I'm doing some computational work and am manually inputting Z-matrices for the molecule (no GUI support for my program). For this, I need bond lengths:

  1. I tried Herzberg's book
  2. I tried a basic optimize in Avogadro and then asked it to generate a Z matrix.
  3. I checked Scifinder for structural properties

However, bond lengths vary quite a bit from these sources (upto 0.4 Angstroms). I understand that these sources are a few decades apart from each other.

Should I just stick to the newest source? Or should I stick to Herzberg because he's like the established standard?

(I'm not sure what are the best tags for this question, please feel free to suggest more/edit them)


  • 1
    $\begingroup$ Your (2) is not a source at all. Also, some bond lengths vary quite a bit in real life, so even the newest sources would not remove the discrepancy. $\endgroup$ – Ivan Neretin Aug 7 '17 at 19:38
  • 2
    $\begingroup$ I'd immediately answer single-crystal analysis data of a given compound with well-established ESDs for atom coordinates, but it's only for the solid state. $\endgroup$ – andselisk Aug 7 '17 at 19:44
  • 1
    $\begingroup$ for small molecules look at NIST databases. $\endgroup$ – porphyrin Aug 8 '17 at 6:48