Thermal Resistance at a Twist Boundary and Semicoherent Heterointerface

TL;DR

This paper develops a theoretical model to analyze phonon scattering at twist boundaries and heterointerfaces, highlighting the significant role of dislocation strain fields in thermal resistance, with implications for device thermal management.

Contribution

It introduces a formalism that separately quantifies the effects of acoustic mismatch and dislocation strain fields on thermal boundary resistance.

Findings

Dislocation strain fields double the thermal resistance in Si-Ge interfaces.

Strain fields dominate thermal resistance in Si-Si twist boundaries.

Model aligns with experimental and simulation data on boundary resistance trends.

Abstract

Traditional models of interfacial phonon scattering, including the acoustic mismatch model (AMM) and diffuse mismatch model (DMM), take into account the bulk properties of the material surrounding the interface, but not the atomic structure and properties of the interface itself. Here, we derive a theoretical formalism for the phonon scattering at a dislocation grid, or two interpenetrating orthogonal arrays of dislocations, as this is the most stable structure of both the symmetric twist boundary and semicoherent heterointerface. With this approach, we are able to separately examine the contribution to thermal resistance due to the step function change in acoustic properties and due to interfacial dislocation strain fields, which induces diffractive scattering. Both low-angle Si-Si twist boundaries and the Si-Ge heterointerfaces are considered here and compared to previous experimental and simulation results. This work indicates that scattering from misfit dislocation strain fields doubles the thermal boundary resistance of Si-Ge heterointerfaces compared to scattering due to acoustic mismatch alone. Scattering from grain boundary dislocation strain fields is predicted to dominate the thermal boundary resistance of Si-Si twist boundaries. This physical treatment can guide the thermal design of devices by quantifying the relative importance of interfacial strain fields, which can be engineered via fabrication and processing methods, versus acoustic mismatch, which is fixed for a given interface. Additionally, this approach captures experimental and simulation trends such as the dependence of thermal boundary resistance on the grain boundary angle and interfacial strain energy.