In the structure of MOFs, both redox metal ions and organic ligands can be used as active sites. The conversion mechanism is to use the valence change of the central metal to achieve effective lithium storage. According to whether the metal element is produced during the charge-discharge cycle, it can be divided into reversible and irreversible conversion mechanisms. In the process of reversible transformation, the central metal only changes its valence and does not produce simple metal, so the reversibility is obviously improved. In the process of irreversible conversion reaction, the formation of metal elements makes the structure of MOFs irreversibly destroyed, and even the electrode structure will be pulverized after only one charge-discharge cycle, which affects the capacity and coulombic efficiency of MOFs to a certain extent. Due to the decomposition and structural collapse of MOFs, the reversible capacity of most MOFs anode materials based on conversion mechanism (such as MOF-177 and Fe-MOF (MIL-88B)) after multiple cycles is only provided by the metal or metal oxide generated by the decomposition of MOFs, and the cycle stability is poor^[7][8]. Therefore, the irreversible conversion reaction is the key factor for the capacity fading of MOFs anode, and the reversible conversion reaction of metal center is the source of electrode capacity stabilization.

In the stable porous structure of some MOFs, organic ligands contain unsaturated bonds, and functional groups such as amino, carboxyl and benzene rings have the ability to store and transfer charges. In the electrochemical reaction process, energy is reversibly stored mainly by opening and reconnecting unsaturated chemical bonds.Different types of unsaturated chemical bonds are used to increase the number of active sites involved in electrochemical reactions, promote the rapid transmission and reversible intercalation/deintercalation of Li⁺, and improve the lithium storage capacity. Such as MIL-101 (Fe), Fe-MOF, and 2D/2D Ti₃C₂T_x/NiCo MOF heterostructures achieve effective lithium storage through this mechanism^[9⇓~11]. Therefore, it is necessary to rationally select organic ligands and optimize the structural design to obtain MOFs anode materials with high specific capacity and cycle stability.

The adsorption/desorption mechanism relies on the electrostatic interaction of the electrode surface to store energy, rather than chemical bonding, and the capacity is usually very low. Although there are few examples of reversible lithium storage using adsorption/desorption mechanism (such as Na⁺storage in hard carbon), it should be paid enough attention to porous MOFs electrode materials with high specific surface area^[12].

Co-MOFs are one of the more widely studied anode materials for lithium-ion batteries. Li et al. Prepared optimized CoBTC-EtOH hollow microspheres by hydrothermal method^[19]. The reversible capacity of CoBTC-EtOH anode is 856 mAh/G after 100 cycles at 100 mA/G. After 500 cycles at a high current density of 2 A/G, it is still 473 mAh/G. Although the Li⁺ can be intercalated into the organic part of Co-BTC, the Co²⁺ always remains ionic and does not directly participate in the charge-discharge cycle. Similarly, after 200 cycles at 100 mA/G, the capacity of the conchoidal Co₂(OH)₂BDC(S-Co-MOF) is as high as 1021 mAh/G; The capacity is still 435 mAh/G after 1000 cycles at 1 A/G^[10]. This good cycling stability and outstanding lithium storage capacity depend on the organic moiety-dominated intercalation/deintercalation mechanism to ensure the efficient transport of Li⁺. In addition, the contribution of pseudocapacitive characteristics to the Li⁺ intercalation/deintercalation ability is also important, that is, the pores of S-Co-MOF absorb an appropriate amount of Li⁺, resulting in a higher capacity than the theoretical value at low rates. Compared with Co-MOFs, the micro-flower-shaped H-Co-MOF anode composed of nanosheets has high pore volume and specific surface area, which can provide sufficient lithium storage sites, and can also further improve the performance with the help of the ion buffer zone in the interface layer^[20]. Obviously, the hierarchical structure is an important factor affecting the performance of most Co-MOFs anodes, which rely on a single metal or ligand to participate in electrochemical reactions^[21]. Ning et al. Synthesized morphology-controlled Co-MOFs (u-CoTDA) ultrathin nanosheets by ultrasonic method, and the capacity, rate capability and long-term cycling stability of this anode material not only depend on the characteristics of more active sites for Li⁺ intercalation/deintercalation, less diffusion barriers and lower mechanical strain,Ultrathin nanosheets can also increase the proportion of coordinatively unsaturated metal active sites exposed on the surface and improve the redox activity of metal centers, that is, the ligands of Co (II) and u-CoTDA participate in the redox process at the same time^[22]. Therefore, it is feasible to rationally design nanosized MOFs to shorten the ion/electron diffusion distance and improve the diffusion kinetics.

In addition, because the coordination compound is easy to decompose during charge-discharge cycles, the construction of one-dimensional chain or multi-dimensional framework structure can effectively improve its solid-state stability and capacity reversibility. The metal ions and the bridging ligands are not connected by π-π stacking interaction or covalent bonds, but are coordinated by weak hydrogen bonds to form a one-dimensional chain structure, which does not affect the size of the Li⁺ diffusion channel of the electrode material and the dissolution of the active organic in the electrolyte, and the capacity is improved mainly through the combination of two active elements, and the coulombic efficiency is up to 98.3%^[23]. It can be seen that the participation of aromatic chelating ligands and metal centers is a new mechanism for lithium storage in coordination compounds. Co₂(DOBDC)(Co-MOFs) anode was prepared by a modified hydrothermal method, and most of the capacity was derived from the pseudocapacitive characteristics including surface or near-surface redox reactions and a large number of ion intercalation, rather than the diffusion-controlled faradaic effect, which may be related to the microporous structure characteristics and high specific surface area^[24]. The reversible capacity at 2 A/G does not decrease significantly and remains at 408.2 mAh/G. Due to the synergistic effect of metal centers and organic ligands and the uniform nanowire morphology during cycling, the carboxylic acid-based one-dimensional coordination polymer (Co-BDCN) anode not only has a high capacity of 1132 mAh/G after 100 cycles at 100 mA/G, but also has excellent cycle stability^[25]. Depending on the metal-ligand coordination mechanism and intermolecular interactions, metal ions (or metal ion clusters) and organic bridging ligands can also assemble into two-dimensional or three-dimensional frameworks. In the serrated worm-like nano-Co-BTC, the chain structure effectively promotes the reversible charge transfer, and the conjugated carboxylate has a strong π-π interaction, which makes the capacity, cycle stability and coulombic efficiency outstanding, but the internal structure transformation and electrochemical mechanism still need to be further explored^[26]. CoC₈H₄O₄(CoTPA) is a kind of lamellar MOFs containing carbonyl groups, and both Co²⁺ and organic moieties containing carbonyl and aromatic structures possess electrochemical active sites, while multi-electron transfer reactions provide high capacity and good capacity retention^[27]. This is inconsistent with the lithium storage mechanism of common MOFs anode. The three-dimensional Co₂(OH)₂BDC with high purity and crystallinity prepared by Gou et al. By solvothermal method is superior to other known BDC-based anode materials in terms of specific capacity, cycle stability and rate capability, especially the effect of solvent ratio on the purity of the product is obvious^[28]. Coordination polymers belong to weak framework structures, which absorb relatively less solvent and are more prone to reversible transformation reactions. Fei et al. Used a simple template-free method to synthesize CoC₆H₂O₅(H₂O)₂ hollow microspheres^[29]. Under the combined action of the crystal structure, the reversible conversion reaction and the oxygen atom in the aromatic ring, the cycle stability is obviously higher than that of the non-hollow microsphere.

In terms of structure, size, compositional functionality, permanent porosity and specific surface area, MOFs are promising electrode materials, but their capacity and cycle stability are not ideal. For example, the initial discharge capacity of MOF-177(Zn₄O(1,3,5-benzenetribenzoate)₂) is 400 mAh/G, and after the first discharge, metal zinc and lithium salts are irreversibly generated, and the charge-discharge cycle is realized through the reversible alloying of zinc and lithium, but the capacity decays rapidly due to the structural collapse in the subsequent cycle process, so it is not suitable as an anode material for lithium-ion batteries^[7]. The formate-bridged rhombic structure MOFs(Zn₃(HCOOH)₆) reacts reversibly with Li during cycling by relying on a turnover mechanism^[30]. Compared with Li₂O, the thermodynamic feasibility of lithium formate to transition metal formate is better, and the matrix involved in the cycling process is lithium formate rather than the typical Li₂O. The reversible formation or regeneration of FOR1 MOF is the key to obtain good lithium storage ability, and its electrochemical performance is significantly better than that of MOF-177. On the one hand, the post-modification of hydrophobic functional framework can be used to construct stable MOFs. On the other hand, internal pores with polar functions can chemically capture and strongly interact with guest atoms or molecules. Lin et al. Designed and synthesized BMOF(Zn(IM)_1.5(abIM)_0.5) with outstanding chemical and thermal stability, which can store lithium reversibly mainly through adsorption mechanism and functionalized pores.At the same time, the host-guest interaction between Li and amine functional groups/N atoms in imidazole can not be ignored, that is, the lithium storage capacity can be further improved by increasing the number of active N-rich functional sites and the ratio of surface area/pore volume^[31]. In addition, the effects of the molar ratio of the initial compounds, reaction temperature and time on the final product were also systematically studied. Therefore, the application of Zn-MOFs anode is largely limited by its low conductivity and capacity, as well as poor cycling stability.

The characteristics of low cost, low toxicity, safety, and environmental friendliness make Mn-MOFs have great potential for energy storage applications. In the [Mn(tfbdc)(4,4'-bpy)(H₂O)₂](Mn-LCP) of the two-dimensional microporous structure, the Mn (II) is weakly antiferromagnetically coupled^[32]. Although the ideal capacity is not obtained due to the conversion mechanism, the electrochemical performance may be improved by self-assembling variable valence metal ions and low molecular weight organic ligands, which provides a new idea for the preparation of MOF anode materials. Using surfactants to control the shape, growth, and initial nucleation stability of building units, Fei et al. Prepared Mn₂(C₆H₂O₄S)·2H₂O microspheres using a soft template method^[33]. The results show that the reversible capacity is 645.7 mAh/G after 250 cycles at 400 mA/G. Coulombic efficiency approaches 100% even after 650 cycles at 500 mA/G. Compared with MOF materials with irregular shapes, the new microsphere anode has better cycle stability and is very potential for lithium-ion batteries. Conjugated carboxylates are weak electron-attracting ligands with strong π-π interactions, which act as redox centers to coordinate with lithium and can maintain structural integrity and crystallinity in lithium-ion batteries. Maiti et al. Synthesized Mn-BTC MOF with coexistence of structural pores and intergranular pores by solvothermal method^[34]. The specific capacity of the anode material is 694 mAh/G at 100 mA/G; It also shows good rate capability when the potential is lower than 2. 0 V. Li⁺ intercalates into the organic part of carboxylic acid group and benzene ring in Mn-BTC MOF, and the electron-donating effect is the main driving force for lithium storage, and its redox reaction does not belong to the conversion mechanism. Xiong et al. Synthesized uniform single crystal Mn-PBA(Mn[Fe(CN)₆]_0.6667·nH₂O) cubes at room temperature^[35]. In addition to the inherent three-dimensional framework structure, which is conducive to Li⁺ transport, the larger free space can accommodate transition metal ions and some small molecules, and also maintain the charge balance and redox reaction through the change of iron valence. Therefore, the capacity, rate capability, cycle stability and coulombic efficiency are all outstanding. Mn-1,4-BDC @ 200 was prepared from Mn-1,4-BDC MOF by heat treatment (200 ℃ for 24 H). Both of them have the same characteristic bands and framework structure, but the morphology is changed, that is, the loose and uniform lamellar structure is split into small pieces^[36]. Reinsch and Stock and Li et al. Considered that the crystallinity and reversible intercalation/deintercalation of Li⁺ of MOFs are closely related to the coordination solvent in the pores^[37][19]. Therefore, the reversible capacity of Mn-1,4-BDC @ 200 anode is as high as 974 mAh/G after 100 cycles at 100 mA/G, further demonstrating the positive effect of removing the coordinating solvent molecules on improving the lithium storage performance. Mn-MOF(Mn₂(OH)₂BDC) is composed of randomly arranged close-packed ultrathin nanosheets, which has the advantages of low metal ion radius and high specific surface area^[38]. During the charge-discharge cycle, the connection between some BDC^2- and surface metal ions is interrupted, resulting in the formation of coordinatively unsaturated metal sites on the exposed surface, and the metal redox activity is improved, and the central metal ions and aromatic ligands BDC^2- participate in lithium storage. Under the combined effect of these factors, the reversible capacity, rate capability and Li⁺ diffusion coefficient of the anode material are high.

F F Érey et al. First used MIL-53 (Fe) as an anode material for lithium-ion batteries, with a capacity of only 75 mAh/G and unsatisfactory cycle stability^[39]. Benzoquinone can absorb electrochemically active molecules and act as a redox mediator to accelerate electron transfer and increase the initial capacity of MIL-53 (Fe), but the exchange of benzoquinone and electrolyte molecules will cause capacity fading in subsequent cycles^[40]. In addition, the reversibility of electrochemical cycling of MIL-68 (Fe) is not as good as that of MIL-53 (Fe) because the pore shape is not conducive to Li⁺ diffusion^[41]. Under the condition of no precursor and pH adjustment, Hu et al. Reasonably selected iron source and directly prepared nano-sized Fe-BTC MOF with carboxylate as ligand^[42]. The electrochemical performance of the Fe-BTC MOF anode, especially the rate capability and coulombic efficiency, which are closely related to practical applications, is quite outstanding due to the heterogeneous catalytic oxidation and stable structure during cycling. In addition, the flexible structure MIL-88A synthesized by solvothermal method belongs to the MIL series of Fe-MOFs materials^[43]. Restricted by its poor conductivity and the formation of SEI layer, it is not conducive to Li⁺ intercalation/deintercalation, and the initial discharge specific capacity is 140. 5 mAh/G. Under the action of external conditions such as guest molecules, the flexible framework structure is easy to cause the expansion and contraction of pores, and some electrolytes and electrode materials are continuously consumed to produce irreversible by-products (lithium hydroxide, lithium oxide and lithium carbonate), which lead to rapid capacity decay in the subsequent cycle process. There is no doubt that improving the reversibility of Li⁺ intercalation/deintercalation process while improving the conductivity of nanostructured materials is an effective way to further improve the performance of electrode materials. Shen et al. Designed and synthesized Fe-MOF (Fe-MIL-88B) nanorods with polyhedral structure, and measured the half-cell and full-cell performance of their composition^[44]. The results show that the higher capacity and cycling stability mainly depend on the transition metal clusters and organic ligands in Fe-MIL-88B anode, which can help to better understand the lithiation/delithiation behavior of small organic molecules in advanced electrode materials. Shin et al. Suggested that the capacity fading of MIL-101 (Fe) anode materials is not because of the collapse of the framework structure, but depends on the time-dependent irreversible oxidation (Fe²⁺→Fe³⁺), that is, the Fe²⁺→Fe³⁺ relaxation promotes the irreversible accumulation of Li in MIL-101 (Fe)^[9]. Obviously, precise control of the electronic structure can effectively improve the reversibility of MOFs electrode materials. The morphology of Fe (Zn) -BDC @ 300 was improved, the specific surface area increased and the charge transfer resistance decreased under the action of the structure and morphology directing agent Zn(NO₃)₂, but the effect of Zn content was negligible. Therefore, the capacity and cycling stability of this anode material are higher^[45].

Li et al. Synthesized waffle-like Ni-MOF(Ni₂(OH)₂BDC) by ultrasonic method^[38]. However, the large atomic radius and steric effect of Ni, the low specific surface area, the lack of unsaturated redox active sites and the small control of surface capacitance are not conducive to the migration and diffusion of electrons/ions, which makes it difficult to maintain the original nanostructure, and the reversible capacity and rate performance are not ideal. On the basis of choosing low-cost terephthalic acid as the ligand, the Ni-MOF synthesized by Zhang et al. By one-step hydrothermal method has a reversible specific capacity of 620 mAh/G after 100 cycles at 100 mAh/G, and has a high rate performance, but the mechanism of lithium storage and its influencing factors have not been studied in detail^[46]. Compared with carboxylate and/or pyridine ligands, MOFs bearing azo groups are chemically and thermally stable^[47]. The H₂Me₄bpz ligand linked to Ni²⁺ provides a strong support for the structural flexibility and stability of Ni-MOF(Ni-H₂Me₄bpz), and the volume change in the process of Li⁺ deintercalation is alleviated to some extent, and the original crystal structure is retained^[14]. Therefore, the specific capacity and cycling performance of the 2D ripple hierarchical structure Ni-H₂Me₄bpz porous anode are very outstanding. It can be seen that although Ni-MOFs materials can effectively alleviate the volume effect and improve the ion transport kinetics, they have obvious shortcomings such as insufficient thermal stability and low charge/discharge potential, and there is still a certain gap in meeting the practical application.

Cu-MOFs contain abundant active components for redox reactions, and the capacity reduction first occurs at the cluster (Cu/Cu²⁺), which may be related to the incomplete redox and insufficient structural stability of Cu²⁺ located at the metal cluster site. The high specific surface area Cu-BDC MOF([Cu₂(C₈H₄O₄)₄]_n) porous anode was prepared based on the optimized in situ electrochemical synthesis conditions, and the first cycle discharge capacity was 227 mAh/G, which was about 95% of the theoretical capacity. Compared with Zn₄O(1,3,5-tribenzoate)₂, the rate capability of MOFs anode with terephthalic acid ligand is higher^[48]. However, under the influence of a large number of pores, amorphization reaction and side reaction between Li⁺ and electrolyte, the initial charge/discharge capacity has a large difference^[7,30]. Among the alkali metal ions, Li⁺ binds to MOFs with the highest affinity, but the number of binding sites provided and the binding mode mainly depend on the coordination tendency with the central metal ion and the Li⁺ content. In Cu₃(BTC)₂, coordinatively unsaturated metal ions can bind to small molecules. The Cu₃(BTC)₂ anode prepared by solvothermal method for the first time has a reversible capacity of 740 mAh/G at 96 mA/G. The capacity retention was nearly 100% after 50 cycles at 383 mA/G^[49]. The results show that the charge storage mechanism may be related to the redox reaction of carboxylate ligands, which can not be explained by the conventional conversion mechanism. Hu Xiaoshi prepared the [Cu₂(cit)(H₂O)₂]_n anode material based on the modified hydrothermal method, and the reversible capacity (950 mAh/G) and cycling stability were higher than those of the previously reported Cu-MOFs^[50]. After the first discharge, the Cu²⁺ of the metal center is almost completely converted to elemental Cu, and the reversibility is gradually enhanced during the charge process, and Cu is oxidized to Cu_xO again, resulting in an abnormal increase in capacity.

In addition to the above transition metal MOFs, researchers have also made useful attempts on other metal-based MOFs anode materials. Wu et al. First synthesized Sn-MOF(C₂₄H₁₆Li₂O₁₄Sn) polyhedra composed of Sn²⁺ and phthalic acid anions by one-step reflux method^[51]. Due to the effective mitigation of structural collapse caused by volume expansion/contraction during cycling and the maintenance of structural stability, this Sn-MOF anode exhibits a high capacity (610 mAh/G) and cycling stability, with a capacity retention of more than 92% even after 200 cycles. On this basis, Wu et al. Further compared the effects of preparation methods on the morphology and lithium storage performance of Sn-MOF anode^[52]. Different from the hydrothermal method, the Sn-MOF prepared by the reflux method has more uniform morphology, smaller particle size, larger specific surface area and lower charge exchange resistance, which can provide higher lithium storage activity and faster Li⁺ diffusion kinetics, and the reversible capacity and coulombic efficiency are correspondingly higher. The agaric-like hierarchical Al-FumA MOFs anode is composed of intertwined ultrathin nanosheets, with a reversible capacity of 392 mAh/G at 37.5 mA/G and no capacity degradation after 100 cycles. The capacity is 258 mAh/G at 37.5 A/G^[53]. Low impedance, fast and reversible deintercalation/intercalation of Li⁺, and structural stability are important factors that determine the good lithium storage performance of Al-FumA MOFs. In addition, whether the framework structure is beneficial to Li⁺ transport and the organic part plays a decisive role in the lithium storage mechanism, or the high thermal stability and specific surface area are beneficial to the reversible migration and storage of Li⁺, the three-dimensional porous Pb-MOF or Cd-MOF anode materials synthesized by solvothermal method show good cycle stability and rate performance^[16,18]. The new Ti-MOF anode has an initial discharge capacity of 1590.24 mAh/G at 100 mA/G, and especially maintains good reversibility after 8000 cycles^[54]. Density functional theory calculations prove that the carboxyl functional group acts as an electrochemical active site to reversibly bind to Li⁺, and the Ti octahedron acts as a stabilizing framework. This is attributed to the stable porous structure and sufficient redox sites of Ti-MOF, but the whole electrochemical process is mainly controlled by the diffusion of Li⁺, which leads to the decrease of rate capability, and it is necessary to increase the pore size and select small molecular electrolyte solvents to improve the capacity and pseudocapacitive behavior.

A large number of studies have shown that monometallic MOFs are generally difficult to meet the requirements of cyclicity and structural stability simultaneously. Bimetallic MOFs or metal-based MOFs mixtures can exert the synergistic effect between different metal ions to obtain more excellent comprehensive properties. For example, Han et al. Prepared Li/Ni-NTC anode materials by rheological phase reaction, which combined the respective advantages of Ni-NTC and Li-NTC in conductivity and structural stability, and its electrochemical capacity and cycle stability were significantly improved, providing an important basis for the development of a new generation of MOFs anode materials by designing and optimizing the molecular structure of aromatic derivatives^[55]. Saravanan et al. Synthesized Zn_1.5Co_1.5(HCO₂)₆ with Zn²⁺ instead of Co²⁺, which significantly reduced the amount of expensive and toxic Co²⁺, improved the capacity and cycle stability of electrode materials under the combined action of conversion and alloying/dealloying reactions and formate, and provided a simple and friendly synthesis method for the preparation of MOFs electrode materials with long life and high thermal stability^[30].

Some literature reported bulk anode materials of MOFs and their electrochemical performances are summarized in Table 1. Although some progress has been made in the research of MOFs bulk materials directly used as negative electrodes for lithium-ion batteries, there are still some problems to be solved. On the one hand, metal-based MOFs themselves are not conductive enough. On the other hand, the structural controllability of metal-based MOFs is poor due to the constraints of preparation methods. In most cases, only a single metal or organic ligand participates in the electrochemical reaction, which can easily lead to the structural collapse of metal-based MOFs. These factors directly or indirectly affect the reversible capacity, cycle stability and rate capability. Therefore, in the preparation of MOFs anode materials, not only the main structure should be kept basically unchanged, but also ligands with active functional groups (such as carboxyl, amino, benzene ring, etc.) And variable valence metals should be selected to increase the number of active sites for lithium storage and theoretical specific capacity, and the porous framework structure should be fully utilized to accelerate Li⁺ transport and reversible intercalation/deintercalation.

Conductive MOF materials have high conductivity, specific surface area and abundant active sites. Compared with other traditional carbon materials and metal oxide materials, the pore structure has more obvious advantages. Cu-CAT nanowires synthesized by solvothermal method show high Li⁺ diffusivity and conductivity^[56]. The specific capacity of the anode material is about 631 mAh/G at a current density of 0. 2 A/G, even close to 381 mAh/G at a high rate of 2 A/G, and the capacity retention rate is about 81% after 500 cycles at 5 A/G. In particular, the energy density of the Cu-CAT nanowire-based full cell is as high as 275 Wh/kg. Similarly, the diffusion coefficient of Li⁺ in the deintercalation process of one-dimensional Ni-CAT nanorods synthesized by hydrothermal method reaches 10^-9~10^-10cm²/s, and the discharge capacity is about 889 mAh/G at the current density of 0.1 A/G, and its rate capability and cycle stability are also quite outstanding, as shown in Fig. 2^[57]. A two-dimensional conducting MOF (Cu-HHTQ) facilitates the Li⁺ and electron transport and ensures the resilience of the electrode, and has a discharge capacity of about 657. 6 mAh/G at 0. 6 a/G^[58]. Half-cells and full-cells constructed with Co-CAT anode materials also exhibit excellent lithium storage performance, which is mainly attributed to the conductive nature of the MOF structure contributing to charge migration, electrolyte penetration, and regulation of volume change^[59]. These studies show that conductive MOF is a very potential anode material, and it is necessary to further select low-cost organic ligands and improve product stability to accelerate the application and development of this new advanced energy storage material.

Considering the limited lithium storage capacity of MOFs bulk anode materials, MOFs derivatives have become a hot research topic in recent years. These derivatives are new functional materials prepared by different synthetic methods using MOFs as templates or precursors, mainly including metal oxides, porous carbon, other metal compounds and nanocomposites synthesized on the basis of them. Due to the high controllability of synthesis conditions and chemical reaction system, the product composition is more uniform, and the structure and morphology are more diverse, these MOFs derivatives and their composites can play a more advantageous role and obtain the expected performance when used as anode materials for lithium-ion batteries.

It is well known that the advantage of metal oxide electrode materials lies in the chemical transformation during charge and discharge, that is, one or more electrons exchange and replace the intercalated Li⁺. The research on monometallic oxides derived from MOFs as precursors or templates started earlier, and the synthesis methods are relatively mature, and the process routes are simple and easy to control. In particular, low-density MOFs are often accompanied by significant volume shrinkage during transformation, which is more suitable for the synthesis of hollow nano-functional materials. Common MOFs-derived mono-oxide metal oxide anode materials include NiO, Co₃O₄, and CuO, etc., and the effect of metal type on their electrochemical performance is particularly obvious.

The reversible capacity of CuO nanoparticles prepared with copper phenylalanine as the precursor is 505. 3 mAh/G after 200 cycles at 100 mA/G, but the capacity is only 111. 6 mAh/G at 1000 mA/G, which shows that the rate capability is low^[61]. Wu et al. Synthesized porous CuO hollow octahedra using Cu-MOF (Cu-btc) as a template^[62]. In addition to the contribution of the porous hollow structure, wide pore size distribution and appropriate specific surface area (49.6 m²/g) to the cycling stability and rate capability, the { 111 } facet is more favorable for Li⁺ transport than other facets, that is, the facet structure also affects the electrode performance^[63]. Although the reversible capacity (538 mA · H/G) of nanostructured spherical CuO synthesized based on pyrolytic Cu-MOF (MOF-199) is significantly better than the reported results, further regulation and optimization of the process are needed to meet the requirements of anode materials^[64]. Obviously, the morphology and porosity of these copper oxides change with the subtle changes of MOF template and pyrolysis process, and the power density or cycle life is very limited. In addition, due to the retention of the morphology and structural characteristics of Cu-BTC, the stability of the main structure is not destroyed during cycling, and the charge storage process controlled by the surface redox pseudocapacitive behavior makes the CuO octahedral anode show ultra-high capacity and rate performance.This pseudocapacitive property is related to the ultrathin fine two-dimensional nanosheet subunit, while the hierarchical porous structure is a prerequisite for the extrinsic pseudocapacitance in the nanosheet subunit^[65]. More importantly, this room temperature solid-state transformation method significantly reduces energy consumption and avoids the release of toxic gases, which is a green synthesis strategy to accelerate the development of micro/nano-structured metal oxide anodes.

The electrode potential of Mn₂O₃ is low and the energy density is high, but the low electrochemical activity and high volume change in the process of Li⁺ intercalation/deintercalation directly affect its use as an anode material. Hierarchical structure Mn₂O₃ porous octahedron is composed of nano-sized units, which can reduce the interfacial contact resistance and avoid agglomeration during cycling, and the higher specific surface area and internal volume in the mesoporous structure provide more active sites and shorten the Li + intercalation/deintercalation diffusion distance^[66]. Affected by the formation of SEI layer and electrolyte decomposition, the Mn-LCP template-based Mn₂O₃ porous anode has a low initial coulombic efficiency (74%), but it has an advantage over micron-sized Mn₂O₃ in charge exchange kinetics, which indicates that the appropriate pore structure can significantly improve the cycle stability and rate capability^[67]. Considering that bulk or common large hollow structural materials are often not conducive to alleviating volume expansion due to lack of internal space or too large internal space, Cao et al. Designed and synthesized a mini-hollow Mn₂O₃ polyhedron.In the internal small pores of the hierarchical nanostructure, the nanoparticles rearranged due to the nanosize effect are uniformly dispersed, which induces more active sites, improves the reaction kinetics and forms more oxidation products^[68]. Therefore, it is very important to reasonably control the size of the hollow structure to improve the electrochemical performance of the conversion mechanism anode materials. The hollow Mn₂O₃ microspheres with hierarchical porous shell structure were generated from the non-uniform shrinkage caused by non-equilibrium heat treatment, and the electrochemical performance changed with the heating rate, although it was improved^[69]. Hu et al. Prepared mesoporous MnO_x micro-cuboids using Mn-MOF-74 as a precursor^[70]. On the one hand, the special 3D hierarchical structure promotes ion migration and reduces the exchange charge resistance. On the other hand, high specific surface area, electrochemical activity and structural stability lay the foundation for good lithium storage performance. Obviously, high-performance metal oxide electrode materials with complex 3D micro/nano structures can be prepared by morphology and dimensionality control, and some breakthroughs have been made in low-cost and pollution-free synthesis technology.

Co₃O₄ with high theoretical capacity is the most concerned Co-based oxide anode material, but it is also plagued by problems such as volume expansion during cycling. Pyrolytic MOFs materials can effectively reduce the surface area, increase the pore size, improve the conductivity and keep the diffusion path unobstructed. Although the specific surface area of agglomerated Co₃O₄ nanoparticles synthesized by Co-MOF([Co₃(HCOO)₆]·(DMF)₄) for the first time is low, the capacity, rate capability and cycle life are obviously improved, and the electrode performance can be further reasonably optimized by agglomerate structure control with variable primary and secondary particle sizes^[71]. Li et al. Employed a template-free self-assembly strategy to fabricate one-dimensional hierarchically porous Co₃O₄ nanorods composed of interconnected nanoparticles^[72]. The modification of nanoparticles and the hierarchical porous structure can shorten the diffusion distance of ions and charges and alleviate the volume change. At the same time, one-dimensional nanorods provide more exposed active sites and larger electrode/electrolyte contact area. Therefore, the capacity of the anode material is 628 mAh/G after 350 cycles at 1000 mA/G; After more than 600 cycles at 5000 mA/G, the capacity fade was only 0.068% per cycle. Compared with bulk materials, porous Co₃O₄ wrinkled nanosheets have higher aspect ratio and surface area, and more available active sites, all of which are beneficial to their use as anode materials with high specific capacity and rate capability^[73]. In addition, the porous Co₃O₄ cuboid anode synthesized by direct pyrolysis of Co-BTC precursor is also improved in lithium storage and rate capability^[74]. Han et al believed that the parallelepiped of porous hollow Co₃O₄ composed of uniform and continuous nanoparticles could significantly improve the reversible capacity, rate performance and cycle stability.The importance of three-dimensional nanostructured electrochemically active materials in the field of energy storage is emphasized, and this simple solid phase conversion method can also be used to prepare other porous hollow metal oxides with fine morphology^[75]. In order to overcome the shortcomings of traditional Co₃O₄ electrodes, Hu et al. Fabricated Co₃O₄ nanocages composed of porous shells and hollow structures loaded with a large number of nanoparticles using the non-equilibrium interdiffusion (Kirkendall effect) of Prussian blue analogue (PBA,Co₃[Co(CN)₆]₂)^[76]. Compared with the dense hollow structure, this unique structure is more conducive to the intercalation/deintercalation of Li⁺ in the active material and accelerates the diffusion rate, and the capacity is stable at 1465 mAh/G after 50 cycles at a high current density (300 mA/G). Mesoporous nanostructured Co₃O₄ was synthesized from MOF-71 template by low temperature (300 ℃) pyrolysis and long time (12 H) reaction, and its electrochemical performance was significantly improved by reasonable pore volume, high specific surface area and low particle size^[77]. Gou et al. Further proved that only appropriate temperature and time can remove the ligand and obtain unique morphology and structure during the pyrolysis of Co-MOFs precursor^[78]. With the combined effect of mesopores and nanoparticles, the process-optimized Co₃O₄ anode has a reversible capacity of 1058.9 mAh/G after 100 cycles at 200 mA/G; The capacity is maintained at 348 mAh/G at 2000 mA/G. In addition, under the same pyrolysis process, porous MOFs derivatives often have different properties due to the retention of precursor morphology. For example, the electrochemical performance of the twin-hemispherical Co₃O₄ anode is more outstanding than that of the flower-shaped Co₃O₄ with a relatively large surface area, and the reversible capacity is also higher than that of the commercial graphite anode (372 mAh/G), due to the smaller particle size of the uniform nanoparticles, the appropriate pore size, and the denser hierarchical structure^[79]. There is also a view that the final morphology of metal oxides derived from pyrolytic MOFs is related to factors such as organic chain type, metal ions and organic amines, and does not depend on their initial morphology^[80]. In contrast, different kinds of Co₃O₄ hollow dodecahedra were derived due to the change of synthesis process even if ZIF-67 precursor was selected, and the formation mechanism was similar to that of core-shell microspheres synthesized by pyrolysis of solid precursor^[81][82]. Due to the advantages of the size, specific surface area and structure of the primary nanoparticles, the utilization of the active material is significantly improved, and the reversible capacity of the hollow dodecahedral mid-sphere Co₃O₄ anode is as high as 1550 mAh/G, and the cycling stability is ideal (1335 mAh/G after 100 cycles). This similar phenomenon has been observed in Co₃O₄ hollow microspheres with complex internal structure^[83]. Hexagonal Co₃O₄ nanorings are pyrolytic derivatives of Co-MOFs assisted by organic amines, and the final morphology is determined by the Co²⁺ release rate of the reaction system and the steric hindrance of the organic chains^[80]. Compared with commercial Co₃O₄ particles, the unique morphology and single crystal characteristics make the reversible capacity and stability of hexagonal Co₃O₄ nanorings anode higher.

Iron oxide has been widely studied as an anode due to its stable structure and high theoretical capacity (1000 mAh/G), but its low conductivity and cycle stability restrict its development. Spindle-shaped α-Fe₂O₃ anode materials can be prepared by using Fe-MOFs as templates or precursors, but different preparation methods will cause changes in the electrode structure^[84,85]. For example, the product of the one-step method is a spherical α-Fe₂O₃ with nanometer size; The products of the two-step method are mesoporous materials composed of small-sized α-Fe₂O₃ nanoparticles, which provide larger specific surface area and more electrochemical active sites, and greatly shorten the Li⁺ diffusion distance, thus significantly improving the lithium storage performance. With the help of microwave irradiation for homogeneous nucleation and the inherent framework structure and porosity of MOFs (MIL-53-Fe) template, Guo et al. Synthesized mesoporous nanostructured Fe₂O₃ with different morphologies by simply adjusting the irradiation time^[86]. Among them, the porous core-shell structure of Fe₂O₃ octahedron performs well in terms of reversible capacity, cycle life and rate capability due to its special structure, which is significantly better than the porous and hollow anode materials with single structure. Metal-based nanomaterials with hollow structures can effectively alleviate the accompanying volume change during the conversion of low-density MOFs into metal oxides. Zhang et al. Prepared Fe₂O₃ microboxes with different shell structures by using the characteristics of Prussian blue cube oxidative decomposition and synchronous growth of iron oxide shell^[87]. Due to the hierarchical hollow porous structure with high crystallinity and structural stability, which helps to maintain the initial nanostructure during cycling, the capacity and cycling stability of this Fe₂O₃ microbox anode are high. Compared with the widely used liquid phase method, this solid phase method is easier to synthesize uniform anisotropic hollow structures on a large scale, but the morphology and structural characteristics of the final products are often different due to the change of calcination temperature. From this point of view, further exploring the coating of the internal space and the construction of core-shell or other hollow structure anode materials are conducive to the realization of structural efficiency and value.

Soundharrajan et al. Found that the reversible capacity of NiO nanoparticle anode was 748 and 410 mAh/G after 100 cycles at 500 and 1000 mA/G, respectively, which mainly depends on its particle size, surface area and structural stability, and its comprehensive performance is comparable to that of electrode materials with other supporting materials (such as graphene or graphene oxide)^[88]. Although the capacity of the nano/microstructured NiO porous anode is high (about 400 mAh/G), attention should be paid to the effects of calcination temperature and time on its structure and morphology^[89]. In addition, restricted by the poor stability of crystal structure, small contact area with electrolyte and low reactivity with Li⁺, the NiO anode derived by two-step calcination of Ni-MOF(Ni-H₂Me₄bpz) leads to capacity fading and reduced cycle performance due to pulverization, while the porous hollow NiO nanospheres can accelerate the volume change in the process of Li⁺ intercalation/deintercalation kinetics and coordination conversion^[14]. The preparation methods of inorganic hollow nano-structured materials generally include Kirkendall effect, electrochemical replacement, chemical etching and template methods. Among them, the template method is an important means to prepare hollow metal oxides with different morphologies by using MOFs. For example, NiO quasi-nanospheres derived from Ni-MOF are typical porous hollow structures, which are composed of a large number of self-aggregation particles with an average diameter of about 40 nm^[90]. As an anode material, it has a reversible capacity of 760 mAh/G after 100 cycles at 200 mA/G, which is even higher than the theoretical capacity of NiO electrode (718 mAh/G) based on the conversion mechanism. These additional capacities are due to the Faradaic effect of electrolyte side reactions and/or the pseudocapacitive behavior of reversible formation of polymer-like films. More importantly, the capacity is still 392 mAh/G at a high current density (3200 mA/G).

Compared with other traditional metal oxides, TiO₂ electrodes are prone to structural collapse and do not form SEI layers during charge-discharge cycling, especially the low rate performance caused by low conductivity and volume effect, which directly affects their industrial application^{[91⇓⇓~94]}. Within the MOFs crystal, the metal and oxygen atoms are arranged in periodic atomic levels, which can be completely transformed into metal oxides without long-range atomic migration and retain permanent pores to some extent. At the same time, the typical hydrophobic pores generated by calcination are conducive to electrolyte penetration and Li⁺ diffusion, which are the basic conditions for the preparation of new porous metal oxide anode materials by MOFs template method, and the technical route is completely different from the traditional inorganic materials synthesis method. For example, Wang et al. First used MIL-125 (Ti) as a precursor to prepare anatase TiO₂ porous anode, and the reversible capacity and rate capability were significantly improved, which provided a new idea for exploring new MOFs-derived TiO₂ materials with different structures and Ti contents^[95]. In addition, the 2D GeO₂ nanosheets also showed an ultrahigh lithium storage capacity during long-term cycling (1393 mAh/G reversible capacity after 350 cycles at 100 mA/G)^[96].

Generally speaking, most of the traditional bimetallic oxide anodes belong to the spinel structure, which mainly improves the performance by the synergistic effect of transition metal atoms with different activities.However, the problems of poor conductivity, low activation energy of electron transfer and ion diffusion coefficient, few vacancies to store Li⁺ and high polarization potential during cycling have seriously restricted its large-scale application. MOFs-derived bimetallic oxides containing different metal ions are expected to solve these problems to some extent. It should be emphasized that these bimetallic oxides are not a mixture of two metal oxides in the physical sense, and there may be a constraint effect between the metal elements. Therefore, in addition to the special porous structure and abundant redox active sites, bimetallic oxides are more conductive than monometallic oxides, with significantly shorter electron transport pathways, faster Li⁺ diffusion and migration, higher reversible capacity, and stable operation even at high current densities^[97]. For example, bimetallic oxides such as ZnFe₂O₄, CoFe₂O₄, and CuFe₂O₄ are only partially reduced during cycling, followed by reoxidation to improve the reversible capacity^{[98][99][100]}.

Benefiting from the unique porous structure and chemical composition, as well as the rapid migration of a large number of Li⁺ and electrons, the electrochemical and rate performances of MOFs-based derived MnCo₂O₄ nanowires are superior to those of traditional spinel-structured MnCo₂O₄ in the long-term charge-discharge process, and the preparation method is also significantly improved^[101]. The specific surface area of two-dimensional nanostructured materials is larger, which has more advantages in alleviating volume change, exposing active surface and shortening the diffusion distance of Li⁺ and electrons. Wang et al. Partially replaced Co in Co₃O₄ with Zn and successfully synthesized Zn_xCo_3-xO₄ nanosheets to improve the lithium storage performance by adjusting the metal element ratio and rationally designing the porous structure, combined with two-dimensional morphology and nano-sized building units^[97]. Similarly, the porous ZnCo₂O₄ nanosheet anode has a reversible capacity of 1640.8 mAh/G after 50 cycles at 100 mA/G; The discharge capacity is still 581.3 mAh/G after 190 cycles at 1500 mA/G^[102]. This structural design method can not only greatly reduce the cost and toxicity of electrode materials, but also be used to prepare two-dimensional porous bimetallic oxide anodes derived from other MOFs, which is expected to achieve commercial applications. With the help of interconnected porous structure, structural stability enhanced by a large number of mesopores, high specific surface area formed by gas evolution, and retained morphological characteristics of MOFs, the capacity of Mn_1.8Fe_1.2O₄ nanocubes was 827 mAh/G after 60 cycles at 200 mA/G^[103]. The mixed valence state (Fe²⁺↔Fe³⁺ and Mn²⁺↔Mn³⁺) of the multimetallic center provides the expected conductivity for effective energy storage, probably due to the defect effect mechanism. This obvious difference just proves the significance of lowering the synthesis temperature for the bimetallic oxide anode. Hollow structural materials generally contain controllable internal pores and functional shells, but most of the single-shell structures are often too simple and relatively single in composition, which makes it difficult to maintain structural integrity in the process of repeated insertion/extraction of Li⁺, which inevitably limits the improvement of cycle performance of MOF-derived hollow metal oxides. Therefore, reasonable structural design of MOF-derived materials is the development direction of high-performance anode. Wu et al. Used bimetallic MOF (Ni-Co-BTC) as a precursor to prepare spinel-type Ni_xCo_3-xO₄ multi-shelled hollow spheres composed of nanosized subunits in the shell and large pore spaces between adjacent shells by a self-sacrificial template method^[104]. The reversible capacity is 1109.8 mAh/G after 100 cycles at 100 mA/G, and the rate capability is outstanding. In addition, adjusting the chemical composition can also improve the electrochemical performance of the Ni_xCo_3-xO₄ anode. Although the synthesis technology of hollow spherical bimetallic oxides has been relatively mature, the design and preparation of non-spherical hollow complex bimetallic oxides still face challenges. Wu et al. Prepared porous spinel-type Zn_xCo_3-xO₄ hollow polyhedron anode materials by pyrolysis-induced transformation of heterogeneous bimetallic Zn-Co-ZIF^[105]. On the one hand, the high-symmetry hollow structure composed of nano-sized subunits in the shell can provide more degrees of freedom and form atomic steps, rather than asymmetric structures. The high-density atomic steps on these atomic planes are beneficial to the reaction between Li and Zn_xCo_3-xO₄, so the cycle stability and rate capability are good. On the other hand, the hollow structure effectively suppresses the initial nanoparticle aggregation and dissolution in the electrolyte. Therefore, this low-cost and simple method is expected to be used to prepare other hetero-hollow structured bimetallic metal oxides and construct advanced energy conversion and storage electrodes.

As mentioned above, transition metal oxides often face the problems of electrode pulverization caused by volume effect, rapid capacity decay and reduced cycle stability. In bimetallic oxides, metal ions react electrochemically in the matrix of different material systems, and the combination of two metals provides abundant redox reaction sites and improves conductivity. Under such conditions, the matrix, either active or inactive for lithium, undergoes a volume change in a progressive manner rather than in a fixed pattern, relying on the unreacted component to coordinate the strain produced by the reacted component. As a precursor or template for the preparation of new nano-functional materials, nano-MOFs have unique hollow structure and high porosity and specific surface area, which make it easier to select and control different metal ions and organic ligands. Especially in the conversion process, the pores and long-range ordered structure of nano-MOFs provide a convenient way for the rapid transport of small molecules and ions. Porous hollow CoFe₂O₄ nanocubes and NiFe₂O₄ nanocages were synthesized by using Co[Fe(CN)₆]_0.667 and Ni₂Fe(CN)₆ as precursors or templates, and the element hybridization characteristics made the two anode materials change in volume periodically during cycling, showing excellent performance in both rate capability and cycling stability without being affected by the increase of current density^[106,107]. Similarly, Co-Ni-O bimetallic oxide is a derivative obtained by further heat treatment based on the microwave-assisted synthesis of Co-Ni bimetallic MOF nanorods^[108]. For example, the reversible capacity of mesoporous Ni_0.3Co_2.7O₄ nanorods reached 1410 mAh/G after 200 cycles at a low current density (100 mA/G) and maintained a high reversible capacity at 5000 mA/G. The excellent electrochemical performance is mainly due to the synergistic effect between the interconnected mesoporous nanostructure and the two active metal oxide components (Co₃O₄ and NiO). Obviously, most of these bimetallic oxides are derived from bimetallic MOFs, in which the bimetallic elements include Co, Ni, Zn, Mn, Fe, etc. Different from this, Chen et al. Used MIL-25 (Ti-MOF) and LiNO₃ as titanium source and lithium source, respectively, to prepare porous Li₄Ti₅O₁₂ anode by solid state sintering method, which not only had good cycle stability and rate performance, but also provided a new idea for the preparation of other electrode materials^[109]. The precursor/template and main electrochemical performance of some MOFs-derived metal oxide anode materials are shown in Table 2.

A large number of studies have shown that metal oxides derived from MOFs are superior to traditional transition metal oxides in terms of specific capacity, cycle stability and rate capability. Inspired by this, many researchers have turned their attention to other metal compounds derived from MOFs, aiming to make full use of the advantages of MOFs and use simple preparation methods to construct nano-anodes with different structural characteristics to meet the needs of lithium-ion batteries in lithium storage capacity. Yu et al. Used a two-step diffusion-controlled method to synthesize hollow prisms composed of interconnected bubble-like nanosized CoS₂ subunits, and the lithium storage performance was significantly better than that of the previously reported Co sulfide-based anode^{[110][111,112]}. The ultra-thin-walled nanobubbles significantly shorten the diffusion distance of the Li⁺ and improve the electrochemical kinetics. More importantly, the multi-level hollow interior of the hierarchical prism can effectively coordinate the structural stress during the repeated charging and discharging process. Metal phosphides have lower redox potentials and higher theoretical capacities (e.g., 926 and 894 mAh/G for FeP and CoP, respectively). The bimetallic NiCoP nanocage retains the quasi-polyhedral characteristics of the ZIF-67 template, and the inner and outer shells are composed of many wrinkled nanosheets with a large number of micropores on the surface and connected with each other. This multi-channel structure effectively shortens the diffusion distance of ions and electrons and relieves the diffusion-induced stress. At the same time, the free space between adjacent nanosheets can coordinate the stress change caused by the volume change^[113]. Therefore, the NiCoP anode material has a complete structure without agglomeration, shrinkage and collapse during cycling, and has good reversible capacity, rate capability and cycle stability.

Only when the electrode reaction is completely reversible during the charge-discharge process, the bulk materials of MOFs directly used as the negative electrode can achieve ideal cycle stability. However, the poor conductivity of MOFs will inevitably affect their performance to a certain extent. Cao Na et al. Used liquid phase deposition method to make MOFs grow in situ on carbon cloth, and prepared Co-MOFs/CF composites without binder^[136]. The initial discharge specific capacity of the anode was 1621.3 mAh/G at 50 mA/G, and the specific capacity remained at 445.1 mAh/G even after 100 charge-discharge cycles, which was related to the synergistic effect between the hierarchical structure of MOFs and the highly conductive carbon fibers, which was beneficial to ion migration and storage. Li et al. Synthesized MIL-101 (Cr)/graphene oxide composites by a simple solvothermal method, that is, MIL-101 (Cr) octahedra were uniformly distributed on the surface of graphene oxide with hierarchical structure, but the combination of graphene oxide did not affect the formation of MIL-102 (Cr)^[137]. The initial discharge capacity of MIL-101 (Cr)/GO anode is almost 2 times higher than that of MIL-101 (Cr) due to the increase of contact area between electrode and electrolyte caused by high specific surface area. The porous structure and the conductive graphene significantly shorten the diffusion time and accelerate the phase change reaction, the agglomeration of the electrode material and the pulverization of the active material caused by volume expansion/contraction in the charge-discharge process are also effectively inhibited, and the rate performance and cycle stability are significantly improved. On the basis of Cu-MOF prepared by coprecipitation method, Gao Guoliang et al. Synthesized Cu-MOF/rGO composite in situ, using carbonyl as binding site to realize reversible intercalation/deintercalation of Li^+[138]. The specific discharge capacity of the anode material is 520 mAh/G after 50 charge-discharge cycles at 50 mA/G. In addition to its own good conductivity, rGO also avoids the direct contact between the electrolyte and Cu-MOF to some extent, inhibiting the SEI layer thickening. In addition, the flexibility of rGO also improves the volume effect and capacity fading caused by the lack of structural stability of MOF, and achieves good results in improving the performance of electrode materials. Considering that F contributes to the effective nucleation of MOFs and improves the reversibility of lithiation/delithiation, Wei et al. Realized the in situ self-assembly of hollow urchin-like F-Co-MOF with rGO via a solvothermal method^[139][140]. The reversible capacity of F-Co-MOF/rGO anode is 1202.0 mAh/G at 100 mA/G and the coulombic efficiency is 99.95% after 50 cycles due to the synergistic effect and the promotion of Li⁺ intercalation/deintercalation. Especially, the full cell composed of nanostructured F-Co-MOF/rGO anode and LiFePO₄ also exhibited good cycling stability. Jin et al. Prepared Fe-MOF/RGO composites composed of two-dimensional RGO nanosheets and their coated Fe-MOF octahedral particles by solvothermal method, and focused on the effect of RGO content on their lithium storage capacity^[141]. The results show that only the Fe-MOF/RGO (5%) anode exhibits the best reversible capacity (2259.3 mAh/G at 500 mA/G), rate capability, and cycle stability. The electrochemical performance of MIL-53 (Fe) anode material mainly depends on the redox reaction of Fe³⁺, but the terephthalic acid ligand does not exert its electrochemical activity due to poor conductivity and thick surface SEI layer. Benefiting from the chemical passivity and surface hydrophilicity of RGO, which can reduce the contact area between the electrolyte and MIL-53 (Fe), avoid the surface reaction of the electrolyte, activate the carboxyl functional groups and improve the conductivity, the three-dimensional porous MIL-53 (Fe) @ RGO anode has better reversible specific capacity and rate capability than MIL-53^[142]. However, due to the limited metal chelating ability of RGO and the absence of any carboxyl functional group on the basal plane, the rate performance improvement of Fe-MOF/RGO anode is not obvious. Co₂(OH)₂BDC is a typical MOF-based electrode material for aromatic chelating ligand controlled Li⁺ intercalation/deintercalation reaction^[10]. Considering that abundant carboxyl functional groups can significantly improve the metal chelation strength of MOF and graphene and form a suitable three-dimensional structure, Li et al. Synthesized covalently reinforced Co₂(OH)₂BDC/ carboxylated graphene (CoCGr) composites by surfactant-free solvothermal method^[143]. The outstanding electrochemical performance of CoCGr-5 anode is mainly attributed to the Li⁺ intercalation/deintercalation reaction controlled by the organic part and the bicontinuous electron/ion transfer path between Co-BDC and CGr. Meanwhile, the CGr loading not only affects the specific surface area and pore volume of CoCGr anode, but also affects the rate capability and cycle stability. In addition, the surface lithium storage reaction is Li⁺ fast intercalation/deintercalation process, which can also provide additional capacity and improve the rate performance^[144]. Using H₂BDC non-covalently functionalized graphene oxide as nucleation sites and structure-directed templates to promote the growth of MOFs, He et al. Successfully fabricated MOF/RGO_n composites^[145]. The reduction of transition metal ions (Mn²⁺) and the subsequent lithiation of organic ligands are the direct reasons for the higher reversible capacity of the MOF/RGO_n anode, and the long-range conducting RGO network can also ensure efficient electron transport and provide mechanical flexibility to the MOF, avoiding drastic volume changes and self-aggregation. Therefore, the electrochemical performance of MOFs-based electrode materials can be improved by relying on other electron-withdrawing groups to bridge organic ligands and metal ions^[146]. The electrochemical performances of MOFs/porous carbon composite anode materials are summarized in Table 4.

Core-shell structure is a typical core-shell structure and partial hollow structure in which one component material is coated with another component material and assembled in a hierarchical and orderly manner. In a broad sense, it is a general term for a typical core-shell structure with no gap between the core and the shell^[147]. The advantages of core-shell MOFs-based composites are: (1) high conductivity of the core material; (2) the porous core-shell structure effectively slows down the volume expansion in the charge-discharge cycle process and provides a certain buffer space; (3) Avoid direct contact of the core material with the electrolyte. Therefore, the reasonable selection of MOFs shell and core materials and the precise control of the size, morphology, thickness and distribution of the shell are undoubtedly very beneficial to improve the structural stability and electrochemical performance of MOFs-based nano-anode materials. The Fe₃O₄@MOF composite synthesized by the hierarchical assembly method consists of a Fe₃O₄ microsphere core with a porous MOFs (HKUST-1) shell^[148]. After 100 charge-discharge cycles at 100 mA/G, the reversible specific capacity of the composite anode is about 1001. 5 mAh/G, which is significantly better than that of pure Fe₃O₄(696 mAh/g). Among the known MOFs, HKUST-1 with π-conjugated benzene ring and carboxylate has high conductivity and capacity retention, and can also provide additional capacity^[34]. HKUST-1 is beneficial to electron and Li⁺ migration, and the excellent rate performance of the Fe₃O₄@MOF composite anode must be related to the improved reaction kinetics. The inner core can be CNTs, metal/nonmetal oxides, organics, and MOFs because it is not limited by composition, dimensionality, size (tens of nanometers to several micrometers), and shape (one-dimensional nanowires to three-dimensional hollow spheres). This opens up the possibility of hybrid hybridization of the ZIF-8 shell with different core materials and obtaining the expected performance. The Zn₂SnO₄/ZIF-8 nanocomposite anode consists of a Zn₂SnO₄ nanoparticle core and a ZIF-8 shell on its surface^[149]. In the process of Li-Sn alloying/dealloying, the nanoporous properties release the stress caused by severe volume expansion, while the conductivity and thickness of ZIF-8 shell directly affect its cycle stability. In addition, MOFs can also serve as growth templates for internal active nanoparticles in core-shell structures. For example, Wang et al. Utilized the porosity of MIL-101 (Cr) to embed SnO₂ nanoparticles inside the pores^[150]. Due to the absence of solvent in the pores, the heat-treated multi-core/shell structured SnO₂@MIL-101(Cr) nanocomposite is a concave octahedron with a smooth surface. The MIL-101 (Cr) protective shell alleviated the volume change of SnO₂ and inhibited the aggregation of powdered nanoparticles, and the lithium storage performance of the SnO₂@MIL-101(Cr) composite anode was significantly improved. It can be seen that although the abundant pores and the redox reaction of metal ions are beneficial to Li⁺ intercalation/deintercalation, it is still necessary to further simplify the synthesis method and improve the conductivity of MOFs/metal oxide composite anode. Affected by the poor overlap of p and d orbitals of metal ions and insulating organic ligands, most MOFs without any structural modification are typically poor conductors. Controlling the composition and morphology and optimizing the rigid-flexible framework structure can improve the conductivity of MOFs. Wang et al. Prepared CuS @ Cu-BTC polyhedra by a simple sulfidation reaction, in which rod-like CuS was distributed on the surface of Cu-BTC^[151]. The active material CuS can improve the conductivity, while Cu-BTC can provide porosity and chemical stability, which can effectively inhibit the conversion of Li₂S into polysulfides by improving the reversibility of the conversion reaction. The capacity of CuS (70 wt%) @ Cu-BTC anode reached 1609 mAh/G (higher than the theoretical capacity) after 200 cycles at 100 mA/G. ~ 490 mAh/G at 1000 mA/G. It can be seen that the content of CuS is not only related to the sulfidation time, but also affects the porosity and morphology of CuS @ Cu-BTC composite anode, and then affects its lithium storage performance.

The conductivity of transition metal oxides is relatively low, which is easy to produce obvious volume effect and serious agglomeration in the cycle process, and even cause electrode pulverization and rapid capacity decay. Only when the lithium storage sites are abundant, the Li⁺ at the electrode/electrolyte interface and the adsorption/desorption table/interface migrate rapidly, and the diffusion distance of Li⁺/ electrons is shortened, the electrode materials can obtain good lithium storage performance. Metal oxide/metal oxide composites are generally composed of dual or multiple metal oxides, which can play the role of different metal oxides at the same time, and can also enhance the mechanical stability, electrochemical reactivity and ionic conductivity of each component by means of special structure and synergistic effect.This requires that the design and synthesis of metal oxide/metal oxide composites should not only consider the properties and functions of each component material, but also provide a conductive network for efficient electron and ion transport by delicately constructing and assembling porous structures to uniformly distribute them on the nanoscale. For example, hollow core-shell metal oxide/metal oxide nanocomposites with different morphologies can maintain the integrity of the electrode to the greatest extent, prevent capacity fading and prolong the cycle life. Therefore, metal oxide/metal oxide nanocomposites have become one of the research hotspots in the field of anode materials for lithium-ion batteries.

Using bimetallic Co-Ce-MOF as a precursor, Kang et al. Designed and synthesized a heterostructured Co₃O₄/CeO₂ composite^[152]. The Co₃O₄/CeO₂ (molar ratio of Co/Ce = 5 ∶ 1) anode has a high reversible capacity (1131.2 mAh/G after 100 cycles at 100 mA/G) and cycle stability (835.3 mAh/G at 2000 mA/G), and the good lithium storage ability is mainly due to the conversion reaction and interfacial capacitance. However, the introduction of semiconductor CeO₂ does not affect the diffusion of Li⁺ in the Co₃O₄, which provides a basis for the modification of electrode materials by rare earth components. The electrochemical performance of hollow MOFs derived anode is related to the interfacial interaction between the core and the shell and the stress distribution, but the effect of the expansion/contraction of the core on the structural stability can not be ignored. Zhang et al. Synthesized Fe₂O₃/SnO₂, Fe₂O₃/GeO₂, and Fe₂O₃/Al₂O₃ multi-component microboxes with complex hollow structures using a controlled template reaction method^[153]. Since the beneficial effect of the hollow hierarchical structure is maximized, the volume expansion/contraction is effectively suppressed, the diffusion distance of the Li⁺ is shortened, and the capacity retention rate of these multi-shell composite anodes is high. In addition to the known advantages of the hollow structure, the synergistic effect of the tightly combined Fe₂O₃ and SnO₂ can also provide additional capacity for the multi-component microbox anode, and the specific lithium storage mechanism needs to be further studied. Li et al. Prepared ZnO/NiO nanospheres composed of nanorod shell and microsphere core by controlling the calcination temperature. The unique core-shell structure retained from bimetallic Zn-Ni MOFs is beneficial to electrolyte diffusion and ion migration. The reasonably hybridized nanosized ZnO and NiO improve the electrochemical kinetics and reactivity with Li⁺, effectively suppressing the volume effect^[154]. Therefore, the high specific surface area ZnO/NiO composite anode shows excellent specific capacity (1008.6 mAh/G after 200 cycles at 0.1 A/G), rate capability (437.1 mAh/G at 2 A/G), and long-term cycling stability (592.4 mAh/G after 1000 cycles at 0.5 A/G). Xu et al. Prepared hollow Co₃O₄/TiO₂ polyhedra using cation exchange and subsequent heat treatment^[155]. During the synthesis, ZIF-67 served both as a host for the exchange of Ti⁺ and as a template for the formation of Co₃O₄/TiO₂ composites. The structurally stable TiO₂ continuously provides high capacity for Co₃O₄ and also provides a conductive path for fast electron transport. After 200 cycles at 500 mA/G, the reversible capacity of the Co₃O₄/TiO₂ anode is 662 mAh/G, which is about 96.9% of the initial capacity, and is significantly better than that of the pure Co₃O₄ anode with the same structure. In particular, this novel synthesis method can also be used to construct more complex nanostructured composites. Hu et al. Fabricated hollow octahedral CuO/Cu₂O composite by pyrolytic [Cu₃(btc)₂]_n template, and analyzed the formation process of porous hollow structure in combination with non-equilibrium interdiffusion, volume decrease, and gas release^[156]. According to the theoretical capacity of CuO (670 mAh/G) and Cu₂O(670 mAh/g), the conversion mechanism and the proportion of CuO, the theoretical capacity of CuO/Cu₂O composite anode was deduced to be 520 mAh/G.

Although the cyclicity and lithium storage performance of CuO/Cu₂O anode materials have been improved due to the hollow porous shell and different components of metal oxides, the problems of low inherent conductivity, poor transport kinetics and electrode pulverization have not been fundamentally solved. In order to solve the volume effect of metal oxide electrodes, Guo et al. Designed and synthesized CuO @ NiO hollow multilayer spheres derived from Cu-Ni-BTC, which are composed of a three-layer sphere-in-sphere structure and interconnected nanoparticles, and the elements Ni and Cu are distributed in a gradient from core to shell^[157]. After 200 charge-discharge cycles, the reversible capacity of CuO @ NiO composite anode is 1061 mAh/G, which is much higher than the theoretical capacity of CuO and NiO anode, which is mainly related to the stepwise intercalation reaction initiated by the commensurate composition from core to shell, the smooth Li⁺/ electron diffusion and the coordinated volume change. Metal-oxygen clusters in MOFs can be directly transformed into porous transition metal oxides without long-range atomic migration during pyrolysis. The Cr₂O₃@TiO₂ nanocomposite is a concave octahedron synthesized by calcination of the MIL-101(Cr)@TiO₂ template in argon and air sequentially, which is a typical multi-core core-shell structure^[158]. Due to the protective effect of the chemically stable TiO₂ shell, the pulverization of internal particles is effectively inhibited, and the structural stability of the electrode material is improved. The reversible capacity of the Cr₂O₃@TiO₂ composite anode is 510 mAh/G after 500 cycles at 0.5 mA/G, which is higher than that of nano Cr₂O₃(200 mAh/g). More importantly, its capacity recovery rate is also outstanding. The synthesis of mesoporous CuO@TiO₂ octahedra involves two stages: uniform growth of TiO₂ shell on Cu-MOF (HKSUT-1) template and controllable temperature pyrolysis^[159]. This core-shell structure design fully combines the advantages of different components and solves the pulverization problem of most metal oxide-based electrodes caused by severe volume change. After 200 cycles at 100 mA/G, the reversible specific capacity of the CuO@TiO₂ composite anode is 692 mAh/G, which is significantly better than that of pure CuO and the previously reported CuO-based nanomaterials. At high current densities of 1600 and 3200 mA/G, the capacities were still 441 and 387 mAh/G, respectively. Prussian blue and Prussian blue materials with FCC structure are common templates or precursors for the synthesis of metal oxides.Its derivatives generally have the characteristics of high specific surface area, connected internal pores and uniform element distribution from inside to outside, and low-cost composite materials with controllable composition and special morphology can be obtained by adjusting the preparation process and precursor composition. Porous Fe₂O₃-CuO cubes are composites prepared by pyrolysis of a bimetallic MOF synthesized based on Prussian blue (Fe₄[Fe(CN)₆]₃)^[160]. After 120 cycles at 500 mA/G, the specific capacity of the Fe₂O₃-CuO composite anode is 744 mAh/G; The discharge capacity at high current density (2000 mA/G) is still higher than 500 mAh/G. In particular, the cycle stability and the rate capability are obviously superior to those of other known hybrid transition metal oxides. Fig. 4 is a schematic diagram of the synthesis of the Fe₂O₃ nanotube @Co₃O₄ composite, and it can be seen that the Co₃O₄ host matrix is uniformly coated by the Fe₂O₃ nanotubes^[161].

The hollow structure electrode can effectively alleviate the stress-induced structural changes during the long-term electrochemical reaction. Compared with the single-shell hollow structure, the multi-shell hollow structure has the advantages of high mass fraction of active material, complex internal pores, large specific surface area and controllable shell composition. The multi-shell Co₃O₄@Co₃V₂O₈ nanobox is pyrolyzed from ZIF-67 @ amorphous -Co₃V₂O₈ in air, and the number and composition of its shells can be precisely controlled by simply adjusting the amount of VOT in the reaction system^[162]. Similarly, NiFe₂O₄/Fe₂O₃ hierarchical nanotubes and Fe₂O₃@NiCo₂O₄ porous nanocages were derived from the pyrolysis of core-shell structured MOF precursors (Fe₂Ni MIL-88/Fe MIL-88 and Co₃[Fe(CN)₆]₂@Ni₃[Co(CN)₆]₂, respectively)^[163,164]. Although the two composite anode materials do not retain the structural characteristics of the precursor, their electrochemical performance is significantly higher than that of the single component metal oxide derived from MOFs.

Zhong et al. Prepared nanostructured Co₃O₄-CoFe₂O₄ composites by pyrolyzing MOF-74-FeCo, whose morphology and composition mainly depend on the Fe³⁺/Co²⁺ molar ratio in the bimetallic MOF^[165]. Compared with pure Co₃O₄ and other Co₃O₄-CoFe₂O₄ composite anodes, only Co₃O₄-CoFe₂O₄-12 exhibited the highest initial discharge capacity (1328 mAh/G), reversible capacity (940 mAh/G after 80 cycles at 100 mA/G), and rate capability. This is because Fe³⁺ promotes the transformation of MOF-74-Co from an irregular morphology to a relatively regular spherical structure, resulting in the formation of fine and dispersed Co₃O₄-CoFe₂O₄ nanoparticles, whose phase and structural stability are not changed by repeated charge-discharge cycles, and the specific capacity hardly decays after 80 cycles. Xu et al. Successfully prepared ZnO/ZnCo₂O₄ nanosheets composed of both primary ZnO and ZnCo₂O₄ nanoparticles with uniform size and interconnection by one-step pyrolysis of Zn-Co-MOF precursor^[166]. A reversible capacity of 1016 mAh/G and a coulombic efficiency of almost 99% were obtained for the three-dimensional hierarchical ZnO/ZnCo₂O₄ porous anode after 250 cycles at 2 a/G. The capacity is still 630 mAh/G at 10 A/G. On the one hand, the size and morphology of MOFs core/shell composites are directly controlled by selecting the appropriate template. On the other hand, special functional groups are used to modify the surface of the core component to control the binding and coating of the MOFs core, and to inhibit the growth of MOFs on the precursor solution and the core surface. Benefiting from the multi-component functional complementation and structural characteristics, the specific capacity, cycle life, and rate capability of the Ni(OH)₂@ZIF-8(Zn,Co) nanowire-derived core-shell NiO@ZnCo₂O₄ composite anode are significantly improved, while this synthesis method provides the possibility for the MOFs shell to grow on the surface of core materials with arbitrary chemical composition, structure, dimensionality, and size^[167]. Hierarchical ZnO/ZnFe₂O₄ mesoporous submicron cubes are composed of uniformly dispersed ZnO and highly electrochemically active ZnFe₂O₄ nanophase^[168]. Compared with ZnFe₂O₄, the excellent electrochemical performance of the ZnO/ZnFe₂O₄ composite anode mainly depends on the unique structure and the synergistic effect of two-component ZnO and ZnFe₂O₄ active phase. The discharge capacity of the porous ZnO/ZnFe₂O₄ composite anode composed of nanoparticles with a particle size of about 5 nm was 804 mAh/G after 500 cycles at 1000 mA/G. 496 mAh/G even after 1000 cycles at 2000 mA/G^[169]. In addition to the porous nanostructure providing additional space to coordinate the volume change, the ZnO phase as a buffer not only spatially separates the coexisting ZnFe₂O₄ nanophase, but also further prevents its self-aggregation during cycling^[170]. The electrochemical properties of the metal oxide/metal oxide composite anode materials are summarized in detail in Table 5.

It is well known that the performance of electrode materials mainly depends on their basic properties, morphology and structure. Low intrinsic conductivity and obvious volume effect seriously restrict the application of transition metal oxides (such as Co, Ni and Fe) with high theoretical capacity in the field of anode materials. On the basis of fully retaining the activity of MOFs derivatives, combined with the advantages of carbon-based materials such as carbon nanotubes, graphene, carbon fibers and amorphous carbon in terms of stability and conductivity, different metal oxide/carbon-based materials can be directly synthesized by using MOFs containing metal ion centers and organic ligands as templates or precursors. Although the control process is complex and there are many influencing factors, the comprehensive performance of anode materials can be improved by optimizing the heat treatment process and accurately controlling the morphology and structure.

Compared with solid materials with the same size, low-density hollow structural materials have clear internal voids and higher specific surface area, especially the characteristics of difficult pulverization, coordinated volume change and micro/nano structure, which make them suitable for energy storage applications. The core-shell hierarchical porous Co₃O₄/C dodecahedral composite anode derived from ZIF-67 by two-step heat treatment exhibited a specific capacity of 1100 mAh/G after 120 cycles at 200 mA/G, which is much higher than the theoretical capacity of Co₃O₄ (890 mAh/G)^[171]. Kinetic analysis shows that most of the stored charge comes from the capacitive process, and the rate-determining step is surface confinement rather than solid-state diffusion^[172]. In addition, the improvement of electrode conductivity by carbon matrix cannot be ignored. In the solid-solid conversion process, the porosity and long-range order of the MOFs template facilitate the rapid entry and exit of small molecules and ions, and the temperature and atmosphere directly affect the structure and composition of the derivatives. MOFs-derived core-shell nanomaterials usually have high specific surface area and stable hollow structure, which will not affect the product morphology due to template removal. By fully combining the advantages of the retained core-shell structure of MOFs and the matrix of different material systems, the multi-component electrochemical reaction occurs on the hybrid matrix of different material systems, which effectively solves the serious pulverization of electrodes and the resulting rapid capacity decay. In this case, the defined matrix causes the volume change in a progressive manner rather than in a fixed pattern, i.e., the strain produced by the reacted phase is coordinated by the unreacted components. At the same time, the combination of carbon-based materials can also provide abundant redox reaction sites for transition metal oxides and improve conductivity^[173]. Similarly, both core-shell ZnO @ C and Co₃O₄@C polyhedra are derived from MOFs^[174]. With the synergistic effect of the three-dimensional porous structure and the surface conductive carbon layer, the reversible capacity of the ZnO @ C and Co₃O₄@C composite anode is 526 and 721 mAh/G after 500 cycles at 250 mA/G, respectively; More than 1000 charge-discharge cycles at 2.0 A/G. The kinetic analysis indicates that the surface-controlled pseudocapacitive behavior directly affects the lithium storage ability of the ZnO @ C and Co₃O₄@C composite anode. It is worth noting that the binding force of MOFs-derived metal oxide/carbon-based nanocomposites is stronger than that of metal oxide directly coated with carbon layer. The mesoporous spindle-shaped CuO/C composite prepared by one-step annealing well retains the Cu-MOF template morphology synthesized by microwave-assisted method^[175]. Although they are all typical hollow structures, the electrochemical performance of porous CuO/C microcube anode is far inferior to that of mesoporous spindle-shaped CuO/C composite anode due to the different composition of Cu-MOF template and pyrolysis temperature^[176]. Peng et al. Prepared mesoporous Mn₃O₄/C composite by one-step calcination of 3D Mn-PBA template and studied the effect of calcination temperature on its surface morphology^[177]. The results show that the reversible capacity and rate capability of the Mn₃O₄/C microsphere anode are significantly better than those of the sponge network Mn₃O₄/C anode^[178]. After 100 cycles at 50 and 500 mA/G, the specific capacity of the 3D rhombic MnO/C composite anode is 835 and 586 mAh/G, respectively, which is much higher than that of the porous carbon and Mn₂O₃ derived from the same Mn-MOFs^[179]. This is because of the beneficial effects of the uniformly distributed MnO nanoparticles on the surface and the conductive carbon matrix containing a large number of pores in increasing the active sites for lithium storage, shortening the ion and electron transport distance, avoiding agglomeration, and alleviating the volume effect. In addition, because the morphology of Mn (PTA) -MOFs changes with the molar ratio of Mn²⁺ and benzoic acid, Sun et al. Synthesized MnO/C composites with different morphologies on the basis of self-sacrifice and thermal conversion^[180]. Only the reversible capacity and rate capability of the spindle-shaped MnO/C microrods were higher, and the capacity did not decay significantly even after 200 cycles at a high current density of 1000 mA/G. Therefore, not only the reasonable selection of organic ligands and their dosage is very necessary, but also the difference of the connection mode between the same metal ion and different organic ligands affects the final morphology and spatial structure of metal oxide/carbon composites, and the electrochemical performance is also different.

Although the controlled heat treatment process can simultaneously synthesize ordered nanostructured transition metal-based nanoparticles and carbon frameworks, there are still many challenges in other methods to uniformly embed nanoparticles into porous carbon frameworks. The hierarchical ZnO/C hollow spheres were derived from the precursor Zn-BTC by one-step annealing treatment, and the ultra-small ZnO nanodots (average particle diameter < 5 nm) were uniformly dispersed in the carbon nanosheet matrix on the surface of the porous hollow structure^[181]. The reversible capacity of ZnO/C composite anode is 919 mAh/G after 100 cycles at 100 mA/G. The reversible capacity is maintained at 741 mAh/G after 120 cycles at 500 mA/G. It should be noted that the irreversible capacity fading at the first cycle is due to the irreversible reduction of Li⁺ during the formation of surface SEI layer, which is a common problem for most anode materials. The ZnOQDs @ C composite is obtained by controlled pyrolysis and structural reorganization of IRMOF-1, which is composed of ZnO QDs without inherent defects or agglomeration and porous carbon matrix^[182]. The reversible capacity of the hierarchical structure anode exceeds the theoretical prediction, and the cycle performance is also better than the reported value^[183]. In the highly long-range ordered structure of the MOF precursor, each ZnO quantum dot is separated by amorphous carbon that lacks long-range order, and interfacial charge storage at the surface of the ZnO quantum dot and weak binding of lithium, reaction with the electrolyte, and charge separation at the SEI layer may occur and provide additional capacity^[184]. In contrast, the known carbon-coated ZnO nanoparticles are more or less agglomerated during charge-discharge cycles, which seems to ignore other possible ways of lithium storage^[185]. Compared with the commercial Fe₃O₄ anode and the spindle-shaped Fe₂O₃ anode prepared by direct carbonization of Fe-MOFs in air, the porous Fe₃O₄/C octahedral anode has a higher capacity, that is, a specific capacity of 861 mAh/G after 100 cycles at 100 mA/G^[186]. The cycling stability and rate performance of the Fe₃O₄/C composite anode are also outstanding due to the porous nanostructure enhancing the diffusion speed of Li⁺ and the highly conductive randomly doped amorphous carbon framework, which effectively prevent the collapse of the electrode structure and inhibit the undesirable growth of SEI layer. The Fe₃O₄@C nanocomposite was prepared from Fe-MOFs with small organic ligands (formic acid) via high-temperature carbonization, and the Fe₃O₄ nanoparticles were coated by carbon layers (< 1 nm)^[187]. This is obviously completely different from other core-shell metal oxide @ C composites with metal oxide as the core. The reversible capacity of the Fe₃O₄@C nanoanode is 1041 mAh/G after 50 cycles at 100 mA/G; The rate capability is lower than that of the Fe₃O₄/C anode synthesized by in situ pyrolysis^[188]. Wang et al. Successfully prepared SnO₂@C nanocomposites by taking advantage of the strong adsorbability of porous MOFs, that is, monodisperse SnO₂ nanoparticles distributed within the three-dimensional porous carbon network^[189]. Especially, the porous carbon structure not only improves the wettability of the electrode/electrolyte and ensures the ion mobility of the electrode reaction, but also prevents the SnO₂ nanoparticles from falling off from the electrode. Therefore, the SnO₂@C nanocomposite anode exhibits good reversible specific capacity, cycling stability, and durability. During the pyrolysis of In-MOF In N₂ atmosphere, the reduction potential of In ions to form In₂O₃ nanoparticles is lower than − 0.27 V. Jin et al. First treated In-MOF (MIL-68 (In)) by one-step calcination to homogeneously embed In₂O₃ nanoparticles into carbon matrix^[190]. The initial discharge capacity of the porous In₂O₃/C nanocomposite anode is about 1410 mAh/G; The reversible discharge capacity is 720 mAh/G after 150 cycles at 100 mA/G. However, the effects of different precursors on the composition, morphology and properties of In₂O₃/C composites need to be investigated. The MOF-derived In₂O₃/HPNC is an ultrastable composite anode material, and the morphological characteristics are shown in Fig. 5^[191]. After 2000 cycles at a current density of 1000 mA/G, the specific capacity of the In₂O₃/HPNC composite anode was still as high as 623 mA H/G, and the capacity fading was only 0.017% per cycle. This is because the N-doped hierarchical porous structure can not only ensure the mechanical strength and coordinate the volume expansion of In₂O₃ nanocrystals, but also provide sufficient electron and Li⁺ interpenetration paths and fast migration during charge-discharge process. Due to the inherent chemical passivity and high hydrophobicity, it is difficult to directly deposit other materials on the surface of multi-walled carbon nanotubes (MWCNTs). The hierarchically porous MWCNTs/Co₃O₄ composite anode containing nano-sized units was prepared by embedding MWCNTs into Co₃O₄ polyhedrons through heat treatment of MWCNTs/ZIF-67 precursor^[192]. Similarly, nano-MWCNTs/ZnO composite anode was synthesized using ZIF-8/MWCNTs as precursor^[193]. Under the joint action of Co₃O₄ or ZnO with high theoretical capacity, MWCNTs with high conductivity and a polyhedral structure, the porous MWCNTs/Co₃O₄ or MWCNTs/ZnO nanocomposite anode is superior to the corresponding metal oxide in specific capacity, cycle stability and rate capability, and shows super lithium storage capacity and wide application prospect. For the NiO/CNT composite anode, CNTs are not only distributed on the surface of NiO microspheres, but also embedded in the interior, thus connecting with each other and forming a 3D conductive network^[194]. Especially, the maximum reversible capacity of NiO/CNTs-10 anode is 812 mAh/G after 100 cycles at 100 mA/G. It is still stable at 502 mAh/G after 300 cycles at a high current density of 2000 mA/G. Obviously, the incorporation of one-dimensional CNTs into metal oxides is an effective way to improve the transport speed of Li⁺ and the lithium storage performance. Although the hierarchical tubular structure CNT/Co₃O₄ composite exhibits excellent rate capability and electrochemical reactivity with a specific capacity of 782 and 577 mAh/G at 1000 and 4000 mA/G, respectively, the two-step synthesis process is significantly more complex^[195]. In addition, one-dimensional carbon fiber is one of the ideal templates for the synthesis of electrode materials because of its fast electron transfer rate, high strength and strong buffer volume effect. For example, the porous CFs@Co₃O₄ composite is composed of carbon fibers and Co₃O₄ polyhedra on its surface, which well retains the structural morphology and size uniformity of the CFs @ ZIF-67 precursor^[196]. Benefiting from the buffering effect of the porous core-shell structure and the synergistic effect between Co₃O₄ with higher theoretical capacity and carbon fiber, the CFs@Co₃O₄ composite anode shows good capacity, cycle stability and rate performance, which is very potential to replace traditional metal oxide and carbon fiber anode materials. Flexible graphene can also be used as a conductive matrix for MOFs to prepare metal oxide/graphene composites due to its special physical, chemical and mechanical properties. Ji et al. Synthesized a free-standing porous 3DGN/CuO composite anode by in-situ self-assembly growth of Mn-MOF on the 3DGN matrix by impregnation method, and then by subsequent heat treatment, that is, CuO octahedral nanoparticles were uniformly distributed in the three-dimensional graphene framework^[197].

Similarly, Shao et al. Synthesized NiO/GF composites by growing Ni-MOF on the surface of graphene foam by solvothermal method and then heat treating at high temperature^[198]. Flexible conductive graphene not only supports active materials, but also prevents NiO particles from diffusing into the electrolyte. It can be seen that precise control of active material loading, rational design of porous hierarchical nanostructure and utilization of interfacial interaction are the key to obtain the best lithium storage performance. Although the electrochemical properties of 3DGN/CuO and NiO/GF binderless anodes are better than those of the corresponding metal oxide and graphene electrodes, the limited active sites and easy pulverization still limit their application to some extent. Reduced graphene oxide (rGO) has no oxygen functional groups in its structure, and has high conductivity, chemical stability, flexibility, specific surface area and porosity. Considering that the organic combination of multifunctional MOFs and graphene fiber structure can obtain more novel functions and physicochemical properties, Zhang et al. Prepared one-dimensional porous Fe₂O₃/rGO composite by two-step annealing treatment of MIL-88-Fe/GO precursor^[199]. Compared with the composite aerogel prepared by common freeze-drying and annealing processes, the Fe₂O₃/rGO composite anode has similar electrochemical active surface area and higher density, and has more excellent specific capacity, cycling stability and rate capability. The initial Coulombic efficiency of the nanoporous rGO/Co₃O₄ composite anode is about 70%, and the rate capability and long-term cycling stability are outstanding due to the SEI layer formation, interfacial lithium storage, partial undecomposed Li₂O, and irreversible decomposition of the electrolyte. Obviously, the positive effect of rGO on improving the conductivity and reducing the charge exchange resistance at the electrode/electrolyte interface cannot be ignored^[200]. In addition, GO/Ni-MOFs were synthesized by the strong electrostatic adsorption of GO on the surface of MOFs, and then the RGO/NiO composite anode was obtained by subsequent heat treatment^[201]. The high specific surface area and medium-size porous structure ensure the effective charge transfer efficiency during charge-discharge process, and the active RGO outer layer protects the electrode structure from destruction.

N is close to the C atom diameter, but the electronegativity is slightly higher. The inessential defects produced by N doping can improve the reactivity and conductivity of carbon^[130]. In addition, the doping of N atoms can also change the electronic characteristics, provide more active sites for the adsorption of Li⁺, enhance the interaction between carbon structure and Li⁺, and improve the diffusion and migration kinetics of Li⁺, which are beneficial to improve the electrochemical performance of electrode materials. Although the pyrolytically synthesized MnO/C-N composite well retains the morphological characteristics of the rutile Mn-MOFs (Mn-PBI) precursor, the size is slightly reduced due to pyrolysis and volume shrinkage^[202]. Temperature is an important parameter affecting the amount of N-doping when MOFs are used as sacrificial templates to prepare high-performance electrode materials. The capacity (higher than the theoretical capacity of MnO, 765 mAh/G), cycling stability and rate capability of the MnO/C-N nanoanode are higher. In particular, the higher N doping content (5. 10 wt%) makes the MnO/C-N-500 anode structure electron-poor, which is easier to combine with Li atoms strongly and accept more charges from the Li^+[203,128]. In order to make all active sites of electrode materials participate in the electrochemical reaction, Chu et al. Made NiO nanocrystals grow on N-doped porous carbon substrate by optimizing the MOFs (Ni-NTA) derivation strategy, which solved the problem of NiO nanoparticles falling off from the electrode surface due to mechanical stress during repeated charge-discharge^[204][205]. Therefore, the rod-like NiO @ N-C composite anode with rough surface has a specific capacity of 1373 mAh/G after 200 cycles at 100 mA/G. After 1000 cycles at 1000 mA/G, the capacity is still as high as 877 mAh/G. These results indicate that the conductivity and electrochemical kinetics of the NiO @ N-C composite anode are good, which is more conducive to the diffusion, transport and storage of Li⁺. In the process of direct pyrolysis of MOFs template, the high thermal stability is helpful for the in situ transformation of organic chains into N-doped carbon materials, avoiding the volatilization of residual organic matter. Therefore, as sacrificial templates and auxiliary carbon and nitrogen precursors, nitrogen-containing MOFs with high thermal stability can be directly calcined in air to synthesize metal oxide/N-doped porous carbon composites. Upon direct calcination of the Fe-ZIF precursor, the catalytic effect of Fe atoms accelerates the decomposition and carbonization of organic ligands, resulting in uniform intercalation of ultrafine Fe₂O₃ particles into the N-doped carbon spherical shell^[206,207]. Compared with the core-shell structure, the utilization rate of atoms in this unique internal hollow structure is higher. The hollow interior and N-doped carbon matrix of the Fe₂O₃@N-C composite anode can be regarded not only as a nanoelectrochemical reactor, but also as a storage for Li⁺ and electrolyte. Therefore, the reversible capacity is still stable at 1142 mAh/G after 100 cycles at 1000 mA/G. The three-dimensional hierarchical NCW@Fe₃O₄/NC composite anode is a nanomaterial synthesized by self-assembling Fe-MOFs on N-doped carbon nanowires and then carbonizing, and the ultra-small Fe₃O₄ nanodots are uniformly embedded in the N-doped carbon nanowires^[208]. After 600 cycles at 1 A/G, the reversible capacity of the NCW@Fe₃O₄/NC composite anode is 1741 mAh/G; The capacity is still 723 mAh/G at a high current density of 10 A/G. Obviously, this self-assembly process is not only suitable for other types of MOFs to uniformly coat one-dimensional carbon nanotubes, two-dimensional graphene and three-dimensional N-doped carbon nanomesh, but also can realize the controllable growth of other functional materials (such as metal oxides, sulfides and selenides) on different carbon nets. Co₃O₄/N-C polyhedron, fish-scale structure Co₃O₄/N-C, Co₃O₄/N-PC dodecahedron, and starfish-like Co₃O₄@N-C are porous nanomaterials synthesized by using Co-TATB, nitrogen-rich Co-MOFs, ZIF-67, and Co-MOFs as sacrificial templates or precursors, respectively, with highly uniform distribution of Co₃O₄ nanoparticles on the N-doped carbon matrix^{[209][210][211][212]}. Without exception, the capacity, rate performance and cycle stability of these composite anodes are significantly improved, but there are still some differences due to the preparation process, morphology and structure, and the amount of N doping. The hollow structure of transition metal oxides is easy to construct, but the size distribution, morphology uniformity, pore volume, shell definition and functional groups are difficult to control. Ding et al. Prepared N-doped Co₃O₄@PC microspheres with multi-shelled hollow structure by thermal oxidation of Co-MOFs precursor at different temperatures^[213]. The coexisting pyrrole and pyridinic N can also provide additional electrochemical active sites, although the amount of N doping is low. The reversible capacity of N-doped Co₃O₄@PC anode is 1701 mAh/G after 60 cycles at 100 mA/G, especially 427 mAh/G at 5000 mA/G, due to the combined effects of the porous shell structure composed of nanocrystals, the gap between the shells and the porous carbon. The results show that there are two ways to combine CNTs with nano-sized metal oxides: one is to form metal oxides on the surface of CNTs, which retains the electron and ion migration path, but the volume expansion can not be effectively alleviated. The other is to intercalate metal oxides into the pores of CNTs, which can alleviate the volume expansion, but hinder the electron and ion migration to some extent. Obviously, the in-situ growth of CNTs and metal oxides is a feasible method to solve the above contradiction. As bifunctional metal-organic precursors and self-sacrificial templates, MOFs enable the construction of heterogeneous atom-doped carbon-based architectures with specific surface properties through controlled pyrolysis or post-treatment. Actually, the interlayer spacing of 2D MOFs is wide and the interlayer interaction is weak, which is favorable for Li⁺ intercalation/deintercalation. Compared with the typical three-dimensional structure, the thin layer can provide shorter ion migration distance and expose more active centers. The heterostructured CoO-NCNTs composite anode is derived from layered Co-rich 2D-MOFs, the active CoO coated on the top of NCNTs provides good ionic conductivity and alleviates the volume change, and the tightly entwined NCNTs constitute an effective conductive network for ion migration^[214]. At 500 mA/G, the cycle life of the CoO-NCNTs composite anode is more than 2000 cycles, and the capacity fading per cycle is only 0.0063%. Making full use of the advantages of MOF and graphene-based aerogel, a large number of Co₃O₄ particles of ZIF-67 @ NGA were embedded into the N-doped graphene network after heat treatment, which solved the problem that the metal oxide particles could only be deposited on the graphene sheets but could not enter the interlayer^[215]. The reversible capacity of this porous hierarchical structure Co₃O₄@NGN composite anode is not only higher than that of graphite and bulk Co₃O₄ anode, but also higher than that of MWCNTs/Co₃O₄ anode^[192]. In addition, the ZnO @ C composite anode is composed of spherical particles with uniform submicron size, and each core-shell composite particle contains a ZnO core, a conformal carbon layer with precisely controlled thickness, and a boundary region with uniform intermingling inside. The good charge-discharge reversibility, cycle life, and rate capability mainly depend on the synergistic effect of the N-doped conductive carbon layer, higher structural/interfacial stability, and the LiF-rich SEI layer to improve the Li⁺ diffusivity^[216].

In addition, doping metal elements is also the main means to improve the lithium storage performance of metal oxide/carbon matrix composites. For example, Co can simultaneously improve the conductivity and carbon graphitization degree of Co-doped ZnO @ C (CZO @ C) composite anode^[217]. After 50 cycles at 100 mA/G, the reversible capacity of the CZO @ C composite anode is 725 mAh/G, which is significantly higher than that of ZnO @ C (725 mAh/G). Similar to ZnO, the reaction of CZO @ C composite anode with Li includes not only the reversible transformation of metal oxide into nano-sized metal and lithium oxide matrix, but also the alloying/dealloying process of lithium and metal. Li et al. Obtained Ti-doped mesoporous CoO @ C octahedra by pyrolyzing bimetallic Co-Ti-MOF to embed ultra-small CoO nanoparticles into the surface and interior of the carbon framework^[218]. As the heterogeneous Ti doping introduces more defects and reduces the nanoparticle size, which is beneficial to the maximization of reactive active sites and the improvement of pseudocapacitive behavior, the CoO @ C nanocomposite anode has a reversible capacity of up to 1180 mAh/G after 150 cycles at 200 mA/G. Although the addition of new metal elements improves the structural stability and reversible capacity of two-component metal oxides, the conductivity is still reduced by agglomeration during cycling. Metal ions enter the structure of MOFs through pores and gaps, and bind to and integrate with the active sites. Therefore, on the basis of the interconnection of metal ions and organic chains, the introduction of heterogeneous metal ions can form a new type of MOFs precursor with controllable structure. Zhang et al. Synthesized C/NiCo₂O₄ composites by heat treatment of bimetallic Ni-Co-MOF precursor^[219]. Due to the cross-connection of NiCo₂O₄ and amorphous carbon to effectively alleviate the volume expansion and improve the conductivity, the specific capacity, rate capability, and cycle stability of the C/NiCo₂O₄ anode material are good. Similar to other metal oxide anodes, the obvious capacity fading and increasing trends of C/NiCo₂O₄ anodes come from the formation of SEI layer and the activity of materials, respectively. The theoretical capacity of hierarchical perovskite structure CoTiO₃ is about 520 mAh/G. Although the structure remains stable during cycling, the low conductivity seriously restricts the application of CoTiO₃ electrode. The amorphous carbon formed after the carbonization of Co-MOF was uniformly coated on the surface of CoTiO₃ nanocrystals under the catalysis of high temperature and metal ions^[220]. In addition to making up for the lack of conductivity of CoTiO₃, the amorphous carbon layer can also avoid the agglomeration of magnetic CoTiO₃ particles and improve the stability of the structure. The reversible capacity of the C/CoTiO₃ composite anode was 610 mAh/G after 1400 cycles at 2000 mA/G. The NiFe₂O₄/N-C composite anode was derived from bimetallic NiFe-MOF calcined in N₂^[221]. The three-dimensional mesoporous nanostructure shortens the electron/ion migration path and improves the diffusion coefficient of the Li⁺, and the NiFe₂O₄ nanoparticles are uniformly distributed and fully utilized during repeated charge-discharge cycles. At the same time, N doping improves the conductivity of the carbon matrix and accelerates the internal electron and ion transport. After 50 cycles at 100 mA/G, the reversible capacity of the NiFe₂O₄/N-C composite anode is 760 mAh/G, and the rate capability is also better than that of NiFe₂O₄/CNTs and NiFe₂O₄/MWCNTs anodes^[222][223]. Due to the complementary effect of the one-dimensional hierarchical porous structure and multi-components, the lithium storage performance of the core-shell CNTs@ZnCo₂O₄ nanowire anode is significantly improved^[167]. On the one hand, the abundant redox sites in metal oxides promote the reversibility of electrochemical reactions and inhibit active material agglomeration. On the other hand, CNTs, as an effective conductive and buffering matrix, improve the conductivity and strain coordination ability. Importantly, the MOFs shell can be coated with surfactants of arbitrary chemical composition, structure, dimensionality, and size to modify the core material, not to mention the dimensionality and size of MOFs. The CoFe₂O₄/GNS composite anode is synthesized by a one-step solvothermal method, and the capacity and the rate capability are obviously improved^[224]. This is not only inseparable from the beneficial effect of higher theoretical capacity CoFe₂O₄, but also fully combines the advantages of graphene nanosheets in conductivity, specific surface area and structural flexibility. Yang et al. First synthesized CoFe₂O₄/GNS nanocomposite using bimetallic MOFs and graphene oxide via one-step chemical precipitation and subsequent pyrolysis^[225]. Compared with pure CoFe₂O₄ nanoparticles, the reversible capacity, capacity retention, and rate capability of the CoFe₂O₄/GNS nanocomposite anode are significantly improved, which is not only related to the synergistic effect of Co and Fe oxides, but also the contribution of graphene nanosheets to coordinate the volume change and enlarge the electrolyte/electrode contact area cannot be neglected. The Zn_xMnO@C composite retains the hollow hexagonal nanodisc structure of the MOF precursor, and the ultrafine Zn_xMnO nanoparticles in each subunit are uniformly embedded into the continuous carbon matrix^[226]. Benefiting from the protective effect of the carbon matrix, the hierarchical hollow structure composed of two-dimensional subunits stacked in parallel, and the fast reaction kinetics derived from the pseudocapacitive characteristics, the Zn_xMnO@C anode has a specific capacity of 1050 mAh/G after 200 cycles at 100 mA/G; The rate capability is 713 and 330 mAh/G at 1 and 10 A/G, respectively. Especially, the stable cycle reaches 120 times when it is combined with the positive electrode of LiMn₂O₄ to form a full battery.

It is difficult to directly derive mixed metal oxide/carbon composites from mixed MOFs, and their development and application are obviously limited. With the uniform porous hierarchical nanostructure, the in situ formation of carbon layer and the strong electron interaction between Fe₃O₄ and MnO, the MnO-doped Fe₃O₄@C nanosphere anode based on Mn-doped MIL-53 (Fe) precursor has a capacity of 1297.5 mAh/G after 200 cycles at 200 mA/G^[227]. The large amount of Co²⁺ in the ZIF-67 structure has a catalytic effect on the formation of conductive sp²- graphitic carbon when heated in N₂. The porous carbon layer of the core-shell structured MoO₂@C composite consists of a large amount of N-doped carbon and CoO nanoparticles^[228]. Pyridine N and low C content (about 3.24 wt%) increase the conductivity and active material ratio, respectively, and MoO₂ and CoO nanoparticles accelerate the Li⁺ diffusion kinetics and provide additional capacity, respectively, which are beneficial to the improvement of the electrochemical performance of the MoO₂@C composite anode. The multilayer ball-in-ball hollow nanostructured Co₃O₄/NiO/C composite was prepared by sequential carbonization and oxidation of CoNi-MOF, in which the shell wall of each ball is composed of uniform nanorods^[229]. After assembling the battery and cycling for 200 times, not only the morphology of the Co₃O₄/NiO/C anode did not change, but also the Co₃O₄ and NiO did not agglomerate, and the electrode structure remained intact. The Fe-Mn-O/C microspheres were derived from the direct calcination of Fe/Mn-MOF-74 in inert atmosphere, and the uniformly dispersed Fe₂O₃ and Mn₃O₄ nanoparticles were coated with thin carbon layers^[230]. The synergetic effect of the hollow spherical structure, hierarchical pores and two-component metal oxides (Fe₂O₃ and Mn₃O₄) makes the Fe-Mn-O/C composite anode have a capacity of 1294 mAh/G after 200 cycles at 100 mA/G; The capacity is stable at 315 mAh/G at 5 A/G. The unique composition, atomic-level structure, and suitable chemical stability facilitate the immobilization of pyrolytically derived transition metal oxides of MOFs in porous carbon frameworks with controllable topology. Wu et al. Modified Co₃O₄@CuO microspheres with carboxyl-functionalized graphene quantum dots (GQDs) to synthesize a Co₃O₄@CuO@GQDs composite anode^[231]. In addition to the core-shell structure, high porosity, and multi-step lithium storage energy affecting their lithium storage performance, the role of GQDs in improving the specific surface area, conductivity, and the number of active sites, as well as enhancing the affinity of Li⁺ is also important. Using Fe^III-MOF-5 as precursor and self-sacrificial template, Zou et al. Successfully constructed uniform nanostructured ZnO/ZnFe₂O₄/C octahedra^[232]. The rate capability, reversible capacity, and cycling performance of the ZnO/ZnFe₂O₄/C composite anode are quite outstanding due to the effects of the hollow internal structure, ultrafine constituent units, conductive elastic buffer framework, and high porosity. This low-cost, convenient and controllable synthesis process is expected to be mass-produced, and can also be used to prepare other porous carbon-stabilized electrode materials with high power and energy density. The core-shell ZnO/Ni₃ZnC_0.7/C composite anode exhibited a stable reversible capacity of 1002 mAh/G after 750 cycles at 500 mA/G^[233]. The coulombic efficiency, cyclicity, and rate capability of the ZnO/ZnFe₂O₄@C composite anode are quite outstanding under the combined effect of factors such as the core-shell hollow ZnO/ZnFe₂O₄ mesoporous nanospheres, the uniformly coated ultrathin N-doped carbon layer, the nanosized heterogeneously connected mixed ZnO-ZnFe₂O₄, and the interconnected carbon matrix throughout the internal ZnO/ZnFe₂O₄^[234]. Niu et al. Fabricated N-doped C@ZnO/ZnCo₂O₄/CuCo₂O₄ nanocomposites using Co-Cu-ZIF-8 as a sacrificial template^[235]. Multicomponent metal oxides provide them with higher conductivity and richer redox kinetics due to their lower activation energy for electron transfer. A large amount of N doping makes the porous hollow structure electron-poor and easier to combine with the Li⁺. The reversible capacity of N-doped C@ZnO/ZnCo₂O₄/CuCo₂O₄ nanoanode is 1742 mAh/G after 500 cycles at 300 mA/G; The capacity is still 1009 and 667 mAh/G at high current densities of 3 and 10 A/G, respectively. Obviously, not all active materials react with Li⁺ at the same time during the repeated charge-discharge process, and other component active materials that do not participate in the reaction shoulder the role of relieving stress, coordinating volume change and inhibiting electrode pulverization. Therefore, the good lithium storage performance of MOFs-based derived hybrid multi-component composite electrodes mainly depends on the mutual synergism and inherent structural advantages of each component. The precursors/templates and main electrochemical performances of some metal oxide/carbon-based composite anode materials are systematically summarized in Table 6.

Metal sulfides generally have lower cost and higher theoretical capacity, higher conductivity than the corresponding metal oxides, and relatively small volume expansion during Li⁺ intercalation/deintercalation. Therefore, metal sulfides have obvious advantages in structural stability and reaction kinetics^[236]. For example, the Co₉S₈ anode with a theoretical specific capacity of 544 mAh/G mainly realizes Li⁺ intercalation/deintercalation through chemical reaction Co₉S₈+16Li⁺+16e^-↔9Co+8Li₂S^[237]. However, metal sulfides are also difficult to avoid electrode pulverization and structural instability caused by volume change, and even more prone to dissolution of polysulfides in electrolyte, resulting in capacity fading, rate capability and cycle stability reduction. Zhang et al. Synthesized H-Co₉S₈@C hexagonal prism using Co-MOF-74 as precursor^[238]. The highly conductive porous carbon adsorbs and captures intermediate reaction products (such as polysulfides) to reduce the interfacial resistance between metal sulfide particles, while the ultrafine and uniform hollow structure Co₉S₈ nanoparticles provide the maximum space to buffer stress changes and avoid agglomeration. After 250 cycles at 100 mA/G, the capacity of the H-Co₉S₈@C composite anode still showed an upward trend and maintained at 900. 5 mAh/G, especially the rate capability and cycle stability were also significantly better than those of S-Co₉S₈@C. The Co₉S₈@NMCN composite nanomaterials retained the size uniformity and morphological structure of the original ZIF-67, and the Co₉S₈ nanoparticles were completely coated by the N-doped carbon layer^[239]. Zeng et al. Synthesized Co₉S₈/N-C nanocomposite by in situ pyrolysis and sulfidation process to make Co₉S₈ uniformly dispersed in N-rich carbon hollow spherical shell^[240]. The discharge capacity of Co₉S₈/N-C anode is 784 mAh/G at 1 C, which is higher than the theoretical capacity of Co₉S₈. The discharge capacity at 4 C is still stable at 518 mAh/G. These results suggest that the N atom changes the charge distribution on the surface of the carbon matrix, and the N-doped carbon matrix provides more active sites for lithium storage reaction, which promotes the Li⁺ and electron migration in the electrode by improving the conductivity and electrochemical activity. In addition, the reversible formation/dissolution of the SEI layer and the polymer-like/gel layer formed in the first cycle also provides additional capacity^[241]. The strong reducing organic ligand in ZIF-67 can react with divalent transition metal ions to form the corresponding metal. Sulfur-containing organic compounds and sulfur powder or H₂S gas are generally selected as sulfur sources in traditional sulfurization reactions. Although transition metal sulfides (such as CdS and ZnS) have high sulfur content and are rarely used as sulfur sources, their decomposition products are easy to control and recycle, which provides the possibility of green synthesis of some multifunctional sulfides with special structures and components. Wang et al. Synthesized Co₉S₈/S-NC nanomaterials by solid-phase ion-exchange and diffusion methods, in which CdS replaced common sulfur sources such as sulfur powder, and ZIF-67 was used as a sacrificial template, reducing agent, and C, N, Co source^[242]. The results show that the capacity of the Co₉S₈/S-NC anode synthesized at the annealing temperature of 800 ℃ is 500 mAh/G after 300 cycles at 1000 mA/G, and the cycle stability is good. Obviously, the preparation of MOFs-derived sulfide/carbon-based composites by sulfurization reaction with additional sulfur source is mostly complex. Chen et al. Directly used the Co-MOF containing sulfonic acid groups as the precursor to prepare the N/S Co-doped 3D composite (Co₉S₈/NSC) for the first time by virtue of the in situ sulfidation reaction of sulfonic acid functional groups and Co²⁺ to generate Co sulfides during pyrolysis, that is, Co₉S₈ nanoparticles were uniformly distributed on the N/S Co-doped porous carbon matrix^[243]. The Co₉S₈/NSC anode material exhibited a reversible capacity of 1179 mAh/G at 100 mA/G and a specific capacity of 789 mAh/G after 1000 cycles at 2000 mA/G. This is because N doping enhances the chemisorption of S atoms by the carbon matrix and improves the cycling stability^[244]. Wu et al. Obtained three-dimensional CoS @ PCP/CNTs polyhedral composites by simultaneous pyrolysis and vulcanization of ZIF-67 template using a one-step method^[245]. During the reaction, Co²⁺ from the pyrolysis of ZIF-67 reacts with sulfur powder to form Co sulfide, which subsequently acts as a catalyst to promote the in situ growth of carbon nanotubes on the surface of carbon polyhedra. The strong coupling of porous hollow interior structure, ultra-small building unit, flexible conductive carbon matrix, and sulfide/carbon interface enables the CoS @ PCP/CNTs anode to exhibit ultrahigh reversible capacity (1668 mAh/G) and rate capability. Wang et al. Prepared N-doped porous carbon/cobalt sulfide (NC/CoS₂) composite by a simple low-temperature sulfurization process using nanocrystalline ZIF-67 as precursor, and studied the effects of CoS₂ and carbon matrix particle size on battery performance^[246]. With the combination of ultrafine CoS₂ particles and N-rich porous carbon layer, the specific capacity of NC/CoS₂ anode is 560 mAh/G after 50 cycles at 100 mA/G; The reversible capacity is maintained at 410 mAh/G at a high current density of 2500 mA/G. The binderless CoS₂-N-C/3DGN composite anode is prepared based on the organic combination of solution immersion reaction and carbonization-sulfurization treatment, and the excellent electrochemical performance mainly comes from the synergistic effect of the CoS₂-N-C with high electrochemical activity, the low-density three-dimensional graphene framework with high conductivity and the macroscopic porous structure^[247]. In addition to providing space for CoS₂ volume expansion and effectively preventing pulverization, host structures such as graphene and porous carbon can also absorb and trap polysulfide intermediates. The reasonably designed hollow structure and the local pseudocapacitance characteristic, which is beneficial to the rapid insertion/extraction of Li⁺, provide a strong guarantee for the excellent rate performance of the CoS₂/NSCNHF anode, with a coulombic efficiency of about 95.0% (except for the first and second cycles) after 100 cycles of charge and discharge, still maintaining a fairly high capacity (845.0 mAh/G)^[248].

MOFs are difficult to combine with carbon-based materials during pyrolysis, so hollow metal sulfide/carbon nanocomposites are rarely synthesized by hydrothermal method. Tian et al. Prepared nanosized Co₃S₄/MNCNT composite from solvothermally synthesized ZIF-67/MWCNT precursor by sulfurization and subsequent heat treatment, in which multi-walled carbon nanotubes penetrated the hollow Co₃S₄ quasi-polyhedron and formed a conductive network^[249]. After 500 cycles at 2000 mA/G, the specific capacity of the Co₃S₄/MNCNT composite anode is 976.5 mAh/G, and the coulombic efficiency is close to 100%, which is better than that of most known cobalt sulfide nanocomposite anodes. Yang et al. First used micro/nanostructured Co-BTC instead of nanocrystalline ZIF-67 polyhedra as the precursor to synthesize the rostonium-like Co_1-xS/C composite through low temperature carbonization and subsequent sulfidation reaction, and the pseudocapacitive characteristics and redox reaction diffusion kinetics provided more contributions to improve the electrochemical performance^[250]. In addition, this simple self-templating method is also suitable for the preparation of other MOFs-derived three-dimensional hierarchical micro/nano structure electrode materials. Heteroatom O can induce a large number of defects, provide additional reaction sites and change the growth kinetics of carbon layer, while S can sometimes promote the dissolution of active material Fe and prevent rapid passivation of the electrode. Therefore, with the synergistic effect of O-doped carbon and special nanosheet morphology, the capacity of FeS/C composite anode is 830 mAh/G after 150 cycles at 100 mA/G. The capacity is 460 mAh/G at 5 A/G^[251]. Combined with chemical etching of self-sacrificial template Fe-MOF (MIL-53) and sulfidation reaction, Yin et al. Fabricated porous FeS₂@POC concave octahedron^[252]. After 100 cycles at 100 mA/G, the reversible specific capacity of the FeS₂@POC nanocomposite anode is 1074 mAh/G, which is about three times that of commercial graphite; The capacity is still 607 mAh/G after 200 cycles at 2000 mA/G. It should be noted that the intact carbon matrix not only provides a stable surface for the formation of the SEI layer, but also inhibits its continuous cracking and re-formation. The core-shell C@Fe₇S₈ nanorods prepared by solid-state chemical sulfidation process contain about 13% carbon and 87%Fe₇S₈, and the specific surface area is up to 277 m²/g^[253]. After 170 cycles at 500 mA/G, the lithium storage capacity of this conversion-type composite anode is 1148 mAh/G, especially the rate capability at 2000 mA/G is more outstanding (657 mAh/G), which is significantly higher than that of Fe₃O₄/C anode. Obviously, S is very important for the preparation of C@Fe₇S₈ core-shell nanorods based hierarchical porous composites, and the additional capacity mainly comes from the continuous formation of electroactive polymer colloid film on the surface of the hierarchical porous structure and the interfacial lithium storage provided by the interfacial pseudo-capacitive lithium storage mechanism. In addition to the high specific surface area, porous structure, and conductive carbon matrix, the intimate combination of metal sulfide nanoparticles and carbon framework contributes to the enhanced conductivity, reaction kinetics, and structural stability of ZnS/PC composite anode^[254]. On the basis of N-doping and carbon coating to improve the conductivity and stabilize the SEI layer, the new ZnS @ NC composite anode has a capacity of 853 mAh/G after 500 cycles at 500 mA/G by making full use of the strong interaction between N-doped carbon materials and Li^+[255]. Although the lithium storage properties of ZnS/carbon composites with two-dimensional or three-dimensional structures have been improved to some extent, the inhomogeneity of ZnS/carbon composites synthesized by solid phase and coprecipitation methods needs to be further improved. For example, the three-dimensional hollow ZnS NR @ HCP composite anode synthesized by simultaneous carbonization and sulfurization of ZIF-8 template not only realizes the controllable growth of crystal morphology, but also ensures the compatibility of different subunits in nano-size^[256].

The sheet-like ZnS/C composite anode was synthesized in situ by pyrolysis of MOFs precursor composed of Zn²⁺ and sulfur-containing organic ligands^[257]. The results show that the pyrolysis temperature is closely related to the pore collapse, carbon content, morphology, polymorphic transformation and crystallinity, which ultimately affect the lithium storage performance of ZnS/C composite anode. Although α-MnS anode material has a high theoretical capacity (616 mAh/G) and a low redox potential, its lithium storage performance is not ideal due to the volume effect. The intercalation of nano-sized transition metal sulfides into porous carbon matrices (or layers) such as carbon nanotubes, amorphous carbon, or graphene is expected to solve these problems, but due to the complexity or high cost of the process, this method has rarely been used to prepare α-MnS based composites, and the mechanism of lithium storage has not yet been clarified. Heteroatoms S and N can improve the wettability of the electrode-electrolyte and provide additional lithium storage sites. In the α-MnS/SCMFs microrod composite, ultrafine α-MnS nanoparticles are homogeneously embedded within the S-doped carbon mesoporous framework^[258]. The specific capacity of the composite anode is 1383 mAh/G after 300 cycles at 200 mA/G, and the kinetic analysis shows that the pseudocapacitive behavior is beneficial to rapid lithium storage. More importantly, the rate capability, long-term stability, and Coulombic efficiency of the α-MnS/SCMFs/Cu//NCM full cell are also quite outstanding.

In addition to the preparation of metal sulfide/C matrix composites by sulfurization reaction with different sulfur sources, the organic combination of MOFs-derived carbon materials and metal sulfides can also effectively compensate for the insufficient conductivity and cycle stability of metal sulfides. The N-doped carbon material derived from ZIF-8 (N content up to 34wt%) has the characteristics of better conductivity, faster Li⁺ diffusion rate, low working voltage and high theoretical capacity, and the core-shell structure MoS₂@NC anode is a composite material synthesized by in-situ growth of the N-doped carbon layer derived from ZIF-8 on the surface of a MoS₂ microsphere^[259]. In contrast, the uniform growth of two-dimensional ultrathin MoS₂ nanosheets on the active surface of N-doped porous carbon dodecahedra derived from direct pyrolysis of ZIF-8 can also be achieved without any treatment and surfactant assistance,The MNCD core not only acts as a Li⁺ reservoir and an electron transport medium, but also facilitates the ultrafast charge/discharge of the outer MoS₂ shell and improves the structure, thereby improving the cycling stability of the porous MNCD@MoS₂ composite anode^[260]. During the carbonization process, the simple confined reaction of MOF contributes to the uniform distribution of MoS₂ nanocrystals within the porous carbon pores, as shown in Fig. 6^[261]. This strong confinement effect can avoid the growth and agglomeration of MoS₂, thus forming abundant active centers and edges. Influenced by the fast migration rate of Li⁺ (~10^-9cm²/s) and the obvious improvement of electrical conductivity, the MoS₂⊂C hybrid shows ultrahigh rate capability and cycling stability. The core-shell NiCo₂S₄@D-NC composite anode prepared with thiourea as the sulfur source is composed of double N-doped carbon layers and their coated NiCo₂S₄ nanoparticles^[262]. The rate performance and cycle stability of the NiCo₂S₄@D-NC composite anode are better than those of the NiCo₂S₄@NC anode, because the outer carbon layer effectively inhibits the dissolution of intermediate products (polysulfides), enhances the structural integrity of the electrode and improves the conductivity. In order to give full play to the advantages of bimetallic sulfides, carbon materials, and hollow structures, Aslam et al. Fabricated porous core-shell ZnCoS@Co₉S₈/NC composite anode, with special emphasis on the importance of in situ kinetic control and ion migration control through intermetallic synergism on the structure-related active sites after synthesis^[263]. After 500 cycles at 500 mA/G, the specific capacity of the ZnCoS@Co₉S₈/NC composite anode is 1813 mAh/G; 1095 mAh/G after 400 cycles at 2000 mA/G. The main electrochemical properties of some typical metal sulfide/carbon-based composite anode materials are listed in detail in Table 7.

Compared with oxides and sulfides, metal selenides have higher energy density and faster electrochemical reaction kinetics. Due to the relatively weak bonding strength between metal and selenium, chemical bond breaking and conversion reactions are prone to occur during charge-discharge cycles, and the electrochemical performance is often reduced due to volume change. Only by optimizing the chemical composition and microstructure to improve the specific surface area and porosity can the capacity and cycle performance requirements be met. Metal selenide/carbon-based composites are usually selenized during the synthesis of precursors or after calcination, which is similar to the synthesis of metal oxide (or sulfide)/carbon-based composites. The CoSe @ C composite anode is synthesized in situ by carbonization of ZIF-67 precursor and subsequent selenization reaction, and the CoSe nanoparticles are uniformly distributed in the porous carbon polyhedron^[264]. Among them, CoSe has higher reactivity with lithium and faster conversion reaction kinetics, and PCP provides conductive support for CoSe and alleviates stress-induced structural changes. Using 3DG/MOF aerogel as a template, Jiang et al. Designed and synthesized hierarchically structured 3DG/Fe₇Se₈@C composites, that is, nanoparticles composed of a Fe₇Se₈ rich inner core and a carbon rich outer shell distributed within a three-dimensional graphene framework^[265]. After 120 cycles at 100 mA/G, the reversible capacity of the flexible 3DG/Fe₇Se₈@C composite anode is 884.1 mAh/G, which is significantly higher than the theoretical capacity of Fe₇Se₈ (418 mAh/G); It is still 815.2 mAh/G after 2500 cycles at 1 A/G.

With the synergistic lithium storage of the two metal selenides, the stable SEI layer and the pseudocapacitive characteristics, the reversible capacity of the mesoporous bundle-like Ni-Co-Se/C composite anode is 2061 mAh/G after 300 cycles at 500 m A/G, which is much higher than the theoretical capacity^[266]. The hierarchical rod-like Bi₂Se₃@C composite anode prepared by the in-situ selenization method still maintains the initial morphology characteristics after 1000 cycles, and the structure is not obviously damaged.The outstanding specific capacity and rate capability are mainly determined by the stable micro/nano structured porous carbon matrix and pseudocapacitive behavior, and the electrochemical lithium storage reaction mechanism indicates that some Bi₂Se₃ transform from rhombohedral to orthorhombic structure after repeated charge-discharge cycles^[267]. As mentioned above, the introduction of heterogeneous N atoms into the carbon matrix produces a large number of lattice defects and forms a disordered carbon structure, which improves the conductivity and accelerates the electron migration. Liu et al. controlled the morphology and size of the ZnSe/NC composite anode by adjusting the particle size of the ZIF-8 precursor, for example, the ZnSe/NC-300 anode had a reversible discharge capacity of 724.4 mAh/G after 500 cycles at 1000 mA/G^[268]. Similarly, Tao et al. First used MOF-Ni as a template to synthesize Ni₂P/NC porous spheres via in situ low-temperature phosphating reaction, and the resulting assembled LiNi_1/3Co_1/3Mn_1/3O₂‖Ni₂P/NC full cell exhibited good rate performance and cycle life^[269]. In addition, the structural characteristics of transition metal phosphide/carbon-based composite anodes are diverse, not limited to the common hollow and spherical ones. For example, the specific capacity, cycle stability and rate performance of the hierarchical porous CoP/C nanobox composite anode are good, and the electrochemical reaction mechanism needs to be further clarified^[270]. Xia et al. Employed a simple MOF-derived self-template method and precisely regulated the pyrolysis temperature to successfully prepare a composition-controlled Co_xP-NC composite, that is, Co_xP nanoparticles embedded in an N-doped polyhedral porous carbon matrix generated in situ by a ZIF-67 template^[271]. The Co_xP-NC-800 composite anode is mainly composed of Co₂P and CoP, and has a reversible capacity of 1224 mA/G after 100 cycles at 100 mA/G; The cycle life is up to 1800 cycles at a high current density of 1000 mA/G. Therefore, it is necessary to study the evolution of composition and structure in the process of electrochemical reaction with the help of advanced technologies such as in situ spectroscopy and in situ electron microscopy. During the carbonization process, the carbon layer generated in situ Mo₃(BTC)₂ of the precursor restricts the growth and serious agglomeration of MoC nanocrystals, and ensures good high-temperature stability. The optimized MoC @ C-700 anode has a reversible capacity of 647.3 mAh/G at 200 mA/G; The capacity is still 509.8 mAh/G after 500 cycles at 2 A/G^[272]. This is not only related to the special pomegranate-like structure to reduce electrolyte consumption and form SEI layer, but also the carbonization temperature, MoC nanocrystal size and carbon layer content affect the electron migration and ion diffusion kinetics of MoC @ C composite anode. Yan Haoran et al. Obtained Ni₂B@C composite materials by carbonizing and boronizing Ni-MOF in turn, and analyzed the relationship between temperature and its morphology and electrochemical performance^[273]. The initial discharge specific capacity of the Ni₂B@C-500 anode is 1033.88 mAh/G, and the capacity is 467.17 mAh/G after 500 cycles at 500 mA/G. The good cycling stability is due to the integrity of the uniform octahedral structure and the difficulty of crushing. Table 8 shows the related electrochemical properties of other metal compound/carbon-based composite anode materials reported in some literatures. Compared with metal oxide/C and sulfide/C composites, there are relatively few studies on MOF-derived metal phosphide/C and metal carbide/C anode materials, especially the mechanism of electrochemical lithium storage needs to be studied in detail.

In MOFs materials, the central metal ions interconnected by organic ligands are in a dispersed and isolated state. The highly carbonized organic ligand can avoid metal agglomeration during heat treatment, which provides the possibility of preparing nanocomposite anode by using porous carbon-coated metal particles and metal oxides, and is expected to significantly improve its cycle stability and life.

Co₃O₄/Co/ carbon nanocages were prepared from Co-MOFs (ZIF-67) precursors by sequential carbonization and oxidation treatments, and Co₃O₄ and Co nanoparticles were dispersively embedded in the N-doped carbon nanocage matrix^[274]. The hollow dodecahedron structure with uniformly distributed pores not only solves the problem of particle agglomeration and nanostructure cracking caused by charge-discharge cycles, but also provides a continuous flexible conductive carbon framework for rapid transmission of ions and electrons. With the advantages of reversible specific capacity, coulombic efficiency, rate capability and cycle stability, Co₃O₄/Co/ carbon nanocages have great potential as anode materials. Zhong et al. Synthesized a series of core-shell structured Co/Co₃O₄@N-C nanocomposites using Co-MOF as precursor at different carbonization temperatures^[275]. Since the carbonization temperature not only affects the N and C content, the degree of carbon graphitization and electrical conductivity, but also promotes the gradual aggregation of Co/Co₃O₄ nanoparticles. Therefore, the optimized Co/Co₃O₄@N-C-700 composite anode has the highest initial discharge capacity (1535 mAh/G), reversible capacity (903 mAh/G after 100 cycles at 100 mA/G), and best rate capability (774 mAh/G after 100 cycles at 1000 mA/G). Chen et al. Prepared the core-shell structure Sn/C-ZnO composite by a two-step method, in which the metal Sn originated from the reduction of ZIF-8 coated SnO₂ particles, and the flow of liquid Sn in the pores of MOFs during high temperature pyrolysis was used to limit the shrinkage of pore size and surface area, effectively retaining the porous structure characteristics of MOFs, as shown in Fig. 7^[276]. The results show that the thicker the C-ZnO shell, the higher and more stable the discharge capacity of Sn/C-ZnO composite anode during charge-discharge cycles. He et al. First prepared SnO₂/Co@C nanocubes using a two-step solvent-free thermal solid state reaction^[277]. On the one hand, Co distributed in the SnO₂ promotes electron conduction and inhibits the migration and subsequent agglomeration of Sn during cycling, which improves the structural stability of the electrode and maintains the activity of Sn. Co, on the other hand, consumes excess Li₂O and provides extra capacity, releasing unusable Sn encapsulated by passive Li₂O and promoting its reversible transformation. Therefore, the reversible discharge capacity of the SnO₂/Co@C anode is 800 mAh/G after 100 cycles at 200 mA/G; The capacity is 400 mAh/G after 1800 cycles at 5 A/G. The NiO/Ni/graphene composite material is prepared by continuous carbonization and oxidation treatment of a Ni-MOF precursor, is composed of conformal graphene and ultrafine NiO/Ni nanoparticles coated with the graphene, and belongs to a porous nano hollow sphere-in-sphere structure^[278]. The NiO/Ni/graphene composite anode exhibits outstanding reversible specific capacity (1144 mAh/G), cyclability (almost no capacity fading after 1000 cycles at 2000 mA/G), and rate capability (805 mAh/G at 2000 mA/G) due to the graphene buffer layer improving the conductivity, promoting the formation of a stable SEI layer, and ensuring the structural stability. The capacity retention of the reasonably designed Ni @ ZnO/CNF composite anode is 88% after 100 cycles at 100 mA/G. The specific capacity is about 497 mAh/G after 10 cycles at 1000 mA/G^[279]. The results show that Ni reacts with Li₂O and Zn to form Li-Zn alloy, and C nanocages and heterogeneous N atoms form a conductive framework and act as a buffer framework for the volume change of ZnO. In addition, the carbon framework retains the shortened rhombohedral morphology and the original pores, which is more conducive to the rapid intercalation/deintercalation of Li⁺. Especially, this simple and low-cost synthesis method is also suitable for the preparation of other high-performance unsupported anode materials. The relevant electrochemical performances of some metal/metal oxide/carbon-based composite anode materials are summarized in Table 9.

The above analysis shows that although MOFs and their derived anode materials have made significant progress in the selection of central metal ions and organic ligands, structural design and optimization, and performance improvement, it is undeniable that these materials still have some gaps in meeting practical applications. The advantages and disadvantages of MOFs and their derived anode materials are summarized in Table 10.