Stephen Campbell: 15 May 2013 - US Department of EnergyNEXAFS from each core level of the target - site specific . AFCC Confidential Information STXM Methodology for Ionomer Imaging

The challenges for PEMFC catalysts in automotive

applications.

Stephen Campbell: 15th May 2013

AFCC Confidential Information

The challenge

Cost

Performance

Durability

Commercialization


The Challenge

Lower cost means less platinum group metals (PGM).

Less PGM means

– higher performance catalyst (in Ag-1 Pt)

– reduced durability (especially SU/SD)

– reduced tolerance to impurities.

Optimization focussed upon:- – Fundamental materials properties – Component integration into MEA

• Catalyst layer structure/ performance relationship – Stack/system trade-offs (for operational durability)


Performance

There is a need to reduce Pt loadings to <300mgcm-2 in order to meet automotive cost targets without any loss in performance.

A internal target was established to reach a 4x improvement in ORR activity for cathode catalysts There have been many studies of Pt-transition metal alloys that have enhanced activity but these have not been supplier-led. AFCC has compared the activity of the catalysts of several suppliers using ex-situ RDE screening in 0.1M HClO4 at 30°C. So far none have met the 4x target but we have been able to make progress internally with a ternary alloy. Supplier scale-up requires more work.


0

100

200

300

400

500

600

700

800

Pt supplier PtCo

supplier PtCo

supplier PtCo

supplier ternary

AFCC ternary

mas

s ac

tivity

/ Ag-1

Pt

4x target

Recent progress with Pt alloy/ carbon catalysts

Ex-situ RDE data for ORR (0.1M HClO4) 30°C


Catalyst durability

Start-up causes high potentials due to hydrogen entering an air-filled fuel flow-field.

Potentials can reach 1.8V which can damage the cathode catalyst. AFCC has pursued two materials approaches to mitigate this damage (independent of system mitigation). 1. Alternative catalyst supports resistant to oxidation.

2. Adding a selectively conducting component to the anode

electrode structure to limit the potential transients.


Non-carbon supports

Oxide supports may be stable to oxidation but must also be:-

– Electrically conductive

– High surface area

– Cost effective

– Potentially interactive with the catalyst (SMSI)

Common options are ITO, ATO but these have been found to

have stability issues.

AFCC has demonstrated baseline performance and enhanced

SU/SD durability with an oxide supported Pt cathode catalyst.


Platinum on oxide support

0

100

200

300

400

500

600

0 200 400 600 800 1000 1200

Cell

volt

age/

mV

Cycles

oxide support

Subscale (50cm2) cell; startup/ shutdown AST cycles; performance at 1.5Acm-2

carbon support

Proof of concept only – not commercially viable


Selective oxide on the anode

Tin oxide (SnO2) may be switched from a conductive state in hydrogen, to a resistive state in air.

The speed of switching may be enhanced by the addition of Pt onto the oxide surface.

Six orders of magnitude!

0.0001

0.001

0.01

0.1

1

10

100

1000

0 5 10 15 20 25

Res

ista

nce

/ MΩ

Time/ min

Hydrogen

Air

Nitrogen

SnO2

5% Pt/SnO2


SnO2 layer in the anode

When an SnO2 layer in placed between the anode catalyst layer and the GDL it is conductive during normal operation.

When in an air state, however, it is resistive, limiting in-plane currents and holding the anode potential lower during the start-up transient.

This keeps the cathode from going to damaging, high potentials and reduces cathode performance degradation.


Fuel cell start-up/shut down cycles

450

500

550

600

650

700

0 400 800 1200 1600 2000 2400 2800 3200

Aver

age

cell

volta

ge/ m

V

Cycles

With SnO2 layer on anode

Baseline


Summary of catalyst progress

In-house ternary alloys have demonstrated >4x ORR mass activity over Pt/C baselines using RDE (ex-situ).

Durability of the catalyst during start-up transients may be enhanced by:-

– an oxide layer in the anode that changes from conductive to

resistive depending upon its environment, and/or

– a conductive oxide catalyst support for the cathode that resists

oxidation.


Catalyst layer characterization/ optimization

In order to design the catalyst layer to provide the required transport properties for a high performing catalyst it is necessary to first be able to characterize the catalyst layer in terms of:- – Catalyst structure – Ionomer structure – Pore structure

Then modeling may be used to design the optimum three phase structure. Methods being developed through academic collaborations are:- – E-tomography (3D imaging with HRTEM) – FIB/ HRSEM (Focussed ion beam sectioning) – STXM (scanning transmission X-ray microscopy)

These will be used as design tools for the next generation of fuel cell stacks.


Methods of catalyst layer characterization

FIB tomography Sample size: 5 x 3 x 2 um Voxel: 20x20x20 nm Speciation: no chemical info

STXM Sample size: 5 x 3 x 0.2 um Voxel: 30x30x200 nm Speciation: C + ionomer + Pt

E-tomography Sample size: 0.5 x 0.2 x 0.1 um Voxel: nm x nm x nm Speciation: Pt + Carbon

E-tomography of catalyst Sample size: 0.2 x0.2 x 0.1 um Voxel: nm x nm x nm Speciation: Pt + C


E-tomography of catalyst particles

TEC10-V50E-C+Pt- 80k-ali.avi • Individual catalyst agglomerates

on TEM grid.

• Sample rotated and multiple

images taken.

• Composite of images produces

3D representation of Pt particles

on carbon support.

• Advantage over 2D image is that

position of individual Pt particles

on support surface can be

determined.

• No ionomer in sample.


Visualization of the catalyst layer structure by FIB tomography

Catalyst layer cross-sections are made by application of focused ion beam (FIB) and imaged by high resolution secondary electron microscope (HRSEM).

X

Z

Y

Sectioning of the CL is done by application of the focused ion beam (FIB). CL is bombarded with accelerated Ga ions. FIB tomography: FIB slicing and subsequent HRSEM imaging to get the 3D CL model.


3D representation of the catalyst layer

Amira

2.0 um

4.7 um

1.4 um

C+Pt+Nafion

Connected pores

Non-connected pores

Solids + pores

Mem

GDL

2.0 um

1.4 um

4.7 um

Pore network

Image J


Electron tomography summary

HRTEM tomography can be used to show 3D structure of the Pt/carbon

structure of the catalyst agglomerates.

It can also be used to generate 3D representations of the catalyst layer

structure showing the Pt and pore distribution.

No chemical speciation can be done so cannot distinguish between carbon

support and ionomer.

FIB tomography can be used to generate larger scale representations of

the pore structure within the catalyst layer.

Can only separate solids and pores.

A method to distinguish between Pt, carbon, ionomer and pores is needed.


Soft X-ray STXM

• Chemical speciation through X-ray absorption spectra (NEXAFS)

• Spatial resolution ~30 nm; Energy resolution < 0.1 eV

• Quantitative chemical mapping (Beer’s law)

• All elements (100-2500 eV)

• Transmission requires thin samples ⇒ microtomed sections (~100 nm thick)

Beamlines 5.3.2 &11.0.2 at the Advanced Light Source, USA

Beamline 10ID1 at the Canadian Light Source, Canada

Photon Energy [eV]100 200 300 400 500 600 700 800

Abso

rptio

n [c

m2 /g

]

C K-edge

N K-edge

O K-edge

ALS Bend Magnet beamline 5.3.2

spherical grating 300 l/mm zone plate

D =200µm

Toroidal Mirror

10.0 m Entrance Slit

174o

173o Exit Slit Pinhole

5.3 Bend Magnet

plan view

detector

sample

ZP OSA

NEXAFS from each core level of the target - site specific


STXM Methodology for Ionomer Imaging in CL

285 295 305

Opt

ical

den

sity

/ m

icro

n

2 µm-1

0

carbon support (/2)

PFSA membrane

Energy (eV)685 695 705

1 µm-1

0

C 1s

F 1s

carbon support (/2)

PFSAmembrane

0.015 nm-1

0.0026 nm-1

(a)

(b)

For mapping of the ionomer in the catalyst layer must record X-ray transmission images at two different X-ray absorption energies for carbon (C) and fluorine (F).

C 1s map provides information on

the carbon support

F map provides info on fluoride-

containing species in the catalyst

layer (ionomer, PTFE)


2D Reconstruction of the catalyst layer

membrane

carbon support

platinum

ionomer

Note: each pixel contains a complete NEXAFS spectrum of the species present.


3D imaging of the catalyst layer using STXM

STXM allows for chemical speciation within the catalyst later thin section.

Can differentiate between carbon, platinum, ionomer, pores.

This 2D capability can be used tomographically to generate a 3D representation of all these species by rotating a FIB section of the catalyst layer and recording multiple images.


CCM sample preparation for the 3D STXM tomography (FIB sample prep.)

TEM

-gri

d ba

r 5

um

3.5

um

270 nm

200 nm

90 nm

Need to have a rotator for +/- 60o sample rotation

FIB sample prep. done by Julia Huang, McMaster

Sample is prepared using FIB.


Now Rotate the CL Sample for +/- 60o and Record the C and F maps every 2 Degrees.

285 295 305

Opt

ical

den

sity

/ m

icro

n

2 µm-1

0

carbon support (/2)

PFSA membrane

Energy (eV)685 695 705

1 µm-1

0

C 1s

F 1s

carbon support (/2)

PFSAmembrane

0.015 nm-1

0.0026 nm-1

(a)

(b)

FIB-1 sample: 2071 TPM BOL cathode, height~8 um, width~4um, thickness~200 nm

Every pixel in these images contains a spectrum characteristic for the chemical species.


Summary of STXM

Thin sections may be analyzed to differentiate between platinum, carbon support, ionomer and pores. By rotating the thin section 3D tomographic representation may be constructed from multiple images. Carbon and ionomer 3D structures have been demonstrated. Work continues.


Conclusions & path forward

Significant progress has been made in the development of MEA components (catalyst, membrane, etc.) that address the challenges that must be overcome for commercialization of PEMFCs for automotive applications.

Tools are being developed to understand the structure of the catalyst layer and these will be critical in the design of the next generation of MEAs.

The fundamental materials and the means to put them together in an optimized way are at hand and the path forward is exciting!


Questions for the future

Will it ever be possible to eliminate PGMs from the cathode of a PEMFC? DOE targets:- – Does an EOL performance and a BOL cost target define the target

sufficiently?

EOL performance

High BOL performance High degradation Makes EOL target but at high cost

Operational lifetime

Low BOL performance Low degradation Makes EOL target at lower cost

This approach tends to focus upon the durability of the catalyst layer and its degradation rate at cost.

Documents

Stephen Campbell: 15 May 2013 - US Department of EnergyNEXAFS from each core level of the target - site specific . AFCC Confidential Information STXM Methodology for Ionomer Imaging