Rotation Entropy for non-Abelian point group sym and CC-methods.
Gets Around
8:24:18 AM PDT - Mon, Mar 16th 2015
Hi.

Rotation entropy of molecule depends on its pont group and symmetry number thereof, as described http://cccbdb.nist.gov/thermo.asp (eq 19), but when using CC-methods only abelian point groups is allowed.

In the case of highly symmetric compounds such as cubane or dodecahedrane incorrect values of rotation entropy are obtained.

Gets Around
8:15:18 PM PDT - Wed, Mar 18th 2015
reusing existing hessian with proper symmetry group solve the problem.

Gets Around
11:34:18 PM PDT - Wed, Mar 18th 2015
Hi Vladimir,

I am glad that you were able to solve this problem. Could you post an example input file demonstrating your solution? I will want to perform similar calculations in the future.

Thanks,
Matt

Gets Around
9:00:38 AM PDT - Thu, Mar 19th 2015
Quote:Mernst Mar 19th 1:34 pm
Hi Vladimir,

I am glad that you were able to solve this problem. Could you post an example input file demonstrating your solution? I will want to perform similar calculations in the future.

Thanks,
Matt

Hi Matt, here is an example.
As described in http://nwchemgit.github.io/index.php/Vibration I reuse the previously computed hessian by specifying "reuse" in the "freq" module input block.
title "N8-cubane CCSD/cc-pVDZ hessian"

memory stack 100 mb heap 100 mb global 100 mb

geometry
 symmetry Oh
 N                     0.76104280     0.76104280     0.76104278
 N                    -0.76104280    -0.76104280     0.76104278
 N                    -0.76104280     0.76104280    -0.76104278
 N                     0.76104280    -0.76104280    -0.76104278
 N                    -0.76104280    -0.76104280    -0.76104278
 N                     0.76104280     0.76104280    -0.76104278
 N                     0.76104280    -0.76104280     0.76104278
 N                    -0.76104280     0.76104280     0.76104278
end

freq
 reuse N8.hess
end

task tce freq


It is important that the order of the atoms during the calculation of hessian and its subsequent reusing were exactly the same. Check it out in the output file.

Gets Around
4:36:10 PM PDT - Tue, Jul 14th 2015
Do not invoke "task tce freq" with no tce input block. The defaults for tce may be far from optimal. Look at the examples and user manual for good settings. For example:
tce

  freeze atomic

  ccsd

  io ga

  2eorb

  2emet 13

  tilesize 30

  attilesize 30

end

Gets Around
9:28:39 PM PDT - Wed, Jul 15th 2015
Hi, Jeff.

Reusing hessian means that no actual TCE calculations were performed in this case, only ZPVE and total entropy were recalculated.
Normally "Total CPU times" of such task is less than 1s.

Vladimir.


Forum >> NWChem's corner >> NWChem functionality