Molcas Forum

huangtf

Member

Registered: 2017-04-23

Posts: 18

Geometry is converged to a Minimum but with imaginary frequencies

Dear team,

I try to run the calculation of CASPT2/ANO-RCC-VTZP (12/12) optimization for singlet NUN (nitrogen-uranium-nitrogen) molecule on the ground state employing Molcas 8.0 (molcas-8.0-15.06.18_CentOS_6.6_x86_64). As shown below, which come from my .log file, geometry is converged in 21 iterations to a minimum structure, however, I gain three imaginary frequencies after numerical differentiation is finished. (During the whole process of my CASPT2 geometry optimization, I often check my active space and find it is always correct. The last time occurs after geometry is converged, I use the final structure and corresponding .RasOrb to generate .grid, and then find that my active space still does not change.)

 19 -28059.69744139 -0.00000065 0.000552-0.002525 nrc006   0.021889  nrc004   -28059.69744660 RS-RFO  BFGS    0  
 20 -28059.69742367  0.00001772 0.005706 0.003292 nrc003   0.020563  nrc004   -28059.69744205 RS-RFO  BFGS    0  
 21 -28059.69744138 -0.00001771 0.000312 0.000174 nrc003   0.000348  nrc004   -28059.69744142 RS-RFO  BFGS    0  

       +----------------------------------+----------------------------------+
       +    Cartesian Displacements       +    Gradient in internals         +
       +  Value      Threshold Converged? +  Value      Threshold Converged? +
 +-----+----------------------------------+----------------------------------+
 + RMS + 0.3020E-03  0.1200E-02     Yes   + 0.1562E-03  0.3000E-03     Yes   +
 +-----+----------------------------------+----------------------------------+
 + Max + 0.3902E-03  0.1800E-02     Yes   + 0.1736E-03  0.4500E-03     Yes   +
 +-----+----------------------------------+----------------------------------+

 Geometry is converged in  21 iterations to a Minimum Structure

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

  Numerical differentiation is finished!

 Observe that the harmonic oscillator analysis is only valid at stationary points!

 Note that rotational and translational degrees have been automatically removed.


 Harmonic frequencies in cm-1

 IR Intensities in km/mol

                        1         2         3         4         5         6

     Frequency:     i1024.97    i36.21    i12.30    368.88   1092.24   1186.89                                          

     Intensity:    0.447E+02 0.533E+02 0.389E+01 0.394E+02 0.485E+03 0.329E+02

Please help me out with your invaluable suggestions. Thank you very much!

Best regards,
huangtf

Last edited by huangtf (2017-07-06 09:21:29)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: Geometry is converged to a Minimum but with imaginary frequencies

Perturb the geometry along the imaginary frequency mode and try optimizing from there. Linear molecules are tricky, numerical differentiations are tricky.

huangtf

Member

Registered: 2017-04-23

Posts: 18

Re: Geometry is converged to a Minimum but with imaginary frequencies

Dear Ignacio,

Thanks a lot for your kind and helpful reply last time! I've tried your method several times, and it really works. Unfortunately, because of limited time, I'm forced to use symmetry (C2v) and obtain a minimum with all real frequencies at last. (see below)

IR Intensities in km/mol

                        1         2         3         4         5         6

     Frequency:       i13.50      3.95     70.00    698.87   1036.59   1093.01                                          

     Intensity:    0.393E+00 0.155E-01 0.974E+02 0.847E+02 0.257E-01 0.511E+03

IR Intensities in km/mol

                        1         2

     Frequency:      1035.37   1093.52                                                                                  

     Intensity:    0.308E-03 0.511E+03

Sincerely yours,
huangtf

Last edited by huangtf (2017-07-24 09:20:45)