RE: pyramidalization of amides

My opinion: the calculations will be valid, if interpreted properly.
 I tend to think of such systems as "obeying" a modified, three-level
 Born-Oppenheimer approximation.  In the standard two-level Born-Oppenheimer
 approximation, the nuclei are frozen, with the electrons treated in quantum
 mechanical fashion, giving the familiar potential energy surface.  However,
 once we have that potential energy surface, we can use it to give the nuclei
 a quantum treatment: that is what we do when calculating zero-point and
 thermal energies.
 However, hydrogen compounds can be better treated with a three way
 distinction between electrons, hydrogen nuclei (especially if they are
 protons, i.e. 1H), and then the heavier nuclei.  One calculates the
 potential energy surface in the usual fashion (freezing all nuclei and
 allowing quantum mechanical treatment of electrons).  However, once that
 potential energy surface is calculated, the hydrogen atoms would be treated
 quantum mechanically while other nuclei are still frozen.  A new heavy atom
 potential energy surface is thus generated, and wavefunctions and
 probability distributions are calculated for the hydrogen nuclei.  There was
 an article in Accounts of Chemical Research a few years ago in which this
 was done for water clusters.
 Back to your question: a planar amide is still distinguished from a
 pyramdialized amide in having a one-minimum or two-minima potential energy
 surface for the amide hydrogen.  These produce one-maximum or two-maxima
 probability distributions (respectively) for the hydrogen nuclei.
 Experimental geometries typically come from diffraction studies, which
 actually measure these probability distributions since the results are
 averaged among the molecules in the sample.  Thus, the experimental
 geometries should correspond to the minima found by B3LYP/6-31G* and the
 --David Shobe
 Süd-Chemie Inc.
 phone (502) 634-7409
 fax     (502) 634-7724
 email  dshobe { *at * }
 Don't bother flaming me: I'm behind a firewall.
 -----Original Message-----
 From: Ulrike Salzner [mailto:salzner { *at * }]
 Sent: Thursday, August 01, 2002 5:59 AM
 To: chemistry { *at * }
 Subject: CCL:pyramidalization of amides
 I am reviewing a thesis which deals with pyramidalization in amides, S,
 Se anlogues and amids in larger ring systems. Substituent effects are
 analyzed and compared with AIM and NBO. All of these systems have
 inversion barriers smaller than ammonia, some of them are planar. Thus,
 like ammonia, these molecules probably violate the Born-Oppenheimer
 approximation. How reliable are standard calculations (B3LYP/6-31G*,
 MP2/6-31G*) of bond angles, i.e. degree of pyramidalization? Should
 trends be ok? I would appreciate oppinions on this.
 Ulrike Salzner
 -= This is automatically added to each message by mailing script =-
 CHEMISTRY { *at * } -- To Everybody  | CHEMISTRY-REQUEST { *at * } -- To Admins
 Ftp:  |  WWW:   | Jan: jkl { *at * }