported to be a better catalyst for the
HER and OER than Co2P. Moreover, the number of reports on
CoP is 3-fold higher than that on Co2P. Co2P has been
reported as nanoneedles and NWs and as a composite with
CNTs and NCNTs for OER and HER applications. The
synthesis of Co2P resembles the synthesis of FexP. One such
recent report with a significantly different synthetic route was
made by Dutta and co-workers, who synthesized Co2P
nanoneedles by a liquid-phase alkylamine-assisted synthesis
but instead of TOP as the P source, PH3 gas produced ex-situ
was purged into the reaction vessel at 230 °C. This method
provided the significant advantage of avoiding the strong
possibility of explosion by other methods in the case when a
little bit of O2 is present inside the reaction chamber.217 In the
case of CoP, there is a cluster of reports on their synthesis and
subsequent applications to the HER and OER. The synthesis of
CoP can be done by any of the above-seen methods for the
synthesis of FexP, viz., liquid-phase metal ion reduction and
Figure 8. (a, b) SEM images of (a) Fe2O3 NAs/CC and (b) FeP NAs/ phosphidization, high-temperature solid−gas-phase reaction, or
CC. (c) TEM and (d) HRTEM images of FeP nanorods. (e) SAED anion exchange reaction. Like the contribution made by Sun
pattern recorded from the FeP nanorod. (f) STEM image and EDX and co-workers in the case of sulfide and selenide catalysts of
elemental mapping of C, P, and Fe for the FeP NAs/CC. Reproduced these three metals, they have done an enormous amount of
from ref 212. Copyright 2014 American Chemical Society. work on cobalt phosphide too by various methodologies and
subsequently used them for both HER and OER applications.
Among them, fabrication of self-supported nanoporous CoP
hydrothermal growth of FeOOH on the Ti plate followed by its nanowire arrays,218 CoP nanowire arrays for sensing and
chemical conversion to an FeP NW array on the Ti plate at low photcatalytic HER,219 templated-assisted synthesis of CoP
temperature. As expected, Liu and co-workers composited N- nanotubes,220 CoP nanowire arrays on Ti mesh,221 3D
doped CNTs to FeP by growing the Fe2O3/NCNT precursor interconnected CoP nanowire arrays,222 and a combined
composite first via a hydrothermal route followed by experimental−theoretical study of CoP with varying Fe dopant
8082 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

contents are the significant works.144 Popczun et al.223 reported Very recently, Zhuo et al.249 reported the effect of Se doping
the HER activity under acidic conditions of CoP NPs prepared onto the NCs of NiP and NiP2 on their HER activities. As an
by the first method mentioned above. Subsequently, a number interesting advancement in the nickel phosphide research, Jiang
of reports appeared in literature with either a significant and co-workers reported the exceptional HER performance of
improvement in the catalytic activity or an improvement NiP2 under both acidic and alkaline conditions.250 The higher P
resulting in a relatively easier synthetic route. CoP has been content in NiP2 fetches an additional corrosion resistance under
reported with a branched starlike morphology by Popczun et harsh alkaline conditions and improves the HER performance
al.,224 as hollow polyhedra by Liu and Li,225 as urchinlike NCs of the same. Similarly, the Ni5P4 NCs were found to be highly
by Yang et al.,226 as porous NR bundles by Niu et al.,227 as stable while catalyzing the HER under both acidic and alkaline
nanosheets on a Ti foil substrate by Pu et al.,228 as self- conditions because of the increased P content, as reported by
supported mesoporous NRs by Zhu et al.,229 as surface-oxidized Laursen and co-workers.251 As a step ahead and above, the
NRs by Chang et al.,230 again as NRs by Huang et al.,231 as same Ni5P4 was later applied to full water splitting by
nanoporous NWs by Gu et al.,232 with a mixed morphology by Ledendecker and co-workers because it holds better stability
Jiang et al.233 and as CoOOH-covered CoP NPs by Ryu et in alkali than other phosphides of Ni.252 Huang and co-workers
al.234 Apart from various morphologies, as expected CoP was showed the catalytic and photocatalytic HER performance of
composited with carbonaceous materials such as CNTs by Liu Ni12P5 NPs.253
et al.235 and with GO by Ma et al.236 via hydrothermal The methodologies and materials applied to the synthesis of
deposition of CoP on GO. One other important study was the phosphides of Fe, Co, and Ni are summarized in Table 4.
reported by Hou et al.,237 who decorated the CNTs with Beyond the monometallic phosphides, studies of phosphides
ultrafine CoP NPs and applied them as a highly active of mixed metals and their electrocatalytic performance toward
bifunctional water splitting catalyst. Beyond this, an interesting water splitting have recently also started to appear in the
report on CoP@C core−shell nanostructures was first made by literature. The first such mixed metallic phosphide of Ni and
Wang’s group, who applied them to HER electrocatalysis.238 Co was reported by Feng and co-workers, who prepared quasi-
Like the FeP case, CoP was also grown on CC by Li et al.239 hollow Ni−Co−P nanocubes and applied them to the HER in
through the two-step universal process of growing metal alkaline solution.254 A little later, a ternary array of Ni−Co−P
phosphides on CC. nanosheets was synthesized by Li and co-workers, who applied
The phosphides of Ni are quite different from those of Fe it to total water splitting under alkaline conditions.255 The
and Co. Unlike the latter, the phosphides of Ni have been advantage of making such bimetallic phosphides is that it helps
reported beyond the mono- and dimetallic centers. The nickel
to achieve better durability performance during both the HER
phosphides reported to date are NiP, Ni2P, Ni5P4, Ni12P5, and
and OER as reported by Li et al.255
NiP2. The P content and its direct influence on the catalytic
Having briefly discussed the synthetic strategies for the
performance in the HER is well-explained in the very recent
sulfides, selenides, and phosphides of Fe, Co, and Ni, we can
review of metal phosphide synthesis for HER applications by
come to the conclusion that there is still room to improve the
Shi and Zhang.8 Moreover, the evolution of nickel phosphides
synthetic routes of these sulfides, selenides, and phosphides to
with varying P content led Kucernak and Sundaram240 to make
reduce the associated experimental, environmental, and health
a linear relationship between the P content and the catalytic
activity of the metal phosphide toward the HER in acidic hazards by minimizing the utilization of more toxic precursors,
media. As observed with Fe and Co, the nickel phosphide with avoiding high-temperature reactions and explosive reaction
high P content (NiP) was found to be a better HER catalyst conditions, and minimizing the waste from solution-based
than the others. However, such a direct correlation between the syntheses. We have also seen that the phosphides of Fe, Co,
P content and the OER activity of metal phosphides cannot be and Ni have been found to be better catalysts than the sulfides
made because the mechanism of the OER is much more and selenides of the same. Now it is time to find simplified,
complicated than that of the HER. Syntheses of nickel easier, one-step, quick synthetic strategies to make the efficient
phosphides were also done by any one of the three known metal phosphides affordable for industrial-level large-scale H2
methods of phosphidization as explained for Co and Fe production more conveniently. Beyond the extensive contribu-
phosphides. Nickel phosphides with various P contents were tions made by Sun and his group members to this particular
obtained simply by changing the molar ratio of the metal ion field of electrocatalysis of water splitting using sulfides,
precursor to the source of P, as was done by Kucernak and selenides, and phosphides of Fe, Co, and Ni, they have newly
Sundaram.240 Among the known nickel phosphides, Ni2P is the reported two other catalysts, namely, a Ni−Mo hollow nanorod
most frequently reported one with various morphologies such array for total water splitting256 and amorphous NiB alloy NPs
as metallic nanosheets by Li and co-workers,241 urchinlike grown on Ni foam for total water splitting as well.257 These
crystals on Ni foam by You and co-workers,242 nanoaggregates reports have basically made us aware that there is still room to
by Li and co-workers,243 W-doped Ni2P microspheres by Jin improve the catalytic efficiencies of these catalysts. From the
and co-workers,111 nanosheets of Ni foam by Shi and co- above discussion, it is quite obvious to expect a review summary
workers,244 monodispersed NCs with different phases by Pan on Co-catalyst-based materials for the electrocatalysis of water
and co-workers,147 nanoflakes (NFs) on graphene/Ni foam splitting from Sun and his group members, and the same has
hybrid electrode for the HER from pH 0 to 14 by Han and co- appeared very recently in the literature.258 This particular
workers,245 and NRs on Ni foam by Wang and co-workers.246 review nicely corroborates the activity trends of various Co
The second most reported nickel phosphide is NiP. NiP has catalysts such as its oxides, sulfides, selenides, and phosphides
been reported by Yu et al.247 as carbon-coated porous for both the HER and OER in acidic, basic, and neutral
nanoplates that were applied in the OER, and Wang and co- environments. However, it should be noted here that Sun’s
workers reported the one-step formation of NiP nanosheet review is centered around only the Co-based catalysts and does
arrays on Ni foam for efficient HER in acidic electrolyte.248 not cover Ni- and Fe-based ones completely. This has been
8083 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
Table 4. Methodologies and Precursors Used in the Synthesis of Metal (Fe, Co, and Ni) Phosphides
s. no. material and morphology metal precursor source of phosphorus method and reaction conditions ref
1 iron phosphide nanotubes Fe(NO3)3·9H2O NaH2PO2 heated at 300 °C for 30 min under Ar flow 208
2 FeP nanorods@CP FeCl3·6H2O red phosphorus phosphorization treatment in P vapor at 500 °C for 209
ACS Catalysis

30 min
3 FeP nanosheets Fe18S25−TETAH trioctylphosphine anion exchange reaction by diffusion at >300 °C 210
4 FeP nanorods FeCl2 NaH2PO2 heated at 350 °C for 2 h under Ar flow 211
5 FeP nanorod arrays FeCl3·6H2O NaH2PO2 heated at 300 °C for 2 h under Ar flow 212
6 FeP nanoparticle film@CC Fe(NO3)3·9H2O NaH2PO2 heated at 300 °C for 2 h under Ar flow 213
7 rugae-like FeP nanocrystal@CC FeSO4 NaH2PO2 phosphidation at 500 °C 214
8 3D FeP nanowires FeCl3 NaH2PO2 heated at 300 °C for 2 h in a static Ar atmosphere 215
9 FeP@NCNT FeCl3·6H2O NaH2PO2 Calcined at 350 °C under N2 flow for 2 h 216
10 surface-oxidized Co2P nanoneedles CoCl2·6H2O Ca3P2 solvothermal method at 230 °C for 30 min 217
11 3D CoP nanowires@CC Co(NO3)2·6H2O NaH2PO2 heated at 300 °C for 1 h in a static Ar atmosphere 218
12 CoP nanowires Co(Ac)2 NaH2PO2 heated at 300 °C for 2 h 219
13 CoP nanotubes CoCl2·6H2O NaH2PO2 heated at 300 °C for 2 h in a static Ar atmosphere 220
14 CoP@Ti mesh Co(NO3)2·6H2O NaH2PO2 heated at 300 °C for 2 h in a static Ar atmosphere 221
15 CoP nanoparticles@CC CoCl2·6H2O NaH2PO2 heated at 300 °C for 2 h in Ar flow 222
17 CoP nanoparticles Co2(CO)8 trioctylphosphine solvothermal method at 320 °C for 1 h 223
18 branched CoP nanostructures cobalt(II) acetylacetonate trioctylphosphine and trioctylphosphine heated at 120 °C for 1 h under vacuum 224
oxide
19 CoP hollow polyhedra Co(NO3)2·6H2O NaH2PO2·H2O heated at 250 °C for 2 h in a static N2 atmosphere 225
20 urchinlike CoP nanocrystals Co(NO3)2·6H2O NaH2PO2 heated at 300 °C for 150 min in a static N2 226

8084
atmosphere
21 CoP nanorod arrays@Ti Co(NO3)2·6H2O NaH2PO2 heated at 300 °C for 2 h in a static Ar atmosphere 227
22 CoP nanosheet arrays@Ti plate Co(NO3)2·6H2O NaH2PO2 heated at 300 °C for 1 h in a static Ar atmosphere 228
23 CoP nanorod arrays@Ni foam CoCl2 NaH2PO2 electrodeposition 229
24 surface-oxidized CoP nanorods CoCl2·6H2O NaH2PO2·H2O heated at 300 °C for 2 h in a static N2 atmosphere 230
25 Co2P nanorods Co(Ac)2·4H2O trioctylphosphine solvothermal method at 120 °C for 30 min in N2 231
flow
27 CoP nanoparticles@C cobalt(II) acetylacetonate trioctylphosphine solvothermal method at 300 °C 234
28 CNTs decorated with CoP Co(Ac)2 NaH2PO2 heated at 300 °C for 2 h 235
nanocrystals
29 CoP nanoparticles@RGO Co(Ac)2·4H2O NaH2PO2 heated at 300 °C for 2 h 236
30 CoP nanoparticles@CNT Co(Ac)2 NaH2PO2 heated at 300 °C for 2 h in Ar flow 237
31 CoP@C core−shell nanocables cobalt(II) acetylacetonate triphenylphosphine heated in a sealed tube at 400 °C for 100 min 238
32 nickel phosphide NiSO4 NaH2PO2 electrodeposition 240
33 Ni2P nanosheets NiCl2·6H2O NaH2PO2 heated at 380 °C for 15 h 241
34 urchinlike Ni2P@Ni foam Ni foam NaH2PO2·H2O heated at 400 °C for 2 h in a static Ar atmosphere 242
35 NixPy Ni(NO3)2·6H2O NaH2PO2·H2O heated at 275−475 °C for 2 h in a static Ar 243
atmosphere
36 Ni2P nanosheets@Ni foam Ni foam trioctylphosphine Solvothermal method at 320 °C for 2 h 244
37 Ni2P nanoflakes@graphene/Ni foam Ni foam KH2PO4 Chemical vapor deposition process 245
38 Ni5P4−Ni2P nanosheet array Ni foam red phosphorus heated at 500 °C for 6 h in N2 flow 246
39 carbon-coated nickel phosphide potassium tetracyanidonickelate(II) and NiCl2·xH2O NaH2PO2 heated at 300 °C for 2 h in a static Ar atmosphere 247
nanoplates
Review

40 Ni foam red phosphorus 248

DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
Ni2P nanorods@Ni foam hydrothermal method at 200 °C for 48 h
ACS Catalysis Review

rectified in the present review, which includes updated

249
250
251
252
253

111
144
ref literature reports.

solvothermal method at 390 °C for for 30 min under


heated at 500 °C for 30 min in Ar flow atmosphere
heated at 300 °C for 1 h in a static Ar atmosphere
8. APPLICATIONS OF GROUP VIII 3D METAL (FE, CO,
AND NI) SULFIDE, SELENIDE, AND PHOSPHIDE
heated at 550 °C for 1 h in inert atmosphere

solvothermal method at 80−90 °C for 2 h


NANOSTRUCTURES IN WATER SPLITTING
method and reaction conditions

solvothermal method at 390 °C for 1·5 h

heated at 300 °C for 2 h under Ar flow


Before the discussion of the applications of sulfide, selenide,
and phosphide nanostructures of Fe, Co, and Ni, it would be
useful to elaborate the most common problems that are faced
during the application of nanostructured catalysts in water
splitting and some recent trends of fabricating electrodes using
these nanostructured catalysts for water splitting.
8.1. Common Problems Encountered with Nano-
structured Catalysts in Water Splitting. One of the
major problems encountered with nanostructured catalysts is
achieving reasonable stability without compromising the
N2

catalytic activity. The sources of stability failure can include


the binder used to tether the nanostructured catalysts to the
substrate/current collector, the pH at which electrolysis is
carried out, high surface energy of the nanomaterials, a decrease
source of phosphorus

in conductivity due to overoxidation of the catalyst (in the case


of the OER), and the effect of loading on the conductivity.
8.1.1. Binders Used To Tether the Nanostructured
trioctylphosphine oxide

Catalysts and Associated Demerits in Water Splitting. In


triphenylphosphine

general, the binder is chosen on the basis of the medium in


red phosphorus

red phosphorus

which the electrolysis is carried out. Since the most reactive


NaH2PO2

NaH2PO2

HER and OER catalysts are found to catalyze these


Na2PO2

electrochemical reactions at either extremely low pH (strongly


acidic) or extremely high pH (strongly alkaline), the most
obvious choice is to go with a binder that can offer excellent
FeCl3·6H2O, Fe(NO3)3·9H2O, CoCl2·6H2O, and Co(NO3)2·6H2O

proton transfer in acidic solutions and good hydroxide


conductivity in alkaline conditions. The common commercial
binder used for this purpose is Nafion, which is actually a
perfluorinated alkyl sulfonate ionomer that is available at
various concentrations ranging from 5% to 40% as a solution in
water or a methanol/propanol mixture. The combination of the
very high fluorine content in the carbon backbone and its high
metal precursor

hydrophobicity with the sufficiently high polarity and hydro-


philicity of the sulfonate functional group helps the protons to
get transferred from the aqueous solutions to the hydrophobic
electrode surfaces easily. As a consequence of these advantages,
Nafion has been widely used as a binder for the HER and OER
under acidic conditions. However, when the electrolysis is
performed under alkaline conditions, this proton transfer
Ni(NO3)2·6H2O
Ni(NO3)2·6H2O

ionomer fails to deliver the same efficiency as it did in the


Ni(acac)2·xH2O

Ni(Ac)2·4H2O

acidic medium. As a result, the resistance of the catalyst−


electrolyte interface is considerably increased, which leads to a
Ni foil

NiSO4

drastic degradation in the catalytic activity and affects the


stability of the nanostructured-catalyst-modified electrodes.
However, the same 5 wt % Nafion solution with additives
such as propyl alcohol or isopropyl alcohol in water in the
volume ratio of 0.5:2.0:7.5 is being used as a binder for many
material and morphology

nanostructured catalysts in alkaline water electrolysis. In this


W-doped NixP microspheres
FexCo1−xP nanowire array

case, the hydroxide conductivity is believed to occur via the


water channels available on the thin film formed while drying
nanocrystalline Ni5P4

Ni12P5 nanoparticles
Ni2P NP films@Ti

this binder solution with the desired catalyst on the substrate


Table 4. continued

electrode surface. Although this modification makes the Nafion


Se-doped NiP2

Ni5P4 Films

applicable in alkaline water electrolysis, the serious problem


associated with it is the hydration of the thin films formed
during drying upon prolonged exposure to strongly acidic and
alkaline conditions. This results in a gradual increase in the
s. no.

thickness of the catalyst layer on the substrate electrode, leading


41
42
43
44
45

46
47

to increased catalyst resistance, which in turn decreases the


8085 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

catalytic activity and affects stability of the catalyst directly. To their counterparts, and when they are exposed to such a high
overcome this problem, attention has now been diverted applied electric potential field, there is a fair chance of
toward some anion exchange ionomers made up of some agglomeration with nearby particles. This will ultimately reduce
indole-based polymers and ionic liquids. However, these are the overall number of available active sites compared with the
quite expensive, and the efforts that have been made to reduce initial stage of the catalytic process. Besides agglomeration,
the overall expenses associated with water electrolysis by there are chances for overoxidation of the nanomaterial surface
making these nanostructured catalysts would become mean- due to these higher surface energies in cases where the
ingless if we must rely on a costly anion exchange ionomer to agglomeration of nearby particles is restricted by high dilution
attain reasonable stability with our catalysts. Other than these, it of the catalyst concentration. As a consequence of over-
is quite common to see reports with some other binders such as oxidation of the catalyst surface, the resistance associated with
poly(tetrafluoroethylene) (PTFE), poly(vinyl pyrrolidine) the catalyst will also increase, which will ultimately result in
(PVP), poly(vinyl alcohol) (PVA), N-methylpyrrolidone reduced catalytic activity. Efforts should be made to avoid all of
(NMP), and dimethylformamide (DMF). However, except the above problems associated with nanostructured catalysts by
for DMF the other binders listed above are nonconductive and optimizing the experimental conditions with care.
require a conductive additive such as carbon black in equal 8.2. Recent Trends in Electrode Fabrication with
proportion to the catalyst material. This could lead to wrong Nanostructured Catalysts for Prolonged Water Electrol-
interpretation in assessing the catalytic activity of the desired ysis. All of the stability problems elaborated in the previous
catalyst by either partially masking the active sites in the catalyst section can be overcome by simple and brilliant electrode
or additionally contributing to the overall activity. Though fabrication methods. The most stable nanostructured-catalyst-
DMF does not require any such conductive additive, the modified interfaces that have recently been reported make use
stability of the catalyst-modified surface is poor compared with of one of the following two methods: electrochemical
other binders, particularly when it is subjected to the deposition or hydrothermal/solvothermal growth of nano-
accelerated degradation test and long-term galvanostatic or structured catalysts on the desired substrate materials.
potentiostatic electrolysis. Very recently, in this concern of 8.2.1. Electrochemical-Deposition-Assisted Improvement
overcoming the problems associated with the binders, we have in the Stability of Nanostructured Catalysts. Obviously, it is
exploited the use of DNA molecular self-assemblies, which are more advantageous to support the catalysts on the desired
cheaper than these anion and proton exchange ionomers, by current collectors by electrochemical deposition than to
anchoring of the ultrasmall Pt NPs for the HER under acidic prepare the same by other methods and depend on suitable
conditions and IrO2 NPs for the OER under alkaline binders. However, this method is not always suitable for all
conditions.17,259 These Pt NP-anchored DNA molecular self- kinds of materials. It is most reported for metal-, metal oxide-,
assemblies have shown extreme stability and increased catalytic and metal hydroxide-based nanostructures. Additionally, there
performance over a commercial Pt/C catalyst under identical are reports for sulfides and selenides too. The common
experimental conditions.17 The mechanism of binding with substrates employed are metal foils such as Ti and Cu, metals
DNA molecular self-assemblies is mainly dependent on the such as Ni in the form of Ni mesh and Ni foam, indium-doped
electrostatic interactions between the DNA molecules and tin oxide (ITO), fluorine-doped tin oxide (FTO), carbon fiber
various charged and polarized entities during polarization of the paper, and carbon cloth. Depending on the material, any of the
electrode in both the anodic and cathodic directions; apart from above substrates can be used for electrodeposition. For
these, the catalyst resistance is not increased by the conductivity example, in photoelectrocatalytic water splitting, ITO and
associated with the DNA as a consequence of free electron FTO are the most obvious choices of substrates. The major
movements by resonance among the purine and pyrimidine advantages of this method are the ability to monitor the
bases in their backboned. This study now has opened up quantity of deposited catalyst and the control over the catalyst
pathways utilizing other biomolecules with charged and film thickness. The disadvantages are the inefficiency in
polarized entities, such as amino acids, peptides, proteins, and producing nanomaterials with various other morphologies and
polysaccharides, as binders for other materials in the future. the requirement for sophisticated instruments such as electro-
8.1.2. pH of the Medium of Water Electrolysis. The chemical workstations.
medium of water electrolysis plays a crucial role in water 8.2.2. Hydrothermal- or Solvothermal-Growth-Assisted
electrolysis, as elaborated earlier under various categories. Here Improvement in the Stability of Nanostructured Catalysts.
the effect of the pH of the water electrolysis solution on the This method has recently been explored by materials scientists,
stability of the nanostructured catalyst is discussed. We have and it is suitable for almost all sort of materials. In this method,
seen that at under strongly acidic or alkaline conditions, non- the substrate on which the catalysts must be grown is taken
noble-metal catalysts such as the ones considered in this review along in the hydrothermal/solvothermal vessel after necessary
are highly prone to corrosion, which affects the stability pretreatments. The most common substrates are Ti foil, Ni
directly. To overcome this issue, many catalysts have now been foam, carbon fiber paper, and carbon cloth. These methods
designed to be active at neutral pH. In this regard, the metal offer the advantage of preparing catalysts with various
phosphides, sulfides, and selenides have more advantages than morphologies of a hierarchical nature, which enables some
the metals, metal oxides, and metal hydroxides, as the oxides of special studies such as morphology- and shape-selective
S, Se, and P are highly resistive toward corrosion. Hence, these catalytic studies. The main disadvantage of this method is the
materials find very limited stability problems due to the lack of control over the quantity of the catalyst grown on the
solution pH of water electrolysis. desired substrate. Both methods are efficient in fabricating
8.1.3. High Surface Energy and Overoxidation of the highly stable 3D catalyst electrodes and should be chosen
Catalyst Surface. These are two other serious problems that according to the material of interest.
reduce the stability of catalysts for water electrolysis. It is 8.3. Sulfides of Fe, Co, and Ni in Water Splitting.
known that the nanomaterials have higher surface energies than Having discussed the methods of evaluating an electrocatalyst
8086 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

for water splitting and the synthesis of the catalyst materials, we


now proceed to a discussion of benchmarking all of the
reported sulfides of Fe, Co, and Ni with respect to the
overpotential at 10 mA/cm2 (η10) and Tafel slope values under
two categories, viz., metal sulfides for the HER and metal
sulfides for the OER. The activity parameters for total water
splitting catalysts are split up and included along with other
catalysts used for the HER and OER. The same trend will be
followed for the selenides and phosphides too. As far as total
water splitting is concerned, no Fe-based sulfides have been
reported to date. There are only few reports on total water
splitting by these metal (Fe, Co, and Ni) sulfides. The report by
Wang et al.168 on CoS supported on CTs that in turn were Figure 9. Benchmarking the metal (Fe, Co, Ni) sulfides with respect
grown on CC is the only one on the total water splitting activity to the HER overpotential at 10 mA/cm2 and the corresponding trend
in the Tafel slopes.
of a Co-based sulfide. However, the literature says that the Ni-
based sulfides are better bifunctional catalysts than the Fe- and
Co-based ones. NiS microspheres grown on Ni foam and Ni3S2 primarily indicates that the thermodynamic activity parameter
nanosheets grown on Ni foam are good in catalyzing both the (i.e., the overpotential) does not affect or only slightly affects
HER and OER under alkaline conditions.171 However, a the kinetic activity parameter (i.e., the Tafel slope). Similar
significant advancement in the bifunctional activity of these trends are also observed with selenide- and phosphide-based
sulfides is achieved when they are present as bimetallic sulfides. catalysts for both the HER and OER, as discussed in the
The electrodeposited Ni−Co−S reported by Liu et al.185 and subsequent sections.
the Ni-promoted formation of CoS2 NW arrays on CC as 8.3.2. Activity Trends of Metal (Fe, Co, Ni) Sulfides in the
reported by Fang et al.186 are examples of this kind. Electrocatalytic OER. Although the sulfides of Fe, Co, and Ni
8.3.1. Activity Trends of Metal (Fe, Co, Ni) Sulfides in the have been devoted more toward the HER, there are also reports
Electrocatalytic HER. The available reports on the sulfides of available on the OER catalytic performance of these sulfides,
Fe, Co, and Ni clearly reveals that they have been employed particularly the sulfides of Co and Ni. There is no data on the
more for the HER than for the OER and total water splitting. It OER and total water splitting activities of Fe-based sulfides
has been highlighted in the synthesis section that to date there alone. However, Fe has been reported by Shen and co-
have been no reports on Fe-based sulfides except the one on workers163 as a bimetallic sulfide with Co as Co−Fe−S on N-
FeS for the HER by Di Giovanni et al.162 and the one on pyrite- doped mesoporous carbon for the OER along with the ORR.
type FeS2 for the HER by Faber et al.164 On the other hand, The story with Co-based sulfides is different from that of Fe-
literature reports on Co-based sulfides are more abundant than based sulfides. Though Co-based sulfides have also been much
for both Fe and Ni. Interestingly, almost 80% of the available devoted to the HER, there are reports on their OER and total
reports on Co-based sulfides were studied for the HER, and the water splitting activities. Another thing to be noticed here is
reports for the OER and total water splitting are limited. that wherever a Co-based sulfide is applied to the OER or total
Among the various Co-based sulfides such as CoS, CoS2, Co3S4, water splitting, it is always applied with a substrate material
and Co9S8, CoS and pyrite-type CoS2 are the ones frequently such as a carbon nanostructure or Ti foil. Among the reported
reported for the HER. In the case of Ni-based sulfides, there are catalysts are Co3S4 on N-doped CNT by Wang et al.,173 CoS
only two reports for the HER, namely, the report on NiS2 by nanosheets grown on Ti by Liu et al.,262 and CoS2 grown on N-
Faber et al.164 and the report on Ni3S2 by Tang et al.187 Apart and S-codoped GO by Ganesan et al.172 In the case of Ni, there
from these, there are reports on the HER activities of bimetallic is a lone report on Ni3S2 nanosheets grown on Ni foam that
sulfides such as Co−Fe−S, Co−Ni−S, Fe−Ni−S, and Ni−Co− showed a low overpotential for the OER under alkaline
S. To have a comparative view of the catalytic activities of these conditions. Other than these, bimetallic NiCo2S4@graphene
sulfides toward the HER, they have been benchmarked with has been reported by Liu et al.184 for the OER and ORR under
respect to their overpotential at 10 mA/cm2 and Tafel slope, as alkaline conditions. As for the HER, as a comparative measure
shown in Figure 9. More information on the electrochemical the plots of η10 and the Tafel slope are shown in Figure 10, and
conditions under which these data were acquired and others are the data are listed in Table 6.
provided in Table 5. From Figure 9 it is clear that the FeNiS As we have seen for the HER, the activity trends of these
catalyst reported by Long and co-workers is the best catalyst metal sulfide catalysts benchmarked with respect to η10 and the
with the lowest HER η10 of 105 mV. In Table 5, other catalysts Tafel slope follow an irregular trend in terms of kinetics. This
are listed in order of increasing η10. again implies the same conclusion that the Tafel slope
Since there had been a serious discussion of the selection of (kinetics) has no direct relation to the overpotential
activity parameters to evaluate an electrocatalyst’s efficiency for (thermodynamics) of a water splitting catalyst.
water splitting, it is fruitful to arrive at a conclusion whether the 8.4. Selenides of Fe, Co, and Ni in Water Splitting. The
use of η10 with the Tafel slope as the primary activity parameter irregularities seen in the trends of selenides of Fe, Co, and Ni
is good for evaluating catalysts. Figure 9 clearly reveals that the are more vigorous for both the OER and HER than those
overpotential has nothing to do with the kinetics of the observed with the sulfides. This indicates the strong influence
electrocatalytic water splitting process, as we can see that some of a ligand with an ionic radius comparable to that of the central
catalysts with low η10 have larger Tafel slopes than some metal atom, which belongs to the 3d series of the periodic table
catalysts with relatively larger η10, such as CoS|P/CNT reported with 4s shells. However, the best correlations among these
by Liu and co-workers,260 which showed an overpotential of metal selenides are given for a comparative evaluation of the
480 mV at 10 mA/cm2 with a Tafel slope of 55 mV/dec. This HER and OER activity trends via the benchmarking plots of η10
8087 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Table 5. Benchmarking the Metal (Fe, Co, Ni) Sulfides with Respect to the HER Overpotential at 10 mA/cm2 (η10)
s. no. catalyst electrolyte η10 (mV) Tafel slope (mV/dec)a ref
1 A-FeNiS 0.5 M H2SO4 105 40 190
2 B-FeNiS 0.5 M H2SO4 117 48 190
3 CoS2 0.5 M H2SO4 128 52 192
4 Ni2.3%CoS2/CC 1 M KOH 136 106 235
5 Co3S4/NCNTs 0.5 M H2SO4 140 70 189
6 NiCoS/CC NSs 1 M KOH 140 96 177
7 CoS2 NS/RGO/CNT 0.5 M H2SO4 142 51 171
8 (Fe0.48Co0.52)S2 0.5 M H2SO4 143 47.5 164
9 NiS/Ni Foam 1 M KOH 150 83 181
10 Ni2.3%CoS2/CC 1 M KOH 150 106 186
11 Fe0.9Co0.1S2/CNT 0.5 M H2SO4 160 46 190
12 Co3S4 0.5 M H2SO4 160 79 173
13 Co9S8−NixSy/NiF 1 M KOH 163 88 188
14 (Co0.59Ni0.41)S2 0.5 M H2SO4 170 50.4 170
15 CoS2/Ti (pH 0.3) 0.5 M H2SO4 190 72 170
16 NiCo2S4 NA/CC 1 M KOH ∼190 141 187
17 Ni3S2@Ni 1 M KOH 195 107 178
18 Fe0.07Ni0.91S2 0.5 M H2SO4 196 58.7 164
19 NiMoS/C 1:1 pH 7 200 85.3 191
20 Zn0.76Co0.24S/CoS on Ti mesh 1 M KOH ∼200 164 175
21 B-NiS 0.5 M H2SO4 202 − 190
22 FeS2 0.5 M H2SO4 217 56.4 164
23 NiS2 0.5 M H2SO4 230 48.8 164
24 NixSy/NiF 1 M KOH 230 87 188
25 MW-CoS (nanoprism) 0.5 M H2SO4 230 76 175
26 ST-CoS (nanoprism) 0.5 M H2SO4 240 90 175
27 NiS on CC pH 7 243 69 179
28 CoS2/Ti (pH 13.37) 0.5 M H2SO4 244 133 183
29 Fe0.1NiS2 NA/Ti 1 M KOH ∼250 108 180
30 Ni3S2−Ni 1 M KOH 270 141 178
31 Co9S8/C phosphate buffer 280 − 174
32 CoS2 NS/RGO 0.5 M H2SO4 280 82 171
33 3D G/CoSx 1 M PBS 330 93 176
34 Co9S8 (all pH) phosphate buffer 340 − 174
35 FeS (pH 7) 0.1 M phosphate buffer 350 150 162
36 Co9S8-700 phosphate buffer 370 − 174
37 CoS|P/CNT 0.5 M H2SO4 480 55 260
a
Dashes indicate that the corresponding data are not available in the respective reports cited here.

selenides for the HER. However, there are some bifunctional


water splitting bi- and trimetallic selenides. On the other hand,
the most reported Co-based selenide, CoSe2 with various
morphologies and substrate materials, has itself shown different
trends in the HER catalytic performance. The same can be seen
when they are benchmarked with respect to η10 and Tafel slope,
as shown in Figure 11.
As all of them are the same material, it is hard to make a
conclusion about the trend of the HER activity of 3d group VIII
metal (Fe, Co, and Ni) selenides. However, the Co0.13Ni0.87Se2/
Ti is the one with the lowest overpotential of 64 mV at 10 mA/
Figure 10. Benchmarking the metal (Fe, Co, Ni) sulfides with respect cm2 for the HER among all other selenide-based HER catalysts,
to the OER overpotential at 10 mA/cm2 and the corresponding trend as reported by Liu and co-workers.262 The selenides of the
in the Tafel slopes. group VIII 3d metals that have been reported to date are listed
in Table 7 in increasing order of η10 along with the
and Tafel slope in Figures 11 and 12 and the data in Tables 7 experimental conditions under which the evaluations were
and 8, respectively. made. As observed with the sulfides, the selenides show similar
8.4.1. Activity Trends of Metal (Fe, Co, Ni) Selenides in the trends in the correlation between the thermodynamic activity
Electrocatalytic HER. As we highlighted in the synthetic parameter (the overpotential) and the kinetic activity parameter
methodologies part, there are no reports on Ni- and Fe-based (the Tafel slope). This once again emphasizes the independ-
8088 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Table 6. Benchmarking the Metal (Fe, Co, Ni) Sulfides with Respect to the OER Overpotential at 10 mA/cm2 (η10)
s. no. catalyst electrolyte η10 (mV) Tafel slope (mV/dec)a ref
1 Fe0.1NiS2 NA/Ti 1 M KOH ∼205 43 180
2 Ni2.3%CoS2/CC 1 M KOH 297 117 262
3 NiS/Ni foam 1 M KOH 300 89 179
4 Zn0.76Co0.24S/CoS on Ti mesh 1 M KOH ∼300 79 175
5 Ni2.3%CoS2/CC 1 M KOH ∼300 119 186
6 Ni3S2@Ni 0.1 M KOH 330 163 178
7 NiCoS/CC NSs 1 M KOH 330 109 177
8 NiCo2S4 NA/CC 1 M KOH 340 89 187
9 NiCo2S4@N/S-RGO 0.1 M KOH 355 − 184
10 CoS2/N−S GO 1 M KOH 370 − 172
11 ALD NiSx 1 M KOH 372 41 182
12 Co3S4 0.1 M KOH 375 − 184
13 Ni3S2−Ni 0.1 M KOH 410 330 178
14 CP/CNT/CoS 1 M KOH 450 − 168
15 Ni3S4 0.1 M KOH 555 − 184
a
Dashes indicate that the corresponding data are not available in the respective reports cited here.

expected, as similar irregular trend in the Tafel slopes is seen


when these selenide-based catalysts are queued according to
increasing overpotential. This too emphasizes again the
independence of the overpotential from the nature of the
HER kinetics and vice versa.
8.5. The Phosphides of Fe, Co, and Ni in Water
Splitting. Similar to the sulfides and selenides, the Activity
trends of the phosphides of Fe, Co, and Ni are comparative
evaluated by plotting the overpotential against the phosphide
catalysts in the increasing overpotential order as shown in
Figure 13 and Figure 14 and the experimental conditions under
Figure 11. Benchmarking the metal (Fe, Co, Ni) selenides with
which the evaluation were made are listed in Table 9 and Table
respect to the HER overpotential at 10 mA/cm 2 and the
corresponding trend in the Tafel slopes. 10 for HER and OER respectively.
8.5.1. Activity Trends of Metal (Fe, Co, Ni) Phosphides in
the Electrocatalytic HER. Unlike the selenides, the phosphides
of Fe, Co, and Ni have been equally reported for the HER.
Where the HER is concerned, the maximum number of reports
(∼40) have been published using the phosphides. The sulfides
come second, and the number of selenide-based HER catalysts
is minor compared with the sulfides and phosphides. The plots
of η10 and the Tafel slope for the metal phosphides for the HER
are shown in Figure 13, and Table 9 provides the experimental
details under which the measurements of the HER activities of
these phosphide catalysts were made. From Figure 13 and
Table 9 an interesting observation can be made: lower HER
Figure 12. Benchmarking the metal (Fe, Co, Ni) selenides with
overpotentials at 10 mA/cm2 were shown predominantly by the
respect to the OER overpotential at 10 mA/cm 2 and the Fe-based phosphides compared with the Co- and Ni-based
corresponding trend in the Tafel slopes. phosphides. This is in sharp contrast to the HER activity trends
of the sulfides and selenides of Fe, Co, and Ni. For the sulfides
and selenides, either Co or Ni or a mixture of Co and Ni
ence of the overpotential from the nature of the HER kinetics occupied the top position in the table of HER and OER activity
and vice versa. trends, in accordance with the activity trends of 3d metal oxides
8.4.2. Activity Trends of Metal (Fe, Co, Ni) Selenides in the and hydroxides predicted by Subbaraman and co-workers
Electrocatalytic OER. As with the HER catalysis, the OER earlier.150
catalysis by these metal selenides is mainly occupied by CoSe2 However, there are some specific phosphides of Ni and Co
and other Co-based sulfides. Some of the significant works with specific P contents, such as Ni5P4, NiP2, and CoP, that
include the CoSe oxohalides reported by Rabbani et al.,263 the showed good HER performance comparable to that of the Fe-
coral-like 3D CoSe2 crystals reported by Liao et al.,202 and the based phosphides. As observed with the sulfides and selenides,
CeO2-doped CoSe2 reported by Zheng et al.204 For NiSe2 there the phosphide catalysts ranked by the overpotential at 10 mA/
is a report by Tang et al.193 As we did for the HER, cm2 show an irregular trend in the Tafel slopes, as catalysts with
benchmarking plots of the activity parameters (i.e., η10 and the low overpotentials were found to show higher Tafel slopes and
Tafel slope) of these selenides are provided in Figure 12, and vice versa. This indicates to us that the phosphides are not
other relevant data on the same are included in Table 8. As exceptions and that they too have independent thermody-
8089 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Table 7. Benchmarking the Metal (Fe, Co, Ni) Selenides with Respect to the HER Overpotential at 10 mA/cm2 (η10)
s. no. catalyst electrolyte η10 (mV) Tafel slope (mV/dec) ref
1 Co0.13Ni0.87Se2/Ti 1 M KOH 64 63 262
2 MoS2/CoSe2 0.5 M H2SO4 68 36 105
3 electrodeposited Ni3Se2 1 M KOH 70 82 195
4 NiSe/NF 1 M KOH 96 120 193
5 Ni3Se2/CF 1 M KOH 100 98 196
6 Ni/NiO/CoSe2 0.5 M H2SO4 100 39 197
7 NiPSe (1.93:0.07) 0.5 M H2SO4 100 43 249
8 CoSe2−NW/CC 0.5 M H2SO4 130 32 205
9 CoSe2/CP 0.5 M H2SO4 137 40 198
10 NiPSe(0.09:1.91) 0.5 M H2SO4 170 44 249
11 NiSe2 0.5 M H2SO4 175 33 249
12 CoSe2-MP/CC 0.5 M H2SO4 193 50 205
13 EG/CoSe-NiFe-LDH 1 M KOH 260 −a 121
14 Co0.13Ni0.87Se2/Ti 1 M KOH 270 94 262
15 CoPSe/MWCNT/GCE 0.5 M H2SO4 370 46 117
a
Dash indicates that the corresponding data are not available in the respective reports cited here.

Table 8. Benchmarking the Metal (Fe, Co, Ni) Selenides with Respect to the OER Overpotential at 10 mA/cm2 (η10)
s. no. catalyst electrolyte η10 (mV) Tafel slope (mV/dec) ref
1 EG/CoSe-NiFe-LDH 1 M KOH 250 57 121
2 coral-like CoSe2 1 M KOH 290 40 172
3 electrodeposited Ni3Se2 1 M KOH 290 82 195
4 CoSe2/Ti mesh 1 M KOH 292 69 199
5 Ni3Se2/CF 1 M KOH ∼300 80 196
6 CeO2/CoSe2 1 M KOH 310 44 204
7 CoSe2 NPs 1 M KOH 350 49 172
8 NiSe/NF 1 M KOH 400 −a 193
9 Mn3O4−CoSe2 0.1 M KOH 450 49 185
a
Dash indicates that the corresponding data are not available in the respective reports cited here.

namics and kinetics of water splitting. However, the shape of


the Tafel slope line in Figure 13 is showing merely a linear
relation between the overpotential and the Tafel slope values,
unlike the sulfides and selenides.
8.5.2. Activity Trends of Metal (Fe, Co, Ni) Phosphides in
the Electrocatalytic OER. Application of phosphides of Fe, Co,
and Ni to the OER is new, and there are relatively few reports
compared with the number of reports available on the same for
HER applications. However, here we have made similar
comparative evaluation of the available reports on the metal
phosphides for OER applications via the benchmarking plots of
Figure 13. Benchmarking the metal (Fe, Co, Ni) phosphides with
η10 and the Tafel slope shown in Figure 14 along with Table 10
respect to the HER overpotential at 10 mA/cm 2 and the carrying information on the experimental conditions under
corresponding trend in the Tafel slopes. which the results on the OER activities of these metal
phosphides were acquired. As expected, there is an irregular
trend in the Tafel slopes when these catalysts are queued in
increasing order of their overpotentials at 10 mA/cm2. Hence,
the same conclusion of the independence of the kinetics of an
electrocatalyst from the thermodynamics of the catalyst can be
drawn.
8.6. Activity Trends of Metal (Fe, Co, Ni) Phospho-
sulfides and Phosphoselenides in Electrocatalytic HER
and OER. Beyond the sulfides, selenides, and phosphides alone
as electrocatalysts, people have now moved forward in
combining the effects of two different ligands such as P and
S and P and Se together with a single metal among Fe, Co, and
Figure 14. Benchmarking the metal (Fe, Co, Ni) phosphides with Ni. The reports where such kinds of materials have been
respect to the OER overpotential at 10 mA/cm 2 and the employed for the HER, OER, and total water splitting to date
corresponding trend in the Tafel slopes. are summarized here. Pyrite-structured CoPS for the HER was
8090 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

Table 9. Benchmarking the Metal (Fe, Co, Ni) Phosphides with Respect to the HER Overpotential at 10 mA/cm2 (η10)
s. no. catalyst electrolyte η10 (mV) Tafel slope (mV/dec)a ref
1 Ni5P4 0.5 M H2SO4 24 27 251
2 FeP/CC nanocrystal (300 °C) 0.5 M H2SO4 34 29.2 214
3 Fe0.5Co0.5P/CC 0.5 M H2SO4 39 30 222
4 NixPy 0.5 M H2SO4 40 46.1 213
5 CoP NPs/CC 0.5 M H2SO4 48 70 221
6 Ni5P4 1 M NaOH 50 14 251
7 CoP MNA 0.5 M H2SO4 54 51 229
8 Ni2P−G@NiF 0.5 M H2SO4 55 30 245
9 FeP NAS/CC 0.5 M H2SO4 58 45 212
10 nanoporous CoP pH 0−14 67 51 218
11 CoP/Ti mesh 1 M KOH 72 65 221
12 NiP2−NS/CC 0.5 M H2SO4 75 51 250
13 3D CoP NWs on Ti 0.5 M H2SO4 ∼80 65 222
14 CoP-NBAs/Ti 0.5 M H2SO4 90 40 227
15 CoP/NCNT 0.5 M H2SO4 92 49 236
16 CoP branched NS on Ti Foil 0.5 M H2SO4 98 − 224
17 Ni2P/Ni/NF H2 sat. 1 M KOH 98 72 242
18 Ni2P−G@NiF 1 M KOH 100 30 245
19 CoP NWs 0.5 M H2SO4 110 54 233
20 FeP/NCNT 0.5 M H2SO4 113 59 162
21 NiP2−NS/CC 1 M KOH 110 64 250
22 CoP@NPC 0.5 M H2SO4 120 69 115
23 Ni5P4−Ni2P NS 0.5 M H2SO4 120 79.1 246
24 CoP/CNT 0.5 M H2SO4 123 54 140
25 CoP/C 0.1 M KOH 130 − 234
26 CoP NTs 0.5 M H2SO4 130 60 220
27 Ni2P-NRs-Ni N2 sat. 0.5 M H2SO4 130 106.1 248
28 Fe2P/NGR 0.5 M H2SO4 138 − 211
29 NiCoP hollow NCs 1 M KOH 145 60.6 254
30 Co2P/NCNT 0.5 M H2SO4 148 62 236
31 Ni2P/NF H2 sat. 1 M KOH 150 93 242
32 CoP hollow polyhedra 0.5 M H2SO4 159 59 225
33 Ni2P−G@NiF 1 M KPi 160 40 245
34 CoP NSs 0.5 M H2SO4 164 61 233
35 CoP/CNT 0.5 M H2SO4 165 68 236
36 CoP 0.5 M H2SO4 168 46 226
37 CoP@C Ar sat 0.5 M H2SO4 170 61 238
38 Co2P/CNT 0.5 M H2SO4 195 74 236
39 CoP NPs 0.5 M H2SO4 221 87 233
40 CoP/RGO (0.30) 0.5 M H2SO4 240 104.8 236
41 CoP 0.5 M H2SO4 383 90 236
42 Co2P 0.5 M H2SO4 406 101 236
a
Dashes indicate that the corresponding data are not available in the respective reports cited here.

Table 10. Benchmarking the Metal (Fe, Co, Ni) Phosphides with Respect to the OER Overpotential at 10 mA/cm2 (η10)
s. no. catalyst electrolyte η10 (mV) Tafel slope (mV/dec)a ref.
1 Ni2P/Ni/NF O2 sat. 1 M KOH 205 − 242
2 Ni2P/NF O2 sat. 1 M KOH 220 − 242
3 Ni2P 1 M KOH 300 64 248
4 Co2P NPs 1 M KOH 310 50 222
5 CoP/Ti mesh 1 M KOH 310 87 221
6 CoP MNA 1 M KOH 310 65 229
7 Ni5P4 1 M KOH 330 − 252
8 CoP/C 0.1 M KOH 358 66 234
9 CP@FeP 1 M KOH 365 63.6 209
10 CoP hollow polyhedra 1 M KOH 400 57 225
a
Dashes indicate that the corresponding data are not available in the respective reports cited here.

8091 DOI: 10.1021/acscatal.6b02479


ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

very recently reported by Liu and co-workers.261 Similarly, Xiao of the lower values of Tafel slopes (approximately up to 55
and co-workers reported exceptional HER performance of Co− mV/dec) were seen when the studies were carried in acidic
P−Se NPs.117 Another interesting report was recently solution. This again indicates that although there are potential
published by Zhuo and co-workers on the HER activity of catalysts being developed for the HER in alkaline media, there
Se-doped pyrite-type NiP2.250 These catalysts are included with is no catalyst with better kinetics than observed in acidic HER
the sulfide and selenide benchmarking plots (Figures 9 and 11) electrocatalysis. One more thing that needs to be emphasized
for comparative evaluation, and the respective experimental here in the benchmarking of catalysts is that the lowest
details are included in Tables 5 and 7, respectively. These kinds overpotential is shown by fabricated electrodes such as
of materials are quite promising and can offer new pathways to nanoarrays on various supports such as Ti mesh, CC, CF, Ni
increase the catalytic activities of these sulfides and selenides. foam, and carbon paper. In these cases the catalyst loading is
Having seen a detailed comparative evaluation by means normally 5−10-fold higher than for the normal nanocatalysts
benchmarking plots and the trends observed with the Tafel used in other studies, which leads us to misunderstand that they
slopes, it is now essential to compare the best HER catalysts are better than the other studies when actually they are not.
among the sulfides, selenides, and phosphides as well as the This again emphasizes the conclusion that the mass-normalized
best OER catalysts among the sulfides, selenides, and current density would be better than the geometrical-area- or
phosphides. Such comparative plots are provided in Figure ECSA-normalized current density for a fair comparison of
15, from which we can conclude that the phosphides of Fe, Co, different catalysts. Having reviewed intensively the current
perspectives in electrochemical water splitting catalysis, we
recommend that the following activity parameters be
compulsorily incorporated in future reports of any such
electrocatalysts for the OER and HER: overpotential at a
defined current density normalized by the geometrical area and
the mass of loaded catalyst, the Tafel slope, the exchange
current density (for the HER), the mass activity, and the
loading.

9. SUMMARY AND OUTLOOK


The increasing pressure due to the faster depletion of
conventional fossil fuels and the negative impacts of the same
on the environment have urged the research community to find
alternative and viable sources of energy to enlighten the
decendants of humankind in the near future. As per the recent
studies of energy consumption and the availability of sources,
by 2015 the world will require 30 TW of power from new and
nonconventional energy sources.1 Though energies harvested
from solar, wind, tide, and other similar natural and endless
resources seem to be the most promising future ways of energy
generation, the major problem associated with those resources
is that they are seasonal and hence require large-scale energy
storage systems beyond the incapable batteries and super-
capacitors. One such indirect way of large-scale energy storage
is water splitting.
As we are aware that conventional combustion engines
Figure 15. Benchmarking plots of the best-ever metal (Fe, Co, Ni) running on conventional carbon fuels are badly contaminating
sulfide, selenide, and phosphide (a) HER and (b) OER catalysts the environment, which is considered as the most serious threat
according to their overpotentials at 10 mA/cm2. for humankind in the near future, people have now diverted
their attention to new zero-carbon-emitting and/or less-carbon-
and Ni are better catalysts than the sulfides and selenides of the emitting engines such as fuel cells. Fuel cell technology is a
same for both the HER and OER. This offers us new pathway greener available way of generating electrical energy from
to modify the catalytic properties of these materials by tuning various sources such as hydrogen, oxygen, methanol, glycerol,
the heteroatom (S, Se, P) content and/or mixing them with a borohydride, formic acid, and more as fuels. Among them, the
single metal atom among Fe, Co, and Ni or with a bi- or fuel cell that uses H2 and O2 as fuel is the greenest way to
trimetallic alloy of the same. With the given comparative view, produce electrical energy via a simple electrochemical redox
the knowledge of the activity trends of the Fe-, Co-, and Ni- reaction between them that produces water as the byproduct.
based sulfide, selenide, and phosphide catalysts is now clearly The efficiency of these fuel cells can be as high as possible if the
exposed. Hence, we can recommend that beyond only the H2 supplied is the purest. The conventional coal reforming
overpotential and the Tafel slope, other activity parameters technology for H2 production is certainly not the one that gives
such as mass activity, specific activity and TOF are to be the purest H2. The H2 obtained by coal reformation contains a
determined to elucidate the real catalytic efficiency of an considerable quantity of CO which is a serious threat to the
electrocatalyst for water splitting. Careful investigation of cathodes of fuel cells, as it will poison them by making strong
Tables 5, 7, and 9 reveals that the anomaly observed in the carbonyl coordination compounds with Pt atoms on the
Tafel slope values with respect to the overpotentials is mainly catalyst’s surface, thereby reducing the number of active sites
due to the pH of the medium of electrolysis. It is clear that all available for the desired electrochemical reaction. Great
8092 DOI: 10.1021/acscatal.6b02479
ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

attention has now been given to water electrolysis for the


production of H2 in the purest form. However, there are several
■ REFERENCES
(1) Habas, S. E.; Platt, H. A. S.; Van Hest, M. F. A. M.; Ginley, D. S.
thermodynamic and kinetic barriers in splitting water electro- Chem. Rev. 2010, 110, 6571−6594.
chemically, and they cause a huge energy loss in this process, (2) Steele, B. C.; Heinzel, A. Nature 2001, 414, 345−352.
which increases the overpotentials of the HER and OER at the (3) Aricò, A. S.; Srinivasan, S.; Antonucci, V. Fuel Cells 2001, 1, 133−
cathode and anode further. Hence, both the HER and OER 161.
need catalysts. (4) Wasmus, S.; Küver, A. J. Electroanal. Chem. 1999, 461, 14−31.
In earlier days, for HER the Pt electrode was used as the (5) Liu, H.; Song, C.; Zhang, L.; Zhang, J.; Wang, H.; Wilkinson, D.
state-of-the-art electrocatalyst, and it is still so used. Similarly, P. J. Power Sources 2006, 155, 95−110.
(6) Wang, Y.; Chen, K. S.; Mishler, J.; Cho, S. C.; Adroher, X. C.
for the OER either Ir or Ru and their compounds are used as Appl. Energy 2011, 88, 981−1007.
the state-of-the-art catalysts. However, we know that they are (7) Kirubakaran, A.; Jain, S.; Nema, R. K. Renewable Sustainable
noble, expensive, and less abundant metals. Hence, we cannot Energy Rev. 2009, 13, 2430−2440.
afford such expensive materials for simple water electrolysis, (8) Shi, Y.; Zhang, B. Chem. Soc. Rev. 2016, 45, 1529−1541.
during which there are potential chances for losing the (9) Vesborg, P. C. K.; Seger, B.; Chorkendorff, I. J. Phys. Chem. Lett.
materials by corrosion as the OER is the process that can 2015, 6 (6), 951−957.
occur at very high anodic overpotential compared with the (10) Xiao, P.; Chen, W.; Wang, X. Adv. Energy Mater. 2015, 5,
HER. To avoid such pitfalls, people have started looking at the 1500985.
3d transition metal oxides and hydroxides for OER electro- (11) Fabbri, E.; Habereder, A.; Waltar, K.; Kötz, R.; Schmidt, T. J.
Catal. Sci. Technol. 2014, 4, 3800−3821.
catalysis and the chalcogenides of W, Mo, Ta, and Ti for HER (12) Casalongue, H. S.; Kaya, S.; Viswanathan, V.; Miller, D. J.;
electrocatalysis. Very recently, people have found that the Friebel, D.; Hansen, H. A.; Nørskov, J. K.; Nilsson, A.; Ogasawara, H.
sulfides, selenides, and phosphides of 3d transition metals, Nat. Commun. 2013, 4, 2817.
particularly the group VIII metals (Fe, Co, Ni), can catalyze (13) Matsumoto, Y.; Sato, E. Mater. Chem. Phys. 1986, 14, 397−426.
both the HER and OER at all pH with almost the same kinetics. (14) Man, I. C.; Su, H.-Y.; Calle-Vallejo, F.; Hansen, H. A.; Martínez,
The advantages of these 3d-transition-metal-based catalysts are J. I.; Inoglu, N. G.; Kitchin, J.; Jaramillo, T. F.; Nørskov, J. K.;
that they highly abundant, cheap, efficient, and easy to design in Rossmeisl, J. ChemCatChem 2011, 3, 1159−1165.
a desired shape and structure depending on the needs for (15) Ibrahim, H.; Ilinca, A.; Perron, J. Renewable Sustainable Energy
electrolysis. Rev. 2008, 12, 1221−1250.
(16) Gao, M.-R.; Xu, Y.-F.; Jiang, J.; Yu, S.-H. Chem. Soc. Rev. 2013,
Hence, to make use of these advantages, we should be aware
42, 2986−3017.
of the activity trends and mechanisms of the OER and HER on (17) Anantharaj, S.; Karthik, P. E.; Subramanian, B.; Kundu, S. ACS
the surfaces of these group VIII 3d metal (Fe, Co, Ni)-based Catal. 2016, 6, 4660−4672.
sulfide, selenide, and phosphide catalysts. In this review, we (18) Zolfaghari, A.; Jerkiewicz, G.; Chrzanowski, W.; Wieckowski, A.
have made such a comparative measurement of group VIII 3d J. Electrochem. Soc. 1999, 146, 4158−4165.
metal (Fe, Co, Ni)-based sulfide, selenide, and phosphide (19) Zeng, Z.; Tan, C.; Huang, X.; Bao, S.; Zhang, H. Energy Environ.
catalysts to enable those who are working on increasing the Sci. 2014, 7, 797−803.
efficiencies of these catalysts for H2 production by water (20) Wang, J. X.; Zhang, Y.; Capuano, C. B.; Ayers, K. E. Sci. Rep.
electrolysis with little loss in electrical energy via designing 2015, 5, 12220−12227.
(21) Wang, C.; Dekrafft, K. E.; Lin, W. J. Am. Chem. Soc. 2012, 134
catalysts with low overpotentials for both the HER and OER.
(17), 7211−7214.
This comparative review enables researchers to acquire (22) Sun, L.; Ca, D. V.; Cox, J. A. J. Solid State Electrochem. 2005, 9,
knowledge of the trends in the activities of these catalysts 816−822.
toward the HER and OER and helps them to formulate new (23) Sun, C. L.; Chen, L. C.; Su, M. C.; Hong, L. S.; Chyan, O.; Hsu,
highly efficient water splitting catalysts for the sake of our C. Y.; Chen, K. H.; Chang, T. F.; Chang, L. Chem. Mater. 2005, 17,
global future energy requirements. 3749−3753.


(24) Song, H. M.; Anjum, D. H.; Sougrat, R.; Hedhili, M. N.;
Khashab, N. M. J. Mater. Chem. 2012, 22, 25003−25010.
AUTHOR INFORMATION (25) Roy, N.; Leung, K. T.; Pradhan, D. J. Phys. Chem. C 2015, 119,
Corresponding Author 19117−19125.
*E-mail: skundu@cecri.res.in and subrata_kundu2004@yahoo. (26) Alves, V. A.; Da Silva, L. A.; Boodts, J. F. C.; Trasatti, S.
co.in. Fax: +91 4565-227651. Tel: +91 4565-241487. Electrochim. Acta 1994, 39, 1585−1589.
(27) Anantharaj, S.; Jayachandran, M.; Kundu, S. Chem. Sci. 2016, 7,
Notes 3188−3205.
The authors declare no competing financial interest. (28) Bi, R.; Wu, X.; Cao, F.; Jiang, L.; Guo, Y.; Wan, L. J. Phys. Chem.


C 2010, 114, 2448.
(29) Burke, M. S.; Enman, L. J.; Batchellor, A. S.; Zou, S.; Boettcher,
ACKNOWLEDGMENTS S. W. Chem. Mater. 2015, 27, 7549−7558.
We acknowledge Dr. Vijayamohanan K. Pillai, Director, CSIR- (30) Chakrapani, K.; Sampath, S. Chem. Commun. 2014, 50, 3061−
CECRI, Karaikudi, India, for his continuous support and 3063.
encouragement. S.A., S.R.E., and K.S. acknowledge CSIR, New (31) McCrory, C. C. L.; Jung, S.; Peters, J. C.; Jaramillo, T. F. J. Am.
Chem. Soc. 2013, 135, 16977−16987.
Delhi, and K.K. acknowledges UGC, New Delhi, for the
(32) Chen, L. Y.; Hou, Y.; Kang, J. L.; Hirata, a.; Fujita, T.; Chen, M.
afforded funding through the Senior and Junior Research W. Adv. Energy Mater. 2013, 3, 851−856.
Fellowship (SRF and JRF) schemes. We also acknowledge The (33) Chen, X.; Chen, G.; Yue, P. L. J. Phys. Chem. B 2001, 105,
Royal Society of Chemistry and The American Chemical 4623−4628.
Society for the licenses given to reproduce figures from their (34) Cherevko, S.; Reier, T.; Zeradjanin, A. R.; Pawolek, Z.; Strasser,
publications. P.; Mayrhofer, K. J. J. Electrochem. Commun. 2014, 48, 81−85.

8093 DOI: 10.1021/acscatal.6b02479


ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

(35) Fuentes, R. E.; Farell, J.; Weidner, J. W. Electrochem. Solid-State (70) Masa, J.; Weide, P.; Peeters, D.; Sinev, I.; Xia, W.; Sun, Z.;
Lett. 2011, 14, E5−E7. Somsen, C.; Muhler, M.; Schuhmann, W. Adv. Energy Mater. 2016,
(36) Karan, H. I.; Sasaki, K.; Kuttiyiel, K.; Farberow, C. a; Mavrikakis, DOI: 10.1002/aenm.201600980.
M.; Adzic, R. R. ACS Catal. 2012, 2, 817−824. (71) Castro, E. B.; Gervasi, C. a; Vilche, J. R. J. Appl. Electrochem.
(37) Kötz, R. J. Electrochem. Soc. 1983, 130, 825−829. 1998, 28, 835−841.
(38) Kötz, R.; Stucki, S. Electrochim. Acta 1986, 31, 1311−1316. (72) Wang, L. P.; Van Voorhis, T. J. Phys. Chem. Lett. 2011, 2, 2200−
(39) Kwon, S. J.; Fan, F.-R. F.; Bard, A. J. J. Am. Chem. Soc. 2010, 2204.
132, 13165−13167. (73) Farrow, C. L.; Bediako, D. K.; Surendranath, Y.; Nocera, D. G.;
(40) Lee, Y.; Ye, B.; Yu, H. k.; Lee, J.; Kim, M. H.; Baik, J. M. J. Phys. Billinge, S. J. L. J. Am. Chem. Soc. 2013, 135, 6403−6406.
Chem. C 2011, 115, 4611−4615. (74) Jiang, N.; You, B.; Sheng, M.; Sun, Y. Angew. Chem., Int. Ed.
(41) Lee, Y.; Suntivich, J.; May, K. J.; Perry, E. E.; Shao-Horn, Y. J. 2015, 54, 6251−6254.
Phys. Chem. Lett. 2012, 3, 399−404. (75) Rasiyah, P.; Tseung, A. C. C. J. Electrochem. Soc. 1983, 130,
(42) Lewerenz, H. J.; Stucki, S.; Kötz, R. Surf. Sci. 1983, 126, 463− 365−368.
468. (76) Wu, G.; Li, N.; Zhou, D.-R. R.; Mitsuo, K.; Xu, B.-Q. Q. J. Solid
(43) Marshall, A. T.; Haverkamp, R. G. Electrochim. Acta 2010, 55, State Chem. 2004, 177, 3682−3692.
1978−1984. (77) Sadiek, I. M.; Mohammad, A. M.; El-Shakre, M. E.; El-Deab, M.
(44) Jiao, F.; Frei, H. Energy Environ. Sci. 2010, 3, 1018−1027. S. Int. J. Hydrogen Energy 2012, 37, 68−77.
(45) Ponce, J.; Rehspringer, J.; Poillerat, G.; Gautier, J. L. Electrochim. (78) Godwin, I. J.; Lyons, M. E. G. Electrochem. Commun. 2013, 32,
39−42.
Acta 2001, 46, 3373−3380.
(79) Córdoba, S. I.; Carbonio, R. E.; Teijelo, M. L.; Macagno, V. A.
(46) Bockris, J. O.; Otagawa, T. J. Phys. Chem. 1983, 87, 2960−2971.
Electrochim. Acta 1987, 32, 749−755.
(47) Masa, J.; Xia, W.; Sinev, I.; Zhao, A.; Sun, Z.; Grützke, S.; Weide,
(80) Singh, A.; Chang, S. L. Y.; Hocking, R. K.; Bach, U.; Spiccia, L.
P.; Muhler, M.; Schuhmann, W. Angew. Chem., Int. Ed. 2014, 53,
Energy Environ. Sci. 2013, 6, 579−586.
8508−8512. (81) Skowroński, J. M.; Ważny, A. J. New Mater. Electrochem. Syst.
(48) Chen, M.; Wu, Y.; Han, Y.; Lin, X.; Sun, J.; Zhang, W.; Cao, R. 2006, 9, 345−351.
ACS Appl. Mater. Interfaces 2015, 7 (39), 21852−21859. (82) Freitas, M. B. J. G. J. Power Sources 2001, 93, 163−173.
(49) Smith, R. D. L.; Prevot, M. S.; Fagan, R. D.; Trudel, S.; (83) Chen, W. F.; Sasaki, K.; Ma, C.; Frenkel, A. I.; Marinkovic, N.;
Berlinguette, C. P. J. Am. Chem. Soc. 2013, 135, 11580−11586. Muckerman, J. T.; Zhu, Y.; Adzic, R. R. Angew. Chem., Int. Ed. 2012,
(50) Trotochaud, L.; Young, S. L.; Ranney, J. K.; Boettcher, S. W. J. 51, 6131−6135.
Am. Chem. Soc. 2014, 136, 6744−6753. (84) Yang, Y.; Ren, Y.; Sun, C.; Hao, S. Green Chem. 2014, 16, 2273−
(51) Fillol, J. L.; Codola, Z.; Garcia-bosch, I.; Gomez, L.; Pla, J. J.; 2280.
Costas, M. Nat. Chem. 2011, 3, 807−813. (85) Bediako, D. K.; Lassalle-kaiser, B.; Surendranath, Y.; Yano, J.;
(52) Wade, W. H.; Hackerman, N. Trans. Faraday Soc. 1957, 53, Yachandra, V. K.; Nocera, D. G. J. Am. Chem. Soc. 2012, 134, 6801−
1636−1647. 6809.
(53) Corrigan, D. A. J. Electrochem. Soc. 1987, 134, 377−384. (86) Li, X.; Walsh, F. C.; Pletcher, D. Phys. Chem. Chem. Phys. 2011,
(54) Hutchings, G. S.; Zhang, Y.; Li, J.; Yonemoto, B. T.; Zhou, X.; 13, 1162−1167.
Zhu, K.; Jiao, F. J. Am. Chem. Soc. 2015, 137, 4223−4229. (87) Gennero de Chialvo, M. R.; Chialvo, A. C. Electrochim. Acta
(55) Jiao, F.; Frei, H. Angew. Chem., Int. Ed. 2009, 48, 1841−1844. 1988, 33, 825−830.
(56) Bocca, C.; Barbucci, A.; Delucchi, M.; Cerisola, G. Int. J. (88) Kamnev, A. A.; Ezhov, B. B. Electrochim. Acta 1992, 37, 607−
Hydrogen Energy 1999, 24, 21−26. 613.
(57) Lu, X. Y.; Zhao, C. J. Mater. Chem. A 2013, 1, 12053−12059. (89) Rashkova, V.; Kitova, S.; Konstantinov, I.; Vitanov, T.
(58) Mcenaney, J. M.; Soucy, T. L.; Hodges, J. M.; Callejas, J. F.; Electrochim. Acta 2002, 47, 1555−1560.
Mondschein, J. S.; Schaak, R. E. J. Mater. Chem. A 2016, 4, 3077− (90) Kibria, M. Int. J. Hydrogen Energy 1996, 21, 179−182.
3081. (91) Dinca, M.; Surendranath, Y.; Nocera, D. G. Proc. Natl. Acad. Sci.
(59) Jin, H.; Wang, J.; Su, D.; Wei, Z.; Pang, Z.; Wang, Y. J. Am. U. S. A. 2010, 107, 10337−10341.
Chem. Soc. 2015, 137, 2688−2694. (92) Fominykh, K.; Feckl, J. M.; Sicklinger, J.; Döblinger, M.;
(60) Yeo, B. S.; Bell, A. T. J. Am. Chem. Soc. 2011, 133, 5587−5593. Böcklein, S.; Ziegler, J.; Peter, L.; Rathousky, J.; Scheidt, E.; Bein, T.;
(61) Kanan, M. W.; Surendranath, Y.; Nocera, D. G. Chem. Soc. Rev. Fattakhova-Rohlfing, D. Adv. Funct. Mater. 2014, 24, 3123−3129.
2009, 38, 109−114. (93) Ghosh, P. K.; Mau, A. W.-H.; Bard, A. J. J. Electroanal. Chem.
(62) Zhong, D. K.; Gamelin, D. R. J. Am. Chem. Soc. 2010, 132, Interfacial Electrochem. 1984, 169, 315−317.
4202−4207. (94) Hu, C. C.; Chen, J. C.; Chang, K. H. J. Power Sources 2013, 221,
(63) Han, X.; Zhang, Z.; Zhang, T.; Li, Y.; Lin, W.; You, W.; Su, Z.- 128−133.
M.; Wang, E.-B. J. Am. Chem. Soc. 2014, 136, 5359−5366. (95) Cibrev, D.; Jankulovska, M.; Lana-Villarreal, T.; Gómez, R. Int. J.
(64) Gardner, G. P.; Go, Y. B.; Robinson, D. M.; Smith, P. F.; Hydrogen Energy 2013, 38, 2746−2753.
(96) Kamath, P. V.; Dixit, M.; Indira, L.; Shukla, A. K.; Kumar, V. G.;
Hadermann, J.; Abakumov, A.; Greenblatt, M.; Dismukes, G. C.
Munichandraiah, N. J. Electrochem. Soc. 1994, 141, 2956.
Angew. Chem., Int. Ed. 2012, 51, 1616−1619.
(97) Gong, M.; Dai, H. Nano Res. 2015, 8, 23−39.
(65) Maiyalagan, T.; Jarvis, K. A.; Therese, S.; Ferreira, P. J.;
(98) Louie, M. W.; Bell, A. T. J. Am. Chem. Soc. 2013, 135, 12329−
Manthiram, A. Nat. Commun. 2014, 5, 3949. 12337.
(66) Nikolov, I.; Darkaoui, R.; Zhecheva, E.; Stoyanova, R.; Dimitrov, (99) Singh, R. N.; Koenig, J.-F.; Poillerat, G.; Chartier, P. J.
N.; Vitanov, T. J. Electroanal. Chem. 1997, 429, 157−168. Electrochem. Soc. 1990, 137, 1408−1413.
(67) Sadiek, I. M.; Mohammad, A. M.; El-Shakre, M. E.; Awad, M. I.; (100) Chen, S.; Duan, J.; Jaroniec, M.; Qiao, S. Z. Angew. Chem., Int.
El-Deab, M. S.; El-Anadouli, B. Int. J. Electrochem. Sci. 2012, 7, 3350− Ed. 2013, 52, 13567−13570.
3361. (101) Matsumoto, Y.; Sato, E. Mater. Chem. Phys. 1986, 14, 397−
(68) Zhu, C.; Wen, D.; Leubner, S.; Oschatz, M.; Liu, W.; Holzschuh, 426.
M.; Simon, F.; Kaskel, S.; Eychmüller, A. Chem. Commun. 2015, 51, (102) Vrubel, H.; Hu, X. Angew. Chem., Int. Ed. 2012, 51, 12703−
7851−7854. 12706.
(69) Kundu, S.; Mukadam, M. D.; Yusuf, S. M.; Jayachandran, M. (103) Gao, M.-R.; Chan, M. K. Y.; Sun, Y. Nat. Commun. 2015, 6,
CrystEngComm 2013, 15, 482−497. 7493−7500.

8094 DOI: 10.1021/acscatal.6b02479


ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

(104) Xiao, P.; Sk, M. A.; Thia, L.; Ge, X.; Lim, R. J.; Wang, J.; Lim, (137) Grigoriev, S. A.; Millet, P.; Fateev, V. N. J. Power Sources 2008,
K. H.; Wang, X. Energy Environ. Sci. 2014, 7, 2624−2629. 177, 281−285.
(105) Gao, M.-R.; Liang, J.; Zheng, Y.; Xu, Y.; Jiang, J.; Gao, Q.; Li, (138) Herraiz-Cardona, I.; Ortega, E.; Pérez-Herranz, V. Electrochim.
J.; Yu, S. Nat. Commun. 2015, 6, 5982−5988. Acta 2011, 56, 1308−1315.
(106) McEnaney, J. M.; Crompton, J. C.; Callejas, J. F.; Popczun, E. (139) Jakšić, J.; Vojnović, M.; Krstajić, N. Electrochim. Acta 2000, 45,
J.; Biacchi, A. J.; Lewis, N. S.; Schaak, R. E. Chem. Mater. 2014, 26, 4151−4158.
4826−4831. (140) Kibler, L. A. Electrochim. Acta 2008, 53, 6824−6828.
(107) Chen, X.; Wang, D.; Wang, Z.; Zhou, P.; Wu, Z.; Jiang, F. (141) Rami, A.; Lasia, A. J. Appl. Electrochem. 1992, 22, 376−382.
Chem. Commun. 2014, 50, 11683−11685. (142) Seto, K.; Iannelli, A.; Love, B.; Lipkowski, J. J. Electroanal.
(108) Ma, L.; Ting, L. R. L.; Molinari, V.; Giordano, C.; Yeo, B. S. J. Chem. Interfacial Electrochem. 1987, 226, 351−360.
Mater. Chem. A 2015, 3, 8361−8368. (143) Kibsgaard, J.; Tsai, C.; Chan, K.; Benck, J. D.; Nørskov, J. K.;
(109) Chang, K.; Chen, W. ACS Nano 2011, 5, 4720−4728. Abild-Pedersen, F.; Jaramillo, T. F. Energy Environ. Sci. 2015, 8, 3022−
(110) Li, Y.; Liu, J.; Wang, Y.; Wang, Z. L. Chem. Mater. 2001, 13, 3029.
1008−1014. (144) Tang, C.; Gan, L.; Zhang, R.; Lu, W.; Jiang, X.; Asiri, A. M.;
(111) Jin, Z.; Li, P.; Huang, X.; Zeng, G.; Jin, Y.; Zheng, B.; Xiao, D. Sun, X.; Wang, J.; Chen, L. Nano Lett. 2016, 16, 6617.
J. Mater. Chem. A 2014, 2, 18593−18599. (145) Rossmeisl, J.; Qu, Z. W.; Zhu, H.; Kroes, G. J.; Nørskov, J. K. J.
(112) Huang, G.; Liu, H.; Wang, S.; Yang, X.; Liu, B.; Chen, H.; Xu, Electroanal. Chem. 2007, 607, 83−89.
M. J. Mater. Chem. A 2015, 3, 24128−24138. (146) Callejas, J. F.; Read, C. G.; Popczun, E. J.; McEnaney, J. M.;
(113) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Schaak, R. E. Chem. Mater. 2015, 27, 3769−3774.
Fundamentals and Applications; Wiley: New York, 2001. (147) Pan, Y.; Liu, Y.; Zhao, J.; Yang, K.; Liang, J.; Liu, D.; Hu, W.;
(114) Bockris, J. O.; Reddy, A. K. N.; Gamboa-Aldeco, M. E. Modern Liu, D.; Liu, Y.; Liu, C. J. Mater. Chem. A 2015, 3, 1656−1665.
Electrochemistry; Plenum Press: New York, 1998. (148) Xiao, P.; Sk, M. A.; Thia, L.; Ge, X.; Lim, R. J.; Wang, J.; Lim,
(115) Zhu, Y.-P.; Xu, X.; Su, H.; Liu, Y.-P.; Chen, T.; Yuan, Z.-Y. ACS K. H.; Wang, X. Energy Environ. Sci. 2014, 7, 2624−2629.
Appl. Mater. Interfaces 2015, 7, 28369−28376. (149) Wired Chemist. Common Bond Energies. http://www.
(116) Bockris, J. O.; Otagawa, T. J. Phys. Chem. 1983, 87, 2960− wiredchemist.com/chemistry/data/bond_energies_lengths.html (ac-
2971. cessed Aug 24, 2016).
(117) Xiao, M.; Miao, Y.; Tian, Y.; Yan, Y. Electrochim. Acta 2015, (150) Subbaraman, R.; Tripkovic, D.; Chang, K.-C.; Strmcnik, D.;
165, 206−210. Paulikas, A. P.; Hirunsit, P.; Chan, M.; Greeley, J.; Stamenkovic, V.;
(118) Gorlin, Y.; Jaramillo, T. F. J. Am. Chem. Soc. 2010, 132, 13612− Markovic, N. M. Nat. Mater. 2012, 11, 550−557.
13614. (151) Bastos, A. C.; Ferreira, M. G.; Simoes, A. M. Corros. Sci. 2006,
(119) Kibsgaard, J.; Jaramillo, T. F.; Besenbacher, F. Nat. Chem. 48, 1500−1512.
(152) Edwards, M.; Hidmi, L.; Gladwell, D. Corros. Sci. 2002, 44,
2014, 6, 248−253.
(120) Reier, T.; Oezaslan, M.; Strasser, P. ACS Catal. 2012, 2, 1765− 1057−1071.
(153) De Lima-Neto, P.; De Araujo, A. P.; Araujo, W. S.; Correia, A.
1772.
N. Prog. Org. Coat. 2008, 62, 344−350.
(121) Hou, Y.; Lohe, M. R.; Zhang, J.; Liu, S.; Zhuang, X.; Feng, X.
(154) Zin, I. M.; Lyon, S. B.; Pokhmurskii, V. I. Corros. Sci. 2003, 45,
Energy Environ. Sci. 2016, 9, 478−483.
777−778.
(122) Lu, Z.; Xu, W.; Zhu, W.; Yang, Q.; Lei, X.; Liu, J.; Li, Y.; Sun,
(155) Raja, P. B.; Sethuraman, M. G. Mater. Lett. 2008, 62, 1602−
X.; Duan, X. Chem. Commun. 2014, 50, 6479−6482.
1604.
(123) Liang, H.; Meng, F.; Caban-Acevedo, M.; Li, L.; Forticaux, A.;
(156) Hong, Y.; Roy, D.; Babu, S. V. Electrochem. Solid-State Lett.
Xiu, L.; Wang, Z.; Jin, S. Nano Lett. 2015, 15, 1421−1427. 2005, 8, G297−G300.
(124) Gong, M.; Li, Y.; Wang, H.; Liang, Y.; Wu, J. Z.; Zhou, J.; (157) Desai, M. N.; Desai, S. M. Corros. Sci. 1970, 10, 233−237.
Wang, J.; Regier, T.; Wei, F.; Dai, H. J. Am. Chem. Soc. 2013, 135, (158) Turcotte, D. E.; Lockwood, F. E. Aqueous Corrosion
8452−8455. Inhibitors. U.S. Patent 09,422,596, May 8, 2001.
(125) Vrubel, H.; Moehl, T.; Grätzel, M.; Hu, X. Chem. Commun. (159) Lamprey, H. Noncorrosive Antifreeze Liquid. U.S. Patent
2013, 49, 8985−8987. 2147409 A, Feb 14, 1939.
(126) Stoerzinger, K. A.; Qiao, L.; Biegalski, M. D.; Shao-horn, Y. J. (160) Lewis, A. E. Hydrometallurgy 2010, 104, 222−234.
Phys. Chem. Lett. 2014, 5, 1636−1641. (161) Joo, J.; Na, H. B.; Yu, T.; Yu, J. H.; Kim, Y. W.; Wu, F.; Zhang,
(127) Hu, W.; Wang, Y.; Hu, X.; Zhou, Y.; Chen, S. J. Mater. Chem. J. Z.; Hyeon, T. J. Am. Chem. Soc. 2003, 125, 11100−11105.
2012, 22, 6010−6016. (162) Di Giovanni, C.; Wang, W. A.; Nowak, S.; Greneche, J. M.;
(128) Xu, J.; Liu, G.; Li, J.; Wang, X. Electrochim. Acta 2012, 59, Lecoq, H.; Mouton, L.; Giraud, M.; Tard, C. ACS Catal. 2014, 4, 681−
105−112. 687.
(129) Murray, C. B.; Sun, S.; Doyle, H.; Betley, T. MRS Bull. 2001, (163) Shen, M.; Ruan, C.; Chen, Y.; Jiang, C.; Ai, K.; Lu, L. ACS
26, 985−991. Appl. Mater. Interfaces 2015, 7, 1207−1218.
(130) Zhang, S.; Gai, S.; He, F.; Ding, S.; Li, L.; Yang, P. Nanoscale (164) Faber, M. S.; Lukowski, M. A.; Ding, Q.; Kaiser, N. S.; Jin, S. J.
2014, 6, 11181−11188. Phys. Chem. C 2014, 118, 21347−21356.
(131) Roy, A.; Srinivas, V.; Ram, S.; De Toro, J. A.; Mizutani, U. Phys. (165) You, B.; Jiang, N.; Sheng, M.; Sun, Y. Chem. Commun. 2015,
Rev. B: Condens. Matter Mater. Phys. 2005, 71, 184443−10. 51, 4252−4255.
(132) Lyons, M. E. G.; Brandon, M. P. J. Electroanal. Chem. 2009, (166) Kornienko, N.; Resasco, J.; Becknell, N.; Jiang, C. M.; Liu, Y.
631, 62−70. S.; Nie, K.; Sun, X.; Guo, J.; Leone, S. R.; Yang, P. J. Am. Chem. Soc.
(133) Kumar, T. N.; Sivabalan, S.; Chandrasekaran, N.; Phani, K. L. 2015, 137, 7448−7455.
Chem. Commun. 2015, 51, 1922−1925. (167) Aslan, E.; Akin, I.; Patir, I. H. Chem. - Eur. J. 2016, 22, 5342−
(134) Guo, S.; Liu, Y.; Bond, A. M.; Zhang, J.; Karthik, P. E.; 5349.
Maheshwaran, I.; Kumar, S. S.; Phani, K. L. N. Phys. Chem. Chem. Phys. (168) Wang, J.; Zhong, H. X.; Wang, Z. L.; Meng, F. L.; Zhang, X. B.
2014, 16, 19035−19045. ACS Nano 2016, 10, 2342−2348.
(135) Karthik, E. P.; Raja, A. K.; Kumar, S. S.; Phani, K. L. N.; Liu, Y.; (169) Faber, M. S.; Dziedzic, R.; Lukowski, M. A.; Kaiser, N. S.; Ding,
Guo, S.-X.; Zhang, J.; Bond, A. M. RSC Adv. 2015, 5, 3196−3200. Q.; Jin, S. J. Am. Chem. Soc. 2014, 136, 10053−10061.
(136) Gao, M.; Sheng, W.; Zhuang, Z.; Fang, Q.; Gu, S.; Jiang, J.; (170) Zhang, H.; Li, Y.; Zhang, G.; Wan, P.; Xu, T.; Wu, X.; Sun, X.
Yan, Y. J. Am. Chem. Soc. 2014, 136, 7077−7084. Electrochim. Acta 2014, 148, 170−174.

8095 DOI: 10.1021/acscatal.6b02479


ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

(171) Peng, S.; Li, N.; Han, X.; Sun, W.; Srinivasan, M.; Mhaisalkar, (203) Ullah, K.; Lei, Z.; Ye, S.; Ali, A.; Oh, W. C. RSC Adv. 2015, 5,
S. G.; Cheng, F.; Yan, Q.; Chen, J.; Ramakrishna, S. Angew. Chem., Int. 18841−18849.
Ed. 2014, 53, 12594−12599. (204) Zheng, Y. R.; Gao, M. R.; Gao, Q.; Li, H. H.; Xu, J.; Wu, Z. Y.;
(172) Ganesan, P.; Prabu, M.; Sanetuntikul, J.; Shanmugam, S. ACS Yu, S. H. Small 2015, 11, 182−188.
Catal. 2015, 5, 3625−3637. (205) Liu, Q.; Shi, J.; Hu, J.; Asiri, A. M.; Luo, Y.; Sun, X. ACS Appl.
(173) Wang, H.; Li, Z.; Li, G.; Peng, F.; Yu, H. Catal. Today 2015, Mater. Interfaces 2015, 7, 3877−3881.
245, 74−78. (206) Park, J.; Koo, B.; Yoon, K. Y.; Hwang, Y.; Kang, M.; Park, J. G.;
(174) Feng, L. L.; Li, G. D.; Liu, Y.; Wu, Y.; Chen, H.; Wang, Y.; Hyeon, T. J. Am. Chem. Soc. 2005, 127, 8433−8440.
Zou, Y. C.; Wang, D.; Zou, X. ACS Appl. Mater. Interfaces 2015, 7, (207) Henkes, A. E.; Vasquez, Y.; Schaak, R. E. J. Am. Chem. Soc.
980−988. 2007, 129, 1896−1897.
(175) Liang, Y.; Liu, Q.; Luo, Y.; Sun, X.; He, Y.; Asiri, A. M. (208) Yan, Y.; Xia, B. Y.; Ge, X.; Liu, Z.; Fisher, A.; Wang, X. Chem. -
Electrochim. Acta 2016, 190, 360. Eur. J. 2015, 21, 18062−18067.
(176) Zhou, W.; Wu, X.-J.; Cao, X.; Huang, X.; Tan, C.; Tian, J.; Liu, (209) Xiong, D.; Wang, X.; Li, W.; Liu, L. Chem. Commun. 2016, 52,
H.; Wang, J.; Zhang, H. Energy Environ. Sci. 2013, 6, 2921. 8711−8714.
(177) Tang, C.; Pu, Z.; Liu, Q.; Asiri, A. M.; Luo, Y.; Sun, X. Int. J. (210) Xu, Y.; Wu, R.; Zhang, J.; Shi, Y.; Zhang, B. Chem. Commun.
Hydrogen Energy 2015, 40, 4727−4732. 2013, 49, 6656−6658.
(178) Ouyang, C.; Wang, X.; Wang, C.; Zhang, X.; Wu, J.; Ma, Z.; (211) Du, H.; Gu, S.; Liu, R.; Li, C. M. Int. J. Hydrogen Energy 2015,
Dou, S.; Wang, S. Electrochim. Acta 2015, 174, 297−301. 40, 14272−14278.
(179) Tang, C.; Pu, Z.; Liu, Q.; Asiri, A. M.; Sun, X. Electrochim. Acta (212) Liang, Y.; Liu, Q.; Asiri, A. M.; Sun, X.; Luo, Y. ACS Catal.
2015, 153, 508−514. 2014, 4, 4065−4069.
(180) Yang, N.; Tang, C.; Wang, K.; Du, G.; Asiri, A. M.; Sun, X. (213) Tian, J.; Liu, Q.; Liang, Y.; Xing, Z.; Asiri, A. M.; Sun, X. ACS
Nano Res. 2016, 9, 3346. Appl. Mater. Interfaces 2014, 6, 20579−20584.
(181) Zhu, W.; Yue, X.; Zhang, W.; Yu, S.; Zhang, Y.; Wang, J.; (214) Yang, X.; Lu, A.-Y.; Zhu, Y.; Min, S.; Hedhili, M. N.; Han, Y.;
Wang, J. Chem. Commun. 2016, 52, 1486−1489. Huang, K.-W.; Li, L.-J. Nanoscale 2015, 7, 10974−10981.
(182) Li, H.; Shao, Y.; Su, Y.; Gao, Y.; Wang, X. Chem. Mater. 2016, (215) Jiang, P.; Liu, Q.; Liang, Y.; Tian, J.; Asiri, A. M.; Sun, X.
28, 1155−1164. Angew. Chem., Int. Ed. 2014, 53, 12855−12859.
(183) Mabayoje, O.; Shoola, A.; Wygant, B. R.; Mullins, C. B. ACS (216) Liu, Q.; Pu, Z.; Asiri, A. M.; Sun, X. Electrochim. Acta 2014,
Energy Lett. 2016, 1, 195−201. 149, 324−329.
(184) Liu, Q.; Jin, J.; Zhang, J. ACS Appl. Mater. Interfaces 2013, 5, (217) Dutta, A.; Samantara, A. K.; Dutta, S. K.; Jena, B. K.; Pradhan,
5002−5008. N. ACS Energy Lett. 2016, 1, 169−174.
(185) Liu, T.; Sun, X.; Asiri, A. M.; He, Y. Int. J. Hydrogen Energy (218) Tian, J.; Liu, Q.; Asiri, A. M.; Sun, X. J. Am. Chem. Soc. 2104,
2016, 41, 7264−7269. 136, 7587−7590.
(186) Fang, W.; Liu, D.; Lu, Q.; Sun, X.; Asiri, A. M. Electrochem. (219) Tian, J.; Cheng, N.; Liu, Q.; Xing, W.; Sun, X. Angew. Chem.,
Commun. 2016, 63, 60−64. Int. Ed. 2015, 54, 5493−5497.
(187) Liu, D.; Lu, Q.; Luo, Y.; Sun, X.; Asiri, A. M. Nanoscale 2015, 7, (220) Du, H.; Liu, Q.; Cheng, N.; Asiri, A. M.; Sun, X.; Li, C. M. J.
15122−15126. Mater. Chem. A 2014, 2, 14812−14816.
(188) Ansovini, D.; Lee, C. J. J.; Chua, C. S.; Ong, J.; Tan, H. R.; (221) Yang, L.; Qi, H.; Zhang, C.; Sun, X. Nanotechnology 2016, 27,
Webb, W.; Raja, R.; Lim, Y.-F. J. Mater. Chem. A 2016, 4, 9744−9749. 23LT01−7.
(189) Wang, D. Y.; Gong, M.; Chou, H. L.; Pan, C. J.; Chen, H. A.; (222) Zhou, D.; He, L.; Zhu, W.; Hou, X.; Wang, K.; Du, G.; Zheng,
Wu, Y.; Lin, M. C.; Guan, M.; Yang, J.; Chen, C. W.; Wang, Y. L.; C.; Sun, X.; Asiri, A. M. J. Mater. Chem. A 2016, 4, 10114−10117.
Hwang, B. J.; Chen, C. C.; Dai, H. J. Am. Chem. Soc. 2015, 137, 1587− (223) Popczun, E. J.; Read, C. G.; Roske, C. W.; Lewis, N. S.; Schaak,
1592. R. E. Angew. Chem., Int. Ed. 2014, 53, 5427−5430.
(190) Long, X.; Li, G.; Wang, Z.; Zhu, H.; Zhang, T.; Xiao, S.; Guo, (224) Popczun, E. J.; Roske, C. W.; Read, C. G.; Crompton, J. C.;
W.; Yang, S. J. Am. Chem. Soc. 2015, 137, 11900−11903. McEnaney, J. M.; Callejas, J. F.; Lewis, N. S.; Schaak, R. E. J. Mater.
(191) Miao, J.; Xiao, F.-X.; Yang, H. B.; Khoo, S. Y.; Chen, J.; Fan, Z.; Chem. A 2015, 3, 5420−5425.
Hsu, Y.-Y.; Chen, H. M.; Zhang, H.; Liu, B. Sci. Adv. 2015, 1, (225) Liu, M.; Li, J. ACS Appl. Mater. Interfaces 2016, 8, 2158−2165.
e1500259−e1500259. (226) Yang, H.; Zhang, Y.; Hu, F.; Wang, Q. Nano Lett. 2015, 15,
(192) Zhu, J.; Koltypin, Y.; Gedanken, A. Chem. Mater. 2000, 12, 7616−7620.
73−78. (227) Niu, Z.; Jiang, J.; Ai, L. Electrochem. Commun. 2015, 56, 56−60.
(193) Tang, C.; Cheng, N.; Pu, Z.; Xing, W.; Sun, X. Angew. Chem., (228) Pu, Z.; Liu, Q.; Jiang, P.; Asiri, A. M.; Obaid, A. Y.; Sun, X.
Int. Ed. 2015, 54, 9351−9355. Chem. Mater. 2014, 26, 4326−4329.
(194) Wang, Z.; Li, J.; Tian, X.; Wang, X.; Yu, Y.; Owusu, K. A.; He, (229) Zhu, Y. P.; Liu, Y. P.; Ren, T. Z.; Yuan, Z. Y. Adv. Funct. Mater.
L.; Mai, L. ACS Appl. Mater. Interfaces 2016, 8, 19386−19392. 2015, 25, 7337−7347.
(195) Shi, J.; Hu, J.; Luo, Y.; Sun, X.; Asiri, A. M. Catal. Sci. Technol. (230) Chang, J.; Xiao, Y.; Xiao, M.; Ge, J.; Liu, C.; Xing, W. ACS
2015, 5, 4954−4958. Catal. 2015, 5, 6874−6878.
(196) Pu, Z.; Luo, Y.; Asiri, A. M.; Sun, X. ACS Appl. Mater. Interfaces (231) Huang, Z.; Chen, Z.; Chen, Z.; Lv, C.; Humphrey, M. G.;
2016, 8, 4718−4723. Zhang, C. Nano Energy 2014, 9, 373−382.
(197) Xu, Y. F.; Gao, M. R.; Zheng, Y. R.; Jiang, J.; Yu, S. H. Angew. (232) Gu, S.; Du, H.; Asiri, A. M.; Sun, X.; Li, C. M. Phys. Chem.
Chem., Int. Ed. 2013, 52, 8546−8550. Chem. Phys. 2014, 16, 16909−16913.
(198) Kong, D.; Wang, H.; Lu, Z.; Cui, Y. J. Am. Chem. Soc. 2014, (233) Jiang, P.; Liu, Q.; Ge, C.; Cui, W.; Pu, Z.; Asiri, A. M.; Sun, X.
136, 4897−4900. J. Mater. Chem. A 2014, 2, 14634−14640.
(199) Liu, T.; Liu, Q.; Asiri, A. M.; Luo, Y.; Sun, X. Chem. Commun. (234) Ryu, J.; Jung, N.; Jang, J. H.; Kim, H. J.; Yoo, S. J. ACS Catal.
2015, 51, 16683−16686. 2015, 5, 4066−4074.
(200) Carim, A. I.; Saadi, F. H.; Soriaga, M. P.; Lewis, N. S. J. Mater. (235) Liu, Q.; Tian, J.; Cui, W.; Jiang, P.; Cheng, N.; Asiri, A. M.;
Chem. A 2014, 2, 13835−13839. Sun, X. Angew. Chem., Int. Ed. 2014, 53, 6710−6714.
(201) Zhang, H.; Yang, B.; Wu, X.; Li, Z.; Lei, L.; Zhang, X. ACS (236) Ma, L.; Shen, X.; Zhou, H.; Zhu, G.; Ji, Z.; Chen, K. J. Mater.
Appl. Mater. Interfaces 2015, 7, 1772−1779. Chem. A 2015, 3, 5337−5343.
(202) Liao, M.; Zeng, G.; Luo, T.; Jin, Z.; Wang, Y.; Kou, X.; Xiao, D. (237) Hou, C. C.; Cao, S.; Fu, W. F.; Chen, Y. ACS Appl. Mater.
Electrochim. Acta 2016, 194, 59−66. Interfaces 2015, 7, 28412−28419.

8096 DOI: 10.1021/acscatal.6b02479


ACS Catal. 2016, 6, 8069−8097
ACS Catalysis Review

(238) Wang, C.; Jiang, J.; Zhou, X.; Wang, W.; Zuo, J.; Yang, Q. J.
Power Sources 2015, 286, 464−469.
(239) Li, Q.; Xing, Z.; Asiri, A. M.; Jiang, P.; Sun, X. Int. J. Hydrogen
Energy 2014, 39, 16806−16811.
(240) Kucernak, A. R. J.; Sundaram, V. N. N. J. Mater. Chem. A 2014,
2, 17435−17445.
(241) Li, Z.; Dou, X.; Zhao, Y.; Wu, C. Inorg. Chem. Front. 2016, 3,
1021−1027.
(242) You, B.; Jiang, N.; Sheng, M.; Bhushan, M. W.; Sun, Y. ACS
Catal. 2016, 6, 714−721.
(243) Li, J.; Li, J.; Zhou, X.; Xia, Z.; Gao, W.; Ma, Y.; Qu, Y. ACS
Appl. Mater. Interfaces 2016, 8, 10826−10834.
(244) Shi, Y.; Xu, Y.; Zhuo, S.; Zhang, J.; Zhang, B. ACS Appl. Mater.
Interfaces 2015, 7, 2376−2384.
(245) Han, A.; Jin, S.; Chen, H.; Ji, H.; Sun, Z.; Du, P. J. Mater. Chem.
A 2015, 3, 1941−1946.
(246) Wang, X.; Kolen’Ko, Y. V.; Bao, X. Q.; Kovnir, K.; Liu, L.
Angew. Chem., Int. Ed. 2015, 54, 8188−8192.
(247) Yu, X.; Feng, Y.; Guan, B.; Lou, X. W. D.; Paik, U. Energy
Environ. Sci. 2016, 9, 1246−1250.
(248) Wang, X.; Kolen’ko, Y. V.; Liu, L. Chem. Commun. 2015, 51,
6738−6741.
(249) Zhuo, J.; Cabán-Acevedo, M.; Liang, H.; Samad, L.; Ding, Q.;
Fu, Y.; Li, M.; Jin, S. ACS Catal. 2015, 5, 6355−6361.
(250) Jiang, P.; Liu, Q.; Sun, X. Nanoscale 2014, 6, 13440−13445.
(251) Laursen, A. B.; Patraju, K. R.; Whitaker, M. J.; Retuerto, M.;
Sarkar, T.; Yao, N.; Ramanujachary, K. V.; Greenblatt, M.; Dismukes,
G. C. Nanocrystalline Ni5P4. Energy Environ. Sci. 2015, 8, 1027−1034.
(252) Ledendecker, M.; Krick Calderón, S.; Papp, C.; Steinrück, H.
P.; Antonietti, M.; Shalom, M. Angew. Chem., Int. Ed. 2015, 54,
12361−12365.
(253) Huang, Z.; Chen, Z. Z.; Chen, Z. Z.; Lv, C.; Meng, H.; Zhang,
C. ACS Nano 2014, 8, 8121−8129.
(254) Feng, Y.; Yu, X. Y.; Paik, U. Chem. Commun. 2016, 52, 1633−
1636.
(255) Li, Y.; Zhang, H.; Jiang, M.; Kuang, Y.; Sun, X.; Duan, X. Nano
Res. 2016, 9, 2251−2259.
(256) Tian, J.; Cheng, N.; Liu, Q.; Sun, X.; He, Y.; Asiri, A. M. J.
Mater. Chem. A 2015, 3, 20056−20059.
(257) Liang, Y.; Sun, X.; Asiri, A. M.; He, Y. Nanotechnology 2016, 27,
12LT01−8.
(258) Wang, J.; Cui, W.; Liu, Q.; Xing, Z.; Asiri, A. M.; Sun, X. Adv.
Mater. 2016, 28, 215−230.
(259) Anantharaj, S.; Karthik, P. E.; Kundu, S. J. Mater. Chem. A
2015, 3, 24463−24478.
(260) Liu, W.; Hu, E.; Jiang, H.; Xiang, Y.; Weng, Z.; Li, M.; Fan, Q.;
Yu, X.; Altman, E. I.; Wang, H. Nat. Commun. 2016, 7, 10771−10779.
(261) Liu, T.; Liang, Y.; Liu, Q.; Sun, X.; He, Y.; Asiri, A. M.
Electrochem. Commun. 2015, 60, 92−96.
(262) Liu, T.; Asiri, A. M.; Sun, X. Nanoscale 2016, 8, 3911−3915.
(263) Rabbani, F.; Svengren, H.; Zimmermann, I.; Hu, S. C.; Laine,
T.; Hao, W. M.; Akermark, B.; Akermark, T.; Johnsson, M. Dalton
Trans. 2014, 43, 3984−3989.

8097 DOI: 10.1021/acscatal.6b02479


ACS Catal. 2016, 6, 8069−8097

You might also like