Different structural topologies affect how π bonds can interact, and thus the energies and stabilities of each molecule. Chemists have spent lots of synthetic effort in identifying the limits of Hückel theory. One area is in making molecules that might be "antiaromatic:" it is predicted that cyclic systems with 4n electrons will be less stable than the localized polyene, and should show alternation of short and long bond lengths. Two molecules that demonstrate this are cyclobutadiene and 1,3,5,7-cyclooctatetraene:

Cyclobutadiene


Cyclobutadiene has a storied history, since it is very unstable (even if aromatic!). Calculations vary in terms of predicting whether the structure should be square or rectangular, but when it was generated in an argon matrix at ultralow temperatures, the IR spectrum fit the predictions for the rectangular species. Finally, in 1991, the molecule was generated as a guest in a host-guest complex, and its NMR spectrum was measured.

Cyclooctatetraene


Cyclooctatetraene is similar though as we saw it was easier to make and isolate. It avoids a planar structure by adopting a "tub" conformation, and again alternates bond lengths. Interestingly, it reacts readily with sodium or potassium to form a dianion; that 10-electron system is planar.
Black background
White background
 Spacefilling model
Wireframe
Ball & Stick

Homoaromaticity

So what happens if we interrupt the cycle of pi bonds? Given the thermodynamics of delocalization, we ought to expect there to be some driving force for two termini of a (4n + 2) π system to interact with each other. This has been observed. The C7 "tropylium" cation is known to be flat, stable (water-soluble, in fact), and has all 7 C-C bonds identical. We can make the "homotropylium" cation by protonating cyclooctatetraene:
Comparison between the tropylium cation (C7H7+: planar, with all C-C bonds 1.47 angstrom) and protonated cyclooctatetraene (homotropylium) and a structurally characterized version made from protonating a bicyclic ketone


The homotropylium ion has not been characterized by X-ray crystallography, but NMR behavior shows there are two different H's in the CH2 group that behave as if they are locked in place. However, the predicted C-C bond length in the cyclopropane is quite long: >2 Å.

The hydroxy compound has been characterized by both X-ray crystallography and NMR, and shows a long but reasonable bond length in the cyclopropane.
A really neat recent report describes synthesis of an all-carbon 18-member ring of alkynes that is planar (and likely aromatic).


Last updated: 12/27/2019