Is it possible to gain the idea of full molecular structure from IR or Raman spectroscopy?

  • 2
    $\begingroup$ Practically speaking: no. While these are very useful methods to show the presence of certain chemical groups, there are many molecular features (eg groups with less characteristic signals, the difference between isomers) that not well represented in the spectrum. $\endgroup$ – Greg Nov 14 '18 at 14:35
  • 1
    $\begingroup$ For a pure "simple" molecules you can pattern match an unknown with spectra from a library of known compounds. But pragmatically it is hard to get "pure" compounds. In fact it often words the other way. You know you have a reasonably pure compound when there are no stray peaks when compared to a known pure sample of the compound. $\endgroup$ – MaxW Nov 15 '18 at 4:37
  • 1
    $\begingroup$ No, you cannot get bond lengths from IR/Raman you need to do microwave spectroscopy and this is only a productive approach for simple molecules. NMR is the only spectroscopic technique, this generally gives topology, i.e. this grouping is next to that, e.g. CH2 next to CH3 etc. Xray diffraction is the only definitive structural method . $\endgroup$ – porphyrin Nov 15 '18 at 10:45