Top Page > Browsing

Mismatch DOS with conductance profile Date: 2019/08/19 14:22 Name: Devanand Dear OpenMx users, I am trying to run the negf example given in work directory under negf_example directory. I have done calculation for 8-ZGNR (Left-Lead and Right-Lead and negf) to calculate conductance. I got the correct profile of conductance (as given in openmx manual) however the DOS is not correct according to the conductance profile. DOS profile is not corroborating the conductance spectra. Could anyone explain to me why is it so. Thank you Regards, Dev

Page: [1]

Re: Mismatch DOS with conductance profile ( No.1 ) Date: 2019/08/23 14:20 Name: Devanand Dear Naoya Yamaguchi, As I have recently installed OpenMX and running some examples to get an idea. Meanwhile, I face the above problem. Could you please help me in this regard? Thank you, Regards, Dev Re: Mismatch DOS with conductance profile ( No.2 ) Date: 2019/08/24 14:40 Name: Naoya Yamaguchi Dear Dev, I don't have enough experiences and knowledges about NEGF calculations, and according to the manual: http://www.openmx-square.org/openmx_man3.8/node116.html , where there is a tip of "The numbers of k-points to discretize the reciprocal vectors b and c are specified by the keyword 'NEGF.Dos.Kgrid'. The set of numbers given by 'NEGF.Dos.Kgrid' tends to be larger than that by 'NEGF.scf.Kgrid' because of computational efficiency." I think that a dense k-grid is required for "computational accuracy" instead of "computational efficiency", and in fact the Japanese version of the manual says so. So, please try to increase the number of k-points for accurate DOS calculations through a keyword "NEGF.Dos.Kgrid". Regards, Naoya Yamaguchi

Page: [1]