Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly  
We are NATO-free zone

Re^4: Too many iterations in CPHF (MP2 gradient calc)

Alex Granovsky

Dear David,

as to search button, you need to open Firefly's main page:

then pressing "Discussion" button will open this forum as a frame
of the main page and there will be buttons on the left including
search button

As to incorrect bond length, you incorrectly specified units=angs
in the $contrl group while your geometry assumes units=bohr

Hope this helps.

Kind regards,

On Fri Mar 25 '16 5:02am, David Shobe wrote
>I have to take some of this back.  I opened the output file in Facio, and all the bonds have been stretched about 2x their normal size.  I don't know how that happened.  Anyway, that likely caused the convergence problem.  I'm deservedly embarrassed about this, but it's hard to know where to look when something goes wrong.

>Thanks for everyone's patience,
>--David Shobe
>On Fri Mar 25 '16 0:49am, David Shobe wrote
>>Thanks, Alex.  (although now I look for the "search" button and can't see it).  Anyway, you wrote, at the link provided:

>>> These problematic cases should be further inspected to find out
>>> whether MP2 problems are caused by the use of inadequate basis
>>> set (and hence, the problems can be solved using better basis
>>> set),
>>> or are inherent to MP2 (e.g., because of milti-configuration
>>> nature of true wavefunction).

>>I doubt (although I could be wrong) that simple alkynes require a multireference model.  That leaves the basis set.  I think of larger basis sets as having more convergence difficulties but there could be exceptions.  I would have liked a triple-zeta basis set, but I also would like to be able to put in third-row elements (Ge-Br) and options are limited.  

>>Anyway, I am only trying to calculate heats of hydrogenation, compared to that of acetylene.  (R-CC-R' + CH3-CH3 --> R-CH2-CH2-R' + HCCH).  Using MP4/DZV(d,p)++//MP2/DZV(d,p) might actually be overkill for an isodesmic reaction like this.  I was in the habit of using CCSD//B3LYP methods (and a different computational chemistry program) for just about everything containing only s- and p- block elements, and it usually worked in a reasonable amount of time.  Now I can't use coupled cluster theory with your program (correct?) so I have to be a bit more judicious.

>>--David Shobe
>>On Thu Mar 24 '16 10:11pm, Alex Granovsky wrote
>>>Dear David,

>>>if you search this site (using "search" button on the left) for mp2 cphf
>>>you will find the following link


>>>Hope this help.

>>>Kind regards,
>>>Alex Granovsky

[ Previous ] [ Next ] [ Index ]           Fri Mar 25 '16 11:43pm
[ Reply ] [ Edit ] [ Delete ]           This message read 526 times