wrong QP correction energy

[Yu Xie]

Dear all,

To get the QP-corrected band structure, I'm running G0W0 calculations for 2D NiSeO, NiTeO, NiSeS, and NiTeS. I have successfully generated the correct band structures for NiSeO, NiTeO, and NiSeS monolayers. But I cannot get the correct one for NiTeS monolayer. I have tried two different combinations of bands and NGsBlkXp, namely 200_4Ry and 300_5Ry. The results look like the following:

200_4Ry:
# K-point Band Eo [eV] E-Eo [eV] Sc|Eo [eV]
#
1 38 -0.398820 0.218203 6.154930
1 39 -0.304721 0.214727 6.243649
1 40 -0.001477 0.042638 5.170319
1 41 1.131394 0.505985 -1.359906
1 42 1.820613 0.006978 -3.157616
1 43 2.373251 0.107993 -3.148714
2 38 -0.396685 0.208641 6.194119
2 39 -0.302747 0.207068 6.171379
2 40 -0.012805 0.020562 5.093807
2 41 1.119537 0.489282 -1.426847
2 42 1.816585 0.026809 -3.095500
2 43 2.359188 0.122458 -3.124007
3 38 -0.394021 0.189293 6.311312
3 39 -0.295914 0.179744 5.982427
3 40 -0.024818 -0.032686 4.884567
3 41 1.093585 0.445260 -1.616055
3 42 1.794522 0.087551 -2.931261
3 43 2.324766 0.159110 -3.064402
4 38 -0.401620 0.176196 6.437999
4 39 -0.283711 0.137038 5.804931
4 40 -0.011454 -0.078962 4.626990
4 41 1.080941 0.388902 -1.872964
4 42 1.725226 0.183740 -2.708758
4 43 2.288188 0.202969 -3.002927
5 38 -0.430250 0.174395 6.562064
5 39 -0.263050 0.107833 5.667802
5 40 0.000000 -0.082984 4.477935
5 41 1.113263 0.336847 -2.090808
5 42 1.600930 0.261788 -2.509344
5 43 2.268354 0.242487 -2.966255

300_5Ry:
# K-point Band Eo [eV] E-Eo [eV] Sc|Eo [eV]
#
1 38 -0.398847 0.232347 6.175115
1 39 -0.304752 0.228924 6.264064
1 40 -0.001503 0.029197 5.152659
1 41 1.131346 0.444725 -1.442808
1 42 1.820566 -0.123118 -3.327800
1 43 2.373260 -0.025128 -3.323077
2 38 -0.396712 0.221830 6.213422
2 39 -0.302777 0.219107 6.189129
2 40 -0.012830 0.003643 5.071308
2 41 1.119488 0.425629 -1.512621
2 42 1.816537 -0.099447 -3.260670
2 43 2.359195 -0.008730 -3.295753
3 38 -0.394047 0.201837 6.329479
3 39 -0.295943 0.186295 5.992393
3 40 -0.024839 -0.058952 4.848932
3 41 1.093531 0.375774 -1.709294
3 42 1.794476 -0.028803 -3.083808
3 43 2.324770 0.032871 -3.229942
4 38 -0.401648 0.188279 6.455839
4 39 -0.283735 0.137718 5.806072
4 40 -0.011466 -0.116325 4.576202
4 41 1.080882 0.312216 -1.975443
4 42 1.725186 0.079264 -2.846224
4 43 2.288187 0.081997 -3.162017
5 38 -0.430282 0.187260 6.581142
5 39 -0.263062 0.104194 5.662995
5 40 0.000000 -0.125753 4.419808
5 41 1.113201 0.253586 -2.201235
5 42 1.600892 0.165611 -2.636062
5 43 2.268345 0.125065 -3.120641

It looks like that all bands are upshifted by hundreds of meV.

Any help is appreciated!

Thank you all.

Best, Yu

[Daniele Varsano]

Dear Yu,

please sign your post with your name and affiliation, this is a rule of the forum, you can do it once for all by filling the signature in your user profile.
I'm not sure I well understood your question, can you repeat more explicitly what is the problem you are facing?

But I cannot get the correct one for NiTeS monolayer.

What do you mean by correct, with respect to previous calculations in the literature, experiments?

It looks like all bands are upshifted by hundreds of meV.

Again, upshifted with respect to what?

Looking at k point number five I can see:

200_4Ry: the gap is 1.533
300_5Ry: the gap is 1.492

so it seems it is not yet completely converged.

Best,
Daniele

[Yu Xie]

Dear Daniele,

Thanks for your reply and sorry for the unclear description.

What I mean is that the shape of the GW band is quite different from the PBE one.

Anyway, I found the problem is that the EXXRLvcs parameter is not converged.

Thank you!