advanced cathode catalysts and supports for pem fuel cells · this presentation does not contain...

1

Advanced Cathode Catalysts and Supports for PEM Fuel Cells

Mark K. Debe 3M Company May 15, 2012

Project ID: FC 001 This presentation does not contain any proprietary, confidential, or otherwise restricted information

2012 Annual Merit Review DOE Hydrogen and Fuel Cells and

Vehicle Technologies Programs

3M Advanced Cathode Catalysts …… ………………... 2012 DOE Hydrogen Program Review, May 14-18

2

Overview Timeline

Project start : April 1, 2007 Project end : (98% complete) Original - March 31, 2011 w/No Cost Ext. – June 30, 2012

Barriers A. Electrode and MEA Durability B. Stack Material & Mfg Cost C. Electrode and MEA Performance

DOE Technical Targets Budget Total Project funding $10.742 MM $8.593 MM DOE and FFRDC $2.148 MM 3M share

Allocated in FY11: $ 450,000 Invoiced in FY11: $ 852,888 Remaining for FY12: $ 244,704

Dalhousie U. (J. Dahn, D. Stevens) JPL (C. Hays) ANL (N. Markovic, V. Stamenkovic) GM (E. Thompson, stack testing)

Partners

Electrocatalyst / MEA 2017 2020

Lifetime w/cycling (Hrs) 5000 5000

% Loss in mass activity < 40 < 40

Mass Activity @ 0.9V (A/mg) 0.44 0.44

Total PGM (g/KW rated) 0.125 0.125

Performance @ Rated Power (W/cm2 ) @ 0.8V

1 0.24

1 0.24

Additional Interactions Nuvera Fuel Cells, Other OEM’s, Proton Onsite, Giner Inc., ANL (Ahluwalia modeling group)

Project Management - 3M (A. Steinbach, M. Kurkowski, S. Hendricks, A. Hester, P. Kadera, G. Vernstrom, M. Debe)

3

Objectives: Development of a durable, low cost, high performance cathode electrode (catalyst and support), that is fully integrated into a fuel cell membrane electrode assembly with gas diffusion media, fabricated by high volume capable processes, and is able to meet or exceed the 2015 DOE targets.

Relevance and Approach

Approach: Development of advanced cathode catalysts and supports based on 3M’s nanostructured thin film (NSTF) catalyst technology platform. Optimize integration with membrane and gas diffusion media for best overall MEA performance, durability and cost.

Primary Focus Topics for Past Year: Short stack testing with PtCoMn based NSTF electrodes

Completion of 1st short stack testing for performance Down-selection and fabrication of final MEA components for 2nd durability stack Initiation of 2nd stack durability cycling protocol

Development work on Pt3Ni7 NSTF catalysts Work towards optimization of deposition and annealing processes Work towards optimization of ex-situ de-alloying and roll-to-roll pilot level scale-up Addressing membrane integration issues


4

5/15/12

Relevance and Approach: Project Timeline and Milestones

Budget Period 1 Budget Period 2 4/1/07 01/01/10

= Go-No Go for Extension of Task

= Go-No Go for Stack Testing

= Go-No Go for Large Area, Single Cell Durability Tests

Task 1.3 – Activity Optimization

Task 1.1 – ECSA Optimization

Task 1.2 – Cat. Fundamentals

Task 2 – Cat. Durability Gains

Task 5.1 PEM Integration, Task 5.2 GDL Integration

Task 3

Task 4 – Support durability

Task 5.3 - Stacks

Task 6 – Start-up, conditioning

Q12 Q16 Q13 Q14 Q15

9 month and 6 month no cost extensions for stack testing.

= Task End

Q17 Q1 – Q11 Q12 – Q21


Short stack testing with PtCoMn based NSTF electrodes (Task 5.3) Completed 1st 29 cell rainbow short stack performance testing at GM to down-select the

MEA configuration. Down-selection from 6 to 1 configurations successfully made. Initiated durability cycling tests with 2nd short stack (20 cells with one type of 3M MEA).

Met 2017 CV cycling and OCV targets with MEA type used in 2nd short stack testing 30,000 CV cycle test: Demonstrated 10 + 7 mV loss at 0.8 A/cm2, 16 + 2% loss of EC

surface area, and 37 + 2 % loss of mass activity (Task 2). Met 3M OCV hold test: 570 hours with OCV loss = 13 % under 50 kPa H2 overpressure

Development work on Pt3Ni7 NSTF catalysts (Task 1) Extended enhanced catalyst deposition process improvement (P1) from pure Pt and

PtCoMn to Pt3Ni7 , obtaining same dramatic gains in Pt(hkl) grain size with simpler, more cost effective coating process.

Screened over 100 different de-alloying conditions for impact on fuel cell performance. Developed 240x faster ex-situ de-alloying process than initial nitric acid bath conditions. Developed roll-to-roll pilot level scale-up with 240x faster dealloying process conditions. Achieved 0.4 – 0.45 A/mg-Pt mass activity with 100% roll-to-roll fabricated, dealloyed

and SET- “annealed” catalyst at 0.121 mg/cm2 loading of Pt3Ni7 for cathode use. Achieved 0.16 gPt/kW at 0.65 V,80 oC and 200kPaa using 0.15 mg/cm2 total Pt in MEA.

Developed model to explain a fundamental NSTF extended surface catalyst property Higher current density at low loadings due to ensemble packing of NSTF ext. surface cat.

5

Major Technical Accomplishments Since Last Review (5/10/11) Technical Accomplishments and Progress


The various configurations showed a wide range of performances which facilitated the MEA configuration down-selection for Stack 2 durability testing.

MEA Configuration 1 performed best, had best stability under all GM tests conditions.(MEA Config. 1 was down-selected, but not ultimately used in Stack 2).

For all GM test conditions, the stack performance of each MEA configuration was significantly lower than the same MEA’s in 50 cm2 cells tested at 3M. 6

Short stack 1 testing (2 week plan, 4.5 months actual) Technical Accomplishments and Progress


Objective: Test 6 MEA configurations and down-select to 1 based on beginning of life performance, for final Stack 2 testing.

Stack 1 : 29 cells Break-in conditioning Extensive “debugging” of

low performance at GM, 3M. Beginning Of Life testing

from 4/10/11 to 6/28/11. Reversible decay tests from

~ 7/21/11 to 8/3/11. Durability test protocol

development completed 8/29/11.

CCM ID PEM Anode Cathode S1622 Cells

3M-24um (w/add. 2) 0.05 P1 PtCoMn

3M-24um (w/add. 1) 0.05 P1 PtCoMn

Config. 2 3M-24um (w/add. 2) 0.05 P1 PtCoMn 0.10 P1 PtCoMn 5-8, 22-25

0.05 P1 PtCoMn

0.05 P1 PtCoMn

Config. 6 0.05 P1 PtCoMn 0.15 P1 PtCoMn 17,18

Config. 7 0.05 P1 PtCoMn 0.10 P1 PtCoMn 19-21

Config. 8 3M-24um (w/add. 1) 0.05 P1 PtCoMn 0.15 P1 PtCoMn 1-4, 26-29

Config. 1

Config. 3

0.15 P4 PtCoMn + SET 9-12

13-16

3M-X

0.15 P1 PtCoMn3M-S

0.0 0.5 1.0 1.5 2.00.4

0.5

0.6

0.7

0.8

0.9

1.0

NSTF 2009 Best of ClassMEA tested at both GM (red)

and 3M (blue) in 50 cm2 Cells (3M HCT conditions).

Cell V

olta

ge (V

olts

)

J (A/cm2)

3M Single Cell "HCT Cond." v. GM Stack 1 "Cond. 1"

Config. 1 MEA's at 3M, 50 cm2 cells (HCT)

Stack 1(GM) Four Config. 1MEA's

Task 5.3 Final Stack testing GM-3M data for stack 1\DOE 6 stack 1 CCM S6 032911-072911-graph 27

Technical Accomplishments and Progress


Break-in conditioning was problematic - test station water identified as an issue. Best MEA performances (MEA Config. 1) were much worse than past history with 50 cm2

single cells at both GM and 3M at low pressures (bottom right figure). Config. 1 MEA’s (same lot as in stacks) also underperformed 3M expected results Catalyst ORR metrics and surface area’s, and CCM’s SEM characterization were normal. Possible reasons for under-performance: (3M high current test (HCT) ≈ GM conditions) More complete conditioning possible with single cells; water purity Differences due to stack vs. single cell flow field effects (see Stack 2 discussion). Membrane ionomer properties or contamination.

Short stack 1 testing: issues 4/1

1/201

1

Initia

l CA

Brea

k-in

4/21/2

011

+12 C

A Cy

cles

5/5/20

11

+12 C

A Cy

cles

5/9/20

11

+12 A

N, 2

CA C

ycles

5/16/2

011

+7hr

Flus

h5/1

8/201

1

24hr

40C

Cycle

s6/2

2/201

1+1

6hr F

lush

6/28/2

011

+16h

r Flus

h, St

and2

0.45

0.50

0.55

0.60

0.65

3287

6

GM Stack 1 Break-in History - Test Condition 1 at 1.20 A/cm2

Cell V

olta

ge (V

olts

)

MEA Configs.

1

P ~ 7.5 psig 80 oC

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.4

0.5

0.6

0.7

0.8

0.9

1.0

C: Task 5.3 Final Stack Testing\3M S1 Stack data - graph 10

5 Cell 3M Short Stack Build 110411 FC12060 359.DAT 50 cm2 Single Cell FC12645 437.DAT 50 cm2 Single Cell

Cell V

olta

ge (V

olts

)

J (A/cm2)

90/79/79 oC; 250/250 kPaa H2/Air, CS(2,100)/CS(2.5, 167); Stack- 3min/pt

22 psig

(B)

8

With configuration 1 MEA, 3M’s 312 cm2 short stack gives similar performance to 50 cm2 single cells at high pressure but not at low pressure (3M Stack flow fields very similar to single 50 cm2cell) (A,B).

3M short stack gives higher performance than GM short stack (C).

These results suggest possible flow field effect contribution to under-performance.

Technical Accomplishments and Progress Short stack 1 testing: comparison with 3M short stack testing


22 psig

Config. 1

5 cell 3M Short Stack Build 110411

0.00 0.15 0.30 0.45 0.60 0.75 0.90 1.05 1.200.4

0.5

0.6

0.7

0.8

0.9

1.0

FC21990 50 cm2 Single Cell

Cell V

oltag

e (V

olts)

J (A/cm2)

75/70/70 oC, 0/0psig H2/Air

5 Cell 3M Short StackBuild 110411

C: Task 5.3 Final Stack Testing\3M S1 Stack data - graph 11

0 psig

Config. 1 (A)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.400.450.500.550.600.650.700.750.800.850.900.951.00

3M Stack (110411)5 Cell Av. of Config. 8153 kPag ; 90/79/79 oCCS (2,2.5)

GM Stack 1 vs 3M Stack - MEA Configuration 8

Cell V

olta

ge (V

)

Current Density (A/cm2) S1622 - Pressure Series.-graph 2

GM Stack 18 Cell Av. of Configuration 8153 kPag; 80/40/42 oCCS (2,2)

MEA configuration 1 average performance improves with pressure similarly in GM stack as in 3M single cell, using ANL test conditions, but stack still lags single cells.

GM stack data is 4 cell average of the best performing MEA configuration type 1. Test conditions are

those from A NL (Ahluwalia) systems modeling group. Strong effect of

pressure consistent with mass transport issues. Cell compression in

GM stack evaluated and believed not to be an issue.

Technical Accomplishments and Progress Short stack 1 testing: GM stack vs 3M 50 cm2 cells - ANL pressure series.


0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

Cell V

olta

ge (V

olts

), HF

R (o

hm-c

m2 )

ANL Condition 2.180/73/73C, 1/1atm H2/Air, CS(2,100)/CS(2.0, 200)

GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)

4 cell average of best performing MEA Configuration

in 29 cell GM stack test.

0 psig

15 psig

7.5 psig


in 29 cell GM stack test.ANL Condition 2.2

80/70/70C, 1.25/1.25atm H2/Air, CS(2,100)/CS(2.0, 200)GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)Ce

ll Vol

tage

(Vol

ts),

HFR

(ohm

-cm

2 )

6, 50 cm2 single cell testsat 3M with same MEA type that performed best in stack.




in 29 cell GM stack test.

ANL Condition 2.480/56/56C, 2.0/2.0atm H2/Air, CS(2,100)/CS(2.0, 200)

GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)

Cell V

olta

ge (V

olts

), HF

R (o

hm-c

m2 )

J (A/cm2)

Case 2.1,2.2,2.4,2.5 - Impact of Pressure

Task 5.3 Final Stack Testing\GM-3M data for Stack 1\ANL Systems Model - graph 13

0.78 W/cm2

0.26 gPt/kW

22.5 psig

ANL Condition 2.580/40/40C, 2.5/2.5atm H2/Air, CS(2,100)/CS(2.0, 200)

GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)Cell V

olta

ge (V

olts

), HF

R (o

hm-c

m2 )

J (A/cm2)

1.0 W/cm2

0.2 gPt/kW

9

10

Effect of contamination at 3M Laboratory water (18 Mohm-cm, but new unit):

High current density performance in galvanodynamic scans dramatically affected. Activity measurements at 900mV show co-adsorption of OH- and probably Cl-.

Technical Accomplishments and Progress Short stack 1 testing: Illustration of water contamination effect

ORR Absolute Activity @ 1050s: Contaminated DI: 7.5mA/cm2-planar Analytical Lab DI: 17.6mA/cm2-planar

0 200 400 600 800 1000 12000.00

0.01

0.02

0.03

0.04

0.05

Activity w/ contaminantfalls off more quickly

C:\Users\US117400\Documents\DOE-6 032212\Dirty water effects\DirtyWaterImpact 040212-[graFC023852PDSHCTORRCV]

J (A

/cm

2 ) @ 0

.900

V

Time (seconds)

80/80/80C, 150 kPa H2/O2

696/1657SCCMPSS(0.900V, 20min)

Activity same immediately upontest start (contaminants desorbedat low E prior to ORR scan start)

Analytical Lab DI (red curve): 1.2 + 0.5 µM Cl- Contaminated waters : 16.9 + 0.3 µM Cl-

(Averages of 5 samples by IC)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

C:\Users\US117400\Documents\DOE-6 032212\Dirty water effects\DirtyWaterImpact 040212-[Graph3]

Analytical Lab DI

Mixture of clean and contaminated waters

J (A/cm2)

80/68/68C, 7.35/7.35psig H2/Air, CS(2,100)/CS(2.5, 167)GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)

Cell V

olta

ge (V

), an

d HF

R (o

hm-c

m2 )

A/C: 0.05/0.10 mgPt/cm2 of PtCoMn


Analytical Lab DI

11



Short stack 2 testing: Objectives, status and issues Objective: Demonstrate 3000 hours of durability load cycling using 29 cell short stack with 20 cells devoted to the down-selected 3M MEA.

CCM production issues and membrane availability forced a change in final MEA configuration from the original down-selected configuration (1 ) to a combination of configurations of 1 and 3, using new, 3M experimental membrane.

CCM’s and GDL’s delivered to GM on 1/15/12, ~ 6 months after plan. New GM stack plates fabricated Stack 2 build completed on 2/1/12 Break-in conditioning completed ~ 3/1/12. Durability testing started 3/16/12. 3M discovered after shipment, that CCM’s sent used an underperforming membrane lot

• Impact is significant loss of high current density performance in 50 cm2 cells. • Likely possibility that stack 2 durability testing will be negatively impacted. • However single cell MEA accelerated stress tests still meet 2017 targets.

Stack 2 Performance to date Stack 2 performance fell below that of stack 1 under most conditions tested. First four sets of stack 2 durability cycling completed, 350 hours, by 3/29/12. Significant decay rates observed during first four sets of 1500 durability cycles/set. GM replaced 3 weak cells at 350 hours. Stack restarted 4/11/12. Some improvement under some conditions after re-start.

Status and Issues:

12



Short stack 2 testing: Issues – Background for Underperforming MEA’s Gradual performance decay over much of 2011 realized in this project’s single cell testing Myriad other CCM component changes masked what was occurring Also question about possible DI water quality in this time period, but not confirmed. Discovery of membrane performance issue after shipment of MEA’s for Stack 2 Problem traced to inadvertent release of experimental PEM lot that had been put on-hold.

Most obvious effect of membrane lot issue is loss of high current density performance.

50 cm2 cell tests of CCM’s from same lot as used for Stack 2 (blue stars) show:

• Catalyst specific and mass activities were a little low in as-made membranes.

• ECSA’s and HFR are both normal.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.4

0.5

0.6

0.7

0.8

0.9

1.0

Anode: PtCoMn, 0.05 mgPt/cm2

Cathode: PtCoMn, 0.15 mgPt/cm2

PEM: Various 3M-experimental

C:\Users\US117400\Documents\DOE-6 032212\Reviews for DOE\2012 Annual Merit Review\DOE6Stack2 3M-S A762411 121611-[Graph11]

Cell V

olta

ge (V

olts

), HF

R (o

hm-c

m2 )

J (A/cm2)

80/68/68C, 7.35/7.35psig H2/Air, CS(2,100)/CS(2.5, 167)GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)Upscan (high->low J) only.

Underperforming Membrane Lots

Good Performing Membrane Lots

CCM Lots used for Stack 2

As-made Acid washed Acid + H2O2 washing0.00

0.05

0.10

0.15

0.20

0.25

0.30

Mass Activity at 900 mv Specific Activity at 900 mV

Membrane Treatment

Mas

s Ac

tivity

at 9

00 m

v

0.0

0.4

0.8

1.2

1.6

2.0

2.4

Spe

cific

Activ

ity a

t 900

mV

Anode: PtCoMn, 0.05 mgPt/cm2

Cathode: PtCoMn, 0.10 mgPt/cm2

PEM: 24 micron, 850 EW

13

Technical Accomplishments and Progress Short stack 2 testing: Issues - Membrane performance issue 50 cm2 single cell tests with best 3M-S experimental PEM (high performing) and similar catalysts show

significantly better performance over that of MEA’s used in stack 2 (left figure). Limiting current density and ORR activities are strongly affected. Membrane cleaning (regular 3M PEM w/ same ionomer lot as stack 2 PEM) has large effect on PtCoMn

ORR specific and mass activity: • Activities in as-made PEM are below historical averages (2011 AMR, slide 17) of 1.75 mA/cm2

Pt and 0.16 A/mgPt by -24% and -16% respectively.

• Acid washing increases the mass and specific activities above the historical values. • Acid + peroxide cleaning increases the specific and mass activities to +40% and +60% above the

historical values respectively. Membrane cleaning does not improve the reduced limiting current density issue!


0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

7.5 psig

Similar MEA as used for Stack 2 but with best 3M-S PEM

C:\Users\US117400\Documents\DOE-6 032212\Reviews for DOE\2012 Annual Merit Review\0811354A DOE Electrocatalyst Durability 030712-[Graph1]

Cell V

olta

ge (V

), HF

R (o

hm-c

m2 )

J (A/cm2)

80/68/68 oC, 7.35/7.35 psig H2/Air or as noted, CS(2,100)/CS(2.5, 167)GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)

Same MEA lotsas used for Stack 2

0.05/0.15 mgPt/cm2

22 psig

14

RH cycle tests of MEAs from same lot of CCM as used for Stack 2 are still being completed: The first CCM failed (developed hole) after

1060 hours or 15,900 cycles, short of 20,000 cycle target (see bottom left and right). Second CCM just starting and has lower leak

rate to start (see bottom left, red curve.) OCV Hold test of MEA from same lot of CCM

as used for Stack 2 has passed 500 hr target (see top right).

Technical Accomplishments and Progress Short stack 2 testing: RH Cycling and OCV Hold lifetime tests with Stack 2 MEA

**RH Cycle Running Conditions: • Wet Stream: 1000 sccm, 150%RH • Dry Stream: 2000sccm, 0%RH • 80°C

Cell Temperature, 0/0psi

• 2 min wet, 2 min dry cycle • Leak check every 5 hours (physical

sccm leak measurement instead of crossover current measurement across MEA)

** Based on: Table D-4 Membrane Mechanical Cycle and Metrics, (Test using a MEA)Table revised December 10, 2009,

Target Tables from U. S. Drive Fuel Cell Technical Team Technology Roadmap.


0

1

2

3

4

5

6

7

8

9

10

0 200 400 600 800 1000 1200

Leak

rate

(scc

m)

Time (hours)

FC023462FC023856

Failure Rate = 10 sccm

4/11/12, 303 hrs

Failed at 1060 hours

15,900 Cycles

Humidity Cycle Lifetime Testing

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 100 200 300 400 500 600

Volta

ge

Time (hours)

OCV Chemical Lifetime Testing, Stack 2 MEAAnode Overpressure OCVStation Shutdown

Cell T = 90 oC A/C Dew Points = 61oC/61 oC200/150 kPaa H2/Air

2 MEA’s : definition (same as used in Stack 2) Cathode Catalyst: 0.15 mg/cm2 PtCoMn (P4+SET) Anode catalyst: 0.05 mg/cm2 PtCoMn (P4) GDL’s: 3M standard 2979 on A/C PEM: 3M-Supported with additive, 18 micron thick

15

Test Protocol (US Drive FC Tech. Team)

30,000 saw-tooth cycles at 50 mV/sec; 0.6 - 1.0 - 0.6 V; 80/80/80 oC, 50 cm2 100/100 kPa H2/N2; 200/200SCCM.

First MEA polarization curves taken periodically using both DOE and 3M (HCT) protocols. Second MEA ran 30,000 cycles without stopping to save time. Both exceed 2017 targets.

Technical Accomplishments and Progress Stack 2: Objectives, status and issues: Cathode CV cycling stress test


0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

After 30k Cycles-3mV @ 0.8A/cm2

FC023466 339.RAW Initial FC023466 462.RAW After 10k FC023466 572.RAW After 20k FC023466 798.RAW After 30k

C:\Users\US117400\Documents\DOE-6 032212\Reviews for DOE\2012 Annual Merit Review\DOE Electrocatalyst Cycle 020111-final update 041012-[graDOEFC23466]

Cell V

olta

ge (V

olts

) and

HFR

(ohm

-cm

2 )

J (A/cm2)

80/59/59C, 7.35/7.35psig H2/Air (OUTLET), CS(1.5,100)/CS(1.8, 167)GDS(0->2->0A/cm2, 0.2A/cm2/step, 180s/pt, 0.4V lim)

DOE Polarization Curve During Electrocatalyst Cycle Durability Testing

Initial

HFR

First MEA

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.60.00.10.20.30.40.50.60.70.80.91.0

After 30k Cycles non-stop-16mV @ 0.8A/cm2

FC023844 503.RAW INITIAL FC0238844 728.RAW AFTER 30K CYCLES FC023844 503.RAW INITIAL FC0238844 728.RAW AFTER 30K CYCLES

C:\Users\US117400\Documents\DOE-6 032212\Reviews for DOE\2012 Annual Merit Review\DOE Electrocatalyst Cycle 020111-final update 041012-[graDOEFC23844]

Cell V

olta

ge (V

olts

), HF

R (o

hm-c

m2 )

J (A/cm2)

80/59/59C, 7.35/7.35psig H2/Air (OUTLET), CS(1.5,100)/CS(1.8, 167)GDS(0->2->0A/cm2, 0.2A/cm2/step, 180s/pt, 0.4V lim)

DOE Polarization Curve During Electrocatalyst Cycle Durability Testing

Initial

HFR

Second MEA

16

ORR and Polarization Curve Metrics vs Number of Cycles (2012 vs 2011)

Surface area losses of (-14, -18)% meet 2017 target ( < 40% loss of initial area). DOE Polarization curve voltage losses at 0.8 A/cm2 of (-3, -16)mV meet target (< 30 mV). Mass activity loss of (-34, -39)% meet 2017 target ( < 40% loss of initial mass activity).

Technical Accomplishments and Progress Stack 2: Objectives, status and issues: Cathode CV cycling stress test


Stack 2 type MEA’s meet all 2017 targets for this test, even with contaminated PEM.

0 5 10 15 20 25 309101112131415

Activity Metrics of Series' MEAs

SEF

(cm

2 -Pt/c

m2 -p

lana

r)

0 5 10 15 20 25 300.0100.0150.0200.0250.030

OR

R A

bsol

ute

Activ

ity(A

/cm

2 -pla

nar)

0.6-1.0V Cycles (Cycles/1000)0 5 10 15 20 25 301.0

1.5

2.0

2.5

OR

R S

peci

fic A

ctiv

ity(m

A/cm

2 -Pt)

0 5 10 15 20 25 300.76

0.77

0.78

0.79

0.80

DOE

Mea

s V

@ 0

.20A

/cm

2 (Vol

ts)

0 5 10 15 20 25 300.60

0.62

0.64

0.66

0.68

C:\Users\US117400\Documents\DOE-6 032212\Reviews for DOE\2012 Annual Merit Review\DOE Electrocatalyst Cycle 020111updated-[Graph4]

DOE

Mea

s V

@ 1

.00A

/cm

2 (Vol

ts)

DOE Polarization Curve Metrics During Electrocatalyst Cycle Durability Testing

0 5 10 15 20 25 300.35

0.40

0.45

0.50

0.55

DOE

Mea

s V

@ 1

.50A

/cm

2 (Vol

ts)

0.6-1.0 kCycles (Cycles/1000)

0 5 10 15 20 25 300.650.660.670.680.690.700.71

DOE

Mea

s V

@ 0

.80A

/cm

2 (Vol

ts)

2011

2012

2012

2012

2012

2012

2011

2011

2011

2011 2011

2011

2012

2012

17



Short stack 2 testing: Beginning of Life Performance

3M Single Cell HCT Cond.: T (oC) Cell/An/Ca DP’s = 80/68/68 oC, H2/air pressure = 150/150 kPaa An/Ca Stoichs. = 2/2.5

Stack Cond.

T (oC) An/CaSt.

An RH in (%)

Can RHout (%)

Pressure

1 ~ 82 ~1.5/1.8 25 82 Variable

2 ~ 75 ~1.5/1.8 30 85 Variable

3 ~ 65 ~2 /1.8 30 >100 Variable

4 ~ 78 ~1.5/1.8 20 65 Variable

5 ~ 78 ~2.0/1.8 >100 >100 Variable

Stack 2 testing beginning of life performance and testing conditions Stack 2 performance was evaluated under five different sets of operating conditions (table). Stack 2 performance was much lower than expected under all conditions. Stack 2 performances did not vary significantly from driest to wettest conditions. Stack 2 performances under-performed 3M 50 cm2 single cell with same MEA lot as in stack. Stack 2 performances were generally below stack 1 performances at start of durability cycling. Average Hupd cathode surface areas were ~ normal at 8.2 m2/g. Shorting resistances were lower than standard MEA’s. Compression paper results normal.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.60.0

0.2

0.4

0.6

0.8

1.0

GM Condition 4, 2/14/12 GM Condition 3, 2/14/12 GM Condition 3 - HFR, 2/14/12 GM Condition 1, 2/17/12 GM Condition 2, 2/17/12 GM Condition 2-HFR 2/17/12 3M Single Cell, Stack 2 MEA, 3M HCT conditions GM Condition 5, 2/14/12 GM Condition 5-HFR , 2/14/12

Cell V

olta

ge (V

) and

HFR

(ohm

-cm

2 )

Current Density (A/cm2)DOE-6 032212\Task 5.3\Stack 2 Data\S1623 Cruves-graph 4

3M Single CellHCT Condition

Stack 2Conditions 1-5

18



Short stack 2 testing: Durability cycling status Stack 2 showed significant degradation after each of four, 1500 cycle sets: Durability cycling modified US Drive Fuel Cell Tech Team recommended protocol:

• Higher pressure, controlled current ramp rate, voltage control, fine tuning… Results after 4 sets of 1500 cycles:

• Crossover leak rate of stack and hydrogen take-over in cells were high • HFR increased with time but cannot account for lost performance • 2 point (beginning and end) decay rates for 4 Cycle sets over 250 hours are much higher

(3x-8x) than expected (bar-graph). • There are significant fluctuations in performances between each of the 4 Cycle sets. • After replacing weak cells 4/12/12, 100mV increase at 1.5 Acm2 under Cond. 2.

0

100

200

300

400

500

600

DOE-6 032212\Task 5.3\Stack 2 Data\S1623 Cruves-graph 31.00.6

Condition 2 Condition 1 Condition 5

Cell V

olta

ge D

ecay

Rat

e (µ

V/Hr

)

Current Density (A/cm2)0.2

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

0.0 0.5 1.0 1.5 2.0

Hig

h Fr

eque

ncy

Res

ista

nce

( Ω.c

m2 )

Aver

age

Cel

l Vol

tage

(V)

Current Density (A/cm2)

NSTF @ BoL

NSTF @ 350h

Performance Decay, Condition 2

19



Impact of flow field: Study started 3/25/12 comparing 1st of 9 different types of flow field Tests are using the same MEA lot as used for Stack 2. Testing on multiple stations. First alternate FF completed and compared to 3M standard 50 cm2 quad-serpentine: Main difference is width of lands and channel dimensions:

• Quad-Serpentine: 4 parallel channels, 10 loops; Channel W, D; Land = 0.8 mm,1 mm; 0.8 mm • First Alternate Serpentine: 6 parallel channels, 2 loops; Channel W,D; L = 2 mm, 0.25 mm; 2 mm

Both types have similar pressure drops (see figure on right). Much poorer performance at 150kPaa pressure with wide lands of first alternate serpentine. Performance with alternate FF improves only slightly with pressure increases (not shown).

Short stack 2 testing: Issues – Investigating possible flow field effects

0.0 0.4 0.8 1.2 1.6 2.00

1

2

3

4

5

Anode, Hex -2 Loop

Cathode, Hex -2 Loop

Cathode, Quad. - 10 Loop

Anode, Quad. - 10 Loop

C:\Users\US117400\Documents\DOE-6 032212\Task 5.3 Final Stack Testing\Flow field comparisons\0811354A FF Sens 040912-[Graph6]

FC023466 291.RAW QuadSerpQ2 FC023910 285.RAW 2mm S21 FC023910 470.RAW 2mm S14 FC022902 289.RAW QuadSerp S6 FC023466 291.RAW QuadSerpQ2 FC023910 285.RAW 2mm S21 FC023910 470.RAW 2mm S14 FC022902 289.RAW QuadSerp S6

Inle

t P (p

sig)

J (A/cm2)

75/70/70C, 0/0psig H2/Air, 800/1800SCCMPDS(0.85V->0.25V->0.85V, 0.05V/step, 10s/step)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.00.10.20.30.40.50.60.70.80.91.0

80/68/68C, 150/150 kPaa H2/Air, CS(2,100)/CS(2.5, 167)GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)Upscan (high->low J) only.

1st AlternateSerp. Wide Lands

Two MEA's

FC023910 301.RAW FC023466 323.RAW FC23949 281.RAW FC22902 353.DAT FC23996 360.DAT FC023844 470.RAW FC23406 322.DAT

C:\Users\US117400\Documents\DOE-6 032212\Task 5.3 Final Stack Testing\Flow field comparisons\0811354A FF Sens 040912-[Graph5]

Cell V

olta

ge (V

), HF

R (o

hm-c

m2 )

J (A/cm2)

Quad SerpentineFour MEA's, two

test stations

Stack 2 Type MEA Configuration

20



Pt3Ni7 Development: Dealloying experiments and process scale-up

Approach: Use ex-situ methods to remove excess Ni from as-made Pt3Ni7 catalysts: Catalyst material factors Loading (0.075, 0.10, 0.125, and 0.15 mg/cm2) Alloy homogeneity - Catalyst deposition process (P1 vs. P4 process) Post-fabrication annealing – SET process

Dealloying processes and conditions investigated Electrochemical (ex-situ CV Cycling) Acid Washing (composition, concentrations, time, temperature)

• Batch treatment (Pt/Ni loadings vs time, acid composition and concentrations) • Requirements for roll-to-roll processing at reasonable web speeds

Roll-to-roll scale-up capability (pilot level) Facilities identification within 3M, equipment modifications, pretrial web-runs Multiple trial runs, correlation of Pt/Ni dissolution rates with batch experiments

Fuel cell testing and property characterization (XRD, XRF, TEM) Correlation of ORR metrics and limiting currents with dealloying/SET processing H2/O2 testing diagnostics of limiting currents Correlation of Pt3Ni7 fcc(hkl) grain sizes, lattice parameters and compositions

Objective: Study NSTF Pt-Ni alloy system to understand how to achieve stable nanoporosity that will increase surface area, specific activity and maintain stable high current density performance while being practicably scalable.

21



Pt3Ni7 Development: Dealloying experiments and process scale-up Results: Fuel cell metrics correlated as “global” summaries Over 100 different combinations of bath compositions, concentrations, times, roll-to-roll dealloy

web speeds, and SET treatment process parameters for four loadings of Pt3Ni7 were tested in duplicate in 50 cm2 fuel cells.

16 ORR relevant kinetic and performance metrics were extracted from the PDS and GDS polarization curves and correlated with materials and process parameters (all proprietary).

Critical metrics may be plotted w/o identifying material/process factors=> 38 scatter-plots. Example scatter plots follow. Show that roll-to-roll dealloying and SET processes were found

which generate Pt3Ni7 ORR specific activities up to 3 mA/cm2Pt (left), and mass activities

between 0.4 and 0.45 A/mgPt (right).

0 5 10 15 20 250.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

ORR

Spe

cific

Activ

ity a

t 900

mV(

1050

sec

) n

ot iR

-cor

rect

ed (m

A/cm

2 plan

ar)

Surface Area Enhancement Factor (cm2Pt / cm2

planar)PtNi Data - all scatter plots - graph 72 0 5 10 15 20 25

0.0

0.1

0.2

0.3

0.4

0.5

ORR

Mas

s Ac

tivity

at

900

mV

(not

iR-c

orre

cted

) (A/

mg)


planar)PtNi Data - all scatter plots - graph 70

0.44 A/mgPt

22

Technical Accomplishments and Progress Pt3Ni7 Development: Dealloying experiments and process scale-up

Results: Fuel Cell testing global summary Surface area enhancement (SEF) factors up to

25 cm2Pt/cm2

planar were generated, with mass specific surface areas of 15-20 m2/gPt common for the highest mass activities (bottom left.)

The absolute activity at 900 mV is highly linear with SEF with a slope of a pseudo-specific activity of 3.6 mA/cm2

Pt (bottom right).


0 5 10 15 20 250

2

4

6

8

10

12

14

16

18

20 0.0750.10 mg-Pt/cm2

0.125

Mass Specific Surface Area (m2-Pt/g)

Mas

s Sp

ecific

Sur

face

Are

a (m

2 /g)



0.15

0 5 10 15 20 250

10

20

30

40

50

60

70

ORR

Abs

olut

e Ac

tivity

at

900

mV(

1050

s) (

mA/

cm2 pl

anar)



Slope = 3.59 mA/cm2-Pt

0 200 400 600 800 1000 12000.00

0.02

0.04

0.06

0.08

0.10

0.12

C:\Users\US117400\Documents\DOE-6 032212\Reviews for DOE\2012 Annual Merit Review\0.125PtNi 030612-[Graph1]

J (A

/cm

2 ) at 0

.900

V

Time (seconds)

0.054 A/cm2-planar @ 1050sSpec. Activity = 2.8 mA/cm2

Pt

Mass Activity = 0.43 A/mgPt

1050 secFC236160.05/0.125 mgPt/cm2 Pt3Ni7 R2R dealloyed, SET "annealed"850 EW, 20 µm: Std. 3M GDL's

80/80/80C, 7.35/7.35psig H2/O2, 696/1657SCCMPSS(0.900V, 20min)

ORR activity measure-ment protocol

5 10 15 20 250.83

0.84

0.85

0.86

0.87

0.88

0.89

0.90

0.91

0.92

HCT Meas V @ ~20mA/cm2 (Volts)

HCT

Cell V

olta

ge a

t 0.0

2 A/

cm2 (V

olts

)



23



Pt3Ni7 Development: Dealloying experiments and process scale-up Results: Fuel Cell global summary Low current voltages correlate with SEF, but

progressively less and less up to 1 A/cm2.

Above 1 A/cm2 the cell voltages depict an inverse dependence on surface area.

5 10 15 20 250.50

0.55

0.60

0.65

0.70

0.75 HCT Meas V @ 1A/cm2 (Volts)

HCT

Cell V

olta

ge a

t 1 A

/cm

2 (Vol

ts)



5 10 15 20 250.68

0.70

0.72

0.74

0.76

0.78

0.80

HCT Meas V @ 0.32A/cm2 (Volts)

HCT

Cell V

olta

ge a

t 0.3

2 A/

cm2 (

Volts

)



5 10 15 20 250.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

HCT Meas J @ 0.5VDS (A/cm2)

HCT

Curre

nt D

ensit

y at

0.5

V (

A/cm

2 )



Excess Ni at higher loadings further reduces limiting currents.

24



0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.5

0.6

0.7

0.8

0.9

1.0

Tested at 3MSt. Paul Lab

Q:\Projects\0046650001_DOE_VI\Shared\Task 1.1.1 Series Summaries\PtNi Noda 032012-[Graph4]

Cell V

olta

ge (V

olts

)

J (A/cm2)

80/68/68 oC, 7.35/7.35 psig H2/Air, CS(2,100)/CS(2.5, 167)GDS(2 to 0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)

Identical CCMTested at 3MSumitomo Lab

Cathode: 0.15 mg/cm2Pt of De-alloyed Pt3Ni7Anode: 0.05 mg/cm2 PtCoMn: PEM: 3M 24 micron

Results: Fuel Cell results for best configuration The limiting current is still much lower than it should be. Some mechanism other than

mass transport is operative, e.g. : • Concentration polarization due to excess Ni in the vicinity of the catalyst surface • Test station effects and conditioning protocols, as demonstrated in the figure below. • Water cleanliness, particularly Cl- that affect GDS maximum curves in 3M labs. • Flow fields and non-uniformity of current density distributions – inlets, to outlets • Membrane properties

Pt3Ni7 Development: Dealloying experiments and process scale-up

Results: Different limiting currents observed at 3M St. Paul and 3M Sumitomo: Identical MEA’s Same 50 cm2 FCT cells Same GDS protocol Different conditioning Different humidification approach

The best overall performance to date is obtained with Pt3Ni7 on the cathode that has been roll-to- roll dealloyed and SET treated with best conditions explored thus far. MEA below used: • Anode = 0.03 mg/cm2 NSTF-Pt ; Cathode = 0.121 + 0.003 mg/cm2 of NSTF Pt3Ni7 (by ICP) • PEM = Standard 3M-24 µm, 850 EW ; GDL’s = 3M standard 2979 ; FF = 3M standard Quad-serpentine. • Inverse Specific Power density = 0.14 – 0.18 gpt/kW over 0.6 – 0.65 V and 150 to 250 kPaa operating

ranges at 80oC , with 0.151 + 0.003 mgPt/cm2 total loading per MEA (right graph). • Pt3Ni7 gives 0.21 to 0.31 mA/cm2 at 0.8 V (“1/4 power”) over 150 to 250 kPaa pressure (outlet control). • These Pt3Ni7 cathodes show limiting current improvements over last year, but there is still opportunity for

further gains. (Temperature sensitivity over 80 – 95oC at 150 kPaa shown in Back-up Slides, # 34)

25



Pt3Ni7 Development: Toward 2012 Best of Class Pt/Pt3Ni7 CCM configuration.

0.50 0.55 0.60 0.65 0.70 0.750.10

0.12

0.14

0.16

0.18

0.20

0.22

0.24

150 kPaa outlets

250 kPaaoutlets

200 kPaaoutlets

150 kPaa outlets

ANODE: 0.050 mgPt/cm2 Pt69Co28Mn3

CATHODE: 0.15 mgPt/cm2 Pt69Co28Mn3 PEM: 3M-S, 18um, 850EW; GDL: 2979/2979

ANODE: 0.030 mgPt/cm2

CATHODE: 0.121 + .003 Pt3Ni7 (DEALLOY+SET)PEM: 3M 24u 850EW; GDL: 2979/2979

C:\Users\US117400\Documents\DOE-6 032212\Reviews for DOE\2012 Annual Merit Review\FC23951_LowTotalLoadingPt3Ni7 g-PtperkW-[Graph2]

Inv.

Spe

c. P

ower

Den

sity

(gPt/k

W)

Cell Voltage (Volts)0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

0.0

0.2

0.4

0.6

0.8

1.0

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.400.70

0.72

0.74

0.76

0.78

0.80

0.82

0.84

0.86

0.88

0.90

200 kPaa

250 kPaa

0.27 0.21

Cel

l Vol

tage

(V)


0.31 A/cm2

150 kPaa

0.80 V

250 kPaa

200 kPaa

Cell T = 80 oC ANODE: 0.030 mg/cm2 pure PtCATHODE: 0.121 + .003 Pt3Ni7PEM: 3M 24u 850EW GDL: 2979/2979

FC23951 478.RAW. 80/68/68C, 150/150kPa, CS2/2.5 FC23951 478.RAW. 80/68/68C, 150/150kPa, CS2/2.5 FC23951 495.RAW. 80/63/63C, 200/200kPa, CS2/2.5 FC23951 495.RAW. 80/63/63C, 200/200kPa, CS2/2.5 FC23951 510.RAW. 80/54/54C, 250/250kPa, CS2/2.5 FC23951 510.RAW. 80/54/54C, 250/250kPa, CS2/2.5

Cell V

olta

ge (V

), HF

R (o

hm-c

m2 )


150 kPaa

26

Collaborations Over Life of Project Subcontractors Dalhousie University : Subcontractor. Focused on Pt3Ni7 studies. Funding ended Dec., 2010. ANL (Markovic/Stamenkovic group): Subcontractor, periodic measurements in 2010, 2011.

NASA-JPL: Subcontractor, periodic interactions in 2010. TEM, co-deposition of Pt3Ni7 in 2010. System Integrators and stack manufacturers (partial list) GM Fuel Cell Activities-Honeoye Falls: Collaboration outside of DOE H2 program with materials

generated at 3M under this contract. Multi-year single cell performance and activity validations, stack testing, cold/freeze start and water management evaluations, PEM and GDL integration, durability testing, fundamental modeling studies. Final short stack testing: Done under this project, Task 5.3.

Nuvera Fuel Cells – Large area short stack testing-combining open flow field with NSTF MEAs – collaborative work under Task 3 concluded by mid-2010.

Proton OnSite – Collaboration outside of DOE H2 program. Performance testing of NSTF MEAs in electrolyzers. Continuous testing and periodic interaction past year.

Giner Inc., LLC – Collaboration outside of DOE H2 program. Performance testing of NSTF MEAs in electrolyzers. Periodic testing and interaction past year.

National Laboratories ANL(Ahluwalia) – Supplied extensive NSTF fuel cell performance data for ANL systems modeling. LBNL, LANL, UTC– Collaborative interactions outside this contract under LBNL project “FC

fundamentals at Low and Subzero temperatures.” NIST – Samples and data supplied to NIST for optical method development for CCM Pt loading

measurement done under FC Manufacturing.


27

Future Work (4/11/12 to 6/30/12)

Final Stack Testing

Complete stack testing at GM – decision of when to stop Understand issues with performance and decay

Complete flow field study Apply load cycling protocol to 50 cm2 cells w/ MEA’s having

high-performing PEM’s.

Characterize Pt3Ni7 dealloyed (R2R) materials under DOE durability accelerated stress tests.

Prepare Final Report


28

Characteristic Units Targets 2017

Status: Values for roll-good CCM w/ 0.15mgPt/cm2 per MEA or as stated

PGM Total Content gPt/kWe rated in stack

0.125 0.14 - 0.18 gPt/kW for cell 0.6 < V < 0.65 at 80 oC and 150kPaa to 250 kPaa outlet.

Pt3Ni7, 50 cm2 cell w/ 0.15 mg/cm2 total Pt.

PGM Total Loading mg PGM / cm2 total

0.125 0.15 to 0.20, A+C with PtCoMn alloy 0.15 A+C with Pt/Pt3Ni7

Mass Activity (150kPa H2/O2 80oC. 100% RH, 1050 sec)

A/mg-Pt @ 900 mV, 150kPa O2

0.44 0.24 A/mg in 50 cm2 w/ PtCoMn ~ 0.43 A/mg in 50 cm2 with R2R Pt3Ni7

Specific Activity (150 kPa H2/O2 at 80oC, 100% RH)

mA/cm2-Pt @ 900 mV

0.720 2.1 for PtCoMn, 0.1mgPt/cm2 2.7-3.0 for R2R Pt3Ni7, 0.125 mgPt/cm2

Durability: 30,000 cycles 0.6 -1.0V, 50mV/sec,80/80/800C, 100kPa,H2/N2

- mV at 0.8 A/cm2

- % ECSA loss - % Mass activity

< 30mV

< 40% < 40 %

10+7mV loss at 0.8 A/cm2 16+2% loss ECSA, PtCoMn 37+2% loss mass activity

Durability: 1.2 V for 400 hrs. at 80oC, H2/N2, 150kPa, 100% RH

- mV at 1.5 A/cm2

% ECSA loss % Mass activity

< 30mV < 40% < 40%

10 mV loss at 1.5 A/cm2 10% loss ECSA

10 % loss mass activity

Durability: OCV hold for 500 hrs. 250/200 kPa H2/air, 90oC, 30%RH

H2 X-over mA/cm2 % OCV loss

< 20

< 20 %

13 + 4 mA/cm2 at 500 hrs (5 MEAs) 12 + 5 % OCV loss in 500 hrs

Durability under Load Cycling (membrane lifetime test)

Hours, T < 80oC Hours, T > 80oC

5000 5000

9000 hrs, 3M PEM (20µm, 850 EW w/ stabilizers), 50cm2 , 80/64/64 oC

2000 hrs (OEM short stack,0.1/0.15)

Project Summary : Status Against DOE Targets – March, 2012 (blue = new)


29

Technical Back-Up Slides

30

Technical Accomplishments and Progress – Back up Slide


XRD results for SET treated and dealloyed Pt3Ni7 FCC(111) crystallite sizes increase with loading and SET treatment of as-made catalysts as

historically seen for PtCoMn and Pt3Ni7 (left figure). SET and dealloying treatment produces smaller crystallites, independent of SET parameters.

Pt3Ni7 Development: Dealloying experiments and process scale-up Results: Hundreds of lab batch dealloying, with XRF Pt/Ni atomic fraction measurements completed. Process chemistry identified to reduce dealloy time by factor of 1/240, feasible for roll-to-roll. Roll-to-roll (R2R) equipment identified and modifications made for down-selected chemistry. Four R2R dealloying experiments completed and optimum conditions determined. Combination of R2R dealloying and SET processing completed multiple times using full width

roll-goods.

0 40 45 5060

70

80

90

100

110

120

0.075 mg/cm2

0.10 mg/cm2

FCC (111)

Pt3Ni7 by P1 ProcessSET treated only

FC(h

kl) G

rain

size

(Ang

stro

ms)

(Ang

stro

ms)

SET Parameter (arb. units)

GID 200441

0.15 mg/cm2

graph 55 0 40 45 5060

70

80

90

100

110

120

0.10 mg/cm2

FCC (111) FCC (111)

Pt3Ni7 by P1 Process;Dealloyed and SET treated

FC(h

kl) C

ryst

allite

size

(Ang

stro

ms)

SET Parameter (arb. units)

GID 201603

0.15 mg/cm2



Pt3Ni7 Development: Dealloying experiments and process scale-up Results: TEM Images Characterization

20 nm20 nm

2211299B-R2R dealloyed

20 nm20 nm

221299B- R2R dealloyed + SET

20 nm20 nm

2211299B- As-made

HAADF images may suggest less density of the PtNi in the R2R dealloyed whiskers Whiskerettes appear more “feathery” after SET treatment ( see bottom three images).

31

Pt and Ni distribution appears uniform (first two images left).

Oxygen is restricted to surface of whisker coatings (last image left).

32


0.000 0.050 0.100 0.150 0.200 0.250

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.184mg/cm2

NSTFPtCoMn

iR -free only

Av. 0.1-0.5 mg/cm2 Pt/CiR and Mass Transport

Corrected Voltage

0.2 mg/cm2 PtCo/CiR and Mass Transport

Corrected Voltage

0.1 mg/cm2 PtML/Pd/CiR-free only V

0.146mg/cm2

NSTFPtCoMn

iR -free only

0.05 mg/cm2

NSTFPtCoMn

iR-free only

0.1 mg/cm2

NSTFPtCoMn

iR-free only

Book 19, graph 76

80 oC , 150 kPa H2/Air

Corre

cted

Cel

l Vol

tage

(V)

SEF Normalized Current Density (A/cm2Pt)

70 mV/dec


Ref: Mark K. Debe, “Effect of Electrode Structure Surface Area Distribution on High Current Density Performance of PEM Fuel Cells,” J. Electrochemical Society 159(1) B54-B67 (2011).

NSTF Fundamentals: Extended Surface Catalyst fundamental properties

Loss of high current density with cathode loadings below 0.2 mgPt/cm2 in Pt/C electrodes is an issue. We show this effect is much less at a given loading with the NSTF catalyst type electrodes (Figs. 1 and 2 below) We develop a model based on elementary kinetic gas theory and known molecule/surface interaction mechanisms that

take place in the Knudsen regime to explain these differences. The close packed nature of the extended NSTF catalysts and their spacing on the order of the Knudsen length enables O2

molecules to make many more surface collisions per unit time (see cartoons on next slide). When modeled the result is an additional pre-exponential scaling factor in the Butler-Volmer equation related to a distance

metric, ρs describing the catalyst surface area distribution, (see bottom graphs on next slide.) The model is able to predict the correct heat of enthalpy for O2 physisorption and the observed ratio of current densities at

V(iR-free) = 0.7 V for NSTF compared to Pt/C dispersed electrodes in the 0.05 to 0.15 mgPt/cm2 range from published data for eleven different catalyst types and cathode loadings below 0.2 mg/cm2, (shown in graph immediately below).

A “universal” polarization curve

Fig. 1 Fig. 2

33



0 1 2 3 4 50

25

50

75

100

125

Book 9,,graph 46

J/(S

EF*A

s) at

0.7

V (i

R-Fr

ee) (

arb.

uni

ts)

ρs (105cm2Pt/cm3 )

30% PtCo/V 50% Pt/V PtML/Pd/C NSTF - 0.146 NSTF - 0.103 NSTF - 0.054 NSTF - 0.184

s

Slope = 2.66 x 10-4 cm3/cm2Pt

0 1x105 2x105 3x105 4x1050.4

0.5

0.6

0.7

0.8

0.9

1.0

Book 12, ,graph 65

ρs (cm2-Pt/cm3)

f (ρ s,λ

(1) )

f = 0.528 + 2.08 x 10-6 ρs - 2.70 x10-12 ρs2

D ~ λProcess

Continues

Vx

Vz

Vz = Vx = Vy

Vz > Vx = Vy

λ ~ 80 nm

Model Development: Extended Surface Catalyst Fundamental Properties

34 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0ANODE: 0.030 mgPt/cm2

CATHODE: 0.121 + .003 Pt3Ni7 (DEALLOY+SET)PEM: 3M 24u 850EW; GDL: 2979/2979

FC23951 478.RAW. 80/68/68C, 150/150kPa, CS2/2.5 FC23951 536.RAW. 85/77/77C, 150/150kPa, CS2/2.5 FC23951 549.RAW. 90/84/84C, 150/150kPa, CS2/2.5 FC23951 562.RAW. 95/91/91C, 150/150kPa, CS2/2.5

C:\Users\US117400\AppData\Local\Temp\notesFCBCEE\FC23951_LowTotalLoadingPt3Ni7-[TSensPolCurve150kPa]

Cell V

olta

ge (V

olts

), HF

R (o

hm-c

m2 )

J (A/cm2)

80/68/68C, 7.35/7.35psig H2/Air, CS(2,100)/CS(2.5, 167)GDS(0.02->2->0.02, 10steps/decade, 120s/pt, 0.4V limit, 0.1maxJstep)

Temperature Sensitivity @ 150 kPaa Reactant Pressure

HFR

Technical Accomplishments and Progress Short stack 2 testing: Issues –

Investigating possible flow field effects

3M Std. Quad-Serpentine 4 channels, 10 loops

1st Alternative FF 6 channels, 2 loops

Pt3Ni7 Development: Toward 2012 Best of Class Pt/Pt3Ni7 CCM Configuration.

Top: Pressure series at 90 oC for 0.03/0.121 Pt/Pt3Ni7 based MEA.

Bottom: 0.03/0.121 Pt/Pt3Ni7 based MEA shows very little sensitivity over temperature range of 80 to 95 oC (inlet RH controlled to give ~ 100% outlet RH at each temperature.)


0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.00.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0 FC23951 549.RAW. 90/84/84C, 150/150kPa, CS2/2.5 FC23951 578.RAW. 90/80/80C, 200/200kPa, CS2/2.5 FC23951 593.RAW. 90/76/76C, 250/250kPa, CS2/2.5

Cell V

olta

ge (V

)Current Density (A/cm2)

DOE-6 032212\Task 5.3\Stack 2 Data\S1623 Cruves-graph 5

Pressure Series at 90 oC