Multiwfn forum

saeed_E

Member

Registered: 2019-12-21

Posts: 158

Inconsistency between sobEDAW and Gaussian results

Dear Tian,

Please suppose a Diels-Alder reaction as B...A+C-->TS in which B is F-Cl bound to diene A and C is a given dienophile. I performed distortion-interaction analysis (DIA) over the TS structure as Delta_E_act=delta_E_dist(total)+delta_E_int where, Delta_E_act, delta_E_dist(total), and delta_E_int indicate activation energy, total distortion energy (sum distortion energies of B...A and C), and interaction energy, respectively. To compute delta_E_int one can write: delta_E_int= E(TS)-E(B...A distorted)-E(C distorted). In this equation, E(TS) is the total electronic energy of the fully optimized TS while E(B...A distorted) and E(C distorted) reveal the energy change along distortion of reactants from fully separately optimized structure to the structure they adopt in fully optimized TS.

On the other hand, I performed sobEDA/W analysis on the TS structure so that one can write: delta_E_int= delta_E_xrep+
delta_E_els+delta_E_orb+delta_E_dis. While DIA as well as sobEDA/W calculations were performed at the B3LYP-GD3(BJ)/6-311+G(2d,p) level, the value of delta_E_int resulted from DIA is significantly different from that of sobEDA/W analysis. If possible, could you please let me know the origin of such a difference while it seems these two values are expected to be same?

In advance, too many thanks for your valuable time and kind attention.

Sincerely yours,
Saeed

saeed_E

Member

Registered: 2019-12-21

Posts: 158

Re: Inconsistency between sobEDAW and Gaussian results

Dear Tian,
Please, once again, accept my highest apology to bother you. It seems the origin of difference was clarified provided that you also agree and make completely confirm.

Indeed, through sobEDA/W analysis I employed (iCP=1 option) indicating the complex basis functions are used for all systems. Thus, The value of interaction energy resulted from sobEDA is quite free of BSSE correction. In other words, taking sobEDA analysis with complex basis functions, one can write: delta_E_int= (CP-delta_E_int) where CP denotes (counterpoise corrected) but within the DIA I used own basis functions for any of reactants and TS. Thus, a BSSE correction needs mandatory. In fact if a BSSE value calculated at B3LYP-GD3(BJ)/6-311+G(2d,p) is added to the value of delta_E_int resulted by DIA, the inconsistency problem would be resolved and these two approaches lead to a quite result. Is my statement a quite logical rationalization about the origin of such difference discussed above?

Sincerely,
Saeed

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,639

Website

Re: Inconsistency between sobEDAW and Gaussian results

Dear Saeed,

I cannot comment without more information on the way of calculating the delta_E_int resulted from DIA. I principel it should be exactly the same as the total interaction energy reported by sobEDA.sh. Note that sobEDA.sh provides energy of complex and each fragment, please compare the data one-by-one to the corresponding data produced by DIA calculation.

Tian

saeed_E

Member

Registered: 2019-12-21

Posts: 158

Re: Inconsistency between sobEDAW and Gaussian results

Dear Tian,

Many manay thanks, my nice friend. If you so kindly allow me, tomorrow I will send you details of a calculation here in the forum or into your email box. Which one do you prefere?

On the other hand, is not my rationalization taking BSSE correct?

Sincerely,
Saeed

Last edited by saeed_E (2023-12-29 20:35:49)

saeed_E

Member

Registered: 2019-12-21

Posts: 158

Re: Inconsistency between sobEDAW and Gaussian results

Dear Tian,
Please be aware that through a much more careful investigation, I found the source of my mistake. Now, the value of delta_E_int is quite identical using DIA as well as sobEDAW analysis. My mistake was due to a bad computation of diene distortion energy. In addition, please be aware that the "iCP=0" must be considered in sobEDAW analysis so that (because of the use of complex basis functions) a semi-BSSE correction is not applied; otherwise a difference (depending on the BSSE magnitude) exists between two values.

Once again, too many thanks for your notifications directing me to recheck computations.

Sincerely yours,
Saeed

Last edited by saeed_E (2023-12-30 13:09:37)