[molpro-user] Trouble computing nonadiabatic couplings with DDR

[Terry Frankcombe]

I haven't looked closely at the orbitals you're using, but we have found
that this type of thing can happen if the orbitals do not form a totally
symmetric set.  By that I mean if you include some p orbitals but not
the complete shell, the wavefunction can develop some directionality
which gets picked up in the starting vectors for subsequent
calculations.

Something to consider.


On Wed, 2009-07-29 at 22:34 +0200, Robert Send wrote:
> Dear Molpro-users,
> 
> I am trying to compute nonadiabatic coupling vectors with the input 
> script below, using the DDR method. However, depending on whether or 
> not I compute the couplings at bond distance r=0.5 before the rest of 
> the distances (1.0,2.0,3.0,4.0), I get different results. In other 
> words,
> r=[0.5,1.0,2.0,3.0,4.0]
> gives different couplings from
> r=[1.0,2.0,3.0,4.0].
> 
> I tried to stick as close as possible to the LiF input script from the 
> molpro online manual.
> 
> Thanks a lot for your help,
> 
> Robert
> 
> Here comes the input:
> 
> memory,160000000
> basis,h=aug-cc-pvdz     r=[1.0,2.0,3.0,4.0]
> dr=0.001               geometry={
> x,y
> h
> h,h,rbh
> }
> 
> hf
> 
> do i=1,#r             rbh=r(i)          hf
> 
>     {casscf
>     closed
>     occ,10,4,4,0
>     wf,2,1,0
>     state,2,1,2
>     orbital,2140.2
>     }
> 
>    {ci;
>    wf,2,1,0
>    state,2,1,2   noexc            save,6000.2;    dm,8000.2}
> 
>    rbh=r(i)+dr
> 
>     {casscf
>     closed
>     occ,10,4,4,0
>     wf,2,1,0
>     state,2,1,2
>     start,2140.2
>     orbital,2141.2
>     }
> 
>    {ci;
>     wf,2,1,0
>     state,2,1,2
>     noexc;
>     save,6001.2}
> 
>    {ci;trans,6000.2,6001.2;       dm,8100.2}
> 
>     rbh=r(i)-dr
> 
>    {casscf
>     closed
>     occ,10,4,4,0
>     wf,2,1,0
>     state,2,1,2
>     start,2140.2
>     orbital,2142.2
>     }
> 
>    {ci;
>     wf,2,1,0
>     state,2,1,2
>     noexc;save,6002.2}
> 
>    {ci;trans,6000.2,6002.2;         dm,8200.2}
> 
>    {ddr,dr,2140.2,2141.2,8100.2}    nacme1p(i)=nacme                 
> {ddr,-dr,2140.2,2142.2,8200.2}   nacme1m(i)=nacme
> 
>    {ddr,2*dr                         orbital,2140.2,2141.2,2142.2;    
> density,8000.2,8100.2,8200.2}     nacme2(i)=nacme
> 
> end do
> 
> nacmeav=(nacme1p+nacme1m)*0.5      
> table,r,nacme1p,nacme1m,nacmeav,nacme2  title,Non-adiabatic couplings 
> for H2
> 
> ---------------------------------------
> 
> Robert Send
> Theoretische Chemie
> Institut für Physikalische Chemie
> Universität Karlsruhe
> 
> +49-721-608-8344
> 
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
-- 
Dr. Terry Frankcombe
Research School of Chemistry, Australian National University
Ph: (+61) 0417 163 509    Skype: terry.frankcombe