Input for (Quasi-)Harmonic Approximation calculations¶

Last updated

Jul 25, 2024

Author

Gianfranco Ulian

The input for HA and QHA calculation used by Quantas is a YAML file (extension .yaml), which must contains serveral data that will be employed during the software run.

Each data type in the input is preceded by a keyword (in lower cases), which describes the kind of information that will be collected. The format employed is reported in the following:

keyword1:    value(s)
keyword2:    [ (array of) values(s) separated by commas ]
...

In the following table, the keywords employed by Quantas are reported, alongside a short description of their functionality.

Mandatory Keywords	Description
natom	Number of atoms in the unit cell
supercell	3x3 matrix related to the unit cell expansion
qpoints	Number of q points used to calculate phonon properties
volume	List of unit cell volumes over over which the phonon properties were calculated
energy	List of energy values related to each unit cell volume
phonon	List of phonon values

phonon sub-keywords	Description
q-position	Vector related to the specific q point
weight	Weight of the phonon band
band	Phonon band(s)

band sub-keywords	Description
# n	Band number
frequency	Array of natom phonon frequency values (float)

For the sake of an example, if you calculated the phonon dispersion relations for a FCC lattice containing 2 atoms in the primitive cell, the input file for the (Q)HA analysis will be something like the following:

natom:   2
supercell:
- [     -2,      2,      2 ]
- [      2,     -2,      2 ]
- [      2,      2,     -2 ]
qpoints: 656
volume: [  V1, V2, V3, ... Vn ]
energy: [  E1, E2, E3, ... En ]
phonon:
- q-position: [ q1_coord_1, q1_coord_2, q1_coord_2 ]
  weight: 1
  band:
  - # 1
    frequency: [ v1_1, v1_2, v1_3, ..., v1_n ]
  - # 2
    frequency: [ v2_1, v2_2, v2_3, ..., v2_n ]
  - # 3
    frequency: [ v3_1, v3_2, v3_3, ..., v3_n ]
  - # 4
    frequency: [ v4_1, v4_2, v4_3, ..., v4_n ]
  - # 5
    frequency: [ v5_1, v5_2, v5_3, ..., v5_n ]
  - # 6
    frequency: [ v6_1, v6_2, v6_3, ..., v6_n ]
- q-position: [ q2_coord_1, q2_coord_2, q2_coord_2 ]
  weight: 1
  band:
  - # 1
    frequency: [ v1_1, v1_2, v1_3, ..., v1_n ]
  - # 2
    frequency: [ v2_1, v2_2, v2_3, ..., v2_n ]
  - # 3
    frequency: [ v3_1, v3_2, v3_3, ..., v3_n ]
  - # 4
    frequency: [ v4_1, v4_2, v4_3, ..., v4_n ]
  - # 5
    frequency: [ v5_1, v5_2, v5_3, ..., v5_n ]
  - # 6
    frequency: [ v6_1, v6_2, v6_3, ..., v6_n ]
...

Note

A command for the automatic generation of the input file used by Quantas for the (quasi-) harmonic approximation analysis of crystalline solids is provided to aid the user. However, at the moment, it is compatible only with the output files of CRYSTAL14 (and above) and phonopy.

Keywords description¶

A detailed description of the data requested for each keyword is reported in the following.

natom¶

This is the number of atoms in the unit cell that was employed to perform the calculation of the phonon properties, either \(\Gamma\)-point only or phonon dispersion relations.

Example 1

You have calculated the phonon properties of NaCl by considering its crystallographic cell. In this case, you should write:

natom: 8

Example 2

Same as for Example 1, but instead you considered the primitive cell. In this case you should write:

natom: 2

Note

natom is an integer type. Each unit cell volume employed to calculate the phonon properties must contain the same number of atoms.

supercell¶

The keyword supercell tells Quantas the expansion matrix used to calculate phonon dispersion relations.

Example 1

If you have performed calculations using a \(2 \times 2 \times 2\) expansion matrix, you should write:

supercell:
- [      2,      0,      0 ]
- [      0,      2,      0 ]
- [      0,      0,      2 ]

Example 2

If the phonon calculations involved only \(\Gamma\)-point (i.e. for a large unit cell), you should write:

supercell:
- [      1,      0,      0 ]
- [      0,      1,      0 ]
- [      0,      0,      1 ]

Note

supercell is an integer type.

qpoints¶

The keyword qpoints represents the number of q points sampled in the reciprocal space and used to calculate phonon properties (i.e., phonon dispersion relations).

Note

qpoints is an integer value.

Warning

The number of q points indicated by qpoints must be consistent with the sum of the weights of each phonon band (see below).

volume/energy¶

The keywords volume and energy are each one followed by an array of \(n\) values.

Note

The values of volume and energy are float type. It is possible to consider a single-volume calculation \((n = 1)\), but at least \((n > 4)\) points are required to perform quasi-harmonic approximation calculations.

phonon¶

phonon represents the block of data containing the phonon properties of the material. phonon has some sub-keywords.

phonon sub-keywords
q-position	Vector related to the specific q point
weight	Weight of the phonon band
band	Phonon band(s)
band sub-keywords
# n	Band number
frequency	Array of phonon frequency values (float)

A brief explanation of each sub-keyword is here presented:

q-position: it represents one building block containing the phonon band calculated at that q point. It is a \(1 \times 3\) array containing the fractional coordinates of the sampled q* point.

weight (integer number): this sub-keyword is a child of q-position, representing the number of times that this phonon band is seen when the phonon properties have been calculated. This value is strictly related to the symmetry of the system under consideration.

band: this sub-keyword is another child of q-position, and begins the declaration of the phonon frequencies calculated at that q-position.

# n: child of band, it is simply a label of the phonon frequency.

frequency: \((1 \times natoms)\) array of float values.