Official Resources
- Homepage: https://www.physics.rutgers.edu/pythtb/
- Documentation: https://www.physics.rutgers.edu/pythtb/usage.html
- Repository: https://github.com/pythtb/pythtb
- License: MIT License (Assumed, check distribution)
Overview
PythTB is a lightweight Python package for constructing and solving tight-binding models. Developed by David Vanderbilt's group at Rutgers University, it is widely used as a pedagogical tool for teaching topological band theory and Berry phase physics. Its simplicity makes it ideal for rapid prototyping of model Hamiltonians, while its specialized routines for calculating Berry phases, Wilson loops, and Chern numbers make it a powerful research tool for topological insulators and semimetals.
Scientific domain: Band Theory, Topology, Berry Phase
Target user community: Students, Educators, and Researchers in topological matter
Theoretical Methods
- Tight-Binding: Slater-Koster or manual hopping parameters.
- Topological Invariants:
- Berry Phase: $\phi = \oint \mathbf{A} \cdot d\mathbf{k}$.
- Berry Curvature: $\Omega(\mathbf{k}) = \nabla \times \mathbf{A}(\mathbf{k})$.
- Chern Numbers: Integration of curvature over the BZ.
- $\mathbb{Z}_2$ Invariants: Via Wilson loop spectra or parity analysis (for inversion symmetric systems).
- Surface States: Slab construction to inspect edge modes.
Capabilities
- Model Construction:
- 0D (molecules) to 3D crystals.
- Spinful and Spinless fermions.
- Complex hopping amplitudes.
- Solvers:
- Exact Diagonalization on k-paths or meshes.
- Analysis:
- Band structures.
- Berry flux through k-space patches.
- Hybrid Wannier charge centers (Wilson loops).
Key Strengths
- Pedagogy: The syntax is extremely cleaner and intuitive (
model.set_hop(...)), making it the "Arduino of tight-binding codes."
- Topology Native: Unlike general TB codes, PythTB has built-in, robust functions specifically for Berry phases and Wilson loops, reflecting the expertise of the Vanderbilt group.
- Pure Python: Zero compilation required; easy to modify and inspect.
Inputs & Outputs
- Inputs: Python scripts.
- Outputs:
- Arrays of eigenvalues/vectors.
- Matplotlib plots (bands, Berry curvature).
Interfaces & Ecosystem
- Dependencies: NumPy, Matplotlib.
- Adoption: Used in widely circulated lecture notes on topological insulators.
Performance Characteristics
- Speed: Pure Python loops can be slow for very large systems or huge k-meshes.
- Scalability: Best for small unit cells (model physics). Not designed for large-scale supercells (use Pybinding or Kwant for that).
Comparison with Other Codes
- vs. Kwant: Kwant is for transport/scattering. PythTB is for bulk band topology and Berry phases.
- vs. Pybinding: Pybinding is high-performance (C++) for large systems. PythTB is simpler and focused on topological invariants of perfect lattices.
Application Areas
- Haldane Model: The classic example of a Chern insulator.
- Weyl Semimetals: Calculating the chirality of Weyl nodes via Berry flux.
- Education: Teaching the concept of the geometric phase.
Community and Support
- Development: Rutgers University (Sinisa Coh, David Vanderbilt).
- Source: Website / GitHub.
Verification & Sources