PSU Neutron Beam Group (NBG)

Study of the nature of Flooding and Drying in Polymer Electrolyte Fuel Cells

Participants:	M. M. Mench, Assistant Prof. of Mechanical and Nuclear Engineering
	J. Brenizer, Prof. of Mechanical and Nuclear Engineering

	Kenan Ünlü, Professor of Mechanical and Nuclear Engineering
	K. Heller, PhD student at Mechanical and Nuclear Engineering

	A. Turhan, PhD student at Mechanical and Nuclear Engineering
	J. J. Kowal, PhD student at Mechanical and Nuclear Engineering

Services Provided:	Neutron Beam Laboratory

Sponsors:	General Motors Corporation
	Radiation Science and Engineering Center

Introduction

Due to its high efficiency, low operating temperature (~30-80 o C), and rapid evolution since over the past decade, the polymer electrolyte fuel cell (PEFC) is currently under intense research and development. Compared to present power systems, such as the internal-combustion engine, fuel cells are advantageous for several reasons and present a promising future. The operating efficiencies can reach as high as 50-90% for units and also pollutants such as nitrous oxides and particulate matter are eliminated, while carbon dioxide and carbon monoxide are reduced to near zero.

Figure 1 shows the basic operation of a hydrogen PEFC. Hydrogen is supplied to the anode of the fuel cell while oxygen, usually taken from the air, is supplied to the cathode. The electrochemical oxidation reaction at the anode produces hydrogen ions and electrons.

Figure 1. Simplified Schematic of PEFC

The flow of the electrons through an external circuit powers a load. The ions produced at the anode are transported through an ionically conductive polymer electrolyte to the cathode, where water is produced. A ~10-20 m m platinum catalyst layer is typically employed at both electrodes to reduce the activation energy for the electrochemical reactions. Covering each electrode is a 200-400 m m porous carbon fiber gas diffusion layer (GDL). The GDL functions to enable reactant transport to, and product from the catalyst layer, while providing conductivity for electron transport.

In a PEFC, the level of water must be precisely balanced. Adequate water vapor must be available maintain high electrolyte ionic conductivity and ensure suitable performance. However, if excessive water is present in the liquid phase, it can block pores in the catalyst and GDLs, hindering the transport of reactants to the catalyst. This phenomenon is known as “flooding”, and greatly diminishes cell performance. Due to the delicate balance between the benefit of saturated flow and the deleterious effects of flooding concomitant with liquid water accumulation, there is extensive ongoing research to more fundamentally understand two-phase water transport in PEFCs, to enable performance and design optimization.

Although there have been numerous models presented in literature that predict the water production and transport phenomenon is fuel cells, there has been little research in experimental visualization and quantification of the liquid water distribution and transport. Neutron radiography and radioscopy are excellent non-intrusive techniques for visualization and quantification of the two-phase flow within the fuel cell in real time or steady-state.

Experimental Setup

The neutron radioscopy system and thermal neutron beam from the Breazeale Nuclear Reactor at the Penn State Radiation Science and Engineering Center was utilized in this study. Specialized image processing hardware was developed for the analysis, storage and presentation of the collected images.

An integrated test station at the Neutron Beam Lab (NBL) was built to control and monitor the fuel cell operating parameters. The NBL Test Station (NBLTS) is isolated from the neutron beam source, as illustrated in Figure 2. The station can accommodate various sized fuel cells (up to 22.8 cm diameter in a single frame) for neutron imaging processes while the following conditions are controlled by the operator on the station's control panel:

• Gas flow rates

• Inlet gas temperature and humidity

• Cell temperature

• Current/Voltage draw

• Operating pressure

• Nitrogen purge

The visualization and quantification of the water distribution in the fuel cell is performed by neutron radioscopy and radiography techniques.

Figure 2. Test station for fuel cell imaging at the NBL

Neutron Radiography Results[1,2]

A series of neutron images were collected using the fuel cell and imaging setup . The cell temperature (80°C) and gas flow back pressure (7.35 psig) were maintained constant throughout the experiments whereas the relative humidity of the anode and cathode were maintained at 100% at 80 o C for all t ests . Figure 3 shows images of two conditions with the same current density but different cell temperature. With the same water production (same current) and relatively similar water mass in the diffusion media, the fuel cell with couple of degrees lower temperature has severe flooding losses. The cell performance curve for each condition is shown in Figure 4. The polarization region for low temperature cell is clearly seen from this figure. Therefore, it is seen that the flooding behavior in the fuel cell is highly dependent on water distribution in the flow channels and diffusion media. Furthermore, the effective porosity change for both conditions was calculated and it was found that the flooding loss is due to a small difference between the porosity values.

Figure 3. Neutron Radiography Image Comparisons for Conditions of Flooding (a) and Not Flooding (b) for 1.0 mm by 1.0 mm Channel/Land Fuel Cell.

Figure 4. Polarization Curves for Conditions of Flooding and Not Flooding for 1.0 mm by 1.0 mm Channel/Land Fuel Cell.

Conclusions

Neutron radiography and radioscopy yield excellent spatial and temporal resolution for the investigation of water transport phenomenon and the measurement of liquid water inside an operating polymer electrolyte fuel cell. Results indicate that, for high current density operation with the catalyst and diffusion media utilized, flooding is a highly localized phenomenon controlled by a small volume of liquid water in the DM. The change in effective porosity is lower than predicted by a Bruggeman segregated phase filled pore model, and greater than that predicted by a thin film transport resistance model. Thus, a new combined film resistance, pore filling physical model of flooding is needed to explain the experimentally observed results.