Electrochemical impedance spectroscopy analysis of Ⅴ-Ⅰ characteristics and a fast prediction model for PEM-based electrolytic air dehumidification

摘要

PEM-based electrolytic air dehumidification is innovative due to its high efficiency, compact size and cleanness. However, high polarization loss and severe performance degradation have been observed, especially at high applied voltages (2.5 V). Understanding the V-I characteristics is critical to performance optimization. This study experimentally investigated the V-I characteristics and internal response of materials under various operating conditions, with in-situ Electrochemical Impedance Spectroscopy (EIS) methods. Real-time mass transfer, electrochemical polarization and reaction dynamics of PEM components during dehumidification were derived by EIS. Then, a fast prediction model was built to directly predict the dehumidification rate and attenuation without any iteration, suitable for online monitoring and adjustment. Compared to other models, this model can take a quick understanding of the impact of operating conditions on the material characteristics inside the PEM element. The deviations of current density, PEM proton conductivity and moisture removal were 3%, 11.2% and 15.3%, respectively, compared to experiment data. Results showed that when the applied voltage changed from 1.5 to 3.5 V, the high-frequency resistance of the PEM element increased from 1.69 to 2.69 Omega, and the PEM proton conductivity decreased by about 38 times. The sharp drop in PEM proton conductivity resulted in a current attenuation. With this model, requirements for key components of PEM dehumidification were also obtained. Analysis of the overpotential distribution showed that increasing the water retention and reducing the dependence of proton conductivity on water molecules of the PEM can effectively improve the performance. This research provides guidance for the performance optimization and material selection of PEM-based dehumidification. (C) 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.