ONETEP

Introduction

ONETEP is a fully-featured density-functional package combining linear scaling with system size, systematic plane-wave accuracy, and excellent parallel scaling. It uses a set of atom-centered local orbitals (denoted NGWFs) which are optimised in situ to enable high accuracy calculations with a minimal number of orbitals.

This interface makes it possible to use ONETEP as a calculator in ASE. You need to have a copy of the ONETEP code (and an appropriate license) to use this interface.

Additionally you will need pseudopotential or PAW dataset files for the combination of atom types of your system.

Environment variables

The environment variable ASE_ONETEP_COMMAND must hold the command to invoke the ONETEP calculation. The variable must be a string with a link to the ONETEP binary, and any other settings required for the parallel execution Example:

You can this environment variable in your shell configuration file:

$ export ASE_ONETEP_COMMAND="export OMP_NUM_THREADS=4; mpirun -n 6 ~/onetep/bin/onetep.arch PREFIX.dat >> PREFIX.out 2> PREFIX.err"

Or within python itself:

>>> environ["ASE_ONETEP_COMMAND"]="export OMP_NUM_THREADS=4; mpirun -n 6 ~/onetep/bin/onetep.arch PREFIX.dat >> PREFIX.out 2> PREFIX.err"

ONETEP Calculator

This is implemented as a FileIOCalculator: most parameters from the ONETEP keyword list: http://www.onetep.org/Main/Keywords can be specified using the calculator’s \(set\) routine.

keyword type default value description
label str None Name of input and output files
cutoff_energy str 1000 eV Energy cutoff of psinc grid
ngwf_radius str 12.0 bohr Cutoff Radius of NGWF
kernel_cutoff str 1000 bohr Cutoff Radius for density kernel

Species Definitions

By default, the calculator will create a “species” definition in the ONETEP input file for each type of element present in the calculation. However, if you add tags to certain atoms (either via the GUI or using the set_tags routine), then a species will be created for each different combination of element and tag value. This automatically propagates through to pseudopotential and pseudoatomic solver blocks.

This provides a useful way to set up (for example) Local Density of States calculations, whereby a certain subset of the atoms are identified as an LDOS group. It is also very useful for defining an atom to have a core hole, for the purposes of EELS.

Examples

Here is an example of setting up a calculation on a water molecule:

# Set up water molecule in box with 6 ang padding.
from ase.build import molecule
wat = molecule('H2O')
wat.center(6)

# Set up a ONETEP geometry optimisation calculation using the PBE functional
from ase.calculators.onetep import Onetep
from os import environ
environ["ASE_ONETEP_COMMAND"]="export OMP_NUM_THREADS=8; mpirun -n 2 ~/onetep/bin/onetep.arch PREFIX.dat >> PREFIX.out 2> PREFIX.err"
calc = Onetep(label='water')
calc.set(pseudo_path='/path/to/pseudos')
calc.set(pseudo_suffix='.PBE-paw.abinit') # use pseudopotentials from JTH library in abinit format
calc.set(task='GeometryOptimization',paw=True,xc='PBE',cutoff_energy='600 eV')
wat.set_calculator(calc)
wat.get_forces()

Here is an example of setting up a calculation on a graphene sheet:

# Set up a graphene lattice with a 9x9 supercell
from ase.lattice.hexagonal import *
index1=9
index2=9
alat = 2.45
clat = 31.85
gra = Graphene(symbol = 'C',latticeconstant={'a':alat,'c':clat},size=(index1,index2,1))

# Set up a ONETEP calculation using PBE functional and ensemble DFT
from ase.calculators.onetep import Onetep
from os import environ
environ["ASE_ONETEP_COMMAND"]="export OMP_NUM_THREADS=4;
    mpirun -n 6 ~/onetep/bin/onetep.arch PREFIX.dat >> PREFIX.out 2> PREFIX.err"
calc = Onetep(label='gra')
calc.set(pseudo_path='/path/to/pseudos')
calc.set(pseudo_suffix='.PBE-paw.abinit') # use pseudopotentials from JTH library in abinit format
calc.set(paw=True,xc='PBE', cutoff_energy='500 eV',ngwf_radius=8,edft='T')

# Run the calculation
gra.get_potential_energy()

Here is an example of setting up an EELS and LDOS calculations on an N-substituted graphene sheet, demonstrating several more advanced functionalities (eg tags, species groups, and overrides to pseudopotentials and atomic solver strings):

# Import modules
from ase.lattice.hexagonal import *
from ase.calculators.onetep import Onetep

# Set up a graphene lattice with a 9x9 supercell
index1=9
index2=9
alat = 2.45
clat = 31.85
gra = Graphene(symbol = 'C',latticeconstant={'a':alat,'c':clat},size=(index1,index2,1))

# find atom near centre of cell to make impurity
j = 80
sym = gra.get_chemical_symbols()
sym[j] = 'N'
gra.set_chemical_symbols(sym)

# define radii for up to 5th nearest neighbour atoms and tag appropriately
tags = gra.get_tags()
tags[j] = -1 # exclude impurity
shell_rad = [1.5,2.5,3.0,4.0,4.5]
for k in range(len(shell_rad)):
    tags = [ k+1 if ((gra.get_distance(i,j)<shell_rad[k]) and
                     (tags[i]==0)) else tags[i] for i in range(len(gra)) ]
tags[j] = 0 # reset impurity tag
gra.set_tags(tags)

# Set up a ONETEP calculation using the PBE functional and ensemble DFT
calc = Onetep(label='gra_Nsub')
calc.set(pseudo_path='/path/to/pseudos')
calc.set(pseudo_suffix='.PBE-paw.abinit') # use pseudopotentials from JTH library in abinit format
calc.set(paw=True,xc='PBE', cutoff_energy='500 eV',ngwf_radius=8,ngwf_radius_cond=9,edft='T')

# Set up a corehole in the nitrogen (change PAW dataset, set core wavefunctions, and set solver string)
calc.set(species_pseudo={"N":"corehole/N.PBE-1s-hole-paw.abinit"})
calc.set(species_core_wf={"N":"corehole/N.PBE-1s-hole-corewf.abinit"})
calc.set(species_solver={"N":"SOLVE conf=1s1 2p4"})

# Set up groups for LDOS: each group is a python list of strings, arranged in a list
calc.set(species_ldos_groups=[['C'],['C1'],['C2'],['C3'],['C4'],['C5'],['N']])

# Write an input file for ONETEP
calc.atoms = gra.copy()
calc.write_input(gra)