Summary:Shielding tensors for H



 Dear CCL Members,
 On Monday I asked a question about NMR shielding tensors for
 H atoms. It's hard to say but I've got only 2 (two) comments and
 few qurries to summarize. It seems that there is no comparative study
  about calculating shielding tensors for H atom. Thus DO EXIST
 publications on ^{11}B, ^{13}C, ^{15}N, ^{17}O, ^{19}F, ^{31}P ....
 Some people have calculated ^{1}H shielding tensors for few molecules
 but there was no systematic study on level of theory needed or basis set
 effects. It really puzzles me ... .
 For those  who asked here is reference to work by Cheeseman et al:
 Cheeseman J.R., Trucks G.W., Keith T.A., Frisch M.J., J. Chem. Phys.
 Vol 104, No. 14 (8 April 1996) pp. 5497-5509.
 Thanks to Georg Schreckenbach and David E. Bernholdt who have shared their
 thoughts.
 Best regards,
 Grzegorz Bakalarski
 ICM|Warsaw University
 POLAND
 grzesb -AatT- asp.biogeo.uw.edu.pl
 ==================  SUMMARY  ==================================
 ----------------- Q U E R R Y ------------------------------------
  Dear Netters,
 Recently I've read article by Cheeseman el al. about calculating  absolute
 and relative NMR shielding tensors from different level of theory ( HF, MP2 and
 DFT) for different basis sets. The article is very interesting but the auhtors
 deal only with more "difficult  cases - C, N and B atoms". I am
 interested
 in  predicting NMR shielding tensors for H atoms in molecules ( it seems
 a easier case). Could anyone point me to similar article where I could find
 a systematic study of predicting NMR shielding tensors for H atom and it
 dependency on level of theory and  basis set ? Any comments are also most
 welcomed!
 Best regards,
 ------------------ R E P L I E S ----------------------------------------
 From schrecke -AatT- zinc.chem.ucalgary.ca
 Hi Grzegorz,
 I am always interested in the NMR stuff, therefore, I would be
 interested in your summary. Further, would you mind providing me with
 the reference for the article that you cite?
 I am not sure why H should be less difficult then C,N or B. In fact,
 I recall a presentation given recently by Don Chesnut where the conclusion
 was that H is MORE difficult than, e.g., Carbon. It depends, of course,
 on the meaning of "difficult". What Chesnut means is that the errors
 in calculated 1H shieldings (or shifts) are bigger than for say Carbon,
 if you consider the relative error w.r.t. the total shift range for
 the given nucleus. Chesnut has written a number of very readable reviews
 on NMR calculations, the most recent ones in "Reviews of Computational
 Chemistry" (I am not sure here) and in "Annual Reports on NMR
 Spectroscopy"
 Vol.29, G.A.Webb (Ed.), Academic Press, New York, 1994. I am not
 sure though whether your particular question is covered there.
 Yours, Georg
 ----------------
 From: bernhold -AatT- npac.syr.edu
 My experience (not yet published) is that H shieldings are harder to
 get right (absolute, not sure about relative).  I don't understand
 why this is; neither does an experimental colleague with extensive
 computational experience with shieldings.  Part of it has to do with
 the small range and shifts typical for H, but I don't think that's the
 whole story.
 --
 David E. Bernholdt                      | Email:  bernhold -AatT- npac.syr.edu
 ______________________
 D. Y. Chen and Ray Crawford asked for summary.
 ======================== T H E   E N D ========================================