The Dangers and Dilemma
Isolating and observing the properties of 1D carbon has been an ongoing challenge in the chemical and materials science community. This interest, in part, stems from their ability to act as model systems for the 1D allotrope of carbon, carbyne, providing insight into its spectral, mechanical and electronic properties.
Because of their violent reactivity, researchers typically need to add bulky end-capping groups to the chains just to avoid accidentally synthesising a benchtop bomb. While these caps keep the molecules stable enough to measure, the choice of end-cap species forces the carbon backbone into either a semi-conducting or metallic form, obscuring the intrinsic properties of a pure chain. To mitigate this, researchers have, for example, grown long chains of 1D carbon inside double-walled carbon nanotubes. This keeps them stable without capping groups, but makes direct electronic characterisation of the material incredibly difficult.
The Australian Job
It was against this backdrop that our lab received a package from Paul Low’s group at the University of Western Australia. Already working with phenyl capped oligoynes, my supervisor Andrea and I immediately saw their potential.
The package contained a series of oligoyne-bridged gold phosphite compounds synthesised by Jarred, along with promising preliminary data from Elena, both Paul’s PhD students at the time. The goal: transmetalate the carbon chains to electrodes in-situ.
However, measuring beyond the C6 compound was challenging. The junctions needed a combination of high stability junction cycling while retaining surface mobility. As luck would have it, I was just wrapping up a full redesign of our flagship STM-based measurement system, Leviathan, with the help of Tom and Adam (both PhDs at Liverpool), the code of which I have since made free and open-source and available here. This new design allowed the high-speed, high-throughput, set and forget style data collection we needed for this project.
With these upgrades in place, I was able to finally complete the measurements which showed semi-conducting decay with length, but with an astoundingly low attenuation and high overall conductance. Excellent! Done!
Extending the Series
But we were all of us deceived, for another set of compounds were being made. In the land of Australia, in the lab of Low, Jarred Potter synthesised in secret, an extension of the series…
Our fate was sealed. Of course we had to measure them.
By the time the compounds (C10 and C12) arrived, I was finalising my thesis so Andrea and I split the work. Andrea took C10 and I took C12. C10 behaved exactly as expected. I sent him the C12 raw data and he processed it. I remember vividly being on a train and looking at the datasets on my phone. I messaged Andrea, thinking I’d made a mistake somewhere, ‘Is C12 higher in conductance than the others?’. After a little while (probably Andrea carefully checking it over as always) I finally got a response: ‘Holy crap! Yes’. We didn’t believe it. Quite literally, we even re-did the experiment to be sure. But it was true, C12 had shot up in conductance, completely defying the conductance decay that we expected.
After Jarred, now thought to have lost many fingers in making these compounds, made the C14 and C16 gold-capped molecules, we got to work measuring their conductance properties. Then, something remarkable happened. The conductance didn’t decay at all. It was showing length independent charge transport, a characteristic associated with the metallic, non-alternating, form of 1D carbon.
This meant I had much work to do. Recovering a grand total of 2 days after my thesis submission, I jumped straight back into the lab. Current-voltage measurements showed the C10 to C12 transition was apparent across a wide voltage range, approaching a linear, metallic regime at C16. At 1 V and around 2 nm (2 billionths of a metre) it even broke a record for the highest ever current (around 40 µA) observed at that length scale.
Where Has My Hat Gone?
To understand this transition, I had to swap my experimentalist hat for my theoretician one. I reparametrised a simple chemical model of carbon chains (SSH model) into a 2-parameter one, consisting of only interfacial interaction strength and bond order (a proxy for the semi-conducting to metallic transition).
Using this parameterisation and one of my weekend side projects, the Tight-Binding Funtime Engine (I know, I need better hobbies), I calculated a series of heuristic conductance trends by varying both the bond order and the interfacial interaction. Only by varying the bond order was the experimental trend reproduced correctly. We also ruled out mid-gap states using DFT with the help of Roberto.
Searching for Spectroscopic Proof
Sceptical of relying solely on conductance and theory, we sought spectroscopic proof: if correct, the oligoyne backbones should turn into cumulenes. The best way to do this was with Raman vibrational spectroscopy.
So, I set about making some gold nanoparticles and conducting the spectroscopy experiments and controls (a big thanks to the Materials Innovation Factory Raman team for providing access to their spectrometer). Comparing the powder Raman to surface enhanced Raman spectra revealed what we expected. Upon depositing to the gold nanoparticles the strong oligoyne peak red-shifted to a more cumulenic signature as length increased!
Explosive Limits and Future Endeavours
So, where does this leave us? By combining far too much experimental data, extensive synthetic work on the Australian end, and a fair chunk of theoretical and computational work, we can finally study pure, end-cap free 1D carbon in a stable junction environment. In doing so we established a reliable, reproducible way to actually make, measure and explain the properties of 1D carbon wires in junctions, including an observed oligoyne-to-cumulene transition.
But, in science, answering one question inevitably creates a bigger one. Now that we have the techniques and the analytical framework to understand it, we are nicely positioned to push these compounds even further. Although Jarred tried synthesising C18 and C20, they rapidly decomposed or, in the case of C20, exploded (Jarred is healthy and well, and fingers still intact, don’t worry). But, hitting the explosive limit in the fume hood just means we have to get a bit more creative in the physical chemistry side of the lab.
With Leviathan purring and the new theoretical models getting better every day, we finally are in the right place to start asking some big questions about 1D carbon and its bulk allotrope carbyne. This paper isn’t the end of the story, it’s just the foundation. I for one can’t wait to see what the data tells us next.