Bandlinescale is not given in input file thus default value is used (pi/a).
Correct me if I'm wrong, but if it were a matter of Bohr/Angstrom conversion we would have lattice vectors and bands lines given with respect to a lattice constant defined with different units, isn't it? As far as I know LatticeVectors block's values as well as BandLines block's values (with bandlinescale=pi/a ) have no unit, they just scale LatticeConstant and pi/LatticeConstant respectively. This means that if the real cell is ~ 2.82 times larger, reciprocal cell should be 1/2.82 shorter, correct? Or is there something elementary I'm missing?! Thanks for your attention, kind regards, Pietro 2011/2/10 Marcos Veríssimo Alves <[email protected]> > > Bohr / Angstrom conversion? > > On Thu, Feb 10, 2011 at 3:47 PM, pietro bonfa <[email protected]> wrote: >> >> Dear Siesta users, >> >> I have a problem that I wasn't able to overcome despite it has been largely >> discussed in this mailing list. >> The following BandLines block is from the input file that have been given >> to me by Prof. Felix Yudarin: >> >> LatticeConstant 1.98 Ang >> >> %block LatticeVectors >> 2.828427 0.000000 0.000000 >> 0.000000 2.828427 0.000000 >> 0.000000 0.000000 4.3387217 >> %endblock LatticeVectors >> >> %block BandLines >> 1 0.35355 0.000000 0.000000 X >> 100 0.00000 0.000000 0.000000 \Gamma >> 100 0.35355 0.353550 0.000000 M >> 100 0.35355 0.000000 0.000000 X >> 100 0.00000 0.000000 0.000000 \Gamma >> 100 0.00000 0.000000 0.230809 R >> %endblock BandLines >> >> >> I can't figure out the calculation leading to 0.35355 for X and M point in >> brillouin zone (and or course 0.230809 for R) . >> >> Here's how I calculate that value: >> >> X is [2*(pi)/a, 0, 0], where a=2.828427*LatticeConstant is real lattice >> constant. >> >> Thus I'm expecting 2/2.828427 = 0.70710 to be the value I should put in >> BandLines block. The 0.35355 value is instead about half my value. >> (first brillouin zone can be found here: >> http://journals.iucr.org/b/issues/2010/01/00/gw5003/gw5003fig1.jpg) >> >> What am I missing? >> >> Before concluding this email I want to thank all the people contributing to >> this mailing list: I couldn't get much far if it hadn't been for all the >> detailed explanation given here. >> >> Best regards, >> Pietro >
