It's possible this might work with de-oxygenated air. As long as the nitrogen does not react with the generated oxygen and release NOx, this could be done with

It's possible this might work with de-oxygenated air. As long as the nitrogen does not react with the generated oxygen and release NOx, this could be done with an adsorption process, which is a lot cheaper and less energy than full air separation. Even then, NOx can be mitigated. This process is fascinating in that unlike most "breakthroughs" this has a semblance of a chance of scaling.