Here’s today’s weird molecule, for sure. A collaboration between IBM-Zürich and Oxford has reported a new allotrope of carbon, this one an 18-membered ring of alternating triple and single bonds (!) People have been speculating about such structures for years, but they appear to be too reactive to spot easily in the wild. There’s some evidence in the gas phase, but not enough to settle arguments about whether such structures even exist, and if they do whether they’re pure alternating-ynes or have some cumulene character (double bond attached right on to double bond, perhaps all the way around the ring). Theoretical calculations are of limited help, since different approaches land on both of the different answers.
As many will have guessed from the IBM team, though, this new work was done at low temperature via scanning-tunneling and atomic-force microscopy. The synthetic scheme is shown at right, with the three remaining alkynes formed via the masking cyclobutenediones. These sorts of compounds had already been prepared twenty years ago in studies towards making the C18 ring by more conventional means, but the cold and delicate touch of the AFM tip (and the isolation of the resulting molecule) are what did the trick this time. The starting material was sublimed onto a copper surface that had “islands” of bilayer NaCl scattered on it, and the individual molecules that landed on these inert salt rafts were the ones selected for the AFM step. There were some partially decarbonylated species found by imaging even at that point, apparently having been produced during the sublimation, but the triangular parent molecule was quite clear (the carbonyls stand out because of their high electron density). The molecule appears to be puckered and not lying completely flat on the surface.
It was then decarbonylated by voodoo. Well, pretty close, anyway. The team raised the AFM tip about three Ångstroms from the molecule and briefly increased the voltage, which provided enough energy (via inelastic tunneling?) for the cyclobutenediones to rearrange and spit out CO molecules. The main species formed were the intermediates where one or two of those events had happened, as in the scheme above. Out of ninety attempts, though, they did get about a dozen molecules of C18 via the triple rearrangement. The triple bonds have higher electron density themselves, of course, and could be seen alternating in both the intermediates and in the final symmetric ring. So we can say that at low temperature, adsorbed onto an inert surface, the alkyne structure is the correct one. You do wonder, though, how much higher in energy the cumulene stuff is, and whether it starts rearranging to that form as it warms up and starts to fall apart. Interestingly, the C18 molecules themselves were pretty weakly bound to the surface. The paper notes that they often jumped around during the manipulations, and were often found having landed on step edges of the surface or next to adsorbed CO molecules, presumably because there was a bit more to grab onto (even at 5 degrees K!)
So put down another weirdo allotrope for carbon. This one (so far) has only been seen as twelve individual molecules, but it can and does exist. Perhaps it (or its higher-ring homologs) are floating around in cold interstellar clouds or something? If you want to see them down here on Earth, though, you’re going to have to work for it. I’m sure it has very interesting properties (the second link in the first paragraph has more on that), but investigating those is not going to be so easy!