3

I am told that because of the poor overlap between a chlorine atom and a carbon atom and chlorine's relatively high electronegativity, there is a strong partial positive character on the central carbon in an acyl chloride molecule.

This makes sense. I am then told that because of this high partial positive character created through the inductive effect of chlorine and oxygen and the poor overlap between chlorine and oxygen, the oxygen's lone pair(s) actually participate in a sort of resonance with the central carbon, thereby giving the C=O bond in an acyl chloride partial triple bond character.

Is this true? Also what exactly is this called? Backbonding?

Jan
  • 67,989
  • 12
  • 201
  • 386
Dissenter
  • 18,815
  • 49
  • 177
  • 341
  • 1
    http://chemistrytextbookcrawl.blogspot.com/2012/12/infrared-spectra-of-acid-chlorides.html The IR stretch is between 1775 and 1810 suggesting greater bond order. – Brinn Belyea Sep 06 '14 at 19:09

1 Answers1

5

The various resonance structures that one can draw for an acid chloride are shown below.

enter image description here

Structure IV, is the "triple bond" structure that you mentioned. The acylium ion ($\ce{R-C#O^{+}}$) is very stable, so structure IV does contribute to the overall description of an acid chloride. Note that the $\ce{C#O}$ bond in IV is not a real triple bond as the carbon remains $\ce{sp^2}$ hybridized, with the attendant 120 degree (non-linear) bond angles. I've also drawn a fifth resonance structure involving hypeconjugation. This is probably a reasonable structure as well since it places negative charge on oxygen and positive charge on the hydrogen. It also helps explain why hydrogens on the carbon alpha to a carbonyl have increased acidity.

Usually the term "backbonding" is used when electron density is transferred from one atom to another through the use of antibonding orbitals. See here for a better description. In resonance structure IV, the lone pair from oxygen is being shared with an empty carbon $\ce{sp^2}$ orbital, the one that had been used to bond to the chlorine. I've never heard that referred to as backbonding.

Community
  • 1
ron
  • 84,691
  • 13
  • 231
  • 320
  • The ion is "stable" by itself or stable when solvated or stable in the gas phase? – Dissenter Sep 06 '14 at 21:07
  • Also @ron how can one calculate the percent resonance contribution for each resonance contributor? – Dissenter Sep 06 '14 at 21:11
  • @Dissenter The acylium ion is stable enough that it can be observed spectroscopically, both in solution and the gas phase, but it is still a reactive intermediate. If you mix an acid chloride with $\ce{AlCl_3}$ in benzene, you can generate the acylium ion and it will react with the benzene via electrophilic aromatic substitution to produce a phenyl alkyl ketone. There's no simple way to estimate the contributions of the various resonance structures, other than through the use of a computer, as far as I know. – ron Sep 06 '14 at 21:27
  • Are there any programs I can use? – Dissenter Sep 06 '14 at 21:28
  • 2
    Martin or Philipp might know the answer to that. I suspect that to get meaningful estimates you'd have to use something fairly advanced. – ron Sep 06 '14 at 21:30
  • why does your triple bonded resonance structure have a ~120 degree bond angle? – Dissenter Sep 07 '14 at 01:56
  • 2
    Because nuclear positions can't change when we draw resonance structures and the carbonyl carbon was approximately sp2 hybridized. – ron Sep 07 '14 at 03:31
  • So the bond angle should probably be between 120 and 180 degrees? – Dissenter Sep 07 '14 at 03:36
  • 1
    Probably closer to 120 – ron Sep 07 '14 at 13:47
  • 1
    @Dissenter To approximate the contributions of a resonance structure you would have to follow the valence bond theory approach. I am not aware of any free programs for this. Although this molecule is quite simple, it will still require a reasonably powerful computer and a day of calculating I guess.|| Another possibility to check if a resonance structure is well enough agreeing with the overall electron density, one can use the natural bond orbital theory approach. I am also not aware that there is free software for it around. Computing is much simpler though. – Martin - マーチン Sep 08 '14 at 08:29