Dear fellow Siesta users,

I am trying to calculate the I-V curve of a zigzag graphene nanoribbon
(ZGNR) system, using spin-polarized LDA xc and DZP basis. I have tried
various schemes of DM mixing, electronic temperature, Occupation Methods and
such, to no avail. I use 3 nanoribbon unit cells are electrode unit cell
(3UC) and 3+5+3 = 13 ZGNR unit cells in the scattering region.

The initial SR converges with the creating of a DM, but the transiesta step
show a tendency such as

TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 392   -28439.8239   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 393   -28439.8248   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 394   -28439.8232   -28439.8262   -28439.8262  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 395   -28439.8249   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 396   -28439.8249   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 397   -28439.8250   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 398   -28439.8250   -28439.8262   -28439.8262  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 399   -28439.8250   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 400   -28439.8250   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 401   -28439.8249   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 402   -28439.8249   -28439.8261   -28439.8261  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 403   -28439.8248   -28439.8262   -28439.8262  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 404   -28439.8251   -28439.8262   -28439.8262  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 405   -28439.8251   -28439.8263   -28439.8263  0.0000 -4.2444
TranSiesta: Qsol,Qtot:  750.000  750.000    1.000
transiesta: 406   -28439.8240   -28439.8264   -28439.8264  0.0000 -4.2444


which seems to be converging nowhere.

The 6-ZGNR has a bandgap of order of 0.25 eV, which appears when spin states
are locailzed at the edges. I start my runs with the edge Carbon atoms spin
polarized, edges polarized Anti-Ferromagnetically.

Does anyone have an idea why such a loop exists and persists when I alter DM
mixing parameter, temperature etc.

I will send my input files and other details for those interested.

Have a nice day.

Kemal Bagci

Responder a