BEOL TDDB model: predicting dielectric reliability across 3D BEOL structures in sub-30nm metal pitch

Objective:
This project aims to implement from scratch the Local E-Field Aware Model (LEFAM) described in the paper “Local Electric Field–Aware 3D TDDB model for BEOL reliability predictions” (see reference) using Python. The goal is to reproduce the model’s predictions of dielectric breakdown time (tBD) and extend its applicability to novel interconnect structures beyond those originally simulated.

Background:
As interconnect dimensions continue to scale below 30 nm, local geometrical variations such as line-edge roughness (LER), via misalignment (VM), and tip-to-tip spacing variations significantly impact backend-of-line (BEOL) dielectric reliability. LEFAM improves upon previous approaches by integrating finite element simulations of the local E-field (Eloc), defect generation dynamics, and percolation theory to model dielectric breakdown more realistically.

Internship Tasks:

• Study and interpret the LEFAM-based TDDB model, including:

  • Finite Element Method (FEM) simulations of Eloc under geometric variability

  • E-field–dependent and defect density–dependent defect generation

  • Local defect clustering and percolation-based breakdown triggering

• Reconstruct the LEFAM framework in Python based on the article

• Validate the model against published figures or synthetic geometries

• Simulate tBD distributions for custom L2L and V2L structures under varying variability settings (e.g., σ_LER, VM)

• Fit clustering distributions to non-Weibull tBD outputs and compare with SSPP model predictions

• Perform sensitivity analysis to assess the influence of geometry and variability on TDDB lifetime metrics

Skills to be Gained:

• Practical experience with reliability modeling of semiconductor interconnects

• Proficiency in Python for FEM simulation, statistical modeling, and Monte Carlo methods

• Insight into the role of geometry-aware modeling in reliability predictions

Requirements:

• Background in electrical engineering, physics, materials science, or a related field

• Familiarity with Python and numerical simulation

• Basic understanding of dielectric breakdown mechanisms is a plus