c2x and Spinors

This page serves to give a trivial example of using Castep, c2x and spinors. GaAs is used as an example. From version 2.30 of c2x the same analysis can be performed on the output of Abinit or QE. The text below describes using Castep, but a corresponding input file for QE would be GaAs.in.

For the .cell file, the following suffices:

 2.8268500  2.8268500  0.0000000
 2.8268500  0.0000000  2.8268500
 0.0000000  2.8268500  2.8268500
%endblock LATTICE_CART

 Ga  0.000000000  0.000000000  0.000000000
 As  0.250000000  0.250000000  0.250000000

As  Asrel.RRKJ3.UPF
Ga  Garel.RRKJ3.UPF
%endblock SPECIES_POT

KPOINT_MP_OFFSET 0.0 0.0 0.0

and for the .param file:

xc_functional   LDA
spin_treatment VECTOR
spin_orbit_coupling true
fix_occupancy   false
popn_calculate  false

One has to provide suitable pseudopotentials, for the OTF ones currently won't do. So they are provided here as Asrel.RRKJ3.UPF and Garel.RRKJ3.UPF. And that is it: one should then be able to run "castep GaAs". The .castep file should say "treating system as non-collinear spin".

Finally, a simple example of c2x interpreting the final output:

$ c2x -B=3 -v --null GaAs.check 
Castep version 16.110
Wavefunction grid     24 24 24
Cell volume 45.179174
natoms      2
Requested cut-off energy 408.171 eV
First FFT grid     24 24 24
spins=1   spinors=2
Found 3D data for band_vs0_k1_b3
  min=2.70772e-05  max=0.108625  sum=158.081  int=0.516635
Found 3D data for band_vs1_k1_b3
  min=2.56075e-05  max=0.101641  sum=147.901  int=0.483365

(Extract, as density, band number three, k-point 1 (implicit, modify with -k=), both spinor components (implict, modify with -S=), produce no output other than the above information about extrema and integrals.) Note that the sum of the two integrals is 1.0 exactly as norm-conserving pseudopotentials were used.

(Note a change in Castep's .check file format introduced in Castep 18 caused the reading of spinor components to fail until c2x was modified at version 2.34c.)