Transition Metal Dichalcogenides (TMDC) have a structure in which one transition metal layer is sandwiched between chalcogen atomic layers. Their physical properties vary and include metals, semimetals, semiconductors, and insulators, depending on their composition and crystal structure. One characteristic feature is the layer number dependence, in which the physical properties change depending on the number of layers. In MoSe₂, which I am focusing on, the electronic transition changes from an indirect transition to a direct transition when it becomes a monolayer; the monolayer structure of MoSe₂ and the band gap transition are shown in Figures 1(a) and 1(b).

Figure 1. (a) Single layer structure of MoSe₂ (b) Band gap transition

Currently, graphene is the mainstay of two-dimensional semiconductors. However, graphene cannot be applied to switching devices as it is because it does not have a band gap. In contrast, TMDC have a finite band gap, depending on the type, and can be used as-is in devices.

Spatial Inversion Symmetry Breaking

TMDC have various crystal structures, and the stable structure differs depending on the composition. The crystal structures are shown in Figure 2.

Figure 2. Various crystal structures of TMDC

In MoSe₂, which I am focusing on, 2H is the most stable structure. 2H cannot be rotated 180° in the in-plane direction back to its original crystal structure, so it is a "space inversion symmetry broken" system. Figure 3 shows this situation.

Figure 3. Broken spatial inversion symmetry system

Phase Transition

Some types of TMDC undergo a phase transition under certain conditions: in MoTe2, a photo-induced phase transition from the Td phase to the 1T' phase is reported to occur under light irradiation at a low temperature of 250K. Since electrical and other properties depend on the crystal structure, the phase transition can change electrical and other properties.

Valleytronics

Transition metals, which constitute TMDC, have strong spin-orbit interactions. In TMDC where the spatial inversion symmetry is broken, this spin-orbit interaction splits the conduction band and valence band, which were previously degenerate into one. Figure 4 shows the valley in the K-space of MoSe₂.

Figure 4. Valley of MoSe₂ in K-space

Since electrons with different spins enter the red and blue valleys in Figure 4, each valley has a different momentum. This relationship is called valley degrees of freedom, and the information processing technology that uses valley degrees of freedom to represent 0s and 1s is called "valleytronics.

　In the case of light, each valley can be accessed by circularly polarizing light, and since the electrons in the two valleys have chirality, for example, left circularly polarized light can excite only the electrons in the red valley and right circularly polarized light can excite only the electrons in the blue valley. This state is called valley polarization. The valley polarization is shown in Figure 5.

Figure 5. Valley Polarization

σ^- :left circular polarization and σ^+: right circular polarization

Valleytronics has been actively studied for applications in the field of quantum computing. A superposition state of valleys is formed by electrons excited with arbitrary polarization, and the direction of polarization of the light emitted upon relaxation can be observed to enable its use as a qubit. The development of this technology is expected to lead to the proposal of new devices.