Recent Advances in Surface Modifications of Elemental Two-Dimensional Materials: Structures, Properties, and Applications

The advent of graphene opens up the research into two-dimensional (2D) materials, which are considered revolutionary materials. Due to its unique geometric structure, graphene exhibits a series of exotic physical and chemical properties. In addition, single-element-based 2D materials (Xenes) have garnered tremendous interest. At present, 16 kinds of Xenes (silicene, borophene, germanene, phosphorene, tellurene, etc.) have been explored, mainly distributed in the third, fourth, fifth, and sixth main groups. The current methods to prepare monolayers or few-layer 2D materials include epitaxy growth, mechanical exfoliation, and liquid phase exfoliation. Although two Xenes (aluminene and indiene) have not been synthesized due to the limitations of synthetic methods and the stability of Xenes, other Xenes have been successfully created via elaborate artificial design and synthesis. Focusing on elemental 2D materials, this review mainly summarizes the recently reported work about tuning the electronic, optical, mechanical, and chemical properties of Xenes via surface modifications, achieved using controllable approaches (doping, adsorption, strain, intercalation, phase transition, etc.) to broaden their applications in various fields, including spintronics, electronics, optoelectronics, superconducting, photovoltaics, sensors, catalysis, and biomedicines. These advances in the surface modification of Xenes have laid a theoretical and experimental foundation for the development of 2D materials and their practical applications in diverse fields.

Borophene possesses some unique physical and chemical properties. Due strong B-B bonds and distinctive atomic structure, borophene exhibits an ultrahig chanical modulus [83,84]. Young's modulus of the Pmmm and 2-Pmmn phases of phene along the armchair direction can reach 574.61 and 398 N m −1 , respectively, than that of graphene. In fact, several factors, including the vacancy concentration, ical modification, layer numbers, and temperature, can affect the mechanical prop of borophene. Young's modulus of borophene was found to decrease with an incre vacancy concentration, layer numbers, and temperature, as well as hydrogenati fluorination [83][84][85][86][87][88][89]. Because of its outstanding flexibility and excellent electronic co tivity, borophene has a wide range of intriguing applications in flexible electronic d [79]. In addition, the thermal conductivity of the 2-Pmmn phase of borophene is dif along the zigzag and armchair directions due to its highly anisotropic atomic stru [90,91]. The electronic band structures of borophene show 1D, nearly free electron
Borophene possesses some unique physical and chemical properties. Due to the strong B-B bonds and distinctive atomic structure, borophene exhibits an ultrahigh mechanical modulus [83,84]. Young's modulus of the Pmmm and 2-Pmmn phases of borophene along the armchair direction can reach 574.61 and 398 N m −1 , respectively, larger than that of graphene. In fact, several factors, including the vacancy concentration, chemical modification, layer numbers, and temperature, can affect the mechanical properties of borophene. Young's modulus of borophene was found to decrease with an increasing vacancy concentration, layer numbers, and temperature, as well as hydrogenation or fluorination [83][84][85][86][87][88][89]. Because of its outstanding flexibility and excellent electronic conductivity, borophene has a wide range of intriguing applications in flexible electronic devices [79]. In addition, the thermal conductivity of the 2-Pmmn phase of borophene is different along the zigzag and armchair directions due to its highly anisotropic atomic structure [90,91]. The electronic band structures of borophene show 1D, nearly free electron states and metallic Dirac fermions [92][93][94][95]. However, the metallic-to-semiconducting transition can be achieved via fluorination and a uniaxial or biaxial strain [85,96,97]. Last but not least, the superconductivity of borophene is the most notable characteristic, sparking plenty of research interest [80,[98][99][100][101][102][103][104][105][106]. The superconducting transition temperatures (T c ) can be tuned with vacancy concentration, doping, strain, and Mg intercalation. T c exhibits a V-shaped function as the hexagon hole density [99], illustrating that T c gradually decreases with a rising boron vacancy concentration of up to ν = 1/9; thereafter, T c steadily increases with the vacancy concentration. Furthermore, tensile strain or hole-doping can increase T c ; in contrast, a compressive strain or electron-doping decreases T c [106]. The suppression induced by electron-doping makes it difficult to experimentally probe superconductivity in substratesupported borophene.
The physical and chemical properties of borophene can be tuned using surface modifications, such as hydrogenation, fluorination, doping, intercalation, strain, etc. Young's modulus of borophene decreases after hydrogenation or fluorination. Furthermore, hydrogenation can lead to Dirac cones with massless Dirac fermions in C2/m, Pbcm, and Pmmn structures, while Cmmm structures exhibit Dirac ring features (Figure 4a-d) [107]. Interestingly, BH sheets have been successfully prepared from MgB2 by using the cation exchange method (Figure 4e,f) [108]. For Mg intercalation, the intercalated bilayer borophene (B2MgB2) can exhibit good phonon-mediated superconductivity with a high Tc of 23.2 K (Figure 4g,h) [102]. Moreover, tensile strain in borophene is beneficial for superconducting [106]. Li-doped borophene-graphene heterostructure shows gas-sensitive properties, and this is promising for borophene-based gas sensors [109].  Gallenene and thallene. The surface modifications of aluminene and indiene have not been reported before, as studies on them are still in the theoretical research stage, without experimental realization. Therefore, this section will focus on gallenene and thallene. In 2018, few-layer gallenene was first obtained using the solid melt exfoliation technique [29]. Thereafter, the epitaxial growth method was used to prepare gallenene. The substrate and the loading amount of gallium can modify the atomic and electronic structures of gallenene. With a low loading amount of gallium, the monolayer gallenene grown on Si(111) displays a 4 × √ 13 superstructure (Figure 5a-d), while the second-layer gallenene exhibits a hexago-nal honeycomb structure with a high loading amount (Figure 5e,f) [110]. The buckled honeycomb gallenene shows metallic properties (Figure 5g) [110]. Nevertheless, the growth behavior of gallenene on a GaN(0001) substrate is totally different, showing a bilayer flat gallenene (Figure 5h,i) [111]. Excitingly, the bilayer hexagonal gallenene exhibits superconducting with a T c of 5.4 K (Figure 5j,k) [111]. However, thallene has rarely been reported.
without experimental realization. Therefore, this section will focus on gallenene and thallene. In 2018, few-layer gallenene was first obtained using the solid melt exfoliation technique [29]. Thereafter, the epitaxial growth method was used to prepare gallenene. The substrate and the loading amount of gallium can modify the atomic and electronic structures of gallenene. With a low loading amount of gallium, the monolayer gallenene grown on Si(111) displays a 4 × √13 superstructure (Figure 5a-d), while the second-layer gallenene exhibits a hexagonal honeycomb structure with a high loading amount ( Figure  5e,f) [110]. The buckled honeycomb gallenene shows metallic properties (Figure 5g) [110]. Nevertheless, the growth behavior of gallenene on a GaN(0001) substrate is totally different, showing a bilayer flat gallenene (Figure 5h,i) [111]. Excitingly, the bilayer hexagonal gallenene exhibits superconducting with a Tc of 5.4 K (Figure 5j,k) [111]. However, thallene has rarely been reported. Recently, honeycomb thallene was successfully formed on a NiSi2/Si(111) substrate [112].

Group IV
Graphene. The sp 2 hybridization of carbon atoms leads to the formation of flat honeycomb graphene with a σ bond between neighboring carbon atoms [113]. The σ bond is the key to the high robustness of graphene, with a Young's modulus of 1T Pa and a fracture strength of 130 GPa [114]. The unhybridized p orbit, perpendicular to the planar structure of graphene, binds covalently with neighboring carbon atoms to form a π band that is half-filled. Graphene is a semimetal, showing linear dispersion bands near the Fermi level with massless Dirac fermions [113]. As a result, graphene displays an ambipolar electric field effect and high carrier mobility [115]. In addition to its excellent transparent properties, graphene becomes a low-cost alternative to indium tin oxides [116]. Furthermore, graphene exhibits impressive thermal properties with a thermal conductivity ranging from 3000-5000 W m −1 K −1 [117].
Graphene's distinct physical characteristics make it potentially useful for field-effect transistors (FET), sensors, transparent conductive films, energy devices, etc., but its intrinsic gapless character still constrains its further applications. First, heteroatom-doping or chemical adsorption can effectively tune the electronic properties of graphene. N-doping

Group IV
Graphene. The sp 2 hybridization of carbon atoms leads to the formation of flat honeycomb graphene with a σ bond between neighboring carbon atoms [113]. The σ bond is the key to the high robustness of graphene, with a Young's modulus of 1T Pa and a fracture strength of 130 GPa [114]. The unhybridized p orbit, perpendicular to the planar structure of graphene, binds covalently with neighboring carbon atoms to form a π band that is half-filled. Graphene is a semimetal, showing linear dispersion bands near the Fermi level with massless Dirac fermions [113]. As a result, graphene displays an ambipolar electric field effect and high carrier mobility [115]. In addition to its excellent transparent properties, graphene becomes a low-cost alternative to indium tin oxides [116]. Furthermore, graphene exhibits impressive thermal properties with a thermal conductivity ranging from 3000-5000 W m −1 K −1 [117].
Graphene's distinct physical characteristics make it potentially useful for field-effect transistors (FET), sensors, transparent conductive films, energy devices, etc., but its intrinsic gapless character still constrains its further applications. First, heteroatom-doping or chemical adsorption can effectively tune the electronic properties of graphene. N-doping can not only induce an n-type-doping effect (Figure 6a-c) [118] but can also give rise to p-type-doping with different configurations of N substitutions. For example, graphitic N induces n-type-doping, but pyridinic N induces p-type-doping [119]. In addition, Bdoping can introduce p-type transfer characteristics [120]. Chemical functional groups can also produce various doping effects. For example, the adsorption of spiropyran and DR1P molecules introduce n-type and p-type-doping to graphene, respectively [36,121].
Moreover, light can reversibly switch the molecular transformations, resulting in the controllable shift of the Dirac point of graphene ( Figure 6d) [36,121]. By seamlessly and precisely stitching the domains of graphene and h-BN (Figure 6e), the hybrid atomic layers of in-plane heterostructures can be applied for intriguing electronic applications [42,122]. Interestingly, in the proper twisted angles of bilayer graphene, the electronic band structure shows flat bands near Fermi energy, resulting in the correlated insulating states at half-filling and unconventional superconductivity with a T c of 1.7 K (Figure 6f,g) [52].
can not only induce an n-type-doping effect (Figure 6a-c) [118] but can also give rise to ptype-doping with different configurations of N substitutions. For example, graphitic N induces n-type-doping, but pyridinic N induces p-type-doping [119]. In addition, B-doping can introduce p-type transfer characteristics [120]. Chemical functional groups can also produce various doping effects. For example, the adsorption of spiropyran and DR1P molecules introduce n-type and p-type-doping to graphene, respectively [36,121]. Moreover, light can reversibly switch the molecular transformations, resulting in the controllable shift of the Dirac point of graphene (Figure 6d) [36,121]. By seamlessly and precisely stitching the domains of graphene and h-BN (Figure 6e), the hybrid atomic layers of inplane heterostructures can be applied for intriguing electronic applications [42,122]. Interestingly, in the proper twisted angles of bilayer graphene, the electronic band structure shows flat bands near Fermi energy, resulting in the correlated insulating states at halffilling and unconventional superconductivity with a Tc of 1.7 K (Figure 6f,g) [52]. Silicene. Silicene, the silicon analog of graphene, was first predicted in a theoretical study [123] and first realized experimentally via epitaxial growth on Ag(110) [20]. Unlike the way the sp 2 hybridization of carbon atoms induces a flat honeycomb structure in graphene, silicon prefers mixed sp 2 -sp 3 hybridization to form low-buckled honeycomb silicene, retaining the existence of Dirac fermions [124,125]. Considering the spin-orbit coupling (SOC) effect, silicene is predicted to have a spin-orbit gap of 1.55 meV, much larger than that of graphene [16]. Owing to its topologically nontrivial electronic structures, silicene exhibits many unique physical properties, including the quantum spin Hall (QSH) effect [16,126], giant magnetoresistance [127,128], the field-tunable bandgap [129,130], and nonlinear electro-optic effects [131]. Hence, silicene shows great potential for device applications, especially for gate-controlled topological FET [132]. Although silicene has been prepared on many substrates, the poor air stability of silicene is the major challenge, Silicene. Silicene, the silicon analog of graphene, was first predicted in a theoretical study [123] and first realized experimentally via epitaxial growth on Ag(110) [20]. Unlike the way the sp 2 hybridization of carbon atoms induces a flat honeycomb structure in graphene, silicon prefers mixed sp 2 -sp 3 hybridization to form low-buckled honeycomb silicene, retaining the existence of Dirac fermions [124,125]. Considering the spin-orbit coupling (SOC) effect, silicene is predicted to have a spin-orbit gap of 1.55 meV, much larger than that of graphene [16]. Owing to its topologically nontrivial electronic structures, silicene exhibits many unique physical properties, including the quantum spin Hall (QSH) effect [16,126], giant magnetoresistance [127,128], the field-tunable bandgap [129,130], and nonlinear electro-optic effects [131]. Hence, silicene shows great potential for device applications, especially for gate-controlled topological FET [132]. Although silicene has been prepared on many substrates, the poor air stability of silicene is the major challenge, requiring the proper encapsulation or passivation of the reactive surface for device fabrication [133,134].
Due to the limitations of the poor air stability of silicene, it is difficult to experimentally perform surface modifications on silicene. Theoretical investigations on the surface modifications of silicene focus on doping, strain, hydrogenation, intercalation, and chem-ical adsorption. Transition metal adsorption can induce various doping effects in silicene (n-type via Cu, Ag, and Au adsorption; p-type via Ir adsorption; and neutral type via Pt adsorption in Figure 7a) [135], and so can applying strain [136]. Moreover, Mn-doping can induce a ferromagnetic state in silicene, which can be transformed into an antiferromagnetic state with the application of biaxial strain (Figure 7b) [137]. Under a certain level of pressure strain, the spin-orbit bandgap of silicene will increase from 1.55 to 2.9 meV [16]. One-side semi-hydrogenation can introduce ferromagnetism to silicene, as well as make it semiconducting with a direct bandgap of 1.74 eV (Figure 7c,d) [138]. Oxygen intercalation into underlying silicene on a Ag(111) surface can effectively reduce the orbital hybridization of the top-layer silicene and Ag substrate, leading to massless Dirac fermions (Figure 7e-g) [139]. However, in K-intercalated bilayer silicene, the Dirac cones are recovered with a small bandgap of 0.27 eV [140]. Chemical adsorption (gas and organic molecules) can tune the electronic properties of silicene, which could be a better candidate for detecting gas and organic molecules compared with graphene [109,141,142]. requiring the proper encapsulation or passivation of the reactive surface for device fabrication [133,134].
Due to the limitations of the poor air stability of silicene, it is difficult to experimentally perform surface modifications on silicene. Theoretical investigations on the surface modifications of silicene focus on doping, strain, hydrogenation, intercalation, and chemical adsorption. Transition metal adsorption can induce various doping effects in silicene (n-type via Cu, Ag, and Au adsorption; p-type via Ir adsorption; and neutral type via Pt adsorption in Figure 7a) [135], and so can applying strain [136]. Moreover, Mn-doping can induce a ferromagnetic state in silicene, which can be transformed into an antiferromagnetic state with the application of biaxial strain (Figure 7b) [137]. Under a certain level of pressure strain, the spin-orbit bandgap of silicene will increase from 1.55 to 2.9 meV [16]. One-side semi-hydrogenation can introduce ferromagnetism to silicene, as well as make it semiconducting with a direct bandgap of 1.74 eV (Figure 7c,d) [138]. Oxygen intercalation into underlying silicene on a Ag(111) surface can effectively reduce the orbital hybridization of the top-layer silicene and Ag substrate, leading to massless Dirac fermions (Figure 7e-g) [139]. However, in K-intercalated bilayer silicene, the Dirac cones are recovered with a small bandgap of 0.27 eV [140]. Chemical adsorption (gas and organic molecules) can tune the electronic properties of silicene, which could be a better candidate for detecting gas and organic molecules compared with graphene [109,141,142]. Germanene. Germanene, similar to silicene, shows a low-buckled honeycomb structure, leading to a topologically nontrivial electronic structure with a spin-orbit bandgap of 23.9 meV due to a greater SOC than that of graphene and silicene [16,143,144]. These characteristics make germanene a promising candidate for applications in high-speed and low-energy-consumption devices since it has high charge-carrier mobility and exhibits QSHE [144]. Germanene has also been successfully prepared via epitaxial growth on various substrates. Germanene. Germanene, similar to silicene, shows a low-buckled honeycomb structure, leading to a topologically nontrivial electronic structure with a spin-orbit bandgap of 23.9 meV due to a greater SOC than that of graphene and silicene [16,143,144]. These characteristics make germanene a promising candidate for applications in high-speed and low-energy-consumption devices since it has high charge-carrier mobility and exhibits QSHE [144]. Germanene has also been successfully prepared via epitaxial growth on various substrates.
Doping various atoms can introduce totally different influences on the physical properties of germanene. For instance, the adsorption of alkali metal atoms makes semi-metallic germanene become metallic, with the Dirac point moving below the Fermi level and an opened small bandgap at the Dirac point, while the adsorption of halogen atoms could lead to relatively large bandgaps, ranging from 0.416 to 1.596 eV, which is promising for optoelectronic applications [145][146][147]. The adsorption of transition metal atoms (e.g., Ti, Sc, V, Cr, Mn, Fe, and Co) can induce magnetism, while nonmagnetic semiconducting states can be realized for Ni, Cu, and Zn adsorption [148,149]. The atomic structures of germanene can be controlled with supported substrate and growth conditions. Directly grown on a Ag(111) surface, two distinct phases of germanene can be observed: one shows a striped phase, a honeycomb lattice partially commensurate with the substrate; the other is a quasi-freestanding phase, a honeycomb lattice incommensurate with the substrate [150]. By epitaxially preparing it on Ag (111) thin-film grown on Ge(111) with a segregation method, the germanene shows a highly ordered long-range superstructure with two types of protrusions (hexagon and line), resulting in a (7 √ 7 × 7 √ 7)R19.1 • supercell with respect to Ag(111) (Figure 8a,b) [151]. However, the single domain (3 × 3) and multidomain ( √ 7 × √ 7)R(±19.1 • ) of germanene can exist simultaneously on an Al(111) surface (Figure 8c) [152]. On MoS 2 substrate, germanene islands can be formed at high deposition rates, whereas Ge atoms prefer to intercalate between MoS 2 layers to form Ge clusters at low deposition rates [147,153]. On a Au(111) surface, honeycomb (1 × 1) germanene with a buckled structure was identified in a ( √ 7 × √ 7) superstructure, exhibiting distinctive vibrational phonon modes and enhancing electron-phonon coupling induced by the tensile strain [154].
Doping various atoms can introduce totally different influences on the physical properties of germanene. For instance, the adsorption of alkali metal atoms makes semi-metallic germanene become metallic, with the Dirac point moving below the Fermi level and an opened small bandgap at the Dirac point, while the adsorption of halogen atoms could lead to relatively large bandgaps, ranging from 0.416 to 1.596 eV, which is promising for optoelectronic applications [145][146][147]. The adsorption of transition metal atoms (e.g., Ti, Sc, V, Cr, Mn, Fe, and Co) can induce magnetism, while nonmagnetic semiconducting states can be realized for Ni, Cu, and Zn adsorption [148,149]. The atomic structures of germanene can be controlled with supported substrate and growth conditions. Directly grown on a Ag(111) surface, two distinct phases of germanene can be observed: one shows a striped phase, a honeycomb lattice partially commensurate with the substrate; the other is a quasi-freestanding phase, a honeycomb lattice incommensurate with the substrate [150]. By epitaxially preparing it on Ag(111) thin-film grown on Ge(111) with a segregation method, the germanene shows a highly ordered long-range superstructure with two types of protrusions (hexagon and line), resulting in a (7√7 × 7√7)R19.1° supercell with respect to Ag(111) (Figure 8a,b) [151]. However, the single domain (3 × 3) and multidomain (√7 × √7)R(±19.1°) of germanene can exist simultaneously on an Al(111) surface (Figure 8c) [152]. On MoS2 substrate, germanene islands can be formed at high deposition rates, whereas Ge atoms prefer to intercalate between MoS2 layers to form Ge clusters at low deposition rates [147,153]. On a Au(111) surface, honeycomb (1 × 1) germanene with a buckled structure was identified in a (√7 × √7) superstructure, exhibiting distinctive vibrational phonon modes and enhancing electron-phonon coupling induced by the tensile strain [154].  Stanene. Analogously, monolayer Sn prefers to form a low-buckled structure due to the mixed sp 2 -sp 3 hybridization of Sn atoms [155]. Stanene has been predicted to have massless Dirac fermions and open a spin-obit bandgap of 0.1 eV at the K point with SOC [155]. The bandgap at the Dirac point can produce a conductive 1D helical edge state with opposite spin polarization, allowing for low-power spintronic applications [13,156].
Double-side-decorated stanene created by chemical functional groups appears in the most stable configuration (Figure 9a). For pristine stanene, SOC can open a bandgap of 0.1 eV at the K point; thus, stanene becomes a QSH insulator (Figure 9c). After hydorgenation or fluorination, the bandgap at the K point is substantially enlarged because of the saturation of the π orbital (Figure 9d,e). Fluorination induces a parity exchange between the occupied and unoccupied bands at the Γ point. In detail, a negative-parity Bloch state shifts downward into valence bands, leaving a positive-parity Bloch state as the conduction band minimum (Figure 9d), leading to a topologically nontrivial band structure. Nonetheless, the band inversion at the Γ point cannot be seen in hydrogenated stanene (Figure 9e). In fact, the band inversion exists for several chemical functional groups (halogen atoms and -OH) (Figure 9b). 0.1 eV at the K point; thus, stanene becomes a QSH insulator (Figure 9c). After hydorgenation or fluorination, the bandgap at the K point is substantially enlarged because of the saturation of the π orbital (Figure 9d,e). Fluorination induces a parity exchange between the occupied and unoccupied bands at the Γ point. In detail, a negative-parity Bloch state shifts downward into valence bands, leaving a positive-parity Bloch state as the conduction band minimum (Figure 9d), leading to a topologically nontrivial band structure. Nonetheless, the band inversion at the Γ point cannot be seen in hydrogenated stanene (Figure 9e). In fact, the band inversion exists for several chemical functional groups (halogen atoms and -OH) (Figure 9b).
Stanene is predicted to show superior sensing performance for small molecules. CO, O2, NO, NO2, and SO2 molecules act as charge acceptors, whereas H2O, NH3, and H2S molecules act as charge donors [157,158]. In addition, molecule adsorption can effectively tune the work function of stanene (Figure 9f) [157]. The edge shapes of stanene play a key role in the physical properties of stanene nanoribbons (NRs). Armchair stanene NRs are nonmagnetic semiconductors with tunable bandgaps by ribbon width, whereas zigzag stanene NRs present antiferromagnetic ground states with an opposite spin order between the two edges [159]. Generally, the energy gap (0.1 eV) of monolayer stanene rules out phonon-mediated superconductivity. Interestingly, doping with Ca (Li) can lead to superconductivity with a low Tc of ~1.3 K (~1.4 K), lower than the value (3.7 K) of bulk βtin [160].  Stanene is predicted to show superior sensing performance for small molecules. CO, O 2 , NO, NO 2 , and SO 2 molecules act as charge acceptors, whereas H 2 O, NH 3 , and H 2 S molecules act as charge donors [157,158]. In addition, molecule adsorption can effectively tune the work function of stanene (Figure 9f) [157]. The edge shapes of stanene play a key role in the physical properties of stanene nanoribbons (NRs). Armchair stanene NRs are nonmagnetic semiconductors with tunable bandgaps by ribbon width, whereas zigzag stanene NRs present antiferromagnetic ground states with an opposite spin order between the two edges [159]. Generally, the energy gap (0.1 eV) of monolayer stanene rules out phonon-mediated superconductivity. Interestingly, doping with Ca (Li) can lead to superconductivity with a low T c of~1.3 K (~1.4 K), lower than the value (3.7 K) of bulk β-tin [160].
Plumbene. Unlike the other four Xenes in main group IV, no Dirac point crosses linearly from the Pb p z orbit at the K point without SOC. Although turning on SOC opens a large bandgap of~400 meV, no Dirac edge state has been observed in the bandgap of plumbene [161]. In addition, resulting from the energy decrease in the s antibonding state from graphene to plumbene, the s antibonding state is lower than all p bonding and an-tibonding states at the Γ point in plumbene, totally different from graphene, silicene, germanene, and stanene [162]. Therefore, plumbene is a normal insulator with a topologically trivial character. However, through electron-doping, plumbene can become a topological insulator with a large bulk gap (~200 meV) [163]. Plumbene decorated by chemical function groups (hydrogen and halogen atoms) can transform from a normal insulator to a QSH insulator with giant bulk gaps from 1.03 to 1.34 eV [161]. Plumbene has been predicted to be magnetic with Ti-, V-, Cr-, Mn-, Fe-, and Co-doping, while Sc-and Ni-doped plumbene is nonmagnetic [164]. It is interesting that plumbene can be successfully grown using segregation on a Pd 1−x Pb x (111) alloy surface [30]. Furthermore, a c(2 × 4) structure of Pb forms on Ir(111) substrate, whereas a flat honeycomb plumbene can be formed on an Fe monolayer on Ir(111) [165].

Group V
Phosphorene. Puckered and buckled structures in phosphorene are the most common allotropic monolayer structures, corresponding to the individual atomic layers of black phosphorous and blue phosphorous crystals, respectively [166]. Puckered monolayer phosphorene is semiconducting with a direct bandgap of 1.83 eV, whereas buckled monolayer phosphorene is a semiconductor with an indirect bandgap of 2.0 eV. Phosphorene can be obtained through mechanical exfoliation, liquid phase exfoliation, electrochemical exfoliation, chemical vapor deposition, epitaxial growth, etc. [166]. To date, the semiconducting character of phosphorene has led to some experimental demonstrations in various applications, including electronics, optoelectronics, photovoltaics, supercapacitors, and catalysis [73,167].
Transition metal-doped black phosphorene possesses dilute magnetic semiconducting properties. In particular, the substitutional doping of Ti, Cr, and Mn can create a spin-polarized semiconducting state, while a half-metallic state is realized via V-and Fedoping (Figure 10a) [168]. Both Fe-doping and N-doping can significantly improve the electrocatalytic activity of black phosphorene for nitrogen reduction reactions [169,170]. For blue phosphorene, B-doping and C-doping can both improve the sensitivity of NH 3 gas molecules, and the sensitivity of CO gas molecules can be enhanced by B-doping [171]. With a monotonic increase in an external electric field, black phosphorene can transition from a normal insulator to a topological insulator and, eventually, to a metal (Figure 10c-e) [172]. In addition, an external electric field can realize reversible potassium intercalation in a blue phosphorene-Au network (Figure 10b) [173]. When axial strain is applied in the zigzag or armchair direction, the bandgap of black phosphorene will exhibit a direct-indirect-direct transition (Figure 10f,g) [51]. Moreover, a topological phase transition of black phosphorene can be realized via the application of compressive biaxial in-plane strain and perpendicular tensile strain [174].
Arsenene. Arsenene mainly has two kinds of allotropic structures, including buckled and puckered [166]. Buckled honeycomb monolayer arsenene, derived from semimetallic gray arsenic, has an indirect bandgap of 2.49 eV [24], while puckered monolayer arsenene, exfoliated from black semiconducting arsenic, possesses an indirect bandgap of 0.831 eV [75]. Both buckled arsenene and puckered arsenene have thickness-dependent bandgaps. The methods to obtain arsenene include top-down (mechanical exfoliation and liquid phase exfoliation) and bottom-up strategies (molecular beam epitaxy, chemical vapor deposition, physical vapor deposition, etc.) [175]. For epitaxial growth, monolayer buckled arsenene was successfully grown on a Ag(111) substrate [176], whereas monolayer armchair arsenene nanochains have been formed on a Au(111) surface [32]. Twodimensional arsenene has been theoretically predicted to exhibit various physical properties, such as an indirect-to-direct bandgap transition, a semimetal-to-semiconductor transition, superconductivity, and a QSH effect, deserving many research efforts [24,177,178]. Recently, few-layer black arsenene was proved to exhibit a particle-hole asymmetric Rashba valley and exotic quantum Hall states due to synergetic effects between the spin-orbit interaction and the Stark effect [179]. Arsenene. Arsenene mainly has two kinds of allotropic structures, including buckled and puckered [166]. Buckled honeycomb monolayer arsenene, derived from semi-metallic gray arsenic, has an indirect bandgap of 2.49 eV [24], while puckered monolayer arsenene, exfoliated from black semiconducting arsenic, possesses an indirect bandgap of 0.831 eV [75]. Both buckled arsenene and puckered arsenene have thickness-dependent bandgaps. The methods to obtain arsenene include top-down (mechanical exfoliation and liquid phase exfoliation) and bottom-up strategies (molecular beam epitaxy, chemical vapor deposition, physical vapor deposition, etc.) [175]. For epitaxial growth, monolayer buckled arsenene was successfully grown on a Ag(111) substrate [176], whereas monolayer armchair arsenene nanochains have been formed on a Au(111) surface [32]. Two-dimensional arsenene has been theoretically predicted to exhibit various physical properties, such as an indirect-to-direct bandgap transition, a semimetal-to-semiconductor transition, superconductivity, and a QSH effect, deserving many research efforts [24,177,178]. Recently, few-layer black arsenene was proved to exhibit a particle-hole asymmetric Rashba valley and exotic quantum Hall states due to synergetic effects between the spin-orbit interaction and the Stark effect [179].
Substrate temperatures can modify the formation of arsenic nanostructures. On Au(111) substrate, the arsenic monolayer formed by 0D tetrahedral As4 clusters can transform into a monolayer formed by 1D armchair arsenene nanochains (Figure 11a) [32]. The application of biaxial tensile strain can effectively tune the band structures of arsenene. Under low tensile strain, arsenene can transform from an indirect to a direct bandgap (Figure 11b,c) [24]. By further enlarging the tensile strain, the direct bandgap will gradually disappear, causing band inversion at the Γ point (Figure 11d,e) [177]. The consideration of SOC can open a spin-orbit bandgap (~131 meV) under a tensile strain of 18.4% (Figure 11f), indicating a QSH effect in arsenene [177]. Intriguingly, under proper biaxial tensile strain and electron-doping, arsenene can be superconducting, with a Tc of 30.8 K [178]. Indeed, 3D-transition-metal-doping can strongly tailor the electronic and magnetic properties of arsenene. Ti-, V-, Cr-, Mn-, and Fe-doping can induce magnetic states for Substrate temperatures can modify the formation of arsenic nanostructures. On Au (111) substrate, the arsenic monolayer formed by 0D tetrahedral As 4 clusters can transform into a monolayer formed by 1D armchair arsenene nanochains (Figure 11a) [32]. The application of biaxial tensile strain can effectively tune the band structures of arsenene. Under low tensile strain, arsenene can transform from an indirect to a direct bandgap (Figure 11b,c) [24]. By further enlarging the tensile strain, the direct bandgap will gradually disappear, causing band inversion at the Γ point (Figure 11d,e) [177]. The consideration of SOC can open a spin-orbit bandgap (~131 meV) under a tensile strain of 18.4% (Figure 11f), indicating a QSH effect in arsenene [177]. Intriguingly, under proper biaxial tensile strain and electrondoping, arsenene can be superconducting, with a T c of 30.8 K [178]. Indeed, 3D-transitionmetal-doping can strongly tailor the electronic and magnetic properties of arsenene. Ti-, V-, Cr-, Mn-, and Fe-doping can induce magnetic states for arsenene [180]. Meanwhile, Ti-and Mn-doping leads to a half-metallic state, while V-, Cr-, and Fe-doping results in a spin-polarized semiconducting state [180]. In addition, doping can further modify chemical properties, making arsenene potentially useful for hydrogen evolution and oxygen evolution reactions (Figure 11g) [181].
Doping with 3d transition metal atoms for antimonene can lead to significant changes in the bandgap and the magnetic moment [187]. For Cr-doped β-antimonene, a spin-polarized semiconducting state was predicted. For Ti-, Mn-, and V-doped β-antimonene, half-metallic behavior was calculated. Similarly, a Cr-doped bismuthene structure leads to a spin-polarized semiconducting state, while V-doped bismuthene can produce a magnetic metal character, and Mn-and Fe-replacing systems result in half-metal features [188]. Additionally, V-doped systems exhibit ferromagnetism (FM) when two V atoms are far apart, but Cr-, Mn-, and Fe-doped bismuthene exhibits anti-ferromagnetism (AFM) when two impurity atoms are close together or far apart [188]. Bivacancy-doping in βantimonene can reduce the bandgap of pristine β-antimonene, but monovacancy-doped β-antimonene exhibits a metallic character [189]. Electron-doping and Ca-intercalation can transform bilayer β-antimonene from a semimetal into a superconductor [190]. Moreover, the physisorption of the organic molecules tetrathiafulvalene and Antimonene and bismuthene. The most stable structures of antimonene and bismuthene are α-form (puckered) and β-form (buckled), which are the monolayers of black and gray bulk allotropes, respectively. Monolayer β-Sb and β-Bi possess indirect bandgaps of 2.28 and 0.99 eV, respectively, whereas the direct bandgaps of monolayer α-Sb and α-Bi are 1.18 and 0.36 eV, respectively [74,182,183]. Notably, the calculated bandgaps may vary depending on the methods used. Both antimonene and bismuthene exhibit tunable bandgaps, high carrier mobility, high stability, and in-plane anisotropy, providing a basis for multifunctional applications in electronics, optoelectronics, sensors, batteries, etc. [184][185][186].
Doping with 3d transition metal atoms for antimonene can lead to significant changes in the bandgap and the magnetic moment [187]. For Cr-doped β-antimonene, a spinpolarized semiconducting state was predicted. For Ti-, Mn-, and V-doped β-antimonene, half-metallic behavior was calculated. Similarly, a Cr-doped bismuthene structure leads to a spin-polarized semiconducting state, while V-doped bismuthene can produce a magnetic metal character, and Mn-and Fe-replacing systems result in half-metal features [188]. Additionally, V-doped systems exhibit ferromagnetism (FM) when two V atoms are far apart, but Cr-, Mn-, and Fe-doped bismuthene exhibits anti-ferromagnetism (AFM) when two impurity atoms are close together or far apart [188]. Bivacancy-doping in β-antimonene can reduce the bandgap of pristine β-antimonene, but monovacancy-doped β-antimonene exhibits a metallic character [189]. Electron-doping and Ca-intercalation can transform bilayer β-antimonene from a semimetal into a superconductor [190]. Moreover, the physisorption of the organic molecules tetrathiafulvalene and tetracyanoquinodimethane can induce n-type-and p-type-doping for antimonene, respectively [191]. Under a monotonic increase in biaxial tensile strain, β-antimonene and β-bismuthene undergo indirect-todirect bandgap and semiconducting to semi-metallic transitions and even topological phase transitions [24,192,193].

Group VI
Selenene and tellurene. Selenene has been predicted to have three allotropic structures, including a 1T-MoS 2 -like structure (t-Se or α-Se), a tiled helical-chain structure (c-Se), and s square structure (s-Se) [194]. Both t-Se and c-Se are semiconductors with indirect bandgaps of 0.71 and 1.74 eV, respectively, while s-Se is semi-metallic. The formation energy of c-Se is the lowest, indicating that c-Se is the most stable phase. In addition, a ring structure of selenene (r-Se) is proposed [58]. However, theoretical investigations predict the existence of three phases of tellurene, including α-, β-, and γ-phases, possessing 1T-MoS 2 -like, rectangle, and 2H-MoS 2 -like structures, respectively [27]. It was found that αand β-Te show semiconducting characteristics with indirect bandgaps of 0.76 and 1.17 eV, respectively, whereas γ-Te is a metal. It is interesting that the band structures of square selenene and tellurene (s-Se and s-Te) show Dirac-cone-like dispersions. A large bandgap (~0.1 eV) opened by SOC makes them become topological insulators and host nontrivial edge states [58]. Therefore, square selenene and tellurene become promising candidates for spintronic applications. The tensile strain can monotonically decrease the bandgaps of α-Se and α-Te [195,196]. Meanwhile, for 2D square tellurene under a strain effect, the system displays three structural phases, buckled square, buckled rectangle, and planar square phases, which exhibit extraordinary topological properties [196]. In particular, the buckled rectangle tellurene can act as a QSH insulator with a bandgap of 0.24 eV.

Applications of 2D Xenes
As mentioned above, 2D Xenes possess various physical and chemical properties, such as flexibility, layer-dependent semiconducting, high carrier mobility, molecule and light sensitivity, topologically nontrivial band structures, etc.
In order to utilize 2D Xenes more effectively, surface modifications become particularly important ( Figure 12) to tune the properties of 2D Xenes. For instance, the FET of acene-type graphene nanoribbons exhibits excellent semiconductor characteristics with an on/off ratio of 88 [197]. To enhance the mid-infrared (MIR) absorption of graphene, the localized surface plasmon resonance of B-doped Si quantum dots (QDs) results in a QD/graphene hybrid photodetector with ultrahigh responsivity, gain, and specific detectivity in the UV-to-MIR region [34]. A 2D bismuthene/Si(111) heterostructure exhibits excellent photodetection performance in the Vis-MIR region due to the promoted generation and transportation of charge carriers in the heterojunction [198]. Solution-exfoliated black phosphorene flakes, as an electron transport layer, can enhance the performance of organic solar cells [199]. In addition, layered black phosphorene exhibits the selective detection of methanol [200]. The thermoelectric power (S) in black phosphorene can be effectively controlled with ion-gating. In the hole-depleted state, the S of black phosphorene can reach +510 µV/K at 210 K, much higher than the bulk single crystal value of +340 µV/K at 300 K [201]. Under the proper electron-doping and biaxial tensile strain, buckled arsenene shows superconductivity with a high T c of 30.8 K [178]. Iodine-decorated stanene exhibits a topologically nontrivial band structure with a larger gap of~320 meV than that of pristine stanene (~100 meV) [155]. Graphene/Pt(111) surfaces can cause CO adsorption/desorption and CO oxidation surface reactions [202]. MoS 2 /graphene hybrids decorated by CdS nanocrystals can act as high-performance photocatalysts for H 2 evolution under visible light irradiation [203]. Moreover, 2D Xenes are also regarded as promising agents for biomedical applications [204]. For example, an ultrathin bismuthene can act as a sensing platform to detect microRNA with a detection limit of 60 PM [205]. Polyethylenecoated antimonene quantum dots can be used as photothermal agents with a high photothermal conversion efficacy of 45.5% for photothermal therapy in cancer theranostics [206]. Overall, thanks to surface modifications, 2D Xenes show great potential for applications in plenty of fields. photothermal conversion efficacy of 45.5% for photothermal therapy in cancer theranostics [206]. Overall, thanks to surface modifications, 2D Xenes show great potential for applications in plenty of fields. Figure 12. Applications of 2D Xenes with surface modifications. Transistors: Reprinted with permission from Ref. [207]. Copyright 2015, American Chemical Society. Optoelectronics: Reprinted with permission from Ref. [36]. Copyright 2012, American Chemical Society. Sensors: Reprinted with permission from Ref. [200]. Copyright 2015, American Chemical Society. Photovoltaics: Reprinted with permission from Ref. [199]. Copyright 2016, Wiley. QSH effect: Reprinted with permission from Ref. [208]. Copyright 2018, American Association for the Advancement of Science. Superconductivity: Reprinted with permission from Ref. [52]. Copyright 2018, Springer Nature Ltd. Thermoelectric: Reprinted with permission from Ref. [201]. Copyright 2016, American Chemical Society. Catalysis: Reprinted with permission from Ref. [202]. Copyright 2014, National Academy of Science.

Conclusions
In total, except graphene, 15 different elemental 2D materials of the main group elements have been experimentally created or theoretically predicted to date. In fact, 14 Xenes, including borophene, graphene, silicene, phosphorene, gallenene, germanene, arsenene, selenene, stanene, antimonene, tellurene, thallene, plumbene, and bismuthene, have been successfully grown on proper substrates using epitaxial methods. Although a lot of research, engineering, and development related to 2D Xenes-most of which were investigated theoretically-has been reported in recent years, experimental studies are still desperately required for the development of synthesis strategies and novel applications. Before realizing the surface modifications of 2D Xenes for the applications of interest, three significant challenges need to be overcome: (i) The synthesis strategies of 2D Xenes must not only ensure reliably large-scale production but also be tailored for the requirements of the application. For example, the  [207]. Copyright 2015, American Chemical Society. Optoelectronics: Reprinted with permission from Ref. [36]. Copyright 2012, American Chemical Society. Sensors: Reprinted with permission from Ref. [200]. Copyright 2015, American Chemical Society. Photovoltaics: Reprinted with permission from Ref. [199]. Copyright 2016, Wiley. QSH effect: Reprinted with permission from Ref. [208]. Copyright 2018, American Association for the Advancement of Science. Superconductivity: Reprinted with permission from Ref. [52]. Copyright 2018, Springer Nature Ltd. Thermoelectric: Reprinted with permission from Ref. [201]. Copyright 2016, American Chemical Society. Catalysis: Reprinted with permission from Ref. [202]. Copyright 2014, National Academy of Science.

Conclusions
In total, except graphene, 15 different elemental 2D materials of the main group elements have been experimentally created or theoretically predicted to date. In fact, 14 Xenes, including borophene, graphene, silicene, phosphorene, gallenene, germanene, arsenene, selenene, stanene, antimonene, tellurene, thallene, plumbene, and bismuthene, have been successfully grown on proper substrates using epitaxial methods. Although a lot of research, engineering, and development related to 2D Xenes-most of which were investigated theoretically-has been reported in recent years, experimental studies are still desperately required for the development of synthesis strategies and novel applications. Before realizing the surface modifications of 2D Xenes for the applications of interest, three significant challenges need to be overcome: (i) The synthesis strategies of 2D Xenes must not only ensure reliably large-scale production but also be tailored for the requirements of the application. For example, the applications of electronics and batteries demand low costs and scalable techniques, while plasmonics and spintronics require high fidelity and reproducible techniques. Hence, reliable synthesis approaches play a crucial role in functional design and practical applications.
(ii) The strategies to enhance the environmental stability and mitigate the degradation of 2D Xenes must be taken into consideration for certain applications. For instance, an optoelectronic device based on black phosphorene must be concerned with ambient stability, requiring appropriate encapsulation or the passivation of the surface.
(iii) Many of the predicted and fascinating properties of 2D Xenes require further efforts to find strategies for their implementation, which may offer opportunities to discover revolutionary technologies. In addition, the exciting and novel properties necessitate not just "one-off" research, but also statistical evaluation to ascertain their viability and accessibility on a commercial scale.
Despite the challenges ahead, the exceptional properties of 2D Xenes will significantly impact future applications in various fields. It is hoped that this review will inspire more exhilarating discoveries and applications in the growing family of 2D Xenes.