Comment by amluto
I find it surprising that the paper has no discussion whatsoever of the thermodynamics of the process. The overall reaction is very endothermic (you can burn ammonia in oxygen as fuel!), so the only way it’s happening at all is that it’s approaching equilibrium, presumably driven by the increase of entropy available by creating a low concentration of ammonia in whatever weird phase it’s created in. Getting high concentrations from a similar process is going to need some energy-consuming step to shift that equilibrium.
Worse, they seem to be using some chilled object to condense ammonia solution from the air, so you’re also paying the energy cost of keeping it cold, which means you’re paying the full cost of producing a lot of water from atmospheric water vapor. Maybe a future improvement could start with liquid water.
The droplet size appears to be critical here.