Table 10 Number of features considered in studies devoted to the prediction of the properties of MXenes by machine learning and, when available, the ones that were considered more important

Rajan et al.⁵⁴

Metal−semiconductor identification

47 | 11

Volume per atom; Average elemental boiling point; Standard deviation of group number in the periodic table; Standard deviation of average elemental boiling point; Formation Enthalpy.

Rajan et al.⁵⁴

Band-gap estimation

47 | 15

Molar volume; Mean elemental boiling point; Standard deviation of mean elemental group number in periodic table; Standard deviation of mean elemental melting point; Formation Energy.

Frey et al.²⁰⁹

Synthesizability

>80 | >80

M-X bond length; Cohesive energy; Bader charge of X; Formation energy; Mass per atom.

Mishra et al.²¹²

Valence band

56 | 7

Average vacuum potential; Conduction band minima; MXene phase; Vacuum potential for the lower surface; Band gap.

Marchwiany et al.²¹⁹

Cytotoxicity

17 | 17

MxOy presence; Li on the surface; Surface modification; Delaminating agent; Lateral size; Cl on the surface.

Venturi et al.²¹⁴

Mechanical strength, band gap and formation energy

NA | NA

Atomic and bond feature vectors were used. Authors did not report the most important features, but selected descriptors as group or period numbers, electronegativity, number of valence electrons or nature of bonded atoms.

Wang et al.²²²

HER

41 | 5

Distance between oxygen and metal atoms in the outermost layer; Distance between the nearest neighbor oxygen atoms; Ionization energy difference; Average value of affinity energy; Number of valence electrons of X.

Zheng et al.²²¹

HER

24 | 20

Electronegativity of T; Atomic mass of T; M-X bond length; Bader charge of T; Bader charge of X.

He et al.²¹⁰

Stability

25 | 16

The most important descriptor, (χ(M)-R_vdW(M))³, where R_vdW(M) and χ(M) correspond to the atomic van der Waals radius and the electronegativity, respectively, of the metal atom (M), was identified via the symbolic regression of the material stability. Other descriptors were: number of electrons in the d orbital of M; electronegativity of M; second ionization energy of M; first ionization energy of M.

Li et al.²²⁶

Energy storage

19 | 16

Categorical descriptors indicating the presence of Mg, K, Ta, Hf, and O.

Song et al.²¹⁸

Saturation magnetization

9 | 9

Number of unpaired electrons of M; Enthalpy of atomization; Atomic number; Polarizability; Density.

Tian et al.²¹⁶

Tensile stiffness

19 | 3

Thickness; Stiffness of M–X bonds; Surface terminations.

Abraham et al.²²⁹

CO₂ activation

18 | 18

d-band center; Surface metal electronegativity; Number of valence electrons of M.

Boonpalit et al.²³⁰

CO sensing

NA | NA

Atomic and bond feature vectors were used.

Chen et al.²²⁴

ORR and OER

NA | 12

Charge transfer; d-band center of the active transition metal.

Jiao et al.²³¹

C−N coupling

13 | 13

Period number; Atomic first ionization-energy; Atomic Pauling electronegativity; Number of valence electrons of M; Atomic weight; Atomic radius of M; Thickness.

Liang et al.²²³

HER

49 | 8

Molar volume of surface element; Atomic radius of single atom; Boiling point of surface element; Boiling point of single atom; Electron affinities of surface element.

Liang et al.²²³

HER

49 | 8

Third ionization potential of surface group element; Ionic radius of surface group element; Atomic mass of surface element; Atomic radius of surface element; Molar volume of surface element

Ma et al.²²⁵

ORR and OER

13 | 13

Hydroxide adsorption energy; Binding energy of Pt on substrate; d-band center of Pt; Oxide formation enthalpy; Lattice constant.

Roy et al.²¹³

Work function

15 | 15

Electronegativity of T atoms; Ionization potential of T atoms; Electron affinity of T atoms; Orbital radius of the outermost orbitals of the T atoms; Orbital radius of the outermost orbitals of the M atoms.