C2H2: Localization of Molecular Orbitals

Sample directory: adf/Cntrs.LocOrb_C2H2/

An illustration of the computation of localized molecular orbitals in C2H2.

The delocalized molecular orbitals as they result from the scf are localized in two different ways. In the first the three σ bonds are recombined only among themselves (no π bonds are mixed in), yielding two equivalent localized CH σ bonds and one localized σ bond. In a second step the localization of the remaining bond (the two π's) is performed, but this produces nothing new since no combination of the two π's is more localized then they are already by themselves.

$ADFBIN/adf << eor
title C2H2, localization Sigma and Pi separately

Atoms
C  0 0 .63
C 0 0 -.63
H 0 0 1.63
H 0 0 -1.63
END

Basis
 Type TZP
 Core Small
End

LocOrb
alfa  4 5
alfa  1 2 3
END

integration 4.0

end input
eor

In the first localization cycle the π-orbitals are left out: #4 and #5 in the list of all occupied valence MOs: first 3 MOs of the first irreducible representation (s), then the 2 from the second irrep (π). In the second localization step the first three (meanwhile localized) orbitals are kept aside.

With densf the local orbitals can be computed in a user-defined grid (for plotting purposes). densf requires a file with name TAPE21.

$ADFBIN/densf << eor

Grid
  0.  -5.  -5.
  100 100
  0. 0. 1. 10.
  0. 1. 0. 10.
End

Orbitals  Local
  1 2 3 4 5
End

END INPUT
eor

The program cntrs is applied to process the densf result file TAPE41.

$ADFBIN/cntrs << eor

SCAN
  0.01 0.02 0.04 0.08 0.16 0.32 0.64 1.28
  -0.01 -0.02 -0.04 -0.08 -0.16 -0.32 -0.64 -1.28
END

file ctr.a1
LocOrb%1 1.00

file ctr.a2
LocOrb%2 1.00

file ctr.a3
LocOrb%3 1.00

file ctr.a4
LocOrb%4 1.00

file ctr.a5
LocOrb%5 1.00

END INPUT
eor

Again, gnuplot may be used to display the result on your screen.

$gnuplot << eor
set term dumb  100 80
set output "outplot"
plot "ctr.a1" using 1:2 with lines
plot "ctr.a2" using 1:2 with lines
plot "ctr.a3" using 1:2 with lines
plot "ctr.a4" using 1:2 with lines
plot "ctr.a5" using 1:2 with lines
eor

cat outplot

This results in 5 pictures, the first one looking like:

                                                                       

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

            

The second illustration of the computation of localized molecular orbitals in C2H2 combines directly all MOs (σ and π). This yields 3 equivalent 'banana' bonds, mixtures of σ and π, and two equivalent pure σ bonds.

$ADFBIN/adf << eor
title C2H2   localization without frozen orbitals

Atoms
C 0 0 .63
C 0 0 -.63
H 0 0 1.63
H 0 0 -1.63
end

fragments
C  t21.C
H  t21.H
end

integration 4.0

locorb
end

end input
eor

 

Copyright      Terms of Use      Privacy Policy
Search:
Home
About
News
Sitemap
Contact
Why ADF?
Brochure
Demos
Trial Version
How to buy
Downloads
FAQ
Newsletters
Documentation
Community