Multiwfn forum

rikaaardoss

Member

Registered: 2020-07-29

Posts: 20

regarding Energy Decomposition Analysis

Dear Forum, I trying to make an Energy Decomposition Analysis (EDA) at the M06-2X/Def2TZVP level and I have the following question regarding dispersion component (E_disp)

we have:

(1) E_int = E_elect + E_Pauli + E_orb + E_disp

and because m06-2x contains dispersion corrections

(2) E_steric = E_elect + E_Pauli + E_disp

so

(3) E_int = E_steric + E_orb

Is Eq. (2) correct? Can I add the dispersion term in the steric component when using a functional that has dispersion corrections?

Thanks in advance for the help

R.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,773

Website

Re: regarding Energy Decomposition Analysis

Please note that M06-2X doesn't have a dispersion correction term (like B3LYP-D3), however, it is still able to reasonably represent dispersion effect because of its XC functional form and training set. In other words, dispersion effect is included in its XC energy. If you hope to isolate the dispersion contribution to interfragment interaction, M06-2X should not be used.

rikaaardoss

Member

Registered: 2020-07-29

Posts: 20

Re: regarding Energy Decomposition Analysis

Dear Tian Lu, thank you for your prompt reply and thank you very much for clarifying the M06-2X functional. I am not looking to separate the dispersion term, on the contrary, if a functional is able to describe by itself (i.e. without using Grimme's corrections) the dispersion effects, is it possible to assume that the term E_disp is inside E_steric, as in equation 2 of my first post?, i.e. it would be correct to assume that equation? (for functionals that can describe reasonably well the dispersion effects).

Best regards

R.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,773

Website

Re: regarding Energy Decomposition Analysis

For M06-2X, since its exchange-correlation energy already sufficiently accounts for dispersion effect, (1) should be E_int = E_elect + E_Pauli + E_orb + E_xc, where E_xc denotes change of exchange-correlation energy during binding. (2) should be E_steric = E_elect + E_Pauli + E_xc. Your (3) is correct in this context.

rikaaardoss

Member

Registered: 2020-07-29

Posts: 20

Re: regarding Energy Decomposition Analysis

thanks for your help!

rikaaardoss

Member

Registered: 2020-07-29

Posts: 20

Re: regarding Energy Decomposition Analysis

Dear Tian Lu, its properly use the first iteration of the CCSD(T) calculation in order to perform at simple-EDA at the CCSD(T) level?

thanks for your help!

r.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,773

Website

Re: regarding Energy Decomposition Analysis

Dear Tian Lu, its properly use the first iteration of the CCSD(T) calculation in order to perform at simple-EDA at the CCSD(T) level?

thanks for your help!

r.

This is in principle inappropriate.

rikaaardoss

Member

Registered: 2020-07-29

Posts: 20

Re: regarding Energy Decomposition Analysis

Thank you for your replay, maybe is more convenient the use of the approximation of the Fock matrix implemented in multiwfn?

r.

Last edited by rikaaardoss (2022-07-02 02:53:26)

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,773

Website

Re: regarding Energy Decomposition Analysis

I don't understand you question. The simple-EDA method is only applicable to pure SCF method, such as HF and DFT (except for double-hybrid functionals)