From structural ceramics to 2D materials with multi-applications: A review on the development from MAX phases to MXenes

MAX phases (Ti3SiC2, Ti3AlC2, V2AlC, Ti4AlN3, etc.) are layered ternary carbides/nitrides, which are generally processed and researched as structure ceramics. Selectively removing A layer from MAX phases, MXenes (Ti3C2, V2C, Mo2C, etc.) with two-dimensional (2D) structure can be prepared. The MXenes are electrically conductive and hydrophilic, which are promising as functional materials in many areas. This article reviews the milestones and the latest progress in the research of MAX phases and MXenes, from the perspective of ceramic science. Especially, this article focuses on the conversion from MAX phases to MXenes. First, we summarize the microstructure, preparation, properties, and applications of MAX phases. Among the various properties, the crack healing properties of MAX phase are highlighted. Thereafter, the critical issues on MXene research, including the preparation process, microstructure, MXene composites, and application of MXenes, are reviewed. Among the various applications, this review focuses on two selected applications: energy storage and electromagnetic interference shielding. Moreover, new research directions and future trends on MAX phases and MXenes are also discussed.


Introduction
MAX phases are layered carbides and nitrides with the general formula of M n+1 AX n (n = 1-4), where M is an early transition metal, A is a group A element, and X is either carbon and/or nitrogen. The typical MAX phases are Ti 3 SiC 2 , Ti 3 AlC 2 , Ti 2 AlC, V 2 AlC, Ti 2 AlN, etc. Currently, there are over 90 MAX phases. The possible constitutive elements of MAX phases that have been experimentally synthesized are shown in Fig. 1. In Fig. 1, the pink elements can be M, the green elements can be A, and the blue element can be X. Additionally, more constitutive elements of MAX phases are theoretically predicted and more MAX phases are being synthesized.
In the crystal structure of MAX phases, M atoms are near-closed packed with the X atoms filling the octahedral sites to form M n+1 X n layers with ceramic properties. The M n+1 X n layers are separated by A layers with metal properties. Thus, MAX phases combine the properties of ceramics and metals. Like metals, MAX phases are thermally and electrically conductive, easy to be machined with conventional tools, and resistant to thermal shock; like ceramics, they have high strength, excellent high temperature strength, and thermal stability.
As early as the 1960s, several MAX phase members, such as Ti 3 SiC 2 , Ti 2 AlC, Ti 2 AlN, were synthesized and researched [1][2][3]. However, most of the researches were ignored until 1996. In 1996, Barsoum and El-Raghy [4] at Drexel University made Ti 3 SiC 2 samples from TiH 2 , SiC, and graphite by reactive hot pressing (HPing) at 1600 . The fully dense ℃ predominantly single-phase samples of Ti 3 SiC 2 were characterized with remarkable properties. Thereafter, Ti 3 SiC 2 and other MAX phases attracted the attention from academia and industry. The scientists around the world have extensively researched the microstructures, properties, and applications of these materials.
Although MAX phases combined the merits of metals and ceramics, they are generally considered as a family of structural ceramics with special properties and applications. Compared with typical structural ceramics, such as SiC, Si 3 N 4 , MAX phases are machinable and electrically conductive. Therefore, MAX phases can be applied as structural ceramics, requiring complex shape to replace typical structural ceramics. Additionally, MAX phases can be applied as heating elements, nuclear materials, pantograph slide plate for high-speed trains, etc.
The unique properties of MAX phases origin from the layered crystal structure, in which M n+1 X n layers are separated by A layers. The atoms in M n+1 X n layers are bonded by strong M-X bonds, which are ionic and/or covalent bonds. The interlayer bonds between M n+1 X n layers and A layers are weak M-A bonds, which are metallic bonds. Therefore, MAX phases are layer materials. Another well-known layered material is graphite. Thus, MAX phases have similar properties with graphite.
In 2004, graphite was mechanically exfoliated, resulting in a new two-dimensional (2D) material, graphene [5]. As a 2D material, graphene has different properties from its three-dimensional (3D) counterpart, graphite, and can be applied in many fields. Thereafter, graphene becomes a hot research area. Since then, many other 2D materials have been discovered, such as transition metal-dichalcogenides (TMDs, e.g., MoS 2 ), hexagonal boron-nitride (h-BN), and black phosphorous or phosphorene. Because MAX phases and graphite have similar structure and properties, some researchers of MAX phases tried to exfoliate MAX phases to obtain a new 2D material. However, the interlayer bonds are metallic bonds, which are weaker than the intralayer bonds of MAX phases, but stronger than the interlayer bonds of graphite, van der Waals bond. Therefore, MAX phases cannot be mechanically exfoliated. All the attempts to exfoliate MAX phases failed until Ti 3 AlC 2 , a typical MAX phase, was exfoliated by selectively etching off Al atoms in HF solution to form 2D Ti 3 C 2 in 2011 [6]. Then, the same group lead by Prof. Barsoum and Prof. Gogotsi at Drexel University exfoliated a series of Al-contained MAX phases to form many 2D materials [7][8][9][10]. The 2D materials are made by selectively removing A layers from MAX phases. The rest MX layers have graphene-like 2D structure and were named as MXenes. The surface of MXenes made in HF solutions was terminated with F/OH/O due to high surface energy. MXenes made by other methods can be terminated by more elements (Cl, S, Br, etc.) [11,12]. Thus, the general formula of MXenes is M n+1 X n T z , where T is the termination of MXenes. The possible constitutive elements of MXenes are shown in Fig. 1. By now, only part of MAX phases can be exfoliated successfully to make MXenes. Thus, in Fig. 1, the pink elements in shading background can be the constitutive elements of MAX phases and MXenes, while the pink elements in solid background only can be the constitutive element of MAX phases. The blue elements (C and N) can be X of both MAX phases and MXene. T can be F, Cl, Br, O (including OH), S, Se, and Te.
As a novel family of 2D materials, MXenes have different property-combination from other 2D materials. Like graphene, MXenes have high electrical conductivity. However, unlike graphene, MXenes are hydrophilic. Thus, MXenes can be dispersed in water to form colloid solutions, which is the characteristic of clay. Therefore, MXenes can be thought as conductive clay. Since the discovery of MXenes, scientists around the world have been attracted to extensively research these novel 2D materials. Many applications of MXenes have been reported. MXenes can be applied as electrodes in metal-ion batteries and supercapacitors, catalysts, adsorbents for chemicals, electromagnetic shielding materials, sensors, etc.
This article reviews the milestones and the latest progress on the research of MAX phases and MXenes. Especially, different from other reviews on MAX phases or MXenes, this article focuses on the conversion from MAX phases to MXenes. Most authors of this article have the research experience on MAX phases before the discovery of MXenes. From the perspective of ceramic science, we tell the stories of MAX phases and MXenes.

Structure and preparation of MAX phases 2. 1 Discovery of MAX phases
MAX phases were firstly discovered by Nowontny and his coworkers during the 1960s and in that time, they were named as H-phases, including more than thirty 211 phases (Ti 2 AlC, Ti 2 AlN, V 2 AlC, etc.) and two 312 phases (Ti 3 SiC 2 and Ti 3 GeC 2 ) [1][2][3]. In the 1980s, another important 312 phase, Ti 3 AlC 2 , was found by Schuster and Nowotny [13] (Schuster was Nowontny's student). However, these MAX phases have not attracted much attention in scientific community, due to their impure and unstable performance, until the middle of the 1990s. In 1996, fully dense predominantly single-phase samples of Ti 3 SiC 2 were prepared by Barsoum's group and their mechanical and oxidation behaviors were firstly evaluated [4]. Later, several bulk MAX phases (Ti 2 AlC, V 2 AlC, Ti 2 SnC, Nb 2 AlC, etc.) were successfully fabricated by the same group and it was found that they had similar properties with those of Ti 3 SiC 2 [14]. Especially in 1999, Barsoum et al. [15] further discovered Ti 4 AlN 3 , which gave the first prototype of 413 phase and led to the birth of the formula M n+1 AX n or MAX for short.
Since then, numerous research groups focused on the synthesizing and characterizing more MAX phases. A large number of new MAX phases, including Nb 4 AlC 3 [16], Ta 6 AlC 5 [17], Ti 3 IrC 2 [18], Ti 2 ZnC [19], and V 2 SnC [20], were successfully fabricated. Phase diagram was used as an indicator to estimate the thermodynamic stabilities of some novel MAX phases [21].
Besides these MAX phases with the formula of M n+1 AX n (X is C/N), many ternary compounds with similar structures were fabricated. They are not MAX phases because (1) the ratio of M:X is not (n+1)/n, (2) A is not 1, or (3) X is B rather than C/N. However, they have MAX-phase-like structure and are researched as www.springer.com/journal/40145 MAX phases. The typical MAX-phase-like compounds are Ti 5 Si 2 C 3 [22], Ti 7 Si 2 C 5 [23], V 12 Al 3 C 8 [24], Ti 3 Au 2 C 2 [18], Mo 2 Ga 2 C [25], Ti 2 InB 2 [26], and Cr 4 AlB 4 [27]. For the compounds that X is B element instead of C/N (such as Ti 2 InB 2 and Cr 4 AlB 4 ), a new term, MAB phase, was created to name the compounds.
Additionally, the investigation on the MAX phase solid solutions achieved some exciting results. For instance, in 2014, Liu et al. [28] reported the successful synthesis of (Cr 2/3 Ti 1/3 ) 3 AlC 2 , which is an out-of-plane ordered, quaternary MAX phase, abbreviated as o-MAX. In 2017, Tao et al. [29] fabricated (Mo 2/3 Sc 1/3 ) 2 AlC phase, which showed an in-plane ordered structure and was labeled as i-MAX. In 2020, Li et al. [30] prepared a serials of V 2 (Sn,A)C (A = Fe, Co, Ni, Mn) solid solutions and V 2 (Mn x Fe y Co z Ni n Sn 1-x-y-z-n )C highentropy MAX phase. Based on these pioneering works, six o-MAX phases and more than thirty i-MAX phases have been discovered in past five years, which greatly enrich the MAX phase family [31,32].

2 Crystal structures of MAX phases and MAX-phase-like compounds
All ternary MAX phases with the formula M n+1 AX n have the same space group (P6 3 /mmc) [31]. As shown in Figs. 2(a)-2(d), the 211, 312, 413, and 514 MAX phases crystallize in a hexagonal structure and their unit cells consist of edge-sharing M 6 X octahedra interleaved with layers of A elements. Meanwhile, the MX layers are twinned with each other, where A layer acts as mirror plane and thus, giving a rise to "zig-zag" structures along c-axis. The difference between these phases lies in the number of M layers separating the A Fig. 2 Illustration of the unit cell projected long (110). Blue, red, and black balls denote the M, A, and X atoms, respectively. Reproduced with permission from Ref. [33], © Elsevier 2020.
layers. Specifically, it is two in the 211 phases, three in the 312 phases, four in the 413 phases, and five in the 514 phases.
There are some compounds that deviate from the formula M n+1 AX n , which can be labeled as M n+2 A 2 X n . Taking Ti 5 Al 2 C 3 for example, its structure can be described as a combination of half unit cell of Ti 2 AlC and half unit cell of Ti 3 AlC 2 phase (Fig. 2(e)). In the same way, Ti 7 Si 2 C 5 phase can be composed of two half unit cells of Ti 3 SiC 2 and Ti 4 SiC 3 phases, with alternating three and four Ti-C layers between the Si layers.
In addition, some other MAX-phase-like compounds, such as Mo 2 Ga 2 C [25], Ti 3 Au 2 C 2 [18], Nb 12 Al 3 C 8 [34], V 12 Al 3 C 8 [24], and Ti 2 InB 2 [26], are not fulfilled with the formula M n+1 AX n nor M n+2 A 2 X n . For Mo 2 Ga 2 C, Mo 6 C octahedral layers are interleaved with two Ga layers and twinned according to the virtual middle plane between the two Ga layers. The structures of Ti 2 Au 2 C and Ti 3 Au 2 C 2 are similar with Mo 2 Ga 2 C, but their twin orientation is dislocated. For Nb 12 Al 3 C 8 and V 12 Al 3 C 8 , they are essentially carbon-vacancy-ordered 413 phases and belong to P6 3 /mcm space group. Ti 2 InB 2 is MAB phase with X is boron rather than carbon/ nitrogen [26]. The space group of Ti 2 InB 2 is P6m2, which is in fact a hexagonal subgroup of P6 3 /mmc.
If element replacement occurs in the M, A, or X site of ternary MAX phases, quaternary MAX phases can be obtained and most of them are disordered solid solutions, which show the similar crystal structure with ternary MAX phases (P6 3 /mmc). It is noteworthy that there exist two kinds of ordered quaternary MAX solid solutions: out-of-plane ordered MAX phases (o-MAX) and in-plane ordered MAX phases (i-MAX). Specifically, o-MAX phases have a general formula of (M',M") n+1 AlC n , where M' and M" denote early transition metals and n is either 2 or 3. For 312 o-MAX phases ( Fig. 3(a)), the outer M' atoms occupy two 4f Wyckoff sites and the inner M" atoms occupy one 2a Wyckoff site. For P6 3 /mmc group, 2a Wyckoff sites are the sites of (0, 0, 0) and (0, 0, 1/2); 4f Wyckoff sites are the sites of (1/3, 2/3, z), (2/3, 1/3, z+1/2), (2/3, 1/3, -z), (1/3, 2/3, -z+1/2). Differently, for 413 o-MAX phases ( Fig. 3(b)), the outer M' and the inner M" atoms occupy two 4e and two 4f Wyckoff sites, respectively. 4e Wyckoff sites are the sites of (0, 0, z), (0, 0, z+1/2), (0, 0, -z), (0, 0, -z+1/2). It should be noted that some of the M' and M" atoms exchange their positions, thus resulting in an imperfect ordered structure. For instance, in (Mo 0.5 Ti 0.5 ) 4 AlC 3 lattice, the outer M' layers were composed of 77% Mo and 23% Ti, whereas the inner M" layers were 86% Mo and 14% Ti. To date, although theoretical calculations predict the existence of many o-MAX, only six such phases have been successfully synthesized in the laboratory. Obviously, the crystal structure of these o-MAX phases is hexagonal (P6 3 /mmc), which is the same as that of ternary MAX phases.

Preparation of MAX phase bulks
The synthesis of MAX phase bulks or MAX composite bulks should be conducted at high temperatures under vacuum or inert atmosphere. It is involved complicated reactions of related elements and compounds and usually results in unwanted secondary phases. Hence, preparing MAX phase bulks with high-purity attracts worldwide attention over last two decades, and several methods, such as hot isostatic pressing (HIP) [4,14,38], HPing [16,17,28,39], pressureless sintering (PLS) [26,29,40,41], self-propagation high-temperature synthesis (SHS) [42][43][44], spark plasma sintering (SPS) [45][46][47][48], and microwave sintering (MWS) [49,50], have been extensively investigated. Although most MAX phase bulks can be prepared by HIP and HP, only small samples with simple shape can be obtained. Meanwhile, the high temperature and long holding time also result in low efficiency and high cost. Compared with HP and  HIP, SPS possesses rapid heating and cooling rates, relative low synthesis temperature, and short holding time (several minutes). SHS is another rapid synthesis method by which the reactions can be accomplished in several or dozens of seconds. The samples made by SHS are generally porous. To make dense bulk samples, a pseudo-hot isostatic pressing (PHIP) process was combined with the SHS process. The SHS-PHIP method was used to make dense Ti 2 AlC [51,52] and Ti 3 SiC 2 [53] with high performance. However, both SPS and SHS are not suitable for complex shape or batch preparation. From this point of view, PLS is much more economical and scalable but usually needs high temperature, and the obtained MAX phase bulks show low density and purity. Moreover, MWS can be recognized as special PLS method with rapid heating rate. As mentioned above, each preparation method for MAX phase bulks has advantages and disadvantages. The strategy should be adopted depends on the specific applications and actual conditions. The key point of preparation for MAX phase bulks should be focused on composition and proportion of raw materials as well as processing parameters.

Preparation of MAX phase films and coatings
Physical vapor deposition (PVD) and chemical vapor deposition (CVD) are usually adopted to prepare MAX phase films or coatings. The PVD for MAX phase synthesis is far from thermodynamic equilibrium and thus can reduce the temperature substantially. Sputtering technique is the primary PVD method that generally performs at 800-1000 . Palmquist ℃ et al. [54] first demonstrated the feasibility of Ti 3 SiC 2 film using sputtering technique from elemental sources and Ti 3 SiC 2 target. Reactive sputtering is suitable for the preparation of MAX phase nitrides. Joelsson et al. [55] synthesized Ti 2 AlN using reactive sputtering in an Ar/N 2 mixture from a 2Ti:Al target. To further decrease deposition temperature, Rosén et al. [56] investigated Ti 2 AlC synthesis using a pulsed cathodic-arc setup from elemental Ti, Al, and C cathodes at a substrate which allowed a deposition temperature of 900 . Furthermore, ℃ pulsed laser deposition (PLD) was developed and the first attempt on Ti 3 SiC 2 synthesis was conducted by Phani et al. [57], who deposited Ti-Si-C films over the temperature range of  . Differently, the ℃ synthesis of MAX phases through CVD technique is relatively few although the first research paper was published in 1972. Of these limited works, CVD was conducted focusing on the preparation of Ti 3 SiC 2 film and high temperatures (1000-1300 ) were required ℃ [58,59]. Compared with CVD, PVD can not only reduce the synthesis temperature, but also help to find new MAX phases, especially low-temperature metastable phases. Meanwhile, the compositions and defects can be easily regulated by PVD and therefore, beneficial to the fundamental research on MAX phases. A detailed and systematic introduction on MAX phase films processing via PVD and CVD can be seen in Ref. [60].

Preparation of MAX phase powders
The production of MAX phase powders provides precursors not only for MAX phase bulks but also for MXenes. Molten salt assisted synthesis is a kind of liquid phase synthesis, in which molten inorganic salt is employed as the reaction medium and MAX phase powders can be easily obtained by this method. For instance, Xiao et al. [61] synthesized Cr 2 AlC using Cr, Al, and C as starting materials and KCl as molten salt. After holding for 60 min at 1250 under argon ℃ atmosphere, high purity Cr 2 AlC powders were obtained. Wang et al. [62] prepared V 2 AlC using V, Al, and C as elemental powders and NaCl as molten salt. They found that the formation of V 2 AlC starts at 1050 , ℃ which is far less than that of the conventional solid phase synthesis. Liu et al. [63] synthesized homogenous Ti 3 AlC 2 powders with particle size of ~100 nm using 3Ti/Al/2C as starting elemental powders in NaCl-KCl molten salt at 900 for 10 h. Recently, Dash ℃ et al. [64] reported a modified method which named as molten salt shielded synthesis/sintering (MS 3 ). They use KBr to achieve gas-tight encapsulation around a specimen before placing it in a KBr salt bed for further heating. By this method, the synthesis can be conducted in air and several highly pure MAX phases, such as Ti 3 SiC 2 , Ti 3 AlC 2 , Ti 2 AlN, Ti 2 AlC, and Cr 2 AlC, have been successfully prepared at relative low temperatures. In addition to the ternary MAX phases as mentioned above, V 2 (A x Sn 1-x )C (A = Fe, Co, Ni, Mn) solid solutions were synthesized by Huang's group using V, Sn, Fe, Co, Ni, Mn, and C as starting materials and NaCl/KCl as molten salt [30]. This work indicates that molten salt assisted synthesis can offer a simple method not only for the preparation of ternary MAX phases but also for MAX phase solid solutions.
The advantages of molten salt assisted synthesis lie in its low synthesis temperature and small grain size in obtained products. For some strongly bonded MAX phases, small particle size is favorable for subsequent selective etching. Therefore, from the perspective of preparing MXenes, molten salt assisted synthesis provides a good low-cost method to prepare the MAX phase precursors. Although some MAX phases have been synthesized by this method up to now, more MAX phases need to be synthesized. The future work should focus on the optimization of molten salt system and processing parameters to control the composition and morphology of obtained MAX phases.

4 Summary on the structure and preparation of MAX phases
In summary, we only emphasized here on the lattice structure of MAX phases and overlooked the microstructure which is actually related to the preparation methods. Although over 155

Crack healing properties and potential applications of MAX phases
MAX phases are called "remarkable" ceramics because they combine the properties of metals and ceramics [4]. Like ceramics, they are elastically stiff, light-weight, resistant to corrosion and oxidization, and retain strength and stiffness at high-temperature strength. Like metals, MAX phases are thermally and electrically conductive, readily machinable, relatively soft, resistant to thermal shock, and damage-tolerant. Elastically, the MAX phases are in general quite stiff and elastically isotropic. They are generally applied as high-temperature structure ceramics. The mechanical properties of MAX phases at room temperature and high temperatures, such as bending strength, toughness, elastic modulus, are extensively researched and reported. This part mainly focused on crack healing properties of MAX phases. High-temperature structure ceramics are always brittle and sensitive to the presence of surface cracks caused by machining, loading, thermal shock, or fatigue. The presence of such cracks causes the loss in performance or even a sudden catastrophic failure of ceramic components during service, especially in thermal cycle www.springer.com/journal/40145 conditions. If ceramics have self-healing ability to repair cracks in situ and thus restore their initial performance, the advantages will be obvious in improving structural integrity and functional reliability.
Here, we present a summary of major research and current understanding of mechanisms on crack healing behavior of MAX, and future opportunities in MAX applications.

1 Crack healing in air
MAX phases can repair damages (pores, scratches, and cracks) mainly through oxidation reaction mechanism in an oxidizing environment. Oxidation reaction induces the formation of oxides of M-O and A-O to fill the damages, thus recovering the properties. Ti 3 AlC 2 was the first MAX phase reported with crack healing ability in 2008 [81]. It healed a crack with a size of 7 mm in length and 5 m in width created by tensile deformation. The crack was completely filled by a primary phase of α-Al 2 O 3 and some rutile-TiO 2 after healing at 1100 for 2 h in air. The formed Al ℃ 2 O 3 layer was well bonded to the matrix, but a larger fraction of TiO 2 formed within the crack gap weakened the adhesion between matrix and oxides. Hence, the hardness and modulus measured in the healed zones in Ti 3 AlC 2 increased by about 14% and 3%, respectively. It should be realized that good bonding between crack-filling phases and healed zones can endow the healed materials with improved performance.
For Al-containing MAX phases, the selective-oxidation formed Al 2 O 3 has strong adhesion to MAX matrices [82]. Therefore, the more Al 2 O 3 is in cracks, the higher the recovery performance achieves. In the MAX phase family, Ti 2 AlC has about 25 at% Al, higher than 17% at% Al for Ti 3 AlC 2 , indicating that more Al 2 O 3 can be formed to fill cracks. In 2012, the crack healing behavior of Ti 2 AlC was firstly investigated [83]. Damages made by Knoop indentations and fatigue test in Ti 2 AlC caused a drop in strength, from 211 MPa of the initial strength to 152 MPa. All damaged samples were fractured along the Knoop dents ( Fig. 4(a)). After healing at 1200 for 2 h in air, the healed samples became ℃ Fig. 4 Optical images of (a) the damaged samples after three-point bending test, which were fractured along the Knoop indents; (b) the healed sample after testing; (c) cross-sectional back-scattered SEM for the healed indent as shown in (b); (d) flexural strength of the virginal, damaged, and healed Ti 2 AlC samples. Error bars represent the standard deviation. Reproduced with permission from Ref. [83], © Elsevier Ltd. 2012. stronger due to the fact that these samples did not fracture through but just outside the damage-healed zones ( Fig. 4(b)). The cross-sectional SEM micrograph demonstrates that the bottom of the Knoop dent was covered and the microcracks were filled by only Al 2 O 3 ( Fig. 4(c)). The Ti 2 AlC ceramic showed an attractive crack healing ability to recover the strength up to 224 MPa, even slightly exceeding the initial strength of 211 MPa (Fig. 4(d)). The above feature indicates that the adhesion between Al 2 O 3 in the crack gap and the Ti 2 AlC matrix is strong. Yang et al. [97] used in situ microcantilever bending test to conveniently evaluate the strength of Ti 2 AlC-Al 2 O 3 interface after healing at 1200 for 1.5 ℃ h in air. The experimental results demonstrated that the bonding strength of Ti 2 AlC-Al 2 O 3 interface is beyond the cleavage strength of Ti 2 AlC grains and the strength of healed grain boundary is two times the strength of original grain boundary, suggesting that the Ti 2 AlC ceramic may achieve full recovery of strength after healing at high temperatures.
Especially, the Ti 2 AlC MAX phase exhibits a repeatable crack healing ability [83]. A through-thickness crack with a length of ~2.0 mm was introduced by single edge notched beam (SENB) method after loading in three-point bending. After healing at 1200 for 2 h ℃ in air, the millimeter-sized crack was completely healed ( Fig. 5(a)). The Ti 2 AlC samples were repeatedly cracked and healed under the same conditions, but the healed crack gap became wider and a part of crack in the healed region was no longer completely filled in 2 h of healing at 1200 in air ( Fig. 5(b)). Seven healing cycles after ℃ repeated cracking at a given location in Ti 2 AlC have been achieved. The fracture toughness (K IC ) recovery in a range of 3-6.7 MPa·m 1/2 was also quantified during the repeatable healing cycles (Fig. 5(c)).
Interestingly, a prolonged healing time up to 100 h at 1200 in air still leads to the complete filling of ℃ the wider crack gap by Al 2 O 3 and TiO 2 ( Fig. 6(a)). The thinner crack gap sections were mainly filled by Al 2 O 3 , but the wider crack sections were filled by a mixture of Al 2 O 3 covering the two opposite crack surfaces and TiO 2 distributing in the center of the healed zone (Figs. 6(b) and 6(c)). Sloof et al. [98] further investigated the rate and position dependence of crack repair in pristine Ti 2 AlC material and in previously healed cracks using the 4D in situ synchrotron X-ray tomographic microscopy. The repeated crack healing test results demonstrated that the healed cracks have sufficient mechanical integrity to make subsequent cracks form elsewhere upon reloading after healing.
Crack healing behavior of ceramics has been investigated over 50 years and demonstrated for single or one-time-only healing event per damage site. So far, the multiple healing performance has been only displayed in self-healing polymers with microvascular networks [99]. The Ti 2 AlC MAX ceramic is such a unique rigid material exhibiting repeatedly repairing ability. The crack healing ability combined with other attractive high-temperature properties makes Ti 2 AlC promising in critical high-temperature applications.  After that, the crack healing behavior of Ti 2 AlC and its solid solutions as filler and coating [86,87,100,101] has been studied. Bei et al. [86] prepared Ti 2 Al 1-x Sn x C solid solutions by substitution of Al by Sn, and found that solution of Sn in Ti 2 AlC reduced the occurrence of oxidation reaction between Al and O from 900 to 460 . They proposed that Ti ℃ 2 Al 1-x Sn x C solid solutions may repair damages at reduced temperatures as exposed to air. Then they further investigated the crack healing in Al 2 O 3 composites reinforced with 5-20 vol% Ti 2 Al 0.5 Sn 0.5 C [87]. The experimental results revealed that the Ti 2 Al 0.5 Sn 0.5 C is an efficient repair filler to trigger oxidation-induced crack healing in Al 2 O 3 ceramic composites at temperatures below 1000 . The crack ℃ section near the surface of Ti 2 Al 0.5 Sn 0.5 C/Al 2 O 3 composite was completely filled with solid oxidation products of TiO 2 , Al 2 O 3 , and SnO 2 , but the crack section in deep was filled with TiO 2 , Al 2 O 3 , and Sn after healing at 900 for 3 h in air. With increasing Ti ℃ 2 Al 0.5 Sn 0.5 C filler content from 5 to 20 vol% and healing time, the recovered strength of healed composites increased correspondingly after healing at 900 . For the 20 vol% ℃ Ti 2 Al 0.5 Sn 0.5 C/Al 2 O 3 composite, the strength recovered fully, even higher than the initial strength level, after healing at 900 for only 0.5 h in air. ℃ Thereafter, Boatemaa et al. [100] also investigated the oxidation-induced crack healing of Al 2 O 3 containing 20 vol% of Ti 2 AlC. The 20 vol% Ti 2 AlC/Al 2 O 3 can repair indentation induced surface cracks of ~80 m in length and 0.5 m in width after healing at temperatures above 800 . At such temperatures, the cracks are ℃ filled by both TiO 2 and Al 2 O 3 resulting from the oxidation reaction of the Ti 2 AlC filler. A fully recovered strength can be achieved after healing at 1000 for 1 ℃ h in air. Wang et al. [101] found that a Ti 2 Al 0.6 Sn 0.4 C solid solution coating can effectively repair microcracks induced by indentation after healing at 700 for 10 h. ℃ The cracks were filled by SnO 2 and some rutile TiO 2 . These studies suggest that Ti 2 AlC and its solid solutions are an efficient reinforcement to not only improve the performance of ceramic matrix composites (CMCs), but also endow CMCs with highly attractive crack healing ability.
Moreover, Ti 3 SiC 2 , Cr 2 AlC, and Ti 2 SnC MAX phases have also demonstrated with crack healing ability. For example, Ti 3 SiC 2 achieved an increase in strength after oxidation at 1000 for 2 h in air [102]. Filling the ℃ surface scratches and micro-pores by oxidation induced oxides of TiO 2 and SiO 2 should be one of the mechanisms for improving flexural strength from the initial value from ~280 to ~650 MPa. Li et al. [85] studied the crack healing behavior of a fine-grained Cr 2 AlC (~2 m) ceramic with different sized cracks after annealing at 1100 in air for 4 ℃ -100 h, and reported that the strength recovery of Cr 2 AlC has a strong dependence on crack length and healing time, with better strength recovery after increasing healing time. The indentation induced crack was completely filled with only α-Al 2 O 3 after healing at 1100 for 4 h in air. Pei ℃ et al. [94] studied the crack healing behavior of Cr 2 AlC MAX phase with X-ray tomography, and further confirmed the above conclusion. The advantage of only α-Al 2 O 3 presenting in the healed zones is the generation of compressive stresses around the crack-healed zone, producing by the thermal expansion mismatch between Cr 2 AlC (12× 10 -6 -13×10 -6 ℃ -1 ) and Al 2 O 3 (7.5×10 -6 -9.6×10 -6 ℃ -1 ) upon cooling down from high temperatures. The compressive stresses mechanically lock the two sides of the crack and impart the healed material with an enhanced strength recovery.
Unlike the filling phases in Ti 3 AlC 2 and Ti 2 AlC, the only filling phase of α-Al 2 O 3 in the healed zones in Cr 2 AlC can bond the crack planes strongly, providing optimal strength recovery after healing. However, once the Al 2 O 3 layers are formed on the crack surfaces, it would retard the outward diffusion of Al from Cr 2 AlC and inward diffusion of O, inducing the slow growth of Al 2 O 3 layers. The slow growth of Al 2 O 3 in the damaged zones causes a long period of healing. Therefore, promotion of the rapid growth of Al 2 O 3 should be considered in the crack healing of MAX materials. Berger and Boucher [103] reported that the addition of yttrium (Y) is one of the ways to promote the growth of Al 2 O 3 in Cr 2 AlC. Y-containing Cr 2 AlC coating exhibits an excellent self-healing ability to repair microcracks of 43 μm in length and 200 nm in width at 900 for ℃ 1 h in air [103]. To further improve the healing efficiency of MAX phases, controlling the crack-filling phases with chemical stability and mechanical match to the MAX phases and promoting growth of such filling phases is necessary to consider.
Ti 2 SnC is another attractive MAX phase with crack healing capability in an oxidizing environment [80]. Ti 2 SnC can repair thermal shock induced cracks of ~3 mm in length at a relatively low temperature of 800 within only 1 h in air. The healed samples ℃ demonstrated that both the low flexural strength and electrical conductivity are almost fully recovered. During healing, a high oxygen partial pressure near the sample surface induces the oxidation of Ti 2 SnC to form TiO 2 and SnO 2 . The large cracks on the surface are completely filled by TiO 2 and SnO 2 , contributing to the recovered mechanical properties. The small cracks especially deep in the material provide a low oxygen partial pressure. Sn easily diffuses from Ti 2 SnC due to its low migration energy of 0.66 eV [104]. Under such condition, Sn diffusion from Ti 2 SnC may not be triggered to oxidation but remains a metallic element to fill the small cracks. The presence of metallic Sn in the healed cracks should play a key role in the recovery of electrical conductivity. The healing temperature for Ti 2 SnC decreased greatly as compared with those of 1100-1200 required for Al ℃ -containing MAX phases (Ti 3 AlC 2 , Ti 2 AlC, and Cr 2 AlC).
On the basis of the above studies, it can be found that Al-containing MAX phases are the promising crack healing materials due to the formation of stable Al 2 O 3 filler which is well adhering to the crack planes. However, it should be noted that the Al 2 O 3 scale is a good barrier for the outward diffusion of M and Al elements and inward diffusion of O. Once the Al 2 O 3 scale is formed on the crack surfaces, healing of wider cracks or multiple healing at the same damage zones requires a long period time. Therefore, the formation of suitable crack filling oxides (e.g., ZrO 2 ), which allow continuous transport of O in MAX phases or solid solutions, may be another way to achieve the complete healing of wide cracks in the MAX phases.
In addition, the MAX phases containing low melting point of A elements (A = Ga, Cd, Sn, In, and Pb) may be of interest to achieve crack healing at relatively lower temperatures for a short time period due to low cohesive bonding energy and migration energy of these low melting point of A metals [104].
As compared with binary ceramics, some MAX phases can heal even millimeter sized cracks at temperatures below 1200 for 1 ℃ -4 h in air. To improve the healing efficiency of MAX phases, relatively low temperatures and short time periods are required. In addition, the filling phases should be stable and have strong bonding with the crack planes to recover original performance and functionality. Furthermore, the above research mainly focused on the crack repair of MAX phases by the oxidation induced healing mechanism in oxidizing atmosphere. It is highly desirable to know whether MAX phases can heal cracks in vacuum or low oxygen partial pressures, and to understand the corresponding healing mechanisms.

2 Crack healing in low oxygen partial pressures
MAX ceramics and their structural components have potential applications ranging from combustion gas turbines to heat elements for industrial vacuum furnaces. The gas turbine generally operates under combustion gas with lower oxygen partial pressure than that in air, and vacuum furnace works under vacuum or inert atmosphere. Under such environments, it is not clear whether MAX phases can repair damages to prolong their service life and maintain the structural integrity and safety. There is little information focusing on the crack healing behavior of MAX phases in low oxygen partial pressures. Cr 2 AlC was the first one in MAX phase family exhibiting crack healing ability in water (with low oxygen pressures) during thermal shock [88]. Upon quenching in water, the formed cracks are simultaneously filled by formation of Al 2 O 3 and Cr 7 C 3 from an intensive oxidation between Cr 2 AlC and H 2 O. The reactants repair thermal shock cracks and adhere well to the crack faces, endowing Cr 2 AlC with abnormal strength increase after certain quenching temperatures. The crack healing mechanism clearly explains the abnormal thermal shock behavior (i.e., the residual www.springer.com/journal/40145 strength of as-quenched samples decreases gradually with increasing quenching temperatures and then unbelievably increases after quenching from even higher temperatures) of MAX phases which puzzles researchers for many years.
Ti 2 SnC was another MAX phase to be demonstrated with crack healing ability in a low oxygen partial pressure of 5 Pa [89]. After healing at 800 for 1 h at ℃ 5 Pa, only metallic Sn was detected cracks introduced by water-quenching (Figs. 7(a) and 7(b)). The low migration energy of Sn may trigger the precipitation of Sn from Ti 2 SnC. The precipitation Sn induced crack healing is the main mechanism for Ti 2 SnC to achieve a completely recovered electrical conductivity at low oxygen partial pressures.
Oxidation-induced crack healing is the main mechanism in MAX phases in air, but precipitation induced crack healing plays a predominant role in MAX phases containing low melting point of A elements (A = Ga, Cd, Sn, In, and Pb) at low oxygen partial pressures. Other crack healing mechanisms may be existed upon healing of MAX phases in oxidizing environments or at low oxygen partial pressures, which still need to be further investigated.

3 Potential applications of MAX phases
The distinct crack healing ability [80,81,83,85,87,89,98,100,103] and other attractive high temperature properties including high strength [47,48], nonsusceptibility to thermal shock [88,90,91], good oxidation resistance [48,62,105], and good ablation resistance [95], make the MAX phases the materials of choice for applications in combustion gas turbines, heat elements, and liners for industrial furnaces, nozzles, hypersonic vehicles, refractories, and coatings. In addition, MAX particles are also good healing agents in ceramic and metal matrix composites to repair damages. Key components made of such MAX phases with self-healing ability are highly expected to repair cracks in situ and thus restore initial mechanical performance, improve device reliability, and decrease replacement and maintenance costs.
Ti 2 AlC, Cr 2 AlC, and Ti 3 AlC 2 are the most studied and developed products. For example, heating elements made of Ti 2 AlC were developed by the company 3-ONE-2, LLC (Fig. 8). Ti 2 AlC heating elements can in situ repair cracks induced by thermal cycling during use at high temperatures in oxidizing environments. Instantaneous crack healing during service may prolong their service life.
Except for applications of MAX phases as structural components, their self-healing coatings as oxidation and thermal barriers also attract much attention. For example, Cr 2 AlC MAX phases have been developed as high-temperature protective coating on refractory alloys [106]. MAX phase self-healing coatings may also be used on gas turbine components because service temperatures (> 1250 ) are high enough to trigger ℃ oxidation-induced crack healing in coatings.
MAX reinforced metal (Cu, Fe, Al, Ni, Ag, etc.) composites have been developed for electrical applications, including pantographs and contact materials. Ti 3 SiC 2 /Cu vacuum contactor materials are used in high power vacuum switches ( Fig. 9) [107]. MAX/Cu composites have been developed as pantographs and applied successfully in high-speed trains with speeds of 250-300 km/h in China [108]. Ti 2 SnC/Ag composites as electrical contacts have also been developed [105]. These key components with MAX self-healing agents may repair cracks by electrochemical oxidation induced crack healing mechanism.

1 HF solution etching
Since the remarkable studies on preparing dense Ti 3 SiC 2 samples by HIP [4], the research of Ti 3 SiC 2 and other MAX phases attracted the worldwide attention from academia and industry. Most MAX phases were applied as structural materials in that period. However, novel functional applications of MAX phases were attempted by many groups. MAX phases have layered structure similar with graphite. Graphite has an important application as anode of lithium-ion-batteries (LIBs). Graphite is the first commercialized anode material of LIB. Considering the similarity between MAX phases and graphite, Prof. Barsoum cooperated with Prof. Gogotsi to apply a project from U.S. Department of Energy to research the application of MAX phases as anode of LIB.
Unfortunately, the research could not obtain desirable results at early stage. The measured capacities were always very low, because the interlayer bonds of MAX phases are too strong. Li ion cannot enter the interlayer space. Interlayer bonds of MAX phases need to be broken. Chemical etching is a common method to break chemical bonds. Many acids were tried to etch many MAX phases. HF acid was selected to etch Ti 3 SiC 2 , because HF is usually used to etch glass and both glass and Ti 3 SiC 2 contain Si. However, HF etching did not work on Ti 3 SiC 2 . However, it worked well on Ti 3 AlC 2 , another common MAX phase. Al was selectively etched off from Ti 3 AlC 2 . The first MXene, Ti 3 C 2 MXene, was successfully prepared [6].
In that work, many bubbles were generated if Ti 3 AlC 2 powders were put into HF solution. After 2 h etching, black powders were separated by centrifugation and rinsing from the etching solution. The black powders were characterized. It was found that Al element was etched off from Ti 3 AlC 2 . The black powders were the Ti 3 C 2 left. However, due to high surface energy, the surface of Ti 3 C 2 was terminated by OHor F -. Ti 3 C 2 T x was used as the chemical formula of the etched product, where T represents OH and/or F. The SEM images of Ti 3 AlC 2 before and after etching are shown in Figs. 10(a) and 10(b), and the obtained sample had the accordionlike structure of stacked multilayer sheets. From the XRD patterns, almost all the peaks of Ti 3 AlC 2 disappeared except the (002) peak shifted low angle. The strong (002) peak means that there is periodicity in c direction due to the stacked multilayer structure. The low-angleshift of this peak means the increasing of lattice parameter c. The c increased because F and/or OH replaced Al in the Ti 3 AlC 2 crystal and the interlayer space between Ti 3 C 2 T x sheets was generated. The multilayer Ti 3 C 2 T x samples are electrically conductive and have interlayer space for Li storage and motion, which are desirable properties for anode of LIB. The reactions to convert Ti 3 AlC 2 to Ti 3 C 2 T x MXene [6] were proposed as Ti 3 AlC 2 + 3HF = AlF 3 + 3/2H 2 + Ti 3 C 2 (1) In Reaction (1), Al is removed from Ti 3 AlC 2 to obtain Ti 3 C 2 2D sheets. In Reactions (2) and (3), the surfaces of Ti 3 C 2 are terminated by -F and/or -OH due to the high surface energy of Ti 3 C 2 2D sheets.
Following this strategy, many MAX phases with Al as A elements, such as Ti 2 AlC, Ta 4 AlC 3 , (Ti 0.5 ,Nb 0.5 ) 2 AlC, (V 0.5 ,Cr 0.5 ) 3 AlC 2 , and Ti 3 AlCN, were etched by HF solution and corresponding 2D MXenes were prepared [8]. The SEM images of the some MXenes (Ti 3 C 2 , Ti 2 C, and Ta 4 C 3 ) are shown in Fig. 10. All the MXenes have stacked multilayer structure. The accordion-like multilayer structure is a feature microstructure of MXenes. The materials with this structure have some periodicity in the direction and only in the direction perpendicular to the layer plane. Thus, the XRD patterns of these materials only have diffraction peaks related with (00l) planes. The microstructure of MXenes is discussed in details in the following section.
In the first two important Refs. [6,8] on the synthesis of MXenes, the etching temperature was claimed to be "room temperature". The "room temperature" is not a fixed temperature. Especially, it is confusing for the researchers working on structural ceramics, such as MAX phases, because the temperature to synthesize or sinter MAX phases is generally above 1300 . The ℃ temperature below 100 can be considered as room ℃ temperature by the researchers working on MAX phases. And these researchers easily ignored the impact of temperature on the etching reactions to synthesize MXenes. Thus, although the method to prepare MXenes was reported in 2011, only a few groups could experimentally prepared MXenes with high purity by the method before 2014.
Thereafter, the impact factors on the etching reactions were researched to clarify the synthesis procedures. At a research in 2013 [109], Ti 3 AlC 2 powders were etched by 50% HF solution at different temperatures for 2 h or at room temperature for different time. The XRD patterns of the etched samples are shown in Fig. 11. The peak with 2θ = 39 corresponds to the (104) peak of Ti 3 AlC 2 . This peak diminishes with increasing temperature and time, disappearing above 50 and ℃ after 15 h, respectively, indicating the completely consuming of Ti 3 AlC 2 . Concomitantly, the intensity of the (002) peak at 2θ ≈ 9, corresponding to the basal planes of the 2D Ti 3 C 2 -based layers, increases. Therefore, higher temperatures, long reaction time, and small initial particle sizes, are favorable for the synthesis of MXenes.
Zhou and his coworkers [110,111] found that Ti 3 AlC 2 was difficult to be selectively etched at "uncontrolled" room temperature. Below 15 , the etched products ℃ always contained some Ti 3 AlC 2 besides the desired Ti 3 C 2 MXene. Therefore, the etching should be carried out in oil bath environment to control temperature. Higher temperature, such as 60 , is favorable for the etching ℃ process. Two kinds of Ti 3 AlC 2 powders were used for Ti 3 C 2 MXene synthesis. Compared with hot-pressed Ti 3 AlC 2 from element powders, pleasureless-synthesized Ti 3 AlC 2 made by tube furnace from TiH 2 as Ti source is a cheaper starting material to make MXene with highly preferred orientation. A suitable condition to synthesize MXene was 24 h at 60 in 49% HF ℃ solution.
The MXene preparation can be expedited by controlling temperature and procedure. Li and his coworkers [112][113][114] reported a method to rapidly prepare Ti 3 C 2 MXene. Ti 3 AlC 2 powders were immersed in 40% HF solution in a polytetrafluoroethylene (PTFE) container. To shorten the etching time and improve the reaction efficiency, a magnetic stirring device was used to maintain stirring and heating throughout the whole etching process. The time was only 0.5 h to prepare Ti 3 C 2 MXene by exfoliating Ti 3 AlC 2 HF solution at 50 .  In Ref. [115], LiF+HCl, instead of HF, was used to make Ti 3 C 2 MXene. This method is high-yield, safer, easier, and provides a faster route to delaminated MXene multilayers. As shown in Fig. 12, the Ti 3 C 2 MXene made by this method was clay-like material. The multilayer structure was fully delaminated by the intercalation of Li + during the etching process. The single layers of MXene are hydrophilic, swell in volume when contact with water, and can be shaped like clay and dried into a highly conductive solid or rolled into films tens of micrometers thick. Therefore, MXene can be deemed as "conductive clay", for a host of applications.
Inspired by the idea that HF is replaced by LiF+HCl, other fluorides are also be used as etchant with HCl to make MXene. Liu et al. [116] reported the preparation of Ti 3 C 2 MXene and Ti 2 C MXene by etching Ti 3 AlC 2 and Ti 2 AlC with etching of LiF, NaF, KF, and NH 4 F with HCl. The influences of temperature and time on the exfoliating process were researched and clarified. Based on theoretical calculation and XPS results, the type of positive ions (Li + , Na + , K + , or NH 4 + ) in etchant solution affects the surface structure of prepared MXenes. Residual Li was detected on the surface of MXenes exfoliated by LiF with HCl. However, almost no residual K and Na was detected on the surface of corresponding MXenes. The difference of residual positive ions, in turn, affects the methane adsorption properties of MXene.
The etchants making MXene have different etching abilities, which can be changed by controlling the ions other than For H + in the etchants. The "etching ability" includes "exfoliate ability" and "corrosivity". Exfoliate ability is the ability to exfoliate MAX phases to make MXenes. Corrosivity is the ability to destroy the MAX phase and obtain the MXenes. Generally, the corrosivity of HF is stronger than fluoride+HCl. Due to the high corrosivity, some MXenes that are difficult to be made by HF etching, however, can be easily made by other etchants. For example, V 2 C MXene, another MXene, made by HF etching was always not pure. This is because HF has very strong corrosivity. Long time etching by HF destroys the structure of obtained V 2 C MXene; however, short time etching cannot fully exfoliate the precursor, V 2 AlC. Zhou and his coworkers [117,118] explored different etchants to make highly pure V 2 C MXene at 90 . As shown in ℃ Fig. 13, only NaF+HCl solution was effective for synthesizing highly pure V 2 C MXene. The existence of sodium (Na + ) and chloridion (Cl − ) in etching solution was essential for the synthesis of V 2 C MXene. The purity of as-synthesized V 2 C MXene was > 90 wt% with a few impurities of Na 5 Al 3 F 14 and V 2 AlC. The V 2 C MXene made by NaF+HCl etching at 90 was ℃ much purer than those made by HF etching at room temperature.

3 Hydrothermal etching
In the etching process, high temperature is generally beneficial to the synthesis of MXenes. However, the temperature cannot be higher than the boiling point of water, 100 . That is why 90 was selected in the ℃ ℃ synthesis of V 2 C MXene in previous works [117,118]. If the etching process is carried out in a hydrothermal reactor, instead of opened environment (OE), such as an oil bath pan, the etching temperature can be higher than 100 . This is ℃ the hydrothermal etching process. By this method, the exfoliating ability of etchant is further improved. And compared with the etching process in OE, the etching proses in closed reactor is safer to the experiment operators. The hydrothermal etching can be used to make MXenes or other 2D materials [119].
In 2016, Wang et al. [120] developed the hydrothermal method to prepare Ti 3 C 2 MXene. NH 4 F was dissolved in deionized water as etching solution instead of HF acid. Then Ti 3 AlC 2 and NH 4 F solution was sealed in a Teflon-lined stainless-steel autoclave, kept at 150 ℃ for 24 h, and then allowed to cool to room temperature naturally. By this method, highly pure Ti 3 C 2 MXene can be made.
The hydrothermal etching can be carried out at higher temperatures and higher pressures, compared with traditional etching in OE. Thus, this method has higher ability to exfoliate MAX phases due to the temperature and pressure. Then mild etchant (NH 4 F) can be used to replace strong etchant (HF).
Moreover, some MXenes that are difficult to be made can be made by this method. By this method, with fluoride+HCl as etching solution, V 2 C MXene was made from V 2 AlC [121,122] and Mo 2 C MXene was made from Mo 2 Ga 2 C [123]. Additionally, the effect of etching environment (opened for traditional method or closed for hydrothermal method) on the synthesis of V 2 C MXene was studied [122]. As shown in Fig. 14, in OE, only NaF+HCl etching solution can be used to synthesize highly pure V 2 C MXene. However, in closed environment (CE), both LiF+HCl and NaF+HCl etchants can be used to prepare V 2 C MXene. Moreover, the V 2 C MXene samples made in CE had higher purity and better layered structure than those made in OE.
Although the purity of V 2 C obtained by LiF+HCl is lower than that of V 2 C obtained using NaF+HCl, it shows better electrochemical performance as anode of LIBs. Therefore, hydrothermal etching is a good method for preparing highly pure MXene.

4 Molten salt etching at high temperatures
In previous sections, aqueous solutions were used as etching solution to make MXene. Some alkalis and salts can be molten at high temperatures and be used to etch MAX phases to make MXenes. In 2013, inspired by the method to make porous active carbon, Sun et al. [124] attempted to etch Ti 3 AlC 2 by KOH at 700 . ℃ However, this attempt failed. The KOH etching method cannot make Ti 3 C 2 MXene. In 2016, Urbankowski et al. [125] used molten fluoride salt (KF/LiF/NaF) to etch Al from a Ti 4 AlN 3 powder precursor at 550 ℃ under an argon atmosphere. By this method, Ti 4 N 3 was successfully made. The MXene made by this method was still terminated by F.
In 2019, Li et al. [19,126] used molten ZnCl 2 as etchant to make MXene without F/OH termination. First, novel MAX phases Ti 3 ZnC 2 , Ti 2 ZnC, Ti 2 ZnN, and V 2 ZnC were synthesized by the replacement reaction at 550 between the Zn element from molten ZnCl ℃ 2 and the Al element in MAX phase precursors (Ti 3 AlC 2 , Ti 2 AlC, Ti 2 AlN, and V 2 AlC) [126]. Then, excess ZnCl 2 was employed to derive Cl-terminated MXenes (such as Ti 3 C 2 Cl 2 and Ti 2 CCl 2 ) by a subsequent exfoliation of Ti 3 ZnC 2 and Ti 2 ZnC due to the strong Lewis acidity of molten ZnCl 2 [19]. Moreover, in 2020, a redox-controlled A-site etching of MAX phases in Lewis acidic melts was proposed and validated by the synthesis of various MXenes from unconventional MAX-phase precursors with A elements, Si, Zn, and Ga [12].
Thus ZnCl 2 etching can be used to make MXenes with Cl termination. Moreover, MXene with other terminations also can be synthesized by the etching of other molten salts. In 2020, Kamysbayev et al. [11] introduced a general strategy to install and remove surface groups by performing substitution and elimination reactions in molten inorganic salts. By this strategy, MXenes with O, NH, S, Cl, Se, Br, and Te surface terminations, as well as bare MXenes (no surface termination), were successfully synthesized. Br-terminated Ti 3

5 Other etching methods to make MXenes
Besides aforementioned methods, many other etching methods were proposed and researched to make MXenes by more etchants from MAX phases. In 2017, electrochemical etching was used to form a layer of Ti 2 CT x MXene on porous Ti 2 AlC electrodes in dilute hydrochloric acid [127]. Thereafter, Yang et al. [128] prepared Ti 3 C 2 MXene by electrochemically etching www.springer.com/journal/40145 method based on the anodic corrosion of Ti 3 AlC 2 phases in an ammonium chloride and TMAOH electrolyte. In 2018, Li et al. [129] used alkaline-etching to synthesize Ti 3 C 2 MXene without F termination from Ti 3 AlC 2 by a high-temperature hydrothermal approach in NaOH solution at 270 for 12 h. In 2020, ℃ Natu et al. [130] etched and delaminated Ti 3 C 2 MXenes in the absence of water, by using organic polar solvents in the presence of NH 4 HF 2 .

Microstructure of MXenes
Investigations into the structure of MXenes play essential roles in understanding the properties of MXenes. The layer structure and surface terminations impose a significant influence on mechanical properties and electronic structures of MXenes. To fully exploit the advantages of MXenes, it is critically important to fully understand the structure of MXenes. Until now, wide ranges of advanced characterization techniques, such as scanning electron microscopy (SEM), transmission electron microscopy (TEM), scanning transmission electron microscopy (STEM), have been widely used for distinguishing the microstructure of MXenes. A summary on the difference in microstructure of MXene from various characterization techniques is stated in the following section.

1 Morphology of MXenes under SEM
Generally, the transformation of MAX phases to MXenes with different shapes, sizes, and morphologies is related to the HF concentration, the HF-containing etchants, the reaction time and temperature, the delamination or intercalation agent (dimethyl sulfoxide, NH 3 ·H 2 O, and urea), the source of available precursors (solid-liquid reaction synthesis method, molten salt method), the drying methods (oven drying, freeze drying, and supercritical drying), and the post-treatment of MXenes (heat treatment in various atmosphere, such as NH 3 , Ar) [109,115,133,134].
MXenes are generally obtained through the selective etching of the reactive A layers from the MAX precursors. Notably, concentration of etching solutions, duration of immersion, and temperature affect the morphology of as-prepared MXenes. For example, high concentration HF solution-etched MXene (usually > 40%) always exhibits a slack accordion-like morphology, while when the diluted HF solution (usually < 10%) was adopted, the morphologies of MXenes changed into thicker multilayer lamellas with fewer inter-stack gaps, as shown in Figs. 15(a) and 15(b). More importantly, the low concentration HF etched-MXene always possesses more open interlayer space and accommodates more charge carrying ions between the interlayers, enabling more charge carrying ions to be more readily accessible to the active sites and deliver higher capacity. Moreover, another one important factor to affect the morphology of MXene is duration of immersion in HF solution and reaction temperature. The longer the etching duration, the thinner the stacked sheets were obtained. When the etching duration is long enough, the thin sheets are separated from each other [109,135]. As the etching temperature increases, the etching time decreases. Normally, MXenes with higher "M" atomic numbers require higher HF concentration or longer etching time to achieve complete conversion.
Although HF etching was widely used as a facile and versatile method to obtain MXenes, it was still a harsh way with high toxicity and danger to handle and probably resulting in plenty of defects on prepared MXenes. Therefore, an in situ formed HF, via mixing fluoride salts (LiF, NaF, and KF) and acids (HCl, H 2 SO 4 ), can be used to selectively etch the A atoms through similar reactions with less dangerousness.
Recently, the as-synthesized multilayer Ti 3 C 2 T x and Mo 2 CT x from the HCl/LiF mixture solution route as initially proposed by Ghidiu et al. [115], can be fully and easily delaminated by mere sonication in water. This dramatically effective delamination has been attributed to intercalated water and possible cations between the multilayer MXenes acting as a spacer and lubricant that facilitated shearing. Consequently, the larger single-or few-layer MXene flakes with fewer defects are obtained readily by hand-shaking without any sonication or further intercalation [139]. The obtained flake sizes and shapes all keep the grain sizes and shapes of the precursor MAX phase powder www.springer.com/journal/40145 [137]. Moreover, the obtained Ti 3 C 2 T x from the HCl/LiF mixture solution route has a structure as thin as cicada's wings by means of microwave irradiation in glycerin, as shown in Fig. 15(c). Thus, one can choose the appropriate MAX phases to develop the required MXenes according to specific research goals. In addition, reducing the initial particle size of the MAX phase precursor is an effective strategy to obtain submicron or nano-sized MXene. Naguib et al. [133] have demonstrated that the required etching duration can be effectively reduced, when V 2 AlC powders were subjected to attrition milling prior to chemical exfoliation. However, the obtained flake sizes and shapes are easy to destroy by that method. Thus, a bunch of efforts have been devoted to explore a molten salt method to synthesize nanosized and fine-grained Ti 3 AlC 2 powders. Nanosized Ti 3 C 2 T x MXene is also easier to be prepared by etching the fine-grained Ti 3 AlC 2 precursor in concentrated HF solution [140]. The SEM image of as-prepared nanosized Ti 3 C 2 T x MXene is shown in Fig. 15(d) [137].
Drying methods and the post-treatment of MXene are another valuable strategy to tailor the morphologies of MXene sheets. Drying methods, including oven drying, freeze drying, and supercritical drying, are one important factor to determine the morphologies of MXene. Many nanoparticles often arise on the surface of MXene via oven-drying. This is probably because residual Al 3+ ions migrate with water to the surface of MXene and form Al 2 O 3 nanocrystals due to oxidation. Freeze drying is a powerful technique that is utilized in preparing fine morphologies of MXene. During the rapid cooling process, liquid water will transform into solid ice and separate from the nanoarchitectures. The porous structures can be maintained without shrinking by the following sublimation of ice. Therefore, the MXene of freeze-drying always has a clean surface without oxide particles on the surface, as shown in Fig. 15(e) [138]. Similarly, supercritical drying has been demonstrated to be an efficient technique to obtain aerogels that maintain their native morphology at micro-and nano-scale [141]. The MXene obtained by supercritical drying displays clearly visible layered structure, as shown in Fig. 15(f).
Additionally, MXene heat treatment in various atmosphere or other solvents can greatly affect the morphologies of MXene. It was noted that annealing Ti 2 CT x in various controlled atmospheres, such as Ar, N 2 , or N 2 /H 2 , retained its initial chemical structure but caused slight thinning of the layers, thus resulting in a favorable mesoporous structure to allow more access to the aqueous electrolyte [142]. A similar phenomenon for Ti 3 C 2 T x was also reported by Refs. [136,143], laying the groundwork for modification and/or elimination of surface groups by using heat treatment in various gas environments to structurally alter the morphology of Ti 3 C 2 T x , as shown in Figs. 15(g)-15(i). Zhao et al. [144] demonstrated a scalable, cost efficient, and rapid strategy to construct 3D porous Ti 3 C 2 T x network by simply adding alkali solutions to colloidal MXene suspensions to induce crumpling of Ti 3 C 2 T x nanosheets.

2 Identification of MXene's microstructure under TEM
TEM, one of the most widely used techniques, is a powerful technique to characterize the size, crystallinity, phase, and exposed facets of 2D materials [145]. Unlike SEM, electrons in TEM are generated from a higher voltage and thus have a shorter wavelength, giving rise to a much higher resolution than that of SEM. Thus, using TEM is highly desirable to fully understand various microstructural features of as-synthesized MXenes. Generally, the size, crystallinity, and exposed facets of MXenes prepared by different strategies would be different. For example, the size of Ti 3 C 2 T x for HCl/LiF etching is more intact than that of preparing by HF. Additionally, the MXene preparing by HCl/LiF always has few defects, and keeps the grain sizes and shapes of the precursor MAX phase powder as shown in Figs. 16(a) and 16(b). If precursor MAX phase was synthesized by molten salt, the obtained MXene often possesses sub-micron or nano-sized shape in the 3D scale, with multilayer laminate structures ( Fig. 16(c)). More recently, Urbankowski et al. [125] managed to delaminate multilayer Ti 4 N 3 T x into a mixture of single-and few-layered sheets using TBAOH as the intercalant, with the aid of sonication. The obtained MXene flakes possess rough edge with many small pieces on surface of MXene flake, as shown in Fig. 16(d).
Due to unsaturated M enabling spontaneous donation of electrons, precursor of metals, such as HAuCl 4 , H 2 PdCl 4 , AgNO 3 , and H 2 PtCl 6 , tailoring the morphologies of MXene sheets would be a fashionable strategy to control their functionality and properties. For instance, Cheng et al. [147] have demonstrated that MXenes with appropriate terminations are capable of reducing noble metal ions into zero-valent metals, and the MXenes still remain their native morphologies at microsacle. There is no obvious change in the cross-section structure of MXene anchored by precious metal (Fig. 16(e)). Similarly, Pt nanoparticles/Ti 3 C 2 T x composites were obtained by the direct self-reduction of H 2 PtCl 6 with Ti 3 C 2 T x due to the strong reductive activity on the unsaturated Ti [148]. As shown in Fig. 16(f), there is a negligible change in the Ti 3 C 2 T x flakes after Pt single atoms anchoring on surface of MXene.

3 Atomic structures of MXene under STEM
The STEM technology provides a very powerful technique to image separate atoms in ultrathin 2D materials, especially after the introduction of aberration-corrected optics in this technique [149]. More importantly, atomic resolution images can be obtained in high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images where the contrast of the atoms is directly proportional to the atomic number of the atoms, thus making it easy to identify the defects, vacancies, and doping atoms in a nanosheet [150]. HF-etched MXene shows a layered structure with a nonuniform interlayer spacing, as verified by HAADF-STEM ( Fig. 17(a)). The possible reason for inhomogeneous interlayer spacing is the different interactions between different terminations, which also imply the coexisting feature of functional groups. Annealing in ammonia or intercalation of cations, for example, NH 4 + can swell the structure of the MXene and increase the homogeneity of the layered domains, which is confirmed by HAADF-STEM image, as shown in Fig. 17(b). In HAADF-STEM images of the annealed Ti 3 C 2 T x MXene, the smooth and flat Ti 3 C 2 T x units are arranged periodically. Similar phenomenon is also observed in the HCl/LiF etched MXene before and after annealing in ammonia [143].
It is experimentally observed that MXenes are not structurally perfect. As a typical example, Sang et al. [154] have demonstrated that terminal Ti atoms are somewhat lost, leaving Ti vacancies, as shown in Fig. 17(c). The HF concentration used in the etching process can tune the concentration of vacancies in single-layer Ti 3 C 2 T x MXene. High concentration structural M vacancies could be generated by selectively etching off M1 (or M2) and A from the MAX phases with double M elements, (M1,M2) m AX m-1 . For instance, most Al and Sc atoms are etched off by immersing (Nb 2/3 Sc 1/3 ) 2 AlC powders in concentrated HF solutions, leading to a great deal of Nb vacancies distributed in Nb 4/3 CT x [151], as shown in Fig. 17(d). The M vacancies are not necessarily disordered, whose distribution highly depends on the distribution status of solution atoms in the MAX phases. In solution-atom ordered (Mo 2/3 Y 1/3 ) 2 AlC [155], (W 2/3 Y 1/3 ) 2 AlC [156], and (Mo 2/3 Sc 1/3 ) 2 AlC [29], Y/Sc  [157] demonstrated the mechanism of ambient degradation of Ti 3 C 2 T x MXene where titanium oxide starts to grow in the vicinity of basal and edge atomic defects. Positively charged titanium vacancies induce the formation of amorphous carbon, while oxidation of Ti cations and the following formation of TiO 2 occur at nearby positions. Meanwhile, Nb adatoms at the Nb 2 CT x MXene surface contact with ambient O, forming oxide nanoclusters that start to destroy local structural order with following degradation of the entire MXene sheet [158]. Zhang et al. [152] identified that electron probes can controllably impart metal atoms in the carbide slabs to the functional group slabs in MXene multilayers (Fig. 17(e)). Defects in MXenes aforementioned usually irreversibly react with active cations and lower the electrochemical performance when they are employed in batteries [159]. Therefore, Cheng et al. [160] have developed an effective strategy to achieve vanadium patching up Ti vacancies, giving rise to an enhancement of lithium ion storage capability. Additionally, Mo 2 TiC 2 T x MXene with Mo vacancies immobilizing single Pt atoms is an efficient catalyst for the hydrogen evolution reaction (Fig. 17(f)) [153]. The above results imply that defects give us a chance to control the surface properties of MXenes for energy storage.

Preparation and characterization of
MXene composites

1 Type of MXene composites
MXene composites can be divided into two types: (1) MXene nanosheets are dispersed in other matrix materials as reinforcement; (2) MXene nanosheets are mixed with other materials to form composites without continuous matrix. The first type of composites are generally in dense bulk shape, including MXene/polymer composites [161,162], MXene/metal composites [113], and MXene/ceramic composites [163,164]. Considering the 2D nanoflake-like microstructure, MXenes are the ideal materials for the structural and functional reinforcements.
The second type of composites can be in the shape of dense/porous bulks, films, or powders. They are made by mixing MXene nanosheets with the second phase.
Because there are many reviews on the second type of MXene composites [173][174][175], and the function applications of MXene are reviewed in Sections 7 and 8, this section mainly reviews the preparation and performance of the MXene composites reinforced by MXene nanosheets (the first type), with a short review on the preparation of MXene composites without matrix (the second type).

2 Preparation and performance of bulk composites reinforced by MXenes
MXenes have high chemical activity and abundant surface oxygen-containing functional groups, so they are easily oxidized at high temperatures. Since high temperatures are often involved in the preparation of metal-matrix or ceramic-matrix composites, MXenes have been most widely reported as reinforcements in the field of polymer-matrix composites.

Polymer-matrix composites
So far, a large number of hydrophilic or hydrolyzable polar polymers (such as poly(vinyl alcohol) (PVA) [176][177][178], polyurethane (PU) [179], polyethyleneimine (PEI) [180], and polyacrylamide (PAM) [181]) are mixed with the aqueous dispersion [176,182] or polar dispersion (e.g., N,N-dimethylformamide (DMF) [183] and dimethylsulfoxide (DMSO) [184]) of MXene by solution blending to prepare composite materials. The solution blending method can be that the polymer is directly added to the dispersion of MXene, or the polymer and MXene are dispersed in the same or different miscible dispersions and finally mixed. For example, Ling et al. [176] prepared the MXene/PVA composite by mixing the water dispersion of MXene and PVA. Naguib et al. [181] dispersed MXene with DMSO and dissolved PAM with water, and then mixed the two mixtures or solutions to obtain the corresponding MXene/PAM composite, as shown in Fig. 18. When the polymer is weakly polar or non-polar, it is necessary to modify the surface of MXene to increase the dispersibility between the materials. For instance, Si et al. [185] modified the surface of MXene with long-chain cationic bromide to make it better dispersed in DMF, For hydrophobic polymers (such as epoxy resin), MXene is often prepared into powder (multilayer MXene powders obtained by vacuum drying or delaminated MXene powders obtained by freeze drying [186,187]), and then mixed with polymer materials. In order to improve the mixing efficiency, sometimes a volatile solvent such as acetone [188] is used to disperse the MXene powder and then mixed with the polymer materials. Wang et al. [189] prepared MXene powder by freeze-drying and then mixed the acetone suspension of MXene powder with epoxy resin to successfully prepare MXene/epoxy composite material. Carey et al. [190] prepared MXene/epoxy composite by adding multilayer MXene directly into an epoxy resin. However, the dried MXene was agglomerated and difficult to be dispersed. It is still a challenge to disperse a few-layered MXene in the obtained composite material. In order to find a better decentralization strategy, Liu et al. [161] put multilayer MXene in an epoxy curing agent for ultrasonic peeling to obtain a few-layered MXene uniformly dispersed curing agent solution, and then mixed with the epoxy resin liquid to prepare MXene/epoxy composite, as shown in Fig. 19. Similar to epoxy resin, the small molecule monomer, initiator, or curing agent is fully mixed with MXene nanosheets, and then the small molecules are polymerized into large molecules in-situ, which is also called in-situ polymerization blending. By this method, MXenes were mixed with many polymers to make composites, such as polyaniline (PANI) [191], polypyrrole (PPy) [192], and polydopamine (PDA) [193]. In-situ polymerization blending could improve the dispersion of MXene in Fig. 19 Preparation process of MXene/epoxy composites. Reproduced with permission from Ref. [161], © The Author(s) 2020.
polymers, strengthen the interfacial adhesion between MXene and the polymer matrix, and improve the electrical, thermal, and mechanical properties of polymers [194,195].
Melt blending technology can also be used to prepare MXene/polymer composites, as shown in Fig. 20. Sheng et al. [196] pretreated the MXene nanosheets with polyethylene glycol (PEG) by using water as the Fig. 20 Schematic of fabrication of MXene/TPU with melt blending. Reproduced with permission from Ref. [196], © Elsevier Ltd. 2019. processing solvent, and prepared a series of MXene/ polyurethane (TPU) nanocomposites showing uniform dispersion of MXene by melt blending TPU and PEG-pretreated MXene. High-speed mechanical agitation has also been used in the preparation of MXene/polymer composites, Zhang et al. [162] mixed ultrahigh molecular weight polyethylene (UHMWPE) powders and Ti 3 C 2 powders with isopropyl(dioctylphosphate) titanate (DN-101) surface-modified homogeneously through a high-speed mixer, and then used a press vulcanizer to prepare the composites to study their thermal and mechanical properties.

Metal-matrix composites
Although MXenes have mostly been employed to reinforce polymer-matrix materials, agglomeration and poor wettability have been a big problematic challenge in metal matrix composites. Ball milling is a simple and efficient powder dispersion process. Hu et al. [113] prepared Ti 3 C 2 T x /Al composites by pressureless sintering followed by hot extrusion technique. Firstly, the mixed powders were cold-pressed into green compacts. Then the green compacts were pressureless sintered at 650 ℃ for 1 h in Ar, and finally subjected to hot extrusion at 460 . The rich surface functional groups of Ti ℃ 3 C 2 T x enhance the wettability between Ti 3 C 2 T x and Al, and improve the mechanical properties and friction resistance of the composite. However, the ball milling process has problems such as deformation of metal particles and damage to the morphology of the enhanced phase. Mai et al. [197] used a facile electrodeposition technique at room temperature to prepare MXene/Cu composite coating materials, which improved the tribological properties of Cu. For metals such as Li that cannot be used with ball milling or electrodeposition techniques, Kamysbayev et al. [198] heated lightweight metals Mg and Li to liquid metals gelled state, and then processed Ti 3 C 2 T x MXene powders in Mg-Li alloy at 700 to ℃ prepare MXene reinforced metal matrix composites. Xue et al. [199] used Sn-Ag-Cu-Ti 3 C 2 composites as textured surfaces of CSS-42L bearing steel to improve the tribological properties at high temperatures. The composites were made from the mixtures of Ti 3 C 2 and Sn-Ag-Cu by vacuum infiltration technology at 500 . ℃

Ceramic-matrix composites
Ceramic preparation processes are generally pressureless sintering, hot pressing sintering, and spark plasma sintering, which often involve high-temperature engineering. MXenes are prone to oxidation at high temperatures, resulting in less reported MXene reinforced ceramic composites. Fei et al. [163] used a ball mill to mix MXene and Al 2 O 3 powders, and prepared MXene-reinforced ceramic materials after sintering; however, MXene was severely oxidized. Wozniak et al. [164] fabricated a MXene-reinforced silicon carbide (SiC) matrix nanocomposite by a powder metallurgy route. Guo et al. [200] successfully prepared MXene-reinforced ceramic materials using cold-press sintering technology, as shown in Fig. 21.
The Ti 3 C 2 T x MXene and ZnO mixed powders were put into the steel die and heated at 300 under the press ℃ of 250 MPa for 1 h, and the densification of ZnO-Ti 3 C 2 T x was obtained. The hardness and elastic modulus of the composite with only 0.5 wt% Ti 3 C 2 T x MXene increased by 40%-50%.
Besides structural ceramics, MXene 2D sheets also can be used to made functional ceramic-matrix composites. Lu et al. [201] used a self-assembly protocol to deliver a homogeneous dispersion of Ti 3 C 2 MXene in (Bi,Sb) 2 Te 3 (BST) matrix. By controlling the pH value to 3.0, the zeta potential for BST nanoplates and MXene nanosheets are +34 and -39, respectively. Therefore, the two different nanoparticles assembled spontaneously. The MXene/BST composite powders were densified by SPS at 693 K. After sintering, the relative density values for all specimens are higher than 94%. The incorporation of MXene into BST can simultaneously achieve the improved power factor and greatly reduced thermal conductivity. A peak thermoelectric figure of merit (ZT) of up to 1.3 and an average ZT value of 1.23 from 300 to 475 K are realized for the 1 vol% MXene/BST composite. Combined with the high-performance composite and rational device design, a record-high thermoelectric conversion efficiency of up to 7.8% is obtained under a temperature gradient of 237 K.

3 Preparation of MXene composites without continuous matrix
MXenes have a good dispersion uniformity in water due to their hydrophilic functional groups such as -OH. Moreover, water is usually used to obtain MXene colloids in the process of preparing few-layer or single-layer MXene. Therefore, MXene is often doped with other materials in aqueous solution or polar solvent. Thereafter, the mixed solutions were processed by vacuum filtration, spraying coating, spin coating, layer-by-layer (LbL) self-assembly, wet-spinning, and printing (inkjet printing and 3D printing) to obtain green bodies with desired shape. Then the green bodies were dried by vacuum drying, freeze-drying, or inert gas blowing to obtain the composites. Stirring or ultrasonic dispersion is the simplest method. Xu et al. [202] used simple ultrasound to mix the GO water dispersion and MXene dispersion for 1 h. GO/Ti 3 C 2 T x dispersion was filtered by vacuum-assisted method, and then reduced to porous rGO/Ti 3 C 2 T x films. Yue et al. [203] mixed GO solution and MXene solution and then further sonicated to further obtain a homogeneous solution. Finally, the solution was freeze-dried to obtain 3D composite aerogel. Hydrothermal method is also a simple method. Xu et al. [204] synthesized the petal-like MoS 2 /MXene composite by one-step hydrothermal method, and the electrode capacity of the composite was significantly improved. In addition, Cao et al. [205] manufactured an ultrathin and flexible carbon nanotubes/ MXene/cellulose nanofibers composite paper through alternate vacuum-assisted filtration process for electro magnetic interference (EMI) shielding research. Layered MXene composites can also be realized by spraying coating, spin coating, and LbL self-assembly, and the thickness is controllable. Shen et al. [206] spin-coated the PEBA solution on the surface of the MXene nanofilm for five cycles to produce gas separation films. MXene solution is negatively charged, so it can self-organize with positively charged materials through electrostatic LbL. Tian et al. [207] used alternately deposition of positively charged material PEI and negatively charged material MXene to manufacture supercapacitors and found that the performance was greatly improved compared with non-LbL structures. Layer thickness can be easily controlled by controlling the number of spin coating or alternate deposition. Yang et al. [208] got flexible and highly conductive MXene/rGO hybrid fibers through wet-spinning method. Printing can realize the establishment of specific pattern and structure. Rastin et al. [209] formulated a new conductive cell-laden bioink composed of Ti 3 C 2 MXene nanosheets/hyaluronic acid/alginate, and demonstrated its performance of extrusion-based 3D bioprinting.

4 Characterization of MXene composites
There are many methods for composite characterization, and herein we only introduce the special methods for the characterization due to the introduction of MXene in the composites. First, the c-lattice parameters (c-LP) for the MXene in the composites are usually considerably in the preparation processing in which the X-ray diffraction (XRD) always can find small changes. Second, the functional groups on the MXene surface are usually changed due to the combination of other phases. The X-ray photoelectron spectroscopy (XPS) is commonly used to track changes in MXene surface functional groups during composite preparation. During composite preparation, phase stability is the main issue to be considered. After the preparation of the MXene composites, structural changes during service life of the composites need to be considered. The structure change affects the performance of the composites. Generally, MXenes have good thermal conductivity. MXene composites, especially MXene-reinforced polymer composites, often need to be tested for thermal properties. Thermal analyzers are commonly used to test thermogravimetry, and electricity heat drum wind-drying oven is used for thermal oxidative aging test. Dynamic thermomechanical analyzer (DMA) was used to detect the creep properties and glass transition temperature of composite materials [210]. The friction coefficient can be tested by universal friction and wear testing machine [162].
MXenes with 2D sheet structures have good rigidity and strength to improve the mechanical strength of the materials. For example, the addition of a small amount of MXene can improve the tensile properties, bending properties, and fracture toughness of epoxy [176], and increase the hardness of ceramic materials [189], etc. Typically, the universal testing machine is used to test the tensile, compression, bending, fracture toughness, and other mechanical properties of composite materials. The hardness tester is used to characterize the hardness of composite materials. DMA also can be considered for the characterization of the storage modulus and other dynamic mechanical properties of MXene composite materials.

Application of MXenes in energy storage
MXenes can be applied in many areas as functional materials with high performance. In this review, we focused in two most important areas: energy storage (Section 7) and electromagnetic interference shielding (Section 8).
For the purpose to apply MAX phases as anode of LIBs, A element was selectively etched off from MAX phase to generate the room for lithium storage. Thus MXenes were created as anode of LIBs [6,7]. Therefore, energy storage (LIBs, sodium-ion batteries, super-capacitors, etc.) is the first and most extensively researched area for MXenes' application.

LIB was first introduced by Sony Corporation in 1991.
LIB is the secondary battery (rechargeable battery) that works by moving lithium ions between the positive 1220 J Adv Ceram 2021, 10(6): 1194-1242 www.springer.com/journal/40145 and negative electrodes. LIBs have been widely used in portable electronics, electric vehicles, and many other fields due to their high energy density and high reliability [211,212]. According to the results of theoretical calculation, the LIB can reach a high theoretical specific energy density of 421 Wh·kg −1 [213]. However, the energy density of commercial LIBs is far from theoretical value. At the same time, the cycling stability of LIBs has been waiting for further improvement. In order to solve these problems, in recent years, researchers devote to exploring and developing new anode materials with good cycling stabilities and higher specific capacities rather than current graphite or lithium titanate anodes, which showed high theoretic specific capacities of 372 and 175 mAh·g -1 , respectively [214,215].
MXenes are considered to be a promising electrode material for LIBs due to their high specific surface area, good conductivity, and fast diffusion rate of lithium ions [8,122]. In 2012, Naguib et al. [7] firstly reported the Ti 2 C MXene as anode materials for LIB with a stable capacity of 225 mAh·g -1 at 0.04 C. Cyclic voltammetry (CV) curves demonstrated lithium intercalation and extraction peaks at 1.6 and 2 V, respectively.
Based on the theoretical calculations, Tang et al. [216] concluded that the theoretical capacity of Ti 3 C 2 MXene (without any functional group) was 320 mAh·g −1 . However, once the surface of MXene was covered with functional groups (F or OH), the theoretical capacity was significantly reduced to 130 or 67 mAh·g −1 . It was suggested that MXene with pure F groups may be a way to further enhance its specific capacities. In the preparation process, MXene inevitably contains O, OH, F, and other surface functional groups. In 2014, Sun et al. [111] synthesized an exfoliated Ti 3 C 2 MXene with F termination as anode achieved 123.6 mAh·g −1 at 1 C, which is close to the theoretical capacity of Ti 3 C 2 F 2 Li (130 mAh·g −1 ). MXenes were prepared by hydrofluoric acid or fluorine salt solution etching, resulting in abundant functional groups (OH, F, or O) on the surface of MXenes. These functional groups have a great influence on the physical and chemical properties of MXenes. Therefore, surface design and structure adjustment are effective strategies to obtain electrode materials with excellent electrochemical performance for LIBs.
These results are encouraging and suggest that the other MXenes such as Ta 4 C 3 , (Ti 0.5 ,V 0.5 ) 2 C, (Ti 0.5 ,V 0.5 ) 3 C 2 , (Ti 0.5 ,Nb 0.5 ) 2 C, (V 0.5 ,Cr 0.5 ) 3 C 2 , (Nb 0.5 ,V 0.5 ) 4 C 3 , Mo 2 TiC 2 T x , Mo 2 Ti 2 C 3 T x , Cr 2 TiC 2 T x , Ti 3 CN, Nb 2 C, V 2 C, Nb 4 C 3 , Mo 2 C, Zr 3 C 2 , Hf 3 C 2 , and Ti 4 N 3 , could be used as Li + intercalation electrode in LIBs. Sun et al. [217] studied the lithium storage performance of eight MXenes (V 2 C, Cr 2 C, Ta 2 C, Sr 2 C, Ti 2 C, Zr 2 C, Nb 2 C, and Hf 2 C) with O group through theoretical calculation, and predicted that V 2 C MXene with −O termination had the best lithium storage performance (735 mAh·g −1 ) in these eight MXenes. Hence, Naguib et al. [133] synthesized V 2 C MXene by selective etching of V 2 AlC with HF etching agent. The obtained V 2 C MXene exhibited great potential as electrode materials for LIBs with capacity of 260 mAh·g −1 at 1 C rate. Subsequently, Liu et al. [118] prepared V 2 C MXene with high purity via etching V 2 AlC with sodium fluoride and hydrochloric acid at 90 ℃ for 72 h. Compared with the HF etching route, the NaF-HCl etching approach is a milder etching method and produces a purer V 2 C MXene. The obtained highly pure V 2 C MXene exhibited superior electrochemical performance as anode of LIBs with specific capacity of 260 mAh·g −1 at the current density of 370 mA·g −1 , as shown in Fig. 22. The purity of V 2 C is also an important parameter affecting the electrochemical performance of the V 2 C MXene-based electrode material. It is an effective strategy to explore a new method to prepare highly pure MXene as anode material for LIBs.
In 2021, Zhang et al. [218] synthesized N/S co-doped V 2 C MXene by an annealing method. As anode for LIB, the obtained N/S co-doped V 2 C MXene demonstrated outstanding electrochemical performance with a highly reversible capacity of 590 mAh·g −1 after 100 cycles at the current density of 0.1 A·g −1 . As increased the current density to 2 A·g −1 , the specific capacity still reaches 298 mAh·g −1 after 300 cycles, which is superior to the previously reported V 2 C MXene-based anode. Not just N and S, but other heteroatoms (e.g., O, B, or P) doping could also decorate the surface chemical functionalities, which is beneficial to reducing the band gap of MXene, thereby reducing its resistance and accelerating the diffusion of electrolyte ions on the surface of MXenes.
Naguib et al. [133] reported a 2D Nb 2 C material, which was synthesized by treating Nb 2 AlC in aqueous HF. The obtained Nb 2 C MXene exhibited excellent reversible specific capacity (170 mAh·g −1 at 1 C rate) at low lithiation potentials. The Nb 2 C MXene could maintain high specific capacity at a high cycling rate In 2015, Jin et al. [219] introduced a new ternary nanolaminated carbide MAX phase, i.e., Mo 2 Ga 2 C, which shows a hexagonal structure with two Ga layers, and exhibits similar chemical and physical properties to other MAX phases. Hence, Mo 2 Ga 2 C MAX phases can be etched to obtain Mo 2 C MXene. Herein, Mei et al. [220] reported a UV-induced selective etching approach to prepare Mo 2 C MXene from Mo 2 Ga 2 C MAX phase. The obtained fluorine-free Mo 2 C MXene as anode materials for LIBs exhibited high specific capacity (~150 mAh·g −1 ), good rate performance, and long cycle life. In addition, the fabricated flexible Mo 2 C MXene-based battery devices demonstrated good specific capacity retention in the state of muti-angle bending. This work introduces a new method for the preparation of fluorine-free Mo 2 C MXene materials by UV-induced selective etching method, which can also be extended to the preparation of other new MXenes.

2 Supercapacitors
Supercapacitors (SCs) have attracted attention due to its high-power density, fast charge-discharge rate, easy maintenance, wide operating temperature range, excellent cycle life, and high security [221,222]. Therefore, SCs have been widely used in electric vehicles, portable electronics, military products, intelligent instruments, and storage power stations. Generally, based on the energy storage mechanism, SCs can be divided into two types: electrical double layer capacitors (EDLC, by electrostatic adsorption) and faradic pseudocapacitors (by rapid surface redox reactions). Besides the above two types of SCs, hybrid supercapacitors are called the third type of SCs, which is a combination of EDLCs and LIBs. This hybrid SC not only has the characteristics of EDLC (fast charging and long cycle life), but also has the advantages of high energy density of LIB.
2D MXenes combine the metallic conductivity of transition metal carbides with the hydrophilic nature of their hydroxyl or oxygen terminated surfaces. These properties make MXene as a promising electrode for SC applications. First-principles calculations based on DFT theory revealed that MXene sheets had high mechanical strength when they were drawn along the basal plane, which prevented structural collapse during charge and discharge processes, resulting in good cyclic stability of MXenes. In 2013, Lukatskaya et al. [223] first explored the application of MXene as an electrode material for SCs. In this work, titanium carbide MXene was investigated in various electrolytes, and delivered high specific volumetric capacitance of 300 F·cm −3 , which is superior than that of carbon materials. This work provides a basis for exploring a big family of 2D carbides and carbonitrides MXenes in electrochemical energy storage applications using single and multivalent ions.
Hu et al. [224] prepared a binder-free Ti 3 C 2 MXene film and evaluated it in 1 M H 2 SO 4 electrolyte, which showed a volumetric capacitance of 226 F·cm −3 at a scan rate of 2 mV·s −1 . The lower resistance of Ti 2 CT x MXene, due to its thinner lamellae than Ti 3 C 2 T x MXene, means that Ti 2 CT x MXene has a higher specific capacity than Ti 3 C 2 T x MXene. Rakhi et al. [142] synthesized Ti 2 CT x MXene by an HF-etching route from Ti 2 AlC max phase and annealed the Ti 2 CT x MXene in air, Ar, N 2 , and N 2 /H 2 atmosphere for 2 h to remove the functional groups on the surface of MXenes. The as-prepared Ti 2 CT x MXene delivered specific capacitance of 51 F·g −1 at 1 A·g −1 in 30 wt% KOH electrolyte.
Generally, most of the electrolyte solutions used in previous studies consist of dangerous, toxic, and expensive organic and inorganic chemicals. However, if these SC devices are leaked during use, they will cause harm to human body and pollute the environment. Moreover, most of the electrode materials in the acid or alkaline electrolytes are not stable for a long time. However, most electrode materials can maintain good chemical stability in neutral solution. Generally, lithium, potassium, and sodium salts provide low ionic resistivity, fast ion diffusion kinetics, and are friendly to human body and environment. Therefore, as a rich reserve of seawater, it has great potential as an electrolyte for SCs.
The electrochemical properties of Ti 3 C 2 MXene were evaluated using seawater as an electrolyte by a three-electrode system, as shown in Fig. 23 [225]. Figure 23(a) shows the CV curves of Ti 3 C 2 MXene at various scan rates in seawater electrolyte. All the CV curves retain nearly rectangular loops, which are caused by the electro-sorption/desorption of Na + cations from seawater electrolyte. The specific capacitances of Ti 3 C 2 MXene were calculated from the discharge curves ( Fig. 23(b)). Specific capacitance reached 67.7 F·g −1 at 1 A·g −1 , corresponding to the volumetric specific capacitance of 121.8 F·cm −3 (Fig. 23(c)). Ti 3 C 2 electrodes were tested at a high current density of 10 A·g −1 for 5000 cycles. The Ti 3 C 2 electrode still exhibits nearly 96.6% retention (Fig. 23(d)), demonstrating excellent cycle stability in seawater electrolyte, which is comparable to that reported MXene-based SCs. This www.springer.com/journal/40145 excellent configuration can still be applied to other MXenes. He et al. [226] developed a NaF+HCl etching to obtain highly pure V 2 C MXene. The V 2 C MXene also delivered a high capacitance of 317.8 F·cm −3 in seawater electrolyte.
There have been many reports of nitrogen-doped graphene used as electrode materials for SCs. Based on the same principle, nitrogen doped MXene has been reported to be used as electrode materials for SCs and obtains enhanced electrochemical performance. Wen et al. [227] demonstrated a facile method to synthesize N-doped Ti 3 C 2 MXene by post-etch annealing Ti 3 C 2 in ammonia. The doping of N to replace C atoms in the MXene properly increased the interlayer spacing between Ti 3 C 2 nanosheets and narrowed the band gap of MXene, and further improved its electrical conductivity. The doping of elemental nitrogen also provided more active sites. These benefits resulted in a 4.6-fold increase in the capacitance of N-doped MXene compared to pristine MXene.
With the development of science and technology, the demand for flexible wearable electronic products is more and more vigorous. Flexible and wearable SC devices have recently attracted more and more attention. Yu et al. [228] demonstrated a route to using 2D Ti 3 C 2 MXene as a flexible, conductive, and electrochemically active binder for one-step assembly of MXene-bonded activated carbon as a flexible electrode for SCs in an organic electrolyte, as shown as Fig. 24. As a result, the flexible MXene-based electrode delivered a high specific capacitance of 126 F·g −1 at 0.1 A·g −1 and with a retention of 57.9% at a high current density of 100 A·g −1 in an organic electrolyte.
The development of high capacitance and low-cost MXene-based electrodes will promote the wide application of such SCs. To explore a method for preparing a low-cost, flexible, tough, and upgradeable SC electrode, the silk was directly carbonized to make a conductive independent textile substrate [229]. By gently baking the surfactant-free Ti 3 C 2 nanosheet suspension loaded on the carbonized silk, a uniform, nanometer thick bonding coating of Ti 3 C 2 flakes was formed on the conductive fabric support, forming a MXene-coated flexible textile electrode. The assembled electrode exhibited a high areal capacitance of 362 mF·cm -2 with excellent cycling stability and outstanding foldability.

3 New secondary battery
As a rechargeable secondary battery, LIBs have achieved great commercial success and are widely used in portable electronic products, electric vehicles, and the defense industry, etc., for their advantages of high operating voltage, high energy density, long cycle life, and environmental friendliness [175,230]. However, shortage of lithium resources can be a future problem. It is urgent to develop a new generation of rechargeable secondary batteries with low cost and high performance [231]. Na and K not only have similar chemical and physical properties to Li, but also have the economic advantages such as abundant reserves and low cost. Therefore, sodium ion batteries (SIBs) and potassium ion batteries (PIBs) are considered as primary alternatives to LIBs. However, the radii of K + (1.38 Å) and Na + (1.02 Å) are much larger than that of Li + (0.76 Å), which leads to slow reaction kinetics and huge volume changes when potassium and sodium ions are repeatedly inserted/extracted from electrodes, causing irreparable damage to their structure. Hence, design and development of suitable electrode materials for SIB and PIB is the key to solve this problem. 2D MXene materials provide many merits as electrode materials for alkali metal-ion batteries due to their large number of accessible active sites, tunable structure, and chemical compositional diversity [232]. MXenes have rapid electrolyte ion diffusion features, whose diffusion kinetics is quite high for many electrolyte ions such as Li + , Na + , K + , and Ca 2+ [233]. Based on DFT calculations, Ti 3 C 2 MXene exhibits great potentials for SIBs and PIBs with a theoretical capacity of 351.8 and 191.8 mAh·g −1 , respectively [234]. Naguib et al. [232] experimentally prepared Ti 3 CNT z MXene as an anode material for PIBs. Figure   25(a) demonstrates a first cycle potassiation and de-potassiation capacity of 710 and 202 mA·h·g −1 , respectively. Figure 25(b) shows three peaks at 1.25, 0.81, and ~0 V during the 1st potassiation cycle, which are corresponding to the de-potassiation peaks of 1.50, 0.80, 0.50 V, respectively. As shown as Fig. 25(d), the MXene material exhibits excellent rate performance with minimal drops at a high current density (500 mA·g −1 ), which should be attributed to the intercalation mechanism into the layered structure. From Fig. 25(e), at the 1st discharge cycle (potassiation), the K ions intercalated between the MXene layers cause an increase in the c-lattice direction parameter. The trapped K ions act as pillars between the MXene layers that decrease large expansion/contraction during charge/discharge process. Lithium sulfur batteries (LSBs) are holding promise for the next-generation energy storage devices that can provide ultrahigh theoretical energy density of 2600 Wh·kg −1 [235]. Dong et al. [236] introduced an MXene-based flexible and integrated sulfur cathode, and Ti 3 C 2 MXene as sandwich layer on a separator (polypropylene, PP). This configuration makes for an LSB with high energy density and good cycling stability. Especially, the Ti 3 C 2 framework with open interconnected microporous structrue and the fully exposed surface ensures a large load of sulfur and rapid ion transportation for improving lithiation/delithiation kinetics. In this novel design, no metal is needed as a current collector, and the integrated Ti 3 C 2 −S/d-Ti 3 C 2 /PP electrode is directly used for LSBs, which is delivered a high capacity of 1062 mAh·g −1 at 0.2 C rate. Furthermore, the all MXene-based integrated cathodes displayed an outstanding rate capacity of 288 mAh·g −1 at 10 C rate and excellent cycle stability.
These novel and interesting 2D MXene materials have been proven to be an excellent electrode material for a variety of electrochemical energy storage devices, including LIBs, SCs, KIBs, NIBs, and LSBs, encouraging further exploration of the applications of MXene in electrochemical energy storage. Therefore, the production of new MAX phases and its further exfoliation into 2D MXenes has become an important strategic direction for the intensive study of MXene. We expect that there are still many MXenes that are more suitable as electrodes for electrochemical energy storage devices, and worth investigating.

Applications of MXenes as electromagnetic interference shielding materials
At the very beginning of the advent of MXenes [6], the interest for them was primarily in electrochemical applications [234,[237][238][239]. With the further exploration on their properties, MXenes were found to be promising in other technologically important domains. In 2016, Shahzad et al. [240] demonstrated that flexible Ti 3 C 2 T x films (1-45 µm) exhibited excellent electrical conductivity and electromagnetic interference (EMI) shielding capabilities (as contained in Figs. 26(a) and 26(b), better than any known synthetic materials with similar thicknesses), and paved the way for MXenes to be used as EMI shielding materials. EMI can disrupt electronic devices and equipment that operated in medicine, telecommunication, military, aerospace, and many other domains [241]. In addition to the impact on devices and systems, EMI might even influence human health [242,243]. An EMI shielding material can protect devices against incident EM waves, and its shielding effectiveness (EMI SE) is expressed as a ratio between the transmitted and incident powers on a logarithmic scale, as shown in Eq. (4): where P and E refer to power and electric field intensity, respectively; their subscripts I and T indicate incident and transmitted EM waves, respectively [244].
The interaction between an incident EM wave from air and shielding material is usually explained by reflection, absorption, and multiple reflection, as illustrated in Fig. 26(d) [240]. Total EMI shielding effectiveness (EMI SE T ) is the sum of the attenuation produced by reflection (SE R ), absorption (SE A ), and multiple reflection (SE M ), as shown in Eq. (5) [244]: The high conductivity of the material can significantly improve the reflection loss, while the porous structure, electron, and magnetic coupling can increase the absorption loss, and the efficient multiple reflection loss can be achieved by constructing a porous or multilayer structure. The merits of MXenes as stated in the forgoing sections just fit the requirements for the electromagnetic shielding materials. In addition to their high electrical conductivity (Fig. 26(c)) and good electronic coupling between the layers, their 2D structure has been proven to play a key role on their outstanding EMI shielding effectiveness. In particular, the fact that MXenes have the exceptionally high absorption of electromagnetic waves is attributed to their layered, metamaterial-like structure [243]. This fact deserves more explorations, and a comprehensive understanding of the fundamental interaction mechanisms between electromagnetic waves and their 2D structure will provide effective guidance for designing advance EMI shielding materials by employing these 2D materials [240,243]. To adapt to various application scenarios and also enhance their EMI shielding performance, a number of MXene-based materials/structures were designed and tested for the EMI shielding application, including solo MXene films, MXene composite films, and foams and aerogels constructed from MXenes. In general, all the formats of reported MXene-based EMI shielding materials fall into two categories: films and 3D porous structures (e.g., foams and aerogels). Herein, representative MXene-based EMI shielding materials are discussed by citing the highly cited references published in recent years.

1 MXene and its composite films
Shahzad et al. [240] synthesized Ti 3 C 2 T x ultrathin films and Ti 3 C 2 T x -SA (sodium alginate) composite films, and demonstrated that the EMI SE of the Ti 3 C 2 T x film (thickness of 45 μm) reached 92 dB (Fig. 26(a)), and the EMI SE of the Ti 3 C 2 T x -SA composite film (8 μm, 10 wt% SA) exhibited 57 dB (Fig. 26(b)). This performance originates from the excellent electrical conductivity of Ti 3 C 2 T x films (4665.1 S·cm -1 , Fig. 26(c)) and multiple internal reflections from Ti 3 C 2 T x flakes in free-standing films (Fig. 26(d)). Such a high EMI SE opened the Ti 3 C 2 T x material in EMI application research, and brought the explosion of research on this topic.
Mechanical performance is the basic requirements for the appropriate service of EMI shielding materials. To enhance its mechanical performance, Wan et al. [245] incorporated Ti 3 C 2 T x with poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate) (PEDOT/PSS) and then treated it with concentrated sulfuric acid to remove the insulating PSS. An introduction of 30 wt% PEDOT:PSS into the Ti 3 C 2 T x -based film rendered it a high tensile strength of 38.5±2.9 MPa, which was 1.55 times that of solo Ti 3 C 2 T x films. And the as-prepared mechanically strong composite film offered an excellent EMI SE up to 40.5 dB (~6.6 μm). Bacterial cellulose (BC) was employed to fabricate ultrathin, strong, and highly flexible Ti 3 C 2 T x /BC composite films by Wan et al. [246], using a scalable in-situ biosynthesis method. They uniformly dispersed Ti 3 C 2 T x nanosheets into the 3D BC network, resulting in a mechanically entangled structure that endowed the Ti 3 C 2 T x /BC composite films with excellent tensile strength (297.5 MPa at 25.7 wt% Ti 3 C 2 T x ) and flexibility. Importantly, a 4 μm-thick Ti 3 C 2 T x /BC composite film with 76.9 wt% Ti 3 C 2 T x content gave a specific EMI SE of 29,141 dB cm 2 ·g -1 , which surpasses those of most previously reported MXene-based polymer composites with similar MXene contents and also, the most reported carbon-based polymer composites.
As the portable and wearable electronics rigorously develop, lightweight, flexible, and stable films with high EMI SE are highly demanded. This is usually achieved by assembling MXenes with other functional nanomaterials by electrostatic interaction, van der Waals (vdW) interaction, and hydrogen bond. LbL assembly is an advanced fabrication method for nanocomposites, and Weng et al. [247] developed an spin spray layerby-layer (SSLbL) technique, which can rapidly assemble MXenes and carbon nanotubes (CNT) together, yielding a composite with hierarchical architectures with desirable cross-functionalities. For the MXene-CNT composite realized by SSLbL route, a high stability and tensile strength was endowed via electrostatically bonded polymer fillers, and the as-prepared semitransparent MXene-CNT composite films also showed high conductivities up to 130 S·cm -1 (the highest values achieved among LbL composite films to date). The authors showed that the MXene-CNT film has a maximum specific EMI shielding effectiveness (SSE/t) of 58,187 dB·cm 2 ·g -1 , which is among the highest reported SSE/t values. Ultrathin, lightweight, and flexible CNT and MXene (Ti 3 C 2 T x ) buckypaper with an enhanced EMI SE was obtained by Yang et al. [248] through a facile electrophoretic deposition process. The prepared Ti 3 C 2 T x @CNT hybrid bucky paper of 100 μm exhibited an outstanding EMI SE of 60.5 dB in the X-band, and the hybrid bulkypaper with a Ti 3 C 2 T x content of 49.4 wt% even exhibited an EMI SE of 50.4 dB across the X-band with a thickness of only 15 μm, which means its EMI SE is 105% higher than that of the pristine CNT buckypaper. And as the thickness of the hybrid paper was decreased to merely 5 μm, it still demonstrated an average specific EMI SE value of 5.7×10 4 dB·cm 2 ·g -1 .
Due to their high electrical conductivity, good mechanical properties, and outstanding EMI SE, a number of metal-based EMI shielding materials have been extensively applied for modern integrated electronic and telecommunication systems ranging from military domains to civilian areas. However, their relatively high density, low flexibility, poor chemical corrosion resistance, and poor processability are hindering their widespread applications, particularly in the emerging technological territories. MXenes' rich terminating groups were harnessed by Ma et al. [249] to fabricate an ultraflexible and mechanically strong aramid nanofiber Ti 3 C 2 T x MXene/silver nanowire (ANF-MXene/AgNW) nanocomposite. Hydrogen-bonding interactions are confirmed between the components of the as-prepared ANF-MXene/AgNW nanocomposite. The authors used a facile two-step vacuum-assisted filtration followed by hot-pressing approach (Fig. 27(a)). The as-fabricated double layered nanocomposite paper with a low MXene/AgNW content of 20 wt% which delivered an excellent electrical conductivity of 922.  outstanding comprehensive properties. Specifically, the nanocomposite papers showed a maximum electrical conductivity of 3725.6 S·cm -1 and EMI SE of ~80 dB ( Fig. 27(b)) at a MXene/AgNW content of 80 wt%. It is noted that this structure has an absorption-dominant shielding mechanism, which can be explained by the massive ohmic losses in the highly conductive MXene/ AgNW layer, multiple internal reflections between Ti 3 C 2 T x MXene nanosheets and polarization relaxation of localized defects, and abundant terminal groups. Liang et al. [250] designed and realized a MXene/Ni chain hybrid with excellent electromagnetic wave absorption and shielding capacity. The hybrid consisted of one-dimensional Ni nanochains and 2D Ti 3 C 2 T x nanosheets (Fig. 27(c)). Simply, by only adjusting the ratio of conductive Ti 3 C 2 T x nanosheets and magnetic Ni nanochains, the authors demonstrated the adjustable EM wave absorption and shielding properties of the MXene/Ni chain hybrid. The minimum reflection loss of −49.9 dB was recorded by a sample loaded with 10 wt% MXene and with a thickness of 1.75 mm at 11.9 GHz, increasing the MXene content to 50 wt%, and the optimal EMI SE reaches 66.4 dB with an SE A of 59.9 dB, showing an evident dependence on the sample thickness (Fig. 27(d)). The interaction mechanism between the incident electromagnetic waves and the hybrid was considered to be severalfold, as illustrated in Fig. 28. The synergistic effect of the conductivity loss, multiple scattering, eddy current loss, dipole polarization, and interface polarization is responsible for the satisfactory EMI SE of the hybrid.
Intelligent garment is a highly promising area of wearable electronics, which are developed by integrating multifunctionalities into textile. Starting from the superb intrinsic electrical conductivity, tunable surface groups, and layered structure of MXenes, via a dip-coating approach, Wang et al. [251] fabricated highly conductive, hydrophobic, and flexible MXene-decorated poly(ethylene terephthalate) (PET) textiles with exceptional EMI shielding efficiency and Joule heating performance. They obtained a multifunctional textile exhibiting high electrical conductivity of 1000 S·m -1 , and simultaneously an exceptional EMI SE of 90 dB at a thickness of 1.3 mm. Particularly, as the society is going to an age of Internet of Things (IoT), the proliferation of wireless devices may raise nonnegligible safety issues for human health. And therefore, it demands stretchable materials that are capable of shielding EMI. Li et al. [252] fabricated an ultrastretchable conductor by depositing a crumble-textured coating composed of 2D Ti 3 C 2 T x nanosheets (MXene) and single-walled carbon nanotubes (SWNTs) onto latex. This ultrastretchable conducting MXene-SWNT (S-MXene)/latex material can be fashioned into wearable antennas and EMI shields with high performance. The device made from this material survived an 800% areal strain and exhibited strain-insensitive resistance profiles during a 500-cycle fatigue test. By incorporating SWNTs into MXene films, the resultant construction gained a synergy for both mechanical and EMI shielding performance. Therefore, a single layer of stretchable S-MXene conductors demonstrated an EMI shielding performance of 30 dB up to 800% areal strain, and the shielding performance was strain-independent; and the shielding performance was further improved to 47 and 52 dB by stacking 5 and 10 layers of the S-MXene conductors, respectively.
Developing electronic ink is highly demanded for applications in printable and flexible electronics. Vural et al. [253] employed tandem repeat synthetic proteins based on squid ring teeth (SRT) as templates of molecular self-assembly to engineer MXene inks, and the binder protein molecules can form sequence controlled assemblies with hydrogen bonding 2D crystals during printing, and therefore assemble MXenes together, as illustrated in Fig. 29(a). The authors demonstrated that the resulting ink formulations could be effectively printed on a number of substrates (e.g., cellulose paper, glass, and PET), yielding electrode patterns. The MXene-electrodes printed on flexible PET substrates had the electrical conductivity of 1080±175 S·cm −1 for films with an average thickness of 2.25 μm (Fig. 29(b)), corresponding to the highest electrical conductivity reported among electrodes that are inkjet-printed using 2D materials. Thanks to the intrinsic excellent conductivity of MXenes and the flexibility of the MXene-ink, the highly conductive electrodes printed on flexible PET substrates remained integrated and maintained their electrical conductivity under bending deformation, and therefore, it provided outstanding EMI SE (50 dB, a film thickness of 1.35 μm, Fig. 29(c)).

2 MXene and its composite foams and aerogels
Foam-like structures can take advantage of impedance matching to take in more EM waves, and multiple reflection and scattering within the materials would cause enhanced EM absorption [254]. Bian et al. [255] demonstrated a facile, freeze-casting route for preparing MXene aerogels starting from MXene colloidal solution ( Fig. 30(a)). The as-prepared MXene aerogels had micro-sized pores inherited the morphology of ice crystals, and remarkably, they showed excellent EMI shielding performance (up to 75 dB), with exceptionally high absorption loss (SE R < 1 dB). In particular, thanks to the ultralow density of the aerogels, their SEE/t can be as high as 9904 dB·cm -3 ·g -1 (Fig. 30(b)).
To guarantee the stability of MXenes in a moist or wet service environment, Liu et al. [256] fabricated hydrophobic MXene foam with reasonable strength by assembling MXene sheets into films and then expanding them via a hydrazine-induced foaming process, as illustrated in Fig. 30(c). And the as-prepared MXene foams, in contrast to well-known hydrophilic MXene materials, exhibited hydrophobic surfaces and outstanding water resistance and durability. Compared with its unfoamed film counterpart (53 dB), a much higher  enhanced EMI SE of 70 dB (Fig. 30(d)) was recorded for the lightweight MXene foam, which is attributed to its highly efficient wave attenuation in the favorable porous structure. Therefore, the MXene foam with excellent EMI SE is highly promising for applications in aerospace and portable and wearable smart electronics, where the weight matters.
Lightweight, compressible, and conductive materials are demanded in EMI shielding gaskets, wearable electronics, sensors, and other specific areas. Based on the superb properties of MXenes, Wu et al. [257] fabricated the compressible and electrically conductive polydimethylsiloxane (PDMS)-coated MXene foams. The fabrication route employed by the authors is schematically shown in Fig. 31(a). 3D MXene aerogel architectures were preformed under the assistance of SA, and followed by coating a thin layer of PDMS to render the porous architecture structural stability and durability. The resultant lightweight MXene/SA hybrid aerogel had an outstanding conductivity of 2211 S·m -1 and a high average EMI SE of 70.5 dB. Furthermore, the PDMS coating endowed the hybrid aerogel with excellent compressibility and durability, and a PDMScoated MXene foam with 6.1 wt% of MXene reserved its high EMI shielding efficiency of 48.2 dB after 500 compression-release cycles. Sun et al. [258] fabricated the highly conductive MXene@PS nanocomposites via electrostatic assembly of the negative MXene nanosheets on the positive PS microspheres followed by compression molding, as illustrated in Fig. 31(b). The interfaces of large PS microspheres were covered by the highly conductive MXene nanosheets, forming a continuous and highly efficient conduction network, affording superb electrical conductivity and excellent EMI shielding performances. A low percolation threshold of 0.26 vol% was achieved and the PS nanocomposite with 1.90 vol% of Ti 3 C 2 T x exhibited a superb electrical conductivity (up to 1081 S·m -1 ), providing the nanocomposites with excellent EMI SE. An EMI SE of above 54 dB over the whole X-band with a maximum www.springer.com/journal/40145 of 62 dB was realized by a PS nanocomposite with only 1.90 vol% of Ti 3 C 2 T x , which was among the best performances yet reported for the polymer nanocomposites.
EMI shielding performance usually refers to the sum of EM wave reflection and absorption, and the conventional EMI shielding materials, e.g., metallic materials, have high EMI SE, and are predominated by reflection loss. And it is a long-sought goal to control the ratio of EM reflection and absorption for an EMI shielding material. Based on MXenes, a series of aerogels with lamellar structure and long-range ordering were constructed by Han et al. [259] via a bidirectional freeze-casting under dual temperature gradients, as illustrated in Fig. 31(c). The as-fabricated aerogels were anisotropic and compressible, with low density, resulting in outstanding EMI shielding performance and, especially, tunable ratio of reflection to absorption. EMI SE of 70.5, 69.2, and 54.1 dB were achieved for Ti 3 C 2 T x , Ti 2 CT x , and Ti 3 CNT x aerogels, respectively, with the compression thickness of 1 mm and a density of only 11.0 mg·cm -3 . An SEE/t of 8818.2 dB·cm -3 ·g -1 was recorded, which was among the best values reported for EMI shielding materials (Fig. 31(d)). More importantly, due to their 3D skeletons with well-arranged cellular structure, as illustrated in Fig. 31(e), these MXene aerogels maintained stable total shielding effectiveness with the increasing ratio of reflection to absorption under compression from 3 to 1 mm.
To take the advantages of outstanding EMI shielding performance of Ag and MXenes, Weng et al. [260] fabricated lightweight AgNW/MXene hybrid sponges via a method of dip-coating and unidirectional freeze-drying, as shown in Fig. 32(a). The light sponges (43.0 mg·cm -3 ) with a novel buckling design exhibited a conductivity of up to 15.3 S·cm -1 and superior EMI SE of above 40.0 dB in the X-band (Fig. 32(b)). The introduction of an irregular honeycomb-like MXene architecture enabled an EMI SE of 52.6 dB associated an enhanced conductivity (24.5 S·cm -1 ) at 49.5 mg·cm -3 . These impressive performances of the sponges manifested the advantages of buckled structures and the synergistic effect of AgNW and MXene (Fig. 32(c)).
To reconcile the lightness and the mechanical robustness, Wang et al. [261] fabricated a waterborne polyurethane/Ti 3 C 2 T x MXene/nickel ferrite (WPU/ MXene/NiFe 2 O 4 ) hybrid aerogel. In this aerogel, there was a strong chemical bonding interaction between an NCO-terminated WPU prepolymer and hydroxyl functionalized MXene nanosheets, as shown in Fig. 32(d). As expected, the resultant aerogels were remarkably light and mechanically strong, and in particular, their high compressive stress far exceeded that of other MXene-based porous materials. Furthermore, synergistic effects of the oriented porous architecture and the multiphase skeleton endowed the hybrid aerogels with a high X-band EMI SE of 64.7 dB at 38.2 mg·cm -3 ( Fig. 32(e)), surpassing most of the reported EMI shielding materials.
CNT was also employed to enhance the mechanical robustness of MXene aerogels by Sambyal et al. [262], and via the bidirectional freezing method, the authors fabricated a 3D porous Ti 3 C 2 T x /CNT hybrid aerogel. The as-prepared CNT-reinforced MXene/CNT hybrid aerogels exhibited satisfactory mechanical robustness and increased the compressional modulus (nearly 100 times that of the pristine MXene aerogel). Furthermore, the synergism of the lamellar and porous structure of the MXene/CNT hybrid aerogels contributed to both their excellent electrical conductivity (9.43 S·cm −1 ) and superior EMI SE (103.9 dB at 3 mm thickness at the X-band frequency). Graphene oxide (GO) was used to form a 3D porous architecture of MXenes via directional-freezing and subsequent freeze-drying route in a successful demonstration by Zhao et al. [263]. The resultant hybrid aerogels exhibited an aligned cellular microstructure, where the GO sheets served as the inner skeleton, while the compactly attached Ti 3 C 2 T x sheets presented as shells of each cell. The highly conductive aerogels (1085 S·m -1 ) were highly efficient to endow epoxy nanocomposite with a high electrical conductivity (695.9 S·m -1 ) and an outstanding EMI SE (> 50 dB in the X-band at a low Ti 3 C 2 T x content of 0.74 vol%).
In conclusion, the novel 2D MXenes have presented great potential in EMI shielding applications with excellent metal conductivity, layered structure, abundant surface defects and functional groups, and easy processability in solution. The ultrathin solo MXene film exhibits high EMI SE, but their poor mechanical www.springer.com/journal/40145 properties, relatively high densities, and large EM secondary pollution caused by high reflection loss hinder their applications in the field of flexible electronics and intelligent garments. Therefore, the MXene composite films have been developed by compounding with functional materials such as polymers and metals, which can significantly improve the mechanical properties, and enhanced EM absorption. The development of MXene coatings and electronic inks has further promoted the application of MXene as flexible electronic devices. Besides, by constructing 3D porous structures (foams and aerogels) of MXenes, it can make the materials improve impedance matching and absorption loss of EM. MXenes have provided a new opportunity for the development of high-performance EMI shielding materials, and they are promising to see commercial applications as the related researches are being exploded.

Summary and outlook
In this review, the developments of MAX phases and MXenes are reviewed. The synthesis, structures, properties, and applications of the two families of materials are summarized. Especially, this review focuses on the conversion process from MAX phases to MXenes. We try to introduce the 2D materials (MXenes) to the scientists on ceramics, and clarify the difference and the connection between the 3D MAX phases and their 2D counterpart MXenes.
MAX phases are generally processed and applied as structure ceramics. The typical method to process ceramics, such as PLS sintering, hot-pressing, SPS, and SHS, can be used to process MAX phase. Compared with typical structure ceramics (Al 2 O 3 , Si 3 N 4 , SiC, etc.), MAX phases are relatively soft, damage tolerant, readily machinable, and electrically conductive. Although many applications, such as structure components, heat elements, and electrical contacts, have been explored, there are still have difficulties for the commoditization of MAX phases as structure ceramics. This is because those structural materials always cost sensitively. MAX phases, like many other new materials, are costly.
MXenes are 2D materials made from MAX phases by removing A elements. MXene can be considered as conductive clay. They can be dispersed in water for forming colloidal solution and processed as slurry or pellet. MXenes have large layer space to store metal ions or air molecules, and they are electrically conductive. Their chemical composition, surface terminations, and microstructure can be regulated to control their properties. Thus, MXene can be applied in the areas of energy storage, catalysis, EMI shielding, sensors, etc.
Different with MAX phases, MXenes are mainly applied as functional materials. Thus, the applications of MXenes are not super sensitive to cost. Then MXenes are easier to be commercialized than MAX phases. Then, the precursor of MXenes is a very possible way to commercialize MAX phases. It has been found that there are some MAX phases in sale in markets as precursors of MXenes.
For the applications of MXenes, the improvement of performance is prior to the reduce of cost. The chemical composition is a key factor to determine the performance of MXene. The chemical composition of precursor MAX phase determines the chemical composition of MXene, in turn, the performance of MXene. For example, Mo-contained MXenes and V-contained MXenes have different properties with the common Ti-contained MXenes (Ti 3 C 2 and Ti 2 C). To obtain those two kinds of MXenes, the related MAX phases, Mo 2 Ga 2 C and V 2 AlC, should be synthesized first. By the way to design MAX phases, MXenes with different chemical composition and properties can be designed and prepared. Therefore, for the purpose to obtian MXenes with desired compositions/performance, it is significant to fully understand the synthesis process of MAX phases and conversion process from MAX phases to MXenes.