Negative value of superconducting gap

[simba2828]

Dear Sir,

Yes, I checked it and found that the Wannier bands are very well fitted with DFT bands. But the spreads for some of the atoms are large. Attaching the spreads and the Wannier bands. Please check.

Code: Select all

 Running Wannier90

     Wannier Function centers (cartesian, alat) and spreads (ang):

     (   0.49925   0.28886   0.98713) :   1.73951
     (   0.53545   0.30514   3.33642) :   2.64991
     (  -0.00461   1.03417   3.25670) :   3.59971
     (  -0.85892   0.16977   2.79665) :   4.87394
     (   0.75361  -0.23561   1.35920) :   9.86762
     (   0.00722   0.56533   0.57935) :   2.71592
     (  -0.03645   0.55981   1.43628) :   2.57573
     (   0.51684  -0.19334   1.32163) :   3.47311
     (  -0.46746   0.29874   0.98754) :   4.32743
     (   0.42678   0.33296   3.27519) :   2.99586
     (   0.01929   0.57143   0.55829) :   2.63779
     (   0.49617  -0.25639   0.59628) :   3.11302
     (  -0.02773   1.18478   2.44713) :   2.85141
     (  -0.07989   0.96564   2.58605) :   4.47229
     (   0.17657   0.67448   1.22342) :  14.09387
     (   0.47358   0.28139   0.98852) :   3.57051
     (   1.01395  -0.52679   2.50908) :   3.37114

[gkafle1]

Hi simba2828,
Although your Wannier bands fit well with the DFT bands, I see some Wannier functions have a very large spread. Please try to get good Wannier functions with a smaller spread and redo the calculation. This may help.

Thanks!

Gyanu

[simba2828]

Hello,
Yes I tried, sometimes I get the minimum spread and sometimes there are large spreads. Even then the negative value of Min Superconducting gap appears. Could you please suggest how can I make them small. It is like a black-box, and if I have to run it again and again it will take forever to get the correct combination. I would like to mention that with muc=0.0 the negative value is not there and it becomes zero, and in that case the spread for some of the Wannier functions are even larger.

With muc= 0.0, the Wannier spreads:

Code: Select all

Wannier Function centers (cartesian, alat) and spreads (ang):

     (   0.99576  -0.55942   0.96698) :   4.75716
     (   0.98895   0.53141   0.63876) :   3.36597
     (   0.06603   1.11252   1.00957) :   6.20847
     (   1.19577   0.47692   2.85311) :   6.64885
     (   1.28888   0.94065   1.21157) :  17.10356
     (  -0.02071   0.58370   0.58690) :   2.71064
     (   0.50465   0.29031   2.43178) :   2.96794
     (   0.00196   0.61274   0.54904) :   2.51828
     (  -0.00549   0.67022   1.32574) :   3.28729
     (   0.43038   0.25077   2.60033) :   3.92173
     (   0.89853   0.50733   1.29144) :   3.46119
     (  -0.53499   0.32661   3.27744) :   2.99358
     (   0.53697   0.29608   3.33856) :   2.67764
     (  -0.15729   1.05387   2.68383) :   4.59482
     (   0.07021   0.57738   1.42165) :   4.19832
     (   0.02683   0.77062   3.02112) :   5.61935
     (  -0.42265   0.28234   1.02944) :   6.54702

Do you think, this negative superconducting gap value can affect my electron-phonon coupling strength, a2F and the superconducting Tc?

Also, in an another thread, I mentioned that my phDOS and a2F are not consistent with each other, what does it signify? I am attaching the figure here.

Shubham

[gkafle1]

Hi,
I think you can try to play with the different projections, coarse k-meshes, and frozen windows to see if you can get good Wannier functions with a smaller spread.
Thanks!

Gyanu