[molpro-user] Convergence Issues With DIP/QUAD

[Gerald Knizia]

Michael Morris wrote:
> I'm having some convergence issues when trying to apply a DIP or QUAD 
> perturbation to my structures. The error arises when MOLPRO gets to 
> performing MP2/CCSD etc calculations: [..]
> //
> /HF/
> /i=1/
> /DIP+,0.0,0.0,0.001/
> /{MP2/
> /CORE,0}/

While I think that the code here is supposed to work (but would 
calculate orbial unrelaxed properties, which here would likely be very 
wrong), I guess that it can be solved by moving the dip+ before the 
Hartree-Fock. Otherwise the perturbation is not included in the 
mean-field orbitals and mysterious things might happen in the 
correlation treatment following it.
-- 
Gerald Knizia