Give me a photon! No, you better make it 2!

It's pretty well known that polymers (and all organic materials for that matter) can fall apart when they are exposed to UV light. The energy of the photons (E = hc/λ) is in the same order of magnitude of the energy needed to break the chemical bonds [1] and so the end result is not too surprising. In some rarer cases, visible light can do the job too on certain bonds, although this happens only with small organic molecules, not full-length polymers. Since visible light has less energy than UV, the bonds being broken are inherently weaker. Trying to build a polymer with such bonds in the backbone would be especially problematic, since with the high number of bonds in a polymer, the odds of one of them breaking (and thereby degrading the molecular weight of the polymer and corresponding mechanical properties) would be pretty high.

So the report of a polymer completely degrading upon exposure to near-IR light (with their even lower energy photons) can be an eye opener [2], unless you know what it really occurring behind the scenes. The trick is not for the bond to absorb one photon, but to simultaneously [3] absorb two of the photons. The total absorbed energy is now 2hc/λ, so if λ = 470 nm (barely in the near-IR), then the energy absorbed in the bond is the equivalent of a single 470/2 = 235 nm photon, and that's plenty energetic to degrade most organic bonds.

The challenge preventing widespread adoption of this technique is that the near-IR light of each individual beam is not strongly absorbed on its path thought the material, so the odds of having the two photons absorbed in the needed time interval is also unfortunately small. But with the challenge also arrives a tremendous processing advantage: with two-photon absorption, the degradation reactions can be localized to very small spaces if the two photon beams arrive at right angles to each other. Furthermore, this local reaction site can be deep within the material. The degradation occurs only at the intersection of the beams and not along their individual paths.

The advantages of using this approach(subscription required) in therapeutic situations should be pretty obvious, even if the specifics are quite worked out just yet.

[1] For instance, an aliphatic C-C bond has a bond energy of 347 kJ/mol, which corresponds to light of 344 nm, and a C-H bond has a bond energy of 414 kJ/mol, corresponding to light of 288 nm.
[2] A scientific blog that forgets to mention that this is two-photon absorption?
[3] "Simultaneously" is too restrictive of a term - the absorptions need to occur within a very small time interval that depends on a number of variables in the experiment.