Experimental and computed bond lengths: The importance of their differences

Abstract State‐of‐the‐art experimental electron diffraction and computational information on the structure of alkaline earth dihalide molecules are in agreement for the shape of these symmetric triatomic molecules (linear/bent/quasi‐linear). However, the computed and measured bond lengths show differences that are not only considerably larger than the experimental error but also have the wrong sign. The physical meaning of experimental bond lengths depends on the physical techniques used in their determination and the ways of averaging over molecular vibrations. The choice of model potentials in the elucidation of experimental information is also important, especially for floppy molecules. When improved computational bond lengths become available, their comparison with experimental information will have to take account of the physical meaning of the experimentally determined bond lengths. The computed equilibrium distance ( r e ) should be smaller than the experimental distance‐average bond length ( r g ). The differences may range from a few thousandths of an Å to a few hundredths with increasing temperature and, especially, with increasing floppiness of the molecule. For truly accurate comparison, experimental bond lengths should be compared with computed ones only following necessary corrections, bringing all information involved in the comparison to a common denominator. © 1992 John Wiley & Sons, Inc.