#1 External electric field perpendicular to thin film by sdub 10.02.2022 23:31

I am willing to set up a set of simple SCF calculations of thin films under constant external electric field.
As a warm-up exercise, I try to reproduce the various band structure topologies of graphene under perpendicular electric field, including the effect of SOC.

To do so, I added the following <fields> tag in the calculationSetup section:
<fields zsigma="0.0" sig_b_1="0.05" sig_b_2="-0.05" autocomp="T"
dirichlet="F" eV="F">
</fields>



But the calculation ends with the error message:
***************ERROR***************
B-fields can only be used in spin-polarized calculations
***************ERROR***************


Could someone help me to figure out what I am doing wrong?
Thanks in advance,

Simon


NOTE:
This is FLEUR version: fleur 35
Its git version is:
described by: MaX-R6.0-3-g5518b73

#2 RE: External electric field perpendicular to thin film by sdub 11.02.2022 13:48

Dear all,

eventually, setting up a spin-polarized calculation solves the problem. I attach the resulting band-structure.
The disadvantage of this approach is that I had to impose a spin-polarized occupation for the carbon atoms in the inp.xml. While self-consistency brings the system towards a non-magnetic solution, the final net magnetic moment is not exactly zero but of the order of 8x10-7...

Parsing the output file, I still have one little question. Equal and opposite charges have been placed on the external sheets 1 and 2. Why does the code reports external electric fields of opposite magnitude on these two planes?

------------------------------------------------------------
parameters for external electric field:
total electronic charge = 12.00000
total nuclear charge = 12.00000
z-z1 of external sheet = 0.00000 a0 = 0.00000 A
charge on external sheet 1 = 0.00500 (surface density= 0.00027 e/a.u.**2)
external field on sheet 1 = -0.17264E+08 V/cm
charge on external sheet 2 = -0.00500 (surface density= -0.00027 e/a.u.**2)
external field on sheet 2 = 0.17264E+08 V/cm
NOTE: The equation for the E field assumes two oppositely charged plates.
You may need to divide by two before summing the external fields to avoid double counting
------------------------------------------------------------

I guess that in this case, the actual field within the capacitor is of 0.17264E+08 V/cm, which is consistent with the computed band-structure. Is that right?

Best regards,

Simon

#3 RE: External electric field perpendicular to thin film by yang 28.02.2022 14:38

I keep reporting this error, have you encountered it? If so, how to solve it

**************juDFT-Error*****************
Error message: z-mirror/inversion symmetry but second e-fieldsheet specified
Error occurred in subroutine: efield
Error from PE:0/1
*****************************************
Last known location:
Last timer:Initialization
Timerstack:
Timer:Total Run
*****************************************

#4 RE: External electric field perpendicular to thin film by Gregor 01.03.2022 13:08

The error message

1
 
 z-mirror/inversion symmetry but second e-fieldsheet specified
 


to me sounds like your specification of an E-field is inconsistent with the symmetries found in the unit cell.
I think the issue here is that the symmetries are detected by the input generator inpgen and the E-field is
typically introduced only once the Fleur input file is available. At least it is not considered in the input
generator symmetry detection.

You thus have to manually make sure that the input generator does not detect these two symmetries
mentioned in the error message. You can either do that by distinguishing between atoms of the same
chemical element in the inpgen input by adding ".01" to one of the atomic numbers, or you can more
generally switch of all symmetry detection by using the input generator with the "-nosym" command line
switch. If you take the latter approach you should be aware that detected symmetry operations are
used to speed-up calculations and are also beneficial to the stability of calculations.

Xobor Einfach ein eigenes Xobor Forum erstellen
Datenschutz