This is package for multi_lattice analysis based on the notion of sites, bonds, and local environments. It supports the conversion from and to CIF file formats.
This is a temporary project. It will eventually been merged into the PyStrucTrans project.
Clone or download/unzip the project into your working directory.
Then to use it in your local python scripts, start with import bonds
or from bonds import *.
A useful short hand may be from bonds import MultiLattice as ML.
Let's generate a multilattice with P 21 space group, and two occupied 2a sites, one for Ti, one for Ni.
Import the package first
from __future__ import absolute_import
from bonds import *
from bonds import MultiLattice as ML
First, we need a base matrix. You can generate it by PyStrucTrans. If it is a monoclinic lattice, here is a helper function in bonds.utils
a = 2.91760
b = 4.04300
c = 4.64030
beta = 97.8320
B = utils.lp2B(a, b, c, beta)
Then we need to define our two sites, using the constructor Site(atom, pos, label, occupancy)
NiSite = Site("Ni", (0.54421, 0.25000, 0.17586), "Ni", 1)
TiSite = Site("Ti", (0.08316, 0.75000, 0.28152), "Ti", 1)
Finally, we can generate a MultiLattice object
syms = ('x, y, z', '-x, y+1/2, -z')
latt = ML(B, [NiSite, TiSite], syms)
Now, you can check its sites:
print(latt.sites())
It will give the two sites you initially assigned,
[Ni(Ni) @ (0.544210, 0.250000, 0.175860), Ti(Ti) @ (0.083160, 0.750000, 0.281520)]
To check if the symmetry works, print all the sites in the unit cell
print(latt.all_sites_by_atom())
The output is
{'Ti': [Ti(Ti) @ (0.083160, 0.750000, 0.281520), Ti(Ti) @ (0.916840, 0.250000, 0.718480)], 'Ni': [Ni(Ni) @ (0.544210, 0.250000, 0.175860), Ni(Ni) @ (0.455790, 0.750000, 0.824140)]}
Clearly, it generates all four atoms in the cell correctly.
Now let's export it to a CIF file. The method to_vesta_cif comes in handy:
# call the phase "New structure", and assign the space group No. 4 (P 21) to it
content = latt.to_vesta_cif('New structure', "P 21", 4)
# write it to a file
file = open("NiTi_4atm_P21.cif", "w")
file.write(content)
file.close()
Now load the generated file in VESTA to view the multi lattice.
In fact, you can also load a structure from a cif file.
lp, syms, sites = cif.read_cif("NiTi_4atm_P21.cif")
B = utils.lp2B(*lp)
latt2 = ML(B, sites, syms)
print(latt2.all_sites_by_atom())
Again, we see
{'Ni': [Ni(Ni) @ (0.544210, 0.250000, 0.175860), Ni(Ni) @ (0.455790, 0.750000, 0.824140)], 'Ti': [Ti(Ti) @ (0.083160, 0.750000, 0.281520), Ti(Ti) @ (0.916840, 0.250000, 0.718480)]}
A powerful tool come with MultiLattice is rebase.
Through rebase, you can change the description of the same lattice,
by change the symmetry operations (space group) or the size of the unit cell.
First, let's look at change symmetry. Say, we want a P1 descrition of the NiTi lattice defined above.
import numpy as np
L = np.eye(3) # not change the unit cell
O = NiSite.pos() # move the origin to Ni site
syms = ['x, y, z'] # P1 symmetry
lattP1 = latt.rebase(L, O, ['x, y, z'])
print(lattP1.all_sites_by_atom())
We get again 4 atoms per cell. Notice that because the two Ti sites are considered different, they have different labels now.
{'Ti': [Ti(Ti1) @ (0.372630, 0.000000, 0.542620), Ti(Ti2) @ (0.538950, 0.500000, 0.105660)], 'Ni': [Ni(Ni1) @ (0.000000, 0.000000, 0.000000), Ni(Ni2) @ (0.911580, 0.500000, 0.648280)]}
Export it to a cif file and check in VESTA
content = lattP1.to_vesta_cif('New structure', "P 1", 1)
file = open("NiTi_4atm_P1.cif", "w")
file.write(content)
file.close()
If we want to enlarge the cell but still keep P 21 symmetry, we can do
L = np.array([[2, 0, 0], [0, 1, 0], [0, 0, 2]]) # not change the unit cell
O = [0, 0, 0] # not move the lattice point
latt_16atm = latt.rebase(L, O) # by default using the original symmetry
Again, export to CIF and check in VESTA
content = lattP1.to_vesta_cif('New structure', "P 21", 4)
file = open("NiTi_16atm_P21.cif", "w")
file.write(content)
file.close()
The reason that the main package is called bonds is that it can calculate all the shortes bonds (nearest neighbors) for each site in a multilattice.
# define the coordination numbers for sites
coords = dict()
coords[('Ti', 'Ti')] = coords[('Ni', 'Ni')] = 6
coords[('Ti', 'Ni')] = coords[('Ni', 'Ti')] = 8
envs = latt.all_local_envs(coords)
for env in envs:
print(env.tostr(show_vec=True))
It outputs
"Ni" site at ( 0.544, 0.250, 0.176) :
=======================================
Ni-Ti bonds:
------------
Ni-Ti: 2.513032, ( 1.376, 0.000, -2.103)
Ni-Ti: 2.513175, (-1.412, -2.022, 0.486)
Ni-Ti: 2.513175, (-1.412, 2.022, 0.486)
Ni-Ti: 2.566964, ( 1.506, -2.022, 0.486)
Ni-Ti: 2.566964, ( 1.506, 2.022, 0.486)
Ni-Ti: 2.603043, ( 0.744, 0.000, 2.494)
Ni-Ti: 2.606943, (-1.541, 0.000, -2.103)
Ni-Ti: 3.308537, (-2.174, 0.000, 2.494)
Ni-Ni bonds:
------------
Ni-Ni: 2.588816, (-0.036, -2.022, -1.617)
Ni-Ni: 2.588816, (-0.036, 2.022, -1.617)
Ni-Ni: 2.917600, (-2.918, 0.000, 0.000)
Ni-Ni: 2.917600, ( 2.918, 0.000, 0.000)
Ni-Ni: 3.662494, (-0.668, -2.022, 2.980)
Ni-Ni: 3.662494, (-0.668, 2.022, 2.980)
"Ti" site at ( 0.083, 0.750, 0.282) :
=======================================
Ti-Ti bonds:
------------
Ti-Ti: 2.917600, ( 2.918, 0.000, 0.000)
Ti-Ti: 2.917600, (-2.918, 0.000, 0.000)
Ti-Ti: 2.949806, (-0.762, -2.022, 2.009)
Ti-Ti: 2.949806, (-0.762, 2.022, 2.009)
Ti-Ti: 3.286712, (-0.129, -2.022, -2.588)
Ti-Ti: 3.286712, (-0.129, 2.022, -2.588)
Ti-Ni bonds:
------------
Ti-Ni: 2.513032, ( 1.376, 0.000, -2.103)
Ti-Ni: 2.513175, ( 1.412, -2.022, -0.486)
Ti-Ni: 2.513175, ( 1.412, 2.022, -0.486)
Ti-Ni: 2.566964, (-1.506, -2.022, -0.486)
Ti-Ni: 2.566964, (-1.506, 2.022, -0.486)
Ti-Ni: 2.603043, ( 0.744, 0.000, 2.494)
Ti-Ni: 2.606943, (-1.541, 0.000, -2.103)
Ti-Ni: 3.308537, (-2.174, 0.000, 2.494)