Voltage jump during polarization of a PEM fuel cell operated at low relative humidities

Abstract

A PEM fuel cell with a Nafion 211 membrane-based membrane electrode assembly (MEA) was tested with an H₂/air stoichiometry of 2/4 at 25%, 50%, 75%, and 100% relative humidities. A voltage jump on the polarization curve was observed when the cell was operated at a lower humidity. This phenomenon may be explained by the water back-diffusion from the cathode into the membrane resulting in both a non-uniform water distribution in the membrane and a liquid-equilibrated interface between the membrane and the anode catalyst layer. Experimental results obtained by AC impedance spectroscopy measuring the MEA resistance (membrane+catalyst ionomer layers) at different current densities as well as collected polarization data at high feed-gas flow rates (or at low backpressures) and high temperatures all confirmed the validity of the proposed water back-diffusion hypothesis.

Introduction

Proton exchange membrane fuel cells (PEMFCs) have been recognized as one of the most promising clean power technologies, mainly due to having few to no emissions when used in transportation and stationary power applications. However, in order to speed up fuel cell commercialization, present technical barriers, such as high cost and low reliability and durability, must be overcome [1].

Simplifying the fuel cell system is expected to be an effective way of reducing costs, as are other approaches such as reducing component–material costs [2]. Usually PEM fuel cells are operated with fully hydrated gases because they provide the best performance. In order to humidify the feed gases, a humidifier is a key system component, especially when the fuel cell is operated at lower temperatures (<80 °C). However, this humidifier not only contributes to the system's weight and cost, but is also a parasitic load, resulting in a reduced efficiency. Thus, reducing the power load of the humidifier by operating the fuel cell at a low relative humidity (RH), or eliminating the humidifier by operating at zero RH, is always desirable. Buchi and Srinivasan [3], Qi and Kaufman [4], and Picot et al. [5] investigated PEMFC single cell and stack operating without external humidification. Cell performance and water symmetry factor were presented. Transient performance of PEMFCs at zero RH was reported by Yu and Ziegler [6] and Ziegler et al. [7]. Hysteresis was observed in the forward and backward voltage scans, which indicates water transport and distribution in the membrane electrode assemblies (MEAs).

Water distribution and transport in the membrane and MEA are important in determining PEMFC performance. A dry anode catalyst layer leads to slow proton transport and slow reaction kinetics, while an excess of water in the cathode side results in flooding, leading to a limiting current [8]. Extensive studies have been reported on modeling and experimental investigation of water management, which was pioneered by Springer, who predicted the net water per proton flux ratio and a way to alleviate the problem of membrane resistant by using thinner membranes [9]. Later, various models such as the diffusion model and the two-phase model have been used to simulate water transport and water distribution [7]. Recent work published by Weber and Newman simulated non-uniform water distribution in the polymer electrolyte membrane [10]. Shah reported a comprehensive non-isothermal one-dimensional model for PEMFCs, where the effect of water on cell performance was modeled and validated by experimental data [11]. Berg presented a simplified model for water management operating under prescribed current with iso-potential plates [12], where co- and counter-flowing air and fuel were considered. Sensitivities of current distribution along the channel to humidities, gas composition, and stoichiometries were reported. Bao analyzed the water and thermal management in a single cell and cell stacks [13]. Stoichiometric ratio and cathode outlet pressure effect on the performance were analyzed. Fuel inlet humidity on the cell performance was modeled by Jang, which was used for designing a baffle blocked flow field [14]. Kraytsberg [15] investigated the water management problems and suggested the removal of water from the anode side as a solution. Eikerling's recent modeling showed that the cathode catalyst layer acted as a watershed, regulating the balance of opposite water fluxes toward the membrane and the cathode outlet [16]. In situ measurements by Teranishi et al. [17] using magnetic resonance imaging (MRI) showed the non-uniform distribution of water in the membrane, and Albertini's results using X-ray diffraction (XRD) showed the dehydration of the polymer electrolyte membrane when it is subjected to in situ XRD studies [18]. Ciureanu and Badita [19] investigated the water balance at both anode and cathode compartments and found that the direction of net water flux changes with current density. Buchi and Scherrer [20] showed that the membrane resistance increased with current density, which indicates a non-uniform water distribution in the membrane.

Water distribution and transport are affected by cell temperature, gas flow rate, and backpressure. Buchi and Srinivasan [3] and Yu and Ziegler [6] reported the cell performance and water removal at different conditions. The temperature investigated was limited to a range less than 60 °C, where water remains in the liquid state or liquid/vapor state, depending on the cell conditions. Thus it is desirable to investigate water transport in a wide temperature range such as below and above 100 °C, where water exists in different phases.

A common observation is that the fuel cell's performance at zero RH is lower than it is at high RH, mainly due to the high ionic resistance of both the membrane (Nafion) and ionomer in the catalyst layers, caused by the low water content [3], [8], [21]. For practical applications, several methods have been introduced to retain water inside the MEA in order to solve the water management problem: (1) the use of self-humidifying membranes [22], [23], [24], [25], [26], [27], (2) the input of water through wicks or hollow fibers [28], [29], and (3) the design of special flow fields [30], [31]. However, it is desirable to decease the RH without compromising the fuel cell performance.

Here we report the effect of RH on the performance of H₂/air PEMFCs. The RHs investigated were 25%, 50%, 75%, and 100%, in the temperature range of 80–120 °C. In the course of running fuel cell polarization (voltage vs. current density curves), an unusual phenomenon was observed: a voltage jump occurred in a certain current density range. The effects of cell temperature, gas flow rate, and backpressure on the voltage jump were also investigated. Although water management and hysteresis between the backward scan and forward scan of PEMFCs have been well documented, to the best of our knowledge, this voltage jump as well as the effect of factors such as cell temperature on this voltage jump over a wide temperature range have not been reported. We present this preliminary report and give some tentative qualitative explanations based on a water distribution hypothesis and experimental verification.