Statistical Process Control Strategies for In-Line Sheet Resistance Uniformity in High-Throughput Phosphorus Diffusion Furnaces

Abstract

Wafer-to-wafer and within-wafer sheet resistance uniformity in phosphorus diffusion remains the dominant contributor to emitter-related yield loss in industrial silicon solar cell manufacturing. We present a statistical process control (SPC) framework that combines four-point-probe inline measurements with multivariate Hotelling T-squared monitoring across thirteen tube-position-and-zone covariates. The framework was developed and validated against archived production data from three independent diffusion-furnace platforms and demonstrates a measurable reduction in false-alarm rate compared to univariate SPC at equivalent sensitivity. We further describe a Cpk-driven recipe-stability scoring rubric that links upstream process drift to downstream cell efficiency and recommend a tiered escalation protocol for production engineering teams. The methodology generalizes to TOPCon emitter formation and to boron-doped p-type rear-junction architectures with minor parameter adaptation.