Dynamical matrix and phonon frequency

Running Abinit Ground state
1

Hi all,

I was doing few phonon calculations at finite-q points. In the output file we get the dynamical matrix

 Dynamical matrix, in cartesian coordinates,
   if specified in the inputs, asr has been imposed
     j1       j2             matrix element
  dir pert dir pert     real part    imaginary part

To get the phonon energy in Hatree, do we just take the square root of the eigen values of the dynamical matrix obtained from the above segement?

The values I obtained were different when I manually tried to diagonalise this matrix compared to the phonon energies obtained in the ABINIT run. Is my understanding correct? I am attaching the ABINIT output file and the python script I was using to get the eigen values here.
phonon.abo (66.0 KB)
dynamical.py.txt (3.6 KB)

Thanks
Dominic

2

I realized the issue was with the unit conventions — the ion masses were given in atomic mass units (amu), while the force constants were expressed in Hartree/bohr². In Hartree units, mass is defined relative to the electron mass, so 1 amu~ 1840 me.
When converting consistently, taking the square root of this mass ratio brings the phonon frequencies into perfect agreement with the values obtained from ABINIT.

3

Hi Dominic,

Just a comment to your post: ABINIT uses atomic units across the board unless specified otherwise.

Good luck!

Bogdan