Molcas Forum

connie_zqq

Member

Registered: 2017-02-14

Posts: 1

TESSERA problems in PCM calculation using Molcas8.0!

Dear All,

     Happy Valentine's Day!
     I currently ran the CASPT2 calculation in the PCM model, but repeatedly encountered the internal inconsistency problems.

     The input is as follows:
     

>>>>>>> EXPORT MOLCAS_MOLDEN=0N       
>>COPY s0_pcm_pt2.InpOrb INPORB                             
&GATEWAY                                                   
title                                                      
   title:s0_pcm_hf                                         
EXPERT                                                     
coord = s0_pcm_hf.xyz; basis = 6-31G*; group = c1          
RF-input                                                   
 PCM-model; solvent= water                                 
 End of RF-input                                           
                                       
&SEWARD                                     
                             
 &RASSCF &END                
Title                        
s0_pcm_pt2                   
Symmetry                     
 1                           
Spin                         
 1                           
Nactel                       
 12 0 0                      
Inactive                     
 79                          
Ras2                         
 10                          
thrs                         
 0.000001 0.001 0.001
Ciroot                       
 4 4                         
 1 2 3 4                     
 1 1 1 1                                 
 RFRoot=1         
Lumorb            
End of input 

 &CASPT2 &END               
Title                       
s0_pcm_pt2                  
Lroot                       
 1                          
Shift                       
 0.2                        
IPEA                        
 0.0                        
MaxIter                     
 100                        
 RFPert                     
End of Input                
                            
 &RASSI &END                
Nrofjobiphs                 
 1     4                    
 1 2 3 4                    
  RFPert                    
End of input                
                                                                      
 &RASSCF &END                
Title                        
s0_pcm_pt2                   
Symmetry                     
 1                           
Spin                         
 1                           
Nactel                       
 12 0 0                      
Inactive                     
 79                          
Ras2                         
 10                          
thrs                         
 0.000001 0.001 0.001        
Ciroot                       
 4 4                         
 1 2 3 4                     
 1 1 1 1                     
NONEquilibrium               
 RFRoot=2        
Lumorb           
End of input                                                         
                             
 &CASPT2 &END                
Title                        
s0_pcm_pt2                   
Lroot                        
 1                           
Shift                        
 0.2                         
IPEA                         
 0.0                         
MaxIter                      
 100                         
RFPErt                       
End of Input                 
                             
 &CASPT2 &END                
Title                        
s0_pcm_pt2                   
Lroot                        
 2                           
Shift                        
 0.2                         
IPEA                         
 0.0                         
MaxIter                      
 100                         
RFPErt                       
End of Input                 
                             

The error information of output likes that:

      Polarizable Continuum Model (PCM) activated                                                            
      Solvent:water                                                                                          
      Version: Conductor                                                                                     
      Average area for surface element on the cavity boundary: 0.4000 Angstrom^2                             
      Minimum radius for added spheres: 0.2000 Angstrom                                                      
      Calculation type    : equilibrium                                                                      
                                                                                                             
                                                                                                             
 TESSERA: too many vertices in a tessera                                                                     
--- Stop Module:  rasscf at Sat Feb 11 22:25:02 2017 /rc= _INTERNAL_ERROR_ ---                               
                                                                                                             
         Some internal inconsistency of the code was detected                                                
                                                                                                             
Non-zero return code - check program input/output                                                            
--- Stop Module:  auto at Sat Feb 11 22:25:02 2017 /rc= _INTERNAL_ERROR_ ---                                 
--- Module auto spent 4 minutes and 19 seconds                                                               

    The process of calculation is terminated before the first RASSCF.
    To exclude the problem of initial orbitals (InpOrb), I also tried the only SCF calculation, but obtained the same information.
    I read the related part in Molcas 8.0 manual, the keyword AAre means the average area of tesserae, but changing the value to 0.2 was of no use.

    I spared not efforts to search the similar prolems online, but found nothing! In fact, this system only contains 40 atoms.
   
    I also have some questions.
    1. Concerning the ground state RASSCF, if the CIROOT and RFRoot=1 are required.       
    2. For &RASSI part, if RFPert is required.     

    Thank you for your sincere reply !

Best Wishes,
QQ

Last edited by connie_zqq (2017-02-14 07:52:06)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: TESSERA problems in PCM calculation using Molcas8.0!

That seems to be a problem when creating the PCM cavity, unrelated to the electronic structure, orbitals, etc. Perhaps your molecule has a "strange" structure or uncommon atoms... but I don't know how to fix that.

Regarding your other questions:

- CIRoot is required if you want to perform a state-average calculation, otherwise you will get the ground state only. RFRoot is necessary whenever you use CIRoot, there will be some default value (probably the highest root), but don't rely on it, always specify RFRoot.

- RFPert in RASSI is needed if you want to include the effect of the reaction field (the solvent). But note that it will be "wrong" anyway. Ideally each state will have a diferent non-equilibrium reaction field, but RASSI can use only a single one (as far as I know).

valera

Administrator

Registered: 2015-11-03

Posts: 129

Re: TESSERA problems in PCM calculation using Molcas8.0!

connie_zqq,
Think about spheres with different size around atoms. If they all large enough - the resulting surface is smooth. But for small ones - the surface is very fragmented.. A while ago the limitation related to the amount of tessera was lifted. So, one solution is to send (as a bug report) complete input, so I can check - does it work in developer's version or not. Another solution - increase the radii of spheres.