Self-healing materials are growing rapidly in popularity, at least as far as research topics go. I mentioned one example a couple of weeks back that was pretty impressive , but a new paper doesn't even come close to reaching what was achieved previously. (More on that in a minute)
From the well-respected Matyjaszewski group (a perennial candidate on Chembark's Nobel Prize predictions) comes this offering: "Introduction of self-healing properties into covalent polymer networks via the photodissociation of alkoxyamine junctions". The chemistry is also very trendy using not only ATRP (this is the Matyjaszewski groups after all) but also "click" chemistry  to build up a network with alkoxyamine junctions. The paper goes into extensive detail (lots of Electron Spin Resonance spectra) about how UV light can be used to breakdown the network thereby increasing the mobility of the chains so that additional UV light can be used to reform the network.
I don't think this paper was refereed very well. The chemistry aspects are all spot on, but when you are talking "self-healing", emphasis needs to be placed just as much on the mechanical properties and how much they recover after the healing is over with. This paper really falls short in that manner, and as I noted in footnote 2, it can be questioned if there was any damage that needed healing in the first place.
So is this polymer really self healing? At this point, I'm not convinced in the least.
 For crying out loud, polymer people, can we decide on what "click" chemistry is? Half the world wants it only refer to the the copper-catalyzed alkyne/azide cycloaddition reaction, (as is the case in this paper that I'm discussing) and the other half wants to include thiol-ene, Diels-Alder and a whole host of other reactions such as was originally proposed by Nobel Prize winner Sharpless. [†] I prefer the latter option myself, but more importantly, I prefer that we make a decision and stick with it.
 Scratched? They put 600 nm indentations into the film using an ATM in tapping mode! Scratched? Wouldn't compressed be a better word? I'm struck by how the "scratch" fills in from the bottom up (as you would expect from elastic rebound), even though the top of the coating receives the most light and should therefore heal faster. Maybe it is a scratch, but you better prove it, or at least make one in a more conventional manner that isn't open to questions.
[†] You don't suppose that the difference in usage is what sets off the Nobel Prize winner from the wannabe, do you?