[molpro-user] Molpro-user Digest, Vol 103, Issue 15

Mon Feb 27 13:34:38 CET 2017

```Hello,
you may perhaps like to consider the following (part of) input, that I used for one of our latest projects on the diatomic CuAu, for diagonalization of the SO matrix on top of the non-rotated MRCI wavefunctions.
What the input does is the following: 1) Perform the spin-free (Lambda-S) MRCI calculations (considering scalar relativistic of course) and switch off the MRCI rotations (SWAP,ROTREF=-1). The rationale for fixing the MRCI eigenstates (SWAP,ROTREF=-1) is to make sure that the diagonal matrix elements (spin-free energies which are taken from a previous spin-orbit free job) in "hlsdiag" are assigned correctly to the desired spin-free MRCI wavefunction. Dynamic correlation might lead to fllipping and/or overlapping between the reference functions. Nonetheless, It is also possible to work with rotated MRCI wavefunctions. But, it requires very much more attention for the correct assignments, because the wavefunctions get rotated and they overlap with each other if large number of states are involved.2) save the non-rotated MRCI wavefunctions using the "save" keyword.3) define the diagonal spin-free MRCI energies that are calculated in a previous (higher-level) job via hlsdiag=[...]. Note that the order of energies given should be identical to the order of eigenstates saved in step one.4) finally, using {ci;...}, diagonal spin-free energies and the wavefunctions to which the diagonal energies are to be assigned need to be written in correct ordering.
I hope it helps.Best wishes.Davood.
input:==========={MRCI,SWAP,ROTREF=-1;wf,108,1,0;state,6,1,3,4,7,10,11;pspace,1;;save,3010.1;}{MRCI,SWAP,ROTREF=-1;wf,108,1,0;ref,4;state,6,2,5,6,8,9,12;pspace,1;save,3020.1;}{MRCI,SWAP,ROTREF=-1;wf,108,4,0;state,2,5,7;pspace,1;save,3030.1;}{MRCI,SWAP,ROTREF=-1;wf,108,4,0;ref,1;state,6,1,2,3,4,6,8;pspace,1;save,3040.1;}{MRCI,SWAP,ROTREF=-1;wf,108,2,0;ref,3;state,8;pspace,1;save,3050.1;}{MRCI,SWAP,ROTREF=-1;wf,108,3,0;ref,2;state,8;pspace,1;save,3060.1;}{MRCI,SWAP,ROTREF=-1;wf,108,1,2;state,6,1,3,5,8,10,11;pspace,1;save,3012.1;}{MRCI,SWAP,ROTREF=-1;wf,108,1,2;ref,4;state,6,2,4,6,7,9,12;pspace,1;save,3022.1;}{MRCI,SWAP,ROTREF=-1;wf,108,4,2;state,2,4,6;pspace,1;save,3032.1;}{MRCI,SWAP,ROTREF=-1;wf,108,4,2;ref,1;state,6,1,2,3,5,7,8;pspace,1;save,3042.1;}{MRCI,SWAP,ROTREF=-1;wf,108,2,2;ref,3;state,8;pspace,1;save,3052.1;}{MRCI,SWAP,ROTREF=-1;wf,108,3,2;ref,2;state,8;pspace,1;save,3062.1;}  ! SO calculation with valence Davidson-corrected MRCI diagonal energies hlsdiag=[-20638.47166,-20638.37106,-20638.24697,-20638.21554,-20638.20991,-20638.33891,-20638.35404,-20638.23661,-20638.22393,-20638.21558,-20638.21383,-20638.31034,-20638.21473,-20638.2009,-20638.35404,-20638.23661,-20638.22393,-20638.21558,-20638.21383,-20638.31034,-20638.36693,-20638.23836,-20638.21633,-20638.21561,-20638.21657,-20638.20795,-20638.2073,-20638.28475,-20638.36693,-20638.23836,-20638.21633,-20638.21561,-20638.21657,-20638.20795,-20638.2073,-20638.28475,-20638.38351,-20638.36899,-20638.27379,-20638.21463,-20638.19863,-20638.23744,-20638.35865,-20638.23992,-20638.22405,-20638.21493,-20638.19964,-20638.31125,-20638.21938,-20638.21563,-20638.35865,-20638.23992,-20638.22405,-20638.21493,-20638.19964,-20638.31125,-20638.37549,-20638.2413,-20638.21816,-20638.22059,-20638.21731,-20638.20911,-20638.20842,-20638.28475,-20638.37549,-20638.2413,-20638.21816,-20638.22059,-20638.21731,-20638.20911,-20638.20842,-20638.28475]  {ci;hlsmat,ls,3010.1,3020.1,3030.1,3040.1,3050.1,3060.1,3012.1,3022.1,3032.1,3042.1,3052.1,3062.1;print,vls=3}

From: "molpro-user-request at molpro.net" <molpro-user-request at molpro.net>
To: molpro-user at molpro.net
Sent: Monday, 27 February 2017, 14:30:01
Subject: Molpro-user Digest, Vol 103, Issue 15

Send Molpro-user mailing list submissions to
molpro-user at molpro.net

To subscribe or unsubscribe via the World Wide Web, visit
http://www.molpro.net/mailman/listinfo/molpro-user
or, via email, send a message with subject or body 'help' to
molpro-user-request at molpro.net

You can reach the person managing the list at
molpro-user-owner at molpro.net

than "Re: Contents of Molpro-user digest..."

Today's Topics:

1. Spin-orbit coupling and HLSDIAG (Hoffman, Gerald)

----------------------------------------------------------------------

Message: 1
Date: Fri, 24 Feb 2017 21:49:51 +0000
From: "Hoffman, Gerald" <GHOFFMAN at edinboro.edu>
To: "molpro-user at molpro.net" <molpro-user at molpro.net>
Subject: [molpro-user] Spin-orbit coupling and HLSDIAG
Message-ID:
<624C71FA8F84914CBF467494BC2F1F459E8F27BE at EX02.eup.edinboro.edu>
Content-Type: text/plain; charset="iso-8859-1"

I am attempting to calculate spin-orbit energies of a system whose total energies are calculated at the level of MRCI including Davidson cluster correction.  However, calculation of the spin-orbit energies are based upon the MRCI energies without cluster correction.  I realize that it is possible to substitute the cluster corrected energies for the uncorrected ones using the variable HLSDIAG, but it is not at all clear how to go about doing so.  An example of what the input should look like would be helpful.

Could anyone please show me how to do this?

Thank you,

Gerry Hoffman

Gerald J. Hoffman, Ph. D.

Associate Professor of Chemistry

Edinboro University of Pennsylvania

Edinboro, PA 16444

814-732-2813

ghoffman at edinboro.edu<mailto:ghoffman at edinboro.edu>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20170224/67a4a856/attachment-0001.html>

------------------------------

Subject: Digest Footer

_______________________________________________
Molpro-user mailing list
Molpro-user at molpro.net
http://www.molpro.net/mailman/listinfo/molpro-user

------------------------------

End of Molpro-user Digest, Vol 103, Issue 15
********************************************

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20170227/9068286c/attachment.html>
```