“…The fitting results for the HF capacitive loop also show that, as the structure of the protective system becomes more complex, the CPE exponent (n) moves away from the ideal unit value. Hence, a normal distribution of time constants within the coating thickness can be associated, as suggested by Hirschorn et al [93], an argument that has been recently used by Prada-Ramirez et al [55][56][57] to explain the impedance response of anodized Al as sealing occurs within the pores. It is important to emphasize that, as extensively discussed in the literature, the oxide layer on top of the Mg surface exposed to different electrolytes is composed of a porous, non-protective layer, mainly Mg(OH) 2 , on top of a thin MgO compact and protective layer [75,83,84].…”