Operation of proton exchange membrane fuel cells (PEMFC) without external humidification of the reactant gases is advantageous for the PEMFC system, because it eliminates the need of a gas‐humidification subsystem. The gas‐humidification subsystem is a burden in the fuel cell system with respect to weight, complexity, cost, and parasitic power. A model for the operation of PEMFC with internal humidification of the gases is presented and the range of operating conditions for a PEMFC using dry /air was investigated. The model predicts that dry air, entering at the cathode, can be fully internally humidified by the water produced by the electrochemical reaction at temperatures up to 70°C. This model was experimentally verified for cell temperatures up to 60°C by long‐term operation of a PEMFC with dry gases for up to 1800 h. The current densities, obtained at 0.6 V, were 20 to 40% lower than those measured when both gases were humidified. The water distribution in the cell, while operating with dry gases, was investigated by measuring the amount of product water on the anode and cathode sides. It was found that the back‐diffusion of product water to the anode is the dominant process for water management in the cell over a wide range of operating conditions. The dominating water back‐diffusion also allows internal humidification of the hydrogen reactant and prevents drying out of the anode.

In the first paper of this series the bulk and surface structural properties of Ti-fiber based porous transport layers (PTL) were characterized and described. In this second part the correlation of structure to the performance in polymer electrolyte water electrolysis cells is analyzed by determination of the three main overpotentials of ohmic, kinetics and mass transport losses. The strongest correlation between the PTL bulk transport properties and cell performance is obtained for heat transport. The current density dependent temperature gradients show good agreement with ex situ determined heat conductivity from part one. However, surface properties of the PTL materials, have a stronger influence on cell performance than the bulk properties. Catalyst layer utilization and ohmic interfacial resistances correlate with the interfacial contact areas reported in part one and performance increases with increasing contact area. This is due to a local mass transport resistance decreasing with increasing catalyst layer utilization.

The in situ resistance of Nafion membranes with different thickness was measured in one-dimensional fuel cells as a function of current density. Except for the thin Nafion 112 membrane, an increase of the ionic resistance with current density (in the range 0 to 1 A/cm²) was found. The thicker the membrane, the stronger the increase in the same current density interval. The resistance distribution across the thickness of membranes was determined by using membranes composed from several thin sheets with interlying thin gold wires as potential probes. It was found that the increase of the resistance is always confined to the membrane sheet contacting the anode electrode. These measurements, combined with the results from experiments with membranes of different water content, lead to the conclusion that the resistance increase at the anode side is due to the insufficient compensation of the electro-osmotic drag by the back transport of water to the anode. Based on a solution diffusion mechanism of the water motion in the membrane, the experimental results may be explained by a mechanism whereby the electro-osmotic drag coefficient is independent of the local membrane hydration and the water diffusion coefficient <mml:math display="inline"><mml:msub><mml:mi>D</mml:mi><mml:mrow><mml:msub><mml:mo>H</mml:mo><mml:mrow><mml:mn>2</mml:mn></mml:mrow></mml:msub><mml:mo>O</mml:mo></mml:mrow></mml:msub><mml:mo>,</mml:mo></mml:math>DH2O, is a strong function of the local membrane water content. The experimental data would, qualitatively, also be in line with a model proposing back transport of water to the anode by convection of water in the submicropores of the membrane. © 2001 The Electrochemical Society. All rights reserved.

In the first paper of this series, the bulk, surface and transport properties of porous transport layer materials (PTL) for polymer electrolyte water electrolysis cells (PEWE) are characterized. A systematic PTL matrix of Ti fiber materials with 3 fiber diameters and 2 nominal porosities as well as a state of the art sintered powder material are investigated to get a better understanding of the governing parameters in electrolysis cells. X-ray tomographic microscopy analysis of ex situ PTL structures and post operando membrane electrode assemblies is performed. On the tomographic structures, bulk (porosity, pore/solid size distributions, fiber orientation), mass transport (diffusivity, permeability, conductivities) and surface parameters (roughness, membrane deformation, PTL surface area and interfacial contact area) are determined. The second paper of this series will correlate the structural parameters with in-depth electrochemical analysis. The new insights into the effect of PTL properties on PEWE performance allow to isolate governing key parameters. From know-how obtained on the fiber based materials fundamental design guidelines for optimization of PTL structures are deduced.