Recent advances of MXenes as electrocatalysts for hydrogen evolution reaction

MXenes, an emerging two-dimensional (2D) transition metal carbides, nitrides and carbonitrides, have exhibited great potential as electrocatalysts for hydrogen evolution reaction (HER) due to the excellent characters, including excellent structural and chemical stability, superior electrical conductivity, and large active surface area. In this comprehensive study, firstly, the preparation advances of MXenes are systematically summarized. Then, the representative applications of MXenes-based HER electrocatalysts are introduced, from experimental and theoretical aspects. Thirdly, the strategies for improving HER catalytic activity of MXenes are demonstrated, such as optimizing active sites by termination modification and metal-atom doping, increasing active sites by fabricating various nanostructures. Finally, the existing challenges and new opportunities for MXenes-based electrocatalysts are also elucidated. This paper provides reference for the future development of new and efficient MXenes-based electrocatalysts for hydrogen production through water-splitting technology.


INTRODUCTION
With the growing problems of energy shortage and environmental pollution in the world, hydrogen has received extensive attention. Hydrogen has several advantages including high energy density, zero pollution emission, no greenhouse gas emission, recycling and so on. It can be well compatible with the green energy storage and conversion fields, with the merits of safety, efficiency, low-cost, sustainability, and environmental friendliness. Hydrogen can be prepared via various approaches, including burning of fossil fuels or biomass feed stock, fermentation of marsh gas or organic waste. Of which, the water-splitting based HER has been an economic and environmentally-friendly approach 1 .
The electrocatalyst is significant in HER for obtaining hydrogen, in which the precious metal-based electrocatalysts (such as Pt, Ru) exhibit good efficiency and stability. However, the ever-increasing demand for precious metal-based electrocatalysts has been hampered by limited resources and rising cost. Therefore, further development of high-efficient non-previous metal (NPM) based HER electrocatalysts for preparing hydrogen has been a pressing challenge [2][3][4][5] . Till now, considerable achievements have been made and following electrode materials have been developed: phosphides 6 , carbides 7 , sulfides 8 , nitrides 9 , oxides 10 , and even metal-free nanocarbons. There are still two main problems: one is the poor electrical conductivity, leading to the thickening of the active coating and sharp decline of activity. The other one is the insufficient electrochemical reaction stability of these NPM based materials in aqueous electrolytes. With the advances of materials, MXenes, known as the graphene-like two-dimensional (2D) transition metal carbides, carbonitrides, or nitrides, show the potential to overcome above-mentioned problems. Attributed to the high carrier mobility and intrinsic layered structure, MXenes have been regarded as a class of promising HER electrocatalysts.
Since Ti 3 C 2 T x were discovered by Gogotsi's group in 2011, MXenes have received increasing attention resulted from the excellent characters, including excellent structural and chemical stability, superior electrical conductivity, and large active surface area [11][12][13] . So far, more than 30 MXene compositions have been synthesized, and more than 100 kinds have been studied through computational methods. In addition, MXenes has also evolved from the initial type containing only one or two transition metals to the current four transition metal high-entropy 2D carbide MXenes (TiVNbMoC 3 , TiVCrMoC 3 ), and its variety has been greatly increased 14 . MXenes can be demonstrated with a general chemical formula: M n + 1 X n T x , where M is early transition metal atoms, such as Ti, Zr, Nb, and V; X is carbon or nitrogen; n presents 1, 2, 3 or 4; T x is the surface termination group, such as -F, -OH, or -Se, further, the M layers cover the X layers in [MX] n M arrangement 15 . MXenes are generally produced from ternary metal carbides or nitride precursors MAX phase. MAX phase can be demonstrated with a general chemical formula: M n + 1 AX n , where A is generally group 13-14 elements (i.e., Al, In, Si, Pb) and late transition metals (e.g., Zn, Cu, Ni, and Co); M layers are terminated with hydrophilic groups (-O, -OH, or -F) 11 . During the preparation, the M n + 1 AX n precursors are treated with concentrated hydrofluoric acid (HF), or a mixed solution of lithium fluoride and hydrochloric acid (HCl). The A layer can be selectively extracted without breaking the M-X bonds. Few-layered MXenes are gained from the exfoliated MXenes, by sonicating in organic solvent such as dimethyl sulfoxide (DMSO). Additionally, fewlayered MXenes possess multilayer structure (less than 1 nm in thickness) and millimeter-scale lateral dimension 16 .

SYNTHESIS STRATEGIES OF MXENES
In recent years, extensive efforts have been invested to develop diverse routes for preparing MXenes in cost and energy-economic strategies. The clean production and green chemistry should be ever-increasingly considered in producing MAX phase materials. So far, MXenes have been generally prepared from MAX phase with fluoride etching, alkali treatment, electrochemical etching, and water-free etching, which will be discussed in this section (Fig.  1). The corresponding comparisons of different etching methods for the synthesis of MXenes are listed in Table 1.

Fluoride etching
There are two general ways for performing fluoride etching. One is direct etching with concentrated HF. The MAX phase is etched at room temperature under constant agitation, then MXenes with multilayer structures are exfoliated by sonicating in methanol and isopropyl alcohol. The other is reaction-generated HF in situ 38 from HCl and fluorides, including LiF 39 , NaF 40 , KF 41 , CsF 39 , CaF 39 , FeF 3 41 , and tetra-n-butyl ammonium (Fig. 2a) 39 . SEM analysis shows Ti 3 AlC 2 (Fig. 2b) has a layered structure, resemble to the 'accordion'. Restacked multi-layers with obvious sheet gaps could be observed in the particles after etching MAX (Fig. 2c). SEM, TEM and high-resolution TEM image of as-prepared monolayered Ti 3 C 2 T x (MXene) nanosheets revealed electron-transparent foils with thin, smooth surfaces and diameters ranging from 100 nm to several micrometers (Fig. 2d-f).
The synthesis of MXenes has been reported by Naguib et al. 12 , referring to the chemical reaction between MAX phase and fluorine-rich aqueous environment. The Al atoms were removed (reaction 1), and the MXenes were terminated with -OH and -F (reaction 2 and 3, respectively). The oxygen termination was also observed in MXenes. In other words, the Ti-Al bond was replaced by Etching with HF or other fluorinated solutions has been successfully applied in the development of MXenes, which has become a standard preparation method for MXenes. However, this method shows several problems. First, the Al atoms in MAXphase precursors show high reactivity with F − and it limited the synthesis of various MXenes due to fluoride-based aqueous solutions mainly effective for Al-containing MAX-phase precursors. Second, the involvement of HF leads to safety and environmental problems. Third, the resulted inert F-terminal reduces material performance, thus hindering the further applications of MXenes. To sum up, it is significant to develop innovative HF-free etching strategies.
Electrochemical etching (1) Low yield (1) Avoid the use of HF [55][56][57] (2) Extend the range of etching techniques and potential compositions of MXenes. single-layer Ti 3 C 2 (OH) 2 were successfully prepared. However, with alkali exposure, the formation of Al oxide/hydroxide layers, such as Al(OH) 3 and AlO(OH), may further hinder the required Al extraction. In xie' s work 50 , only the aluminum atoms were removed from the outer surface of Ti 3 AlC 2 . Besides, Zou et al. 52 reported that some Na/K-Ti-O compounds rather than MXene were produced while the alkali treatment removed part of the aluminum. Therefore, these oxidation layers must be removed. NaOH assisted hydrothermal etching of Ti 3 AlC 2 was reported to prepare Ti 3 C 2 T X (T = OH, O) ( Fig. 3a-e) 54 . This hydrothermal treatment was carried out in an argon atmosphere to minimize the oxidation of the sample. The multilayer Ti 3 C 2 T X could be obtained by selective removal of Al in Ti 3 AlC 2 with NaOH solution at 270°C. The purity of product with OH and O terminal groups was 92 wt%, without observable F terminal group. The SEM image of Ti 3 C 2 T x (270°C, 27.5 M NaOH) shows a compact layered structure (Fig. 3f). The morphologies of these Ti 3 C 2 T x (MXene) nanosheets were analyzed by TEM and HAADF-STEM ( Fig. 3g-h). It indicates that Ti 3 C 2 T x sheets have a nonuniform interlayer spacing of ∼1.2 nm, which is larger than the pristine Ti 3 AlC 2 (0.93 nm). The mass capacitance and volume capacitance of the Ti 3 C 2 T X thin film electrode was up to 314 F·g −1 and 511 F cm −3 . Compared with the multilayer Ti 3 C 2 T X prepared by HF, the mass capacitance was increased by 214%. This alkali etching method succeeds to remove the amphoteric or acidic atoms from original MAX phase. It provides a HF-free strategy for developing innovative MXenes.

Electrochemical etching
Chemical etching is based on the reactivity difference between M-Al and M-C bond. However, electrochemical etching refers to the direct transfer of charge. Sun et al. 55 demonstrated the successful electrochemical extraction of Al on porous Ti 2 AlC electrode in diluted HCl solution. A layer of Ti 2 CT x MXenes can be obtained on Ti 2 AlC (Fig. 4a). Compared to chemical etching using HF or LiF/HCl, the electrochemical etching method did not involve any F ion, thus obtaining MXenes with only -Cl, -O and -OH groups. However, the problems in electrochemical etching may be the over-etching of MAX phase into carbines-derived carbon (CDC). A core-shell model was proposed to interpret how the Ti 2 AlC was electrochemically etched to Ti 2 CT x and CDC (Fig. 4b). It indicated that etching parameters should be carefully balanced to generate MXenes while avoiding overloading. The electrochemical etching has extended the range of etching techniques and potential compositions of MXenes.
Yang et al. 56   can be selectively removed, and then replaced with hydroxyl groups. Thus, single-layer or double-layer Ti 3 C 2 T x nanosheets can be formed, with large average size and a yield higher than 90%. SEM images (Fig. 4f) reveal that Ti 3 C 2 T x sheets have irregular edges and mostly dispersed size distribution generally ranging from 1 to 5 μm, due to an exfoliation-fragmentation mechanism. Under mild etching conditions, the diameters of some larger slices can reach 18.6 μm (Fig. 4g). TEM image (Fig. 4h) displays a typical Ti 3 C 2 T x sheet with well-defined and clean edges. The results were similar or even superior to chemical etching techniques. The obtained Ti 3 C 2 T x thin film was applied to assemble solid supercapacitors for energy storage. The area capacitance can be up to 220 mF cm −2 . In addition, neither dangerous fluorinated agent was required, nor tough etching conditions. It is suitable for the preparation of innovative MXenes.
A thermal-assisted electrochemical etching has been reported 57 . MXenes were prepared under HF-free conditions, including Ti 2 CT x , Cr 2 CT x , and V 2 CT x . A three-dimensional (3D) composite electrode was applied. In the presence of diluted HCl as etchant, gentle heating sped up the etching reaction of MAX phase, thus greatly improving the etching efficiency of MXenes. Since Ti-based MXenes have been widely prepared, Ti 2 CT x was selected as a typical example in this study. Ti 2 CT x was prepared by electrochemical etching following a two-stage strategy. In the stage 1, Al atoms were firstly removed from layered structure under the effects of applied voltage, since the Ti-Al bond was easier to be broken compared to the Ti-C bond. The MXenes can be obtained. In the stage 2, Al and Ti atoms would be completely removed, leaving only monolayer carbon atoms. By controlling the etching time, temperature, and electrodes, the MXenes with different morphologies can be obtained. With this method, another two MXenes (V 2 C and Cr 2 C) can be successfully obtained, which were generally considered difficult to prepare. In previous reports, V 2 C required HF corrosion for more than two days, with a HF concentration of up to 50%. The Cr 2 C had not been successfully prepared. The thermal-assisted electrochemical etching has not only proposed a universal strategy for the synthesis of MXenes, but also fabricated MXenes difficult to prepare. It opens a window for preparing MXenes in a rapid, facile and safe way.
Water-free etching At present, the methods of non-aqueous etching include molten salt etching (Lewis acid etching) and iodine assisted etching. The reaction of MAX phase in molten salt (Lewis acid) was similar to that of in HF, such as the reaction of Ti 3 AlC 2 in molten ZnCl 2 58 . The Ti 3 ZnC 2 MAX phase was obtained by substitution reaction, and then it was transformed into Ti 3 C 2 Cl 2 MXenes by increasing the ratio of MAX: ZnCl 2 . After optimizing the reaction, a class of M n + 1 ZnX n phases were prepared through substitution reaction, thus obtaining corresponding MXenes (M n + 1 X n Cl 2 ) with Cl group on the surface.
Following the report of MXenes synthesis by molten Lewis acid in 2019, Huang et al. 59  research, then proposed the strategy of preparing MXenes with general molten salt. The redox potential of Lewis acid cations was so high as to oxidize the A-layer atoms in the MAX phase. Under the guidance of this redox potential-based rule, a variety of molten salts (CdCl 2 , FeCl 2 , CoCl 2 , CuCl 2 , AgCl, NiCl 2 ) were further applied to etching different MAX phases (Ti 2 AlC, Ti 3 AlC 2 , Ti 3 AlCN, Nb 2 AlC, Ta 2 AlC, Ti 2 ZnC, Ti 3 ZnC 2 ) to prepare the corresponding MXenes. These results suggested that in the traditional MAX phase, A element can be replaced with the late transition metal halide. It has greatly expanded the range of MAX phase as precursors of MXenes. As a HF-free method, Lewis acid etching provides a green and feasible way for preparing MXenes. The range of Lewis acid etching can be further expanded. In the study of Dmitri V. Talapin 60 , MAX phase was first treated with molten CdCl 2 to prepare Ti 3 C 2 Cl 2 , Ti 2 CCl 2 , and Nb 2 CCl 2 with Clterminated group. Then, the Br-terminated MXenes were also prepared using CdBr 2 molten salts. Concerning the surface groups, the bond energy of the Ti-Cl bond (405 kJ mol −1 ) or the Ti-Br bond (373 kJ mol −1 ) in these MXenes was much weaker than that of the Ti-F bond (569 kJ mol −1 ) or the Ti-O bond (666 kJ mol −1 ) in the classical MXenes. The termination of transformation chemistry was further explored. Ti 3 C 2 Br 2 was dispersed into molten CsBr/KBr/LiBr, then Li 2 Te or Li 2 S was added to obtain Ti 3 C 2 Te or Ti 3 C 2 S, respectively. Similarly, they also converted Ti 3 C 2 Cl 2 to Ti 3 C 2 Te, Ti 3 C 2 S, and Ti 3 C 2 (NH). Interestingly, LiH was able to reduce Ti 3 C 2 Br 2 and Ti 2 CBr 2 to Ti 3 C 2 □ 2 and Ti 2 C□ 2 (where □ indicated defect) at 300°C. Above samples were further treated with n-butyl lithium (n-BuLi) for obtaining the product with Li intercalation layer, which could be further dispersed in polar organic solvents to form monolayered product (Fig. 5a). TEM image of Ti 3 C 2 Cl 2 momolayered displays crystallinity and hexagonal symmetry of the individual flake (Fig. 5b-c) and negative zeta potential (−29.3 mV). Termination transformation was accompanied by the dramatic change of lattice structure (Fig. 5d-e), indicating that surface modification made a great impact on all aspects of MXenes performance. The electron transport operation of the Nb 2 C MXenes was examined ( Fig. 5f-g). The superconducting transition temperature of Nb 2 CCl 2 was about 6 K, while the bulk phase of Nb 2 AlC was not superconductive.
Recently, Shi et al. 61 reported a non-aqueous iodine (I 2 ) assisted etching route for the synthesis of 2D MXene with oxygen-rich terminal groups. The Ti 3 C 2 I x was firstly prepared through iodine etching in anhydrous acetonitrile at 100°C, followed by the stratification in HCl solution to further transformed Ti 3 C 2 I x into Ti 3 C 2 T x flakes with moderate sizes (ca. 1.8 μm), over 71% of which were thinner than 5 nm (Fig. 6a, b). In addition, the obtained Ti 3 C 2 T x flakes showed an excellent ambient stability in dispersions for at least 2 weeks. The mechanism of the synthetic process was proposed as below: According to the X-ray diffraction (XRD) patterns (Fig. 6c), the characteristic (104) peak of Ti 3 AlC 2 (2θ = 39°) in the I 2 -etched MAX disappeared, and the (002) peak (2θ = 9.8°) shifted to a lower angle (2θ = 6.1°), which reflected the expansion of interlayer spacing from 9.3 Ǻ to 14.4 Ǻ, thus indicating the successful Al etching. Due to the subsequent washing and delamination, the (002) peak of the 2D IE-MXene shifted to an even lower angle (2θ = 5.2°), indicating the further expansion of interlayer spacing from 14.4 Ǻ to 17.4 Ǻ. The SEM images (Fig. 6d-f) also clearly revealed the morphological changes.

MXENES-BASED ELECTROCATALYSTS FOR HER
Development of hydrogen energy offers an efficient solution to today's environmental and energy issues 62,63 . HER is an important strategy for water splitting and hydrogen production. It is critical to develop HER catalysts with superior performance in the proposed hydrogen economy, including good conductivity, stability, selectivity etc. The high-performance electrocatalyst can minimize the overpotential necessary for the HER thus improving efficiency 64,65 . In theory, density functional theory (DFT) calculations can be used to determine the Gibbs free energy of hydrogen adsorption (ΔG H ), which has been proved to be a descriptor of HER activity as a first approximation [66][67][68][69][70][71] . And when ΔG H is closest to thermal-neutral, the HER activity is the best.
Compared with other NPM based HER electrocatalysts, MXenes has higher potential, as well as other excellent physical and chemical properties: In recent years, MXenes-based HER electrocatalysts have attracted increasing attention, and there have been numerous experimental and theoretical research. The MXenes have been optimized intrinsically and extrinsically via structural engineering, from the aspects of termination modification, metal-atom doping, nanostructure fabrication, and hybridization. MXenes and MXenebased composites as HER electrocatalysts display potential progress in replacing Pt-based catalysts ( Table 2).

Termination modification
Termination modification of MXenes is considered beneficial for enhancing HER performance, via improving the conductivity and surface termination 71 . It is well recognized that the termination modification of MXenes can optimize the electronic structure and promote HER activity constitutionally. In recent years, termination modification of MXenes has been widely investigated with experimental and theoretical studies.
2D MXenes with O*/OH* terminal groups exhibited metallic properties experimentally, thus supporting excellent charge transfer 72 . Consistent with the experimental results, the surface Pourbaix diagrams also proved that MXenes with O*/OH* terminal groups show the best surface chemical stability. The performance of O*/OH*-terminated MXenes was also optimized, manifested as dramatically improved exchange current and promoted hydrogen evolution ( Fig. 7a-b). The ΔG H was calculated, demonstrating that surface oxygen atoms were active sites for the HER. These surface oxygen atoms promoted the interaction between H* and 2D MXenes. Further study demonstrated that the HER process of O-terminated MXenes followed the Heyrovsky mechanism. The HER activity of MXenes with diverse surface groups (-O, -OH, -F) was exploited (Fig. 7d-f) 73 . The O-terminated surface was much more effective than OH-and F-terminated surface. Above studies have proved that the catalytic activity of 2D MXenes based HER catalysts can be adjusted with termination modification.
The theoretical analysis has been further performed. Ling et al. 74  Several other studies on termination modification have been performed. Jiang et al. 75 designed an efficient HER electrocatalyst based on the ultrathin O-functionalized Ti 3 C 2 MXenes. The counterpart F-functionalized MXenes on the basal plane was detrimental to HER, depressing hydrogen adsorption kinetics. Ti 3 C 2 O x was prepared by dispersing Ti 3 C 2 T x in KOH aqueous solution, and the F-termination was reduced with OH groups 76 . Then, the Ti 3 C 2 (OH) x was calcinated under 450°C in Ar atmosphere, and the OH groups would be transformed into O-terminal group via dehydration reaction. As shown in Fig. 7g-h, Ti 3 C 2 O x nanosheets was proved to be a superior HER electrocatalyst. It provided an overpotential of 190 mV at current density of 10 mA cm −2 , which was much lower than that of Ti 3 C 2 (OH) x (217 mV) and Ti 3 C 2 T x -450 (266 mV). Further, the Tafel slope was 60.7 mV dec −1 , resulted from the highly active O-sites on the basal plane of Ti 3 C 2 O x MXenes. This work lays the foundation for improving the performance of MXenes-based HER electrocatalysts by modifying their surface terminal groups.
In short, the catalytic activity of MXenes in HER is influenced by the functional groups on the basal plane. For Ti 3 C 2 , Mo 2 C, Mo 2 Ti 2 C 3 MXenes, higher F content on the basal plane indicated lower HER activity 77 . The O-terminated Ti 3 C 2 MXenes is an ideal electrocatalyst for HER 70,73 . Thus far, MXenes with other surface groups (such as -S, -Cl, -Br) have been synthesized, but their HER performance has not been discussed.  80 , etc. to improve the performance of HER electrocatalysts, from the aspects of tuning the electronic structure, modifying the elemental composition, and handling surface chemistry 81,82 . In the 2D material lattice and coordination environment, the single-atom catalysts can stabilize single metal atoms with strong covalent bonds, which can function as active sites for HER (Fig. 8a) 83 . Inspired by this phenomenon, the HER catalytic activity of MXenes can be enhanced experimentally and theoretically by confining single metal atom.
The structural and computational model of transition metal (TM)-promoted V 2 CO 2 was established (Fig. 8b) 84 . Four surface sites with different activity could be obtained under varied hydrogen coverage rate. These sites were labeled as T 0 , T 1 , T 2 , and T 3 , in which 'T' indicated 'top site of surface O atom' and the subscript was the number of surrounding V atoms that connected to active O atom and the promoter atom. By using first-principles calculation (Fig. 8c), ΔG H was generally greater in TM-promoted V 2 CO 2 , compared to that of pristine V 2 CO 2 . The bonding strength of H-O on the surface of the TM-V 2 CO 2 can be reduced, thus improving HER performance. Moreover, by adjusting the adsorption concentration of TM atoms on the surface, the catalytic activity can also be effectively regulated. Most importantly, ΔG H of T 3 (25% TM-Ni), T 0 (16.7% TM-Fe), T 1 (16.7% TM-CO), T 1 (16.7% TM-Fe), and T 2 (12.5% TM-CO) was only −0.01, −0.04, −0.05, 0.03, and −0.03 eV, respectively, showing performance superior to commercial Pt/C (≈0.09 eV) 85 .
The effects of metal atoms (Ir, Os, Re, Rh, Fe and Zn) modification on HER activity were also explored, with Ti 3 CNO 2 as a model (Fig. 8d) 86 . On each side of Ti 3 CNO 2 MXenes, three O-atom sites were systematically selected (labeled as S 1 , S 2 and S 3 ), and the corresponding ΔG H was calculated. All three kinds of O sites on the C-side of Ir, Rh, Zn modified system aided in the enhancement of HER activity. Further, at least one O site on the C-side of Re, Os and Fe doped Ti 3 CNO 2 surface exhibited HER activity (Fig. 8e). Li et al. 87 also demonstrated that the catalytic activity of MXenes materials could be optimized by surface modification with TM atoms. These theoretical results provide vital insights into catalytic mechanism of TM-doped MXenes for HER. It assists in designing and applying TM-doped MXenes in electrochemical reactions.
Apart from the theoretical basis, several strategies have been developed experimentally to produce monometallic atom doped MXenes, including electrochemical exfoliation 88 , pyrolysis 89 , MAX precursor doping 90 . A stable single-atom catalyst Pt 1 /Ti 3−x C 2 T y was prepared by using ultrathin 2D Ti 3−x C 2 T y MXenes, which provided many Ti defect vacancies and high reduction capacity 91 . A strong metallic carbon bond was formed between the single atom and Ti 3−x C 2 T y . Zhang et al. 88 designed single Pt atom-MXenes catalyst (Mo 2 TiC 2 T x -Pt SA ) with a high mass activity. Electrochemical exfoliation of double transition metal MXenes (Mo 2 TiC 2 T x ) was firstly performed, followed by immobilizing Pt single atoms in the Mo vacancies ( Fig. 8f-g). As a result, Mo 2 TiC 2 T x -Pt SA delivered an overpotential of 30 mV at 10 mA cm −2 , much better than that of the pristine Mo 2 TiC 2 T x and Mo 2 TiC 2 T x -V Mo (Fig.  8h). The Density functional theory (DFT) calculations showed that the ΔG H values of Mo 2 TiC 2 T x -Pt SA was -0.08 eV, which was lower compared to that of Mo 2 TiC 2 T x (-0.19 eV) and Pt (−0.1 eV). Besides, Mo 2 TiC 2 T x -Pt SA showed excellent stability as HER electrocatalyst, resulted from the strong interaction of Pt atoms with Mo 2 TiC 2 T x , thus preventing the surface diffusion and coarsening. Single site Co-substituted 2D molybdenum carbide (Mo 2 CT x : Co) can be obtained from a Co-substituted Mo 2 GaC MAX precursor 90 . The Mo 2 CT x : Co delivered an overpotential of 180 mV at 10 mA cm −2 , better than that of Mo 2 CT x (230 mV). The DFT result suggested that Co-substituted Mo 2 CT x promoted the adsorption of hydrogen on the MXenes surface, thus enhancing the HER kinetics.
Nanostructuring Nanostructure architecting has been another strategy for improving HER electrocatalytic activity of MXenes, such as nanoribbons and nanodots. The MXenes with certain nanostructures possess quick electrochemical response, special electronic properties, and abundant active sites, which can be potential HER electrocatalysts. Nanostructuring expands the application scope of MXenes and assists in assembling active components.
Twelve kinds of MXenes nanoribbon systems were constructed 92 . DFT calculations revealed that the edges of MXenes nanoribbons containing metal and C(N) atom could capture hydrogen species, which can be applied as the reaction sites for hydrogen evolution with fast kinetics. These MXenes nanoribbons displayed different ΔG H (Fig. 9a-c). Notably, the nanoribbons of nanostructure is an innovative strategy for enhancing the HER activity of MXenes. Several synthetic strategies have been reported to prepare MXenes nanoribbons or nanodots, such as hydrothermal treatment 93 , shaking treatment 94 , or ball-milling 95 . Ti 3 C 2 T x nanofibers (NFs) can also be fabricated from hydrolyzed bulk Ti 3 AlC 2 and HF etching (Fig. 9d) 96 . As shown in Fig. 9e-f, the obtained Ti 3 C 2 T x NFs exhibited high HER catalytic activity, manifested as an overpotential of 169 mV at 10 mA cm −2 , which was better than Ti 3 C 2 T x flakes (385 mV). The Tafel slope was 188 and 97 mV dec −1 for the Ti 3 C 2 T x flakes and NFs, respectively. Meanwhile, the MXenes NFs provided much higher specific surface area (up to 58.5 m 2 g −1 ) compared to that of MXenes flakes (8.5 m 2 g −1 ), thus exposing more active sites (Fig. 9g). Hence, the high HER catalytic performance of proposed Ti 3 C 2 T x NFs was mainly resulted from the improved specific surface area and increased active sites.
A 3D interconnected porous framework of alkalized Ti 3 C 2 (a-Ti 3 C 2 ) MXenes nanoribbons were prepared via continuously shaking Ti 3 C 2 MXenes in alkalized solution (KOH). Because of the increased interlayer spacing and narrowed nanoribbons width, the reaction kinetics and morphology stability can be improved 93 . Such alkalization strategy can be further broadened to prepare other MXenes-based materials, which are applicable to a large family of MAX phases. Water-soluble and monolayered Ti 3 C 2 MXenes quantum dots (MQDs) were prepared with hydrothermal treatment 94 . Upon etching the A elements in the Ti 3 AlC 2 MAX phase with 48% HF acid, bulk layered Ti 3 C 2 MXenes were cut. Fewlayered Ti 3 C 2 T x nanodots (TNDs) (2-5 nm in thickness, ≈ 6 nm in size) were synthesized using ball-milling with red phosphorus (P) 95 . In theory, MXenes would be increased in practice by reducing their lateral size to the nanoscale. The surface O groups of Ti 3 C 2 T x would interact with P strongly and chemically, thus forming nanodots under the ball-milling shear force. Notably, almost all raw Ti 3 C 2 T x was converted to TNDs, and the sizes can be controlled by ball-milling with other elements, including carbon, sulfur, and silicon.
In short, MXenes nanostructures can directly affect the HER activity, from the aspects of catalytic process, active sites, and charge transfer 97,98 . Several methods have been reported to construct nanostructures of MXenes, from 1D to 3D architectures 99,100 . More than 30 kinds of MXenes with various nanostructures have been experimentally synthesized and applied as catalysts. These MXenes are promising candidates for HER electrocatalyst.

Hybridization
The hybridization of MXenes with other active materials has been demonstrated to synergistically promote HER activity, such as chalcogenides [101][102][103][104][105][106][107][108][109][110] , layered hydroxides 111 , phosphides 112,113 , metal nanoparticles/alloys [114][115][116][117][118][119][120][121] , carbides 122,123 and even metalfree black phosphorus 124 . The ΔG H of MoS 2 can be precisely tuned 125 by integrating with various 2D materials, such as graphene, h-BN, phosphorene, transition metal dichacolgenides, MXenes, and their derivatives. The catalytic performance of MoS 2 can be enhanced at low S vacancy concentration with firstprinciples calculations. It was worth noting that the optimal free energy ΔG H = 0 could be realized at S vacancy concentration as low as~2.5%, and there was high porosity and tunability in MoS 2 / MXenes-OH heterostructures. Recent experimental studies have also confirmed above findings. As displayed in Fig. 10a, a quick-freezing method was designed to prepare the 'nanoroll'-like MoS 2 /Ti 3 C 2 T x composite 101 . After liquid nitrogen-freezing and annealing, a hierarchical MoS 2 / Ti 3 C 2 T x can be obtained, with a diameter of~200 nm and length of several microns. The rapid freezing functioned in rolling Ti 3 C 2 T x nanowires and forming vertically oriented MoS 2 microcrystals. This MoS 2 /Ti 3 C 2 T x composite provided abundant reaction sites and promoted charge transfer in electrocatalytic reactions. Good catalytic performance was achieved, manifested as an initial overpotential of 30 mV and a low overpotential of 168 mV at 10 mA cm −2 . The Tafel slope was 70 mV dec −1 . In addition, the exchange current density was increased by over 25 times compared to that of MoS 2 (Fig. 10b-c). By mounting a few carbon coated MoS 2 nanocrystals on the carbon stabilized Ti 3 C 2 MXenes, the spontaneous oxidation can be hindered 93 117 reported Pt 3 Ti intermetallic compound (IMC) NPs on Ti 3 C 2 T x MXenes, in which Pt interacted with Ti on the Ti 3 C 2 T x MXenes via in-situ co-reduction. The in-situ X-ray absorption spectra showed that Pt atom was transformed to IMC with increased temperature (Fig. 10d). The obtained Pt/Ti 3 C 2 T x -550 showed excellent performance, requiring a potential of only 32.7 mV at 10 mA cm −2 (Fig. 10e). The mass activity and specific activity of Pt/Ti 3 C 2 T x -550 was better than that of Pt/Vulcan (Fig. 10f). Wang et al. 121 synthesized a class of composite FeNi@MXene (Mo 2 TiC 2 T x )@nickel foam (NF) through introducing Fe 2 + ions and in-situ combining with surface nickel atoms on nickel foam. FeNi@Mo 2 TiC 2 T x @NF exhibited high HER activity with an overpotential of 165 mV at 10 mA·cm −2 due to the synergetic effect of Mo 2 TiC 2 T x and FeNi nanoalloys. Du et al. 114 reported a Ti 3 C 2 T x MXenes based hybrid material modified by doping Nb and surface Ni/Co alloy (Fig. 10g). DFT calculation show that, after Nb doping, the Fermi energy level was shifted toward the conduction band, thus improving the conductivity. In addition, Ni/Co alloy modified M-H bond on the surface of catalyst, thus adsorbing the H with the lowest ΔG H (Fig.  10h). The synthesized Ni 0.9 Co 0.1 @NTM (Nb-doped Ti 3 C 2 T x ) hybrid exhibited excellent HER activity, requiring an overpotential of only 43.4 mV at 10 mA cm −2 (Fig. 10i). Further, long-term stability can be obtained.

CONCLUSION AND OUTLOOK
The development of NPM, environmentally friendly and highperformance electrode materials is significant for energy storage, conversion, and utilization. Because of the large surface area, excellent metallic conductivity, and good hydrophilicity, MXenes have broad prospects as electrode material [126][127][128] . This review has focused on the synthesis strategies and application advances of MXenes as electrocatalysts for HER. First, the synthesis methods of MXenes are summarized, including alkali treatment, fluoride etching, electrochemical etching, and water-free etching. It should be noted that different etching environments will generate different morphology and adjustable surface chemistry of layered MXenes products. Thus, appropriate methods should be selected according to our research objectives. Then promoting HER performance via termination modification, metal-atom doping, nanostructure architecting and hybridization have been systematically introduced (Fig. 11). Although extensive progress has been made in developing MXenes-based electrocatalysts during the past few decades, it remains challenging to develop highly active electrocatalysts that are superior to Pt-based materials and can be commercialized. For instance, (a) Etching in aqueous solution (fluoride etching, alkali etching, electrochemical etching) is detrimental to obtaining high-quality MXene because the dissolved oxygen gas in aqueous etchants induces extra structural defects to MXene sheets and promotes their degradation into TiO 2 . (b) Non-aqueous etching by molten salt requires high temperature, making it difficult to industrialize. (c) The surface functional groups of MXene are still difficult to tune, for which depend on the synthesis method. Based on this, the catalytic mechanism remains controversial due to complex surface environment. (d) The yield of MXene nanostructure is low, and can limit the practicality of MXenes on a large scale.
Based on the above synthesis and modification strategies, the hydrogen production efficiency of MXenes-based electrocatalysts in the future can be improved from the following aspects:

DATA AVAILABILITY
Data sharing was not applicable to this paper, as no original data were generated for this review. Fig. 11 Strategies for optimizing the MXenes-based HER electrocatalysts. Schematic diagram summarizing different approaches for optimizing the MXenes-based HER.