Elsevier

Applied Energy

Volume 254, 15 November 2019, 113712
Applied Energy

Localised electrochemical impedance spectroscopy investigation of polymer electrolyte membrane fuel cells using Print circuit board based interference-free system

https://doi.org/10.1016/j.apenergy.2019.113712Get rights and content

Highlights

  • Interference-free localised electrochemical impedance spectroscopy diagnostic system.

  • Segment gas diffusion layer does not be needed in a localised impedance measurement.

  • Anode inlet gas relative humidity is essential in the increase of catalyst utilisation.

  • The optimum reaction site shifts as the current density increases.

  • Cathode outlet suffers mass transport problem even at a medium current density.

Abstract

Polymer electrolyte membrane fuel cells are promising power sources for vehicle and other portable applications due to their high energy efficiency and zero pollution emission during operation. To improve the performance and reliability of polymer electrolyte membrane fuel cells, effective and accurate diagnostic tools are urgently needed for polymer electrolyte membrane fuel cells practical applications. Different from the previous diagnostic methods that may damage the fuel cell structure, a novel interference-free diagnostic system based on the printed circuit board is proposed in this study. Fuel cell localised electrochemical impedance spectroscopy at different current density is observed. It is found that the activation impedance near inlet decreases sharply when current density increase. In addition, it is also found the flooding problem and the mass transport problem can occur at a medium current density due to the non-uniform behaviour of the polymer electrolyte membrane fuel cells. The proposed diagnostic system is demonstrated to be an effective tool to improve efficiency and robust of polymer electrolyte membrane fuel cells.

Introduction

Polymer electrolyte membrane fuel cells (PEMFCs) are promising power sources for vehicles and various portable electronic applications due to their high efficiency (over 50%) and low pollution emission in operation [1]. As reactions in a fuel cell are complex and various problems often happened during operation [2], it is necessary to develop some diagnostic methods to find the problems in time and remove or avoid them [3]. Among all the diagnostic methods, the electrochemical impedance spectroscopy (EIS) is proved to be an effective in-site and off-site way to monitor the operation state of a PEMFC.

However, traditional EIS could only illustrate information for the whole cell [4]. As a PEMFC often has a large active area, the electrochemical reactions in different regions may be different [5]. For example, Amir Amirfazli et al. [6] investigated the manifold geometry effect on stack temperature uniformity by a modelling work. It was found that the temperature distribution tended to be non-uniform while the stack became bigger. Chi-Young Jung et al. [7] also studied the temperature distribution of fuel cell with three different Nafion membranes. Their result showed that the temperature distribution tends to be more uneven when the exchange membrane is thicker.

Researchers also developed many diagnostic methods to monitor the fuel cell spatially. One of the most useful methods is the current density distribution measurement [8], [9]. The measurement can illustrate the unevenness of current density, which means the reaction in different parts is quite different [9].

But the current density distribution only shows the performance decrease (for example, the current density drops at the same voltage), so the reason for the performance decrease remains uncertain. The causes of performance decrease may be flooding [10], drying [11], or starvation [12]. That means, only with the current density distribution data, it is difficult for researchers to optimise the operating parameters to avoid these problems.

The localised electrochemical impedance spectroscopy (Local EIS) could be a solution to this issue [4]. Earlier in 2003, Brett et al. [13] has developed a local EIS system. They applied this technology in a single-channel solid oxide fuel cell. The characteristics of impedance changes in the direction of the flow channel have been observed. However, in their work, the cell was separated entirely, and all the 10 segments have an independent electric load. The system is equal to 10 individual cells arranged side by side, which is far different from the actual cell situation.

After decades of developments, multiple Local EIS solutions have been proposed. Schneider et al. [14] developed the system with segmented graphite in 2005. Gerteisen et al. [15], [16] developed a 50-channel system in 2011, and the system consisted of graphite and spring contacts. Roduner et al. [17], [18] performed a Local EIS test for different Nafion membranes in 2012 and their system was consisted of a point-like electrode made by Pt-coated tip. Hinds et al. [19] reported a new Local EIS system in 2015. In that system, a Nafion tube salt bridge was located in the segment, and the EIS information was achieved via the tube. In their recently published work [20], the Nafion tube system was applied in a 50 cm2 single cell.

Reshetenko TV et al. [21] studied the effect of Gas Diffusion Layer (GDL) on PEMFC performance with current density distribution and Local EIS technology. They exchanged the standard cathode GDL (25BC) with 25BA at segment 4 to introduce a GDL defect. They proved that current density distribution and Local EIS allow to detect and localise the GDL defects.

However, all the solutions above suffered from a common problem: their systems were far away from the actual fuel cell systems. For instance, in Hinds et al.’s work [20], the fuel cell contained a graphite plate with nine holes, which changed the structure of the cell. As the graphite is hydrophilic, the water produced during the reaction may gather near the holes. That means the conclusions in this work might be still away from the real application.

The Printed Circuit Board (PCB) has applied to measure the fuel cell since 1998 [22]. And it has been proved as a less disturbing system for fuel cells [23]. And Schulze et al. [24] even tried to implement the Local EIS measurement using the PCB technology. However, in that pioneering work, they could not obtain accurate diagnostic results due to the immature testing system and the use of too large embedded electrodes. The effective and accurate interference-free diagnostic system is still lacking, which is extremely important to ensure efficient and reliable operation of fuel cells for practical applications.

To fill this research gap, a novel interference-free diagnostic system based on local EIS technique was designed and evaluated. The system contained a specially designed PCB. However, the impedance of the embedded electrode was reduced [24], and the Local EIS result has become more accurate. As the defect of GDL structure has a great influence on the performance of fuel cells [25], an experiment with 2 home-made Membrane Electrode Assemblies (MEA) was performed to determine whether the Gas Diffusion Layer (GDL) should be segmented or not. After the experiment, the Local EIS system was finally designed, and the GDL is completely preserved. Finally, the Local EIS of a fuel cell with commercial MEA at different current density has been investigated. The result shows that the Local EIS system could observe the EIS of different segments and thus diagnosing possible problems in each segment. This novel system enables reliable and accurate diagnosis of fuel cells in practical operation.

Section snippets

Local EIS system

The new system is developed based on the segmented current density distribution monitor system [26]. The working electrode and the working electrode sensor are specially designed and are embedded in the PCB. The schematic of the system is illustrated in Fig. 1. Two segments whose Local EIS can be measured (EIS Seg 1 and EIS Seg 2) are presented in Fig. 1. In fact, the cell has 49 (7 × 7) segments and each of them has the local current monitor system as was described before [26]. Local EIS can

Comparison between MEA 1 and MEA 2

As could be observed in Fig. 5, the performance of MEA 1 and MEA 2 was quite different. MEA 1 (The red curves in Fig. 5) maintained higher voltage than that of the MEA 2 (The green curves in Fig. 5) at the same current density. The highest current density of MEA 1 was 864 mA·cm−2. However, the highest current density of MEA 2 was only 637 mA·cm−2, which was only 73.7% of that of MEA 1. The active area without GDL only accounts for 1.36% of the total active area. It could be concluded that if

Conclusions

In this work, a Printed Circuit Board-based localised electrochemical impedance spectroscopy test system was designed. The feature of the new system is interference-free to the structure of a fuel cell. During the localised electrochemical impedance spectroscopy test, a direct current electronic load is applied to the tested segments to keep the working condition of the fuel cell the same as the condition before the test.

Two kinds of membrane electrode assemblies were placed in the system and

Acknowledgement

The authors gratefully acknowledge financial support by the National Natural Science Foundation of China (No. 21776222) and Key Technology Research and Development Program of China (No. 2017YFB0102803).

References (33)

Cited by (20)

  • Study of internal performance of commercial-size fuel cell stack with 3D multi-physical model and high resolution current mapping

    2022, Applied Energy
    Citation Excerpt :

    T. Miao et al. investigated the local current density and temperature of a stack with 108 cm2 active area for different cathode flow fields [21]. R. Lin et al. studied the fuel cell voltage reversal process and localized electrochemical impedance spectroscopy with a 7 × 7 PCB based segmented cell device [22,23]. Although numerous experimental approaches have been conducted for fuel cell internal parameter distributions, there remains two main challenges in terms of the resolution and precision for the measurement.

  • Lattice Boltzmann method modeling and experimental study on liquid water characteristics in the gas diffusion layer of proton exchange membrane fuel cells

    2022, International Journal of Hydrogen Energy
    Citation Excerpt :

    As shown in Fig. 11(b and c), a typical flooding feature --- “inductive loop” [9,10] is formed below the real axis. The fuel cell with a wider channel and rib gain a higher low-frequency arc, which made the mid-frequency arc invisible [11,65]. This indicated that wider channel and rib block the liquid water pathway and cause water accumulation in porous layers, including CL, MPL and GDL.

  • Investigation of real-time changes and recovery of proton exchange membrane fuel cell in voltage reversal

    2021, Energy Conversion and Management
    Citation Excerpt :

    The 7x7 segments and reactants at both inlet and outlet were shown in Fig. 1(b). Local EIS was performed based on the monitoring system of segmented current density distribution [31]. Gamry reference 3000 was also used during Local EIS tests.

View all citing articles on Scopus
View full text