nature reviews chemistry https://doi.org/10.

1038/s41570-024-00622-1

Review article Check for updates

Manipulating disorder within

cathodes of alkali-ion batteries
Seongkoo Kang1,4, Suwon Lee1,4, Hakwoo Lee2 & Yong-Mook Kang 1,2,3

Abstract Sections

The fact that ordered materials are rarely perfectly crystalline is widely Introduction

acknowledged among materials scientists, but its impact is often Disorder in solid-state
overlooked or underestimated when studying how structure relates materials

to properties. Various investigations demonstrate that intrinsic Origin of disorder in cathodes

and extrinsic defects, and disorder generated by physicochemical Impact of structural disorder
reactions, are responsible for unexpectedly detrimental or beneficial in cathodes

functionalities. The task remains to modulate the disorder to produce Future perspectives
desired properties in materials. As disorder is often correlated with
local interactions, it is controllable. In this Review, we explore the
structural disorder in cathode materials as a novel approach for
improving their electrochemical performance. We revisit cathode
materials for alkali-ion batteries and outline the origins and beneficial
consequences of disorder. Focusing on layered, cubic rocksalt and
other metal oxides, we discuss how disorder improves electrochemical
properties of cathode materials and which interactions generate the
disorder. We also present the potential pitfalls of disorder that must
be considered. We conclude with perspectives for enhancing the
electrochemical performance of cathode materials by using disorder.

Symmetric Li/TM
distortion Disorder mixing

Asymmetric TM
distortion migration

Vacancy Oxygen
formation evolution

Interlayer Property Li/TM site

disorder disorder

A full list of affiliations appears at the end of the paper. e-mail: dake1234@korea.ac.kr

Nature Reviews Chemistry

Review article

Introduction materials that can be strategically targeted to control disorder (Fig. 1).
The properties of solid-state materials are influenced by their struc- We mainly present specific case studies of face-centred cubic (FCC)
tures. Crystallography has established a fundamental understanding oxide cathode materials for alkali-ion batteries, illustrating how disor-
of atomic structures1,2 by projecting the structure into the reciprocal der can be effectively modulated to control electrochemical property,
space, based on the periodicity of the unit cell and lattice. Unfortu- and to establish the disorder–property relationship to be used as
nately, the conventional crystallographic approach has limitations in possible future design strategies. We also highlight risks that may be
detecting structural disorder of materials, as they often oversimplify encountered when designing disorder into cathode materials.
the real structure of the material in perfect crystals, which requires
careful interpretation of diffraction patterns to extract information of Disorder in solid-state materials
defective structures3. However, solids often contain varying degrees Disorder is inherent to various materials, and investigations of this
of disorder, which leads to unexpected ion or electron conducting, property have relied on crystallographic approaches. However, these
magnetic, and optical properties4–6. Therefore, it becomes imperative approaches have shown limitations in the analysis of structural disor-
to attribute properties to the disorder that accompanies solid-state der that deviates from ideal and perfect crystals. Studies have tried
materials. to propose an adequate system to comprehend disordered materi-
Disorder can be generated through reactions and/or synthesis of als4,17–22. Among the investigations, correlated disorder describes the
materials and also through external stimuli, such as magnetic fields, disorder that exhibits correlations or dependencies between different
pressure or electric environments7–9. The electrode materials of alkali- parts or variables within the system. In other words, although disorder
ion batteries serve as a representative example of these disordered typically implies randomness or lack of crystallographic periodic-
materials in the field of energy storage and battery materials. Alkali- ity, a correlated disorder suggests the presence of relationships or
ion batteries electrochemically store electrical energy via the redox dependencies between the random elements. The crystallography of
reactions of their electrodes. During charge/discharge, the electrode correlated disorder has emphasized the pivotal role of total scattering
material undergoes the extraction and insertion of alkali ions balanced methods for depicting this disorder and establishing a relationship
with the loss and gain of electrons through their redox reactions. This with material properties22. From a statistical mechanics standpoint,
underscores why well-crystallized layered transition metal (TM) oxides, correlated disorder can be described by three considerations: degrees
featuring facile diffusion pathways of alkali ions through interlayer of freedom, interactions between degrees of freedom and lattice4. To
spacings and stable redox centres, have been widely used as cathode illustrate this, we turn to the classic example of the Ising model in sta-
materials in alkali-ion batteries. However, the repetitive creation and tistical mechanics, which involves Ising spin orientations with values
removal of alkali-ion vacancies, balanced by charge compensation of + 1 or −1 (degrees of freedom), magnetic interactions and the lattice
through redox reactions, causes local disorder and even triggers irre- structure, to determine the order or disorder of the system.
versible phase transitions10. This phenomenon leads to detrimental Correlated disorder can be attributed to a unique atomic arrange-
electrochemical properties, as it hinders the reversibility of redox ment arising from an incompatibility between competing interactions
reactions, and thus shortens a battery’s lifetime. Achieving the greatest and the lattice, or the presence of dominating interactions leading to
theoretical energy densities of cathode materials is limited to maximiz- multiple ordering patterns. This results in two ordering problems of
ing the reversibility of redox reactions. Past endeavours to enhance geometrical frustration: frustrated overconstraint and configurational
the electrochemical reversibility of layered cathodes have predomi- underconstraint22,23. Geometrical frustration has been extensively
nantly revolved around mitigating disorder. Although this approach explored in magnetism, water ice and spin ice24–27. For instance, in the
contributed to the commercialization of layered cathode materials, it case of water ice, the oxygen atoms of H2O molecules are tetrahedrally
necessitates a compromise in the theoretical energy density. As a result, coordinated to four other atoms—two hydrogen atoms forming near-
new approaches are required to achieve high reversibility and energy covalent bonds and two hydrogen atoms forming more distant hydrogen
density simultaneously. Investigations into cathode materials have bonds between two oxygen atoms (Fig. 2a, left). This leads to the ordered
increasingly focused on characterizing atomic disorder and utilizing arrangement of water molecules. However, in certain conditions, the ori-
the disorder to engineer the material’s properties. The understand- entation of water molecules with changes in local interactions involving
ing of the intricate interplay between disorder and electrochemical hydrogen bonds can result in structural disorder in water ice influencing
properties will provide a new strategy to design advanced cathode its volume28. A similar issue arises with stacking disordered ice, composed
materials11–16. Nevertheless, the concept of controlling disorder remains of cubic and hexagonal stacking driven by correlated disorder stemming
elusive and even paradoxical, as it entails manipulating randomness. from the interaction energy between different orientations4,29. Frustrated
Consequently, it is imperative to establish definable parameters for magnetism in pyrochlore materials is analogous to pyrochlore cubic ice,
regulating disorder, and how to deliberately integrate disorder into in which the O–H bond distances can be referred to as spin orientations,
crystal structures. Disordered materials seldom embrace absolute and they are thereby termed spin ice states24,25,30 (Fig. 2a, right).
randomness, but they predominantly manifest local correlations4. The concept of frustration has garnered attention in ionic conduc-
In this Review, our primary objective is to expand the concept tion31–33. Ba doping in CaF2 introduces geometric frustration, leading
of disorder within cathode materials for alkali-ion batteries. We aim to structural disorder in the fluorite lattice due to the incompatibility
to introduce a novel strategy involving characterizing the disorder, of interatomic forces arising from the differences in size between Ca
understanding the interactions generating it and fine-tuning these and Ba ions. This results in a fivefold increase in ionic conductivity,
parameters to push the electrochemical properties of cathode mate- facilitated by snake-like transport pathways and an expanded lattice31.
rials closer to their theoretical limits, without impacting reversibility. Medium-entropy and high-entropy alloys are materials composed of
Initially, we revisit different cases of disorder in materials, drawing five or more metals; this composition increases the configurational
upon insights from well-established examples in other fields. Subse- entropy, resulting in enhanced strength, stability and reactivity34.
quently, we provide an overview of the various factors within cathode These alloys display intriguing mechanical properties, such as Shockley

Nature Reviews Chemistry

Review article

Disorder introduced during synthesis

Symmetric, asymmetric distortion Li/TM mixing Intralayer disorder

Jahn–Teller/pseudo Jahn–Teller distortion Steric/magnetic interaction Electrostatic interaction

Li/TM site disorder Interlayer disorder

Steric/electrostatic interaction Steric/electrostatic interaction

Disorder that arises in response to electrochemical stimuli

TM migration Oxygen evolution

CFSE/electrostatic interaction Bond covalency/dimerization

Alkali ion
TM
Oxygen
Substituted anion
Crystal water
Vacancy

Fig. 1 | Schematic summarizing the types of disorder in metal oxide cathode The disorder leads to an atomic arrangement, which gives rise to new or
materials and their chemical origins. Disorder within metal oxides for enhanced properties. However, the relationship between the disorder and
cathode materials fall into two categories: disorder introduced during synthesis the resulting properties is not fully understood. The dashed circles and ovals
(top) and disorder that arises in response to electrochemical stimuli (bottom). are guides to the eye and indicate where changes have occurred or will occur.
Each type of disorder is described in bold text, and the interactions that CFSE, crystal field stabilization energy; TM, transition metal.
induce the disorder are described in normal text below the relevant bold text.

partial dislocations — a crystallographic defect in which a disloca- Another instance of correlated disorder is observed in the anti-
tion splits into two partial dislocations, creating a stacking fault that perovskite Li2FeSO (ref. 37), in which configurational undercon-
alters the material’s properties — which are determined by the balance straint plays a considerable role. According to the point-charge
between the elastic strain energy and the stacking fault energy, and electrostatic model — which calculates the electrostatic energy
are properties associated with the energetic cost of creating stacking considering each ion as point-charge with a formal charge — the
faults35. An inspection of the plasticity using a medium-entropy alloy, octahedra centred on oxygen atoms (OLi 4Fe 2) should favour a
NiCoCr, showed a discrepancy between theoretical and experimental trans configuration, in which Fe2+ ions occupy opposing octahe-
results, as a finite dissociation was observed, rather than an infinite one. dral vertices to minimize electrostatic interactions by maximizing
The discrepancy in the experimental value was attributed to a dominant their separation. However, its X-ray total scattering data are better
effect from the positive stacking fault free energy36. described by a structural model with cis-OLi4Fe2. Computational

Nature Reviews Chemistry

Review article

studies predict that the anion-polar coordination contributes to Origin of disorder in cathodes
the structural energy by altering the anion–cation distance, reduc- Ionic crystals consist of assemblies of polyhedra composed of cations
ing the net electrostatic energy of the system. The stronger oxy- and anions. The arrangement of atoms in crystalline materials can be
gen polarization of cis-OLi4Fe2 indicates its stability relative to the predicted by the numerous principles used in crystallography and
trans configuration. As a cis-OLi4Fe2 octahedron allows for four materials science39–41. Most cathode materials are ionic crystals: layered
possible configurations with its neighbouring octahedra (Fig. 2b), TM oxides with rocksalt structures, which consist of close-packed ani-
configurational underconstraint leads to long-range disorder37. onic sublattices with FCC stacking in which metal centres and oxygen
The structure of α-tetragonal boron has also been debated due to centres are octahedrally coordinated. These MO6 (M stands for metal)
its simultaneous contributions from frustrated overconstraints octahedra are arranged in edge-sharing geometries. Their intriguing
and configurational underconstraints. The overconstraint arises performances arise from alkali-ion diffusion, which originates from
from unsatisfied covalent conditions whereas the configurational the distinct layered arrangement of TM oxides and alkali ions in an FCC
underconstraint is influenced by the many possible arrangements oxygen framework with rhombohedral distortion42. Different layered
of partially occupied interstitial sites38. structures were categorized with nomenclature combining a letter that
The examples mentioned above highlight the presence of cor- defines the alkali ion site (O is octahedral, and P is trigonal prismatic)
related disorder in various material families and their influence on and a number referring to the number of MO2 layers in the repeating
material functionality. This concept provides us with a fresh perspec- cells. The conditions for structural stability and cationic ordering are
tive, allowing us to shift our view of disorder from mere defects within precisely understood40,43–45. However, the materials display multiple
a periodically ordered system to an important design parameter that types of disorder depending on local and long-range interactions3,46.
helps to enhance a material’s properties. The key components of cor- In addition, the ordered layered structures exhibit a weak structural
related disorder — degrees of freedom, interactions between degrees order during the extraction/intercalation of alkali ions47, with aniso-
of freedom and lattice — can be linked to chemical parameters, ena- tropic lattice changes or defects in the TM and oxygen sites. To use
bling the control of correlated disorder for practical material design disorder as a design element to enhance a certain property of a mate-
strategies4. Based on these aspects, we investigate the disorder of rial, the chemical origins of the disorder have to be understood. In the
cathode materials in the following sections. We explore the under- following section, we discuss different origins of structural disorder
lying interactions and chemistry that drive disorder, and discuss in cathode materials.
the strategies for tuning disorder to enhance the electrochemical
properties of cathode materials in alkali-ion batteries. Size effect
The size effect plays a crucial role in predicting structural stability. The
foremost effect stabilizing the layered structure is the ionic radius
ratio between the alkali ions and TMs, permitting independent bond
a Water ice Spin ice lengths between AO6 (A, alkali ion) and MO6 octahedra48 (Fig. 3a).
The formation of a layered structure requires the presence of TMs
of considerably small size relative to the alkali ions, exemplified by
r(Li+) = 0.76 Å and r(Na+) = 1.02 Å (Fig. 3a). This size disparity underpins
Hydrogen the stabilization by impeding TM migration, particularly when Na+ is
Oxygen used as the alkali ion. TMs such as Co3+ (r = 0.545 Å) and Ni3+ (r = 0.56 Å)
Electron spin tend to be stabilized in Li-incorporating layered structures, enabling
the wide application of these structures as cathode materials49,50. The
presence of Ni2+ (r = 0.69 Å) in LiNiO2 due to the instability of Ni3+ thus
b explains Li/Ni mixing.
Beyond its role as a stabilizer in layered structure, the size effect is
a pivotal consideration in atomic substitutions. A successful substitu-
tion requires an ionic size difference of less than 15%39. The cation:anion
radius ratio is equally important, as it predicts the coordination num-
ber and shape of the polyhedra, thereby predicting the atomic struc-
ture. Pauling’s first rule predicts that for a rocksalt structure to be
structurally stable, a value for the TM:O radius ratio (rTM:rO) in the
Configurational underconstraint → Long-range disorder
0.42 ≤ rTM:rO ≤ 0.72 range must be respected40. The size effect offers
Fig. 2 | Examples of correlated disorder. a, Water ice displays an ordered guidance on the appropriate selection of substituting atoms to allow
arrangement in which water molecules adopt specific orientations. In certain or disallow structural order.
situations, such as at low temperatures or under high pressure, the hydrogen
bonds between water molecules may be subject to geometric frustration.
Electrostatic interactions
Pyrochlore spin ice is approximately equivalent to water ice in which the
Electrostatic interactions constitute another critical parameter govern-
length of the O–H bond can be referred to as a spin orientation under magnetic
interactions. b, Possible local configurations of cis-OLi4Fe2 in antiperovskite
ing crystal structure, and they are often interconnected with the size
Li2FeSO. As cis-OLi4Fe2 octahedra allow the adoption of four possible effect. Earlier works claimed the necessity of maintaining overall charge
configurations, the structure exhibits long-range disorder by configurational neutrality in ionic crystals40. Pauling’s second rule mathematically
underconstraint. Part a adapted with permission from ref. 24, Springer Nature defines the criterion of electroneutrality between the metal and oxygen
Limited. Part b is adapted from ref. 37, CC BY 3.0 DEED (https://creativecommons. to ensure the structural stability of a metal oxide. When a TM is utilized
org/licenses/by/3.0/). as the cationic centre, the chemical composition of the material can be

Nature Reviews Chemistry

Review article

a Size effect b Electrostatic interaction

Disorder formation to satisfy Interlayer repulsion weakened
charge neutrality by inserting guest molecules
TM slab rTM

Electrostatic
repulsion
A slab rA

y
z z

x z x x
y y

c Superexchange interaction d Jahn–Teller distortion Pseudo Jahn–Teller distortion

Pristine After Li/Ni mixing

Pristine Elongation Off-centering

x2 – y2 Electron
eg eg
z2 x2 – y2 from O 2p
z2
xy
t2g
90° interaction 180° superexchange interaction xy xz yz t2g
xz yz

TM Oxygen Crystal water

A Substituted anion Vacancy

Fig. 3 | Chemical origins of disorder in metal oxide cathode materials. 180° configuration (right) rather than 90° (left), which describes one of the
a, A considerable difference in size between a transition metal (TM) and reasons that Li/Ni mixing is favourable in LiNiO2. d, Orbital splitting with
oxygen atoms is critical for maintaining the layered structure, due to distinct Jahn–Teller distortions remove orbital degeneracy to minimize electrostatic
cation–anion bond lengths. b, Electroneutrality must be maintained within an interactions. The increase in covalency via charge transfer from a ligand
ionic crystal. Accordingly, electrostatic interactions lead to the formation of orbital with pseudo Jahn–Teller distortions eliminates pseudodegeneracy.
atomic vacancies or charge transfer to keep an electrostatic neutrality (left). The dashed circles are guides to the eye and indicate where changes have occurred
Additionally, interlayer repulsion and attraction influence the stacking order or will occur. A, alkali ion; rA, alkali element ionic radius; rTM, transition metal
(right). c, Superexchange interactions stabilize TM–O–TM bonds in a linear ionic radius.

altered. For instance, Ni3+ in LiNiO2 is easily reduced to Ni2+, resulting be weakened by the insertion of guest molecules. A prominent exam-
in the Li-deficient Li1−x−yNi1+xO2 (ref. 51). In layered LiNixCoyMn1−x−yO2 (NCM), ple of this scenario can be found in the presence of interlayer water,
the introduction of Mn preserves the Li:TM ratio within the layered frequently encountered in clay materials and various TM oxides. In
structure by facilitating the reduction of Ni3+ to Ni2+ at the expense these cases, hydrogen bonds form between the water molecules and
of Mn4+, as the Mn3+/4+ band lies above the Ni2+/3+ band52. Furthermore, the metal hydroxides, altering the prevailing electrostatic interactions,
the charge balance should be affected by the atomic composition as resulting in an expanded interlayer space64–66.
well as the atomic arrangements. In Ni-rich NCM, experimental stud-
ies reported controversial results regarding the oxidation state of Mn Superexchange interactions
(whether Mn3+ can exist instead of Mn4+)53,54, and a computational study Size, electroneutrality and coordination number are important con-
predicted the reduction of Mn4+ to Mn3+ among Mn clusters, which stituents of Pauling’s rules for predicting structural stability. However,
supported experimental observations55. Aside from charge transfer numerous materials do not fully conform to Pauling’s rules41. Therefore,
between ions, charge neutrality can also be achieved through the crea- it is necessary to consider a variety of interactions, including magnetic
tion of site vacancies56. These vacancies arise through atomic substitu- interactions, to understand the disorder in materials. The rocksalt
tions involving cations or anions with different valences from the initial layered oxide structure is composed of MO6 octahedra, with shared
constituents57–62 (Fig. 3b, left; black square). On a larger scale, layered edges to form layers of TM slabs. Within this arrangement, each TM is
stacking arrangements are stabilized by electrostatic interactions bridged by oxygen, resulting in an M–O–M bond angle of 90° (Fig. 3c,
(Fig. 3b, right). This is evident when a TM oxide is paired with alkali left). When a TM migrates to an alkali ion site, the M–O–M angle changes
ions63; in such a situation, the relevant electrostatic interactions can to 180° (ref. 67) (Fig. 3c, right). Superexchange, involving magnetic

Nature Reviews Chemistry

Review article

interactions via ligand orbitals, is enhanced in the linear configuration, force, they often play a role as competing or co-existing secondary
through the σ bond between the metal eg orbital and the O p orbital, effects that induce disorder.
following the Goodenough–Kanamori rules68,69. The Goodenough–
Kanamori rules help to predict the strength and type (antiferromag- Jahn–Teller distortion
netic or ferromagnetic) of magnetic interactions between two ions The Jahn–Teller effect ( JTE) and the pseudo JTE (PJTE) observed in
based on the angle between them and the nature of their orbitals. Within TM oxides are two of the major interactions that generate disorder
this context, Ni2+ and Mn4+ adopt a robust Ni2+–O–Ni2+/Mn4+ configura- in TM oxides, and they are related to the spontaneous distortions of
tion due to the presence of unpaired electrons. By contrast, Co3+, which MO6 octahedra that remove degeneracy and pseudo-degeneracy in
lacks unpaired electrons, has a weaker superexchange, thus suppress- energy levels71. Their degrees of freedom correspond to distortion
ing Li/Ni mixing70. It is worth noting that magnetic interactions make coordinates, representing the directions of spontaneous distortions
a weaker contribution than other factors, such as electrostatic forces. that serve to lower the potential energy of the electronic states.
Although magnetic interactions may not serve as the primary driving The JTE in MO6 octahedra is accompanied by elongation or con-
traction of the M–O bond along the z-axis72 (Fig. 3d, left). As the eg orbit-
als have better overlap with the O 2p orbitals, the TMs with occupied
a CFSE and kinetics of migration eg orbital most affected in cathode materials are Ni3+ and Mn3+. LiMnO2
TM migration energy barrier: or NaMnO2 with Mn3+ undergo cooperative Jahn–Teller distortion due
octahedral site stabilization energy (CFSEOh − CFSETd) to charge, orbital and magnetic interactions between the neighbour-
ing JTE-distorted centres, which then crystallize into orthorhombic
or monoclinic structures73,74. In a system with multiple TM centres,
CFSEOh cooperative distortion can be suppressed by neighbouring TMs. The
JTE also causes disproportionation of Mn3+ to Mn2+ and Mn4+ owing
to the acid–base and/or redox reaction between the axial oxygen of
distorted MO6 octahedra and H+ ions, leading to Mn dissolution75–77.
CFSETd
However, the PJTE leads to asymmetric distortions, such as off-
z
centre ferroelectric displacement in perovskite materials, resulting
in increased covalent bonding overlap78 (Fig. 3d, right). Given that
x electrostatic interactions typically account for general structures of
y
rocksalt materials, the PJTE plays a crucial role in several structural
distortions, especially in the cases in which a d0 TM is incorporated45.
b Oxygen oxidation and evolution Li-excess, disordered rocksalt (DRX) contains distorted octahedra due
Oxygen oxidation Oxygen evolution
to the large charge disparity between Li and the TMs. Computational
O2
investigations anticipate that d0 TMs in DRX can undergo displace-
Fermi level e– ment from the octahedral centre to buffer the distortion of adjacent
edge-shared octahedra, thus lending stability to the DRX within the
Overlap between TM d
cubic lattice79. This off-centred site distortion exhibits parallels with
O 2p and TM d
orbital by covalent phenomena observed in ferroelectric materials48, allowing the possi-
bond bility of PJTE distortions in DRX78. Moreover, a study suggests that the
PJTE could be a general structural origin of electrochemical instability
O 2p in layered cathode materials, offering a new perspective on the cor-
relation between local distortion and the intrinsic electrochemical
properties of cathode materials80.

Oxygen TM Crystal field stabilization energy

Vacancy As layered cathodes charge, alkali ion vacancies emerge in the inter-
layer space. Consequently, oxygen atoms directly face one another
Fig. 4 | Chemical origins of the disorder of metal oxide cathode materials in across the vacated interlayer space and form strong covalent bonds
response to electrochemical stimuli. a, The stability of transition metals (TMs) with the TMs, resulting in an anisotropic bond distribution. Therefore,
in atomic sites is predicted by crystal field stabilization energy (CFSE). When TM migration to vacant alkali ion sites is favoured to make the charge
TM migration occurs through the O–T–O (where O is the octahedral site and distribution homogeneous81. Theoretical studies predict that a TM
T the tetrahedral site) pathway, it is facilitated by a low TM migration energy will migrate through a tetrahedral site on its way from the original
barrier or a low octahedral site stabilization energy, which can be calculated
octahedral site to a vacant octahedral site within the interlayer (O–T–O
from the differences between the CFSE of the TM at the octahedral site and at
diffusion) (Fig. 4a). The TM migration to the interlayer tetrahedral site
the intermediate tetrahedral site. The schematic shows the O–T–O migration
is facilitated by an associated Li migration to a tetrahedral site inside
pathway of a TM from a TM site to a Li site after delithiation. b, Charge transfer
from the O 2p to the TM d band leaves holes in the O 2p band (left). Oxidized
an interlayer neighbouring the TM layer with the vacancy, forming a
oxygen subsequently interacts with a neighbouring oxygen atom, ultimately Li–TM dumbbell82–85. Greater stabilization of the original octahedral
leading to the irreversible loss of O2 from the material (right). The dashed oval site increases the kinetic barrier for migration to the tetrahedral site,
is a guide to the eye and indicates where changes have occurred or will occur. efficiently curbing TM migration. This stabilization can be predicted
CFSEOh, octahedral CFSE; CFSETd, tetrahedral CFSE. Part a from ref. 159, Springer by calculating the crystal field stabilization energy (CFSE). TMs with
Nature Limited. half-filled or filled t2g orbitals, such as low-spin Co3+/4+, Mn4+, Ni3+/4+ and

Nature Reviews Chemistry

Review article

high-spin Cr3+, show high stabilization in the octahedral site82,83. This capacity of around 140 mAh g−1 (ref. 92). To achieve a higher capacity, the
observation reinforces the widespread use of Co, Mn and Ni in cath- use of Ni and Mn as alternative redox centres to Co has been explored.
ode materials48. The octahedral site stabilization energy, calculated However, LiNiO2 suffers from sluggish Li+ diffusion caused by Li and Ni
from the difference between the CFSEs of octahedra and tetrahedra, mixing93, and LiMnO2 is characterized by an inherent structural instabil-
indicates higher stability in the octahedral site in the following order: ity induced by Jahn–Teller distortion and Mn dissolution77,94. A break-
Co3+ > Ni3+ > Mn3+ (ref. 52). through emerged with the compositional disorder in NCM, in which
both Ni and Mn were incorporated into LCO95. Consequently, NCMs
Oxygen oxidation and evolution have been widely commercialized as cathode materials, and further
The redox reactions of cathode materials depend on the energy levels of investigations have been dedicated to enhancing their performance by
each redox couple, which are determined by their electronic structures. increasing their Ni content96. The compositional and configurational
Detailed electronic configurations and their theoretical background disorder altered local configurations generating specific long-range
are described elsewhere86–88. The electronic level is complex, stem- patterns, which activated additional capacities from the oxygen redox
ming from diverse factors. Notably, the hybridization between the chemistry in Li-based and Na-based systems97,98. Subsequent investiga-
metal d band and the O 2p band splits the eg energy level into bonding tions aimed to mix alkali ions and TM sites, yielding DRX cathodes, to
and antibonding states, exhibiting oxygen and metal characteristics, allow for more options for TM oxides with higher energy densities99,100.
respectively. Little hybridization between the metal and oxygen bands Beyond Li-ion and Na-ion batteries, disorder is providing new strategies
gives rise to nonbonding states. The electrochemical deintercalation to solve the sluggish diffusion issues of multivalent ion batteries101.
of alkali ions considerably alters the electronic structure. At a high It is thus necessary to revisit other derivatives from rocksalt struc-
state of charge, the metal d band can overlap with the O 2p band thus tures or different structures, such as birnessite or anatase, that contain
increasing the covalency of the M–O bond. The modified covalency increased disorder for next-generation cathode materials45,101–104.
facilitates the charge transfer from oxygen to the empty metal d band,
eventually leaving holes in the O 2p band. This is known as ligand-hole Modulating the redox mechanism. A redox couple is a pair of chemical
chemistry, which involves creating electron holes in metal complexed species involved in oxidation and reduction reactions under equilib-
ligands, leading to novel reactivity and enabling unique oxidation rium at a specific potential called the redox potential. In Li-ion bat-
states and coordination geometries89 (Fig. 4b, left). Along with the teries, TMs within cathode materials act as redox couples with their
characteristics of the local environment, the oxidized oxygen evolves redox potential synchronized with Li+ extraction and intercalation. The
to form O–O dimers or even leads to O2 release — leaving behind oxygen voltage (V) at equilibrium corresponds to the difference in the chemical
vacancies and structural distortions with the possibility of side reac- potential of Li (μLi) between the cathode and the anode, which can be
tions (Fig. 4b, right). The oxygen stability is even more intricate within calculated using the Nernst equation (equation 1):
alkali-ion-rich materials, with the involvement of unhybridized O 2p
nonbonding states in oxygen redox reactions. One theoretical study μLi (Cathode) − μLi (Anode)
Veq = − (1)
attempted to predict the stability of oxygen with the number of holes e
per oxygen (h0)90. According to this study, the h0 should be kept below
one-third to avoid oxygen release, which emphasizes the importance in which e is the elementary charge. The Li chemical potential of the
of the TM redox couple before oxygen oxidation to limit the number of cathode is related to the Gibbs free energy (Gcat) with respect to Li
oxygen holes. concentration (χLi) by the following equation:

Impact of structural disorder in cathodes μLi (Cathode) =

dGcat
(2)
Disorder can be linked to chemical interactions, as previously men- dxLi
tioned in the context of metal oxide cathode materials. Therefore, it
is important to understand how to manipulate chemical parameters Accordingly, the shape of the voltage curve is the slope of the Gibbs
to deliberately introduce disorder, to use disorder as a key material free energy versus the Li concentration. The voltage plateau indicates
design parameter. Here, we present several examples of disorder in the coexistence of two phases in equilibrium at a particular voltage,
layered and 3D TM oxides, not only in conventional rocksalt-type oxides given that the chemical potential is constant along the common-tangent
but also in different structures, which demonstrate an apparent link construction between Gibbs free energy minima in this region. A solid-
between controlled disorder and improved electrochemical properties. solution-type behaviour manifests as a sloping voltage curve between
Furthermore, we explore specific considerations and area in which cau- high and low Li content resulting from the interlayer Li site vacancy
tion is advised associated with disorder, drawing upon a few instances disorder86,105. Therefore, the voltage profile is highly related to the
in which it gives rise to adverse or opposed material properties. structure of the intercalation material as well as to the interactions
between Li+ and the local environment inside the electrodes, such
Increasing energy and power density as coordination, bond strength with anions and site occupancies
Since the discovery of the first Li-ion battery cathode material in the nearby106. In the case of an ordered layered structure, the slope of the
1980s91, numerous investigations have been dedicated to elevating voltage is dependent on the Li–Li Coulombic interactions, which are
the energy density of cathode materials. Interestingly, this pursuit has screened by oxygen–oxygen repulsion to provide a moderately sloping
led to an increase in disorder within the materials, which is at odds with profile. Given that the Li site energy, which is described by the Li–TM
the conventional belief that disorder undermines battery performance. interactions, is constant in an ordered layered structure, the interac-
LiCoO2 (LCO) is an extensively studied cathode material for Li-ion bat- tions between adjacent Li sites, or effective Li–vacancy interactions
teries and the first to be commercialized91. Early studies revealed that ( JLi-Va), predict the steepness of the voltage slope. Along with Li/TM
LCO allows reversible shuttling of 0.5 Li+ only, equivalent to a specific cation disorder, a larger statistical distribution of TM environment

Nature Reviews Chemistry

Review article

around Li sites leads to non-uniform Li–TM interactions, which results Cation site mixing thus offers a potential approach for adjusting
in a steeper voltage slope during extraction/intercalation of Li+ due the voltage profile of DRX cathode materials towards increased energy
to the combination of Li–TM and Li–vacancy interactions107 (Fig. 5a). densities. This configurational disorder leads to a high-temperature
metastable phase that can be obtained by the high balance of entropy
(TΔS) and/or change of energy (ΔE) to accommodate the disorder48.
a Voltage slope change due to Li/TM disorder
The high TΔS can be achieved by high temperature or high-energy
Li/TM ordered Li/TM disordered ball milling108. As the disordered phase involves site mixing between
cations with large differences in radii and electronegativities, the mate-
rials undergo octahedral distortions. This highlights the importance
of the d0 TM, which can stabilize electrostatic interactions through
Voltage

Voltage

Slope Slope off-centred distortion, potentially due to the PJTE79,80, as discussed in

the previous section. d0 TMs can lower energy costs for disorder by
accommodating the distortions. The DRX structure, resulting from
the cooperative correlation of disorder, can be identified by X-ray dif-
fraction (XRD) patterns. For DRX, Fig. 5a shows an increased net voltage
slope due to the combined effects of site energy distribution and JLi-Va
1.0 0.5 0 1.0 0.5 0
as cation disorder creates diverse local Li environments. However,
x in LixTMO2 (0 ≤ x ≤ 1) x in LixTMO2 (0 ≤ x ≤ 1)
substantial Li+ extractions vacate the octahedral sites of the Li4TM0
clusters that are face-sharing with tetrahedral sites. This leads to the
b Generation of nonbonding O 2p by orbital configuration remaining Li+ ions becoming energetically stable within the tetrahedral
Voltage sites, requiring higher voltages for their removal. Given that the aver-
A–O–A configuration EF
TM d age voltage is thermodynamically dependent on the Gibbs free energy,
Li+ removal at high voltages results in voltage decreases in other
ranges107.

Nonbonding O 2p
Activating anionic redox reactions. Activation of the oxygen redox
chemistry in cathode materials to enhance the energy density by using
an additional redox couple has garnered considerable attention109.
Bonding O 2p Initial investigations demonstrated that the oxygen oxidation of LCO
is a consequence of the electron transfer from oxygen to the oxidized
TM, thereby not affecting the capacity89. However, Li-rich metal oxides,
with the excess Li occupying the transition sites, exhibit an increased
energy capacity. The atomic site disorder, introduced in layered LiMO2-
c Oxygen oxidation coupled with TM vacancies
type structures, is correlated with each other for a homogeneous
Oxygen A
Voltage distribution of electrostatic interactions between cations, resulting
TM d EF
in a honeycomb superstructure (Li2MO3) with monoclinic symmetry
discernible from the characteristic low-intensity superstructure peaks
Oxygen A in XRD patterns. Although the debate regarding Li-rich layered mate-
Oxygen B rials and whether they are the composites of Li2MnO3 and LiMO2 or a
TM solid solution has been ongoing110, an important point that has to be
Oxygen B
migration
considered in this controversy is the coherence length of the Li2MnO3-
Bonding O 2p type domain, which is possibly the length of the pattern generated from
the perspective of correlated disorder, depending on the synthesis and
chemical composition111. In such local environments, oxygen atoms
in the octahedral sites are coordinated to four Li and two TM centres
A TM Oxygen generating an Li–O–Li configuration, instead of three Li and three TM
centres, such as in the conventional layered cathodes112. According to
Fig. 5 | Disorder improves energy density by modulating cationic and anionic theoretical investigations, Li 2s orbitals are not predicted to hybridize
redox mechanisms. a, Modulation of the transition metal (TM) redox mechanism. with O 2p orbitals, due to the large energy gap, so the Li–O–Li configu-
During electrochemical cycling, the Li/TM disordered structure exhibits a steeper ration generates nonbonding O 2p orbitals pinned at an energy state
voltage curve (blue curve, right) than the Li/TM ordered structure (red curve, left).
higher than the O 2p bonding states (Fig. 5b). Once the nonbonding O
b,c, Activation of oxygen redox reactions. The atomic illustration describes local
2p states are at a comparable energy level as the TM band, oxygen redox
configurations around oxygen in an octahedral site generating an unhybridized
is activated at highly delithiated states, providing higher energy densi-
nonbonding O 2p orbital, which contributes to additional capacities (part b). Oxygen
redox can also be activated by a dynamic process with cation migration. More
ties87,113. This concept can be extended to high Li:TM and O:TM ratios as
electronic states close to the Fermi level (EF) for oxygen centres adjacent to vacancies well as the Li/TM site mixtures in DRX114–118. Although Na-based layered
(oxygen A), which allow for the activation of oxygen redox reactions (part c). cathodes also exhibit oxygen redox behaviour119,120, this is unlikely
The dashed circles and ovals are guides to the eye and indicate where changes have to be described by Na–O–Na configurations, as Na cannot occupy
occurred or will occur. Alkali ion, A. Part b adapted from ref. 113, Springer Nature the octahedral sites in TM slabs due to size difference (r(Na+) = 1.02 Å
Limited. Part c adapted from ref. 125 under a Creative Commons licence CC BY 4.0. > r(Li+) = 0.76 Å). The unhybridized O 2p state is produced as a result

Nature Reviews Chemistry

Review article

of the substitution with other cations of smaller sizes (such as can screen the attraction between alkali ions and oxygen, thereby
Mg, Zn or Cu) or simply cationic vacancies, stemming from the size promoting ion diffusivity even for highly polarizable divalent ions
effect and electrostatic interactions among cations121–124. (such as Mg2+ and Zn2+) that exhibit notoriously typical sluggish diffu-
Furthermore, the activation of the oxygen redox chemistry is sion kinetics138. Currently, intriguing results on ion diffusivity and ion
coupled with TM vacancies generated during charge and discharge intercalation kinetics have been reported for various TM chalcogenides
cycling. Investigations of Li1.17−xNi0.21Co0.08Mn0.54O2 in which the density and layered inorganic materials using interlayer molecules139–141, and it
of states with X-ray absorption analysis were coupled demonstrated is thus important to expand this application to other layered materials.
that TM migration was triggered by Li vacancies during charging, The cubic rocksalt has received relatively limited atten-
and it is relevant to the change of oxygen redox potential in Li-rich tion due to its poor Li+ diffusion kinetics compared with layered
layered cathode materials125. Moreover, the formation of vacancies
in the TM site triggers oxygen redox. Notably, oxygens close to cation
vacancies in the TM slab arising from TM migration (Fig. 5c, oxygen A) a Charge screening effect by the expansion of interlayer spacings
can more easily activate oxygen redox due to the electronic states
closer to the Fermi level than those of oxygens close to the migrated
TMs (Fig. 5c, oxygen B), as confirmed by density of states calculations
(Fig. 5c). The migration of TMs or substituted cations is governed
by the electrostatic repulsion between oxygen and cations, as pre-
dicted by computational studies. This explains why oxygen redox is not
activated by Mg2+ or Zn2+ substitution in Li-stoichiometric cathodes,
unlike Na-stoichiometric or deficient cathodes, given that substituted z
cations cannot migrate into vacant interlayer sites126. These strategies
show that modifying the cationic configuration by modulating the TM x
y
site occupancies through chemical compositional changes is a viable
approach for activating anionic redox. Studies have discovered that
b
reversible oxygen redox occurs in rhombohedral layered structures,
Li3TM1 cluster in layered oxide Li4TM0cluster in DRX
including LiNiO2, LiNi1/3Co1/3Mn1/3O2 (NCM111) and LiNi0.8Co0.15Al0.05O2
(NCA)127–129. Moreover, contrary to the consensus on oxygen oxidation
in Li2MnO3, an experimental study indicated the absence of lattice
oxygen redox at all cycling stages130. These unexpected findings suggest
that cationic configurations are not the only driving force activating
oxygen redox, but structural disorder should also be considered when
z
explaining reversible and irreversible oxygen redox. This implies that
a fundamental understanding of local correlation for the disorder is
x
essential. y

Facilitating ion diffusion. Layered materials exhibit decent ion c TM vacancy-mediated alkali ion diffusion
mobility through simple 2D pathways inside the interlayer space52.
As a result, modifying the interlayer space is an efficient approach to TM

improving the electrochemical performance. One example is the pres- Alkali ion
ence of intercalated water or organic molecules inside the interlayer Oxygen
space131,132. Interlayer crystal water, which is often coordinated to alkali
Manganese
ions through electrostatic interactions, increases the interlayer space
Substituted anion
and leads to a disordered arrangement of the hydrated cations, which is
difficult to characterize. This interlayer engineering allows for intrigu- Crystal water
ing characteristics, including enhanced ion mobility. For instance, in y z Tetrahedral site
several layered Mn, Ti or V oxides, an increased interlayer space and a
Vacancy
reduced electrostatic interactions between the charge carrier ions and x
the lattice ions reportedly enhance the electrochemical performance131.
Figure 6a shows this unusual interlayer configuration, which impacts Fig. 6 | Disorder facilitating ion diffusion. a, The presence of interlayer water
alkali ion–oxygen electrostatic interactions and ion mobility. The dif- reduces electrostatic interactions between an alkali ion and an anionic sublattice by
widening the interlayer space (double arrows) and by inducing a charge-screening
fusion of alkali ions increases with the interlayer spacing because the
effect (dashed lines), which accelerates alkali-ion diffusion. b, The illustration
electrostatic attraction of oxygen decreases133. Consequently, the large
shows the Li3TM1 cluster in an ordered layered structure (left), which has three
interlayer space of NaxMnO2·nH2O birnessite (~7 Å), as shown by XRD
Li atoms and one transition metal (TM) around tetrahedral site, and the Li4TM0
patterns, looks advantageous for alkali ion diffusion. Moreover, the cluster in a disordered rocksalt (DRX) structure (right), which only has four Li
interlayer space can be further enlarged to 9.6 Å by increasing the water atoms around its tetrahedral site due to Li/TM site disorder. The Li4TM0 cluster in a
content or substituting water with the larger tetramethylammonium DRX structure renders O–T–O (where O is the octahedral site and T the tetrahedral
ions, thereby improving alkali ion diffusion134. This strategy is more site) Li+ hopping more favourable due to alleviated electrostatic interactions in the
generalized in ‘beyond Li-ion’ systems, such as batteries with aqueous tetrahedral site. c, Diffusion through a TM vacancy, as TM vacancies provide a low
electrolytes or multivalent ion batteries64,135–137. Notably, crystal water barrier for migration and increase the diffusivity of alkali ions.

Nature Reviews Chemistry

Review article

cathode materials48,142. In layered and DRX materials, the mechanism of Improving reversibility
Li+ migration towards neighbouring vacant octahedral sites involves In a layered cathode, the extraction of alkali ions gives rise to alkali ion
the use of face-shared tetrahedral sites as local diffusion pathways, vacancies and TM oxidation, both of which thermodynamically drive
via the O–T–O path (where O is the octahedral site and T the tetrahedral the in-plane rearrangement of the remaining alkali ions and vacancies.
site). These tetrahedral sites are clusters of Li and TM species occupy- This subsequently triggers atomic or electronic phase transitions146–148.
ing four nearby face-sharing octahedra, so-called Li xTM4−x (0≤x≤4) The thermodynamically driven alkali ion-vacancy ordering has been
clusters, which strongly influence the Li+ migration barrier. In principle, corroborated through calculations of formation energy of alkali ion-
Li3TM1 or Li1TM3 are the only clusters that exist in the layered materi- vacancy arrangement at a specific composition in LixCoO2 with diverse
als. Given the fact that Li+ diffusion within these clusters is affected by alkali ion-vacancy arrangement models149,150. Furthermore, the reduced
electrostatic interactions with adjacent TM slabs, only Li3TM1 clusters screening effect of alkali ions between negatively charged oxygen lay-
favour diffusion (Fig. 6b, left). Contrary to this, the Li/TM site mixing ers induces an increase in the c-lattice parameter due to strong O–O
in DRX, allowed by the presence of d0 TMs buffering the distortion of repulsion. The continued removal of alkali ions destabilizes the cathode
adjacent edge-sharing octahedra, enables the emergence of a variety material, finally triggering the gliding of O–TM–O slabs that trans-
of clusters, including Li4TM0 (Fig. 6b, right), with reduced electrostatic form from trigonal prismatic to octahedral packing to minimize the
repulsion in the tetrahedral site. According to computational studies, instability of the resulting O–O repulsion150.
the presence of these Li4TM0 clusters — also known as 0-TM channels — In addition, TM migration and oxygen evolution — both of which
in DRX facilitates local Li+ hopping. A subsequent excess of Li and are highly correlated with the interactions previously discussed —
compositional fine-tuning to change the short-range order allow a occur during battery operation. During the charging of Li-ion battery
continuous 0-TM percolation network to be established143. cathodes, TM migration takes place towards the interlayer space via
The introduction of vacancies, which emerge to preserve electro- an O–T–O diffusion pathway. This migration is governed by a migra-
neutrality after anion or cation substitution in pristine materials, repre- tion energy barrier, which can be determined with the CFSE and also
sents another effective strategy for accelerating the diffusion kinetics depends on multiple factors, such as geometry, ligand character and
of alkali ions in 3D structures56,57,144 (Fig. 6c). Additionally, this strategy spin pairing energy. Collective TM migration, closely linked to the inter-
efficiently increases the diffusivity of multivalent ions, known for actions with Li and O vacancies, results in phase transitions to spinel
their poor diffusivity due to strong electrostatic interactions with the structures that are accompanied by voltage decays and the genera-
anionic sublattice of the oxide cathode. The cation-deficient anatase tion of Li-deficient cubic rocksalt phases that restrain the extraction/
TiO2, Ti0.78□0.22O1.12F0.40(OH)0.48 (□ = Ti vacancy), containing Ti vacan- intercalation of Li+ by blocking the diffusion channel151,152. Moreover,
cies to ensure electroneutrality after F substitution, effectively over- the lattice mismatch between the original layered phase and spinel/
comes the limitation of low Mg2+ diffusivity using vacancy-mediated rocksalt is one of the main reasons for strain evolution during cycling,
diffusion, as shown by transmission electron microscopy and X-ray which leads to the formation of cracks in the electrodes. Meanwhile,
total scattering refinement101. A density functional theory (DFT) cal- oxygen oxidation leads to the formation of unstable oxidized oxygen
culation revealed that Ti vacancies serve as thermodynamically stable or superoxide species, which tend to result in oxygen vacancies upon
sites for intercalated multivalent ions with higher coordination and can evolution. Oxygen vacancies tend to aggregate to form cavities, low-
decrease the migration barrier by tuning the electrostatic interactions ering mechanical strength and worsening cracks153. Finally, oxygen
between the multivalent ions and the TM. It is worth noting that the release causes unwanted parasitic reactions with the electrolyte dur-
vacancies should be homogeneously distributed to create an intercon- ing cycling. Therefore, suppressing phase transitions or enhancing
nected network that facilitates the diffusion of multivalent cations. structural change reversibility is crucial for boosting electrochemi-
In the context of correlated disorder, the vacancy distribution can cal performance154–158. Here, we focus on the strategies to control the
be controlled by electrostatic interactions between vacancies, which reversibility of TM migration and oxygen evolution that appear during
will either favour or hinder cluster formations. Although this work101 electrochemical cycling.
used 22% Ti vacancies, a higher concentration of vacancies could help
to create a percolated network. However, the increase in TM vacancies Transition metal reversibility. Reversible and irreversible TM migra-
implies a decrease in redox sites and a decrease in structural stabil- tions in Li-rich layered cathode materials cause voltage hysteresis
ity, which opens the question of the optimum number of vacancies. and voltage decay, respectively. To prevent irreversible TM migra-
The optimum concentration could be surmised from the computational tion, engineering the stacking order from an O3 to an O2 structure
calculation in randomly packed spheres in which the site-percolation was performed by ionic exchanges from interlayer Na+ to Li+. Due to
threshold is predicted to be 18.3%145. the decreased interlayer distance in the O2 structure, TMs migrat-
The above-described strategies favour alkali-ion diffusion by ing to the vacant interlayer site will undergo stronger electrostatic
influencing local pathways through disorder. Cation site disorder, repulsions from the adjacent TM sites (Fig. 7a). Consequently, strong
such as cation vacancy and Li/TM mixing, can be stabilized by tuning electrostatic repulsion in the O2 structure increases the energy barrier
the electrostatic interactions or incorporating d0 TMs that are prone for TM migration from the tetrahedral site to the interlayer octahedral
to distort in their octahedral sites. Expansion of the interlayer space site along the O–T–O pathway and fosters reversible TM migration back
has been achieved through the presence of hydrated interlayer cations to the octahedral site of the TM layer. This means that the reversibility
in which the interlayer water has another role—to screen the charge of of TM migration might be ensured by modulating the CFSE through
ions and accelerate their diffusion. However, connections between the electrostatic interactions influenced by the stacking sequence159.
local pathways must be created for long-range diffusion, and the con- A comparative study of layered Li1.2Cr0.4Mn0.4O2 (LCMO) and par-
nections will be influenced by the correlation between disorders, which tially disordered LCMO discovered that cation disorder can contribute
should be carefully controlled. The discrepancy between short-range to the inhibition of collective TM migration160. After charging, the d0
and long-range orders is discussed later. Cr6+ with zero octahedral CFSE readily migrates to the tetrahedral

Nature Reviews Chemistry

Review article

site, forming the Cr–Li dumbbell (Fig. 7b), for both materials. In the of the honeycomb superstructure. A superstructure that can hinder
case of layered LCMO, the creation of the Cr–Li dumbbell triggers in-plane TM migration can thus mitigate irreversible oxygen evolution
the creation of another dumbbell, due to the reduced slab spacing and voltage hysteresis.
and the size of the tetrahedral site, ultimately leading to a collective
TM migration during discharge. Conversely, the partially disordered a O3 structure O2 structure
LCMO homologue has its Cr–Li dumbbells randomly orientated, which
helps to prevent the collective evolution of Cr–Li dumbbells. As such, Reversible
TM migration
reversible Cr migration was confirmed in the X-ray absorption spectra
of partially disordered LCMO. The cation site disorder between Li, Mn
and Cr inhibited collective TM migration. z
Modifying the lattice can be an effective method to tune the cor-
relation between disorder. Ab initio calculations have demonstrated
x
that Na-birnessite is a layered polymorph of sodium manganese oxide, y
and it tends to be stabilized by the interlayer crystal water and alkali Weak repulsion Strong repulsion
ions161. In addition to its role in improving ion mobility, the presence of
crystal water contributes to the stability of the cathode material162,163. b Ordered Disordered
Upon desodiation in Na-ion battery cathodes, Mn migrates to the
tetrahedral sites and coordinates with the oxygen of crystal water in Cr–Li
the interlayer site to form a spinel-like state (Fig. 7c). According to dumbbell
DFT calculations, this stable tetrahedral Mn geometry in Na-birnes-
site (Na0.27MnO2·0.5H2O) exhibits a considerably lower Mn migration
energy than the highly crystalline layered Na0.27MnO2 without crystal z
water, which emphasizes the role of the interlayer water. The spinel-
like phase more easily reverts to its original layered structure with x Aligned Cr–Li dumbbells Random distribution of
y
the re-intercalation of Na+ (Fig. 7c). Furthermore, an optimal amount Cr–Li dumbbells
of crystal water (Na0.27MnO2∙0.09H2O) leads to reversible Mn migra-
tion, given that too high a water content increases the Mn migration c Spinel-like Layered
energy barrier and hinders the formation of a spinel-like phase due to
the interference between water molecules163. Therefore, the existence
of crystal water in the interlayer of birnessite and the resulting lattice
Charge
change can serve as a representative illustration of how to control the
disorder induced by battery cycling. Discharge
By changing the synthesis processes and the chemical composition,
z
the above-mentioned examples illustrate the effective manipulation
of local correlations, which, in turn, impact the migration of TMs. These
methods involve cation site disorder and the incorporation of crystal x
y
water, ultimately resulting in the induction of reversible TM migration. Reversible phase transition between spinel-like and layered structure

Consequently, the incorporation of novel atomic patterns within the

TM Manganese Vacancy
lattice emerges as a potent strategy for enhancing the reversibility of
Alkali ion Crystal water
TM diffusion pathways.
Cation with site disorder Oxygen

Oxygen reversibility. During charging, in-plane TM migration in Fig. 7 | Disorder improving TM reversibility. a, Changes to the stacking
layered cathode materials can trigger the formation of O–O dimers sequence suppress transition metal (TM) migration. The illustration shows
susceptible to irreversible oxygen loss. This TM migration is related irreversible TM migration via O–T–O (O, octahedral site; T, tetrahedral site)
to the electrostatic interactions between the migrating TM and sur- diffusion in the O3-type structure (left) and reversible TM migration in the
rounding ions along its migration path to a vacant site, which varies O2-type structure (right), in which the TM returns to its original octahedral site
with specific superstructure types in pristine structures (Fig. 8a), such due to stronger repulsion that destabilizes the vacant octahedral site (empty
as honeycomb (P2-Na0.75Li0.25Mn0.75O2), ribbon (P2-Na0.6Li0.2Mn0.8O2) and circle). b, Schematic of ordered and disordered Li1.2Cr0.4Mn0.4O2 structures
mesh-like (Na2Mn3O7) arrangements. The honeycomb superstructure after charging. During charging, oxidized Cr6+ migrates to a tetrahedral
has the highest number of vacancies, which is less dispersed in the site, a process accompanied by the formation of Cr–Li dumbbells. In the
cation-ordered structure, the Cr6+ at tetrahedral sites is stabilized due to the
structure thereby facilitating in-plane TM migration, while the ribbon
aligned Cr–Li dumbbells. Increased stabilization of Cr6+ initiates a collective
and mesh-like structures have a lower number of vacancies, which are
TM migration leading to the creation of additional dumbbells. Conversely,
more uniformly dispersed164. When layered honeycomb and ribbon
collective TM migration is absent in the cation-disordered structure due to
superstructures are compared, their XRD patterns are different, even the random orientation of the Cr–Li dumbbells. c, Schematic of the reversible
if both are of the P2 stacking sequence. This discrepancy arises from phase transition in birnessite. During charging, Mn migrates to a tetrahedral
the unique arrangements within each superstructure that ensures site, forming a spinel-like structure. The dashed circle is a guide to the eye and
uniform electrostatic interactions164. Although oxygen oxidation was indicates where changes have occurred or will occur. Part a adapted from ref. 159,
observed by resonant inelastic X-ray scattering in both superstructures, Springer Nature Limited. Part b adapted from ref. 160, Springer Nature Limited.
a large voltage decay due to oxygen loss was observed only in the case Part c adapted from ref. 163 under a Creative Commons licence CC BY 4.0.

Nature Reviews Chemistry

Review article

a Honeycomb Ribbon Mesh Fig. 8 | Impact of disorder on the reversibility

of oxygen redox reactions. a, In-plane transition
metal (TM) migration pathways in honeycomb,
ribbon and mesh-like superstructures leading to
the formation of O2. The energy and number of
migrated TMs required for O2 evolution decrease in
the following order: mesh, ribbon and honeycomb
superstructures; this supports the higher stability of
mesh-like structures. b, Schematic representation
of MO6 (where M stand for metal) octahedra after
delithiation in ordered and Li/TM disordered
Li-rich cathodes. When oxygen is oxidized, oxygen
dimerization occurs along the direction of the
O–O axis so that axial symmetry is maintained.
Energy required to form O2 by in-plane TM migration: By contrast, oxygen dimerization is suppressed by
mesh > ribbon > honeycomb Li/TM disordering, which breaks the axial symmetry
and brings about a distortion of the octahedra.
b Ordered Pristine Li/TM disordered
c, Band structure illustrating the TM–O bond
character. Strong covalent TM–O bonds stabilize
oxide ions with the formation of holes. However,
relatively weak covalent and ionic TM–O bonds
destabilize oxide ions, which can then form
superoxide or oxygen gas. The dashed ovals are
guides to the eye and indicate where changes
have occurred or will occur. Part a adapted from
ref. 164, Springer Nature Limited. Part b adapted
c Covalent bond Weak covalent bond from ref. 116 under a Creative Commons licence
CC BY 4.0. Part c adapted with permission from
ref. 166, ACS.
TM eg TM eg

Fermi level
e– e–
TM t2g
TM t2g

O 2p
O 2p

Ligand-hole formation O2 gas release

TM Oxygen

The adjustment of local symmetry around the oxygen in layered for regular Li2RuO3, explaining the stable cycling performance of intra-
compounds contributes to oxygen reversibility by suppressing O–O layer-disordered Li2RuO3. The formation of an O–O dimer was also
dimerization. Li2RuO3 exhibits a honeycomb superstructure with restrained via the cationic disordering in Li1.19Ni0.26Mn0.55O2 through
Li occupation in the TM layer, similar to that of Li2MnO3. The intralayer Li/Ni mixing and Li vacancies165.
Li/Ru site disorder was introduced through ball milling116. Although Compositional disorder through atomic substitutions can affect
the honeycomb superstructure peak was evident in the XRD pattern the TM–O bond, and the stability of oxidized oxygen can be predicted,
of regular Li2RuO3, the intralayer-disordered Li2RuO3 exhibited weaker contingent upon whether the bond exhibits ionic or covalent charac-
superstructure peaks due to the random occupation of Li/Ru sites, ter166. For example, in Li1.33TM0.67O2, the TM–O bond is more covalent for
which broke the axial symmetry. The Li/Ru site disorder impeded the Rh than for Mn, and this effectively stabilizes oxidized oxygen ions by
approach of neighbouring oxygen ions in RuO6 octahedra towards each contributing to a stable ligand-hole formation (Fig. 8c, left). However,
other after delithiation, thus preventing O–O dimerization. Instead, the moderate covalency of Li1.33Mn0.67O2 destabilizes oxygen ions, giving
this disorder promoted the movement of oxygen ions towards the rise to the generation of superoxide species accompanied by oxygen
TM ions along the O–Ru–O bonds, enabling reversible oxygen redox loss (Fig. 8c, right). In this context, the manipulation of Li/TM mixing
chemistry (Fig. 8b). This phenomenon occurred as a result of different in Li-rich cathodes has been utilized to regulate the covalency of the
charge compensation behaviours within the Ru–O bond depending TM–O bond. This deliberate adjustment of the bond character by
on the oxidation of the oxygen ions in different local configurations disordering the atomic arrangement serves to alleviate oxygen loss167.
due to Li/Ru disorder. A computational study showed that oxygen Na-based 3d TM oxide layered cathodes, which activate oxygen
release energies were greater for intralayer-disordered Li2RuO3 than redox due to configurations that have incorporated alkaline earth

Nature Reviews Chemistry

Review article

metal ions or cationic vacancies, as previously mentioned, lead to all pairs of nearest-neighbour spins cannot be minimized, as the TM
decent reversibility of oxygen redox reactions119,120. Stable oxygen spins arranged in 2D triangular lattices are separated by Li layers176.
redox chemistry during the cycling of these Na-deficient cathodes The presence of unpaired spins in Ni2+ results in at least one unknown
can be interpreted as stable in-plane ordering or as suppressed oxygen spin within the triangular lattice of the TM layer, leading to magnetic
release due to the alkaline earth metal ions that remain in the cathode frustration via interplane superexchange reactions that destabilizes
without taking part in the intercalation/extraction processes. As there is the lattice. Increased Li/Ni mixing can relieve the magnetic frustration
a high reversibility of oxygen redox, the limited capacity of Na-deficient and lead to the formation of a stable antiferromagnetic arrangement,
cathodes could be resolved. To reach even higher capacities, the limi- illustrating that optimal disorder can stabilize the lattice of layered
tation on the number of exchangeable Na ions must be targeted by cathode materials and subsequently enhance their electrochemical
investigating oxygen redox reactions in Na-stoichiometric compounds. performance67 (Fig. 9c). This example shows three functionalities
However, reversible oxygen redox reactions in Na-stoichiometric introduced by cation site disorder: hindered Li+ transport, a stabilized
compounds have been experimentally demonstrated in relatively few layered structure due to the pillar effect and decreased magnetic frus-
examples168–172. tration. It is thus important to determine the appropriate amount of
The examples described above show that chemical compositions Li/Ni mixing for specific systems that is needed to maximize advantages
and synthetic procedures can tune disorder to improve the reversibility and minimize harmful effects.
of oxygen oxidation. These insights suggest that changing the atomic Substituting oxygen with fluorine has been widely adopted as a
configurations and electronic structures are viable strategies for facili- way to enhance the electrochemical properties of DRX. This disorder
tating reversible oxygen redox reactions. However, a consensus on the leads to a conflict between properties that simultaneously enhance
mechanism and origin of stable reactions is yet to be achieved, and a and impair the electrochemical performance. Although there are
viewpoint involving disorder is likely helpful as we move towards a full challenges associated with incorporating F into layered cathode mate-
understanding of the oxygen redox reactions. rials177, Li-rich and TM-poor local environments induced by cation
disorder enable effective bulk fluorination178. F substitution has certain
Challenges when taking advantage of disorder advantages, such as increasing the redox voltage through the inductive
Disorder introduces electrochemical functionalities that can be ben- effect179,180 or facilitating TM redox reactions by changing the valence
eficial to the redox reactions of TM oxides, such as increasing their state of the TM and suppressing Jahn–Teller distortions181,182. Notably,
energy density and improving their reversibility. However, the disorder F substitution can control Li+ transport in DRX. According to DFT cal-
can also lead to intricate interactions, with simultaneous interplays culations coupled with a Monte Carlo simulation, most of the F atoms
of advantageous and disadvantageous properties that should not be are likely to be surrounded by five or six neighbouring Li atoms183.
overlooked. In addition, local disorder with useful contributions might This Li-rich local environment with multiple Li4TM0 clusters around a
not exhibit this properly throughout the entire material. Therefore, highly electronegative F enhances Li+ diffusion kinetics184 (Fig. 9d, left).
to make novel cathode materials inspired by disordered structures, However, it is important to note that the ‘Li-locking effect’, also stem-
we must understand the interplay and balance between phenomena ming from the high electronegativity of F, can render some fraction of
stemming from various forms of disorder. This section discusses Li+ unextractable (Fig. 9d, right). In this scenario, under-coordinated F
some examples that explain the trade-offs of multiple local disorder atoms strongly capture neighbouring Li atoms, impeding the full use
in electrochemical properties and how local ordering exerts an overall of Li ions for charge/discharge. In conclusion, optimal F substitution
electrochemical influence on cathode materials. facilitates Li+ mobility without impeding Li extraction by the Li-locking
effect185.
Trade-off between electrochemical properties. A single type of
disorder can give rise to multiple electrochemical properties. These Relationship between local disorder and long-range properties.
properties can sometimes be detrimental and beneficial, simultane- A noteworthy question regarding the relationship between disorder
ously. This section highlights two examples, layered metal oxides and and the properties of cathode materials encompasses whether the
DRX, to explain the necessity of trade-offs when utilizing disorder. influence of local disorder is effective in the entire structure. This
Li/Ni mixing is a representative form of disorder in Ni-containing problem is a key issue in DRX structures, in which Li4TM0 clusters that
layered cathode materials, as the chemically unstable Ni3+ spontane- allow fast Li+ hopping must be interconnected throughout the entire
ously reduces to Ni2+, which can occupy the Li site. This triggers com- cubic structure to form a 0-TM Li percolation network, providing higher
petition between the different interactions, impacting the stability of Li+ diffusivity186 (Fig. 9e). The percolation network can be precisely
the materials and thereby making it necessary to optimize the system modulated by their short-range order (SRO), and limited Li percolation
to achieve the best performance. The primary negative effect of Li/Ni implies a lower extractability of Li+, decreasing the capacity of the cath-
mixing on Li+ transport results from a collapsing interlayer structure ode. For instance, although both Li1.2Mn0.4Zr0.4O2 and Li1.2Mn0.4Ti0.4O2
following the oxidation of Ni2+ in the Li site to Ni3+ upon delithiation173. contain the same number of Li+ and Li4TM0 clusters187, they exhibit
Furthermore, Li/Ni mixing stimulates additional cation migration, different SROs, depending on the choice of the d0 TM188. The initial
causing an irreversible phase transition174 (Fig. 9a). Consequently, Li/Ni capacity is lowered with Zr4+ due to the hindered connectivity of Li4TM0
mixing is considered disadvantageous to cycling performance. How- clusters. Other reports showed that in high-entropy DRX utilizing a
ever, a small amount of Li/Ni mixing can delay phase transitions through large number of elements, SRO is effectively suppressed189,190. When
the pillar effect, in which interlayer Ni2+ ions act as a pillar, enhancing more TM species are added to the high-entropy alloy, the energy den-
structural stability by limiting additional cation migration because of sity and rate capability of high-entropy DRX increase at the expense of
the electrostatic repulsion of Ni2+ in the Li layer175 (Fig. 9b). Moreover, SRO, which decreases190.
Li/Ni mixing in NCM can alleviate magnetic frustration within the lat- Multiple investigation sought to alter collective local ordering
tice, improving structural stability. In NCM, the interaction energy for in a material to enhance the connectivity of Li+ diffusion pathways.

Nature Reviews Chemistry

Review article

a Interlayer collapse with Ni migration b Pillar effect

Migrated Ni2+
Additional oxidation to Ni3+ Electrostatic repulsion
TM migration → Interlayer collapse of migrated Ni2+ suppresses
TM migration
z z

x x
y y

c Relieved magnetic frustration after Li/Ni mixing d Controlled Li+ transport by F substitution
Without Li/Ni mixing Li/Ni mixing in Ni layer

Magnetic frustration Stable antiferromagnetic arrangement Introducing multiple Li4TM0 clusters Li-locking effect

e 0-TM connectivity for Li percolation

Alkali ion
TM
Oxygen
Fluorine (F, brown)
Nickel
Vacancy

Without 0-TM connectivity Li percolation through 0-TM percolation network

Fig. 9 | Challenges in establishing relationships between disorder and arrows) (part c, right) in the Ni2+ site alleviates magnetic frustration (part c,
properties. a–c, Advantages and drawbacks of Li/Ni mixing. After Li/Ni mixing, right). d, Advantages and drawbacks of F substitution. The presence of F creates
the interlayer space is reduced as a result of the oxidation of Ni2+ in the Li layer. Li clusters, which form a Li percolation network. However, Li can be captured
This process triggers additional transition metal (TM) migration, resulting in an by F, which reduces the amount of redox-active Li. e, Li4TM0 clusters aid fast
irreversible phase transition (part a). By contrast, electrostatic repulsion (double local hopping with lower electrostatic interactions. However, the connectivity
arrow) of Ni2+ in the Li layer after Li/Ni mixing can also suppress additional between these clusters (present in the right structure) within the long-range
TM migration (part b). The presence of unknown spins of Ni2+, symbolized by structure is crucial for high ionic diffusivity through the 0-TM percolation
the ‘?’ symbol in the 2D triangular lattice of the TM layer (part c, left), induces network. Part c adapted with permission from ref. 176, Elsevier. Part d adapted
magnetic frustration between spins (black arrow). Non-magnetic Li+ (lack of with permission from ref. 184, Wiley.

One example consists of layered cation ordering domains in DRX identi- of Li+ diffusion channels. The formation of interconnected layered-like
fied through polarized Raman spectroscopy191. This study revealed that, cation ordering in the cubic lattice creates more Li+ transport channels,
although the crystal structure of DRX shows cubic isotropy stemming enhancing the electrochemical properties. Moreover, spinel-like order-
from Li/TM site disorder, Li-rich and TM-rich layered-like anisotropy is ing enables more 0-TM percolation networks, providing an advanta-
randomly dispersed throughout the material and aligned along various geous perspective for designing a partially disordered spinel187,192.
directions at the local scale, resulting in the lack of interconnectivity To conclude, the presence of advantageous local disorder could be

Nature Reviews Chemistry

Review article

insufficient when it comes to exhibiting large-scale effects, but this 7. He, Z., Taniyama, T., Kyômen, T. & Itoh, M. Field-induced order-disorder transition in the
quasi-one-dimensional anisotropic antiferromagnet BaCo2V2O8. Phys. Rev. B 72, 172403
must be examined to enable the advantageous functionalization of (2005).
disorder in materials. 8. Raja, A. et al. Dielectric disorder in two-dimensional materials. Nat. Nanotechnol. 14,
832–837 (2019).

Future perspectives 9. Yu, Z. et al. Pressure effect on order–disorder ferroelectric transition in a

hydrogen-bonded metal–organic framework. J. Phys. Chem. Lett. 11, 9566–9571
In this Review, we suggest an approach for improving the electrochemi- (2020).
cal performance of cathode materials by introducing structural dis- 10. Maier, J. Review — battery materials: why defect chemistry? J. Electrochem. Soc. 162,
A2380–A2386 (2015).
order as a design parameter by modulating chemical interactions. 11. Shan, X. et al. Structural water and disordered structure promote aqueous sodium-ion
Although this is still a burgeoning field in battery materials, we estab- energy storage in sodium-birnessite. Nat. Commun. 10, 4975 (2019).
lished a relationship between structural disorder and associated ben- 12. Gao, A. et al. In operando visualization of cation disorder unravels voltage decay in
Ni‐rich cathodes. Small Methods 5, 2000730 (2021).
eficial properties, utilizing several examples of cathode materials. First, 13. Ashton, T. E. et al. Stoichiometrically driven disorder and local diffusion in NMC
we attempted to encapsulate the diverse types of chemical origins that cathodes. J. Mater. Chem. A 9, 10477–10486 (2021).
14. Zhu, H. et al. Bridging structural inhomogeneity to functionality: pair
induce disorder. Although size and electroneutrality effects are piv-
distribution function methods for functional materials development. Adv. Sci. 8,
otal in predicting structural order, they fall short of comprehensively 2003534 (2021).
describing all underlying structural interactions within cathode materi- 15. Huang, W. et al. Elastic lattice enabling reversible tetrahedral Li storage sites
in a high‐capacity manganese oxide cathode. Adv. Mater. 34, 2202745 (2022).
als. Second, we discussed magnetic interactions, Jahn–Teller effects,
16. Szymanski, N. J. et al. Modeling short-range order in disordered rocksalt cathodes by pair
and the destabilization of TM and O after delithiation as additional distribution function analysis. Chem. Mater. 35, 4922–4934 (2023).
sources of correlated disorder in cathode materials. Further interac- 17. Bragg, W. L. & Williams, E. J. The effect of thermal agitation on atomic arrangement in
alloys. Proc. R. Soc. A 145, 699–730 (1934).
tions that influence structural order must be identified, understood
18. Bethe, H. A. Statistical theory of superlattices. Proc. R. Soc. A 150, 552–575 (1935).
and complemented with precise characterization103,104,193–195. We have 19. Onsager, L. Crystal statistics. I. A two-dimensional model with an order-disorder
summarized the cases in which disorder promotes the electrochemi- transition. Phys. Rev. 65, 117–149 (1944).
20. Cowley, J. M. An approximate theory of order in alloys. Phys. Rev. 77, 669–675 (1950).
cal properties of the cathode materials and focused on how they can 21. Welberry, T. R. & Butler, B. D. Diffuse X-ray scattering from disordered crystals. Chem. Rev.
be chemically controlled. Numerous investigations have studied how 95, 2369–2403 (1995).
to manipulate disorder to increase ion mobility, modulate redox pro- 22. Keen, D. A. & Goodwin, A. L. The crystallography of correlated disorder. Nature 521,
303–309 (2015).
cess, and improve the electrochemical and structural reversibility of This paper first introduces the concept of correlated disorder of solid-state materials.
TM and O. However, the influences on disorder remain complex and 23. Moessner, R. & Ramirez, A. P. Geometrical frustration. Phys. Today 59, 24–29 (2006).
sometimes in conflict with each other. As disorder can have multiple 24. Snyder, J., Slusky, J. S., Cava, R. J. & Schiffer, P. How ‘spin ice’ freezes. Nature 413, 48–51
(2001).
effects, discerning which effect is predominant is a paramount con- 25. Fennell, T. et al. Magnetic Coulomb phase in the spin ice Ho2Ti2O7. Science 326, 415–417
sideration. Moreover, rigorous investigations must define whether (2009).
local disorder operates collectively to reshape the functionalities of 26. Fabrèges, X. et al. Spin-lattice coupling, frustration, and magnetic order in
multiferroic R MnO3. Phys. Rev. Lett. 103, 067204 (2009).
an entire material. 27. Wen, J.-J. et al. Disordered route to the Coulomb quantum spin liquid: random transverse
The novel strategy of using structural disorder has a high poten- fields on spin ice in Pr2Zr2O7. Phys. Rev. Lett. 118, 107206 (2017).
tial to make important contributions in cathode design and perfor- 28. Salzmann, C. G., Radaelli, P. G., Slater, B. & Finney, J. L. The polymorphism of ice:
five unresolved questions. Phys. Chem. Chem. Phys. 13, 18468–18480 (2011).
mance, but the critical challenge is how to properly characterize it. 29. Playford, H. Y., Whale, T. F., Murray, B. J., Tucker, M. G. & Salzmann, C. G. Analysis of stacking
Unravelling structural disorder cannot be achieved through a single disorder in ice I using pair distribution functions. J. Appl. Crystallogr. 51, 1211–1220 (2018).
characterization technique; instead, various physicochemical analysis 30. Bramwell, S. T. & Gingras, M. J. P. Spin ice state in frustrated magnetic pyrochlore
materials. Science 294, 1495–1501 (2001).
methods need to be synergistically combined, and they should ide- 31. Düvel, A. et al. Is geometric frustration-induced disorder a recipe for high ionic
ally be implemented under operando conditions. In addition, a clear conductivity? J. Am. Chem. Soc. 139, 5842–5848 (2017).
standard for categorizing the various types of structural disorder is 32. Di Stefano, D. et al. Superionic diffusion through frustrated energy landscape. Chem 5,
2450–2460 (2019).
essential to document their occurrences, characteristics and impacts 33. Wang, S., Liu, Y. & Mo, Y. Frustration in super‐ionic conductors unraveled by the density of
on the properties of materials beyond the currently limited framework. atomistic states. Angew. Chem. 135, e202215544 (2023).
In parallel, a suitable synthesis method must be developed to obtain 34. Hsu, W.-L., Tsai, C.-W., Yeh, A.-C. & Yeh, J.-W. Clarifying the four core effects of high-
entropy materials. Nat. Rev. Chem. 8, 471–485 (2024).
targeted disordered structures. 35. Aubry, S. & Hughes, D. A. Reductions in stacking fault widths in fcc crystals:
semiempirical calculations. Phys. Rev. B 73, 224116 (2006).
36. Baruffi, C., Ghazisaeidi, M., Rodney, D. & Curtin, W. A. Equilibrium versus non-equilibrium
Published online: xx xx xxxx
stacking fault widths in NiCoCr. Scr. Mater. 235, 115536 (2023).
37. Coles, S. W. et al. Anion-polarisation-directed short-range-order in antiperovskite
References
Li2FeSO. J. Mater. Chem. A 11, 13016–13026 (2023).
1. Bragg, W. L. The structure of some crystals as indicated by their diffraction of X-rays.
38. Uemura, N., Shirai, K., Eckert, H. & Kunstmann, J. Structure, nonstoichiometry, and
Proc. R. Soc. A 89, 248–277 (1913).
geometrical frustration of α-tetragonal boron. Phys. Rev. B 93, 104101 (2016).
2. Rietveld, H. M. A profile refinement method for nuclear and magnetic structures. J. Appl.
39. Goldschmidt, V. M. Die Gesetze der Krystallochemie. Naturwissenschaften 14, 477–485
Crystallogr. 2, 65–71 (1969).
(1926).
3. Reynaud, M., Serrano-Sevillano, J. & Casas-Cabanas, M. Imperfect battery materials:
40. Pauling, L. The principles determining the structure of complex ionic crystals. J. Am.
a closer look at the role of defects in electrochemical performance. Chem. Mater. 35,
Chem. Soc. 51, 1010–1026 (1929).
3345–3363 (2023).
41. George, J. et al. The limited predictive power of the Pauling rules. Angew. Chem. Int. Ed.
This paper categorizes various defects in battery materials.
59, 7569–7575 (2020).
4. Simonov, A. & Goodwin, A. L. Designing disorder into crystalline materials. Nat. Rev.
42. Delmas, C., Fouassier, C. & Hagenmuller, P. Structural classification and properties of the
Chem. 4, 657–673 (2020).
layered oxides. Physica B+C 99, 81–85 (1980).
This paper highlights the design principles to control the correlated disorder in a wide
43. Brunel, M., De Bergevin, F. & Gondrand, M. Determination theorique et domaines
range of materials.
d’existence des differentes surstructures dans les composes A3+B1+X22− de type NaCl.
5. Yu, S., Qiu, C.-W., Chong, Y., Torquato, S. & Park, N. Engineered disorder in photonics.
J. Phys. Chem. Solids 33, 1927–1941 (1972).
Nat. Rev. Mater. 6, 226–243 (2021).
44. Hauck, J. Short-range order and superstructures of ternary oxides AMO2, A2MO3 and
6. Bennett, T. D., Cheetham, A. K., Fuchs, A. H. & Coudert, F.-X. Interplay between
A5MO6 of monovalent A and multivalent M metals related to the NaCl structure. Acta
defects, disorder and flexibility in metal-organic frameworks. Nat. Chem. 9, 11–16
Crystallogr. A 36, 228–237 (1980).
(2017).

Nature Reviews Chemistry

Review article

45. Jacquet, Q. et al. Electrostatic interactions versus second order Jahn–Teller distortion as 79. Urban, A., Abdellahi, A., Dacek, S., Artrith, N. & Ceder, G. Electronic-structure origin of
the source of structural diversity in Li3MO4 compounds (M = Ru, Nb, Sb and Ta). Chem. cation disorder in transition-metal oxides. Phys. Rev. Lett. 119, 176402 (2017).
Mater. 30, 392–402 (2018). 80. Zhang, J. et al. Regulating pseudo-Jahn–Teller effect and superstructure in layered
This paper compares the effect of electrostatic interactions and pseudo Jahn–Teller cathode materials for reversible alkali-ion intercalation. J. Am. Chem. Soc. 144,
(second order Jahn–Teller) distortions on the stability of rocksalt-type structures. 7929–7938 (2022).
46. Mather, G. C., Dussarrat, C., Etourneau, J. & West, A. R. A review of cation-ordered rock 81. Delmas, C., Carlier, D. & Guignard, M. The layered oxides in lithium and sodium‐ion
salt superstructure oxides. J. Mater. Chem. 10, 2219–2230 (2000). batteries: a solid‐state chemistry approach. Adv. Energy Mater. 11, 2001201 (2021).
47. Yang, R. et al. Intercalation in 2D materials and in-situ studies. Nat. Rev. Chem. 8, 82. Reed, J. & Ceder, G. Charge, potential, and phase stability of layered Li(Ni0.5Mn0.5)O2.
410–432 (2024). Electrochem. Solid-State Lett. 5, A145–A148 (2002).
48. Clément, R. J., Lun, Z. & Ceder, G. Cation-disordered rocksalt transition metal oxides 83. Reed, J. & Ceder, G. Role of electronic structure in the susceptibility of metastable
and oxyfluorides for high energy lithium-ion cathodes. Energy Environ. Sci. 13, 345–373 transition-metal oxide structures to transformation. Chem. Rev. 104, 4513–4534 (2004).
(2020). 84. Ma, X., Kang, K., Ceder, G. & Meng, Y. S. Synthesis and electrochemical properties of
49. Hewston, T. A. & Chamberland, B. L. A Survey of first-row ternary oxides LiMO2 (M = Sc-Cu). layered LiNi2/3Sb1/3O2. J. Power Sources 173, 550–555 (2007).
J. Phys. Chem. Solids 48, 97–108 (1987). 85. Vinckevičiūtė, J., Radin, M. D., Faenza, N. V., Amatucci, G. G. & Van Der Ven, A.
50. Wu, E. J., Tepesch, P. D. & Ceder, G. Size and charge effects on the structural stability Fundamental insights about interlayer cation migration in Li-ion electrodes at high states
of LiMO2 (M = transition metal) compounds. Philos. Mag. B 77, 1039–1047 (1998). of charge. J. Mater. Chem. A 7, 11996–12007 (2019).
51. Dutta, G., Manthiram, A., Goodenough, J. B. & Grenier, J.-C. Chemical synthesis and 86. Radin, M. D. et al. Narrowing the gap between theoretical and practical capacities in
properties of Li1−δ−xNi1+δO2 and Li[Ni2]O4. J. Solid State Chem. 96, 123–131 (1992). Li‐ion layered oxide cathode materials. Adv. Energy Mater. 7, 1602888 (2017).
52. Manthiram, A. A reflection on lithium-ion battery cathode chemistry. Nat. Commun. 11, 87. Okubo, M. & Yamada, A. Molecular orbital principles of oxygen-redox battery electrodes.
1550 (2020). ACS Appl. Mater. Interfaces 9, 36463–36472 (2017).
53. Lee, K.-S., Myung, S.-T., Amine, K., Yashiro, H. & Sun, Y.-K. Structural and electrochemical 88. Abakumov, A. M., Fedotov, S. S., Antipov, E. V. & Tarascon, J.-M. Solid state chemistry for
properties of layered Li[Ni1−2xCoxMnx]O2 (x = 0.1–0.3) positive electrode materials for Li-Ion developing better metal-ion batteries. Nat. Commun. 11, 4976 (2020).
batteries. J. Electrochem. Soc. 154, A971 (2007). 89. Li, B. & Tarascon, J.-M. in Comprehensive Inorganic Chemistry III (eds Reedijk, J. &
54. Lee, W. et al. New insight into Ni‐Rich layered structure for next‐generation li Poeppelmeier, K. R.) 6–45 (Elsevier, 2023).
rechargeable batteries. Adv. Energy Mater. 8, 1701788 (2018). 90. Ben Yahia, M., Vergnet, J., Saubanère, M. & Doublet, M.-L. Unified picture of anionic redox
55. Liang, C. et al. Unraveling the origin of instability in Ni-Rich LiNi1–2xCoxMnxO2 (NCM) in Li/Na-ion batteries. Nat. Mater. 18, 496–502 (2019).
cathode materials. J. Phys. Chem. C 120, 6383–6393 (2016). 91. Mizushima, K., Jones, P. C., Wiseman, P. J. & Goodenough, J. B. LixCoO2 (0 < x < −1): a new
56. Hahn, B. P., Long, J. W. & Rolison, D. R. Something from nothing: enhancing cathode material for batteries of high energy density. Mater. Res. Bull. 15, 783–789 (1980).
electrochemical charge storage with cation vacancies. Acc. Chem. Res. 46, 1181–1191 92. Rozier, P. & Tarascon, J. M. Review—Li-rich layered oxide cathodes for next-generation
(2013). Li-ion batteries: chances and challenges. J. Electrochem. Soc. 162, A2490–A2499 (2015).
57. Ruetschi, P. Cation‐vacancy model for MnO2. J. Electrochem. Soc. 131, 2737–2744 (1984). 93. Molenda, J., Wilk, P. & Marzec, J. Structural, electrical and electrochemical properties of
58. Bouet, L., Tailhades, P., Rousset, A., Domenichini, B. & Gillot, B. Mixed valence states of LiNiO2. Solid State Ion. 146, 73–79 (2002).
iron and molybdenum ions in MoxFe3−xO4 magnetites and related cation deficient ferrites. 94. Huang, Z.-F., Meng, X., Wang, C.-Z., Sun, Y. & Chen, G. First-principles calculations on the
Solid State Ion. 52, 285–286 (1992). Jahn–Teller distortion in layered LiMnO2. J. Power Sources 158, 1394–1400 (2006).
59. Bach, S. et al. Electrochemical proton insertion in Mn2.2Co0.27O4 from aqueous borate 95. Yabuuchi, N. & Ohzuku, T. Novel lithium insertion material of LiCo1/3Ni1/3Mn1/3O2 for
solution. Electrochim. Acta 45, 931–934 (1999). advanced lithium-ion batteries. J. Power Sources 119–121, 171–174 (2003).
60. Ménétrier, M., Shao-Horn, Y., Wattiaux, A., Fournès, L. & Delmas, C. Iron substitution 96. Yoon, C. S. et al. Extracting maximum capacity from Ni-rich Li[Ni0.95Co0.025Mn0.025]O2
in lithium-overstoichiometric “Li1.1CoO2”: combined 57Fe Mössbauer and 7Li NMR cathodes for high-energy-density lithium-ion batteries. J. Mater. Chem. A 6, 4126–4132
spectroscopies studies. Chem. Mater. 17, 4653–4659 (2005). (2018).
61. Li, W. et al. High substitution rate in TiO2 anatase nanoparticles with cationic vacancies 97. Thackeray, M. M., Johnson, C. S., Vaughey, J. T., Li, N. & Hackney, S. A. Advances in
for fast lithium storage. Chem. Mater. 27, 5014–5019 (2015). manganese-oxide ‘composite’ electrodes for lithium-ion batteries. J. Mater. Chem. 15,
62. Ma, J. et al. Layered lepidocrocite type structure isolated by revisiting the sol–gel 2257 (2005).
chemistry of anatase TiO2: a new anode material for batteries. Chem. Mater. 29, 98. Strobel, P. & Lambert-Andron, B. Crystallographic and magnetic structure of Li2MnO3.
8313–8324 (2017). J. Solid State Chem. 75, 90–98 (1988).
63. Miyamoto, N. & Yamamoto, S. in Supra-Materials Nanoarchitectonics Ch. 7 (eds Ariga, K. 99. Yabuuchi, N. et al. High-capacity electrode materials for rechargeable lithium batteries:
& Aono, M.) 173–192 (Elsevier, 2017). Li3NbO4 -based system with cation-disordered rocksalt structure. Proc. Natl Acad. Sci.
64. Nam, K. W. et al. The high performance of crystal water containing manganese birnessite USA 112, 7650–7655 (2015).
cathodes for magnesium batteries. Nano Lett. 15, 4071–4079 (2015). 100. Kobayashi, T. et al. Nanosize cation‐disordered rocksalt oxides: Na2TiO3–NaMnO2 binary
65. Ferrage, E. Investigation of the interlayer organization of water and ions in smectite from system. Small 16, 1902462 (2020).
the combined use of diffraction experiments and molecular simulations: a review of 101. Koketsu, T. et al. Reversible magnesium and aluminium ions insertion in cation-deficient
methodology, applications, and perspectives. Clays Clay Miner. 64, 348–373 (2016). anatase TiO2. Nat. Mater. 16, 1142–1148 (2017).
66. Yang, J. et al. Size‐independent fast ion intercalation in two‐dimensional titania 102. Boyd, S. et al. Effects of interlayer confinement and hydration on capacitive charge
nanosheets for alkali‐metal‐ion batteries. Angew. Chem. Int. Ed. 58, 8740–8745 (2019). storage in birnessite. Nat. Mater. 20, 1689–1694 (2021).
67. Zheng, J. et al. Ni/Li disordering in layered transition metal oxide: electrochemical 103. Stiles, J. W., McClure, E. T., Bashian, N. H., Tappan, B. A. & Melot, B. C. Reversible
impact, origin, and control. Acc. Chem. Res. 52, 2201–2209 (2019). intercalation of Li ions in an earth-abundant phyllosilicate clay. Inorg. Chem. 61,
68. Kanamori, J. Superexchange interaction and symmetry properties of electron orbitals. 5757–5761 (2022).
J. Phys. Chem. Solids 10, 87–98 (1959). 104. Wu, V. C. et al. Unlocking new redox activity in alluaudite cathodes through
69. Goodenough, J. B. Magnetism and the Chemical Bond (R. E. Krieger, 1976). compositional design. Chem. Mater. 34, 4088–4103 (2022).
70. Zheng, J. et al. Role of superexchange interaction on tuning of Ni/Li disordering in 105. Van Der Ven, A., Bhattacharya, J. & Belak, A. A. Understanding Li diffusion in
layered Li(NixMnyCoz)O2. J. Phys. Chem. Lett. 8, 5537–5542 (2017). Li-intercalation compounds. Acc. Chem. Res. 46, 1216–1225 (2013).
71. Bersuker, I. B. Jahn–Teller and pseudo-Jahn–Teller effects: from particular features to general 106. Abdellahi, A., Urban, A., Dacek, S. & Ceder, G. The effect of cation disorder on the
tools in exploring molecular and solid state properties. Chem. Rev. 121, 1463–1512 (2021). average Li intercalation voltage of transition-metal oxides. Chem. Mater. 28, 3659–3665
72. Freitag, R. & Conradie, J. Understanding the Jahn–Teller effect in octahedral transition- (2016).
metal complexes: a molecular orbital view of the Mn(β-diketonato)3 complex. J. Chem. 107. Abdellahi, A., Urban, A., Dacek, S. & Ceder, G. Understanding the effect of cation disorder
Educ. 90, 1692–1696 (2013). on the voltage profile of lithium transition-metal oxides. Chem. Mater. 28, 5373–5383
73. Armstrong, A. R. & Bruce, P. G. Synthesis of layered LiMnO2 as an electrode for (2016).
rechargeable lithium batteries. Nature 381, 499–500 (1996). This paper clarifies the causal relationship between cation disorder and the voltage
74. Manzi, J. et al. Monoclinic and orthorhombic NaMnO2 for secondary batteries: profile based on Li–vacancy interactions.
a comparative study. Energies 14, 1230 (2021). 108. Kitchaev, D. A. et al. Design principles for high transition metal capacity in disordered
75. Liu, T. et al. Correlation between manganese dissolution and dynamic phase stability in rocksalt Li-ion cathodes. Energy Environ. Sci. 11, 2159–2171 (2018).
spinel-based lithium-ion battery. Nat. Commun. 10, 4721 (2019). 109. Lee, G., Lau, V. W., Yang, W. & Kang, Y. Utilizing oxygen redox in layered cathode materials
76. Asl, H. Y. & Manthiram, A. Reining in dissolved transition-metal ions. Science 369, 140–141 from multiscale perspective. Adv. Energy Mater. 11, 2003227 (2021).
(2020). 110. Lin, T., Seaby, T., Huang, X. & Wang, L. On the disparity in reporting Li-rich layered oxide
77. Yaghoobnejad Asl, H. & Manthiram, A. Proton-induced disproportionation of Jahn– cathode materials. Chem. Commun. 59, 2888–2902 (2023).
Teller-active transition-metal ions in oxides due to electronically driven lattice instability. 111. Zhang, M. et al. Pushing the limit of 3d transition metal-based layered oxides that
J. Am. Chem. Soc. 142, 21122–21130 (2020). use both cation and anion redox for energy storage. Nat. Rev. Mater. 7, 522–540
78. Bersuker, I. B. Pseudo Jahn-Teller origin of perovskite multiferroics, magnetic- (2022).
ferroelectric crossover, and magnetoelectric effects: the d0−d10 problem. Phys. Rev. Lett. 112. Sathiya, M. et al. Reversible anionic redox chemistry in high-capacity layered-oxide
108, 137202 (2012). electrodes. Nat. Mater. 12, 827–835 (2013).

Nature Reviews Chemistry

Review article

113. Seo, D.-H. et al. The structural and chemical origin of the oxygen redox activity in layered 146. Reimers, J. N. & Dahn, J. R. Electrochemical and in situ x‐ray diffraction studies of lithium
and cation-disordered Li-excess cathode materials. Nat. Chem. 8, 692–697 (2016). intercalation in LixCoO2. J. Electrochem. Soc. 139, 2091–2097 (1992).
114. Perez, A. J. et al. Approaching the limits of cationic and anionic electrochemical activity 147. Shao-Horn, Y., Levasseur, S., Weill, F. & Delmas, C. Probing lithium and vacancy ordering
with the Li-rich layered rocksalt Li3IrO4. Nat. Energy 2, 954–962 (2017). in O3 layered LixCoO2 ( x ≈ 0.5 ) : an electron diffraction study. J. Electrochem. Soc. 150,
115. Song, J. et al. A high‐performance Li–Mn–O Li‐rich cathode material with rhombohedral A366–A373 (2003).
symmetry via intralayer Li/Mn disordering. Adv. Mater. 32, 2000190 (2020). 148. Marianetti, C. A., Kotliar, G. & Ceder, G. A first-order Mott transition in LixCoO2. Nat. Mater.
116. Ning, F. et al. Inhibition of oxygen dimerization by local symmetry tuning in Li-rich 3, 627–631 (2004).
layered oxides for improved stability. Nat. Commun. 11, 4973 (2020). 149. Wolverton, C. & Zunger, A. First-principles prediction of vacancy order-disorder and
This paper demonstrates that intralayer Li/Ru mixing suppresses oxygen dimerization intercalation battery voltages in LixCoO2. Phys. Rev. Lett. 81, 606–609 (1998).
and subsequent oxygen loss. 150. Van Der Ven, A., Aydinol, M. K., Ceder, G., Kresse, G. & Hafner, J. First-principles
117. McColl, K. et al. Transition metal migration and O2 formation underpin voltage hysteresis investigation of phase stability in LixCoO2. Phys. Rev. B 58, 2975–2987 (1998).
in oxygen-redox disordered rocksalt cathodes. Nat. Commun. 13, 5275 (2022). 151. Lee, W. et al. Advances in the cathode materials for lithium rechargeable batteries.
This paper elucidates that TM migration is necessary for the formation of O2 trapped Angew. Chem. Int. Ed. 59, 2578–2605 (2020).
in the bulk. 152. Zou, L. et al. Revealing cycling rate-dependent structure evolution in Ni-rich layered
118. Cao, X. et al. Stabilizing anionic redox chemistry in a Mn‐based layered oxide cathode cathode materials. ACS Energy Lett. 3, 2433–2440 (2018).
constructed by Li‐deficient pristine state. Adv. Mater. 33, 2004280 (2021). 153. Zhang, H. et al. Layered oxide cathodes for Li-ion batteries: oxygen loss and vacancy
119. Yabuuchi, N. et al. A new electrode material for rechargeable sodium batteries: P2-type evolution. Chem. Mater. 31, 7790–7798 (2019).
Na2/3[Mg0.28Mn0.72]O2 with anomalously high reversible capacity. J. Mater. Chem. A 2, 154. Zhao, R. et al. The origin of heavy element doping to relieve the lattice thermal vibration
16851–16855 (2014). of layered materials for high energy density Li ion cathodes. J. Mater. Chem. A 8,
120. Maitra, U. et al. Oxygen redox chemistry without excess alkali-metal ions in 12424–12435 (2020).
Na2/3[Mg0.28Mn0.72]O2. Nat. Chem. 10, 288–295 (2018). 155. Oh, P. et al. New ion substitution method to enhance electrochemical reversibility
121. Dai, K. et al. High reversibility of lattice oxygen redox quantified by direct bulk probes of of co‐rich layered materials for Li‐ion batteries. Adv. Energy Mater. 13, 2202237 (2023).
both anionic and cationic redox reactions. Joule 3, 518–541 (2019). 156. Chen, J. et al. Structure/interface coupling effect for high‐voltage LiCoO2 cathodes.
122. Wang, Y. et al. Ultralow‐strain Zn‐substituted layered oxide cathode with suppressed Adv. Mater. 34, 2204845 (2022).
P2–O2 transition for stable sodium ion storage. Adv. Funct. Mater. 30, 1910327 (2020). 157. Zhao, W. et al. Optimized Al doping improves both interphase stability and bulk
123. Zheng, W. et al. Oxygen redox activity with small voltage hysteresis in Na0.67Cu0.28Mn0.72O2 structural integrity of Ni-Rich NMC cathode materials. ACS Appl. Energy Mater. 3,
for sodium-ion batteries. Energy Storage Mater. 28, 300–306 (2020). 3369–3377 (2020).
124. Song, B. et al. Understanding the low-voltage hysteresis of anionic redox in Na2Mn3O7. 158. Nayak, P. K. et al. Al doping for mitigating the capacity fading and voltage decay of
Chem. Mater. 31, 3756–3765 (2019). layered Li and Mn-rich cathodes for Li-ion batteries. Adv. Energy Mater. 6, 1502398
125. Gent, W. E. et al. Coupling between oxygen redox and cation migration explains unusual (2016).
electrochemistry in lithium-rich layered oxides. Nat. Commun. 8, 2091 (2017). 159. Eum, D. et al. Voltage decay and redox asymmetry mitigation by reversible cation
126. Lai, Y. et al. Ion‐migration mechanism: an overall understanding of anionic redox activity migration in lithium-rich layered oxide electrodes. Nat. Mater. 19, 419–427 (2020).
in metal oxide cathodes of Li/Na‐ion batteries. Adv. Mater. 34, 2206039 (2022). 160. Huang, J. et al. Inhibiting collective cation migration in Li-rich cathode materials as a
127. Li, N. et al. Unraveling the cationic and anionic redox reactions in a conventional layered strategy to mitigate voltage hysteresis. Nat. Mater. 22, 353–361 (2023).
oxide cathode. ACS Energy Lett. 4, 2836–2842 (2019). This paper proposes cation-disordered structures to enhance reversibility, effectively
128. Lebens-Higgins, Z. W. et al. Revisiting the charge compensation mechanisms in constraining the collective effects of cation migration.
LiNi0.8Co0.2−yAlyO2 systems. Mater. Horiz. 6, 2112–2123 (2019). 161. Zhai, X.-Z. et al. Layered birnessite cathode with a displacement/intercalation
129. Lee, G. et al. Reversible anionic redox activities in conventional LiNi1/3Co1/3Mn1/3O2 mechanism for high-performance aqueous zinc-ion batteries. Nanomicro Lett. 12, 56
cathodes. Angew. Chem. Int. Ed. 59, 8681–8688 (2020). (2020).
130. Zhuo, Z. et al. Cycling mechanism of Li2MnO3: Li–CO2 batteries and commonality on 162. Nam, K. W. et al. Critical role of crystal water for a layered cathode material in sodium ion
oxygen redox in cathode materials. Joule 5, 975–997 (2021). batteries. Chem. Mater. 27, 3721–3725 (2015).
131. Lee, H. J., Shin, J. & Choi, J. W. Intercalated water and organic molecules for electrode 163. Jo, M. R. et al. Triggered reversible phase transformation between layered and spinel
materials of rechargeable batteries. Adv. Mater. 30, 1705851 (2018). structure in manganese-based layered compounds. Nat. Commun. 10, 3385 (2019).
132. Mitchell, J. B., Chagnot, M. & Augustyn, V. Hydrous transition metal oxides for 164. House, R. A. et al. Superstructure control of first-cycle voltage hysteresis in oxygen-redox
electrochemical energy and environmental applications. Annu. Rev. Mater. Res. 53, 1–23 cathodes. Nature 577, 502–508 (2020).
(2023). 165. Zhang, J. et al. Tuning oxygen redox chemistry in li‐rich Mn‐based layered oxide cathodes
133. Yang, J. H., Kim, H. & Ceder, G. Insights into layered oxide cathodes for rechargeable by modulating cation arrangement. Adv. Mater. 31, 1901808 (2019).
batteries. Molecules 26, 3173 (2021). 166. Yamamoto, K. et al. Charge compensation mechanism of lithium-excess metal oxides
134. Zhao, F. et al. Synergistically controlled mechanism of sodium birnessite with a larger with different covalent and ionic characters revealed by operando soft and hard X-ray
interlayer distance for fast ion intercalation toward sodium-ion batteries. J. Phys. Chem. C absorption spectroscopy. Chem. Mater. 32, 139–147 (2020).
124, 28431–28436 (2020). This paper establishes the effect of bond character on the stability of oxygen redox.
135. Kundu, D., Adams, B. D., Duffort, V., Vajargah, S. H. & Nazar, L. F. A high-capacity and long- 167. Myeong, S. et al. Understanding voltage decay in lithium-excess layered cathode
life aqueous rechargeable zinc battery using a metal oxide intercalation cathode. Nat. materials through oxygen-centred structural arrangement. Nat. Commun. 9, 3285 (2018).
Energy 1, 16119 (2016). 168. Wang, Q. et al. Unlocking anionic redox activity in O3-type sodium 3d layered oxides via
136. Charles, D. S. et al. Structural water engaged disordered vanadium oxide nanosheets for Li substitution. Nat. Mater. 20, 353–361 (2021).
high capacity aqueous potassium-ion storage. Nat. Commun. 8, 15520 (2017). 169. Zhao, C. et al. Rational design of layered oxide materials for sodium-ion batteries.
137. Nam, K. W., Kim, H., Choi, J. H. & Choi, J. W. Crystal water for high performance layered Science 370, 708–711 (2020).
manganese oxide cathodes in aqueous rechargeable zinc batteries. Energy Environ. Sci. 170. Ma, J. et al. Ordered and disordered polymorphs of Na(Ni2/3Sb1/3)O2: honeycomb-ordered
12, 1999–2009 (2019). cathodes for Na-ion batteries. Chem. Mater. 27, 2387–2399 (2015).
138. Levi, E., Gofer, Y. & Aurbach, D. On the way to rechargeable Mg batteries: the challenge 171. Susanto, D. et al. Anionic redox activity as a key factor in the performance degradation of
of new cathode materials. Chem. Mater. 22, 860–868 (2010). NaFeO2 cathodes for sodium ion batteries. Chem. Mater. 31, 3644–3651 (2019).
139. Lukatskaya, M. R. et al. Ultra-high-rate pseudocapacitive energy storage in two- 172. Yu, Y. et al. Triggering reversible anion redox chemistry in O3-type cathodes by tuning
dimensional transition metal carbides. Nat. Energy 2, 17105 (2017). Na/Mn anti-site defects. Energy Environ. Sci. 16, 584–597 (2023).
140. Zhang, K. et al. Pseudocapacitive behavior and ultrafast kinetics from solvated ion 173. Kang, K. & Ceder, G. Factors that affect Li mobility in layered lithium transition metal
cointercalation into MoS2 for its alkali ion storage. ACS Appl. Energy Mater. 2, 3726–3735 oxides. Phys. Rev. B 74, 094105 (2006).
(2019). 174. Yu, H. et al. Study of the lithium/nickel ions exchange in the layered LiNi0.42Mn0.42Co0.16O2
141. Park, J., Kim, S. J., Lim, K., Cho, J. & Kang, K. Reconfiguring sodium intercalation process cathode material for lithium ion batteries: experimental and first-principles calculations.
of TiS2 electrode for sodium-ion batteries by a partial solvent cointercalation. ACS Energy Energy Environ. Sci. 7, 1068 (2014).
Lett. 7, 3718–3726 (2022). 175. Sun, G. et al. The effect of cation mixing controlled by thermal treatment duration on the
142. Reimers, J. N., Rossen, E., Jones, C. D. & Dahn, J. R. Structure and electrochemistry of electrochemical stability of lithium transition-metal oxides. Phys. Chem. Chem. Phys. 19,
LixFeyNi1-yO2. Solid State Ion. 61, 335–344 (1993). 29886–29894 (2017).
143. Lee, J. et al. Unlocking the potential of cation-disordered oxides for rechargeable lithium 176. Xiao, Y. et al. Insight into the origin of lithium/nickel ions exchange in layered
batteries. Science 343, 519–522 (2014). Li(NixMnyCoz)O2 cathode materials. Nano Energy 49, 77–85 (2018).
This paper explores possible local Li diffusion and long-range percolation paths in This paper investigates the advantageous effects of Ni2+/Li+ mixing to alleviate
rocksalt materials. magnetic frustration, thereby enhancing structural stability.
144. Ruetschi, P. & Giovanoli, R. Cation vacancies in MnO2 and their influence on 177. Ménétrier, M. et al. NMR evidence of LiF coating rather than fluorine substitution in
electrochemical reactivity. J. Electrochem. Soc. 135, 2663–2669 (1988). Li(Ni0.425Mn0.425Co0.15)O2. J. Solid State Chem. 181, 3303–3307 (2008).
145. Powell, M. J. Site percolation in randomly packed spheres. Phys. Rev. B 20, 4194–4198 178. Richards, W. D., Dacek, S. T., Kitchaev, D. A. & Ceder, G. Fluorination of lithium‐excess
(1979). transition metal oxide cathode materials. Adv. Energy Mater. 8, 1701533 (2018).

Nature Reviews Chemistry

Review article

179. Chen, R. et al. Li+ intercalation in isostructural Li2VO3 and Li2VO2F with O2− and mixed 194. Flores, E., Mozhzhukhina, N., Aschauer, U. & Berg, E. J. Operando monitoring the
O2−/F− anions. Phys. Chem. Chem. Phys. 17, 17288–17295 (2015). insulator–metal transition of LiCoO2. ACS Appl. Mater. Interfaces 13, 22540–22548 (2021).
180. Takeda, N. et al. Reversible Li storage for nanosize cation/anion-disordered rocksalt-type 195. Chen, Q. et al. Zircon structure as a prototype host for fast monovalent and divalent ionic
oxyfluorides: LiMoO2 – x LiF (0≤x≤2) binary system. J. Power Sources 367, 122–129 (2017). conduction. Chem. Mater. 35, 6313–6322 (2023).
181. Lee, J. et al. Mitigating oxygen loss to improve the cycling performance of high capacity
cation-disordered cathode materials. Nat. Commun. 8, 981 (2017). Acknowledgements
182. Lun, Z. et al. Improved cycling performance of Li‐excess cation‐disordered cathode This work was supported by the National Research Foundation of Korea (NRF) grant funded
materials upon fluorine substitution. Adv. Energy Mater. 9, 1802959 (2019). by the Korea government Ministry of Science and ICT (MSIT) (NRF-2022R1A2B5B03001781)
183. Clément, R. J., Kitchaev, D. & Lee, J. & Ceder, G. Short-range order and unusual modes and the Nano & Material Technology Development Program through the NRF funded by the
of nickel redox in a fluorine-substituted disordered rocksalt oxide lithium-ion cathode. MSIT (RS-2023-00282389). Y.-M.K acknowledges the partial financial support from LG Energy
Chem. Mater. 30, 6945–6956 (2018). Solution.
184. Jiao, S. et al. The mechanism of fluorine doping for the enhanced lithium storage
behavior in cation‐disordered cathode oxide. Adv. Energy Mater. 13, 2301636 (2023). Author contributions
185. Lun, Z. et al. Design principles for high-capacity Mn-based cation-disordered rocksalt S.K. and S.L. contributed equally. All authors contributed to the discussion of the content and
cathodes. Chem 6, 153–168 (2020). the writing of the manuscript.
186. Sun, Y. et al. Expandable Li percolation network: the effects of site distortion in cation-
disordered rock-salt cathode material. J. Am. Chem. Soc. 145, 11717–11726 (2023). Competing interests
187. Urban, A., Lee, J. & Ceder, G. The configurational space of rocksalt-type oxides for The authors declare no competing interests.
high-capacity lithium battery electrodes. Adv. Energy Mater. 4, 1400478 (2014).
188. Ji, H. et al. Hidden structural and chemical order controls lithium transport in Additional information
cation-disordered oxides for rechargeable batteries. Nat. Commun. 10, 592 (2019). Peer review information Nature Reviews Chemistry thanks the anonymous reviewers for their
189. Lun, Z. et al. Cation-disordered rocksalt-type high-entropy cathodes for Li-ion batteries. contribution to the peer review of this work.
Nat. Mater. 20, 214–221 (2021).
190. Squires, A. G. & Scanlon, D. O. Understanding the limits to short-range order suppression Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
in many-component disordered rock salt lithium-ion cathode materials. J. Mater. Chem. A published maps and institutional affiliations.
11, 13765–13773 (2023).
191. Wang, Y. et al. Unraveling the nature and role of layered cation ordering in cation- Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
disordered rock-salt cathodes. J. Am. Chem. Soc. 144, 19838–19848 (2022). article under a publishing agreement with the author(s) or other rightsholder(s); author self-
192. Yang, J. H. & Ceder, G. Activated internetwork pathways in partially‐disordered spinel archiving of the accepted manuscript version of this article is solely governed by the terms of
cathode materials with ultrahigh rate performance. Adv. Energy Mater. 13, 2202955 (2023). such publishing agreement and applicable law.
193. Rutt, A. et al. Novel structural motif to promote Mg-ion mobility: investigating
ABO4 zircons as magnesium intercalation cathodes. ACS Appl. Mater. Interfaces 15, © Springer Nature Limited 2024
34983–34991 (2023).

Department of Materials Science and Engineering, Korea University, Seoul, Republic of Korea. 2Department of Battery-Smart Factory, Korea University,
1

Seoul, Republic of Korea. 3KU-KIST Graduate School of Converging Science and Technology, Korea University, Seoul, Republic of Korea. 4These authors
contributed equally: Seongkoo Kang, Suwon Lee.

Nature Reviews Chemistry