3D Current Reconstruction for Failure Analysis with Quantum Diamond Microscopy

Abstract

Reconstructing electrical current flow within integrated circuits is a powerful capability for electrical failure analysis (EFA), as it reveals the actual conduction paths associated with defects. Quantum Diamond Microscopy (QDM) enables non-destructive, vector-resolved magnetic field imaging, providing the basis for quantitative current reconstruction even in buried structures. In this work, we extend conventional approaches by introducing a data-driven framework for three-dimensional (3D) current reconstruction. Vector magnetic field measurements are combined with a U-Net-based model trained on physically representative current distributions to infer volumetric current density from a single measurement plane. The approach is validated using custom-designed test structures, and real device measurements, demonstrating accurate depth-resolved reconstruction across a range of conditions. These results highlight the potential of combining QDM with machine learning to enhance current reconstruction capabilities for advanced EFA applications.