Cp2k SCF convergence problem
Matt W
MattWa... at gmail.com
Fri Jul 8 17:39:08 UTC 2016
Does your system have a band gap?? If so then stick with OT.
With direct mixing, try reducing alpha (&scf / &mixing / alpha) ot 0.2 or
0.1
Matt
On Friday, July 8, 2016 at 3:57:34 PM UTC+1, Felipe Zapata wrote:
>
> Dear all,
>
> I have been trying to do single point calculations for a MD simulation of
> a quantumdot (Cd68Se55_26HCOO). The MD simulation was done with another
> packages and now I want to compute the molecular orbitals for each point.
> For most of the points the SCF converge sucessfully but for other points
> I'm having some converge problem.
>
> I first do a single point calculation using the OT method and then using
> that initial gues I use the following input to calculate the Molecular
> orbitals:
>
>
> &FORCE_EVAL
> &DFT
> BASIS_SET_FILE_NAME ./Cp2k/cp2k_basis/BASIS_MOLOPT
> &MGRID
> CUTOFF 400
> NGRIDS 4
> &END
> POTENTIAL_FILE_NAME ./cp2k_basis/GTH_POTENTIALS
> &PRINT
> &MO
> ADD_LAST NUMERIC
> &EACH
> QS_SCF 0
> &END
> EIGENVALUES
> EIGENVECTORS
> FILENAME
> ./Quantumdots/classicTrajectory_Cd68Se55/classic_Cd68Se55/point_405/mo_coeff_405.out
> NDIGITS 36
> OCCUPATION_NUMBERS
> &END
> &END
> &QS
> METHOD GPW
> &END
> &SCF
> &OUTER_SCF
> EPS_SCF 1.0E-7
> MAX_SCF 500
> &END
> ADDED_MOS 100
> &DIAGONALIZATION
> JACOBI_THRESHOLD 1e-06
> &END
> EPS_SCF 5e-05
> MAX_DIIS 7
> MAX_SCF 200
> &MIXING
> ALPHA 0.4
> METHOD BROYDEN_MIXING
> NBROYDEN 8
> &END
> SCF_GUESS RESTART
> &END
> WFN_RESTART_FILE_NAME
> ./Quantumdots/classicTrajectory_Cd68Se55/plams.29958/plamsjob/point_405-RESTART.wfn
> &XC
> &XC_FUNCTIONAL PBE
> &END
> &END
> &END
> &SUBSYS
> &CELL
> ABC [angstrom] 33.0 33.0 33.0
> PERIODIC XYZ
> &END
> &KIND C
> BASIS_SET DZVP-MOLOPT-SR-GTH-q4
> POTENTIAL GTH-PBE-q4
> &END
> &KIND CD
> BASIS_SET DZVP-MOLOPT-SR-GTH-q12
> POTENTIAL GTH-PBE-q12
> &END
> &KIND H
> BASIS_SET DZVP-MOLOPT-SR-GTH-q1
> POTENTIAL GTH-PBE-q1
> &END
> &KIND O
> BASIS_SET DZVP-MOLOPT-SR-GTH-q6
> POTENTIAL GTH-PBE-q6
> &END
> &KIND SE
> BASIS_SET DZVP-MOLOPT-SR-GTH-q6
> POTENTIAL GTH-PBE-q6
> &END
> &TOPOLOGY
> COORD_FILE_NAME
> ./Quantumdots/classicTrajectory_Cd68Se55/classic_Cd68Se55/point_405/
> coordinates_405.xyz
> COORDINATE XYZ
> &END
> &END
> &END
>
> &GLOBAL
> PRINT_LEVEL LOW
> PROJECT point_405
> RUN_TYPE ENERGY_FORCE
> &END
>
>
> The problem is that the SCF never converge it keep on Jumping like
>
> SCF WAVEFUNCTION OPTIMIZATION
>
> Step Update method Time Convergence Total energy
> Change
>
> ------------------------------------------------------------------------------
> 1 NoMix/Diag. 0.40E+00 6.3 0.84522228 -4649.6390448163
> -4.65E+03
> 2 Broy./Diag. 0.40E+00 7.6 0.96131110 -4649.6699305732
> -3.09E-02
> 3 Broy./Diag. 0.40E+00 7.7 0.11609686 -4649.5832054595
> 8.67E-02
> 4 Broy./Diag. 0.40E+00 7.8 0.07416114 -4649.6039417400
> -2.07E-02
> 5 Broy./Diag. 0.40E+00 7.8 0.01729699 -4649.6443739516
> -4.04E-02
> 6 Broy./Diag. 0.40E+00 7.9 0.02445440 -4649.6348925566
> 9.48E-03
> 7 Broy./Diag. 0.40E+00 7.9 0.03103822 -4649.6351411954
> -2.49E-04
> 8 Broy./Diag. 0.40E+00 8.0 0.00114658 -4649.6416145735
> -6.47E-03
> 9 Broy./Diag. 0.40E+00 8.1 0.60426960 -4649.6373542748
> 4.26E-03
> 10 Broy./Diag. 0.40E+00 7.5 0.68281399 -4649.6701726108
> -3.28E-02
> 11 Broy./Diag. 0.40E+00 7.7 0.08554781 -4649.5897153631
> 8.05E-02
> 12 Broy./Diag. 0.40E+00 7.7 0.05092342 -4649.6084580537
> -1.87E-02
> 13 Broy./Diag. 0.40E+00 7.8 0.01359178 -4649.6371949470
> -2.87E-02
> 14 Broy./Diag. 0.40E+00 7.8 0.02071287 -4649.6307287482
> 6.47E-03
> 15 Broy./Diag. 0.40E+00 7.9 0.03120689 -4649.6326753881
> -1.95E-03
> 16 Broy./Diag. 0.40E+00 8.0 0.00705844 -4649.6379286529
> -5.25E-03
> 17 Broy./Diag. 0.40E+00 8.0 0.82343001 -4649.6320356443
> 5.89E-03
> 18 Broy./Diag. 0.40E+00 7.6 0.91596279 -4649.6678385355
> -3.58E-02
> 19 Broy./Diag. 0.40E+00 7.7 0.11249971 -4649.5839525229
> 8.39E-02
> 20 Broy./Diag. 0.40E+00 7.7 0.07057385 -4649.6014268377
> -1.75E-02
> 21 Broy./Diag. 0.40E+00 7.8 0.01673834 -4649.6398071673
> -3.84E-02
> 22 Broy./Diag. 0.40E+00 7.8 0.02373111 -4649.6307821358
> 9.03E-03
> 23 Broy./Diag. 0.40E+00 7.9 0.02421883 -4649.6313626871
> -5.81E-04
> 24 Broy./Diag. 0.40E+00 8.0 0.00625369 -4649.6360887050
> -4.73E-03
> 25 Broy./Diag. 0.40E+00 8.1 0.81451803 -4649.6310548261
> 5.03E-03
> 26 Broy./Diag. 0.40E+00 7.6 0.90433045 -4649.6683294326
> -3.73E-02
> 27 Broy./Diag. 0.40E+00 7.7 0.11107016 -4649.5835546482
> 8.48E-02
> 28 Broy./Diag. 0.40E+00 7.7 0.06947376 -4649.6010024424
> -1.74E-02
> 29 Broy./Diag. 0.40E+00 7.8 0.01655600 -4649.6389548196
> -3.80E-02
> 30 Broy./Diag. 0.40E+00 7.8 0.02337783 -4649.6301460537
> 8.81E-03
> 31 Broy./Diag. 0.40E+00 7.9 0.02373579 -4649.6307737334
> -6.28E-04
> 32 Broy./Diag. 0.40E+00 8.0 0.00729403 -4649.6353776821
> -4.60E-03
> 33 Broy./Diag. 0.40E+00 8.1 0.80710648 -4649.6309035058
> 4.47E-03
> 34 Broy./Diag. 0.40E+00 7.6 0.89657936 -4649.6684316925
> -3.75E-02
> 35 Broy./Diag. 0.40E+00 7.7 0.11017153 -4649.5833813599
> 8.51E-02
> 36 Broy./Diag. 0.40E+00 7.7 0.06877835 -4649.6008281933
> -1.74E-02
> 37 Broy./Diag. 0.40E+00 7.8 0.01645468 -4649.6384845655
> -3.77E-02
> 38 Broy./Diag. 0.40E+00 7.9 0.02326432 -4649.6297693665
> 8.72E-03
> 39 Broy./Diag. 0.40E+00 7.9 0.02365685 -4649.6304524367
> -6.83E-04
> 40 Broy./Diag. 0.40E+00 8.0 0.00731630 -4649.6350605681
> -4.61E-03
> ................
>
>
>
> If I try to use the Direct Mixing method, then I got the following error,
>
>
> SCF WAVEFUNCTION OPTIMIZATION
>
> Step Update method Time Convergence Total energy
> Change
>
> ------------------------------------------------------------------------------
> 1 P_Mix/Diag. 0.40E+00 4.8 0.84522228 -4649.6390448163
> -4.65E+03
> 2 P_Mix/Diag. 0.40E+00 5.7 0.61513185 -4649.5851421835
> 5.39E-02
> 3 P_Mix/Diag. 0.40E+00 5.7 1.17283140 -4649.5611028412
> 2.40E-02
> 4 P_Mix/Diag. 0.40E+00 5.7 2.67694470 -4649.0561208282
> 5.05E-01
> 5 P_Mix/Diag. 0.40E+00 5.7 3.42934913 -4643.4740477837
> 5.58E+00
> 6 P_Mix/Diag. 0.40E+00 5.7 230.21389660 -4410.1611504328
> 2.33E+02
> 7 P_Mix/Diag. 0.40E+00 5.7 156.53481655 -1382.0737857003
> 3.03E+03
> 8 P_Mix/Diag. 0.40E+00 5.7 94.91552555 -659.8356955460
> 7.22E+02
> 9 P_Mix/Diag. 0.40E+00 5.7 188.09969747 -1310.6533078835
> -6.51E+02
> 10 P_Mix/Diag. 0.40E+00 5.7 216.36335400 -959.3789707307
> 3.51E+02
> 11 P_Mix/Diag. 0.40E+00 5.7 140.36776849 -365.5503317119
> 5.94E+02
> 12 P_Mix/Diag. 0.40E+00 5.7 142.08262186 -2168.4593416849
> -1.80E+03
> 13 P_Mix/Diag. 0.40E+00 5.7 228.71605005 -1268.5148398924
> 9.00E+02
> 14 P_Mix/Diag. 0.40E+00 5.7 194.24691123 -465.4361346950
> 8.03E+02
> 15 P_Mix/Diag. 0.40E+00 5.7 153.73485990 -719.5330431343
> -2.54E+02
> 16 P_Mix/Diag. 0.40E+00 5.7 144.60489986 -1913.6921969993
> -1.19E+03
> 17 P_Mix/Diag. 0.40E+00 5.7 167.82643407 -1020.8436844043
> 8.93E+02
> 18 P_Mix/Diag. 0.40E+00 5.7 157.13131120 -1296.9022888403
> -2.76E+02
> 19 P_Mix/Diag. 0.40E+00 5.7 157.83912304 -1853.6888015754
> -5.57E+02
> 20 P_Mix/Diag. 0.40E+00 5.7 110.14805222 -650.2493834771
> 1.20E+03
>
>
> Can someone please help me?
>
> Best,
>
> Felipe
>
> --
> Felipe Zapata
> Postdoctoral research fellow
> Vrije Universiteit Amsterdam
> Division of Theoretical Chemistry
> De Boelelaan 1083
> NL-1081 HV Amsterdam
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20160708/5f26bf1d/attachment.htm>
More information about the CP2K-user
mailing list