[CP2K-user] [CP2K:14503] Re: TDDFPT calculations with CP2K7.1 using the range-separated functionals
Lucas Lodeiro
eluni... at gmail.com
Mon Jan 11 20:02:03 UTC 2021
Hi Liu,
I did not run TDDFT calculations, but I did some tests between CP2K and
other programs like G09. As Vladimir says, your basis sets are not the
same, and some difference could appear due to this reason. But in your case
the differences are big, 1.6 eV approx. I found that some default settings
of convergences criterium are differents, for example the *EPS_SCF* which
is 1E-8 in G09, you could tight your convergence criterion, *EPS_SCF, *
*EPS_DEFAULT, **EPS_SCHWARZ* to -8, -12 and -8 to get results with similar
convergences in both programs. Also, you are using a PBC calculation in a
big cell, but maybe it is no sufficient to mimic the isolated molecule as
in G09... and yout cutoff radius for HF is a little bit short, if you run a
non-periodic calculation, you can use just the long range potential without
the truncation.
Finally, just to speed up, you can use OT instead of diagonalization
method, with it you can use *ADMM_PURIFICATION_METHOD *MO_DIAG.
In order to have the same basis sets, as vladimir says, you could explore
to use the same basis sets in both programs, you can get basis sets to both
programs from: https://www.basissetexchange.org/
And the Auxiliary basis set is the minimum one, you could explore FIT and
pFIT basis set to check if the result is sensitive.
Regards
Lucas Lodeiro
El lun, 11 ene 2021 a las 13:01, Vladimir Rybkin (<rybk... at gmail.com>)
escribió:
> Dear Xiang-Yang Liu,
>
> most importantly: you are using different basis sets in Gaussian and CP2K
> (CP2K also use pseudopotentials). With this difference in mind you
> differences are within reasonable. Generally, difference below 0.1 eV for
> TDDFT implementations is not "great".
>
> Yours,
>
> Vladimir
>
> понедельник, 11 января 2021 г. в 14:49:47 UTC+1, lx... at gmail.com:
>
>> Dear All,
>>
>> I have tried to use the TDDFPT method implemented in CP2K7.1 to do
>> excited state calculations with the range-separated functionals such as
>> wB97XD.
>> However, after several tests with a small molecule, namely H2Pc, I found
>> that the first two excitation energies are greatly underestimated in
>> comparison with the LR-TDDFT results obtained in GAUSSIAN09 (ca. 0.51 and
>> 0.60 eV (CP2K7.1) vs. 2.18 and 2.26 eV (GAUSSIAN09)).
>> I wonder whether there are some mistakes with my input file or there are
>> some problem of TDDFPT for such calculations?
>>
>> The input file used in my calculations is written as follows:
>>
>> *&GLOBAL*
>> * PROJECT tddfpt*
>> * RUN_TYPE energy*
>> * PRINT_LEVEL medium*
>> *&END GLOBAL*
>>
>> *&FORCE_EVAL*
>> * METHOD Quickstep*
>>
>> * &PROPERTIES*
>> * &TDDFPT*
>> * NSTATES 5 # number of excited states*
>> * &END TDDFPT*
>> * &END PROPERTIES*
>>
>> *&DFT*
>> * BASIS_SET_FILE_NAME GTH_BASIS_SETS*
>> * POTENTIAL_FILE_NAME POTENTIAL*
>> * BASIS_SET_FILE_NAME BASIS_ADMM_MOLOPT*
>> * BASIS_SET_FILE_NAME BASIS_ADMM*
>>
>> * &AUXILIARY_DENSITY_MATRIX_METHOD*
>> * METHOD BASIS_PROJECTION*
>> * ADMM_PURIFICATION_METHOD NONE *
>> * EXCH_CORRECTION_FUNC BECKE88X*
>> * &END AUXILIARY_DENSITY_MATRIX_METHOD*
>>
>> * CHARGE 0 *
>> * &MGRID*
>> * CUTOFF 400*
>> * &END MGRID*
>> * &QS*
>> * METHOD gpw*
>> * EPS_PGF_ORB 1e-12*
>> * &END QS*
>> * &SCF*
>> * MAX_SCF 100*
>> * EPS_SCF 1e-5*
>> * SCF_GUESS atomic*
>>
>> * &DIAGONALIZATION*
>> * ALGORITHM STANDARD*
>> * &END DIAGONALIZATION*
>>
>> * &MIXING T*
>> * ALPHA 0.5*
>> * METHOD PULAY_MIXING*
>> * NPULAY 5*
>> * &END MIXING*
>> * &END SCF*
>> * &POISSON*
>> * PERIODIC XYZ*
>> * &END POISSON*
>>
>> * &XC*
>> * &XC_FUNCTIONAL*
>> * &LIBXC*
>> * FUNCTIONAL HYB_GGA_XC_WB97X_D*
>> * &END*
>> * &END XC_FUNCTIONAL*
>> * &HF*
>> * &SCREENING*
>> * EPS_SCHWARZ 1.0E-6*
>> * &END*
>> * &MEMORY*
>> * MAX_MEMORY 100*
>> * &END*
>> * &INTERACTION_POTENTIAL*
>> * POTENTIAL_TYPE MIX_CL_TRUNC*
>> * OMEGA 0.2*
>> * SCALE_LONGRANGE 0.777964*
>> * SCALE_COULOMB 0.222036*
>> * CUTOFF_RADIUS 5.0*
>> * T_C_G_DATA t_c_g.dat*
>> * &END*
>> * &END*
>> * &XC_GRID*
>> * XC_DERIV SPLINE2_SMOOTH # this is needed for the 2nd
>> derivatives of the XC functional*
>> * &END XC_GRID*
>> * &END XC*
>>
>> * &END DFT*
>> * &SUBSYS*
>> * &TOPOLOGY*
>> * &CENTER_COORDINATES*
>> * &END CENTER_COORDINATES*
>> * &END TOPOLOGY*
>> * &CELL*
>> * ABC 30.0 30.0 30.0*
>> * PERIODIC XYZ*
>> * &END CELL*
>> * &COORD*
>> * N 0.01468800 2.00544200 -0.29683900*
>> * N -2.37562900 2.41155000 -0.13186300*
>> * N -2.41200400 -2.37377100 -0.17491600*
>> * N 2.41112400 2.37299900 -0.12658700*
>> * N -0.01307800 -2.00591400 -0.34375500*
>> * N 2.37788800 -2.41418900 -0.17065800*
>> * C -4.17222800 0.73645800 -0.00309900*
>> * C -0.67412900 4.18239500 0.01934000*
>> * C -4.18380100 -0.67562700 -0.02014000*
>> * C 0.73808000 4.17070300 0.02178600*
>> * C -2.78102300 1.14403300 -0.18692600*
>> * C -1.10192300 2.79691600 -0.16605600*
>> * C -2.79906300 -1.10173400 -0.21220500*
>> * C 1.14375300 2.77826000 -0.16260500*
>> * C -0.73890800 -4.16630900 -0.00019300*
>> * C 4.18311400 0.67096100 0.00115100*
>> * C 0.67355900 -4.17905700 0.00226800*
>> * C 4.17314700 -0.74140800 -0.01277100*
>> * C -1.14329800 -2.77792600 -0.20621200*
>> * C 2.79688300 1.09912600 -0.17512500*
>> * C 1.10326800 -2.79777700 -0.20275300*
>> * C 2.78178900 -1.14729500 -0.19912900*
>> * C -5.34524900 1.46617200 0.20504700*
>> * C -1.38645100 5.36877700 0.20954200*
>> * C -5.36863500 -1.39063500 0.17089500*
>> * C 1.47037200 5.34443400 0.21437700*
>> * C -6.53049700 0.75063900 0.38667600*
>> * C -0.65342400 6.54324100 0.39406800*
>> * C -6.54209900 -0.66018000 0.36971000*
>> * C 0.75719100 6.53149500 0.39645600*
>> * C -1.47251300 -5.33425600 0.22345700*
>> * C 5.36943600 1.38205000 0.19903500*
>> * C 1.38473700 -5.36039200 0.22884400*
>> * C 5.34798200 -1.47459500 0.17084800*
>> * C -0.76071300 -6.51537100 0.44153000*
>> * C 6.54461400 0.64821000 0.37381100*
>> * C 0.65044900 -6.52827600 0.44412700*
>> * C 6.53422100 -0.76263800 0.36007100*
>> * H -5.31982100 2.55648400 0.22741200*
>> * H -2.47726000 5.36199000 0.21499900*
>> * H -5.36047700 -2.48141400 0.16783300*
>> * H 2.56090800 5.31878300 0.22274500*
>> * H -7.46690300 1.28881400 0.54987000*
>> * H -1.17840000 7.48942700 0.54335800*
>> * H -7.48727000 -1.18662900 0.52012100*
>> * H 1.29740200 7.46875700 0.54746900*
>> * H -2.56294400 -5.30668600 0.23110100*
>> * H 5.36158400 2.47267500 0.21857200*
>> * H 2.47541900 -5.35281000 0.24119000*
>> * H 5.32277600 -2.56515200 0.16738300*
>> * H -1.30192600 -7.44735900 0.61932200*
>> * H 7.49034300 1.17231500 0.52877000*
>> * H 1.17373400 -7.47001200 0.62407400*
>> * H 7.47217800 -1.30348900 0.50426500*
>> * H -0.95410508 0.01210298 -0.52371821*
>> * H 0.94993522 -0.00596112 -0.48792422*
>> * N -2.00779500 0.01729500 -0.33767900*
>> * N 2.00585900 -0.01636500 -0.31524900*
>> * &END COORD*
>> * &KIND N*
>> * BASIS_SET DZVP-GTH*
>> * POTENTIAL GTH-PBE*
>> * BASIS_SET AUX_FIT cFIT3*
>> * &END KIND*
>> * &KIND C*
>> * BASIS_SET DZVP-GTH*
>> * POTENTIAL GTH-PBE*
>> * BASIS_SET AUX_FIT cFIT3*
>> * &END KIND*
>> * &KIND H*
>> * BASIS_SET DZVP-GTH*
>> * POTENTIAL GTH-PBE*
>> * BASIS_SET AUX_FIT cFIT3*
>> * &END KIND*
>> * &END SUBSYS*
>> *&END FORCE_EVAL*
>>
>> *Best wishes, *
>> *Xiang-Yang Liu*
>>
> --
> You received this message because you are subscribed to the Google Groups
> "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to cp... at googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/cp2k/4897d4d1-a074-44f6-89ff-3b2c613eab83n%40googlegroups.com
> <https://groups.google.com/d/msgid/cp2k/4897d4d1-a074-44f6-89ff-3b2c613eab83n%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210111/cb7c4be2/attachment.htm>
More information about the CP2K-user
mailing list