[Pw_forum] Monolayer MoS2 band Structure

Giovanni Cantele giovanni.cantele at spin.cnr.it
Thu Sep 24 10:48:57 CEST 2015

Dear Cameron,

I don’t think that this strange behavior is due to mixing_beta or conv_thr, so you can remove those flags from the input. Moreover, the vacuum space you use (~27 A) is a bit large, you can slightly reduce it to speedup the calculation.

As far as I can see you should: 1) check if the lattice parameter is correct, for some materials it can happen that if you use a lattice parameter quite different from the equilibrium one the band structure undergoes more or less significant changes; 2) check your k-point path, are you sure you gave the correct k-points in the correct units? Maybe seeing the input for the nscf calculation would be useful as well.

Finally, even though I do not think that this might affect your band structure calculation, if you include one or more ultrasoft pseudo potentials in you calculations, as you do, typical values of ecutrho, depending on the atomic element and on the pseudo potential, can range from ~ 6*ecutwfc to even 12*ecutwfc, so your choice 300 Ry ~ 4.3 * 70 Ry might produce not fully converged results. 

Last, if I were you, I would replace  the norm conserving pseudo potential for S with an ultrasoft pseudo potential, which allows you to use much lower values of ecutwfc.

I attach a band structure of monolayer MoS2 that has been done using QE 5.2.0, where the direct gap at K (~1.74 eV) is evident.


> On 24 Sep 2015, at 08:20, Cameron Foss <cjfoss at umass.edu> wrote:
> Hello,
> I am attempting to obtain the band structure of monolayer MoS2, however my results so far do not agree well with what has been documented for monolayer (ML) MoS2. I consistently get an indirect band gap for ML MoS2 caused by the 'Q-valley' minimum seen in bulk MoS2.
> I have tested the code with interplanar distances of 15 and 30 angstroms to eliminate any interplanar interactions. There have been several papers that report the band structure for ML MoS2 from first principles, they have used functionals from LDA, GGA, and an HSE hybrid functional. My best success has been with GGA-PW91(see input file below). Note in all my simulations I have measured a band gap of ~1.9eV which is in agreement with reported values.
> I am uncertain as to what the issue could be? I have tested 3 different PPs. 
> %%%%%%%%%%%%%%%%%%%%
> Details on simulations:
>  Specifically, my simulations continue to show that the 'Q-valley' is slightly lower (16meV~30meV) than the minimum at the K point, suggesting an indirect band gap. However ML mos2 has been widely shown to have a direct band gap at the K point where this 'Q-valley' along the conduction band is suppressed above the K-point minimum.
> I run a scf (pw.x) calculation first, followed by an nscf (pw.x) run, and a bands calculation with pw.x (and of course bands.x to retrieve the energy values). I specify 16 bands for nbnd, kpoints along symmetry paths, and a MP grid of 8x8x1 with a 1 1 1 offset. 
> Input file: 
> &control
>     calculation='scf'                      !nscf, then bands
>     restart_mode='from_scratch',
>     !pseudo_dir='directory where pseudopotentials are stored/',
>     !outdir='directory where large files are written/'
>     pseudo_dir='/home/cameron/QE/espresso-5.1/pseudo/',
>     outdir='/home/cameron/QE/espresso-5.1/2dout'
>     prefix='mos2-gga',
>  /
>  &system
>     ibrav=4, celldm(1)=5.8364, celldm(3)=10,
>     nat=3, ntyp=2, ecutwfc =70, ecutrho=300         ! specified nbnd=16
>  /
>  &electrons
>     conv_thr =  1.0d-15                   ! reduced to 1.0d--12 for nscf, bands
>     mixing_beta = 0.5                     ! increased to 0.7 for nscf, bands
>  /
>  Mo  95.94    Mo.pw91-n-van.UPF
>  S   32.065   S.pw91-n-mt.UPF
> S        0.00002  -0.00001   0.000000000
> Mo       2.91822   1.68481   2.929886569
> S        0.00002  -0.00001   5.859870374
> K_POINTS automatic
>  8 8 1 1 1 1
> I ran into some convergence issues with all the bands with this conv_thr in the nscf and bands calculations, so i had to reduce it to 1.0d-12 for those calculations. I also used a mixing beta of 0.7 for both the nscf and bands calculations. I am uncertain as to whether changing mixing_beta's between calculations is of importance or not as well, but I do not think this is the case.
> Best regards,
> Cameron
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum


Giovanni Cantele, PhD
c/o Dipartimento di Fisica
Universita' di Napoli "Federico II"
Complesso Universitario M. S. Angelo - Ed. 6
Via Cintia, I-80126, Napoli, Italy
e-mail: giovanni.cantele at spin.cnr.it
Phone: +39 081 676910
Skype contact: giocan74

ResearcherID: http://www.researcherid.com/rid/A-1951-2009
Web page: http://people.na.infn.it/~cantele

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20150924/3a2b65ec/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1d_2--MoS2_1L_ecut45_ecutrho455_kpt8x8x1-GM-M-K-GM.bands.png
Type: image/png
Size: 26149 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20150924/3a2b65ec/attachment.png>

More information about the users mailing list