Multiwfn forum / ESP of a charged molecule

Re: ESP of a charged molecule

dummy@example.com (Quark) — Mon, 16 Nov 2020 01:37:27 +0000

Dear Dr. Tian Lu,
Thanks for answering my questions.

Best regards
Eugene

Re: ESP of a charged molecule

dummy@example.com (sobereva) — Mon, 16 Nov 2020 00:52:45 +0000

You should optimize geometry and perform ESP analysis respectively for the cation and anion.

Re: ESP of a charged molecule

dummy@example.com (Quark) — Mon, 16 Nov 2020 00:38:03 +0000

Dear Dr. Tian Lu,
Thank you very much for your reply. As for my first question, I cannot understand the algorithm for constructing surfaces ESP for charged molecules. If I use the whole molecule [A- and K +], then I have a single ESP surface covering part of the K+ molecule. Please tell me how I can build two surfaces ESP separate for ions A- and K +. Should I optimize separately anion and cation or the whole molecule [A- and K +], and then make single calculations of the wave function for the obtained geometry of anions and cations? Thus, I planned to evaluate the sigma holes on the iodine atom

Best regards
Eugene

Re: ESP of a charged molecule

dummy@example.com (sobereva) — Sun, 15 Nov 2020 23:32:32 +0000

I don't fully know why you intend to "use separate cation and anion" to calculate ESP. For a charge system, in otherwords, a cation or anion, you can use exactly the same procedure to conduct quantum chemistry calculations and then use Multiwfn to perform ESP analysis. The only point that should be paid attention to is the choice of basis set, for anion you must use a basis set containing diffuse functions, otherwise the loosely bounded electrons cannot be properly represented. Using def2-TZVPD is fully reasonable for anion; decreasing to def2-TZVP is usually acceptable for optimization and frequency analysis purposes (however, some critical referees may make comment on this treatment. A safer way is using 6-31+G* instead, which also contains diffuse functions and is significantly cheaper than def2-TZVPD and def2-TZVP).

ESP of a charged molecule

dummy@example.com (Quark) — Sun, 15 Nov 2020 00:04:50 +0000

Dear Dr. Tian Lu,
I would be grateful if you answer my questions
How to build ESP for a charged molecule? Is it correct to optimize the whole molecule, and to use separate cation and anion to calculate the ESP surface? As far as I know, the basis function should contain diffuse functions, so I plan to use PBE0 dеf2-TZVPD to calculate ESP. However, for optimization, can I use a def2-TZVP to shorten the computation time?

Best regards
Eugene