Water Transport Exploratory Studies Office of Hydrogen, Fuel Cells, and Infrastructure Technologies 2007

kickoff meeting

February 13-14, 2007DOE Forrestal Building

Rod BorupMukundan Rangachary, Bryan Pivovar, Yu Seung Kim, John

Davey, David Wood, Tom Springer, Muhammad Arif , Ken Chen, Simon Cleghorn, Will Johnson, Karren More, Peter

Wilde, Tom Zawodzinski

Los Alamos National LabThis presentation does not contain any proprietary or confidential information

Objectives• Develop understanding of water transport in PEM

Fuel Cells (non-design-specific) • Evaluate structural and surface properties of materials

affecting water transport and performance • Develop (enable) new components and operating

methods • Accurately model water transport within the fuel cell• Develop a better understanding of the effects of

freeze/thaw cycles and operation • Present and publish results

Technical Targets/Barriers

N. Vanderborgh, 1994 “...... it’s all about water management ...”

≤1.550kPa (absolute)

Inlet water vapor partial pressure

20102004UnitsCharacteristic

Technical Targets:

Status

Membranes

Approach• Develop understanding of water transport

– Experimental measurement and testing – Characterization – Modeling

• Evaluate structural and surface properties of materials affecting water transport and performance – Measure/model structural and surface properties of material

components – Determine how material properties of GDL, MPL, catalyst layers

& interfaces affect water transport (and performance) – Determine properties change during operation (degradation effects)

• Develop (enable) new components and operating methods – Prevent flooding (high power operation) – Prevent dehumidification (low RH operation - transportation)

• Develop a better understanding of the effects of freeze/thaw cycles and operation – Help guide mitigation strategies.

Prior Work

• Prior relevant work includes (LANL):– Modeling of mass transport losses – Freeze/thaw – GDL characterization (durability)

• More team relevant prior work – Modeling – Electro-osmotic drag measurements

– MEA / GDL manufacturing (etc.) – Characterization – Neutron imaging

0.5

0.6

0.7

0.8

0.9

1.0

MEA Overpotential Losses (Modeling to delineate degradation losses)

• Increase in cathode GDL mass-transport Overpotential

• Hydrophobicity loss

Constant Current Operation @ 0.90 A/cm2

η

η

η

Vi

( )

⇒T⇒ / 15

⇒⇒ /( /cm2

Catalyst-Layer Mass-Transport Improvement from More

Hydrated Electrode Ionomer

Time, tx,GDL

Time, tx,elec

Time, ORR

Time, R-free

Solid Symbols = 0 hr Beginning of LifeGray Symbols = 536 hr Durability Testing

White Symbols = 1014 hr Durability Testing

cell = 80°C Gas Pressure = 15

psig (A/C) RH = 75% / 75% (A/C) Flow Rates = 1.2 2.0

A/C) × 1.5 A equiv.

0.45

• Decrease in cathode electrode mass transport overpotential

• Decrease in HFR

0.40

0.35

0.30

• Increase in ORR overpotential

• S.A. decrease η (V

)

0.25

0.20

0.15

0.10

0.05

0.00

-0.05 0.00 0.25 0.50 0.75 1.00 1.25 1.50

Current Density (A/cm2)

V i R

-free

(V)

Changes in mass-transport (Æ water management)

0.4

GDL Fiber Chemistry and Contact Angle96

78

80

82

84

86

88

90

92

94

Unaged N2, 60°C,

)

f

77

81

85

89

93

97

(θ °)

Hy

Unaged Unaged N2, 80°C, Air, 60°C, Air, 80°C,

Con

tact

Ang

le (d

eg Hydrophobic

Hydrophilic

Increasing Aggressiveness oAging Environment

Con

tact

Ang

le (d

eg)

5 wt% PTFE 20 wt% PTFE

Plain TGP-H Fiber = 80.3

Increasing Graphitization Temperature

drophobic Hydrophilic

SGL GDL 24 Spectracorp 2050-C Toray TGP-H TGP-H 120 TGP-H 090 TGP-H 090 460 hr; 460 hr; 680 hr; 680 hr; GDL Graphite Fiber Type (Plain) #1 (17.2 #2 (17.0 TGP-H 060 TGP-H 090 TGP-H 060 TGP-H 090

wt% FEP) wt% FEP) (17.2 wt% (16.7 wt% (16.9 wt% (17.0 wt% FEP) FEP) FEP) FEP)

• Fiber graphitization can increase single-fiber contact angle ~ 10°

• Both graphitization T and PTFE loading can change the liquid-water wetting regime of the GDL substrate.

Before/After SEM Comparison of PTFE Microstructure of GDL

Fresh After Testing

• Loss of definition delineating the individual PTFE particles – (typically shaped like and ellipsoid and 200-500 nm diameter).

• Features could be associated with changes in liquid water transport characteristics or carbon fiber surface changes.

• Features are also affected by sintering time and temperature.SIGRACET GDL 24BC

Freeze/Thaw cycling

GDL: carbon paper (SGL), Active area: 5 cm2

Confocal laser imaging Dr. Mike Hickner Sandia National Lab

• No systematic degradation observed up to 45cycles • Mechanical Failure @ 45 cycles • Fiber breakage at land/channel edge is a potential failure mechanism

• More tests required to investigate failure mechanisms

Water Transport Exploratory Studies Project initiated in FY2007 for 4 years

Segmented cell studies/

transient operation

Water balance thawing

Conductivity of

down to -40 oC

water content steady-state and transient conditions

2007 2008 2009 2010

Quasi-3D model for simulating electrode layers

in situ water/ice imaging neutron

Start GDL surface properties

aging effects

100 freeze/thaw

crystallinity

MEA properties < 0 oC / effect of freeze/thaw cyclingshutdown

cycles on ionomer Component water removal during

Qtr. 4 based on Reasonable 1D Prediction)

Qtr. 6

Qtr. 10

Qtr. 10

Go/No-Go

o

oC) will be explored in further studies.

Demonstrate 1-D model validity in case studies. (Go/No Go Decision to Proceed to 3D

Demonstrate start up fuel cells from -20 C. Only materials and conditions that meet this decision criterion (start up from –20

z-direction imaging capability by neutrons. If not capable of performing tomography at this point, we will conclude the neutron imaging work.

Electro-osmotic drag coefficient measurements. Activity gradient cells vs. electrophoretic NMR measurements.

Organizations / Partners

• Los Alamos National Lab – (Lead: experimental measurements, modeling)

• Sandia National Laboratory (modeling) • Case Western Reserve University

(characterization, modeling) • W. L. Gore and Associates, Inc. (MEAs) • SGL Carbon Group (GDLs, MPLs) • Oak Ridge National Lab (characterization)• National Institute of Standards and Technology

(neutron imaging)

Budget

DOE Cost Share Recipient Cost TOTAL Share

$6,550,000 $290,811 $6,840,811

96% 4% 100%

FY07 LANL $1000k

Industrial Partners (Gore/SGL/CWRU) $300k Other National Labs (SNL/ORNL) (?) $350k FY07 Total 1650

Input / Needs• Automotive OEM (FC Tech Team) definition of

inlet RH as a function of power/time during operating drive cycles

Wat

ts

Volts

20 min

AM

PS

W

atts

Volts

W

atts

Am

ps

6 min

VO

LTS