Low-cost polyanion cathodes for sodium ion batteries: challenges, strategies, and progress

Abstract

As one of the most promising battery technologies for commercial applications after lithium-ion batteries, sodium-ion batteries have attracted extensive attention from researchers due to their abundant resources, high safety and wide operating temperature range. They have shown strong competitiveness in large-scale energy storage and some price-sensitive applications. However, the related technical research is not very mature. Among them, as a key component in determining the battery's performance, the choice of cathode materials is crucial. In this regard, the polyanionic cathode material has the characteristics of high working potential and good safety and becomes a promising candidate. However, the low electronic conductivity hinders its further development. Therefore, this review will focus on the modification measures of polyanionic cathode materials, including element doping, surface modification, mophology control, crystal plane design and phase engineering, pay attention to their improvement effect on the performance of active materials, and discuss the modification mechanism and the latest development trend. At the same time, the challenges in the commercialization of polyanionic cathode SIBs are pointed out. This review aims to provide potential modification ideas for researchers and promote the development of polyanion cathode materials for sodium ion batteries.

---

1. Introduction

In 2020, renewable energy (mainly hydropower, wind and solar) and nuclear energy will account for 12.6% and 4.3% of total energy consumption in the global energy mix. The remaining majority (83%) still comes from the burning of non-renewable fossil fuels, which also brings well-known problems,^1,2 namely, the depletion of non-renewable energy reserves and environmental issues.^3–5 To solve the global climate problem and ensure the normal energy supply of human beings, it is necessary to reduce the use of fossil fuels and increase the proportion of renewable energy.^6–9 However, the production of renewable energy sources such as wind energy and solar energy is difficult to predict and has the characteristics of strong intermittence, which leads to the instability of the power supply of the power grid and makes it difficult to be directly used by human society.^10–14 Therefore, the development of energy storage systems (ESSs) to regulate the power supply of the power grid is particularly important for the large-scale utilization of renewable energy technologies.^15–17 Studies have pointed out that by 2024, the global energy storage device market will increase to 1095 GW/2850 GWh, an increase of more than 120 times compared with the 9 GW/17 GWh deployed in 2018.¹⁸ Therefore, in the context of the global pursuit of energy conservation and emission reduction, the demand for ESSs will rise sharply, and it is imperative to develop a suitable ESS.^19–22 Among the various energy storage schemes, rechargeable batteries are one of the most competitive schemes due to their long cycle life and easy maintenance.^23–29 It is worth noting that the fixed batteries used in ESSs usually work under harsh environmental conditions (such as large temperature differences between day and night or high humidity). Therefore, the equipment parameters that need to be considered are significantly different from the battery parameters in common electronic products. First of all, the cost of battery production must be low enough, which is the fundamental premise of large-scale application. Secondly, in terms of performance, it needs to have sufficient energy density and be able to charge and discharge quickly and continuously at high rates to achieve rapid response to the power demand; thirdly, it is necessary to have a large enough working temperature range to adapt to the complex and changeable working environment. Finally, when it reaches the end of life, it can be recycled in a green way to avoid environmental pollution.³⁰

At present, lithium-ion batteries (LIBs), sodium-ion batteries (SIBs), lead–acid batteries (LABs), redox flow batteries (RFBs) and supercapacitors (SCs) are expected to be used as large-scale energy storage devices. Fig. 1d compares the characteristics of the above several types of energy storage devices. It can be seen that SIBs are not the best in each performance index, but they are the most suitable in the performance framework required by ESSs. In terms of cost, abundant and easily available sodium, iron, and manganese resources and economical battery components make SIBs have lower production costs.^31–34 In terms of performance, SIBs can work in a very wide temperature range (−40–80 °C) and can withstand complex and variable working conditions. At the same time, the open framework of NASICON-type materials provides sufficient space for Na⁺ diffusion and supports the high-rate charge–discharge behavior of SIBs.^35–37 In terms of recycling, since Al does not react with Na, Al can be used instead of Cu as the negative current collector of SIBs, which simplifies the battery components and is easier to recycle. Moreover, the recovery of SIBs does not involve the complex separation process of rare materials, which not only helps to improve the recovery efficiency but also reduces the use of chemicals and the negative impact on the environment.^30,38–43 These advantages make SIBs not only show strong competitiveness in the field of large-scale energy storage applications, but also show broad application prospects in the fields of low-speed electric vehicles, communication base stations and home energy storage. In the past decade, many companies have also launched corresponding commercial SIBs. The UK's Faradion, the US's Natron Energy, France's Tiamat, and China's HiNa and CATL have all launched the first-generation SIB technology, setting off a global wave. Among them, CATL announced that it is expected to launch a second-generation sodium-ion battery that can discharge normally in a cold environment of −40 °C in 2025.^44–49 It can be seen that after LIBs, SIBs may become the next large-scale application battery technology.⁵⁰

Fig. 1 (a) Working principle of sodium ion batteries. Reproduced with permission from ref. 1 Copyright 2017, Elsevier. (b) Earth crust element abundance. (c) Application scenarios of SIBs. (d) Evaluation of SIBs, LIBs, LABs, SCs, and RFBs.

Cathode materials are an important part of the battery, directly affecting the specific capacity, cycle life and operating voltage. A variety of cathode materials have been systematically explored. They are mainly layered transition metal oxides, polyanionic compounds and Prussian blue analogues. Among them, polyanionic compounds have a more stable institutional framework, higher operating voltage and better safety. Therefore, they are considered to be the most promising cathode materials for SIBs. However, some obstacles lie in their application, namely, low specific capacity and low intrinsic electronic conductivity. Therefore, it is necessary to modify the design to improve the performance. This review analyzes the structure of polyanionic compounds and their performance advantages. At the same time, it also highlights the inherent defects brought by its structure. Therefore, the material is modified by element doping, surface modification, morphology adjustment, crystal plane design and phase engineering to solve the main problems of low intrinsic electronic conductivity and low specific capacity. We pay attention to the mechanism of these modification measures, and also pay attention to the development trend of these modification measures. Through the discussion in this review, we hope to give researchers a basic understanding of the current research progress of polyanionic cathode materials, to help the development of SIBs.

2. Choice of cathode material and challenges

As a key component of the battery module, the choice of cathode material is crucial.^48,51–54 Currently, there are primarily three types of sodium ion cathode materials: layered transition metal oxides (TMOs), Prussian blue analogues (PBAs), and polyanionic compounds.^55–58 Among them, TMOs have high theoretical capacity, and the layered structure provides an open two-dimensional ion diffusion channel, which is conducive to the rapid deintercalation of sodium ions and has good rate performance. However, TMOs will undergo an irreversible phase transition during charge and discharge. After multiple cycles, the phase transition will inevitably lead to the accumulation of mechanical stress and strain inside the material, resulting in lattice distortion and crystal cracking. In addition, poor air stability makes it difficult to manufacture and transport the electrodes composed of TMOs, which hinders their practical application.^59–62 For PBAs, the three-dimensional open framework structure is conducive to the rapid transport of Na⁺. However, due to its aqueous synthesis environment, there is a certain amount of crystal water in the structure of PBAs. The presence of crystal water not only reduces the specific capacity of the material but also hinders the transmission of Na⁺ and affects the electrochemical performance of the material. More critically, the crystal water escaping from the material will have side reactions with the organic electrolyte and even produce gas (H₂O, NH₃, and N₂), which seriously endangers the safety performance of the battery.^63–65 In contrast, the polyanionic compound cathode material has a stable framework structure, good safety, and long life; the main raw materials are cheap and environmentally friendly such as sodium, iron and manganese, and the raw material cost is low. Therefore, it is expected to be applied in the field of large-scale energy storage. Polyanionic compounds Na_xM_y(XO₄)_n (X = S, P, Si, B, As, W; M = Fe, Mn, V and other transition metals) are composed of an anionic tetrahedron (XO₄) and transition metal octahedron (MO₆). According to the different anionic groups, it can be divided into phosphate, pyrophosphate, mixed phosphate, fluoride transition metal salt, sulfate, silicate and borate. Due to the presence of high electronegativity F⁻, fluorine transition metal salts exhibit higher redox potential under the action of the “induction effect”.⁶⁶ Similarly, the high electronegativity of S also brings a higher redox potential to sulfate.⁶⁷ In pyrophosphates, the higher stability of the P₂O₇⁴⁻ group leads to better structural stability, and the larger cell parameters make pyrophosphates have a higher Na⁺ diffusion rate.⁶⁸ In silicate, the strong Si–O bond formed between Si and O not only provides better thermal stability, but also reduces the volume change in the material during charging and discharging.⁶⁹ Different anionic groups provide different characteristics. However, in general, due to the presence of anionic groups, polyanionic compounds still have significantly different characteristics from the other two materials, as follows:

(1) High redox potential: as the most prominent feature of polyanionic compounds, it is generally believed that their high redox potential can be attributed to the “induction effect”. As shown in Fig. 2a, the covalent interaction between the transition metal M and the oxygen atom O splits the molecular orbital into an anti-bonding orbital (near vacuum) and a bonding orbital (away from vacuum). When the covalency between M–O is enhanced, the energy level difference between the two orbitals becomes larger, so that the energy level difference (Δ) between the antibonding orbital and the vacuum becomes smaller, showing a lower redox potential. In polyanionic compounds, the introduction of a strong electronegative atom X forms an X–O–M bond, and the “induction effect” plays a role. The strong X–O bond introduces ionicity in the M–O bond, thereby reducing the covalency between M–O, making Δ larger, which also leads to a higher redox potential.⁷⁰ In contrast, in layered transition metal oxides (such as NaCoO₂), the induction effect of O²⁻ is weaker, the anti-bonding orbital is higher, the Δ is smaller, and the redox potential is relatively lower.

Fig. 2 (a) Schematic diagram showing the influence of the M–O covalency on the orbital energy level. Δ is the energy needed to drive one electron to the vacuum state. Δ is proportional to the voltage. (b) Comparison of average voltage and energy density of different polyanionic cathodes. (c) Performance comparison of the three cathode materials for SIBs.

(2) High security: the anionic tetrahedron (XO₄) and transition metal octahedron (MO₆) in polyanionic compounds form a stable and open framework, which is conducive to the rapid transmission of Na⁺. Also, the strong X–O covalent bond between the strong electronegative atom X and the oxygen atom greatly improves the stability of the oxygen atom in the lattice, reduces the volume change and complex phase transition reaction caused by the repeated insertion/extraction of Na⁺, and ensures the structural stability of polyanionic compounds under various environmental conditions, thus showing significantly better safety than layered transition metal oxides and Prussian blue analogues.⁷¹

(3) Low intrinsic electronic conductivity and low specific capacity: the inactive macromolecular anionic groups inevitably reduce the tap density and specific capacity of the material. Moreover, because the MO₆ octahedra are isolated from each other and separated by the XO₄ group, the –M–O–M– electron delocalization cannot be carried out directly, and the electrons involved in the electrochemical reaction are difficult to transmit quickly in the electrode material, resulting in the low electronic conductivity of the material, which greatly limits its rate performance.⁷²

3. Element doping

As mentioned in Section 2, the main disadvantage of polyanionic compounds is their low electronic conductivity. Element doping can adjust the crystal structure of the material and improve the performance from the inside of the material, such as improving the intrinsic electronic conductivity and stabilizing the crystal structure. Different doping elements have different effects on the improvement of material properties. Through the purposeful selection of doped elements, the crystal structure parameters of the bulk material are finely adjusted to improve its electrochemical performance.⁷³ At present, the doping modification of polyanion compounds includes the incorporation of heteroatoms at Na sites, transition metal sites and polyanion sites.

3.1. Na sites

For the Na sites, the current mainstream method is to select the K element with a larger ion radius as the doping element. On the one hand, K⁺ can expand the unit cell volume, reduce the diffusion energy barrier of Na⁺, and increase the diffusion rate of Na⁺. On the other hand, K⁺ can be used as a pillar ion to stabilize the structure of the material during the charge and discharge process and improve the structural stability.^74–76

As shown in Fig. 3a, Shen et al.⁷⁷ prepared the Na_2.95K_0.05V₂(PO₄)₃(NVP-K_0.05)material by the sol–gel method (Fig. 3a). The introduction of K⁺ expands the cell volume and broadens the migration channel of Na⁺, which not only improves the diffusion rate of Na⁺ but also significantly enhances the structural stability of the material. Consequently, the NVP-K_0.05 material exhibits excellent electrochemical performance at low temperature (−25 °C), room temperature (25 °C) and high temperature (40 °C), and ensures the safety of the battery. It is worth noting that NVP-K_0.05 still has a reversible capacity of 72 mAh g⁻¹ at −25 °C and 2C, while that of the undoped Na₃V₂(PO₄)₃ (NVP) is almost 0. However, it should be noted that a higher K⁺ doping level is not always better. Excessive K⁺ doping will reduce the electronic and ionic conductivity of the material and reduce the sodium storage performance of the material. Therefore, exploring the appropriate heteroatom doping amount can effectively improve the material performance.⁷⁸ In addition to K⁺, some researchers have also chosen to dope Li⁺ at the sodium sites to improve the electrochemical performance of the material.^79,80 Zheng et al.⁸¹ studied the mechanism of Li⁺ doping in NVP and synthesized a series of Na_3−xLi_xV₂(PO₄)₃/C (x = 0, 0.01, 0.05, 0.1, 0.5, 0.7 and 1.0) compounds. The results show that when the x value is low, Li⁺ tends to mainly enter the Na2 site due to the low substitution energy of Li⁺ doping on the Na2 site. When the x value is high (x > 0.1), Li⁺ will occupy both Na1 and Na2 sites due to the repulsion between similar ions. It is worth noting that although the introduced Li⁺ does not exhibit electrochemical activity, due to the small exchange energy difference between Li⁺ and Na⁺, rearrangement occurs between the two, which activates Na⁺ at the Na1 site to participate in the electrochemical cycle, resulting in additional specific capacity. By the same strategy, Cong et al.⁸² prepared Li⁺ doped Na_3−xLi_xV₂(PO₄)₃/C to improve the electronic conductivity of NVP (Fig. 3bI). The results show that an appropriate amount of Li⁺ (x = 0.2) doping reduces the band gap of the material (Fig. 3bII–IV), reduces the electron localization phenomenon, and also activates Na⁺ at the Na1 position, which not only significantly improves the electronic conductivity and ionic conductivity of the material, but also leads to high capacity and high stability. The Na_2.8Li_0.2V₂(PO₄)₃/C cathode exhibits a reversible capacity of 116.9 mAh g⁻¹, almost reaching the theoretical capacity (117 mAh g⁻¹). Besides doping K⁺ and Li⁺, Lu³⁺ doped Na₃V₂(PO₄)₃ (NVP/C@Lu-1%) has been prepared recently.⁸³ The introduced high-valent multi-charged Lu³⁺ occupies the Na1 site, which enhances the structural stability and also leads to slight lattice distortion, thereby inducing more active sites to store Na⁺. At the same time, the successful doping of Lu³⁺ forms a stable Lu–O bond between the crystal planes, which plays the role of anchor pillars between the crystal planes, thereby reducing the lattice strain during the Na⁺ deintercalation process and making the crystal structure more stable. Finally, the NVP/C@Lu-1% cathode exhibits excellent cycle stability (Fig. 3c).

Fig. 3 (a) Schematic illustration of the synthetic processes of NVP-K_0.05 samples; graphical representations of NVP and NVP-K_0.05. Reproduced with permission from ref. 77 Copyright 2023, Elsevier. (b) (I) Schematic illustration of the synthetic processes of Na_3−xLi_xV₂(PO₄)₃/C. Optimized crystal structure and the density of states of (II) NVP and Na_2.8Li_0.2V₂(PO₄)₃ with Li⁺ occupying (III) Na1 site and (IV) Na2 site. Reproduced with permission from ref. 82 Copyright 2024, Elsevier. (c) The best discharge ability of sample NVP/C@Lu-1% at 1C for different cycles; NVP/C@Lu-1% cycled at 30C for 7500 cycles. Reproduced with permission from ref. 83 Copyright 2024, Elsevier.

3.2. Anionic sites

Anionic groups constitute the main component of the polyanionic crystal structure. As a framework in the crystal structure, anion site doping has a significant effect on improving sodium ion transport kinetics and structural stability.^84–86 For example, in order to solve the problem of low conductivity of iron-based mixed phosphate Na₄Fe₃(PO₄)₂P₂O₇ (NFPP), Liu et al.⁸⁷ prepared Na_4.05Fe₃(PO₄)_1.95P₂O₇(SiO₄)_0.05 (NFPP-Si_0.05) by replacing PO₄³⁻ with SiO₄⁴⁻ (Fig. 4a). Since the Si–O bond is longer than the corresponding P–O bond at the same position, the introduction of a larger SiO₄⁴⁻ group is beneficial for the rapid transport of Na⁺. Moreover, the high stability of the Si–O tetrahedron strengthens the crystal structure and reduces the volume change during the charge and discharge process, which improves the structural stability of the material. Finally, NFPP-Si_0.05 still has a capacity retention rate of 92.1% (initial capacity of 80 mAh g⁻¹) after 3000 cycles at 20C. In addition, the use of SiO₄⁴⁻ substituted PO₄³⁻ groups can also effectively solve the problems of capacity loss and cycle stability reduction caused by Mn²⁺ occupying Na⁺ vacancies in Na₃MnTi(PO₄)₃. Lu et al.⁸⁸ prepared the Na₃MnTi(PO₄)_2.7(SiO₄)_0.3 (NMTP-Si_0.3) compound. Due to the additional negative charge brought by SiO₄⁴⁻, Mn²⁺ is partially oxidized to Mn³⁺, which has a low preference for Na⁺ vacancies, thus effectively inhibiting the occupation of Na⁺ vacancies by Mn²⁺, and the voltage hysteresis is significantly suppressed (Fig. 4b), thereby improving the long-term cycle stability of the material.

Fig. 4 (a) The method by which SiO₄⁴⁻ is doped into NFPP. Reproduced with permission from ref. 87 Copyright 2023, American Chemical Society. (b) Illustration of the voltage gap between charge and discharge profiles for NMTP and NMTP-Si_0.3. Reproduced with permission from ref. 88 Copyright 2024, American Chemical Society. (c) Structural illustrations of Na₄MnCr(PO₄)₃ and F doped Na₄MnCr(PO₄)₃; total and partial DOS patterns of NMCPF and NMCP. Reproduced with permission from ref. 93 Copyright 2023, Chemical Science. (d) Schematic for the crystal structure of NFSF; schematic diagram of the CEI for NFSF and NFS. Reproduced with permission from ref. 94 Copyright 2025, American Chemical Society. (e) The crystal structure and the corresponding Na⁺-diffusion paths of (I) NFS and (II) NFSP-0.3. (III) The diffusion energy barrier of different Na⁺ migration paths in in NFS and NFSP-0.3. Reproduced with permission from ref. 95 Copyright 2024, Elsevier. (f) (I) Schematic representation of the structural changes in the Na₃VP_2−xS_xOF cathode during the charging and discharging processes. (II) Conductivity test for VP and VPS-1. (III) VPS-1 in ex situ XRD spectra in 5C–1C mode. Reproduced with permission from ref. 96 Copyright 2025, Elsevier.

In addition to the incorporation of the SiO₄⁴⁻ group, halogen doping can also significantly improve the material properties.^89–91 Shao et al.⁹² studied the effect of halogen doping (F, Cl, and Br) on the properties of phosphate Na₂VTi(PO₄)₃ (NVTP). Halogen doping destroys the hexagonal symmetry of the Na sites, enhances the interaction between anions (O and halogen elements) and the Na sites, and improves the transport speed of Na⁺. In addition, halogen doping narrows the band gap of NVTP, which facilitates the electron transition to the conduction band and improves the electronic conductivity. Therefore, NVTP exhibits significantly improved rate performance. Among the halogen elements, F has been widely used in doping modification of materials due to its strong bonding strength, excellent dielectric properties, high electronegativity and low relative atomic mass. For example, Zhang et al.⁹³ doped F (NMCPF) into manganese-rich phosphate Na₄MnCr(PO₄)₃ (NMCP) to improve its poor structural stability (Fig. 4c). The introduction of F not only significantly reduces the material's band gap (Fig. 4c) but also improves the electronic conductivity. In addition, the overall chemical bonding of Mn–O/F bonds is enhanced locally, triggering the “Mn-locking” effect in NMCPF, inhibiting the dissolution of Mn²⁺, and significantly enhancing the structural stability of the material. By the same doping strategy, Zhang et al.⁹⁴ prepared Na_2.2Fe_1.75(SO_3.9F_0.1)₃ (NFSF) (Fig. 4d) by doping F into iron-based sulfate. The charge redistribution caused by F doping makes iron tend to exist in the form of Fe²⁺, which avoids the capacity reduction caused by partial oxidation of iron. And after F substitution, the migration barrier of sodium ions in different paths is reduced, which promotes ion transport. More importantly, F reduces the electron density of the crystal plane, thereby reducing the oxidative decomposition of the solvent and the intense nucleophilic reactions at the interface at high voltage, promoting the formation of a stable and uniform Chemical–Electrochemical Interface (CEI), and further accelerating the Na⁺ transfer at the cathode interface, which can be seen intuitively in Fig. 4d. As mentioned in Section 2, different anionic groups have different characteristics. SO₄²⁻ has a higher redox potential, but poor air stability. PO₄³⁻ is the opposite. Considering that S and P are adjacent in the periodic table of elements and have similar properties, the ionic radii of PO₄³⁻ (238 pm) and SO₄²⁻ (258 pm) are also similar. Therefore, the researchers chose to combine the two and learn from each other. Liu et al.⁹⁵ introduced PO₄³⁻ into Na_2.6Fe_1.7(SO₄)₃ (NFS) to improve water sensitivity, and prepared Na_2.9Fe_1.7(SO₄)_2.7(PO₄)_0.3 (NFSP-0.3). The introduction of PO₄³⁻ not only reduces the band gap of the material (3.44 eV to 2.41 eV), but also reduces the diffusion energy barrier of Na⁺ (Fig. 4eIII), thus effectively improving the electronic and ionic conductivity of the material. In addition, PO₄³⁻ doping also effectively weakened the adsorption capacity of NFSP-0.3 and H₂O. After exposure to air with a humidity of about 50% for one week, the structure of the NFSP-0.3 electrode did not change significantly, showing better air stability. Wang et al.⁹⁶ partially replaced PO₄³⁻ with SO₄²⁻ in Na₃V₂(PO₄)₂O₂F (VP) to prepare Na₃V₂(PO₄)_1.95(SO₄)_0.05O₂F (VPS-1) (Fig. 4fI). The introduction of SO₄²⁻ can reduce the band gap of VP, and reduce the migration energy barrier. Compared with VP, VPS-1 exhibits reduced polarization at different current densities, the average diffusion coefficient is about twice that of VP, and the overall internal resistance is lower (Fig. 4fII). As a result, the rate performance (80C, 75.5 mAh g⁻¹) and cycle stability (111.0 mAh g⁻¹ after 1000 cycles at 10C) of VPS-1 are improved. As shown in Fig. 4fIII, all peaks returned to the original position after the reaction, confirming the high stability and high reversibility of VPS-1 during the fast charging process.

3.3. Transition metal sites

Transition metal site doping has been widely reported as an effective strategy for customizing the structure of polyanionic compounds. Various elements have unique ionic radius, valence state, solid solubility and other characteristics. By incorporating different elements, the electrochemical properties of the matrix material can be improved, such as conductivity and structural stability.^97,98 For example, in Na₄Mn_1.5Fe_1.5(PO₄)₂P₂O₇ doped with Nb, by introducing a high charge density Nb–O bond, the length of the Mn–O bond is shortened and the lattice stability is improved. The lattice distortion caused by the Jahn–Teller effect is reduced, and the structural stability of the material is significantly improved.⁹⁹ In addition, researchers often use Al doping to improve the structural stability and electrical conductivity of materials.^100–102 Gao et al.¹⁰³ prepared the microporous Na_3.9Fe_2.9Al_0.1(PO₄)₂(P₂O₇) (NFAPP) cathode material (Fig. 5a) by doping Al³⁺ into Na₄Fe₃(PO₄)₂P₂O₇. The introduction of Al³⁺ reduces the band gap of the material (2.05 eV to 1.63 eV) and broadens the diffusion channel of Na⁺ (Fig. 5a), thus giving NFAPP a faster electron transport rate and Na⁺ mobility. Moreover, the strong Al–O–P bond caused by Al³⁺ doping alleviates the volume vibration during (de)sodiation, enabling highly reversible valence variation and structural evolution. Consequently, the NFAPP cathode exhibits excellent cycle stability and ultrafast rate capability.

Fig. 5 (a) Crystal structures of NFPP and optimized NFAPP; the DOS of NFPP and NFAPP. Reproduced with permission from ref. 103 Copyright 2024, Elsevier. (b) Schematic illustration of the crystal structure of the NFPF-0.07Zr sample along the a-axis; initial charge/discharge curves for NFPF and NFPF-0.07Zr samples; the discharge curves and corresponding platform capacity distribution at different rates. Reproduced with permission from ref. 104 Copyright 2023, Elsevier. (c) (I) The migration energy barrier from Na1 → Na2 → Na1 → Na2 sites of NFPF and NFMPF. (II) The corresponding schematic crystal structure from different directions for NFMPF. Reproduced with permission from ref. 105 Copyright 2024, Elsevier. (d) Schematic illustration of the structural change in the NVP-Mn(0.5) electrode on cycling; DOS and corresponding PDOS diagram of NVP-Mn(0) and NVP-Mn(0.5). Reproduced with permission from ref. 112 Copyright 2023, Wiley-VCH.

Since Na₂FePO₄F (NFPF) has two different Na sites, denoted as Na1 and Na2, respectively, Na2 has electrochemical activity and determines the capacity of the material. The Na1 site affected by the semi-unstable oxygen atom is inert, resulting in a lower specific capacity of NFPF. In order to improve the specific capacity of the material, He et al.¹⁰⁴ incorporated Zr into the NFPF material (NFPF-0.07Zr) (Fig. 5b). The introduction of Zr increases the average bond length of the Na2 site, and the average bond length of the Na1 site decreases accordingly (Fig. 5b). The elongated Na2 exhibits more electrochemical activity, while the shortened Na1 stabilizes the structural framework, and the latter provides expanded Na⁺ migration channels, favoring diffusion kinetics. In addition, the introduction of Zr improves the electronic conductivity. Based on the double improvement of electronic and ionic conductivity, as shown in Fig. 5b, the first voltage platform of NFPF-0.07Zr is significantly extended, contributing more capacity. In addition to improving the utilization of the Na2 site, the Na1 site can also be activated to increase the specific capacity of the material. Liu et al.¹⁰⁵ proposed an effective d⁰ orbital Mg doping method to improve the electrochemical performance of NFPF. Different from the 3d orbital of Fe, the Mg in the d⁰ orbital mainly contributes to the p and s orbitals in the Mg–O bond, which leads to a change in the electronic state of the O atom and reduces the barrier from Na1 to Na2, thus opening the channel for the transfer of inert Na1 to active Na2 (Fig. 5cI and II) and activating the Na1 site. Therefore, Na₂Fe_0.93Mg_0.07PO₄F (NFMPF) exhibits a higher reversible capacity (121.4 mAh g⁻¹vs. 108.7 mAh g⁻¹). In addition, the introduction of Cu weakens the effect of uncoordinated O2 around Na1, thereby reducing the deintercalation potential of Na sites, which can also lead to the activation of an inert Na1 site.¹⁰⁶

In addition, elements with low cost advantages, such as Mn and Fe, are often used to replace elements of transition metal sites.^107–110 Zhou et al.¹¹¹ prepared Na_3.5V_1.5Fe_0.5(PO₄)₃ (NVFP). Fe substitution not only reduces the cost, but also activates the V⁴⁺/V⁵⁺ redox pairs, achieving the reversible multi-electron redox reactions of Fe²⁺/Fe³⁺, V³⁺/V⁴⁺ and V⁴⁺/V⁵⁺ redox pairs. Finally, the NVFP cathode delivers a discharge capacity of 148.2 mAh g⁻¹ at 0.5C. More importantly, in terms of application feasibility, the HC//NVFP full battery exhibits excellent cycle stability (63.5% capacity retention after 3000 cycles at 50C) and a material-level energy density of 304 Wh kg⁻¹, showing excellent performance. Chen et al.¹¹² prepared Na₃V_1.5Mn_0.5(PO₄)₃ (NVP-Mn(0.5)) (Fig. 5d) by the Mn doping strategy. After adding Mn³⁺, the band gap of the material is significantly reduced, which greatly improves the electronic conductivity. In addition, Mn doping also activates the V⁴⁺/V⁵⁺ redox pair, thereby achieving continuous conversion of four stable oxidation states (V²⁺/V³⁺, V³⁺/V⁴⁺ and V⁴⁺/V⁵⁺) and forming a reversible three-electron reaction (Fig. 5d). Finally, NVP-Mn(0.5) delivers a high reversible discharge capacity of 170.9 mAh g⁻¹ at 0.5C with an ultra-high energy density of 577 Wh kg⁻¹.

The strong Ti–O bond formed between Ti and O can enhance the ionic covalent properties and effectively stabilize the crystal structure of the material. Zhu et al.¹¹³ designed Ti-doped Na_2.5V_1.5Ti_0.5(PO₄)₃/C (NVTP-0.5) hierarchical microspheres. The introduction of Ti not only improves the irreversible phase transformation and structural degradation of the matrix material but also enhances the structural stability of the material. It also helps to reduce the band gap (Fig. 6a) and improve the electronic conductivity. More importantly, the continuous redox pairs of V²⁺/V³⁺, Ti³⁺/Ti⁴⁺ and V³⁺/V⁴⁺ (Fig. 6a) endow the NVTP-0.5 cathode with an ultra-high discharge specific capacity of 192.42 mAh g⁻¹ (3.2 electron reaction) and a high energy density of 497.3 Wh kg⁻¹, showing excellent electrochemical performance. Beyond doping transition metal elements, some researchers have also used alkali metal element doping to improve electrochemical performance. Shen et al.¹¹⁴ explored the substitution of alkali metal elements (Li, Na, and K) at the V position in NVP. As shown in Fig. 6b, the doping of alkali metal elements makes V³⁺ partially converted to V⁴⁺, resulting in a small polaron, which significantly reduces the band gap, thus effectively enhancing the intrinsic electronic conductivity of the material. Among them, K-doped Na₃V_1.94K_0.06(PO₄)₃ (NVP-K_0.06) exhibits the best performance. The discharge capacities of NVP-K_0.06 at 55 °C, 30 °C and −30 °C are 121.3, 113.2 and 45.8 mAh g⁻¹, respectively, while those of NVP are 82.9, 76.2 and 9.2 mAh g⁻¹, respectively, which is mainly due to the improvement of crystal structure stability. Fig. 6b also confirms that NVP-K_0.06 undergoes a highly reversible phase transition during charge and discharge, and the volume change is only 0.82% after one cycle, which contributes to the excellent cycle performance of NVP-K_0.06 at various operating temperatures.

Fig. 6 (a) Density of states of NVTP-0 and NVTP-0.5; schematic structural change of NVTP-0.5 during charge/discharge. Reproduced with permission from ref. 113 Copyright 2024, American Chemical Society. (b) In situ XRD patterns at a current rate of 0.1C and corresponding charge/discharge curves; calculated density of states of NVP, NVP-Li, NVP-Na and NVP-K cathodes. Reproduced with permission from ref. 114 Copyright 2023, Elsevier. (c) The DOS of NMTP and HE-NMTP. Reproduced with permission from ref. 125 Copyright 2024, Elsevier. (d) (I) Schematic illustration of the crystal structure of HE-NASICON. (II) Galvanostatic discharge profiles at a current density of 0.1C of HE-NASICON. SEM cross-sectional images after 500 cycles: (III) HE-NASICON and (IV) Na₄MnCr(PO₄)₃. Reproduced with permission from ref. 126 Copyright 2024, Elsevier. (e) In situ XRD 2D contour plot and corresponding charge–discharge curves of (I) NVP and (II) HE-NVP-0.2 electrodes in the voltage range of 2.5–4.3 V; (III) the DOS of NVP and HE-NVP-0.2 materials based on the DFT calculations. Reproduced with permission from ref. 127 Copyright 2024, Elsevier.

The previous discussions are all based on single ion doping, but considering that each element has its unique properties, in fact, researchers often use a variety of element co-doping methods to comprehensively improve the electrochemical performance of the material. Wang et al.¹¹⁵ prepared Co/Zr co-doped Na₃V_1.86Co_0.07Zr_0.07(PO₄)₃/C composites with the p–n-type doping effect. Co²⁺ replaces V³⁺ to introduce p-type doping and generate holes. The introduction of the same amount of Zr⁴⁺ can introduce n-type doping and balance the charge. The combination of the two effectively improves the electronic conductivity of the material. Previous studies have shown that the increase in entropy can stabilize the crystal structure and significantly improve the performance of electrode materials. Therefore, some medium-entropy¹¹⁶ and high-entropy materials have emerged in recent years.^117–122 According to thermodynamic theory, a mixture with a high configuration entropy greater than 1.61R (where R is the gas constant) can be called a high-entropy material. There are many crystal phases in high-entropy materials, and the multiphase provides additional sites for ion insertion/extraction or redox processes, which makes high-entropy materials exhibit higher capacity and stability.¹²³ In recent years, there have been more and more studies on high-entropy polyanionic compound cathodes. Among them, Gu et al.¹²⁴ first introduced the concept of high entropy into polyanion compounds and successfully prepared a carbon-free Na₃V_1.9(CaMgAlCrMn)_0.1(PO₄)₂F₃ (HE-NVPF) high-entropy phosphate cathode. Compared with the original Na₃V₂(PO₄)₂F₃, the favorable high entropy effect in HE-NVPF completely suppresses the bad phase transition behavior at the low voltage platform (=3.4 V), significantly improves the average operating voltage (from 3.67 V to 3.81 V), and also improves the energy density (from 395.78 to 445.42 Wh kg⁻¹). In addition, the introduced high-entropy doping changes the electronic states near the Fermi level, and the conductivity of HE-NVPF is significantly enhanced even without any carbon doping. Recently, Zhang et al.¹²⁵ also prepared a high-entropy Na_3.12MnTi_0.9(VFeMgCrZr)_0.02(PO₄)₃ (HE-NMTP) cathode by the high-entropy strategy. By introducing a high-entropy doping strategy, HE-NMTP not only reduces the band gap of the material (Fig. 6c), but also promotes the excitation and transfer of electrons. It also eliminates the significant trap states in the NMTP band gap, thereby reducing the electron transport loss. After high entropy doping, the oxygen ligand framework is enhanced, thereby improving the overall structural stability and preventing the dissolution of Mn²⁺ in long-term cycling. Finally, the HE-NMTP cathode exhibits a high reversible capacity of 169.6 mAh g⁻¹ at 0.1C (1C = 176 mA g⁻¹) and a high energy density of more than 500 Wh kg⁻¹. Similarly, Zhu et al.¹²⁶ selected five elements (Cr, Fe, Mn, V, and Al) to incorporate into the material to construct a high-entropy Na₄Cr_0.7Fe_0.4Mn_0.3V_0.3Al_0.2(PO₄)₃ material (HE-NASICON) (Fig. 6dI). Among them, Cr, Fe, Mn and V are mainly used for charge compensation, which helps to improve the material capacity, while Al is used to stabilize the crystal structure and suppress the Jahn–Teller effect of Mn³⁺. Consequently, HE-NASICON exhibits a high reversible capacity of 165.0 mAh g⁻¹ at 0.1C (Fig. 6dII). It is worth noting that the cell volume change of HE-NASICON during the deintercalation of Na⁺ is only 1.45%, which effectively alleviates the crystal structure cracking and degradation during the long-term cycling. It can also be seen from the SEM image and the structural evolution diagram (Fig. 6dIII and IV) that many obvious cracks appeared in the Na₄MnCr(PO₄)₃ electrode sheet after 200 cycles. In contrast, the HE-NASICON electrode sheet did not change significantly, showing better cycle stability. Zhou et al.¹²⁷ introduced high-entropy elements (Cr, Mn, Fe, Zn, and Al) into NVP to prepare a carbon-free Na₃V_1.8(CrMnFeZnAl)_0.2(PO₄)₃ cathode (HE-NVP-0.2). In addition to showing a significant improvement in electrical conductivity, it is more noteworthy that due to the reconstruction of the crystal structure, HE-NVP-0.2 generates a Na₂V₂(PO₄)₃ intermediate phase during the charge–discharge process, achieving a three-phase reaction (Fig. 6e). Different from the two-phase reaction of NVP, the existence of the intermediate phase effectively alleviates the lattice mismatch between Na₃V₂(PO₄)₃ and Na₁V₂(PO₄)₃, thereby promoting rapid phase transition and providing better rate performance and cycle stability.

In general, for the doping of transition metal sites, researchers have explored dozens of doping elements, each of which has its unique properties. Among them, Al, Fe, Mn, and Cr are the most common. The lower price makes them improve the material properties without increasing the cost of raw materials, which is more in line with the needs of large-scale applications and has attracted more attention.

3.4. Multiple sites

The above discussions are all about doping heteroatoms at the same site, but in recent years, the strategy of doping heteroatoms at two sites has been adopted by more and more researchers.^128,129 As shown in Fig. 7a, Li et al.¹³⁰ prepared a Cr and Si co-doped Na₃V_1.9Cr_0.1(PO₄)_2.9(SiO₄)_0.1 material (NVP-CS) by the solid-state method. Among them, the introduced Cr³⁺ provides additional electrons, activates the V⁴⁺/V⁵⁺ redox couple, and obtains a higher operating voltage. In addition, the inert Cr³⁺ plays a supporting role, which improves the structural stability of the material. The introduction of Si changes the electronic structure of V and O, thereby promoting the redox ability of V³⁺/V⁴⁺ and exerting a higher capacity. Under the synergistic effect of these two elements, a lower migration barrier (2.19 eV to 2.08 eV) was obtained in NVP-CS (Fig. 7a), showing a significant improvement in electrochemical performance. Similarly, for the doping of transition metal sites and anion sites, Zhao et al.¹³¹ introduced Al and F (NVMP-Al&F) into Na₄VMn(PO₄)₃, and the “double-pinning effect” effectively inhibited the Jahn–Teller effect. It can be seen from Fig. 7b that the cell volume change of NVMP-Al&F is only 4.43% during the charge and discharge process, which is less than 7.06% that of Na₄VMn(PO₄)₃ (NVMP), showing better structural stability. In addition, it can be seen from Fig. 7b that the Na⁺ diffusion energy barrier of NVMP-Al&F is lower than that of NVMP (0.43 eV vs.1.02 eV), which enhances the Na⁺ diffusion kinetics. Finally, NVMP-Al&F achieved excellent structural stability with a capacity retention of 86.1% after 1000 cycles at 5C.

Fig. 7 (a) Schematic illustration of the synthesis and action mechanism of NVP containing heteroatom composition; local density of states plot of NVP and NVP-CS. Reproduced with permission from ref. 130 Copyright 2023, American Chemical Society. (b) Crystal structure of NVMP-Al&F; Na⁺ diffusion energy barriers of NVMP and NVMP-Al&F. Reproduced with permission from ref. 131 Copyright 2024, Elsevier. (c) Cycling curves of the NVP/C@KMn0.01 sample at 15C, 60C, and 90C. Reproduced with permission from ref. 135 Copyright 2024, Elsevier. (d) Schematic illustration of the cycling processes of NVP and KSi-NVP cathode particles. Reproduced with permission from ref. 132 Copyright 2023, Elsevier.

Beyond the simultaneous doping of transition metal sites and anion sites, researchers have also chosen to simultaneously regulate alkali metal sites and anion sites. Dou et al.¹³² prepared Na_3.24K_0.10V_2.01(PO₄)_2.94(SiO₄)_0.14 (KSi-NVP). The introduction of K⁺ as a supporting ion at the Na1 site alleviates the shrinkage/expansion of the cell during the cycle, avoids the grain microcracks and particle rupture caused by the accumulation of lattice strain during the repeated cycle, and improves the cycle stability of the material (Fig. 7d). At the same time, the substitution of SiO₄⁴⁻ for PO₄³⁻ with the same structure induces additional Na⁺ incorporation and increases the capacity. Consequently, the electrochemical performance of KSi-NVP is improved. Some researchers also chose to dope alkali metal sites and transition metal sites.^133,134 Wang et al.¹³⁵ synthesized K⁺/Mn²⁺ co-substituted Na_2.97K_0.04V_1.99Mn_0.01(PO₄)₃/C@CNTs composites (NVP/C@KMn0.01). Among them, K⁺ mainly occupies Na⁺ at the Na1 site, which acts as a supporting pillar without reducing capacity. Mn²⁺ occupies the V³⁺ site and further stabilizes the crystal structure of NVP. In addition, the divalent Mn²⁺ occupies the trivalent V³⁺ site to produce a p-type doping effect, introducing more hole carriers, and thereby increasing the intrinsic conductivity of the material. Finally, the rate performance and cycle performance of NVP/C@KMn0.01 have been significantly improved (Fig. 7c).

In addition to co-doping two sites, researchers have doped three sites at the same time. Sun et al.¹³⁶ doped K/La/Si elements at three sites in NVP. Compared with Na, V and P, the radius of K, La and Si ions is larger, which expands the migration channel of Na⁺ and reduces the resistance in the adjacent coordination environment, thus improving the stability of the crystal framework. At the same time, the ternary substitution makes the lattice slightly distorted, and increases the active sites that can be used for reversible deintercalation of Na⁺, which is beneficial to supplement the discharge capacity during the structural phase transition. Thus, the electrochemical performance of NVP is improved.

4. Surface modification

Surface modification involves the formation of a coating layer on the surface of the active material through physical or chemical processes, or to combine with other materials to purposefully improve the electrochemical performance of the active material. As mentioned above, the polyanionic compound cathode material has a lower intrinsic electronic conductivity due to the presence of larger anionic groups. Carbon-based materials are widely used in the surface modification of polyanionic compounds due to their high electronic conductivity. The introduction of carbon-based materials can not only form an external conductive framework on the surface of the active material but also improve the electronic conductivity. It can also isolate the active material from the electrolyte and reduce the occurrence of side reactions. In addition, carbon-based materials can also inhibit the particle agglomeration and expansion of active materials during the synthesis process, reduce the particle size, and increase the ion diffusion coefficient. Therefore, carbon-based materials are currently used for surface modification.

By adding carbon sources such as glucose, citric acid, and polyvinylpyrrolidone (PVP) to the precursor solution/slurry of the active material, and then carbonizing the carbon source during the heat treatment process, a thin in situ carbon coating can be formed on the surface of the active material to achieve the purpose of improving the electrochemical performance of the material.^137–139 Gu et al.¹⁴⁰ selected agarose (AG) as a carbon source and coated a uniform carbon coating on Na₃V₂(PO₄)₂F₃ (NVPF) nanoparticles. As shown in Fig. 8aI, the carbon coating not only enhances the electronic conductivity of the NVPF material but also inhibits the growth of NVPF particles during high-temperature calcination, thereby shortening the migration path of Na⁺. It is worth noting that the P–O and V–F bonds are strengthened and the V–O bond is weakened after carbon coating, as shown in Fig. 8aII. This means that the ionicity of the NVPF lattice is improved due to the enhancement of the F-induction effect during the carbon coating process, thereby effectively increasing the operating voltage of the phosphate cathode material (from 3.59 V to 3.71 V). This provides a new method for the development of other high-performance materials. Similarly, in order to improve the conductivity of NVPF, Zhang et al.¹⁴¹ proposed a low-cost and scalable amylopectin-assisted synthesis of in situ carbon-coated NVPF nanosheets (NVPFC-NS), as shown in Fig. 8b. Amylopectin can not only induce NVPF to nucleate along its skeleton to form two-dimensional nanostructures, but also serve as a source of amorphous carbon for the formation of in situ carbon coatings on the surface of active materials. The unique two-dimensional nanostructure and in situ carbon coating effectively improve the ionic and electronic conductivity of NVPFC-NS.

Fig. 8 (a) (I) HRTEM image (inset: FFT image) of Na₃V₂(PO₄)₂F₃@C. (II) XPS of NVPF@C and p-NVPF: high-resolution XPS spectra and the corresponding deconvolution results for the orbital peaks of P 2p, O 1s, F 1s, and V 2p. Reproduced with permission from ref. 140 Copyright 2020, Elsevier. (b) Schematic illustration of the preparation of NVPFC-NS. Reproduced with permission from ref. 141 Copyright 2022, American Chemical Society. (c) The formation mechanism diagram of cathode materials of NMVP@C–N. Reproduced with permission from ref. 146 Copyright 2022, Elsevier. (d) SEM images of (I) P-NC15@NVP and (II) U-NC15@NVP samples. (III) Schematic diagram of the diffusion path of Na⁺/e⁻ in NVP/NC and the transport process of the N-doped carbon layer. Reproduced with permission from ref. 147 Copyright 2024, Elsevier. (e) Flow diagram of the synthesis of the Na₃V₂(PO₄)₃/C samples; schematic diagram of ClO₄⁻ as a carrier in the charging and discharging process. Reproduced with permission from ref. 150 Copyright 2023, Elsevier. (f) Simulation analysis of charge density of N/Cl modified carbon; the changing trend of the crystal plane. Reproduced with permission from ref. 151 Copyright 2025, Elsevier.

Although the carbon coating has effectively improved the electronic conductivity of the material, based on the pursuit of high-performance cathodes, researchers often choose to dope non-metallic elements (N, S, B, P, Cl, etc.) in the carbon coating to adjust the electronic structure of the carbon coating and introduce additional electrochemical active sites and external defects, thereby improving the potential and electronic conductivity of the cathode material to store Na⁺ during charge and discharge.^142–145 At present, N-doped carbon coating is one of the most effective and widely used modification strategies. As shown in Fig. 8c, Wang et al.¹⁴⁶ coated a layer of N-doped carbon on the surface of Na₄MnV(PO₄)₃ by the sol–gel method. The introduction of N produces more defects and active sites in the carbon layer, thereby improving the electronic conductivity and Na⁺ diffusion rate of the material. However, it should be noted that different nitrogen sources and different nitrogen contents have different effects on the properties of the materials. Zhu et al.¹⁴⁷ studied the effects of different nitrogen sources on the electrochemical properties of materials. They used urea and polyvinylpyrrolidone (PVP) as N sources to synthesize N-doped carbon-coated NVP composites U-NC15@NVP and P-NC15@NVP, respectively. Compared with PVP, urea is pyrolyzed at high temperature to produce volatiles, and N is doped into the carbon layer while making the material loose and porous (Fig. 8dI and II). As a result, many external defects and active sites are generated, which not only help to reduce the energy bandwidth of the carbon layer, but also improve the conductivity of the carbon layer. It also effectively promoted the storage, diffusion and adsorption of Na⁺ (Fig. 8dIII). Zhang et al.¹⁴⁸ studied the effect of carbon coatings with different N contents on the properties of materials. Appropriate N content indeed improves the degree of graphitization of carbon, enhances the electronic conductivity, and significantly reduces the electrode polarization. However, a high content of N doping will increase the disorder of carbon and reduce the electronic conductivity. In addition, excessive N leads to an increase in the number of pyridinic N defects and intercepts Na⁺, thereby inhibiting the absorption of Na⁺ by the graphite structure. Beyond using a single element doped carbon coating, some researchers have also selected a double-doped carbon coating to produce more active sites. Kang et al.¹⁴⁹ synthesized Na₃V₂(PO₄)₃ (NVP@SNC) with S–N co-doped carbon coating by the sol–gel method combined with heat treatment. During the preparation process, thiourea was used as an S and N source, and acted as a surfactant, reducing the agglomeration of the composite material and increasing the porosity, thereby promoting the penetration and transport of the electrolyte. In addition, S–N co-doping induces a large number of lattice defects in the carbon coating, which not only effectively reduces the Na⁺ diffusion energy barrier but also provides sufficient Na⁺ migration channels, significantly improving the rate performance of NVP@SNC.

Different from the above two, Li et al.¹⁵⁰ successfully synthesized Na₃V₂(PO₄)₃/C composites in situ by a simple sol–gel method using chitosan quaternary ammonium salt hydrogel (CHACC) as the substrate, as shown in Fig. 8e. The hydrogel substrate can not only disperse NVP particles in the synthesis process and effectively prevent agglomeration and improve crystallinity, but also form N–Cl co-doped carbon coating after carbonization, which significantly improves the performance of the carbon coating. On the one hand, nitrogen doping is beneficial to improve the disorder degree of the carbon layer. On the other hand, chlorine doping can accelerate the electron transport in the carbon layer, and some chlorine is also converted into active ClO⁴⁻, acting as additional charge carriers, increasing the electron transport speed and providing some additional capacity (Fig. 8e). Finally, the electrochemical performance of NVP was significantly improved under the synergistic effect of N and Cl. Similarly, Dong et al.¹⁵¹ prepared N–Cl co-doped carbon-coated NVP composites (CZP-2). It is worth noting that after the N/Cl atoms in the CZP-2 carbon coating occupy the C site, the charge accumulates at the N/Cl site (Fig. 8f). The enriched charge is relatively attractive for an O atom, which induces the formation of O vacancies inside NVP. And the offset of the O atom lengthens the P–O and V–O bonds, contributing to the expansion of the VO₆ and PO₄ framework and thus enhancing the stability of the crystal structure and providing the foundational conditions for the multivalent reaction system between V^3+/4+/5+, inducing the in situ extraction of Na⁺ under high-voltage platforms. Therefore, in addition to the improvement of conductivity, CZP-2 also has good cyclic reversibility and near-zero strain performance (Fig. 8f). This work discovered the regulation of the internal crystal structure of the material by using the carbon coating, which provided a new idea for later research work.

In addition, the combination of active materials with carbon materials such as carbon nanotubes (CNTs) and redox graphene (rGO) can also achieve the purpose of improving material properties.^152–155 Zhang et al.¹⁵⁶ prepared NVPF@3%CNT composites. The introduction of CNTs enhances the electronic conductivity of the material and avoids partial electrochemical insulation inside the NVPF particles, leading to better rate performance. In addition, the TOF-SIMS mapping images of the original NVPF (p-NVPF) and NVPF@3%CNT under Cs⁺ sputtering and their corresponding counting curves (Fig. 9aI) show that the CEI film on the surface of NVPF@3%CNT is thinner and denser, indicating that the introduction of CNTs enhances the compactness of the CEI and effectively inhibits the external degradation of NVPF. The TEM image in Fig. 9aII and III also proves this conclusion. Like CNTs, rGO is also often used to combine with active materials to improve performance. Sun et al.¹⁵⁷ prepared a small amount (0.5 wt%) of rGO added Na₃V₂(PO₄)₂F₂O/rGO composites by the in situ solid phase synthesis. The preparation process is shown in Fig. 9b. Due to the high-energy ball milling and high-temperature treatment during the preparation process, the amount of sp³ carbon and defective carbon in the carbon layer increased (Fig. 9b), thus providing more active sites for Na⁺ diffusion during charging/discharging. It should be noted that too little rGO content cannot effectively enhance the conductivity of the composite material, nor can it provide sufficient electrochemically active sites. However, too much rGO may be recombined through the interaction of van der Waals forces during the preparation of composites, which may reduce the conductivity of composites to a certain extent. Therefore, moderate content of rGO can effectively improve the material properties.

Fig. 9 (a) (I) TOF-SIMS mapping images of NaF₂⁻ for p-NVPF and NVPF@3%CNT cathodes. TEM images of the (II) p-NVPF and (III) NVPF@3%CNT cathodes after 500 cycles at 5C. Reproduced with permission from ref. 156 Copyright 2023, American Chemical Society. (b) Schematic illustration of the facile in situ preparation of NVPFO/rGO composites; schematic illustration of the hybrid structure between NVPFO and rGO in the composites. Reproduced with permission from ref. 157 Copyright 2023, Elsevier. (c) Illustration of the encapsulation mechanism of the three types of carbonaceous precursors. Reproduced with permission from ref. 161 Copyright 2024, American Chemical Society. (d) The carbon layer model of different doping elements and its composite model with an NVP cathode. The schematic diagram of Na⁺ interlayer migration and overall migration of NVP/NSFC and its corresponding Na⁺ migration energy barrier. Reproduced with permission from ref. 162 Copyright 2024, Elsevier.

In addition to the above carbon-based materials, carbon dots (CDs), a zero-dimensional carbon-based material, have been increasingly used by researchers to improve the performance of battery materials in recent years.^158,159 CDs are characterized by a size of less than 10 nm and consist of a carbon core encapsulated by a surface passivation layer. And due to their disposable carbon core and shell composition as well as the diversity of functional groups, CDs have unique configurations and tunable functions that are not found in other carbon-based materials, giving them a unique advantage in enhancing material properties.¹⁶⁰ Li et al.¹⁶¹ selected three types of carbonaceous precursors (Super P/carbon dots/citric acid) to prepare NVP/SP, NVP/CDs, and NVP/CA composites (Fig. 9c) to compare the effects of different carbonaceous precursors on cathode performance. Among them, the CD carbonaceous precursor has an appropriate amount of O/H-containing functional groups. On the one hand, it effectively increases the bonding sites with the positive electrode and provides a reduction advantage in the carbothermal reduction process. On the other hand, it avoids the uneven coating caused by a large amount of gas produced during the sintering process due to excessive functional groups. Therefore, compared with the island carbon coating of NVP/SP and the non-uniform carbon coating of NVP/CA, the uniform and dense carbon layer covered on the surface of NVP/CDs not only provides a physical barrier for the active material, but also promotes the graphitization of the carbon coating due to the moderate H/O atom ratio in the coating, thus effectively improving the material performance. Inspired by the high-entropy strategy, Li et al.¹⁶² selected CDs as a precursor to construct a carbon layer with multiple heteroatoms (H, O, N, S, and F)/defects, that is, a defect configuration entropy (DCE) carbon layer, and prepared a defect configuration entropy enhanced Na₃V₂(PO₄)₃/carbon composite (NVP/NSFC). The carbon layer model is shown in Fig. 9d. The presence of a heterogeneous interface affected by the defect configuration entropy enhances the stability of the NVP cathode phase and interface, greatly reduces the Na⁺ diffusion energy barrier (Fig. 9d), and enhances the electronic conductivity, laying a foundation for excellent low-temperature performance. In addition, the CEI film of NVP/NSFC is rich in NaF, which effectively promotes the diffusion of Na⁺, enhances the interfacial mechanical integrity within the CEI, and reduces electrolyte decomposition, providing support for high temperature stability. Consequently, NVP/NSFC exhibits excellent electrochemical performance in the temperature range of −30–60 °C. After 300 stable cycles at −20 °C (1C), the capacity remains 69.6%, and after 1500 cycles at 60 °C (10C), the capacity remains 92.9%.

In order to further improve the electrochemical performance of electrode materials, in recent years, researchers have often adopted a variety of carbon coating synergistic modification strategies to combine the characteristics of different carbon materials to achieve the purpose of comprehensively improving the performance of materials. Xu et al.¹⁶³ constructed a 3D rGO shell-supported NVPF nanocuboid (NVPF@C@rGO) by a two-step method with the assistance of PVA. Among them, PVA plays a vital role in the preparation process. It regulates the interaction between the active NVPF particles and the rGO layer through the hydrogen bond between its hydroxyl group and rGO, resulting in a 3D framework composed of an NVPF nanocuboid and conformal coating rGO layer (Fig. 10aI–IV). With this unique structure, NVPF@C@rGO exhibits excellent sodium storage performance and cycle stability. In order to improve the poor cycle stability caused by the low conductivity of Na₄MnCr(PO₄)₃ and the Jahn–Teller effect of Mn³⁺, Chen et al.¹⁶⁴ selected a multi-carbon layer strategy to prepare NMCP@C@PVP@CNTs composites (Fig. 10bII–IV). In this strategy, the excellent conductivity of CNTs can effectively improve the capacity and rate performance of the material. PVP not only improves the conductivity of the material but also helps to improve the cycle stability of the material. On the one hand, the incomplete combustion of PVP forms a carbon coating with abundant defects. On the other hand, PVP, as a non-ionic surfactant, forms micelles and disperses particles in aqueous solution, which effectively regulates the shape and size of particles and alleviates the lattice stress caused by Jahn–Teller distortion of Mn³⁺ during charging and discharging. Finally, NMCP@C@PVP@CNTs exhibited a higher sodium diffusion coefficient (D_Na⁺ about 2.3 × 10⁻¹² cm² s⁻¹), conductivity (R_p = 103 Ω) and cycle stability (Fig. 10bI).

Fig. 10 (a) (I and II) SEM and (III and IV) TEM images of NVPF@C@rGO. Reproduced with permission from ref. 163 Copyright 2024, Elsevier. (b) (I) Long-term cycle performance of NMCP series samples at a high rate of 5C. (II) SEM, (III)TEM, and (IV) HR-TEM images of NMCP@C@PVP@CNTs. Reproduced with permission from ref. 164 Copyright 2024, Elsevier. (c) Schematic illustrations of the electrode/electrolyte interface mechanism. Reproduced with permission from ref. 168 Copyright 2023, Elsevier.

Beyond coating carbon-based materials on the surface of active materials, researchers have also begun to try to improve the electrochemical performance of materials through non-carbon material coatings in recent years.^165–167 The structural degradation of Na₄MnV(PO₄)₃ during cycling was studied. Wang et al.¹⁶⁸ prepared Na₄MnV(PO₄)₃/C@Al(PO₃)₃ composites by the wet process (Fig. 10c). On the one hand, the coating is composed of thermally stable Al(PO₃)₃ and NaPO₃ with high ionic conductivity formed during sintering, which improves the stability of the material and provides high ionic conductivity. On the other hand, a small amount of Al³⁺ doping during the sintering process effectively suppresses the Jahn–Teller effect of Mn³⁺ and stabilizes the material structure. Finally, Na₄MnV(PO₄)₃/C@Al(PO₃)₃ still has a stable specific capacity of 76.2 mAh g⁻¹ after 4000 cycles at a high rate of 10C in the 2.5–3.8 V voltage window. Facing the same problem, Song et al.¹⁶⁹ proposed another strategy to deposit NASICON-type solid electrolyte Na_1.3Al_0.7Ti_1.3(PO₄)₃ (NATP) on the surface of the Na₄MnV(PO₄)₃ (NMVP) material. As an interface compatible coating of NMVP, the NATP coating acts as a protective layer. More importantly, the gradient doping of Ti⁴⁺ and Al³⁺ from NATP improves the lattice matching between NMVP and NATP, and enhances the interface and structural stability. It also extends the lattice spacing of NMVP and suppresses the Jahn–Teller effect. The perfect lattice matching of NVMP and NATP improves the interface and structural stability during long-term cycling.

5. Mophology control

As mentioned above, polyanionic compounds have a problem of low conductivity due to the presence of anionic groups. By adjusting the morphology of the material particles, especially nanocrystallization, the problem of low conductivity can be effectively solved. The reduction of the particle size increases the contact area between the material and the electrolyte, promotes the electrolyte penetration and shortens the Na⁺ diffusion distance, which effectively improves the electronic conductivity and ionic conductivity of the material. In addition, the reduction of particle volume can effectively alleviate the volume deformation during charging and discharging, and further improve the cycle stability of the material.^170–172

Xu et al.¹⁷³ prepared nano-sized and micron-sized carbon-coated Na₃V₂(PO₄)₂O₂F (NVPOF), and compared the performance differences between the two. In contrast, nano-NVPOF has a smaller particle size (Fig. 11b), which facilitates the conduction of charge through the carbon network, thus ensuring efficient redox reactions during charging/discharging. Therefore, nano-NVPOF exhibits higher specific capacity and rate performance. In the voltage window of 3–4.5 V, the specific capacity of nano-NVOPF at 1C is more than 70 mAh g⁻¹, while that of micro-NVOPF is only about 20 mAh g⁻¹. As shown in Fig. 11aI, Zheng et al.¹⁷⁴ adopted a microfluidic-based continuous flow strategy to prepare nano-sized Na₃V₂O₂(PO₄)₂F (NVOPF) materials with different shapes (spherical, flake or rod) by adjusting the pH value (Fig. 11aII–IV). This method can not only prepare cathode materials at low cost, fast and on a large scale, but the prepared cathode materials also have excellent electrochemical performance. In the absence of other modification strategies, the nanospherical NVOPF can still maintain a discharge specific capacity of 114.0 mAh g⁻¹ at 10C. The nanosheet NVOPF exhibits excellent cycle stability (97.3% capacity retention after 1000 cycles at 10C). It should be noted that active materials with low-dimensional structures, such as nanospheres and nanosheets, also have some disadvantages, including low tap density, being more likely to lead to irregular aggregation and more serious adverse surface side reactions, which lead to low-volume energy density and poor cycle stability.^175–178 In order to avoid the above problems, Cao et al.¹⁷⁹ prepared three-dimensional Na₃V₂(PO₄)₃/C microspheres (NVP/C-MSs) assembled by using interconnected nanosheets by the hydrothermal method. As shown in Fig. 11c, the average diameter of the microspheres is ∼10 μm, and the microspheres are composed of nanosheets with a thickness of about 20–30 nm. This structure can not only improve the shorter electron/ion path and larger electrode electrolyte contact area, but compared with nanomaterials, it can also provide higher tap density. At the same time, the robust structural stability reduces the volume deformation during the frequent deintercalation of Na⁺. Finally, NVP/C-MSs still have a capacity retention rate of 79.1% after 10 000 cycles at a high rate of 20C.

Fig. 11 (a) (I) Scheme for the synthesis of Na₃V₂O₂(PO₄)₂F in a PTFE tube microreactor. TEM images of Na₃V₂O₂(PO₄)₂F prepared at different pH values: (II) pH 5.7, (III) pH 3.6, and (IV) pH 1.3. Reproduced with permission from ref. 174 Copyright 2021, Elsevier. (b) SEM images of (I) nano-NVOPF and (II) micro-NVOPF. Reproduced with permission from ref. 173 Copyright 2019, Elsevier. (c) (I) Schematic illustration of the formation process for hierarchical microspheres with various sizes and building blocks. (II) FESEM images of NVP/C-MSs. Reproduced with permission from ref. 179 Copyright 2019, Elsevier.

Based on the nanocrystallization of material particles, the design of active materials with unique morphology is also an effective strategy to improve the performance of the positive electrode.^180,181 Gao et al.¹⁸² prepared a mesoporous material Na₄Fe₃(PO₄)₂(P₂O₇) (NFPP@C@rGO) with a cross-linked double-carbon skeleton by a one-step solid-state method (Fig. 12aI), showing a 3D mesoporous sponge-like morphology as shown in Fig. 12aII and III. The porous structure effectively increases the contact area between the NFPP material and the electrolyte, shortens the diffusion distance of Na⁺ and improves the electronic conductivity. Finally, NFPP@C@rGO achieved excellent cycle stability at room temperature (86.7% capacity retention after 30 000 cycles at 20C) and high temperature (98.2% capacity retention after 200 cycles at 1C at 60 °C). As shown in Fig. 12b, Sun et al.¹⁸³ synthesized a unique Craspedia globosa-shaped Na₃V₂(PO₄)₃ (NVP-180) by the methanol hydrothermal method. It can be seen from Fig. 12b that this unique surface covers multiple interlaced paths. It not only increases the contact area between the active material and the electrolyte, but also provides sufficient paths for accelerating electron transport and effectively improves the electronic conductivity. Finally, NVP-180 still has a capacity retention rate of 88.7% after 1200 cycles at 10C due to its unique morphology.

Fig. 12 (a) (I) Scheme for the synthesis of the NFPP@C@rGO sample by the facial solid-state route. (II and III) SEM images of the NFPP@C@rGO sample. Reproduced with permission from ref. 182 Copyright 2023, Elsevier. (b) Schematic diagram of the synthesis of the NVP-180 sample by the hydrothermal method; the conductive mechanism diagram of NVP-180. Reproduced with permission from ref. 183 Copyright 2022, Elsevier. (c) (I) Flow diagram of the synthesis of CHACC-CNTs-NVP. (II and III) TEM images of etched CHACC-CNTS-NVP. Reproduced with permission from ref. 184 Copyright 2023, American Chemical Society.

Similarly, aiming at the problem of low conductivity of NVP, Li et al.¹⁸⁴ prepared dandelion-like Na₃V₂(PO₄)₃ (CHACC-CNT-NVP) modified by using cross-linked chitosan quaternary ammonium hydrogel (CHACC) and CNTs by the sol–gel method, as shown in Fig. 12cI. Among them, CHACC is carbonized to form an N–Cl co-doped carbon coating with a multi-defect structure. The doping of N and Cl atoms makes the surrounding carbon atoms positively charged. At the same time, the broken CNT tube end has a slight negative charge. Under the action of strong electrostatic adsorption, the two combine to form a unique and stable rod-like structure (Fig. 12cII and III). Thus, the rate performance and cycle stability of CHACC-CNT-NVP are effectively enhanced. Amorphous iron phosphate (FePO₄) has high theoretical specific capacity and excellent electrochemical reversibility, but it is limited by low rate performance and obvious capacity decay.^185,186 In order to solve this problem, Zhang et al.¹⁸⁷ prepared amorphous yolk–shell FePO₄ nanospheres (FePO₄-YSNSs) by the template method. The mesoporous nano-egg yolk structure effectively promoted the penetration of the electrolyte and enhanced the charge transfer and Na⁺ diffusion ability. In addition, the strong nanoshells ensure the structural integrity of FePO₄-YSNSs, thereby greatly improving the cycle stability of the material. The synergistic effect of the yolk–shell structure and nanoparticle structural unit effectively reduced the internal stress of the Na⁺ repeated deintercalation process. This unique structure allows FePO₄-YSNSs to maintain a discharge specific capacity of 97.1 mAh g⁻¹ after 1000 cycles at 100 mA g⁻¹, with a capacity retention rate of 91.3%. Besides the above morphology, researchers have prepared active materials such as coral-like and mulberry-like, which have shown significantly improved electrochemical performance.^188–191

6. Crystal plane design

The crystal plane of the cathode material acts as a reaction site between the electrolyte and the cathode material. The structure of the crystal plane determines the adsorption and desorption of the reactants (including Na⁺) and the charge transfer rate between the reactants on the crystal plane. Some crystal planes are densely distributed, which will hinder the migration of ions, while other crystal planes have abundant ion diffusion channels with open structures. Therefore, according to the characteristics of the crystal structure, it is necessary to find a suitable preparation method to control the directional growth of the crystal and design a crystal structure with specific crystal plane exposure, which can not only improve the performance of the cathode material, but also avoid the new problems of low density and high impurity content caused by the traditional modification strategy.^192,193 The nature of the crystal itself is an internal factor that determines its growth process and final morphology, but the external environment of its growth also has an important influence on its morphology. Therefore, the same crystal will have different morphologies in different growth environments. It is generally believed that the difference in crystal morphology is mainly caused by the anisotropic growth process of the crystal. In this process, the growth rate of the high-energy surface of the crystal is higher than that of the low-energy surface, resulting in a preferential growth orientation in the high-energy surface direction of the material. The presence of foreign substances can change the surface energy of different crystal planes of the crystal, so the growth rate of different crystal planes will also change, thus changing the particle morphology.^194,195

Li et al.¹⁹⁶ prepared more (002) active-surface-exposed Na₃V₂(PO₄)₂F₃ (NVPF-(002))by introducing clusters of carbon atoms during Na₃V₂(PO₄)₂F₃ (NVPF) crystal growth. As shown in Fig. 13a, the clusters of carbon atoms act as adsorbents during the crystal growth process, and because the adsorption energy of the (002) crystal plane is lower, the clusters of carbon atoms are more inclined to be adsorbed on the (002) crystal plane, which impedes the growth of crystals perpendicular to this crystal plane and forms two-dimensional particles. At the same volume, the area of the (002) crystal plane exposed to the electrolyte in the NVPF-(002) sample was 2.36 times larger than that of the pure NVPF sample, and the exposed (002) crystal plane had more stable Na⁺ storage sites, which significantly lowered the diffusion energy barrier for Na⁺ (0.43 vs. 0.66 eV (pure NVPF)). As a result, the NVPF-(002) cathode exhibited higher reversible capacity (132 mAh g⁻¹) and energy density (482 Wh kg⁻¹), and the performance advantage was gradually enhanced with the increase in current (Fig. 13a). In addition to the use of carbon atom clusters to regulate the crystal growth process of NVPF, NaX (X = F, Cl, Br) can be introduced as a crystal facet inducer to regulate the surface energy of different crystal facets and induce the dominant growth of specific crystal facets, thus obtaining materials with more active crystal facets as a means to improve the Na⁺ diffusion rate.¹⁹⁷

Fig. 13 (a) Schematic diagram of the growth of the (I) NVPF-(002) sample; (II) NVPF sample. (III) Rate performance comparison of NVPF cathodes at 0.1C–5C and corresponding charge–discharge curves (IV). Reproduced with permission from ref. 196 Copyright 2024, The Royal Society of Chemistry. (b) Schematic orientation illustration of NVOPF-PE single-crystal lamella. (II) AFM image and the corresponding (III) height profiles of NVOPF-PE plates. Reproduced with permission from ref. 197 Copyright 2024, American Chemical Society.

In view of the poor kinetic performance of Na₃(VOPO₄)₂F, Xu et al.¹⁹⁸ prepared a Na₃(VOPO₄)₂F cathode material with the {001} active surface dominated by the topochemical synthesis route (NVOPF-PE). As shown in Fig. 13b, a favorable sodium ion diffusion path is along the c-axis direction, which is perpendicular to the NVOPF-PE layer with preferred exposed {001} facets. Therefore, the exposure of more {001} facets and the micron-scale single crystal lamellar structure (Fig. 13b) significantly shorten the diffusion distance of Na⁺, thereby enhancing the sodium ion diffusion kinetics. Finally, without further material modification or electrolyte optimization, the NVOPF-PE electrode has a reversible capacity of 129 mAh g⁻¹ at a high rate of 10C, which is very close to the theoretical capacity of 132 mAh g⁻¹, achieving a high energy density of 452 Wh kg⁻¹ and a high power density of 4660 W kg⁻¹.

7. Phase engineering

The phase structure directly affects the insertion/extraction process of Na⁺ ions, which in turn affects the electrochemical performance of the material. Each phase has its advantages and disadvantages. By regulating the phase of the material, we can maximize its advantages and reduce the impact of its disadvantages, to obtain excellent performance, which is phase engineering. This strategy has been widely used in layered oxide cathode materials and has also shown significant performance improvement. However, there are relatively few applications in polyanionic compounds.^199–203 At present, there are mainly two kinds of phase, the non-stoichiometric phase and amorphous phase.

7.1. Non-stoichiometric phase

Non-stoichiometric phase materials refer to compounds whose atomic or ionic ratios of constituent elements do not conform to a simple integer ratio. The chemical composition of these compounds is variable within a certain range, which allows them to design high-performance materials by regulating defect structures. For example, Na₄Fe₃(PO₄)₂P₂O₇, as a representative mixed iron-based polyanion cathode material, has a high theoretical specific capacity of 129 mAh g⁻¹ and a stable voltage platform of 3.1 V. However, electrochemically inert NaFePO₄ impurities will be generated during the synthesis process, which will affect the material properties. Therefore, how to inhibit the generation of impurities has become the key. Recently, Fan et al.²⁰⁴ used a solid-solution strategy to decompose Na₄Fe₃(PO₄)₂P₂O₇ into 2NaFePO₄·Na₂FeP₂O₇ from the perspective of molecular composition. The nonstoichiometric pure phase material Na_3.4Fe_2.4(PO₄)_1.4P₂O₇ was prepared by adjusting the initial ratio of NaFePO₄ and Na₂FeP₂O₇ during the synthesis process. Thanks to the elimination of impurities, Na_3.4Fe_2.4(PO₄)_1.4P₂O₇ still has a discharge specific capacity of 50.5 mAh g⁻¹ at an ultra-high rate of 100C. Moreover, the capacity retention rate is still as high as 83.1% after 14 000 cycles at 20C. For the problem of NaFePO₄, in addition to inhibiting its formation, its particle size can also be reduced to the nanoscale, thereby stimulating electrochemical activity. To this end, Xu et al.²⁰⁵ reported a new type of iron-based phosphate composite material Na_4.5Fe_3.5(PO₄)_2.5(P₂O₇), which is composed of NaFePO₄ and Na₄Fe₃(PO₄)₂(P₂O₇) phases with a molar ratio of 0.5 : 1. As shown in Fig. 14a, many NaFePO₄ nanodomains are formed surrounded by the Na₄Fe₃(PO₄)₂(P₂O₇) phase. This nanocrystallization at the crystal structure level and the penetration of the two-phase interface can induce the transport of Na⁺ through the NaFePO₄ phase. Therefore, the electrochemical activation of the NaFePO₄ phase can be achieved even without any extreme processing (such as ultrahigh-voltage constant charging and nanoengineering by high-energy ball milling). Finally, Na_4.5Fe_3.5(PO₄)_2.5(P₂O₇) with a two-phase heterogeneous structure exhibits long-term cycle stability of more than 11 000 cycles and low temperature performance of −40 °C. Similarly, given the attenuation of Na₂FeP₂O₇ (original phase) at a high voltage platform, Wang et al.²⁰⁶ proposed to partially replace Na with Fe, and constructed the Na_1.4Fe_1.3P₂O₇ cathode material (iron-rich phase). The increased Fe ions in the iron-rich phase can effectively enhance the reversibility of the intermediate phase β-NaFeP₂O₇ during the charge–discharge process. Fig. 14b shows the phase transition energy between the structural phase and β-NaFeP₂O₇ after removing 0.5 Na⁺ in the original phase and the iron-rich phase, respectively. It can be seen that in the iron-rich phase, the energy barrier of the two-phase transition is lower, so it is easier for the reversible phase change to occur, and the Fe³⁺ reduction of the iron-rich phase is more complete under high pressure, thus effectively inhibiting the attenuation of the high voltage platform and improving the cycle stability of the material. After 650 cycles at 1C, the capacity and average voltage retention of the iron-rich phase are 84% and 95%, respectively, which are much higher than those of the original phase (12% and 61%).

Fig. 14 (a) Schematic illustration diagram of the NFPP-4.5 cathode with a two-phase intergrown heterostructure of NaFePO₄ and Na₄Fe₃(PO₄)₂(P₂O₇). Reproduced with permission from ref. 205 Copyright 2024, American Chemical Society. (b) The energy change of the phase transition at 3.5 V; the crystal structure of β-NaFeP₂O₇. Reproduced with permission from ref. 206 Copyright 2023, Elsevier. (c) Na_3+xV₂(PO₄)₃/C preparation process flowchart; the crystal structure of NaV₂(PO₄)₃. Reproduced with permission from ref. 209 Copyright 2023, American Chemical Society. (d) GCD profiles and dQ/dV curves of (I and II) N30TMP and (III and IV) N35TMP at 0.1C within the voltage range between 1.5 and 4.2 V. Reproduced with permission from ref. 208 Copyright 2021, American Chemical Society. (e) Stepwise fabrication process of NVP/C-0.6 and illustration of enhanced performance for bi-phase synergy of the amorphous and crystalline phases. Reproduced with permission from ref. 212 Copyright 2024, Elsevier. (f) Schematic illustration of the two synthesis routes for NVP/C. Reproduced with permission from ref. 214 Copyright 2025, Elsevier.

Na₃TiMn(PO₄)₃ (NTMP) with a NASICON structure has two high discharge voltage platforms of 3.5 V and 4.0 V (relative to Na⁺/Na). However, due to the existence of a low-voltage discharge platform (2.5 V), not only is the capacity of the 3.5 V and 4.0 V high-voltage discharge platforms significantly reduced, but a serious voltage lag is also caused. Studies have shown that the blocking effect of Mn²⁺ is located in the Na⁺ path, and the change of the induction effect of the P–Mn–O bond leads to a significant voltage hysteresis and capacity loss in the effective voltage range of 2.5–4.2 V.²⁰⁷ To this end, Zhang et al.²⁰⁸ synthesized sodium excess Na_3.5TiMn(PO₄)₃ (N35TMP) by a non-stoichiometric strategy. Even if no other modification measures are taken, the discharge platform and voltage hysteresis at 2.5 V are significantly suppressed, as shown in Fig. 14d. In addition, N35TMP also exhibits excellent cycle stability, with a capacity retention rate of 91.7% after 2000 cycles at 2C. As shown in Fig. 14c, Cong et al.²⁰⁹ prepared sodium-excess Na_3.4V₂(PO₄)₃/C (Fig. 14c) by the solid-phase method. The sodium excess significantly increased the Na⁺ diffusion rate during the entire reaction process, and a highly reversible two-phase reaction occurred between Na_3.4V₂(PO₄)₃ and NaV₂(PO₄)₃, resulting in excellent electrochemical performance. Finally, Na_3.4V₂(PO₄)₃/C delivers an ultra-high specific capacity (132.4 mAh g⁻¹) exceeding the theoretical specific capacity and maintains a capacity retention rate of 96% after 300 cycles.

7.2. Amorphous phase

In addition, considering the relatively low binding energy of atoms on the surface of amorphous materials, a large number of unsaturated coordination sites and dangling bonds provide abundant ion storage sites, and a large number of defects in the amorphous phase are conducive to the rapid conduction of ions.^210,211 Therefore, researchers intend to solve the problem of the low conductivity of polyanionic compounds by combining the amorphous phase and the crystalline phase. As shown in Fig. 14e, Zhou et al.²¹² successfully realized the homogeneous hybridization of crystalline and amorphous phases (NVP/C-0.6) in Na₃V₂(PO₄)₃/C by the sol–gel method. The amorphous phase around the crystal effectively increases the transport rate of Na⁺, reduces the volume deformation during the cycle, and also increases the content of Na2 activation sites and improves the reversible capacity. In addition, by reducing the bond length during the synthesis process and introducing an amorphous phase at the grain boundary (Fig. 14e), NVP/C-0.6 achieves excellent cycle stability. Finally, under the synergistic effect of the two phases, the NVP/C-0.6 cathode achieves a high discharge specific capacity of 82 mAh g⁻¹ at 10C rate, and the capacity can still be maintained at 94.5% after 300 cycles at 5C. In addition, previous studies have shown that amorphous phase materials can promote the generation of pseudocapacitance and improve material properties.²¹³ Based on this, Hao et al.²¹⁴ used the flash Joule heating (FJH) process shown in Fig. 14f to perform a short-time annealing at 1000 °C for 60 s to synthesize a special NVP material with a mixed texture of nanocrystalline and amorphous phases (F-NVP/C-1000). Compared with the control sample (R-NVP/C-800) prepared by conventional heat treatment, the particle size of F-NVP/C-1000 is significantly reduced and the oxygen vacancies are increased due to the advantages of short sintering time and fast cooling of the FJH process. These advantages, together with the nanocrystalline/amorphous mixed texture of NVP, are conducive to obtaining more pseudocapacitive charge storage behavior and enhancing electron/ion transport kinetics. Consequently, the F-NVP/C-1000 electrode exhibits a reversible capacity of 65.6 mAh g⁻¹ at a high rate of 40C.

8. Commercialization challenges

Relying on abundant sodium reserves, SIBs exhibit significant cost advantages, complemented by the performance benefits of high safety and high redox potential of polyanionic cathodes, which demonstrate strong competitiveness in the field of large-scale energy storage. However, there are still two key problems to be solved before commercial application, namely the production cost and process.

In order to achieve commercial application, the first thing to be solved is the cost of raw materials. Fig. 15 compares the cost ratio of each component of SIBs with polyanion cells and LIBs with LiFePO₄ cells. It can be seen that because the raw materials required for the polyanionic cathodes are low-cost Na, Fe and Mn, the proportion of the cathode in the cost of its battery is only 21.51%, which is much lower than the proportion of the cathode in the cost of the LiFePO₄ battery (41.17%). Although the abundance of raw materials brings significant cost advantages, the volume capacity density of polyanion batteries is lower than that of LiFePO₄ batteries, which requires more consumables and packaging. As a result, the cost of manufacturing, shell and accessories is almost twice as high as that of LIBs (38.6% vs.19.7%).^215,216 Therefore, the total cost of the battery is not only affected by the price of raw materials, but also by battery performance, accessory prices and battery structure, which will lead to changes in the total cost. In summary, even if the raw material cost of SIBs is low, the cost is still high due to factors such as battery performance and technological maturity.

Fig. 15 Cost proportion of each component of the SIB polyanionic core and LIB LiFePO₄ core.

In addition to the cost problem, another problem to be solved is the process problem in large-scale production. As mentioned above, a variety of modification strategies have been adopted to improve the low intrinsic electronic conductivity and low specific capacity of polyanionic cathodes, and remarkable results have been achieved. However, it is worth noting that most of the current research is still in the small-scale preparation stage, and it is still unclear whether the active material can maintain excellent electrochemical performance when produced on a large scale. For example, whether the doping element can be uniformly distributed when the element is doped, whether the coating material can be uniformly wrapped on the material particles when the surface is coated, and the regulation of particle morphology in large-scale production are significant problems. Therefore, rational design of modification strategies to ensure performance improvement while adapting to large-scale production is crucial for the commercial application of polyanionic SIBs.

9. Conclusion and perspectives

Sodium-ion batteries show great application potential due to the cost advantages provided by abundant resources. Especially in the field of large-scale energy storage applications, its cost advantage is further amplified. In addition, the replacement of lead–acid batteries in the field of low-speed electric vehicles has further broadened the application market of sodium-ion batteries. Among the cathode materials, polyanionic compounds exhibit significantly better safety than other cathode materials due to their unique structure, which highlights their advantages in large-scale energy storage applications. However, polyanionic compounds are still limited by low specific capacity and low intrinsic electronic conductivity. So far, a variety of effective modification strategies have been reported. In this review, we introduce these modification strategies for the improvement of material properties, while focusing on their most advanced progress. Table 1 and Fig. 16 provide a comprehensive summary of the advantages and disadvantages, modification strategies and electrochemical properties of typical polyanion electrode materials.

Table 1 Summary of modification strategies of polyanionic electrode materials

Modifying methods	Preparation method	Composite	Rate performance	Cycle performance	Ref.
K doping	Sol–gel method	Na2.95K0.05V2(PO4)3	80.2 mAh g^−1 at 100C (40 °C)	80.4% after 500 cycles (50C, 40 °C)	77
Li doping	Hydrothermal assisted sol–gel method	Na2.8Li0.2V2(PO4)3/C	102 mAh g^−1 at 20C	99.82% after 500 cycles (0.2C)	82
Lu doping	Sol–gel method	NVP/C@Lu-1%	75.5 mAh g^−1 at 80C	84% after 800 cycles (1C)	83
SiO4^4− doping	Sol–gel method	Na3.95Fe3(PO4)1.95P2O7(SiO4)0.05	56.8 mAh g^−1 at 100C	92.1% after 3000 cycles (20C)	87
SiO4^4− doping	Citric acid-assisted sol–gel method	Na3MnTi(PO4)2.7(SiO4)0.3	81 mAh g^−1 at 10C	79.2% after 1000 cycles (initial 3 cycles at 0.5C and later at 5C)	88
F doping	Sol–gel method	Na3.85MnCr(PO3.95F0.05)3	60.4 mAh g^−1 at 40C	55.1% after 1000 cycles (20C)	93
F doping	Solid-phase ball milling	Na2.2Fe1.75(SO3.9F0.1)3	121.5 mAh g^−1 at 12 mA g^−1	78.8% after 1000 cycles (600 mA g^−1)	94
PO4^3− doping	Solid-phase ball milling method	Na2.9Fe1.7(SO4)2.7(PO4)0.3	86.7 mAh g^−1 at 30C	85.5% after 6000 cycles (30C)	95
SO4^2− doping	Solvothermal method	Na3V2(PO4)1.95(SO4)0.05O2F	75.5 mAh g^−1 at 80C	111 mAh g^−1 after 1000 cycles (10C)	96
Al doping	Freeze drying method	Na3.9Fe2.9Al0.1(PO4)2(P2O7)	41.7 mAh g^−1 at 200C	85.1% after 10 000 cycles (50C)	103
Zr doping	Sol–gel method	Na1.86Fe0.93Zr0.07PO4F	73.78 mAh g^−1 at 5C	60% after 2000 cycles (5C)	104
Mg doping	—	Na2Fe0.93Mg0.07PO4F	61.5 mAh g^−1 at 20C	73.8% after 1000 cycles (20C)	105
Fe doping	One-step solid-state method	Na3.5V1.5Fe0.5(PO4)3	68 mAh g^−1 at 100C	84% after 10 000 cycles (100C)	111
Mn doping	High-temperature solid-state reaction	Na3V1.5Mn0.5(PO4)3	79.4 mAh g^−1 at 10C	110.1 mAh g^−1 after 100 cycles (1C)	112
Ti doping	Spray-drying-assisted annealing method	Na2.5V1.5Ti0.5(PO4)3/C	90.31 mAh g^−1 at 5 A g^−1	94.1% after 1000 cycles (1 A g^−1)	113
Alkali metal doping	Sol–gel method	Na3V1.94K0.06(PO4)3	100.5 mAh g^−1 at 20C	99.1% after 3000 cycles (20C)	114
Co/Zr co-doping	Sol–gel method	Na3V1.86Co0.07Zr0.07(PO4)3/C	69.11 mAh g^−1 at 100C	85.4% after 2000 cycles (100C)	115
High-entropy	—	Na3V1.9(CaMgAlCrMn)0.1(PO4)2F3	71.4 mAh g^−1 at 50C	80.4% after 2000 cycles (20C)	124
High-entropy	Sol–gel method	Na3.12MnTi0.9(VFeMgCrZr)0.02(PO4)3	99.92 mAh g^−1 at 20C	81.7% after 1000 cycles (5C)	125
High-entropy	Sol–gel method	Na4Cr0.7Fe0.4Mn0.3V0.3Al0.2(PO4)3	49.6 mAh g^−1 at 20C	70.67% after 2000 cycles (10C)	126
High-entropy	One-pot, solid-state reaction	Na3V1.8(CrMnFeZnAl)0.2(PO4)3	75.6 mAh g^−1 at 50C	80% after 3000 cycles (10C)	127
Cr, Si co-doping	Solid-state method	Na3V1.9Cr0.1(PO4)2.9(SiO4)0.1	94.1 mAh g^−1 at 5C	90% after 300 cycles (1C)	130
Al, F co-doping	—	NVMP-Al&F	83 mAh g^−1 at 15C	86.1% after 1000 cycles (5C)	131
K, Si co-doping	—	Na3.24K0.10V2.01(PO4)2.94(SiO4)0.14	78.5 mAh g^−1 at 40C	64% after 10 000 cycles (20C)	132
K, Mn co-doping	Sol–gel method	Na2.97K0.04V1.99Mn0.01(PO4)3/C@CNTs	106.1 mAh g^−1 at 15C	77.76% after 5800 cycles (15C)	135
K, La, Si co-doping	Sol–gel method	Na3.03V1.93La0.07(PO4)2.9(SiO4)0.1	103.8 mAh g^−1 at 10C	83.2% after 1500 cycles (10C)	136
Carbon coating	Wet ball-milling method	Na3V2(PO4)2F3@C	87.3 mAh g^−1 at 10C	101.6 mAh g^−1 after 200 cycles (0.5C)	140
Carbon coating	—	NVPFC-NS	93 mAh g^−1 at 60C	93% after 500 cycles (10C)	141
N doped carbon-coating	Sol–gel method	NMVP@C-20N	60.7 mAh g^−1 at 20C	86.1% after 2000 cycles (5C)	146
N doped carbon-coating	Sol–gel method	U-NC15@NVP	56.5 mAh g^−1 at 20C	92.1% after 400 cycles (1C)	147
S–N co-doped carbon coating	Sol–gel method combined with calcination	NVP@SNC	58.6 mAh g^−1 at 50C	70.6% after 4000 cycles (50C)	149
N–Cl co-doped carbon coating	Sol–gel method	CHACC-NVP-3	81.6 mAh g^−1 at 120C	88.4% after 3500 cycles (60C)	150
Composite with CNTs	Sol–gel method	NVPF@3%CNT	96.2 mAh g^−1 at 10C	84.5% after 1000 cycles (10C)	156
Composite with rGO	In situ solid-state approach	NVPFO/rGO	81 mAh g^−1 at 20C	72.3% after 300 cycles (5C)	157
Carbon coating of CDs as a carbonaceous precursor	Straightforward solid-phase ball milling combined with a sintering method	NVP/CDs	88.7 mAh g^−1 at 50C	92% after 10 000 cycles (50C)	161
Defect configuration entropy-strengthened carbon coating	Solid-phase ball milling coupled with the sintering method	NVP/NSFC	88.9 mAh g^−1 at 200C	90.2% after 5000 cycles (100C)	162
PVA-assisted construction of a 3D rGO network	Two-step method	NVPF@C@rGO	64 mAh g^−1 at 100C	98.3% after 700 cycles (50C)	163
Carbon-coated composite with PVP and modified CNTs	A simple sol–gel method and freeze-drying technology	Na4MnCr(PO4)3@C@PVP@CNTs	63 mAh g^−1 at 15C	59.3% after 600 cycles (5C)	164
Al(PO3)3 coating	Wet process and heat treatment	NMVP/C@Al(PO3)3	61 mAh g^−1 at 50C	88.5% after 3000 cycles (5C)	168
Na1.3Al0.7Ti1.3(PO4)3 coating	Sol–gel method	NMVP@2%NATP	56.3 mAh g^−1 at 10C	44.8% after 1000 cycles (5C)	169
Nanospheres and nanosheets	Microfluidic-based continuous-flow strategy	Na3V2O2(PO4)2F nanospherical and nanosheets	Nanospheres: 114 mAh g^−1 at 10C	Nanosheets: 97.3% after 1000 cycles (10C)	174
Nanoflake-constructed porous hierarchical microspheres	Hydrothermal method	NVP/C-MSs	99.3 mAh g^−1 at 100C	79.1% after 10 000 cycles (20C)	179
Mesoporous sponge-like	Facial solid-state route	Mesoporous sponge-like NFPP@C@rGO	80.7 mAh g^−1 at 20C	86.7% after 30 000 cycles (20C)	182
Craspedia globosa-shaped	Methanol hydrothermal method	NVP-180	99.3 mAh g^−1 at 10C	88.7% after 1200 cycles (10C)	183
Dandelion-shaped	Sol–gel method	CHACC-CNT-NVP	81.5 mAh g^−1 at 120C	75.27% after 4000 cycles (10C)	184
Yolk–shell-structured	Multi-step templating strategy	FePO4-YSNSs	74.3 mAh g^−1 at 1000 mA g^−1	91.3% after 1000 cycles (100 mA g^−1)	187
Dominantly exposed (002) active facets	—	NVPF-(002)	98 mAh g^−1 at 5C	89% after 1400 cycles (2C)	196
Dominantly exposed {001} active facets	Topochemical synthesis route	NVOPF-PE	129 mAh g^−1 at 10C	94.5% after 6000 cycles (10C)	198
Nonstoichiometry	Sol–gel method	Na3.4Fe2.4(PO4)1.4P2O7	50.5 mAh g^−1 at 100C	83.1% after 14 000 cycles (20C)	204
Nonstoichiometry	Sol–gel method	Na4.5Fe3.5(PO4)2.5(P2O7)	103 mAh g^−1 at 20C	88% after 11 000 cycles (20C)	205
Nonstoichiometry	Sol–gel method	Na1.4Fe1.3P2O7	—	84% after 650 cycles (1C)	206
Nonstoichiometry	Sol–gel method	Na3.5TiMn(PO4)3	61 mAh g^−1 at 5C	91.7% after 2000 cycles (2C)	208
Nonstoichiometry	High-temperature solid-phase method	Na3.4V2(PO4)3/C	—	96% after 300 cycles (20 mA g^−1)	209
Homogeneous hybridization of amorphous and crystalline phases	Sol–gel method	NVP/C-0.6	82 mAh g^−1 at 10C	94.5% after 300 cycles (5C)	212
Nanocrystal/amorphous phase-mingled texture	Flash Joule heating process	F-NVP/C-1000	65.6 mAh g^−1 at 40C	69.3% after 2000 cycles (1C)	214

---

Fig. 16 Modification measures of polyanion cathodes for sodium ion batteries.

At present, the most important modification strategies for polyanionic compounds are element doping, surface modification and morphology adjustment, which are also commonly used modification methods for other cathode materials. In addition, in recent years, some researchers have also modified the polyanionic compounds through crystal surface design and phase engineering, which also shows a good performance improvement.

(1) For element doping, different elements have different properties such as ionic radius, alternative potential energy, valence, and solid solubility. The structure of the matrix material is adjusted by adding other elements to enhance its electrochemical performance. Although much research has been carried out and good results have been achieved, there may still be undiscovered available doping elements (combinations). There is no doubt that determining the “best” doping elements (combinations) of various cathode materials through experiments is an extremely labor-intensive and material-intensive task. In addition, the influence of different contents of doping elements (combinations) on the performance of materials is also significantly different. Therefore, it is necessary to determine the content of the “best” doping element (combinations) while determining the “best” doping element (combinations), which undoubtedly further increases the difficulty of research work. Therefore, we propose to combine machine learning algorithms such as deep learning and neural networks with theoretical calculations to reveal hidden possibilities and predict appropriate doping schemes to minimize resource requirements to provide researchers with the “best” solution, thereby accelerating the development of high-performance polyanionic compound cathode materials.

(2) Surface modification can effectively improve the conductivity of polyanionic compounds. Among them, carbon materials are widely used because of their low cost and good electrical conductivity. In recent years, many strategies of composite modification of carbon materials have been adopted, which have brought more significant performance improvement. Different from carbon materials, which focus on the improvement of electronic conductivity, non-carbon-based materials have a variety of characteristics (high ionic conductivity, high thermal stability, etc.), resulting in various performance improvements. However, there are few related studies at present, and it is necessary to further explore the appropriate materials and explore its mechanism. In general, no matter what kind of material, the surface coating is required to be thin and uniform to achieve the best improvement effect under the influence of reducing the energy/volume density of the positive electrode. At the same time, based on the application purpose of large-scale energy storage, the content and cost of coating materials must be limited to avoid excessive material costs.

(3) For the regulation of morphology, the specific surface area of the material can be effectively increased by nanocrystallization or designing special morphology, while shortening the ion migration path and improving the conductivity of the material. However, it should be noted that nanoparticles are easy to agglomerate during high-temperature sintering and subsequent electrochemical cycling, which reduces the electrochemical performance of the material. Moreover, whether it is nanocrystallization or design of special morphology, an increase in specific surface area will lead to a decrease in tap density and an increase in surface side reactions while improving the conductivity of the material. Therefore, it is very important to control the appropriate specific surface area and find a balance point.

(4) For crystal surface design and phase engineering, there are relatively few studies on these two types of modification strategies, and there is a lack of relevant in-depth mechanism research. For example, the growth mechanism of the crystal and the theoretical model of the crystal plane are still unclear. Similarly, there are few theoretical studies on phase engineering, and it is still in the experimental exploration stage. Therefore, we believe that these two types of modification strategies should focus on the corresponding mechanism research, to provide clear theoretical support for related research, to improve the material performance more effectively.

In general, the above-mentioned modification strategies each have a unique improvement effect on the performance of active materials. As mentioned above, the primary issue with polyanionic cathodes is their low intrinsic electronic conductivity and low specific capacity, which are attributed to the large anionic groups in the structure. Element doping can improve the electronic conductivity and specific capacity of the material from the crystal structure level by doping different elements into the matrix material. At the same time, a large number of studies have focused on finding suitable doping elements to improve material performance, with more comprehensive experimental data. Therefore, in the above-mentioned modification strategies, we believe that element doping has a more comprehensive effect on the performance improvement of polyanionic cathodes.

For the commercialization of polyanionic SIBs, production cost and process are two major problems that must be solved. Reducing production costs requires multiple synergies. In addition to trying to improve the performance of polyanionic cathodes, it is also possible to optimize the battery structure from the perspective of battery structure design, reduce non-essential components, and increase the energy density of the battery. At the same time, the cooperation between upstream and downstream enterprises is strengthened in the production chain to reduce the cost of raw materials. In terms of the production process, we believe that we can learn from the production experience of LIBs, regulate the crystal structure of the material and coat an appropriate amount of coating to ensure the electrochemical performance of the material while adapting to large-scale production.

Although there has been more and more related research in recent years and some progress has been made, the practical application of polyanionic cathode materials is still a big problem. Therefore, we believe that while exploring appropriate modification strategies, we should pay more attention to relevant theoretical calculations and structural analysis. At present, we can prepare electrode materials with excellent performance by some modification methods, but the deep mechanism is not very clear, such as the structural evolution and reaction mechanism caused by some doping elements entering the host material, and the interaction between the coating material and the active material, which is difficult to be targeted. Therefore, we suggest observing the crystal structure change, phase transition and lattice constant change of the cathode electrode during the charge and discharge process through relevant theoretical calculations and intuitive and accurate in situ testing technology, such as in situ XRD. The detailed structural information of the intermediate products produced during the redox reaction of the electrode was observed by in situ Raman spectroscopy. The valence state changes of elements in the reaction process were observed by in situ XANES. In addition, in situ electrochemical mass spectrometry and in situ optical fiber sensors can effectively observe the changes of each component of the battery during charge and discharge in real time, so as to provide more in-depth and accurate insights into the related mechanisms of polyanionic compounds.

In terms of the full-cell configuration and optimization, it is necessary to explore compatible electrolytes and anode materials to achieve synergy and maximize full-cell performance. For electrolytes, it is essential to adapt to high-voltage electrodes while maintaining stability in various climatic environments. Because in the high voltage region, the high valence transition metal may catalyze the decomposition of the electrolyte, and the future demand for battery performance will inevitably require the stability of the electrolyte at high voltage. At the same time, considering the application scenarios of large-scale energy storage, higher requirements will be placed on the stable operation of electrolytes under different environmental conditions (high temperature, low temperature, humidity, etc.). For the negative electrode material, it should have sufficient Na⁺ adsorption active sites, suitable working potential, a short Na⁺ diffusion path and high electronic conductivity to meet the needs of fast charging of the battery. In addition, high safety is also an essential point to avoid the problem of thermal runaway due to the increase in internal temperature during the rapid charging and discharging process of the battery. Finally, the synthesis process of anode materials is optimized to reduce production costs, so as to ensure the cost advantage of SIBs and help their development in the field of large-scale energy storage applications.

Finally, based on the purpose of large-scale energy storage applications, the preparation methods of materials also need to be paid attention to. It should not be limited to small-scale preparation in the laboratory stage, but should consider whether it is suitable for industrial production. The performance of modified materials should be ensured while preparing on a large scale. Therefore, we suggest that we should pay attention to the preparation method while exploring the material modification methods, and learn from the development experience of lithium-ion batteries to help the development of sodium-ion batteries.

Conflicts of interest

There are no conflicts to declare.

Data availability

No primary research results, software or code has been included and no new data were generated or analysed as part of this review.

Acknowledgements

This work was supported by the National Key Research and Development Program of China (no. 2023YFB3809300) and the Fundamental Research Funds for the Central Universities (no. 2024ZYGXZR099).

References

1. J. L. Holechek, H. M. E. Geli, M. N. Sawalhah and R. Valdez, Sustainability, 2022, 14, 4792.
2. A. Bauer, J. Song, S. Vail, W. Pan, J. Barker and Y. H. Lu, Adv. Energy Mater., 2018, 8, 1702869.
3. E. Vine, Energy Effic., 2008, 1, 49–63.
4. W. G. Santika, M. Anisuzzaman, P. A. Bahri, G. M. Shafiullah, G. V. Rupf and T. Urmee, Energy Res. Soc. Sci., 2019, 50, 201–214.
5. M. Amin, H. H. Shah, A. G. Fareed, W. U. Khan, E. H. Y. Chung, A. Zia, Z. U. R. Farooqi and C. H. Y. Lee, Int. J. Hydrogen Energy, 2022, 47, 33112–33134.
6. F. Rizzi, N. J. van Eck and M. Frey, Renewable Energy, 2014, 62, 657–671.
7. A. Qazi, F. Hussain, N. Abd Rahim, G. Hardaker, D. Alghazzawi, K. Shaban and K. Haruna, IEEE Access, 2019, 7, 63837–63851.
8. A. I. Osman, L. Chen, M. Y. Yang, G. Msigwa, M. Farghali, S. Fawzy, D. W. Rooney and P. S. Yap, Environ. Chem. Lett., 2023, 21, 741–764.
9. A. Rahman, O. Farrok and M. M. Haque, Renewable Sustainable Energy Rev., 2022, 161, 112279.
10. R. Baños, F. Manzano-Agugliaro, F. G. Montoya, C. Gil, A. Alcayde and J. Gómez, Renewable Sustainable Energy Rev., 2011, 15, 1753–1766.
11. R. Kardooni, S. B. Yusoff and F. B. Kari, Energy Policy, 2016, 88, 1–10.
12. A. Raheem, S. A. Abbasi, A. Memon, S. R. Samo, Y. H. Taufiq-Yap, M. K. Danquah and R. Harun, Energy Sustainability Soc., 2016, 6, 16.
13. Y. Y. Zeng, T. Zhou, T. Wang, M. Zhang, S. P. Zhang and H. R. Yang, Energies, 2025, 18, 466.
14. S. Chu and A. Majumdar, Nature, 2012, 488, 294–303.
15. A. Poullikkas, Renewable Sustainable Energy Rev., 2013, 27, 778–788.
16. M. Müller, L. Viernstein, C. N. Truong, A. Eiting, H. C. Hesse, R. Witzmann and A. Jossen, J. Energy Storage, 2017, 9, 1–11.
17. J. L. Li, Z. H. Liang and S. H. Xu, Energy Rep., 2022, 8, 584–596.
18. A. A. Kebede, T. Kalogiannis, J. Van Mierlo and M. Berecibar, Renewable Sustainable Energy Rev., 2022, 159, 112213.
19. C. K. Ekman and S. H. Jensen, Energy Convers. Manage., 2010, 51, 1140–1147.
20. B. Dunn, H. Kamath and J. M. Tarascon, Science, 2011, 334, 928–935.
21. S. Hameer and J. L. van Niekerk, Int. J. Energy Res., 2015, 39, 1179–1195.
22. T. M. Gür, Energy Environ. Sci., 2018, 11, 2696–2767.
23. M. Perrin, Y. M. Saint-Drenan, F. Mattera and P. Malbranche, J. Power Sources, 2005, 144, 402–410.
24. H. S. Chen, T. N. Cong, W. Yang, C. Q. Tan, Y. L. Li and Y. L. Ding, Prog. Nat. Sci., 2009, 19, 291–312.
25. V. Etacheri, R. Marom, R. Elazari, G. Salitra and D. Aurbach, Energy Environ. Sci., 2011, 4, 3243–3262.
26. J. Cho, S. Jeong and Y. Kim, Prog. Energy Combust. Sci., 2015, 48, 84–101.
27. M. Li, Z. W. Chen, T. P. Wu and J. Lu, Adv. Mater., 2018, 30, 1801190.
28. J. Xie and Q. C. Zhang, Small, 2019, 15, 1805061.
29. R. Usiskin, Y. X. Lu, J. Popovic, M. Law, P. Balaya, Y. S. Hu and J. Maier, Nat. Rev. Mater., 2021, 6, 1020–1035.
30. T. F. Liu, Y. P. Zhang, Z. G. Jiang, X. Q. Zeng, J. P. Ji, Z. H. Li, X. H. Gao, M. H. Sun, Z. Lin, M. Ling, J. C. Zheng and C. D. Liang, Energy Environ. Sci., 2019, 12, 1512–1533.
31. Y. M. Li, Y. X. Lu, C. L. Zhao, Y. S. Hu, M. M. Titirici, H. Li, X. J. Huang and L. Q. Chen, Energy Storage Mater., 2017, 7, 130–151.
32. A. Eftekhari and D. W. Kim, J. Power Sources, 2018, 395, 336–348.
33. J. F. Peters, A. P. Cruz and M. Weil, Batteries, 2019, 5, 10.
34. Y. Y. Zhu, Y. H. Wang, Y. T. Wang, T. J. Xu and P. Chang, Carbon Energy, 2022, 4, 1182–1213.
35. D. Kundu, E. Talaie, V. Duffort and L. F. Nazar, Angew. Chem., Int. Ed., 2015, 54, 3431–3448.
36. F. Feng, S. L. Chen, S. Q. Zhao, W. L. Zhang, Y. G. Miao, H. Y. Che, X. Z. Liao and Z. F. Ma, Chem. Eng. J., 2021, 411, 128518.
37. Z. B. Liu, C. Peng, J. Wu, T. T. Yang, J. Zeng, F. K. Li, A. Kucernak, D. F. Xue, Q. Liu, M. Zhu and J. Liu, Mater. Today, 2023, 68, 22–33.
38. J. Peters, D. Buchholz, S. Passerini and M. Weil, Energy Environ. Sci., 2016, 9, 1744–1751.
39. Y. Q. Wang, N. An, L. Wen, L. Wang, X. T. Jiang, F. Hou, Y. X. Yin and J. Liang, J. Energy Chem., 2021, 55, 391–419.
40. H. Zhang, Y. Gao, X. H. Liu, L. F. Zhou, J. Y. Li, Y. Xiao, J. Peng, J. Z. Wang and S. L. Chou, Adv. Energy Mater., 2023, 13, 2300149.
41. Y. Zhao, Y. Q. Kang, J. Wozny, J. Lu, H. Du, C. L. Li, T. Li, F. Y. Kang, N. Tavajohi and B. H. Li, Nat. Rev. Mater., 2023, 8, 623–634.
42. S. Zhang, B. Steubing, H. K. Potter, P. A. Hansson and A. Nordberg, Resour., Conserv. Recycl., 2024, 202, 107362.
43. H. Rostami, J. Valio, P. Suominen, P. Tynjälä and U. Lassi, Chem. Eng. J., 2024, 495, 153471.
44. P. K. Nayak, L. T. Yang, W. Brehm and P. Adelhelm, Angew. Chem., Int. Ed., 2018, 57, 102–120.
45. K. Chayambuka, G. Mulder, D. L. Danilov and P. H. L. Notten, Adv. Energy Mater., 2020, 10, 2001310.
46. E. Goikolea, V. Palomares, S. J. Wang, I. R. de Larramendi, X. Guo, G. X. Wang and T. Rojo, Adv. Energy Mater., 2020, 10, 2002055.
47. A. Rudola, A. J. R. Rennie, R. Heap, S. S. Meysami, A. Lowbridge, F. Mazzali, R. Sayers, C. J. Wright and J. Barker, J. Mater. Chem. A, 2021, 9, 8279–8302.
48. X. H. Liang, J. Y. Hwang and Y. K. Sun, Adv. Energy Mater., 2023, 13, 2301975.
49. K. Wu, X. W. Dou, X. X. Zhang and C. Y. Ouyang, Engineering, 2023, 21, 36–38.
50. Y. J. Liu, X. Cui, Y. Liu and Y. Y. Xia, Small, 2023, 19, 2302972.
51. C. L. Zhao, L. L. Liu, X. G. Qi, Y. X. Lu, F. X. Wu, J. M. Zhao, Y. Yu, Y. S. Hu and L. Q. Chen, Adv. Energy Mater., 2018, 8, 1703012.
52. X. W. Pang, B. G. An, S. M. Zheng and B. Wang, J. Alloys Compd., 2022, 912, 165142.
53. C. X. Peng, X. J. Xu, F. K. Li, L. Xi, J. Zeng, X. Song, X. H. Wan, J. W. Zhao and J. Liu, Small Struct., 2023, 4, 2300150.
54. M. Y. He, S. M. Liu, J. T. Wu and J. L. Zhu, Prog. Solid State Chem., 2024, 74, 100452.
55. M. Morant-Giner, R. Sanchis-Gual, J. Romero, A. Alberola, L. García-Cruz, S. Agouram, M. Galbiati, N. M. Padial, J. C. Waerenborgh, C. Martí-Gastaldo, S. Tatay, A. Forment-Aliaga and E. Coronado, Adv. Funct. Mater., 2018, 28, 1706125.
56. D. X. Zuo, C. P. Wang, J. J. Han, J. W. Wu, H. J. Qiu, Q. Zhang, Y. Lu, Y. J. Lin and X. J. Liu, ACS Sustain. Chem. Eng., 2020, 8, 16229–16240.
57. X. Z. Wang, Y. T. Zuo, Y. B. Qin, X. Zhu, S. W. Xu, Y. J. Guo, T. R. Yan, L. Zhang, Z. B. Gao, L. Z. Yu, M. T. Liu, Y. X. Yin, Y. H. Cheng, P. F. Wang and Y. G. Guo, Adv. Mater., 2024, 36, 2312300.
58. L. Wang, M. Y. Sun, L. Deng, Y. Q. Zheng, X. Y. Li, Y. S. Jiang, L. Zhao and Z. B. Wang, J. Energy Storage, 2024, 81, 110299.
59. Q. N. Liu, Z. Hu, M. Z. Chen, C. Zou, H. L. Jin, S. Wang, S. L. Chou and S. X. Dou, Small, 2019, 15, 1805381.
60. X. Tan, J. Zeng, L. Y. Sun, C. X. Peng, Z. Li, S. H. Zou, Q. Shi, H. Wang and J. Liu, Infomat, 2025, e12636,  DOI:10.1002/inf2.12636.
61. Y. J. Guo, R. X. Jin, M. Fan, W. P. Wang, S. Xin, L. J. Wan and Y. G. Guo, Chem. Soc. Rev., 2024, 53, 7828–7874.
62. J. Q. Wang, Y. F. Zhu, Y. Su, J. X. Guo, S. Q. Chen, H. K. Liu, S. X. Dou, S. L. Chou and Y. Xiao, Chem. Soc. Rev., 2024, 53, 4230–4301.
63. Q. N. Liu, Z. Hu, M. Z. Chen, C. Zou, H. L. Jin, S. Wang, S. L. Chou, Y. Liu and S. X. Dou, Adv. Funct. Mater., 2020, 30, 1909530.
64. J. Peng, W. Zhang, Q. N. Liu, J. Z. Wang, S. L. Chou, H. K. Liu and S. X. Dou, Adv. Mater., 2022, 34, 2108384.
65. Y. Xiao, J. Xiao, H. K. Zhao, J. Y. Li, G. L. Zhang, D. Y. Zhang, X. Guo, H. Gao, Y. Wang, J. Chen, G. X. Wang and H. Liu, Small, 2024, 20, 2401957.
66. S. T. Xu, Y. Yang, F. Tang, Y. Yao, X. Lv, L. Liu, C. Xu, Y. Z. Feng, X. H. Rui and Y. Yu, Mater. Horiz., 2023, 10, 1901–1923.
67. Y. Gao, H. Zhang, X. H. Liu, Z. Yang, X. X. He, L. Li, Y. Qiao and S. L. Chou, Adv. Energy Mater., 2021, 11, 2101751.
68. Y. B. Niu, Y. Zhang and M. W. Xu, J. Mater. Chem. A, 2019, 7, 15006–15025.
69. W. H. Guan, B. Pan, P. Zhou, J. X. Mi, D. Zhang, J. C. Xu and Y. Z. Jiang, ACS Appl. Mater. Interfaces, 2017, 9, 22369–22377.
70. P. Barpanda, L. Lander, S. Nishimura and A. Yamada, Adv. Energy Mater., 2018, 8, 1703055.
71. Q. Ni, Y. Bai, F. Wu and C. Wu, Adv. Sci., 2017, 4, advs.201770012.
72. N. Zhao, T. Zhang, H. L. Zhao and Y. L. Hou, Mater. Today Nano, 2020, 10, 100072.
73. X. Wang, Y. Xu, Y. Xi, X. Yang, J. Wang, X. Huang, W. Li, K. Xu, K. Zhang, R. Duan, D. Liu, N. Hou, Z. Yang, H. Wang and X. Li, J. Mater. Chem. A, 2024, 12, 9268–9295.
74. Y. J. Cao, Y. Liu, D. Q. Zhao, J. X. Zhang, X. P. Xia, T. Chen, L. C. Zhang, P. Qin and Y. Y. Xia, J. Alloys Compd., 2019, 784, 939–946.
75. H. Zhou, L. Y. Yong, X. Cheng, G. Ai, X. C. Zhao, L. J. Tong, T. T. Xu and W. F. Mao, J. Energy Storage, 2024, 93, 112358.
76. J. Z. Sun, X. H. Meng, Y. T. Hui, Q. Shen, M. Q. Fan, J. Y. Zhang, X. Wang, H. M. Jiang, Q. L. Kang, L. J. Yan, C. B. Wan and T. L. Ma, Nano Energy, 2025, 134, 110587.
77. L. Shen, Y. Li, C. Hu, Z. Huang, B. Wang, Y. Wang, Q. Liu and F. Ye, Mater. Today Chem., 2023, 30, 101506.
78. Q. Wu, Y. Z. Ma, S. Q. Zhang, X. Chen, J. B. Bai, H. Wang and X. J. Liu, Small, 2024, 20, 2404660.
79. J. Liu, L. L. Zhang, X. Z. Cao, X. Lin, Y. Shen, P. Zhang, C. Wei, Y. Y. Huang, W. Luo and X. L. Yang, ACS Appl. Energy Mater., 2020, 3, 7649–7658.
80. R. H. Wang, S. Q. Wu, F. Zhang, X. Zhao, Z. J. Lin, C. Z. Wang and K. M. Ho, Phys. Chem. Chem. Phys., 2020, 22, 13975–13980.
81. Q. Zheng, X. Ni, L. Lin, H. M. Yi, X. W. Han, X. F. Li, X. H. Bao and H. M. Zhang, J. Mater. Chem. A, 2018, 6, 4209–4218.
82. J. Cong, S. H. Luo, P. Y. Li, X. Yan, L. X. Qian and S. X. Yan, Appl. Surf. Sci., 2024, 643, 158646.
83. Y. R. Zhao, C. H. Qian, C. C. Liu, Q. Huang and Y. J. Chen, J. Energy Storage, 2024, 94, 112536.
84. M. Y. Wang, J. Z. Guo, Z. W. Wang, Z. Y. Gu, X. J. Nie, X. Yang and X. L. Wu, Small, 2020, 16, 1907645.
85. X. Liu, J. Gong, X. J. Wei, L. Ni, H. Y. Chen, Q. J. Zheng, C. G. Xu and D. M. Lin, J. Colloid Interface Sci., 2022, 606, 1897–1905.
86. H. Zhou, Z. T. Cao, Y. F. Zhou, J. X. Li, Z. H. Ling, G. Z. Fang, S. Q. Liang and X. X. Cao, Nano Energy, 2023, 114, 108604.
87. M. Z. Liu, M. Li, B. L. Zhang, H. M. Li, J. X. Liang, X. Y. Hu, H. M. Liu and Z. F. Ma, ACS Sustain. Chem. Eng., 2023, 11, 18102–18111.
88. S. T. Lu, Y. X. Cai, Y. Y. Li, X. Y. Du, J. J. Wang, Y. Y. Liu, K. Z. Cao and Y. Fan, ACS Appl. Mater. Interfaces, 2024, 16, 38092–38100.
89. B. Y. Xing, J. K. Ren, P. Hu, W. Luo, B. Mai, H. W. Cai, J. H. Wu, X. F. Wu, X. B. Chen, Z. H. Deng, W. C. Feng and L. Q. Mai, Small, 2024, 20, 2310997.
90. J. J. Wang, H. B. Jing, X. M. Wang, Y. Q. Xue, Q. H. Liang, W. H. Qi, H. Yu and C. F. Du, Adv. Funct. Mater., 2024, 34, 2315318.
91. J. C. Wang, X. Q. Lai, Z. X. Li, P. F. Wang, J. H. Zhang and T. F. Yi, J. Ind. Eng. Chem., 2024, 135, 356–363.
92. J. C. Shao, Q. Y. Meng, H. Z. Chi, W. Zhang and H. Y. Qin, Small, 2025, 21, 2409731.
93. W. Zhang, Y. L. Wu, Y. H. Dai, Z. M. Xu, L. He, Z. Li, S. H. Li, R. W. Chen, X. Gao, W. Zong, F. Guo, J. X. Zhu, H. B. Dong, J. W. Li, C. M. Ye, S. M. Li, F. X. Wu, Z. Zhang, G. J. He, Y. Q. Lai and I. P. Parkin, Chem. Sci., 2023, 14, 8662–8671.
94. T. S. Zhang, P. C. Zhang, J. Liu, L. F. Zhang, Y. W. Zheng, X. W. Shen, Y. J. Qian, X. Zhou, J. Q. Zhou, T. Qian and C. L. Yan, Nano Lett., 2025, 25, 2150–2158.
95. C. Y. Liu, K. A. Chen, F. M. Li, A. L. Zhao, Z. X. Chen, Y. J. Fang and Y. L. Cao, Nano Energy, 2024, 125, 109557.
96. X. Y. Wang, F. Zhang, X. Y. Zhou, Q. Wang, C. Y. Liu, Y. Y. Liu, H. Wang and X. J. Liu, J. Energy Chem., 2025, 100, 509–521.
97. M. T. Deng, Y. L. Sui, K. X. Rao, Y. Wang, W. B. Fei, K. Y. Sun, Z. L. Yang, X. P. Zhang and L. Wu, Chem. Eng. J., 2024, 499, 156347.
98. F. J. Ji, K. Chen, F. M. Li, A. L. Zhao, Y. J. Fang and Y. L. Cao, Energy Storage Mater., 2024, 68, 103371.
99. J. Y. Zeng, J. Q. Gao, W. S. Jian, H. J. Wang, W. Y. Li, N. Y. Hong, B. C. Zhang, J. N. Huang, B. Song, C. L. Gan, S. Yasar, W. T. Deng, G. Q. Zou, H. S. Hou and X. B. Ji, Adv. Funct. Mater., 2024, 34, 2410992.
100. C. Sun, Y. J. Zhao, Q. Ni, Z. Sun, X. Y. Yuan, J. B. Li and H. B. Jin, Energy Storage Mater., 2022, 49, 291–298.
101. C. L. Xu, R. J. Xiao, J. M. Zhao, F. X. Ding, Y. Yang, X. H. Rong, X. D. Guo, C. Yang, H. Z. Liu, B. H. Zhong and Y. S. Hu, ACS Energy Lett., 2022, 7, 97–107.
102. Z. C. Li, C. Sun, M. Li, X. Y. Wang, Y. Li, X. Y. Yuan, H. B. Jin and Y. J. Zhao, Adv. Funct. Mater., 2024, 34, 2314019.
103. J. Q. Gao, J. Y. Zeng, W. S. Jian, Y. Mei, L. S. Ni, H. J. Wang, K. Wang, X. Y. Hu, W. T. Deng, G. Q. Zou, H. S. Hou and X. B. Ji, Sci. Bull., 2024, 69, 772–783.
104. L. He, X. C. Ge, X. Wang, J. Li, X. D. Li, S. M. Li, Y. Q. Lai and Z. Zhang, Energy Storage Mater., 2023, 61, 102905.
105. Y. Liu, S. Y. Li, Z. Y. Gu, Y. L. Heng, H. Y. Lü, J. L. Yang, M. Du, X. T. Wang, J. Z. Guo, F. L. Dong, K. Li and X. L. Wu, Energy Storage Mater., 2024, 68, 103319.
106. H. Q. Huang, Y. F. Xia, Y. C. Hao, H. S. Li, C. Y. Wang, T. T. Shi, X. Y. Lu, M. W. Shahzad, B. B. Xu and Y. Z. Jiang, Adv. Funct. Mater., 2023, 33, 2305109.
107. J. Zhang, X. D. Zhao, Y. Z. Song, Q. Li, Y. C. Liu, J. Chen and X. R. Xing, Energy Storage Mater., 2019, 23, 25–34.
108. N. Wang, J. C. Ma, Z. L. Liu, J. Xu, D. Q. Zhao, N. Wang, C. Yang, Y. J. Cao, J. Lu and J. X. Zhang, Chem. Eng. J., 2022, 433, 133798.
109. C. L. Xu, J. M. Zhao, Y. A. Wang, W. B. Hua, Q. Fu, X. M. Liang, X. H. Rong, Q. Q. Zhang, X. D. Guo, C. Yang, H. Z. Liu, B. H. Zhong and Y. S. Hu, Adv. Energy Mater., 2022, 12, 2200966.
110. L. J. Yue, C. Peng, C. L. Guo, X. Y. Zhou, G. Li, N. N. Wang, J. S. Zhang, J. Q. Liu, Z. C. Bai and X. S. Zhao, Chem. Eng. J., 2022, 441, 136132.
111. Y. F. Zhou, G. F. Xu, J. D. Lin, Y. P. Zhang, G. Z. Fang, J. Zhou, X. X. Cao and S. Q. Liang, Adv. Mater., 2023, 35, 2304428.
112. Y. X. Chen, Q. P. Li, P. Wang, X. Y. Liao, J. Chen, X. Q. Zhang, Q. J. Zheng, D. M. Lin and K. H. Lam, Small, 2023, 19, 2304002.
113. L. Zhu, M. M. Wang, S. Xiang, L. Fu, D. Sun, X. B. Huang, Y. X. Li, Y. G. Tang, Q. Zhang and H. Y. Wang, ACS Nano, 2024, 18, 13073–13083.
114. X. Shen, M. Han, Y. F. Su, M. Wang and F. Wu, Nano Energy, 2023, 114, 108640.
115. B. B. Wang, B. F. Zhang and Y. J. Chen, J. Alloys Compd., 2025, 1010, 177328.
116. M. Li, C. Sun, X. Y. Yuan, Y. Li, Y. F. Yuan, H. B. Jin, J. Lu and Y. J. Zhao, Adv. Funct. Mater., 2024, 34, 2314019.
117. M. Li, C. Sun, Q. Ni, Z. Sun, Y. Liu, Y. Li, L. Li, H. B. Jin and Y. J. Zhao, Adv. Energy Mater., 2023, 13, 2203971.
118. Z. Q. Hao, X. Y. Shi, W. Q. Zhu, Z. Yang, X. Z. Zhou, C. C. Wang, L. Li, W. B. Hua, C. Q. Ma and S. L. Chou, ACS Nano, 2024, 18, 9354–9364.
119. X. P. Hu, S. Q. Liang, J. D. Lin, W. Ren, S. Q. Fu, Z. T. Cao, T. Zhang, L. Zhang and X. X. Cao, Adv. Energy Mater., 2024, 2404965,  DOI:10.1002/aenm.202404965.
120. N. Jiang, X. Y. Wang, H. R. Zhou, Y. C. Wang, S. Y. Sun, C. Yang and Y. Liu, Small, 2024, 20, 2308681.
121. S. Y. Wang, T. D. Xu, H. T. Leng, S. Y. Liang, W. Zhang, Y. H. Jin, J. X. Qiu and S. Li, J. Mater. Chem. A, 2024, 12, 33617–33623.
122. X. Y. Liao, Y. J. Li, B. Xie, M. Xie, X. Tan, Q. J. Zheng, L. Li, X. X. Zhao, Z. Y. Gu, S. C. Smith, J. X. Zhao, D. M. Lin and X. L. Wu, Energy Storage Mater., 2025, 74, 103920.
123. S. El Moutchou, N. Sabi, N. Oueldna, V. Trabadelo, H. Aziam and H. Ben Youcef, J. Energy Storage, 2024, 98, 113078.
124. Z. Y. Gu, J. Z. Guo, J. M. Cao, X. T. Wang, X. X. Zhao, X. Y. Zheng, W. H. Li, Z. H. Sun, H. J. Liang and X. L. Wu, Adv. Mater., 2022, 34, 2270110.
125. N. Zhang, X. R. Dong, Q. Yan, J. Y. Wang, F. Jin, J. X. Liu, D. L. Wang, H. K. Liu, B. Wang and S. X. Dou, Energy Storage Mater., 2024, 72, 103734.
126. P. Zhu, Y. P. Wang, J. Li and Y. Jin, Energy Storage Mater., 2024, 71, 103650.
127. Y. F. Zhou, G. F. Xu, J. D. Lin, J. Zhu, J. A. Pan, G. Z. Fang, S. Q. Liang and X. X. Cao, Nano Energy, 2024, 128, 109812.
128. Y. Liu, C. Sun, Q. Ni, Z. Sun, M. Li, S. Ma, H. B. Jin and Y. J. Zhao, Energy Storage Mater., 2022, 53, 881–889.
129. S. Y. Li, Q. M. Yin, Z. Y. Gu, Y. Liu, Y. N. Liu, M. Y. Su and X. L. Wu, J. Colloid Interface Sci., 2024, 664, 381–388.
130. P. Li, M. Gao, D. Wang, Z. Z. Li, Y. L. Liu, X. H. Liu, H. Y. Li, Y. Sun, Y. Liu, X. B. Niu, B. H. Zhong, Z. G. Wu and X. D. Guo, ACS Appl. Mater. Interfaces, 2023, 9475–9485,  DOI:10.1021/acsami.2c22038.
131. J. W. Zhao, W. T. Yan, S. J. Li, S. Y. Li, W. H. Wang and Y. Bai, Nano Energy, 2024, 119, 109002.
132. M. Y. Dou, Y. X. Zhang, J. Wang, X. Y. Zheng, J. Y. Chen, B. Han, K. S. Xia, Q. Gao, X. X. Liu, C. G. Zhou, R. M. Sun and Z. Cai, J. Power Sources, 2023, 560, 232709.
133. A. Das, S. B. Majumder and A. R. Chaudhuri, J. Power Sources, 2020, 461, 228149.
134. Z. Y. Tian, Y. N. Chen, S. Q. Sun, H. L. Liu, C. Wang, Q. Huang, C. C. Liu, Y. Z. Wang and L. Guo, J. Colloid Interface Sci., 2022, 613, 536–546.
135. B. B. Wang, Y. R. Zhao, S. Q. Liu, B. F. Zhang and Y. J. Chen, J. Energy Storage, 2024, 99, 113305.
136. S. Q. Sun, Y. J. Chen, Q. Bai, Q. Huang, C. C. Liu, S. N. He, Y. X. Yang, Y. Z. Wang and L. Guo, J. Mater. Chem. A, 2022, 10, 11340–11353.
137. X. C. Ge, X. H. Li, Z. X. Wang, H. J. Guo, G. C. Yan, X. W. Wu and J. X. Wang, Chem. Eng. J., 2019, 357, 458–462.
138. M. F. Su, J. W. Shi, Q. L. Kang, D. W. Lai, Q. Y. Lu and F. Gao, Chem. Eng. J., 2022, 432, 134289.
139. W. Ren, Y. Y. Wang, X. P. Hu, Z. T. Cao, Y. Q. Xu, Y. F. Zhou, X. X. Cao and S. Q. Liang, Chem. Eng. J., 2024, 487, 150492.
140. Z. Y. Gu, J. Z. Guo, Z. H. Sun, X. X. Zhao, W. H. Li, X. Yang, H. J. Liang, C. D. Zhao and X. L. Wu, Sci. Bull., 2020, 65, 702–710.
141. D. Jia, Y. Zhang, W. Song, P. C. Y. Tang and Y. Huang, ACS Appl. Mater. Interfaces, 2022, 40812–40821,  DOI:10.1021/acsami.2c07897.
142. W. Li, Z. J. Yao, Y. Zhong, C. A. Zhou, X. L. Wang, X. H. Xia, D. Xie, J. B. Wu, C. D. Gu and J. P. Tu, J. Mater. Chem. A, 2019, 7, 10231–10238.
143. W. Li, Z. J. Yao, C. A. Zhou, X. L. Wang, X. H. Xia, C. D. Cu and J. P. Tu, Small, 2019, 15, 1902432.
144. Y. S. Li, X. H. Liang, G. B. Zhong, C. Wang, S. J. Wu, K. Q. Xu and C. H. Yang, ACS Appl. Mater. Interfaces, 2020, 12, 25920–25929.
145. D. Yan, S. H. Xiao, X. Y. Li, R. Wu, J. X. Jiang, X. B. Niu and J. S. Chen, ChemSusChem, 2022, 15, e202201121.
146. Y. S. Wang, Q. C. Wang, X. Y. Ding, M. Wang, Y. H. Xin and H. C. Gao, Appl. Surf. Sci., 2022, 601, 154218.
147. Y. Q. Zhu, H. Xu, J. Ma, P. D. Chen and Y. Chen, Surf. Interfaces, 2024, 45, 103888.
148. H. Zhang, I. Hasa, D. Buchholz, B. S. Qin and S. Passerini, Carbon, 2017, 124, 334–341.
149. J. Kang, L. Zhu, F. Y. Teng, S. Q. Wang, Y. G. Huang, Y. H. Xiang, Z. Chen and X. W. Wu, Rare Met., 2023, 42, 1570–1582.
150. J. H. Li, Y. J. Chen, S. N. He, Y. X. Yang, Y. Z. Wang and L. Guo, Chem. Eng. J., 2023, 452, 139311.
151. H. D. Dong, C. C. Liu, Q. Huang, L. Guo and Y. J. Chen, Energy Storage Mater., 2025, 75, 104070.
152. C. D. Zhao, J. Z. Guo, Z. Y. Gu, X. X. Zhao, W. H. Li, X. Yang, H. J. Liang and X. L. Wu, J. Mater. Chem. A, 2020, 8, 17454–17462.
153. S. Kadam, R. Kate, U. Chothe, P. Chalwadi, J. Shingare, M. Kulkarni, R. Kalubarme and B. Kale, ACS Appl. Mater. Interfaces, 2023, 15, 34651–34661.
154. L. Li, Y. Y. Qin, S. S. Zhang, H. Y. Zhao, J. Zhao, X. Y. Li, J. Y. Zhao, H. Wu, Y. Q. Su and S. J. Ding, J. Power Sources, 2023, 576, 233226.
155. J. D. Guan, S. L. Zhou, J. J. Zhou, F. D. Wu, X. Y. Shi, J. L. Xu, L. Y. Shao, Z. Q. Luo and Z. P. Sun, ACS Appl. Mater. Interfaces, 2024, 16, 20559–20569.
156. Q. Q. Zhang, X. G. Sun, K. Liu, Q. Xu, S. J. Zheng and S. Dai, ACS Sustain. Chem. Eng., 2023, 11, 12992–13001.
157. X. F. Sun, Z. K. Wang, A. Ndahimana, Y. B. Han, Q. H. Bo, X. T. Gu, J. L. Cui and X. S. Mei, J. Energy Storage, 2023, 72, 108400.
158. S. Y. Liu, X. X. Cao, Y. P. Zhang, K. Wang, Q. Su, J. Chen, Q. He, S. Q. Liang, G. Z. Cao and A. Q. Pan, J. Mater. Chem. A, 2020, 8, 18872–18879.
159. R. Huang, D. Yan, Q. Y. Zhang, G. W. Zhang, B. B. Chen, H. Y. Yang, C. Y. Yu and Y. Bai, Adv. Energy Mater., 2024, 2400595,  DOI:10.1002/aenm.202400595.
160. C. Hu, M. Y. Li, J. S. Qiu and Y. P. Sun, Chem. Soc. Rev., 2019, 48, 2315–2337.
161. Y. J. Li, Y. Mei, Y. J. Huang, X. Zhong, Z. L. Geng, Z. D. He, H. R. Ding, W. T. Deng, G. Q. Zou, T. C. Liu, X. B. Ji, K. Amine and H. S. Hou, ACS Nano, 2024, 18, 25053–25068.
162. Y. J. Li, Y. Mei, H. X. Liu, H. R. Ding, Y. J. Huang, X. Zhong, Z. L. Geng, Z. D. He, W. T. Deng, G. Q. Zou, X. B. Ji and H. S. Hou, Nano Energy, 2024, 130, 110107.
163. S. L. Xu, Y. Zhu, X. Y. Li, Y. M. Wang, D. Yan, X. B. Niu, J. X. Jiang, R. Wu and J. S. Chen, J. Energy Chem., 2024, 99, 100–109.
164. R. Y. Chen, D. S. Butenko, S. L. Li, X. Y. Zhang, G. S. Li, I. Zatovsky and W. Han, Chin. Chem. Lett., 2024, 35, 108358.
165. Y. Zhang, J. H. Xun, K. Y. Zhang, B. Zhang and H. Y. Xu, J. Mater. Chem. A, 2022, 10, 11163–11171.
166. S. Chakrabarty, T. Sharabani, S. Taragin, R. Yemini, A. Maddegalla, I. Perelshtein, A. Mukherjee and M. Noked, J. Energy Storage, 2024, 84, 111507.
167. C. H. Qian, M. N. Shi, C. F. Fan, C. C. Liu, Q. Huang and Y. J. Chen, J. Colloid Interface Sci., 2024, 664, 573–587.
168. K. Wang, X. B. Huang, C. C. Luo, Y. M. Shen, H. Y. Wang and T. Zhou, J. Colloid Interface Sci., 2023, 642, 705–713.
169. C. R. Song, S. Y. Li and Y. Bai, Nano Res., 2024, 17, 2728–2735.
170. Y. R. Qi, Z. Z. Tong, J. M. Zhao, L. Ma, T. P. Wu, H. Z. Liu, C. Yang, J. Lu and Y. S. Hu, Joule, 2018, 2, 2348–2363.
171. Y. J. Cao, Y. Liu, T. Chen, X. P. Xia, L. C. Zhang, J. X. Zhang and Y. Y. Xia, Ionics, 2019, 25, 1083–1090.
172. Z. H. Chen, W. C. Ma, Q. L. Fan, Y. H. Liu, M. Sun and S. Wang, Mater. Res. Bull., 2024, 170, 112594.
173. J. L. Xu, J. Z. Chen, L. Tao, Z. L. Tian, S. Zhou, N. Zhao and C. P. Wong, Nano Energy, 2019, 60, 510–519.
174. L. M. Zheng, D. T. Zhang, X. Y. Wang and G. S. Guo, Appl. Mater. Today, 2021, 23, 101032.
175. X. H. Liu, W. H. Lai and S. L. Chou, Mater. Chem. Front., 2020, 4, 1289–1303.
176. H. Y. Wu, X. Zou and X. L. Wu, Nano Futures, 2020, 4, 042001.
177. J. W. Hu, X. W. Li, Q. Q. Liang, L. Xu, C. S. Ding, Y. Liu and Y. F. Gao, Nano–Micro Lett., 2025, 17, 33.
178. S. T. Xu, S. M. Yang, C. C. Liu, Y. Yao, Y. Yang, X. H. Rui and Y. Yu, Small, 2025, 2407285,  DOI:10.1002/smll.202407285.
179. X. X. Cao, A. Q. Pan, B. Yin, G. Z. Fang, Y. P. Wang, X. Z. Kong, T. Zhu, J. Zhou, G. Z. Cao and S. Q. Liang, Nano Energy, 2019, 60, 312–323.
180. R. M. Zhan, Y. Q. Zhang, H. Chen, Q. J. Xu, Q. R. Ma, W. Gao, T. T. Yang, J. Jiang, S. J. Bao and M. W. Xu, ACS Appl. Mater. Interfaces, 2019, 11, 5107–5113.
181. B. J. Wang, Y. Hu, X. P. Zhang, Z. H. Shi, L. Wu and Y. L. Sui, Ceram. Int., 2023, 49, 1061–1068.
182. J. Q. Gao, Y. Tian, Y. Mei, L. S. Ni, H. J. Wang, H. Q. Liu, W. T. Deng, G. Q. Zou, H. S. Hou and X. B. Ji, Chem. Eng. J., 2023, 458, 141385.
183. S. Q. Sun, Y. J. Chen, Z. Y. Tian, X. M. Jiang, J. H. Li, Z. Tian, Q. Huang, C. C. Liu, Y. Z. Wang and L. Guo, J. Alloys Compd., 2022, 909, 164719.
184. J. H. Li, Y. J. Chen, Q. Bai, S. N. He, Y. X. Yang, C. Zheng, Y. Z. Wang and L. Guo, ACS Sustain. Chem. Eng., 2023, 11, 12631–12645.
185. S. Y. Duan, J. Y. Piao, T. Q. Zhang, Y. G. Sun, X. C. Liu, A. M. Cao and L. J. Wan, NPG Asia Mater., 2017, 9, e414.
186. Z. Z. Zhang, Y. Han, J. M. Xu, J. H. Ma, X. S. Zhou and J. C. Bao, ACS Appl. Energy Mater., 2018, 1, 4395–4402.
187. Z. Z. Zhang, Y. C. Du, Q. C. Wang, J. Y. Xu, Y. N. Zhou, J. C. Bao, J. Shen and X. S. Zhou, Angew. Chem., Int. Ed., 2020, 59, 17504–17510.
188. Y. Hou, K. Chang, Z. Y. Wang, S. Gu, Q. Liu, J. J. Zhang, H. Cheng, S. L. Zhang, Z. R. Chang and Z. G. Lu, Sci. China Mater., 2019, 62, 474–486.
189. W. Li, Z. J. Yao, Y. Liu, S. Z. Zhang, X. L. Wang, X. H. Xia, C. D. Gu and J. P. Tu, Chem. Eng. J., 2022, 433, 133557.
190. Y. J. Chen, Z. Y. Tian, J. H. Li and T. Zhou, Chem. Eng. J., 2023, 472, 145041.
191. C. H. Qian, M. N. Shi, C. C. Liu, Q. Huang and Y. J. Chen, Chem. Eng. J., 2024, 494, 153087.
192. S. Y. Zhou, T. Mei, X. B. Wang and Y. T. Qian, Nanoscale, 2018, 10, 17435–17455.
193. L. Zhu, L. Fu, K. X. Zhou, L. X. Yang, Z. Tang, D. Sun, Y. G. Tang, Y. X. Li and H. Y. Wang, Chem. Rec., 2022, 22, e202200128.
194. Y. N. Xia, Y. J. Xiong, B. Lim and S. E. Skrabalak, Angew. Chem., Int. Ed., 2009, 48, 60–103.
195. J. J. Li and F. L. Deepak, Chem. Rev., 2022, 122, 16911–16982.
196. Z. Z. Li, L. Qiu, P. Li, H. Liu, D. Wang, W. B. Hua, T. Chen, Y. Song, F. Wan, B. H. Zhong, Z. G. Wu and X. D. Guo, J. Mater. Chem. A, 2024, 12, 7777–7787.
197. L. Zhu, Q. Zhang, D. Sun, Q. Wang, N. N. Weng, Y. G. Tang and H. Y. Wang, Mater. Chem. Front., 2020, 4, 2932–2942.
198. C. L. Xu, Q. Fu, W. B. Hua, Z. Chen, Q. H. Zhang, Y. Bai, C. Yang, J. M. Zhao and Y. S. Hu, ACS Nano, 2024, 18, 18758–18768.
199. Q. Huang, J. T. Liu, L. Zhang, S. Xu, L. B. Chen, P. Wang, D. G. Ivey and W. F. Wei, Nano Energy, 2018, 44, 336–344.
200. Y. Xiao, P. F. Wang, Y. X. Yin, Y. F. Zhu, X. N. Yang, X. D. Zhang, Y. S. Wang, X. D. Guo, B. H. Zhong and Y. G. Guo, Adv. Energy Mater., 2018, 8, 1800492.
201. Y. Xiao, Y. F. Zhu, W. Xiang, Z. G. Wu, Y. C. Li, J. Lai, S. Li, E. Wang, Z. G. Yang, C. L. Xu, B. H. Zhong and X. D. Guo, Angew. Chem., Int. Ed., 2020, 59, 1491–1495.
202. N. Jiang, Q. N. Liu, J. W. Wang, W. F. Yang, W. S. Ma, L. Q. Zhang, Z. Q. Peng and Z. H. Zhang, Small, 2021, 17, 2007103.
203. X. Liang and Y. K. Sun, Adv. Funct. Mater., 2022, 32, 2206154.
204. Z. W. Fan, W. D. Song, N. Yang, C. J. Lou, R. Y. Tian, W. B. Hua, M. X. Tang and F. Du, Angew. Chem., Int. Ed., 2024, 63, e202316957.
205. C. L. Xu, L. Zhou, T. Gao, Z. Chen, X. Y. Hou, J. Zhang, Y. Bai, L. R. Yang, H. Z. Liu, C. Yang, J. M. Zhao and Y. S. Hu, J. Am. Chem. Soc., 2024, 146, 9819–9827.
206. J. Wang, W. H. Zeng, J. W. Zhu, F. J. Xia, H. Y. Zhao, W. X. Tian, T. T. Wang, Y. X. Zhang, S. J. Zhang and S. C. Mu, Nano Energy, 2023, 116, 108822.
207. C. L. Xu, W. B. Hua, Q. H. Zhang, Y. Liu, R. B. Dang, R. J. Xiao, J. Wang, Z. Chen, F. X. Ding, X. D. Guo, C. Yang, L. R. Yang, J. M. Zhao and Y. S. Hu, Adv. Funct. Mater., 2023, 33, 2302810.
208. J. S. Zhang, C. F. Lin, Q. B. Xia, C. Wang and X. S. Zhao, ACS Energy Lett., 2021, 6, 2081–2089.
209. J. Cong, S. H. Luo, P. Y. Li, K. Li, P. W. Li, S. X. Yan, L. X. Qian and X. Liu, ACS Sustain. Chem. Eng., 2023, 11, 16341–16353.
210. S. D. Feng, Z. Q. Song, Y. H. Zhang, Z. J. Li, L. M. Wang and R. P. Liu, Sci. China Mater., 2023, 66, 4143–4164.
211. T. Hakari, M. Deguchi, A. Sakuda, M. Tatsumisago and A. Hayashi, J. Power Sources, 2023, 562, 232739.
212. Y. J. Zhou, X. C. Yang, M. J. Hou, L. Q. Zhao, X. Y. Zhang and F. Liang, J. Colloid Interface Sci., 2024, 667, 64–72.
213. H. B. Li, M. H. Yu, F. X. Wang, P. Liu, Y. Liang, J. Xiao, C. X. Wang, Y. X. Tong and G. W. Yang, Nat. Commun., 2013, 4, 1894.
214. L. Y. Hao, Z. K. Li, L. J. Cao, H. G. Lv and Y. Zhao, J. Energy Storage, 2025, 111, 115378.
215. Z. Hao, X. Shi, Z. Yang, X. Zhou, L. Li, C.-Q. Ma and S. Chou, Adv. Mater., 2024, 36, 2305135.
216. Y. Wang, Z. Cao, Z. Du, X. Cao and S. Liang, Acta Phys.-Chim. Sin., 2025, 41, 100035.

---