Oxygen concentrators make use of adsorption, a phenomenon in which gas molecules stick temporarily to surfaces. The forces that bind gas molecules to surfaces are weaker than ordinary chemical bonds, but they are strong enough to keep some molecules in place for seconds, minutes, or even longer. Most surfaces are decorated with adsorbed molecules and the only really clean surfaces are those prepared in nearly perfect vacuum conditions. Some molecules stick better than others to each surface, a fact that makes it possible to use adsorption to separate various molecules from one another.
The heart of an oxygen concentrator is a porous material known as a zeolyte. With a vast network of tiny holes, it resembles a miniature Swiss cheese and presents an enormous amount of surface area on which gas molecules can adsorb. At ordinary temperatures and pressures, air molecules stick occasionally to the surfaces of some of the pores. Nitrogen molecules stick more often than oxygen molecules because nitrogen molecules bind somewhat more strongly to the zeolyte surface than do oxygen molecules. Zeolytes tend to concentrate oxygen in the air by removing most of the nitrogen molecules.
To extract as much nitrogen as possible, an oxygen concentrator pumps pressurized air into a container of zeolytes. The zeolytes adsorb most of the nitrogen in this air, leaving nearly pure oxygen for breathing. But the zeolytes eventually saturate with nitrogen molecules and can adsorb no more. At that time, the oxygen concentrator removes all gas from the container and the nitrogen molecules gradually leave the zeolyte surface. After this regeneration cycle, the container of zeolytes is ready to begin concentrating oxygen again. To keep oxygen flowing at all times, a typical oxygen concentrator has two separate zeolyte-filled containers. At any given time, one container is providing oxygen for breathing while the other is regenerating by releasing its stored nitrogen into the air.
Answered by Louis A. Bloomfield of the University of Virginia