The ABINIT Users Mailing List ( CLOSED )

[Zeila Zanolli <zeila.zanolli@uclouvain.be>]

Dear Wang,

yes that gives the gap. I was just asking if the LDA gap remains 0 also when the system is described with the metallic occupation. 

(how much is it???)

If it is not 0, then there is hope: you could increase the k-mesh in the LDA calculation (and trying a non-shifted grid as well), 

and increase the number of bands, ecutwfn and ecuteps in the screening and GW correction datasets, 

or attempt some selfconsistency for the GW part...

By the way, if the LDA gap is 0 without the In 4d states, it will be 0  (or, to be more precise, "negative") also with the  In 4d states.

Adding the "d" electrons is necessary to accurately describe the system, but this will make worse the LDA gap problem,

as it has been explained in the PRB paper I've mentioned you.

I know that you can get a gap with VASP but not with ABINIT, but this is mostly due to the different pseudopotential used.

A that time, I've also tried the scissor operator method with abinit, but i could not get a gap in LDA.

Despite I have published the work with the VASP calculation, I have to say that the ABINIT result is the correct one:  no gap in LDA for InAs.

Indeed, I have also performed all-electrons calculation for InAs with the WIEN2K code:

I have obtained a negative gap in LDA for scalar relativistic (as you did in ABINIT and VASP) and almost fully relativistic calculations.

The only situation where you can have a gap is for non relativistic calculations.

All these things are reported in another paper:

 arXiv:cond-mat/0610066 

 All-electron study of InAs and GaAs wurtzite: structural and electronic properties 

The point is that you need to start the GW calculation from an LDA  one (or from an intermediate calculation, as in PRB 75, 245121)

with a gap for the InAs and the correct occupation of the states, otherwise GW cannot do miracles for you as it did not make them for me! 

All the best,

Zeila

On 30 Mar 2008, at 23:09, wxw079000@utdallas.edu wrote:

Dear zeila,

Thank you for the reply, and I also noticed this paper, which is a very good one.PhysRevB vol 75 (2007) 245121 "Model GW band structure of InAs and GaAs in the wurtzite phase"

In my calculation, I use 3 electrons as valence without 4d electrons. As you said, I will try 13 valence electron which is really timeconsuming. 

To find the LDA gap in my calculation, I check it from the following information, Is that the right way to find it?

##################

Perturbative Calculation

 k =    0.000   0.000   0.000

 Band     E0 <VxcLDA>   SigX SigC(E0)      Z dSigC/dE  Sig(E)    E-E0       E

    4  -0.452 -10.890 -11.914   0.900   0.800  -0.249 -10.989  -0.099  -0.551

    5  -0.452 -10.890 -10.499  -0.507   0.798  -0.253 -10.983  -0.093  -0.545

 E^0_gap          0.000

 E^GW_gap         0.006

 DeltaE^GW_gap    0.006

############

Best wishes,

Weichao Wang

---------------------------------------------------------------------------------------------

Dr. Zeila Zanolli

Université Catholique de Louvain (UCL)

Unité Physico-Chimie et de Physique des Matériaux (PCPM) 

Place Croix du Sud, 1 (Boltzmann)

B-1348 Louvain-la-Neuve, Belgium

Phone: +32 (0)10 47 3501 

Mobile: +32 (0)487 556699

Fax: +32 (0)10 47 3452

e-mail: zeila.zanolli@uclouvain.be

web: http://www.nano2hybrids.net/

---------------------------------------------------------------------------------------------