Improve the quality of the FS (task 102)

[Lizard]

Hi all,

I did scf calculation on 10x10x10 mesh, but I would like to generate FS on much more denser k-point set. After SCF I ran one self-consistent loop (set maxscl to 1 and tasks to 1) with a larger k-point set, but FS is generated on 10^3 mesh (taks 102). Is there any other way to increase the number of k-point after one loop? What is the max number of K-point that one can be used?

thanks
Lizard

[Lars Nordström]

Dear Lizard,
I do not understand, task 102 should use your given k-point mesh. How does your elk.in file look like? There is no limit of number of k-points what I am aware of, only those dictated by available time and memory ...
Best wishes /Lars

[Lizard]

Dear Lars,

After scf calculation on mesh 21^3, I changed tasks

tasks
  1
  102

and added

maxscl
2

then I change

ngridk
 33 33 33

Then, after 2 scf I get in INFO.OUT

.
.
.
Total number of atoms per unit cell :    2

Spin treatment :
 spin-unpolarised

Number of Bravais lattice symmetries :   48
Number of crystal symmetries         :   48
Crystal has inversion symmetry
Real symmetric eigensolver will be used

k-point grid :     33    33    33
k-point offset :   0.5000000000      0.5000000000      0.5000000000
k-point set is reduced with full crystal symmetry group
Total number of k-points :     3281

Muffin-tin radius times maximum |G+k| :    9.000000000
 using average radius
Maximum |G+k| for APW functions       :    3.461538462
Maximum (1/2)|G+k|^2                  :    5.991124260
Maximum |G| for potential and density :    24.00000000
Constant for pseudocharge density :   16
Radial integration step length :    4

G-vector grid sizes :     64    64    64
Number of G-vectors :    92217
.
.
.
+--------------------+
| Loop number :    2 |
+--------------------+

Reached self-consistent loops maximum

Energies :
 Fermi                       :     0.273279555693
 sum of eigenvalues          :     -5923.50019274
 electron kinetic            :      10334.3776926
 core electron kinetic       :      10129.9734098
 Coulomb                     :     -19923.3064971
 Coulomb potential           :     -15892.0217717
 nuclear-nuclear             :     -844.353641376
 electron-nuclear            :     -22265.8839397
 Hartree                     :      3186.93108402
 Madelung                    :     -11977.2956112
 xc potential                :     -365.856113663
 exchange                    :     -275.787902496
 correlation                 :     -4.73159785184
 electron entropic           :    -0.265789444102E-04
 total energy                :     -9869.44833140

Density of states at Fermi energy :    8.220376897
 (states/Hartree/unit cell)

Estimated indirect band gap :   0.9413990180E-04
 from k-point   1027 to k-point    579
Estimated direct band gap   :   0.2010689840E-02
 at k-point   1026

Charges :
 core                        :    64.00000000
 valence                     :    26.00000000
 interstitial                :    4.011900266
 muffin-tins (core leakage)
  species :    1 (Zn)
   atom    1                 :    49.23009223     (  0.8015955301E-03)
  species :    2 (La)
   atom    1                 :    36.75800750     (  0.8929363570E-05)
 total in muffin-tins        :    85.98809973
 total calculated charge     :    89.99894505
 total charge                :    90.00000000
 error                       :   0.1054950480E-02

+------------------------------+
| Self-consistent loop stopped |
+------------------------------+

Wrote STATE.OUT

Timings (CPU seconds) :
 initialisation                        :       275.91
 Hamiltonian and overlap matrix set up :      6762.60
 first-variational eigenvalue equation :      1389.72
 charge density calculation            :   -211322.72
 potential calculation                 :        14.29
 total                                 :   -202880.21

+----------------------------+
| Elk version 4.3.06 stopped |
+----------------------------+

Hovewer, FERMISURFACE.bxfs is still created on mesh: 21 21 21
0.000000000 0.000000000 0.000000000
0.5394967580 0.5394967580 -0.5394967580
0.5394967580 -0.5394967580 0.5394967580
-0.5394967580 0.5394967580 0.5394967580
using 1027 to k-point from scf calculations.

Best
Lizard

[Lars Nordström]

OK, it is still unclear when you say "I change ... and I add ...".
What is the final elk.in file and are you sure that a new FERMISURFACE.bxfs was created?

Are you aware that if you have a STATE.OUT and EFERMI.OUT from a scf run, you can calculate just task 102 with whatever k-mesh you like?

Best /Lars

[Lizard]

scf calculation:

tasks
  0
  102
  ngridk
 21 21 21

The quality of FS is poor, so after scf calculation I modificated elk.in in the same folder

tasks
  1
  102
  maxscl
2
ngridk
 33 33 33

and run elk. Both FERMISURFACE.bxfs files have the same number of lines, the mesh 21 21 21 is also the same, however the values inside are different.

[Lars Nordström]

Dear Lizard!

Sorry I remembered wrong:
For task 102 the k-mesh is set up by np3d not ngridk.
You set it up through plot3d (see manual and the Fe-FS example).
If you do not give a mesh, the deafult np3d mesh is 20,20,20 which corresponds to a 21,21,21 mesh for the output to xcrysdens (well, an old bug in xcrysdens which is now a feature ..).

Good luck /Lars

[Lizard]

Dear Lars,

Thank you very much!! You have saved my life :)

Thank you also for a such great software.

with regards
Lizard