Novel technique for measuring oxygen crossover through the membrane in polymer electrolyte membrane fuel cells

https://doi.org/10.1016/j.ijhydene.2013.04.142Get rights and content

Highlights

  • The exact amount of oxygen crossover that reacts with hydrogen is investigated.

  • Effective oxygen crossover ratio is in a range between 0.927 and 0.933.

  • Only 93% of the entire oxygen crossover can affect membrane degradation.

Abstract

In this study, the exact amount of oxygen crossover that reacts with hydrogen has been investigated using a mass spectrometer system. By measuring the amount of oxygen crossover that reacts with hydrogen, the exact amount of oxygen crossover that affects membrane degradation and/or water generation can be calculated under the fuel cell operating conditions. The amount of oxygen crossover that reacts with hydrogen is expressed as an effective oxygen crossover ratio, which is in a range between 0.927 and 0.933 under the fuel cell operating temperature conditions. This means that approximately 93% of the entire oxygen crossover through the membrane can affect membrane degradation and/or water generation at the anode catalyst layer. Thus, the effective oxygen crossover ratio should be considered as a novel index of oxygen crossover because it represents the exact amount of oxygen crossover that reacts with hydrogen.

Introduction

In recent times, thin membranes with thickness ranging from 18 to 50 μm have been widely adopted in polymer electrolyte membrane fuel cells (PEMFCs) to improve fuel cell efficiency and allow high current density operations. However, the use of thin membranes aggravates gas crossover and may decreases fuel cell durability. It is well known in the literature [1], [2] that hydrogen crossover from the anode side to the cathode side may deteriorate the durability of PEMFC. On the other hand, as shown in Fig. 1, oxygen can also permeate from the cathode side to the anode side. Oxygen crossover also causes several problems related to limiting the durability of PEMFCs. Willsau et al. [3] reported that oxygen crossover influenced the electrochemical carbon corrosion in a cathode gas diffusion layer (GDL). It is well known that the air/fuel boundary is created at the anode side if oxygen is transported from the cathode side to the anode side. This increases the potential of the cathode to a value greater than the open circuit voltage and quickly corrodes the cathode carbon layer [4], [5]. Another problem caused by oxygen crossover is membrane degradation [6], [7], [8]. Oxygen crossover does provide a means for the formation of hydrogen peroxide and thus oxygen-containing radicals, which can deteriorate the membrane durability.

As oxygen crossover causes the degradation of the membrane, several techniques including the time-lag method [9], [10], [11], the volumetric method [12], the electrochemical monitoring method [13], [14], [15], [16], [17], [18], [19] and the direct gas detection method using gas chromatography (GC) [20], [21] or mass spectrometry (MS) [22], [23], [24], [25], [26], [27] have been used to measure the oxygen crossover rate. Both the time-lag and volumetric methods apply higher pressure at one side of the membrane, then the gas molecules start sorbing into the high-pressure side of the membrane (upstream), and subsequently diffuse to the opposite low-pressure side of the membrane (downstream). The time-lag method involves monitoring the transient accumulation of species due to permeation on a fixed volume in a downstream reservoir [11], whereas the volumetric method obtains the flow rate in the downstream [12]. These two methods offer a simple and effective technique for determining the oxygen permeation rate, but they are not suitable for the fuel cell operating conditions (i.e., wide ranges of humidity and temperature). Therefore, the use of results from these methods may be limited in both numerical and experimental studies of PEMFCs. For in-situ measurements of oxygen crossover under the fuel cell operating conditions, electrochemical monitoring and direct gas detection methods are employed. In the electrochemical monitoring technique, one of the sides is exposed to an acid solution with a counter electrode while a reactive gas is supplied to the other side of the membrane. Current is then generated due to gas crossover and the crossover rate through the membrane can be estimated by measuring the current over time [17]. However, the electrochemical monitoring technique is still limited under actual operating conditions of PEMFCs because one side of the membrane should be exposed to the acid solution which is not a reactant gas for the fuel cell operation. Another way to measure the oxygen crossover rate of PEMFCs is to use direct gas detection methods. Recently, direct gas detection methods using GC or MS were newly suggested to measure the gas crossover rate under the actual fuel cell operating conditions. Broka et al. [20] and Mohamed et al. [21] reported oxygen permeability through a Nafion® 117 membrane by using the GC system at different values of temperature and relative humidity (RH). Shim et al. [23] performed oxygen crossover measurements using the MS system. An illustration in Baik and Kim [22] shows that the direct gas detection system is connected at the exit of the conventional fuel cell system, implying that the oxygen crossover rate under the actual fuel cell conditions can be measured. The measurements of the oxygen crossover rates under conditions that are as close as possible to actual conditions of PEMFCs appear to be important; therefore, the direct gas detection method using GC or MS is more suitable for oxygen crossover measurements.

During actual operation of PEMFCs, the amount of oxygen that reacts with hydrogen is very important because it affects membrane degradation and/or water management of anode side. Oxygen that permeates from the cathode side can be divided into two parts under the fuel cell operating conditions, as shown in Fig. 1. First, some amount of the oxygen reacts with hydrogen at the anode catalyst layer. However, not all of the oxygen gas that permeated from the cathode side reacts with hydrogen at the anode catalyst layer, which means there is a residual amount of oxygen that is released via the anode outlet [2]. Some reports mentioned that there is no residual oxygen in the anode exit stream because all the oxygen reacted with the hydrogen at the anode catalyst layer [23]. However, the fact that there was a small amount of residual oxygen at the anode side was repeatedly checked in this study, indicating that not all oxygen that permeated from the cathode side reacted with hydrogen at the anode side. Thus, it is important to distinguish the amount of oxygen that reacts with the hydrogen at the anode catalyst layer from the total amount of oxygen crossover.

In this study, quantitative measurements of oxygen crossover that reacts with hydrogen have been conducted using a MS system. The amount of oxygen crossover that reacts with hydrogen (specific oxygen permeability) was measured and compared with the conventional normal oxygen permeability. Lastly, an effective oxygen crossover factor that is a correlation between the normal and specific oxygen permeabilities was suggested as a novel index of oxygen crossover.

Section snippets

Single fuel cell preparation

A single PEMFC with an active area of 25 cm2 was used for the quantitative measurements of oxygen crossover in this study. The single PEMFC consisted of end plates, graphite bipolar plates, gas diffusion layers (GDLs), gaskets and a membrane electrode assembly (MEA). Graphite bipolar plates with conventional five-serpentine flow fields were used for both the anode and the cathode. A GDL sample with a thickness of 325 μm was obtained from a commercial manufacturer, and it consisted of both a

Gas selectivity measurement

The effect of gases that have different molecular weights, such as hydrogen, helium, and nitrogen, on oxygen crossover could be different, so gas selectivity tests were conducted to identify the effect of the molecular weight of each gas. For configuration of the gas selectivity, non-reactive gases such as helium, nitrogen, and argon were used as supplying gases at the anode side, whereas oxygen was supplied at the cathode side. Hydrogen was not used in this test because it can react with

Conclusion

Quantitative measurements of the amount of oxygen crossover that reacts with hydrogen have been conducted using a mass spectrometry (MS) system. Oxygen permeability values estimated in this study show similar trends to those estimated by other studies. To identify the effect of gases with different molecular weights on oxygen crossover, gas selectivity tests were conducted. Results from these tests show that the oxygen concentration shows no significant dependence on molecular weight. Thus, the

Acknowledgments

This work was supported by the IAMD of Seoul National University. Additional support by the World Class University (WCU) and National Research Laboratory (NRL) program of the Ministry of Education, Science and Technology is greatly appreciated. The support from the New & Renewable Energy of the Korea institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea Government Ministry of Knowledge Economy (2011301003008A) is also appreciated.

References (34)

Cited by (19)

  • A review of membranes in proton exchange membrane fuel cells: Transport phenomena, performance and durability

    2021, Renewable and Sustainable Energy Reviews
    Citation Excerpt :

    The experimental studies on the gas crossover in membranes mainly focus on the influence of the shape, number and size of the pinholes manufactured by hand on the performance and the specific degradation process of pinhole formation [190,191]. Many measurement methods for gas crossover can be found in the literature [192,193], and most of the studies on crossover phenomena are carried out under open-circuit conditions through in-situ measurements [194–197]. For example, the gas crossover rate of the membrane can be experimentally studied by measuring the gas permeability coefficient or measuring the hydrogen crossover current density [198].

  • Numerical study for in-plane gradient effects of cathode gas diffusion layer on PEMFC under low humidity condition

    2020, International Journal of Hydrogen Energy
    Citation Excerpt :

    These assumptions are commonly used in PEMFC modeling [37–42]. With literature survey about reactant gas crossover [43–45], it is concluded that reactant gas (oxygen, hydrogen, and nitrogen) crossover rate through membrane is much lower than flow rate of main rate, and crossover through membrane is negligible like other studies [37, 42, 51]. It is a one-dimensional, isothermal, isobaric, and steady-state process.

  • Proton exchange membrane fuel cell reversible performance loss induced by carbon monoxide produced during operation

    2016, Journal of Power Sources
    Citation Excerpt :

    It was believed that hydrogen peroxide is formed and attacks carbon to generate surface oxides and then carbon dioxide. This hypothesis is supported by the precise determination of the formation of hydrogen peroxide with a Rotating Ring-Disc Electrode (RRDE) by Paulus et al. [30] and by the crossover of oxygen measured by many techniques [31,32]. Moreover, H2O2 is known to introduce defects in carbon [33].

  • Water distribution in high temperature polymer electrolyte fuel cells

    2016, International Journal of Hydrogen Energy
View all citing articles on Scopus
View full text