Optimization of any bulk crystal

[Frank123]

hello sir,
            I am a beginner. I want to know the full procedure to optimize any bulk crystal using ATK.

[Ulrik G. Vej-Hansen]

Please use our extensive documentation - this was found with a simple search: https://docs.quantumwise.com/tutorials/geometry_optimization/geometry_optimization.html

[Frank123]

Thank you for your help.
 I have optimized the internal coordinates of a bulk within a fixed volume using the optimizer . i want to do c/a optimization, and then the energy minimization by taking different volumes using the optimized c/a ratio. But my question is how can i optimize c/a ratio. please help

[Ulrik G. Vej-Hansen]

Why do you want to split it up like that? The easiest way to optimize the volume is to do it at once, using stress. You can do that by un-ticking "Constrain Lattice Vectors" in the Optimize Geometry widget. In this way, you can optimize both the internal coordinates and the length of the cell vectors at once, while keeping the Bravais Lattice constrained.

[Frank123]

Yes Sir i have tried the one you are suggesting. But i need to fit the Murnaghan EOS with different volumes.

[Jess Wellendorff]

Consider HCP ruthenium as an example:

Code

# Set up lattice
lattice = Hexagonal(2.7058*Angstrom, 4.2816*Angstrom)

# Define elements
elements = [Ruthenium, Ruthenium]

# Define coordinates
fractional_coordinates = [[ 0.333333333333,  0.666666666667,  0.25          ],
                          [ 0.666666666667,  0.333333333333,  0.75          ]]

# Set up configuration
bulk_configuration = BulkConfiguration(
    bravais_lattice=lattice,
    elements=elements,
    fractional_coordinates=fractional_coordinates
    )

# Get c/a
lattice = bulk_configuration.bravaisLattice()
a = lattice.a()
c = lattice.c()
c_over_a = c/a
print c_over_a

which yields c/a = 1.5823785941311257 in this experimental geometry.

[Petr Khomyakov]

But i need to fit the Murnaghan EOS with different volumes.

You may then do a set of ion relaxation calculations for your crystal structure with different unit cell volumes, keeping the optimized c/a fixed; you certainly have to first calculate this ratio, as suggested by Ulrik. It means that you only need do ion relaxation for a given unit cell volume. In this case, you have to keep "Constrain Lattice Vectors" ticked in geometry optimization settings.

[Frank123]

Thank you. I will try that.