Recent advances of metal fluoride compounds cathode materials for lithium ion batteries: a review

As the most successful new energy storage device developed in recent decades, lithium-ion batteries (LIBs) are ubiquitous in the modern society. However, current commercial LIBs comprising mainly intercalated cathode materials are limited by the theoretical energy density which cannot meet the high storing energy demanded by renewable applications. Compared to intercalation-type cathode materials, low-cost conversion-type cathode materials with a high theoretical specific capacity are expected to boost the overall energy of LIBs. Among the different conversion cathode materials, metal fluorides have become a popular research subject for their environmental friendliness, low toxicity, wide voltage range, and high theoretical specific capacity. In this review, we compare the energy storage performance of intercalation and conversion cathode materials based on thermodynamic calculation and summarize the main challenges. The common conversion-type cathode materials are described and their respective reaction mechanisms are discussed. In particular, the structural flaws and corresponding solutions and strategies are described. Finally, we discussed the prospective of metal fluorides and other conversion cathode materials to guide further research in this important field.

Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence.Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

Introduction
Spurred by the rapid development of alternative energy technology, lithium-ion batteries (LIBs) have become the most important electrochemical energy sources on account of the large energy density, high working voltage, and environment-friendliness [1][2][3][4][5].Applications span mobile intelligent devices to hybrid/electric vehicles and largescale complementary energy storage devices, LIBs [6][7][8][9][10].However, new applications demand better energy sources and LIB technology is facing challenges including the energy density, safety, and cost.In particular, the overall energy density is the biggest bottleneck barrier limiting expanded applications of LIBs [11][12][13][14][15][16].This is because current commercial LIB cathode materials often have a lower discharge specific capacity than the corresponding anode materials and the imbalance or mismatch makes it difficult to the higher energy demand by new applications [17][18][19][20][21].According to Liebig's Law of the Minimum, the crux of breaking the shackles of the energy density of LIBs lies in the technological breakthrough of the electrode materials, especially the cathode materials [22].Therefore, development of high-performance cathode materials with a high theoretical specific capacity, wide voltage range, high energy density, and low cost is crucial to the development of LIBs [23][24][25][26].
The cathode materials in commercial LIBs are mainly intercalation-type cathode materials [27][28][29].However, they are seriously hampered by the limited theoretical capacity [28,[30][31][32][33]. Recently, Li-rich cathode materials with high theoretical capacities have attracted attention, especially improving the cycling and rate performance [34][35][36][37].In addition to reaction-based cathode materials, economical Li-free cathode materials have high theoretical capacities and can be operated in a wide voltage range [38][39][40].Conversion cathode materials have received less attention than insertion cathode materials due to issues such as the intrinsic conductivity of the materials and other technical challenges.Therefore, research on conversion cathode materials is still insufficient.Considering the operating voltage, energy density, and cost, metal fluorides are suitable conversion cathode materials for new-generation of LIBs boasting advantages such as sustainable production, safety, environmental friendliness, and high energy density [41][42][43][44][45].In this review, the electrochemical conversion mechanism and structural defects of metal fluoride cathode materials are described.Issues such as elemental substitution, interfacial modification, composite electrode materials, and so on are discussed in detail.Finally, the challenges and future are proposed to provide guidance to future research for this important topic.

Method of calculation
In general, compared with intercalated cathode materials, conversion cathode materials have a higher theoretical specific capacity because the conversion reaction makes full use of the valence state change in the oxidation-reduction reaction.The typical conversion reaction for LIBs composed of metal fluoride can be represented by the following equation [14,40,46,47]: MF x + yLi ↔ yLiF x/y + M(M = transition metal), (1) where M is a transition metal (Fe, Co, Ni, etc) and the anion can be a reducing element (O, S, P, Cl, etc) except F. The energy storage capacity of different conversion-type materials depends on the anion species and oxidation state of the transition metals.The standard reaction Gibbs free energy in reaction ( 1) is calculated by the following equation: Under the reversible reaction conditions, the maximum electrical work provided by the reaction corresponds to the Gibbs free energy of the reaction as shown in the following: where n is the number of electrons in the cell reaction for 1 mol, F is Faraday's constant (96 485 C mol −1 ), and U is the cell voltage.The theoretical specific capacity follows the following relationship: Theoretical specific capacity = nF/3.6M.( The unit of theoretical specific capacity is mAh g −1 , and M is the relative molecular mass of the electrode material. The energy density of the battery can be derived in terms of gravimetric energy density (TGED Wh kg −1 ) and volumetric energy density (Wh L −1 ).The gravimetric energy density is calculated as follows: where ∑ 2 i =A,B a i M i is the sum of the molar mass of the reactants.Similarly the gravimetric energy density can be derived according to E as shown in equations ( 3) and (4) as follows: The volumetric energy density (TVED) can be calculated by the following equation: where ∑ 2 i a i V i is the sum of the molar volumes of the reactants and ρ is the materials density.
According to equations ( 1)- (7), the working voltage, theoretical specific capacity, gravimetric energy density, and volumetric energy density of batteries containing metal fluoride cathodes and lithium metal anodes are summarized in figures 1(a)-(c).Compared to traditional intercalated cathode materials, conversion cathode materials clearly show theoretical capacities and energy densities with higher application potential.However, in recent decades, the development of metal fluorides has been hampered by its own physical and material properties, such as the dark blue cloud in figure 1(c).Under this cloud, researchers have still made representative results and breakthroughs in the mechanism.Therefore, it is necessary to understand the challenges to be faced when using conversion cathode materials, especially metal fluorides, as electrode materials.

Low conductivity.
Under optimal conditions, electrode materials should have favorable ionic and electronic conductivity to achieve superior energy storage utilization.Unfortunately, all-metal fluorides are insulators with poor intrinsic conductivity [39,48] mainly due to the strong ionic interactions between the transition metal elements and fluoride ions resulting in a wide band gap [49][50][51][52][53].The low electrical conductivity impacts the reversibility, Coulombic efficiency, reaction kinetics, and rate capability of the materials.Although some metal fluorides contain metal nanoparticles formed by the conversion reaction and the conductivity improve, the overall conductivity is normally unsatisfactory [54,55].

Large voltage hysteresis.
Conversion-type metal fluoride cathode materials exhibit severe overpotential and voltage hysteresis [40,[56][57][58].The overpotential is mainly the voltage deviation between the resting electrode potential and actual potential in the reaction as shown in figure 2 for FeF 3 , where η represents the deviation between the theoretical potential and actual potential of FeF 3 .The overpotential arises from a series of polarization steps.As for the voltage hysteresis of metal fluorides, the SEI film on the electrode surface is the culprit [39,40].In metal fluoride conversion reactions, chemical bonds are broken during charging and discharging and LiF and transition metal nanoparticles with different structures and lower free energy than the original phase are formed.However, when the conversion reaction proceeds repeatedly, these new phases need to overcome higher activation potential barriers during formation and separation.Since metal fluorides have higher activation potential barriers than other conversion materials, voltage hysteresis and worse reaction kinetics result.

Volumetric expansion.
In the electrode reactions, the mechanism of metal fluorides is more complex, especially when new phases are generated in the reversible reactions which can change the crystal and material structure [55].Besides, shuttling of Li + changes the lattice causing volume expansion or contraction [40,59,60].Compared to the large volume expansion of sulfur cathodes in Li-S batteries, the volume change of metal fluoride is sufficient to negatively affect the battery properties generally in the range of 2% to 50% [24,39,40,61,62].The volume change affects the mechanical and electrochemical stability of the materials.During charging and discharging, the excessive volume change causes the electrode materials to fragment and pulverize and the active particles detach from the binder and collector, consequently increasing the contact surface with the electrolyte and raising the probability of adverse side reactions [63,64].In addition, expansion of the materials inevitably lead to larger thicknesses, which will trigger cell expansion causing deterioration in the cycling stability, voltage hysteresis, Coulombic efficiency, as well as reversibility.

Side reactions with electrolytes.
Capacity degradation of the intercalated cathode materials can be attributed side reactions with the electrolyte.The typical spinel-type LiMn 2 O 4 cathode in an electrolyte containing LiPF 6 can be attacked chemically by HF as a result of the reaction between LiPF 6 and water and the side reaction becomes more violent at a higher temperature [11,[65][66][67].Meanwhile, owing to the disproportionation reaction of Mn 3+ , the reduced Mn 2+ is deposited on the electrode surface in the form of compounds or atomic deposits.The surface resistance increases and the deposits may further catalyze decomposition of the electrolyte to affect the electrochemical characteristics [11,[68][69][70][71]. Unfortunately, almost all conversion cathode materials are soluble in polar organic solvents to some extent and side reactions similar to those on the intercalated cathodes [72][73][74][75].Even though the reactivity of most metal fluoride cathode materials is modest thus avoiding to some extent electrolyte decomposition at high electrode potentials, the lower working potential may lead to a reduction of the electrolyte on the metal fluoride surface [39,74].Meanwhile, during the conversion reaction, the reduced transition metal nanoparticles with high surface energy and strong catalytic activity further catalyze decomposition of the electrolyte gradually forming a cathode electrolyte interface (CEI) on the cathode [76].The CEI seriously affects movement of lithium ions and as the conversion reaction continues, the thickness of the interfacial layer increases.Consumption of the electrolyte continues and the electrochemical properties deteriorate significantly.
The physical and electrochemical properties of common transition metal fluorides are summarized in table 1. Examples include the theoretical capacity of the material, the theoretical working potential, the band gap, the theoretical volume expansion, the voltage hysteresis, and the stability in the electrolyte.

FeF 2
The rutile-type FeF 2 belongs to the tetragonal crystal system with a P4 2 /mnm space group as shown in figure 3(a).The anions are arranged in a regular octahedron by compression of the a-axis and the metal ions are surrounded by a twisted octahedron consisting of six fluoride ions.The theoretical specific capacity of rutile FeF 2 is 571 mAh g −1 and the discharge potential plateau is about 2.66 V [77,78].Unlike most fluorides, rutile FeF 2 does not undergo reversible conversion of alkali metal ions during charging and discharging but instead direct phase conversion reactions as shown in the equation [79][80][81]: The charge storage mechanism of FeF 2 was further investigated using in situ magnetometry by Hu et al.It was found that FeF 2 did not completely complete the conversion reaction in the voltage interval of 1.0-4.0V, in which the reaction is shown below: Step Step 2 : nLi + + Fe/LiF+ne − ↔ Fe n− /nLi + /LiF.(10) It is shown that the spin-polarized surface capacitance and conversion reactions of FeF 2 dominate the charge storage at low and high operating voltages, respectively, and that these two behaviors together determine the electrochemical performance of FeF 2 .
As shown in figure 3(g), the conversion reaction begins on the surface of FeF 2 and lithiation proceeds by layer-bylayer propagation [82].The FeF 2 conversion reaction is slow  Reprinted with permission from [54].Copyright (2011) American Chemical Society.(g) Schematic illustration of the diffusion of the reaction front in a single FeF 2 particle, using the 'layer-by-layer' reaction as the mechanism.Reproduced from [82], with permission from Springer Nature.
compared to the fast intercalation reaction due to the low diffusivity of Li + in the 110 channel in FeF 2 [83].As schematic figure 3(b) and structural characterization figures 3(c)-4(f) show in the reaction between FeF 2 and Li + , tiny Fe nanoparticles (<5 nm) nucleate around the LiF phase because of the low mobility of Fe 2+ .These Fe nanoparticles are interconnected in the form of almost the same frame structure as the original one to form a bi-continuous conductive network that provides local electron transport channels for insulating LiF [54].In addition, the lattice on the nanoscale solid-phase interface provides effective ion transport channels in the phase transition, thus allowing the crystal structure of FeF 2 to maintain a certain degree of reversibility during discharging [55].
However, the unmodified FeF 2 cathode does not show levels consistent with the theoretical capacity on account of the insulating properties of difluoride.

FeF 3
The ReO 3 -type FeF 3 belongs to the R 3c space group with a trigonal crystal system as shown in figure 4(a).The crystal structure is similar to that of the perovskite-type ABX 3 , which is interconnected by co-vertex FeF 6/2 octahedra to form a threedimensional (3D) tunnel-like structure [84,85].Fe 3+ is located at the (102) crystal plane of the rhombic structure and the vacant sites located in the (204) crystal plane are available for insertion of lithium ions [85].In general, electrochemical lithium storage in ReO 3 -FeF 3 proceeds by two steps [52,[86][87][88]: and Reaction (11) proceeds by insertion of lithium in the operating voltage range of 4.5-2.0V. Li + is reversibly inserted into the ReO 3 -FeF 3 lattice and after 0.5 Li + is inserted, the lattice begins to change.With gradually decreasing operating voltages (2.0-1.5 V), the mechanism changes from Li + insertion to a conversion reaction (reaction (12)) and LiF with poor conductivity is formed [89].In fact, as shown in figure 4(b), FeF 3 undergoes a more complex phase transition with the change of potential during the electrochemical reaction, with the consequent formation of more thermodynamic and nonthermodynamic structural phases [41].Overall, three electrons are involved in the two-step reaction and so the theoretical specific capacity of ReO 3 -type FeF 3 is 712 mAh g −1 .However, FeF 3 does not have the desired electrochemical properties due to the poor conductivity, complex reaction, as well as significant voltage hysteresis arising from the complex phase change.

FeFx • mH 2 O
Iron-based fluoride (FeF x • mH 2 O) containing crystalline water is mainly FeF 3 • 0.33H 2 O in the hexagonal tungsten bronze phase and pyrochlore type FeF 3 • H 2 O and FeF 3 • 3H 2 O.These fluorides have different stoichiometric ratios of water of crystallization leading to higher conductivity, so that they have slightly higher electrochemical activity than the pure-phase iron fluoride [42,90].The voltage range of 1.7-4.5 V is a common choice for iron-based fluoride cathode materials containing water of crystallization.Different from the pure phase FeF x , the electrochemical mechanism of FeF x • mH 2 O is dominated by intercalation/deintercalation.For instance, the reactions of FeF 3 • 0.33H 2 O are shown in the following [91][92][93]: and In most cases, the electrochemical reaction is carried out based on equation ( 13) in order to avoid the negative impact of water molecules in the electrolyte.

HTB-FeF
The hexagonal tungsten bronze phase of FeF 3 • 0.33H 2 O belongs to the orthorhombic crystal system with a Cmcm space group and lattice parameters of a = 7.423 Å, b = 12.73 Å, and c = 7.526 Å as well as cell volume V = 711.17Å 3 [94].Figure 5(a) shows the crystal structure of FeF 3 • 0.33H 2 O.In the structure, one Fe 3+ is connected to six F − to form a regular FeF 6 n-octahedron and on the (001) crystal plane, six FeF 6 noctahedra are connected by sharing F − vertices to form a large hexagonal cavity [95].The cavity has an average van der Waals radius of 1.8 Å, which is larger than that of the H 2 O molecule (1.5 Å) and also the ion channel radius of Li + (0.76 Å) [96].The hexagonal cavities in FeF 3 • 0.33H 2 O constitute a unique one-dimensional (1D) tunnel structure which facilitates transport and storage of lithium ions in the crystal.Hence, FeF 3 • 0.33H 2 O has garnered attention as ironbased fluoride materials.Unlike ReO 3 -FeF 3 , FeF 3 • 0.33H 2 O forms a single-phase solid solution Li x FeF 3 • 0.33H 2 O through reversible intercalation of Li + during discharging.The inserted Li + is reversibly removed during charging giving rise to electrochemical conversion [97].
The crystal structure and crosssectional tetrahedra diagram of pyrochlore-type FeF 3 • 0.5H 2 O is shown in figures 5(b) and (c).Fe 3+ is bonded to six F − to form the ortho-octahedral FeF 6 and six ortho-octahedral FeF 6 are linked by one F atom to form a giant hexagonal cavity similar to that in the FeF 3 • 0.33H 2 O structure.Four closely packed FeF 6 orthoctahedra are linked to form one (FeF 6 ) 4 orthotetrahedron, whereas four (FeF 6 ) 4 orthotetrahedral units are connected vertically to form the pyramidal (FeF 6 ) 16 oversized orthotetrahedral structure with interconnected cavities, which together with the interlocking atoms form a continuous 3D crossed microporous framework [99].This structure helps to unblock electron transfer channels caused by water molecules and facilitates Li + transport.The large cell volume (1127-1131 Å 3 ) also bodes well for embedding large particles such as Na + (Na + radius 1.02 Å), thereby allowing rapid transfer in the structure [98,100].
belongs to the tetragonal crystal system of the P4/n space group with a = 7.832 Å, c = 3.877 Å, and unit cell volume V = 237.87Å 3 as shown in figures 5(d) and (e).In the unit cell, one Fe 3+ is surrounded by four F − and two H 2 O molecules combining to form irregular Fe(F 4 O 2 ) octahedra.The octahedral units are not connected in the [001] direction and four octahedra form a small tetrahedral cavity with a van der Waals radius of 1.2 Å [96].The H 2 O molecule at the center of the tetrahedral cavity interacts with the surrounding ligands by hydrogen bonding to maintain the stability of the structure.In the [100] direction, the irregular Fe(F 4 O 2 ) octahedra form a 1D octahedral chain in a shared F − vertex fashion [101].Because of severe distortion and polarization of the Fe(F 4 O 2 ) octahedra, the tetragonal cavities appear distorted to some extent.FeF 3 (H 2 O) 2 • H 2 O is not very suitable as electrode materials due to the large amount of crystal water [102].Instead, FeF 3 (H 2 O) 2 • H 2 O is commonly used as a precursor in the preparation of iron-based fluoride materials containing different but smaller ratios of crystalline water.

FeOF
Since the Fe-F bond in the Fe-F x series cathode materials causes low conductivity due to the wide bandgap, Fe-O bonds have been used to replace part of the Fe-F bond to reduce the bandgap and increase the intrinsic conductivity [103,104].At the same time, elemental substitution enhances the stability of the crystal structure to improve the electrochemical properties.The crystal structures of FeF 2 in rutile phase and FeOF after being replaced by O atoms are shown in figures 6(a)-(c).In particular, substitution of fluoride by O atoms produces mixed anionic iron fluoride with the general formula FeO x F 2−x .FeO x F 2−x has the advantage of working in a wider working voltage and having better electrical conductivity and cyclic stability.Moreover, FeOF has a larger theoretical capacity (885 mAh g −1 ) than FeF 3 .In electrochemical reactions, O/F in FeOF provides stable lithiation sites, as shown in figures 6(d)-(g).In addition, the mechanism of FeOF is dominated by intercalation (15), expulsion ( 16), ( 17), and conversion (18) according to the following reactions and phase changes [105][106][107]: Despite the obvious advantages of FeOF as electrode materials, the main problem with FeOF is the complex and non-scalable synthesis process compared to other iron-based fluoride materials (FeF 2 , FeF 3 , and FeF 3 • 0.33H 2 O).Even though substitution of oxygen can improve the conductivity of FeF x , it is still difficult to satisfy the requirements for cathode materials.Furthermore, reactions ( 15)-( 18) involve complex structural and phase changes during charging and discharging, especially the conversion reaction in reaction (18), consequently leading to severe cohesion failure and pulverization.

CuF 2
By using CuF 2 in transition metal fluoride as the cathode materials for LIBs, the theoretical voltage reaches 3.55 V.The theoretical specific capacity is 528 mAh g −1 and the energy density is 1874 Wh kg −1 , which is at least twice that of the theoretical specific capacity of traditional intercalated electrode materials [112,113].The crystal structure is shown in figures 6(h) and (i).Despite the higher operating voltage, the actual energy density is only 25% of the theoretical energy density.The poor electrochemical reversibility of CuF 2 without crystalline water hampers further development for LIBs.Omenya et al [111] have prepared Fedoped Cu 1−y Fe y F 2 /C (y = 0, 0.2, 0.5, 1) nanocomposites by high-energy ball milling and investigated the electrochemical reconversion mechanism for Fe-substituted CuF 2 revealing challenges with CuF 2 as cathode materials in Li-ion batteries.The electrochemical reaction occurring in CuF 2 is shown in the following: It is challenging to improve the electrochemical reversibility of CuF 2 for LIBs because of the following reasons: (1) oxidation potential of Cu being less than 3.5 V, which is significantly lower than the CuF 2 reconversion voltage, (2) excessive diffusion of Cu ions, and (3) the ability of Cu ions to move through the electrolyte.Coating CuF 2 may not be a viable solution since Cu ions can still diffuse into the electrolyte through the coating and then be deposited reductively on the anode.This results in severe loss of active materials and impacts the electrochemical reversibility (figure 6(j)).Therefore, the modification scheme should focus on finding better electrolytes, especially solid-state electrolytes that preferentially transport Li ions rather than oxidized Cu ions.Nevertheless, the diffusivity of Cu ions and similarity between the ionic radii of Cu and Li ions are practical challenges for solid-state electrolytes.In addition, even if a better electrolyte can be identified to specifically transport lithium, copper may still separate from LiF in the cathode instead of forming a continuous copper network in LiF forming Cu particles and inhibiting the conversion reaction of CuF 2 .
Recently, Xia et al [114] used water as the solvent and sodium alginate (SA) as the binder in the preparation of CuF 2 electrode materials, which led to the formation of a Cu 2+coordinated Cu-SA layer on the surface of CuF 2 particles, which had the effect of inhibiting the dissolution of copper, and the process of the preparation of the material and the coordination structure of the Cu-SA are shown in figures 7(a) and (c), respectively.In the aqueous solvent, the in situ reaction between SA and copper ions generates a strong Cu-SA layer on the surface of CuF 2 particles.The Cu-SA layer is selectively permeable to Li + but impermeable to Cu 2+ , which effectively reduces the degree of solubilization of copper ions and their negative effects in the carbonate electrolyte environment (figure 7(b)), and at the same time provides a certain degree of structural stability.The reversible capacity was 420.4 mAh g −1 after 50 cycles at a current density of 0.05 C.Although the electrochemical performance of CuF 2 -SA is still not ideal compared with other fluoride electrode materials, it is a promising beginning for the study of CuF 2 as an electrode material.

Others
In addition to iron and copper metal fluorides, other transition metal fluorides have been studied as cathode materials for LIBs, for instance, TiF 3 [115], MnF 2 [116,117], NiF 2 [118,119], CoF 2 [78,120,121], BiF 3 [122][123][124], etc.These transition metal fluorides have potential as cathode materials for LIBs in terms of theoretical capacity and energy density, but they are also severely constrained by issues plaguing large-scale applications in the field of energy storage.In order to solve the problems of low conductivity, significant voltage hysteresis, and volume expansion of electrode materials of transition metal fluorides, materials modification has been proposed.
For example, CoF 2 , which has excellent theoretical capacity (553 mAh g −1 ) and energy density (2038 mAh cm −3 ), has also been widely studied and reported as an cathode material for LIBs [125,126].Similar to FeF 2 , the source of the electrochemical capacity contribution of CoF 2 is mainly purely a conversion reaction mechanism, and the commonly reported equations for CoF 2 conversion reactions are as follows [127]: However, due to the larger bandgap of the CoF 2 (4.44 eV) itself and the LiF with a larger bandgap generated by the conversion reaction, both of which will lead to CoF 2 as an electrode material will exhibit obvious defects such as poor reaction reversibility and low conductivity.Similarly, MnF 2 in transition metal fluoride faces the same problem.Although MnF 2 has a high theoretical capacity of 557 mAh g −1 and a power density of 1519 Wh kg −1 , the intrinsic bandgap of rutile-phase MnF 2 is larger than that of CoF 2 (7.3-10 eV), which severely impacts the depth of its application as a conversion-type cathode material for Li-ion batteries [116,117,128,129].NiF 2 has the same disadvantages, although NiF 2 has a higher theoretical reaction voltage of 2.96 V vs. Li/Li + compared to other transition metal fluorides, the same intrinsic band gap problem will lead to low conductivity and cause voltage hysteresis, which will seriously hamper the smooth electrochemical reaction of NiF 2 [118,130].
Compared to FeF 3 , other transition metal trifluorides such as TiF 3 and BiF 3 have been relatively less studied for applications in the field of LIBs.According to the current study, the electrochemical reaction mechanism of TiF 3 and BiF 3 is similar to that of FeF 3 , and theoretically, if a near or complete three-electron conversion reaction occurs, the embedded deembedding reaction is usually dominant at the early stage of the reaction (1Li), and the conversion reaction will be the dominant one in the subsequent reaction (>1Li).As an example, the overall reaction process of TiF 3 is as follows [131]: Initial charge process: From the performance index of theoretical capacity alone, TiF 3 and BiF 3 have certain application potentials in the field of LIB cathode materials, but the few researches so far have not adequately solved the problems such as extremely fast capacity decay and low Coulombic efficiency of TiF 3 and BiF 3 during cycling.Although the weak volume change of BiF 3 during charging and discharging can indeed give it a certain advantage in the application field of electrode materials, the poor electrochemical performance of the material as a whole is often not only attributed to the material itself, because the electrochemical performance is a manifestation of the concentration of the integrated factors of the entire battery reaction system.For example, undesirable side reactions between the active materials and the electrolyte, by-products from these surface reactions, can progressively impede and deteriorate the electrochemical reactions, making it more difficult for the material to exhibit acceptable electrochemical properties.

Preparation of metal fluorides
Metal fluoride conversion cathode materials can be fabricated by different techniques such as the solid-state method [132], precipitation [100,133], solvothermal method [134], sol-gel method [135], ball milling method [136], electrochemical method [115], chemical de-alloying [137], and so on.Among them, the solid-state method is a common and simple method to synthesize cathode materials on a large scale.If the conversion fluoride cathode materials are synthesized by a one-step solid-state method, it is often necessary to add some conducting materials such as carbon to improve the conductivity.However, if controlled the particle size is tactic to improve the electrochemical performance, then the shortcomings of traditional solid-state methods are obvious.Compared to traditional solid-state methods, chemical precipitation, high-energy ball milling, and solvothermal and sol-gel methods have advantages in the synthesis of nanoscale electrode materials.To prepare fluoride cathode materials by chemical precipitation, the reaction between the fluorine-containing precipitant and metal salt is commonly adopted and the particle size can be adjusted by controlling the precipitation conditions.High-energy ball milling is often employed to prepare fluoride/carbon nanocomposites.In the solvothermal method, the physicochemical conditions can be changed to influence crystal growth so that the particle morphology and size can be tailored [138].In addition, lowtemperature synthesis by calcination-free procedures and controlling microstructures is widely used in the preparation of fluoride cathode materials.An example of a low-temperature ionic liquid method using BMIMBF 4 as a fluorine source is shown in figure 8(a) [139].The non-aqueous chemical synthesis proceeds at a mild reaction temperature to obtain crystalline nanostructured materials.Compared with corrosive fluorine sources, ionic liquid fluorine sources are environmentally friendly and safe to operate.The unique properties can influence particle aggregation and growth, making it easier to obtain monodispersed nanocrystals.Li et al [140] have prepared FeF 3 • 0.33H 2 O/graphene nanosheets by the in situ ionic liquid-assisted method using BMIMBF 4 as the fluorine source.The interaction between imidazole cations in BMIMBF 4 and π-cations in the graphene nanosheets leads to good dispersion and better anchoring and prevention of irregular agglomeration of FeF 3 • 0.33H 2 O particles by graphene nanosheets.Electrochemical tests reveal the excellent rate properties of the FeF 3 • 0.33H 2 O/graphene nanosheet nanomaterials such as a discharge capacity of 74 mAh g −1 at a high current density of 40 C. Lai et al [141] prepared porous iron fluoride cathode materials for LIBs by deep eutectic solvent method using mild NH 4 HF 2 as a fluorine source.This method does not require harsh reaction conditions, is inexpensive and easy to operate.The preparation process and electrochemical properties are shown in figure 8(b).The rate performance test shows that the material exhibits excellent structural and electrochemical stability.
In addition to low-temperature synthesis, large-scale synthesis of ordered particles has been expanded.Fan et al [142] have used spray pyrolysis to prepare pomegranate-like nanostructured FeCo/LiF/C (figure 9(a)).The nanostructure is composed of a carbon layer with a thickness of 2-3 nm, metal nanoparticles with a diameter of about 10 nm, LiF nanoparticles with a diameter of about 20 nm (figure 9(e)), and porous carbon spheres with a thickness of 100-1000 nm (figures 9(b)-(d)).This nanostructure increases the effective  contact area between the active particles and conductive carbon and shortens the diffusion distance of Li + .The 3D carbon matrix possesses a certain degree of flexibility to alleviate the mechanical stress generated by the volume expansion of the electrode materials during the electrochemical reaction and inhibits irregular aggregation of metal particles.Hence, the materials maintain high structural integrity during cycling.The garnet-like nanoparticles have excellent cycling characteristics including a high specific discharge capacity of 300 mAh g −1 after 100 cycles at a current density of 30 mA g −1 .Spray pyrolysis has the advantage of large-scale scalable preparation of powder materials and the products can have a large interfacial area and small particle size consequently enhancing the charge transfer capability, shortening the ion transport distance, and improving the reversibility in the conversion reactions.
The effects of different synthesis methods on the electrochemical properties of fluoride cathode materials are obvious and proper selection of the fluorine source is crucial.High-temperature fluorination requires heat treatment in a corrosive atmosphere such as F 2 and NF 3 [143,144].If fluorination is completed before the high-temperature treatment, more fluoride can be selected as the source.It is different from the selection of the lithium source in the lithiation process of conventional intercalation-type cathode materials.More inorganic and organic fluorinated substances are used in the preparation of fluorinated materials.Fluorinated inorganic substances include NH 4 F [145,146], HF [92,147,148], CF x [44], NH 4 HF 2 [149,150] [156], trifluoroacetic acid (TFA) [157], metal salts of hexafluoroacetylacetone, and 2-thenoyltrifluoroacetone [78].Fluorinated salts with different phases and compositions often show unexpected effects and therefore the appropriate fluorine source must be selected according to the reaction mechanism.As for organic fluoride salts, some functional groups and elements play a vital role in the crystallinity, dispersibility, and even electrochemical properties of the final product.Table 2 summarizes the metal fluoride cathode materials prepared from different fluorine sources and the electrochemical properties exhibited respectively.Due to the low intrinsic electronic conductivity caused by the large bandgap in metal fluoride cathode materials, elemental doping has become a direct means to solve the conductivity and energy band problems [159].Doping is conducted with metal cations and anions.Metal cations such as Co [160][161][162][163], Ti [164], Mn [165,166], Cu [130,167], and Ni [168,169] 10(c)).Wang et al [167] have substituted Cu into the FeF 2 lattice to form a Cu y Fe y−1 F 2 solid solution by reversible oxidation of Cu and Fe.Cu has a lower energy barrier for nucleation of rutile-like iron fluoride compared to conversion to CuF 2 due to structural similarities, so that the voltage hysteresis improves significantly.
Cesar Villa et al [130] have investigated the effects of Cu substitution of NiF 2 on the electrochemical properties and observed the correlation between structural evolution and electrochemical properties of ternary metal fluorides with different Cu/Ni ratios by in situ transmission electron microscopy (TEM).Proper Cu substitution reduces volume expansion due to the lithium-embedded debonding reaction (or conversion reaction) and also fluorine loss during the debonding process in the synthesis of NiF 2 .As the conversion reaction proceeds, Cu nanoparticles provide the conductive framework to foster the conversion reaction.
Liu et al [168] have prepared Ni-doped FeF 3 • 0.33H 2 O nanospheres by a solvothermal method as shown in figure 11(a).Ni substitution for Fe produces F vacancies without changing the inherent crystal structure, a decrease in bandgap increases the conductivity and expands the ion channels to promote ion and electron transport.8% Ni doping leads to good reversibility, longer-lasting cycling stability, and better rate properties.The first discharge capacity is 412 mAh g −1 in the voltage range of 1.5-4.5 V and current density of 0.25 C.After 100 cycles, Fe 0.92 Ni 0.08 F 3 • 0.33H 2 O still shows a discharge capacity of 264 mAh g −1 and at a current density of 2 C, it has a discharge capacity of 248 mAh g −1 .The Nbsubstituted Fe cathode with a rod-like morphology is prepared by the ionic liquid-assisted solvothermal method as shown in figures 11(b)-(e) [158].The total energy of the crystal structure decreases after 3% Nb substitution and the structural stability improves.The carbon layer formed after ionic liquid carbonization improves the conductivity and suppresses volume expansion resulting in excellent rate and cycling characteristics.The significant improvement in electrochemical performance arises from the synergistic effect of Nb doping and conductivity.

Nonmetallic elements and functional groups.
In addition to metal cations, elemental doping of fluoride materials with anions has been studied.Some nonmetallic elements and functional groups have been introduced to fluoride materials to tailor the ionic interactions between Fe-F [170][171][172][173]. Lee and Kang [172] have prepared nitrogen-doped FeF 3 /C nanocrystals using melamine as the nitrogen source.The strong ionicity of Fe-F is weakened by N doping consequently improving the electrochemical activity, especially the rate performance such as discharge capacities of 114 mAh g −1 at 16.9 C

Surface modification.
In fluoride cathode materials, free transition metals and metallic Lithium are exposed to the electrolyte thus degrading the capacity.The poor electrical conductivity of fluoride materials is also an obvious obstacle and therefore, depositing a coating with an appropriate thickness can overcome the difficulty.Metal oxides, polymeric conductive compounds, fluoride, and carbon materials have been proposed [174][175][176][177][178][179][180][181][182].Surface coatings can stabilize the electrode-electrolyte interface, improve the conductivity, inhibit phase changes, and prevent drastic volume changes during charging and discharging.As shown in figures 13(a) and (b), Zhou et al [183] deposited a layer of amorphous carbon on FeF 2 by chemical vapor deposition, which effectively improved the electron and ion transport properties of the material and suppressed the negative effects of the CEI layer while promoting the positive enhancement of the CEI layer in terms of mechanical properties.The thickness of the  Recently, polymer-derived carbon encapsulated nano FeF 2 (FeF 2 @PDC) electrode materials were prepared by Su et al [184].The PDC shell inhibited electrolyte decomposition and CEI generation.At the first discharge, a Fe 3 O 4 shell was formed in situ on the surface of the positive electrode material, which provided good mechanical strength and protection.The FeF 2 @PDC electrode exhibits excellent multiplicity performance and long cycling performance.After 500 cycles at high current densities of 30 C and 60 C, the material still exhibits reversible capacities of 120 mAh g -1 and 107 mAh g -1 , respectively (1 C = 571 mAh g -1 ).

Composite materials
Fluoride compounds with high conductivity composed of for example, fluoride and carbon are electrode materials.Highenergy mechanical milling is utilized to prepare FeF 3 /C composites to improve the electrochemical properties compared to pristine FeF 3 [185].A variety of carbon-based materials such as carbon black, acetylene black [131,186,187],  carbon nanofibers, carbon nanotubes [188,189], 3D mesoporous carbon [190][191][192][193][194], carbon nanohorns [195], graphite oxide [196], reduced graphene oxide [93,197,198], carbon nanosheets, and bio-substrates [171,199] have been used.Different types of carbonaceous materials have unique structures and properties.For example, pseudocapacitive charge storage can be realized by modulating two-dimensional carbon nanosheets and active particles to improve the electrochemical and kinetic properties of electrode materials, as shown in figure 14(a) [200].The nanosheets with a thickness of about 40 nm form a honeycomb porous interconnected structure that is effectively combined with Fe 3 C@C precursor particles (figure 14(b)), and this fluent interconnected structure reduces the diffusion energy barrier of Li and enhances the pseudocapacitance of redox.The resulting composites exhibit excellent cycling and rate characteristics in both half-cell and full-cell electrochemical tests.
In addition to 1D carbon nanotubes, microporous or mesoporous structures composed of 3D mesoporous carbon show improved diffusion of lithium ions and conductivity.Li et al [192] have synthesized FeF 3 • 0.33H 2 O@CMK-3 nanocomposites by impregnation.The discharge capacity of 111 mAh g −1 at 10 C decreases to 78 mAh g −1 at a high current density at 50 C in addition to excellent stability.The interconnected ordered structure of CMK-3 provides high-speed transfer for embedding and removal of lithium ions while suppressing irregular growth and agglomeration of nanoparticles.Song et al [191] have incorporated FeF 2 -based materials into ordered mesoporous carbon (CMK-3) to produce FeF 2 nanomaterials.In situ transformation between Fe 2 O 3 as a precursor and ordered mesoporous carbon with high electrical conductivity form a well-connected 3D network which is then fluorinated to produce the FeF 2 /CMK-3 composite.The FeF 2 /CMK-3 composite is tested for 1000 cycles at a constant current and the capacity decay per cycle does not exceed 0.3‰.It also has a discharge specific capacity of 500, 400, and 320 mAh g −1 after 100 cycles at rates of 500, 2000, and 4000 mA g −1 in the potential range of 1.5-4.5 V, respectively.The FeF 2 /CMK-3 electrode has an energy density almost three times that of the commercial LiCoO 2 attributable to the synergistic effects rendered by the porous structure.
Since biomass-based carbon materials are widely available, economical, sustainable, and environmentally friendly and can enhance the electrochemical performance of the electrode materials, it is an effective strategy to retaining the heteroatoms components in biomass-based carbon materials and compounding them with fluoride cathode materials.Zhang et al [199] have prepared N, O-doped 3D porous carbon materials as a substrate for FeF 3 • 0.33H 2 O using nori as the feedstock (figure 15(a)).Ding et al [171] have prepared N-S co-doped porous carbon from antibacterial residues by acid leaching filtration and alkaline activation and incorporated the FeF 3 • 0.33H 2 O nanoparticles into porous carbon by wet impregnation (figure 15(b)).The conserved heteroatom components change the interfacial properties and the porous structure provides fast channels for electron migration to stabilize the structure.
Graphene is used in a myriad of applications due to the large surface-to-mass ratio and electrical conductivity stemming from sp 2 -hybridized carbon atoms.Qiu et al [197] [201] have constructed a micro-nanostructured composite with a large bulk energy density consisting of porous graphene, carbon nanotubes, and FeF 3 • 0.33H 2 O nanoparticles by a solvothermal method (figure 16(c)).The FeF 3 • 0.33H 2 O nanoparticles condense on the surface and between the carbon nanotubes and graphene as shown in the TEM image of figure 16(d).The 1D carbon nanotubes prevent stacking caused by π-π interactions between graphene layers and increase the longitudinal ion/electron conductivity.After rGO + CNTs is incorporated with FeF 3 • 0.33H 2 O nanoparticles, the overall energy density increases.Chen et al [202] have prepared 3D porous graphene/FeF 3 • 0.33H 2 O nanocomposites by a hydrothermal process and in situ fluorination.The conductive network composed of 3D porous graphene improves the conductivity and alleviates particle aggregation, whereas the multiple channels entail rapid migration of Li + .Hence, solid-phase diffusion is enhanced and the concentration polarization decreases (figure 16(e)).Compared to 2D graphene, 3D porous graphene more effectively avoids blockage to electrolyte channels and slow ion migration due to the strong π-π conjugation effect.
In addition to carbonaceous materials, metal oxides and fluorides can be utilized to improve the electrochemical properties [91,[203][204][205][206][207][208].Metal compounds can activate the conversion reactions, provide channels for electron transfer, reduce the charge transfer resistance, enhance the reaction kinetics, and mitigate voltage hysteresis.Metal compounds can be used to encapsulate the active powder to inhibit irregular aggregation and pulverization during the electrochemical reaction, whereas transition elements can trigger the redox reaction and contribute to the capacity in some metal oxides.Some fluorides can also be used in the conversion reactions to provide stable phase transition and enhance the electrochemical capacity.The electrochemical properties of composite electrode materials consisting of different transition metal fluorides and different materials are shown in table 3.

Morphological regulation
Incorporation of active materials into ordered aggregates or single crystals by physical or chemical methods can shorten the diffusion distance of lithium ions and increase the effective contact area with electrolyte [209].Examples include nanofibers [210,211], nanosheets [200,212], nanoflowers [42,52,213,214], nanorods [105,[215][216][217], micron-nano-spheres [149,[218][219][220][221][222][223][224], microcubes [225] and core-shell or hollow structures, etc [97,144,[226][227][228][229][230][231][232].Fu et al synthesized FeF 3 /C composite nanofibers by electrospinning with iron acetylacetonate (C 15 H 21 FeO 6 ) as the iron source and polyacrylonitrile (PAN) as the carbon source as shown in figure 17(a).After spinning, the precursor fiber was made into an electrode that is carbonized at different temperature (500 • C-700 • C) to prepare Fe/C nanocomposite fibers.Finally, the FeF 3 /C nanocomposite fibers are prepared by fluorinating the Fe/C nanocomposite fibers in NF 3 .Carbonization and fluorination do not destroy the 1D morphology but with increasing carbonization temperature, the FeF 3 nanoparticles in the nanocomposite fibers coarsen gradually and the FeF 3 nanoparticles are confined to the nanocomposite fibers consequently suppressing the irreversible phase transition of FeF 3 as well as segregation of Fe and LiF.They thus play a role in suppressing the rapid increase of electrode resistance in the battery.The assembly does not require binders and additives and the FeF 3 /C nanofibers can be used as independent electrodes in the battery assembly.The conductive nature of the interconnected nanocomposite fibers not only improves the electrical conductivity, but also provides excellent mechanical stability for long-term cycling (figure 17(b)).
Zhang et al [230] have prepared self-supporting, binderfree vertical nanosheet arrays of FeF 3 • 0.33H 2 O modified by graphene quantum dots (GQDs) by solvothermal and electrophoretic methods.The titanium foil is etched with H 2 O 2 to form a honeycombed surface and then attached to the inner wall of a high-pressure autoclave.The FeF 3 • 0.33H 2 O nanoarrays are prepared from iron, fluorine, and anhydrous ethanol (figure 17(c)).Finally, the GQD modified FeF 3 • 0.33H 2 O vertical nanosheet arrays are prepared by electrophoretic deposition using GQDs as the electrolyte as shown in figure 17(d) SEM image.Figure 17(e) AFM images shows that the average size of the quantum dots is 5.65 nm and the thickness is 1.5 nm.The attached GQDs inhibit the growth of film on the electrode surface and the array morphology is preserved, resulting in enhanced electrode stability and lithium ion transport.
To overcome the limitations of metal fluoride, Wu et al [144] have prepared FeF 3 @C with a 3D honeycomb morphology by evaporation, carbonization, and low-temperature fluorination using water-soluble polymer poly (vinylpyrrolidone) (PVP) as the carbon source mixed with metal nitrate as shown in figure 18(a).SEM images figures 18(b) and (c) shows that these materials exhibit a honeycomb shape after carbonization.The shape formed by bubbles generated by highly viscous PVP during carbonization is preserved after low-temperature fluorination (figures 18(d) and (e)).Similarly, honeycomb FeF 3 @C cathode materials were prepared by Wang et al [233].The preparation process is shown in figure 18(f).The carbon walls in the honeycomb structure can accelerate electron transfer, while the ordered hexagonal pores can facilitate lithium ion transport.The active particles loaded in the honeycomb structure also avoid the negative agglomeration effect to some extent.The honeycomb form of the cathode material showed good discharge capacity and cycle stability in the electrochemical cycle test after assembling FeF 3 @C as cathode material into a full battery (figure 18(g)).
In contrast to lithiation of lithium-ion transition metal oxide cathode materials, the fluorination of metal fluoride materials usually takes place in an oxygen-free atmosphere and so the metal-organic frameworks (MOFs) can be introduced to prepare fluoride cathode materials [127,190,[234][235][236].By controlling the topology and diversity, adjusting the porosity, enlarging the specific surface area, and introducing multifunctional active sites, fluoride cathode materials can be improved.Cheng et al [127] have designed zeolite imidazolate framework-derived CoF 2 /Fe 2 O 3 heterostructures.The in situ

Conclusion
In this review, the basic principles pertaining to metal fluoride conversion cathode materials are described by focusing on the reaction mechanisms, inherent defects, and improvement schemes adopted in recent years.Compared to traditional intercalated LIB cathode materials, metal fluoride conversion cathode materials have advantages in terms of the preparation cost and charge storage density but at the same time, are hampered by some inherent constraints.To offset the nearly insulating nature, significant voltage hysteresis, and slow reaction kinetics, methods such as elemental substitution, optimized synthetic techniques, morphological control, and composite formation have been proposed, and these modification strategies have improved the electrochemical performance of metal fluorides.This review not only summarizes the recent research progress of metal fluorides as cathode materials for LIBs, but also provides certain inspirations for further improvement and expansion of research applications.

Future perspectives
Although the great achievements on metal fluoride cathode materials have been made, the synthesis route of metal fluoride cathode materials is slightly more complicated and expensive than that of conventional intercalation-type cathode materials, and there is still less research on methods to synthesize metal fluoride cathode materials that can be sustainably expanded for large-scale applications.At the same time, the difference in chemical activity between fluorine and lithium sources means that fluorination reactions are potentially dangerous, and it is currently difficult to achieve a good balance between high-cost and stable organic fluorine sources and low-cost and highly corrosive inorganic fluorine sources for large-scale fluorination, in terms of safety and cost.Therefore, the sustainable and expanded synthetic routes as well as eco-friendly improvement method have to be further investigated.
At present, there is still much uncertainty about the conversion mechanism for conversion cathode materials during the electrochemical reaction, and different studies have proposed different reaction mechanisms.For the study of the reaction mechanism of conversion cathode materials, a number of recent studies have begun to gradually adopt more advanced characterization techniques and material evaluation methods, which will facilitate researchers to gain a deeper and more accurate understanding of the reaction mechanism of this type of materials.Currently, the number of similar studies is gradually increasing, which will also bring more new insights and challenges.
In addition to the conversion cathode materials that have been reported so far, nanomaterials with different dimensional microstructures consisting of transition metal fluoroammonium fluorides and transition metal hydroxyfluorides with fluorinated chalcogenide structures of (NH 4 ) x M n+ F x+n have shown intriguing electrochemical behaviors in sodium-ion, lithium-ion, and potassium-ion batteries.Therefore, exploring novel conversion cathode materials of fluoride or other compounds can provide us with more potential candidates with desirable properties.
Currently, LiPF 6 -EC (ethylene carbonate)/DMC (dimethyl carbonate) electrolytes are extremely widely used in LIBs, but metal fluorides in electrolyte environments consisting of EC and DMC tend to perform poorly in terms of electrochemical cycling stability due to the dissolution of cations.This means that the composition of the electrolyte must be considered as an important factor prior to electrochemical testing of LIBs using metal fluoride as the cathode material, although some electrolyte combinations are widely used and stable.
According to part of current research results, some conversion-type cathode materials show stable long-cycle performance without using traditional electrolyte solvents.When the fluoroethylene carbonate replaces ethylene carbonate in the electrolyte, metal fluorides exhibits more stable cycling performance, which has already been explored to a certain extent.In fact, this also explains part of the research in the traditional electrolyte combinations when attempting to modify the metal fluoride material tends to obtain a significant increase in the actual discharge capacity rather than an increase in cycling stability, even if it finally shows some improvement in cycling stability, this may only be a mathematical result obtained after calculations in the context of the actual high discharge capacity, therefore, the effect of different electrolyte systems on the overall performance of the metal fluoride is extremely obvious.Hence, the improvement strategy of electrolyte interfacial layer of electrode materials is equally important, such as the construction of stable electrochemical interfaces and the detailed study of their mechanisms.
As for the anode materials, the current commercial graphite cannot exchange ions with metal fluorides, so developing the corresponding anode materials is quite important.In fact, the suitable anode materials could improve the overall energy density of energy storage devices, especially some lithiumcontaining cathode materials.Future research is expected to overcome the constraints of conversion cathode materials and broaden their scope and applications.It is expected that conversion cathode materials will play an important role in the next-generation energy storage devices.

Figure 1 .
Figure 1.(a) Theoretical operating voltages and capacities of some intercalation and conversion LIB cathode materials.(b) Theoretical gravimetric energy densities.(c) Theoretical volumetric energy densities.(d) Defects of metal fluorides as electrode materials and representative studies.

Figure 3 .
Figure 3. (a) Crystal structure of FeF 2 .(b) Schematic diagram of the conversion of FeF 2 into a bicontinuous network of Fe nanoparticles and LiF during the first lithiation.(c)-(f) Morphology and spatial distribution of the phases in the initial FeF 2 -C nanocomposite electrode: (c) BF TEM and (d) Elemental maps of C (blue) and FeF 2 (yellow).(e) BF TEM and (f) Elemental maps of Fe (green) and LiF (red).Reprinted with permission from[54].Copyright (2011) American Chemical Society.(g) Schematic illustration of the diffusion of the reaction front in a single FeF 2 particle, using the 'layer-by-layer' reaction as the mechanism.Reproduced from[82], with permission from Springer Nature.

Figure 4 .
Figure 4. (a) Crystal structure of FeF 3 .(b) Simplified Li-Fe-F ternary phase diagram and reaction paths of FeF 3 -FeF 2 system in different states.Reproduced from [41], with permission from Springer Nature.

Figure 6 .
Figure 6.(a) Rutile FeF 2 (P4 2 /mnm) in projection along [001].(b) Fe octahedral structure in FeOF, Fe deviates from the central position tending to move toward the O atoms.(c) Nonprimitive cell of the lowest energy FeOF structure.Lowest-energy superstructures of (d) FeOF.(e) Li 0.25 FeOF.(f) Li 0.5 FeOF.and (g) Li 0.75 FeOF in the [001] projection.Reprinted figure with permission from [108], Copyright (2013) by the American Physical Society.(h) Structure of highly distorted octahedral coordination of CuF 2 .Reprinted from [109], Copyright © 2010 Elsevier Inc. Published by Elsevier Inc.All rights reserved.(i) Crystal structure of CuF 2 in the monoclinic unit cell.Reprinted from [110], Copyright © 2012 Elsevier Ltd.All rights reserved.(j) Electrochemical curves of CuF 2 and EDS elemental maps of Li metal before and after charging CuF 2 to 4.5 V. Reprinted with permission from [111].Copyright (2019) American Chemical Society.

Figure 7 .
Figure 7. (a) Preparation process of CuF 2 -SA electrode materials.(b) Mechanism of selective permeation and inhibition of copper dissolution in the Cu-SA layer.(c) Cross-linking effect and coordination structure between Cu 2+ and SA.[114] John Wiley & Sons.© 2022 Wiley-VCH GmbH.
replace M in MF to weaken bonding between M-F and change the bandgap and crystal structure, especially broadening the cavity structure in fluoride materials.The change in the cavity structure improves transport of lithium ions.Ding et al [165] have prepared Mn-doped Fe 1−x Mn x F 3 • 0.33H 2 O/C (x = 0, 0.06, 0.08, and 0.10) nanocomposites by hydrothermal and ball milling and used firstprinciples calculation to investigate the effects of Mn doping on the crystal structure and electronic structure of FeF 3 •H 2 O.As shown in figures 10(a) and (b), the calculation shows that Mn doping enlarges the hexagonal cavity and reduces the bandgap of FeF 3 •H 2 O for better conductivity.The composite shows a discharge capacity of 140.3 mAh g −1 after 200 cycles at a current of 200 mA g −1 (figure
and 103 mAh g −1 at 21.1 C. Qiu et al [173] have synthesized oxygen-substituted FeF 2.67 O 0.33 • 0.33H 2 O conversion cathode materials.Substitution of oxygen reduces the ionic properties between Fe-F and oxygen participates in the reversible anionic redox reaction after oxygen substitutes some fluorine atoms.In the electrochemical reaction, not only does Fe undergo the redox reaction Fe 3+ ↔ Fe 2+ , but oxygen also participates in the reaction of O − ↔ O 2− .Therefore, compared to the pristine FeF 3 • 0.33H 2 O, the reversible storage capacity of FeF 2.67 O 0.33 • 0.33H 2 O after oxygen substitution increases from 150 mAh g −1 -225 mAh g −1 (figure 12(a)).Computation and experiments have revealed that the stabilized low-coordinated anions are prerequisite to the anionic redox electrochemistry.To improve the reversible storage capacity of TMF 3 , p-d hybridization and O-H bonds are implemented to stabilize low-coordination O-ions, as shown in figures 12(b)-(d).

Figure 14 .
Figure 14.(a) Schematic illustration of the preparation of FeF 3 • 0.33H 2 O@CNS nanocomposites.(b) SEM images of the FeF 3 • 0.33H 2 O@CNS nanocomposites.Reproduced from[200] with permission from the Royal Society of Chemistry.

Figure 19 .
Figure 19.(a) Schematic of the synthesis of the MOF-shape CoF 2 @C nanocomposite.(b) SEM of the Co-MOF-67; (c) SEM of the Co@C composites.(d) and (e) TEM of the Co@C composite.(f) HR-TEM of CoF 2 @C.Reprinted with permission from [234].Copyright (2021) American Chemical Society.

Table 1 .
Theoretical capacity, operating voltage, band gap, volume expansion, voltage hysteresis, and stability in electrolytes of common metal fluorides.

Table 2 .
Summary of electrochemical properties of metal fluorides synthesized using different fluorine sources.

Table 3 .
A mini-summary of electrochemical properties of different transition metal fluoride composites.