halomethanes



 Dear Netters:
 A few days ago I turned to the CCL with the following questions:
 Original request:
 > Dear Netters:
 >
 > I would like to calculate the optimized structures and
 > normal modes for a set of related compounds (halomethanes).
 >
 >           Cl2-C-F2 (Dichlorodifluoromethane)
 >            /      \
 >        Cl-C-F2*   Cl2-C-F2+
 >        (radical)  (cation)
 >        /      \
 >    Cl-C-F2+   C-F2
 >    (cation)   (carbene)
 >                 \
 >                 C-F
 >
 >What type of basis-set would you recommend, that
 >treats all of these species in a balanced way?
 >
 >My idea was to use polarization and diffuse functions
 >to catch the ionic or radical character of these species.
 >
 >I belive, that Hartree-Fock (HF) will do a poor job for some
 >of these molecules and one has to include electron-correlation
 >even during optimization! Do you think that it is sensible
 >to start from a HF minimum structure and to reoptimze with MP2?
 >
 >Calculating normal modes with MP2 will be quite expensive
 >in terms of CPU and disk usage! Any idea about the "precision"
 >that can be expected from such a calculation?
 >
 >I have little experience with the calculation of ionic or
 >radical species and every comment, hint, or suggestion
 >is very much appreciated! I will summarize all the responses.
 I think it is time to summarize the replies. I am very grateful
 for the valuable information I obtained from:
 Dr. Luigi Cavallo, Dr. Peter R. Schreiner, Errol Lewars, and
 Russell D. Johnson III. Their original messages will be appended
 to this posting.
 Basically, I decided to compare three levels of theory
 (HF and Post-HF) for the problem stated in the original
 posting:
 1) QCISD/6-311G(d,p)
 2) MP2/6-311++G(d,p)
 3) B3LYP/cc-pVTZ
 The results of these calculations will be compared to
 available experimental data (geometries). The effect
 of correlation and/or size of basis set should be
 clearly visible from the results.
 I tried to visit the URL given in the fourth reply, but
 could only get through to the parent page:
 http://fluid.nist.gov/
 everything else seems to be unreachable for me?!
 Cheers,
 Ferenc
 Answers to the original posting:
 ----------------------------------------------------------------------
 On Oct 9,  8:56am, CAVALLO()at()CHEMNA.DICHI.UNINA.IT wrote:
 > Subject: Re: CCL:halomethanes
 >
 >
 > dear ferenc,
 >
 > I tried to make some calculations on substituted haloalkanes, and I found
 > quite a strange behaviour indeed.
 >
 > I tried to optimize the geometry of the following molecules:
 > CF2Cl-CH3  and the corresponding anion CF2Cl-CH2(-).
 > Well, the parent molecule gave me any problem, while the anion simply
 > broke apart. The chlorine atom ran away like an anion and left the
 > CF2=CH2 alkene. I.E. I obtained the van der waals complex of
 > 1,1-difluoroethylene and chloride anion. More or less similar
 > results were obtained with similar molecules.
 >
 > I was using classical ab-initio without any correlation. The
 > basis set I used was 631G + 1 polar + 1 diffuse on all the atoms
 > (H included). Unhappy of this behaviour, I also tried the simple
 > mopac and I also tried DFT methods with NL corrections included.
 > (This last approach should include good correlation effects) but
 > the results were always the same.
 >
 > The only reference I could find on the subject is:
 > Bernardi et al. Gazz. Chim. Ital. 1990, 120, 301.
 >
 > I know that you were asking for cations and radicals instead
 > of anions, but I think that my experience could be of some
 > help in warning you with respect to strange behaviours in
 > haloalkanes.
 >
 > yours
 > luigi
 >
 > ------------------------------
 > Dr. Luigi Cavallo
 > Dept. Of Chemistry
 > Univ. Of Naples
 > ITALY
 > Email cavallo()at()chemna.dichi.unina.it
 > ------------------------------
 >-- End of excerpt from CAVALLO()at()CHEMNA.DICHI.UNINA.IT
 On Oct 9,  9:12am, Peter R. Schreiner wrote:
 > Subject: halomethanes
 >
 > > I would like to calculate the optimized structures and
 > > normal modes for a set of related compounds (halomethanes).
 > >
 > >             Cl2-C-F2 (Dichlorodifluoromethane)
 > >             /      \
 > >         Cl-C-F2*   Cl2-C-F2+
 > >         (radical)  (cation)
 > >         /      \
 > >     Cl-C-F2+   C-F2
 > >     (cation)   (carbene)
 > >                  \
 > >                  C-F
 > >
 > > What type of basis-set would you recommend, that
 > > treats all of these species in a balanced way?
 > >
 > > My idea was to use polarization and diffuse functions
 > > to catch the ionic or radical character of these species.
 > >
 > > I belive, that Hartree-Fock (HF) will do a poor job for some
 > > of these molecules and one has to include electron-correlation
 > > even during optimization! Do you think that it is sensible
 > > to start from a HF minimum structure and to reoptimze with MP2?
 > >
 > > Calculating normal modes with MP2 will be quite expensive
 > > in terms of CPU and disk usage! Any idea about the
 "precision"
 > > that can be expected from such a calculation?
 >
 >
 > Dear Ferenc,
 >
 > The best compromise is probably DFT, preferrably B3LYP with a flexible
 > basis set, say 6-311+G* for geometry optimizations. For energies,
 > you will need a better basis for very accurate results. Others and
 > my experience would favor correlation-consistent basis sets of
 > triple zeta quality (e.g., cc-pVTZ) or higher, if you can afford it.
 > Ideally, you'd used CCSD(T) for the single points, but this may be too
 > expensive. Thus, the B3LYP/cc-pVTZ//B3LYP/6-311+G* + ZPVE should
 > be an acceptable compromise.
 >
 > For a mini-review on the performance of various methods with carbenes
 > see: J. Org. Chem. 1996, 61, 7030.
 >
 > Hope this helps,
 > Peter
 >
 >
 > --
 >   Dr. Peter R. Schreiner           Computer-Chemie-Centrum
 >   prs()at()organik.uni-erlangen.de      Universitaet Erlangen-Nuernberg
 >   Phone:    +49-(0)9131-856533     Naegelsbachstr. 25
 >   FAX:      +49-(0)9131-856566     D-91052 Erlangen, Germany
 >
 > NOTE: NEW ADDRESS from November 1, 1996:
 >
 >   Institut fuer Organische Chemie
 >   Universitaet Goettingen
 >   Tammannstr. 2			   Phone: +49-(0)551-393291
 >   D-37077 Goettingen, Germany      FAX:   +49-(0)551-399475
 >
 >   http://www.ccc.uni-erlangen.de/ecc/private
 > --
 >   Dr. Peter R. Schreiner           Computer-Chemie-Centrum
 >   prs()at()organik.uni-erlangen.de      Universitaet Erlangen-Nuernberg
 >   Phone:    +49-(0)9131-856533     Naegelsbachstr. 25
 >   FAX:      +49-(0)9131-856566     D-91052 Erlangen, Germany
 >
 > NOTE: NEW ADDRESS from November 1, 1996:
 >
 >   Institut fuer Organische Chemie
 >   Universitaet Goettingen
 >   Tammannstr. 2			   Phone: +49-(0)551-393291
 >   D-37077 Goettingen, Germany      FAX:   +49-(0)551-399475
 >
 >   http://www.ccc.uni-erlangen.de/ecc/private
 >-- End of excerpt from Peter R. Schreiner
 On Oct 9, 10:21am, E. Lewars wrote:
 > Subject: HALOMETHANES
 > Hello,  some useful refs are:
 > 1.   Theoretical Enthalpies of Formation of CHmCln
 > 	: Neutral Molecules and Cations
 > 	C. F. Rodrigues, D. K. Bohme, A. C. Hopkinson
 > 	J Phys Chem  1996, _100_ 2942-2949
 > 2.   Photochemical Isomerization of Dichloromethane in Argon Matrices
 > 	J Am Chem Soc 1990, _112_ 5117-5122.
 > Other papers by Hopkinson, in JPC and JACS, in the last 10 years, may also
 > be helpful.
 > Good Luck
 > ====
 > Errol Lewars
 > ===
 >-- End of excerpt from E. Lewars
 On Oct 9, 12:09pm, Russ Johnson wrote:
 > Subject: halomethane structures
 > Dear Ferenc,
 >         First , there is some information on structures of fluoro methanes
 > and radicals at:
 >      http://fluid.nist.gov/~drb/cgi/SpeciesData.html
 >
 > I would recommend a density functional method, as I've found it usually
 > gives very good geometries and frequencies for radicals and cations. Large
 > basis sets are usually not necessary so that B3LYP/6-31G* should
 > do quite well.
 >
 > You may have trouble with Cl2-C-F2+ , as both CF4+ and CCl4+ have non-
 > classical structures, they are more loosely bound complexes such as
 > CF3+ --- F.
 >
 > Good luck,
 > Russ
 >
 >
 >
 >
 > Russell D. Johnson III
 > Research Chemist
 > Physical and Chemical Properties Division
 > National Institute of Standards and Technology
 > Gaithersburg, MD 20899
 > voice 301+975-2513     fax  301+975-3670
 > email: russell.johnson()at()nist.gov
 >-- End of excerpt from Russ Johnson
 Ferenc Molnar
 ---------------------------------------------------------------------------
 Institut fuer Physikalische und Theoretische Chemie
 - Lehrstuhl Prof. Dick -                   Tel.:  (+49) 941 943-4466 /-4486
 Universitaet Regensburg                    Fax.:  (+49) 941 943-4488
 Universitaetsstrasse 31
 D-93053 Regensburg
 Deutschland / Germany
 ---------------------------------------------------------------------------
 EMail (SMTP):  Ferenc.Molnar()at()chemie.uni-regensburg.de
 ---------------------------------------------------------------------------
  Those who do not archive the past are condemned to retype it!
   -- Garfinkel & Spafford, Practical Unix Security
 ---------------------------------------------------------------------------