Carbon capture's energy problem is not thermodynamic. The energy required to reverse CO₂ adsorption — to break the bond between a sorbent and a captured molecule — is small. The energy required to heat a bed of sorbent material to the temperature at which that bond breaks is large. The gap between these two numbers is the gap between moving one molecule and heating everything near it.
Conventional carbon capture uses temperature swing adsorption: heat the sorbent to release captured CO₂, cool it to re-adsorb. Amine scrubbing, the current industrial standard, consumes 2.1 to 2.6 gigajoules per tonne of CO₂ in thermal regeneration alone — enough to consume a quarter to two-fifths of a power plant's output. Most of this energy goes not to breaking CO₂-sorbent bonds but to raising the temperature of the entire material bed, its housing, and the gas stream flowing through it. The useful work is a small fraction of the total energy input.
Feringa's group at Groningen, working with collaborators in Milan and Warsaw, built a porous aromatic framework with azobenzene pendants that captures and releases CO₂ using visible light. Green light at 530 nanometers bends the azobenzene molecules into the Z isomer — a compact, twisted conformation that protrudes into the pore space and physically displaces adsorbed gas. Blue light at 420 nanometers straightens them back to the E isomer, opening the pores for re-capture. The framework backbone — built from irreversible carbon-carbon bonds between aromatic building blocks — survives temperatures above 400 degrees Celsius and resists acids, bases, and organic solvents.
The critical advance is not the switching itself — azobenzene photoswitches have been known for decades. It is that the switching occurs throughout the bulk of the material, driven by visible light. Previous approaches used ultraviolet light, which is absorbed strongly at the surface and cannot penetrate deeper. Only the outer shell of the material participated; the interior remained unswitch. The ortho-fluorine substitution on the azobenzene red-shifts the absorption into the visible range, where penetration is deeper. Solid-state NMR confirmed near-complete isomerization across the full volume.
The energy economics change qualitatively, not just quantitatively. Thermal regeneration delivers energy indiscriminately — every atom in the sorbent bed receives kinetic energy, regardless of whether it is near a CO₂ molecule. Photoisomerization delivers energy to specific molecular bonds — the azobenzene chromophore absorbs the photon and changes shape. Nothing else heats up. The material stays at ambient temperature throughout the cycle. There is no cooling step afterward because nothing was heated.
The difference is analogous to the difference between heating a house and flipping a light switch. Both deliver energy, but the light switch delivers it to the device that needs to change state. The furnace delivers it to everything in the room, and the device changes state incidentally, as a side effect of the entire environment reaching the required temperature.
The practical implications are distant — this is lab-stage research, with industrial deployment perhaps decades away. The synthesis uses expensive palladium-catalyzed cross-coupling, the light delivery at scale requires engineering that doesn't exist, and the actual CO₂ swing capacity hasn't been publicly characterized. But the principle established is not incremental. It is a different category of energy delivery. When the fundamental limit of a process is not thermodynamic but logistic — not how much energy is needed but how precisely it can be delivered — then the improvement comes not from better materials within the existing paradigm but from changing how energy reaches the molecules that need it. The paradigm shift is from heating the room to flipping the switch.