xespresso

Quantum ESPRESSO Calculator for Atomic Simulation Environment (ASE).

For the introduction of ASE , please visit https://wiki.fysik.dtu.dk/ase/index.html

Features

• Support all QE packages, including: pw, band, neb, dos, projwfc, pp ...
• Spin-polarized calculation
• LD(S)A+U
• Automatic submit job
• Automatic check a new calculation required or not
• Automatic set up "nscf" calculation
• Read and plot dos, pdos and layer resolved pdos
• Plot NEB

Author

• Xing Wang xingwang1991@gmail.com

Dependencies

• Python
• ASE
• numpy
• scipy
• matplotlib

Installation using pip

pip install --upgrade --user xespresso

Installation from source

You can get the source using git:

git clone --depth 1 https://github.com/superstar54/xespresso.git

Add xespresso to your PYTHONPATH. On windows, you can edit the system environment variables.

export PYTHONPATH="/path/to/xespresso":$PYTHONPATH
export ASE_ESPRESSO_COMMAND="/path/to/PACKAGE.x  PARALLEL  -in  PREFIX.PACKAGEi  >  PREFIX.PACKAGEo"
export ESPRESSO_PSEUDO="/path/to/pseudo"

Examples

Automatic submit job

A example of setting parameters for the queue. See example/queue.py

queue = {'nodes': 4,
         'ntasks-per-node': 20,
         'partition': 'all',
         'time': '23:10:00'}
calc = Espresso(queue = queue)

Automatic check a new calculation required or not

Before the calculation, it will first check the working directory. If the same geometry and parameters are used, try to check whether the results are available or not. Automatic check input parameters with Quantum Espresso document.

calc = Espresso(label = 'scf/fe')

Show debug information.

calc = Espresso(debug = True)

Add new species

Some atoms are special:

• atoms with different starting_magnetization
• atoms with different U values
• atoms with special basis set

For example, Fe with spin state AFM. See example/spin.py

atoms.new_array('species', np.array(atoms.get_chemical_symbols(), dtype = 'U20'))
atoms.arrays['species'][1] = 'Fe1'

Setting parameters with "(i), i=1,ntyp"

Hubbard, starting_magnetization, starting_charge and so on. See example/dft+u.py

input_ntyp = {
'starting_magnetization': {'Fe1': 1.0, 'Fe2': -1.0},
'Hubbard_U': {'Fe1': 3.0, 'Fe2': 3.0},
}
input_data['input_ntyp'] = input_ntyp,

Setting parameters for "Hubbard_V(na,nb,k)"

Hubbard, starting_magnetization, starting_charge and so on. See example/dft+u.py

input_data = {
'hubbard_v': {'(1,1,1)': 4.0, '(3,3,1)': 1.0},
}

Control parallelization levels

To control the number of processors in each group: -ni, -nk, -nb, -nt, -nd) are used.

calc = Espresso(pseudopotentials = pseudopotentials,
                 package = 'pw',
                 parallel = '-nk 2 -nt 4 -nd 144',  # parallel parameters
                 }

Non self-consistent calculation

A example of nscf calculation following the above one.

# start nscf calculation
from xespresso.post.nscf import EspressoNscf
nscf = EspressoNscf(calc.directory, prefix = calc.prefix,
                occupations = 'tetrahedra',
                kpts = (2, 2, 2),
                debug = True,
                )
nscf.run()

Calculate dos and pdos

A example of calculating and plotting the pdos from the nscf calculation.

from xespresso.post.dos import EspressoDos
# dos
dos = EspressoDos(parent_directory = 'calculations/scf/co',
            prefix = calc.prefix,
            Emin = fe - 30, Emax = fe + 30, DeltaE = 0.01)
dos.run()
# pdos
from xespresso.post.projwfc import EspressoProjwfc
projwfc = EspressoProjwfc(parent_directory = 'calculations/scf/co',
            prefix = 'co',
            DeltaE = 0.01)
projwfc.run()

Calculate work function

from xespresso.post.pp import EspressoPp
pp = EspressoPp(calc.directory, prefix = calc.prefix,
                plot_num = 11,
                fileout = 'potential.cube',
                iflag = 3,
                output_format=6,
                debug = True,
                )
pp.get_work_function()

Restart from previous calculation

calc.read_results()
atoms = calc.results['atoms']
calc.run(atoms = atoms, restart = 1)

NEB calculation

See example/neb.py

from xespresso.neb import NEBEspresso
calc = NEBEspresso(
                 package = 'neb',
                 images = images,
                 climbing_images = [5],
                 path_data = path_data
                 )
calc.calculate()
calc.read_results()
calc.plot()

Workflow

Oxygen evolution reaction (OER) calculation

The workflow includes four modules:

• OER_bulk
• OER_pourbaix
• OER_surface
• OER_site

The workflow can handle:

• Generate surface slab model from bulk structure
• Determine the surface adsorption site
• Determine the surface coverage(, O, OH*), Pourbaix diagram
• Calculate the Zero-point energy

oer = OER_site(slab,
               label = 'oer/Pt-001-ontop',
               site_type = 'ontop',
               site = -1,
               height=2.0,
	           calculator = parameters,
               molecule_energies = molecule_energies,
               )
oer.run()

To do lists:

• add qPointsSpecs and Line-of-input for phonon input file