ASE is short for the atomistic simulation environment, a Python package to simplify the process of setting up, running and analysing results from atomistic simulations across different programs. Extremely powerful in this respect are the routines this code provides for setting up complicated systems (including surface-adsorption scenarios, defects, nanotubes, etc). See also the ASE installation instructions.
This example shows how to use ASE to setup a particular gallium arsenide surface and run the resulting calculation in DFTK. If you are less interested in having access to the full playground of options in DFTK, but more interested in performing analysis in ASE itself, have a look at asedftk. This package provides an ASE-compatible calculator class based on DFTK, such that one may write the usual Python scripts against ASE, but the calculations are still run in DFTK.
The particular example we consider the (1, 1, 0) GaAs surface separated by vacuum with the setup slightly adapted from 1.
D. Raczkowski, A. Canning, and L. W. Wang Thomas-Fermi charge mixing for obtaining self-consistency in density functional calculations Phys. Rev. B 64, 121101(R).↩
Parameters of the calculation. Since this surface is far from easy to converge,
we made the problem simpler by choosing a smaller
Ecut and smaller values
More interesting settings are
Ecut = 15 and
n_GaAs = n_vacuum = 20.
miller = (1, 1, 0) # Surface Miller indices n_GaAs = 2 # Number of GaAs layers n_vacuum = 4 # Number of vacuum layers Ecut = 5 # Hartree kgrid = (4, 4, 1); # Monkhorst-Pack mesh
Use ASE to build the structure:
using PyCall ase_build = pyimport("ase.build") a = 5.6537 # GaAs lattice parameter in Ångström (because ASE uses Å as length unit) gaas = ase_build.bulk("GaAs", "zincblende", a=a) surface = ase_build.surface(gaas, miller, n_GaAs, 0, periodic=true);
Get the amount of vacuum in Ångström we need to add
d_vacuum = maximum(maximum, surface.cell) / n_GaAs * n_vacuum surface = ase_build.surface(gaas, miller, n_GaAs, d_vacuum, periodic=true);
Write an image of the surface and embed it as a nice illustration:
pyimport("ase.io").write("surface.png", surface * (3, 3, 1), rotation="-90x, 30y, -75z")
to convert to DFTK datastructures.
These two functions not only support importing ASE atoms into DFTK,
but a few more third-party datastructures as well.
Typically the imported
atoms use a bare Coulomb potential,
such that appropriate pseudopotentials need to be attached in a post-step:
using DFTK atoms = load_atoms(surface) atoms = [ElementPsp(el.symbol, psp=load_psp(el.symbol, functional="pbe")) => position for (el, position) in atoms] lattice = load_lattice(surface);
We model this surface with (quite large a) temperature of 0.01 Hartree
to ease convergence. Try lowering the SCF convergence tolerance (
temperature to see the full challenge of this system.
model = model_DFT(lattice, atoms, [:gga_x_pbe, :gga_c_pbe], temperature=0.001, smearing=DFTK.Smearing.Gaussian()) basis = PlaneWaveBasis(model, Ecut; kgrid=kgrid) scfres = self_consistent_field(basis, tol=1e-4, mixing=KerkerMixing());
n Free energy Eₙ-Eₙ₋₁ ρout-ρin Diag --- --------------- --------- -------- ---- 1 -16.60904549584 NaN 2.60e-01 4.8 2 -16.72296863664 -1.14e-01 8.41e-02 1.0 3 -16.66425534939 5.87e-02 5.14e-02 2.3 4 -16.73220156006 -6.79e-02 1.93e-02 2.0 5 -16.73631937580 -4.12e-03 5.13e-03 1.8 6 -16.73635765509 -3.83e-05 3.00e-03 3.0
Energy breakdown: Kinetic 5.8405521 AtomicLocal -105.5505203 AtomicNonlocal 2.3482321 Ewald 35.5044300 PspCorrection 0.2016043 Hartree 49.5126002 Xc -4.5931249 Entropy -0.0001312 total -16.736357655094