返回文章

Figure 1.  Simulation setup and thermal bath configurations. (a) Schematic of the Si–Ge heterojunction NEMD simulation system showing key parameters. (b) Representative temperature profile illustrating the interfacial temperature drop ΔT_int used for thermal resistance calculation. (c) Schematic diagrams of the three thermal bath configurations examined: defect-free thermal bath, thermal bath with 20% atomic vacancies, and thermal bath with 20% volume fraction of triangular prism-shaped voids.

Figure 2.  Dependence of Si/Ge ITR on system dimensions for the three indicated thermal bath types. (a) ITR versus L_sys with L_bath = 20 nm and τ = 1 ps. (b) Variation of ITR with 1/L_sys. (c) ITR versus L_bath with L_sys = 100 nm and τ = 1 ps. (d) Variation of ITR with 1/L_bath. For comparison, ITR calculated using the more realistic Stillinger–Weber potential is shown for the pristine structure.

Figure 3.  Effects of thermostat coupling. (a) Thermal conductance versus coupling constant τ for all bath types. (b) Averaged temperature difference between thermal baths as a function of coupling constant τ, with insets showing temperature profiles at τ = 0.1 ps, 2 ps, and 10 ps for the pristine bath system.

Figure 4.  Spectral thermal conductance versus phonon frequency for varying L_sys. (a) Pristine thermostat, (b) thermostat with 20% atomic vacancies, and (c) thermostat with triangular prism-shaped voids. Shaded regions indicate the relative increase in conductance between L_sys = 10 nm and 100 nm.

Figure 5.  Spectral thermal conductance versus phonon frequency for varying L_bath. (a) Pristine thermostat, (b) thermostat with 20% atomic vacancies, and (c) thermostat with triangular prism-shaped voids. Shaded regions indicate the relative increase in conductance between L_bath = 3 nm and L_bath = 15 nm.