Pseudopotential optimization and TB09 functional

[Chris K. Lee]

Dear all,


1.
Can I use CP2K with TB09 in libxc?
I read that TB09 has some problems in libxc v2.0.1 in this forum.
But they fixed it in v2.0.3, and now I compiled the CP2K with libxc 3.0.0.

1-1.
If it's possible, which correlation function should I use? I only find the 
MGGA_X_TB09 in libxc.

1-2.
There's a parameter "c" to set for TB09. How can I set it in my input file?


2.
CP2K already has many Pseudopotentials, but not for some functionals.
So I'm trying to optimize it by using "PSEUDOPOTENTIAL_OPTIMIZATION" option.
I read the manual(https://manual.cp2k.org/trunk/CP2K_INPUT/ATOM.html) and 
some examples on this forum and ./tests/ATOM/regtest-pseudo.

2-1
There's no optimization examples for the metal like Ti.
I find "ELECTRON_CONFIGURATION [Ar] 3d2 4s2" from ./test/ATOM/regtest-1
Is it ok to optimize with that configuration to get q12 result?

2-2
At &POTENTIAL,
What does confinement option means and how it affects?
Some of the examples are using it like "CONFINEMENT 0.0 4.0 2.0"
But I can't understand it with the explanation on the manual.
"Definition of parameters for confinement potential"
2-3
At &POWELL,
The examples in the regtest-pseudo are using "WEIGHT_PSIR0 0.1". but not 
for the examples from this forum. Should I use this option?
And the default of "ACCURACY" is "1.00000000E-006". What's the limit of the 
accuracy?

3.
Is there anything wrong with the attached file except the GTH pp parameters?

4.
Is it important which pseudopotential to get start with?
May I get different results, starting with GTH-BP-q12 or GTH-PADE-q12 ?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20170411/12b84c39/attachment.htm>
-------------- next part --------------
&GLOBAL
  PROGRAM_NAME ATOM
  PRINT_LEVEL DEBUG
&END GLOBAL
&ATOM
  RUN_TYPE PSEUDOPOTENTIAL_OPTIMIZATION
!-----------------------------ATOM------------------------------!
!  ATOMIC_NUMBER 22						!
!  ELEMENT Ti							!
!  CORE [Ar]							!
!  ELECTRON_CONFIGURATION [Ar] 3d2 4s2				!
!---------------------------------------------------------------!
  COULOMB_INTEGRALS ANALYTIC
  EXCHANGE_INTEGRALS ANALYTIC
  &METHOD
     METHOD_TYPE  KOHN-SHAM
     RELATIVISTIC DKH(2)
     &XC
!---------------------------option 1----------------------------!
!       &XC_FUNCTIONAL BLYP					!
!       &END XC_FUNCTIONAL					!
!---------------------------------------------------------------!
!---------------------------option 2----------------------------!
!       &XC_FUNCTIONAL						!
!         &LIBXC						!
!           FUNCTIONAL XC_MGGA_X_M06_L XC_MGGA_C_M06_L		!
!         &END LIBXC						!
!       &END XC_FUNCTIONAL					!
!---------------------------------------------------------------!
     &END XC
  &END METHOD
  &OPTIMIZATION
    EPS_SCF 1.e-10
  &END
  &AE_BASIS
     BASIS_TYPE GEOMETRICAL_GTO
  &END AE_BASIS
  &PP_BASIS
     BASIS_TYPE GEOMETRICAL_GTO
  &END PP_BASIS
  &POTENTIAL
    PSEUDO_TYPE GTH
!?    CONFINEMENT 0.0 4.0 2.0	!?
!---------------------------option 1----------------------------!
!    POTENTIAL_FILE_NAME C_blyp_gth_q1				!
!    POTENTIAL_NAME C						!
!---------------------------------------------------------------!
!---------------------------option 2----------------------------!
!    &GTH_POTENTIAL						!
!    2    4    0    0						!
!    0.241474531885       2  -16.691805741058    2.494605958440	!
!       1							!
!    0.220838245905       1   18.355844903927			!
!    &END 							!
!---------------------------------------------------------------!
  &END POTENTIAL
  &POWELL
     ACCURACY   1.00000000E-010
     STEP_SIZE  0.005
!?     WEIGHT_PSIR0 0.1	!?
  &END
&END ATOM