<html><body><p>Dear Wien2k users,<br></p><p>I came across a problem with equilibrium atomic volume of </p><p>the BiNi compound. The experimental lattice is hexagonal <br></p><p>with a = 4.079 Angstroem, c = 5.359 Angstroem <br></p><p>(P. Villars, Pearson's Handbook: Crystallographic Data for Intermetallic Phases)</p><p>However, the equilibrium volume turns out to be more <br></p><p>than 16 % higher than the experimental one. <br></p><p>I wonder since the equilibrium volume of <br></p><p>pure Bi and Bi3Ni comes out with much better agreement with <br></p><p>experiment (about 4 to 5 % deviation). <br></p>I used GGA (no spin orbit coupling), <p>Rmt*Kmax = 8.8, lmax = 10, Gmax = 16, 5000 k-points in the <br></p><p>whole Brillouin zone. I enclosethe structure file in attachment. <br></p><p>I tried LDA that gives better agreement with experiment <br></p><p>(about 10 % deviation) but I think this is still too much. I have tried <br></p><p>to use gaussian smearing instead of the tetrahedron method,<br></p>increase Rmt*Kmax to 11, increase k-points to 20 000 in the whole <br>Brillouin zone but nothing helped. <br>In the mailing list I found someone had similar problem with a more <br>complicated structure containing bismuth, but that was not solved:<br>http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg10479.html<br>Do you have any idea? <br>Thank you in advance <br>With best regards <br>Tomas Kana <br>Institute of Physics of Materials,<br>Academy of Sciences of the Czech Republic <br>Brno, Czech Republic <br></body></html>