Ab initio synthesis of single-layer III-V materials

The discovery of a novel material requires the identification of the material's composition as well as of suitable synthesis conditions. We present a data-mining approach to identify suitable substrates for the growth of two-dimensional materials and apply the method to the recently predicted two-dimensional III-V compounds. We identify several lattice-matched substrates for their epitaxial growth, stabilization, and functionalization. Density-functional calculations show that these substrates sufficiently reduce the formation energies of the metastable two-dimensional materials to make them thermodynamically stable. We show that chemical interactions of the two-dimensional materials with the substrates shift the Fermi level of these materials, resulting in doping. The large adsorption energies and strong doping indicate that these metals should provide good electrical contact to enable transport measurements and electronic applications.

Figure 1

Lattice mismatch (a) for the 2D tetragonal III-V materials and the (100) surfaces of transition metals and (b) for the 2D hexagonal III-V materials and the (0001) surfaces of hcp transition and rare-earth metals. The values for the lattice mismatches are provided in the Supplemental Material [29].

Figure 2

Energy cost $E_{\mathrm{strain}}$ for straining the 2D III-V materials with respect to the ground-state structure. The energy cost to strain these materials by $\pm 4\%$ is less than 150 meV/atom.

Figure 3

(a), (b) show the schematic representation (top view) of the different possible adsorption sites on Ho(0001) and Pd(100) surfaces, respectively. The large spheres represent the metal atoms in the topmost and the smaller spheres represent the atoms in second to topmost layer. Sites on each metal surface are marked by yellow spheres.

Figure 4

(a) Top and side views of isolated 2D tetragonal AlAs and adsorbed on Pt(100). (b) Top and side views of isolated 2D hexagonal InN and adsorbed on Ho(0001) surface. The formation energies are shown for various combinations of substrates and 2D materials with lattice mismatches up to 4% for (c) the tetragonal and (d) the hexagonal structures. Formation energies of the 2D III-V materials relative to their 3D bulk phases for the isolated unstrained 2D materials $E_{\mathrm{vac}}^{\mathrm{f}}$ (red), the strained 2D materials absorbed on the substrates as calculated using the PBE functional neglecting van der Waals interactions $E_{\mathrm{ads}}^{\mathrm{f}},\mathrm{PBE}$ (black symbols), and as calculated using the vdW–DF functional that includes van der Waals interactions $E_{\mathrm{ads}}^{\mathrm{f}},\mathrm{vdW}$ (blue symbols). The shaded region shows the regime where 2D materials are stable on these substrates. Electrons transferred from substrate surface atoms to the 2D materials, a measure of the doping of the 2D materials, for (e) tetragonal 2D materials adsorbed on fcc transition metals, and (f) hexagonal 2D materials on hcp transition and rare-earth metals. Red bars denote electrons lost from the substrate surface atoms, filled blue bars show electrons gained by the substrate for semiconducting hexagonal and tetragonal 2D materials, and open blue bars show electrons gained by substrate surface atoms for metallic tetragonal 2D materials.

Figure 5

Charge densities distribution for (a) isolated tetragonal AlAs, (b) tetragonal AlAs adsorbed on Pt(100), (d) isolated hexagonal InN, and (e) hexagonal InN adsorbed on Ho(0001). The cross section for AlAs is taken along the $\left(100\right)$ plane passing through the As atoms closest to Pd and for InN it is taken along the $\left(\overline{1}10\right)$ plane. The charge density is in units of electrons/Å${}^3$. Plots of the plane-averaged electron density difference $\Delta \rho$ between the interacting and isolated 2D material and substrate pair along the direction perpendicular to the interface for the fixed geometry of the adsorbed case are shown in (c) for AlAs on Pt and (f) for InN on Ho. The red and blue colors indicate electron accumulation and depletion, respectively.

Figure 6

(a) Strain effect on the band gap of 2D tetragonal AlN, AlP, hexagonal GaSb, and InN. Positive strains refer to biaxial tensile and negative to biaxial compressive strains. (b) Shifted density of states of strained 2D III-V material in vacuum (black dotted line), and unshifted density of states of 2D III-V material adsorbed on substrate (red solid line) for hexagonal InN on Ho(0001) and (d) for tetragonal AlN on Ni(100). The density of states has the units of states per eV per unit cell of 2D material. The shaded gray region corresponds to the conduction and valence band regions of the strained 2D material. (c) Shift in Fermi level of the 2D material $\Delta E_{\mathrm{Fermi}}$ versus the work function difference between the metals and 2D materials ${\varphi }_{\mathrm{M}}-{\varphi }_{2\mathrm{D}}$. Square symbols indicate the 2D tetragonal materials and diamonds the hexagonal ones.

Figure 7

Electronic band structure projected on group-V atoms of (a) InN, (b) InN on Ho(0001), (d) AlN, (e) AlN on Ni(100), (g) AlAs, and (h) AlAs on Pt(100). Band structure projected on group-III atoms of (c) InN on Ho(0001), (f) AlN on Ni(100), and (i) AlAs on Pt(100). The valence band maximum is set to zero. The symbol sizes and colors denote the weights of the specific site.

Figure 8

Work function of 2D III-V hexagonal materials, diamond symbols, and 2D III-V tetragonal materials, square symbols compared to their Mulliken electronegativities.