Electrode Stability

Xiaoping Wang, Deborah Myers, and Romesh KumarArgonne National Laboratory

May 16-19, 2006

This presentation does not contain any proprietary or confidential information

Project ID # FC 17

Work sponsored by U.S. Department of Energy, Hydrogen, Fuel Cells and Infrastructure Technologies Program

Overview

BarriersBarriers addressed A. DurabilityC. Electrode Performance

InteractionsOak Ridge and Los Alamos National LaboratoriesRegularly providing updates and soliciting feedback from the FreedomCAR Fuel Cell Technical TeamCollaborations with Argonne’s BES-funded groups on characterization

TimelineFormal project start data: October FY’06Project end data: Open Percentage complete: n/a

BudgetTotal funding: ($500 K)Funding for FY06: $350 K

3

Objective

Elucidate rates and mechanisms of loss of electrochemically active surface area (EASA) of polymer electrolyte fuel cell platinum electrodes – Proposed mechanisms of EASA loss

• Platinum dissolution and re-deposition to form larger particles• Platinum dissolution with loss of platinum into the membrane, GDL,

product water, or other electrochemically inaccessible portion of the MEA

• Particle migration and coalescence or transport of neutral Pt onsupport

• Loss of electrochemically accessible Pt due to restructuring of electrode layer (e.g., loss of contact with proton-conducting phase) or impurity adsorption

4

ApproachDetermine mechanisms and rates of Pt dissolution– Dissolution measurements as a function of

• Dissolution time• Potential• Type of electrolyte (adsorbing and non-adsorbing)• Form of Pt electrode and alloying (Pt wire, Pt/C, Pt3Co/C)• Potential cycling• Temperature

– Mechanism of the platinum dissolution reaction using RRDE

Determine particle size, oxidation state, and distribution of platinum• Quantifying the extent of platinum loss in membrane-electrode

assemblies using electrochemical and ex situ analyses• In situ X-ray fluorescence and wide-angle diffraction studies of

membrane-electrode assemblies

Progress vs. FY ’06 Milestones

Determine effect of potential and potential cycling ondissolution of Pt and Pt-Co supported on carbon (06/06)– Polycrystalline Pt for baseline performance

• Completed study of dissolution behavior as a function of both potential and potential cycling

– Pt/C • Completed determination of effect of potential and determined

particle size change• Potential cycling study is near completion

– Pt-Co/C • In progress

Determine effect of temperature on dissolution of Pt and Pt-Cosupported on carbon (09/06)– Design and construction of test apparatus for elevated temperatures is

underway

6

Pt dissolution at 0.90 V reached equilibrium in >48h

10 wt% Pt/C supported on carbon cloth (E-Tek), imbibed with Nafion®

in 0.57 M HClO4

Pt wire - similar behavior

0.0

0.2

0.4

0.6

0.8

1.0

0 20 40 60 80

Time (h)

Dis

solv

ed P

t x 1

09(g

/cm

2 )

0.0

0.2

0.4

0.6

0.8

1.0

0 20 40 60 80

Time (h)

Dis

solv

ed P

t x 1

09(g

/cm

2 )

7

Dissolution behavior indicates that Pt/C is not passivated at >1.2 V

1.E-09

1.E-08

1.E-07

1.E-06

0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5

Potential (Volts vs. SHE)

Dis

solv

ed P

t con

c. (M

)

10

20

30

40

Cur

rent

(µA

/cm

²)

Concentration of dissolved Pt correlates with the current-potential behavior

There exists a potential shift between the concentration and current-potential curves

Equilibrium concentration of dissolved Pt depends on potential

1.E-10

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5

Potential (Volts vs. SHE)

Dis

solv

ed P

t con

c. (

M)

Pt/C-80oC, GM-MITANL-Pt wire (updated with new data)ANL-Pt/C-Nafion-23oC (new data)From Pourbaix

Dissolved Pt equilibrium concentration

– Less dependent on potential than predicted by Pourbaix diagram (in agreement with GM-MIT, Pt/C data)

– Higher than predicted by Pourbaix diagram at < 0.95 V, less at > 0.95 V

Both polycrystalline Pt and Pt/C showed a maximum Pt equilibrium concentration, at 1.1 V and 1.0 V, respectivelyAt >1.2 V, dissolved Pt concentration decreases for polycrystalline Pt, but continues to increase for Pt/C

9

Polycrystalline Pt dissolution behavior explained by complex oxide layer formation

1 .E -09

1 .E -08

1 .E -07

1 .E -06

1 .E -05

0 .6 0 .7 0 .8 0.9 1 .0 1.1 1 .2 1 .3 1 .4 1 .5

P oten tia l (V olts vs . SH E )

Dis

solv

ed P

t con

cent

ratio

n ( M

)

0

50

10 0

15 0

Cur

rent

(µA/

cm²)

1 .E -09

1 .E -08

1 .E -07

1 .E -06

1 .E -05

0 .6 0 .7 0 .8 0.9 1 .0 1.1 1 .2 1 .3 1 .4 1 .5

P oten tia l (V olts vs . SH E )

Dis

solv

ed P

t con

cent

ratio

n ( M

)

0

50

10 0

15 0

Cur

rent

(µA/

cm²)

1 .E -09

1 .E -08

1 .E -07

1 .E -06

1 .E -05

0 .6 0 .7 0 .8 0.9 1 .0 1.11 .E -09

1 .E -08

1 .E -07

1 .E -06

1 .E -05

0 .6 0 .7 0 .8 0.9 1 .0 1.1 1 .2 1 .3 1 .4 1 .5

P oten tia l (V olts vs . SH E )

Dis

solv

ed P

t con

cent

ratio

n ( M

)

0

50

10 0

15 0

Cur

rent

(µA/

cm²)Pt = Pt2+ + 2e-

Pt = PtO + 2e-

PtO + H+ = Pt2+ + H2O

oPt Pt Pt Pt Pt Pt

Pt Pt Pt Pt Pt

o ooPt

Pt Pt

Pt

PtPt Pt Pt Pt Pt

oPt

o o

o Pt PtPt

Pt PtPt Pt Pt Pt Pt Pt

oo oo

oo

o

10

TEM analysis shows growth of Pt particles most significant during electrode prep or cycling pre-treatment

As-received: 1-3 nm, 5 nm

All other samples:- Increase in Pt size- 2-5 nm, 10 nm

No strong effect of potential hold on particle growth

as-received, no Nafion

20 nm

Cycled, held at 0.65 V, 72 h

20 nm

Cycled, held at 0.9 V, 467 h

20 nm

Cycled, held at 1.4 V for 72 h

20 nm

Cycled, 0 to 1.4 V

20 nm

11

The upper limit of potential cycling strongly affects the rate of Pt dissolution

0

20

40

60

80

100

120

140

0 500 1000 1500 2000No. Cycles

Dis

solv

ed P

t con

c. (p

pb)

0.4 - 0.9

0.4 - 1.15 V

0.4 - 1.10

0.4 - 1.05 V

0.4 - 1.00

0.4 - 1.20 VDissolved Pt conc.

– Approaches steady-state value

– Increases with the upper limit potential

10% Pt/C-Nafion0.57 M HClO4Room temperature 10 mV/s

12

Pt dissolution rate is accelerated by potential cycling to >1.0 V

Dissolution rate for potential cycling: Pt/C >> Pt wire

0.0E+00

2.0E-13

4.0E-13

6.0E-13

8.0E-13

1.0E-12

1.2E-12

1.4E-12

0.90 0.95 1.00 1.05 1.10 1.15 1.20E/V vs. SHE

Dis

solu

tion

rate

(g/c

m2 s)

1.0E-16

1.0E-15

1.0E-14

1.0E-13

1.0E-12

1.0E-11

0.90 0.95 1.00 1.05 1.10 1.15 1.20E/V vs. SHE

Dis

solu

tion

rate

(g/c

m2 s)

Potential cycling --- Pt/C-Nafion

Potentiostatic --- Pt/C-Nafion

Potential cycling --- Pt wire

13

Potential cycling to higher potentials showed more extensive Pt particle growth

20nm 0.4 to 1.0 V20 nm

0.4 to 0.9 V

20 nm

0.4 to 1.05 V

• 2 to 4 nm, 6 nm

20 nm

0.4 to 1.2 V

• 2 to 6 nm, 8 nm

• 2 to 6 nm, 10 nm

• 2 to 5 nm, 6 nm

14

Summary

Under potentiostatic conditions– Dissolved Pt equilibrium concentration is less dependent on potential

than predicted by Pourbaix diagram (in agreement with GM-MIT data)

– Dissolved Pt equilibrium concentration is higher than predicted by Pourbaix diagram at < 0.95 V, less at > 0.95 V

– Both Pt wire and carbon-supported Pt nanoparticles (Pt/C) showed a maximum Pt equilibrium concentration, at 1.1 V and 1.0 V, respectively

– At >1.2 V, dissolved Pt concentration decreases for Pt wire, but continues to increase for Pt/C

– The dissolution rates at 0.9 V for Pt wire and Pt/C are comparable

15

Summary (cont.)

Under potential cycling conditions

– Pt/C: concentration of dissolved Pt increased with increasing upper potential limit

– Pt/C and Pt wire: higher dissolution rate with increasing upperpotential limit

Potential Cycling vs. Potentiostatic– Dissolution rate increased with the upper limit potential for potential

cycling, but showed a maximum for potentiostatic condition– At > 1.02 V, higher dissolution rate for potential cycling than for

potentiostatic condition– At < 1.02 V, lower dissolution rate for potential cycling than for

potentiostatic condition– Pt/C yielded much higher Pt dissolution rates than Pt wire

16

Response to FY ’05 Reviewers’ Comments

“Project should expand to include drier conditions and polymer electrolyte interface“

“Probably needs to find ways to do this with more realistic conditions next”– These will be addressed in the second phase of the project

• Membrane-electrode assemblies

• In sulfonic acid medium

“Extrapolations of the present corrosion results for real systems are required”– Work scope will include a modeling effort to correlate Pt dissolution

data with published single cell and stack data (e.g., LANL data)

Future workPt/C high surface area electrode (E-Tek) in HClO4

– Finish determining potential cycling effect with higher upper-limit potentials (1.3 – 1.5 V)• TEM for Pt particle size and support change

– Effect of atmosphere (air and hydrogen) – Effect of temperature (room temperature to 120°C)

Pt3Co/C high surface area electrode (Tanaka) in HClO4

– Potential (0.65-1.4 V)– Potential cycling (0.4 V to 1.4 V)– Atmosphere (air and hydrogen)– Temperature (room temperature to 120°C)

Particle size, oxidation state, and distribution of platinum in MEA under varied experimental conditions (e.g, potentiostatic, potential cycling, different atmosphere, and temperature)

18

Acknowledgments

Jennifer Mawdsley and Nancy Kariuki for TEM work

Y. Tsai and Argonne’s Analytical Chemistry Laboratory for ICP-MS analyses

19

Publications/Presentations

“Effect of Voltage on Dissolution of Pt, Relevance to Polymer Electrolyte Fuel Cells”, Xiaoping Wang, Romesh Kumar, and Deborah J. Myers, Electrochem. and Solid-State Letters, 9 (5), A225-A227 (2006)

“Fundamental Studies of Platinum Electrocatalyst Degradation”, Xiaoping Wang, Deborah J. Myers, and Romesh Kumar, Fuel Cell Durability 2005, Washington, DC, December 8-9, 2005

“Polymer Electrolyte Fuel Cell Cathode Electrocatalysts”, Xiaoping Wang, Romesh Kumar, and Deborah J. Myers, Poster and Abstract, 2005 Fuel Cell Seminar, Palm Springs, CA, November 14-18, 2005