🔬 VLSI Virtual FabLab: Process Predictive Dashboard

A multi-module Python application built with Streamlit to simulate and visualize core semiconductor manufacturing processes. This project focuses on physical modeling for bridging the gap between theoretical VLSI equations and real-world fab results.

Project Overview

Originally developed as a Thermal Oxidation validator against the BYU Cleanroom "Gold Standard," this tool evolved into a comprehensive suite covering the three fundamental pillars of the IC fabrication cycle: Additive, Modificative, and Subtractive processes.

Thermal Oxidation (Additive)
Physics: Implements the Deal-Grove Model with Arrhenius temperature dependence for linear ($B/A$) and parabolic ($B$) growth.
Validation: Calibrated to match BYU benchmarks (e.g., 71.1 nm at $1000^\circ\text{C}$ for 60 min, Dry $O_2$).
Visualization: Maps oxide thickness to a dynamic Thin-Film Interference color profile (Tan/Brown, Purple, Blue, etc.).

Ion Implantation (Modificative)
Physics: Models dopant concentration using a Gaussian Distribution based on LSS Theory.
Parameters: Supports Boron, Phosphorus, and Arsenic species with variable Energy (keV) and Dose ($atoms/cm^2$).
Visualization: Features a Semi-Logarithmic Profile ($Log(C)$ vs. Depth) to accurately identify electrical junction depths ($x_j$).

Etch Rate & Selectivity (Subtractive)
Physics: Simulates material removal for $SiO_2$ and $Si_3N_4$ using various chemistries like Buffered HF, KOH, and Plasma RIE.
Metrics: Calculates Selectivity (S) ratios and provides critical feedback on Substrate Loss during over-etching.
Visualization: A real-time cross-sectional "stack" view showing the physical thinning of layers.

Roadmap & Future Development

Photolithography Module (In Development)

The final pillar of the "Virtual Fab" is currently being engineered. This module will move the dashboard into the Optical Domain, including the implementation of the Rayleigh Criterion ($R = k_1 \frac{\lambda}{NA}$) to determine minimum feature sizes, calculating process windows ($DOF = k_2 \frac{\lambda}{NA^2}$) for different light sources (I-line, KrF, ArF), and a dynamic "Reticle-to-Wafer" scaling visualization to show how wavelength affects chip density.

Quality of Life (QoL) & UI Enhancements

To make the "Virtual Fab" more comprehensive and reactive, the following upgrades are on the way:

Interactive Cross-Sections: Real-time animations in the Etch Lab that respond instantly to slider movements for a "live etch" feel.
Process Chaining: A feature to pass the output of one tab (e.g., Oxide Thickness) as the starting input for the next (e.g., Etch Target).
Advanced Error Handling: "Safety Guardrails" that trigger visual warnings if process parameters (like Temperature or Dose) exceed physically realistic bounds.
Exportable Fab Reports: A one-click "Download CSV/PDF" button to save simulation results and plots for academic lab reports.

Tech Stack

Language: Python 3.10+.
Framework: Streamlit (Web UI).
Math/Stats: NumPy, SciPy.
Visualization: Matplotlib.
Other Tools: BYU Cleanroom Process Calculator(s), Google Gemini AI.