D01.01: Summary report on modelling and experimental ... · PUMA MIND - 303419 D01.01 - Summary...

PUMA MIND - 303419

D01.01 - Summary report on modelling and experimental studies on PEMFC physicochemical and ageing mechanisms

D01.01: Summary report on modelling and experimental studies on PEMFC physicochemical and ageing mechanisms

FCH JU Grant Agreement no. 303419

Project acronym: PUMA MIND

Project title: Physical bottom Up Multiscale Modeling for Automotive PEMFC Innovative performance and Durability optimization

Funding scheme: Collaborative Project, Small or medium-scale focused research project

Area: SP1-JTI-FCH.2011.1.3: Improvement of PEMFC performance and durability through multi-scale modeling and numerical simulation

Start date of project: 17-12-2012 Duration 36 months

Project Coordinator: CEA, Pascal Schott

Tél: +33 4 38 78 48 40 ; Fax : +33 4 38 78 41 39 ; email : [email protected]

Date of the latest version of Annex I against which the assessment will be made: 16-10-2012

Document properties

WP1: Specifications and Technology Watch WP leader: DLR Author (Partner) Thomas Jahnke (DLR) Approval by author

Other Authors (Partner):

Georg Futter (DLR) Arnulf Latz (DLR) Thomas Malkow (JRC) Georgios Papakonstantinou (JRC) Georgios Tsotridis (JRC) Pascal Schott (CEA) Mathias Gérard (CEA) Manuelle Quinaud (CEA) Matias Quiroga (LRCS) Alejandro Franco (LRCS) Kourosh Malek (SFU) Federico Calle-Vallejo (ENSL) Torsten Kerber (ENSL) Philippe Sautet (ENSL) David Loffreda (ENSL) Stephan Strahl (CSIC) Maria Serra (CSIC) Pierpaolo Polverino (UNISA) Cesare Pianese (UNISA) Manik Mayur (HSO) Wolfgang Bessler (HSO) Costis Kompis (VODERA)

Approval by partners

Approval by WP leader

Approval by coordinator

Dissemination level Public

Distribution JTI Nikolaos Lymperopoulos e-mail e-mail e-mail e-mail e-mail e-mail

Revisions

PUMA MIND - 303419


Version Date of changes Author Changes 1.0 16-07-2014 Thomas Jahnke Merging of all contributions

Table of contents

NOMENCLATURE .................................................................................................................................................... 3

LIST OF FIGURES ............................................................................................. ERREUR ! SIGNET NON DEFINI.

1. INTRODUCTION .............................................................................................................................................. 5

1 MODELING OF PEMFC PERFORMANCE ................................................................................................. 5

1.1 First-principle approaches at the atomic level ...................................................................................... 5

1.2 Microscale simulations ......................................................................................................................... 6

1.3 Cell level ............................................................................................................................................... 8

1.4 Stack and system level ........................................................................................................................ 12

1.5 Coupling between the scales ............................................................................................................... 12

2 MODELING DEGRADATION MECHANISMS IN PEMFC ..................................................................... 14

2.1 Membrane ........................................................................................................................................... 14

2.2 Catalyst layer ...................................................................................................................................... 15

2.3 Gas diffusion layer .............................................................................................................................. 17

2.4 Bipolar plates ...................................................................................................................................... 17

3 EXPERIMENTAL VALIDATION ................................................................................................................ 18

3.1 In-situ methods ................................................................................................................................... 18

3.2 Ex-situ methods .................................................................................................................................. 24

4 CONCLUSION ................................................................................................................................................. 25

REFERENCES .......................................................................................................................................................... 25

PUMA MIND - 303419


NOMENCLATURE

BPP Bipolar plate

CCL Cathode Catalyst Layer

CE Counter electrode

CFD Computational Fluid Dynamics

CG Coarse grained

CGMD Coarse grained molecular dynamics

CI Current interruption

CL Catalyst layer

CM Continuum Models

CNT Carbon nanotubes

CV Cyclic voltammetry

DFT Density Functional Theory

ECSA Electrochemical surface area

EDL Electrochemical double layer

EEC Equivalent electrical circuit

EIS Electrochemical Impedance Spectroscopy

FC Fuel cell

FER Fluoride emission rate

GDE Gas diffusion electrode

GDL Gas diffusion layer

GFC Gas flow channel

HOR Hydrogen Oxidation Reaction

kMC Kinetic Monte Carlo

LBM Lattice Boltzmann method

PUMA MIND - 303419


LSV Linear sweep voltammetry

MEA Membrane Electrode Assembly

MPL Micro Porous Layer

NEB Nudged Elastic Band

ORR Oxygen Reduction Reaction

PBC Periodic boundary conditions

PEMFC Polymer Electrolyte Membrane Fuel Cells

PFSA Perfluorosulfonic acid

PNM Pore Network Modeling

PTFE Polytetrafluoroethylene

RE Reference electrode

RH Relative humidity

RHE Reversible hydrogen electrode

SEM Scanning electron microscopy

WE Working electrode

PUMA MIND - 303419


1. INTRODUCTION Achieving the goal of providing predictive tools to model PEMFC performance, durability and degradation is a challenging task requiring the development of detailed and realistic models reaching from the atomic/molecular scale over the meso scale of structures and materials up to components, stack and system level. In addition an appropriate way of coupling the different scales is required. Finally, suitable characterization methods and measurement techniques have to be applied for experimental validation of the models. In the following we provide an overview of the state of the art in modeling of PEMFC, covering all relevant scales from atomistic up to system level as well as the coupling between these scales. Furthermore, we focus on the modeling of PEMFC degradation and review the state of the art in experimental characterization methods and measurement techniques applicable to model validation. This report is a condensed version of the review paper which has been written by the PUMA MIND consortium within the scope of Task 1.1. Currently, this paper is under review at Journal of Power Sources and contains in total almost 500 references.

1 MODELING OF PEMFC PERFORMANCE

1.1 First-principle approaches at the atomic level

During the last decade, numerous theoretical works have been proposed to tackle the controversial interpretations regarding the elementary steps of the ORR [1–6]. Firstly, several groups have elucidated simple elementary steps at the level of gas/Pt(111) models in periodic boundary conditions (PBC), starting from molecular oxygen dissociation [7–10], water formation through an *O + *H pathway [11] and then, disproportionation routes [12,13] and water formation through an *OOH pathway [14]. Secondly, more systematic studies have been proposed to combine these competitive elementary steps and to conclude about the preferential elementary mechanism [3,15]. During the past few years, significant efforts have been devoted to probe the influence of the aqueous solution on the ORR mechanism [16–24], either by considering continuum models [16] or by treating explicitly a small number of water molecules [17,19,22–24] such as ice bilayers [18,20], or confined static water [21]. At the same time, the influence of the applied electric field has been evoked in the literature [2,3,25–30] either by thermodynamic approaches such as the computational hydrogen electrode [2,3,25,27,28,30–34], by first-principles based mean-field or multiscale models [5,6,35,36] or by self-consistently minimized DFT schemes [26,29]. Computational electrochemistry is a field still in development that is providing significant insight into the ORR mechanism on Pt, but the answers to several important questions are still elusive. In spite of contradicting observations, the field is slowly reaching consensus about the reaction intermediates and their relative importance and correlations between adsorption energies [37], while the number of effects included in the calculations keeps increasing. Recent works have started to theoretically address the influence of the

PUMA MIND - 303419


nanoparticle size on the adsorption energies and the ORR activity [38,39], but this kind of studies is not widespread yet, due to the large computational resources required. Solvation, sintering and degradation effects remain open issues and there are no studies that consider all these possible effects coupled to the electrocatalytic behavior within a single simulation [40]. Instead, what is found in the literature are collections of deconvolved effects that lack a sense of generality. Thus, the major challenges in the computational modeling of the ORR are currently the inclusion of the largest possible number of effects and the improvement of their descriptive and predictive power, so that the theoretical representation of the electrode-support-electrolyte interface is as realistic as possible [31]. Currently, a given surface property, namely an electronic [41], geometric [39] or thermodynamic parameter [37,41], is used to describe the trends in adsorption energies and/or catalytic activities. The latter can either be theoretically estimated as the free energy of reaction of the potential-determining steps or experimentally measured current densities. Ideally, when a wide range of materials is covered, a volcano-type curve should appear in which the activity initially grows up to a maximum, after which it decays. Thus, optimal catalysts exhibit a compromise in the adsorption strength of reactants and products, in line with the Sabatier principle. More sophisticated and consequently more accurate volcano plots should include the effects of pH and electrolyte, applied electric field and catalysts’ size, morphology and composition, in order to confirm the predictions of cruder approaches or make different predictions.

1.2 Microscale simulations

In the frame of the multiscale modeling, the so-called electrochemical double layer (EDL) model plays an important role in the in the redox kinetic processes [42–45]. This EDL represents the interface between the electrode surface and the electrolyte. The EDL is generally assumed to be formed by two regions, the Diffuse Layer and the Compact Layer. The Diffuse Layer model describes the proton transport, affected by the water and/or the morphology of the charged polymer (e.g. Nafion®) at the vicinity of the catalyst surface. The Compact Layer model describes the coverage evolution of the intermediate reaction species (involved in the REDOX reactions) as well as adsorbing/desorbing water molecules on the substrate surface. Significant efforts have been made in scaling up the EDL modeling from atomistic and molecular level properties to overall performance cell models [6,36,43,46,47]. Proton transport in the EDL is usually modeled within the Poisson-Nernst-Planck framework, not considering the role of water in the effective proton transport properties. Recently, a more complete multiscale model describing in detail the interplay between the EDL structure and DFT-based elementary kinetics has been developed [48,49] where the detailed influence of the Nafion® structure and the finite size of both water and solvated protons were considered to calculate the electrode potential as function of the imposed current density. The Kinetic Monte Carlo (kMC) method plays an important role in the description of surface dynamics, coverage evolution, adsorption, desorption and reaction rates involved in chemical surface. In principle kMC is an algorithm that computes random hops from one surface state to another. The complexity of the chemical events that takes place on the surface makes the study under first-principle electronic structure calculations unfeasible. Otherwise, first-principle theories, like DFT, are able to study isolated events and provide kinetic parameters to be implemented in kMC. This methodology is well described by Reuter [50].

PUMA MIND - 303419


The most used kMC algorithms are: the First Reaction Method (FRM [85] and FRMb [87]) Variable Step Size Method (VSSM [51] and VSSMb [87]); the Random Selection Method (RSM) [52]; the fast FRM (fFRM [53]). In the context of PEM fuel cells, Zhdanov [54] utilized MC simulation to study the kinetics of oxygen electrochemical reduction on Pt and found that this reaction may be nearly first order in oxygen under a wide range of reaction conditions even if the O coverage is appreciable and the O mobility is low. Zhdanov proposed that the reactant coverage is not first order in the O2 reaction. Abramova et al. [55] studied the O2 adsorption on fcc(111) metals by kMC. By taken into account seven elementary steps, they report the total oxygen coverage profile in terms of σ (charge density) and sticking probability reporting a strong coverage dependency in the pre-exponential factor. The Monte Carlo simulation was used to study the field fluctuations in electrochemical reactions [56]. This determines fluctuation of the kinetic rates, which implies that the Mean Field approach cannot be implemented. The kMC approach was also used to study Pt oxide growth [57] and to study the ORR in Pt cathode for solid oxide fuel cells (SOFCs) [58]. Self-organization phenomena in PEMFCs Despite recent progress in developing multiscale modeling approaches [59] enormous challenges remain in bridging atomistic simulations of realistic structures and continuum models that describe the operation of functional materials for PEMFC applications. Molecular-based simulations at mesoscale level provide insights into segregation, structural correlations and dynamical behavior of different phases in complex CLs. They contribute to furnishing reliable relations between structure, transport properties, and reactivity [60,61]. Despite an enormous number of phenomenological models, less effort was put into exploring effects of the microstructure of the CL on transport and reactivity. Advanced experimental and theoretical tools can address the correlation of transport-reaction processes with structural details at meso-to-micro- and down to atomistic scale. Coarse-grained (CG) molecular simulation techniques can describe the system at the micro-to-meso level, while still being able to capture the morphology at long time and length scales [62]. The application of CG models, however, requires special care. One should note that due to the reduced number of degrees of freedom, CG simulations may not be able to accurately predict physical properties that directly rely upon time correlation functions (e.g., diffusion).

A significant number of mesoscale computational approaches has been employed to understand the phase-segregated morphology and transport properties of water-swollen Nafion® membranes [48,62–64]. Because of computational limitations, full atomistic models are not able to probe the random morphology of these systems. However, as demonstrated by these simulations and applications to other random composite media, mesoscale models are computationally feasible to capture their morphology. Several approaches have been used, such as termed cellular automata and coarse-grained meso-dynamics based on self-consistent mean field theory. For Nafion®, most of these simulations support the idea that narrow water-filled channels and irregularly shaped, nanometer-size clusters of ionic head groups and water form the proton-conducting network that is embedded into the hydrophobic matrix. Structural complexity is even more pronounced in CLs, since they consist of a mixture of Nafion® ionomer, Pt clusters supported on carbon particles, solvent and water. Independent computational strategies are generally needed in order to simulate the microstructure formation in CLs. For instance, the structure of the ionomer-phase in CLs cannot be trivially inferred from that in

PUMA MIND - 303419


membrane simulations, as there are distinct correlations between Nafion®, water and carbon particles in CLs. The computational approach based on coarse-grained molecular dynamics (CGMD) simulations is developed in two major steps [62,65,66]. In the first step, Nafion® chains, water and hydronium molecules, other solvent molecules and carbon/Pt particles are replaced by corresponding spherical beads with pre-defined sub-nanoscopic length scale. In the second step, parameters of renormalized interaction energies between the distinct beads are specified. CGMD data can be used to parameterize microstructurally resolved models [45,66,67] of transport processes in PEMFC electrodes.

1.3 Cell level

1.3.1 Single cell models

Springer et al. [68] developed a one-dimensional steady-state model of a single PEM fuel cell. The novelty of this model lies in the experimental relationship which correlates the membrane water content, at a certain temperature, to the water vapor activity. Maggio et al. [69] presented a simple but innovative PEMFC model in which an original calculation of diffusion losses was proposed. They stated that the assumption of a constant gas porosity in the electrode diffusional layer hinders the identification of the real limiting current density Ilim involved in the calculation of diffusion losses. For this reason an innovative expression has been exploited to evaluate the limiting current density. A similar approach has been assumed by Costamagna [70] who studied how to ensure proper membrane hydration and avoid local temperature peaks developing a three-dimensional PEM fuel cell model. The main drawback of the approach applied in these two studies is that they neglect the presence of liquid water inside the porous media and at the interface between gas diffusion layer (GDL) and gas flow channel (GFC). To overcome this limitation, Mazumder et al. [71] proposed a Computational Fluid Dynamics (CFD) model able to simulate liquid water transport inside a 3D PEMFC, observing that the assumption of no liquid water formation leads to an overestimation of the cell voltage, especially at high current densities. Um et al. [72,73] developed a 3D computational fuel cell dynamics (CFCD) model of a PEMFC with a Nafion® 117 membrane. This model allows comparing the effects of straight and inter-digitated flow fields on oxygen and vapor concentration at the cathode side and also on the current density distribution. The model embeds also a detailed Membrane Electrode Assembly (MEA) sub-model where the membrane water content and catalyst layer reaction rate and ionic resistance are computed. As a result, the authors stated that the interdigitated flow fields at the cathode side improved the cell performance at high current densities due to the enhanced mass transfer of the oxygen. The model developed by Um et al. has been extended by Wang et al. [74,75] in order to simulate the transient processes occurring in a PEMFC due to a step change in the steady-state operating condition. The single-channel PEMFC model considers Gore® 18 µm and Nafion® 112 membranes to highlight the correlations between transient responses in terms of cell voltage and water content variation at low reactant humidity. From this study, the two main observations are: on the one hand, the ohmic resistance has been identified as predominant on the cell performance; on the other hand, membrane water content has been recognized as the chief part of the water management during transient maneuvers.

PUMA MIND - 303419


1.3.2 Membrane

The most important and popular materials for use as an electrolyte in PEMFC are perfluorinated polymers. Therefore, in this review, we will stick to models describing transport in perfluorosulfonic acid (PFSA) membranes. The macro-scale transport models for PFSA membranes presented in the literature can be roughly divided into three different approaches, depending on the driving forces that are considered. These are gradients of species concentration, pressure or, in the most general case, chemical potential. Two early models for the transport of protons and water in the membrane were developed by Springer et al. [68] and Bernardi and Verbrugge [76,77], which are both based on dilute solution theory, the first being of diffusive, the second of hydraulic type. Another hydraulic model by Eikerling et al. [78] considers electro-osmotic drag for the transport of water in the membrane counterbalanced by Darcy flow. The model by Fuller and Newman [79] uses Stefan-Maxwell equations derived from concentrated solution theory. In this approach, which has been adopted several times [80–82], the chemical potential acts as driving force for the transport. One example is the model of Janssen [80]. Also based on [79], Weber and Newman developed a model [81] which represents a combination of the diffusive Springer [68]- and the convective Bernardi/Verbrugge [76,77]-model and allows a physical description of Schröder’s paradox[83]. Thampan et al. also adapted the approach of Fuller and Newman for their model [82] and applied the “dusty-fluid model” [84] for species interactions. As shown by Fimrite et al. [85] this approach was erroneous as extra viscous forces are introduced and therefore the binary friction model, developed in [86], is the favorable approach. Consequently Fimrite et al. presented a new model [87] for transport and conductivity based on the binary friction model. Baschuk and Li [88] also used generalized Stefan-Maxwell equations to describe transport in the membrane with potential and partial density of water in the electrolyte as driving forces combined with non-equilibrium thermodynamics. Apart from the approaches used above, Choi, Jalani and Datta [89] presented a conductivity model, accounting for transport of protons via en masse diffusion, surface diffusion and Grotthuss mechanism.

1.3.3 Electrodes

The catalyst layers (CL) are made of catalyst (Platinum), supported by an electronic conductive material (carbon grains support) and ionomer impregnated around the agglomerates to transport the ionic current (H+). Moreover the CL must be porous to transport the reactive gases (hydrogen and oxygen) and evacuate the produced water [90]. Macrohomogeneous models consider that all phases are homogeneously mixed. A generalized Darcy law is used to describe transport phenomena [91]. Currently, local properties of the CL like contact angle or porosity, cannot be measured locally and might be different from the average properties that are used in the macrohomogeneous models. Therefore, macrohomogeneous models may lack accuracy and should be limited to modeling the performance at cell level [92]. Agglomerate models consider phenomena that happen within the agglomerate length scale. In these models, gradients of concentration and potential can occur within the agglomerates [93]. Modeling can be based on the Butler-Volmer equation, were the efficiency factor enables to account for the geometry while the Thiele modulus accounts for the ratio between reaction kinetics and reagents diffusion [94]. The models show that

PUMA MIND - 303419


the thickness of Nafion® around the agglomerate is a key parameter and should be no more than a few nm to promote reagents diffusion towards the Pt. Another way to model the catalyst layers is to consider a thin interface between the membrane and the GDL [95]. Transport phenomena within the CCL cannot be considered in this approach. Harvey et al. [94] compared the agglomerate model and the thin-film model and concluded that the agglomerate model is able to predict with accuracy all the domains of the polarization curves. Cetinbas et al., [96] proposed an original geometrical model of discrete Pt particles on carbon support. The results reveal that the discretely-distributed particle approach can capture diffusion losses due to particle interactions and that particle-level diffusion plays an important role in the global diffusion. This model is also able to demonstrate the strong impact of Pt loading on the overall performance. Moreover, authors demonstrate that performances at high current densities are increased when Pt particles are located at the periphery of the agglomerates, thanks to reduced diffusion limitations.

More recently, Direct Numerical Simulations (DNS) have been developed to investigate the influence of the structure of the CCL on transport properties [97,98]. This approach is focused on transport through the secondary pores. However, reconstruction of the structure is first needed.

Pore Network Modeling (PNM) is well adapted to two phase transport in porous media. El Hannach developed a porous network model (PNM) to calculate gas and liquid transport properties in the CCL, allowing to analyze the effect of local properties of the CCL on the two-phase transport mechanism [99]. A detailed model of the PEMFC cathode has been recently developed by Strahl, Husar and Franco allowing to study the impact of the carbon morphology (pore size distribution) and operation temperature dynamics on the electrode hydratation dynamics and cell potential variation [100].

1.3.4 Gas Diffusion Layer (GDL)

Diffusion layers are called GDL (Gas Diffusion layer) and MPL (Micro Porous Layer). The GDL aims at distributing gas from the channels of the bipolar plates to the catalyst layer, removing the generated water and conducting the electrons to the bipolar plate ribs. The GDL is made of carbon fibers covered with polytetrafluoroethylene (PTFE) to increase its hydrophobicity, thus promoting removal of liquid water from the GDL. The MPL aims at improving water management [101–103]. The MPL has been particularly studied by Carrigy et al. [104]. They demonstrated the paramount contribution of Knudsen diffusion in the MPL that can be up to 80% of the effective diffusivity, due to small pore sizes (50 nm). Since the MPL protrudes to a large extent into the GDL, Knudsen diffusivity is relevant in the GDL as well. To study two phase transport in the GDL, four methods have been established: Continuum Models (CM), Direct Numerical Simulation (DNS), Pore Network models (PNM) and the Lattice Boltzmann method (LBM). Continuum models consider a continuum media with (volume) averaged properties. They rely on generalized Darcy’s law and phenomenological relationships [105]. However, due to the lack of length scale separation between the pore sizes (about 50 µm) and the thickness of the GDL (about 300 µm), CM can lead to inaccurate conclusions and are not appropriate to study of the GDL in detail [106].

PUMA MIND - 303419


Direct Numerical Simulation models aim at computing the two-phase flow directly at the pore scale. However due to time consuming calculations, this method may not be the most appropriate one to study the GDL [106]. Pore Network Modeling is a powerful approach that enables to compute transport parameters in the GDL, to be then used in macroscopic models. PNM relies on capillary forces to describe liquid water invasion[99]. The network is made of a distribution of pores connected through throats. Effective transport properties are computed from both structural information (porosity, pore size distribution) and from physico-chemical properties (wettability). PNM can also give insights for optimizing the structure of the GDL by improving the understanding of the phenomena that occur inside the GDL [106]. PNM enables a two phase modeling where evaporation and condensation (GDL are usually cooler than CL) are also taken into account. Recently, the Lattice Boltzmann method (LBM) was also developed for calculating properties in porous structures. LBM simulations consider a collection of particles, each one consisting of multiple molecules. LBM is well adapted when simulating a flow through a complex geometry. Therefore the real structure of the GDL can easily be considered for the computation. Moreover, LBM are highly versatile and numerically stable and the employed algorithm is inherently parallel in computation. Although examples of two-phase flow modeling using LBM are scarce in the literature [107,108], LBM has been found to be particularly relevant for multiphase flows in porous media where interfacial dynamics play a major role [109].

1.3.5 Interface between Gas Diffusion Layer and Gas Flow Channel

Several authors focus on the liquid water impact on the PEMFC behavior, investigating the effect of the operating conditions and the cell components characteristics on the liquid water removal, particularly at the cathode. Some authors coupled experimental activity and model analysis [110–115], whereas others exploit only experimental data available in literature to validate the developed mathematical models [116–118]. Kumbur et al. [113] modeled the detachment of single droplet through a lumped static force balance. They also built a non-reactive cell experimental apparatus to support and validate their model. Also other authors, such as Chen et al. [110] and Theodorakakos et al. [114], followed the same approach. On the same line, Cho et al. [111] developed an analytical model of a water droplet deformation and detachment accounting for both a 3D (sphere shape) and a 2D case (cylindrical shape). Among others, He et al. [117] developed a Two-Fluid (TF) model for the representation of the PEMFC two-phase flow, validated through a comparison with the experimental data of Wang et al. [119]. Interesting observations were made by Mortazavi et al. [120], who investigated the effect of different polytetrafluoroethylene (PTFE) content in the GDL on water droplet emerging and detachment. Some publications exploit the Volume of Fluid (VOF) method, as done by Chen [116]. Another approach, based on the Mean Value Model (MVM), has been followed by Esposito et al. [121,122] to investigate the liquid water transport mechanisms at the cathode. An in-depth experimental investigation of the effects of droplet oscillation behavior on its detachment characteristics has also been performed [123]. They highlighted the importance of the oscillation mechanism on droplet detachment. Based on the observation by Esposito et al., the mathematical model of Polverino et al. [118] describes the main influence of the oscillation mechanism on the droplet detachment by means of a dynamic lumped force balance among drag, surface tension and inertia. The

PUMA MIND - 303419


model has been validated through the exploitation of experimental data available in the literature [114], showing good agreement with the experimental results.

1.4 Stack and system level

In order to be able to produce energy, it is necessary to integrate the fuel cell stack with other components to form a fuel cell-based power generation system. Fuel cell system modeling has played a decisive role in system design and developing, optimizing and testing of fuel cell control strategies. Since the control strategy for a real system has to take into account important actuator and peripheral system dynamics, the cell level model has to be upgraded to make it suitable for the controller design. However, modeling and controlling PEM fuel cell based systems is a particularly challenging task due to the interactions between physical phenomena of different nature and scale and the presence of nonlinear structures. Although most physical phenomena occurring in a PEM fuel cell can be incorporated in the macroscopic CFD models at cell level, it leads to time-consuming simulations with high computational costs. To improve performance and durability of fuel cell systems, the design of controllers that optimize these aspects under all the expected operating conditions is required. However, the design of good controllers is highly dependent on the available dynamic models. Therefore, the task of obtaining control-oriented models at system level has become important. Control-oriented PEM fuel cell stack and system modeling has been studied by several authors using different approaches. An example for a dynamic PEMFC model specially developed for control engineering is the model presented by Pukrushpan et al. [124] to study two main control problems: (i) the amount of oxygen supplied at the cathode side to the fuel cell system and (ii) the hydrogen flow supplied at the anode side and provided by a fuel processor system. The model describes the transient behavior of the air compressor, the manifold filling dynamics, the reactant partial pressures and the membrane humidity. However, the model neglects the electrochemical reaction kinetics and stack temperature is treated as a constant parameter due to its slow time constant. However, the simulation results using this type of models are limited because either they use an electrochemical model that is not experimentally validated or experimental models that do not take into account the electrochemistry. Therefore, more research on multiscale modeling is needed to include these effects not only in CFD modeling but also in control-oriented modeling, considering mitigation strategies for degradation mechanisms and water management related performance improvement.

1.5 Coupling between the scales

As discussed by Franco [44,45,125,126], two approaches are possible to develop multiscale models of electrochemical power generators: the direct one and the indirect one. Direct multiparadigm multiscale models consist in coupling “on-the-fly” models developed in the frame of different paradigms. For example, continuum equations describing transport phenomena of multiple reactants in a porous electrode can be coupled with kinetic Monte Carlo (kMC) simulations describing electrochemical reactions among these reactants. Computationally speaking, these direct multiparadigm methods are very expensive; therefore, the indirect multiparadigm method represents a feasible alternative. This method consists in extracting data from a lower-scale and using it as input for the

PUMA MIND - 303419


upper-scale via their parameters. For example, activation barriers for elementary reaction steps can be extracted from DFT by Nudged Elastic Band (NEB) calculations [127], and then injected into Eyring’s expressions to estimate the kinetic parameters, these expressions are used for the calculation of the individual reaction rates vi at the continuum level [6] and is in turn used for the calculation of the evolution of the surface or volume concentrations of the reaction intermediates. Molecular Dynamics (CGMD) for the calculation of the materials structural properties (e.g. tortuosity τ and porosity ε) as function of the materials chemistry, which are used in turn for the estimation of the effective diffusion coefficient used in continuum reactants transport models [128]

0DDeff τε= , (1)

with D0, the molecular (binary) diffusion coefficient of the diffusing species, e.g., hydrogen-vapor water mixture at the anode or air-water vapor mixture at the cathode. In this way one could then imagine building up an electrochemical power generator model with macroscopic equations based on material parameters extracted from atomistic and molecular level calculations. Within this sense, in 2002, Franco has invented the multiscale simulation package of PEMFC called “MEMEPhys”, from the French spelling Modèle Multiéchelle Physique [65,129]. The model is an indirect multi-paradigm, multiscale approach: a continuum model describing electrochemical and transport mechanisms with parameters extracted from ab initio databases (for the reactions kinetic parameters as function of catalyst chemistry and morphology) and CGMD calculations (for the materials structural properties as function of their chemistry). More recently, Franco has presented his new simulation package MS LIBER-T (Multiscale Simulator of Lithium Ion Batteries and Electrochemical Reactor Technologies) [45,130,131]. This computational software is a flexible multiparadigm, multiscale and multiphysics code devoted to the numerical simulation of electrochemical devices for energy conversion (fuel cells, electrolyzers) and storage (batteries, supercapacitors). It introduces significant progress compared to MEMEPhys which depends on commercial numerical solvers (Simulink). MS LIBER-T is programmed in C and Python and therefore is highly flexible, modular and portable. MS LIBER-T is designed to support direct multiparadigm calculations, for instance, simulations coupling on the fly the numerical resolution of continuum models (e.g. describing charge transport in the electrode volume) with discrete models (e.g. Monte Carlo module describing the elementary reaction kinetics). Another novelty introduced by MS LIBER-T is its capability of integrating phase field models describing phase formation, separation and evolution, mechanisms highly relevant for material microstructure simulations [125]. Models proposed by Franco et al. are able to capture the impact of the chemical and structural properties of the materials onto the cell macroscopic observables which can be measured at the laboratory scale (e.g. polarization curves, charge-discharge curves, impedance spectra, cyclic voltammetry, etc.). The models represent explicitly the different physical phenomena as nonlinear sub-models in interaction. Such a developed model is a multi-level one in the sense that it is made of a set of interconnected sub-models describing the phenomena occurring at different levels in the PEMFC. However, this description remains macroscopic (suitable for engineering applications) in the sense that it is based on irreversible thermodynamic concepts as they are extensively used in chemical engineering: use of conservation laws coupled to closure equations. Such an approach allows to easily modify the sub-models and to test new assumptions keeping the mathematical structure of the model and the couplings [65].

PUMA MIND - 303419


2 MODELING DEGRADATION MECHANISMS IN PEMFC

2.1 Membrane

Deterioration of PFSA membrane properties such as gas separation, proton conductivity and electrical insulation is governed by thermal, mechanical and (electro-) chemical degradation [132]. For state of the art membranes under typical operating conditions, chemical degradation is the most important degradation mechanisms. Therefore, in the following, the existing models for chemical degradation will be reviewed. Two types of models can be distinguished. To gain insight in the basic mechanisms of degradation and to interpret experimental results, kinetic 0D (zero dimensional) models are applied [133–136]. Further, 1D models for chemical degradation of the membrane were embedded in a simulated fuel cell environment [137,138]. A kinetic model to determine the contribution of backbone unzipping and side chain cleavage to the overall chemical degradation process in various degradation environments was developed by Xie and Hayden [133]. Correlation of infrared spectroscopy measurements to this kinetic model gives the basis to determine the rate constants ratio under different experimental conditions. Chen and Fuller conducted membrane durability tests at different temperatures [135]. To explain the experimental results a kinetic model was developed. In this model, four steps of membrane degradation are proposed: initiation via Fenton’s reaction, propagation by oxygen attack and unzipping, termination and side group degradation. Assuming steady state, an expression for HF formation is formulated. Based on this expression, a rate expression for fluorine loss is derived taking into account the contributions of Fenton’s- and side group degradation reaction. Another kinetic model aiming to elucidate the mechanisms of chemical degradation was presented by Gubler et al. [134]. In their model, the probability of H2O2 undergoing different reaction pathways is investigated. The impact of Fe-ion and COOH-group concentrations as well as radical attack on the polymer under Fenton test and cell conditions was simulated. The model allows for the calculation of H·, OH· and OOH· concentrations in the membrane and fluoride emission rates. This model gives detailed insight in the interaction of different reactions supposed to occur during membrane degradation. Concerning the 1D models, Kundu et al. investigated degradation of a Gore™ PRIMERA® series 5510 catalyst coated membrane [139]. A semi-mechanistic 1D transient model was developed and compared to the experimental investigation of fluoride emission rate (FER), crossover current and OCV. In this model, degradation is assumed to start at the cathode and to advance in a wavelike manner to the anode causing thinning of the membrane. Crossover current, H2O2-formation, subsequent radical formation and membrane degradation is proposed to depend on the flux of H2 from anode to cathode. Cumulative fluoride release of anode and cathode, crossover current and the evolution of OCV are simulated. A 1D fuel cell model incorporating chemical degradation of the membrane was presented by Shah, Ralph and Walsh [138]. The degradation model describes H2O2 formation, evolution of radicals from Fenton`s reaction and direct formation at the anode. Chemical degradation is assumed to proceed via unzipping of the backbone, side chain cleavage and decomposition of “molecule A”. The model results suggest that degradation propagates from the anode to the cathode and the influence of oxygen concentration at OCV, membrane thickness, load operation, temperature, water activity and reaction rate constants was investigated.

PUMA MIND - 303419


A second 1D continuum model was developed by Gumalla et al. [137] to simulate the membrane and the electrodes. Diffusive transport and reaction of crossover gases, radical formation and attack on the membrane with subsequent HF release under OCV conditions is incorporated into the model. Radicals can be formed and quenched at Pt-particles which are assumed to be uniformly distributed inside the membrane. The model results are fitted to the measured FERs data from the OCV-tests conducted in [140]. Influence of the Pt-particle size and the spacing between them, concentration of oxygen at the cathode, relative humidity and membrane thickness on the degradation rate was simulated. The first model to incorporate the effect of chemical degradation on the macroscopic properties of the membrane was developed by Coulon, Bessler and Franco [141,142]. In [141], an elementary kinetic model is presented simulating H2O2 formation at the anode, OH· formation via Fenton`s reaction and radical attack which leads to side chain cleavage and loss of sulfonic acid sites. It couples a new model for the membrane to the multiscale electrode models [36,43] of MEMEPhys® [40,143]. For the first time, a feedback between chemical degradation and the transport properties of the membrane is established.

2.2 Catalyst layer

2.2.1 Platinum dissolution

Darling and Meyers [144] proposed a model for platinum dissolution and movement in the MEA that claims to fit reasonably well with experimental data from literature. The model takes into account three electrochemical reactions involved in platinum dissolution. Bi and Fuller [145] proposed a similar model that includes also the formation of a Pt band in the membrane, but it appears that the model is not predicting Pt degradation rate very well, which leads to the conclusion that additional mechanisms are involved, such as nanoparticle coarsening. To enhance the stability of the Pt-based catalysts, Pt-alloys are currently developed. Platinum-Titanium (Pt-Ti), Platinum-Ruthenium (Pt-Ru), and Platinum-Cobalt (Pt-Co) are suggested candidates. Pt-Co showed increased stability compared to conventional Pt-catalysts. A multiscale mechanistic model has also been developed by Franco et al.,[35] for Pt-Co catalyst performance and degradation rate estimation. The model allows ranking the degradation rate among Pt-Co, Pt3Co, Pt and PtCo3 catalysts depending on the cell operating conditions.

2.2.2 Coarsening and coalescence

Electrochemical Ostwald ripening that occurs in PEMFC both involves transport of Pt2+ in the electrolyte and transport of electrons in the carbon support : led by the difference in the chemical potential of two close Pt particles, Pt2+ ions dissolve from the smaller particle and precipitate on the larger one [146]. Simultaneously, electrons are transported via the carbon support from the smaller particle to the larger one. Pt particle growth also occurs through coalescence following the migration of small particles on the carbon support. This mechanism is expected to be promoted by the presence of liquid water in the catalyst layers [147]. A modeling approach to describe the coalescence of catalyst crystallites is presented in [148].

PUMA MIND - 303419


A model describing the coupling between electrochemical double layer effects, Pt dissolution and sintering and associated particle size change kinetics was for the first time proposed in [149].

2.2.3 Carbon corrosion

Carbon corrosion is of major concern for long-term durability of PEMFC. Indeed, carbon corrosion leads to reduced carbon support and Pt particles can detach from their support, agglomerate or become isolated. Carbon corrosion mainly occurs at the cathode side during start-up and shutdown. The corrosion rate of an electrode peaks at the beginning of life (BoL) and decays with time. Also it has been shown that carbon corrosion leads to significant collapse and compaction of the CLs, thus blocking transport pathway for the reactant gas through the CLs. Moreover, carbon corrosion entails increase of hydrophilic properties of the support, thus promoting flooding of the electrodes [150]. Thus, carbon corrosion contributes to CL degradation through multiple mechanisms. Takeuchi et al. [151] developed a model for carbon corrosion that studies the influence of the operating conditions on the corrosion rate. Carbon corrosion is expected to be significantly decreased with recent reinforced membranes that are much more resistant to mechanical stress. Therefore, modeling of the carbon corrosion should be limited to severe conditions like high cathode potential and start-up/shutdown procedures.

2.2.4 Modeling the coupling between performance and degradation

In his recent review [40], Franco underlines that in current PEMFC models, the instantaneous feedback between performance and aging is not described. They describe which operating conditions enhance a given degradation process (e.g. carbon corrosion) but do not describe the impact of this degradation process on the instantaneous performance (a prediction of the transient behavior of the PEMFC MEA electrochemical observables, such as the cell potential degradation or durability, is not provided). According to Franco, the main drawbacks of current PEMFC degradation models are: • Aging mechanisms addressed in a separated (uncoupled) way. • Potentiostatic-potentiodynamic simulations: In most of the available kinetic degradation models, the Butler–Volmer electrode potential is the input variable, the output being a material corrosion rate and the cell current. Implicitly, it is assumed that the potential of the nanomaterials is equal to the external/macroscopic applied potential. • Use of the classical Butler–Volmer theory: The use of such a macroscopic Butler–Volmer theory cannot be really justified for describing electron transfer reactions on nanomaterials with an evolving structure.

Franco et al. proposed several multiscale models to overcome some of these issues, including descriptions of coupled electrochemical aging processes (e.g. Pt and PtxCoy oxidation/dissolution/ripening, carbon catalyst support corrosion, PEM degradation) [40,45,49,128,141,143,149,152–157]. The approach describes the feedback between the instantaneous performance and the intrinsic material aging processes, thus durability of individual components and of the entire cell can be predicted. Within this context, the model has provided interesting information on the competition of aging phenomena, by demonstrating that anode contamination by CO can be used, under some particular current-cycle conditions, to mitigate cathode carbon corrosion and PEM degradation

PUMA MIND - 303419


[154,158].

2.3 Gas diffusion layer

Among the various degradation mechanisms that lead to decreasing cell performance, degradation of the GDL plays a major role [159]. Degradation is mainly caused by the loss of hydrophobicity due to PTFE degradation [160]. It is commonly stated that loss of hydrophobicity entails flooding of the GDL, leading to a decrease of cell performance. Pauchet et al., [161] have recently studied GDL degradation by achieving a one-way coupling between Pore Network Models (PNM) and Performance Model (PM). In their models, gas diffusion is calculated by PNM for various amounts of liquid water and then serves as input for the PM. Loss of hydrophobicity of the gas diffusion layer can lead to cell performance losses comparable to the one observed experimentally. This suggests that degradation of the PTFE plays an important role in the overall degradation of the GDL. From a different point of view, an interesting analysis of the impact of water flooding on the diffusivity at cathode side is presented by Casalegno et al. [162]. The main outcome of this model consists in the definition of the GDL effective diffusivity as a function of a local flooding coefficient (representative of the flooding magnitude), temperature and water concentration. The results obtained with the model, in agreement with the data gathered during a dedicated experimental activity, highlight that the pore obstruction caused by flooding does strongly affect the GDL without an MPL, whereas not much if the MPL is included, with consequent less fluctuation of water transport.

2.4 Bipolar plates

BPPs have several important functions for FC operation, such as, uniform fuel and oxidant distribution over the active areas, water and heat removal, current transport from cell to cell, and leak proofing of reactants and coolants. There is a large body of experimental literature on BPP degradation via corrosion and contact resistance analysis [163]. There are various factors inherent to PEMFC operations that support BPP corrosion, including high humidity (presence of moisture), low pH due to acidic membrane, and elevated temperatures (as compared to “standard” environmental corrosion). Most importantly, as corrosion is an electrochemical process, it shows a strong interaction with electrochemical FC operation [164]. Despite broad experimental evidence, a detailed theoretical approach towards understanding BPP degradation is still missing. The pathway towards a BPP degradation model should be carried out in two steps, that is, (i) describe the physical origin of the interfacial contact resistance (without degradation) [165][166] and (ii) include corrosion through protective oxide growth and metal dissolution [167][168].

PUMA MIND - 303419


3 EXPERIMENTAL VALIDATION

3.1 In-situ methods

3.1.1 Polarization measurements

The most common method used to evaluate a PEMFC performance is the polarization curve measurement (fuel cell voltage vs.current or current density) under operating conditions, e.g. temperature, pressure, relative humidity (RH), reactant composition and flow rate. It can be performed under galvanostatic or potentiostatic control. The performance losses ignite already at OCV which is always lower than its theoretical value (or reversible potential) expressed by combining the Nernst equation for the two electrode reactions (eq. (2), for ideal gases) and corresponds to the Gibbs free energy change during the reaction of H2 with O2 to produce H2O. At 25oC it is 1.229 V for liquid H2O production. Apparently from eq. (2), it depends T and P.

22

2

2/1log

2

303.2229.1

HO

OHrev PP

P

F

RTOCV −= (2)

Also other factors affect the measured OCV, i.e. the internal current being considered the H2 crossover current (anodic current produced on the cathode by oxidation of molecular H2 permeated through the electrolyte) and the electrical short circuiting caused by small electronic conductivity of the electrolyte as well as the Pt catalyst surface state. The hydrogen cross over current influence can be calculated by the Butler-Volmer equation for the O2 reduction reaction (ORR) simplified to eq. (3), where o

ci and cb are the

exchange current density and Tafel slope, respectively, of the ORR and ni the internal current density.

−=

oc

ncrevmeas i

ibOCVOCV log (3)

However, the actual measured OCV is still lower than predicted by eq. (3) due to carbon support corrosion, H2O2 formation and surface Pt oxidation, see eq. (4) [169].

NHE vs.V 0.88=E ,2e+2H+PtOOH+Pt oPt/PtO

-2

+↔ (4) The reported mixed cathode potential is 1.06 V at standard conditions in acidic electrolytes and its dependence on O2 partial pressure is 15 mV/atm [170,171]. Upon closing the circuit, a voltage drop linked to the activation energy of the two half-cell reactions and the charge transfer to and from the electrodes, named activation overpotential, acη occurs expressed through the Butler-Volmer equation (for ORR) [172]:

−

−−=RT

Fna

RT

Fnii ccccccocc

ηηαexp

)1(exp (5)

where oci is the exchange current density (reflecting intrinsic kinetics), ca the electron

transfer coefficient, cn the electron number involved in the rate determining step (rds) of

ORR and cη the cathode electrode activation overpotential. At high overpotential (> 2.303RT/F ≈ 60 mV) it becomes the tafel equation:

PUMA MIND - 303419


=−=

oan

an

nan

oananan i

i

Fn

RTEE log

303.2

ααη (6)

with slope (revealing the reaction mechanism [173]), c

cc

bFn

RT =− )1(

303.2

α Similar expressions

hold for the much faster H2 oxidation reaction (HOR), oci ~ 10-5 A/cm2 of Pt [174]. The Tafel

slope for HOR on Pt is ca. 30 mV [175], implying that the rds is the Tafel reaction (H2 dissociative adsorption) while for ORR on Pt is ca. 60 mV [176], implying that the rds is the protonation of the molecularly adsorbed O2 [177]. Usually, the activation voltage losses are omitted to result in the following Tafel expression for the ORR [174]:

−=

occatPtc

cc iAuL

ib logη (7)

where cL (g/cm2) is the Pt loading, catelPt AAu /= is the Pt utilisation, elA (cm2 Pt/g Pt) the

Pt surface area participating in the electrochemical reaction and catA (cm2 Pt/g Pt) the surface area of the Pt catalyst. The limiting current density, iL, reads

δo

L

nFDCi = (8)

where D is the diffusion coefficient of the reactant species diffusing through a thin film of thickness δ and oC the concentration of the reacting species at the surface of the thin film. The mass transfer or concentration overpotential is:

−=

Lconc i

i

nF

RT1log

303.2η (9)

where n is the number of electrons exchanged in the reaction. In reality, mass transfer is an intercourse of convective flow inside the flow fields to also extent inside the GDM controlled by Knudsen diffusion especially in the micro-porous layer (MPL) and CL [172,178]. There is need for understanding what governs multiphase flow in the porous electrodes [68,179,180]. Semi-empirical account of I-V curve fits fairly well known [181–183]. Though they lack physical meaning [184] they are useful in practice. Accounting for all phenomena, we find the overall cell voltage loss

concnoccatPtc

nco

ccatPtc

ncrevcell rii

iAuL

iib

iAuL

ibOCVV η−+−

+−

−= )(loglog (10)

It was shown [174,185] that by using pure O2 as oxidant at high stoichiometry, mass transfer phenomena can be eliminated and kinetic parameters for the ORR can be extracted from eq. (10)[186].

3.1.2 EIS measurements

Electrochemical Impedance Spectroscopy (EIS) measurements [187–192] are often used to obtain fuel cell impedance (or admittance) as function of perturbation frequency upon applying a small AC perturbation [193–199]. Usually, EIS spectra are recorded at selected DC potentials representative for activation polarization (kinetic regime), the Ohmic regime and concentration polarization (mass transfer limitation regime) to probe the various physico-chemical phenomena and transport processes [195,198,200].

PUMA MIND - 303419


Eventually, fitting EIS data to an appropriate equivalent electrical circuit (EEC) model [199,201–219], transmission line (TL) model [199,201,202,220–225] or numerical model [226–234], for example, by means of a complex non-linear least square (CNLS) procedure [235,236] allows to obtain characteristic fuel cell parameters ( ohmic resistance, polarisation resistance, charge transfer resistance and double layer capacitance, etc) to consistently explain experimental findings eventually to guide development and optimisation. The estimation of such fitted parameters using EEC (or TL) models was recently called into doubt given the oversimplification of models [193,237]. It was argued that EIS measurements need to be complemented by other preferably transient techniques and ex-situ methods and not to aim at good fits of experimental data [238]. Nowadays, EIS measurements are seemingly routinely performed [202,204,221,238–247] and many EIS studies on fuel cells have been recently reviewed [193,195,196,198,199,237]. In essence, the effect of operating and test conditions such as reactant compositions, temperature including sub-zero conditions and humidity, flooding, flow field geometry, membrane type, microporous layer (MPL), contaminants, and catalyst surface chemistry are documented and the results are reasonably in line with results from other electrochemical performance measurements. Notably, the high frequency (100 Hz and above) and low frequency (100 mHz and below) regions usually represent respectively the charge transfer in the catalyst layer (CL) and mass transport in the GDL (electron conduction, water flow, convection / diffusion), catalyst and MPL and the membrane (proton conduction and water diffusion / permeation). At low current, charge transport in the CL dominates and at high current , mass transport phenomena are of greater significance often exhibited by an increase in the polarisation resistance [237]. Further, inductive features (positive imaginary part) at high frequencies are most likely due to wire inductances. At medium and low frequencies inductive loops are sometimes observed and often attributed to surface reactions (i.e. competitive adsorption) [237,248,249].

3.1.3 Current interruption method

In the current interruption (CI) method, the current is suddenly interrupted for a cell under load and the transient voltage response is measured [250,251]. The ohmic loss (IR drop) due to the membrane resistance almost instantaneously disappears with a duration of 0.5 and 10 ns [250] upon the interruption followed by activation and mass transfer losses [252,253]. It is used to measure overall FC stack ohmic resistance [251,254,255] and that of individual cells [256]. Usually, it is implemented in parallel to polarization curve measurements to monitor the cell voltage evolution at each current step [257] and to obtain IR-free or IR-corrected polarization curves and when fitted to the Butler–Volmer equation, concentration losses can be separated from activation losses thereby gaining extra information from exciting fundamental cell phenomena [258] with this time domain method [252]. CI was also used to study electrode processes [253,258,259]. Jaouen et al. [253] used it to investigate the effect of O2 partial pressure and humidity on the dynamic response of the cathode to determine the Tafel slope, exchange current density, double-layer capacitance (DLC), ionomer O2 solubility and the effective diffusivity [253]. The experimental curves were fitted to an analytical model assuming that the gas diffusion electrode (GDE) structure consists of spherical agglomerates [260]. Ceraolo et al. [261] developed a simplified dynamic model with empirical equations to reveal information on the proton concentrations and to discern GDE tortuosity and porosity. Sugiura et al. [259]

PUMA MIND - 303419


used a simple equivalent circuit of a single resistance, a parallel single resistance and a single capacitor to examine the effect of cell temperature, flooding phenomenon and gas composition/flow rate and linked seven parameters to the IR drop, anode/cathode gas diffusion resistance, reactive resistance, three-phase interfacial resistance and electrolyte properties. Similarly, Rubio et al. [258] diagnosed cathode flooding as accumulated amount of H2O from the membrane resistance, double-layer capacitance, diffusion resistance, diffusion time constant and charge transfer resistance. Generally, The CI method overestimates the IR drop by 20-30% relative to EIS [262] attributed to the non-uniform potential distribution within the porous electrodes, resulting in the appearance of ionic current within the electrolyte and electronic current within the electrode matrix with almost equal magnitude of ionic and electronic conductivity of the porous electrode. Small magnitude interruptions may avoid the main drawback of CI, its strong interference with cell operation [263]. Electrochemical parameter identification (EPI) is a model-based method for the characterisation of cell impedances in the time domain [264] with sinusoidal, step or rectangular excitations [264]. Evaluating EEC parameters, ohmic and polarisation resistance and double layer capacitance are determined [264,265]. The greatest advantages of this technique are short measurement time (1-5 s) and short calculation time (1-7 s) providing essentially drift-free measurements [264] usually corresponding to measurements in the frequency domain in the range of 0.5 Hz to 10 kHz [263,264].

3.1.4 Voltammetry measurements

Linear sweep voltammetry (LSV) and cyclic voltammetry (CV) are used for studying redox reactions to provide information on the steps [266]. In LSV, the electrode potential is ramped with constant rate between two limits, the anodic (high potential) and cathodic (low potential), or vice-versa, while electrode current is recorded. CV is commonly used in a three-electrode system comprised of the working electrode (WE), which is the electrode of interest, counter electrode (CE) that enables the current passage and a reference electrode (RE) against which the WE potential is measured. This configuration has been used in numerous studies of HOR [267–269], ORR [173,177,270], CO oxidation [271–273] and small organic molecules oxidation [271] in aqueous acidic electrolytes. The presence of adsorbed surface poisons [271] and the effectiveness of various catalyst combinations in tolerance to impurities [269,274,275] has been also studied. CV has been utilized to study the performance dependence on crystal lattice structure [276] and catalyst particle size effects on ORR [277] and COad electrooxidation activity [278]. The Pt electrochemical surface area (ECSA) [269,270,273,277], ORR mass and specific activity by Rotating Disk Electrode (RDE) [270,279,280] and catalyst stability [281–283] can be evaluated by CV. CV in PEMFCs is performed in the absence of RE, although there are examples of RE utilization within the MEA arrangement [274,284]. CV technique is quite valuable in determining the ECSA of Pt catalysed GDEs and catalyst utilisation within the GDE and MEA configuration [174]. The method involves the integration of the charge under the voltammetric peaks associated to H2 adsorption and/or desorption on Pt, corrected for double-layer charging (the capacitive component) [285]. Within GDEs, the whole Pt surface area is usually not accessible for the electrochemical reactions due to either insufficient contact with the electrolyte or electrical isolation of catalyst particles from each other by a non-conductive layer [286]. Therefore, the Pt ECSA evaluated by CV may be substantially smaller than the real catalyst surface area.

PUMA MIND - 303419


Sometimes, CV on a thin film of catalyst supported on a glassy carbon disk and immersed in aqueous acidic electrolytes is used [174,269,270,273,277,279,286] as well adsorption of probe molecules from gas phase, i.e., H2 and CO (chemisorption), N2 and Ar (physisorption), X-ray line broadening analysis or electron microscopy [287]. An alternative way of determining the ECSA is by CO stripping voltammetry, which lies in the oxidation of adsorbed CO and operating under the same principle as CV [288]. Another aspect making the ECSA evaluation ambiguous is CO bonding on the catalyst surface is either linear (on a-top sites) or multi-fold [289,290]. CO stripping voltammetry can be a reliable diagnostic tool to access information on PtRu anode electrode surface state by monitoring the CO oxidation peak potential, used as an indicator for degradation, during accelerated testing of reversal potential caused by fuel starvation [291]. Cyclic voltammetry variations have been used in half-cell and MEA configurations to test the catalyst stability by applying triangular and/or rectangular potential waves replicating Fuel Cell Vehicle (FCV) operation [281,292]. Half-cell measurements are usually performed at RT, while essentially elevated T (60-80oC) is a requisite to accelerate degradation. The latter can be achieved in MEA configuration, while N2 can be used instead of O2 for cathode catalyst stability tests [281]. It is hypothesized that the interruption or reduction of relatively corrosion-resistant steady-state oxide or hydroxide coverage of Pt during voltage transient to low potentials leads to accelerated production of oxidized Pt species mobile in the ionomer phase upon subsequent voltage cycling to high potentials [293]. On the other hand, the carbon support is vulnerable to surface oxidation and corrosion, especially at high potentials (> 1 V vs. RHE) causing the detachment of Pt particles from the support and the collapse of the catalyst layer pore structure [294]. The latter may occur on cathode through the reverse current mechanism [295] due to anode fuel starvation [296] and the formation of H2/Air fronts in the anode during start/stop [291] leading to high cathode potentials up to 1.5 V. For start/stop durability, triangular waves are adopted with fast sweeps and potential range 1 – 1.5 V vs. RHE, whereas rectangular and/or triangular waves with fast sweeps are used to simulate load-cycle operation with potential range 0.6 – 1 V vs. RHE [281]. Pt surface area losses were monitored using CV over the life of short stacks subjected to voltage cycling (0.6 - 1 V) with 20 mV/s under fully humidified H2/N2 at 80oC [293]. 46 wt% Pt/C catalyst was used in the cathode of the catalyst coated membranes CCMs. After 10000 cycles, 63% of the initial ECSA was lost, unacceptable for the automotive targets. 10000 cycles of triangular voltage wave (0.4 – 1 V) with sweep rates of 50 and 200 mV/s were also performed on MEAs using about 50 wt% Pt/C catalysts [297]. The durability of carbon supported PtCo alloy was also tested under supersaturated N2 as the cathode gas, involving voltage cycling (0.4 – 1 V) with 200 mV/s sweep rate for up to 50000 cycles [298]. The PtCo alloy catalyst exhibited greater resistance to sintering, followed by ECSA determination by CV, than Pt/C [297] under supersaturated RH conditions. At 80oC under fully humidified H2 in the anode and N2 in the cathode, 75% ECSA losses were noticed for a carbon supported PtCo alloy catalyst [299]. Both mass and specific activity drop with cycling, suggesting that specific activities of PtCo alloys seem to decrease, while those of Pt/C catalysts increase (particle size effect) leading to the effect that both types of catalysts approach the same values after extensive voltage cycling [300]. The carbon support corrosion was studied in CCM configuration for two different cathode catalysts by voltage cycling using rectangular wave with potential holds at the lower (0.6 V for 30 sec) and varied upper potential limit (0.9 – 1.4 V for varying dwell times) [301]. It was proposed that in O2 presence the place exchange of O2 in Pt crystal and the formation of higher order Pt oxides are increased during longer dwell times at upper voltage limit, resulting in greater de-stabilization of the Pt structure, surface roughening and greater Pt dissolution

PUMA MIND - 303419


during the subsequent downward cycle, opposing the commonly suspected anodic and/or chemical dissolution processes typical under potential hold experiments [302]. It was suggested that the loss of performance at 1 A/cm2 occurs at or just before the maximum Cdl is reached allowing the prediction of the onset of performance losses due to carbon support corrosion as a function of T, RH and upper potential limit [301]. The carbon support corrosion was also studied at MEA level for 46 wt% Pt/C (TKK) catalysts by applying square wave cycling (0.9 – 1.3 V vs. RHE, 30 sec at each potential limit) and triangular wave cycling (1 – 1.5 V vs. RHE, 500 mV/s) operated at 80oC with 100% RH H2/N2 [294]. By decreasing the lower voltage limit of the cycles, carbon corrosion becomes more severe due to more complete depassivation of Pt [294]. Linear Sweep Voltammetry (LSV) technique is widely used to determine the H2 crossover current and therefore the H2 crossover rate from anode to cathode through the polymer electrolyte in PEMFCs, on single cells [303] and on individual cells of a stack [304]. The flow and electrode configurations are similar to CV measurements. The technique is based on the principle of limiting current, i.e., the H2 permeating through the membrane is electro-oxidized on the electrode flushed with inert gas, serving as cathode under FC operation, generating a limiting current at sufficiently high electrode potential, 0.4 – 0.5 V vs. RHE, a potential range that is free from the effects of H2 adsorption/desorption on Pt [305]. The maximum swept potential value should be limited at 0.8 V vs. RHE to avoid Pt oxidation [306] and the sweep rate should be low enough, typically below 5 mV/s, to assure pseudo steady state conditions and minimize double layer capacitance contributions. It should be mentioned that under these conditions the H2 crossover rate measured represents the case at OCV rather than at load [307]. Since H2 crossover depends on the partial pressure of H2, under load the H2 crossover rate is expected to decrease from its OCV value due to fuel consumption [307,308]. In addition, LSV allows the detection of any internal electronic short circuit between anode and cathode [308]. Evidently, LSV can quantitatively separate the effects of crossover from electrical shorting, both of which lower the OCV as well as the cell voltage [178]. It was shown that correcting the measured current density by taking into consideration the H2 and O2 permeability [309] produces linear correlation to the Tafel line at the lower current densities, supporting the assumption that the crossed reactants are consumed electrochemically [308]. In fact, correction only for H2 crossover is enough to assess more accurately the activity of the catalysts [286], as O2 crossover in the anode doesn't affect significantly the anode overpotential due to fast H2 oxidation kinetics [310]. To develop and explore new materials for PEMs used in FCs, LSV can help by detecting the gas crossover rate. LSV has been employed to characterise H2 permeability of novel composite membranes such as H3PO4/Nafion®-PBI [311] and Nafion®/HPA [306]. Furthermore, the durability of nanocomposites included inside PEMs, acting as hydrogen peroxide scavengers, has been explored through LSV during OCV durability testing. Membranes incorporating MnO2 and MnO2/SiO2-SO3H [312] and ZrO2 nanoparticles [313] inside Nafion® membranes showed enhanced durability with minimal H2 crossover increase during OCV testing. On the contrary, Nafion® membranes showed increased crossover and even short circuit inside the PEM under similar testing conditions [313]. In a similar study, CeO2 nanoparticles were incorporated in the MEA structure, either as a uniform CeO2/Nafion® composite membrane or facing anode or cathode electrode only [314].

PUMA MIND - 303419


3.1.5 In-situ imaging methods

Small Angle Scattering (SAS) is used to study the water content in the membrane. The water content is investigated by the displacement of the ionomer peak and can be measured at various locations in the cell plane. Also, degradation of the membrane can be assessed by comparing data corresponding to bigger structures in the intensity spectra, before and after an aging test [315].

3.2 Ex-situ methods

3.2.1 Rotating Ring-Disc Experiments (RRDE)

RRDE experiments were first developed by Frumkin and Nekrasov in 1959 [316]. RRDE use a rotating disc electrode and a concentric ring electrode, separated by an insulator. The rotating electrode generates a convective flow of reagents, so that the chemical reaction that takes place at the ring electrode does not depend on diffusion mechanisms, but only on kinetics of the reaction [317].

RRDE experiments can give access to the intrinsic kinetics constants, thus being able to validate mesoscopic models.

Also, RRDE experiments can be used as an H2O2 probe [318].

3.2.2 Half-cell experiments

Half-cells are made of one GDL/MPL/C-Pt electrode surrounded by impregnated Nafion® and in contact with a liquid electrolyte. A gas inlet provides either H2 or O2 to the GDL. The liquid electrolyte is under agitation so that the concentration of reagents and products is homogeneous. Temperature of the electrolyte can be regulated. Half-cells aim at having a less complex system than the one of the whole PEMFC. Especially, anodic and cathodic contributions can be distinguished. 3.2.3 Ex-situ methods for degradation characterization

Scanning Electron Microscopy (SEM) has a nm resolution that is well adapted to the characterization of the structure of both CL and MPL .Transmission Electron Microscopy (TEM) offers a sub-nm resolution. Thus, it is well adapted to characterization of the catalyst nanoparticles. It enables determination of the catalyst particle size distribution and their growth with aging, due to both coalescence and Ostwald ripening. Given its sub-nm resolution, X-rays Diffraction (XRD) is the other main technique to study the catalyst size and distribution changes [319]. The advantage of the XRD is to give a good statistics of the size distribution of the MEA catalysts [320]. However the location of the particles in the catalyst layer is not available. X-rays Computed Tomography (CT) allows acquiring 2D tomographic images (virtual slices) around an axis of rotation. Due to its relatively low resolution (µm), CT enables investigation of larges structures and is very well adapted to the determination of the structure of the GDL [321]. Using a Focus-Ion-Beam (FIB), the sample is sliced and the cross-section is imaged using the associated SEM column. 3D reconstruction of a sample volume can be achieved by serial sectioning. This is a slice-and-view process. X-rays Photoelectron Spectroscopy (XPS) is a surface sensitive spectroscopy technic that enables determination of the

PUMA MIND - 303419


elemental composition of the surface of a material [159]. It is used for the investigation of surface composition of catalyst layers, GDL, and bipolar plates. In SEM or TEM, X-rays Energy Dispersion Spectroscopy (X-EDS) is used for the determination of the chemical composition of a sample.

4 CONCLUSION

Summarizing, it can be said that a lot of work has been done so far in the modeling of PEMFC components performance and degradation on all relevant scales as well as on the development of experimental methods suitable for model validation. The complexity of the available models on all scales increases consistently by taking into account more and more relevant mechanisms. In addition, progress is made in coupling on the one hand the models at different scales and on the other hand the performance and degradation models, but still a lot of work has to be done to pave the way to real predictive multiscale PEMFC models.

REFERENCES

[1] J. Greeley, I.E.L. Stephens, A.S. Bondarenko, T.P. Johansson, H.A. Hansen, T.F. Jaramillo, J. Rossmeisl, I. Chorkendorff, J.K. Nørskov, Nat. Chem. 1 (2009) 552.

[2] V. Tripkovic, E. Skúlasona, S. Siahrostamia, J.K. Nørskov, J. Rossmeisl, Electrochim. Acta 55 (2010) 7975.

[3] D.C. Ford, A.U. Nilekar, Y. Xu, M. Mavrikakis, Surf. Sci. 604 (2010) 1565.

[4] I.E.L. Stephens, A.S. Bondarenko, F.J. Perez-Alonso, F. Calle-Vallejo, L. Bech, T.P. Johansson, A.K. Jepsen, R. Frydendal, B.P. Knudsen, J. Rossmeisl, I. Chorkendorff, J. Am. Chem. Soc. 133 (2011) 5485.

[5] R. Jinnouchi, K. Kodama, T. Hatanaka, Y. Morimoto, Phys. Chem. Chem. Phys. 13 (2011) 21070.

[6] R.F. de Morais, P. Sautet, D. Loffreda, A.A. Franco, Electrochim. Acta 56 (2011) 10842.

[7] A. Eichler, J. Hafner, Phys. Rev. Lett. 79 (1997) 4481.

[8] A. Eichler, F. Mittendorfer, J. Hafner, Phys. Rev. B 62 (2000) 4744.

[9] P. Gambardella, Z. Sljivancanin, B. Hammer, M. Blanc, K. Kuhnke, K. Kern, Phys. Rev. Lett. 87 (2001) 4.

[10] Y. Xu, A. V Ruban, M. Mavrikakis, J. Am. Chem. Soc. 126 (2004) 4717.

[11] A. Michaelides, P. Hu, J. Am. Chem. Soc. 122 (2000) 9866.

PUMA MIND - 303419


[12] A. Michaelides, P. Hu, J. Am. Chem. Soc. 123 (2001) 4235.

[13] A. Michaelides, P. Hu, J. Chem. Phys. 114 (2001) 513.

[14] L. Qi, J. Yu, J. Lia, J. Chem. Phys. 125 (2006) 8.

[15] S. Kandoi, A.A. Gokhale, L.C. Grabow, J.A. Dumesic, M. Mavrikakis, Catal. Lett. 93 (2004) 93.

[16] Y. Sha, T.H. Yu, Y. Liu, B. V Merinov, W.A. Goddard, J. Phys. Chem. Lett. 1 (2010) 856.

[17] Y. Wang, P.B. Balbuena, J. Phys. Chem. B 108 (2004) 4376.

[18] T. Jacob, W.A. Goddard, J. Am. Chem. Soc. 126 (2004) 9360.

[19] Y. Wang, P.B. Balbuena, J. Phys. Chem. B 109 (2005) 14896.

[20] Z. Gu, P.B. Balbuena, J. Phys. Chem. C 111 (2007) 17388.

[21] M. Otani, I. Hamada, O. Sugino, Y. Morikawa, Y. Okamotode, T. Ikeshoji, Phys. Chem. Chem. Phys. 10 (2008) 3609.

[22] L. Ou, F. Yang, Y. Liu, S. Chen, J. Phys. Chem. C 113 (2009) 20657.

[23] J.M.M. de la Hoz, D.F. Leon-Quintero, P. Hirunsit, P.B. Balbuena, Chem. Phys. Lett. 498 (2010) 328.

[24] R. Callejas-Tovar, W. Liao, J.M.M. de la Hoz, P.B. Balbuena, J. Phys. Chem. C 115 (2011) 4104.

[25] J.K. Nørskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J.R. Kitchin, T. Bligaard, H. Jónsson, J. Phys. Chem. B 108 (2004) 17886.

[26] J.-S. Filhol, M. Neurock, Angew. Chem. Int. Ed. 45 (2006) 402.

[27] G.S. Karlberg, J. Rossmeisl, J.K. Nørskov, Phys. Chem. Chem. Phys. 9 (2007) 5158.

[28] A.U. Nilekar, M. Mavrikakis, Surf. Sci. 602 (2008) L89.

[29] M.J. Janik, C.D. Taylor, M. Neurock, J. Electrochem. Soc. 156 (2009) B126.

[30] J.A. Keith, G. Jerkiewicz, T. Jacob, ChemPhysChem 11 (2010) 2779.

[31] F. Calle-Vallejo, M.T.M. Koper, Electrochim. Acta 84 (2012) 3.

[32] E. Skulason, V. Tripkovic, M.E. Björketun, S. Gudmundsdottir, G. Karlberg, J. Rossmeisl, T. Bligaard, H. Jonsson, J.K. Nørskov, J. Phys. Chem. C 114 (2010) 18182.

[33] A.S. Bondarenko, I.E.L. Stephens, H.A. Hansen, F.J. Perez-Alonso, V. Tripkovic, T.P. Johansson, J. Rossmeisl, J.K. Nørskov, I. Chorkendorff, Langmuir 27 (2011) 2058.

PUMA MIND - 303419


[34] K.-Y. Yeh, M.J. Janik, J. Comput. Chem. 32 (2011) 3399.

[35] A.A. Franco, S. Passot, P. Fugier, E. Billy, N. Guillet, L. Guetaz, E. De Vito, S. Mailley, J. Electrochem. Soc. 156 (2009) B410.

[36] A.A. Franco, P. Schott, C. Jallut, B. Maschke, Fuel Cells 7 (2007) 99.

[37] I.C. Man, H.Y. Su, F. Calle-Vallejo, H.A. Hansen, J.I. Mart’nez, N.G. Inoglu, J. Kitchin, T.F. Jaramillo, J.K. Norskov, J. Rossmeisl, ChemCatChem 3 (2011) 1159.

[38] G.A. Tritsaris, J. Greeley, J. Rossmeisl, J.K. Norskov, Catal. Lett. 141 (2011) 909.

[39] F. Calle-Vallejo, J.I. Martínez, J.M. García-Lastra, P. Sautet, D. Loffreda, Angew. Chemie Int. Ed., Press (2014).

[40] A.A. Franco, in:, C. Hartnig, C. Roth (Eds.), Polym. Electrolyte Membr. Direct Methanol Fuel Cell Technol., Woodhead, UK, 2012.

[41] J.K. Norskov, T. Bligaard, J. Rossmeisl, C.H. Christensen, Nat. Chem. 1 (2009) 37.

[42] A.N. Frumkin, Acta Physicochim. URSS 18 (1943) 23.

[43] A.A. Franco, P. Schott, C. Jallut, B. Maschke, J. Electrochem. Soc. 153 (2006) A1053.

[44] A.A. Franco, in:, I. Kondov, G. Sutmann (Eds.), Multiscale Model. Methods Appl. Mater. Sci., CECAM & FZ Jülich, Germany, 2013.

[45] A.A. Franco, in:, R. Savinell, K.I. Ota, G. Kreysa (Eds.), Encycl. Appl. Electrochem., Springer, UK, 2013.

[46] R. He, S. Chen, F. Yang, B. Wu, J. Phys. Chem. B 110 (2006) 3262.

[47] C.O. Laoire, S. Mukerjee, K.M. Abraham, E.J. Plichta, M.A. Hendrickson, J. Phys. Chem. C 114 (2010) 9178.

[48] M.A. Quiroga, K.-H. Xue, T.-K. Nguyen, M. Tulodziecki, H. Huang, A.A. Franco, J. Electrochem. Soc. 161 (2014) E3302.

[49] A.A. Franco, in:, I. Kondov, G. Sutmann (Eds.), Multiscale Model. Methods Appl. Mater. Sci., CECAM & FZ Jülich, Germany, 2013.

[50] K. Reuter, in:, O. Deutschmann (Ed.), Model. Simul. Heterog. Catal. React. From Mol. Process to Tech. Syst., Weinberg: Wiley-VCH, 2011.

[51] A.P.J. Jansen, Comput. Phys. Commun. 86 (1995) 1.

[52] J. Lukkien, J. Segers, P. Hilbers, R. Gelten, a. Jansen, Phys. Rev. E 58 (1998) 2598.

[53] V. Rai, H. Pitsch, A. Novikov, Phys. Rev. E 74 (2006) 46707.

PUMA MIND - 303419


[54] V.P. Zhdanov, J. Electroanal. Chem. 607 (2007) 17.

[55] L. a. Abramova, A. V Zeigarnik, S.P. Baranov, E. Shustorovich, Surf. Sci. 565 (2004) 45.

[56] V.P. Zhdanov, B. Kasemo, Appl. Surf. Sci. 219 (2003) 256.

[57] D.A. Harrington, J. Electroanal. Chem. 420 (1997) 101.

[58] X. Wang, K.C. Lau, C.H. Turner, B.I. Dunlap, J. Power Sources 195 (2010) 4177.

[59] W. Goddard III, B. Merinov, A. Van Duin, T. Jacob, M. Blanco, V. Molinero, S.S. Jang, Y.H. Jang, Mol. Simul. 32 (2006) 251.

[60] M. Eikerling, A.A. Kornyshev, A.A. Kulikovsky, Electrochem. Eng. 5 (2007) 447.

[61] M. Eikerling, A.S. Ioselevich, A.A. Kornyshev, Fuel Cells 4 (2004) 131.

[62] K. Malek, M. Eikerling, Q. Wang, T. Navessin, Z. Liu, J. Phys. Chem. C 111 (2007) 13627.

[63] J.T. Wescott, Y. Qi, L. Subramanian, T.W. Capehart, J. Chem. Phys. 124 (2006) 134702.

[64] D. Damasceno Borges, K.. Malek, S. Mossa, G. Gebel, A.A. Franco, ECS Trans. 45 (2013) 101.

[65] A.A. Franco, A Physical Multiscale Model of the Electrochemical Dynamics in a Polymer Electrolyte Fuel Cell - An Infinite Dimensional Bond Graph Approach, PhD Thesis, Université Claude Bernard Lyon-1, France, 2005.

[66] K.-H. Xue, T.-K. Nguyen, A.A. Franco, J. Electrochem. Soc. 161 (2014) E3028.

[67] K. Malek, T. Mashio, M. Eikerling, Electrocatalysis 2 (2011) 141.

[68] T.E. Springer, T.A. Zawodzinski, S. Gottesfeld, J. Electrochem. Soc. 138 (1991) 2334.

[69] G. Maggio, V. Recupero, L. Pino, J. Power Sources 101 (2001) 257.

[70] P. Costamagna, Chem. Eng. Sci. 56 (2001) 323.

[71] S. Mazumder, J. V. Cole, J. Electrochem. Soc. 150 (2003) A1510.

[72] S. Um, C.Y. Wang, Proc. ASME Fuel Cell Div. (2000) 7.

[73] S. Um, C.Y. Wang, J. Power Sources 125 (2004) 40.

[74] Y. Wang, C.Y. Wang, Electrochim. Acta 50 (2005) 1307.

[75] Y. Wang, C.Y. Wang, Electrochim. Acta 51 (2006) 3924.

[76] D.M. Bernardi, M.W. Verbrugge, AIChE J. 37 (1991) 1151.

PUMA MIND - 303419


[77] D.M. Bernardi, M.W. Verbrugge, J. Electrochem. Soc. 139 (1992) 2477.

[78] M. Eikerling, J. Electrochem. Soc. 145 (1998) 2684.

[79] T.F. Fuller, J. Newman, J. Electrochem. Soc. 140 (1993) 1218.

[80] G.J.M. Janssen, J. Electrochem. Soc. 148 (2001) A1313.

[81] A.Z. Weber, J. Newman, J. Electrochem. Soc. 150 (2003) A1008.

[82] T. Thampan, S. Malhotra, H. Tang, R. Datta, J. Electrochem. Soc. 147 (2000) 3242.

[83] P. V. Schroeder, Z. Phys. Chem. 45 (1903) 75.

[84] E.A. Mason, A.P. Malinauskas, Gas Transp. Porous Media. Dusty-Gas Model (1983) Amsterdam.

[85] J. Fimrite, H. Struchtrup, N. Djilali, J. Electrochem. Soc. 152 (2005) A1804.

[86] M.J. Moran, H.N. Shapiro, Fundamentals of Engineering Thermodynamics, John Wiley & Sons, New York, 1996.

[87] J. Fimrite, B. Carnes, H. Struchtrup, N. Djilali, J. Electrochem. Soc. 152 (2005) A1815.

[88] J.J. Baschuk, X. Li, Int. J. Hydrogen Energy 35 (2010) 5095.

[89] P. Choi, N.H. Jalani, R. Datta, J. Electrochem. Soc. 152 (2005) E123.


[91] M. Eikerling, J. Electrochem. Soc. 153 (2006) E58.

[92] P. Schott, P. Baurens, J. Power Sources 156 (2006) 85.


[94] D. Harvey, J.G. Pharoah, K. Karan, J. Power Sources 179 (2008) 209.

[95] Y.W. Rho, S. Srinivasan, Y.T. Kho, J. Electrochem. Soc. 141 (1994) 2089.

[96] F.C. Cetinbas, S.G. Advani, A.K. Prasadz, J. Electrochem. Soc. 160 (2013) F750.

[97] A. Suzuki, U. Sen, T. Hattori, R. Miura, R. Nagumo, H. Tsuboi, N. Hatakeyama, A. Endou, H. Takaba, M.C. Williams, A. Miyamoto, Int. J. Hydrogen Energy 36 (2011) 2221.

[98] G. Wang, P.P. Mukherjee, C.-Y. Wang, Electrochim. Acta 52 (2007) 6367.

[99] M. El Hannach, Simulation et Analyse Des Mécanismes de Transfert Diphasique Dans Les Couches Actives Des Piles À Combustible PEMFC, Institut National Polytechnique de Toulouse, 2011.

PUMA MIND - 303419


[100] S. Strahl, A. Husar, A.A. Franco, Int. J. Hydr. Energy, Press (2014).


[102] H. Meng, Int. J. Hydrogen Energy 34 (2009) 5488.

[103] Oberholzer P., Boillat P., Siegrist R., Kästner A., Lehmann E. H., Scherer G. G., A. Wokaun, Electrochem. Commun. 20 (2012) 67.

[104] N.B. Carrigy, L.M. Pant, S. Mitra, M. Secanella, J. Electrochem. Soc. 160 (2013) F81.

[105] J. Hermann, C. Ziegler, J. Electrochem. Soc. 155 (2010) B1066.

[106] M. Rebai, M. Prat, J. Power Sources 192 (2009) 534.

[107] X.D. Niu, T. Munekata, S.A. Hyodoa, K. Suga, J. Power Sources 172 (2007) 542.

[108] Y. Tabe, Y. Lee, T. Chikahisa, M. Kozakai, J. Power Sources 193 (2009) 24.

[109] P.P. Mukherjee, Q. Kanga, C.Y. Wang, Energy Environ. Sci. 4 (2010) 346.

[110] K.S. Chen, M.A. Hickner, D.R. Noble, Int. J. Energy Res. 29 (2005) 1113.

[111] S.C. Cho, Y. Wang, K.S. Chen, J. Power Sources 206 (2012) 119.

[112] S.C. Cho, Y. Wang, K.S. Chen, J. Power Sources 210 (2012) 191.

[113] E.C. Kumbur, K. V. Sharp, M.M. Mench, J. Power Sources 161 (2006) 333.

[114] A. Theodorakakos, T. Ous, M. Gavaises, J.M. Nouri, J. Colloid Interface Sci. 300 (2006) 673.

[115] F.Y. Zhang, X.G. Yang, C.-Y. Wang, J. Electrochem. Soc. 153 (2006) A225.

[116] K.S. Chen, Proc. Fuel Cell (2008) 797.

[117] G. He, P. Ming, Z. Zhao, A. Abudula, Y. Xiao, J. Power Sources 163 (2007) 864.

[118] P. Polverino, A. Esposito, C. Pianese, Int. J. Hydrogen Energy 38 (2013) 8934.

[119] Z.H. Wang, C.Y. Wang, K.S. Chen, J. Power Sources 94 (2001) 40.

[120] M. Mortazavi, K. Tajiri, J. Power Sources 245 (2014) 236.

[121] A. Esposito, A. Montello, C. Pianese, Y.G. Guezennec, J. Fuel Cell Sci. Technol. 7 (2010) 031021.

[122] A. Esposito, C. Pianese, Y.G. Guezennec, J. Power Sources 195 (2010) 4149.

[123] A. Esposito, A.D. Montello, C. Pianese, Y.G. Guezennec, J. Power Sources 195 (2010) 2691.

PUMA MIND - 303419


[124] J.T. Pukrushpan, A.G. Stefanopoulou, H. Peng, Control of Fuel Cell Power Systems: Principles, Modeling, Analysis, and Feedback Design, Springer, 2004.

[125] A.A. Franco, RSC Adv. 3 (2013) 13027.

[126] A.A. Franco, K.-H. Xue, ECS J. Solid State Sci. Technol. 2 (2013) M3084.

[127] D. Sheppard, R. Terrell, G. Henkelman, J. Chem. Phys. 128 (2008) 134106.

[128] K. Malek, A.A. Franco, J. Phys. Chem. B 115 (2011) 8088.

[129] A.A. Franco, A Multiscale Modeling Framework for the Transient Analysis of Electrochemical Power Generators - From Theory to the Engineering Practice, Habilitation to become Research Director (H.D.R.) manuscript, Université Claude Bernard-Lyon I (France), 2010.

[130] A.A. Franco, Multiscale Modeling of Direct Alcohol Fuel Cells, Book Chapter in: Direct Alcohol Fuel Cells Technologies: Research and Development, E. Gonzalez and H. Corti, Springer, USA, 2013.

[131] http://www.modeling-electrochemistry.com/, (2013).

[132] A. Collier, H. Wang, X. Ziyuan, J. Zhang, D. Wilkinson, Int. J. Hydrogen Energy 31 (2006) 1838.

[133] T. Xie, C. a. Hayden, Polymer (Guildf). 48 (2007) 5497.

[134] L. Gubler, S.M. Dockheer, W.H. Koppenol, J. Electrochem. Soc. 158 (2011) B755.

[135] C. Chen, T.F. Fuller, J. Electrochem. Soc. 156 (2009) B1218.

[136] C. Chen, T.F. Fuller, Polym. Degrad. Stab. 94 (2009) 1436.

[137] M. Gummalla, V. V. Atrazhev, D. Condit, N. Cipollini, T. Madden, N.Y. Kuzminyh, D. Weiss, S.F. Burlatsky, J. Electrochem. Soc. 157 (2010) B1542.

[138] A.A. Shah, T.R. Ralph, F.C. Walsh, J. Electrochem. Soc. 156 (2009) B465.

[139] S. Kundu, M.W. Fowler, L.C. Simon, R. Abouatallah, N. Beydokhti, J. Power Sources 183 (2008) 619.

[140] T. Madden, D. Weiss, N. Cipollini, D. Condit, M. Gummalla, S. Burlatsky, V. Atrazhev, J. Electrochem. Soc. 156 (2009) B657.

[141] R. Coulon, W.G. Bessler, A.A. Franco, ECS Trans. 25 (2010) 259.

[142] R. Coulon, Modeling of Chemical Degradation in Polymer Electrolyte Membrane Fuel Cells, PhD thesis, Université de Grenoble, 2012.

PUMA MIND - 303419


[143] A.A. Franco, R. Coulon, R.F. de Morais, S.K. Cheah, A. Kachmar, M.A. Gabriel, ECS Trans. 25 (2009) 65.

[144] R.M. Darling, J.P. Meyers, J. Electrochem. Soc. 150 (2003) A1523.

[145] W. Bi, T.F. Fuller, J. Power Sources 178 (2008) 188.

[146] P. Parthasarathy, A. V. Virkar, J. Power Sources 234 (2013) 82.

[147] T. Ous, C. Arcoumanis, J. Power Sources 240 (2013) 558.

[148] E. Ruckenstein, B. Pulvermacher, J. Catal. 29 (1973) 224.

[149] A.A. Franco, M. Tembely, J. Electrochem. Soc. 154 (2007) B712.

[150] J.D. Fairweather, D. Spernjak, A.Z. Weber, D. Harvey, SilviaWessel, D.S. Hussey, D.L. Jacobson, K. Artyushkova, R. Mukundan, R.L. Borup, J. Electrochem. Soc. 160 (2013) F980.

[151] N. Takeuchi, T.F. Fuller, J. Electrochem. Soc. 155 (2008) B770.

[152] S.K. Cheah, O. Lemaire, P. Gélin, A.A. Franco, ECS Trans. 25 (2010) 275.

[153] A.A. Franco, S. Passot, P. Fugier, C. Anglade, E. Billy, L. Guétaz, N. Guillet, E. De Vito, S. Mailley, J. Electrochem. Soc. 156 (2009) B410.

[154] A.A. Franco, M. Guinard, B. Barthe, O. Lemaire, Electrochim. Acta 54 (2009) 5267.

[155] A. a. Franco, M. Gerard, J. Electrochem. Soc. 155 (2008) B367.

[156] A.A. Franco, K. Malek, Prep. (2014).

[157] A.A. Franco, J. Electrochem. Soc., Submitt. (2014).

[158] V. Parry, G. Berthomé, J.-C. Joud, O. Lemaire, A.A. Franco, J. Power Sources 196 (2011) 2530.

[159] M. Schulze, N. Wagner, T. Kaz, K.A. Friedrich, Electrochem. Acta 52 (2007) 2328.

[160] R. Borup, J. Meyers, B. Pivovar, Y.S. Kim, R. Mukundan, N. Garland, D. Myers, M. Wilson, F. Garzon, D. Wood, P. Zelenay, K. More, K. Stroh, T. Zawodzinski, J. Boncella, J.E. McGrath, M. Inaba, K. Miyatake, M. Hori, K. Ota, Z. Ogumi, S. Miyata, A. Nishikata, Z. Siroma, Y. Uchimoto, K. Yasuda, K.-I. Kimijima, N. Iwashita, Chem. Rev. 107 (2007) 3904.

[161] J. Pauchet, M. Prat, P. Schott, S.P. Kuttanikkad, Int. J. Hydrogen Energy 37 (2012) 1628.

[162] A. Casalegno, F. Bresciani, G. Groppi, R. Marchesi, J. Power Sources 196 (2011) 10632.

[163] R.A. Antunes, M.C.L. Oliveira, G. Ett, V. Ett, Int. J. Hydrogen Energy 35 (2010) 3632.

PUMA MIND - 303419


[164] D.A. Shores, G.A. Deluga, in:, W. Vielstich, H.A. Gasteiger, A. Lamm (Eds.), Handb. Fuel Cells, Wiley, New York, 2003, pp. 273–285.

[165] P. Zhou, C.W. Wu, G.J. Ma, J. Power Sources 159 (2006) 1115.

[166] V. Mishra, F. Yang, R. Pitchumani, J. Fuel Cell Sci. Technol. 1 (2004) 2.

[167] P. Marcus, ed., Corrosion Mechanisms in Theory and Practice, CRC Press, 2011.

[168] R.S. Munn, ed., Computer Modeling in Corrosion, ASTM International, 1992.

[169] J. Zhang, Y. Tang, C. Song, H. Wang, J. Power Sources 163 (2006) 532.

[170] J.P. Hoare, J. Electrochem. Soc. 109 (1962) 858.

[171] R. Thacker, J.P. Hoare, J. Electroanal. Chem. Interfacial Electrochem. 30 (1971) 1.

[172] D.-Y. Zhang, X. Yuan, Z.-F. Ma, in:, H. Wang, X.-Z. Yuan, H. Li (Eds.), PEM Fuel Cell Diagnostic Tools, CRC Press, Boca Raton, 2012.

[173] J.X. Wang, N.M. Markovic, R.R. Adzic, J. Phys. Chem. B 108 (2004) 4127.

[174] H.A. Gasteiger, W. Gu, R. Makharia, M.F. Mathias, B. Sompalli, in:, W. Vielstich, A. Lamm, H.A. Gasteiger (Eds.), Wiley, Chichester, 2003, pp. 593–610.

[175] B.N. Grgur, N.M. Markovic, P.N. Ross, Electrochim. Acta 43 (1998) 3631.

[176] A. Damjanovic, J.O. Bockris, Electrochim. Acta 11 (1966) 376.

[177] N.M. Markovic, R.R. Adzic, B.D. Cahan, E.B. Yeager, J. Electroanal. Chem. 377 (1994) 249.

[178] S.S. Kocha, in:, W. Vielstich, A. Lamm, H.A. Gasteiger (Eds.), Wiley, Chichester, 2003, pp. 538–565.

[179] M.B. Ji, Z.D. Wei, Energies 2 (2009) 1057.

[180] K. Jiao, X. Li, Prog. Energy Combust. Sci. 37 (2011) 221.

[181] J. Kim, S.M. Lee, S. Srinivasan, J. Electrochem. Soc. 142 (1995) 2670.

[182] L. Pisani, G. Murgia, M. Valentini, B. D’Agurno, J. Power Sources 108 (2002) 192.

[183] M. V. Williams, H.R. Kunz, J.M. Fenton, J. Electrochem. Soc. 152 (2005) A635.

[184] C. Siegel, Energy 33 (2008) 1331.

[185] H.A. Gasteiger, J.E. Panels, S.G. Yan, J. Power Sources 127 (2004) 162.

PUMA MIND - 303419


[186] M.F. Mathias, R. Makharia, H.A. Gasteiger, J.J. Conley, T.J. Fuller, C.J. Gittleman, S.S. Kocha, D.P. Miller, C.K. Mittelsteadt, T. Xie, S.G. Yan, P.T. Yu, Electrochem. Soc. INTERFACE 14 (2005) 24.

[187] C. Gabrielli, Use and Applications of Electrochemical Impedance Techniques, Farnborough, 1997.

[188] C. Gabrielli, Identification of Electrochemical Processes by Frequency Response Analysis, Farnborough, 1998.

[189] N.D. Cogger, N.J. Evans, An Introduction to Electrochemical Impedance Measuerments, Farnborough, 1999.

[190] E. Barsoukov, J.R. Macdonald, eds., Impedance Spectroscopy: Theory, Experiment, and Applications, 2nd ed., Wiley & Sons, Ltd, Chichester, 2005.

[191] M.E. Orazem, B. Tribollet, Electrochemical Impedance Spectroscopy, 1st ed., John Wiley & Son, Hoboken, New Jersey, 2008.

[192] V.F. Lvovich, Impedance Spectroscopy: Applications to Electrochemical and Dielectric Phenomena, John Wiley & Son, Hoboken, New Jersey, 2012.

[193] S.M.R. Niya, M. Hoorfar, J. Power Sources 240 (2013) 281.

[194] M.A. Travassos, V. V. Lopes, R.A. Silva, A.Q. Novais, C.M. Rangel, Int. J. Hydrogen Energy 38 (2013) 7684.

[195] J. Zhang, J. Wu, H. Zhang, J. Zhang, eds., PEM Fuel Cell Testing and Diagnosis, Elsevier, Amsterdam, 2013.

[196] A. Arvay, E. Yli-Rantala, C.-H. Liu, X.-H. Peng, P. Koski, L. Cindrella, P. Kauranen, P.M. Wilde, A.M. Kannan, J. Power Sources 213 (2012) 317.

[197] V.M. Huang, S.-L. Wu, M.E. Orazem, N. Pébèreb, B. Tribollet, V. Viviera, Electrochim. Acta 56 (2011) 8048.

[198] X.-Z. Yuan, C. Song, H. Wang, J. Zhang, eds., Electrochemical Impedance Spectroscopy in PEM Fuel Cells Fundamentals and Applications, Springer, London, Dordrecht, Heidelberg, New York, 2010.

[199] X. Yuan, H. Wang, J.C. Sun, J. Zhang, Int. J. Hydrogen Energy 32 (2007) 4365.

[200] A. Lasia, in:, B.E. Conway, J. Bockris, R.E. White (Eds.), Mod. Asp. Electrochem., Kluwer Academic/Plenum Publishers, New York, NY, 1999, pp. 143–248.

[201] J.J. Quintana, A. Ramos, I. Nuez, Math. Probl. Eng. 2013 (2013) 807034.

[202] H. Nara, T. Momma, T. Osaka, Electrochim. Acta 113 (2013) 720.

[203] N. Wagner, Bulg. Chem. Commun. 44 (2012) 371.

PUMA MIND - 303419


[204] M. Maidhily, N. Rajalakshmi, K.S. Dhathathreyan, Int. J. Hydrogen Energy 36 (2011) 12352.

[205] R.A. Antunes, M.C.L. de Oliveira, G. Ett, Int. J. Hydrogen Energy 36 (2011) 12474.

[206] J. Lee, J.-H. Lee, W. Choi, K.-W. Park, H.-Y. Sun, J.-H. Oh, J. Power Sources 195 (2010) 6001.

[207] M.-C. Yang, C.-H. Hsueh, J. Electrochem. Soc. 153 (2006) A1043.

[208] B. Andreaus, G. Scherer, Solid State Ionics 168 (2004) 311.

[209] B. Andreaus, A.J. McEvoy, G. Scherer, Electrochim. Acta 47 (2002) 2223.

[210] M. Ciureanu, S.D. Mikhailenko, S. Kaliaguine, Catal. Today 82 (2003) 195.

[211] N. Wagner, M. Schulze, Electrochim. Acta 48 (2003) 3899.

[212] N. Wagner, J. Appl. Electrochem. 32 (2002) 859.

[213] P.D. Beattie, F.P. Orfino, V.I. Basura, K. Zychowska, J. Ding, C. Chuy, J. Schmeisser, S. Holdcroft, J. Electroanal. Chem. 503 (2001) 45.

[214] M. Ciureanu, R. Roberge, J. Phys. Chem. B 105 (2001) 3531.

[215] C.A. Schiller, F. Richter, E. Gülzow, N. Wagner, Phys. Chem. Chem. Phys. 3 (2001) 2113.

[216] N. Wagner, E. Gülzow, J. Power Sources 127 (2004) 341.

[217] M. Ciureanu, H. Wang, J. New Mater. Electrochem. Syst. 3 (2000) 107.

[218] E. Gülzow, M. Schulze, N. Wagner, T. Kaz, R. Reissner, G. Steinhilber, A. Schneider, J. Power Sources 86 (2000) 352.

[219] M. Ciureanu, H. Wang, Z. Qi, J. Phys. Chem. B 103 (1999) 9645.

[220] J. Bisquert, Phys. Chem. Chem. Phys. 2 (2AD) 4185.

[221] O. Reid, F.S. Saleh, E.B. Easton, Electrochim. Acta, Vol. 114, 30 December 2013, Pages 278-284 114 (2013) 278.

[222] J. Hou, W. Song, H. Yu, Y. Fu, L. Hao, Z. Shao, B. Yi, J. Power Sources 176 (2008) 118.

[223] M. Kheirmand, A. Asnafi, Int. J. Hydrogen Energy 36 (2011) 13266.

[224] M. Eikerling, A.A. Kornyshev, E. Lust, J. Electrochem. Soc. 152 (2005) E24.

[225] M. Eikerling, A.A. Kornyshev, J. Electroanal. Chem. Interfacial Electrochem. 475 (1999) 107.

PUMA MIND - 303419


[226] J. Deseure, J. Power Sources 178 (2008) 323.

[227] H. Kuhn, A. Wokaun, G.G. Scherer, Electrochim. Acta 52 (2007) 2322.

[228] C. Montella, J.P. Diard, J. Electroanal. Chem. 625 (2009) 156.

[229] C. Montella, J.P. Diard, J. Electroanal. Chem. 623 (2008) 29.

[230] S. Chevalier, D. Trichet, B. Auvity, J.C. Olivier, C. Josse, M. Machmouma, Int. J. Hydrogen Energy 38 (2013) 11609.

[231] A.A. Kulikovsky, J. Electroanal. Chem. 669 (2012) 28.

[232] A.A. Kulikovsky, M. Eikerling, J. Electroanal. Chem. 691 (2013) 13.

[233] J. Mainka, G. Maranzana, A. Thomas, J. Dillet, S. Didierjean, O. Lottin, Fuel Cells 12 (2012) 848.

[234] J.J. Baschuk, X. Li, J. Power Sources 86 (2000).

[235] B.A. Boukamp, Solid State Ionics 169 (2004) 66.

[236] M.E. Orazem, B. Tribollet, Electrochim. Acta 53 (2008) 7360.

[237] P.M. Gomadam, J.W. Weidner, Int. J. Energy Res. 29 (2005) 1133.

[238] S.K. Roy, H. Hagelin-Weaver, M.E. Orazem, J. Power Sources 196 (2011) 3736.

[239] G. Maranzana, J. Mainka, O. Lottin, J. Dillet, A. Lamibrac, A. Thomas, S. Didierjean, Electrochim. Acta, Vol. 83, 30 Novemb. 2012, Pages 13-27 83 (2012) 13.

[240] G. Mousa, F. Golnaraghi, J. DeVaal, A. Young, J. Power Sources 246 (2014) 110.

[241] S.J. Andreasen, J.R. Vang, K.S. Knudsen, Int. J. Hydrogen Energy 36 (2011) 9815.

[242] M. Mamlouk, K. Scott, Electrochim. Acta 56 (2011) 5493.

[243] M. Maidhily, N. Rajalakshmi, K.S. Dhathathreyan, Renew. Energy 51 (2013) 79.

[244] M.S. Kondratenko, M.O. Gallyamov, A.R. Khokhlov, Int. J. Hydrogen Energy 37 (2012) 2596.

[245] S. Cruz-Manzo, R. Chen, P. Rama, J. Hydrog. Energy 38 (2013) 1702.

[246] S. Cruz-Manzo, R. Chen, J. Electroanal. Chem. 694 (2013) 45.

[247] S. Cruz-Manzo, R. Chen, J. Electroanal. Chem. 702 (2013) 45.

[248] A. Sadkowski, Electrochim. Acta 49 (2004) 2653.

PUMA MIND - 303419


[249] A. Sadkowski, Solid State Ionics 176 (2005) 1987.

[250] F.N. Buchi, A. Marek, G.G. Scherer, J. Electrochem. Soc. 142 (1995) 1895.

[251] L.J. Bonville, H.R. Kunz, Y. Song, A. Mientek, M. Williams, A. Ching, J.M. Fenton, J. Power Sources 144 (2005) 107.

[252] M.A. Rubio, A. Urquia, S. Dormido, in:, H.L. H. Wang X.-Z. Yuan (Ed.), PEM Fuel Cell Diagnostic Tools, CRC Press, Boca Raton, 2012.

[253] F. Jaouen, G. Lindbergh, K. Wiezell, J. Electrochem. Soc. 150 (2003) A1711.

[254] F. Weng, B. Jou, A. Su, S.H. Chan, P. Chi, J. Power Sources 171 (2007) 179.

[255] Y. Park, J. Caton, J. Power Sources 179 (2008) 584.

[256] T. Mennola, M. Mikkola, M. Noponen, T. Hottinen, P. Lund, J. Power Sources 112 (2002) 261.

[257] R. O’Hayre, S.-W. Cha, W. Clella, F.B. Priz, Fuel Cells Fundamentals, John Wiley & Son, Hoboken, New Jersey, 2005.

[258] M.A. Rubio, A. Urquia, S. Dormido, J. Power Sources 171 (2007) 670.

[259] K. Sugiura, M. Yamamoto, Y. Yoshitani, K. Tanimoto, A. Daigo, T. Murakami, J. Power Sources 157 (2006) 695.

[260] F. Jaouen, G. Lindbergh, J. Electrochem. Soc. 150 (2003) A1699.

[261] M. Ceraolo, C. Miulli, A. Pozio, J. Power Sources 113 (2003) 131.

[262] K.R. Cooper, M. Smith, J. Power Sources (2006).

[263] M.A. Rubio, A. Urquia, R. Kuhn, S. Dormido, J. Power Sources 183 (2008) 118.

[264] M.A. Danzer, E.P. Hofer, J. Power Sources 183 (2008) 55.

[265] W.-Y. Chang, Int. J. Electr. Power Energy Syst. 53 (2013) 584.

[266] G. Bontempelli, R. Toniolo, in:, J. Garche (Ed.), Encycl. Electrochem. Power Sources, Elsevier, Amsterdam, 2009, pp. 643–654.

[267] J.X. Wang, T.E. Springer, R.R. Adzic, J. Electrochem. Soc. 153 (2006).

[268] S. Chen, A. Kucernak, J. Phys. Chem. B 108 (2004) 13984.

[269] H.A. Gasteiger, N.M. Markovic, P.N. Ross, J. Phys. Chem. 99 (1995) 8290.

[270] N.M. Markovic, H.A. Gasteiger, B.N. Grgur, P.N. Ross, J. Electroanal. Chem. 467 (1999) 157.

PUMA MIND - 303419


[271] N.M. Markovic, P.N. Ross Jr., Surf. Sci. Rep. 45 (2002) 117.

[272] N.M. Markovic, T.J. Schmidt, B.N. Grgur, H.A. Gasteiger, R.J. Behm, P.N. Ross, J. Phys. Chem. B 103 (1999) 8568.

[273] V. Climent, R. Gomez, J.M. Orts, J.M. Feliu, J. Phys. Chem. B 110 (2006) 11344.

[274] G. Papakonstantinou, F. Paloukis, A. Siokou, S.G. Neophytides, J. Electrochem. Soc. 154 (2007) B989.

[275] S. Mukerjee, J. McBreen, J. Electrochem. Soc. (1999).

[276] Z. Jusys, J. Kaiser, R.J. Behm, Electrochim. Acta 47 (2002) 3693.

[277] K.J.J. Mayrhofer, B.B. Blizanac, M. Arenz, V.R. Stamenkovic, P.N. Ross, N.M. Markovic, J. Phys. Chem. B 109 (2005) 14433.

[278] A.N. Gavrilov, E.R. Savinova, P.A. Simonov, V.I. Zaikovskii, S.V. Cherepanova, G.A. Tsirlina, V.N. Parmon, Phys. Chem. Chem. Phys. 9 (2007) 5476.

[279] T.J. Schmidt, H.A. Gasteiger, G.D. Stab, P.M. Urban, D.M. Kolb, R.J. Behm, J. Electrochem. Soc. 145 (1998) 2354.

[280] G. Zhang, Z.-G. Shao, W. Lu, F. Xie, H. Xiao, X. Qin, B. Yi, Appl. Catal. B Environ. 132-133 (2013) 183.

[281] A. Ohma, K. Shinohara, A. Iiyama, T. Yoshida, A. Daimaru, ECS Trans. 41 (2011) 775.

[282] K. Ono, Y. Yasuda, K. Sekizawa, N. Takeuchi, T. Yoshida, M. Sudoh, Electrochim. Acta 97 (2013) 58.

[283] A. Kumar, V. Ramani, J. Electrochem. Soc. 160 (2013) F1207.

[284] Z. Siroma, R. Kakitsubo, N. Fujiwara, T. Ioroi, S.-I. Yamazaki, K. Yasuda, J. Power Sources 156 (2006) 284.

[285] S. Trasatti, O.A. Petrii, Pure Appl. Chem. 63 (1991) 711.

[286] H.A. Gasteiger, S.S. Kocha, B. Sompalli, F.T. Wagner, Appl. Catal. B Environ. 56 (2005) 9.

[287] S. Mukerjee, J. Appl. Electrochem. 20 (1990) 537.

[288] H.N. Dihn, X.M. Ren, F.H. Garzon, P. Zelenay, S. Gottesfeld, J. Electroanal. Chem. 491 (2000) 222.

[289] A. Cuesta, A. Couto, A. Rincon, M.C. Perez, A. Lopez-Cudero, C. Gutierrez, J. Electroanal. Chem. 586 (2006) 184.

[290] M. Heinen, Y.X. Chen, Z. Jusys, R.J. Behm, Electrochim. Acta 52 (2007) 5634.

PUMA MIND - 303419


[291] A. Taniguchi, T. Akita, K. Yasuda, Y. Miyazaki, J. Power Sources 130 (2004) 42.

[292] U.S. Department of Energy, Office of Energy Efficiency & Renewable Energy, Vehicle Technology Office, U.S. DRIVE (Driving Research and Innovation for Vehicle Efficiency and Energy Sustainability), Fuel Cell Technical Team Roadmap, Fuel Cell Technical Team, 2013.

[293] P.J. Ferreira, G.J. la O’, Y. Shao-Horn, D. Morgan, R. Makharia, S. Kocha, H.A. Gasteiger, J. Electrochem. Soc. 152 (2005) A2256.

[294] Y.-C. Park, K. Kakinuma, M. Uchida, D.A. Tryk, T. Kamino, H. Uchida, M. Watanabe, Electrochim. Acta 91 (2013) 195.

[295] C.A. Reiser, L. Bregoli, T.W. Patterson, J.S. Yi, J.D. Yang, M.L. Perry, T.D. Jarvi, Electrochem. Solid-State Lett. 8 (2005) A273.

[296] Y. Yu, G. Wang, Z. Tu, Z. Zhan, M. Pan, Electrochim. Acta 71 (2012) 181.

[297] S. Ohyagi, T. Matsuda, Y. Iseki, T. Sasaki, C. Kaito, J. Power Sources 196 (2011) 3743.

[298] S. Ohyagi, T. Sasaki, Electrochim. Acta 102 (2013) 336.

[299] S. Chen, H.A. Gasteiger, K. Hayakawa, T. Tada, Y. Shao-Horn, J. Electrochem. Soc. 157 (2010) A82.

[300] K.C. Neyerlin, R. Srivastava, C.F. Yu, P. Strasser, J. Power Sources 186 (2009) 261.

[301] A.P. Young, V. Colbow, D. Harvey, E. Rogers, S.. Wessel, J. Electrochem. Soc. 160 (2013) F381.

[302] M. Matsumoto, T. Miyazaki, H. Imai, J. Phys. Chem. C 115 (2011) 11163.

[303] K. Panha, M. Fowler, X.-Z. Yuan, H. Wang, Appl. Energy 93 (2012) 90.

[304] J. Wu, X.-Z. Yuan, J.J. Martin, H. Wang, D. Yang, J. Qiao, J. Ma, J. Power Sources 195 (2010) 1171.

[305] M. Inaba, T. Kinumoto, M. Kiriake, R. Umebayashi, A. Tasaka, Z. Ogumi, Electrochim. Acta 51 (2006) 5746.

[306] V. Ramani, H.R. Kunz, J.M. Fenton, J. Membr. Sci. 232 31 232 (2004) 31.

[307] X. Cheng, J. Zhang, Y. Tang, C. Song, J. Shen, D. Song, J. Power Sources 167 (2007) 25.

[308] S.S. Kocha, J.D. Yang, J.S. Yi, AIChE J. 52 (2006) 1916.

[309] T. Sakai, H. Takenaka, N. Wakabayashi, Y. Kawami, E. Torikai, J. Electrochem. Soc. 132 (1985) 1328.

[310] A.Z. Weber, J. Electrochem. Soc. 155 (2008) B521.

PUMA MIND - 303419


[311] Y. Zhai, H. Zhang, Y. Zhang, D. Xing, J. Power Sources 169 (2007) 259.

[312] D. Zhao, B.L. Yi, H.M. Zhang, H.M. Yu, J. Memb. Sci. 346 (2010) 143.

[313] S. Xiao, H. Zhang, C. Bi, Y. Zhang, Y. Ma, X. Li, H. Zhong, J. Power Sources 195 (2010) 8000.

[314] S. Xiao, H. Zhang, C. Bi, Y. Zhang, H. Dai, Z. Mai, X. Li, J. Power Sources 195 (2010) 5305.

[315] G. Gebel, O. Diat, S. Escribano, R. Mosdale, J. Power Sources 179 (2008) 132.

[316] A.N. Frumkin, L.N. Nekrasov, J. Electrochem. Soc. 126 (1959) 115.

[317] E.O. Barnes, G.E.M. Lewis, S.E.C. Dale, F. Marken, R.G. Compton, Analyst 137 (2011) 1068.

[318] T.J. Schmidt, U.A. Paulus, H.A. Gasteiger, R.J. Behm, J. Electroanal. Chem. 508 (2001) 41.

[319] M. Loster, D. Balzar, K.A. Friedrich, J. Garche, J. Phys. Chem. C 111 (2007) 9583.

[320] R.L. Borup, J.R. Davey, F.H. Garzon, D.L. Wood, M.A. Inbody, J. Power Sources 163 (2006) 76.

[321] J.P. James, H.-W. Choi, J.G. Pharoah, Int. J. Hydrogen Energy 37 (2012) 18216.

D01.01: Summary report on modelling and experimental ... · PUMA MIND - 303419 D01.01 - Summary...

Documents

Transcript of D01.01: Summary report on modelling and experimental ... · PUMA MIND - 303419 D01.01 - Summary...