This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision | ||
smeagol_usage [2011/12/02 12:07] prokop |
smeagol_usage [2012/11/30 15:21] |
||
---|---|---|---|
Line 1: | Line 1: | ||
- | |||
- | ====== Tutorial in PDF ====== | ||
- | |||
- | |||
- | |||
- | There is simple presentation of smeagol usage presented on Smeagol Workshop in Hungary | ||
- | |||
- | {{:smeagol:smeagol_tutorial.pdf|}} | ||
- | |||
- | (no much text explanatiobn included) | ||
- | |||
- | |||
- | The input data files for this tutorial could be downloaded here | ||
- | |||
- | {{:smeagol:smeagol_h2_tutorial.zip|}} | ||
- | |||
- | ====== Wiki Tutorial ====== | ||
- | |||
- | ===== Intro ===== | ||
- | |||
- | Smeagol computation is split to 2 parts and 4 independent runs | ||
- | |||
- | - LEADs computation | ||
- | - Fireball SCF (converge equlibrium density of LEADs) | ||
- | - Export LEADs files | ||
- | - SYSTEM computation | ||
- | - Fireball SCF (converge equlibrium density of molecule with leads) | ||
- | - Smeagol computation (Get current, conductivity, transmission spectra) | ||
- | |||
- | Currently there are 3 versions of smeagol implementation | ||
- | - non-SCF smeagol - there you use equlibrium density from Fireball. It's much faster than nonequlibrium SCF-loop in smeagol. Currently this is the only version which looks to work fine. | ||
- | - SCF with Kohn-Sham grid - This should be almost identical to siesta implementation of smeagol. Currently it looks working in principle but there are problems with discontinuity on bonundary of leads. | ||
- | - McWeda smeagol - There is a problem with tranformation of overlap matrix, so consider it as non-working | ||
- | |||
- | As simplest example I will show computation of hydrogen molecule in between hydrogen leads | ||
- | |||
- | ===== LEADS computation ===== | ||
- | |||
- | Let's use this lead geometry | ||
- | |||
- | answer.bas | ||
- | <code> | ||
- | 6 | ||
- | 1 3.000000 3.000000 1.000000 | ||
- | 1 3.000000 3.000000 2.000000 | ||
- | 1 3.000000 3.000000 3.000000 | ||
- | 1 3.000000 3.000000 4.000000 | ||
- | 1 3.000000 3.000000 5.000000 | ||
- | 1 3.000000 3.000000 6.000000 | ||
- | </code> | ||
- | |||
- | cel.lvs | ||
- | <code> | ||
- | 20.000000000 0.000000000 0.000000000 | ||
- | 0.000000000 20.000000000 0.000000000 | ||
- | 0.000000000 0.000000000 6.000000000 | ||
- | </code> | ||
- | |||
- | because you have to converge density of lead is infinite system you should use good sampling in z-direction | ||
- | (**NOTE:** Smeagol always expect current to flow in z-direction !!!) | ||
- | |||
- | input.kpts | ||
- | <code> | ||
- | 100 | ||
- | 0.0000000000 0.0000000000 -0.5183627873 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.5078908118 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4974188363 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4869468608 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4764748853 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4660029098 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4555309343 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4450589588 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4345869833 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4241150078 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4136430323 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.4031710568 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3926990813 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3822271058 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3717551303 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3612831548 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3508111793 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3403392038 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3298672283 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3193952528 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.3089232773 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2984513018 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2879793263 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2775073507 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2670353753 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2565633998 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2460914243 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2356194488 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2251474733 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2146754978 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.2042035223 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1937315468 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1832595713 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1727875958 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1623156203 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1518436448 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1413716693 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1308996937 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1204277182 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.1099557427 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0994837672 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0890117917 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0785398163 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0680678407 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0575958652 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0471238897 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0366519142 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0261799387 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0157079633 0.0100000000 | ||
- | 0.0000000000 0.0000000000 -0.0052359878 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0052359878 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0157079633 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0261799388 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0366519143 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0471238898 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0575958653 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0680678408 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0785398163 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0890117917 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.0994837672 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1099557427 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1204277182 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1308996937 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1413716692 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1518436448 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1623156203 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1727875958 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1832595713 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.1937315468 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2042035223 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2146754977 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2251474732 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2356194487 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2460914242 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2565633997 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2670353753 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2775073508 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2879793263 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.2984513018 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3089232773 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3193952528 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3298672283 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3403392038 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3508111793 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3612831548 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3717551303 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3822271058 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.3926990813 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4031710568 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4136430323 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4241150078 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4345869833 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4450589588 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4555309343 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4660029098 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4764748853 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4869468607 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.4974188362 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.5078908117 0.0100000000 | ||
- | 0.0000000000 0.0000000000 0.5183627872 0.0100000000 | ||
- | </code> | ||
- | |||
- | first you have to run SCF calculation to converge charges, use this input file | ||
- | Fireball.in (scf) | ||
- | <code> | ||
- | &OPTION | ||
- | basisfile = answer.bas | ||
- | lvsfile = cel.lvs | ||
- | icluster = 0 | ||
- | nstepf = 1 | ||
- | sigmatol = 0.0000000001 | ||
- | max_scf_iterations = 100 | ||
- | iqout = 1 | ||
- | ismeagol = 0 | ||
- | ifixcharge = 0 | ||
- | &END | ||
- | |||
- | &OUTPUT | ||
- | iwrtHSrho = 0 | ||
- | iwrteigen = 1 | ||
- | iwrtdos = 0 | ||
- | &END | ||
- | </code> | ||
- | |||
- | then fix charges, activate iwrtHSrho = 1 and run computation again to export LEADs density | ||
- | <code> | ||
- | &OPTION | ||
- | basisfile = answer.bas | ||
- | lvsfile = cel.lvs | ||
- | icluster = 0 | ||
- | nstepf = 1 | ||
- | sigmatol = 0.0000000001 | ||
- | max_scf_iterations = 100 | ||
- | dt = 0.5 | ||
- | iqout = 1 | ||
- | ismeagol = 0 | ||
- | ifixcharge = 1 | ||
- | &END | ||
- | |||
- | &OUTPUT | ||
- | iwrtHSrho = 1 | ||
- | iwrteigen = 0 | ||
- | iwrtdos = 0 | ||
- | &END | ||
- | </code> | ||
- | |||
- | to do this you have to specify k-point sampling of the SYSTEM you want to use the LEADs in. This kpoints file is called MOLECULE.kpts. In our case it has just gamma point, however in general it could have any sampling in x and y direction. (never in z!!). | ||
- | |||
- | MOLECULE.kpts | ||
- | <code> | ||
- | 1 | ||
- | 0.0000000000 0.0000000000 0.0000000000 1.0000000000 | ||
- | </code> | ||
- | |||
- | after the computation you get file called ELECTRODE where k-space representations of Hamiltonian, Density matrix and overplap matrix are stored. | ||
- | |||
- | ===== SYSTEM computation ===== | ||
- | |||
- | Your system should contain the geometry of leads in it's geometry description, it should also contain some region where charge redistribution is screened in order to smoothly align with LEADs, because LEADs itselfs represent infinite bulk and their chrage distribution is fixed during the computation. For example: | ||
- | |||
- | answer.bas | ||
- | <code> | ||
- | 24 | ||
- | 1 3.000000000 3.000000000 1.000000000 #start LEAD.left | ||
- | 1 3.000000000 3.000000000 2.000000000 | ||
- | 1 3.000000000 3.000000000 3.000000000 | ||
- | 1 3.000000000 3.000000000 4.000000000 | ||
- | 1 3.000000000 3.000000000 5.000000000 | ||
- | 1 3.000000000 3.000000000 6.000000000 # end LEAD.left | ||
- | 1 3.000000000 3.000000000 7.000000000 # screening region (left) | ||
- | 1 3.000000000 3.000000000 8.000000000 | ||
- | 1 3.000000000 3.000000000 9.000000000 | ||
- | 1 3.000000000 3.000000000 10.000000000 | ||
- | 1 3.000000000 3.000000000 11.000000000 | ||
- | 1 3.000000000 3.000000000 13.000000000 # molecule itselfs | ||
- | 1 3.000000000 3.000000000 14.000000000 | ||
- | 1 3.000000000 3.000000000 16.000000000 # screening region (right) | ||
- | 1 3.000000000 3.000000000 17.000000000 | ||
- | 1 3.000000000 3.000000000 18.000000000 | ||
- | 1 3.000000000 3.000000000 19.000000000 | ||
- | 1 3.000000000 3.000000000 20.000000000 | ||
- | 1 3.000000000 3.000000000 21.000000000 #start LEAD.right | ||
- | 1 3.000000000 3.000000000 22.000000000 | ||
- | 1 3.000000000 3.000000000 23.000000000 | ||
- | 1 3.000000000 3.000000000 24.000000000 | ||
- | 1 3.000000000 3.000000000 25.000000000 | ||
- | 1 3.000000000 3.000000000 26.000000000 # end LEAD.left | ||
- | </code> | ||
- | |||
- | cel.lvs | ||
- | <code> | ||
- | 20.000000000 0.000000000 0.000000000 | ||
- | 0.000000000 20.000000000 0.000000000 | ||
- | 0.000000000 0.000000000 26.000000000 | ||
- | </code> | ||
- | |||
- | Your kpoint sampling MUST be the same as MOLECULE.kpts in LEADs calculation. Otherweis ELECTRODE files are incompatible. | ||
- | input.kpts | ||
- | <code> | ||
- | 1 | ||
- | 0.0000000000 0.0000000000 0.0000000000 1.0000000000 | ||
- | </code> | ||
- | |||
- | first you should run SCF run with fireball (with smeagol off), it's much faster than doing smeagol directly from neutral atom charges. What's more, currently smeagol SCF implementation in fireball is not perfect. | ||
- | |||
- | fireball.in (scf) | ||
- | <code> | ||
- | &OPTION | ||
- | basisfile = answer.bas | ||
- | lvsfile = cel.lvs | ||
- | icluster = 0 | ||
- | nstepf = 1 | ||
- | sigmatol = 0.0000000001 | ||
- | max_scf_iterations = 100 | ||
- | iqout = 1 | ||
- | ismeagol=0 | ||
- | &END | ||
- | |||
- | &OUTPUT | ||
- | iwrtHSrho = 0 | ||
- | &END | ||
- | </code> | ||
- | |||
- | After convergence of charges, you need to copy ELECTRODE files from LEADs calculation () |