Wet Oxide Growth Testing

At RIT SNL I worked on optimizing LPCVD wet oxide growth in a Tystar Tytan 3600. The standard recipes for thick oxide growth on the Tystar furnace were suffering from excessive cycle times, creating a bottleneck in fabrication throughput. The goal was to characterize and optimize a "Dry-Wet-Dry" oxidation process to achieve a target thickness of >5µm while strictly maintaining wafer uniformity. I used Deal–Grove model to model the oxidation growth rates to predict the exact time required for the rapid wet oxidation step versus the slower, high-quality dry oxidation cap layers. This modeling allowed me to set precise dry-wet-dry segment times to maximize growth speed without sacrificing film density, optimize temperature ramp rates, and validate the model by mapping films with reflectometry, iterating the recipe based on actual growth data.
    I was able to achieve a >20% reduction in total process time while maintaining quality with <2% non-uniformity across the wafer (Batch mean: ~5540 nm).