[CP2K:7719] Transition State Optimization with Dimer Method

[luca]

Dear Natalie
the dimer method is a tool to search
TS structures and it can be used to refine (or validate) TS provided by
other methods.
I have just looked your input, the dimer method section is ok; however,
the calculation of the frequency is very sensitive, so,  I would suggest
you to increase the accuracy,  decreasing RMS displacement and so on.
This is my usual set-up:
    RMS_DR         1.0E-04 #   1.0E-04
    MAX_DR         3.0E-04 #   3.0E-05
    RMS_FORCE      1.0E-04 #   1.0E-04
    MAX_FORCE      3.0E-04 #   3.0E-05
Actually, to obtain a single ifreq more tight  parameters are required.
Pay attention, the dimer method shows a slow convergence, so it requires
many steps.
In the attached file you can find parameters that I used in
DOI:10.1021/jo3023439

Cheers,
Luca

On Thu, 2016-05-05 at 17:27 -0700, Natalie Austin wrote:
> Hello,
> 
> I recently attempted to calculate a transition state for CO2
> dissociation using CI-NEB. Unfortunately, the NEB did not converge
> after 66 steps ( It ran once up to 32 steps then I had to resubmit it
> to run another 34 steps) and so I stopped the calculation because I
> noticed that the MAX FORCE for the NEB calculation was just
> fluctuating(see below). Would anyone be able to indicate why this is
> the case? I used the same input for a located a transition from
> physisorbed CO2 to chemisorbed CO2 on the same metal system and I did
> not have this problem. 
> 
> I did come across a case in the steps where all criteria was met
> except for the MAX force as shown below: 
> 
>  RMS DISPLACEMENT =   0.00013     [ 0.00100]
> (YES)
>  MAX DISPLACEMENT =   0.00225     [ 0.00300]
> (YES)
>  RMS FORCE        =   0.00025     [ 0.00100]
> (YES)
>  MAX FORCE        =   0.00407     [ 0.00300]
> ( NO)
> 
> So I took that structure and ran vibrational analysis on it and found
> that it had 3 imaginary frequencies as shown below:
> 
> 
> As a previous user suggested I then tried to use the Transition State
> Dimer method to see if I could guide this structure to a transition
> state by providing the vectors in DIMER _VECTOR for the mode
> corresponding to the state that I was looking for. I was able to
> locate a potential transition state. I've providing the transition
> state coordinates below. Does it make sense to use CI-NEB and
> reoptimize with the Transition state search in cases where it is
> difficult to converge the NEB?  Is this a common occurrence? I'm just
> trying to make sure that I'm locating transition states in an
> appropriate way. 
> 
> Thanks,
> 
> Natalie
>  
> 
> -- 
> 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 cp2k+uns... at googlegroups.com.
> To post to this group, send email to cp... at googlegroups.com.
> Visit this group at https://groups.google.com/group/cp2k.
> For more options, visit https://groups.google.com/d/optout.

-------------- next part --------------
   &GEO_OPT
    TYPE TRANSITION_STATE
    OPTIMIZER      CG
    MAX_ITER       2000
    RMS_DR         0.10E-04
    MAX_DR         0.30E-04
    RMS_FORCE      0.10E-04
    MAX_FORCE      0.30E-04
    &CG                                                                         
      MAX_STEEP_STEPS 0                                                         
      &LINE_SEARCH                                                              
       TYPE 2PNT                                                                
       &2PNT                                                                    
         MAX_ALLOWED_STEP 0.20                                                  
       &END                                                                     
      &END                                                                      
    &END CG
  
     &TRANSITION_STATE                                                               
      METHOD DIMER                                                                  
      &DIMER                                                                        
        DR [angstrom] 0.01                                                          
        ANGLE_TOLERANCE [deg] 2.0                                                   
        INTERPOLATE_GRADIENT  T                                                     
        &ROT_OPT                                                                    
          OPTIMIZER CG                                                              
          MAX_ITER  12                                                               
          &CG                                                                       
            MAX_STEEP_STEPS 0                                                       
            &LINE_SEARCH                                                            
              TYPE 2PNT  
               &2PNT		
	          MAX_ALLOWED_STEP 0.10
	       &END   
            &END                                                                    
          &END                                                                      
        &END                                                                        
      &END                                                                          
    &END 
  &END GEO_OPT