Modulation of interface modes for resonance-induced enhancement of the interfacial thermal conductance in pillar-based Si/Ge nanowires

The interfacial thermal conductance (ITC) plays a crucial role in nanoscale heat transfer, and its enhancement is of great interest for various applications. In this study, we explore the influence of resonance on the interfacial modes in pillar-based Si/Ge nanowires through nonequilibrium molecular dynamics simulations, employing both empirical and machine-learning potentials. Our results reveal a significant enhancement in the ITC by introducing pillars in the nanowire structure. The resonance-induced enhancement of the matching degree of the phonon density of states together with the calculation results of the phonon transmission coefficient indicate a significant improvement in both elastic and inelastic phonon transport at the interface. Moreover, we demonstrate the effective utilization of resonance to modulate the interfacial modes in pillar-based Si/Ge nanowires, resulting in improved phonon transport efficiency. This modulation is achieved by strategically repositioning the Si and Ge walls near the interface, leading to the development of the ATI-wall structure. Remarkably, the ATI-wall structure exhibits an unprecedented increase in the ITC compared to the original pillar-based design. To provide additional support for our conclusion, we conduct supplementary simulations using graphics processing unit molecular dynamics in conjunction with the neuroevolution potential to calculate the ITC. Our findings highlight the significance of interfacial mode modulation in enhancing the heat transfer in nanoscale systems and provide valuable insights for the design and optimization of thermal management devices and materials.

Figure 1

Schematic of the pillar-based resonant structures and NEMD simulation model. The side views depict the periodic unit cell for (a) pristine Si/Ge NW, (b) two-side branched Si/Ge NW, (c) surrounding wall Si/Ge NW, and (d) the NEMD simulation model used for calculating the ITC.

Figure 2

Effect of the number and height of pillars on the interfacial phonon transport and related phonon characterization. (a) Impact of the number of pillars, ranging from pristine to four-side branched and further to surrounding wall NWs, on the ITC, with the $L_{\mathrm{h}}=2.17\mathrm{nm}$. The right $y$ axis represents the corresponding overlap factor. (b) The dispersion relations for the pristine, two-side and surrounding wall NWs corresponding to the configurations in panel (a). (c) The effect of the geometry, such as pillar height $L_{\mathrm{h}}$, pillar length $L_{\mathrm{ps}}$, and edge length $L_{\mathrm{e}}$ for surrounding wall NWs on the ITC. (d) The transmission coefficients of the partial structures corresponding to the pillar configurations shown in panel (a).

Figure 3

The comparison of the ITC and corresponding PDOS for three representative structures. (a) The comparison of the ITC (calculated by the empirical SW potential and NEP machine-learning potential) among three structures: pristine, surrounding wall, and ATI-wall Si/Ge NWs. The inset shows the schematic of ATI-wall. (b) The difference in interfacial modes for the three structures depicted in panel (a), where the PDOS is calculated from the velocity of one layer of atoms at the interface.

Figure 4

The variation of the ITC with increasing temperatures in different structures, including (a) pristine and surrounding wall Si/Ge NWs and (b) ATI-wall Si/Ge NWs.