The experiment proceeds in a vacuum environment (so that the electrons can travel freely) where the sample is excited by a high energy electron beam (typically 5 kilovolts and 20 microamps). These high energy electrons kick out core level electrons from atoms on the surface. The surface atoms, now in highly excited states, quickly relax by having higher level core or valence electrons fall down to the excited core hole. The excess binding energy still present in the atom is removed by yet a third electron (but now near the valence shell) that is ejected, carrying with it a unique kinetic energy determined by the relative energies of the three states involved. Hence, the escaping electron's kinetic energy is a unique fingerprint of the atom from which it came. Therefore, AES is a means of identifying the chemical nature of atoms on the surface.
Here is a schematic representation of an AES experiment with a retarding field analyzer (RFA). The incident electron beam impinges upon the surface, ejecting the core electrons. Some of the secondary electrons that are subsequently emitted are Auger electrons. They are energy selected by potentials applied to the grids and finally measured as a current at the collection surface.
Here is a typical AES spectrum taken on a sample of silicon shortly after being placed inside the vacuum chamber. Only adsorbed contaminants (mainly CO, CO2, and H2O) are in the surface layer and only peaks associated with carbon and oxygen atoms are observed.