Hi,
The use of HDLOs gives you the ultimate solution for the total energy
with virtually "exact" (for the given DFT functional) result. However,
HDLOs have not been implemented for all "properties".
For instance, you cannot use HDLOs for optics or for wien2wanner (you
can use, however, one HELO (a LO at higher energy).
Am 10.04.2024 um 08:32 schrieb 夏宇阳:
Dear Prof.Blaha,
Thanks for your reply. It really helps me a lot.
Actually,i did the initialization with -nodstart before, and there was an error
in the lapw2 calculation later.Now I know the reason.
After plot the DOS and bandstructure of KFe2As2 slabs, i tried to use
wien2wannier for it.
I followed the guide step by step and when i do
x wannier90 -up
An error came out. After check the file generated by "x w2w -up", i found that all of the data in "*.mmnup" and the data at the bottom of "*.amnup" are "NaN".
The possible reason for this is in the "*.outputwfup", the LO COEFFICIENT of Fe and As is
NaN and "***********"
ATPAR
ATOMIC PARAMETERS FOR Fe1
ENERGY PARAMETERS ARE 0.02 0.02 0.05 0.02 0.02 0.02
L U(R) U'(R) DU/DE DU'/DE NORM-U'
0 -0.551142E+00 0.124267E+00 0.109667E+00 0.340139E+00
1 0.640043E+00 0.161089E+00 -0.789075E-01 -0.334058E+00
2 0.186063E+00 -0.174896E+00 -0.572407E+00 -0.542898E+00
3 0.825200E+00 0.874714E+00 -0.593096E-01 -0.306556E+00
4 0.950759E+00 0.151315E+01 -0.398060E-01 -0.274852E+00
5 0.105105E+01 0.218939E+01 -0.301429E-01 -0.254103E+00
LO COEFFICIENT: l,A,B,C 0 0.86181 4.33109 0.00000 0.00000
LO COEFFICIENT: l,A,B,C 0 0.06758 0.00000 0.99586 0.00000
LO COEFFICIENT: l,A,B,C 1 0.74147 6.01428 0.00000 0.00000
LO COEFFICIENT: l,A,B,C 1 0.10317 0.00000 0.98454 0.00000
LO COEFFICIENT: l,A,B,C 2 0.97213 0.31600 0.00000 0.00000
LO COEFFICIENT: l,A,B,C 2 ************ 0.00000************ 0.00000
number of rad. functions per L: 3 3 3 2 2 2
ATPAR
ATOMIC PARAMETERS FOR As2
ENERGY PARAMETERS ARE 0.42 0.02 0.02 0.02 0.02 0.02
L U(R) U'(R) DU/DE DU'/DE NORM-U'
0 -0.229133E+00 0.708995E+00 0.212786E+00 0.312661E+00
1 0.523646E+00 -0.223201E+00 -0.137651E+00 -0.366257E+00
2 -0.722238E+00 -0.403473E+00 0.544474E-01 0.338522E+00
3 0.867710E+00 0.895087E+00 -0.617087E-01 -0.320081E+00
4 0.101233E+01 0.163812E+01 -0.401997E-01 -0.284836E+00
5 0.112453E+01 0.241620E+01 -0.300851E-01 -0.262492E+00
LO COEFFICIENT: l,A,B,C 0 0.97291 1.04766 0.00000 0.00000
LO COEFFICIENT: l,A,B,C 0 1.87772 0.00000 -0.92893 0.00000
LO COEFFICIENT: l,A,B,C 1 0.88419 3.36358 0.00000 0.00000
LO COEFFICIENT: l,A,B,C 1 NaN 0.00000 NaN 0.00000
LO COEFFICIENT: l,A,B,C 2 0.44520 5.90551 0.00000 0.00000
LO COEFFICIENT: l,A,B,C 2 0.09624 0.00000 0.97840 0.00000
number of rad. functions per L: 3 3 3 2 2 2
What cause it? And What should i do next?
Looking forward to your reply.
With redards.
Yuyang Xia
----- 原始邮件 -----
发件人: "Peter Blaha" <[email protected]>
收件人: "wien" <[email protected]>
发送时间: 星期三, 2024年 4 月 10日 下午 1:26:52
主题: Re: [Wien] Question of bad E(TOP) with no error
Hi,
The "crazy" E-top is NOT a problem.
While for very low lying semicore states we expect to find E-top AND E-bottom
(STOP in the corresponding line of case.in1), this is not necessary for higher
states. (CONT in in1). We simply do not search for E-top much above EF, since
this could lead to an E-parameter (expansion energy of the radial wavefunction)
which is higher than EF and thus in the unoccupied region. Obviously, during
scf we are interested on an accurate description of the OCCUPIED states below
EF.
Just one further remark (provided you did not have RMTs above 2.3 in the
original setup with -prec 1):
init_lapw -perc 1 -ecut -7.0 -sp
runsp_lapw -i 1000 -fc 3 -p
save_lapw unrelaxed
runsp_lapw -i 1000 -min -fc 1 -p
save_lapw relaxed
init_lapw -prec 2 -ecut -7.0 -sp -nodstart <----- this would save a lot
of cpu time
runsp_lapw -i 1000 -cc 0.0001 -p
-nodstart will not produce a new density, but you would continue with the
already converged one (from the previous runsp_lapw). Only if you had large
spheres (above 2.3), prec 2 will reduce these spheres automatically and you
must start over with dstart since the radial mesh has been changed.
I am doing a project about KFe2As2 surface.
After relaxing the surface slab in prec 1, i did a prec 2 calculation with cc
0.0001.
There is no error in the process, but when i check the scf1up and scf1dn, i
found that E(TOP) is not good.
ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM As1
:e__0004: OVERALL ENERGY PARAMETER IS 0.0158
OVERALL BASIS SET ON ATOM IS LAPW
:E2_0004: E( 2)= 0.0158
APW+lo
:E2_0004: E( 2)= -2.3962 E(BOTTOM)= -2.455 E(TOP)= -2.337 0 1 120
LOCAL ORBITAL
:E0_0004: E( 0)= 0.4158
APW+lo
:E0_0004: E( 0)= -0.8200 E(BOTTOM)= -1.495 E(TOP)= -520.000 3 -1 175
LOCAL ORBITAL
:E1_0004: E( 1)= 0.0158
APW+lo
:E1_0004: E( 1)= 0.0158
LOCAL ORBITAL(SECDER)
ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM As2
:e__0005: OVERALL ENERGY PARAMETER IS 0.0158
OVERALL BASIS SET ON ATOM IS LAPW
:E2_0005: E( 2)= 0.0158
APW+lo
:E2_0005: E( 2)= -2.4026 E(BOTTOM)= -2.462 E(TOP)= -2.344 0 1 124
LOCAL ORBITAL
:E0_0005: E( 0)= 0.4158
APW+lo
:E0_0005: E( 0)= -0.8200 E(BOTTOM)= -1.506 E(TOP)= -520.000 3 -1 175
LOCAL ORBITAL
:E1_0005: E( 1)= 0.0158
APW+lo
:E1_0005: E( 1)= 0.0158
LOCAL ORBITAL(SECDER)
Is there any influence to the results? if is,what should i do?
The structure is 1X1X1 supercell with 25 vacuum in z(for less time),rmt reduced
3%.
init_lapw -perc 1 -ecut -7.0 -sp
runsp_lapw -i 1000 -fc 3 -p
save_lapw unrelaxed
runsp_lapw -i 1000 -min -fc 1 -p
save_lapw relaxed
init_lapw -prec 2 -ecut -7.0 -sp
runsp_lapw -i 1000 -cc 0.0001 -p
in outputst,the As is
As RHFS
FOR SPIN 1 R-MT= 2.12000 R*V= -3.20524 V= -1.51191
FOR SPIN 2 R-MT= 2.12000 R*V= -2.68025 V= -1.26427
TOTAL CHARGE FOR SPIN 1 : 18.000000000108162
TOTAL CHARGE FOR SPIN 1 INSIDE SPHERE: 15.842689296044156
TOTAL CHARGE FOR SPIN 2 : 15.000000000108740
TOTAL CHARGE FOR SPIN 2 INSIDE SPHERE: 14.639384914097263
TOTAL CHARGE in sigma FOR SPIN 1 :
0.0000000000000000
TOTAL CHARGE in sigma FOR SPIN 1 INSIDE SPHERE:
0.0000000000000000
TOTAL CHARGE in sigma FOR SPIN 2 :
0.0000000000000000
TOTAL CHARGE in sigma FOR SPIN 2 INSIDE SPHERE:
0.0000000000000000
E-up(Ry) E-dn(Ry) Occupancy q/sphere core-state
1S -861.070611 -861.062271 1.00 1.00 1.0000 T
2S -109.198830 -109.195361 1.00 1.00 1.0000 T
2P* -97.557529 -97.550205 1.00 1.00 1.0000 T
2P -94.870103 -94.863015 2.00 2.00 1.0000 T
3S -13.958022 -13.947526 1.00 1.00 1.0000 T
3P* -10.047612 -10.035593 1.00 1.00 0.9999 T
3P -9.678540 -9.666373 2.00 2.00 0.9999 T
3D* -2.998617 -2.979436 2.00 2.00 0.9965 F
3D -2.945248 -2.925727 3.00 3.00 0.9963 F
4S -1.117588 -0.948715 1.00 1.00 0.6527 F
4P* -0.448722 -0.278655 1.00 0.00 0.4165 F
4P -0.428475 -0.258969 2.00 0.00 0.3984 F
TOTAL CORE-CHARGE: 18.000000
TOTAL CORE-CHARGE INSIDE SPHERE: 17.999236
TOTAL CORE-CHARGE OUTSIDE SPHERE: 0.000764
TOTAL ENERGY (RYD): -4521.980893
SUM OF EI:-2.6347206E+03 NUC:-1.0916132E+04 COUL:-7.0407530E+03
V-XC SPIN 1:-1.1047898E+02 E-XC SPIN 1:-8.4909182E+01
V-XC SPIN 2:-1.0750097E+02 E-XC SPIN 2:-8.2641601E+01
In addition, in the bulk calulation, the As still has a bad E(TOP).
Looking forward to your reply.
With regards!
Yuyang Xia
_______________________________________________
Wien mailing list [ mailto:[email protected] |
[email protected] ] [
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien |
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien ] SEARCH the
MAILING-LIST at: [
http://www.mail-archive.com/[email protected]/index.html |
http://www.mail-archive.com/[email protected]/index.html ]
--
-----------------------------------------------------------------------
Peter Blaha, Inst. f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-158801165300
Email: [email protected]
WWW: http://www.imc.tuwien.ac.at WIEN2k: http://www.wien2k.at
-------------------------------------------------------------------------
_______________________________________________
Wien mailing list
[email protected]
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/[email protected]/index.html
