2008 DOE Hydrogen Program ReviewJune 9-13, 2008

Rod Borup, John Davey, Hui Xu, Axel Ofstad, Fernando Garzon, Bryan Pivovar

Los Alamos National Laboratory

PEM Fuel Cell Durability

FC26 This presentation does not contain any proprietary, confidential, or otherwise restricted information

Overview

Budget• FY04: $900k• FY05: $950k• FY06: $1000k• FY07: $0 $300• FY08: $300

Timeline2001: Project started as Fuel Cell

Stack Durability on Gasoline Reformate

2004: Changed focus PEM H2Durability

2007: Ended. Restarted at $300k

Barriers

Collaborators• No Formal Partners• ORNL (Karren More)• LANL H2 Storage Center• Analysis:

• Univ. New Mexico, Augustine Scientific, LANL MPA-MC

• Materials:• Gore, SGL, Cabot Fuel Cells

• Durability• Cost • Electrode Performance

• Define degradation mechanisms• Design materials with improved durability• Identify and quantify factors that limit PEMFC Durability

• Measure property changes in fuel cell components during life testing• Life testing of materials

• Examine testing conditions, incl. drive cycle• Membrane-electrode durability• Electrocatalyst activity and stability• Electrocatalyst and GDL carbon corrosion• Gas diffusion layer hydrophobicity• Bipolar plate materials and corrosion products

• Develop/apply methods for accelerated and off-line testing• Improve durability

Objectives:Quantify and Improve PEM Fuel Cell Durability

2010-2015 Technical Target: 5000 hours Durability (with cycling)

Approach to Durability Studies• Fuel Cell MEA Durability Testing and Study

• Constant voltage/current/power and power cycling (drive cycle)• VIR / cell impedance• Catalyst active area• Effluent water analysis

• in situ and post-characterization of MEAs, catalysts, GDLs• SEM / XRF / XRD (ex situ and in situ) / TEM / ICP-MS / neutron scattering / H2 adsorption / Inverse Gas Chromatography / Contact Angle / totalporosity / hydrophillic vs. hydrophobic porosity

• Develop and test with off-line and accelerated testing techniques• Potential cycling • Environmental component aging, testing and characterization• Component interfacial durability property measurements

Durability Testing Issues• Testing times can be lengthy (and costly)

– 5,000 hrs = ~ 7 months (automotive target)– 40,000 hrs = ~ 4.6 years (stationary system target)– Need relevant accelerated testing – Need to close ‘field / lab gap’ or ‘transfer function’

• Lab single cell ‘real’ stacks field data

• Operating variables effect not fully understood– Many degradation mechanisms likely yet undefined– Power transients - vehicle fuel cell/battery hybridization– Transient power, temperature, RH– Shut down / start-up

• Materials still being developed and improved– Need relevant accelerated testing

Comparison of Accelerated Testing MethodsUSFCC Accelerated Catalyst Test #1

Step vs. Triangle Potential Cycle

-100

-80

-60

-40

-20

0

20

40

60

80

100

0 200 400 600 800 1000 1200

Voltage / V

Cur

rent

0 Cycles400 Cycles1600 Cycles2800 Cycles4000 Cycles8000 Cycles

0

0.25

0.5

0.75

1

1.25

0 50 100 150 200

Time / sec

Volta

ge /

V

• Accelerated catalyst testing by potential cycling in H2 / Air

• Voltage cycling: 0.6 and 0.96V (H2/Air)

Step Potential Cycle Triangle Potential Cycle

• H2/Air requires load bank for high current (can’t use potentiostat).• Some MEAs do not reach 0.96V OCP with standard load control• Triangle potential sweep shows much faster degradation

-400

-300

-200

-100

0

100

200

300

0 200 400 600 800 1000 1200

Voltage / V

Cur

rent

0 Cycles400 Cycles1200 Cycles2400 Cycles4000 Cycles8000 Cycles16000 Cycles

0

0.25

0.5

0.75

1

1.25

0 50 100 150 200

Time / sec

Volta

ge /

V

Comparison of Accelerated Testing ProtocolsTest #1. Voltage cycling: 0.6 and 0.96V (H2/air)

Step vs. Triangle Potential Cycle

0.80

0.83

0.86

0.89

0.92

0.95

0 5000 10000 15000 20000

# Cycles

OC

P / V

0.80

0.83

0.86

0.89

0.92

0.95

0 5000 10000 15000 20000

# Cycles

OC

P / V

• OCP decreases with cycles• Varies potential limits

• Increase in sample HFR • H2/Air cycling not just

catalyst degradation

• Difficulties with this test being consistent and repeatable• Does not separate catalyst durability from other components

0.00.10.20.30.40.50.60.70.80.91.0

0.0 0.5 1.0 1.5Current Density / A/cm2

Volta

ge /

V

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

HFR

0 Cycles400 Cycles1600 Cycles2800 Cycles4000 Cycles8000 CyclesHFR 0 CyclesHFR 400 CyclesHFR 1600 CyclesHFR 2800 CyclesHFR 4000 CyclesHFR 8000 Cycles

Step Potential Cycle Triangle Potential Cycle

0.00.10.20.30.40.50.60.70.80.91.0

0.0 0.5 1.0 1.5

Current Density / A/cm2

Volta

ge /

V

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

HFR

0 Cycles400 Cycles1600 Cycles2800 Cycles4000 Cycles8000 Cycles16000 CyclesHFR 0 CyclesHFR 400 CyclesHFR 1600 CyclesHFR 2800 CyclesHFR 4000 CyclesHFR 8000 CyclesHFR 16000 Cycles

OCP vs. Cycles OCP vs. Cycles

VIR curves over 16,000 Cycles VIR curves over 8,000 Cycles

Shut-down/Start-up Effects

Modified from: Sathya Motupally, UTC Power

• ‘Reverse Current’ degradation• Non-homogeneous mixture of H2 on anode • H2/air portion of cell drives ‘reverse current’ elsewhere

Stop-Start Cycling Effect on Carbon Corrosion

0

5

10

15

20

25

0 1000 2000 3000 4000Time / sec

CO

2 Out

let /

ppm

CO2 (ppm)Anode:Flowrate: 100 sscm H2

Purge: 50 sccm Air

66.5

386.1

50 cm2

25 oC

Start-up

Shut-down

0

5

10

15

20

25

0 500 1000 1500 2000 2500Time / sec

CO

2 Out

let

/ ppm

CO2 (ppm)

13125.5

Anode:Flowrate: 400 sscm H2

Purge: 200 sccm Air

50 cm225 oC Start-up

Shut-down

• Operation• OCV and dry air (250 sccm)

continuously to cathode • Shut-down: anode dry air purge: 5 min.• Start-up: flow dry H2 to anode: 5 min. • Measure CO2 (and CO) evolution at

cathode by NDIR (Non-dispersive Infrared)

• Results• Increasing anode gas change-over rate

decreases CO2 evolution • More CO2 evolution at start-up

compared to shut-down, 25 oC• Small amounts of CO produced

Anode Purge Rate Comparison

Purge time (1 turn-over) = 3.7 sec

Purge time (1 turn-over) = 15.0 sec

Temperature Effect on Carbon Corrosion During Stop-Start Cycling

0

5

10

15

20

25

30

35

40

0 500 1000 1500 2000 2500Time / sec

CO

2 Out

let C

onc.

/ pp

m CO2 (ppm)60 oC

Shut-downStart-up 7700 3160

Anode Flowrate: 100 sscm H2

Purge: 50 sccm Air50 cm2

05

10152025303540

0 500 1000 1500 2000 2500Time / sec

CO

2 Out

let C

onc.

/ pp

m CO2 (ppm)80 oC

Shut-down

Start-up2659

2604

Anode Flowrate: 100 sscm H2

Purge: 50 sccm Air50 cm2

•CO2 Evolution at Slow Purge Rate•25 oC

• Higher at start-up than shut-down• Much lower evolution than at 60 oC

•60 oC• Greatest evolution• Higher evolution at shut-down

• 80 oC• Non-zero steady-state evolution• ~ Equal shut-down/start-up evolution

0

5

10

15

20

25

30

35

40

0 1000 2000 3000 4000Time / sec

CO

2 Out

let C

onc.

/ pp

m CO2 (ppm)

Anode Flowrate: 100 sscm H2

Purge: 50 sccm Air50 cm2

66.5

386.1Start-up

Shut-down

25 oC

GDL DurabilityContact Angle Changes

130132134136138140142144146148150

0 100 200 300 400 500 600

Time After Drop Placement (seconds)

Con

tact

Ang

le (d

egre

es)

Untreated

DN NaCl-01-Cathode

DN NaCl -01-Anode

DN NaCl-03-Cathode

DN NaCl-03-Anode

Contact angles with aging (Single Fiber Measurements)

• GDLs lose hydrophobicity with aging• Exposure to NaCl make GDLs more hydrophobic

• Also slows rate of water uptake

78

80

82

84

86

88

90

92

94

96

UnagedTGP-H 120

(Plain)

UnagedTGP-H 090

#1 (17.2wt% FEP)

UnagedTGP-H 090

#2 (17.0wt% FEP)

N2, 60°C,460 hr;

TGP-H 060(17.2 wt%

FEP)

N2, 80°C,460 hr;

TGP-H 090(16.7 wt%

FEP)

Air, 60°C,680 hr;

TGP-H 060(16.9 wt%

FEP)

Air, 80°C,680 hr;

TGP-H 090(17.0 wt%

FEP)

Con

tact

Ang

le (d

eg)

Hydrophobic

Hydrophilic

Increasing Aggressiveness of Aging Environment

Contact angles with NaCl exposure(Paper Sessile Drop Measurements)

Surface Analysis of GDL Material• Confirm –COOH surface species

– Observe –OH and C=O IR

• Confirm acyl chloride – Reduction of –OH, and/or C-Cl

MPA-MC, John Rau, Clay Macomber

DRIFTS Spectra of Aged GDL(Diffuse Reflectance Infrared Transmission Spect.)

• -OH species identified• Not yet satisfactorily identified

surface species• Using DRIFTS, will also explore

Raman

-OH

Spatial Resolution of Durability:Individual Fuel Cell Segments: VIRs over Time

Segment 01

Current density / mA cm-20 100 200 300 400 500 600 700 800 900

Vol

tage

/ m

V300

400

500

600

700

800

900

100024 h96 h264 h432 h600 h792 h

Segment 04

Current density / mA cm-20 100 200 300 400 500 600 700 800 900

Vol

tage

/ m

V

300

400

500

600

700

800

900

100024 h96 h264 h432 h600 h792 h

Segment 10

Current density / mA cm-20 100 200 300 400 500 600 700 800 900

Vol

tage

/ m

V

300

400

500

600

700

800

900

100024 h96 h264 h432 h600 h792 h

Segment 07

Current density / mA cm-20 100 200 300 400 500 600 700 800 900

Vol

tage

/ m

V

300

400

500

600

700

800

900

100024 h96 h264 h432 h600 h792 h

• Performance degradation greater at fuel cell inlet and near outlet

Segmented six parallel channel serpentine flow-field (left side) and segmented cathode hardware including gasket and current collector plates.Segment-- 7.7 cm2

Total– 77 cm2

Segmented Cell Hardware

Crossover Current DensityElectrocatalyst Surface Area

Seg

01

Seg

02

Seg

03

Seg

04

Seg

05

Seg

06

Seg

07

Seg

08

Seg

09

Seg

10

Cro

ssov

er c

urre

nt d

ensi

tyat

0.2

5 V

/ m

A c

m-2

0

5

10

15

20

25

30

24 h 96 h 264 h 432 h600 h 792 h

Note: Because of segmented flowfield traversing in series, each subsequent segment cross-over is cumulative for all previous segments

• H2 cross-over per segment is ~ constant

• Unclear about segment 10

Seg

01

Seg

02

Seg

03

Seg

04

Seg

05

Seg

06

Seg

07

Seg

08

Seg

09

Seg

10

EC

A /

cm2 P

t cm

-2

10

20

30

40

50

60

70

80

90

100

24 h 96 h264 h 432 h 600 h792 h

•Loss of electrocatalyst surface area (ECA) predominately at cathode outlet • (higher water content)

•Loss of ECA at inlet doesn’t explain significant performance loss at inlet

RH Effect on Membrane Degradation

Time (hours)0 50 100 150 200

Cur

rent

Des

ity (m

A/c

m2 )

0

4

8

12

16

20

24un-humid. 20% RH 60% RH 100%RH

H2, 500 sccm, 26psi; cathode: air/N2, 1000 sccm, 26psi.

• Increase in H2 crossover at medium RHs (20-60%)

Time (hours)0 50 100 150 200

OC

V

0.6

0.7

0.8

0.9

1.0un-humid.20% RH60% RH100% RH

• Stable OCP at 100% RH• OCP degrades at 20 & 60% RH

• More H2 and O2 crossover results in greater H2O2formation

Hydrogen Crossover Open-Circuit (OCP)

Fluoride Emission Rate (FER)

Time (hours)0 50 100 150 200

FER

, μm

ol/(h

.cm

2 )

10-4

10-3

10-2

10-1

100

un-humid.20% RH 60 %RH 100% RH

Time (hours)0 50 100 150 200

FER

, μm

ol/(h

.cm

2 )10-4

10-3

10-2

10-1

100

un-humid.20% RH 60 %RH 100% RH

CathodeAnode

• Highest fluoride ion emission rate at 60% RH at anode and cathode• 20% RH rate similar to 100% RH rate

Anode: H2, 500 sccm, 26psi; cathode: air, 1000 sccm, 26psi.

Durability Test with Hydrogen from Chemical Hydride

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 0.5 1 1.5 2 2.5 3Time / hr

Cur

rent

(A) /

Vol

tage

(V)

0

50

100

150

200

250

Flow

rate

/ sc

cm

Cell VoltsCell AmpsH2 FlowrateAir Flowrate

H2 from Tank

H2 f

Out of Hydrogen

0

0.2

0.4

0.6

0.8

1

1.2

0 20 40 60 80 100Time / hr

Cur

rent

(A) /

Cel

l Vol

tage

(V)

Cell VoltsCell Amps

Switched to H2 from H2 Storage Material

H2 from Tube Trailer

(small recovery)

Tube Trailer H2

5 cm2 Cell80 oC100 % Inlet

Cell reaches zero current within ~ 3 hours

Test #1 Test #2

• Immediate decrease in cell performance upon switching to H2 from H2 Storage Material

• Complete failure in 3 hours• Gas analysis suggests B-N

species

• Cell gradually recovered ~ 80% over several days

• Used carbon filter in H2 line• No immediate decrease in

performance• Simple filtration may work

Supporting LANL H2 Storage COE

MilestonesPEM Fuel Cell Durability

Mon Yr Milestone May 07 Shut-down / start-up protocol comparison of degradation

rates Dec 07 Electrocatalyst particle size growth measurements

performed on 2010 and 2015 DOE target loadingsJan 08 Comparison of off-line potential square-wave cycling

with fuel cell operation with square-wave cycling

Jun 08 Segmented Cell OperationSept 08 Peroxide formation results as function of Temperature,

Operating potential and Electrocatalyst

Carboncorrosion

DOE/USFCCH2/AirH2/N2

S.S.

Summary - Durability Testing• Durability testing remains difficult and time intensive

• Time constraints led to accelerated type testing• Decay mechanisms required to define accelerated test protocols

• Need to understand all degradation mechanisms• Need to correlate accelerated testing with real fuel cell life

• Operational variables important to component durability• RH, temperature, potential and potential cycling• Shut-down / start-up variations important to corrosion

• Components• Electrocatalyst (Particle growth)• Membrane (Chemical and mechanical degradation)• GDL (Hydrophobicity loss/gain, porosimetry losses)

• Not sure of future funding status (>FY08)• MEA durability measurements

– Drive cycle testing, operating effects (shut-down), spatial distribution– Identification of degradation mechanisms

• Accelerated testing and durability correlation– Correlate accelerated durability tests to fuel cell performance– Continue to develop accelerated tests for degradation mechanisms

• Component interfacial durability property measurements– GDL / MEA catalyst layer material interfacial contact

Remainder of FY08:– Evaluate surface species leading to hydrophobicity changes

• (both decreasing and increasing)– Evaluate mechanisms leading to change in hydrophobicity

• Examine Nafion / PTFE degradation and carbon bonding

Future Activities