[molpro-user] Discrepancy in energy values
Peterson, Kirk
kipeters at wsu.edu
Sun Jan 8 17:13:02 CET 2017
Dear Bhargava,
presumably this is just due to the different initial guesses to multi in these two cases. In a single point, the default for multi will be to take the orbitals from the preceding hf calculation since there are no other orbitals available. In the loop calculation, it will take its guess from the multi orbitals of the previously computed geometry. Since this geometry is near dissociation, I would imagine the hf orbitals to be a very poor starting guess for your casscf calculation.
regards,
-Kirk
PS - if you turn on orbital printing you’ll probably see that the output orbitals from multi in these two cases are quite different.
> On Jan 7, 2017, at 10:20 PM, Bhargava Anusuri <bhargava.anusuri at gmail.com> wrote:
>
> Dear users,
> I am computing potential energy surface of CCH+ using Molpro. I have observed that there is a difference in energy values for the same geometry when it was computed in a loop and when it is computed separately as a single point.
>
> r(C-H) = 2.4 a.u , r(C-C) = 21.9 a.u, C-C-H angle = 179 degree : -75.74572157 hartree (Single point)
> -75.78211095 hartree (in a loop)
>
> My input for loop is
>
> ***,CCH+
> memory,70,m
> gthresh,energy=0.32d-6
>
> r=1.0d0
> rc=2.4d0
> theta = 179 deg
>
> geometry={
> C1;
> C2,C1,r;
> H,C2,rc,C1,theta}
>
> basis=aug-cc-pVQZ
>
> do i=1,56
>
> {hf;wf,12,2,2}
>
> {multi;
> maxiter,40;
> pspace,0.3;
> wf,12,2,2}
>
> {ci;
> option,maxit=90
> option,maxiti=9000;
> !option,nstati=2;
> option,cluster=2;
> wf,12,2,2}
>
> e(i)=energy
> ed(i)=energd2
>
> if(i.lt.3) then
> r=r+0.2d0
> elseif(i.ge.3).and.(i.lt.24) then
> r=r+0.1d0
> elseif(i.ge.24).and.(i.lt.36) then
> r=r+0.2d0
> elseif(i.ge.36).and.(i.lt.44) then
> r=r+0.5d0
> else
> r=r+1.0d0
> endif
>
> {table,sr,capr,e,ed
> save,179d.dat,new
> }
>
> enddo
>
> Is there something I am doing wrong? Any help is greatly appreciated. Thanks in advance.
>
> Bhargava,
> Dept. of Chemistry,
> IIT Madras.
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.molpro.net_mailman_listinfo_molpro-2Duser&d=DgICAg&c=C3yme8gMkxg_ihJNXS06ZyWk4EJm8LdrrvxQb-Je7sw&r=mqaXWsZ0ghQR-pXvPLjY_g&m=4mWmtOdPKgtxS8lxfgEPm1Bx4i3Cd8aKzwv1Zkd6Uis&s=ce1LAGDAJpxRqo2oyrDlkXN5z9TwwtObjQCKM9-3pRg&e=
More information about the Molpro-user
mailing list