Skip to main content
European Commission logo print header

Predictive Modelling for Innovative Unit Management and Accelerated Testing Procedures of PEFC

Final Report Summary - PREMIUM ACT (Predictive Modelling for Innovative Unit Management and Accelerated Testing Procedures of PEFC)

Executive Summary:
A general objective is to contribute to the improvement of stationary fuel cell systems durability, keeping in mind that the target required is 40000h. In this sense, the success of the project should help to overcome one of the main bottlenecks for European providers of stationary fuel cell systems and will contribute to cross cutting issues relevant for European R&D and fuel cell industry development. Thus, reliable systems corresponding to the technical specifications of the energy global market could be widespread, which will change the end-user habits towards the stationary energy management and will help to reduce greenhouse gas emission.
Premium Act specific objectives are to propose a reliable method to predict lifetime, based on validated accelerated degradation tests, to benchmark components and to improve operating strategies of real systems. The project addresses two types of technology: Direct Methanol (DMFC) and Proton Exchange Membrane Fuel Cells (PEMFC) operating with reformate hydrogen.
As far as the degradation understanding is concern, focus of Premium Act is on the core of the fuel cells: the Membrane Electrode Assemblies (MEA). The structure of the project is based on the following scheme: (i) initial information is requested from the real systems developed by the industry partners in order to (ii) define the reference conditions to which (iii) the MEA have to be submitted in controlled devices, namely short stacks and single cells, to (iv) check the nominal loss of the fuel cell performance and the degradation of components. Thereby, detailed in-situ and ex-situ analyses are conducted to characterize the degradation. In-situ tests are providing information about reversible or permanent performance loss as well as modifications, in electrochemical behaviour and properties, during ageing. Ex-situ methods are used to assess chemical, physical, and structural properties of the components with lateral and vertical resolution allowing a correlation of local degradation with heterogeneous conditions and operation in the stacks and cells. Segmented cell devices are also developed and used to directly relate local operating conditions to local performance. In order to enhance interpretation of non-homogeneous operation and understanding of local degradation phenomena, modelling is used mainly to enable better description and prediction of coupling phenomena. Models are developed from electrochemical local level for mainly degradation mechanisms description to more macroscopic cell level with mainly fluid transport and current distribution simulation.
The core idea of the project is to use the information coming from these degradation investigations to reach the final results of the project. Main differences compared to state of the art at the beginning of the project are related to the fuels used (methanol and polluted hydrogen) and to the anode catalyst containing both platinum and ruthenium, much less studied than pure hydrogen and pure platinum cases. New phenomena and mechanisms have been identified and deeply analysed for better understanding of the degradations inherent to these parameters, with regard to the fuel cell systems considered. As final outcomes, specific degradation mechanisms related to catalyst degradation with particularly new evidences concerning ruthenium dissolution and re-deposition, or to local conditions, for both technologies; to polymers losses and transport properties particularly for methanol, have been identified but also qualitatively and quantitatively related to operating parameters (temperature, fuel composition) or methods (stops, air break, load cycles).
Thanks to this better understanding, accelerating features but also stabilizing conditions have been proposed to enable developing and validating on one hand, operating strategies for improving voltage stability and hence lifetime and, on the other hand, relevant specific accelerated stress tests to qualify more rapidly MEAs versus degradation for DMFC or PEMFC operating with reformate fuel cell systems.

Project Context and Objectives:
Context and key Objectives of the project
A general objective is to contribute to the improvement of stationary PEFC systems durability, one of the main hurdles to overcome before successful market development, knowing that the target required is 40000h.
Premium Act specific objectives are to propose a reliable method to predict lifetime, to benchmark components and to improve operating strategies of real systems, in order to reach the following achievements: relative prediction of durability (the ranking of durability predicted between different MEAs is in agreement with ranking obtained in real conditions); Absolute prediction of durability (the method is successful in predicting the observed lifetime with reduced testing duration thanks to accelerated tests); Innovative unit management strategies (the strategies allow a measurable durability increase in the stacks and systems of the industrial partners without negative impact on the customer needs’ fulfilment).

Challenges/issues addressed
Main issue is the durability of fuel cell system for micro Combined Heat and Power applications.
It is necessary to first identify and understand the causes of fuel cells degradation in real conditions. Degradation is mainly related to the degradation of the Membrane Electrodes Assemblies (MEAs), core of the fuel cells where electrical power is produced. This degradation is double: decrease of the electrical power with time and degradation of the components. As multiple factors induce degradation (materials used for the catalysts or for the membrane, operating conditions such as cell temperature, gases composition), it is also necessary to understand the link between the different degradation mechanisms.
The challenges of Premium Act are related to the approach which is to combine specific experimental and modelling tools to study the degradation at different scales from fuel cell performance down to microstructure of materials.
Additional issue is the consideration of two strategic fuel cell technologies for stationary markets: DMFC (Direct Methanol Fuel cell) power generators and CHP systems fed by reformate hydrogen.

Technical approach/objectives
The technical approach has been built on technical tasks interconnected in the right way to complete the challenging objectives described above.
First step is to identify and understand the causes of degradation, particularly of Membrane Electrode Assemblies, in real conditions (with a focus on the accelerating features). These conditions are thus reproduced on small devices to estimate MEAs’ lifetime. Then, analyses are conducted on the components, to elucidate how their microstructure or their properties are degraded during fuel cell operation.
In parallel, multi-physics models are developed to enable the description of the phenomena appearing during the ageing at the different scales of the cell and for different conditions, considering as well the decrease of the electrochemical performance as the alteration of the MEA materials features (catalysts, electrodes and membranes); models validation being conducted through specific single cell tests.
The core technical part of the project will be to combine all the information coming from the experimental tests or analyses and from the modelling to propose and validate relevant accelerated tests able to couple various degradation factors and to assess different MEAs’ lifetime more rapidly than with normal tests. Final expected outcomes are operating strategies able to improve the lifetime of the systems considered and a methodology to predict the life time of their MEAs.

The project objectives have been declined for the two executive periods.

For the first period, general objective was to start with all technical activities planned in order to get enough information about parameters influencing fuel cell degradation at the beginning of the second period which will be dedicated to the final goals of the project, namely the development of operating strategies and of accelerated degradation tests as a basis of a lifetime prediction methodology.
All technical objectives listed below are of course related to the two technologies considered in the project, for the same type of micro-CHP application: Proton Exchange Membrane and Direct Methanol Fuel Cells.
Considering the organisation in specific work packages, their individual objectives were the following:
• For WP1 (Specifications): to define all components and protocols to be used at least for the first period: MEAs, cells and stacks, ageing conditions and current profiles, methods for in-situ and ex-situ analyses.
• For WP2 (Experiments under real operating conditions): to get information from field tests data and systems specifications to enable defining particular ageing conditions for smaller scale tests (single cells and stacks), to manufacture and provide reference MEA, to perform ageing tests following selected protocols in nominal representative conditions, or critical conditions to acquire experimental knowledge about the impact of components or operating parameters on performance or on materials degradation. The development and use of specific methods or tools, particularly of segmented cell for local in-situ measurements for better understanding was also part of the first period goals.
• For WP3 (Quantification of components degradation): selection, if necessary development or adaption, and application of characterization techniques to analyse first the non aged reference components, to evaluate when relevant the interest of the method, and then to start with the analysis of the aged samples in order to identify composition or microstructure modifications. One important issue was also to perform these analyses with a vertical and a horizontal resolution in order to relate degradation to local conditions to get more accurate information for better understanding of mechanisms and relevant support to degradation models development.
• For WP4 (Predictive Modelling), to improve existing models describing cells and MEA operation, implementing specific conditions as defined by the systems considered, to develop models for a local description of the performance, to develop degradation models describing specific mechanisms for the technologies considered, to start integrating degradation phenomena in the performance models, in order to perform first evaluation of the selected mechanisms impact on local performance evolution.
• For WP5 (Lifetime prediction methodology), to start sensitivity studies for the evaluation of operating parameters impact, when coupled, towards degradation, with the aim to get information about particularly acceleration of performance degradation for enabling first proposal of composite accelerated tests. This WP includes also modelling validation with specific tests but, as for accelerated tests, main work was planned for the second period.
• For WP6 (Dissemination), to communicate about project aims and advancement as general aspect but, as major purpose, to organise a workshop on the characterisation of fuel cells MEA degradation.
• For WP7 (Management), to conduct coordination actions related to contractual, financial, administrative and scientific management, including all exchanges with the FCH-JU, ending the period by the preparation of the mid-term review (incl. financial report, this first periodic report and the meeting with reviewers).

The two major technical goals of the second period were:
o The development of operating strategies to improve the lifetime
o The development of accelerated degradation tests as a basis of a lifetime prediction methodology
The general aim was to use the main outcomes of period one, particularly all information about parameters influencing fuel cell degradation, to develop the ageing protocols enabling to stabilize the performance or on the contrary to accelerate the performance losses as requested to achieve the final goals of the project.
All technical objectives listed below are of course related to the two technologies considered in the project, for the same type of micro-CHP application: Proton Exchange Membrane and Direct Methanol Fuel Cells.
• For WP1 (Specifications): no specific activities but some up-date to be considered for the definition of benchmark components used in the second part of the project (reported in WP2).
• For WP2 (Experiments under real operating conditions): to continue with systems data collection or development for further applicability of strategies; to manufacture and provide additional reference MEA or stack, but also benchmark MEA (mainly defined as presenting different behaviour from the reference); to perform ageing tests following selected protocols in representative conditions as during period one to get in-situ data and give samples for ex-situ analyses on reference samples, but also on the benchmark components, to enable their comparison; to propose and start validating operating strategies
• For WP3 (Quantification of components degradation): mainly application of the characterization techniques decided previously as the most relevant; to continue with local analyses to relate degradation to local operation conditions for samples aged following different protocols including accelerated tests; to further identify or confirm particular mechanisms mainly responsible for performance degradation and needed to be enhanced by accelerated test.
• For WP4 (Predictive Modelling), to develop additional description of mechanisms or of operational functions and to implement those in performance cell models to enable simulating performance evolution; to enable the interpretation of electrodes or cell behaviour in correlation with the mechanisms supposed; to help developing accelerated tests and interpreting the acceleration.
• For WP5 (Lifetime prediction methodology): to continue with sensitivity studies; to use all other WPs data and sensitivity results about parameters able to increase the components degradation and to accelerate cell voltage losses to propose accelerated protocols; to develop and validate thanks to in-situ measurements and specific experiments, degradation rates modelling predictions, to confirm degradation mechanisms in accelerated tests evaluating ex-situ analyses results; to propose a general methodology for lifetime prediction including these accelerated tests.
• For WP6 (Dissemination), to communicate about project results in the events organized by FCH-JU, to exchange with other projects dedicated to similar topic as suggested by reviewers, to enhance scientific dissemination of the results and to consider further exploitation as recommended by the mid-term reviewers.
• For WP7 (Management), to continue the general coordination actions, with two main issues being the finalization of actions decided at the mid-term review, and the ending of the project with last meeting, and last reports.

Project Results:

1.3.1 Introduction

For this final report, it has been decided to give a synthesis of Premium Act main outcomes by type of activities, more than by WPs as it is already done in the periodic reports.
The aim is hence to give here a better overview of the project outcomes during the first and second period considering particularly the points outlined below:
• New or original aspects of each activities and results
• Specific outcomes regarding degradation mechanisms with regard to existing knowledge
Final outcomes related to the two main goals: operating strategies and accelerated degradation for life time prediction methodology

Selected illustrations of results are also presented.
1.3.2 General context at the beginning of the project:
• Few data are available on the 2 technologies considered in the project: DMFC and reformate H2 PEMFC
Comment: most of the mechanisms and ageing results described in the literature concern pure H2 and automotive application.
• R&D work conducted in Premium Act is largely based on the consortium knowledge
Comment: partners involved have all specific background at state of the art level, directly applicable to complete the technical and scientific objectives of the project.
As a consequence, technical activities conducted and scientific results obtained on the two types of technologies considered present new and original aspects compared to existing data and knowledge.

1.3.3 Synthesis of new and original aspects of the activities and results Outcomes related to systems
Possibility to exploit for durability and degradation understanding data of very long term tests available for the project with state of the art IRD DMFC components and systems:
• Availability of aged samples from real systems after degradation due to different critical conditions
• Availability of data for a systems operated more than 20000 hours in stationary conditions, one system operated more than 3500 hours with a degradation of10 – 15µV/h).
• In addition, for the DMFC systems the use of very specific electrode structures for MEAs allowed original information from analyses of both new and aged samples.

Design adaptation of the SOPRANO PEMFC CHP system balance of plant: flexible tool particularly suitable to validate innovative protocols in real operative conditions
• Targeted complete system lifetime of 10.000 hours initially while being able to reach the targeted 40.000 hours in future evolutions
• Design implementations focusing on net efficiency, overall product size, and components cost
• Design modification to enable air bleeding
• Development of an accurate process control
• Process control implementation to enable innovative control strategies

• Production of a compact Programmable Logic Controller for the process control of the PEMFC CHP system able to operate in harsh environment (EN61131-2 standard). It is composed of a base including a Digital Signal Processor and 4 modules which can be analog, digital inputs/outputs or special functions like relay modules, measurement system (See Figure 5).

Detailed information on the operation of the ICI PEMFC CHP system vs. fuel reformer management based on 1000 hours of operation: effect of fuel composition at system level useable for further studies at cell level
During the tests and normal utilization, data was collected and analyzed in order to better understand stacks behavior, which is influenced mainly by a set of variables like air and syngas thermodynamic properties, hydrogen and oxygen stoichiometry, cooling DI water characteristics.

From data analysis, CO concentration in the syngas seems to have an important role in the stack performance, identifying the fuel processor as the most critical component in system operations. Results show lower performance of the stack when fed with high concentration CO syngas.

The introduction of some devices improved the system stability and therefore syngas quality; in particular the implementation of an air bleed system seems to enhance stack performance.
Benefit of air bleeding: the performance rises from under 250V to 265V with CO concentration over 20 ppm, and up to 293V with about 5 ppm. Furthermore, if a system component failure occurs (i.e. water circuit pump), and the CO concentration rises abruptly up over 20ppm, there’s a loss in performance (10V/h ca.) which results completely recoverable when the CO concentration decreases.

Another recoverable loss in performance is connected to cathode oxygen concentration. When it is about 10.5% in cathode exhaust, doesn’t result any problem, but if it rises the loss occurs because the membranes are not correctly hydrated and the recovering process requires many hours. So right now an improvement in cathode air control is in progress.
Best operation condition achieved are for: low CO concentration (under 5 ppm), regular and stable stack voltage over 300 V. A small performance loss during operation occurs.

Another improvement on Sidera30 system is the development of a 110 single cell volt measurement. It makes possible find out failure on each stack’s cell. Outcomes related to DMFC in-situ testing

• Original insights on DMFC degradation test method in single cells:
• parallel study of anode half cells and single cells used to identify anode/cathode contribution on both temporary/permanent degradation
• real-time monitoring of cathode effluent composition, to evaluate methanol and water transport though the MEA
• Original insights on DMFC temporary degradation:
• existence of anode temporary degradation: its origin is related to water and methanol concentration reduction at anode side, due to CO2 accumulation in electrode and GDL, hindering reactants transport
• temporary methanol crossover reduction is observed as a CO2 content variation in cathode exhaust, coherently with the proposed interpretation of anode temporary degradation, while water content in cathode exhaust remain constant
• cathode temporary degradation is most probably caused by a temporary reduction of ECSA, due to Pt or Ru oxides formation during operation at high potential (Ru presence at cathode is confirmed by ex-situ measurements)
• proposal and validation of refresh cycles: dramatic reduction of temporary degradation; the proposed refresh cycles permit a recovery of temporary degradation at both anode, permitting CO2 removal from GDL/electrode, and cathode, determining the reduction of Pt/Ru oxides

• Original insights on DMFC permanent degradation (long term test >1000 h):
• from EIS spectra, CV, polarisation curves: considerable reduction of cathode active area, mainly due to Pt dissolution/re-deposition, determining the major contribution to permanent degradation (about 70%)
• from EIS spectra, anode polarisation curves: reduction of anode catalyst effectiveness, mainly due to Ru dissolution/migration, determining the remaining contribution to permanent degradation (about 30%)
• membrane transport properties decrease is very limited
• comparing reference and benchmark MEAs: methanol crossover and water transport/flooding influences considerably the relation between cathode ECSA loss and performance decay Outcomes related to PEMFC in-situ testing
• Original insights on the behaviour of PEMFC in single cells and short stacks: all ageing studies conducted in load cycling mode instead of stationary
• Identification of reversible and permanent degradation for a single cell operated with load cycles and reformate fuel (1000 hours): carbon monoxide concentration has stronger impact on the reversible degradation,
• Permanent degradation is much lower than reversible and related to catalyst layers activity decrease; no membrane degradation while using reinforced references (from EIS and cyclic voltammetry)
• Evidence of an increase in reversible degradation when operating under reformate fuel compared to pure hydrogen, first direct and in-situ evidence that the anode tolerance can be the main issue.
• Confirmation that too high CO content with non aged catalyst, or decrease of CO tolerance due to degradation, can lead to voltage oscillating behaviour related to a threshold allowing full coverage of the anode catalyst surface, causing fast increase of anode potential up to CO oxidation potential and total cleaning before starting again with pollutant adsorption.
• Confirmation that load cycles lead to cathode catalyst ECSA losses and evidence by cyclic voltammetry and also by EIS (on high frequency anode activity resistance) of anode catalyst (PtRu) modifications of structure and properties.
• Severe increase of performance degradation at high current density when operating with more dynamic load cycles or with reformate at higher CO concentration: high CO concentration and dynamic cycles identified as major parameters for further developments of accelerated tests

• With segmented cell measurements applied in a 25cm² single cells it has been found that small amounts of CO in the fuel lead to a homogeneous degradation of the current density. In contrast, humidity gradients unambiguously lead to accelerated performance loss in areas with low local humidity, when considering more cathode degradation.

• Availability and first operation of PEMFC stack with same components and same load cycles as in a single cell: systematic study of fuel composition enabling to get additional information compared to single cell for more accurate interpretation of degradation. Use of electrochemical methods for each cell analysis in a stack allowing further knowledge about local conditions impact at stack level (possibility to exploit for degradation understanding the possible heterogeneities among cells).
• Evidence of conditions enabling better voltage stability such as air bleeding, and temperature increase for fuel tolerance, and stops and restart for temporary degradation limitation; but with always a controlled use of the parameter to avoid additional degradation phenomena particularly related to membrane.

• Validation on different MEAs in single cell and also at stack level of the accelerated ageing protocols proposed with more dynamic and higher load cycles, with increased CO centration stages during nominal load cycles, or combined protocols including both features. Identification of the accelerating effect of an increase water content (or partial flooding) for the degradation of anode catalyst. Outcomes related to ex-situ analyses
New information retired from ex-situ analyses techniques:
• Vibrational spectroscopy techniques: Large area FTIRAS mapping and FTIR-ATR microscopy established for screening of all MEA components allowing for spatially resolved degradation of polymers. Yet, due to high scatter of the data the data evaluation needs statistical analysis procedures to provide reliable information. In the case of PEFC the GDL seem to be stable under and the PTFE does not change during aging. For DMFC MEAs, on the other hand, the PTFE distribution changes upon aging indicating a change of the hydrophobicity in accordance with XPS measurements

• Electron microscopy techniques:
• Application of standard and high resolution transmission electron microscopy (TEM) allowing to detect very small quantities of catalysts particles showing the very beginning of ageing (Ruthenium precipitates in the membrane of PEMFC)
• First use of last generation analytical TEM (FEI-Osiris) allowing accurate spatially resolved chemical analysis by X-EDS: identification of platinum core in Ru particles detected in the membranes. New approach for improved determination of mechanisms. In addition, use of high resolution TEM allowing to identify the atomic structure to ensure metallic phase of the particles detected.
• Identification of a new mechanism leading to Pt-core / metallic Ru petals involving the common Pt-band in the reduction of Ru species coming from anode Ru dissolution.

• Enhanced knowledge of Ru dissolution mechanisms: evidence of occurrence whatever the operating conditions, even in cases with low anode potentials and identification of parameters influencing qualitatively and quantitatively the dissolution and further re-deposition of the ruthenium (in the membrane or anode microporous layer particularly).

• Of particular interest for the project, all mechanisms related to catalysts degradation (cathode or anode) have been confirmed to be enhance quantitatively when applying selected accelerated ageing and also, no additional mechanism is provoked by these AccST.

• Electron microscopy/spectroscopy techniques:
• A new depth profiling method in combination with XPS has been established to analyse porous media, such as MPL and CL. Thereby, the upper surface layers are successively removed by delamination with adhesive tape with subsequent XPS analysis, see . The analysed depth reaches from around 0.5 μm to a few tens of micrometers (i.e. macrostructure). Moreover, this approach is chemically non-destructive providing chemically non-falsified depth profiles, in opposite to ion etching (decomposition of polymers). In the Premium Act project it has been successfully used to analyse catalyst degradation; especially, with this technique the Ru cross-over in DMFC was analysed in very detail showing a uniform redistribution of anodic Ru in the cathode CL. Moreover the data show that in DMFC Ru cross-over occurs mainly at the beginning of operation and proceeds only very slowly upon further operation,

• Standard XPS has been used to study all components of MEAs. For instance, in the case of GDL it clearly shows the decomposition of PTFE within the gas diffusion layer (Figure 20). Specifically, in the C1s spectra of the anode GDL the increased signal in the range between 285 and 290 eV clearly indicates a chemical decomposition of the PTFE. In contrast, the PTFE in the GDL on the cathode is not chemically decomposed.
• The loss of polymers, especially at the CL , can be cleary observed.

A comparison of degradation effect found in an 1100h aged DMFC MEA and a DMFC MEA operated at 200h at an AST, as presented in the paragraph about DMFC in-situ testing, has been conducted showing good correspondance of conclusions obtained by XPS analyses.

• Ex-situ analyses of the GDLs transport properties
Preliminary Accelerated Stress test has been defined and applied for DMFC Gas Diffusion Layers: it has been shown that a variation of material properties could imply a modification in mass transport regime Outcomes related to DMFC modelling
New insights on modelling in DMFC:
• Single cell model including a complete and accurate description of water management and flooding effects. The developed model is validated over a wide range of operating conditions, with respect to different typologies of measure and with different configurations of GDLs properties. ;
• Detailed two-phase mass transport phenomena through anode porous components revealed that the main methanol transport mechanism is gas diffusion. The modeling result also implies the possibility to design an optimized DL with balanced mass transport loss and reduced methanol crossover.

• New approaches and models developed for the simulation of anode impedance behaviour, permitting a better interpretation of DMFC anode operation (Methanol Oxidation Reaction) and temporary degradation.;
• New model of Ruthenium dissolution: mechanism included into the MEA model and effect studied spatially resolved within the anode catalyst layer (quicker dissolution near the membrane).
• New model of platinum particle growth based on Ostwald ripening and coalescence developed and included into MEA model. Validation of degradation model by aging tests.

• Empirical correlation based on detailed degradation model derived and included into single cell model for lifetime predictions, evidencing the possibility of severe local degradation at channel inlet during AST tests.
• Study on propagation of uncertainty indicates that knowing the initial particle size distribution is important for accurate lifetime predictions. Outcomes related to PEFC modelling
New insights on modelling in PEMFC:
• New approach based on the interaction of a mechanistic model describing degradation mechanisms and a fluidic model computing the local conditions for the description of electrodes and performance degradation. The coupling has been completed and tested for platinum dissolution, but can be extended to other mechanisms;
• Implementation of local active area degradation in PEMFC fluidic model: evidence of an increase of the initial heterogeneities with horizontal and vertical resolution (from gas inlet to outlet, in the rib and channels and in the thickness of electrode from the membrane to the gas diffusion layer)
Qualitative experimental validation of the predicted degradation at the stack level. Comparison of experimental degradation rates with simulation prediction for various operating conditions.

The simulation results show :
• that the degradation at the MEA level may be very heterogeneous between rib and channel, and also through the thickness of the calalyst layer (depending on the operating conditions);
• that the cell level model is able to predict, at least qualitatively, the inhomogeneous degradation from the inlet to the outlet, and the consequence on the current distribution.

1.3.4 Specific outcomes regarding degradation mechanisms with regard to existing knowledge List of main mechanisms recognized to occur from literature and consortium knowledge
Irreversible degradation (SoA):
Catalyst layer
• Platinum dissolution + migration
• Growing of larger Pt particles by Electrochemical Ostwald ripening
• With PtRu catalysts used anode side: Ruthenium dissolution + migration
• With PtCo catalysts used cathode side: Cobalt dissolution + migration
• Carbon corrosion
• Site blocking (by non reversible contaminants)
• Mechanical damage (related to local hydrothermal cycles, or to chemical degradation)
• Chemical degradation leading to chain scission (incl. mechanism due to Fe contamination)
Gas Diffusion Layers
• Reduction of hydrophobicity (PTFE dissolution)

Reversible degradation (SoA):
Catalyst layer
• Hydrocarbon based contaminations: can be removed by refresh cycle
• Site blocking (mainly CO)
• Catalyst oxidation
• Drying out (locally at air inlet for PEMFC only when partially humidified air is used)
Gas Diffusion Layers
• Water accumulation List of mechanisms identified by the project in aged MEAs of Premium Act systems
Non reversible (Premium Act outcomes):
- Electrochemical Ostwald ripening/coalescence (?): growing of Pt particles identified in PEMFC and in DMFC aged cathode samples after single cell test
Mechanism causing the first permanent performance losses visible on polarisation curves and electrochemical impedance spectra and the evolution of the cyclic voltammograms cathode side towards more define peaks.

- Ru dissolution in PEMFC:
1-Evidence of Pt-core/Ru-petals flower shape particles detected in the membrane
2-Ru deposition occuring in the anode MicroPorous Layer in accelerated conditions
Ru dissolution in MEAs aged under load cycles is the mechanism causing the additional permanent performance losses visible on polarisation curves under reformate hydrogen and responsible for the evolution of the anode electrochemical properties as visible on one hand by cyclic voltammetry with reduced capacitive charge and in worse cases sharpening of hydrogen adsorption/desorption peaks reflecting more Pt behaviour and by impedance spectroscopy with an increased high frequency contribution due to poorer anode activity under reformate. It can be reminded that this degradation does not negatively impact the operation under pure hydrogen for the duration considered: this appears consistent with the common knowledge claiming the reversibility of degradation due to reformate and carbon monoxide.

- Ru dissolution in DMFC: detection of Ruthenium species at cathode side and membrane of an MEA aged in a DMFC system
Mechanism causing probably a part of the anode permanent losses; a probable influence on cathode temporary degradation is suspected
According to XPS analysis Ru cross-over occurs mainly at the beginning of the operation (during conditioning of the MEA), and proceeds only slowly upon further operation.

- PTFE decomposition in DMFC GDLs: Long-time operation leads to gradual chemical decomposition of PTFE in anodic GDL likely leading to a change of hydrophobicity. With XPS fluorine depleted components can be observed after long time operation (Figure 20)
- Loss of polymers in the CL and MPL on both, the anode and the cathode side: a slight loss of polymers is observed (no decomposition) that occurs predominantly at the beginning of operation (Figure 21).

Reversible (Premium Act outcomes):
- CO pollution in PEMFC: tests conducted in single cells and short stacks with reformate fuel containing various CO concentration
Mechanism causing, and measured by, cell voltage decrease at low and high current density. Reversibility confirmed with polarisation curves under pure Hydrogen.
CO identified and confirmed as causing non reversible degradation under reformate operation because the consequence is the loss of ruthenium: impact of CO on permanent degradation clarified as an accelerating feature when increasing its concentration because direct increase of anode potential and of dissolution
- Anode CO2 accumulation in DMFC (electrode/GDL): identified by electrochemical impedance spectra and mass transport analysis
Mechanism causing anode temporary degradation. Reversibility confirmed with different refresh cycles.
- Pt/Ru oxides formation in DMFC cathode catalyst: identified by electrochemical impedance spectra and specific test with different operating strategies
Mechanism causing cathode temporary degradation. Reversibility confirmed with different refresh cycles.

1.3.5 Specific outcomes regarding operating strategies

Operating strategies are addressing the temporary (or reversible) degradation. The aim is find methods enabling to stabilize the performance, or to reduce the cell voltage losses during ageing tests. Operating strategies related to DMFC systems

• Validation of operating strategies, based on refresh cycles (duration: few minutes), able to mitigate temporary degradation have been identified; refresh cycles can be optimize to minimize temporary degradation

• Validation of an operating strategy (duration: some hours) to fully recover performance has been developed

• Possible phenomenon: the minimization of temporary degradation determines a positive effect on permanent degradation both in single cell s and in systems Operating strategies related to PEMFC systems

Experimental results
• Confirmation of the positive impact of stops during load cycling. Local in-situ conditions are supposed to be modified towards less heterogeneity during stops of the load, thus allowing higher average power when restarting current production.

• Validation at stack level of the positive impact of continuous air bleeding (less than 1% for low CO concentration of 10 ppm in the reformate) to first increase the performance but most of all to decrease the voltage degradation rates (minus 25% at high current during day/night type load cycles at 65°C)

• Validation at stack level of the positive impact of increasing temperature from 65 to 80°C again to first increase the performance but mainly to decrease the voltage degradation rates (minus 60% at high current during day/night type load cycles without air bleeding).

Simulation results
Simulation results show that some operating parameters impact strongly the degradation of the cells. The system operating strategy should be designed to remain as close as possible to optimal operating conditions in terms of lifetime. This is especially the case if we consider an hybrid system, where the battery gives more freedom for the power demand to the fuel cell during a cycle. The simulations show that an balance can be found between minimizing the hydrogen consumption and preserving the state of health of the stack

1.3.6 Specific outcomes regarding lifetime prediction methodology and accelerated tests Accelerated tests related to DMFC systems

• Qualification of several AST protocols have been tested and analysed both in single cells and in systems

• AST protocols based on load cycle are affected by a complex fuel cell behaviour: anode/cathode potentials are uncertain as well as temporary degradation effects

• Development and validation of a new AST, based on OCV and air break periods: it permits to accelerate cathode ECSA loss by a factor 5 approximately; performance decay acceleration factor, generally about 6-7, strongly depends on MEA properties

• XPS analysis show that for DMFC MEAs the applied AST partially lead to similar degradation as standard aging. Specifically, regarding the degradation of CL and MPL the observed changes in AST and aged samples agree. The degradation observed in GDLs, however, is completely different for the AST and the 1100h aged DMFC MEA.

See Table summarizing the ex-situ results. Accelerated tests related to PEFC systems
• First generation of AccSt protocols have been developed to check the negative impact of increasing CO content during lmoad cycles and to check the potential negative impact of more dynamic load cycles, including higher maximum current to increase the voltage amplitude during cycles.
• Based on performance (voltage) evolution during the cycles, calculated degradation rates, but also in-situ diagnostics of anode and cathode evolution, and in parallel ex-situ analyses of aged samples, proposed accelerating protocols have been validated: meaning that they do accelerate the degradation of the performance (higher rates in µV/h) but enhancing the original mechanisms, what is looked for, and not provoking new mechanisms or uncontrolled failure what would invalidate the protocol.
• Combined AccST including higher CO content and also more dynamic and higher load cycles have been thus developed as final proposal for PEMFC operating under reformate.
• AccST protocols have been tested and analysed both in single cells and stacks
• It has been shown that the MEA initial components, such as anode catalyst loading has a non negligible impact on the behaviour of the MEA versus acceleration of degradation. The conclusion is thus that the accelerating protocol can be tuned in order to be more efficient depending on the initial characterization of the components.
• Thanks to tests in stacks, water content in the anode has been also identified asa possible accelerating feature. Conclusion about lifetime prediction methodology

A lifetime prediction methodology has been developed; it is composed of:
• a systematic experimental characterization of the MEA in different operating conditions
• testing of the MEA in reference condition for a short period, performing complete diagnostics at meaningful time periods
• AST is performed for a short period equivalent to a long term testing
• detailed models of degradation mechanisms are calibrated on both AST and reference condition testing results. Simple correlations are tuned to the simulations and integrated in MEA models
• MEA models are used to predict the decay of performance and MEA lifetime
The lifetime prediction methodology is validated positively on 2 DMFC MEAs

The same methodology could be applied to PEMFC, but to be effective it needs that the simulations take also into account the anode degradation.

Potential Impact:


The project contributes directly to the development of the European fuel cell activities at least related to the three industrial partners involved.
Premium Act aims at developing a methodology adaptable to the multiple PEFC technologies for the improvement of systems durability. In this sense, the success of the project should help to overcome one of the main bottlenecks preventing fuel cells market development for European providers of stationary fuel cell systems and will contribute to cross cutting issues relevant for European R&D and fuel cell industry development. Thus, reliable systems corresponding to the technical specifications of the energy global market could be widespread, which will change the end-user habits towards the stationary energy management and will help to reduce greenhouse gas emission.

Main dissemination actions have been:
• first classical, with the production of communications and publications in international conferences and scientific journals (see tables);
• to contribute to reporting and dissemination events organized or promoted by the FCH-JU;
• and as specific actions, to co-organize a common workshop with FCH-JU projects KEEPEMALIVE (lead. Sintef) and STAYERS (lead. Nedstack) dealing with stationnary application “Degradation of PEM Fuel Cells”, held in Oslo in April 2013, where degradation studies conducted in the project have been presented on both PEMFC and DMFC;
• and particularly to organize a dedicated Premium Act workshop about “characterization and quantification degradation”.

The international Workshop on “characterization and quantification of MEA degradation processes” was organised on September 26-27, 2012 in Grenoble at CEA – Minatec. The purpose of this workshop was to show the contribution of different characterisation methods for better understanding and quantification of the degradation mechanisms of PEMFC components. Around 50 people attended the workshop, people coming either from industrial or academic institutions inside the Europe community. Five speakers, internationally recognized, have been invited to present their work:
o Atsushi Ohma from Nissan Research Center in Japan,
o Karren More from Oak Ridge National Laboratory in USA,
o Hans Bettermann from University of Düsseldorf in Germany,
o Laetitia Dubau from LEPMI-University of Grenoble in France
o Arnaud Morin from LITEN CEA-Grenoble in France.

Nineteen other presentations were given by the other participants of the workshop. They gave a good overview of the state of the art of research on this topic. Many presentations focused on the degradation mechanisms of both the active layers and the membrane whereas other presentations showed different techniques and strategies developed in order to localise, across the MEA surface area (in plane or through plane) or between the different MEA in a stack, the zone where the MEA degradation is the most severe. All these presentations gave rise to fruitful discussions and probably will emerge on new collaborations between the participants.

List of Websites: