Pressure-induced gap closing and metallization of $\mathrm{Mo}{\mathrm{Se}}_2$ and $\mathrm{Mo}{\mathrm{Te}}_2$

Layered molybdenum dichalchogenides are semiconductors whose gap is controlled by delicate interlayer interactions. The gap tends to drop together with the interlayer distance, suggesting collapse and metallization under pressure. We predict, based on first-principles calculations, that layered semiconductors ${2\mathrm{H}}_c$-${\mathrm{MoSe}}_2$ and ${2\mathrm{H}}_c$-${\mathrm{MoTe}}_2$ should undergo metallization at pressures between 28 and 40 GPa (${\mathrm{MoSe}}_2$) and 13 and 19 GPa (${\mathrm{MoTe}}_2$). Unlike ${\mathrm{MoS}}_2$ where a ${2\mathrm{H}}_c$ $\to$ ${2\mathrm{H}}_a$ layer-sliding transition is known to take place, these two materials appear to preserve the original ${2\mathrm{H}}_c$ layered structure at least up to 100 GPa and to increasingly resist lubric layer sliding under pressure. Similar to metallized ${\mathrm{MoS}}_2$, they are predicted to exhibit a low density of states at the Fermi level, and presumably very modest superconducting temperatures, if any. We also study the $\beta$-${\mathrm{MoTe}}_2$ structure, metastable with a higher enthalpy than ${2\mathrm{H}}_c$-${\mathrm{MoTe}}_2$. Despite its ready semimetallic and (weakly) superconducting character already at zero pressure, metallicity is not expected to increase dramatically with pressure.

Figure 1

Structures of (a) ${2\mathrm{H}}_c$-${\mathrm{MoSe}}_2$ and (b) ${2\mathrm{H}}_a$-${\mathrm{MoSe}}_2$. (c) Brillouin zone for both structures [18].

Figure 2

Pressure dependence of the calculated lattice parameters $a$ (upper panel) and $c$ (lower panel) of ${2\mathrm{H}}_c$-${\mathrm{MoSe}}_2$, together with experimental data from Ref. [9].

Figure 3

Pressure dependence of the enthalpy difference between the ${2\mathrm{H}}_a$ and ${2\mathrm{H}}_c$ structures of ${\mathrm{MoSe}}_2$ and ${\mathrm{MoTe}}_2$, respectively. Note the increase with pressure, indicating increased stability of ${2\mathrm{H}}_c$ in both materials, unlike ${\mathrm{MoS}}_2$ where ${2\mathrm{H}}_a$ prevails at high pressure [11].

Figure 4

Band structure of ${2\mathrm{H}}_c$-${\mathrm{MoSe}}_2$ at $p=10$ GPa (upper panel) and $p=30$ GPa (lower panel). Red line denotes the Fermi level.

Figure 5

Pressure dependence of the band gap of ${2\mathrm{H}}_c$-${\mathrm{MoSe}}_2$ and ${2\mathrm{H}}_c$-${\mathrm{MoTe}}_2$.

Figure 6

Density of states per unit cell of ${2\mathrm{H}}_c$-${\mathrm{MoSe}}_2$ at $p=10$ and $p=30$ GPa. Red line denotes the Fermi level.

Figure 7

Pressure dependence of the calculated lattice parameters $a$ (upper panel) and $c$ (lower panel) of ${2\mathrm{H}}_c$-${\mathrm{MoTe}}_2$.

Figure 8

Band structure of ${2\mathrm{H}}_c$-${\mathrm{MoTe}}_2$ at $p=10$ GPa (upper panel) and $p=13$ GPa (lower panel). Red line denotes the Fermi level.

Figure 9

Density of states per unit cell of ${2\mathrm{H}}_c$-${\mathrm{MoTe}}_2$ at $p=10$ and $p=30$ GPa. Red line denotes the Fermi level.

Figure 10

Structure of (a) the $\beta$ form of ${\mathrm{MoTe}}_2$ and (b) its Brillouin zone.

Figure 11

Band structure of $\beta$-${\mathrm{MoTe}}_2$ at $p=40$ GPa (upper panel) and density of states (per unit cell) at $p=20$ and $p=40$ GPa and for the experimental unit cell at normal conditions (lower panel). In the band structure graph the red line denotes the Fermi level. In the DOS graph, the Fermi level is at zero.