Unlike convention polymer electrolyte fuel cell (PEMFC), open-cathode polymer electrolyte membrane fuel cells (OC-PEMFC) use oxidant supply fed directly from the atmosphere, owing to its simple balance of plant (BOP) and integration. Whereas concerns such as lower performance compared to PEMFC, needs to be studied carefully to understand mechanisms and parametric optimization for efficient operation of OC-PEMFC stack.The performance of OC-PEMFC is restricted due to proton conductivity, reaction mechanism, and mass transport limitation1. The effective flow-field (FF) designs can mitigate the detrimental performance degradation arising from dehydration in OC-PEMFC by effective cooling rate2. The simplified designs with optimum pressure drop enhances the performance with balanced effect of parameters such as gas distribution, water removal/retention, heat removal/accumulation.To enhance the performance of OC-PEMFC, we need to understand the physical and chemical mechanisms of the system. These mechanisms help to understand the losses involved caused due to low protonic conductivity, Oxygen reduction reaction, mass transfer, operating conditions, and nature/composition of materials3 etc. Among all electrochemical techniques, Electrochemical impedance spectroscopy (EIS) stands out as a powerful technique that provides a vast extent of information in a short period of time.The presented work investigates the performance of three distinct novel cathode channels with various cross-section designs (CSDs) such as square, boot, and triangular for four cells OC-PEMFC stack of active area 100 cm2. The analysis incorporates EIS examination for critical parameters, including airflow rate, current, temperature, H2 gas condition (humidified and dry). Furthermore, a parametric optimization technique is developed based on EIS and Artificial intelligence (AI) random forest regression (RFR) algorithm, the flow chart illustrating the RFR algorithm is presented in Fig. 1(a).The findings show that, for square design the minimum low frequency resistance 7 Ω cm2 is achieved at airflow rate 3 m3/sec, current 10-15 A, and stack temperature ~ 39 ̊C, as shown in Fig 1(b). Moreover, the AI algorithm successfully predicted the suitable operating condition with root mean square error value of 98%, as illustrated in Fig. 1(c). References J. C. Kurnia, B. A. Chaedir, A. P. Sasmito, and T. Shamim, Appl Energy, 283, 116359 (2021).S. Thapa, V. Ganesh, H. Agarwal, and A. K. Sahu, J Power Sources, 603, 234398 (2024).Z. Tang, Q. Huang, Y. Wang, F. Zhang, W. Li, A. Li, L. Zhang, and J. Zhang, J Power Sources, 468, 228361 (2020). Figure 1
Read full abstract