The simulation was carried out using PWSCF package of Quantum Espresso [3] simulation tool. After geometry optimization and relaxation of different stacking configurations, self-consistent field (SCF) calculation was done to determine electronic configuration for the relaxed atomic structure followed by bandstructure calculation using the same package. We used normconserving pseudopotenial functions in our simulation framework and 24x24x1 Monkhorst-Pack k-point grids for sampling in the brillouin zone. After extracted of geometrically relaxed structure, biaxial strain was applied to the lattice and the effects on band structure and carrier effective mass were observed. The carrier effective mass at the brillouin zone symmetry points was calculated by simple parabolic fitting to the extracted ab-initio band structure. At first, we performed structural geometry optimization for ABA and ACA stacking configurations. We extracted the optimized values of lattice constant and interlayer separation. Our extracted results showed good agreement with the values reported in [1].
Fig 2. Shows the effect of variation of electron effective mass at K point in the brillouin zone for the simulated lattice structure under strain. The figure reveals lowering of effective mass at K points in the brillouin zone as we move from compressive to tensile strain regime. It is also seen that, for stacking configuration AAA in all three lattice structures, the simulation reveals a slightly lower carrier effective mass at K point for high compressive strain. However, as we move towards tensile strain regime, the stacking configurations reveal almost similar effective masses at K point. The gradual lowering of carrier effective mass at K point in the Brillouin zone is consistent with the results reported in literature[4].
Fig.3 shows the effect of strain on bandgap of different stacking configurations of multilayer TMDC materials as extracted from simulation. In all three material structures used in this study, we see a gradual lowering of bandgap with strain and therefore a gradual shift from semiconducting to metallic properties which is consistent with the results reported in [5] for bilayer TMDC heterostructures. Of the three structures used in this study, MoS2/WSe2/MoS2 shows the lowest bandgap and therefore the quickest transition to metallic state under tesile strain as seen from fig. 3(c). As observed from our simulation results, at unstrained condition, MoS2/WSe2/MoS2 trilayer showed a direct bandgap for all three stacking configurations. On the other hand, unstrained MoS2/MoSe2/MoS2 and MoS2/WS2/MoS2 trilayers showed indirect bandgap for all three stackings. The AAA stacking configuration is observed to be showing the lowest bandgap when compressive strain is being applied to the lattice.
Some results extracted from our simulation for unstrained lattice configuration is listed below:
Lattice | Stacking | Lattice Constant(A0) | Bandgap, Eg(eV) | Effective mass,mK(m0) |
AAA | 3.239 | 0.707 | 0.5035 | |
MoS2/MoSe2/MoS2 | ABA | 3.24 | 0.729 | 0.5129 |
ACA | 3.238 | 0.752 | 0.5142 | |
AAA | 3.186 | 1.293 | 0.5753 | |
MoS2/WS2/MoS2 | ABA | 3.183 | 1.074 | 0.5886 |
ACA | 3.186 | 1.147 | 0.5877 | |
AAA | 3.241 | 0.5035 | 0.5035 | |
MoS2/WSe2/MoS2 | ABA | 3.237 | 0.542 | 0.5084 |
ACA | 3.243 | 0.562 | 0.5097 |
In this work, we have provided a detailed simulation results on the effects of intercalation on electronic properties of TMDC materials. The findings of this study would be helpful in modeilng electronic devices using these material structures.
References:
[1] Ning Lu, et al., Nanoscale 6.9 (2014): 4566-4571.
[2] Sheneve Z Butler., et al., ACS nano 7.4 (2013): 2898-2926.
[3] Paolo Giannozzi, et. al., Journal of Physics: Condensed Matter 21.39 (2009): 395502.
[4] Blanca Biel, et. al., Microelectronic Engineering 147 (2015): 302-305