This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. containing either a thin PL porous layer (40 µm), and/or a DS dense membrane (500 µm), and/or a PS porous substrate (500 µm). Activated carbon (i.e. porogen agent) at 20 wt% was mixed with BSCF to form porous structures. The membranes were processed by dry pressing followed by sintering at high temperatures. The best oxygen fluxes of 2.86 mL min -1 cm -2 at 850 °C were achieved with the PL-DS membrane configuration, which delivered a 42% higher oxygen flux than the DS dense membrane. This improvement was attributed to the larger contact area with the air feed conferred by the thin PL porous layer. However, the PL-DS-PS configuration resulted in a 21% decrease of oxygen flux as compared to the PL-DS membrane which strongly suggested that the PS porous substrate contributed extra resistance to the transport of oxygen. CT scan analysis of the PS porous substrate revealed that the pore volume was concentrated in the center of the PS porous substrate, with a reduced pore volume contribution closer to the external surface. 3D 2 analysis revealed the connectivity of the regions closer to the external surface was over three orders of magnitude lower than that of the center of the PS porous substrate, thus creating a bottle-neck.Although all pores were large (> 1 µm), the expected resistance should be very low, however, the bottle-neck region closer to the external surface provided additional resistance for the transport of oxygen from the membrane interface to the permeate side.