<div dir="ltr"><div>Dear Michal</div><div><br></div><div>Thanks for your reply.</div><div><br></div><div><span style="text-align:left;color:rgb(0,0,0);text-transform:none;text-indent:0px;letter-spacing:normal;font-family:arial,sans-serif;font-size:13.33px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;float:none;background-color:rgb(255,255,255)">I have not tried the newest version. Although I did look for emails stating problems including Efield in older versions of QE and I did not find any. So I'm assuming for now that my problem is not due to the QE version.
<span style="text-align:left;color:rgb(0,0,0);text-transform:none;text-indent:0px;letter-spacing:normal;font-family:arial,sans-serif;font-size:13.33px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;float:none;background-color:rgb(255,255,255)">I have tried several pseudopotentials:<br>1) PAW PBE: Mo.pbe-spn-kjpaw_psl.1.0.0.</span><span style="text-align:left;color:rgb(0,0,0);text-transform:none;text-indent:0px;letter-spacing:normal;font-family:arial,sans-serif;font-size:13.33px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;float:none;background-color:rgb(255,255,255)">UPF, S.pbe-n-kjpaw_psl.1.0.0.UPF<br>2) PAW PBEsol: Mo.pbesol-spn-kjpaw_psl.1.0.0.</span><span style="text-align:left;color:rgb(0,0,0);text-transform:none;text-indent:0px;letter-spacing:normal;font-family:arial,sans-serif;font-size:13.33px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;float:none;background-color:rgb(255,255,255)">UPF, S.pbesol-n-kjpaw_psl.1.0.0.UPF<br>3) USPP PBE: Mo.pbe-spn-rrkjus_psl.1.0.0.</span><span style="text-align:left;color:rgb(0,0,0);text-transform:none;text-indent:0px;letter-spacing:normal;font-family:arial,sans-serif;font-size:13.33px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;float:none;background-color:rgb(255,255,255)">UPF, S.pbe-nl-rrkjus_psl.1.0.0.UPF<br>4) PW LDA: Mo.pw-mt_fhi.UPF, S.pw-mt_fhi.UPF<br>5) USPP LDA: Mo.pz-spn-rrkjus_psl.0.2.UPF, S.pz-n-rrkjus_psl.0.1.UPF
<br>All these trials tend to agree relatively well, meaning that the change in bandgap with Efield from all the simulations is within 10-20%.<br>The change in bandgap with Efield for LDA pseudopotentials tends to be a bit slower than that for GGA pseudopotentials.<br><br>The cut off energy I'm using (80 Ry) is around or above what other calculations in the literature are using. So I'm assuming that increasing the cut off energy will not be enough to increase the change in bandgap with Efield that I'm getting by four times (what is obtained with VASP). Also, as I mention before, my QE calculations agree decently with other independent calculations in the literature (Ref [3] and [4] of the initial email).
<span style="text-align:left;color:rgb(0,0,0);text-transform:none;text-indent:0px;letter-spacing:normal;font-family:arial,sans-serif;font-size:13.33px;font-style:normal;font-variant:normal;font-weight:400;text-decoration:none;word-spacing:0px;display:inline;white-space:pre-wrap;float:none;background-color:rgb(255,255,255)">I choose the PBE results because the bandstructrure at Efield=0 resembles other bandstructures in the literature.
Is there any other issue that you think may be causing the discrepancy?
Thanks
Carlos
Scientist at Sivananthan Laboratories</span>
</span></span><b></b><i></i><u></u><sub></sub><sup></sup><strike></strike><b></b><i></i><u></u><sub></sub><sup></sup><strike></strike><br></div><div><b></b><i></i><u></u><sub></sub><sup></sup><strike></strike><b></b><i></i><u></u><sub></sub><sup></sup><strike></strike><br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;padding-left:1ex;border-left-color:rgb(204,204,204);border-left-width:1px;border-left-style:solid">
<br>
Dear Carlos,<br>
First of all, version 5.4 is ancient. Have you tried with the latest QE?<br>
Assuming both calculations are performed correctly, the difference may stem<br>
from different pseudopotentials. Are you using some old norm-conserving<br>
pseudos? Are you sure your cutoff is large enough? Have you tried SSSP<br>
pseudos?<br>
Anyway, are you expecting reasonable results for this system from PBE?<br>
Best regards,<br>
Michal Krompiec, Merck Electronics<br>
<br>
On Mon, Mar 8, 2021 at 6:40 PM Carlos Polanco Garcia <<a href="mailto:cap3fe@virginia.edu" target="_blank">cap3fe@virginia.edu</a>><br>
wrote:<br>
<br>
> Dear all<br>
><br>
><br>
><br>
> *Problem in brief:* My calculations of the change in the bandgap with<br>
> electric filed in bilayer MoS2 using Quantum Espresso (QE) version 5.4.0 do<br>
> not agree with similar calculations in the literature that used VASP.<br>
><br>
><br>
><br>
> *Questions:*<br>
><br>
> 1. Have anyone found discrepancies of the change of dispersion with<br>
> electric filed from QE and VASP in slab systems including Electric fields<br>
> with the sawtooth potential?<br>
> 2. Is there any problem with my input?<br>
><br>
><br>
><br>
> *Details of the simulation:* An input for bilayer MoS2 with an Electric<br>
> field of 2V/nm is below. I set the discontinuity of the sawtooth potential<br>
> to be in vacuum and I set the magnitude of the electric filed using the<br>
> conversion factor 1 a.u. = 51.4220632*10^10 V/m.<br>
><br>
><br>
><br>
> &control<br>
><br>
> calculation='scf',<br>
><br>
> outdir = '/home/tmp/MoS2.2L.E2.01/'<br>
><br>
> prefix='MoS2.2L',<br>
><br>
> pseudo_dir = '/home/pseudo/',<br>
><br>
> restart_mode='from_scratch',<br>
><br>
> tprnfor=.true.<br>
><br>
> tefield=.true.<br>
><br>
> /<br>
><br>
> &system<br>
><br>
> ibrav=4,<br>
><br>
> celldm(1)= 6.02820958<br>
><br>
> celldm(3)= 6.629172038<br>
><br>
> nat=6,<br>
><br>
> ntyp= 2,<br>
><br>
> ecutwfc= 80.0,<br>
><br>
> nbnd=24<br>
><br>
> vdw_corr='DFT-D'<br>
><br>
> edir=3<br>
><br>
> eamp = 0.00388938<br>
><br>
> emaxpos=0.60<br>
><br>
> eopreg=0.10<br>
><br>
> /<br>
><br>
> &electrons<br>
><br>
> conv_thr= 1.0d-9,<br>
><br>
> mixing_beta= 0.6,<br>
><br>
> /<br>
><br>
><br>
><br>
> ATOMIC_SPECIES<br>
><br>
> Mo 95.94 Mo.pbe-mt_fhi.UPF<br>
><br>
> S 32.065 S.pbe-mt_fhi.UPF<br>
><br>
><br>
><br>
> ATOMIC_POSITIONS {alat}<br>
><br>
> Mo 0.000000000 0.000000000 0.000000000<br>
><br>
> S 0.500000000 0.288675135 -0.492491069<br>
><br>
> S 0.500000000 0.288675135 0.492515463<br>
><br>
> Mo 0.500000000 0.288675135 1.969804069<br>
><br>
> S -0.000000000 -0.000000000 1.477442788<br>
><br>
> S -0.000000000 -0.000000000 2.462449934<br>
><br>
><br>
><br>
> K_POINTS {automatic}<br>
><br>
> 13 13 1 0 0 0<br>
><br>
><br>
><br>
> *Details of the problem:* Overall the dispersion for Efield=0 agrees<br>
> decently with other calculations in the literature including those using<br>
> VASP. However when I increase the electric field, my calculations predict a<br>
> change in the Bandgap with electric field about four times weaker than that<br>
> predicted from VASP calculations. For instance, from E0=0V/nm to E2=2V/nm,<br>
> my bandgap (Gamma to K) decreases by 0.19 eV. On the other hand, similar<br>
> VASP calculations predict that the bandgap decreases by 0.89 eV (Figure 2,<br>
> red dispersions from [1]), and by 0.73 eV (Figure 5a from [2]). My<br>
> calculations agree better with calculations that do not relay on VASP. An<br>
> independent calculation with QE predicts a similar slope in the change of<br>
> Bandgap with Efield for Efileds larger than 1V/nm (red dots Fig 3a [3]). A<br>
> calculation with Siesta predicts a decrease in bandgap of 1.04 eV when<br>
> increasing Efiled from 0V/nm to 5V/nm (Fig 4 b,d orange lines [4]), while<br>
> my calculations predict a change in bandgap of 0.78 eV (20% difference).<br>
> The agreement of the dispersions at Efield=0V/nm points that the problem is<br>
> in how the Efiled is being included. My input for QE may be missing<br>
> something, or there may be a factor difference in how Efield is included<br>
> between VASP and QE.<br>
><br>
><br>
><br>
> *References:*<br>
><br>
> [1] Ramasubramaniam A. et al, ?Tunable band gaps in bilayer<br>
> transition-metal dichalcogenides?, PHYSICAL REVIEW B 84, 205325 (2011)<br>
><br>
> [2] Chu T. et al, ?Electrically Tunable Bandgaps in Bilayer MoS2?, Nano<br>
> Lett. 2015, 15, 8000?8007<br>
><br>
> [3] Nguyen C. et al, ?Band Gap Modulation of Bilayer MoS2 Under Strain<br>
> Engineering and Electric Field: A Density Functional Theory?, Journal of<br>
> ELECTRONIC MATERIALS, Vol. 45, No. 8, 2016<br>
><br>
> [4] Santos E., et al, ?Electrically Driven Tuning of the Dielectric<br>
> Constant in MoS2 Layers?, ACS NANO, VOL. 7, NO. 12, 10741?10746 ? 2013<br>
><br>
><br>
><br>
> Carlos Polanco<br>
><br>
> Scientist at Sivananthan Laboratories<br><br></blockquote></div>-- <br><div class="gmail_signature" dir="ltr"><div dir="ltr"><div><div dir="ltr"><div dir="ltr"><div>Carlos Andrés Polanco García</div></div></div></div></div></div></div>