in situ PEM Fuel Cell Water Measurements

International Symposium onDiagnostics Tools for Fuel Cell Technologies

Trondheim, Norway, June 23-24 2009

Presented by: Rod Borup

Los Alamos National LabRod Borup, Rangachary

Mukundan, John R Davey, Jacob

Spendelow, Partha

Mukherjee, Roger Lujan

National Institute of Standards and TechnologyDavid Jacobson, Daniel Hussey, Muhammad Arif

Approach•

Experimentally measure water in situ operating fuel cells–

Neutron Imaging of water

–

HFR, AC impedance measurements–

Segmented Cells (coupled with AC impedance)

–

X-Ray tomography

•

Characterization of materials responsible for water transport–

Evaluate structural and surface properties of materials affecting water transport •

Measure/model structural and surface properties of material components

•

Determine how material properties affect water transport (and performance)•

Evaluate materials properties before/after operation

•

Modeling of water transport within fuel cells–

Water profile in membranes, catalyst layers, GDLs

–

Water movement via electro-osmotic drag, diffusion, migration and removal

Neutron Imaging & PEMFCs

I = Io exp(-t)Io = reference (dry) imageI = attenuated (wet) imagem = attenuation coefficient of waterT = water thickness

•

Subtle changes in the water distribution inside a fuel cell impact performance and durability

•

Neutron Imaging measures small changes at video frame rate –

(amorphous Si Detector)

0123456789

10

0 200 400 600 800 1000 1200Time (sec)

Wat

er C

onte

tnt (

mg/

cm2 )

-0.2

0

0.20.4

0.6

0.8

11.2

1.4

1.6

Res

ista

nce

( cm

2 )

Channel Water (MEA + GDLs + Channels)Land Water (MEA + GDLs)HFR

0.5 amps 34 amps

Wetting TransientNeutron Images of Cell Wetting

During Current Transient

Minimum flows at 82cc/min and 333cc/min at the anode and cathode @ 0.5A. At 34A it is 1.2 and 2.0 stoich flows.

Wetting from 0.5A to 34A 40C and 0/0 inlet RH

Integrated Water Content from Neutron Images During Current

Transient

0

1

2

3

4

5

6

7

8

0 500 1000 1500 2000 2500

Time (sec)

Wat

er C

onte

tnt (m

g/cm

2 )

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Resistan

ce (

cm2 )

Channel Water (MEA + GDLs + Channels)

Land Water (MEA + GDLs)

HFR

34 amps

0.5 amps

Drying TransientDrying from 34A to 0.5A 40C and 0/0 inlet RH

Neutron Images of Cell Drying During Current Transient

Integrated Water Content from Neutron Images During Current

Transient

•

Wetting response is faster (10 –

30 sec) than the reciprocal drying response (~ minutes) •

Wetting response is the result of water produced at cathode which quickly back diffuses to into the membrane.

•

Drying response requires water to move out of the MEA through wetted GDLs.

Neutron Imaging with MCPs

•

10B or natGd

in wall absorbs neutron•

Reaction particles initiate electron avalanche down channel

•

Charge cloud detected with position sensitive anode

•

Spatial resolution limited by channel size and range of charged particle

•

Ultimate resolution about 1 micron

10B 7Li + 4He + Q (2.79 MeV)

25 micron 10 micron

GDL Teflon Loading Effect on Water Content Monitored by Neutron Imaging and AC Impedance

Co-Flow, 80 oC, 172 kPa (abs)Anode: 1.1 stoich. / 50

% RHCathode: 2.0 stoich / 100 % RH

• Charge transfer resistance• Decreases with increasing current• Greater for GDL with 23% PTFE in MPL

• Mass transfer resistance• Increases with increasing current• Greater for GDL with 23% PTFE in MPL

GDL Variation-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Real (z') - Ohm-cm2

GDL AGDL BGDL C

AC ImpedanceCross-section Neutron Imaging

GDL A = 5% Substrate23% MPL PTFE Loading

GDL B = 5% Substrate10% MPL PTFE Loading

GDL C = 20% Substrate10% MPL PTFE Loading

Cathode

Anode

OutletsInlets

Channel

Land

Cathode

Anode

OutletsInlets Incr

easi

ng w

ater

con

tent

5% PTFE Substrate, 23% PTFE MPL

5% PTFE Substrate, 10% PTFE MPL

Dry Image

Channel Land

•

More PTFE in the MPL results in more water in GDLs and channels

•

Mass transport limitations consistent with lower performance of fuel cells with high MPL Teflon loading at high current densities

Water Profiles Nafion 212 Water content comparison for different operating conditions

• Variation of water content as a function of current density/anode stoichiometry• Anode stoich = 3 (simulating anode recycle), dry cathode has lower water content• Anode stoich = 1.2, dry cathode similar water content to fully humidified cell

• Measured Water content in Nafion lower than expected

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 20 40 60 80 100 120 140 160Pixel

Wat

er D

ensi

ty /

mm

H2O

N212_80_80_0_0p5A_3stN212_80_80_0_1p0AN212_80_80_0_1p0A_3stN212_80_80_80_0p5AN212_80_80_80_1p0A

Membrane/CLGDL Edge

Flowfield Land

FlowfieldChannel Edge

•

Membrane/Catalyst Layer is only ~ 5 pixels wide

•

~ 3 pixels for thinner MEAs•

1 pixel = 14.7 microns

GDL H2

O condensation

•

Low constant stoich (1.1/2.0)•

Simulating anode recycle (3.0)•

Flowfield co-flow

•

Anode channel/GDL water:•

With const. anode stoich ~ 1.1•

Disappears with anode recycle •

Anode GDL water may be water condensation (heat pipe effect)

Anode Cathode

Cell Length Water Profiles Co-flow vs. Counter flow

100 / 0 % RH 1.2 / 2.0 St.

MEA GDL/MEA/GDL

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20Cell Length / mm

Water Den

sity / m

m H2O

Counter(An = 1.2St) MEA

Co‐Flow (An = 1.2St) MEA

00.10.20.30.40.50.60.70.8

0 5 10 15 20Cell Length / mm

Water Den

sity / m

m H2O Counter(An = 1.2St) GDL/MEA/GDL

Co‐Flow (An = 1.2St) GDL/MEA/GDL

Counter Flow : I = 1.49 A/cm2; V = 0.27 VHFR = 0.064 Ohm.cm2

Co-flow : I = 1.41 A/cm2; V = 0.095 VHFR = 0.10 Ohm.cm2

• Higher membrane water with counter flow• Membrane water correlates to lower HFR and higher performance with counter flow

AnodeInlet

CathodeInlet

Profile direction

Counter Flow

CathodeProfile direction

Coflow Flow

Anode

GORE™

PRIMEA®

MEA Series 57110GORE, and PRIMEA are trademarks of W. L. Gore & Associates, Inc.

00.1

0.20.30.4

0.50.6

0.70.8

0 5 10 15 20Cell Length / mm

Water Den

sity / m

m H2O Counter(An = 1.2St)

GDL/MEA/GDLGravity Counterflow (Anode ontop = 1.2st) GDL/MEA/GDL

Cell Length Water Profiles Orientation comparison

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20Cell Length / mm

Water Den

sity / mm H2O Counter(An = 1.2St) MEA

Gravity Counterflow (Anode on top = 1.2st)MEA

MEA GDL/MEA/GDL

Counter Flow Inverted: 1.2StI = 1.39; V = 0.385; HFR = 0.067

Counter Flow : 1.2StI = 1.49; V = 0.27V; HFR = 0.064

100 / 0 % RH 1.2/2.0 St. Orientation inverted

AnodeCathodeProfile

direction

Counter Flow

CathodeProfile direction

Counter Flow InvertedAnode

• Membrane water content similar• Cathode on top shows flooding (gravity effect) and loss of performance• Cathode on bottom GDL water lower water content

GORE™

PRIMEA®

MEA Series 57110GORE, and PRIMEA are trademarks of W. L. Gore & Associates, Inc.

In situ Measurement of Membrane Water3 x Nafion

117

0

0.5

1

1.5

2

2.5

3

3.5

4

0 50 100 150 200

pixels

wat

er th

ickn

ess

OCV0.5 A, pump0.59 A, pump-0.57 A, pumpH2/O2, 0.15 A

40 Mil Nafion

MembraneN117= 14

(40 mil ~ 1000 micron)

•

Interface between membranes is hydrophobic•

Water peaks in the middle of each Nafion

117 membrane slice

•

Interface between membrane and catalyst layer maybe hydrophilic

•

Water peaks near each of the catalyst layers (liquid water in catalyst layer pores)

•

Can clearly distinguish membrane profiles•

FWHM (≈

100 m) is much smaller than membrane thickness (≈

585 –

1000 m)•

Water gradient formed at saturated conditions by H2

pump

= 7.5 = 15.3

= 18.7

= 6

Water Gradient

(80

oC

175% RH)

Wat

er T

hick

ness

Wat

er T

hick

ness

Wat

er T

hick

ness

• Membrane conductivity is a f(water

content)(

= # of Water molecules per # of sulphonic

acid sites)• Large literature base measuring water in Nafion• Vast majority of studies were conducted ex situ

•

Prior Neutron Imaging and modeling based off of literature results show large discrepancy (Factor of 4 difference) A.Z.Weber, M.A. Hickner, Electrochimica

Acta

53 (2008) 7668.

Comparison

0

4

8

12

16

20

24

40 60 80 100Relative Humidity

Lam

bda

(H2O

/SO

4-- )

40 C - THIS WORK80 C - THIS WORK40 C (Liquid) - THIS WORK80 C (Liquid) - THIS WORKTom Z (30 C)Tom Z (30 C - Liquid)Hinatsu (80C)

• Cathode under-saturated -

Membrane water increases from

≈

6 to

≈

10 with current• Membrane water gradients observed, much less than in modeling literature• Reasonable agreement with some literature data at low RH

• Do not measure absolute reliance on ‘membrane thermal history’• Membranes will equilibrate at different for 100% RH and liquid water•Observe higher

at equivalent water activity at higher temperatures• Other recent literature on in situ measurements of membrane water content:

Measurement Summary:

~ 10-12 at 100% RH

~ 18-24 liquid H2

OFuel Cell current increases

:at 50/50 % RH’s

5 8.8at 100/100% 8.3 11.5(depends upon current)

Tsushima, Shoji, Water Transport Analysis by Magnetic Resonance Imaging, LANL/AIST Meeting, San Diego, 2008

A. Isopo, V. Rossi Albertini, An original laboratory X-ray diffraction method for in situ investigations on the water dynamics in a fuel cell proton exchange membrane, Journal of Power Sources 184 (2008) 23–28

20 mil Nafion

Membrane (RH Equilibration, Fuel Cell and H2

Pump Operation)

0

0.5

1

1.5

2

2.5

3

-150 -100 -50 0 50 100 150

Pixel

mm

H2O

OCV0.1 A0.2 A

= 8.3

= 11.5

Fuel Cell Operation 100/100 % RH 40 C

•

Water profile is not flat at OCV.

due to edge effects.•

Middle 7 pixels vary by <4% for the 50/50 case and <5% for the 100/100 case.

•

Single Nafion®

20 mil electrolyte•

6mg/cm2

Pt on cathode and anode•

SGL Sigracet®

24 AA (Hydrophilic no MPL)

•

Vertical setup

0

0.5

1

1.5

2

2.5

3

-150 -100 -50 0 50 100 150pixels

mm

/ H

2O

OCV0.1 A pump0.2 A pump0.5 A pump

= 8.3

= 11.5

Hydrogen Pump

•

Membrane hydration is observed to be•

= 5.0 (50/50%) and 8.3 (100/100%)•

increases with water production (Fuel Cell) and with H2

pump (liquid water formation, electro-osmotic drag)•

Observed differences with 40 mil membrane at 80 oC

& super-saturated conditions•

At 100/100, 0.1 A there is no discernible difference between pump operation and normal operation

•

At 100/100, 0.2 A cathode flow field become wet when the cell is operated in fuel cell, however not in H2

pump mode•

Increasing pump current causes decreasing water on the anode and increasing water on the cathode side.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

-50 -30 -10 10 30 50

Pixel

mm

H2O

OCV0.1 A0.2 A

GDL GDL

= 8.8

50/50 % RH, 40 C –

Fuel Cell Operation

= 5

= 7.1

Segmented Polarization Data for

Different Cathode GDLs

•

Cathode GDL:–

GDL 24BC (5/23 wt% PTFE substrate/MPL)

–

GDL 24B”C”5 (5/5 wt% PTFE substrate/MPL)

•

Anode GDL: –

GDL 24BC (5/23 wt% PTFE substrate/MPL)

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.0 0.4 0.8 1.2 1.6Current Density (A/cm2)

Volta

ge (V

)

Segment 1 (GDL 24BC)Segment 5 (GDL 24BC)Segment 10 (GDL 24BC)Segment 1 (GDL 24BC5)Segment 5 (GDL 24BC5)Segment 10 (GDL 24BC5)

100% / 100% RH (A/C)

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2Current Density (A/cm2)

Volta

ge (V

)

Segment 1 (GDL 24BC)Segment 5 (GDL 24BC)Segment 10 (GDL 24BC)Segment 1 (GDL 24BC5)Segment 5 (GDL 24BC5)Segment 10 (GDL 24BC5)

50% / 50% RH (A/C)

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2Current Density (A/cm2)

Volta

ge (V

)

GDL 24BC, 100% / 100% RH (A/C)GDL 24BC5, 100% / 100% RH (A/C)GDL 24BC, 50% / 50% (A/C)GDL 24BC5, 50% / 50% RH (A/C)

Total Cell Performance

Best performance at inlet

Best performance at outlet

Segmented Cell Measurements show where Mass Transport losses dominate versus where IR losses, and kinetic occur

Performance of 2 GDLs~ identical

AC Impedance Data of Different Segments for Different Cathode GDLs

•

GDL 24BC5 maintains higher water content in catalyst layer at high V and high RH (i.e. lower impedance).

•

GDL 24BC has better water management at low V and low RH.

-0.10

0.10

0.30

0.50

0.70

0.90

0.0 0.2 0.4 0.6 0.8 1.0 1.2Zr (Ω·cm2)

-Zj (Ω

·cm

2 )

GDL 24BC, Seg 1GDL 24BC5, Seg 1GDL 24BC, Seg 10GDL 24BC5, Seg 10

10 kHz120 mHz

12 Hz

Vcell = 0.84 V100% / 100% RH (A/C)

-0.10

0.10

0.30

0.50

0.70

0.90

0.0 0.3 0.6 0.9 1.2Zr (Ω·cm2)

-Zj (Ω

·cm

2 )

GDL 24BC, 0.84 VGDL 24BC5, 0.84 VGDL 24BC, 0.54 VGDL 24BC5, 0.54 V

100% / 100% RH (A/C)

Total Cell

10 kHz 120 mHz

12 Hz

-0.1

0.1

0.3

0.5

0.7

0.9

0.0 0.4 0.8 1.2

Zr (Ω·cm2)

-Zj (Ω

·cm

2 )

GDL 24BC, Seg 5GDL 24BC5, Seg 5GDL 24BC, Seg 10GDL 24BC5, Seg 10

10 kHz

120 mHz

8.3 Hz

Vcell = 0.54 V50% / 50% RH (A/C)

Charge x-ferResistance

Mass Transport problems

Less Mass Transport problems

Charge x-ferResistance

MTResistance

Neutron Imaging of Ice Formation During Operation at -10 oC

•

Neutron imaging of ice formation in a 50 cm2

fuel cell operated at 0.5 V at -10 oC.

•

Calculated/measured water/ice accumulation from current and neutron imaging in the fuel cells track

Water/Ice accumulation @ -10 oC

02468

101214

0 200 400 600 800 1000Time (sec)

Wat

er A

ccum

. (cm

3 cm

-2)

0.0E+00

2.0E-04

4.0E-04

6.0E-04

8.0E-04

1.0E-03

1.2E-03

1.4E-03

1.6E-03

Cumulative Current Density

Measured water thickness

Calculated water volumeTota

l cha

rge

(C)

Incr

easi

ng w

ater

con

tent

0 -

100 sec 800 -

900 sec

Freeze: High resolution imaging

• Location of frozen water (ice) depends on operating temperature and current density•

Water distribution closer to the cathode catalyst layer with:

•

Decreasing temperature•

Increasing current

•

Greater water formation possible at higher temperatures and lower current densities

SerpentineCell

Distance (m)

Wat

er (m

m)

‐0.06

0.04

0.14

0.24

0.34

1000 1500 2000Pixel

mm W

ater

m10C,0.225A/cm2

m20C,0.043A/cm2Anode

GDL CathodeGDL

MEA

1000

0.3

0.2

0.1

0.0

X-RAY Tomograph

Image of water in GDL Pores

GDL

KaptonWater

MPL side against KaptonKapton

•

Evaluate pore size density and distribution

•

Observe GDL compression•

Observe/monitor water diffusion

•

Evaluate water wetting and water diffusion pathways

Water Inside SGL GDL Pores

Water

(23% PTFE MPL, 20% Substrate) (5% PTFE MPL, 5% Substrate)

Water

MPL

Water

Significantly more water in GDL pores as cross-

section approaches water layer

Summary•

Varying MPL and substrate Teflon loadings and cell operating conditions

•

Neutron imaging, AC impedance, HFR, X-Ray Tomography•

Equilibrium water content in the membrane, how membrane water content changes with RH, T, current and water production

•

Segmented cell operation•

Response of GDL and membrane water to transients

•

Fast membrane wetting•

Slow GDL de-wetting, followed by membrane drying

•

Freeze

Future Work•

Experimental and Characterization

•

3-D X-Ray tomography during operation observing water transport in GDL pores

–

Identify hydrophobic pores vs. hydrophillic

pores–

Identify liquid water pathways in GDLs

•

Incorporate 3-D X-ray tomography PSD into Capillary Pressure Simulation

Thanks to

•

U.S. DOE -

EERE Hydrogen, Fuel Cells and Infrastructure Technologies Program for financial support of this work–

Program Manager: Nancy Garland