Mechanism studies of LiFePO₄ cathode material: lithiation/delithiation process, electrochemical modification and synthetic reaction

Abstract

Olivine-structured lithium ion phosphate (LiFePO₄) is one of the most competitive candidates for fabricating energy-driven cathode material for sustainable lithium ion battery (LIB) systems. However, the high electrochemical performance is significantly limited by the slow diffusivity of Li-ion in LiFePO₄ (ca. 10⁻¹⁴ cm² s⁻¹) together with the low electronic conductivity (ca. 10⁻⁹ S cm⁻¹), which is the big challenge currently faced by us. To resolve the challenge, many efforts have been directed to the dynamics of the lithiation/delithiation process in Li_xFePO₄ (0 ≤ x ≤ 1), mechanism of electrochemical modification, and synthetic reaction process, which are crucial for the development of high electrochemical performance for LiFePO₄ material. In this review, in order to reflect the recent progress ranging from the very fundamental to practical applications, we specifically focus on the mechanism studies of LiFePO₄ including the lithiation/delithiation process, electrochemical modification and synthetic reaction. Firstly, we highlight the Li-ion diffusion pathway in Li_xFePO₄ and phase translation of Li_xFePO₄. Then, we summarize the modification mechanism of LiFePO₄ with high-rated capability, excellent low-temperature performance and high energy density. Finally, we discuss the synthetic reaction mechanism of high-temperature carbothermal reaction route and low-temperature hydrothermal/solvothermal reaction route.

---

Feng Yu

Feng Yu received his B.Sc. degree in applied chemistry from the University of Jinan in 2003, and obtained his Ph.D. degree in physical chemistry from the Technical Institute of Physics and Chemistry, Chinese Academy of Sciences (TIPC, CAS) in 2010, working on olive-typed LiMPO4 (M = Fe, Mn, Co and Ni) electrode materials for lithium ion batteries (LIBs) as electrochemical energy storage (EES) devices. He then joined Nanyang Technological University (NTU) as a research fellow and the Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research (ICES, A*STAR), as a research scientist working on the synthesis and electrochemical performance of functional electrode materials. In 2013, he joined Shihezi University as an associate professor to undertake research on advanced electrode materials for high power EES devices.

Lili Zhang

Lili Zhang received her B. Eng. in Chemical and Biomolecular Engineering from the National University of Singapore (NUS) in 2004. After two years of industrial experience in Micron, she continued her Ph.D. study in the same department at NUS and received her Ph.D. degree in 2011. She worked as a research engineer from 2010 to 2011 at NUS and a research fellow in Professor Ruoff's group at the University of Texas at Austin from 2011 to 2012. Now, she is a research scientist in the Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research (ICES, A*STAR), Singapore. Her research interest is developing high performance energy storage materials and systems.

Yingchun Li

Yingchun Li received her B.Sc. degree from Shihezi University in 2003, and M.Sc. degree from Xi'an Jiaotong University in 2006. She continued her Ph.D. study in Martin Luther University of Halle-Wittenberg and received her Ph.D. degree in 2011. Then, she joined Shihezi University as a scientist of the “Recruitment Program of Global Experts” (1000 Talent Plan). Now, Prof. Li continued her research in Shihezi University, and her research field involves the development of advanced functional materials, especially sp^2-type carbon materials including carbon nanotubes and graphene. Additionally, she has developed advanced sensors and molecularly imprinted polymer for serving pharmaceutical analysis, screening of pharmaceutical activity and modernization of traditional Chinese medicine.

Yongxin An

Yongxin An obtained his B.Sc. and M. Sc. Degrees from Heilongjiang University in 2002 and 2007, respectively. Then, he received his Ph.D. degree from Harbin Institute of Technology in 2010. He then joined the Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research (ICES, A*STAR), as a research scientist working on the electrochemical performance of electrolytes for lithium ion batteries (LIBs). In 2013, he joined China Energine International (Holdings) Limited as a director of Graphene & Energy Storage Technology Research Center.

Mingyuan Zhu

Mingyuan Zhu received his B.Sc. degree from Shandong University in 2003, and obtained his Ph.D. degree from the Dalian Institute of Chemical Physics, Chinese Academy of Sciences (DICP, CAS) in 2009. He then joined Shihezi University. He is currently an associate professor in Shihezi University. His research focuses on advanced functional materials for powerful energy storage including fuel cells and supercapacitors.

Bin Dai

Bin Dai obtained his Ph.D. degree in applied chemistry from the Dalian University of Technology (DUT) in 2002. He is the director of Key Laboratory for Green Processing of Chemical Engineering of Xinjiang Bingtuan and a professor at Shihezi University. Prof. Dai has over 20 years of experience in advanced functional materials, working for government, industry and university. His research interests have included porous micro-nano-structured materials for powerful energy storage, plasma catalysis, catalyst for acetylene hydrochlorination, and heterocyclic carbene catalysis.

---

1 Introduction

As excellent electrochemical energy storage (EES) devices, lithium ion batteries (LIBs) have recently attracted significant attention, since the reversible lithiation/delithiation reaction of LiCoO₂ was discovered in 1980 and the use of LiCoO₂ as cathode materials for LIBs in 1990.^1,2 As compared to conventional lead–acid, nickel–cadmium and nickel–metal hydride batteries, rechargeable LIBs possess high working voltage and superior energy density.^3,4 LIBs are not only widely used in consumer electronics, such as cell phones, cameras, toys and laptops, but also used to power increasingly emerging large-scale applications such as electric vehicles (EVs) and hybrid electric vehicles (HEVs).^5,6 Nowadays, LIBs would facilitate the regulation of imbalance in electrical power grids and are the closest to coupling the ultimate requirement of advanced energy storage technologies for renewable energy sources including solar power, wind, and ocean waves.^7,8 Noticeably, LIBs are efficiently meeting the exigent demands of modern energy technology (ET), which is urgently needed to the thrust area closely linked to combustion engines and environmental pollution.^9,10 When combined with renewable energy sources, LIB-based ET is an imperative step to replace the inevitably vanishing non-renewable fossil fuels and avoid negative effects from the current combustion-based ETs on global energy and environmental problems.^11,12 LIBs have become the most viable and promising candidates for EES devices, which strongly minimize the environment impact and maximize energy and resources utilization.^13,14

Although LIBs are well positioned to satisfy the needs of modern society and emerging ecological concerns, one of the greatest challenges is unquestionably the cathode materials where the Li-ions extracting/inserting process occurs.^15,16 Scientists have been focused on the crystal structures and the electrochemical performance of potential cathode materials, such as olivine-structured LiMPO₄ (M = Fe, Co, Ni, Mn), α-NaFeO₂ layered LiMO₂ (M = Co, Ni, Mn), monoclinic structured Li_1+xV₃O₈, orthogonal structured Li₂MSiO₄ (M = Fe, Mn), spinel structured LiMn₂O₄ and NASCION structured Li₃V₂(PO₄)₃.^17,18 Among the aforementioned cathode materials, layer-structured LiCoO₂ with two-dimensional (2D) Li-ion transport has been extensively utilized in LIBs. However, it suffers from high toxicity, inferior safety and high cost.¹⁹ Recently, cubic-spinelled LiMn₂O₄ supporting three-dimensional (3D) Li-ion transport has also been widely used in high powerful EES devices. However, its poor cycle life due to the Jahn–Teller effect remains a big concern.^20,21

Because of the groundbreaking work conducted by Goodenough and co-workers,^22,23 phosphate polyanionic compound LiFePO₄ has attracted tremendous attention and has been used as cathode material in LIBs because of its fantastic performance, such as high theoretical capacity (170 mA h g⁻¹), acceptable operating voltage (3.45 V vs. Li⁺/Li), long cycle life (>2000 cycles), superior safety, low cost, low toxicity, abundant resources, and environmental benignity.^24,25 In the orthorhombic olivine-structured LiFePO₄, the oxygen atoms are located in a hexagonal close-packed and slightly distorted arrangement.^26,27 The phosphorus atoms are on tetrahedral sites and forming PO₄ tetrahedra with oxygen atoms. Lithium atoms from LiO₆ octahedra occupy edge-sharing octahedral positions, while iron atoms from FeO₆ octahedra occupy corner-sharing octahedral positions. Due to the edge-sharing chains along the [010] direction (i.e., b-axis) created by the LiO₆ octahedra, one-dimensional (1D) Li-ion transport is formed in the [010] direction. At the common corners in the bc plane, one FeO₆ octahedron is chained with four FeO₆ octahedra resulting in zigzag planes parallel to the [001] direction (i.e., c-axis). Each FeO₆ octahedron shares one edge with PO₄ tetrahedra and two LiO₆ octahedra, respectively, while PO₄ tetrahedra has two common edges with LiO₆ octahedra.²⁸ This special olivine-structured LiFePO₄ results in spectacular niches including excellent structural flexibility and superior thermal stability, and it provides excellent cycling capabilities and safe characteristics superior to other cathode materials.^29–31

However, there are some limitations of high rate capability due to simplex 1D Li-ion transport of the olive-structured LiFePO₄, which is different from the layer-structured LiCoO₂ providing 2D Li-ion transport and the cubic-spinelled LiMn₂O₄ supporting 3D Li-ion transport.^18,32 Its high-rate performance is significantly limited by the slow diffusivity of the Li-ion in LiFePO₄ (ca. 10⁻¹⁴ cm² s⁻¹) together with the low electronic conductivity (ca. 10⁻⁹ S cm⁻¹).^23,33 The poor power density of LiFePO₄ cathode for LIBs limits its applications in power-demanding EVs and HEVs.^15,34 The development of cathode materials for LIBs with high-rate capability, low-temperature performance, good cycle life and high energy density are still challenging. Efforts have been devoted to the understanding of the dynamics of the lithiation/delithiation process Li_xFePO₄ (0 ≤ x ≤ 1), the electrochemical reaction mechanism, and optimization of synthetic approach, which are crucial for the development of high performance LiFePO₄ material.

In this review, we highlight the mechanism studies of LiFePO₄ including the lithiation/delithiation process, electrochemical property modification and synthetic approach, which are necessary to fully employ its great potential for practical applications. Firstly, we introduce Li-ion diffusion pathway in Li_xFePO₄ and phase translation of Li_xFePO₄ to provide fundamentals for the development of high electrochemical performance LiFePO₄ material. Secondly, we focus on the crucial parameters representing the performances of LiFePO₄, which mainly include specific capacity, rate capability, tap density and low temperature performance. Comprehensive improvements of all those parameters are desired and are also the targets for future research. Finally, we discuss the synthetic reaction mechanism of carbothermal reaction (CTR) route and hydrothermal/solvothermal reaction route, which are two of the most important synthesis routes that need to be improved urgently. We believe this review not only benefits the research of LiFePO₄ but also provides an additional strategy for other materials potentially used in EES devices.

2 Lithiation/delithiation process in LiFePO₄

2.1 Li-ion diffusion pathways of LiFePO₄

2.1.1 One-dimensional Li-ion diffusion along [010] direction. Owing to the fairly compact olivine structure, LiFePO₄ exhibits intrinsic thermal stability in the fully charged state, which makes a major contribution to LIBs safety. Moreover, LiFePO₄ presents excellent structural flexibility, which provides remarkably good cycling stability when compared with cathode materials. However, there are still some obstacle and limitations of its high rate capability. Thus, it has been critical to explore Li-ion diffusion pathway and explain the phase transition during the lithiation/delithiation process of LiFePO₄.^27,35 In recent years, first principle-based modeling has become an important tool to study the reactions during charge and discharge. In 2004, Ceder and his co-workers³⁶ firstly employed first-principles method to study Li-ion transport direction in LiFePO₄. They demonstrated that the Li-ion diffusion coefficient was successfully calculated to be several orders of magnitude higher in the [010] direction (i.e. b-axis) than in the [001] direction (i.e. c-axis). The result of D_[001]/D_[010] ≈ 10⁻³⁷ clearly shows that the [001] direction hardly makes any contribution to the Li-ion motion. Li-ion diffuses through 1D channels along the [010] direction with low energy barriers to cross between the channels due to the FeO₆ octahedral transition state in the [001] direction being face-sharing with two PO₄ tetrahedra. Islam's group^26,37 further designed the structural modeling of LiFePO₄ and used atomistic simulation method to investigate Li-ion migration energy in LiFePO₄. They found that the energy of Li-ion migration is E_mig([010]) = 0.55 eV lower than E_mig([001]) = 2.89 eV and E_mig([101]) = 3.36 eV, which strongly indicates a preference for Li-ion transport along the [010] direction and confirms the results of Ceder's group. Some more theoretical calculations have also been reported and suggested that Li-ion transport is along the [101] direction due to the lower Li-ion migration energy than the other directions.^38,39

Moreover, Li-ion migration was preferential down the [010] channels following a curved trajectory, which is confirmed by Yamada's group.⁴⁰ As shown in Fig. 1, ellipsoids representing Li-ions were refined with 95% probability by Rietveld analysis and motions of Li atoms evolve from vibration to diffusion as an expected curved one-dimensional continuous chain. Moreover, Tse's group⁴¹ found that Li-ion diffusion is a dominant process through a series of jumps from one site to another, resulting into a zigzag pathway along the crystallographic [010] direction. As mentioned above, the Li-ion diffusion pathway mainly occur in the [010] direction, which creates simplex one-dimensional Li-ion transport pathway and is different from that of 2D layer-structured LiCoO₂ and 3D cubic-spinelled LiMn₂O₄. A 1D Li-ion transport tunnel is considered as the main cause of slow Li-ion diffusion, which significantly restricts the high rate performance of LiFePO₄.⁴²

Fig. 1 (a) Anisotropic harmonic lithium vibration in LiFePO₄ shown as green thermal ellipsoids and the expected curved one-dimensional diffusion pathway (dashed lines show how the motions of Li atoms evolve from vibrations to diffusion). (b) Three-dimensional Li nuclear density data shown as blue contours. The brown octahedra represent FeO₆ and the purple tetrahedral represent PO₄ units. (Equi-value 0.15 fm Å⁻³ of the negative portion of the coherent nuclear scattering density distribution.) (c) Two-dimensional contour map sliced on the (001) plane at z = 0.5 and (d) the (010) plane at y = 0, whereas Fe, P and O atoms remain near their original positions.⁴⁰ (Copyright © 2008, Rights Managed by Nature Publishing Group.)

In order to experimentally prove whether Li-ion diffusion pathway is one-dimensional along the [010] direction, ionic conductivity (or ionic mobility) is generally measured and used to calculate ionic diffusivity. Various techniques including cyclic voltammetry (CV), galvanostatic intermittent titration technique (GITT), potentiostatic intermittent titration technique (PITT), electrochemical impedance spectroscopy (EIS), etc., can be used to measure ionic conductivity under an applied voltage.⁴³ Vaknin's group⁴⁴ employed AC impedance spectroscopy to study the Li-ion conductivity of three principal directions (i.e. [100], [010] and [001]) in single crystal LiFePO₄ as a function of temperature (325–525 K). Their results indicate that ionic conductivity along the [010] direction is much higher than that in both [100] and [001] directions, which agrees well with the computational analysis. Meanwhile, Yamada's group⁴⁰ provided clear experimental visualization of Li-ion transport Li_xFePO₄ (x = 0.6) by combining high-temperature (620 K) powder neutron diffraction and the maximum entropy method (MEM), as shown in Fig. 1b–d. To a large extent, lithium delocalizes along the continuous curved 1D chain along the [010] direction on account of the probability density of lithium nuclei. The result is consistent with the aforementioned computational predictions. Therefore, the lithium diffusion constant in nanoparticles with shortened transport paths is much faster than that in bulk.⁴⁵

2.1.2 Two-dimensional Li-ion diffusion along both [010] and [001] directions. The second possible direction of Li transport has been postulated due to the chains separated by octahedral interstitial sites along the [001] direction at elevated temperatures, which suggest that 2D Li-ion transport exist along both [010] and [001] directions.^23,46 As shown in Fig. 2a, Maier's group⁴⁷ investigated ionic and electronic conductivities in single crystalline LiFePO₄ as a function of crystallographic orientation over an extended temperature range, and they found that the activation energies obtained for ionic conductivities σ_Li⁺ along the [010] and [001] orientations are (i.e., E_act([001]) = E_act([010]) = 0.62 eV) smaller than that along the [100] direction (i.e., E_act([100]) = 0.74 eV), which corresponds to effectively two-dimensional Li⁺ conduction. In addition, their study suggested that the activation energies presented for Li-ion diffusion DLi^δ along the [010] and [001] directions are comparable (i.e., E_act([001]) = 0.75 eV, E_act([010]) = 0.70 eV) distinctly less than that of the [100] direction (i.e., E_act([100]) = 0.96 eV) (see in Fig. 2b), which indicated a preferential 2D Li-ion chemical diffusion in the b–c plane. The phenomenon accounted for the effective two-dimensional pattern of Li-ion conductivity and diffusion (i.e., isotropic in the b–c plane). Their study suggested that the ionic conductivity magnitudes in both the [010] and [001] directions and the diffusion coefficient are similar within the temperature range of 140–147 °C, indicating that 2D Li-ion transport must occur within these temperatures.

Fig. 2 (a) Comparison of ionic and electronic conductivities along different crystallographic orientations. (b) Chemical diffusion coefficient along different directions.⁴⁷ (Copyright © 2008 Elsevier B.V. All rights reserved.) (c) For the anti-site defect shown, Fe sits in the channel at site M1 and Li sits in the FePO₄ lattice at site M2. Dashed lines represent the diffusion pathways for Li ions in the lattice. (d) Energy landscape for diffusion around the defect resulting in cross-channel diffusion of Li ions (red) and vacancies (blue).⁴⁹ (Copyright © 2011, American Chemical Society.)

Furthermore, Nazar's group⁴⁸ has also reported that the small polar on the carrier mobility is predicted to be “two-dimensional” with the motions of Li ions as well as electrons being correlated using Mössbauer spectroscopy measurements. They gave experimental evidence for a strong correlation between electron and lithium delocalization events suggesting they are coupled. In 2011, Henkelman's group⁴⁹ reported the various components of Li-ion kinetics in LiFePO₄ calculated from the density functional theory (DFT). As shown in Fig. 2c, there are different kinetic pathways of Li-ion diffusion on the surface, in the bulk, in the presence of defects, and in varying local environments. It is observed that surface diffusion had high barriers resulting into slow kinetics in LiFePO₄. Moreover, the slow bulk diffusion was possibly affected by strain and Li concentration. The slow vacancy diffusion in LiFePO₄ was explained by anti-site defects, which has a barrier of 0.71 eV (Fig. 2d) as compared to 0.29 eV in defect-free channels. Intriguingly, a concerted Li-ion diffusion in FePO₄ exhibited a low barrier of 0.35 eV, allowing for facile cross-channel diffusion at room temperature.

2.2 Phase transition between LiFePO₄ and FePO₄

2.2.1 Two-phase transformation between LiFePO₄ and FePO₄. The lithiation/delithiation process in Li_xFePO₄ (0 ≤ x ≤ 1) is commonly proposed as a two-phase reaction mechanism, which includes various models: shrinking core (i.e., core–shell) model,^23,50 Laffont's (i.e., new core–shell) model,⁵¹ mosaic model,^52,53 domino-cascade model,^54,55 phase transformation wave model,^56,57 and many-particle model.⁵⁸ Generally, the two-phase growth process involves the coexistence of LiFePO₄ and FePO₄, particularly for large particles (e.g., particle size >100 nm).

In 1997, Goodenough's group²³ firstly reported that Li-ion exhibited a two-phase transport between LiFePO₄ and FePO₄ due to the flat charge/discharge profile and introduced the “core–shell” model. With delithiation, LiFePO₄ changes into FePO₄ and yields two phases, forming a two-phase interface. This model is generally named the “core–shell” model. This model can explain the reaction. However, it cannot explain the continuous deviation of the open circuit voltage (OCV) from 3.45 V at the start and end of discharge. Subsequently, Srinivasan's group⁵⁰ proposed a “shrinking core” model to describe the lithiation of FePO₄. This model showed that outside the two-phase coexistence region, there would be a corresponding single phase region, where lithiation proceeds from the surface of a particle moving the two phase interface. With delithiation in the charging process, the FePO₄ shell formed and the FePO₄/LiFePO₄ interface migrated into each particle. Inefficient delithiation from the uncovered LiFePO₄ at the center of the larger particles easily leads to capacity loss. Simultaneously, this shrinking-core model can successfully describe the electrochemical charge/discharge profiles at various rates.⁵⁹

Similar to the shrinking-core model, Thomas's group^52,53 proposed a “mosaic model” by introducing a new concept, i.e., lithiation/delithiation starting at different nucleation sites. These two models are generally considered as “the conventional two-phase mechanism”. However, these two models do not take into account any anisotropy arising from the 1D Li-ion motion within LiFePO₄, which is powerless to describe the lithiation/delithiation process in LiFePO₄. Otherwise, the “mosaic model” is still not supported by direct and convincing experimental evidence.

Laffont's group⁵¹ and Richardson's group⁶⁰ found that the interfacial region between the LiFePO₄ and FePO₄ domains lie in the a–c plane (see in Fig. 3). Then, Laffont's group⁵¹ updated the “core–shell” model and proposed a “new core–shell” model based on the studies of thin Li_xFePO₄ platelets (b-axis normal to the surface) with high resolution electron energy loss spectroscopy (HREELS). Different from the previous shrinking core model, the results suggested that the interface is constituted of FePO₄ and LiFePO₄. There is no Li_xFePO₄ solid solution observed with the gradient of x ranging from 0 to 1. The schematic views of the interfacial region between LiFePO₄ and FePO₄ phases are provided in Fig. 3. According to this new core model, Li-ion diffusion in the [010] direction is asynchronous, and the Li_xFePO₄ particles always maintain the structure with the shell of FePO₄ and core of LiFePO₄, as has also been suggested by Prosini's group.⁶¹ Moreover, this new core–shell model is unambiguously supported by Lemos's group⁶² who observed the existence of both FePO₄ and LiFePO₄ phases at the interface of Li_0.11FePO₄.

Fig. 3 (a) TEM image of a Li_0.5FePO₄ crystal, showing the domains of LiFePO₄ and FePO₄ aligned along the c-axis.⁶⁰ (Copyright 2006, The Electrochemical Society.) (b) TEM and (d) HRTEM images of the delithiated Li_0.45FePO₄ sample. The phase determination present in the edge, core, or interface of the particle, respectively. (c) Sketch of the Li_0.45FePO₄ particle. (e–g) Schematic views of the interfacial region between LiFePO₄ and FePO₄ phases.⁵¹ (Copyright © 2006, American Chemical Society.)

However, Delmas' group^54,55 found that the interface consisted of a single-domain of Li_xFePO₄ rather than LiFePO₄ and FePO₄. Based on the results from X-ray diffraction (XRD) and high resolution transmission electron microscope (HRTEM), a “domino-cascade” model was successfully established to explain the phase transformation process of LiFePO₄ nanoparticles. The results illustrated that the growth of FePO₄ phase at the expense of the LiFePO₄ phase is considerably faster than the nucleation of a new domain. According to the “domino-cascade” model, lithiation/delithiation occurred completely and rapidly in some of the LiFePO₄ nanoparticles with charging/discharging. Thus, the partially intercalated/de-intercalated LiFePO₄ particle generally supported the coexistence of LiFePO₄ and FePO₄ particles that are single-domain.

In 2013, Zhou's group⁶³ studied the phase transition in Li_xFePO₄ by electrochemical impedance spectroscopy (EIS) and firstly proposed a hybrid phase-transition model combining the core–shell model and domino-cascade model. Based on this model, the delithiation of Li_xFePO₄ starts with the domino-cascade model confirmed by a small angle of 30° for the linear Warburg region of EIS, and then turns into a core–shell model approved by a traditional angle of 45°. This hybrid phase-transition model gives attributions to the strong anisotropy in the Li_xFePO₄ particles and could be potentially extended to some other two-phase active electrode materials.

Recently, a “many-particle model” has also been suggested. Gaberscek's group⁵⁸ found that Li-ions were inserted/extracted from LiFePO₄ particles one by one forming a two-phase system. Meanwhile, Zaghib's group⁶⁴ found that both LiFePO₄ and FePO₄ phases exist in the surface layer by using electron microscopy and Raman spectroscopy. This is intended to explain the fact that the lattice coherence length is the same in the process of lithiation/delithiation, while it invalidates both core–shell model and the domino-cascade model. Each particle would be single-phased, either FePO₄ or LiFePO₄, which explains how one particle chooses to be in one phase while the other one remains in the other phase without violating the causality principle. Subsequently, Orikasa's group⁶⁵ observed a transient phase change in the two phase reaction during the lithiation/delithiation process by applying the time-resolved X-ray measurement. It is found that the nonequilibrium phase state during the voltage plateau gradually changes and finally reaches the thermodynamically stable state by the analysis of the X-ray absorption near the edge structure (XANES).

Based on the observation and demonstration by Delmas' group⁵⁴ and Ceder's group,⁴⁵ Chueh's group⁶⁶ also observed the overwhelming majority of Li_0.5FePO₄ particles (i.e., LiFePO₄ electrode in a 50% state-of-charge) were either almost completely delithiated LiFePO₄ particles or lithiated FePO₄ particles in 2013. With the help of scanning transmission X-ray microscopy (STXM), it is clearly found that the delithiation did not appear faster in smaller particles than the larger ones resulting into weak correlation between the particle size and delithiation sequence. Therefore, the mosaic (particle-by-particle) lithiation/delithiation pathway indicated that the rate performance was limited by the rate of phase-transformation initiation rather than the phase-boundary velocity. Moreover, the model can be used to predict the equilibrium between LiFePO₄ and FePO₄ phases, inherent hysteretic behaviour, sequential charging/discharging mechanism and two-phase disappearance behaviour. Porous LiFePO₄ electrodes were also investigated by Bai's group⁶⁷ using a mathematical phase-field method. It is found that the population dynamics of active LiFePO₄ nanoparticles showed non-monotonic transient currents always misinterpreted as the nucleation and growth mechanism by the Kolmogorov–Johnson–Mehl–Avrami (KJMA) theory. The LiFePO₄ nanoparticles were not simultaneously transformed. The results decoupled the roles of nucleation and surface reaction, which are always considered to be affected by a special activation rate and the mean active particle-filling speed. Ichitsubo's group⁶⁸ also applied the phase-field computer simulations to investigate the coherent elastic-strain energy that played a crucial role in the kinetics of phase separation during the lithiation/delithiation process. However, it is found that the nucleation of the new phase is fundamentally unlikely formed in terms of the elastic strain energy, except in the vicinity of the particle surface. The simulation results illustrated that the solid-solution reaction easily occurred by reducing the particle size, but the phase separation by nucleation is quite difficult to carry out. Meanwhile, Dargaville' group⁶⁹ employed the 2D Cahn–Hilliard-reaction (CHR) equation to examine the intercalation process and suggested that the phase separation only occurred at very low discharge rates.

All these aforementioned models are expected to understand the lithiation/delithiation process in LiFePO₄ and provide directions to improve the rate capability of LiFePO₄ for high power EES applications. However, these models invalidate each other and are still under debate. Nevertheless, it is still unclear whether the staging phenomenon refers to a thermodynamically metastable or stable situation in Li_xFePO₄ nanoparticles. Theoretical and experimental clarifications should be further undertaken to clarify the formation mechanism of the lithium staging. Thus, as mentioned above, a single unified model is urgently required to depict the natural lithiation/delithiation process in Li_xFePO₄, and a computational model of the lithiation/delithiation process in LiFePO₄ is essential to clarify the above question. When compared with conventional generalized gradient approximation (GGA)^70,71 and local-density approximation (LDA) methods,^72,73 GGA + U method can avoid large errors especially for transition metals with strong localized or f-orbitals metals and can effectively improve the results.^74–76 In the future, we anticipate that the lithiation/delithiation process in LiFePO₄ will be simulated by a computational model using the GGA + U method. This natural model can not only offer answers to experimental results obtained at moderate or high rates but also provide the direction to prepare LiFePO₄ for high power LIBs.

2.2.2 Quasi-single-phase transformation between LiFePO₄ and FePO₄. Different from two-phase reaction mechanism models, the quasi-single-phase transformation between LiFePO₄ and FePO₄ has also attracted considerable attention. It is doubtful that an intermediate phase exists during the lithiation/delithiation process of Li_xFePO₄. In 2011, Li's group⁷⁷ directly observed Li-ions in LiFePO₄ at an atomic resolution by an aberration-corrected annular-bright-field scanning transmission electron microscopy (ABF-STEM) technique (Fig. 4a), which is capable of resolving Li-ions directly. It was found that the remaining Li-ions in partially delithiated LiFePO₄ preferably occupy every second layer along the [010] direction (i.e., b axis) (Fig. 4b and c). Obviously, this finding challenged the previously proposed LiFePO₄/FePO₄ two-phase reaction mechanisms. In 2011, Ceder's group⁷⁸ demonstrated that Li_xFePO₄ may transform through a single-phase path instead of a two-phase progress. As shown in Fig. 4d, the calculated single-particle voltage hysteresis is no more than 30 mV, where 0.05 < x < 0.9 in Li_xFePO₄. It is clearly shown that the Li_xFePO₄ solid solution formed and avoided phase separation in the charging and discharging process. The solid-solution behavior in the Li_xFePO₄ system is also observed by Richardson's group via X-ray powder diffraction (XRD).⁷⁹ In the Li_0.6FePO₄ sample, there was an intermediate of a line phase with Li_xFePO₄ (x = 0.60 ± 0.04) composition during the transformation from two-phase mixtures to single phase. In 2011, Bazant's group⁸⁰ proposed a novel electrochemical phase-field model. Based on the model, nucleation or spinodal decomposition easily leads to moving phase boundaries at small currents, whereas the particles fill homogeneously and the spinodal disappears above a critical current density, as shown in Fig. 5. This model can effectively explain the long cycle life and superior rate capability of LiFePO₄ nanoparticles.

Fig. 4 (a) Schematic of the annular-bright-field (ABF) imaging geometry. A demonstration of lithium sites within a LiFePO₄ crystal is shown in (b) with the corresponding line profile acquired in the box region shown in (c) to confirm the lithium contrast with respect to oxygen.⁷⁷ (Copyright © 2011, American Chemical Society.) (d) The single-particle voltage within 0.5 < x < 0.9 in Li_xFePO₄.⁷⁸ (Copyright © 2011, Rights Managed by Nature Publishing Group.) (e) Illustration of a solid solution Li_xFePO₄ upon heating the compound to 485 K. On cooling, the parent LiFePO₄ and FePO₄ phases recrystallize, and nucleation and growth take place everywhere within the crystal.⁴⁸ (Copyright © 2006, American Chemical Society.)

Fig. 5 Schematic model of a Li_xFePO₄ nanoparticle at low overpotential. (a) Lithium ions are inserted into the particle (blue arrows) from the active (010) facet with fast diffusion and no phase separation in the depth (y) direction, forming a phase boundary of thickness λ between full and empty channels. (b) Numerical simulation of phase transformation triggered by wetting of the particle boundary.⁸⁰ (Copyright © 2011, American Chemical Society.)

Recently, the single phase of a solid solution Li_xFePO₄ has been found at high temperature above 200 °C.⁸¹ As shown in Fig. 4e, Nazar's group⁴⁸ found that single Li_xFePO₄ solid solution phase formed at an elevated temperature of 212 °C and the lithiation/delithiation process were strongly enhanced. Owing to the single phase, the electrochemical performance is improved by a strong coupled correlation between electron and lithium delocalization events, whereas the power characteristics could be diminished in a two-phase mixture on account of the low mobility of the phase boundary. All these results are in good agreement with the report of Masquelier's group,^82,83 who found that the single phase of Li_xFePO₄ becomes multiphase as the temperature decreases and eventually transforms into the phases of LiFePO₄ and FePO₄ on aging at room temperature.

Excitedly, Masquelier's group^84,85 provided in situ experimental evidence that Li_xFePO₄ can be described as a single phase from a well-established two-phase lithiation process by modifying the particle size and cation ordering. It is can be found that the single intermediate phases of Li_0.5FePO₄ and Li_0.75FePO₄ might be stabilized at room temperature. Simultaneously, the Li_xFePO₄ solid solution may be stabilized below the critical size of 45 nm, and completely achievable below about 15 nm at room temperature.⁸⁶ It is worth mentioning that the lithiation/delithiation mechanism is still controversial and has attracted more and more attention. A series of conditions including particle morphologies, synthesis conditions, charge/discharge rates, especially state particles sizes, are still being debated.^87,88

3 Electrochemical property modification of LiFePO₄

LiFePO₄ delivers high capacities of more than 150 mA h g⁻¹ at slow charging/discharging rates (e.g., 0.1 C). However, high-rated charge/discharge properties and low temperature performance are severely limited due to the poor electronic conductivity and the low Li-ion diffusion efficiency. Namely, electronic conductivity and ionic diffusion rate are two of the most important characteristics that need to be improved urgently. It has been demonstrated to enhance the Li-ion diffusion by controlling the crystal growth orientation (along the a–c plane) and reducing the particle size of LiFePO₄ (e.g., nano-sized^89,90 and 3D porous architectures^91,92). It is also important to ensure an excellent electronic conductivity by coating conductive materials (e.g., amorphous carbon,^93,94 graphene,⁹⁵ carbon nanotubes (CNTs),⁹⁶ silver,⁹⁷ NiP alloy,⁹⁸ metallic Fe₂P,^99,100 Li₃PO₄,¹⁰¹ etc.). Both ionic diffusion and electronic conductivity can be employed to improve the kinetic properties of LiFePO₄ by doping in M_Li and M_Fe sites using the cations of metallic elements (e.g., Cu⁺, Mg²⁺, Mn²⁺, Al³⁺, Ti⁴⁺, Zr⁴⁺, Nb⁵⁺, W⁶⁺, etc.).^62,102,103 Comprehensive improvements of all the effective strategies are desired and are also the targets for high-rated charge/discharge properties and low temperature performance, which are two important parameters for future research and EES applications (especially for EV/HEV). Moreover, the volume energy density of LiFePO₄ is another important parameter, which is critical towards maximizing the space utilization of LIBs. In the following section, we will introduce representative reports on how to improve the performances of LiFePO₄. The reports are categorized into the improvement of (a) rate capability, (b) low temperature performance and (c) tap density.

3.1 Morphology control

3.1.1 Crystal growth orientation along the a–c plane. The Li-ion diffusion rates vary along different directions in LiFePO₄ lattices. If we can deliberately control the synthesis of LiFePO₄ crystals with certain exposed facets for faster Li-ion intercalation, the performances, especially rate capabilities, can be significantly improved. Previous reports showed that Li-ion diffused the fastest along the [010] direction in the orthorhombic LiFePO₄ lattice.⁴⁰ Islam's group¹⁰⁴ revealed that there were mainly three crystal styles agreeing with the results from Franger's group,¹⁰⁵ Nazar's group¹⁰⁶ and Richardson's group,⁶⁰ as shown in Fig. 6a–d. The crystal growth of anisotropic cathode LiFePO₄ is addressed as a key factor controlling rapid Li-ion diffusion. Several surface properties of olivine-structure LiFePO₄ were investigated via first principles calculations within the GGA + U framework by Ceder's group.¹⁰⁷ The calculated Li redox potential for the (010) surface was 2.95 V, which is significantly lower than the bulk value of 3.55 V. The study revealed that it is important to control the morphology of LiFePO₄ crystal growth orientation and expose more surface of (010) with high Li⁺ diffusion rate.¹⁰⁸ Thus, if LiFePO₄ crystals are fabricated with more (010) facets, such as thin nanoplates with (010) surfaces, superior performances can be expected due to the fast Li⁺ insertion and extraction. Richardson's group⁶⁰ reported the hydrothermal synthesis of LiFePO₄ microcrystals with 85% (010) surface areas, but no electrochemical testing was performed. For further improvements, like the synthesis of smaller and thinner nanoplates with more (010) surfaces, coating with a conductive carbon layer may be considered to fully explore the potential of those nanoplates.¹⁰⁹

Fig. 6 (a) Theoretical crystal morphologies of LiFePO₄ with equilibrium morphology from relaxed surface energies. Schematic diagrams of the observed crystal morphologies of LiFePO₄ produced under different experimental conditions: (b) block-shape, (c) rectangular prism and (d) hexagonal platelet.¹⁰⁴ (Copyright © 2008, Royal Society of Chemistry.) SEM images, HRTEM images (inset) and the corresponding SAED patterns (inset) for (e) rectangular prism nanoplates¹¹⁰ (Copyright © 2011, The Electrochemical Society) and (f) hexagonal platelet.¹¹² (Copyright © 2010, Royal Society of Chemistry.) (g) SEM (the inset is an enlarged SEM image) and (h) TEM morphologies of nanoplates.¹¹³ (Copyright © 2008, Royal Society of Chemistry.)

In 2011, a combination of LiFePO₄ rectangular nanoplates was prepared by a low temperature solvothermal synthesis (STS) route reported by Kim's group,¹¹⁰ as shown in Fig. 6e. It can be seen that the growth of the LiFePO₄ nanoplates is identified to be along the [100] and [010] crystallographic directions, resulting into a large surface area of (001) rather than (010). Thus, the as-prepared LiFePO₄ nanoplates yielded the first discharge capacity of only 131 mA h g⁻¹ at 0.25 C (Fig. 7), on account of the relatively low (101) surface area. In particular, the as-prepared LiFePO₄ nanoplates exhibited a capacity decline of about 38 mA h g⁻¹ until 8 C. Wang's group and Wu's group¹¹¹ found that the obtained LiFePO₄/C composite exhibited high rate capability using carbon coating and yielded 104 mA h g⁻¹ and 95 mA h g⁻¹ at 8 C and 12 C, respectively.

Fig. 7 High rate capacity performance of LiFePO₄ particles with different observed crystal morphologies.

Fig. 6f shows LiFePO₄ hexagonal nanoplates (100 nm thick and 800 nm wide) prepared by a solvothermal method in a H₂O-PEG binary solvent.¹¹² When compared with LiFePO₄ hexagonal microplates (300 nm thick and 3 mm wide), carbon coated LiFePO₄ nanoplates exhibited the calculated Li-ion diffusion coefficient of 4.2 × 10⁻⁹ cm² S⁻¹ instead of 2.2 × 10⁻⁹ cm² S⁻¹, and delivered a discharge capacity of more than 155 mA h g⁻¹ instead of 110 mA h g⁻¹ at 0.1 C. In particular, LiFePO₄/C nanoplates could perform a high discharge capacity of 87 mA h g⁻¹ at 60 C (i.e., fully discharged within 30 s), when the content of conductive carbon was increased to 30 wt%.

Recently, Balaya and co-workers^113–115 employed a solvothermal method to control the crystal growth orientation along the a–c plane and reduce the b-axis to the smallest possible thickness. As shown in Fig. 6g and h, the thickness along the b-axis of LiFePO₄ nanoplates is found to be 30–40 nm, while the thickness along the a–c plane is in the range of 500–800 nm. The selected area electron diffraction (SAED) pattern viewed along [010] (i.e., b-axis) reveals that the plate is a single crystal with the b-axis along the thinnest direction. It is indicated that the smallest dimension of the nanoplates is the b-axis, which facilitates the diffusion of Li-ions.¹¹² The as-obtained LiFePO₄/C nanoplates could store Li-ions comparable to its theoretical capacity of 87 mA h g⁻¹ at 0.1 C. Even at 30 C, they could deliver a discharge capacity of up to 87 mA h g⁻¹. All the results revealed that the excellent high rate performance of LiFePO₄ nanoplates is ascertained to the relatively thin thickness along the b-axis and large (101) surface area. In other words, size reduction (similar to 30 nm) at the b-axis and crystal growth orientation along the a–c plane could provide fast Li-ion diffusion. Moreover, uniform carbon coating (similar to 5 nm) throughout the LiFePO₄ nanoplates favors an electronically conducting path for electrons.

3.1.2 Nanosized LiFePO₄ particles. To date, the most widely adopted strategy to improve the rate capability is to use LiFePO₄ structures with smaller sizes and coat the structures with carbon to improve the electronic conductivity.¹⁰⁸ In particular, nano-structured LiFePO₄ particles are in favor of reducing the 1D diffusion length and advantageous for Li-ions transport, as the intrinsic diffusion constant is scale dependent and significantly reduced for large particle sizes.⁴⁵ Various carbon-modified LiFePO₄ nanostructures have been studied, such as nanoparticles, nanorods, nanoflowers, porous microspheres, nanoplates, nanowires/fibers, template mesoporous materials, etc.¹¹⁶ Smaller structures mean shorter diffusion paths for Li-ion, and hence, better utilization of the active materials.¹¹⁷ For example, assuming that for a certain time, Li-ions can diffuse 50 nm within LiFePO₄ lattices, if the electrode is made of a single crystal LiFePO₄ film with a thickness of 1 mm, then only the 50 nm on the surface can be utilized. However, if the electrodes are made of nanoparticles with less than 100 nm diameters and proper electrolyte wrapping and electron conduction is ensured, the entire nanoparticle, and hence the electrode materials, can be utilized.¹¹⁸ This is why nanosized materials are favored to obtain high specific capacity and high rate capability.^119,120 As discussed above, at a high charge/discharge rate, less time is allowed for ion diffusion. Consequently, nanostructures possess a better capability for complete “reactions” with Li-ion within a short time. These nano-sized LiFePO₄ materials could provide short diffusion length, yield better rate performances than bulk materials because of smaller diffusion lengths, and be beneficial for Li-ion batteries, as shown in Fig. 8a.

Fig. 8 (a) Schematic illustration for active LiFePO₄ at different interfaces of electrolyte and various LiFePO₄ particles.¹³⁸ (b) TEM image of LiFePO₄/C nanopaticles.¹²² (Copyright © 2008 WILEY-VCH Verlag GmbH %26 Co. KGaA, Weinheim.) (c) SEM and TEM (inset) images of hollow LiFePO₄.¹²³ (Copyright © 2008, American Chemical Society.) (d) SEM and TEM (inset) images of porous and coarse LiFePO₄/C composite.¹³⁶ (Copyright © 2013, American Chemical Society.) (e) SEM image of an individual LiFePO₄/C microsphere consisting of nanoplates.⁹² (Copyright © 2011, American Chemical Society.) (f) FIB images allowing 3D visualization information obtained from the 3D porous LiFePO₄/C microsphere. (g) SEM image of A area (indicated by a rectangle in panel f). (g) SEM image of B area (indicated by a rectangle in panel f). (i) Scheme showing the structure of LiFePO₄/C nanocomposites in porous microspheres.¹³⁸

The ball milling method is usually employed to synthesize nano-sized LiFePO₄/C composites. It is found that LiFePO₄/C from ball milling in acetone displayed a spherical shape with a size of 60 nm, similar to the size of LiFePO₄/C observed from dry ball milling (DBM) but with a more irregular morphology.¹²¹ Although the LiFePO₄/C nanocomposites prepared from wet ball milling (WBM) in acetone and DBM exhibited the similar discharge capacities of 153 mA h g⁻¹ and 120 mA h g⁻¹ at rates of 0.1 C and 10 C, respectively (Fig. 9a), the previous sample yielded much lower polarization resulting in high energy density. A core–shell structured LiFePO₄/C nanocomposite was also successfully designed and prepared by Zhou's group.¹²² An in situ polymerization restriction method was firstly used for the synthesis of FePO₄/PANI. Then, a highly crystalline LiFePO₄ core was formed with a size of about 20–40 nm, and the polymer shell was transformed into a semi-graphitic carbon shell with a thickness of about 1–2 nm during the heat treatment process at 700 °C, as shown in Fig. 8b. The as-obtained LiFePO₄/C yielded excellent rate performance. Even at the high rate of 60 C, it still delivered a capacity of 90 mA h g⁻¹. Additionally, Cho's group¹²³ reported that LiFePO₄ nanowires were synthesized using the two-dimensional hexagonal SBA-15 silica as the hard template. The LiFePO₄ nanowires delivered excellent rate performance. Even at 10 C and 15 C, LiFePO₄ nanowires yielded 144 mA h g⁻¹ and 137 mA h g⁻¹, respectively, corresponding to 93% and 89% retention of the initial capacity.

Fig. 9 High rate capacity performance of LiFePO₄ particles with (a) nano-sized and (b) 3D porous morphologies.

Taniguchi's group¹²⁴ employed a combination technique of spray pyrolysis (SP) with WBM to prepare LiFePO₄/C nanoparticles. It is found that the LiFePO₄/C nanoparticles possessed a geometric mean diameter of 146 nm. A thin carbon layer endowed LiFePO₄ with excellent electrochemical performance, because LiFePO₄ nanoparticles exhibited an extremely high stability with the carbon coating.¹²⁵ LiFePO₄/C yielded 118 mA h g⁻¹ and 105 mA h g⁻¹ at 10 C and 20 C, respectively. Even at 60 C, LiFePO₄/C yielded 75 mA h g⁻¹ without capacity fading after 100 cycles. Amazingly, Ceder's group¹²⁶ found LiFe_1−2yP_1−yO_4−δ/C (y = 0.05) nanoparticles prepared by WBM in acetone exhibited outstanding high rate performance. The LiFe_0.9P_0.95O_4-δ/C off-stoichiometry nanoparticles showed a specific discharge capacity of 163 mA h g⁻¹ and 152 mA h g⁻¹ at 10 C and 20 C, respectively. Significantly, the specific capacity still remains above 120 mA h g⁻¹ at a discharge rate as high as 197 C, when changing the working electrodes fabrication of the active material, carbon black and binder in a weight ratio from 80 : 15 : 5 to 30 : 65 : 5. This result is a remarkable progress over the previous studies. This presents a new direction on the improvement of LFP cathode performances, although more work is needed to further improve the repeatability and mechanism understanding of this method.

3.1.3 Three-dimensional porous architectures with micro-nano structures. 3D porous LiFePO₄ materials not only enhance the surface-to-volume ratio and reduce the Li-ion transport length but also offer a potential solution to reduce the interfacial energy and agglutinant in electrode, all of which are much better than those measured for the nano-sized LiFePO₄ and hollow micro-spherical LiFePO₄ (Fig. 8c), improve the electrochemical performance and practical applications of LiFePO₄.^123,127 In particular, LiFePO₄ cathodes with 3D porous architectures, spherical architectures, and micro-nano-structures have been extensively and intensively studied for high powerful LIBs. With the aid of template technology, a variety of 3D porous LiFePO₄ materials have been synthesized.^128–130 In 2013, Zhang's group¹³¹ successfully prepared 3D mesoporous LiFePO₄ using Baker's yeast cells both as a structural template and a biocarbon source. The as-obtained LiFePO₄ exhibited a high discharge capacity of about 153 mA h g⁻¹ at 0.1 C.

Besides the soft template, Li₃PO₄/graphene oxide (GO) microspheres are also employed as sacrificial templates to synthesize porous LiFePO₄/graphene microspheres.¹³² The as-obtained LiFePO₄/graphene microspheres (2 μm) were assembled by nanoparticles and wrapped by graphene nanosheets. The graphene has been proven to be advantageous for EES applications on account of its superior electrical conductivities and high surface area. Moreover, the porous microspherical structure exhibited an effective way to achieve high power density for LiFePO₄. The as-obtained LiFePO₄/graphene yielded excellent rate performance, i.e., 101.8 mA h g⁻¹ at 10 C, which holds 72% of the initial capacity at 0.1 C, as shown in Fig. 9b.

Without employing templates, hydrothermal synthesis (HTS) as a low-temperature liquid phase thermal (LPT) synthesis has also been employed to prepare 3D porous LiFePO₄ microspheres. In the HTS process, the LiFePO₄ precursor nanoparticles deposit on account of their small solubility in the solution. Then, the nano-sized LiFePO₄ precursor particles self-assemble to form densely packed microspheres necessary for reducing the surface tension of the dispersed particles. Subsequently, with a dissolution-deposition process, the agglomerated LiFePO₄ precursors evolve into 3D porous microspheres.¹³³ The HTS yields several advantages such as simplicity, morphology control, homogeneous particle size distribution, and low cost.^134,135 In 2010, Cao's group¹³³ first prepared 3D porous LiFePO₄ microspheres, which consisted of nanoparticles, by the hydrothermal process. These LiFePO₄/C microspheres exhibited high Brunauer–Emmett–Teller (BET) surface areas of 38.6 m² g⁻¹ and delivered excellent high rate capability of 115 mA h g⁻¹ at 10 C. Even at 20 C and 30 C, they yielded 93 mA h g⁻¹ and 71 mA h g⁻¹, respectively. In 2011, Goodenough's group⁹² successfully accomplished the self-assembly of LiFePO₄ nanoplates for the porous microspheres using the STS route, as shown in Fig. 8e. It can be seen that the 3D porous LiFePO₄/C microspheres (1–3 μm) consist of nanoplates (80 nm), which interweave together forming a flowerlike structure giving a high BET surface area of 32.9 m² g⁻¹. The as-obtained flowerlike LiFePO₄/C microspheres exhibited a discharge capacity of 72 mA h g⁻¹ at 10 C. Due to the conductive polymer PPy processing macromolecule sp²-type carbon, LiFePO₄/(C + PPy) yielded excellent rate performance, e.g., high discharge capacity of 92 mA h g⁻¹ at 10 C. Additionally, Eom's group and Kwon's group¹³⁶ provided a growth technology to synthesize porous and coarse LiFePO₄/C composites (5–10 μm) using LiFePO₄ nanoparticles (100–200 nm) as seed crystals for the 2nd crystallization process, as shown in Fig. 8d. The SEM and TEM images of LiFePO₄/C composites are shown in Fig. 8d. The as-obtained 3D porous LiFePO₄/C composites delivered a discharge capacity of 100 mA h g⁻¹ at 10 C, which is 65% of the discharge capacity at 0.1 C. All the excellent rate performances strongly fulfil the requirements of rechargeable lithium batteries for high power applications.

It is worth noting that a versatile spray drying methodology is usually employed to prepare 3D porous spheres with micro-nano-superstructures.¹³⁷ Zhang's group¹³⁸ reported a novel and simple template-free concept and the synthesis of 3D porous LiFePO₄/C microspheres via a sol–gel-spray drying (sol–gel-SD) method. As shown in Fig. 8f–i, it is clearly seen that the as-obtained LiFePO₄/C had a large specific surface area of 20.2 m² g⁻¹, an average nano-size of 32 nm and a main pore diameter of 45 nm. The as-obtained 3D porous LiFePO₄/C microspheres easily contact with the electrolyte, which facilitate the Li-ion diffusion, gave a high Coulombic efficiency of 97.2%, and presented an excellent capacity retention rate close to 100% after 50 cycles. However, it only provided 100 mA h g⁻¹ at the high rate of 10 C, due to lower carbon coating and low porosity.¹³⁹ When using citric acid as the carbon source, 3D porous LiFePO₄/C microspheres have been successfully synthesized.^140,141 Polyvinyl alcohol (PVA) gel has been also chosen as the carbon source for its excellent film-forming properties in the spray drying process.¹⁴² The carbon layers transformed from the carbon sources were effectively coated around LiFePO₄ and showed excellent cyclability and superior rate capability.

3.2 Low temperature performance

LiFePO₄ cathode with an excellent power output makes it a superior candidate for HEV and EV applications. However, one key challenge is the poor low temperature performances due to LiFePO₄ itself and the electrolyte in LiFePO₄-based batteries, as compared to the performances at room temperature. In 2013, Shin's group¹⁴³ found that pristine LiFePO₄ exhibited a higher cycling stability at a lower operating temperature of −20 °C than room temperature. It is suggested that LiFePO₄ experiences much milder degradation of the surface and the bulk at low temperatures. To date, LiFePO₄ itself and the electrolyte have been both responsible for the poor low-temperature performance, which is the key challenge for wider applications of LiFePO₄ in EES devices.¹⁴⁴

Generally, carbon coating is used to improve the electron conductivity of the electrodes. The processes are associated with electron conductions.^27,145 Good electronic conductivity is of crucial importance to facilitate electron conduction and achieve high performances.^146,147 Carbon is a common low-cost material with good conductivity and is widely used for conductivity improvement. Carbon layers can be facilely coated on the nanostructure surfaces by many low-cost and scalable methods, such as organic chemical polymer coating and annealing.¹⁴⁸ Carbon coating is quite successful in improving the cathode conductivity, but the weight ratio of carbon has to be carefully controlled.^149,150 To be extreme, a cathode made of pure carbon will simply turn the battery into a double-layer supercapacitor, which has very low energy density and unstable output voltage. Higher carbon content also results in lower volumetric energy density due to the smaller mass density of carbon. Apart from the enhancement of conductivity, carbon also serves as a capping material that can effectively reduce the particle growth during crystallization processes. Simultaneously, the reducing atmosphere of carbon also prevents the oxidation of Fe(II) during high temperature annealing.

In 2011, Tu's group¹⁵¹ successfully prepared carbon-coated LiFePO₄ materials using polystyrene (PS) spheres (50–300 nm) as the carbon source. The results illustrated that LiFePO₄/C with 3.0 wt% carbon content exhibited excellent electrochemical capability at a low temperature of −20 °C, which yielded 147 mA h g⁻¹ and 79.3 mA h g⁻¹ at 0.1 C and 1 C, respectively (Fig. 10). Even after 100 cycles at 1 C, the LiFePO₄/C still exhibited almost 100% capacity retention. It can be attributed to the optimal carbon coating thickness (2.5 nm) and good carbon coating morphology. Using citric acid as a carbon source, Wang's group¹⁵² reported a kind of 3D LiFePO₄/C microspheres. The electronic conductivity of LiFePO₄/C microspheres is between 7.5 × 10⁻² and 10⁻¹ S cm⁻¹ from −40 °C to 23 °C, which yielded attractive low temperature discharge capacity of 110 mA h g⁻¹ and 64 mA h g⁻¹ at −20 °C and −40 °C. Alternatively, Scrosati's group and Sun's group¹⁵³ provided double-carbon-coated LiFePO₄/C porous microspheres, with an excellent specific discharge capacity of 70 mA h g⁻¹ (1 C) at −20 °C. Additionally, polyacene (PAS) was also employed to optimize LiFePO₄.¹⁵⁴ The as-obtained LiFePO₄/PAS microspheres yielded outstanding discharge capacity of 88 mA h g⁻¹ at 1 C at the low temperature of −20 °C. This favorable electrochemical performance is attributed to the homogeneous morphology, the small particles inside, the porous surface, and the conductive PAS (8.5 wt%).

Fig. 10 Rate capacity performance of LiFePO₄ at various temperatures.

Besides coating conductive carbon, metal ion doping is another common method to improve the conductivity of LiFePO₄. Various metal dopants have been studied, such as Mg²⁺, Al³⁺, Cr³⁺, Ti⁴⁺, Zr⁴⁺, Nb⁵⁺, W⁶⁺, etc. A metal dopant will replace the Li⁺ (M1 doping) or Fe²⁺ (M2 doping) in LiFePO₄ lattice to generate an n-type semiconductor.³⁷ The conductivity can be enhanced by over seven orders of magnitude after doping. It was believed that the dopants that replaced metal ions in LiFePO₄ lattice could enhance the electronic and ionic conductivities.²⁴ In 2011, Liao's group¹⁵⁵ optimized LiFePO₄ by slight Mn-substitution. The as-obtained LiFe_0.98Mn_0.02PO₄/C delivered 99.8 mA h g⁻¹ (1 C) at −20 °C, which is superior to 90.7 mA h g⁻¹ for LiFePO₄/C. Even at the low temperature of −40 °C, LiFe_0.98Mn_0.02PO₄/C still gave 70.5 mA h g⁻¹ at 1 C.

Recently, the mixtures of LiFePO₄ and Li₃V₂(PO₄)₃ (i.e., xLiFePO₄ + yLi₃V₂(PO₄)₃, x > 0, y > 0) exhibited excellent electrochemical performance and have recently attracted much attention.^156,157 Chen's group¹⁴⁴ found that LiFePO₄/C delivered a discharge capacity of only 45.4 mA h g⁻¹ (0.3 C) at −20 °C, being 31.5% of the capacity obtained at 23 °C. When compared with LiFePO₄/C, the Li₃V₂(PO₄)₃/C could give a discharge capacity of 108.1 mA h g⁻¹ (0.3 C) at −20 °C, which is 86.7% of the capacity at 23 °C. It is found that the activation energies of LiFePO₄ and Li₃V₂(PO₄)₃ are calculated to be 47.48 kJ mol⁻¹ and 6.57 kJ mol⁻¹, respectively. The Li-ion transformation in Li₃V₂(PO₄)₃ is much easier than that in LiFePO₄. Recently, many mixtures of LiFePO₄ and Li₃V₂(PO₄)₃ have been studied and given excellent rate performances, such as 100 mA h g⁻¹ at 10 C for 2LiFePO₄/C + Li₃V₂(PO₄)₃/C,¹⁵⁸ 116.8 mA h g⁻¹ at 10 C for 3LiFePO₄/C + Li₃V₂(PO₄)₃/C,¹⁵⁹ 114 mA h g⁻¹ at 10 C for 5LiFePO₄ + Li₃V₂(PO₄)₃,¹⁶⁰ and 93.3 mA h g⁻¹ at 10 C for 9LiFePO₄/C + Li₃V₂(PO₄)₃/C.¹⁶¹ However, up to now, the low temperature performance has still been rarely reported. It is worthwhile to examine the influence of the combination of LiFePO₄ and Li₃V₂(PO₄)₃ on the performance of Li-ion battery, and the low temperature of LiFe_xV_2/3-2x/3PO₄/C (0 < x < 1) solid state materials.

An electrolyte is another main reason in addition to LiFePO₄ itself. The main reason might be the poor performances of conventional LiPF₆/EC + DMC + EMC electrolyte system at low temperatures. The electrolyte becomes more viscous at low temperatures, resulting in a low diffusion coefficient of lithium ions in the electrolyte. Performance degradation can be expected since fast Li-ion diffusion (i.e., shorter intercalation time and better utilization of the active material) is of crucial importance for high performance cathodes. The inherent properties of olivine LiFePO₄ might also contribute to the degradations, but it is less likely to be the main reason. In the near future, more efforts are needed to improve the low temperature performances. Huang's group³⁵ reported that LiPF₆ salt in EC (the high freezing point of 36.4 °C) easily leads to poor Li-ion diffusion at low temperatures, particularly below −20 °C. Thus, the exploration of novel electrolyte systems with superior low temperature properties might be a promising research direction. The viscosity increase of the electrolyte at low temperatures may be the reason.²⁷ Ma's group¹⁶² discussed the 1.0 M LiPF₆/EC + DMC + DEC + EMC (1 : 1 : 1 : 3, v/v) electrolyte (Table 1). LiFePO₄/C composite could yield 90 mA h g⁻¹ and 69 mA h g⁻¹ at −20 °C and −40 °C, respectively. Moreover, Zhang's group¹⁶³ found that the LiFePO₄/C composite could operate over a wide temperature range (−50 to 80 °C) using new lithium salts LiBF₄–LiBOB in a solvent mixture of PC + EC + EMC (1 : 1 : 3). Although the LiBOB-based electrolyte has high conductivity above −10 °C and high temperature performance even up to 90 °C, but it fails to perform below −40 °C. Fortunately, the mixture of LiBF₄ and LiBOB helps the electrolyte process a large working temperature range from −50 °C to 90 °C due to the high conductivity of LiBF₄ below −10 °C. With the help of LiBF₄–LiBOB-based electrolytes, LiFePO₄ delivered 62 mA h g⁻¹ and 43 mA h g⁻¹ even at −40 °C and −50 °C.

Table 1 Various LiFePO₄ and the corresponding electrolyte responsible for the excellent low-temperature performance

Materials	Synthesis method	Low temperature range/°C	Electrolyte	Ref.
LiFePO4/C	Solid state reaction method	−20	1 M LiPF6/EC + DEC (1 : 1)	144
LiFePO4/C	Solid state reaction method	−20	1 M LiPF6/EC + DEC (1 : 1)	151
LiFePO4/C	Solid state reaction method	−40	1 M LiPF6/EC + DEC (1 : 1)	152
LiFePO4/C	Co-precipitation method	−20	1 M LiPF6/EC + DEC (1 : 1)	153
LiFePO4/PAS	Co-precipitation method	−20	1 M LiPF6/EC + DEC (1 : 1)	154
LiFe0.98Mn0.02PO4/C	Solid state reaction method	−40	1 M LiPF6/EC + DEC (1 : 1)	155
LiFePO4/C	Solid state reaction method	−40	1 M LiPF6/EC + DEC + DMC + EMC (1 : 1 : 1 : 3)	162
LiFePO4/C	Solid state reaction method	−50	1 M (0.9LiBF4 + 0.1LiBOB)/PC + EC + EMC (1 : 1 : 3)	163

---

3.3 Volumetric energy density

The high volumetric energy density of batteries can make the batteries smaller in volume while exhibiting the same performances.^35,149 When compared with the mass density (5.1 g cm⁻³) of the dominant LiCoO₂ cathode materials, the theoretical density of LiFePO₄ is only 3.6 g cm⁻³.¹⁶⁴ However, the tap density of LiFePO₄ usually lies in the range of 0.8–1.5 g cm⁻³, which seriously limits the volumetric energy density and becomes one main obstacle preventing LiFePO₄ to be acceptable in commercial applications.^35,165 Optimizing the particle morphology and distribution has proven to be helpful for superior performance. In fact, the spherical architecture of LiFePO₄ is preferred to optimize the tap density. When compared with irregular particles, spherical LiFePO₄ particles can decrease the vacant space between the particles and improve the fluidity of the particles. Furthermore, less binder (polyvinylidene fluoride (PVDF) or PTFE) can be stuck to spherical LiFePO₄ particles in fabricating the cathode film, which can enhance the volumetric energy density. In addition, the coating carbon with a density of 2.2 g cm⁻³ consumes volumetric energy density while efficiently enhancing the gravimetric energy density.²⁷ Thus, it is imperative to consider the weight ratio and optimize the balance to get a high gravimetric energy density without sacrificing the volume energy density for LIBs.

Recently, the co-precipitation method,¹⁶⁶ molten-salt method,¹⁶⁷ HTS method,¹⁶⁸ and microwave assisted water-bath reaction (MW-WBR)¹⁶⁹ have been employed to prepare LiFePO₄ with high tap density, as shown in Table 2. The co-precipitation method (including the controlled crystallization method) is considered as one of the most versatile techniques to produce spherical LiFePO₄ particles. Firstly, spherical amorphous FePO₄·xH₂O powders are prepared by the co-precipitation method. Ying's group¹⁶⁶ adopted a reactor (Fig. 11a) to prepare microspherical amorphous FePO₄·xH₂O powders (Fig. 11b) with 8 μm by the controlled crystallization method following the reaction: Fe(NO₃)₃ + H₃PO₄ + 3NH₃ + xH₂O = FePO₄·xH₂O + 3NH₄NO₃. It was found that the pH value should be fixed on 2.1, avoiding the hydrolyzation of PO₄³⁻ into HPO₄²⁻ and H₂PO₄⁻ as the solution's pH decreases and protecting Fe³⁺ from hydrolyzing into Fe(OH)₂⁺, Fe(OH)₂⁺ and Fe(OH)₃ when the solution's pH increases. Secondly, the Li source, carbon source and some doping elements were added with FePO₄·xH₂O and then heat treated at high temperatures. Ying's group¹⁶⁶ synthesized spherical Li_0.97Cr_0.01FePO₄/C (Fig. 11c) with a tap density of 1.8 g cm⁻³, which is superior to 1.37 g cm⁻³ prepared by ball milling^170,171 and similar to 1.82 g cm⁻³ synthesized by a secondary ball milling method.¹⁷² The Li_0.97Cr_0.01FePO₄/C composite exhibited 163 mA h g⁻¹ at 0.005 C, but had low discharge capacity of 142 mA h g⁻¹ at 0.1 C and unsatisfactory rate capability of 110 mA h g⁻¹ at 1 C (Fig. 12a) with the corresponding volumetric discharge capacity of 198 mA h cm⁻³ (Fig. 12b). This is due to the large amount of inert Li_0.97Cr_0.01FePO₄ at the heart of the solid Li_0.97Cr_0.01FePO₄ microsphere, which contacted poorly with electrolytes, as shown in Fig. 8a.¹³⁸

Table 2 Various samples with high volumetric energy density

Materials	Synthesis methods	Tap density (g^3 cm^−3)	Ref.
LiFePO4/Fe2P	Ball milling	1.37	170 and 171
LiFePO4/C	Ball milling	1.82	172
Li0.97Cr0.01FePO4/C	Controlled crystallization method	1.8	166
LiFePO4/PAS	Controlled crystallization method	1.6	154
LiFePO4/C	Co-precipitation method	1.5–1.6	153, 173 and 175
LiFePO4/C	Co-precipitation method	1.8	174
LiFePO4	Molten-salt method	1.55	167
LiFePO4/C	Hydrothermal method	1.4	168
LiFePO4	Spray drying method	1.7	176
LiFePO4	Microwave assisted water-bath reaction	2.0	169

---

Fig. 11 (a) Schematic diagram of the reactor for controlled crystallization process. The SEM images panorama and individual (inset) for (b) FePO₄·xH₂O and (c) Li_0.97Cr_0.01FePO₄/C particles prepared by controlled crystallization method.¹⁶⁶ (Copyright © 2005 Elsevier B.V. All rights reserved.) (d) Porous sponge like LiFePO₄ and the corresponding cross-sectional SEM image (inset).¹⁷⁵ (Copyright 2009, The Electrochemical Society.) (e) SEM images of LiFePO₄/C containing 7.0 wt% carbon synthesized by coprecipitation method using self-produced high-density FePO₄ pressure-filtrated at 20 MPa.¹⁷⁴ (Copyright © 2009 Elsevier Ltd. All rights reserved.) (f) SEM images for panorama and individual (inset) quasi-spherical FePO₄·2H₂O precursor synthesized by the hydrothermal method.¹⁶⁸ (Copyright © 2011, Royal Society of Chemistry.) (g) SEM micrographs (the inset is an enlarged SEM image) of LiFePO₄ with a fairly high tap density of 2.0 g cm⁻³ obtained by microwave assisted water-bath reaction method.¹⁶⁹ (Copyright © 2013 World Scientific Publishing Co.)

Fig. 12 (a) Rate capacities for various samples and (b) their volumetric energy densities.

In order to simultaneously achieve a high gravimetric energy density and tap density, the conductive polymer PAS was employed to improve the rate capability of LiFePO₄ using the controlled crystallization method.¹⁵⁴ They prepared LiFePO₄/PAS with a tap density of 1.6 g cm⁻³, which exhibited a rate capability of 129 mA h g⁻¹ and 97 mA h g⁻¹ at 1 C and 5 C respectively, corresponding to the volumetric discharge capacities of 206.4 mA h cm⁻³ and 155.2 mA h cm⁻³ at 1 C and 5 C, respectively. In 2010, Sun's group and co-workers¹⁷³ prepared spherical LiFePO₄/C using polyvinylpyrrolidone (PVP) as a carbon source with a particle size of 6 μm and high tap density of 1.6 g cm⁻³. The as-obtained LiFePO₄/C yielded discharge capabilities of 132 mA h g⁻¹ (211.2 mA h cm⁻³) and 108 mA h g⁻¹ (172.8 mA h cm⁻³) at 1 C and 5 C, respectively. Chang's group¹⁷⁴ improved the carbon content of LiFePO₄/C to 7.0 wt% and synthesized LiFePO₄/C with a high tap density of 1.8 g cm⁻³ by the co-precipitation method using self-produced high-density FePO₄ pressure-filtrated at 20 MPa (Fig. 11e). Due to the denser structure, LiFePO₄/C delivered volumetric discharge capability of 300.6 mA h cm⁻³ at 0.1 C. Even at 1 C and 5 C, the LiFePO₄/C also exhibited high capacities of 257.4 mA h cm⁻³ and 176.9 mA h cm⁻³, respectively.

To further improve the high rate performance, 3D porous sponge-like LiFePO₄/C particles with a high tap density of 1.5 g cm⁻³ were successfully prepared by the co-precipitation method using pitch as the carbon source (Fig. 11d).¹⁷⁵ Although each sponge-like LiFePO₄/C particle with a micron size of 6 μm consisted of nanoscale 200–300 nm primary LiFePO₄/C particles, the sponge-like LiFePO₄/C particle gave volumetric discharge capacity of 214.5 mA h cm⁻³ and 186.0 mA h cm⁻³ at 1 C and 5 C, respectively, which is 2.5 times that of nano-sized LiFePO₄/C with a tap density of 0.6 g cm⁻³. In 2010, double carbon coating (sucrose and pitch as the carbon source) was employed to synthesize LiFePO₄/C, which possessed a tap density of 1.5 g cm⁻³ and yielded volumetric discharge capacities of 225.0 mA h cm⁻³ and 193.5 mA h cm⁻³ at 1 C and 5 C, respectively.¹⁵³ Even at 10 C and 20 C, the LiFePO₄/C delivered 116 mA h g⁻¹ (174.0 mA h cm⁻³) and 79 mA h g⁻¹ (118.5 mA h cm⁻³), respectively.

In addition to the co-precipitation method, Zhou's group¹⁶⁷ prepared microspherical LiFePO₄ with a tap density of 1.55 g cm⁻³ via a molten-salt method, which showed accelerated reaction rate and controllable particle morphology. However, the LiFePO₄ product delivered a low capacity of 130.3 mA h g⁻¹ (201.5 mA h cm⁻³) at 0.1 C due to the large amounts of inert LiFePO₄ at the heart of the solid LiFePO₄ microsphere.¹³⁸ In 2011, quasi-microspheres of LiFePO₄/C composed of many densely compact nanoplates (with 100 nm size and 30 nm thickness) were synthesized by Zhang's group (Fig. 11f).¹⁶⁸ Due to the FePO₄·2H₂O nanoplates assembled quasi-microspherical precursors via the hydrothermal process, the as-obtained LiFePO₄/C quasi-microspheres possessed a tap density of 1.4 g cm⁻³ and showed excellent high rate performance. Even at 30 C current rate, the discharge capacities can reach 75 mA h g⁻¹ (105 mA h cm⁻³). In 2014, Kim's group¹⁷⁶ successfully prepared 3D porous LiFePO₄ microspheres with a high tap density of 1.7 g cm⁻³ using a spray drying (SD) method. The as-obtained 3D porous LiFePO₄ microspheres with the carbon content of 3.3% delivered high initial discharge capacity of 160 mA h g⁻¹ (272 mA h cm⁻³) at 0.1 C, corresponding to 94% of the theoretical capacity, as shown in Fig. 12.

Interestingly, the MW-WBR method was employed to synthesize spherical LiFePO₄/C with a fairly high tap density of 2.0 g cm⁻³, which will strongly benefit the enhancement of volumetric energy density (Fig. 11g).¹⁶⁹ This spherical LiFePO₄/C exhibited a tap density higher than 1.0 g cm⁻³ (Tianjin Sterlan-Energy Ltd. China), 1.4 g cm⁻³ (Phostech Lithium Inc., Canada), and 1.5 g cm⁻³ (Valence Technology Inc., US).¹⁷⁴ Furthermore, spherical LiFePO₄/C yielded excellent discharge capacities of 131 mA h g⁻¹ (262.0 mA h cm⁻³) and 105 mA h g⁻¹ (210 mA h cm⁻³) at 1 C and 5 C, respectively.

4 Synthetic reaction mechanism

4.1 Carbothermal reduction method

Recently, synthesis approaches and equipment have been used towards fabricating cathode material LiFePO₄, such as solid phase thermal (SPT) routes and liquid phase thermal (LPT) routes.^177,178 It is well known that the carbothermal reduction (CTR) method is one of the promising SPT methods. The CTR method is considered as an effective approach for improving the electrochemical performance of LiFePO₄ materials.^127,179 The CTR method not only provides a special environment favorable for the reduction of Fe(III) and the formation of the LiFePO₄/C material but also provides a continuous conductive carbon film as an electron conductor that enhances the electronic conductivity of LiFePO₄.⁹⁴ Moreover, the evenly distributed carbon can also prevent particle coalescence. Recently, the CTR method has been defined as a feasible, low-cost and environment-friendly method for the fabrication of LiFePO₄/C composites and has been largely employed to synthesize high performance LiFePO₄/C composites.

There are mainly two steps in the CTR method. At the first step, carbon sources are evenly distributed in precursor aggregation using various routes, such as ball milling,^180,181 sol–gel,^182,183 co-precipitation,^184,185 spray drying,^138,186 etc. Carbon sources include inorganic carbon (e.g., carbon black + surface activator), organic sucrose, small molecular acid (e.g., citric acid), big molecular polymer (e.g., PEG10000)¹⁸⁷ and sp²-type carbon (e.g., carbon nanotube (CNT) and graphene).^188,189 At the second step, LiFePO₄/C is formed due to vigorous gas evolution (mainly CO and CO₂) during the degradation and carbonization of the carbon sources. During the heat treatment of an appropriate precursor, elementary carbon is deposited on the walls of primary nanoparticles as a degradation product. Meanwhile, the continuous conductive carbon film can overcome the electronic and lithium-diffusion limitations, improve the rate of insertion/extraction and optimize the electrochemical performance under high-current regimes.

Noticeably, the electrochemical performance of LiFePO₄/C material is influenced by various carbon sources and raw materials. Iron phosphides, such as FeP and Fe₂P, were reported to have effects on the improvement of the rate performance. Obviously, the impurities, especially for Fe₂P, easily appeared as by-products in the CTR synthetic process. It is difficult to control the content of Fe₂P, which has effects on the electrochemical performance of LiFePO₄. Why Fe₂P was synthesized by carbothermal reaction at a different temperature? It is noteworthy that the phase of Fe₂P could take place in the following conditions, such as the self-deoxidization reaction of LiFePO₄/C composite, reducing agent iron nanoparticles, and stoichiometric excess of Fe₂O₃. Moreover, Fe₂P could also appear when using the microwave assisted CTR (MW-CTR) method. In this part, we focus on the synthetic reaction mechanism in CTR synthetic conditions.

4.1.1 Effects of Fe(II) and Fe(III) raw material sources. Usually, Fe(III) sources including FePO₄, Fe₂O₃, Fe(NO₃)₃ and Fe(NH₄)₂(SO₄)₂·6H₂O are employed to prepare LiFePO₄ via the CTR routes.¹⁹⁰ In our previous research,⁹⁴ it is found that LiFePO₄/C can be strongly formed at around 450 °C by an in situ CTR method whether using various carbon sources except for inorganic carbon sources (postponed to 539 °C due to the preparation of Li₃Fe₂(PO₄)₃, as shown in Fig. 13a and b), similar to the reports by Garbarczyk's group¹⁹¹ and Jiao's group.¹⁹² The self-deoxidation reaction of LiFePO₄/C takes place at 840 °C (Table 3), which was in accordance with the results of Wang's group.¹⁹³ The crystallization of Fe₂P and Li₄P₂O₇ occurs at 930 °C, suggesting the severe decomposition of LiFePO₄/C. Normally, the in situ CTR synthetic temperature should be carefully controlled below 840 °C to avoid the Fe₂P by-product during the sintering process. However, it was found that a suitable content of Fe₂P would improve the electrochemical performance of LiFePO₄. Kim's group^194,195 and Lee's group¹⁹⁶ found that the Fe₂P content increased with the increasing carbon content at the temperature of 900 °C, as shown in Fig. 13c. As the carbon content was above 12 wt%, a large amount of LiFePO₄/C disappeared and Fe₂P was predominantly produced.

Fig. 13 (a) TG-DTA curves of the Fe(III) precursor using carbon black as carbon sources, and (b) corresponding XRD patterns of the precursor and LiFePO₄/C samples after heating at various temperatures.⁹⁴ (Copyright © 2009 Elsevier Ltd. All rights reserved.) (c) XRD profiles of the LiFePO₄/Fe₂P composites synthesized using different amounts of excess carbon.¹⁹⁵ (Copyright © 2006 Elsevier B.V. All rights reserved.) (d) XRD patterns of samples prepared from a slightly oxidized Fe(II) precursor using various reductive heat treatments.²⁰⁸ (Copyright © 2003 Elsevier Science B.V. All rights reserved.)

Table 3 Reaction mechanism of deoxidization reaction of LiFePO₄ in various conditions

Reduction condition	Chemical reaction equation	Ref.
Self-deoxidization reaction of LiFePO4/C composite (at 840 °C)	4LiFePO4 + 7C (or 14C) = 2Fe2P + 2Li2O + 2P + 7CO2 (or 14CO)	94 and 195
Severe decomposition of LiFePO4/C composite (at 938 °C)	4LiFePO4 + C (or 2C) = 2Fe2P + 2Li4P2O7 + P + CO2 (or 2CO)	94
Reducing agents (Fe3O4, Fe3C, Fe nanoparticles) from FeC2O4·2H2O (at 675 °C)	8FeC2O4·2H2O + 4LiFePO4 + 24C (or 48C) = 4Fe2P + 16H2O + 24CO2 (or 48CO)	199 and 200
Reducing agent iron nanoparticles	8Fe + 4LiFePO4 + 7C (or 14C) = 4Fe2P + 2Li2O + 7CO2 (or 14CO)	202
Reductive H2 atmosphere	6LiFePO4 + 16H2 = 2Fe2P + 2Li3PO4 + 2FeP + 16H2O	99
Microwave assisted reaction	8LiFePO4 = 2Li4P2O7 +4Fe2P +9O2	215
Microwave assisted reaction (without atmosphere of inert gases)	12LiFePO4 + 3O2 = 4Li3Fe2(PO4)3 +2Fe2O3	213 and 222

---

Besides Fe(III) raw material source, FeC₂O₄ was generally employed as a Fe(II) raw material source to synthesize LiFePO₄. It can be seen that the presence of Fe₂P appeared at 675 °C.^197,198 Zboril's group¹⁹⁹ and Molenda's group²⁰⁰ found that the reducing agents including α-Fe nanoparticles, magnetic Fe₃O₄ and Fe₃C could be obtained during the thermal decomposition of FeC₂O₄·2H₂O in a dry argon flow. These reducing agents, especially for Fe particles, could react with LiFePO₄/C and generate Fe₂P around 700 °C.^201,202 It was worth noting that iron phosphides increased with the increase of excessively stoichiometric iron source and temperature.^203,204

4.1.2 Reductive atmosphere of hydrogen. In addition to the self-deoxidization reaction of LiFePO₄/C composite and reducing agents from Fe raw materials sources, the reductive atmosphere of hydrogen also reduced LiFePO₄ to prepare Fe₂P at 600 °C according to the work of Nazar's group.⁹⁹ The Fe₂P can be observed after the sintering process at 600 °C for 4 h in a gas atmosphere of N₂/H₂ = 97/3 (vol%) reported by Taniguchi's group.^205,206 Furthermore, the phase of FeP and Fe₃P could also be observed.²⁰⁷ Wohlfahrt-Mehrens's group²⁰⁸ studied various gas atmospheres and considered that some traces of Fe(III) could be confirmed when using Fe(II) raw sources but insufficient reductive additives (e.g. N₂/H₂ = 99/1 (vol%), no carbon as the reductive additive). According to the results, the authors suggested that a considerable amount of Fe₂P was generated as the hydrogen raised to 10 vol%, as shown in Fig. 13d. The amount of Fe₂P would increase with the carbon addition. Additionally, it is important to note that LiFePO₄ will decompose into Fe₂P, Li₃PO₄, Li₄P₂O₇ and Li₃P₇ when the heating temperature reaches 850° C under a stream of a gas mixture of Ar/H₂ = 95/5 (vol%).²⁰⁹

4.1.3 Microwave assisted carbothermal reduction method. When compared with a conventional CTR method, MW-CTR method has recently attracted much attention.¹⁷⁷ Microwave irradiation (2.54 GHz) can provide “inert and instant heating”, and exhibit outstanding superiority such as fast response, time effectiveness, low energy consumption, and environmentally benignity.^210,211 Moreover, microwave irradiation was also adopted to optimize the physical and electrochemical performance of LiFePO₄.²¹² Under Ar atmosphere, Higuchi's group²¹³ found that iron acetate acts as a microwave susceptor rather than iron lactate. This result illustrates that LiFePO₄ was successfully prepared using iron acetate as the Fe sources at 500 W for 10 min, whereas LiFePO₄ was not observed using iron lactate as Fe raw sources even up to 30 min. Furthermore, the microwave irradiation time has an important effect on the Fe₂P by-products, which could increase with the increasing time due to the work of Kwon's group.²¹⁴ All the phases of LiFePO₄, Fe₂P and Li₄P₂O₇ were observed when the amount of Fe₂P is above 1.0 wt%.²¹⁵ In addition to Ar atmosphere, carbon materials and carbon sources were both employed to produce vigorous CO gas, which is necessary for the preparation of LiFePO₄.^216,217 Because of this strategy, the electrochemical performance of LiFePO₄ improved considerably by the MW-CRT method using carbon coating,²¹⁸ multi-walled CNTs (MWCNTs) doping,²¹⁹ Mo and La doping.^220,221 However, it is worthwhile to note that most of the carbon could be consumed after extended heating treatment. The transformation from Fe(II) to Fe(III) would take place, leading to impurity phases of Fe₂O₃ and Li₃Fe₂(PO₄)₃.^222,223 Therefore, the microwave-derived and phase-pure LiFePO₄ without impurities could be successfully synthesized via the MW-CTR method under certain proper conditions.

4.2 Low-temperature liquid phase thermal synthesis

In addition to the high-temperature CTR method, the low-temperature LPT routes, such as HTS, STS and ionothermal synthesis (ITS), provided more effective pathways to synthesize LiFePO₄. HTS and STS are generally LPT synthetic methods, which exhibit several advantages including morphology control, homogeneous particle size distribution, relatively low temperature (≤300 °C), low cost and simplicity.^224,225 Recently, HTS and STS were both extensively employed to prepare special morphologic LiFePO₄, such as crystalline growth, nanostructured LiFePO₄ and 3D porous LiFePO₄ architectures with micro-nano-structures (see Section 3.1.1 in this article).^127,226 In order to improve traditional intermittent reactors for HTS and STS, Teja's group²²⁷ and Aimable's group²²⁸ designed the continuous hydrothermal synthesis, which poise great promise in practical industrial applications. Different from HTS (using water as the main mineralizer) and STS (using organic solvents as the main mineralizers), room temperature ionic liquids (RTILs) are regarded as a novel class of mineralizers that have been employed in the ITS route to synthesize cathode materials, especially LiFePO₄ with nano-structures.^229–231 Moreover, microwave irradiation is expected to assist HTS, STS and ITS due to energy- and time-saving advantages of microwaves. In this part, we focus on the effects of stoichiometric ratio of raw sources, pH values of solution, additive mediations (e.g., organic acids, surfactants and solvents), and microwave irradiation.

4.2.1 Stoichiometric ratio of raw sources. Usually, the crystalline powders of LiFePO₄ were synthesized by mixing amounts of the reactants LiOH, FeSO₄ and H₃PO₄ in the stoichiometric ratio of n_Li : n_Fe : n_P = 3 : 1 : 1, as shown in Table 4.²³² An advisable addition of LiOH should be changed with replacing the H⁺ with NH₄⁺ in H₃PO₄. Kanamura's group^233,234 found that the stoichiometric ratio of 2.5 : 1 : 1 would be better when using (NH₄)₃PO₄ instead of H₃PO₄. The stoichiometric ratio of 1 : 1 : 1 and 2 : 1 : 1 were both appropriate for using (NH₄)₂HPO₄. As shown in Fig. 14, our group investigated the products when using NH₄H₂PO₄ as the PO₄³⁻ source. It can be seen that a large amount of Fe₃(PO₄)₂·H₂O as an impurity was obtained in the case of x = 1.0. As the x value changed to 1.5, additional crystallinity of LiFePO₄ was observed. The value of x = 2.0 is obviously in favor of the hydrothermal crystallization of pure LiFePO₄. The further addition of LiOH (e.g., x = 2.5 and 3.0) induced impurities in Li₃PO₄. All these results strongly suggest that the addition of LiOH should be changed as the PO₄³⁻ source was changed. However, the stoichiometric ratio should be changed back to 3 : 1 : 1 when the other Li sources including LiCl and CH₃COOLi were used instead of LiOH.^235,236 It is noteworthy that LiFePO₄ with a pure phase was also prepared by using other Fe(II) salts such as FeCl₂ and (NH₄)₂Fe(SO₄)₂ instead of FeSO₄, but the stoichiometric ratio remained unchanged.^237,238 Furthermore, LiFePO₄ was also successfully synthesized using fresh Fe₃(PO₄)₂·5H₂O and Li₃PO₄.²³⁹

Table 4 Summary of representative reaction mechanism of LiFePO₄ prepared by low-temperatured LPT synthesis. In the table, mediator includes organic acids or surfactant compounds

Synthesis type	Reaction reagents	Stoichiometric ratio	Mediator	Solvent	Condition	Morphology	Ref.
HTS	LiOH, FeSO4, H3PO4	3 : 1 : 1	—	H2O	120 °C, 5 h	Hexagonal platelets	232
HTS	LiOH, FeSO4, (NH4)3PO4	2.5 : 1 : 1	—	H2O	170 °C, 12 h	—	233
HTS	LiOH, FeSO4, (NH4)2HPO4	1 : 1 : 1&2 : 1 : 1	—	H2O	170 °C, 12 h	Nanoplates	234
HTS	LiOH, FeSO4, NH4H2PO4	2 : 1 : 1	—	H2O	180 °C, 5 h	Spindle-like particles	This work
HTS	LiOH, FeCl2·4H2O, P2O5	6 : 3 : 2	—	H2O	170 °C, 3 days	Hexagonal platelets	237
HTS	LiOH, FeSO4, H3PO4	3 : 1 : 1	CTAB	H2O	120 °C, 5 h	Nanoparticles	251–253
HTS	LiOH, (NH4)2Fe(SO4)2·6H2O, H3PO4	3 : 1 : 1	P123, D230	H2O	220 °C, 24 h	Nanoparticles	106
HTS	LiOH, (NH4)2Fe(SO4)2·6H2O, H3PO4	3 : 1 : 1	Citric acid	H2O	180 °C, 10 h	Spindle-like particles	250
HTS	LiOH, FeSO4, NH4H2PO4	2 : 1 : 1	NTA	H2O	180 °C, 20 h	Nanowires	254
HTS	LiAc, FeCl3, NH4H2PO4	1 : 1 : 1	SDS + EN	H2O	180 °C, 10 h	Spheres assembled from nanorods	This work
STS	LiOH, FeSO4, H3PO4	3 : 1 : 1	—	TEG + H2O	190 °C, 5 h	Rectangular nanoplatelets	111
STS	LiOH, FeSO4, H3PO4	3 : 1 : 1	Citric acid	MPG + H2O	140 °C, 12 h	Needles assembled nanorods	255
STS	LiOH, FeSO4, H3PO4	3 : 1 : 1	—	PEG400 + H2O	150 °C, 3 h	Needle-like particles	256
STS	LiOH, FeSO4, H3PO4	3 : 1 : 1	—	PEG400 + H2O	180 °C, 9 h	Hexagonal platelet	112
ITS	LiH2PO4, FeC2O4·2H2O	1 : 1	—	CN-based IL	250 °C, 24 h	Needles assembled Lego blocks	229
ITS	LiOH, FeSO4, H3PO4	3 : 1 : 1	SDBS + ascorbic acid	IL + H2O	240 °C, 20 h	Nanorods	258
MW-HTS	LiOH, FeSO4, H3PO4	3 : 1 : 1	—	H2O	200 °C, 5 min	Globular particles	259
MW-STS	LiOH, FeAc2, H3PO4	1 : 1 : 1	—	TEG	300 °C, 5 min	Nanorods	262

---

Fig. 14 XRD patterns of the samples prepared using different values of x in a molar ratio of x : 1 : 1 for the starting materials including LiOH, FeSO₄ and NH₄H₂PO₄.

Additionally, Yu's group²⁴⁰ suggested that LiFePO₄ should only be obtained under neutral and slightly basic conditions. Fe disorder onto the Li sites (ca. 3.5%) generally happened at a pH value of 6.30. Even at a synthetic condition temperature below 175 °C, there would be some amounts of Fe on Li sites.^241,242 Moreover, the synthetic time also had effects on the LiFePO₄ nanocrystalline structure. In 2013, Ou's group²⁴³ found that the peak intensities of (020) X-ray diffraction line of LiFePO₄ nanoplates increased with the reaction time prolonging from 0 to 6 h. Meanwhile, the relative peak intensities of (020) to (111) also increased due to the results calculated from the XRD patterns of LiFePO₄ nanoplates. All these results demonstrated that the LiFePO₄ nanoplates have a preferred crystal orientation with the large (010) face, which facilitates the good electrochemical performance of LiFePO₄.²⁴⁴

4.2.2 Additive mediations of organic acids, surfactants and solvents. In addition to the synthetic temperature, residence time, reactant concentration and pH value,²⁴⁵ additive mediations including organic acids, surfactants and solvents also have effects on the purity, morphology control and electrochemical performance of LiFePO₄.²⁴⁶ Firstly, organic acid not only affects the pH value but also can be used as a reduction mediator. Whittingham's group²⁴⁷ provided that both of the reduction from Fe(III) into Fe(II) and the formation of the desired LiFePO₄ compound carried out with the addition of L-ascorbic acid. Moreover, it was found that the pH value and reaction time played several roles in self-assembled mesoporous LiFePO₄ with hierarchical spindle-like architectures.²⁴⁸ When the value of pH increased from 7.0 to 10.0, primary nanocrystals disappeared and spindle-like hierarchical LiFePO₄ particles formed. The full transformation from Li₃PO₄ to LiFePO₄ nanocrystal needs a reaction time of over 5 h. Furthermore, in the presence of an organic acid, e.g. citric acid or ascorbic acid, well-crystallized LiFePO₄ nanoparticles have been directly synthesized.²⁴⁹ Chung's group and Grey's group²⁵⁰ successfully controlled the morphological transformations of LiFePO₄ particles during the hydrothermal reaction in the presence of citric acid and ammonium ions (NH⁴⁺).

Secondly, organic surfactants or polymers including O,O-bis(2-aminopropyl)polypropyleneglycol (D230), EO20PO70EO20 (P123), cyltrimethylammonium bromide (CTAB), sodium dodecyl sulfate (SDS), poly(ethylene glycol) (PEG), etc. have been successfully employed to control the particle growth and size distribution. The results from Nazar's group¹⁰⁶ showed the particle size of LiFePO₄ became smaller and much more homogeneous as compared to that of products using molecular reducing agents, if non-ionic surfactants, including pluronic P123 and Jeffamine D230 were added to the reactors. Recently, Gerbaldi and his co-workers^251–253 investigated the effects of the organic surfactant compound CTAB on LiFePO₄ characteristics and electrochemical behavior. It is clearly found that CTAB considerably influences the synthesis and electrochemical properties of LiFePO₄. Moreover, the CTAB micelles are beneficial for controlling the grain size and surface area. Additionally, nitrilotriacetic acid (NTA) surfactant has been also employed to synthesize LiFePO₄ nanowires.²⁵⁴ In 2010, Zhang's group¹⁶⁸ used SDS to assist the synthesis of LiFePO₄ via a hydrothermal process. The as-obtained LiFePO₄/C microspheres are composed primarily of LiFePO₄/C nanoplates with the size of 100 nm and thickness of 30 nm. With the addition of SDS and ethylenediamine (EN), our group successfully synthesized LiFePO₄/C microspheres assembled from nanorods, as shown in Fig. 15c and d.

Fig. 15 (a) SEM images of a LiFePO₄ hemisphere, demonstrating the microsphere consisting of nanofibers.²⁵⁵ (Copyright © 2013, Royal Society of Chemistry.) (b) SEM images of LiFePO₄.¹¹¹ (Copyright © 2011, Springer Science+Business Media, LLC.) (c and d) SEM images of a LiFePO₄ hemisphere, and the LiFePO₄ nanorods. (e and f) SEM image and lattice fringe (the insets of ED patterns and TEM image) of LiFePO₄ nanorods.²⁵⁸ (Copyright © 2011 Elsevier B.V. All rights reserved.) (g–i) TEM images of the large LiFePO₄ needle assembling like Lego blocks, and the corresponding SAED pattern. The white arrow shows the “cementing” zone.²²⁹ (Copyright © 2009, American Chemical Society.)

Thirdly, the hybrids of water and organic solvents have also been employed in the STS route. In 2013, Liao's group²⁵⁵ successfully prepared high-performance LiFePO₄ microspheres consisting of nanofibers (Fig. 15a) by a high pressure alcohol-thermal approach in a water and 1,2-propanediol (PD) composite solvent. With the effect of the PD solvent, the LiFePO₄ crystalline nuclei were linearly elongated to form nanofibers. LiFePO₄ nanofibers aggregated together and formed LiFePO₄ microspheres. Tetraethylene glycol (TEG) has been also used as a cosolvent with water.¹¹¹ As shown in Fig. 15b, rectangular LiFePO₄ nanoplatelets were formed with a very small thickness of 50–80 nm, favouring fast Li-ion diffusion. Moreover, PEG has been usually used to assist the morphology control of LiFePO₄.^256,257 Meanwhile, Liu's group¹¹² provided a water-PEG400 binary solvent to assist the solvothermal method and prepare LiFePO₄ nanoparticles (50 nm in size), hexagonal nanoplates (800 nm wide and 100 nm thick) and hexagonal microplates (3 mm wide and 300 nm thick).

Recently, RTILs have been also been used as the solvent and template to synthesize LiFePO₄. With this new and promising ITS method, Tarascon's group^178,229 enabled LiFePO4 crystalline growth with controlled morphology and size at around 250 °C. As shown in Fig. 15g–i, large LiFePO₄ needles were assembled similar to Lego blocks. As shown in Fig. 15e and f, Teng and co-workers²⁵⁸ successfully prepared LiFePO₄ nanorods in an ionic liquid in the presence of sodium dodecyl benzene sulfonate (SDBS) surfactant and ascorbic acid. Thus, super architectures with micro-nano-structured nanocrystals and morphology control can be obtained with the addition of organic acids, organic surfactants and organic solvents.

4.2.3 Microwave-assisted liquid phase thermal routes. Undoubtedly, microwave irradiation has been employed to assist LPT routes, because microwave irradiation assisted synthesis (MIAS) exhibited overwhelming superiorities such as heating the precursor considerably and quickly, and also provide an efficient and suitable approach for large-scale syntheses in order to be used in practical applications of electrode materials. In our previous work,¹⁷⁷ we exhaustively reviewed MW-LPT including MW-HTS,^259,260 MW-STS²⁶¹ and MW-ITS.¹⁷⁸ Li-ion diffusion can be improved considerably by controlling the morphology of LiFePO₄ via MW-LPT. Moreover, electronic conductivity can be optimized by coating conductive materials such as carbon,^262,263 CNT^264,265 and conductive polymers, e.g., poly(3,4-ethylenedioxythiophene)(PEDOT).²⁶⁶

5 Conclusion and perspective

In summary, the pathway of Li-ion diffusion in Li_xFePO₄ is mainly along the [010] direction. Sometimes, 2D Li-ion transport occurs along both the [010] and [001] directions at elevated temperatures. Undoubtedly, the phase transformation of LiFePO₄ and FePO₄ are affected by various conditions, such as particle size, charge/discharge rate, temperature, etc. To date, many phase transformation models, including shrinking core (i.e., core–shell) model, Laffont's (i.e., new core–shell) model, mosaic model, domino-cascade model, phase transformation wave model, and many-particle model, have been proposed to explain the lithiation/delithiation process and the phase transformation in Li_xFePO₄. However, these models are still in disagreement with each other. Generally, the two-phase transformation models can be used to explain a two-phase growth process involving the coexistence of LiFePO₄ and FePO₄ for the large particle, below a critical current and at a relative low temperature. Simultaneously, the quasi-single-phase transformation models are dominantly used in the solid-solution reaction where Li_xFePO₄ particles do not separate into Li-rich and Li-poor phases during the lithiation/delithiation process.

In order to endow LiFePO₄ with excellent electrochemical performance including high rate capability and low temperature performance, reducing the pathway of Li-ion diffusion, particularly the [010] direction, is an effective route. Usually, fast Li-ion diffusion is achieved by morphology controlling, such as crystal growth orientation along the a–c plane, nano-sized morphologies and 3D porous architectures with micro-nano-structures. Meanwhile, the electronic conductivity is improved by coating conductive materials, e.g., amorphous carbon, CNT, graphene, PAS, PPy, Fe₂P, etc. However, achieving both high rate performance and volumetric energy density is still a big challenge. Recently, carbon coated LiFePO₄ microspheres with the highest tap density of 2.0 g cm⁻³ have been successfully synthesized by the MW-WBR method. They can deliver outstanding discharge capacities of 93 mA h g⁻¹ and 78 mA h g⁻¹ even at 10 C and 20 C, respectively. This additional strategy provides a facile and novel route for obtaining high tap density from LiFePO₄, and it holds the potential to be extended to EES devices.

In addition, it is noteworthy that synthetic reactions play an important role to improve the performance of LiFePO₄ cathodes. When using the CTR method, the purity of LiFePO₄ is significantly affected by the Fe sources, carbon sources and the reductive atmosphere. Obviously, the impurities, particularly for Fe₂P, can be easily separated as by-products. To date, controlling the content of Fe₂P is still a major obstacle and issue, where Fe₂P either provides the high conductive networks for LiFePO₄ or blocks the pathway of Li-ion diffusion. Low-temperature LPT synthesis can be also employed to prepare LiFePO₄ with much reduced impurities. However, both the crystalline and electrochemical performances of LiFePO₄ still need to be enhanced by subsequent heat treatment. Intriguingly, on account of “inert and instant heating”, MIAS can be used to assist both the CTR and LPT routes. We believe that MIAS provides an extremely efficient and suitable route for the large-scale synthesis of LiFePO₄ electrode materials.

Acknowledgements

The authors highly acknowledge the authors and publications data of the original materials. We would like to thank Dr Chaoyi Yan, Prof. Haibin Su (School of Materials Science and Engineering, Nanyang Technological University (MSE, NTU)) and Prof. Jianyi Lin (Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research (ICES, A*STAR)) for their helpful discussions. This work was financially supported by Scientific Research Start-up Fund for High-Level Talents, Shihezi University (no. RCZX201305), the Doctor Foundation of Bingtuan (no. 2014BB004), the Program for Changjiang Scholars and Innovative Research Team in University (no. IRT1161) and the Program of Science and Technology Innovation Team in Bingtuan (no. 2011CC001).

Notes and references

1. K. Mizushima, P. C. Jones, P. J. Wiseman and J. B. Goodenough, Mater. Res. Bull., 1980, 15, 783–789.
2. T. Nagaura and K. Tazawa, Prog. Batteries Sol. Cells, 1990, 9, 209–217.
3. B. Xu, D. N. Qian, Z. Y. Wang and Y. S. L. Meng, Mater. Sci. Eng., R, 2012, 73, 51–65.
4. M. Wagemaker and F. M. Mulder, Acc. Chem. Res., 2013, 46, 1206–1215.
5. J. Cabana, L. Monconduit, D. Larcher and M. R. Palacin, Adv. Mater., 2010, 22, E170–E192.
6. X. P. Gao and H. X. Yang, Energy Environ. Sci., 2010, 3, 174–189.
7. M. R. Palacin, Chem. Soc. Rev., 2009, 38, 2565–2575.
8. S. Chu and A. Majumdar, Nature, 2012, 488, 294–303.
9. J. M. Tarascon and M. Armand, Nature, 2001, 414, 359–367.
10. R. Marom, S. F. Amalraj, N. Leifer, D. Jacob and D. Aurbach, J. Mater. Chem., 2011, 21, 9938–9954.
11. M. Wakihara, Mater. Sci. Eng., R, 2001, 33, 109–134.
12. A. S. Arico, P. Bruce, B. Scrosati, J. M. Tarascon and W. Van Schalkwijk, Nat. Mater., 2005, 4, 366–377.
13. M. Gong and G. Wall, Exergy: An International Journal, 2001, 1, 217–233.
14. K. Zaghib, A. Mauger and C. M. Julien, J. Solid State Electrochem., 2012, 16, 835–845.
15. J. B. Goodenough and Y. Kim, Chem. Mater., 2010, 22, 587–603.
16. M. S. Islam and C. A. J. Fisher, Chem. Soc. Rev., 2014, 43, 185–204.
17. M. S. Whittingham, Chem. Rev., 2004, 104, 4271–4302.
18. F. Yu, J. J. Zhang, C. Y. Wang, J. Yuan, Y. F. Yang and G. Z. Song, Progr. Chem., 2010, 22, 9–18.
19. Y. Gu, D. Chen and X. Jiao, J. Phys. Chem. B, 2005, 109, 17901–17906.
20. K. Y. Chung and K. B. Kim, Electrochim. Acta, 2004, 49, 3327–3337.
21. J. Vetter, P. Novak, M. R. Wagner, C. Veit, K. C. Moller, J. O. Besenhard, M. Winter, M. Wohlfahrt-Mehrens, C. Vogler and A. Hammouche, J. Power Sources, 2005, 147, 269–281.
22. A. K. Padhi, K. S. Nanjundaswamy, C. Masquelier, S. Okada and J. B. Goodenough, J. Electrochem. Soc., 1997, 144, 1609–1613.
23. A. K. Padhi, K. S. Nanjundaswamy and J. B. Goodenough, J. Electrochem. Soc., 1997, 144, 1188–1194.
24. S. Y. Chung, J. T. Bloking and Y. M. Chiang, Nat. Mater., 2002, 1, 123–128.
25. J. W. Fergus, J. Power Sources, 2010, 195, 939–954.
26. C. A. J. Fisher, V. M. H. Prieto and M. S. Islam, Chem. Mater., 2008, 20, 5907–5915.
27. Y. G. Wang, P. He and H. S. Zhou, Energy Environ. Sci., 2011, 4, 805–817.
28. M. S. Islam and P. R. Slater, MRS Bull., 2009, 34, 935–941.
29. D. D. MacNeil, Z. H. Lu, Z. H. Chen and J. R. Dahn, J. Power Sources, 2002, 108, 8–14.
30. K. Zaghib, P. Charest, A. Guerfi, J. Shim, M. Perrier and K. Striebel, J. Power Sources, 2004, 134, 124–129.
31. S. P. Ong, V. L. Chevrier, G. Hautier, A. Jain, C. Moore, S. Kim, X. H. Ma and G. Ceder, Energy Environ. Sci., 2011, 4, 3680–3688.
32. W. J. Zhang, J. Power Sources, 2011, 196, 2962–2970.
33. P. P. Prosini, M. Lisi, D. Zane and M. Pasquali, Solid State Ionics, 2002, 148, 45–51.
34. D. W. Liu and G. Z. Cao, Energy Environ. Sci., 2010, 3, 1218–1237.
35. L. X. Yuan, Z. H. Wang, W. X. Zhang, X. L. Hu, J. T. Chen, Y. H. Huang and J. B. Goodenough, Energy Environ. Sci., 2011, 4, 269–284.
36. D. Morgan, A. Van der Ven and G. Ceder, Electrochem. Solid-State Lett., 2004, 7, A30–A32.
37. M. S. Islam, D. J. Driscoll, C. A. J. Fisher and P. R. Slater, Chem. Mater., 2005, 17, 5085–5092.
38. C. Y. Ouyang, S. Q. Shi, Z. X. Wang, X. J. Huang and L. Q. Chen, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 69, 104303.
39. X. F. Ouyang, M. L. Lei, S. Q. Shi, C. L. Luo, D. S. Liu, D. Y. Jiang, Z. Q. Ye and M. S. Lei, J. Alloys Compd., 2009, 476, 462–465.
40. S. Nishimura, G. Kobayashi, K. Ohoyama, R. Kanno, M. Yashima and A. Yamada, Nat. Mater., 2008, 7, 707–711.
41. J. J. Yang and J. S. Tse, J. Phys. Chem. A, 2011, 115, 13045–13049.
42. J. B. Goodenough, J. Power Sources, 2007, 174, 996–1000.
43. M. Park, X. C. Zhang, M. D. Chung, G. B. Less and A. M. Sastry, J. Power Sources, 2010, 195, 7904–7929.
44. J. Y. Li, W. L. Yao, S. Martin and D. Vaknin, Solid State Ionics, 2008, 179, 2016–2019.
45. R. Malik, D. Burch, M. Bazant and G. Ceder, Nano Lett., 2010, 10, 4123–4127.
46. B. L. Ellis, K. T. Lee and L. F. Nazar, Chem. Mater., 2010, 22, 691–714.
47. R. Amin, J. Maier, P. Balaya, D. P. Chen and C. T. Lin, Solid State Ionics, 2008, 179, 1683–1687.
48. B. Ellis, L. K. Perry, D. H. Ryan and L. F. Nazar, J. Am. Chem. Soc., 2006, 128, 11416–11422.
49. G. K. P. Dathar, D. Sheppard, K. J. Stevenson and G. Henkelman, Chem. Mater., 2011, 23, 4032–4037.
50. V. Srinivasan and J. Newman, J. Electrochem. Soc., 2004, 151, A1517–A1529.
51. L. Laffont, C. Delacourt, P. Gibot, M. Y. Wu, P. Kooyman, C. Masquelier and J. M. Tarascon, Chem. Mater., 2006, 18, 5520–5529.
52. A. S. Andersson, B. Kalska, L. Haggstrom and J. O. Thomas, Solid State Ionics, 2000, 130, 41–52.
53. A. S. Andersson and J. O. Thomas, J. Power Sources, 2001, 97–98, 498–502.
54. C. Delmas, M. Maccario, L. Croguennec, F. Le Cras and F. Weill, Nat. Mater., 2008, 7, 665–671.
55. R. Dedryvere, M. Maccario, L. Croguennec, F. Le Cras, C. Delmas and D. Gonbeau, Chem. Mater., 2008, 20, 7164–7170.
56. D. Burch, G. Singh, G. Ceder and M. Z. Bazant, in Theory, Modeling and Numerical Simulation of Multi-Physics Materials Behavior, ed. V. Tikare, G. E. Murch, F. Soisson and J. K. Kang, 2008, vol. 139, pp. 95–100.
57. G. K. Singh, G. Ceder and M. Z. Bazant, Electrochim. Acta, 2008, 53, 7599–7613.
58. W. Dreyer, J. Jamnik, C. Guhlke, R. Huth, J. Moskon and M. Gaberscek, Nat. Mater., 2010, 9, 448–453.
59. N. Meethong, H. Y. S. Huang, S. A. Speakman, W. C. Carter and Y. M. Chiang, Adv. Funct. Mater., 2007, 17, 1115–1123.
60. G. Y. Chen, X. Y. Song and T. J. Richardson, Electrochem. Solid-State Lett., 2006, 9, A295–A298.
61. P. P. Prosini, J. Electrochem. Soc., 2005, 152, A1925–A1929.
62. D. X. Gouveia, V. Lemos, J. A. C. de Paiva, A. G. Souza, J. Mendes, S. M. Lala, L. A. Montoro and J. M. Rosolen, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 72, 024105.
63. D. Li, T. Zhang, X. Z. Liu, P. He, R. W. Peng, M. Wang, M. Han and H. S. Zhou, J. Power Sources, 2013, 233, 299–303.
64. C. V. Ramana, A. Mauger, F. Gendron, C. M. Julien and K. Zaghib, J. Power Sources, 2009, 187, 555–564.
65. Y. Orikasa, T. Maeda, Y. Koyama, H. Murayama, K. Fukuda, H. Tanida, H. Arai, E. Matsubara, Y. Uchimoto and Z. Ogumi, Chem. Mater., 2013, 25, 1032–1039.
66. W. C. Chueh, F. El Gabaly, J. D. Sugar, N. C. Bartelt, A. H. McDaniel, K. R. Fenton, K. R. Zavadil, T. Tyliszczak, W. Lai and K. F. McCarty, Nano Lett., 2013, 13, 866–872.
67. P. Bai and G. Y. Tian, Electrochim. Acta, 2013, 89, 644–651.
68. T. Ichitsubo, K. Tokuda, S. Yagi, M. Kawamori, T. Kawaguchi, T. Doi, M. Oishi and E. Matsubara, J Mater Chem A, 2013, 1, 2567–2577.
69. S. Dargaville and T. W. Farrell, Electrochim. Acta, 2013, 94, 143–158.
70. L. Wang, T. Maxisch and G. Ceder, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 73, 195107.
71. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
72. M. S. A. Rasiman, F. W. Badrudin, T. I. T. Kudin, M. K. Yaakob, M. F. M. Taib, M. Z. A. Yahya and O. H. Hassan, Integr. Ferroelectr., 2014, 155, 71–79.
73. S. Leoni, M. Baldoni, L. Craco, J. O. Joswig and G. Seifert, Z. Phys. Chem., 2012, 226, 95–106.
74. M. E. A. Y. de Dompablo, N. Biskup, J. M. Gallardo-Amores, E. Moran, H. Ehrenberg and U. Amador, Chem. Mater., 2010, 22, 994–1001.
75. S. Q. Yang, T. R. Zhang, Z. L. Tao and J. Chen, Acta Chim. Sin., 2013, 71, 1029–1034.
76. N. Marx, L. Croguennec, D. Carlier, A. Wattiaux, F. Le Cras, E. Suard and C. Delmas, Dalton Trans., 2010, 39, 5108–5116.
77. L. Gu, C. B. Zhu, H. Li, Y. Yu, C. L. Li, S. Tsukimoto, J. Maier and Y. Ikuhara, J. Am. Chem. Soc., 2011, 133, 4661–4663.
78. R. Malik, F. Zhou and G. Ceder, Nat. Mater., 2011, 10, 587–590.
79. G. Y. Chen, X. Y. Song and T. J. Richardson, J. Electrochem. Soc., 2007, 154, A627–A632.
80. P. Bai, D. A. Cogswell and M. Z. Bazant, Nano Lett., 2011, 11, 4890–4896.
81. J. L. Dodd, R. Yazami and B. Fultz, Electrochem. Solid-State Lett., 2006, 9, A151–A155.
82. C. Delacourt, J. Rodriguez-Carvajal, B. Schmitt, J. M. Tarascon and C. Masquelier, Solid State Sci., 2005, 7, 1506–1516.
83. C. Delacourt, P. Poizot, J. M. Tarascon and C. Masquelier, Nat. Mater., 2005, 4, 254–260.
84. A. Yamada, H. Koizumi, S. I. Nishimura, N. Sonoyama, R. Kanno, M. Yonemura, T. Nakamura and Y. Kobayashi, Nat. Mater., 2006, 5, 357–360.
85. P. Gibot, M. Casas-Cabanas, L. Laffont, S. Levasseur, P. Carlach, S. Hamelet, J. M. Tarascon and C. Masquelier, Nat. Mater., 2008, 7, 741–747.
86. N. Meethong, H. Y. S. Huang, W. C. Carter and Y. M. Chiang, Electrochem. Solid-State Lett., 2007, 10, A134–A138.
87. J. Maier, Nat. Mater., 2005, 4, 805–815.
88. X. L. Wu, L. Y. Jiang, F. F. Cao, Y. G. Guo and L. J. Wan, Adv. Mater., 2009, 21, 2710–2714.
89. P. G. Bruce, B. Scrosati and J. M. Tarascon, Angew. Chem., Int. Ed., 2008, 47, 2930–2946.
90. Y. Wang and G. Z. Cao, Adv. Mater., 2008, 20, 2251–2269.
91. F. Yu, J. J. Zhang, Y. F. Yang and G. Z. Song, J. Power Sources, 2010, 195, 6873–6878.
92. C. W. Sun, S. Rajasekhara, J. B. Goodenough and F. Zhou, J. Am. Chem. Soc., 2011, 133, 2132–2135.
93. G. X. Wang, L. Yang, S. L. Bewlay, Y. Chen, H. K. Liu and J. H. Ahn, J. Power Sources, 2005, 146, 521–524.
94. F. Yu, J. J. Zhang, Y. F. Yang and G. Z. Song, Electrochim. Acta, 2009, 54, 7389–7395.
95. Y. Ding, Y. Jiang, F. Xu, J. Yin, H. Ren, Q. Zhuo, Z. Long and P. Zhang, Electrochem. Commun., 2010, 12, 10–13.
96. B. Jin, E. M. Jin, K. H. Park and H. B. Gu, Electrochem. Commun., 2008, 10, 1537–1540.
97. Z. L. Liu, S. W. Tay, L. A. Hong and J. Y. Lee, J. Solid State Electrochem., 2011, 15, 205–209.
98. C. S. Li, S. Y. Zhang, F. Y. Cheng, W. Q. Ji and J. Chen, Nano Res., 2008, 1, 242–248.
99. Y. H. Rho, L. F. Nazar, L. Perry and D. Ryan, J. Electrochem. Soc., 2007, 154, A283–A289.
100. P. S. Herle, B. Ellis, N. Coombs and L. F. Nazar, Nat. Mater., 2004, 3, 147–152.
101. S. X. Zhao, H. Ding, Y. C. Wang, B. H. Li and C. W. Nan, J. Alloys Compd., 2013, 566, 206–211.
102. S. Hamelet, M. Casas-Cabanas, L. Dupont, C. Davoisne, J. M. Tarascon and C. Masquelier, Chem. Mater., 2011, 23, 32–38.
103. J. Molenda and J. Marzec, Funct. Mater. Lett., 2008, 1, 97–104.
104. C. A. J. Fisher and M. S. Islam, J. Mater. Chem., 2008, 18, 1209–1215.
105. S. Franger, F. Le Cras, C. Bourbon and H. Rouault, J. Power Sources, 2003, 119, 252–257.
106. B. Ellis, W. H. Kan, W. R. M. Makahnouk and L. F. Nazar, J. Mater. Chem., 2007, 17, 3248–3254.
107. L. Wang, F. Zhou, Y. S. Meng and G. Ceder, Phys. Rev. B: Condens. Matter Mater. Phys., 2007, 76, 165435.
108. Y. S. Meng and M. E. Arroyo-de Dompablo, Energy Environ. Sci., 2009, 2, 589–609.
109. C. Y. Nan, J. Lu, L. H. Li, L. L. Li, Q. Peng and Y. D. Li, Nano Res., 2013, 6, 469–477.
110. J. Lim, V. Mathew, K. Kim, J. Moon and J. Kim, J. Electrochem. Soc., 2011, 158, A736–A740.
111. H. F. Xiang, D. W. Zhang, Y. Jin, C. H. Chen, J. S. Wu and H. H. Wang, J. Mater. Sci., 2011, 46, 4906–4912.
112. S. L. Yang, X. F. Zhou, J. G. Zhang and Z. P. Liu, J. Mater. Chem., 2010, 20, 8086–8091.
113. K. Saravanan, M. V. Reddy, P. Balaya, H. Gong, B. V. R. Chowdari and J. J. Vittal, J. Mater. Chem., 2009, 19, 605–610.
114. K. Saravanan, P. Balaya, M. V. Reddy, B. V. R. Chowdari and J. J. Vittal, Energy Environ. Sci., 2010, 3, 457–464.
115. P. Balaya, K. Saravanan and S. Hariharan, in Energy Harvesting and Storage: Materials, Devices, and Applications, ed. N. K. Dhar, P. S. Wijewarnasuriya and A. K. Dutta, Spie-Int Soc Optical Engineering, Bellingham, 2010, vol. 7683.
116. Z. Y. Bi, X. D. Zhang, W. He, D. D. Min and W. S. Zhang, RSC Adv., 2013, 3, 19744–19751.
117. K. T. Lee and J. Cho, Nano Today, 2011, 6, 28–41.
118. C. M. Julien, A. Mauger and K. Zaghib, J. Mater. Chem., 2011, 21, 9955–9968.
119. W. J. Zhang, J. Electrochem. Soc., 2010, 157, A1040–A1046.
120. X. Zhao, B. M. Sanchez, P. J. Dobson and P. S. Grant, Nanoscale, 2011, 3, 839–855.
121. D. Zhang, R. Cai, Y. K. Zhou, Z. P. Shao, X. Z. Liao and Z. F. Ma, Electrochim. Acta, 2010, 55, 2653–2661.
122. Y. G. Wang, Y. R. Wang, E. J. Hosono, K. X. Wang and H. S. Zhou, Angew. Chem., Int. Ed., 2008, 47, 7461–7465.
123. S. Y. Lim, C. S. Yoon and J. P. Cho, Chem. Mater., 2008, 20, 4560–4564.
124. M. Konarova and I. Taniguchi, J. Power Sources, 2010, 195, 3661–3667.
125. J. J. Wang, J. L. Yang, Y. J. Tang, J. Liu, Y. Zhang, G. X. Liang, M. Gauthier, Y. C. K. Chen-Wiegart, M. N. Banis, X. F. Li, R. Y. Li, J. Wang, T. K. Sham and X. L. Sun, Nat. Commun., 2014, 5, 3415.
126. B. Kang and G. Ceder, Nature, 2009, 458, 190–193.
127. F. Yu, S. G. Ge, B. Li, G. Z. Sun, R. G. Mei and L. X. Zheng, Curr. Inorg. Chem., 2012, 2, 194–212.
128. F. Cheng, Z. Tao, J. Liang and J. Chen, Chem. Mater., 2008, 20, 667–681.
129. M. R. Hill, G. J. Wilson, L. Bourgeois and A. G. Pandolfo, Energy Environ. Sci., 2011, 4, 965–972.
130. L. Dimesso, C. Forster, W. Jaegermann, J. P. Khanderi, H. Tempel, A. Popp, J. Engstler, J. J. Schneider, A. Sarapulova, D. Mikhailova, L. A. Schmitt, S. Oswald and H. Ehrenberg, Chem. Soc. Rev., 2012, 41, 5068–5080.
131. X. D. Zhang, X. G. Zhang, W. He, C. Y. Sun, J. Y. Ma, J. L. Yuan and X. Y. Du, Colloids Surf., B, 2013, 103, 114–120.
132. Y. Ma, X. L. Li, S. F. Sun, X. P. Hao and Y. Z. Wu, Int. J. Electrochem. Sci., 2013, 8, 2842–2848.
133. J. F. Qian, M. Zhou, Y. L. Cao, X. P. Ai and H. X. Yang, J. Phys. Chem. C, 2010, 114, 3477–3482.
134. X. Qin, X. H. Wang, H. M. Xiang, J. Xie, J. J. Li and Y. C. Zhou, J. Phys. Chem. C, 2010, 114, 16806–16812.
135. D. G. Tong, Y. L. Li, W. Chu, P. Wu and F. L. Luo, Dalton Trans., 2011, 40, 4087–4094.
136. D. W. Han, W. H. Ryu, W. K. Kim, S. J. Lim, Y. I. Kim, J. Y. Eom and H. S. Kwon, ACS Appl. Mater. Interfaces, 2013, 5, 1342–1347.
137. A. Carné-Sánchez, I. Imaz, M. Cano-Sarabia and D. Maspoch, Nat. Chem., 2013, 5, 203–211.
138. F. Yu, J. J. Zhang, Y. F. Yang and G. Z. Song, J. Mater. Chem., 2009, 19, 9121–9125.
139. S. Yu, S. Kim, T. Y. Kim, J. H. Nam and W. I. Cho, J. Appl. Electrochem., 2013, 43, 253–262.
140. F. Yu, J. J. Zhang, Y. F. Yang and G. Z. Song, J. Power Sources, 2009, 189, 794–797.
141. J. Liu, T. E. Conry, X. Y. Song, M. M. Doeff and T. J. Richardson, Energy Environ. Sci., 2011, 4, 885–888.
142. B. F. Zou, Y. Q. Wang and S. M. Zhou, Mater. Lett., 2013, 92, 300–303.
143. W. Chang, S. J. Kim, I. T. Park, B. W. Cho, K. Y. Chung and H. C. Shin, J. Alloys Compd., 2013, 563, 249–253.
144. X. H. Rui, Y. Jin, X. Y. Feng, L. C. Zhang and C. H. Chen, J. Power Sources, 2011, 196, 2109–2114.
145. H. Q. Li and H. S. Zhou, Chem. Commun., 2012, 48, 1201–1217.
146. D. Zane, M. Carewska, S. Scaccia, F. Cardellini and P. P. Prosini, Electrochim. Acta, 2004, 49, 4259–4271.
147. S. Xin, Y. G. Guo and L. J. Wan, Acc. Chem. Res., 2012, 45, 1759–1769.
148. J. J. Wang and X. L. Sun, Energy Environ. Sci., 2012, 5, 5163–5185.
149. F. Y. Kang, J. Ma and B. H. Li, New Carbon Mater., 2011, 26, 161–170.
150. M. Inagaki, Carbon, 2012, 50, 3247–3266.
151. Y. Zhou, C. D. Gu, J. P. Zhou, L. J. Cheng, W. L. Liu, Y. Q. Qiao, X. L. Wang and J. P. Tu, Electrochim. Acta, 2011, 56, 5054–5059.
152. X. D. Yan, G. L. Yang, J. Liu, Y. C. Ge, H. M. Xie, X. M. Pan and R. S. Wang, Electrochim. Acta, 2009, 54, 5770–5774.
153. S. W. Oh, S. T. Myung, S. M. Oh, K. H. Oh, K. Amine, B. Scrosati and Y. K. Sun, Adv. Mater., 2010, 22, 4842–4845.
154. H. M. Xie, R. S. Wang, J. R. Ying, L. Y. Zhang, A. F. Jalbout, H. Y. Yu, G. L. Yang, X. M. Pan and Z. M. Su, Adv. Mater., 2006, 18, 2609–2613.
155. L. J. Zeng, Q. Gong, X. Z. Liao, L. He, Y. S. He and Z. F. Ma, Chin. Sci. Bull., 2011, 56, 1262–1266.
156. J. C. Zheng, X. H. Li, Z. X. Wang, J. H. Li, L. Wu, L. J. Li and H. J. Guo, Acta Phys.-Chim. Sin., 2009, 25, 1916–1920.
157. G. Yang, C. Y. Jiang, X. M. He, J. R. Ying and J. Gao, Ionics, 2013, 19, 1247–1253.
158. M. R. Yang, W. H. Ke and S. H. Wu, J. Power Sources, 2007, 165, 646–650.
159. L. Wu, J. J. Lu and S. K. Zhong, J. Solid State Electrochem., 2013, 17, 2235–2241.
160. J. C. Zheng, X. H. Li, Z. X. Wang, D. M. Qin, H. J. Guo and W. J. Peng, J. Inorg. Mater., 2009, 24, 143–146.
161. Z. P. Ma, G. J. Shao, X. Wang, J. J. Song and G. L. Wang, Ionics, 2013, 19, 1861–1866.
162. X. Z. Liao, Z. F. Ma, Q. Gong, Y. S. He, L. Pei and L. J. Zeng, Electrochem. Commun., 2008, 10, 691–694.
163. S. S. Zhang, K. Xu and T. R. Jow, J. Power Sources, 2006, 159, 702–707.
164. A. Yamada, S. C. Chung and K. Hinokuma, J. Electrochem. Soc., 2001, 148, A224–A229.
165. Z. H. Li, D. M. Zhang and F. X. Yang, J. Mater. Sci., 2009, 44, 2435–2443.
166. J. R. Ying, M. Lei, C. Y. Jiang, C. R. Wan, X. M. He, J. J. Li, L. Wang and J. G. Ren, J. Power Sources, 2006, 158, 543–549.
167. J. F. Ni, H. H. Zhou, J. T. Chen and X. X. Zhang, Mater. Lett., 2007, 61, 1260–1264.
168. X. M. Lou and Y. X. Zhang, J. Mater. Chem., 2011, 21, 4156–4160.
169. P. K. Shen, H. L. Zou, H. Meng and M. M. Wu, Funct. Mater. Lett., 2011, 4, 209–215.
170. Y. H. Yin, M. X. Gao, J. L. Ding, Y. F. Liu, L. K. Shen and H. G. Pan, J. Alloys Compd., 2011, 509, 10161–10166.
171. Y. H. Yin, M. X. Gao, H. G. Pan, L. K. Shen, X. Ye, Y. F. Liu, P. S. Fedkiw and X. W. Zhang, J. Power Sources, 2012, 199, 256–262.
172. Z. P. Ma, G. J. Shao, W. Wang and C. Y. Zhang, Asian J. Chem., 2013, 25, 3579–3582.
173. S. W. Oh, S. T. Myung, S. M. Oh, C. S. Yoon, K. Amine and Y. K. Sun, Electrochim. Acta, 2010, 55, 1193–1199.
174. Z. R. Chang, H. J. Lv, H. W. Tang, H. J. Li, X. Z. Yuan and H. J. Wang, Electrochim. Acta, 2009, 54, 4595–4599.
175. S. W. Oh, S. T. Myung, H. J. Bang, C. S. Yoon, K. Amine and Y. K. Sun, Electrochem. Solid-State Lett., 2009, 12, A181–A185.
176. J. K. Kim, CrystEngComm, 2014, 16, 2818–2822.
177. F. Yu, L. L. Zhang, M. Y. Zhu, Y. X. An, L. L. Xia, X. G. Wang and B. Dai, Nano Energy, 2014, 3, 64–79.
178. J. M. Tarascon, N. Recham, M. Armand, J. N. Chotard, P. Barpanda, W. Walker and L. Dupont, Chem. Mater., 2010, 22, 724–739.
179. O. Toprakci, H. A. K. Toprakci, L. W. Ji and X. W. Zhang, KONA Powder Part. J., 2010, 50–73.
180. S. M. Oh, H. G. Jung, C. S. Yoon, S. T. Myung, Z. H. Chen, K. Amine and Y. K. Sun, J. Power Sources, 2011, 196, 6924–6928.
181. J. Wang, Z. B. Shao and H. Q. Ru, Ceram. Int., 2014, 40, 6979–6985.
182. J. J. Tan and A. Tiwari, Nanosci. Nanotechnol. Lett., 2011, 3, 487–490.
183. Y. Y. Liu, D. W. Liu, Q. F. Zhang and G. Z. Cao, J. Mater. Chem., 2011, 21, 9969–9983.
184. D. Jugovic, M. Mitric, M. Kuzmanovic, N. Cvjeticanin, S. Skapin, B. Cekic, V. Ivanovski and D. Uskokovic, J. Power Sources, 2011, 196, 4613–4618.
185. C. Miao, J. Zhou, S. Q. Sun and X. Y. Wang, Rare Met. Mater. Eng., 2013, 42, 379–382.
186. Y. Zhang, Q. Y. Huo, P. P. Du, L. Z. Wang, A. Q. Zhang, Y. H. Song, Y. Lv and G. Y. Li, Synth. Met., 2012, 162, 1315–1326.
187. F. Yu, J. J. Zhang, Y. F. Yang and G. Z. Song, Chin. J. Inorg. Chem., 2009, 25, 42–46.
188. J. P. Jegal and K. B. Kim, J. Power Sources, 2013, 243, 859–864.
189. J. L. Yang, J. J. Wang, Y. J. Tang, D. N. Wang, X. F. Li, Y. H. Hu, R. Y. Li, G. X. Liang, T. K. Sham and X. L. Sun, Energy Environ. Sci., 2013, 6, 1521–1528.
190. P. P. Prosini, M. Carewska, S. Scaccia, P. Wisniewski and M. Pasquali, Electrochim. Acta, 2003, 48, 4205–4211.
191. P. Jozwiak, J. E. Garbarczyk, M. Wasiucionek, I. Gorzkowska, F. Gendron, A. Mauger and C. M. Julien, Solid State Ionics, 2008, 179, 46–50.
192. L. Yang, L. F. Jiao, Y. L. Miao and H. T. Yuan, J. Solid State Electrochem., 2009, 13, 1541–1544.
193. J. Liu, J. W. Wang, X. D. Yan, X. F. Zhang, G. L. Yang, A. F. Jalbout and R. S. Wang, Electrochim. Acta, 2009, 54, 5656–5659.
194. C. W. Kim, M. H. Lee, W. T. Jeong and K. S. Lee, J. Power Sources, 2005, 146, 534–538.
195. C. W. Kim, J. S. Park and K. S. Lee, J. Power Sources, 2006, 163, 144–150.
196. K. T. Lee and K. S. Lee, J. Power Sources, 2009, 189, 435–439.
197. Y. B. Xu, Y. J. Lu, L. Yan, Z. Y. Yang and R. D. Yang, J. Power Sources, 2006, 160, 570–576.
198. S. B. Lee, S. H. Cho, S. J. Cho, G. J. Park, S. H. Park and Y. S. Lee, Electrochem. Commun., 2008, 10, 1219–1221.
199. M. Hermanek, R. Zboril, M. Mashlan, L. Machala and O. Schneeweiss, J. Mater. Chem., 2006, 16, 1273–1280.
200. W. Ojczyk, J. Marzec, K. Swierczek, W. Zajac, M. Molenda, R. Dziembaj and J. Molenda, J. Power Sources, 2007, 173, 700–706.
201. H. Liu, J. Y. Xie and K. Wang, Solid State Ionics, 2008, 179, 1768–1771.
202. H. W. Liu and D. G. Tang, Solid State Ionics, 2008, 179, 1897–1901.
203. Y. Lin, J. B. Wu and Y. B. Yu, in Multi-Functional Materials and Structures Iii, Pts 1 and 2, Trans Tech Publications Ltd, Stafa-Zurich, 2010, vol. 123–125, pp. 1227–1230.
204. M. X. Gao, Y. Lin, Y. H. Yin, Y. F. Liu and H. G. Pan, Electrochim. Acta, 2010, 55, 8043–8050.
205. M. Konarova and I. Taniguchi, Mater. Res. Bull., 2008, 43, 3305–3317.
206. M. Konarova and I. Taniguchi, J. Power Sources, 2009, 194, 1029–1035.
207. Y. Lin, M. X. Gao, D. Zhu, Y. F. Liu and H. G. Pan, J. Power Sources, 2008, 184, 444–448.
208. G. Arnold, J. Garche, R. Hemmer, S. Strobele, C. Vogler and A. Wohlfahrt-Mehrens, J. Power Sources, 2003, 119, 247–251.
209. Y. Z. Dong, Y. M. Zhao, H. Duan, L. Chen, Z. F. He and Y. H. Chen, J. Solid State Electrochem., 2010, 14, 131–137.
210. S. Balaji, D. Mutharasu, N. Sankara Subramanian and K. Ramanathan, Ionics, 2009, 15, 765–777.
211. K. Cherian, Adv. Mater. Processes, 2011, 169, 23–27.
212. X. F. Guo, H. Zhan and Y. H. Zhou, Solid State Ionics, 2009, 180, 386–391.
213. M. Higuchi, K. Katayama, Y. Azuma, M. Yukawa and M. Suhara, J. Power Sources, 2003, 119, 258–261.
214. M. S. Song, Y. M. Kang, J. H. Kim, H. S. Kim, D. Y. Kim, H. S. Kwon and J. Y. Lee, J. Power Sources, 2007, 166, 260–265.
215. M. S. Song, Y. M. Kang, Y. I. Kim, K. S. Park and H. S. Kwon, Inorg. Chem., 2009, 48, 8271–8275.
216. K. S. Park, J. T. Son, H. T. Chung, S. J. Kim, C. H. Lee and H. G. Kim, Electrochem. Commun., 2003, 5, 839–842.
217. H. L. Zou, G. H. Zhang and P. K. Shen, Mater. Res. Bull., 2010, 45, 149–152.
218. L. Wang, Y. D. Huang, R. R. Jiang and D. Z. Jia, Electrochim. Acta, 2007, 52, 6778–6783.
219. L. Wang, Y. D. Huang, R. R. Jiang and D. Z. Jia, J. Electrochem. Soc., 2007, 154, A1015–A1019.
220. D. Li, Y. D. Huang, N. Sharma, Z. X. Chen, D. Z. Jia and Z. P. Guo, Phys. Chem. Chem. Phys., 2012, 14, 3634–3639.
221. D. Li, Y. D. Huang, D. Z. Jia, Z. P. Guo and S. J. Bao, J. Solid State Electrochem., 2010, 14, 889–895.
222. Y. Zhang, H. Feng, X. B. Wu, L. Z. Wang, A. Q. Zhang, T. C. Xia, H. C. Dong and M. H. Liu, Electrochim. Acta, 2009, 54, 3206–3210.
223. X. F. Guo, Y. T. Zhang, H. Zhan and Y. H. Zhou, Solid State Ionics, 2010, 181, 1757–1763.
224. S. Feng and L. Guanghua, in Modern Inorganic Synthetic Chemistry, ed. X. Ruren, P. Wenqin, W. P. Qisheng HuoA2-Ruren Xu and H. Qisheng, Elsevier, Amsterdam, 2011, pp. 63–95.
225. X. Q. Ou, S. Z. Xu, G. C. Liang, L. Wang and X. Zhao, Sci. China, Ser. E: Technol. Sci., 2009, 52, 264–268.
226. M. K. Devaraju and I. Honma, Adv. Energy Mater., 2012, 2, 284–297.
227. C. B. Xu, J. Lee and A. S. Teja, J. Supercrit. Fluids, 2008, 44, 92–97.
228. A. Aimable, D. Aymes, F. Bernard and F. Le Cras, Solid State Ionics, 2009, 180, 861–866.
229. N. Recham, L. Dupont, M. Courty, K. Djellab, D. Larcher, M. Armand and J. M. Tarascon, Chem. Mater., 2009, 21, 1096–1107.
230. N. Recham, J. N. Chotard, L. Dupont, K. Djellab, M. Armand and J. M. Tarascon, J. Electrochem. Soc., 2009, 156, A993–A999.
231. P. Barpanda, N. Recham, J. N. Chotard, K. Djellab, W. Walker, M. Armand and J. M. Tarascon, J. Mater. Chem., 2010, 20, 1659–1668.
232. M. C. Tucker, M. M. Doeff, T. J. Richardson, R. Finones, J. A. Reimer and E. J. Cairns, Electrochem. Solid-State Lett., 2002, 5, A95–A98.
233. K. Shiraishi, K. Dokko and K. Kanamura, J. Power Sources, 2005, 146, 555–558.
234. K. Dokko, S. Koizumi, K. Sharaishi and K. Kanamura, J. Power Sources, 2007, 165, 656–659.
235. K. Kanamura, S. H. Koizumi and K. R. Dokko, J. Mater. Sci., 2008, 43, 2138–2142.
236. K. Dokko, S. Koizumi and K. Kanamura, Chem. Lett., 2006, 35, 338–339.
237. S. F. Yang, P. Y. Zavalij and M. S. Whittingham, Electrochem. Commun., 2001, 3, 505–508.
238. M. Koltypin, D. Aurbach, L. Nazar and B. Ellis, J. Power Sources, 2007, 174, 1241–1250.
239. S. Franger, F. Le Cras, C. Bourbon and H. Rouault, Electrochem. Solid-State Lett., 2002, 5, A231–A233.
240. J. L. Liu, R. R. Jiang, X. Y. Wang, T. Huang and A. S. Yu, J. Power Sources, 2009, 194, 536–540.
241. J. Chen, S. Wang and M. S. Whittingham, J. Power Sources, 2007, 174, 442–448.
242. J. J. Chen and M. S. Whittingham, Electrochem. Commun., 2006, 8, 855–858.
243. X. Q. Ou, H. C. Gu, Y. C. Wu, J. W. Lu and Y. J. Zheng, Electrochim. Acta, 2013, 96, 230–236.
244. R. G. Mei, X. R. Song, Y. F. Yang, Z. G. An and J. J. Zhang, RSC Adv., 2014, 4, 5746–5752.
245. J. Lee and A. S. Teja, J. Supercrit. Fluids, 2005, 35, 83–90.
246. K. Dokko, S. Koizumi, H. Nakano and K. Kanamura, J. Mater. Chem., 2007, 17, 4803–4810.
247. J. J. Chen, M. J. Vacchio, S. J. Wang, N. Chernova, P. Y. Zavalij and M. S. Whittingham, Solid State Ionics, 2008, 178, 1676–1693.
248. Y. Xia, W. K. Zhang, H. Huang, Y. P. Gan, J. Tian and X. Y. Tao, J. Power Sources, 2011, 196, 5651–5658.
249. J. F. Ni, M. Morishita, Y. Kawabe, M. Watada, N. Takeichi and T. Sakai, J. Power Sources, 2010, 195, 2877–2882.
250. Z. G. Lu, H. L. Chen, R. Robert, B. Y. X. Zhu, J. Q. Deng, L. J. Wu, C. Y. Chung and C. P. Grey, Chem. Mater., 2011, 23, 2848–2859.
251. G. Meligrana, C. Gerbaldi, A. Tuel, S. Bodoardo and N. Penazzi, J. Power Sources, 2006, 160, 516–522.
252. C. Gerbaldi, J. Nair, C. B. Minella, G. Meligrana, G. Mulas, S. Bodoardo, R. Bongiovanni and N. Penazzi, J. Appl. Electrochem., 2008, 38, 985–992.
253. S. Bodoardo, C. Gerbaldi, G. Meligrana, A. Tuel, S. Enzo and N. Penazzi, Ionics, 2009, 15, 19–26.
254. G. X. Wang, X. P. Shen and J. Yao, J. Power Sources, 2009, 189, 543–546.
255. Y. M. Jiang, S. J. Liao, Z. S. Liu, G. Xiao, Q. B. Liu and H. Y. Song, J. Mater. Chem. A, 2013, 1, 4546–4551.
256. S. Tajimi, Y. Ikeda, K. Uematsu, K. Toda and M. Sato, Solid State Ionics, 2004, 175, 287–290.
257. Z. H. Xu, L. Xu, Q. Y. Lai and X. Y. Ji, Mater. Res. Bull., 2007, 42, 883–891.
258. F. Teng, M. D. Chen, G. Q. Li, Y. Teng, T. G. Xu, S. I. Mho and X. Hua, J. Power Sources, 2012, 202, 384–388.
259. G. Yang, H. M. Ji, H. D. Liu, K. F. Huo, J. J. Fu and P. K. Chu, J. Nanosci. Nanotechnol., 2010, 10, 980–986.
260. G. Yang, H. M. Ji, X. W. Miao, A. Q. Hong and Y. Yan, J. Nanosci. Nanotechnol., 2011, 11, 4781–4792.
261. D. Carriazo, M. D. Rossell, G. B. Zeng, I. Bilecka, R. Erni and M. Niederberger, Small, 2012, 8, 2231–2238.
262. A. V. Murugan, T. Muraliganth and A. Manthiram, J. Phys. Chem. C, 2008, 112, 14665–14671.
263. A. V. Murugan, T. Muraliganth and A. Manthiram, J. Electrochem. Soc., 2009, 156, A79–A83.
264. T. Muraliganth, A. V. Murugan and A. Manthiram, J. Mater. Chem., 2008, 18, 5661–5668.
265. A. V. Murugan, T. Muraliganth, P. J. Ferreira and A. Manthiram, Inorg. Chem., 2009, 48, 946–952.
266. A. V. Murugan, T. Muraliganth and A. Manthiram, Electrochem. Commun., 2008, 10, 903–906.

---