YANG Jia-lu, QIAN Yue, WANG Ke, YUAN Hua-dong, NAI Jian-wei, LIU Yu-jing,WANG Yao, LUO Jian-min, TAO Xin-yong

（School of Materials Science and Engineering, Zhejiang University of Technology, Hangzhou 310014, China）

Abstract： To tackle the issues of rapid electrode degradation and severe safety issues caused by the uncontrollable growth of lithium dendrites in Li metal anodes (LMAs), two-dimensional transition metal carbides/nitrides (MXenes) with a high electrical conductivity, excellent mechanical properties, and abundant surface functional groups have been used as hosts to induce uniform Li nucleation and alleviate the volume changes, eventually inhibiting the formation of Li dendrites. Recent advances in the use of MXenebased nanomaterials in LMAs are summarized. The problems with using LMAs are first considered, and the ways of using MXenebased nanomaterials for suppressing Li dendrite growth and constructing stable LMAs are then summarized. These include the use of MXenes, MXene-metal hybrids, MXene-carbon hybrids, and MXene derivatives as hosts for the anodes and as additives to modify the electrolyte compositions to increase ionic conductivity and inhibit polymer crystallization. Finally, the challenges and prospects for using MXene-based nanomaterials in next-generation LMAs are briefly discussed.

Key words: Lithium metal anodes；MXene；Dendrites；Lithium metal batteries

1 Introduction

With the rapidly growing requirements for remote monitoring stations, electric vehicles and portable electronic devices, the development of high-energy and cost-effective rechargeable batteries is urgently needed, which propels the exploration and development of new battery materials beyond the conventional lithium (Li)-ion batteries (LIBs)[1-3]. Lithium metal batteries (LMBs) have attracted extensive research attention due to their higher energy densities compared to existing Li-ion technology. For example,the Li-S and Li-O2batteries deliver high theoretical specific energies of 2 567 and 3 582 Wh kg-1, respectively, which are much higher than the graphite/Li-CoO2(387 Wh kg-1) LIBs[4]. Although Li metal anode (LMA) exhibits exceptional advantages of high theoretical capacity (3 860 mAh g-1) and low redox potential (-3.04 Vvsstandard hydrogen electrode),the Li dendrite formation and unstable solid electrolyte interphase (SEI) causes rapid electrode degradation and short circuit-related safety issues[5-6]. To overcome these issues, a series of strategies including constructing artificial solid electrolyte interface (SEI)layer1 (Li3PO4[7-8], Li3N[9-10]an1 LiF-containing SEI[11-13]), modifying electrolyte compositions (solid electrolytes[14-15]and hybrid electrolytes[16-17]), and employing skeleton matrixes to host Li metal (nitride@carbon fiber[18], rGO, porous nickel foam[19],and Cu3P nanowires[20], etc.) have been adopted.Nanoengineering is a general method to address the technical challenges of LMAs. Some newly developed nanomaterials have emerged as an opportunity to overcome these challenges in recent years.

MXene, as a new family of two-dimensional(2D) layered early transition metal carbides and carbonitrides, was first reported in 2011[21]and has been widely applied in energy storage[22]. The general chemical formula of MXene is Mn+1XnTx(M, early transition metal including Sc, Ti, Zr, Hf, V, Nb, Ta,Cr, and Mo; X, carbon or/and nitrogen; Tx, surface groups;n=1-4). MXenes are usually synthesized by etching “A” in the precursor of MAX phase, where M is an early transition metal, A is an III or IV A-group element, and X is the C or N[21,23-24]. To date, more than 30 types of MXenes with different compositions,including Ti3C2, Ti2C, Ti4N3and Mo2C have been successfully synthesized[23,25-26]. In terms of development history, researchers have been committed to improving the synthesis strategies of MXenes from the initial fluoride-based aqueous solutions (such as HF,NH4HF2, LiF and HCl mixtures)[27]to electrochemical methods[27-30], an alkali-assisted hydrothermal method[31], water-free etching in polar organic solvents[32], molten salt etching methods[25,33], salt-templated methods[34], thermal reduction method[35], and chemical vapor depositio1 (CVD) method[36]at present. The MXene synthesis conditions have a significant impact on the morphology, quality, surface groups, and electrochemical performance as well[37-38].

MXenes hold great potential for serving as the hosts of LMAs because of their unique properties.Firstly, MXenes have high conductivity renders them capable of facilitating an efficient electron transfer rate for the electrode. Secondly, MXenes exhibit exceptional mechanical robustness and durability,thereby facilitating the assurance of the electrodes’structural integrity. Thirdly, the presence of lithiophilic functional groups (—O, —F and —OH) on the surface of MXene are able to effectively coordinate with Li ions and induce Li uniform nucleation. Therefore,MXenes could be utilized to enhance the cycle stability and electrochemical performance of LMAs. The first study on MXenes for LMAs was reported in 2017[26]. Since then, MXenes have been extensively used in LMAs. However, there is a shortage of specialized reviews focusing on the progress of MXenes used for LMAs.

In this review, we systematically discuss the latest advances in MXene-based nanomaterials in state-of-the-art LMAs. In the first section, we discuss the technical challenges of LMAs from 3 aspects: the intrinsic features of Li negative electrode, the growth of Li dendrites, and the SEI formed on LMAs. In the second section, we introduce 3 kinds of modification strategies, including designing MXene-based LMA hosts, constructing MXene-based artificial SEI films and modifying MXene-based electrolyte additives(Fig. 1). Specifically, MXene-based hosts are divided int1 3 parts: MXenes, MXene-based hybrids(MXene/metal hybrids, MXene/carbon hybrids, and MXene/oxides hybrids) and MXene derivatives. Finally, an overview of future challenges and opportunities of MXene-based nanomaterials for LMBs are briefly discussed.

Fig. 1 Schematic diagram of the applications of MXene-based nanomaterials for lithium metal anodes

2 The technical challenges of lithium metal anode

Lithium, located at the top of alkali metals in the periodic table. As such, it has the smallest metal atomic radius, the lightest metal density (0.53 g cm-3), the lowest redox potential, a low melting temperature, and exceptionally high atomic mobility of all metalsmetals[39]. All these properties endow LMBs to exhibit sufficient energy density and power performance[40-42]. Nevertheless, th1 LMBs experience volume expansion during the cycle, which is highly correlated with dendrite growth and unstable SEI[43].In this context, uncovering and comprehending the mechanisms of dendritic nucleation and growth, as well as the composition of SEI film, aids in the rational design of high-performance LMBs.

2.1 Infinite volume changes of Li metal anode

For Li metal anode, due to its hostless nature, the volume expansion during operation is infinite, which is much greater than that of intercalated anodes(～10% for graphite), and alloy-type anodes (～400%for Si)[41]. Additionally, porous Li formed by dendrite growth causes a greater volume change issue during plating and stripping processes. Regrettably, the capacity degradation of lithium metal batteries is due to the absence of a specific host to trap and restrain Li depositions, resulting in an unstable electrode interface, internal stress fluctuation, and the electrical disconnected Li metal[43].

2.2 Growth of Li dendrites

Lithium tends to form dendrites on the anode current collectors[44]. The growth of dendrites incessantly consumes electrolytes and punctured the separator, which causes an irreversible and permanent loss of capacity, short circuits, as well as potential safety problems[6,45-46]. Lithium dendrites involve in 2 processes: nucleation and growth. So far, 4 models, including the space-charge model, deposition and dissolution model, heterogeneous nucleation model, and stress-driven dendrites growth model, have been proposed to describe the formation and growth of dendrites.

The space-charge theory proposed by Chazalviel et al.[47]emphasizes that the concentration of anions close to the negative electrode decreases when lithium ions are rapidly deposited in a dilute solution. The consumption of Li ions in the near-surface area of the electrode will generate a huge spatial charge and electric field at the electrode/electrolyte interface, eventually induce lithium dendrite formation. The initial time of dendrites growth was named as Sand’s time. It has been confirmed that augmenting the Li-ion mobility or reducing the effective current density could lead to an increase in Sand’s time[48]. The thermodynamically and kinetically process of heterogeneous nucleation was proposed by Ely et al. through numerical simulation[49]. In this theory, the initial stages of lithium-ion deposition is a heterogeneous nucleation process, including: lithium ions gaining electrons and Li depositing on current collectors. The nucleation of lithium will be affected by the topography and roughness of current collectors. The morphology and uniformity of the initial deposited lithium will directly influence the subsequent lithium metal deposition behaviors and morphology, eventually affecting the formation and growth of lithium dendrites. The growth of Li dendrites could be dramatically suppressed if the Li nuclei are stably established, and the growth rate remains constant until the final form. The deposition and dissolution model reveal that the lithium is extruded through the resulting holes on the protective film broken by the mechanical stress at certain points and grows in the form of whiskers[50]. The stress-driven dendrites growth model was proposed by Jiang et al., namely undesirable filamentary Li dendrite growth is originated from plating induced compressive stress,surface passivation by the SEI, and the presence of subsurface planar defects during Li plating/stripping process. In this case, Li dendrites could be suppressed through effective control of stress[51].

By manipulating the Li nucleation and growth, it is possible to regulate and manage the growth of Li dendrites to achieve a planar morphology, as suggested by these models.

2.3 The SEI on Li metal anode

The properties of the solid-liquid interface have important effects on the charging and discharging efficiency, energy density, power density, service life,safety, self-discharge performance and other characteristics of lithium metal batteries. The solid-electrolyte interphase (SEI) was first defined by Peled in 1979 as a result of irreversible reactions between lithium and non-aqueous liquid electrolytes. This phenomenon is attributed to the ultrahigh chemical reactivity of Li, which leads to the creation of an ionicconductive but electronic-insulating layer[52-53]. The prevailing depiction of the SEI configuration is the mosaic model. This model assumes that the insoluble reduction products formed by the electrolyte are mosaicked with each other, while lithium ions are in different phases. The boundary between migrations indicating the surface is not homogeneous[54]. Our groups use Cryo-TEM technology to demonstrate the existence of this model[13].

Researching the SEI film’s composition is important because of the crucial role of SEI in influencing the overall performance and safety of LMBs. Experimental and simulation studies have shown that SEI of LMAs comprises degraded inorganic salt,salts/anions such as PF6-, CF3SO3-, Li3N, Li2CO3,NS(O2CF3)2-, C(SO2CF3)3-and LiX (X = F, Cl etc.),and reduction products of organic solvents such as(CH2OCO2Li)2, CH3CH(OCO2Li), CH2OCO2Li,CH3OLi CH3OCO2Li, HCOOLi[49-59]. However, a practical SEI is often uneven with poor mechanical performance, which reduces charge and discharge efficiency, increases internal resistance, cause inferior electrochemical performance. Therefore, building a stable (electrochemical and mechanically) SEI is critical.

3 The application of MXene in Li metal anodes

During the Li plating and stripping process, promoting uniform Li deposition is crucial to achieving stable and high-performance LMAs. Extensive materials have been developed to obtain dendrites-free LMAs to date, in which functionally ionic/electronic conductive materials (e.g., carbon nanomaterials[60-62],metal foams[63], and conductive polymers[64]) occupy an important position. By comparison, MXenes exhibit good metallic conductivity, impressive mechanic properties, tunable surface chemistry and rapid ion diffusion ability. All these properties have endowed MXenes with great potential in the applications for advanced LMAs. On the one hand, the high electronic conductivity and fast ion transfer rate of MXenes with abundant functional groups could not only promote uniform Li deposition, but also deliver fast electrochemical kinetics, which suppresses the growth of dendrites effectively. Moreover, the remarkable mechanical properties and facile assembly characteristics of 2D MXenes make them promising candidates for processing into diverse 3D hosts for metal anodes, which could effectively mitigate the volume changes.

3.1 Designing MXene-based hosts for lithium metal anodes

The LMAs exhibit a notable and boundless alteration in volume due to their characteristic of lacking a host, which may damage the SEI film, form lithium dendrites, comsume electrolytes, and reduce the Coulombic efficiency during the repetitive lithium plating and stripping procedure[65]. Designing a proper host for the Li-metal anode would be a great strategy to address these issues. MXenes exhibit excellent mechanical strength and good mechanical flexibility, which could alleviate volume fluctuation during cycling. Besides, MXenes exhibit exceptional metallic conductivity, high Li-ion diffusion coefficient, and versatile surface groups for regulating the electric field distribution, minimizing polarization, and facilitating homogeneous Li nucleation and deposition. Considerable research has been directed towards the utilization of MXene-based nanomaterials for constructing the host of lithium metal anodes since 2017. Efficient LMA hosts include not only MXenes themselves, but also hybrid nanomaterials derived from MXenes and MXene derivatives.

3.1.1 MXene

Yang’s group[66]was the first to corroborate the MXenes as the effective Li hosts. They used a roll-toroll mechanical approach to fabricate Ti3C2MXene(graphene, BN)-metallic lithium film. The synthetic procedure comprised 2 steps as illustrated in Fig. 2a.Initially, the Ti3C2MXene, graphene, or boron nitride nanosheets was subjected to rolling with a lithium plate to achieve elongation. Subsequently, the hybrid plates underwent folding to augment the layer count and obtained a pliable plate with a smooth surface(Fig. 2b-c). The composite anode exhibited fast electron and Li-ion transportation owing to the presence of nucleation sites and the highly conductive networks of MXene (Fig. 2d). Regrettably, the cycle life of this anode was not satisfactory. To this end, Yang’s group produced the parallelly aligned MXene layers(PA-MXene) through a simple self-assembly procedure in 2019[67]. The morphology and direction of lithium nucleation and growth on the horizontal plane were efficiently directed by the parallel alignment of MXene (Ti3C2Tx) layers (Fig. 2e-f). As a result, the cycle life of this material was significantly increased,reaching up to 900 h.

Fig. 2 (a) Schematic illustration of the synthesis of Ti3C2 MXene (graphene, BN)-lithium films[66]; (b) SEM image of the Ti3C2-Li composite anode[66];(c) Cross-sectional SEM image of the Ti3C2-Li composite anode[66]; (d) Schematic showing the sufficient transport pathways for electron and Li-ion in Ti3C2-Li composite anode[66]; (e) Schematic illustration of lithium plating on bare lithium and parallelly aligned MXene (PA-MXene) layers[67]; (f) Typical SEM images of PA-MXene layers on lithium. Inset in (f) is the top view of PA-MXene[67]; (g) Two-side press of MXene stacks onto a thin Li host[69]; (h) SEM images of the ILC-Li electrode upon plating at 3 mA cm-2 and 3 mAh cm-2[69]; (i) SEM images of the ILC-Li electrode upon stripping at 3 mA cm-2 and 3 mAh cm-2[69].Reprinted with permission

Subsequently, the morphological configuration of MXene nanosheets has undergone continuous development towards structures with low curvature. This modification provides additional internal space for expansion, which effectively mitigates physical strain and preserves structural integrity under high current density. For example, in 2022, Yang and colleagues fabricated sine-wave analogous MXene (Ti3C2Tx)(SWA-MXene) layers through the self-assembly of MXene layers onto the metallic coil, followed by drying at room temperature[68]. The composite anode relied on its unusually low-curvature construction to effectively eliminate the tip-induced nucleation effect on the surface of SWA-MXene electrodes by uniforming the lithium ions and electric field distribution and lowing the Li-ion concentration gradient. As a result,this anode exhibited a deep plating-stripping capacity of up to 40 mAh cm-2, a long life of up to 1 250 h, and a low nucleation overpotential.

Several studies indicated that when enlarge its interlayer spacing, the expanded MXene exhibited superior ionic and electronic conductivity, as well as excellent volume tolerance, compared to their original MXene nanosheet counterparts. Based on this background, Niu’s groups[69]developed an interlayer-calated thin Li metal (ILC-Li) electrode (Fig. 2g-i). The 2D lamination configuration of this composite electrode provided a significant amount of available space, resulting in the overpotential of about 120 mV after 2 200 h at a high current density of 10 mA cm-2and a high capacity of 10 mAh cm-2. Furthermore, the vertically aligned MXene nanosheet arrays could augment conductivity and address the volumetric challenges associated with lithium metal anodes. In contrast to a horizontal structure, a vertical structure offered a greater volume for Li-ion transport, superior mechanical flexibility, and improved electronic conductivity. For this, Yang and Gong[70]utilized an ice template-assisted blade coating technique to fabricate host electrodes consisting of vertically aligned arrays of Ti3C2Txnanosheets (v-Ti3C2Tx) in a facile manner,as illustrated in Fig. 3a. As depicted in Fig. 3b, the synthesized v-Ti3C2Txwas comprised of nanosheet arrays that were vertically aligned, with a gap width of approximately 15 mm (Fig. 3c) and a thickness of approximately 40 mm (Fig. 3d). A full cell constructed with LiFePO4(1.8 mAh cm-2) cathode and lithiated Li@v-Ti3C2Txanode exhibited specific capacities of 135 mAh g-1and 120 mAh g-1at the rates of 1 and 2 C, much higher tha1 Li@h-Ti3C2Txelectrodes(Fig. 3e). Similarly, Yang’s groups[71]developed a brand-new diversified perpendicular MXene-Li array by ice template-assisted blade coating method. Fig. 3f shows the stripping/plating behaviors of Li on the perpendicular MXene-Li arrays. Taking advantage of the tunable MXene walls and space, and flexible structure, the growth of lithium was observed to occur predominantly on the top surface of the arrays, as opposed to within the gaps of the reduced graphene oxide (rGO) structure (Fig. 3g-h). As such, this electrode could be stable under the high-rate capability of up to 20 mA cm-2.

Fig. 3 (a) The synthesis schematic diagram of v-Ti3C2Tx electrodes[70]; (b-c) Top view SEM image of v-Ti3C2Tx electrodes[70]; (d) Cross-section SEM image of v-Ti3C2Tx electrodes[70]; (e) Coulombic efficiencies of v-Ti3C2Tx and h-Ti3C2Tx electrodes at 1.0 mA cm-2 for 1.0 mAh cm-2[70]; (f) Schematic illustration of the stripping and plating states of perpendicular MXene-Li and rGO-Li arrays. The left bottom in a) is the SEM images of as-prepared perpendicular MXene-Li arrays. The middle bottom in panel (a) is the SEM image of perpendicular MXene-Li arrays after lithium stripping for 20 mAh cm-271; (g) Top view SEM images of perpendicular MXene-Li arrays[71]; (h) Top view SEM images of rGO-Li[71]; (i) Scheme of the fabrication of Ti3C2Tx MXene and a high-concentration MXene ink (～300 mg mL-1)[75]; (j) Scheme of 3D printing MXene arrays and lattices to guide the nucleation and growth of lithium[75].Reprinted with permission

Additive manufacturing technology, commonly referred to as 3D printing, is a nascent technology that could be used to produce a wide range of three-dimensional objects and intricate structures using appropriate inks[72]. The favorable viscosity and shear-thinning rheological behavior of MXene inks makes 2D MXene nanosheets more suitable for being incorporated into 3D architectures through printing[73-74]. For instance, Yang’s groups[75]used extrusion-type 3D printing successively developed a highly concentrated MXene ink (Fig. 4i). In the 3D printing arrays, lithium was grown towards the cobblestone rather than dendrites (Fig. 3j). Remarkably, 3D printing MXene(3DP-MXene) arrays-Li delivered a stable capacity of 111.1 mAh g-1at 30 C. From the above discussion, it could be inferred that MXene arrays with abundant interspaces homogenized both lithium-ion flux and electric field, thereby effectively suppressed the formation of lithium dendrites and exhibited admirable electrochemical performance.

Fig. 4 (a) The Top view SEM images of MXene paper[81]; (b) The Top view SEM images of MXene@Au paper[81]; (c) Cross-sectional SEM image of MXene@Au paper[81]; (d) TEM image of Ti3C2Tx@Au paper[81]; (e) HRTEM image of Ti3C2Tx@Au paper[81]; (f) Cross-sectional SEM image a of Ti3C2Tx MXene@Zn paper[83]; (g) Top view SEM images of MXene@RP paper[85]; (h) TEM images of MXene@RP paper[85];(i) SEM image of NS-Nb2C[86]. Reprinted with permission

As summarized above, the MXene-Li composite anodes could be produced using mechanical rolling,blade coating, and electro-deposition methods. During the assembly process, MXene experiences restacking problems that significantly impact its performance and limit its potential applications. Therefore, it is crucial to prevent the aggregation of MXene. Despite the discovery of various MXenes, Ti3C2is the most extensively studied host material for LMAs.However, research on the MXene-based hosts for LMA is still in its early stages compared to other carbon-based nanomaterials.

3.1.2 MXene-based hybrids

As mentioned above, MXene could be served as an ideal host for LMA due to its exceptional metallic conductivity, high ion diffusion coefficient, and multifunctional surface groups. However, similar to other ultrathin 2D nanomaterials, MXene nanosheets could be easily agglomerated and restacked during the cycling process[76-78]. Therefore, the multifunctional surface groups are limited to be exposed, which eventually affects the morphology of Li on MXene. To overcome this challenge, lithiophilic seeds have been introduced to enhance the lithiophilicity of MXene,which helps regulate the burgeoning and nucleation of Li on the host material. The morphology of Li deposition on MXene could be further regulated by the synergistic effect of multifunctional surface groups on MXene with the introduced lithiophilic seeds. Besides, the introduction of carbon nanomaterials could effectively prevent the agglomeration and sustain the structural integrity of MXene nanosheets. Therefore,the incorporation of lithiophilic seeds and carbon nanomaterials could enable a stable and high-performance battery system based on MXene-hosted LMAs.

(1)MXene/0D hybrids

Metal nanoparticles act as lithiophilic nuclear agents uniformly deposited on layered MXene paper to ensure homogeneous lithium deposition and regulate the lithium dendrite’s growth pathways and directions. Such as, Au could be dissolved into Li and then form a LixAu alloy phase, which could effectively minimize Li nucleation barriers when used as a host material for LMA[79-80]. Feng’s group[81]designed a flexible MXene@Auviaion sputtering (Fig. 4a-e).The TEM image of MXene@Au indicated that the Au nanoparticles were confined on the thin, slightly crumple and self-supporting MXene nanosheets. As a result, this flexible anode showed significantly enhanced reversible capacity with high capacity retention of 98.47% even after 200 cycles.

In addition, Zn has the potential to be the lithiophilic nuclear agent to produce dendrites-free lithium anodes. Yang and co-workers made single zinc atoms fixed on MXene layers (Zn-MXene) as the host for lithium metal anodes[82]. The Zn-MXene-Li anode exhibited a stable and extended cycling life of 1 200 h at 1.0 mAh cm-2, surpassing the cycling lives of MXene-Li (930 h) and Cu-Li (600 h) anodes. The Zn-MXene films and their hybrid lithium anodes possessed an advantage of facile manufacturing through the utilization of established rolling-to-rolling and spray coating technologies, thereby resulting in a significant reduction in costs. This, in turn, contributes to developing commercial lithium batteries. At the same time,Tian and his co-workers designed a freestanding Ti3C2TxMXene@Zn paper as the lithium metal anode host to address the challenges of dendrites growth(Fig. 4f)[83]. Over 600 cycles at 1 mAh cm-2area capacity and 1 mA cm-2current density, the modified anode displayed an average Coulombic efficiency of 97.69%, because of the outstanding mechanical flexibility, high electrical conductivity and lithiophilicity of the MXene@Zn paper. Moreover, ZnO was utilized by Li’s group to modify the MXene-based host of lithium metal batteries due to its strong lithiophilic properties[84].

Apart from the metal nanoparticles, nonmetal element could also be used as lithiophilic seeds. Recently, a flexible paper composed of MXene@iodinedoped red phosphorus (MXene@RP) was designed by Na and Huang[85]through a straightforward method of vapor condensation reduction. The distribution of red phosphorus nanoparticles doped with iodine was uniform across both the surface and interlayer of the MXene matrix. During the vapor condensation process, the integrity of the MXene paper was well preserved at the same time (Fig. 4g-h). The anchorage of iodine-doped red phosphorus onto MXene was identified as a viable approach to prevent dendrite formation during lithium deposition. This was attributed to the improved conductivity of red phosphorus resulting from iodine doping, which in turn enhanced the kinetics of charge transfer. Hierarchical architectures were designed to enhance the ability of electroactive materials against volume fluctuations, while also facilitate ion diffusion pathways for lithium uptake. The utilization of a composite anode expands the applications of MXene and offers direction for developing the lithium hosts free of dendrites.

Except for the typical Ti3C2-MXene, nitrogen and sulfur co-doped Nb2C MXene (NS-Nb2C) anode has been synthesized by Zhang and Chen et al[86]. The obvious laminated structure of NS-Nb2C can be seen in Fig. 4i. N and S doping sites could not only uniform Li deposition but enhance the electroconductivity of Nb2C. The introduced heteroatoms could serve as extrinsic defects and lithiophilic active sites, open up the interlayer spacing and improve the conductivity of Nb2C to suppress the dendrite growth. As such,the voltage-time curve of NS-Nb2C demonstrated a notable degree of long-term stability with minimal fluctuation. Additionally, the resistance values of NSNb2C were significantly lower than those of Nb2C after undergoing 50 cycles. The NS-Nb2C MXene has the potential to stimulate novel ideas for advanced LMAs.

The incorporation of lithiophilic zero-dimensional (0D) nanoparticles or nanodots into the MXene framework could facilitate the nucleation and growth of lithium, resulting in a homogeneous deposition of the element. In the process of Li plating, the preferential Li nucleation occurs around lithiophilic seeds.Uniform deposition of Li metal could be facilitated by the confinement of Li nucleation and growth inside the MXene matrix, which effectively suppresses the formation of “hot spots” and reduces favourable sites for dendrite growth. In additive to the examples listed above, other 0D materials, such as liquid metal[87],Ag[88], TiO2[89], Sn[90-91]etc, have also been introduced into MXene as lithiophilic seeds for homogeneous Li deposition.

(2)MXene/1D hybrids

Several one-dimensiona1 (1D) nanomaterials have been documented to be combined with MXene nanosheets to act as hosts for layered metal alloys(LMAs). The integration of 1D nanomaterials has been shown to improve mechanical properties and hinder the aggregation of MXene nanosheets[92]. As a typical example, the construction of a lithium host with a lamellar structure that incorporates carbon nanotubes (CNTs), MXene nanosheets, and SnO2nanoparticles has been reported by Yan’s research groupviaa simple step-by-step vacuum filtration method(Fig. 5a-b)[93]. The obtained multilayer stacked structure composite anode community of enough porosity,large surface area and high conductivity, showed the Coulombic efficiencies stabilized at 96%-98% at a high deposition capacities of 4 mAh cm-2and delivered only 4.6% volume expansion during the highcapacity loading of 8 mAh cm-2. Consequently, the designed lamellar structure was demonstrated as an efficient approach to buffer the volume variation. Furthermore, Wang[94]and his co-workers reported a particular topological structure by mixing cellulose nanofiber (CNF) with Ti3C2TxMXene as Li host (Fig. 5cd). Such topological architecture is often prepared by vacuum filtration and rotational evaporation process.The utilization of interlocked topological microstructure, aided by CNF, facilitated the formation of microspheres between MXene nanosheets through intermolecular hydrogen bonding. This significantly contributed towards augmenting the mechanical strength and flexibility of the MXene@CNF film. The Li metal anodes demonstrated remarkable long-term stability, as evidenced by a flat polarization voltage of 47 mV at 0.5 mA cm-2for a duration exceeding 1 300 h without any indication of short-circuiting.

Fig. 5 (a) Schematic illustration of Li deposition in the lamellar-structured CNT/MXene/SnO2 composite host based on lithiophilic gradient[93]; (b) Cross-sectional SEM image of the CNT/MXene/SnO2 host[93]; (c) Schematic illustration of lithium plating on MXene@CNF film[94]; (d) Magnified SEM image of MXene@CNF film[94]; (e) Schematic illustration of the C-MXene/AgNW scaffold fabrication[95]; (f) Magnified SEM image showing a joint of the C0.5-MXene/AgNW scaffold[95]; (g) The schematic diagram for the Li stripping-plating process of Ti3C2-LiB-Li hybrid and LiB-Li[96];(h) The corresponding cross-section image of the Ti3C2-LiB-Li[96]. Reprinted with permission

Not limited to CNFs, a conductive and lithiophilic scaffold was constructed by Liang’s group[95]using polysiloxane and silver nanowire (AgNW). The former by cross-linking the MXene nanosheetsviahydrolysis, polycondensation, and silylation reactions significantly enhanced the mechanical stability of the MXene-based scaffold. The later further improved the electrical conductivity and lithiophilicity of the MXene-based scaffold (Fig. 5e). The SEM images depicted the hierarchical structure of the C0.5-MXene/AgNW scaffold, showcasing an average macropore size of approximately 10 μm and a mass density of roughly 0.154 8 g cm-3(Fig. 5f). Mechanical strength and resilience were improved in the covalently crosslinked MXene host compared to the MXene scaffold.The utilization of cushioning mechanisms was beneficial in mitigating the significant internal stress fluctuations that arise from the rapid and deep plating-stripping of lithium. This approach also promoted the preservation of the integrated frame structure throughout the extended charging cycle. The lithium composite anode offered remarkable cyclic stability for lithium plating and stripping, resulting in an exceptional lifespan of 3 000 h. This stability was achieved even at an ultrahigh current density of 20 mA cm-2and an areal capacity of 10 mAh cm-2.

Furthermore, Zhao employed a technique of iteratively rolling and folding MXene-Ti3C2nanosheets in conjunction with vertically aligned lithium-boron nanofibers (LiB) to produce a lithium hybrid with a multistorey corridor structure (Fig. 5g-h)[96]. The implementation of conductive Ti3C2layers had the potential to impede the vertical expansion of lithium dendrites, while simultaneously enhanced the electrical conductivity of the mixed anode. This, in turn, could lead to a decrease in the local current density and uniform electric field. The lithophilicity Li7B6nanofibers could be used as pillars between the Ti3C2layers to prevent the multilayered corridor structure from collapsing and act as nucleating sites to induce the homogeneous deposition of lithium. The hybrid lithium anode was capable of functioning in exceeding settings,with a rate of 20 mA cm-2and an area capacity of 10 mAh cm-2.

(3)MXene/2D hybrids

Graphene oxide (GO) with good gelation ability is always used in assembing 3D structure from low-dimensional nanomaterials[97]. In 2018, Luo and his coworkers explored the MXene aerogel as a scaffold for Li metal anodes for the first time to improve the mechanical stability of traditiona1 MXene nanosheets[98]. The plating/stripping process of Li in the MXene scaffold is depicted in Fig. 6a. The as-prepared framework inherited the porous structure and functional groups of the rGO aerogel (Fig. 6b). The utilization of MXene aerogel was observed to enhance the charging/discharging efficiency of LMA in the presence of high current densities.

Fig. 6 (a) The Ti3C2 MXene aerogel scaffolds for Li metal anodes[98]; (b) SEM image of the MXene/rGO aerogel. Inset: a photograph of the aerogel[98];(c) Schematic diagram of Li-Ti3C2Tx-rGO preparation[99]; (d) SEM cross-section images of Ti3C2Tx-rGO[99]; (e) Schematic illustration of the fabrication process of the 3D MG-Li anode and corresponding photographs of the MGO film, MG film, and MG-Li anode[100]; (f) SEM image of the MG film. Inset is a photograph of the bent MG film[100]; (g) Schematic illustration of the fabrication process of the 3D MXene-MF for the alkali-metal anode[101];(h) SEM image of MXene-MF[101]. Reprinted with permission

In addition to 3D MXene aerogel anode, another important type of lithiophilic and flexible 3D porous anode Ti3C2Tx-rGO films was reported by Cao and his co-workers (Fig. 6c-d)[99]. The compact Ti3C2Tx-GO film was first fabricated by vacuum filtrating a certain amount of Ti3C2Txand GO solution, then contacting this membrane with molten Li by a spark reaction.Attributed to this innovative method, molten metal could be filled and sealed into the films, the composite film matrix along with expanding spacing and porosity performs excellent mechanical stability and flexibility. Thus, nucleation of molten Li was uniformed. Dendrites could be segregated and enclosed in separate interlayer cells and the MXene as well during the constant and repeating stripping/plating process. Consequently, the composite anode exhibited a remarkable specific capacity of 3 086 mAh g-1and the cycling time up to 1 000 h at 2 C. The full cell of LiFePO4|| Li-Ti3C2Tx-rGO exhibited superior capacity maintenance at 96.6% in comparison to the unmodified Li cells (90.1%).

The large gap between theoretical and practical electrochemical properties for the lithium metal battery system should be filled by adopting a series of design architectures towards withstand large current density and high Li loading. For this, Wu’s[100]group proposed a lightweight, porous, freestanding, mechanically stable MXene/graphene (MG) framework as a host for LMAs (Fig. 6e-f). The 3D MG anode had a rich interconnected pore structure with a high surface area (259 m2g-1) and capacity as high as 3 560 mAh g-1coupled with Coulombic efficiencies (＞99%)without Li dendrites. The composite materials exhibited exceptional cycling stability for 120 h at 20 mA cm-2without obvious voltage decay and short-circuit. Soon afterwards, they designed a 3D flexible and conductive Ti3C2TxMXene-melamine foam (MXene-MF) as a host for the LMA (Fig. 6g-h)[101]. This 3D composite anode showed both high electrochemical performance and high durability.

Apart from the carbon-based materials, covalent organic frameworks (COFs), a category of porous crystalline materials that consist of recurring organic building units that are interconnected by covalent bonds, have received great attention for application in modified lithium metal anode, since they have the properties of durable stability, big porosity, and low density, sufficient specific surface area. For example,Lai[102]and his co-workers covalently assembled COFs on MXene nanosheets to enable fast-charging lithium hosts by amination andin-situdeposition.After embellishing with the ultrathin 2D COF-LZU1,the MXene@COF showed high crystallinity, hierarchical porosity, and conductive frameworks, greatly improved lithium storage and charge transport, resulting in the long-term cyclability of 106 mAh g-1after 150 cycles.

Based on the above discussion, it could be inferred that the incorporation of lithiophilic moieties,such as Au and Zn,viasurface/interface modification,represented a highly efficacious approach to enhancing the affinity towards LMAs. As the hosts for LMAs, a combination of MXene and 1D nanomaterials (CNF, CNT and AgNW) showed extraordinarily long-term reversible Li plating-stripping at ultrahigh current densities with ultrahigh areal capacities while maintaining the structural integrity of scaffold across thousands of cycles. Besides, researchers succeed to explore the MXene-based aerogel as a high-performance scaffold for LMAs that delivered a high specific capacity. The fabrication of MXene/2D hybrid nanomaterials through introducing other 2D nanomaterials improved the lithiophilicity or mechanical properties.Meanwhile, it could provide enlarged interlayer spaces or necessary porous channels for Li plating/infusion. Constructing MXene-based hybrid nanoarchitectures has become a promising strategy to develop advanced stable and dendrite-free LMAs. The applications of MXene for LMAs and their corresponding electrochemical performances have been summarized in Table 1.

Table 1 Summary of the MXene-based nanomaterials for Li-metal anodes

3.1.3 MXene derivatives

In addition to the development of MXene-based hybrid nanoarchitectures, MXene-derived hybrid nanoarchitectures have been developed as Li hosts to facilitate the Li infusion and uniform deposition. Recently, Peng’s group[103]prepared a new MXene derivative containing pure rutile TiO2and N-doped carbon.Firstly, heated the precursor MXene under a mixing gas of NH3: Ar (1∶9) at 800 °C and then formed the TiO2/N-doped C hybrid. Secondly, injected molten Li into it. After Li infusion, 3D layered TiO2/N-doped C and TiO2/C samples were achieved. Thus, the Li cells with the LiTiO2-Li3N-C electrode delivered an extremely long cycle life of 2 000 h at 1 mA cm-2and anoverpotential of only 30 mV.

3.2 Constructing MXene-based artificial SEI films

A desirable SEI ought to have a proper thickness,high ionic conductivity, strong mechanical performance, low interfacial resistance, extraordinary stability and steady chemistry during long-term cycles and deep plating-stripping for safe and efficient LMBs.Regretfully, the traditional SEI does not present the above characteristics. There are a lot of strategies that have been investigated to create an artificial and resilient layer to prevent uncontrolled processes during SEI formation. For example, Gong’s group[104]synthesized 2D/2D Ti3C2Tx/g-C3N4nanosheets heterojunction within-situgrown g-C3N4by pyrolysis process and used it as artificial SEI film to inhibit the growth of Li dendrites (Fig. 7a). The synthesis method not only built stable connection points between Ti3C2Txand g-C3N4, but also avoided the agglomeration of Ti3C2Txnanosheets. The as-prepared Ti3C2Tx/g-C3N4hybrid showed a crinkled sheet-like morphology with a thickness of 3.2 nm (Fig. 7b). The obtained composite electrode enabled the conformal Li deposition in the plating process, efficiently increasing the rapid transport of Li-ion. In addition. The primary constituent in the inner layer of SEI was Li3N and LiF deriving from Li interaction with N element of g-C3N4and-F surface termination groups of Ti3C2Txrespectively.The Ti3C2Tx/g-C3N4electrode showed a reduced nucleation overpotential of 14.0 mV , which was significantly less than the electrode materia1 Ti3C2Tx(Fig. 7c). The full-cell when coupled with the cathode of NCM811 (with an areal capacity of 1.2 mAh cm-2),exhibited an initial specific capacity of 167.2 mAh g-1at 0.3 C. This capacity remained stable over 110 cycles with capacity retention of 81.2%.

Fig. 7 (a) Schematic illustration of the fabrication process of Ti3C2Tx/g-C3N4 hybrid[104]; (b) SEM image of Ti3C2Tx/g-C3N4[104]; (c) Voltage-time profiles of Li plating/stripping in different symmetric cells with the capacity of 0.5 mAh cm-2 at 0.5 mA cm-2 and Detailed voltage profiles for selective cycles in panel[104];(d) Schematic diagram of the in-situ generation process of MA-SEI in the Li metal anode interface[105]; (e) Schematic illustration for the Li deposition and migration processes occurring at the Nb2CTX Li-In alloy anode[106]. Reprinted with permission

Liu et al.[105]fabricated a synthetic polymer/2D Ti3C2Tx(MXene) SEI on a lithium metal substrate.Fig. 7d depicts the procedure for producing a lithium metal anode with an artificial solid electrolyte interphase (SEI). Initially, the MXene powders were amalgamated with 1,3-dioxolane (DOL) and 1,2-dimethoxyethane (DME). Subsequently, the suspension was supplemented with particles of Li bistrifluoromethanesulfonimide (LiTFSI) and lithium hexafluorophosphate (LiPF6). Finally, the homogeneous distribution of monomers was expeditiously applied to the lithium foil. The polymer layer offered good interface contact and excellent adaptability to reduce interface resistance. The 2D MXene with a low Li nucleation energy barrier was advanced to uniformly deposited Li and restrain the side reaction. The Li/SSE/Li configuration exhibited a consistent polarization voltage of around 50 mV when operating at a capacity of 0.5 mAh cm-2and kept this exceeding performance for 1 000 h.

Lithium alloy is a promising candidate for LMAs. In 2022, Won[106]used a Li-In alloy as an anode material and combined it with Nb2CTxMXene as an artificial solid-electrolyte interphase, which was referred to as Nb2CTxLi-In. Fig. 7e presents a schematic representation of the Li deposition and migration mechanisms observed on both the bare Li anode and the Nb2CTxLi-In alloy anode. The former demonstrated a notable affinity for Li and facilitates the appropriate migration of Li ions, thereby promoting the stable electrodeposition of Li. Moreover, pouch cells consisted of Nb2CTxLi-In alloy anodes and NCM811 cathodes delivered an energy density of 272 Wh kg-1and 500 Wh L-1. The Nb2CTxLi-In alloy anode could proficiently alleviate the regional expansion caused by Li dendrites and inhibiting side reactions.

The separator is crucial in rechargeable batteries as it prevents an internal short circuit by separating the cathode and anode. The development of effective separators for high-performance lithium metal batteries is a crucial driving force for the prevention of lithium dendrite growth[107-112]. For example, Han[113]and his co-workers designed an engineered interphases separator modified by two-dimensional layered Ti3C2Txand the solid-state electrolyte Li1.3Al0.3Ge1.7(PO4)3(LAGP) synthesized via solidstate reaction method for homogenization Li plating.The homogeneous transmission of lithium ions and electrons was facilitated by the extensively blended conductive properties of LAGP/MXene. The morphological changes of Li metal in full cells was revealed that the Li metal anodes experienced a volume expansion of 257% when paired with PP separators. This expansion was approximately nine times greater than that observed when PP/LAGP+MXene separators were used, which exhibited a volume expansion of only 28%. In addition, the in situ formed Ge, Li3PO4,and LiF interphases, which were derived from the reduction of LAGP, contributed to the stabilization of SEI. The symmetric cells utilizing PP/LAGP+MXene separators exhibited a consistent voltage for Li plating/stripping for a duration of more than 700 h when subjected to a current density of 1 mA cm-2within an area capacity of 1 mAh cm-2. The utilization of the modified separator in the full-cell configuration led to a notable improvement in its cycling stability. The analysis of the action mechanism of the material on the separator provided a new insight for prolonging the life of lithium metal anodes.

3.3 MXene-based electrolyte additives

To overcome the safety problems caused by the flammability of the organic electrolyte in the case of overcharging or short-circuiting, all-solid-state lithium metal batteries (LMBs) have been proposed by the research community. Different species of solid electrolytes could be classified into 4 sections, including polymer solid electrolytes, inorganic solid electrolytes, solid polymer electrolytes, and composite solid electrolytes[114]. Nanocomposite polymer electrolytes(CPEs) exhibited potential as viable materials for allsolid-state lithium metal batteries (LMBs) owing to their improved ionic conductivities, outstanding viscoelasticity, fast film formation, lightweight and stability to the lithium anode[115]. Christopher Y Li and his group studied the critical role of MXenes as an electrolyte addictive on the electrochemical performance of all-solid-state lithium metal batteries at room temperature. To create MXene-based CPEs (MCPEs),they employed a green, facile aqueous solution blending process to uniformly spread small amounts of Ti3C2Txwithin a poly (ethylene oxide)/LiTFSI complex (PEO20-LiTFSI). Their research proved that MXenes as an electrolyte additive could inhibit polymer crystallization, improve the peristalsis of polymer segments, and increase the concentration of carriers. Meanwhile, MXenes nanosheets with negative charges could attract lithium cations to improved transference number and ionic conductivities which are 6 orders of magnitude higher than the CPEs.Meanwhile, it was encouraging to see that the highrate partial state of charge capacity of the MCPEs cell under 1 C rate was 92 mAh g-1. The temperature-dependent conductivity data was tested using the modified Vogel-Tammann-Fulcher (VTF) equation, the results indicated that the CPE with 3.6% (mass fraction) MXene showed a higher ionic conductivity at room temperature (2.2×10-5S m-1at 28 °C)[116]than its counterparts.

4 Summary and outlook

Lithium metal battery is a type of promising energy storage device due to the high energy density.But inherent issues including uncontrollable dendrites and low Coulombic efficiency (CE) impede their commercialization. MXene, as an emerging 2D material, has demonstrated significant potential in addressing the inherent issues associated with metal anodes.This is due to their exceptional mechanical properties,abundant surface functional groups, high electronic and ionic conductivity. In conclusion, we summarized the working principle, construction, and a few important advances of 2D MXene utilized for stable and dendrites-free Li metal anodes in this review, including designing MXene-based LMA hosts, constructing MXene-based artificial SEI films, and modifyin1 MXene-based electrolyte additives. The surface/interface modification involving lithiophilic groups (e.g. Au, Zn, N and S) presented as a promising approach to enhance the affinity for Li metal anode. This strategy could facilitate the creation of additional nucleation sites, promote uniform nucleation,and significantly enhance the stability of LMAs. In addition, designing nanosheet arrays of MXene as lithium metal host electrodes significantly improves the cycling stability and rate capability of LMAs.

The analysis shows that MXene, an emerging 2D material, has been extensively used for stabilizing the LMAs. By modifying the LMAs with 2D MXene, the issues of dendrites growth, large volume changes, and an unstable SEI could be effectively addressed. Despite the improvements in electrochemical performance, commercializing LMAs still face several challenges. To further enhance their praticalibility, future work should prioritize the following aspects:

(1) Enhance the efficiency and decrease the cost of MAX phases and MXenes during the large-scale synthesis, which requires more efforts from both fundamental studies and scale-up engineering techniques.

(2) Combine theoretical studies and experimental research in depth to study and apply various kinds of MXenes besides the Ti-based MXenes.

(3) Although MXene works efficiently on Li metal anodes in preventing dendrite growth and volume changes, more research is required to develop MXene-based skeletons/SEI layers/additives for Nametal, K-metal, and Zn-metal anodes in the future.

(4) Considering the unclear inhibition mechanism of lithium dendrites with MXene-based nanomaterials, advanced characterization techniques, such as cryogenic electron microscopy (cryo-EM), Auger electron spectroscopy (AES), and nuclear magnetic resonance (NMR), etc, could be applied to understand the inherent link among lithium metal nucleation,dendrite growth, defects, crystal morphology, and MXene-based skeletons/SEI layers/additives.

Acknowledgements

The authors acknowledge financial support by the National Natural Science Foundation of China(52225208, U21A20174, 51972285 and 52202314).