[18a-A21-1~11] 23.1 Joint Session N "Informatics"

11:15 AM - 11:30 AM JST(2:15 AM - 2:30 AM UTC)

[18a-A21-9] Thermal Boundary Resistance between Interconnect Unitary Metals for Beyond 2nm Logic Nodes and SiO₂ Dielectric: Molecular Dynamics Calculation based on Neural Network Potential

〇Shuichiro Hashimoto¹, Yusuke Nishimura², Takanobu Watanabe^1,2 (1.Wasda Univ. SEES, 2.Waseda Univ. FSE)

Keywords:

interconnect,thermal resistance,machine learning molecular dynamics calculation

Beyond 2nm logic nodes expect interconnect booster replacing traditional Cu interconnect metal with high melting point interconnect metals including Ru . In very-large-scale integration (VLSI), the self-heating effect, which causes electromigration (EM) leading to interconnect failure, varies depending on the interconnect metal and the interlayer dielectric. Specifically, at Ru/SiO₂ with interlayers, thermal boundary resistance (TBR) changes according to the bonding behavior. Therefore, in miniaturized interconnects, thermal transport between different materials plays a critical role. This study conducted machine learning molecular dynamics calculations based on neural network potentials to investigate the interfacial thermal resistance between SiO₂ interlayer dielectric films and unitary interconnect metals with high-melting point such as Ru.