Facilitating the proper handling of water is one of the main challenges to overcome when trying to improve fuel cell performance. Specifically, enhanced removal of liquid water from the porous gas diffusion layers (GDLs) holds a lot of potential, but has proven to be non-trivial. A main contributor to this removal process is the gaseous transport of water following evaporation inside the GDL or catalyst layer domain. Vapor transport is desired over liquid removal, as the liquid water takes up pore space otherwise available for reactant gas supply to the catalytically active sites and opens up the possibility to remove the waste heat of the cell by evaporative cooling concepts. To better understand evaporative water removal from fuel cells and facilitate the evaporative cooling concept developed at the Paul Scherrer Institute, the effect of gas speed (0.5–10 m/s), temperature (30–60 °C), and evaporation domain (0.8–10 mm) on the evaporation rate of water from a GDL (TGP-H-120, 10 wt% PTFE) has been investigated using an ex situ approach, combined with X-ray tomographic microscopy. An along-the-channel model showed good agreement with the measured values and was used to extrapolate the differential approach to larger domains and to investigate parameter variations that were not covered experimentally.

State-of-the-art liquid cooling of polymer electrolyte fuel cells (PEFC) requires a complex multi-layer design of bipolar plates [1] which is responsible for about 75 % of stack volume and approximately 80 % of its mass [2]. Additionally, an external humidifier is often required to ensure a proper humidification and thus ionic conductivity of the proton exchange membrane [3] at elevated temperatures.Evaporative cooling shows the potential to reduce the stack volume, mass, complexity and cost by up to 30 % by simplifying the design of bipolar plates. Additionally, it allows a better internal humidification which enables higher operating temperatures without external humidification [4]. In our concept, liquid water is fed to each cell through dedicated water channels in the anode flow-field. Subsequently, the water is distributed in a specially designed gas diffusion layer with a mixed hydrophilic and hydrophobic pattern (Figure 1) [5]. Once in contact with the gas flow, the water in the hydrophilic lines evaporates, cools the cell and is eventually released as vapor with the exhaust gases.In our previous work [6] we have elaborated the potentials and limits of this concept on the stack and system level. In this study, we focus on the experimental and numerical investigation of water evaporation from GDLs with patterned wettability [7] in a technical cell environment. Ex-situ evaporation experiments have been carried out with a patterned Toray 060 GDL (7 x 3 cm2, 14 hydrophilic lines, pattern: 500 µm hydrophilic, 940 µm hydrophobic) and nitrogen as a carrier gas. The operating temperature (22 to 90 °C) as well as the gas velocity (0.2 to 7 m/s) has been varied whereas the pressure is kept constant (ambient at cell outlet).In order to further foster the understanding of the underlying evaporation phenomena, a simplistic 2-D water vapor transport model has been developed in this study. Fickian diffusion is taken into account in the GDL and gas channel, while convection is only considered in the channel. Water evaporation is modeled according to the Hertz-Knudsen-Schrage equation [8]. The projected contact area between hydrophilic lines and gas channel is an input parameter to the model and has been determined by ex-situ X-Ray tomographic imaging (XTM).Figure 3 shows that the evaporation rate and thus the related cooling power increases significantly with temperature. This fact is attributed to the higher water saturation pressure as well as to the increased diffusivity at higher temperature. Further, it can be seen that the evaporation rate increases with gas speed. We attribute the strong increases at low gas speed to the fact that the evaporation is only limited by the saturation of the gas channels in this regime. At higher gas speeds, however, the increase in evaporation rate is less pronounced. We hypothesize that the water vapor transport through the GDL is limiting the evaporation rate. Evaporation kinetics have not been a limiting factor in the investigated parameter space, since the interface between hydro...