The paper in Nature Chemistry is here: go.nature.com/2KpqKdr
Przemyslaw Gawel, University of Oxford:
In Oxford, we are exploring new ways to protect fragile polyyne chains so as to be able to investigate and exploit their electronic properties. So far, our approach has mainly involved protection by supramolecular encapsulation, where polyyne chains were threaded through protective macrocycles (J. Am. Chem. Soc. 2016, 138, 1366). However, we are searching for new protection strategies. I first met Leo Gross at the Surface Science Sumer School, which I co-organised at EPFL in Lausanne, Switzerland in 2015 where Leo was a speaker. There, I became fascinated with possibilities of modern non-contact AFM techniques and I kept following Leo’s research. Then, I met him at the ElecMol conference in Paris in 2016, where he presented their latest results on reversible Bergman cyclization (Nature Chem. 2016, 8, 220). During his talk, I had the idea to perform the FBW rearrangement on surface using the same technique, since both these reactions are initiated by breaking the C-Br bond.
After I came back to Oxford, we started working on this ‘side project’ immediately. After initial lack of success on more complicated macrocyclic systems, we decided to take a step back and prepare a simple model system to check our hypothesis: ‘Can 1,2-shifts be initiated by atom manipulation on a surface?’ These experiments produced beautiful images showing intermediate radical and final triyne. It was so exciting to see that this type of rearrangement can be initiated by an AFM tip and monitored with atomic resolution. Next, we decided, together with researchers from IBM’s Zurich lab, to make precursors of longer polyynes to check the scope of this transformation. To our delight, all these systems worked beautifully.
Atom manipulation and nc-AFM imaging are performed at very low temperature (5 K) on an inert surface (NaCl). This enables the formation and characterisation of highly unstable species, such as very long sp-carbon chains. The method of polyyne formation we developed and described in this paper opens new possibilities in the chemistry of carbon-rich structures including carbyne. Currently, we are working on other sp-carbon-containing structures generated on surface using this transformation.
Figure 1. a) Carbyne in polyyne (top) and cumulene (bottom) forms; b) FBW rearrangement; c) Reaction of the first model system; d) AFM images of all steps of rearrangement on surface.
Leo Gross, IBM Research - Zurich:
I was extremely excited when Przemek (Przemyslaw Gawel) proposed his idea and curious if it would be possible. One reason for the excitement was that the product we were trying to create was very interesting: The polyynes are archetypical molecular wires, yet they are very challenging to obtain on surfaces. Moreover, the reaction that Przemek proposed was very interesting and challenging. It involves a molecular skeletal rearrangement, i.e. changing the connectivity within the carbon backbone of the molecule. Such a complicated rearrangement to date has not been performed by atom manipulation, that is by using the tip of the microscope to ignite the chemical reaction. And it was not clear at all if this could work out. It was clear that demonstrating and controlling such a reaction would bring us one step closer to making custom molecules and molecular networks by atom manipulation.
It turned out that the proposed rearrangement by atom manipulation did not work on a copper substrate. However, the idea worked on thin films of sodium chloride presumably because it is less reactive. We were fascinated and immediately tried to get to more complex series of reactions employing the explored mechanism. Our co-workers from Oxford University made precursor molecules of increasing complexity, which required two consecutive 1,2 shifts (skeletal rearrangements) to be performed ending up in polyyne wires of up to 16 sp carbon atoms (8 consecutive triple bonds) in length.
By atom manipulation we created wires of different length and we could characterize the intermediate steps of the reaction with atomic resolution. Finally, we employed scanning tunnelling spectroscopy to characterize the electrical properties of the molecular wires created.