"derivatives bigger than 2 not implemented" for truncated PBE0 functional in TDDFPT

Weibin Chu

unread,

May 1, 2021, 2:32:42 PM5/1/21

to cp2k

Dear CP2k users and developers,

I am using CP2K 8.1 to an LR-TDDFPT calculation with truncated PBE0 functional for a periodic system, and I get this error. I am wondering if anyone also run into the same issue and could help me with it.

I tried Matt's suggestion in another post that adding EXCH_CORRECTION_FUNC PBEX, but it still gives the same error.

The calculation can go through without &PBE_HOLE_T_C_LR section or a pure PBE0 functional without any truncation or a PBE functional.

I have attached my input at the end.

Best,

Weibin

*******************************************************************************

* ___ *

* / \ *

* [ABORT] *

* \___/ derivatives bigger than 2 not implemented *

* | *

* O/| *

* /| | *

* / \ xc/xc_xpbe_hole_t_c_lr.F:364 *

*******************************************************************************

&FORCE_EVAL

METHOD Quickstep

&PROPERTIES

&TDDFPT

NSTATES 10

MAX_ITER 100

CONVERGENCE [eV] 1.0e-5

&MGRID

CUTOFF 400 # separate cutoff for TDDFPT calc

&END

&END TDDFPT

&END PROPERTIES

&DFT

UKS

CHARGE -1

BASIS_SET_FILE_NAME GTH_BASIS_SETS

BASIS_SET_FILE_NAME BASIS_ADMM

POTENTIAL_FILE_NAME GTH_POTENTIALS

&MGRID

CUTOFF 600

REL_CUTOFF 50

&END MGRID

&AUXILIARY_DENSITY_MATRIX_METHOD

ADMM_PURIFICATION_METHOD NONE

EXCH_CORRECTION_FUNC PBEX

&END

&SCF

SCF_GUESS RESTART

EPS_SCF 5.0E-7

MAX_SCF 30

&OT

MINIMIZER DIIS

PRECONDITIONER FULL_SINGLE_INVERSE

&END

&OUTER_SCF

EPS_SCF 5.0E-7

MAX_SCF 10

&END

&END SCF

&PRINT

&MO_CUBES

WRITE_CUBE .FALSE.

NLUMO 20

NHOMO 2

&END MO_CUBES

&END

&XC

&XC_FUNCTIONAL

&PBE

SCALE_X 0.6

SCALE_C 1.0

&END

&PBE_HOLE_T_C_LR

SCALE_X 0.4

CUTOFF_RADIUS 7.5

&END

&END XC_FUNCTIONAL

&XC_GRID

XC_DERIV SPLINE3_SMOOTH # this is needed for the 2nd

&END XC_GRID

&HF

FRACTION 0.4

&SCREENING

EPS_SCHWARZ 1.0E-10

SCREEN_ON_INITIAL_P FALSE

&END

&MEMORY

MAX_MEMORY 9000

EPS_STORAGE_SCALING 0.1

&END

&INTERACTION_POTENTIAL ! Sets up interaction potential between the POTENTIAL_TYPE TRUNCATED ! the potential is truncated

CUTOFF_RADIUS 7.5

T_C_G_DATA ./t_c_g.dat ! external file with parameters needed to

&END

&VDW_POTENTIAL

POTENTIAL_TYPE NON_LOCAL

&NON_LOCAL

TYPE RVV10

KERNEL_FILE_NAME rVV10_kernel_table.dat

CUTOFF 600

&END

Leili Rassouli

unread,

May 9, 2022, 5:24:41 PM5/9/22

to cp2k

Hi Weibin,

I have the same problem, did you figure out how to solve this issue?

Best,

Lili

Matt Watkins

unread,

May 10, 2022, 5:30:58 AM5/10/22

to cp2k

I think that you'll have to remove the &PBE_HOLE_T_C_LR section for TDDFPT calculation at the moment.

With a 7.5A truncation cutoff it is not going to make a significant difference in any case I don't expect.

Reply all

Reply to author

Forward