wrong energy ordering???

Various technical topics such as parallelism and efficiency, netCDF problems, the Yambo code structure itself, are posted here.

Moderators: Davide Sangalli, andrea marini, Daniele Varsano, andrea.ferretti, Conor Hogan, myrta gruning

Post Reply
Posts: 14
Joined: Tue Oct 06, 2009 9:36 am

wrong energy ordering???

Post by giacsport » Thu Sep 22, 2011 5:14 pm

Dear All,
I am calculating the Hartree-Fock Self-energy and Vxc contribution for GaN.
I am using a s2p1 pseudo for Ga and s2p3 for N.

I honestly do not understand why I get this "disordered" ordering of the states (from 5 to 10, in this case) :?

I would expect indeed that deeper states should have lower energy. But in my case 5th state at HF level has higher energy than the 6th (still at HF)

And moreover, the 8th state at LDA is my Valence Band and the 9th is my CB. Should I expect the energy difference between these two states corresponding to the bandgap (i.e, about 2.35 eV)

Thanks in advance,
my very best.


[xc] Functional Perdew & Zunger (xc)

XC HF and LDA [eV] @ K [1] (iku): 0.000 0.000 0.000
<5|HF|5> = -19.96117 -.8670E-8 <5|LDA|5> = -16.25794 -.4089E-8
<6|HF|6> = -20.25342 -.1870E-7 <6|LDA|6> = -16.98509 0.4142E-8
<7|HF|7> = -20.34970 0.1262E-8 <7|LDA|7> = -17.03668 0.1352E-7
<8|HF|8> = -20.34970 0.9239E-8 <8|LDA|8> = -17.03672 0.2222E-8
<9|HF|9> = -8.991189 0.6053E-9 <9|LDA|9> = -13.03069 0.4721E-8
<10|HF|10> = -6.722379 0.1330E-8 <10|LDA|10> = -11.55897 0.6082E-9
<11|HF|11> = -7.562684 0.5999E-9 <11|LDA|11> = -14.41783 -.1089E-7
<12|HF|12> = -4.280390 -.7542E-9 <12|LDA|12> = -9.997887 0.8340E-9

Giacomo Giorgi, PhD.
The University of Tokyo.

User avatar
myrta gruning
Posts: 305
Joined: Tue Mar 17, 2009 11:38 am

Re: wrong energy ordering???

Post by myrta gruning » Fri Sep 23, 2011 11:05 am

Dear Giacomo

What you reported are not the energies, but the matrix elements of the Hartree-Fock self energy <i|HF|i> and of the xc potential <i|LDA|i>.
They are ordered following the KS energies ordering coming out from the DFT calculations. The KS eneries are also reported in the r_* file, together with the LDA gap.

If you want you can get a perturbative HF energy for the state i by calculating:
E_HF^i = E_KS^i -<i|LDA|i> + <i|HF|i>
those are not reported since HF does not give a good estimate for band gap anyway (far too large)
and this is just a step needed for getting the GW corrections.

I hope this clarifies things

Dr Myrta Grüning
Atomistic Simulation Centre
Queen's University Belfast - Northern Ireland


Posts: 14
Joined: Tue Oct 06, 2009 9:36 am

Re: wrong energy ordering???

Post by giacsport » Fri Sep 23, 2011 12:16 pm

Dear Myrta,
it definitely clarifies all my doubts.

Thanks for the prompt and kind reply.

My best,

Post Reply