2 nd International Workshop on Degradation Issues of Fuel Cells Thessaloniki, Greece

Transcription

1 Overview of the FP7 Project Results and Recommendations K. Andreas Friedrich 2 nd International Workshop on Degradation Issues of Fuel Cells Thessaloniki, Greece Degradation Workshop, Thessaloniki, 21 September September 211

2 Project Project general General information Information Project full title: Coordinator: Project major partners: Starting Date: Ending Date: Understanding of Degradation Mechanisms to Improve Components and Design of PEFC K. A. Friedrich, DLR Opel, Volvo, SGL, Solexis, DANA, CEA, ZSW, JRC, Uni. Erlangen, Chalmers Uni. Budget Total/Funding: 5.5 MEUR / 3.7 MEUR Type of project: CP

3 Motivation and Goals First main goal of is to assess the relevance of the degradation processes of polymer electrolyte fuel cell based on the extensive analysis performed in Second goal is to identify and implement improvements for fuel cell durability based on: Understanding of degradation processes Improved materials Improved operation conditions Third goal is the development of prediction tool for degradation based on modelling (different modelling approaches) Degradation Workshop, Thessaloniki, 21 September 211 Page 3

4 Interactions of Work Packages

5 Identification and Ranking of CCM Degradation Mechanisms Mechanism Structural degradation Mechanical degradation of the membrane Loss of electrochemical activity at the cathode Loss of electrochemical activity at the anode Chemical degradation Importance ? Degradation Workshop, Thessaloniki, 21 September 211 Page 5

6 Mechanical + Chemical Stress: Edge Failure Simple gasket design: Degradation issues: Gas X-over on the edges leads to chemical degradation Mechanical shear stress during dynamic operation (membrane expansion/shrinkage) Membrane exposed to GDL fiber puncture Suitable edge/gasket designs will avoid these failure modes Degradation Workshop, Thessaloniki, 21 September 211 Page 6

7 Cell Voltage (V) Hydrogen Crossover (ma/cm²) Stabilized Aquivion TM Membrane Open Circuit Voltalge at 75 C Accelerated aging test for membranes 5. 7 C 5 % RH std standard grades 7 C 5 % RH stab 9 C 5 % RH std 9 C 5 % RH stab stabilized grades 1, Polarisation curves E79-3S + LT25EW (25cm² cell).6 V constant (=~1A/cm²) 1% reactant humidification - 75 C Bar abs, OCV 25 duration (hrs) 3,8,7,6,5,4,,2,4,6,8 1, 1,2 1,4 Current Density (A/cm²) BOL Hour 1 Hour 2 Hour 3 Hour 4 Degradation Workshop, Thessaloniki, 21 September 211 Page 7

8 Stabilized Aquivion TM Membrane Experimental MEA made by CEA using unstabilized AQUIVION membrane without edge protection (29) Experimental MEA made by CEA using reinforced AQUIVION membrane with edge protection (21) Degradation Workshop, Thessaloniki, 21 September 211 Page 8

9 voltage [mv] Ageing of MEAs in single cell Ageing test of CCB MEA (E87-5S + sub-gaskets) + Segmented-cell (DLR) Constant load i = 676 ma.cm -2 Current density distribution Counter-flow 1/1%RH 676 ma/cm 2 ma/cm 2 Cell voltage at OCV, low and high current densities BAqPP923 conditions: counter-flow; single serpentine cell+segm. cell; 15 mbar at cell outlet; 8 C; RH 1/1%; stoich 1.7/2. 1 current density [ma/cm2] Time after start of constant conditions [h] long term OCV operation >1 min No OCV degradation Faster reversible degradation Small irreversible degradation 19 h: 833 h: Current density distribution: H2 average current density = 676 ma/cm 2 Air h 574h + 1min OCV Possible recovery strategy for flooded electrodes: min OCV operation Final Workshop March 24 rd, 211 Page 9

10 Dynamic Test of Membrane and Electrodes Results with ELAT electrodes (SLX) E87-5S edge protected Dry conditions Evidence of better mechanical stability with increased membrane crystallinity & edge protection Degradation Workshop, Thessaloniki, 21 September 211 Page 1

11 Electrode Characterization TEM observations (CEA) Active layers degradation: after cycling and membrane damaged Cathode side: more degradation H2 inlet Air outlet Fresh MEA Pt particles growth by Ostwald ripening at air outlet XRD : 3 nm XRD : 6,3 nm H2 outlet Air inlet 6,3 nm Cathode side Anode side,5 µm C corrosion and massive Pt dissolution + reduction in AL or membrane at air inlet 5 nm Degradation Workshop, Thessaloniki, 21 September 211 Page 11

12 Characterization: Conductive AFM Solexis E87-5S Comparison baseline after 24 h and after 2 h stationary Operation left: reference outlet baseline, 24 h Other Methods in : EIS, CV, LSV Raman spectroscopy XPS right: 2 h stationary operation size / nm Nafion 112 Solexis 2 h stationary Mean structure size Solexis 2 h cycling air outlet Solexis 2 h cycling air inlet Page 12 Solexis 24 h reference

13 Identification and Ranking of GDL Degradation Mechanisms Mechanism Importance Chemical degradation Loss of hydrophobicity Carbon / structure corrosion Structural degradation Change in (gas phase) transport parameters Observed, but influence on performance limited Change in wetting behaviour Page 13

14 Chemical Degradation of Electrodes and GDL Loss of hydrophobicity Partial decomposition of PTFE identified by XPS PTFE decomposition mainly on the anode Decrease of hydrophobicity Changed water balance Reversible loss of performance Degradation Workshop, Thessaloniki, 21 September 211 Page 14

15 Chemical Degradation - Carbon Corrosion Carbon corrosion could be detected No fluoride was found -> No PTFE decomposition for this chemical degradation experiment Mass loss study for quantification Chemical reactions: C 2H2O2 CO2 2H2O Ca( OH) 2 CO2 CaCO 3 H2O CaCO 3 H2CO3 Ca(HCO3 ) 2 In : Comparison of naturally aged and artifically aged GDLs! Degradation Workshop, Thessaloniki, 21 September 211 Page 15

16 Hydrohead Measurements for Testing Hydrophobicity p fluid g h Other Characterization methods: Mercury porosimetry Capillary flow porosimetry EDX mappings Contact angle Degradation Workshop, Thessaloniki, 21 September 211 GDL type New GDL 25BC Naturally aged GDL 25BC for 1h p (mbar) 86.3 ± ± 6.2 Artificially aged GDL 76.2 ± 25BC for 24h in H Modified new GDL 25BC 78.9 ± 5.3 Modified artificially aged 79.4 ± GDL 25BC for 24h in 2.1 H 2 2 New GDL 25BA 15.7 ± 1.2 Page 16

17 IR Spectroscopy Inhomogenity in the Intensity of the C-F Vibration Recommendation: Improve homogeneity of hydrophobic agent distribution Degradation Workshop, Thessaloniki, 21 September 211 Page 17

18 In Situ Single Cell Tests (SGL) Artificial and natural ageing Degradation Workshop, Thessaloniki, 21 September 211 Page 18

19 T [ C] U [V] Short Stack Long Term Test Temperature Cycling Test 25 Voltage time chart over 7 h ave T KW Ein [ C] standard U1 standard U2 25 BC, modified 769 neu U3 used (15 h) U4 used (15 h) U5 aged standard U6 aged standard U7 769 aged 25 aged BC, mod. U8 769 aged 25 aged BC, mod. U9 standard neu U BC, modified neu U11 standard neu U time [h] º Very low degradation of cells with modified GDLs compared to cells with standard GDLs

20 Identification and Ranking of Bipolar Plate Degradation Mechanisms Mechanism Contamination of the Ionomer from external sources via port region Change of contact resistance Water accumulation in areas of low flow and low pressure difference Potential MEA contamination from the plates Release of silicon from the seal material Importance ? Page 2

21 Corrosion products Corrosion products: nickel, iron, chromium degradation products (composite) degradation products (AISI316L) Ni Fe Total sum of metallic cations in the stack 968 ppm Total sum of metallic cations in the stack 1172 ppm comparable metallic contamination of both materials can this be possible?

22 Distribution of contaminants peaks are allocated to the coolant inlet and coolant outlet region direct contact of the ionomer to the medias trough the port cut-outs design proposal elaborated to avoid this contamination Degradation Workshop, Thessaloniki, 21 September 211 Page 22

23 - Stack Contaminations Contamination of the ionomer from external sources via port region Step one introduce Solvicore 5 Layer MEA (Membrane Solexis, Catalyst, Sub gasket, Membrane extended to the edge of the bipolar plate Step two change of MEA design to Ionomer free Sub gasket, Port area Degradation Workshop, Thessaloniki, 21 September 211 Page 23

24 Stack Tests with Improved Stack Design Conclusions of WP6 durability runs: Comparable behavior between new and old MEA configuration Higher cell voltage with conductive coating, irregular cell behavior Durability run with AISI316L blank and new MEA with old configuration at DANA Durability run with AISI316L blank and new MEA with new configuration at DANA Durability run with conductive coating and new MEA configuration Durability run with modified conductive coating, new MEA design and further developed conditions Modified coating and further developed conditions with excellent performance results Degradation Workshop, Thessaloniki, 21 September 211 Page 24

25 Contaminations in MEA Corrosion products: nickel, iron, chromium 24 (AISI316L bipolar plates with organic coating, new MEA Design and new operating conditions) µv/h 15 (AISI316L bipolar plates) 6µV/h Degradation Workshop, Thessaloniki, 21 September 211 Page 25

26 Modelling and Lifetime Prediction Degradation Workshop, Thessaloniki, 21 September 211 Page 26

27 Modelling and Lifetime Prediction

28 Modelling activities and results Membrane and Electrodes: Multiscale elementary kinetics simulation with coupling to microscopical structure Porous media: Molecular Dynamics Lattice Boltzmann Monte-Carlo Performance modelling Bipolar Plates: CFD Movement of droplets by VOF (volume of fluid) Final Workshop March 24 rd, 211 Page 28

29 Summary Improvement achieved by materials: Reinforced membrane with higher crystalinity Modified gas diffusion layer Improvement achieved by design: Edge protection of membrane Blocking of external contermination by new sealing concept Improvement achieved by operation conditions: Avoiding liquid water phase Excursion to open circuit conditions to recover reversible voltage losses Different models with life time prediction capability Degradation Workshop, Thessaloniki, 21 September 211 Page 29

30 THANK YOU FOR YOUR ATTENTION Acknowledgement to the partners of : M. Schulze, A. Haug, E. Gülzow, K.A. Friedrich, Investigation of Local Degradation Effects, ECS Transactions 26 (21) K. Seidenberger, F. Wilhelm, J. Scholta, Monte-Carlo-Simulation - Wasserhaushalt in der GDL einer PEM-Brennstoffzelle article (German), HZwei (April 211), pages S Pulloor Kuttanikkad, J.Pauchet, M.Prat; Pore-network simulations of twophase flow in a thin porous layer of mixed wettability, Journal of Power Sources 196 (211) 1145 K. Seidenberger, F. Wilhelm, T. Schmitt,W. Lehnert, J. Scholta, Estimation of water distribution and degradation mechanisms in polymer electrolyte membrane fuel cell gas diffusion layers using a 3D Monte Carlo model J. Power Sources 196 (211) 5317 M. Holber, P. Johansson and P. Jacobsson, Raman spectroscopy of an aged low temperature polymer electrolyte fuel cell membrane, Fuel Cells, 211, accepted J. Pauchet, M. Prat, P. Schott, S. Pulloor Kuttanikkad, Analysis of the effect of hydrophobicity loss of GDL on performance of PEMFC by coupling pore network and performance modelling, Submitted to the Journal of Power Sources Degradation Workshop, Thessaloniki, 21 September 211 Page 3

31 AFM Analysis of Solexis E87-5S, 2 h outer side Cathode side inner side Anode side inner side outer side Division of membrane in 2 layers of 15 μm after storage in water Height Height Phase Height Overall high density of particles with small phase shift: probably no Pt Platinum particles with high phase shift and high DMT modulus (~elasticity) at inner side of anode with mean diameter of 21 nm (11 nm - 44 nm) Height profile DMT Phase 4 nm 2 nm DMT Degradation Workshop, Thessaloniki, 21 September 211 Page 31