Summary:DFT for hydrogen bond



I sent a message about hydrogen bond calculation by DFT to CCL several
 days ago. I've got many helpful replies. Now I summarize them below.
 Hi, everyone,
 I know most of DFT methods are not good to the calculation of Hydrogen
 Bonding. But a friend of mine told me somebody developed a DFT method
 that can be used in HB calculations. What is this method?
 Are there some papers using different methods (DFT methods preferred) to
 calculate HB and systematically compare them? Any advice is helpful.
 Thanks in Advance
 Yubo Fan
 --
 ============================================================
 Dr. Yubo Fan               Email: yubofan.,at,.mail.chem.tamu.edu
 Department of Chemistry    Tel:   1-979-845-7222
 Texas A&M University
 College Station, TX 77843
 ============================================================
 Hello. I had a couple of such publications a few years back. They are as
 follows:
 A. T. Pudzianowski, "A Systematic Appraisal of Density Functional
 Methodologies
 for Hydrogen Bonding in Binary Ionic Complexes,"  J. Phys. Chem.
 100(12),
 4781-4789 (1996)
 A. T. Pudzianowski, "Current Computational Approaches to The Strong
 Hydrogen
 Bond," Rec. Res. Devel. in Physical Chem. 1, 81-97 (1997).
 Both of these deal with the strong H-bond, but there are plenty of
 references to
 other situations. The latter reference is a review article, so there is
 lots of
 related information. It appears in a volume published in India, so it
 may be
 hard to find. Let me know if you'd like reprints and I'll send them
 along.
 Regards,
 Andrew Pudzianowski
 ----------------------------------------
 Andrew T. Pudzianowski, Ph.D.
 Computer-Assisted Drug Design
 Bristol-Myers Squibb PRI
 Box 4000
 Princeton NJ 08543-4000
 (609) 252-4248 (office)
 (609) 252-6030 (fax)
 -----------------------------------------
 hi Yubo,
 check B(M)LYP functional which was developed for such calculations
 (references in the paper enclosed). It's essentially B3LYP whith HF
 exchange
 increased to 35%. I've tested it on OH- hydration and it performed quite
 well.
 Cheers,
 Piotr
  --
  Prof. Piotr Paneth
  temporary address through 5/3/01:
  Department of Chemistry, Smith Hall  box I11, room 224
  University of Minnesota
  Minneapolis, MN 55455-0431
  207 Pleasant St. SE
  phone: (612) 626-0259, fax: (612) 624-9320 or (612) 626-9390
  E-mail: paneth.,at,.comp.chem.umn.edu or panet001.,at,.umn.edu
 Hi
 Re: your question to CCL concerning h-bonding and dft.
 We have been working on a number of H-bonded clusters involving
 water and have found that the (generally popular) B3LYP method
 and others give too high "binding".
 It is suggested that the HTCH functional could be good - this is
 probably the one you are thinking of.  A ref you might find useful is:
 Tuma, Christian; Daniel Boese, A.; Handy, Nicholas C..
 Predicting the binding energies of H-bonded complexes: A
 comparative DFT study.    Phys. Chem. Chem. Phys.  (1999),
 1(17),  3939-3947.  CODEN: PPCPFQ  ISSN:1463-9076.  CAN
 131:342144  AN 1999:538455    CAPLUS
 and other papers by Handy describing the development of new
 exchange-correlation functionals.
 I would be grateful if you summarize or email me any interesting
 information you get.
 Cheers,
 Tim
 -------------------------------
 Tim Robinson
 Vikings Group
 Department of Chemistry
 University of Otago
 tel 64 3 479 7929
 fax 64 3 479 7906
 timr.,at,.alkali.otago.ac.nz
         Yubo,
         actually some of the hybrid DFT does pretty well
 for H-bonds. Something like B3LYP/6-31+G* gets the
 water dimer interaction right to about 0.5 kcal/mol.
         Frank
 --------------------------------------------------
 | Frank Jensen, Department of Chemistry          |
 | SDU Odense University, DK-5230 Odense, Denmark |
 | FAX +45 66 15 87 80 , Voice +45 65 50 25 07    |
 | http://www.sdu.dk/Nat/Chem/STAFF/sci/FrjE.html |
 | http://bogense.chem.ou.dk/~frj                 |
 --------------------------------------------------
 Try the functionals by Nic Handy (there is one in Gaussian) and also the
 ones
 by Baerends.
 see www.gaussian.com, DFT methods for details or better consult the
 outstanding
 book by
 Koch, Holthausen, VCH 2000
 bye
 Hi Yubo,
 take a look at the following papers:
 - Kumar et al., Low barrier hydrogen bonds: Ab initio and DFT
 Investigations, J. Comp. Chem. 1998, 19, 1345-1352.
 - C. Tuma et al., Predicting the binding energies of H-Bonded complexes:
 A comparative DFT study, Phys. Chem. Chem. Phys.
 (PCCP) 1999, 1, 3939- 3947.
 -Novoa & Soso, Evaluation of the Density Functional Approximation on the
 Computation of H-Bond Interactions, J. Phys. Chem.
 1995, 99, 15837-15845.
 - Molecular Dynamics and DFT Studies of Intermolecular Hydrogen Bonds
 between Bifunctional Heteroazaaromatic Molecules
 and Hydroxylic Solvents, J.P.C. A; 2000; 104(42); 9542-9555.
 - Comparative Study of BSSE Correction Methods at DFT and MP2 Levels of
 Theory (of Hydrogen-bonded clusters),
 J.Comp.Chem. 1998, 69, 575-584 (and references therein).
 - Density Functional Theory and Molecular Clusters, J. Comp. Chem. 1995,
 16, 1315-1325.
 - A Systematic Appraisal of Density Functional Methodologies for
 Hydrogen Bonding in Binary Ionic Complexes," J. Phys.
 Chem. 100, 4781-4789 (1996).
 Hopefully you will find something useful.
 All the best
 Marc Walter
 Dear Yubo,
 I'd say it depends a bit on the system.
 If the nature of the hydrogen bond is mainly electrostatic,
 then gradient-corrected and hybrid functional do a reasonably good job.
 If the nature of the hydrogen bond is mainly
 dispersive, then DFT won't work.
 Best regards
 Klaus
 --
 Dr. Klaus Stark
 Application Scientist
 Molecular Simulations Inc.
 Inselkammerstr. 1
 82 008 Unterhaching
 Germany
 Phone : ++49-89-61459-420
 Mobile : ++49-172-936-3380
 Fax : ++49-89-61459-400
 E-Mail : kstark.,at,.msi.de
 Web Page : http://www.msi.com
 Dear Yubo,
         actually, DFT methods are not really bad for hydrogen bonding,
 they are just really bad in treating dispersion. They overestimate the
 electrostatic attraction in order to compensate for the lack of proper
 treatment of electron correlation. For H-bonded systems, using the
 B3LYP density functional and the 6-31G(d,p) basis set, you can obtain
 with quite reasonable geometries and interaction energies.
 Usually H-bonds are about 0.03A shorter than the they would be when
 you optimise them with MP2 method and interaction energies are about
 5% larger.
 Density functional calculations on H-bonded systems have been
 carried out in:
 M. Meyer, T. Steinke, M. Brandl, J. Sühnel J. Comp. Chem. 2001, 22,
 109-124
 Density Functional Study of Guanine and Uracil Quartets and of
 Guanine/Metal Ion Complexes
 Best regards,
 Maria
 Dear Yubo Fan:
 The following paper might be of interest to you:
 J. Spanget-Larsen: "Infrared absorption and Raman scattering of
 (Z)-3-hydroxypropenal. A density functional theoretical study",
 Chemical Physics 240 (1999) 51-61.
 A number of standard DFT procedures (BLYP, BPW91, B3LYP, B3PW91) are
 applied in
 a study of molecular and vibrational structure of 3-hydroxypropenal
 (malonaldehyd enol), probably the simplest species with an
 intramolecular O..H-O
 hydrogen-bond.
 Jens >--<
 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 JENS SPANGET-LARSEN         Phone: +45 4674 2000 (RUC)
 Department of Chemistry            +45 4674 2710 (direct)
 Roskilde University (RUC)   Fax:   +45 4674 3011
 P.O.Box 260                 E-Mail: JSL.,at,.virgil.ruc.dk
 DK-4000 Roskilde, Denmark   http://virgil.ruc.dk/~jsl/
 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=