From owner-chemistry@ccl.net Mon Oct 26 10:14:00 2020 From: "Jens Spanget-Larsen spanget]=[ruc.dk" To: CCL Subject: CCL:G: C-13 hyperfine constants Message-Id: <-54195-201026101159-21131-yaS/2tjI1lIOl7hYGCP5Qg||server.ccl.net> X-Original-From: Jens Spanget-Larsen Content-Language: da-DK Content-Type: multipart/alternative; boundary="_000_A94E15A372E6194CA8719D62642F674401951F872DMBX2adrucdk_" Date: Mon, 26 Oct 2020 14:11:50 +0000 MIME-Version: 1.0 Sent to CCL by: Jens Spanget-Larsen [spanget(!)ruc.dk] --_000_A94E15A372E6194CA8719D62642F674401951F872DMBX2adrucdk_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Dear CCL! We are presently calculating C-13 isotropic hyperfine constants for a highl= y charged semiquinone radical anion in aqueous solution. We use Gaussian16 = and apply B3LYP with Barone's basis set EPR-III, which is adapted for calcu= lation of EPR parameters with this functional. To account for the solvation= , we apply the PCM models IEFPCM or IPCM, and we experiment with discrete-c= ontinuum models to estimate the impact of specific solute-solvent interacti= ons. We are able to obtain a reasonable linear correlation with experimenta= l values, but we are puzzled by the circumstance that the computed C-13 con= stants are consistently 20-30% lower than the observed ones. Is there any r= easonable explanation for this? Is the strong solvation field possibly affe= ctig the Fermi contact analysis, or something like that? Yours, Jens >--< ________________________________________ Fra: owner-chemistry+spanget=3D=3Druc.dk/./ccl.net [owner-chemistry+spanget= =3D=3Druc.dk/./ccl.net] på vegne af Stefan Grimme grimme#,#thch.uni-bonn= .de [owner-chemistry/./ccl.net] Sendt: 23. september 2020 17:54 Til: Jens Spanget-Larsen Emne: CCL: Intramolecular H-bonding and DFT-D Sent to CCL by: "Stefan Grimme" [grimme(-)thch.uni-bonn.de] Dear Jens, I'm not aware of a benchmark study to this specific question but please allow me to give a general answer. Dispersion corrections very generally improve the PES for inherently overrepulsive (dispersion devoid) functionals giving better energies, structures and frequencies. There is no physical reason why the intramolecu= lar situation is different from the intermolecular one. However, functionals which at intermediate distances account for some "dispersion type" correlation like e.g. PBE or LDA systematically yield too strong hydrogen bonds and consequently too large vibrational red shifts (partially due to artificial CT). In such cases adding the "-D" in whatever form would further (wrongly) strengthen the bonding. For proper choices like BLYP, TPSS, B3LYP or similar I don't expect any problem. Best Stefan -=3D This is automatically added to each message by the mailing script =3D-http://www.ccl.net/cgi-bin/ccl/send_ccl_messagehttp://www.ccl.net/chemistry/sub_unsub.shtmlhttp://www.ccl.net/spammers.txt--_000_A94E15A372E6194CA8719D62642F674401951F872DMBX2adrucdk_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Dear CCL!

We are presently calculating C-13 isotropic hyperfine constants for a highl= y charged semiquinone radical anion in aqueous solution. We use Gaussian16 = and apply B3LYP with Barone's basis set EPR-III, which is adapted for calcu= lation of EPR parameters with this functional. To account for the solvation, we apply the PCM models IEFPCM o= r IPCM, and we experiment with discrete-continuum models to estimate the im= pact of specific solute-solvent interactions. We are able to obtain a reaso= nable linear correlation with experimental values, but we are puzzled by the circumstance that the computed C-13 cons= tants are consistently 20-30% lower than the observed ones. Is there any re= asonable explanation for this? Is the strong solvation field possibly affec= tig the Fermi contact analysis, or something like that?

Yours, Jens >--<

________________________________________
Fra: owner-chemistry+spanget=3D=3Druc.dk/./ccl.net [owner-chemistry+s= panget=3D=3Druc.dk/./ccl.net] på vegne af Stefan Grimme grimme#,#thc= h.uni-bonn.de [owner-chemistry/./ccl.net]
Sendt: 23. september 2020 17:54
Til: Jens Spanget-Larsen
Emne: CCL: Intramolecular H-bonding and DFT-D

Sent to CCL by: "Stefan Grimme" [grimme(-)thch.uni-bonn.de]
Dear Jens,
I'm not aware of a benchmark study to this specific question but
please allow me to give a general answer.
Dispersion corrections very generally improve the PES for inherently
overrepulsive (dispersion devoid) functionals giving better energies,
structures and frequencies. There is no physical reason why the intramolecu= lar
situation is different from the intermolecular one.
However, functionals which at intermediate distances
account for some "dispersion type" correlation like e.g. PBE or L= DA
systematically yield too strong hydrogen bonds and consequently
too large vibrational red shifts (partially due to artificial CT). In such<= br> cases adding the "-D" in whatever form would further (wrongly) st= rengthen
the bonding.
For proper choices like BLYP, TPSS, B3LYP or similar I don't
expect any problem.
Best
Stefan

-=3D This is automatically added to each message by the mailing script =3D-=
Subscribe/Unsubscribe:
Job: http://www.ccl.net/jobs

--_000_A94E15A372E6194CA8719D62642F674401951F872DMBX2adrucdk_--