[molpro-user] Does Molpro support analytical gradients with relativistic correction?

[Ryan Zaari]

Dear Molpro community,

Does Molpro support analytical gradients with relativistic correction?

Included below are 2 basis sets, one with relativistic Douglas-Kroll (Sapporo-DKH3-DZP-2012) and one without (Sapporo-DZP-2012). 
When calculations are done using analytic gradients and relativistically corrected basis sets the gradients produced are incorrect (last table).
Gradients seem to be correctly computed always when using numerical gradients and only when using non-relativistic basis sets + analytical gradients.
Correct results are also produced for analogous runs using TZVP and numerical or analytical gradients.

Any help or input would be appreciated,
Ryan Zaari



> ANALYTIC: Sapporo-DZP-2012
>  Atom          dE/dx     	   	       dE/dy           		dE/dz
>    1         0.000000000         0.000000000        -0.060235214
>    2         0.000000000         0.022641585         0.030117607
>    3         0.000000000        -0.022641585         0.030117607
> 

> NUMERICAL: Sapporo-DZP-2012
>  Atom          	dE/dx           	    dE/dy            		  dE/dz               
>    1         0.000000000        -0.000000000        -0.060306358            
>    2         0.000000000         0.022879742         0.030153179           
>    3        -0.000000000        -0.022879742         0.030153179       
> 

> NUMERIC: Sapporo-DKH3-DZP-2012
> Atom          dE/dx            			dE/dy                dE/dz
>    1         0.000000000        -0.000000000        -0.059500333
>   2         0.000000000         0.022180681         0.029750167
>   3        -0.000000000        -0.022180681         0.029750167 
> 

> ANALYTIC: Sapporo-DKH3-DZP-2012
>  Atom          dE/dx    		            dE/dy           		 dE/dz
>    1         0.000000000         0.000000000        -0.486146591
>    2         0.000000000         0.231072957         0.243073296
>    3         0.000000000        -0.231072957         0.243073296

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20131114/18d61a4a/attachment.html>