Problem with numerical frequency calculation
Clicked A Few Times
2:31:25 PM PST - Wed, Mar 2nd 2016
Hi NWChem users,

I am trying to calculate the vibration frequency of a configuration where adatom moving near the optimized position on CNT surface. I first optimized the whole stucture (221 atoms), and then restart the job with:

restart cnt_5_10-p1vib

set gen_hess:actlist 221 # 221st atom is the adatom
set hessian:compress .true.
vib
reuse

temp 1 300

animate
end
task dft freq numerical

I have two questions:

1. Even thought I put those two set directives in the restart job (as suggested in http://nwchemgit.github.io/Special_AWCforum/st/id518/printing_hessian_with_greater...), I am still getting all atom frequencies with a 663x663 matrix. Is there any trick here to freeze the atoms in CNT and actually have only adatom moving?

2. I also tried to restart again but this time I changed the temperature to 2000 K. However, the output frequencies of normal modes are exactly the same with those at 300 K. The only difference in the output file is in this block:

Temperature                      =   300.00K

frequency scaling parameter      =   1.0000



Zero-Point correction to Energy  = 1852.825 kcal/mol  (  2.952665 au)

Thermal correction to Energy     = 1856.996 kcal/mol  (  2.959312 au)

Thermal correction to Enthalpy   = 1857.592 kcal/mol  (  2.960261 au)



Total Entropy                    =   58.653 cal/mol-K

  - Translational                =   49.102 cal/mol-K (mol. weight =2322.2348)

  - Rotational                   =    0.000 cal/mol-K (symmetry #  =        1)

  - Vibrational                  =    9.551 cal/mol-K



Cv (constant volume heat capacity) =   54.633 cal/mol-K

  - Translational                  =    2.979 cal/mol-K

  - Rotational                     =    1.986 cal/mol-K

  - Vibrational                    =   49.668 cal/mol-K



Temperature                      =  2000.00K

frequency scaling parameter      =   1.0000



Zero-Point correction to Energy  = 1852.825 kcal/mol  (  2.952665 au)

Thermal correction to Energy     = 3162.591 kcal/mol  (  5.039911 au)

Thermal correction to Enthalpy   = 3166.564 kcal/mol  (  5.046242 au)



Total Entropy                    = 1191.330 cal/mol-K

  - Translational                =   58.522 cal/mol-K (mol. weight =2322.2348)

  - Rotational                   =    0.000 cal/mol-K (symmetry #  =        1)

  - Vibrational                  = 1132.808 cal/mol-K



Cv (constant volume heat capacity) = 1095.424 cal/mol-K

  - Translational                  =    2.979 cal/mol-K

  - Rotational                     =    1.986 cal/mol-K

  - Vibrational                    = 1090.458 cal/mol-K

Why are those frequencies the same at different temperature? Is there anything wrong in my procedure?

Thank you so much!
Longtao

Forum Regular
5:40:47 PM PST - Wed, Mar 2nd 2016
Please, try the following input:


restart cnt_5_10-p1vib



set gen_hess:actlist 221 # 221st atom is the adatom

set hessian:compress .true.

task dft hessian numerical



geometry

adatom  adatom_x adatom_y adatom_z

end



vib

reuse

temp 1 300

animate

end



task dft freq


You need to first calculate the Hessian only for the active atom(s). Then if you want to do the frequency analysis for just that, you need to define a new geometry that contains just the active atom(s), otherwise you will get an end-of-file error.

In regards to your second question, the frequencies are calculated in a harmonic approximation and there is no temperature dependence so the only change in output when the temperature is changed should be the thermochemistry analysis.

Clicked A Few Times
2:56:19 PM PDT - Wed, Mar 16th 2016
Thank you for your reply, Sean.

I just got chance to try today. I found if I define a new geometry with just active atom, the normal mode frequencies will all be zero. From the xyz file it's clear that the only atom in the geometry is not moving at all.
The script in my first post can do the job, but it will generate thousands of xyz files. That's why it made me think if I did something wrong.

I understand now to the second question, thank you!


Forum >> NWChem's corner >> Running NWChem