Article

Cite This: Acc. Chem. Res. 2018, 51, 273−281 pubs.acs.org/accounts

Understanding Conversion-Type Electrodes for Lithium

Rechargeable Batteries
Published as part of the Accounts of Chemical Research special issue “Energy Storage: Complexities Among
Materials and Interfaces at Multiple Length Scales”.
Seung-Ho Yu,†,‡ Xinran Feng,† Na Zhang,† Jeesoo Seok,† and Héctor D. Abruña*,†
†
Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853, United States
‡
Cornell High Energy Synchrotron Source, Cornell University, Ithaca, New York 14853, United States
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Downloaded via CHONNAM NATL UNIV on February 1, 2023 at 08:34:36 (UTC).

CONSPECTUS: The need/desire to lower the consumption of fossil fuels and its environmental consequences has reached
unprecedented levels in recent years. A global effort has been undertaken to develop advanced renewable energy generation and
especially energy storage technologies, as they would enable a dramatic increase in the effective and efficient use of renewable
(and often intermittent) energy sources. The development of electrical energy storage (EES) technologies with high energy and
power densities, long life, low cost, and safe use represents a challenge from both the fundamental science and technological
application points of view. While the advent and broad deployment of lithium-ion batteries (LIBs) has dramatically changed the
EES landscape, their performance metrics need to be greatly enhanced to keep pace with the ever-increasing demands imposed
by modern consumer electronics and especially the emerging automotive markets.
Current battery technologies are mostly based on the use of a transition metal oxide cathode (e.g., LiCoO2, LiFePO4, or
LiNiMnCoO2) and a graphite anode, both of which depend on intercalation/insertion of lithium ions for operation. While the
cathode material currently limits the battery capacity and overall energy density, there is a great deal of interest in the develop-
ment of high-capacity cathode materials as well as anode materials.
Conversion reaction materials have been identified/proposed as potentially high-energy-density alternatives to intercalation-
based materials. However, conversion reaction materials react during lithiation to form entirely new products, often with dra-
matically changed structure and chemistry, by reaction mechanisms that are still not completely understood. This makes it
difficult to clearly distinguish the limitations imposed by the mechanism and practical losses from initial particle morphology,
synthetic approaches, and electrode preparations.
Transition metal compounds such as transition metal oxides, sulfides, fluorides, phosphides, and nitrides can undergo conversion
reactions yielding materials with high theoretical capacity (generally from 500 to 1500 mA h g−1). In particular, a number of tran-
sition metal oxides and sulfides have shown excellent electrochemical properties as high-capacity anode materials. In addition,
some transition metal fluorides have shown great potential as cathode materials for Li rechargeable batteries.
In this Account we present mechanistic studies, with emphasis on the use of operando methods, of selected examples of con-
version-type materials as both potentially high-energy-density anodes and cathodes in EES applications. We also include examples of
the conceptually similar conversion-type reactions involving chalcogens and halogens, with emphasis on the Li−S system. In this
case we focus on the problems arising from the low electrical conductivities of elemental sulfur and Li2S and the “redox shuttle”
phenomena of polysulfides.
In addition to mechanistic insights from the use of operando methods, we also cover several key strategies in electrode materials
design such as controlling the size, morphology, composition, and architecture.

Received: September 30, 2017

Published: January 26, 2018

© 2018 American Chemical Society 273 DOI: 10.1021/acs.accounts.7b00487

Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research
1. INTRODUCTION
For the past decades, lithium-ion batteries (LIBs) have been
regarded as the dominant technology for powering portable
devices and consumer electronics.1 Commercial LIBs, based on
intercalation/insertion-type electrodes, are generally quite stable
during long cycles with minimal self-discharge. However, they have
low specific energy, and only limited (incremental/evolutionary)
improvements can be envisioned with the use of intercalation-
type electrode materials. Thus, there is an urgency to improve
battery properties and performance metrics to satisfy the growing
energy demands of consumer electronic devices and electric
vehicles. To achieve this, it is essential to design new electrode
materials and architectures based on an understanding of the
reaction mechanisms of the electrode materials.
Among the candidates to replace current intercalation-type
electrode materials, conversion-type electrode materials are very
promising because of their high theoretical capacity and low cost.
Generally, the conventional/classical concept of a conversion
reaction can be expressed as2,3
TMaX b + (b × c)Li+ + (b × c)e− ⇄ aTM 0 + b LicX
(1)
where TM denotes a transition metal and X an anionic species.
Transition metal compounds such as transition metal oxides, sul-
fides, fluorides, phosphides, and nitrides can undergo such con-
version reactions (reaction 1), yielding high-theoretical-capacity
materials (generally from 500 to 1500 mA h g−1).3 In particular,
various transition metal oxides and sulfides have shown excellent
electrochemical properties as high-capacity anode materials, and
some transition metal fluorides have shown great potential as
promising cathode materials for Li rechargeable batteries. Con-
ceptually similar to, and an extension of, conversion reactions are
reactions of the type
X + a Li+ + ae− ⇄ LiaX (2)
Many chalcogens (e.g., S, Se, and Te) and halogens (e.g., Br and I)
undergo this reaction (reaction 2) to form Li2X and LiX, respec-
tively. Among them, S and Se have been recently considered as
particularly attractive candidates as cathode materials for Li
rechargeable batteries. Sulfur has a very high theoretical gravi-
metric capacity of 1672 mA h g−1 and volumetric capacity of
3467 mA h cm−3. However, there are several problems that limit
its practical use. The low electrical conductivities of elemental
sulfur and Li2S result in low utilization. In addition, polysulfides
can diffuse back and forth between the anode and cathode, being
chemically reduced at the anode and subsequently oxidized at
the cathode. In the Li−S battery literature, this is known as the
so-called “polysulfide shuttle”, which causes severe self-discharge.
Selenium has electrochemical properties similar to those of sul-
fur, with a comparably high theoretical volumetric capacity of
3253 mA h cm−3. Although selenium has a lower theoretical
specific capacity than sulfur (675 mA h g−1), it is still much higher
than those of intercalation-type cathodes. Additionally, Se has a
higher electrical conductivity than sulfur, and polyselenides are
less prone to shuttle effects compared to polysulfides.
In this Account, we discuss representative examples of conven-
tional conversion reactions as well as sulfur cathodes as a second-
type (extended concept) of conversion reaction electrodes,
focusing on reaction mechanisms, especially through operando
experimental methods. We also cover several key strategies in
electrode materials design such as controlling the size, morphology,
composition, and architecture.
274
Article
2. CONVERSION-TYPE ELECTRODES FOR LI-ION
BATTERIES
2.1. Conversion-Type Anodes and Studies of Their Reaction
Mechanism
Conversion-reaction-based anode materials, especially transition
metal oxides, have shown promising properties as replacements
for the currently employed graphite anodes. Our group has
employed operando synchrotron X-ray absorption spectroscopy
(XAS) and X-ray diffraction (XRD) techniques to understand/
unravel the reaction mechanism of Mn3O4 anodes in LIBs.4
Operando measurements have several advantages compared
to ex situ measurements. Since operando measurements are
performed during battery operation in the cell, they can avoid
contamination by air/moisture as well as minimize relaxation
effects. In addition, much larger amounts of sample are needed to
study the whole reaction by ex situ measurements.
The voltage profile of Mn3O4 during the initial cycle is shown
in Figure 1a. The operando XRD patterns for phase I (Figure 1b)
revealed the reduction of Mn ions and the formation of LiMn3O4
(“lithiated” Mn3O4) at the beginning of phase I and the for-
mation of MnO after 0.5 e−/Mn3O4. The decrease in the XAS
edge energy confirmed the reduction of the Mn ions (Figure 1c).
Meanwhile, the intensity of the pre-edge decreased, indicating
that Mn ions were transferred from tetrahedral to octahedral
sites. These trends in the XAS spectra matched well with the
structural changes from LiMn3O4 to MnO. The edge energy con-
tinued to decrease to lower values, corresponding to the forma-
tion of metallic Mn (Figure 1d). The isosbestic points indicate
that Mn ions changed only between MnO and Mn in their local
structure. However, the XAS spectra also indicated that Mn ions
were not involved in the redox reactions in the early stage of
phase II or at the end of discharge (overlapping black and green
lines, respectively, in Figure 1d).
Voltage profiles during the initial delithiation and subsequent
lithiation of Mn3O4 exhibited sloping features, while those during
the initial lithiation showed well-defined plateaus. This is the case
because the domain size of the active materials becomes smaller
with a largely amorphous character after the initial lithiation. The
charge capacity (about 800 mA h g−1) arose from the oxidation of
Mn0 to Mn2+ in MnO as well from non-Mn-centered redox reac-
tions. Combining all of these data, a conversion reaction mecha-
nism for the first cycle was proposed, as shown in Figure 1e. The
changes in the Mn ions follow the sequence Mn3O4 → LiMn3O4 →
MnO → Mn → MnO. The redox reaction between Mn0 and
MnO is reversible and plays a key role regarding the reversibility
of the conversion reaction; in fact, it is responsible for the
reversible charge/discharge processes. Our study not only pro-
vided great insights into the overall mechanism of the Mn3O4
conversion reaction but also emphasized the need to understand
the role of Li-containing species and reactions.
Grey and co-workers reported on the origin of additional
capacity delivered during the voltage slope close to the end of
lithiation.5 In their report, RuO2 was investigated, and lithium-
centered reactions were monitored by ex situ 6Li, 17O, and 1H
solid-state NMR spectroscopy. According to their study, the
initial discharge started with the formation of a lithiated ruthenate
phase and solid−electrolyte interphase (SEI) formation. LiOH
appeared shortly after, supporting the notion that the non-metal-
centered reaction started at the early stages of discharge. The
LiOH later decomposed into Li2O and LiH throughout most of
the discharge, providing additional capacity. They also stated that
the reaction from LiRuO2 to Li2O was another main contributor
DOI: 10.1021/acs.accounts.7b00487
Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research Article

Figure 1. (a) Electrochemical voltage profile for Mn3O4 during the initial cycle. (b) Operando XRD patterns during phase I. (c) Operando X-ray
absorption near edge structure (XANES) spectra of the Mn K-edge during phase I. The inset shows the pre-edge region. Red and black lines were
recorded during the charge intervals 0.0−1.0 and 1.0−2.0 e−/Mn3O4, respectively. (d) XANES spectra of Mn K-edge during phases II and III, colored to
show distinct trends over specified charge intervals. (e) Proposed mechanism of Mn3O4 through conversion reaction. Reproduced with permission from
ref 4. Copyright 2013 Royal Society of Chemistry.

to the extra capacity. While these two systems are structurally and
compositionally distinct, they nevertheless show some common
mechanistic features such as the early generation of a lithiated form
of the starting material and eventual formation of a Li2O matrix.
In situ transmission electron microscopy (TEM) has been
employed to reveal atomistic mechanisms of lithiation into tran-
sition metal oxides.6,7 In many transition metal oxides, such as
Fe3O4 and RuO2, intercalation reactions occur to some extent
before the conversion reaction takes place. He et al.6 observed
the phase evolution of Fe3O4 during intercalation and subse-
quent conversion reactions using in situ TEM. Thermodynami-
cally, these two reactions in Fe3O4 are non-overlapping processes.
However, they found that the initial intercalation reaction followed
a two-phase reaction, initially forming the rock salt LiFe3O4, and
this initial process overlapped with the subsequent conversion
reaction to form Fe + Li2O in a single particle.
2.2. Design of Nanostructures for High-Performance Anodes
Conversion reactions generally yield much higher capacities
compared to insertion reactions. However, they often suffer from
275
problems like low lithium ion mobility, high volume expansion,
and large cell polarization.2 A variety of morphology designs have
been employed to overcome these problems by providing void
space in the structure to accommodate the volumetric expansion
that accompanies lithiation.
Hollow structures are among the most frequently used archi-
tectures for conversion reactions. They are believed to withstand
high volumetric expansion and stress, and their high surface to
volume ratio greatly enhances the Li diffusion rate. However, the
generally low volumetric energy density of these structures is an
issue that remains largely unresolved. In addition, hollow mate-
rials are usually prepared through template-assisted methods, but
the low yields typical for this type of synthesis limit their applica-
tion in large-scale manufacture.
Our group developed a template-free method to synthesize
hollow-structured Co3O4 nanoparticles dispersed on a carbon
support.8 As shown in Figure 2a, pores ranging from 10 to 50 nm
are clearly evident. The Co3O4/C nanoparticles showed excellent
performance as conversion anodes. They exhibited high capacity
DOI: 10.1021/acs.accounts.7b00487
Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research
Figure 2. (a) Bright-field scanning transmission electron microscopy
(BF-STEM) image of hollow Co3O4/C. Reproduced from ref 8.
Copyright 2015 American Chemical Society. (b) Transmission electron
microscopy (TEM) image of Fe3O4 MIONCs clusters. Reproduced
from ref 9. Copyright 2013 American Chemical Society. (c) TEM image
of carbon-coated Mn3−xFexO4 nanocrystals. Reproduced with permis-
sion from ref 10. Copyright 2013 American Association for the Advance-
ment of Science. (d) TEM image of Mn3O4 dispersed on RGO.
The inset shows the electron diffraction pattern. Reproduced from
ref 11. Copyright 2010 American Chemical Society.
retention, with a capacity of 880 mA h g−1 after 50 cycles (the
theoretical capacity is 890 mA h g−1). They also delivered
reversible capacities of 850, 750, 600, and 450 mA h g−1 at 100,
500, 1000, and 2000 mA g−1, respectively.
Lee et al.9 synthesized Fe3O4 mesoporous iron oxide nanopar-
ticle clusters (MIONCs) using a bottom-up self-assembly
approach (Figure 2b). These clusters had voids distributed inside
the structure, which provided buffer space/volume during the
conversion reaction, as well as shortened lithium diffusion paths.
The SEI layer formed mostly on the surface of the MIONCs,
generating a shell that helped stabilize and support the structure.
As a result of the robust three-dimensional (3D) architecture, a
specific capacity of 867 mA h g−1 was achieved after 80 cycles.
The reaction voltage of conversion-reaction-based transition
metal oxides is closely related to the cation, so it can be tuned by
adjusting the cation ratio in multimetal oxides. Oh et al.10 syn-
thesized Mn3−xFexO4 nanocages via a galvanic replacement reac-
tion (Figure 2c). The reaction started with growth of γ-Fe3O4 on
the surface of Mn3O4 nanoparticle clusters (NCs). These Mn3O4
NCs were later reduced by Fe2+ ions and dissolved into solu-
tion during the replacement reaction. These multimetallic oxide
nanoparticles showed high cycle stability, with a reversible capa-
city of ∼1000 mA h g−1 for 50 cycles without apparent capacity
fade. Their study also demonstrated that the reaction voltage
could be easily tuned by adjusting the ratio of Mn and Fe ions in
the Mn3−xFexO4 composite.
Incorporation with/on carbonaceous materials, such as in carbon-
coated metal oxides and metal oxides anchored on graphene
layers, has also shown greatly enhanced performance. For
example, Wang et al.11 prepared Mn3O4 nanoparticles dispersed
276
Article
on reduced graphene oxide (RGO) through a two-step solution-
phase reaction (Figure 2d). These hybrid materials delivered a
capacity of about 900 mA h g−1 at a current density of 40 mA g−1.
At the high current density of 1600 mA g−1, a capacity of about
390 mA h g−1 was delivered. This improvement in the elec-
trochemical performance of Mn3O4 nanoparticles on RGO
compared to bare Mn3O4 is due at least in part to enhanced
electrical conductivity and decreased aggregation arising from
good dispersion of the Mn3O4 nanoparticles on the RGO.
2.3. Conversion-Type Cathodes
The reaction voltage of conversion-reaction-based materials is
strongly dependent on both the transition metal and the anionic
species.12 The calculated electromotive force (emf) values of
binary transition metal compounds with transition metals from
Ti to Cu and anionic species (F, O, S, and N) for conversion
reactions are shown in Figure 3a. As a general trend, for a given
anionic species, the emf increases as the atomic number of the
transition metal increases. For a given metal, the emf values for
these anions increase in the following order: F > O ≈ S > N (the
order of the emfs for oxides and sulfides changes depending on
the metal). The high ionicity of the bond between the metal and
fluoride anions leads to high reduction potentials for the
conversion reaction. Therefore, some transition metal fluorides,
including transition metal oxyfluorides, have been investigated as
potential cathode materials and have shown promising behavior.
Fan et al. synthesized poly-(3,4-ethylenedioxythiophene)
(PEDOT)-coated FeOF (FeOF@PEDOT) nanorods.13 In the
discharge profile (Figure 3b), two plateaus (regions I and II)
were observed, corresponding to intercalation and conversion
reactions, respectively. The composite delivered a high reversible
capacity of about 560 mA h g−1 at a current density of 10 mA g−1.
The material also showed excellent cycling performance (94%
capacity retention) during 150 cycles at 50 mA g−1, after a five-
cycle activation. Wang et al.14 prepared ternary metal fluorides
(CuyFe1−yF2) by ball-milling. Among these, Cu0.5Fe0.5F2 showed
highly reversible behavior, in contrast to the irreversible elec-
trochemical reaction of CuF2 (Figure 3c). It was also found that
the voltage hysteresis of Cu0.5Fe0.5F2, obtained by the galvano-
static intermittent titration technique (GITT), was only 148 and
∼200 mV for the Cu and Fe redox processes, respectively.
Although the conversion-reaction-based binary/ternary tran-
sition metal fluorides/oxyfluorides deliver higher capacities than
conventional intercalation-based cathodes, they do not contain
lithium, which is a critical drawback for cathode materials. How-
ever, recent emerging cathode materials, LiF + TMOs/TMFs,
which are prepared by blending lithium fluoride with transition
metal oxides/fluorides, exhibited high capacity with LiF serving
as the Li source. Kim et al.15 initially reported this strategy and
proved their concept using LiF + FeF2 composites. Well-blended
LiF + FeF2 composites exhibited a capacity of 190 mA h g−1 at
50 mA g−1. Fan and co-workers prepared Fe+LiF+C and
FeM+LiF+C (M = Co, Ni) nanocomposites, and they exhibited
very high capacities of above 400 and 300 mA h g−1, respec-
tively.16,17 Dimov et al.18 expanded this approach to transition
metal oxides (LiF + MnOx). They demonstrated a high reversible
reactivity of LiF + MnOx in the voltage range between 1.5 and
4.8 V. It is worth noting that manganese oxides are electro-
chemically inactive at high potentials, although they can deliver
high capacities at low potentials (<1.5 V) through classical con-
version reactions. Recently, Jung et al.19 carried out an in-depth
analysis of the reaction mechanism of LiF + TMO (TM = Mn,
Fe, Co) nanocomposites (Figure 3d). The reaction mechanism
DOI: 10.1021/acs.accounts.7b00487
Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research
Figure 3. (a) Calculated emfs of binary transition metal compounds for conversion reactions. Reproduced with permission from ref 12. Copyright 2011
Royal Society of Chemistry. (b) Voltage profiles of FeOF@PEDOT nanorods. Reproduced from ref 13. Copyright 2015 American Chemical Society.
(c) Voltage profiles of Cu0.5Fe0.5F2 for the initial five cycles. Reproduced with permission from ref 14. Copyright 2015 Nature Publishing Group.
(d) Schematic of cathode material design by blending LiF and transition metal monoxide. Reproduced with permission from ref 19. Copyright 2017
Nature Publishing Group.
of a LiF + MnO nanocomposite has been suggested to occur on
the surface of the Mn as follows (“surface conversion reaction”):
LiF + Mn IIO ⇄ Li+ + e− + Mn IIIO−F (3)
3. SULFUR CATHODES FOR LI−S BATTERIES
3.1. Reaction Mechanism Studies of Sulfur Cathodes
The theoretical capacity of a sulfur cathode is 1675 mA h g−1
assuming that the starting material is sulfur (S8) and that it is fully
converted into lithium sulfide (Li2S). This value is much higher
than that of state-of-the-art intercalation-based cathodes, such
as LiCoO2. In the typical discharge profile of a Li−S cell with
liquid electrolytes, two plateaus are clearly defined (Figure 4a).
Insoluble S8 is reduced to soluble (higher) polysulfide inter-
mediates in the upper plateau, and soluble polysulfides are
further reduced into insoluble lithium sulfide in the lower
plateau.20 In terms of electrolytes for the Li−S battery system,
since carbonate-based electrolytes irreversibly react with polysulfides,
ether-based solvents such as 1,3-dioxolane (DOL), 1,2-dimeth-
oxyethane (DME), and tetra-(ethylene glycol) dimethyl ether
(TEGDME) are frequently used. On the other hand, only one
plateau is typically observed in the discharge profile when a
solid electrolyte is used or when sulfur contained in micropores
(<2 nm, so that only small molecules, S2−4, are incorporated/
confined) is used as the starting material.21−23 When sulfur con-
fined in micropores is used as the starting material, the carbonate
electrolytes can also be used.
The reactions occurring in a typical Li−S battery system are
very complex, and various soluble polysulfides are produced as
intermediates (Figure 4b).21 Although the detailed mechanism is
not fully understood, much progress has been made. Among the
many analytical techniques available, in situ/operando XRD is
277
Article
especially well suited and powerful to study the evolution of
crystalline structures during battery operation. In our previous
operando XRD-based study, we demonstrated that orthorhom-
bic α-S8 disappeared at the upper plateau and Li2S appeared
toward the end of discharge (Figure 4c).24 In addition, during
charging, the Li2S peak gradually decreased and disappeared at
the end of the lower plateau, and monoclinic β-S8 was formed at
the end of the charge. The formation of β-S8 is interesting
because this phase (β) is known to be stable only above 95 °C.25
However, the origin of the formation of the β phase rather than
the α phase is unclear, though nucleation from polysulfides to
metastable β-phase sulfur might be energetically favorable.
XAS has also been applied to Li/S cells in an effort to identify
the polysulfide intermediates during cycling. We have studied the
evolution of polysulfide species using in situ XAS at the S K-edge
region (Figure 4d). During discharge, the prepeak at about
2471.1 eV, related to S3− radical anions, appeared at the end of
the upper plateau. As the voltage reached the lower plateau, the
main peak shifted to higher energies. The positions of these two
peaks remained essentially constant during the lower plateau
without the appearance of any other signal. On the basis of this
XAS result, it was suggested that long-chain polysulfides chemi-
cally dissociate to form S3−, which is electrochemically reduced to
S32− and subsequently reacts with the longer polysulfides. During
the recharge, the XAS peaks evolved in reverse sequence at the
same position, indicating that the (microscopic) reverse reaction
occurs during charge. Recently, Zhang et al.26 investigated the
reaction mechanism of a Li2S electrode using in situ/operando
XAS and suggested two different pathways during the charge
process when the starting material was Li2S instead of S. During
the first cycle, Li2S was oxidized to elemental sulfur without poly-
sulfide formation. However, during the second charge, residual
polysulfides remained after the first discharge, affecting the
DOI: 10.1021/acs.accounts.7b00487
Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research Article

Figure 4. (a) Typical discharge profile of a Li−S battery showing two clear plateaus. Reproduced with permission from ref 20. Copyright 2013 Nature
Publishing Group. (b) Schematic illustration of stepwise reduction reactions from elemental sulfur to lithium sulfide generating intermediate
polysulfides. Reproduced with permission from ref 21. Copyright 2014 Elsevier. (c) Voltage profile of the Li−S battery during cycling (top) and peak
heights of S8 and Li2S from the corresponding in situ XRD results (bottom). (d) In situ XAS spectra selected from specific points (left) and variations of
the fluorescence signal and cell voltage as a function of time (right). Reproduced with permission from ref 24. Copyright 2014 Royal Society of
Chemistry.

charge processes of Li2S, so that Li2S was converted to poly-
sulfides and then oxidized to elemental sulfur.
3.2. Strategies To Improve the Electrochemical Performance
of Li−S Batteries
Physical entrapment of polysulfides is the most popular strategy
to enhance the capacity and cycling performance of Li−S bat-
teries. Nazar’s group first demonstrated that a highly ordered
mesoporous carbon/sulfur composite cathode exhibited enhanced
cycling stability and high capacity (Figure 5a).27 Sulfur was infil-
trated into mesoporous carbon (CMK-3) by a melt−diffusion
method. The carbon framework acted as both an electronic
conductor and a mini-electrochemical reaction chamber, which
ensured that a more complete redox process would take place.
An extra polymer coating on the surface of the composite further
helped mitigate diffusion of the polysulfides out into the elec-
trolyte. The composites delivered a discharge capacity of up to
1320 mA h g−1 (though normalized to the mass of sulfur).
This work spurred rapid growth in the field of lithium−sulfur
batteries.
Various carbon materials have been used to increase the elec-
trode conductivity and provide physical confinement/entrap-
ment of lithium polysulfides. Among them, microporous carbon/
sulfur composites were found to undergo a new reaction mecha-
nism without forming long-chain polysulfides (Figure 5b).22
They found that small sulfur allotropes, S2−4, could be confined
in a microporous carbon matrix with a pore size of ∼0.5 nm.
These small S2−4 molecules could not be converted into long-
chain polysulfides because of the limited space within the pores,
resulting in a single output plateau at ∼1.9 V. The composite
278
displayed a high specific capacity of 1670 mA h g−1 and excellent
cycling stability during 200 cycles. Although the use of micro-
porous carbons greatly improved the stability, the sulfur load-
ing in the composite was relatively low (<50%) because of the
limited surface area of the microporous carbon materials. It is
worthwhile noting that in a previous study we established the key
parameters governing the energy density (especially volumetric)
of Li−S batteries, particularly in comparison to metal oxide
cathodes, making a direct comparison with LiCoO2.28 In that
study, we demonstrated that the sulfur loading played a critical
role in determining the volumetric energy density and that in
order for Li−S batteries to be competitive with a typical graphite/
LiCoO2 cell, the sulfur loading had to be at or beyond 70% by weight.
Graphene and graphene oxide (GO) have been extensively
studied as sulfur host materials. The polysulfides were wrapped
by graphene/GO, which suppressed diffusion of the polysulfides
into the electrolyte. Further improvement was achieved by using
oxygen-rich GO as the sulfur host.29 The functional groups in
GO play an important role in immobilizing the otherwise-free
lithium polysulfides. However, the superficial sulfur cannot be
fully trapped, and the open channels among the GO layers pro-
vide large space for polysulfides to escape. A soft polymer,
amylopectin (Amy), was employed to wrap the GO/S composite
and form cross-linked 3D structures to immobilize the polysulfide
species and accommodate the volumetric expansion (Figure 5c).30
With the help of the three-dimensionally cross-linked structure,
the Li−S battery exhibited much improved cycling stability and
Coulombic efficiency compared to the unwrapped composite.
DOI: 10.1021/acs.accounts.7b00487
Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research Article

Figure 5. (a) Schematic of sulfur confined in mesoporous carbon. Reproduced with permission from ref 27. Copyright 2009 Nature Publishing Group.
(b) Schematic of S2−4 confined in microporous carbon and its charge/discharge process. Reproduced from ref 22. Copyright 2012 American Chemical
Society. (c) Schematic of GO/S/Amy composite. Reproduced from ref 30. Copyright 2013 American Chemical Society. (d) Schematic of the synthesis
of the S−Pani yolk−shell composite. Reproduced from ref 31. Copyright 2013 American Chemical Society. (e) Schematic of oxidation of polysulfide by
δ-MnO2 to form thiosulfate on the surface. Reproduced with permission from ref 32. Copyright 2015 Nature Publishing Group. (f) Schematic
configuration of a Li−S cell with a bifunctional microporous carbon interlayer. Reproduced with permission from ref 33. Copyright 2012 Nature
Publishing Group.

Conducting polymers represent a novel and interesting class
of encapsulation materials for sulfur cathodes. They can act
as an effective physical barrier to confine polysulfide species.
Our group prepared sulfur−polyaniline (S−Pani) yolk−shell
nanostructures by heat treatment of a core−shell nanostructure
(Figure 5d).31 The capacities of the S−Pani yolk−shell com-
posite were 1101 and 920 mA h g−1 at 0.2C and 0.5C, respec-
tively. Also, stable, reversible capacities of 765 mA h g−1 at 0.2C
and 628 mA h g−1 at 0.5C after 200 cycles were obtained.
The yolk−shell structure can encapsulate the sulfur inside the
polymer shells with a buffer void. The internal void space can
accommodate the large volumetric changes of sulfur while main-
taining the structural integrity of the composite, whereas the
core−shell structure is destroyed during cycling.
Chemical interaction is a new kind of strategy to suppress
diffusion of polysulfides. Metal oxides, nitrides, and sulfides have
been widely studied, as they can exhibit strong affinity for poly-
sulfide species by polar−polar interactions.32 MnO2 was inves-
tigated as a sulfur host material by the Nazar group.32 They
reported that MnO2 can significantly improve the cycling stability
of Li−S batteries through a thiosulfate−polythionate conversion
(Figure 5e). The surface Mn4+ oxidizes polysulfides, forming
Mn2+ and S2O32−. These species anchor higher-order polysul-
fides Sn2− (n > 4) by catenating them into the S−S bond in thio-
sulfate to form polythionate complexes, while the lower-order
polysulfides are in contact with the MnO2 host.
Apart from modification of the cathode materials and architec-
tures, some other strategies have been proposed. For example,
the introduction of a microporous carbon interlayer between the
cathode and the separator was reported by the Manthiram group
(Figure 5f).33 The interlayer helps prevent lithium polysulfides
from diffusing into the electrolyte and also improves the utiliza-
tion of active material by serving as an additional (upper) current
collector. As a result, a stable capacity of over 1000 mA h g−1 was
achieved after 100 cycles at 1C. Modification of the separator
279
is another strategy to enhance the cycling stability of Li−S
batteries. A layer of carbon black was coated onto the polymer
separator.34 This provides much higher capacity than that of the
unmodified one, indicating an enhanced utilization of the active
material. A high initial capacity of 1350 mA h g−1 was achieved,
and the capacity retention reached 55% after 500 cycles at 0.5C.
4. CONCLUSIONS AND OUTLOOK
We have broadly reviewed conversion-type electrode materials
for LIBs with a focus on understanding the reaction mechanisms
involved. Anodic conversion reactions often have different reac-
tion paths between charge and discharge processes associated
with the formation of new phases. Understanding these different
reaction pathways and the roles of Li-centered reactions could
provide additional solutions to improve the round-trip efficiency.
To achieve a significant breakthrough in the performance of
Li−S batteries, it is necessary to have a full understanding of both
the fundamental electrochemistry of Li−S batteries and the
interactions of lithium polysulfides with host materials and lithium
anodes. Furthermore, to make Li−S batteries more comparable/
competitive with commercial LIBs, one needs to increase the
sulfur mass loading to above 70%.
Although we have focused on LIB applications, the general
concepts of conversion-type electrode materials can be applied to
sodium-ion batteries (SIBs). Since traditional graphite and sili-
con are known to be inappropriate as SIB anodes, we anticipate
that conversion-type electrodes will be attractive candidates for
SIB anodes. In addition, many strategies to improve the perfor-
mance of Li−S batteries could be adapted to Na−S batteries.
However, the reaction mechanisms of electrodes can be quite
different in SIBs and Na−S batteries because of different phase
stabilities. Through charge/discharge, sodium compounds form
various stable sulfides (Na2S, Na2S2, Na2S4, and Na2S5) and oxides
(Na2O, Na2O2, and NaO2), while lithium-based materials are
known to form Li2S, Li2O, and Li2O2. Thus, conversion reactions
DOI: 10.1021/acs.accounts.7b00487
Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research
in SIBs and Na−S batteries may have more intermediates.
In addition, the reaction pathways can be different, although the
overall reactions are similar. For example, it was recently demon-
strated that the nanoscale reaction pathways for the reactions of
Li and Na with NiO are different in detail but follow a similar
mechanism.35 The mechanism in SIBs could be more complicated
but can also provide additional venues to explore. Thus, it is crucial
to perform systematic studies on conversion materials in SIBs or
Na−S batteries to reveal their reaction mechanisms.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: hda1@cornell.edu.
ORCID
Seung-Ho Yu: 0000-0003-4833-3181
Na Zhang: 0000-0003-1978-4469
Héctor D. Abruña: 0000-0002-3948-356X
Notes
The authors declare no competing financial interest.
Biographies
Seung-Ho Yu is a postdoctoral associate in the Department of
Chemistry and Chemical Biology at Cornell University and Cornell
Laboratory for Accelerator-Based Sciences & Education (CLASSE).
He received B.S. (2008) and Ph.D. (2013) from the School of Chemical
and Biological Engineering at Seoul National University. He worked as a
researcher at the Research Institute of Advanced Materials (RIAM) and
as a senior researcher at the Center for Nanoparticle Research in the
Institute for Basic Science (IBS) at Seoul National University.
His current research focuses on revealing reaction mechanisms of
electrode materials for Li-ion and post-Li-ion batteries.
Xinran Feng is a Ph.D. candidate in the Department of Chemistry and
Chemical Biology at Cornell University and a graduate research assistant
at CLASSE. She received her B.S. in Chemistry from Pennsylvania State
University (2014). Her current research focuses on understanding
conversion reaction mechanisms using operando X-ray methods.
Na Zhang is a Ph.D. candidate in the Department of Chemistry and
Chemical Biology at Cornell University. She obtained her B.S. degrees
from Jilin University (2014). Her current research focuses on
developing cathode materials for Li−S batteries.
Jeesoo Seok is a Ph.D. candidate in the Department of Chemistry and
Chemical Biology at Cornell University. She obtained her B.S. (2012)
and M.S. (2014) from Ewha Womans University. Her research involves
experimental solid-state chemistry for sodium-ion batteries and organic
materials for redox flow batteries.
Héctor D. Abruña is the Émile M. Chamot Professor in the Department
of Chemistry and Chemical Biology at Cornell University, which he
joined in 1983. He was Department Chair from 2004 to 2008 and is
currently the Director of the Energy Materials Center at Cornell (emc2).
He considers his 52 Ph.D. students and 65 postdoctoral associates to
date as his most important professional achievement.
■ ACKNOWLEDGMENTS
S.-H.Y. acknowledges support from the Cornell High Energy
Synchrotron Source (CHESS) and the Energy Materials Center
at Cornell (emc2). X.F. was supported by CHESS under NSF
MPS Grant DMR-1332208.
280
Article
■ REFERENCES
(1) Tarascon, J.-M.; Armand, M. Issues and Challenges Facing
Rechargeable Lithium Batteries. Nature 2001, 414, 359−367.
(2) Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J.-M.
Nano-Sized Transition-Metal Oxides as Negative-Electrode Materials
for Lithium-Ion Batteries. Nature 2000, 407, 496−499.
(3) Yu, S.-H.; Lee, S. H.; Lee, D. J.; Sung, Y.-E.; Hyeon, T. Conversion
Reaction-Based Oxide Nanomaterials for Lithium Ion Battery Anodes.
Small 2016, 12, 2146−2172.
(4) Lowe, M.; Gao, J.; Abruña, H. D. In operando X-Ray Studies of the
Conversion Reaction in Mn3O4 Lithium Battery Anodes. J. Mater. Chem.
A 2013, 1, 2094−2103.
(5) Hu, Y.-Y.; Liu, Z.; Nam, K.-W.; Borkiewicz, O. J.; Cheng, J.; Hua,
X.; Dunstan, M. T.; Yu, X.; Wiaderek, K. M.; Du, L.-S.; Chapman, K. W.;
Chupas, P. J.; Yang, X.-Q.; Grey, C. P. Origin of Additional Capacities in
Metal Oxide Lithium-Ion Battery Electrodes. Nat. Mater. 2013, 12,
1130−1136.
(6) He, K.; Zhang, S.; Li, J.; Yu, X.; Meng, Q.; Zhu, Y.; Hu, E.; Sun, K.;
Yun, H.; Yang, X.-Q.; Zhu, Y.; Gan, H.; Mo, Y.; Stach, E. A.; Murray, C.
B.; Su, D. Visualizing Non-Equilibrium Lithiation of Spinel Oxide via In
Situ Transmission Electron Microscopy. Nat. Commun. 2016, 7, 11441.
(7) Luo, L.; Wu, J.; Li, Q.; Dravid, V. P.; Poeppelmeier, K. R.; Rao, Q.;
Xu, J. Reactions of Graphene Supported Co3O4 Nanocubes with
Lithium and Magnesium Studied by In Situ Trasmission Electron
Microscopy. Nanotechnology 2016, 27, 085402.
(8) Wang, D.; Yu, Y.; He, H.; Wang, J.; Zhou, W.; Abruña, H. D.
Template-Free Synthesis of Hollow-Structured Co3O4 Nanoparticles as
High-Performance Anodes for Lithium-Ion Batteries. ACS Nano 2015,
9, 1775−1781.
(9) Lee, S. H.; Yu, S.-H.; Lee, J. E.; Jin, A.; Lee, D. J.; Lee, N.; Jo, H.;
Shin, K.; Ahn, T.-Y.; Kim, Y.-W.; Choe, H.; Sung, Y.-E.; Hyeon, T. Self-
Assembled Fe3O4 Nanoparticle Clusters as High-Performance Anodes
for Lithium Ion Batteries via Geometric Confinement. Nano Lett. 2013,
13, 4249−4256.
(10) Oh, M. H.; Yu, T.; Yu, S.-H.; Lim, B.; Ko, L.-T.; Willinger, M.-G.;
Seo, D.-H.; Kim, B. H.; Cho, M. G.; Park, J.-H.; Kang, K.; Sung, Y.-E.;
Pinna, N.; Hyeon, T. Galvanic Replacement Reactions in Metal Oxide
Nanocrystals. Science 2013, 340, 964−968.
(11) Wang, H.; Cui, L.-F.; Yang, Y.; Sanchez Casalongue, H.;
Robinson, J. T.; Liang, Y.; Cui, Y.; Dai, H. Mn3O4-Graphene Hybrid as a
High-Capacity Anode Material for Lithium Ion Batteries. J. Am. Chem.
Soc. 2010, 132, 13978−13980.
(12) Zu, C.-X.; Li, H. Thermodynamic Analysis on Energy Densities of
Batteries. Energy Environ. Sci. 2011, 4, 2614−2624.
(13) Fan, X.; Luo, C.; Lamb, J.; Zhu, Y.; Xu, K.; Wang, C. PEDOT
Encapsulated FeOF Nanorod Cathodes for High Energy Lithium-Ion
Batteries. Nano Lett. 2015, 15, 7650−7656.
(14) Wang, F.; Kim, S.-W.; Seo, D.-H.; Kang, K.; Wang, L.; Su, D.;
Vajo, J. J.; Wang, J.; Graetz, J. Ternary Metal Fluorides as High-Energy
Cathodes with Low Cycling Hysteresis. Nat. Commun. 2015, 6, 6668.
(15) Kim, S.-W.; Nam, K.-W.; Seo, D.-H.; Hong, J.; Kim, H.; Gwon, H.;
Kang, K. Energy Storage in Composites of a Redox Couple Host and a
Lithium Ion Host. Nano Today 2012, 7, 168−173.
(16) Fan, X.; Zhu, Y.; Luo, C.; Suo, L.; Lin, Y.; Gao, T.; Xu, K.; Wang,
C. Pomegranate-Strcutrued Conversion-Reaction Cathode with a Built-
in Li Source for High Energy Li-Ion Batteries. ACS Nano 2016, 10,
5567−5577.
(17) Fan, X.; Zhu, Y.; Luo, C.; Gao, T.; Suo, L.; Liou, S.-C.; Xu, K.;
Wang, C. In Situ Lithiated FeF3/C Nanocomposite as High Energy
Conversion-Reaction Cathode for Lithium-Ion Batteries. J. Power
Sources 2016, 307, 435−442.
(18) Dimov, N.; Kitajou, A.; Hori, H.; Kobayashi, E.; Okada, S.
Electrochemical Splitting of LiF:A New Approach to Lithium-Ion
Battery Materials. ECS Trans. 2014, 58, 87−99.
(19) Jung, S.-K.; Kim, H.; Cho, M. G.; Cho, S.-P.; Lee, B.; Kim, H.;
Park, Y.-U; Hong, J.; Park, K.-Y.; Yoon, G.; Seong, W. M.; Cho, Y.; Oh,
M. H.; Kim, H.; Gwon, H.; Hwang, I.; Hyeon, T.; Yoon, W.-S.; Kang, K.
Lithium-Free Transition Metal Monoxides for Positive Electrodes in
Lithium-Ion Batteries. Nat. Energy 2017, 2, 16208.
DOI: 10.1021/acs.accounts.7b00487
Acc. Chem. Res. 2018, 51, 273−281
Accounts of Chemical Research Article

(20) Su, Y.-S.; Fu, Y.; Cochell, T.; Manthiram, A. A Strategic Approach
to Recharging Lithium-Sulphur Batteries for Long Cycle Life. Nat.
Commun. 2013, 4, 2985.
(21) Busche, M. R.; Adelhelm, P.; Sommer, H.; Schneider, H.; Leitner,
K.; Janek, J. Systematical Electrochemical Study on the Parasitic Shuttle-
Effect in Lithium-Sulfur-Cell at Different Temperatures and Different
Rates. J. Power Sources 2014, 259, 289−299.
(22) Xin, S.; Gu, L.; Zhao, N.-H.; Yin, Y.-X.; Zhou, L.-J.; Guo, Y.-G.;
Wan, L.-J. Smaller Sulfur Molecules Promise Better Lithium−Sulfur
Batteries. J. Am. Chem. Soc. 2012, 134, 18510−18513.
(23) Guo, J.; Xu, Y.; Wang, C. Sulfur-Impregnated Disordered Carbon
Nanotubes Cathode for Lithium−Sulfur Batteries. Nano Lett. 2011, 11,
4288−4294.
(24) Lowe, M. A.; Gao, J.; Abruña, H. D. Mechanistic Insights into
Operational Lithium−Sulfur Batteries by In Situ X-Ray Diffraction and
Absorption Spectroscopy. RSC Adv. 2014, 4, 18347−18353.
(25) Meyer, B. Elemental Sulfur. Chem. Rev. 1976, 76, 367−388.
(26) Zhang, L.; Sun, D.; Feng, J.; Cairns, E. J.; Guo, J. Revealing the
Electrochemical Charging Mechanism of Nanosized Li2S by In Situ and
Operando X-Ray Absorption Spectroscopy. Nano Lett. 2017, 17, 5084−
5091.
(27) Ji, X.; Lee, K. T.; Nazar, L. F. A Highly Ordered Nanostructured
Carbon−Sulphur Cathode for Lithium−Sulphur Batteries. Nat. Mater.
2009, 8, 500−506.
(28) Gao, J.; Abruña, H. D. Key Parameters Governing the Energy
Density of Rechargeable Li/S Batteries. J. Phys. Chem. Lett. 2014, 5,
882−885.
(29) Ji, L.; Rao, M.; Zheng, H.; Zhang, L.; Li, Y.; Duan, W.; Guo, J.;
Cairns, E. J.; Zhang, Y. Graphene Oxide as a Sulfur Immobilizer in High
Performance Lithium/Sulfur Cells. J. Am. Chem. Soc. 2011, 133, 18522−
18525.
(30) Zhou, W.; Chen, H.; Yu, Y.; Wang, D.; Cui, Z.; DiSalvo, F. J.;
Abruña, H. D. Amylopectin Wrapped Graphene Oxide/Sulfur for
Improved Cyclability of Lithium−Sulfur Battery. ACS Nano 2013, 7,
8801−8808.
(31) Zhou, W.; Yu, Y.; Chen, H.; DiSalvo, F. J.; Abruña, H. D. Yolk−
Shell Structure of Polyaniline-Coated Sulfur for Lithium−Sulfur
Batteries. J. Am. Chem. Soc. 2013, 135, 16736−16743.
(32) Liang, X.; Hart, C.; Pang, Q.; Garsuch, A.; Weiss, T.; Nazar, L. F. A
Highly Efficient Polysulfide Mediator for Lithium−Sulfur Batteries. Nat.
Commun. 2015, 6, 5682.
(33) Su, Y.-S.; Manthiram, A. Lithium−Sulphur Batteries with a
Microporous Carbon Paper as a Bifunctional Interlayer. Nat. Commun.
2012, 3, 1166.
(34) Yao, H.; Yan, K.; Li, W.; Zheng, G.; Kong, D.; Seh, Z. W.;
Narasimhan, V. K.; Liang, Z.; Cui, Y. Improved Lithium−Sulfur
Batteries with a Conductive Coating on the Separator to Prevent the
Accumulation of Inactive S-Related Species at the Cathode−Separator
Interface. Energy Environ. Sci. 2014, 7, 3381−3390.
(35) He, K.; Lin, F.; Zhu, Y.; Yu, X.; Li, J.; Lin, R.; Nordlund, D.; Weng,
T.-C.; Richards, R. M.; Yang, X.-Q.; Doeff, M. M.; Stach, E. A.; Mo, Y.;
Xin, H. L.; Su, D. Sodiatoin Kinetics of Metal Oxide Conversion
Electrodes: A Comparative Study with Lithiation. Nano Lett. 2015, 15,
5755−5763.

281 DOI: 10.1021/acs.accounts.7b00487

Acc. Chem. Res. 2018, 51, 273−281