ristikko 0.1.2

Classes and utils for tight-binding models

ristikko

A library for tight-binding simulations, with a particular focus on realistic systems, disorder and impurities.

Some examples are found in the examples/ directory.

For quantum transport, consider using kwant or pyqula.

Features

• Single-particle Hamiltonians, 0d, 1d, 2d (3d possible but not yet implemented).
• Spinless/spinful/multiorbital basis
• Arbitrarily complicated superlattice/sublattice structure
• Arbitrarily distant hoppings
• Arbitrary number of impurities
• Disorder, can be applied to onsite potentials or hoppings
• Superconductivity, magnetism, spin-orbit coupling
• Electronic structure
• Local density of states, resolved over the full basis
• Chern numbers, winding numbers, Z2 invariants
• Build from DFT models

Brief examples

Installation

pip install ristikko, or clone this repo, cd into and pip install . to run the master version.

Create a square lattice and plot

params = Params(dict(
    N1 = 16,
    N2 = 16,
    mu = 0,
    t = 1,
    pbc = False,
))

lattice = models.Square()
sys = System(Space.RealSpace, lattice, params)
sys.plot()

Calculate the local density of states

omega = np.linspace(-1, 1, 1001)
ldos = sys.calc_ldos(omega)

Calculate the spectra

points = np.array([
    [0, 0],
    [0.5, 0],
    [0.5, 0.5],
    [0, 0.5],
    [0, 0]
]) * 2 * np.pi
Ks = np.array(k_path(101, points))

sys = System(Space.KSpace, lattice, params)
Ek = sys.diagonalise(k=Ks)

Phase diagram for a toy model for a topological superconductor

@vectorize_parallel
def calc(mu, J):
    params = dict(
            t = 1,
            mu = mu.tolist(),
            Jz = J.tolist(),
            alpha = 0.5,
            Delta = 1,
            direction = "a2",
    )
    params = Params(params)

    lattice = models.Square()
    sys = System(Space.KSpace, lattice, params)
    M = sys.calc_majorana()
    return M

Mus = np.linspace(-4, 4, 301)
Js = np.linspace(-4, 4, 303)

M = calc(Mus[None, :], Js[:, None], mem=mem, progress=True)

fig, axis = plt.subplots(figsize=(4, 4))
cmap = colours.BlackWhite()
im = axis.pcolormesh(Mus, Js, M, cmap=cmap)
add_colourbar(fig, axis, im)

Shiba chain on a substrate

params = dict(
    N1 = 11,
    N2 = 6,
    t = 1,
    mu = -2,
    alpha = 0.5,
    Delta = 1,
    impurities = dict(
        t = 1,
        mu = 0.5,
        J = 2.5,
    ),
    pbc = False,
)
params = Params(params)

t = Symbol("impurities.t")
mu = Symbol("impurities.mu")
J = Symbol("impurities.J")
lattice = models.Square()
realspace = System(Space.RealSpace, lattice, params)
idxs = np.arange(params.N1*params.N2).reshape(params.N1, params.N2)
y = 3
for x in range(params.N1-1):
    idx = realspace.add_impurity(coord=(x+0.5, y-0.5))
    imp = realspace.impurities[idx]
    idxs[x, y] = idx

    imp.add_hopping(imp, 0, 0, mu + J)
    imp.add_hopping(imp, 1, 1, mu + (-1 * J))

    for i in range(2):
        if x > 0:
            imp1 = realspace.impurities[idxs[x-1, y]]
            imp.add_hopping(imp1, i, i, t)

        imp.add_hopping(Site(x, y, 0, 0), i, i, t)

        imp.add_hopping(Site((x+1)%params.N1, y, 0, 0), i, i, t)
        imp.add_hopping(Site(x, y-1, 0, 0), i, i, t)

        imp.add_hopping(Site((x+1)%params.N1, (y-1)%params.N1, 0, 0), i, i, t)

realspace.plot(aspect=6/11, axis_scale=5)

Disordered lattice

params = Params(dict(
    N1 = 16,
    N2 = 16,
    mu = 0,
    t = 1,
    pbc = False,
    seed = 0,
    scale = 1,
))

lattice = models.Square()
sys = System(Space.RealSpace, lattice, params)

def potential_disorder(params, ts, ij):
    rng = numpy.random.default_rng(params.seed)
    mu = rng.normal(size=(params.N1, params.N2), scale=params.scale)
    for n in range(ts.size):
        x0, y0, x1, y1, i, j = ij[n]
        if i == j and x0 == x1 and y0 == y0:
            ts[n] = mu[x0, y0]
    return ts, ij


H = sys.hamiltonian(postprocess_hoppings=potential_disorder)

Homepage

https://git.sr.ht/~dcrawford/ristikko

Licence

GPLv3+