How to define fragments when studying Mn-Mn bond in Mn2(CO)10?

[Oscar Ac]

Hi,

I've been having a lot of problems with an L-VBSCF/6-31G* calculation to study the Mn-Mn bond in the Mn2(CO)10 molecule. I suspect my main problem is that I can't define the proper fragments to study that bond as if it were a "diatomic" case, treatin Mn(CO)5 as a group. I've proposed the following:

$frag
11 11 11 11
spzdxxdyydzzfzzzfxxzfyyz 1 3 4 5 6 7 8 9 10 11 12; frag sz of Mn1
spzdxxdyydzzfzzzfxxzfyyz 2 13 14 15 16 17 18 19 20 21 22; frag sz of Mn2
pxpydxydxzdyzfxxxfyyyfxyyfxxyfxzzfyzzfxyz 1 3 4 5 6 7 8 9 10 11 12 ; frag xy of Mn1
pxpydxydxzdyzfxxxfyyyfxyyfxxyfxzzfyzzfxyz 2 13 14 15 16 17 18 19 20 21 22 ; frag xy of Mn2
$end

where both Mn atoms (1 and 2) lie on the z axis, every atom being numbered as in the geometry appearing in the LOG file I adjoin.

I know my calculation is wrong (whose XMO file I adjoin) because the "COEFFICIENTS OF STRUCTURES" section indicates that one of the ionic structures (1:94  96  96) has a major contribution to the VB wavefunction, instead of the covalent one (1:94  95  96). I also suspect this could be due to an incorrect choice of MOs to the $gus section and/or an inadequate Lewis structure for this complex.

Could somebody please help me identifying how to improve the fragments I proposed in the $frag section, in case my Lewis structure is adequate? Or could you please advise me on how to make a proper correspondence between the MO to the VB orbitals in the $gus section, in case I defined correctly the fragments (and made an approximate Lewis structure for this molecules in the $orb section)?

I've honestly made too many attempts of changing the $gus section with different MOs that I'm feeling very lost on how to fix this calculation. Most of the times I get this error:

ERROR RETURN OF LINE SEARCH: INFO=  3

POSSIBLE CAUSES: FUNCTION OR GRADIENT ARE INCORRECT

Thank you very much in advance.

Oscar

[应富鸣]

Dear Oscar,

I just saw your mail, sorry for the delay.

I will check and answer you.

Best regards,

Fuming Ying

在 2019年7月1日星期一 UTC+8下午7:28:40，Oscar Ac写道：

[应富鸣]

OK.

Oscar Ac <o.aguil...@gmail.com> 于2019年8月13日周二 下午11:26写道：

Dear Fuming Ying,

I would also be very thankful if you could please write your reply to Sebastian Kraus too (whose e-mail address is also written on the receivers), since he has been helping us to install both, Gamess and XMVB. We think that a successful compilation of this joint version with both programs could be possible when having access to the source files. If there is a possibility to do so, we would be very pleased to follow your instructions in order to do so, because VB calculations are essential for us now. Otherwise, we would be grateful for your precise instructions on how to link both programs to avoid the error(s) pointed on my last e-mail.

Best regards,

Oscar Aguilar Cuevas