Question regarding frequency analysis

Jörg Saßmannshausen jorg.sassm... at strath.ac.uk
Mon Mar 9 11:34:26 UTC 2009


Dear all,

I am fairly new to cp2k and after the workshop in Zurich I am beginning 
slowly to find my way around the program. I thought I started off a bit 
simple and run a geometry optimisation of 2 molecules in one box which I 
have done before using a different program. After successful convergence 
of that run, I decided to run the frequency calculation to see whether 
or not I actually reached a minimum (global or local does not matter 
here right now). So I had a look at the regtest files, picked out one 
which has a frequency calculation in it, amended my input file so it is 
using the restart files as an input. Judging from the output, it was 
reading the new geometry as I got that warning in the logfile:

  *** 18:44:39 WARNING in atoms_input:read_atoms_input err=-300 
      ***
  *** Overwriting coordinates. Active coordinates read from &COORD 
section.  ***
  *** Active coordinates READ from &COORD section atoms_input.F line 
130  ***

I assume that was ok?

I am somehow confused about the output. As it is rather long, I only 
copy the confusing bits in here. The run finished without any errors.

After the calculation of the Hessian I get:

     VIB| Cartesian Low frequencies --- -15.825     -13.679     -12.802 
     -2.2949
VIB| Cartesian Low frequencies --- -1.3729    -0.44402    -0.38034 
-0.21622
VIB| Cartesian Low frequencies ----0.15963
VIB| Frequencies after removal of the rotations and translations
*) VIB| Internal  Low frequencies --- -13.696     -12.335     -9.2308 
   -1.6738
VIB| Internal  Low frequencies ----0.86027    -0.37350    -0.23108E-01 
0.80471E-02
VIB| Internal  Low frequencies --- 0.45472E-01 0.13125     0.27273 
0.33375
VIB| Internal  Low frequencies --- 0.37453     0.42756     0.49331 
0.55433
VIB| Internal  Low frequencies --- 0.65046     0.85933      1.1011 
1.1978


Ok, I am aware that the first 6 frequencies can be dumped as they are 
usually due to rotation and translation of the molecule. That is what 
the program did for me, fine, that makes sense.
So from the next line onwards, which I marked with a *), I should get 
the frequencies which matter in my case. Surprise, surprise, I get 
negative ones, so I am not really in the miniumum but somewhere else.

However, further down in the logfile I will find that:


VIB|                        NORMAL MODES - CARTESIAN DISPLACEMENTS
VIB|
VIB|                         1                    2                    3
VIB|Frequency (cm^-1)  -812.246483          -770.814156          -666.811278
VIB|Intensities           0.000000             0.000000             0.000000
VIB|Red.Masses (a.u.)    13.143487            15.458038            10.885859
VIB|Frc consts (a.u.)    -0.014647            -0.013971            -0.005510

[...]

VIB|                         4                    5                    6
VIB|Frequency (cm^-1)  -283.949590          -203.564861          -134.130900
VIB|Intensities           0.000000             0.000000             0.000000
VIB|Red.Masses (a.u.)     2.055003             2.986810             2.007608
VIB|Frc consts (a.u.)    -0.000034            -0.000013            -0.000002


[...]

VIB|                         7                    8                    9
VIB|Frequency (cm^-1)   -33.363199            19.688092            46.801179
VIB|Intensities           0.000000             0.000000             0.000000
VIB|Red.Masses (a.u.)     3.570177             3.101716             2.496785
VIB|Frc consts (a.u.)     0.000000             0.000000             0.000000


Ok, doing the same as I did above, ignoring the first 6 ones would still 
leave me with the first one I am interested in. As I wanted to have the 
output in the molden-format so I can actually find out which part of the 
molecule is causing the negative frequency, I opened that file in 
molden, only to find out that I am having the 19.68 as the lowest 
frequency but not the -33.36.

So, now I am completly confused. Apparently, my molecule has not 
converged to a minimum as I got negative frequencies, but they are not 
in the molden file so how can I find out which part of the molecule is 
causing the negative frequency?
Is there something I did wrong here?

I have attached input files (input, forceeval and subsys) at the end of 
this email. I took the liberty to omit the coordinates and cut out some 
part of the subsys.inc as they are not part of the problem. As I said, 
the program runs ok, so there is more a confusion or a problem in the 
input file. I can, if requested, provide them of cause.

Many thanks for you help!

Best wishes from Glasgow!

Jörg

input-file:
! Possibly restart (if explicitly requested)
@IF 1
&EXT_RESTART ON
   RESTART_DEFAULT F
   RESTART_FILE_NAME FLUORGEOM-1.restart
   RESTART_POS T
   RESTART_COUNTERS T
&END EXT_RESTART
@ENDIF

&GLOBAL
   PROJECT  FLUORGEOM
   RUN_TYPE NORMAL_MODES
   PRINT_LEVEL LOW
&END GLOBAL

&VIBRATIONAL_ANALYSIS
    NPROC_REP 4
    DX 0.001
!   INTENSITIES
    &PRINT
      &PROGRAM_RUN_INFO ON
      &END
      &MOLDEN_VIB
       FILENAME=frequencies.mol
      &END
    &END

&END

@INCLUDE 'force_eval.inc'

force_eval.inc file:

@SET SCF_CONV   1.0E-6
@SET CUTOFF        450
@SET FUNCTIONAL    BLYP
@SET DECOUPLING     MT
@SET CHARGE          1

&FORCE_EVAL
   &PRINT
     &FORCES OFF
     &END
   &END

  &MIXED
   &PRINT
    &DIPOLE
    &END
   &END
  &END

   METHOD QS

   &DFT
     CHARGE ${CHARGE}
     &MGRID
       CUTOFF ${CUTOFF}
       NGRIDS 5
     &END MGRID
     &QS
       EPS_DEFAULT 1.0E-12
     &END QS
     &SCF
       SCF_GUESS ATOMIC
       EPS_SCF       ${SCF_CONV}
       MAX_SCF      30
       &PRINT
         &RESTART
           LOG_PRINT_KEY T
           FILENAME =./WFN_restart.wfn
           &EACH
             QS_SCF  0
           &END EACH
           ADD_LAST NUMERIC
           COMMON_ITERATION_LEVELS 1
         &END RESTART
       &END PRINT
       &OT ON
         MINIMIZER DIIS
            ! MINIMIZER CG
            ! PRECONDITIONER FULL_ALL
            ! PRECONDITIONER FULL_KINETIC
         PRECONDITIONER FULL_SINGLE_INVERSE
            !ENERGY_GAP  0.001
       &END OT
       &OUTER_SCF
         EPS_SCF ${SCF_CONV}
         MAX_SCF 5
       &END OUTER_SCF
     &END SCF
     &XC
       &XC_FUNCTIONAL ${FUNCTIONAL}
       &END XC_FUNCTIONAL
     &END XC
@IF ( ${DECOUPLING} == NONE )
     &POISSON
     &END
@ENDIF
@IF ( ${DECOUPLING} == MT )
     &POISSON
        POISSON_SOLVER MT
        PERIODIC NONE
        &MT
        &END
     &END
@ENDIF
@IF ( ${DECOUPLING} == WAVELET )
     &POISSON
        POISSON_SOLVER WAVELET
        PERIODIC NONE
        &WAVELET
          SCF_TYPE 60
        &END
     &END
@ENDIF
   &END DFT

   ! Include subsys
@INCLUDE 'subsys.inc'
&END FORCE_EVAL


subsys.inc file:
   &SUBSYS
     &CELL
#      ABC [angstrom] 23 15 10.0
       ABC [angstrom] 25 17 12
       PERIODIC NONE
     &END CELL
     &TOPOLOGY
       CONNECTIVITY OFF
       COORDINATE   XYZ
       COORD_FILE_NAME fluor.xyz
       CENTER_COORDINATES
     &END TOPOLOGY
     &KIND C
       POTENTIAL GTH
       &POTENTIAL
         # C GTH-BLYP-q4
     2    2
      0.33806609    2    -9.13626871     1.42925956
     2
      0.30232223    1     9.66551228
      0.28637912    0
       &END POTENTIAL
       &BASIS
         # C DZVP-GTH-BLYP
   2
   2  0  1  4  2  2
         4.3451975957   0.1511172750   0.0000000000  -0.0883417890 
0.0000000000
         1.2765552224  -0.0330272512   0.0000000000  -0.2797032766 
0.0000000000
         0.4047139801  -0.6994401873   0.0000000000  -0.4686400677 
0.0000000000
         0.1166096360  -0.3842706677   1.0000000000  -0.4085514115 
1.0000000000
   3  2  2  1  1
         0.6000000000   1.0000000000
       &END BASIS
     &END KIND
     &KIND F
       POTENTIAL GTH
       &POTENTIAL
         # F GTH-BLYP-q7
     2    5
      0.21384014    2   -21.90241518     3.27597572
     2
      0.19459028    1    23.79658295
      0.18786326    0
       &END POTENTIAL
       &BASIS
         # F DZVP-GTH-BLYP
   2
   2  0  1  4  2  2
        10.7284760684   0.1539953252   0.0000000000  -0.1034591141 
0.0000000000
         3.1496622427  -0.0503062083   0.0000000000  -0.3116781877 
0.0000000000
         0.9671794608  -0.7107682323   0.0000000000  -0.4776960745 
0.0000000000
         0.2640315133  -0.3648266470   1.0000000000  -0.3688578623 
1.0000000000
   3  2  2  1  1
         1.0000000000   1.0000000000
       &END BASIS
     &END KIND
     &KIND H
       POTENTIAL GTH
       &POTENTIAL
         # H GTH-BLYP-q1
     1
      0.20000000    2    -4.19596147     0.73049821
     0
       &END POTENTIAL
       &BASIS
         # H DZV-GTH-BLYP
[...]

       &END BASIS
     &END KIND
   &END SUBSYS


-- 
*************************************************************
Jörg Saßmannshausen
Research Fellow
University of Strathclyde
Department of Pure and Applied Chemistry
295 Cathedral St.
Glasgow
G1 1XL

email: jorg.sassm... at strath.ac.uk
web: http://sassy.formativ.net

Please avoid sending me Word or PowerPoint attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html





More information about the CP2K-user mailing list