From the development of anion exchange membrane fuel cells it is known that airborne CO2 is easily absorbed in this type of membranes and has a detrimental effect on the conductivity of the membrane1. It was also found, however, that at sufficiently high current densities carbonate and hydrogen carbonate ions are flushed out of the membrane and ion conduction occurs predominantly via hydroxyl ions1. This behavior led to the idea of using AEM based cells for gas separation for two different type of applications. So, during the first waves of COVID-19 medical oxygen became a scarce resource. Local production of oxygen by separation from air is an interesting approach to overcome this issue. This should be feasible using an anion exchange membrane cell with an air side oxygen reduction electrode and an oxygen evolution electrode. The hope is that in combination of these two reactions the required cell voltage can be kept much below typical water electrolysis voltages so that operation with simple PV cells or modules is possible. At the same time the question is if current densities can be achieved at which the OH- transport and thus the anodic release of O2 dominates, as an enrichment of CO2 must be avoided for this application. On the opposite there are worries about the effect of increasing CO2 concentration in occupied confined rooms like vehicles. Here a device that removes CO2 from air without consuming much oxygen is desirable. The removal of CO2 from the air supply to AEM fuel cells by an additional cell placed upstream of the cathode inlet and downstream of the anode outlet was demonstrated by Matz at al.2. As for CO2 separation only, no hydrogen is available for the anode reaction here again the operation against an oxygen evolution electrode is required, optimized, however, towards the CO2 transport minimizing O2 transport.In this study for one type of AEM membrane and ionomer, namely Fumapem FAA3 and Fumion (both Fumatech BWT, Germany) the effect of different cathode electrodes and electrode materials as well as the effect of operation conditions towards the selectivity of oxygen or CO2 transport were investigated. For that purpose, test set-ups were used were the cathode side was supplied with synthetic air, compressed air or synthetic air with added CO2 and the anode side with an inert gas N2 or helium. Nass spectrometry and dedicated CO2 concentration sensors were used to determine the anodic release of O2 and CO2 as well as the reduction of O2 and CO2 concentrations on the cathode side.First measurements of the oxygen transfer using a cell with a silver cathode (Oxag, Gaskatel, Germany) and a nickel foam anode (Gaskatel, Germany) it was found that the oxygen transfer had a promising faradaic efficiency exceeding 80% (cf. Fig. 1). The cell voltage of 1.5 V at 90 mA cm-2 was however still too high for the application. Replacing the Oxag silver electrode by an industrial ORR electrode for alkaline cells did not yield an improvement (cf. fig 2). Increasing humidification to 100 % r.H. improved the current density (cf. Fig 3). Using high humidification conditions, the CO2 effect was studied by switching from synthetic air to pressured air and back and 160 mA cm-2. A rapid increase of the anodic CO2 concentration to values exceeding 2000 ppm was observed (cf. Fig. 4). Comparing CO2 transfer for cells with Oxag cathode and the commercial ORR cathode again showed that the latter would not improve the situation. It proves however, that the CO2 transfer rate can be influenced by the cathode catalyst. Regarding the CO2 separation test with synthetic air with added 3.85 Vol% of CO2 as cathode feed have shown that under such conditions, mainly CO2 is transported (cf. Fig. 5). In total it will be shown that CO2 separation with state-of-the-art materials is already possible. The use of O2 separation regarding onside production of medical O2 today fails due to still too high CO2 selectivity. The fact that CO2 selectivity seems to be dependent on the cathode catalyst may allow further optimization also valuable for AEMFC fuel cell developments.Acknowledgment: The work received financial support in parts from the Fraunhofer-Gesellschaft by the Demo-medVer project part project e3C-O2 as part of the Fraunhofer cs. Corona program. M. Inaba, Y. Matsui, M. Saito, A. Tasaka, K. Fukuta, S. Watanabe and H. Yanagi, Electrochemistry, 79, 322–325 (2011). 2. S. Matz, B. Setzler, C. M. Weiss, L. Shi, S. Gottesfeld and Y. Yan, Meet. Abstr., MA2020-02(34), 2227 (2020). Figure 1