Author Topic: DCOSMO_RS  (Read 4746 times)

martijn

  • Full Member
  • ***
  • Posts: 63
  • Karma: +0/-0
DCOSMO_RS
« on: February 18, 2022, 05:39:18 PM »
Hi,

Am I reading the manual correctly and does one use the default epsilon (epsilon infinity, metallic boundary conditions) for every solvent when running DCOSMO_RS calculations rather than selecting the specific dielectric perimittivity of the specific solvent?

Thanks in advance,

Martijn

diedenho

  • Developers
  • Newbie
  • *
  • Posts: 8
  • Karma: +0/-0
Re: DCOSMO_RS
« Reply #1 on: February 25, 2022, 10:55:38 AM »
Dear Martijn,

the idea is to use the epsilon of the solvent. The pre-defined  potential files (e.g. benzene_25.pot) contain the dielectric constant of the solvent.
It s used in eq. 6 of https://pubs.acs.org/doi/10.1021/jp511158y because the parabolic part of the potential is described well by the COSMO model.

Hope that helps,

Michael

martijn

  • Full Member
  • ***
  • Posts: 63
  • Karma: +0/-0
Re: DCOSMO_RS
« Reply #2 on: February 26, 2022, 01:47:21 AM »
Thanks Michael. That makes sense and is the dielectric permittivity value is in the pot file I don't have to separately define it.

Best,

Martijn

martijn

  • Full Member
  • ***
  • Posts: 63
  • Karma: +0/-0
Re: DCOSMO_RS
« Reply #3 on: August 25, 2023, 03:30:01 PM »
Hi,

I wanted to revist this question as the Turbomole output when using DCOSMO-RS is still not 100% clear to me. When I run a calculation with dcosmo-rs for toluene in Turbomole 7.5 using:

$cosmo
$dcosmo_rs file=toluene_25.pot


in the control file, the output is:

==============================================================================
 DCOSMO-RS RESULTS
 ==============================================================================
  PARAMETER:
    nppa:      1082
    nspa:        92
    nsph:        32
    nps:       3605
    npspher:   1224
    disex:   10.0000   
    disex2:  4656.71   
    rsolv [A]: 1.3000
    routf:     0.8500
    phsran:    0.0   
    ampran:   0.10E-04
    cavity:  closed
    epsilon: infinity
    refind:     1.300
    fepsi:      1.0000000
    DCOSMO-RS pot. file: toluene_25.pot
    fcrs:      0.4827586
    c0 [kcal A**2/mol/e**2]:   1447.2400000
    polscal:      1.0000000
    number of splines per pot.:  86
    data intervals per spline:   3
    sum_x_vol [A**3]:    131.8148000
    sum_x_area [A**2]:    140.5495000
    lambda_3:     -0.1506584
    T[K]:   298.10


And even though the toluene_25.pot has an entry "epsilon=2.4" the turbomole output suggests a value of epsilon infinity was used.

Does this mean (i) that Turbomole reads the correct value from the pot file just for some reason doesn't print it correctly in the output file, or (ii) even the epsilon value is in the pot file turbomole doesn't read it and the user still manually has to add it to the control file, as you would do for a normal cosmo calculation, and else in the absence of an explicit epsilon value in the control file a value of infinity is used, as again, what would happen for a normal cosmo calculation?

Thanks,

Martijn

martijn

  • Full Member
  • ***
  • Posts: 63
  • Karma: +0/-0
Re: DCOSMO_RS
« Reply #4 on: August 29, 2023, 11:02:04 PM »
Hi,

In a follow up to my last question I did some trial calculations in which I copied the toluene_25.pot file and changed the relative dielectric constant in it to different values, going from 1 to 80.1, and run single point calculations with these hacked pot files. The good news is that the energies from these clearly unphysical calculations varies with the modified dielectric constant in the hacked pot files and that this variation is linear with the f(epsilon) factor as defined in the dcosmo-rs paper. Based on this I assume that everything works fine and that the epsilon printed in the output (always infinity) is wrong and set by a bit of code that is not used in the dcosmo-rs calculation. However, it would be great if someone with knowledge of the actual code can confirm this.

Thanks,

Martijn