Pipeline architecture
QE-to-TCAD orchestrates a complete workflow from the Materials Project database to TCAD export, via Quantum ESPRESSO.
Data flow
Main steps
1. Structure acquisition
- Query the Materials Project API by chemical formula (
Si,C,Ge…) - Download the most stable crystal structure (CIF/POSCAR)
- Auto-download ONCV-PBE pseudopotentials (SG15 library)
2. QE input generation
The qe_input_generator.py module produces .in files for SCF, NSCF, and epsilon.x calculations.
3. Execution and convergence
qe_runner.py runs pw.x and epsilon.x with MPI. convergence_manager.py optimizes ecut, k-points and lattice.
4. Extraction and export
parse_tcad_parameters.py parses QE outputs into structured JSON in parsed_data/.
5. Visualization
plotter.py and plot_convergence.py generate terminal and 300 DPI PNG figures.
Entry points
| Command | Role |
|---|---|
qe-bridge <formula> | Full pipeline (recommended) |
qe-plot <formula> | ε(ω) plot |
qe-tcad <formula> diode|transistor | 1D device simulation |
python3 fetcher.py <formula> | Direct orchestrator (source repo) |
python3 convergence_manager.py | Convergence only |
python3 qe_runner.py <file.in> | Low-level QE execution |