PAOFLOW.models.models#
Functions#
|
Build a generalized Slater-Koster tight-binding model (two-center, up to 3NN). |
|
Build an environment-dependent Slater-Koster tight-binding model (up to 3NN). |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Module Contents#
- PAOFLOW.models.models.Slater_Koster(data_controller, params)[source]#
Build a generalized Slater-Koster tight-binding model (two-center, up to 3NN).
This routine populates the PAOFLOW data containers with a real-space tight-binding Hamiltonian constructed in the two-center approximation, up to third nearest neighbors. It uses a simple orbital basis per atom (s,p and d) and derives hopping matrix elements from the standard Slater-Koster direction-cosine expressions.
High-level workflow:
Read lattice vectors and atomic positions from params and set up basic attributes (number of atoms, orbitals, hoppings).
Compute reciprocal lattice vectors and the cell volume.
Build a 3x3x3 supercell of atomic positions to identify neighbors and determine a first-neighbor cutoff.
Construct the Dnm matrix (orbital-position differences) for gradient calculations.
Allocate the real-space Hamiltonian array HRs and fill on-site energies and hopping terms using direction cosines and SK parameters.
Data structure expectations (params):
params[‘model’][‘a_vectors’]: 3x3 lattice vectors.
params[‘model’][‘atoms’]: dict keyed by string indices (“0”, “1”, …), each containing:
‘name’: atomic species label
‘tau’: fractional/cartesian position in lattice units
‘orbitals’: list of orbital labels used to map on-site terms
on-site energies keyed by orbital label
params[‘model’][‘hoppings’]: either a flat dict of SK parameters with keys ‘sss’, ‘sps’, ‘pps’, ‘ppp’, etc., or a shell dict with ‘nn’ and optional ‘nnn’/’nnnn’ blocks containing those keys.
Output side-effects (data_controller):
arry[‘a_vectors’], arry[‘b_vectors’], arry[‘tau’], arry[‘atoms’], arry[‘shells’], arry[‘norbitals’], arry[‘sctau’], arry[‘Dnm’], arry[‘HRs’]
attr[‘alat’], attr[‘omega’], attr[‘natoms’], attr[‘nawf’], attr[‘bnd’], attr[‘nbnds’], attr[‘nk1’], attr[‘nk2’], attr[‘nk3’], attr[‘nkpnts’], attr[‘nspin’], attr[‘dftSO’], attr[‘shift’], attr[‘cutoff’]
Notes:
The neighbor search uses a 3x3x3 supercell for nn only, 5x5x5 when nnn is enabled, and 7x7x7 when nnnn is enabled. Shells are determined from distinct neighbor distances and mid-point cutoffs.
Only the unpolarized case is supported; spin-orbit is disabled here.
- PAOFLOW.models.models.SK_EDTB(data_controller, params)[source]#
Build an environment-dependent Slater-Koster tight-binding model (up to 3NN).
Extends the two-center
Slater_Kosterwith environment-dependent screening corrections in the style of Porezag & Frauenheim.For each bond (i, j) the two-center SK integrals are modulated by a scalar factor that depends on the local atomic environment:
V_λ^eff(i,j) = V_λ^(2c) · exp( -γ_λ · S_ij )
where S_ij = Σ_{k ≠ i,j} f_c(d_ik) · f_c(d_jk) is a screening sum over nearby mediating atoms k, and f_c is a smooth cutoff function.
- The γ_λ screening strengths can be specified:
per SK channel (‘sss’, ‘pps’, …)
per l-pair (‘ss’, ‘sp’, ‘pp’, ‘sd’, ‘pd’, ‘dd’)
as a single global value
Additional input (on top of
Slater_Kosterrequirements):- params[‘model’][‘screening’]dict
- r_cutfloat
Cutoff radius (same units as a_vectors) for mediating atoms.
- gammafloat or dict
If float → single global screening strength. If dict → per-channel {‘sss’: γ1, ‘pps’: γ2, …} or per-l-pair {‘ss’: γ1, ‘sp’: γ2, …}.
- onsite_shiftdict, optional
Per-orbital environment-dependent on-site correction strength. Keys are orbital labels (‘s’, ‘p’, ‘d’); values are floats. Each on-site energy receives Δε = η · Σ_k f_c(d_ik).