$CONICL group               (relevant if RUNTYP=CONICAL)                        
                                                                                
   This group governs a search for the lowest energy on the                     
3N-7 dimensional "seam" of intersection of two electronic                       
potential energy surfaces of the same spin and space                            
symmetry.  Such Conical Intersections (CI) are important in                     
photochemistry, where they serve as "funnels" for the                           
transfer from an excited state to a lower state.  See                           
RUNTYP=MEX and the $MEX input for the simpler case where                        
the two surfaces differ by either space or spin symmetry.                       
                                                                                
   Three search procedures are given, one of which requires                     
the non-adiabatic coupling matrix element (NACME), and two                      
others which do not require NACME information.  The conical                     
intersection search is available only for MCSCF (for which                      
NACME are available) or for TD-DFT potential surfaces                           
(where NACME are not available).  The TD-DFT must be used                       
in the Tamm/Dancoff approximation (see TAMMD in $TDDFT),                        
but can be either conventional or spin-flip.                                    
                                                                                
   The search utilizes some of the options of $STATPT, but                      
note that the Schlegel stepper and HESS=CALC are not                            
permitted.  It may be reasonable to try the RFO stepper                         
sometimes.  The search can only be run in Cartesian                             
coordinates.  Restarts are possible only by updating the                        
coordinates in $DATA.                                                           
                                                                                
   At present, the only solvation model that is supported                       
is conventional TD-DFT with EFP1.                                               
                                                                                
OPTTYP = search procedure choice, see references below!                         
       = GPWNAC  Gradient Projection with NACME, so this                        
                 is only available for MCSCF.                                   
       = BPUPD   branching plane updating method (default)                      
       = PENALTY penalty-constrained optimization method                        
                                                                                
Note that for MCSCF surfaces, if state-averaging is used,                       
the program executes the code needed to produce NACME                           
vectors, to producing the state-averaged gradients.  There                      
is essentially no extra time required to produce also the                       
NACME, hence the GPWNAC stepper might as well be used.                          
                                                                                
IXROOT = array of two states whose CI point is sought.                          
         For example, this might be IXROOT(1)=2,3                               
         The roots are counted exactly the same as IROOT in                     
         the $DET or $TDDFT input groups.  For the latter                       
         case, set IXROOT to 0 if you want the ground state                     
         to be one of the two surfaces searched on.                             
         There is no default for IXROOT!                                        
                                                                                
SYMOFF = flag to switch off point group symmetry,                               
         the default is .TRUE.                                                  
                                                                                
DEBUG  = flag to print debugging info, default is .FALSE.                       
                                                                                
The following are meaningful only for OPTTYP=PENALTY:                           
                                                                                
TOLSTP = energy difference tolerance                                            
         default=1d-6 Hartree                                                   
                                                                                
TOLGRD = gradient convergence tolerance                                         
         default=5d-3 Hartree/Bohr                                              
                                                                                
ALPHA  = parameter ensuring a singularity free penalty,                         
         default=0.02 Hartree                                                   
                                                                                
SIGMA  = Lagrange multiplier for the penalty term.  In                          
         case the energy gap between the states is not                          
         acceptable at the CI point, increase the value.                        
         default = 3.5 (unitless)                                               
                                                                                
An understanding of the search procedures can be gained by                      
reading the following papers:                                                   
                                                                                
  Gradient Projection with NACME:                                               
    M.J.Bearpark, M.A.Robb, H.B.Schlegel                                        
      Chem.Phys.Lett. 223, 269(1994)                                            
  Branching Plane Updating method:                                              
    S.Maeda, K.Ohno, K.Morokuma                                                 
      J.Chem.Theor Comput. 6, 1538(2010)                                        
  Penalty constrained update method:                                            
    B.G.Levine, C.Ko, J.Quenneville, T.J.Martinez                               
      Mol.Phys. 104, 1039(2006)                                                 
    B.G.Levine, J.D.Coe, T.J.Martinez                                           
      J.Phys.Chem.B 112, 405(2008)                                              
                                                                                
A comparative study of the first two procedures is                              
  T.W.Keal, A.Koslowski, W.Thiel                                                
  Theoret.Chem.Acc. 118, 837(2007)                                              
                                                                                
===========================================================                     
                                                                                
                                                                                
                                                                                
===========================================================                     

generated on 7/7/2017