<div dir="ltr"><div><br></div><div>Dear Wannier90 experts, </div><div><br></div><div>I just posted this question on quantum-espresso's archive but it might be more helpful to ask it again here, </div><div><br></div><div>I am trying to calculate the maximally localised wannier functions of defective bilayer graphene using quantumEspresso and Wannier90 codes, </div><div><br></div><div>however I am getting a band gap of about 40 to 60 meV for pristine bilayer graphene using a variety of NCPP+LDA and NCPP+PBE, the band gap and the interlayer distance are slightly improved (10 meV and 0.1 A ) after applying DFT-D3 but not significantly, I wonder how I can fix this issue (please see the sample setup that follows) ?</div><div><br></div><div>Please also note that I get an ideal zero band gap when using ultrasoft pseudopotential but it seems that Wannier90 has a preference for NCPP and I cannot get a nice fit to the band structure calculated using USPP, while the calculated MLWF for
NCPP+PBE are amazingly perfect (total Omega = 6 A^2)</div><div><br></div><div>Best regards,<br></div><div>Roozbeh Anvari,</div><div>PDF, Physics Dept, Queen's University, Canada </div><div><br></div><div>Here is a shortened sample of the setup I use for vc-relax/scf calculations</div><div><br></div><div>&CONTROL<br>...</div><div> forc_conv_thr=1.0d-6,<br><br>&SYSTEM<br>...</div><div>occupations='smearing', smearing='cold', degauss=0.02<br>vdw_corr='grimme-d3' , input_dft='PBE'<br><br></div><div>&ELECTRONS<br>startingwfc='random', diagonalization='cg', conv_thr = 1.0e-12<br>mixing_beta = 0.2<br>/<br> &IONS<br> ion_dynamics='bfgs', upscale=20.0<br><br>&CELL<br> press_conv_thr = 0.5D0, press = 0.D0<br>cell_dynamics = 'bfgs',cell_dofree = '2Dxy', cell_factor = 1.5D0<br><br>ATOMIC_SPECIES<br>C 12.011 C.pbe-mt_gipaw_NC_PBE.UPF </div><div><br>K_POINTS automatic<br>36 36 1 0 0 0<br></div><div><br></div><div>Cheers <div class="gmail-yj6qo"></div><div class="gmail-adL"><br></div></div></div>