reading

Cover
Title Page
Preface
1. The Rationale for the Book and Introduction to Its Contents
About the Editor
List of Contributors
1 The Evolution of Fuel Cells and Their Components
1. 1.1 Overview: A Personal Perspective of Recent Developments
2. 1.2 Basics of Fuel Cell Operation
3. 1.3 Types of Fuel Cells
4. 1.4 Low Temperature Fuel Cells: Components
5. 1.5 Summary
6. Acknowledgments
7. References
2 Degradation Mechanism of Perfluorinated Membranes
1. 2.1 Introduction
2. 2.2 Fluoride Release Rate
3. 2.3 Nuclear Magnetic Resonance
4. 2.4 Fourier Transform Infrared Spectroscopy
5. 2.5 Electron Spin Resonance
6. 2.6 Conclusions
7. Acknowledgments
8. References
3 Ranking the Stability of Perfluorinated Membranes to Attack by Hydroxyl Radicals
1. 3.1 Introduction
2. 3.2 The Chemical Stability of Perfluorinated Ionomers
3. 3.3 Electron Spin Resonance Studies of PFSAs Exposed to Hydroxyl Radicals
4. 3.4 Conclusions
5. Acknowledgments
6. References
4 Stabilization of Perfluorinated Membranes Using Ce³⁺ and Mn²⁺ Redox Scavengers
1. 4.1 Introduction
2. 4.2 Oxidant Chemistry
3. 4.3 Degradation Mechanisms of PFSA
4. 4.4 Mitigation of Chemical Degradation by Redox Quenchers
5. 4.5 Conclusions
6. Acknowledgments
7. References
5 Hydrocarbon Proton Exchange Membranes
1. 5.1 Introduction
2. 5.2 Radical Intermediates in Fuel Cells
3. 5.3 Hydrocarbon Membranes
4. 5.4 Chemical Stabilization by Antioxidants
5. 5.5 The Challenge of Regeneration
6. 5.6 Concluding Remarks
7. References
6 Stabilization of Perfluorinated Membranes Using Nanoparticle Additives
1. 6.1 Nanoparticle Additives as a Stabilizer for Perfluorinated Membranes
2. 6.2 CeO₂ and Modified CeO₂ Nanoparticles as FRSs
3. 6.3 Platinum‐Supported Ceria as FRS
4. 6.4 Manganese Oxide and Manganese Oxide Composite as FRSs
5. 6.5 Metal Nanoparticles as FRSs
6. 6.6 Experimental Techniques for the Detection of Free Radicals and Measurement of the Membrane Degradation Rates
7. 6.7 Conclusions
8. Acknowledgments
9. References
7 Degradation Mechanisms in Aquivion^® Perfluorinated Membranes and Stabilization Strategies
1. 7.1 Introduction
2. 7.2 Properties of SSC Ionomers
3. 7.3 Properties of Aquivion Ionomers
4. 7.4 The Need for High Stability of PFSA Membranes
5. 7.5 PFSA Membrane Degradation in Fuel Cell
6. 7.6 Generation of Radical Species in the Fuel Cell Environment
7. 7.7 Degradation Studies on Aquivion Membranes
8. 7.8 Stabilization Procedures on Aquivion Membranes
9. 7.9 Conclusions
10. References
8 Anion Exchange Membranes
1. 8.1 Introduction
2. 8.2 Chemical Degradation Mechanisms
3. 8.3 Synthetic Approaches
4. 8.4 Conclusions
5. Acknowledgments
6. References
9 Profiling of Membrane Degradation Processes in a Fuel Cell by 2D Spectral–Spatial FTIR
1. 9.1 Introduction
2. 9.2 Optical Images of Nafion Cross Sections
3. 9.3 Line Scan Maps of the Membranes
4. 9.4 FTIR Spectra of Nafion MEAs
5. 9.5 Abstraction of a Fluorine Atom on a Carbon in the Nafion Main Chain by H^•
6. 9.6 Conclusions
7. Acknowledgments
8. References
10 Quantum Mechanical Calculations of the Degradation in Perfluorinated Membranes Used in Fuel Cells
1. 10.1 Introduction
2. 10.2 Computational Methods
3. 10.3 Results and Discussion
4. 10.4 Summary
5. Acknowledgments
6. References
Index
End User License Agreement

List of Tables

Chapter 03
1. Table 3.1 Hyperfine splittings of DMPO adducts in the ionomer dispersions.
2. Table 3.2 Reaction rates of fluorinated ionomers with hydroxyl radicals.
Chapter 04
1. Table 4.1 Thermodynamic properties of major fuel cell radicals [7, 8].
2. Table 4.2 Bond enthalpies of fuel cell species [8].
3. Table 4.3 Kinetic parameters for fuel cell reactions of H^• and HO^• [9].
4. Table 4.4 Kinetic parameters for H₂ and H₂O₂ reactions with HO^• at 95 °C [14, 15].
5. Table 4.5 Concentration of reactive species employed in steady‐state kinetic analysis of 50 µm, EW 1100 membrane.
6. Table 4.6 Hydroxyl radical consumption sinks at a function of [Ce³⁺].
7. Table 4.7 Summary of cerium stack performance parameters for 80 °C/85% RH out polarization curve.
8. Table 4.8 MEA composition and OCV durability at 110 °C/15% RH.
9. Table 4.9 Operation and composition space of OCV experimental design.
Chapter 05
1. Table 5.1 Standard electrode potentials relative to the normal hydrogen electrode (pH 0).
2. Table 5.2 Comparison of kinetic simulation (cf. Ref. 25) results of HO^• formation and attack on PFSA and PSSA ionomer, respectively, using a H₂O₂ concentration of 0.5 mM.
Chapter 10
1. Table 10.1 Products of Nafion^® degradation with the experimental references and theoretical mechanisms of their formation.
2. Table 10.2 The DFT predicted reaction energetics involved in the formation of OH and H radical on a Pt surface.
3. Table 10.3 DFT predicted reaction barriers^a (eV) for Nafion^® degradation of different end groups in the main chain and side chain for four environments.

List of Illustrations

Chapter 01
1. Figure 1.1 Schematic of fuel cell single unit cell with an overview of processes in PEM single cell.
2. Figure 1.2 Some representative membrane structures. (a) Typical shot side‐chain structure (3M ionomer, Aquivion^®), (b) Nafion^® generic structure, and (c) SDAPP (sulfonated Diels–Alder poly(phenylene) structure).
3. Figure 1.3 Electrospinning of membranes. (a) Physical process of electrospinning, (b) membrane formation process, and (c) micrographs of electrospun membranes.
4. Figure 1.4 Performance of a fuel cell operating with NPMC cathodes, polarization, and power curves.
5. Figure 1.5 The fuel cell catalyst layer. Top, left side: TEM micrograph of C‐supported Pt particles. Right side: schematic of structure showing components. Bottom: blowup of functional schematic emphasizing the meeting of molecules, ions, and electrons at a site.
Chapter 02
1. Figure 2.1 Nafion^®, 3M, and Aquivion^® membranes.
2. Figure 2.2 Average FRR.
3. Figure 2.3 Fluoride evolution from model compounds as a function of test time (top, full curve; bottom, low concentration range).
4. Figure 2.4 (a) FRR as a function of MEA construction and (b) FRR as a function of anode catalyst geometric surface area for a range of catalysts, both carbon supported and unsupported.
5. Figure 2.5 FRR as a function of time for different Mn²⁺ loadings in a 3M ionomer with equivalent weight 825.
6. Figure 2.6 Evolution of ¹⁹F NMR (a) and ¹H NMR (b) spectra of Nafion^® 112 aged at 80 °C and 80% relative humidity.
7. Figure 2.7 ¹⁹F NMR (left) and ¹³C NMR spectra (right) of Nafion^® membranes, before (A) and after (B) in situ tests. Asterisks indicate spinning sidebands.
8. Figure 2.8 ¹⁹F NMR spectra of the extracts after hydrogen peroxide vapor exposure tests. Left, Nafion^® 112 (conventional). Right, CSNafion^® NRE‐212 (chemically stabilized). (1) 100 °C and 30% RH, (2) 90 °C and 30% RH, (3) 90 °C and 70% RH, (4) 90 °C and 100% RH, and (5) reference membrane.
9. Figure 2.9 ¹⁹F NMR spectra of (A) fresh H⁺‐Nafion^® and (B–D) Fe²⁺‐Nafion^® after durability test in 30% H₂O₂ at 80 °C for: (B) 0.5 days, (C) for 1.5 days, and (D) 2.5 days. Symbols a–g′ show the positions of corresponding fluorine atoms in the chemical structure of Nafion^®.
10. Figure 2.10 ATR‐FTIR spectra of dry Nafion^® before (a) and after (b) cathode‐side degradation tests in the range of 900–1400 cm⁻¹.
11. Figure 2.11 ATR‐FTIR spectra of degraded membranes in the 1350–900 cm⁻¹ range.
12. Figure 2.12 FTIR spectra of the H₂O₂/metal ions after the Nafion^® ionomer was immersed in the solution for various period of time. The inset is the decomposed F⁻ content.
13. Figure 2.13 FTIR transmission spectrum of 3.5 µm thick Nafion^® membrane. Inset: fingerprint region spectra of the ionized (solid line) and associated acid (dashed line) forms of the membrane.
14. Figure 2.14 FTIR transmission spectra of the proton (dashed line) and ammonium salt (solid line) forms of a 3.5 µm thick Nafion^® membrane. Inset: fingerprint region of both spectra.
15. Figure 2.15 FTIR transmission spectra of the K⁺ salt forms of 50 µm thick NRE212 membranes after 96 h treatment with 30 ppm hydrogen peroxide vapor at the indicated RH values. The untreated membrane is shown as a dot–dash trace. Inset: SEC average molecular weight values of treated and untreated membranes.
16. Figure 2.16 Experimental ESR spectrum at 260 K of the “quartet” radical in Nafion^®/Fe(II)/H₂O₂/H₂O.
17. Figure 2.17 Evolution of the “quintet” ESR spectrum at 200 K as a function of UV irradiation time in Nafion^® 10% neutralized by CuCO3 and evacuated for 1 day prior to irradiation. (b) Transformation of the “quintet” into the “quartet” ESR spectrum upon annealing above 200 K.
18. Figure 2.18 Experimental and simulated ESR spectra at 77 K in UV‐irradiated Nafion^®/Fe(II)/H₂O₂ and in Nafion^®/Fe(III).
19. Figure 2.19 Spin trapping by 5,5‐dimethyl‐1‐pyrroline‐N‐oxide (DMPO).
20. Figure 2.20 Schematic drawing of the in situ fuel cell.
21. Figure 2.21 ESR spectra of DMPO adducts detected at the cathode side in the in situ FC when the spin trap was added just before CCV operation.
22. Figure 2.22 (a) Time evolution of the ESR spectra of DMPO adducts when the spin trap was added after indicated time at the cathode side, CCV operation in the in situ FC. DMPO/OH adduct is indicated by ○. (b) Experimental and simulated ESR spectra of adducts when the spin trap was added after 360 min of operation.
23. Figure 2.23 Anode side, comparison of H₂ (a) and D₂ (b). Appearance of the DMPO/H adduct on CCV and OCV conditions and of the DMPO/OOH adduct only on OCV conditions. The spin trap was added just before FC operation.
24. Figure 2.24 Experimental and simulated ESR spectra of adducts when DMPO was added at the cathode side (CCV (a) and OCV (b) operations) and at the anode side (CCV (c) and OCV (d) operations). Membrane thickness: 25 mm. In the experimental spectra (a) and (c), the signals from DMPO/H are indicated by ▼ and for DMPO/OOH by ▲.
25. Figure 2.25 Experimental and simulated ESR spectra of adducts when DMPO was added at the cathode side (CCV (a) and OCV (b) operations) and at the anode side (CCV (c) and OCV (d) operations). Membrane thickness: 100 mm.
26. Figure 2.26 Relative intensity of the DMPO/H adduct as a function of membrane thickness. The margin of error is ~5%.
Chapter 03
1. Figure 3.1 Weight of side chain fragments in the membranes generated during hydrogen peroxide vapor test.
2. Figure 3.2 Fluoride release rate in mg F⁻/g of polymer for standard extruded Hyflon^® ion, stabilized extruded Hyflon^® ion, Nafion^® N112 (extruded), and Nafion^® NR112 (cast) membranes.
3. Figure 3.3 Solid‐state ¹⁹F NMR spectra of Nafion^® 112 and Hyflon^® after 12 h ex situ Fenton tests with 0.0002 M Fe(II) in 30 v/v% H₂O₂.
4. Figure 3.4 Fluoride release rate for Nafion^® 112 and Hyflon^® membranes after Fenton tests with 0.0002 M Fe(II) in 30 v/v% H₂O₂.
5. Figure 3.5 Plot of (V/v) − 1 versus the concentration ratio [PFEESA]/[DMPO] at pH 7. The value of k_PFEESA, (3.7 ± 0.1) × 10⁶ M⁻¹s⁻¹, was determined from the slope = k_PFEESA/k_DMPO = 8.13 × 10⁻⁴, using k_DMPO = 4.6 × 10⁹ M⁻¹s⁻¹. (V and v are initial trapping rates of HO^• radicals in the absence and presence of PFEESA.)
6. Figure 3.6 Experimental ESR spectra of DMPO adducts detected at 300 K in aqueous dispersions of Nafion^® (a), 3M (b), and Aquivion^® (c) membranes containing 3% H₂O₂ and corresponding simulated spectra. UV irradiation times were 5 min for Nafion^® and 3M membranes and 60 min for the Aquivion^® membrane.
7. Figure 3.7 (a) ESR spectra at 300 K of DMPO adducts recorded in the absence (top) and in the presence (bottom) of the indicated methanol concentrations after 30 s of UV irradiation. Arrow indicates the peak of the DMPO/OH spectrum that was monitored, while asterisks show the carbon‐centered adduct, DMPO/CCR, derived from methanol. (b) Determination of the k_MeOH/k_DMPO ratio using the data shown in (a). The value of k_DMPO, 3.6 × 10⁹ M⁻¹s⁻¹, was based on the literature value for k_MeOH, 9.7 × 10⁸ M⁻¹s⁻¹.
8. Figure 3.8 ESR spectra of the DMPO/OH adduct for the indicated Nafion^® concentrations in the range 0–10 wt% are shown in the inset of (a); the downward arrow indicates the peak that was monitored in the time scans. The intensity of the ESR signal from the DMPO/OH adduct as a function of polymer concentration is shown in the time scans: (a) Nafion^®, (b) CSNafion^®, (c) 3M membrane, and (d) Aquivion^®. The thicker black line in each case shows the results in the absence of ionomers; data collected in the presence of ionomers were normalized to the intensity measured in the absence of each ionomer after 30 s of UV irradiation at 300 K.
9. Figure 3.9 Plots of (V/v) − 1 versus the concentration ratio [M]/[DMPO] for Nafion^® (), CSNafion^® (), 3M (∆), and Aquivion^® () using the inhibition data of the DMPO/OH adduct shown in Figure 3.2. Note that the polymer membrane concentration [M], is expressed in terms of the number of backbone carbons. Data for each system were fitted by assuming that the straight lines extrapolate to the origin.
10. Figure 3.10 (a) Effect of Ce(III) on the intensity of the DMPO/CCR adduct in Nafion^® dispersions exposed to hydroxyl radicals. (b) Intensity of the DMPO/CCR adduct in Nafion^® dispersions as a function of the degree of neutralization.
11. Figure 3.11 Ce(III) as a competitor for hydroxyl radicals: inhibition of DMPO/OH adduct formation by the indicated concentration of Ce(III) ions in water solution during UV irradiation at 300 K.
Chapter 04
1. Scheme 4.1 PFSA unzipping mechanism.
2. Scheme 4.2 Generalized degradation mechanisms of PFSA.
3. Figure 4.1 The 50 cm² OCV chemical durability tests of as‐received and cerium‐stabilized 50 µm Nafion^® NRE212CS MEAs: H₂/Air. 95 °C/50% RH. Solid lines represent voltage and symbolled lines fluoride release rate.
4. Scheme 4.2a Fuel cell chemistry of Ce⁴⁺ reduction.
5. Scheme 4.3 Steady‐state kinetic treatment of PFSA unzipping.
6. Figure 4.2 Fractions of HO^• consumed by reactions of Table 4.3 as a function of Ce³⁺ concentration according to the steady‐state kinetic analysis of Scheme 4.3 and parameters of Table 4.5. The practical Ce³⁺ mitigation zone is roughly bound between 2 and 15% proton replacement values.
7. Figure 4.3 Hydrogen atom abstraction rates from PFSA carboxylic acid end groups as functions of Ce³⁺ and Mn²⁺ concentrations according to steady‐state kinetic analysis of Scheme 4.3 and parameters of Table 4.5.
8. Figure 4.4 Anode and cathode fluoride concentrations emitted from cerium loading stack summarized in Table 4.7. Ce = 0 parts failed after 288 h of operation and were replaced.
9. Figure 4.5 FTIR spectra of Ce³⁺ –doped NRE212 membranes after 95‐23‐95% RH H₂O₂ vapor test.
10. Figure 4.6 In situ ESR spectra of DMPO spin trap adducts obtained from MEA/H and MEA/Ce when DMPO was applied to the cathode electrode immediately prior to fuel cell operation.
11. Figure 4.7 Log(FRR₂₀₀) values obtained at 110 °C/15% RH for MEAs summarized in Table 4.8. The dot‐dash curve represents predicted log(FRR₂₀₀) values for [Ce]/[Fe] ratios between 6 and 14.
12. Figure 4.8 Log(FRR₂₀₀) values of MEAs (Ce = 3.5 µg cm⁻², Fe = 1.0 µg cm⁻²) obtained under various OCV conditions.
13. Figure 4.9 Summary of empirical OCV degradation model derived from experimental data conducted over the design space summarized in Table 4.9.
14. Figure 4.10 FRR profiles produced during OCV degradation tests conducted at 95 °C/50% RH.
15. Figure 4.11 ICP cerium and iron concentrations found at three locations within a large‐scale MEA before and after durability testing.
16. Figure 4.12 Voltage degradation profiles at 1.5 A cm⁻² obtained from cerium‐mitigated stack membranes with and without intramembrane Pt deposits. Inset: steady‐state OCV FRR values obtained at 110 °C/25% RH.
Chapter 05
1. Figure 5.1 Estimated reactivity of polymer substrate with HO^•, obtained from a kinetic simulation of radical formation and reaction (Ref. 25). Ionomer types: PFSA, perfluoroalkylsulfonic acid and PSSA, poly(styrenesulfonic acid). The data points indicated by a symbol correspond to [H₂O₂] = 0.5 mM, a typical concentration under fuel cell operating conditions. Top: rate of attack on polymer by HO^•, r(polymer), as a function of the formation rate of HO^•, r(HO^•). The area‐specific rate in mol/(cm²·h) is determined from the reaction rate (in M/s) assuming a membrane thickness of 50 µm. Bottom: fate of the formed HO^• at the reference H₂O₂ concentration of 0.5 mM. In case of PFSA, most of the HO^• react with H₂ (to form HOO^•), whereas in case of PSSA almost all of the HO^• react with the polymer.
2. Figure 5.2 Mechanism of HO^• attack on PSSA oligomer (cf. Refs. 25, 33) as a model system for hydrocarbon ionomers, involving H‐abstraction and addition of HO^• to the ring as initial reaction. At low pH, the radical adduct (compound B) can undergo acid‐catalyzed H₂O elimination, which increases the yield of benzyl radicals (compound C), which is an intermediate to chain scission. A fraction of the HO‐adduct can yield a stable hydroxylated product (compound D).
3. Figure 5.3 Classes of sulfonated hydrocarbon‐based polymers, each encompassing a large number of variations, considered as proton‐conducting materials for fuel cell application. Typical “main chain” groups X include ─O─ (ether), ─(C═O)─ (ketone), or ─(SO₂)─ (sulfone). R and Ar represent an aliphatic and aromatic moiety, respectively.
4. Figure 5.4 Top: history plot of a fuel cell test at a constant current density of 0.2 A/cm² employing a sulfonated multiblock poly(arylene ether sulfone ketone) (SPESK) membrane. Bottom: history plot of two single‐cell tests (H₂/O₂) with radiation‐grafted membrane (cell data at 0.5 A/cm²) at 80 °C. Filled symbols: steady‐state operation at 0.5 A/cm². open symbols: test using dynamic load profile. Cell resistance is governed by the ohmic resistance of the membrane.
5. Figure 5.5 Comparison of chemical stabilities of hydrocarbon membrane (BPSH, biphenyl sulfone hydrocarbon) and PFSA membrane (Nafion^®) in ex situ and in situ tests [53]. Top: ex situ test in a Fenton solution (partially Fe²⁺‐exchanged Nafion^® immersed in 3% H₂O₂) at 80 °C for 24 h (BPSH) and 96 h (Nafion^®), respectively. Bottom: in situ single‐cell accelerated stress test (H₂/O₂) at open‐circuit voltage (OCV) at 100 °C and 25% relative humidity. The crossover current is a measure of membrane integrity.
6. Figure 5.6 Mechanism of oxidative aging of a hydrocarbon polymer via a radical‐initiated chain reaction (autoxidation, peroxidation). Cycle 1 is the chain propagation reaction, while cycle 2 is where new radicals are created and injected into the reaction sequence.
7. Figure 5.7 Left: mechanism of regenerative scavenging of HO^• based on the redox cycling of cerium ions. Ce⁴⁺ reacts with H₂O₂ present in the fuel cell within ~1 ms to restore Ce³⁺. Right: a similar mechanism based on the phenol/phenoxyl radical couple is conceivable. Yet, kinetic limitations are expected.
8. Figure 5.8 Top: structure of polymer‐bound phenol‐type antioxidant (tyramine) prepared by radiation‐induced grafting (x ≅ 0.7) [79]. Bottom: increase of ohmic resistance (HFR, high frequency resistance) and loss of ion exchange capacity (IEC) after an in situ accelerated stress test (H₂/O₂, OCV, 80 °C, 100% RH, 4 h) of membrane with antioxidant, GMA(Tyr), reference with hydrolyzed comonomer, GMA(diol), and styrene‐only‐based membrane.
9. Figure 5.9 In living cells exposed to oxygen, oxidative stress is predominantly caused by the superoxide radical, O₂^•−, which is created in processes involving oxygen (photosynthesis, cell respiration) as a result of side reactions via electron leakage to O₂. Important defense measures of the cell involve superoxide dismutase, peroxidases, and catalases, which decompose O₂^•− to harmless end products.
10. Figure 5.10 Left: superoxide dismutase (SOD), exemplified by Fe‐SOD, and its working principle to remove O₂^•− in living cells (ping‐pong mechanism). Right: electrode potential of Fe‐SOD: Refs. 91, 92; electrode potentials involving O₂^•−: Ref. 16.
11. Figure 5.11 Top: sketch showing the regeneration of spent antioxidant via reduction using electrons from the fuel cell anode (E ≅ 0 V vs. NHE). Bottom: SEM cross section of a proton‐conducting membrane with one‐sided incorporation of an electronically conducting polymer (polypyrrole, PPy) to transport electrons from the anode into the bulk of the membrane without electrically shorting anode and cathode.
12. Figure 5.12 Top: cyclic voltamogramm of a Nafion^® 212 membrane doped with hydroquinone (H₂Q) in contact with a carbon electrode, showing hydroquinone oxidation/quinone reduction peaks around 0.55 V. Bottom: cyclic voltamogramm using the same configuration, but a Nafion^® membrane with PPy composite layer facing the working electrode. The peak shift between oxidation and reduction currents indicates significant transport limitations in the PEM/PPy composite layer. Scan rate: 0.1 V/s.
Chapter 06
1. Figure 6.1 FERs (μmol F⁻h⁻¹cm⁻²) measured for Nafion^® and Nafion^®–CeO₂ MEAs during accelerated durability testing. Conditions during the accelerated test: cell temperature, 90 °C; humidifier saturator temperatures, 60 °C; relative humidity (RH), 30%; anode gas, H₂; cathode gas, O₂; flow rates, 0.2 slpm at the anode and the cathode; duration of the accelerated test, 24 h.
2. Figure 6.2 TEM micrograph of CeO₂ nanoparticles (T48).
3. Figure 6.3 Concentration of oxygen vacancies in CeO₂ and CeO₂‐doped nanoparticles as a function of grain size.
4. Figure 6.4 Concentration of oxygen vacancies in CeO₂ and CeO₂‐doped nanoparticles as a function of Ce³⁺ surface concentration.
5. Figure 6.5 Macroscopic rate of composite membrane degradation (in an operating fuel cell) as a function of Ce³⁺ surface concentration of the added oxide. CeO₂ loading: 0.15 mg cm⁻² membrane.
6. Figure 6.6 (a) Normalized XANES spectra of CeO₂ and N‐doped CeO₂ samples. (b) The Fourier‐transformed EXAFS oscillations for all the CeO₂ samples.
7. Figure 6.7 In situ fluorescence spectroscopic estimation of regenerative ROS scavenging efficacy within the PEM of an operating PEFC upon addition of (i) 2 wt% of undoped commercial CeO₂, (ii) 0.5 wt% of undoped high‐surface‐area CeO₂, (iii) 5 wt% nitrogen‐doped commercial CeO₂, and (iv) 1 wt% of nitrogen‐doped high‐surface‐area CeO₂. All experiments were performed at 80 °C and 75% RH under flowing H₂/O₂ at a flow rate of 0.1 slpm. Catalyst loadings at the cathode and anode were 0.4 mg_Pt cm⁻² and 0.2 mg_Pt cm⁻².
8. Figure 6.8 In situ fluorescence spectroscopic estimation of regenerative ROS scavenging efficacy within the PEM of an operating PEFC upon addition of (i) 5 wt% N‐doped commercial CeO₂ and (ii) 1 wt% N‐doped high‐surface‐area CeO₂. The experiments were performed at 90 °C and 50% RH under flowing H₂/O₂ at a flow rate of 0.1 slpm.
9. Figure 6.9 Total FER measured for composite Nafion^® membrane‐based MEAs during accelerated durability tests at 90 °C and 30% RH.
10. Figure 6.10 Nyquist plots showing electrochemical impedance spectroscopy results for Nafion^®, MnO₂/Nafion^®, and MnO₂/SiO₂─SO₃/Nafion^® composite membranes at 80 °C. High frequency intercept with the real axis (left side) corresponds with membrane resistance.
11. Figure 6.11 Fuel cell performances for Nafion^®, MnO₂/Nafion^®, and MnO₂/SiO₂─SO₃/Nafion^® composite membranes at 80 °C and 100% RH of the inlet gases.
12. Figure 6.12 Comparison of fuel cell performance of Nafion^® and MnO₂/SiO₂─SO₃/Nafion^® composite membranes at increasingly lower RHs.
13. Figure 6.13 Fluoride emission rate (FER) of Nafion^® membrane and composite membrane with MnO₂ particles and MnO₂/SiO₂─SO₃H nanocomposites in Fenton test.
14. Figure 6.14 Comparisons of durability with different MEAs in OCV test.
15. Figure 6.15 FER measurements for Nafion^®, MnO₂/Nafion^®, and MnO₂/SiO₂─SO₃H/Nafion^® membranes under OCV accelerated PEM durability testing.
16. Figure 6.16 Total fluoride emission rates of composite membrane‐based MEAs (Nafion^®–metal nanoparticles) during accelerated durability testing at 90 °C and 30% relative humidity of the inlet gases.
17. Figure 6.17 Total fluoride emission rates for composite membrane‐based MEAs (Nafion^®–metal nanoparticles supported on silica) during accelerated durability testing at 90 °C and 30% relative humidity of the inlet gases.
Chapter 07
1. Figure 7.1 (a) and (b) The “SSC” monomer: perfluoro‐5‐sulfonylfluoride‐3‐oxa‐1‐pentene (SFVE).
2. Figure 7.2 Synthesis of the SFVE monomer.
3. Figure 7.3 Chemical formulas of the most common PFSA ionomers.
4. Figure 7.4 Heats of second fusion of Aquivion precursor polymers versus Nafion^® as a function of equivalent weight (EW).
5. Figure 7.5 Temperature of complete fusion of SSC precursor polymers as a function of EW.
6. Figure 7.6 Dynamic mechanical spectrum of 850 EW Aquivion^® acid form membrane.
7. Figure 7.7 Dynamic mechanical spectrum of 1100 EW Nafion^® acid form membrane (N117).
8. Figure 7.8 Water uptake from liquid water at 100 °C for extruded Aquivion^® and Dow membranes as a function of EW.
9. Figure 7.9 Scheme of an operating fuel cell: H₂O₂ generated at the cathode and the formation of ^.hydroxyl radicals.
10. Figure 7.10 Scheme of an operating fuel cell: Effect of the hydrogen crossover.
11. Figure 7.11 Scheme of an operating fuel cell: Effect of the oxygen crossover and generation of hydrogen radicals.
12. Figure 7.12 Correlation between F⁻ emission and ─COOH chain ends.
13. Figure 7.13 Hydrogen crossover current density (mA cm⁻²) of Aquivion^® extruded membrane in OCV test with different reactants.
14. Figure 7.14 Hydrogen crossover current density (mA cm⁻²) of Aquivion^® extruded membrane in OCV test (70 °C, 50% reactant humidification) compared with the low permeability membrane.
15. Figure 7.15 Comparison of hydrogen crossover current density (mA cm⁻²) of Aquivion^® extruded membranes in OCV tests with different operating temperatures and reactant humidification.
16. Figure 7.16 Comparison of hydrogen crossover current density (mA cm⁻²) of Aquivion^® standard extruded membranes and stabilized extruded membranes in OCV tests with different operating conditions.
17. Figure 7.17 Comparison of cyclic voltammetry scan (100 mV s⁻¹) on cathode electrode at the beginning of life and after 1000 h of the OCV test with stabilized extruded membrane at 70 °C and 100% humidification of reactants.
18. Figure 7.18 Fluoride emission in the Fenton test: short side chain (SSC) and long side chain (LSC).
19. Figure 7.19 Cumulative fluoride ion emission in prolonged Fenton test for standard and stabilized SSC (⧫, standard; ▲, stabilized).
20. Figure 7.20 Lifetime of Aquivion^® E79‐03S and Aquivion^® E79‐03S + manganese scavenger.
21. Figure 7.21 The proposed decomposition mechanism.
Chapter 08
1. Figure 8.1 Chemical degradation routes of benzyl ammonium groups: (a) nucleophilic displacement (S_N2), (b) ylide‐intermediated rearrangements, and (c) Hofmann (E2) elimination.
2. Figure 8.2 Stability order of tetramethylammonium, 3‐methoxy‐benzyltrimethylammonium, benzyltrimethylammonium (BTMA), and 3‐nitrobenzyltrimethylammonium.
3. Figure 8.3 Alkaline stability order of N‐methyl‐N,N‐dicyclohexylbenzylammonium (MCH), N‐methyl‐N,N‐dipropylbenzylammonium (MnPr), and N‐methyl‐N,N‐diisopropylbenzylammonium (MiPr).
4. Figure 8.4 Chemical degradation route of imidazolium.
5. Figure 8.5 Chemical degradation route of pyridinium.
6. Figure 8.6 Chemical degradation routes of mono‐quaternized DABCO (top) and bis‐quaternized DABCO (bottom).
7. Figure 8.7 Chemical degradation routes of (a) piperidinium, (b) pyrrolidinium, and (c) morpholinium.
8. Figure 8.8 Chemical stability of 6‐azoniaspiro[5,5]undecane (ASU), N,N‐dimethylpiperidinium (DMP), N,N‐dimethylpyrrolidinium (DMPy), 5‐azoniaspiro[4,4]nonane (ASN), 1‐methyl‐4‐aza‐1‐azonia‐bicyclo[2,2,2]octane (MAABCO), and 1‐benzyl‐4‐aza‐1‐azonium‐bicyclo[2,2,2]octane (BAABCO).
9. Figure 8.9 Chemical degradation routes of phenyl guanidinium (top) and sulfonyl guanidinium (bottom).
10. Figure 8.10 Chemical degradation route of benzyltrimethylphosphonium.
11. Figure 8.11 Chemical degradation route of benzyldimethylsulfonium.
12. Figure 8.12 Chemical structures of (a) bis(terpyridine)Ru(II) cation and (b) permethylcobaltocenium.
13. Figure 8.13 Dehydrofluorination of poly(vinylidene fluoride).
14. Figure 8.14 Aryl ether bond cleavage of BTMA‐functionalized poly(arylene ether) under high‐pH conditions.
15. Figure 8.15 Hydrolysis of polyacrylate by hydroxide ion.
16. Figure 8.16 Stability order of backbone hydrolysis of polyacrylates.
17. Figure 8.17 Base‐catalyzed hydrolysis of polybenzimidazole.
18. Figure 8.18 (a)–(e) Chemical degradation routes of quaternized perfluoroalkyl ether side chains of perfluorinated AEMs.
19. Figure 8.19 Stability of the amide linkage of perfluorinated polymers.
20. Figure 8.20 (a) Quaternary ammonium‐functionalized polybutadiene‐b‐poly(4‐methylstyrene), (b) phosphonium‐functionalized polyethylene, and (c) quaternary ammonium‐functionalized polypropylene.
21. Figure 8.21 Stability order of polystyrenes with benzylic trimethylammonium and imidazolium.
22. Figure 8.22 Polystyrene‐co‐polyacrylonitrile‐co‐polypyrrolidinium.
23. Figure 8.23 Polystyrene AEMs cross‐linked with (a) N,N,N′,N′‐tetramethylhexanediamine (TMHDA) and (b) bis(2‐dimethylaminoethyl) ether (AMINOETHER).
24. Figure 8.24 Polystyrene‐b‐poly(ethylene‐co‐butylene)‐b‐polystyrene (SEBS) triblock copolymers with (a) benzyltrimethylammonium and (b) hexyltrimethylammonium, (c) poly(2,6‐dimehyl‐1,4‐phenylene oxide)‐b‐poly(vinylbenzyltrimethylammonium) (PPO‐b‐PVBTMA).
25. Figure 8.25 (a) Poly(N‐vinyl carbazolium) and (b) blended poly(N‐vinyl pyrrolidinium) with poly(vinyl alcohol).
26. Figure 8.26 Blended polybenzimidazole with polystyrene functionalized with benzimidazolium and DABCO.
27. Figure 8.27 (a) Quaternary ammonium‐containing polyfluorene, and polyphenylenes with (b) imidazolium‐, (c) BTMA‐, (d) hexyltrimethylammonium‐, and (e) poly(arylene)‐containing alkyl‐tethered trimethylammonium.
28. Figure 8.28 Polyphenylene‐based cross‐linked AEM.
29. Figure 8.29 Multication‐tethered (a) poly(arylene ether sulfone) and (b) poly(phenylene oxide).
30. Figure 8.30 Polybenzimidazoliums with cations in the (a) polymer backbone (mQPBI‐X), (b) side chain (sQPBI‐X), (c) both the backbone and side chain (msQBPI‐X) and (d) sterically bulky mesityl substituent.
31. Figure 8.31 Blended polybenzimidazolium with polybenzimidazole.
32. Figure 8.32 Polynorbornene AEM.
33. Figure 8.33 Chemical structures of (a) phenylpentamethylguanidinium‐functionalized and (b) morpholinium‐functionalized perfluorinated polymers.
Chapter 09
1. Figure 9.1 Optical images of cross sections for Nafion^® membranes after removal of the catalyst layer: non‐degraded (a), degraded for 52 h (b), and degraded for 180 h (c) at open‐circuit voltage (OCV) conditions and 90 °C. Images of degraded MEAs were collected with the optical camera of the Perkin‐Elmer Spotlight 200 microscope system after removal of the catalyst layer.
2. Figure 9.2 2D spectral–spatial maps of Nafion^® membrane: non‐degraded (a), degraded during 52 h (b), and degraded during 180 h (c). Yellow ovals in (c) show the appearance of strong bands at a depth of 82 µm from the cathode as a result of membrane degradation. Weaker bands at the same frequency are also visible near the cathode. The acquisition time for each 2D spectral–spatial map was 150 min.
3. Figure 9.3 In‐depth profiling: FTIR spectra of the Nafion^® membrane as a function of depth from the cathode: non‐degraded membrane (a), degraded for 52 h (b), and degraded for 180 h (c). In (c) the spectra in thick bold show the highest intensities of degradation species: at 82 and 22 µm from the cathode. All spectra were recorded in reflectance mode with the Perkin‐Elmer Spotlight 200 microscope system.
4. Figure 9.4 Comparison of the FTIR spectra of non‐degraded Nafion^® MEA (a) with the FTIR spectrum recorded at a depth of 82 µm for the membrane degraded during 180 h (b).
5. Figure 9.5 The intensity variation of the C─H and C═O absorption bands as a function of depth from the cathode for the MEA degraded during 180 h. The band at 1740 cm⁻¹ corresponds to the vibration of C═O group, and bands at 2856, 2928, and 2962 cm⁻¹ are due to C─H vibrations. Cathode and anode sides are indicated by gray bars.
Chapter 10
1. Figure 10.1 Chemical structure of Nafion^®. Nafion^® 117 has an average composition of x = 6.5, y = 1, z = 1. N indicates the nonpolar monomeric units, while P indicates the polar monomeric units.
2. Figure 10.2 (a)–(d) Nudged elastic band (NEB) reaction paths of HOOH formation and hydroxyl radical formation from DFT, corresponding to solvent energies in Table 10.2.
3. Figure 10.3 Potential energy map of ORR with barriers of cathode formation in solvent phase. The labels a1 to c3 represent coordinate geometries shown in Figure 10.2 of the newly calculated mechanisms. The other energies were calculated previously [27b]. OH_ad formation is difficult in the solvent phase (barrier 0.97 eV). It was shown that O_ad can react with water to form two OH_ad much easier (0.50 eV), which is depicted in the bottom.
4. Figure 10.4 Illustration on the effect of high concentration of H₂ on the surface reactions of O₂. In a surface covered with H, O_2ad ⇒ 2O_ad is hindered. However OOH_ad is not. This explains why in experiments with a high concentration of H₂ gas, Nafion^® degradation occurs at a higher rate.
5. Figure 10.5 (a) Generation of H radical from hydroxyl radical and H₂ crossover gas. (b) Mechanisms to generate H^• from at the anode. The energetics of the solvated and gas phases are described by regular and italic numbers, respectively. H, adsorbed on the Pt catalyst, can break off directly from the surface with a barrier of 3.69 eV. (c) As an alternative mechanism, H₂ can form H_ad + H^• with a lower barrier of 2.27 eV as shown. The high barriers show the energetic difficulty of generating a radical at the anode from H₂ and Pt.
6. Figure 10.6 Initiation of the Nafion^® polymer degradation for a ─CF₂H end group with a low barrier of 0.02 eV (B3LYP).
7. Figure 10.7 Degradation initiation in the concentrated HO^• environment of the R─CF═CF₂ end group. The mechanism proceeds through either steps (a‐i) and (b‐i) or (a‐ii) and (b‐ii). It continues through (c) and (d). The rate‐determining step is (d) with a barrier of 0.22 eV (B3LYP).
8. Figure 10.8 Degradation initiation of the R─COOH end group with a rate‐determining step barrier (a) of 0.31 eV (B3LYP) for HO^• attachment followed by H₂CO₃ detachment (b).
9. Figure 10.9 Degradation propagation of the polymer main chain in concentrated HO^• environment starting (a) from the products of Figures 10.6, 10.7d, and 10.8b, with a rate‐determining step barrier (b) of 0.09 eV (B3LYP). The product of (c‐i) can continue to degrade by starting from reactant (a). The alternative reaction with water in (b‐ii) to (c‐ii) is likely, but has a higher barrier of 0.31 and 0.42 eV. The R─COOH product of (c‐ii) was found in H₂O₂ vapor tests [13c] and can continue to degrade by starting from the reactant of Figure 10.8a.
10. Figure 10.10 Mechanism of degradation of Nafion^® sulfonic acid group proposed by Ghassemzadeh et al. [6a]. Energetics are from our DFT calculations. Starting with dehydration by HO^• (a) followed by SO₃ detachment (b). We consider steps (b), (d) and (e) to be unlikely at normal fuel conditions.
11. Figure 10.11 Proposed degradation mechanism involving hydroxyl radical attacking Nafion^® sulfonic acid group. (a) Radical substitution, (b) epoxide formation, and (c) epoxide hydrolysis.
12. Figure 10.12 Initiation mechanism for degradation in the fuel cell environment of the R─CF═CF2 end group. In the fuel cell environment, the hydroxyl radical attacks and breaks the double bond followed by O₂ gas attacking the radical (a) and subsequent reaction with H₂ gas (b) followed by dehydration (c). The rate‐determining step is (b) with a barrier of 0.74 eV (B3LYP). The difference between this mechanism and the concentrated HO^• environment mechanism in Figure 10.7 is that it is limited to only one HO^• reactant.
13. Figure 10.13 Mechanism for the propagation of polymer main chain degradation in the fuel cell environment starting from the products of the three defect group initiation mechanisms: Figures 10.6, 10.7d, and 10.8b. After (c), the product of Figure 10.9c‐i can continue to degrade by starting from reactant (a). The rate‐determining step is (b) with a barrier of 0.59 eV (B3LYP). The difference between this mechanism and the concentrated HO• environment mechanism in Figure 10.9 is that it is limited to only one HO^• reactant.
14. Figure 10.14 Proposed degradation mechanism involving H radical attacking Nafion^® side chain. (a) H radical attack F on tertiary carbon and (b) ketone formation.
15. Figure 10.15 Initiation mechanism in the fuel cell environment of the R─CF═CF₂ end group at the cathode without H₂ crossover. The reaction starts with the end group reacting with O₂ to form the R─OO^• as described in Figure 10.12a. The R─OO^• reacts with H₂O to form R─OOH + HO^• (a), which is the rate‐determining step with a barrier of 1.20 eV (B3LYP). The O─O bond cleaves (b) and the resulting R─O^• product breaks from the chain to form a ketone as described in Figure 10.7d. An alternative initiation mechanism for R–CF=CF₂ end group initiation with a slightly higher barrier of 1.31 eV (B3LYP) is described in (c).
16. Figure 10.16 Propagation mechanism of main chain degradation in the fuel cell environment at the cathode without H₂ crossover. The reaction starts with carbon radical R─CF₂^• reacting with O₂ to form R─OO^• as described in Figure 10.13a. The R─OO^• reacts with H₂O (a) to form R─OOH + HO^•, which is the rate‐determining step with a barrier of 0.92 eV (B3LYP). The O─O bond cleaves (b) and the resulting R─O• product breaks from the chain to form a ketone as described in Figure 10.9c‐i. An alternative propagation mechanism without H₂ is described in Figure 10.17b.
17. Figure 10.17 Initiation and propagation mechanisms in the fuel cell environment of the R─CF₂H end group at the anode without O₂ crossover. (a) The initiation starts with H radical formed at the anode attacking the ─CF₂H end group to form ─CF₂^• and H₂ with a barrier of 0.31 eV (B3LYP). (b) Degradation propagation can continuously occur with the breaking of the C─C bond to form CF₂═CF₂, which is the rate‐determining step with a barrier of 1.41 eV (B3LYP).
18. Figure 10.18 Initiation mechanisms in the fuel cell environment of the R─CF═CF₂ end group at the anode without O₂ crossover. (a) H radical formed at the anode attacks the end group breaking the double bond without barrier. (b) The ethylene group can break off the main chain with a barrier of 1.24 eV. The main chain can continue to degrade with the reaction described in Figure 10.17b.
19. Figure 10.19 Initiation mechanisms in the fuel cell environment of the R─COOH end group at the anode without O₂ crossover. (a) H radical formed at the anode attacks the end group forming R─COOH─H with a rate‐determining barrier of 0.62 eV. (b) This group can break off the main chain with a barrier of 0.13 eV. The main chain can continue to degrade with the reaction described in Figure 10.17b.

The Chemistry of Membranes Used in Fuel Cells

Degradation and Stabilization