The Fuel Cell R&D Activities at PSI, a collaborative effort of electrochemists, polymer chemists, physicists, material scientists, and mechanical engineers, was established in 1990. Since then, our polymer electrolyte fuel cell activities have grown tremendously, due to collaborations with industry as well as with other research organizations on a national and international scale.
Work in the area of the Polymer Electrolyte Fuel Cells (PEFCs) is carried out in the two research groups Fuel Cells and Fuel Cell Systems.
Materials Research Aspects (proton-conducting membranes, electrocatalysts) and Spatially Resolved in situ Diagnostic Methods (transient and imaging techniques) for PEFCs are the research domains of the Fuel Cells Group. Combining our know how of these various research aspects allows us to develop a comprehensive understanding of the polymer electrolyte fuel cell on all relevant length scales of single cells.
The Fuel Cell group concentrates on Material and Structural Research Aspects of PEFC in the research domains
“Our aim is to promote the application of transient techniques in polymer electrolyte fuel cell diagnostics for the identification of inhomogeneities and limiting/harmful processes on a local scale”
In a combined experimental and modeling effort, we are developing/using spatially resolved characterization methods for PEFCs that are based on ac impedance spectroscopy, cyclic voltammetry and step response. In addition, we are using steady state imaging techniques based on neutron radiography.
Our current experimental and modeling work focuses on:
Liquid water detection in PEFCs
In situ measurement of local parameters during PEFC operation
Local transient response of PEFCs
Fundamental aspects of the impedance response of PEFCs
Characterization of materials using ac techniques
LIQUID WATER DETECTION IN PEFCs
Through plane neutron radiography (neutron beam perpendicular to the membrane plane) can resolve the distribution of liquid water over the fuel cell area. This technique has been applied for several years at PSI. To resolve the distribution of water across the fuel cell structure, in plane imaging has to be used, which implies very high requirements on resolution. Recent improvements of the imaging method, pixel resolution of a low as 2.5 μm could be reached.
Figure 1:Through plane neutron radiography can resolve the distribution of liquid water over the fuel cell area. In plane imaging allows us to resolve the distribution of water across the fuel cell structure.
IN SITU MEASUREMENT OF LOCAL PARAMETERS
liquid water content
current density
impedance
catalyst utilization
Local current density (CD) measurements on different length scales allow the identification of inhomogeneities in local cell performance. However, they do not allow the identification of the underlying limiting or even harmful processes. Therefore, we have developed novel spatially resolved in situ methods that are based on AC Impedance (EIS) and Cyclic Voltammetry (CV). These spatially resolved techniques are used along with neutron radiography (NR) to observe and to quantify the distribution of liquid water inside an operating fuel cell in a completely non-invasive way. This combination of methods provides unique insights into the impact of water on the local properties of the cell. Figure 2 shows a sketch of the experimental setup for simultaneous NR/EIS/CD measurements in PEFCs on a local scale and experimental results obtained with a linear 200cm2 H2/air PEFC.
Figure 2: Experimental setup for spatially resolved measurements in PEFCs (top) and experimental results obtained with a linear H2/air PEFC (bottom).
LOCAL TRANSIENT RESPONSE OF PEFCs
Transients are changes in operating state, e.g. rapid increase in current density when power demand increases instantly. Associated with transients are possible degradation effects e.g., reactant starvation (fuel or oxidant), cell reversal and formation of local ‘hot-spots’, leading to electrode and / or membrane degradation. Important factors to consider in this context are rate of change and spatial dimension: How does cell response change with the transient magnitude and slope ? Over which length scale do effects occur, along the flow field or even locally on a sub-mm scale ? Which potentials do occur at anode and cathode ? Within the fuel cells group models to describe the local transient response of PEFCs are being developed. In parallel, a portfolio of novel methods is being developed and used for the characterization of parameters such as local current density, local electrode potential or the local membrane resistance on all relevant length and time scales. Recent progress allows spatially resolved measurement of the local membrane resistance with millisecond resolution (Figure 3).
Figure 3: Spatially resolved transient response of a linear ninefold segmented H2/O2-PEFC to a current step under sub saturated conditions (local current response not shown here).
AC RESPONSE OF PEFCs - FUNDAMENTAL ASPECTS
In the majority of cases, the underlying physical processes of both equivalent circuit models or even continuum mechanistic models used to describe PEFC impedance response are either based on assumption or on experiments which corroborate an expected trend in the spectrum, e.g., when changing operating conditions. However, fuel cells are complex three dimensional electrochemical systems (Figure 4), and as a consequence, this kind of approach provides some indication, yet it is not necessarily sufficient to prove the validity of a theory. As a consequence, modeling results are not unique and can lead to ambiguous or even erroneous explanations of the underlying physical processes determining the fuel cell impedance response. Key experiments are needed to clearly attribute features observed in PEFC impedance spectra to the respective physical processes occurring in the cell. Our strategy is to get insight by resolving
anode and cathode contributions
spatial inhomogeneities along and perpendicular to the channels
across MEA and “in plane” contributions along and perpendicular to the channels
The continuum mechanistic models for PEFC ac impedance response developed within the fuel cells group are backed-up by the results of these key experiments.
A novel pseudo reference electrode arrangement for PEFCs allows the separation of anode and cathode contributions to the overall impedance response of a PEFC. This pseudo reference electrode is used in so called one-dimensional cells (Figure 5). A fixed potential micro-reference electrode is under development for potential measurements on all relevant length scales of single cells.
Figure 5: Separation of anode and cathode contributions to the overall impedance response of a PEFC at different cell currents.
Locally resolved EIS makes use of segmented cells and allows the identification of spatial inhomoge-neities in the impedance response of a PEFC along the flow field. The novel idea to combine local ac measurements in PEFCs with sectioned electrodes enables us to separate “across the MEA” and “in plane” contributions along the flow field (Figure 6). By using this technique, we were able to show that reactant depletion and water formation in upstream parts of the cell can significantly contribute to the formation of low frequency capacitive (Figure 6) and low frequency inductive loops in PEFC impedance spectra.
Figure 6: Locally resolved EIS is used in combination with sectioned electrodes to separate “across the MEA” and “in plane” contributions.
MATERIAL PROPERTIES
Novel ac techniques are being developed to characterize membrane materials or porous media, e.g. Gas Diffusion Layer (GDL) materials ex situ and during operation. Interdigitated electrode sensors have been developed within a European Union project to characterize membrane materials (Figure 7).
Figure 7: Interdigitated electrode sensors are used for the characterization of membrane materials.
ELECTROCATALYSIS
MISSION
We focus on preparation and characterization of catalysts for polymer electrolyte fuel cells (PEFCs).
Our aim is to develop Pt-based catalysts with high mass activity and sufficient stability for automotive applications.
Our experimental methods include:
Cyclic Voltammetry
Rotating ring disc electrode measurements
Fuel Cell tests
Impedance measurements
Scanning electron microscopy
Transmission electron microscopy
X-ray Photoelecton spectroscopy
X-ray Diffractometry
X-ray Absorption Spectroscopy
Magnetron sputtering
Chemical nanoparticle preparation
Electrocatalysts for polymer electrolyte fuel cells
Low-temperature fuel cells are in need of improved electrocatalysts for both the anode and the cathode reactions. Since most of the electrochemical performance losses can be attributed to the sluggish kinetics of the oxygen reduction reaction, one of our goals in the pursue for advanced electrocatalysts is the development of oxygen reduction catalysts with reduced noble metal content and good stability.
Figure 1: Scheme of a fuel cell cathode
Catalyst preparation by co-sputtering
One approach to prepare low Pt fuel cell electrodes is
The sputter deposition technique. In this way it is
possible to prepare thin catalyst layers, well attached
to the polymer electrolyte and therefore with a high
catalyst utilization.
Platinum and carbon are co-sputtered as a composite
Material onto gas diffusion media or Nafion.
This one-step procedure also simplifies the preparation of fuel cell electrodes.
Figure 2: Preparation of fuel cell electrodes by magnetron co-sputtering of platinum and cobalt.
Co-sputtered Pt/C catalyst layers are about one order of magnitude thinner than commercial catalyst layers, prepared from Pt/C-ionomer inks. Various sputter substrates are possible like carbon cloth, carbon fiber or Nafion.
Figure 3: Co-sputtered 400nm thick Pt/C film on a Nafion membrane.
Fuel cell tests with co-sputtered anodes show similar performance compared to commercial electrodes. No ionomer impregnation is needed for co-sputtered electrodes.
Figure 4: Fuel cell test with co-sputtered anodes and commercial cathodes. T=80°C
Activity and stability of platinum alloy catalyst for oxygen reduction
We could demonstrate that commercial Pt-alloy catalysts show up to 3 times higher activities for oxygen reduction compared to pure platinum.
Figure 5: Mass activity for oxygen reduction of Pt and Pt alloy catalysts measured with the rotating disc electrode, T=60°C
To increase the mass activity further we aim to develop stable core-shell catalysts.
Figure 6: TEM image of a PtCo/C catalyst
Mikro Fuel Cells
Micro fuel cells are of interest due to their relatively high energy density compared to batteries.
We work on a simplified concept which uses micro-structured flow fields, so a gas diffusion layer can be set aside and the number of components can be significantly reduced. Figure 7 shows the concept of the cell as well as a current-voltage curve.
Figure 7: Concept of the micro-fuel cell and respective current-voltage curve.
RADIATION GRAFTED FUEL CELL MEMBRANES
MISSION
“Major factors limiting PEFC commercialization are cost and durability. We need cost-effective and more durable and reliable fuel cell membranes and membrane electrode assemblies (MEAs) to attain this goal.”
The PSI Membrane is an innovative fuel cell membrane prepared via the radiation grafting technique. The membranes are designed to yield optimum performance and durability in the PEFC. Membrane synthesis, characterization, and evaluation in the fuel cell are our main focus of our work.
Current topics are:
developing new generations of radiation grafted membranes via smart combinations of base polymer, monomers, crosslinkers, and process steps that allow better control over the membrane architecture
establishing correlations between membrane composition / structure and properties / fuel cell performance
understanding limitations in lifetime by studying membrane aging, in application-near and clever model experiments
Radiation Grafted Fuel Cell Membranes
Radiation grafting offers a versatile method to introduce a functional property, in this case, proton conductivity, into a preformed base polymer film. In the process, the desired component is grafted onto activated centers in the base polymer through radical polymerization, thereby forming a graft copolymer.
Advantages:
cost-effective polymer functionalization
no film formation process step requiredflexible choice of base polymer and monomers
Figure 1: Radiation grafting is a polymer film modification technique, whereby a disired functionality is introduced into the base polymer by graft copolymerization
Figure 2: The base polymer, such as FEP, ETFE, or PVDF film (typical thickness 25 micron) is activated by irradiation. In the grafting step, a second component (e.g.. polystyrene) is introduced into the polymer. The graft copolymer is sulfonated to introduce ion exchange sites for the proton conduction functionality in the fuel cell.
Influence of Crosslinking
Crosslinkers are monomers with more than one vinyl unit, such as dvinylbenzene (DVB). Introducing crosslinkers into the grafting solution results in the formation of a crosslinked graft co-polymer structure.
Crosslinked radiation grafted membranes have a higher chemical stability, yet excessive crosslinking reduces proton conductivity notably and yields brittle membranes.
The crosslinker content in the grafting solution has to be carefully balanced. Using styrene and DVB monomer mixtures in ETFE (25 micron) grafted membranes, optimum performance was found at 5 % DVB (Figure 3).
Figure 3: Influence of the dininylbenzene (DVB, = crosslinker) content in the grafting solution with respect to styrene on the resulting fuel cell performance. The optimum is around 5 % DVB, lower extents of crosslinking result in a poor membrane-electrode interface, higher crosslinked films exhibit low proton conductivity.
Direct Methanol Fuel Cell (DMFC)
Radiation grafted membranes based on 75 micron FEP film were optimized for application in the DMFC. The key advantage over perfluorinated membranes, such as Nafion 117, is the possibility of adjusting the extent of crosslinking. As a consequence, the methanol crossover to the cathode can be reduced by a factor of 2 without taking a hit in performance (Figure 4). In addition, the water permeation is reduced: with our optimized radiation grafted membrane, an electroosmotic drag coefficient of 1.7 is obtained, compared to 5.5 for Nafion 117.
Figure 4: Performance in the direct methanol fuel cell and methanol crossover. The suitable crosslinked radiation grafted membrane based on 75 micron FEP film shows the same performance as a standard Nafion 117 membrane at considerably lower methanol crossover, yielding considerably improved fuel efficiency.
Durability
A key requirement for application of a membrane in the PEFC is durability. Our 25 micron based radiation grafted membrane showed stable operation under stationary conditions (0.5 A/cm2) for over 4’000 hours (Figure 5).
There is, however, still room for further improvements in lifetime, particularly under application-relevant conditions involving load changes, excursions to OCV, transients, and start-stop cycles.
Figure 5: Polarization curves at different lifetimes of a cell comprising optimized radiation grafted membrane based on 25 micron FEP and a graft level of 18 %. H2 / O2, temperature: 80 °C.
Advanced Monomers
A promising approach to improve lifetime of radiation grafted membranes is the use of monomers with higher intrinsic chemical stability. Alpha-methylstyrene (AMS) is a promising candidate in this respect, because its a-position is less sensitive to oxidative degradation.
Second generation PSI Membranes based on AMS show a higher chemical stability compared to styrene based membranes (Figure 6). Particularly promising stability data is observed for crosslinked membranes.
Figure 6: Rate of ion exchange capacity (IEC) loss as a measure for chemical membrane stability for two monomer systems based on styrene and a-methylstyrene, respectively. Base polymer is 25 micron FEP. For crosslinked membranes, divinylbenzene was used as crosslinker.
Post Mortem Analysis
Understanding degradation and how it depends on the local operating conditions requires us to analyze the extent of membrane degradation on a local scale. In the fuel cell flow field, areas associated with gas channels and current collection lands can be discerned. We found that membrane degradation is not uniform on the channel-land scale (Figure 7).
Characterizing the extent of degradation as a function of (local) operating conditions of the membrane electrode assembly is subject of current investigations.
Figure 7: Post mortem degradation analysis on the scale of flow field channels and lands by means of transmission FTIR spectroscopy of an artificially aged radiation grafted membrane based on 25 micron FEP.
Model Compounds
Understanding polymer aging in the context of a fuel cell membrane and single cell operation is a complex subject matter. The studying of fundamental degradation processes in simplified and well-controlled model systems (Figure 8) helps to identify mechanisms and reactions, and provides insights for finding monomers with higher stability.
Figure 8: Model systems for studying the chemical stability of poly(styrene sulfonic acid) derived compounds and building blocks with different substituents in a simulated fuel cell environment, i.e. in the presence of hydroxy- and hydroperoxy-ions.
