Official Resources
- Repository: https://github.com/wannier-utils-dev/symclosestwannier
- Documentation: (In repository README/examples)
- License: MIT License
- Developers: wannier-utils-dev team (J. Wang, et al.)
Overview
symclosestwannier is a Python library that implements the Symmetry-Adapted Closest Wannier (SymCW) method. It addresses a key limitation in standard Wannier90 workflows: the difficulty of ensuring that the resulting Maximally Localized Wannier Functions (MLWFs) fully respect the crystalline symmetry of the material. By projecting Bloch states onto a Symmetry-Adapted Multipole Basis (SAMB), this tool constructs high-quality tight-binding models that are naturally symmetric without requiring iterative minimization.
Scientific domain: Condensed matter physics, topological materials, tight-binding modeling.
Target user community: Researchers needing rigorous symmetry preservation in Wannier models (e.g., for topological analysis).
Theoretical Methods
- Closest Wannier Formalism: Analytical construction of Wannier functions closest to a set of trial orbitals in a least-squares sense.
- Symmetry-Adapted Multipole Basis (SAMB):
- Expands the Hamiltonian in terms of bases belonging to the identity representation of the crystal point group to ensure symmetry.
- Utilizes "site-symmetry" adaptation.
- Non-Iterative Projection: Determines model parameters via direct matrix projection rather than iterative disentanglement/minimization.
Capabilities
- Parameter-Free Construction: Avoids the "trial orbital" guessing game and local minima issues of iterative MLWF schemes in many cases.
- Symmetry Restoration: Guarantees the resulting tight-binding Hamiltonian transforms correctly under all crystal symmetry operations.
- Multipole Analysis: Can evaluate hidden electronic multipole degrees of freedom.
- Connectivity: Interfaces with standard DFT codes (Quantum ESPRESSO, VASP, WIEN2k) via their Wannier interfaces.
Key Strengths
- Rigorous Symmetry: Prevents slight symmetry breakings that can occur in numerical MLWF minimization, which is critical for topological invariant calculations.
- Efficiency: Significantly faster than iterative schemes for complex unit cells because it uses a direct projection.
- Robustness: Reduces human error in selecting initial projections for Wannier90.
Inputs & Outputs
- Inputs:
- Wavefunction overlaps usually generated for Wannier90 (
.mmn, .amn or equivalent).
- Symmetry information of the crystal structure.
- Outputs:
- A symmetrized tight-binding Hamiltonian (often in Wannier90
_hr.dat format or internal format).
- Analysis of orbital characters.
Interfaces & Ecosystem
- Wannier90 Compatible: Can function as a pre-processing or alternative step to standard Wannier90 runs.
- DFT Codes: Compatible with any code that generates Wannier90 interface files (e.g.,
pw2wannier90.x in Quantum ESPRESSO).
Computational Cost
- Low: The projection operation is algebraic and very fast compared to the Self-Consistent Field (SCF) cycle of DFT or the iterative minimization of MLWFs in large systems.
Comparison with Other Codes
- vs [Wannier90](file:///home/niel/git/Indranil2020.github.io/scientific_tools_consolidated/TightBinding/4.1_Wannier_Ecosystem/Wannier90.md): Wannier90 uses iterative minimization ($ \Omega $ functional) which is general but can break symmetry; SymClosestWannier uses projection onto symmetry-adapted bases for guaranteed symmetry.
- vs [WannierTools](file:///home/niel/git/Indranil2020.github.io/scientific_tools_consolidated/TightBinding/4.1_Wannier_Ecosystem/WannierTools.md): WannierTools is for analyzing the TB model (surface states, etc.); SymClosestWannier is for constructing the model.
Application Areas
- Topological Materials: Where symmetry eigenvalues at high-symmetry points are crucial for topology (e.g., TCI, Weyl semimetals).
- Phonon-Electron Coupling: Where symmetry affects selection rules.
- Automated TB Construction: For high-throughput databases.
Verification & Sources
- Primary Source: GitHub Repository
- Citation: Wang, J. et al., "Symmetry-adapted closest Wannier functions", (ArXiv/Related publications, e.g., Phys. Rev. B).
- Verification Status: ✅ VERIFIED (Research code).