I have been taught that ring expansion stabilises smaller cyclic compounds to a great extent. So why does cyclopropyl methyl carbocation shows such type of resonance

enter image description here

rather than expanding its ring

enter image description here

enter image description here

thereby decreasing it's angle strain, further it will be stabilized by hyperconjugation effect and inductive effect.

  • $\begingroup$ perhaps the angles in the 4-ring are bad for hyperconjugation? the intermediate 4-ring is even more badly strained than the 3-ring? the resonance takes a lot of strain from the cyclic structure? (last point is definitively true) $\endgroup$
    – Karl
    Sep 25 '18 at 7:28
  • $\begingroup$ It can't be said that the cyclobutane carbocation does not occur. March's Advanced organic chemistry, 10.C.i Neighboring-group participation by π and σ bonds: nonclassical carbocations, tries to explain the stabilization, but also cites that (chloromethyl)cyclopropane by hydrolysis provides 48% cyclopropylmethanol, 47% cyclobutanol, and 5% "allylmethanol". $\endgroup$
    – mykhal
    Sep 25 '18 at 9:40
  • 1
    $\begingroup$ I have found a similar question already answered: chemistry.stackexchange.com/questions/19090/… $\endgroup$ Sep 25 '18 at 9:47
  • $\begingroup$ I already checked this answer but this doesn't explain why ring expansion is not preferred $\endgroup$
    – user66707
    Sep 25 '18 at 9:59
  • 1
    $\begingroup$ Question itself does - it's exceptionally stable (while cyclobutyl isn't) $\endgroup$
    – Mithoron
    Sep 25 '18 at 18:45

Unfortunately, the premise of your question is wrong.

There is no cyclopropylcarbinyl or cyclobutyl cation. They're the same thing. Unlike cations which you might be more familiar with, for example, t-butyl cation, the cation you're looking at here, also called homoallyl, is a non-classical cation.

The open form and the two structures you drew in the second line are not distinct. They're just resonance forms of each other.

You ask does this ion undergo ring expansion? Well, yes and no. Yes, in that there is a contribution from the 4-member ring to the resonance structure. No, in that you can't expand something that doesn't exist.