GNR Quantum Prototype

GNR Quantum Prototype is an early open-source proof of concept for a reproducible benchmark workflow at the intersection of computational chemistry and near-term quantum methods. It explores small Hückel Hamiltonians with both fixed and structure-aware parameterization, estimates HOMO, LUMO, and bandgap values, and compares Jordan–Wigner and compact encoding in terms of qubit requirements.

Overview

This project is designed as a lightweight educational and research demo for exploring how interpretable Hamiltonian models can connect to quantum-oriented workflows. Rather than treating bandgap prediction as a black-box problem, the prototype focuses on simple, physically meaningful parameterization and resource-aware comparison.

Current Features

• Construction of toy Hückel Hamiltonians
• Classical solution of eigenvalues and frontier orbitals
• Estimation of HOMO, LUMO, and bandgap
• Simple structure-aware alpha parameterization
• Comparison of Jordan–Wigner and compact encoding
• Lightweight Streamlit demo for interactive exploration

Why It Matters

This prototype provides a small but practical benchmark workflow for students and researchers interested in computational chemistry, Hamiltonian modeling, and near-term quantum computing. It is intended as an initial step toward a broader open-source toolkit for studying how compact quantum methods may connect to real chemistry and materials problems in a reproducible and interpretable way.

Project Structure

• app.py – Streamlit demo interface
• test_run.py – simple test script for the prototype
• src/huckel.py – Hückel Hamiltonian construction and bandgap estimation
• src/encoding.py – encoding comparison utilities
• requirements.txt – project dependencies

Running the Demo

Install the required packages:

pip install -r requirements.txt

Run the Streamlit app:

python -m streamlit run app.py

Demo Preview

These screenshots show the current lightweight Streamlit demo for exploring toy Hückel Hamiltonians, bandgap estimates, and compact encoding comparisons.

Future Directions

• Add graph-based descriptors
• Add learned alpha and beta prediction
• Add VQE-based experiments
• Expand toward selected graphene nanoribbon model systems
• Package the workflow as a more reusable benchmark toolkit