[molpro-user] Number of electrons in MRCC

Fri Nov 14 05:09:24 GMT 2014

 Dear Peterson,

I have added rhf before the mrcc calculation and have performed
the test calculation for ionization potential of Hf atom.

The MRCC requires ignore_error option since the RHF gives not
converged result.

The IP calculated by using CCSDT/aVTZ-PP is slightly different
from the reported value in JCP, 130, 164108, 2009.

The CCSDT/aVTZ-PP gives the ionization potential of 150.42
kcal/mol, therefore the delta CCSDT is obtained as 0.33 kcal/mol, which is
different from the reported value of 0.10 kcal/mol.

The output shows that the RHF energy is slightly changed even the
"maxit,1" is used. (SCF stops after 3 cycles)

I think this is the possible problem. What other options are required to
stop SCF just after the first cycle?

I have attached the input and output files of CCSDT ionization
potential calculation according to your comments.

** When I manually modified the MRCC input file, fort.56, I can
exactly reproduce the same results in your paper of 0.10 kcal/mol.

But Kallay gives me the comment that there are some problems of
integral list which is passed from Molpro to MRCC when the multi concifugration
is used as the reference.

I also attached 'Hf-CCSDT-IP1.zip' file which is manually modified
MRCC input and corresponding results.

Could you please check them?

Yours sincerely,

-----------------------원본메일-----------------------
보낸사람 : "Peterson, Kirk" <kipeters at wsu.edu>
받는사람: Hyungjun Kim <hyungjun96 at kaist.ac.kr>
참 조: molpro-user molpro-user <molpro-user at molpro.net>
보낸날짜: 2014-11-13 18:27:37 GMT +0900 (Asia/Seoul)
제목: Re: [molpro-user] Number of electrons in MRCC

Dear Hyungjun,

since the last program run before mrcc was multi, where the occ and closed cards do not correspond to a single determinant, mrcc got a little confused.  Before the call to mrcc, simply insert:

{rhf;
   occ,4,1,1,0,1,0,0;
closed,2,1,1,0,1,0,0;
    wf,12,1,2;
    start,2140.2;maxit,1}

Note that I've use the multi orbitals as initial guess and set the maximum number of iterations to 1. This will result in orbitals unchanged from your symmetry equivalenced ones from multi. You can then leave out all the options to mrcc except
 method=ccsdt.

regards,

-Kirk

On Nov 13, 2014, at 1:28 AM, Hyungjun Kim <hyungjun96 at kaist.ac.kr> wrote:

p{margin:0;padding:0;}

Dear Peterson, 

These are input and output files.

Actually the input file for Zr atom is made by the simple modification on the input file you sent to me for Hf atom.

Yours sincerely,

-Hyungjun

---------------------------------------------------
Hyungjun Kim, Ph. D. Student,
Quantum Chemistry Laboratory, Department of Chemistry,
Korea Advanced Institute of Science and Technology [KAIST],
Daejeon 305-701, Korea
e-mail: hyungjun96 at kaist.ac.kr or
hyungjun96 at gmail.com

phone : +82-42-350-2861

mobile phone : +82-10-8537-5051

web : 
http://qclab.kaist.ac.kr

-----------------------원본메일-----------------------
보낸사람 : "Peterson, Kirk" <kipeters at wsu.edu>
받는사람: Hyungjun Kim <hyungjun96 at kaist.ac.kr>
참 조: "molpro-user at molpro.net" <molpro-user at molpro.net>
보낸날짜: 2014-11-12 21:43:09 GMT +0900 (Asia/Seoul)
제목: Re: [molpro-user] Number of electrons in MRCC

Dear Hyungjiun,

Can you post your actual input? There are some tricks to using orbitals from multi in mrcc. 

Regards,

Kirk

Sent from my iPhone

On Nov 12, 2014, at 6:54 AM, Hyungjun Kim <hyungjun96 at kaist.ac.kr> wrote:

Dear Molpro developers,

I have reported the same problem to the MRCC forum and the developer of MRCC, Mihaly Kallay, suggests me to report this error to Molpro developer.

This is my post on the MRCC forum.

***

I have tried to calculate the electronic structure of Zr atom using CCSDT in MRCC program interfaced to Molpro. Stuttgart ECP is
 employed during this calculation. (Therefore only 12 electrons are left.)

I have symmetrized HF orbitals and have done CCSDT calculation using these orbitals.

While 4s and 4p orbitals are left as core electrons, 2 electrons in 5s, and 2 electrons in 4d orbitals are distributed, respectively.

When MRCC reads CASSCF MO, there are 4 electrons.
Frozen orbitals: 4 ( 1 1 1 0 1 0 0 0)
Active orbitals: 6 ( 3 0 0 1 0 1 1 0)
Active electrons: 4
Spin quantum number: 1.0

However, the output created by MRCC gives there are only 2 electrons (1alpha and 1beta).
Number of alpha electrons: 1
Number of beta electrons: 1

I don't know why the number of electrons is reduced from 4 to 2. Could you give me a guide to include '4d' electrons in MRCC calculation?

Any comments are welcome.

Yours sincerely,
***

The reply of Kallay to this question is as follows:

***

Dear Hyungjun,
The occupation vector and the specification of active orbitals passed over to mrcc is not correct. You should contact
 the molpro developers about this issue. The quick solution is to specify the directory where mrcc is executed, that is, the corresponding line in the molpro input should be extended, e.g., as
{MRCC,method=CCSDT,spatial=1,HF=0,nacto=0,dir='/scr/mrcc';
Then the scratch files required by mrcc will be saved to that directory. To correct the error edit file fort.56, and
 replace the last two line by the following ones:
2 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 1 0 0 0
Then execute dmrcc in that directory. Note that in this case an MRCCSDT calculation will be run. If you want to perform
 a simple CCSDT calculation, the last line of fort.56 must be deleted.
***

I have fixed the problem referring to Kallay's comment.

Could you please check this problem?

Yours sincerely,

---------------------------------------------------
Hyungjun Kim, Ph. D. Student,
Quantum Chemistry Laboratory, Department of Chemistry,
Korea Advanced Institute of Science and Technology [KAIST],
Daejeon 305-701, Korea
e-mail: hyungjun96 at kaist.ac.kr or
hyungjun96 at gmail.com
phone : +82-42-350-2861

mobile phone : +82-10-8537-5051

web : 
http://qclab.kaist.ac.kr

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

<zr-atom.inp><zr-atom.out>

<div style="padding:10px 10px;border:1px solid #CCC;width:650px;">
<div style="font-size:9pt;font-weight:bold;padding:3px 0;">대용량파일</div>
<table cellpadding="0" cellspacing="0" border="0" style="border:1px solid #BBB;font-size:9pt;width:100%;border-collapse:collapse;">
	<col style="width:340px;" />
	<col style="width:100px;" />
	<col style="width:150px;" />
	<tr style="background-color:#EFEFEF;">
 <th scope="col" style="font-size:9pt;font-weight:bold;padding:2px;border:1px solid #BBB;height:20px;line-height:20px;">파일 이름</th>
 <th scope="col" style="font-size:9pt;font-weight:bold;padding:2px;border:1px solid #BBB;height:20px;line-height:20px;">파일 용량</th>
 <th scope="col" style="font-size:9pt;font-weight:bold;padding:2px;border:1px solid #BBB;height:20px;line-height:20px;">다운로드 가능기간</th>
</tr>
<tr>
  <td style="font-size:9pt;padding:2px 5px;height:18px;line-height:18px;"><div style="width:340px;overflow:hidden;text-overflow:ellipsis;white-space:nowrap;"><a href='https://mail.kaist.ac.kr/Mail?act=BIGFILE_DOWNLOAD&ukey=546591193fd858f6f4a644f6&mhost=kaist.ac.kr&language=ko' target="_blank">Hf-CCSDT-IP1.zip</a></div></td>
  <td align="right" style="font-size:9pt;padding:2px 5px;height:18px;line-height:18px;">19.6 MB</td>
  <td align="center" style="font-size:9pt;padding:2px 5px;height:18px;line-height:18px;">2014-12-04</td>
</tr></table>
</div>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20141114/3b913bf9/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: hf-ccsdt-ip1.inp
Type: application/octet-stream
Size: 1638 bytes
Desc: not available
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20141114/3b913bf9/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: hf-ccsdt-ip1.out
Type: application/octet-stream
Size: 95234 bytes
Desc: not available
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20141114/3b913bf9/attachment-0003.obj>