2005 DOE HYDROGEN PROGRAM REVIEW May 2326, 2005, Washington, DC
NonPrecious Metal Catalysts
Radoslav T. Atanasoski 3M Company May 24, 2005
Novel Approach to
Project ID#: FC14
This presentation does not contain any proprietary or confidential information
Overview
• Project start date: September 1, 2003 • Project end date: August 31, 2006 • Percent complete: ~40
(consistent with spending)
Timeline
• Total Project funding: $3.6 million DOE share: $2.9 million Contracter share: $0.7 million • Funding received in FY04: $500,000 • Funding for FY05: $700,000
Budget
• O. Stack Material and Manufacturing Cost • P. Durability • Q. Electrode Performance • (Technical targets: See next slide) • Dalhousie University
Barriers
Partners
• Brookhaven National Lab
Prof. J. Dahn; Highthroughput catalyst synthesis and basic characterization
• University of Missouri – Kansas City
Prof. D. Wieliczka; UPS at University of Wisconsin Synchrotron Radiation Center
Dr. X.Q. Yang and Dr. W.S. Yoon X Ray Absorption Spectroscopies Dr. R. Adzic – Exploratory
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
2
Project Goal and Objectives
Goal: Develop new, lowercost, nonprecious metal (NPM) cathode catalysts for replacement of Pt in PEM fuel cells. DOE Objectives/Targets:
• Reduce dependence on precious metals (Pt). • Perform as well as conventional precious metal catalysts currently in use in
MEA’s.
• Cost 50% less compared to a target of 0.2 g Pt/peak kW. • Demonstrate durability of >2000 hours with <10% power degradation.
Specific Objectives for 2005:
•
Synthesize and characterize high catalytically active sites for oxygen reduction reaction (ORR) by
• Introducing new synthetic routes • Understanding and overcoming the cause of the high impedance
•
Preserve process compatibility with high volume manufacturability
3
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
Approach
Physicochemical Characterization
• XPS • XRF • SEM • XRD • Modeling • Ellipsometry • XAS (Brookhaven National Lab) • SIMS • UPS
Catalyst Synthesis Functional Characterization
Nanotechnology
Modeling
Vacuum
New materials
• Fuel Cell 50 cm2 Segmented • Gas Diffusion Electrode
• Catalyst synthesis is carried out via two complementary and interactive approaches: Vacuum Processes: Variety of vacuum processes including mapping via high throughput approach; Nanotechnology: Dispersed catalyst on high stability carbon substrate. • Modeling work is done to guide and verify the synthesis efforts. • Extensive physicochemical analytical characterization is carried out both at 3M and in collaboration with other institutions when appropriate. • MEA fabrication and 50cm2 FC evaluation readily scalable to pilot plant level.
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
4
Technical Accomplishments
Catalyst Synthesis
• Broadened space of catalyst synthesis New area: Nanotechnologybased multicomponent dispersed catalysts • Catalytic activity orders of magnitude higher than previously reported on this project (interim milestone #2 near completion)
Catalyst Characterization
• Insights into catalytic sites based on Modeling: Thermodynamically most favorable CNxFe sites indicated Stateoftheart characterization/analytical techniques • Progress in addressing and overcoming the catalyst highimpedance issue (interim milestone #1 completed) • Compositional areas of stability and activity of CNxFe space mapped • New screening methods for catalyst activity
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
5
Nanotechnology: Component Characterization
Catalysts based on Fe and Sx as ‘matrix/support’ on dispersed carbon have been synthesized. Electrochemically, CV for the combination Fe/S2/C catalyst appears as additive of the CVs of the individual components. However,... …a closer scrutiny via UPS 0.012 indicates some interaction 0.008 between the components. S2 + Fe catalyst
I (A m p s /c m 2 )
0.004 0 0.004 0.008 0.012 0 0.2 0.4
50cm2 FC; 75ºC; 0/0 psig; 500 sccm H2/500 sccm N2; 150%/150%RH; CVs at 50 mV/s
Fe only S2 only
S2+Fe/C: Valence Electrons
Mixed S2/Fe dstates & pstates
Low d DOS
Photon Energy = 55 eV
1.2
0.6 E (Volts)
0.8
1
To emphasize catalytic activity, low surface area, high stability carbon was used as support.
20
15
10
5
0
5
Binding Energy
6
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
Nanotechnology: ORR Component Synergism
• While CV baselines are additive, the oxygen performance is enhanced relative to the addition of the two single component oxygen response curves when a combination S2 and Fe catalyst is used. This points to a possible synergy between S2 and Fe. • Increases in the S2 + Fe catalyst loading, determined both by weight basis (before testing) and XRF measurements (after test), result in a linear increase in current at 0.6 V.
0.02 0
Fe only
Current at 0.6V (mA/cm2)
7 6 5 4 3 2 1 0 0 1 2 3 4
Normalized loading (measured) Normalized Fe (XRF) Normalized S2 (XRF)
I (Am ps/cm2 )
0.02 0.04 0.06 0.08 0.1 0 0.2 0.4 0.6 E (Volts) 0.8
S2 only
S2 + Fe catalyst
50cm2 FC; 75ºC cell, 0/0 psig, 500 sccm H2/500 sccm O2; 150%/150%RH; CVs at 5 mV/s
y = 2.6944x 0.9675 R2 = 0.989 y = 2.7724x 1.1934 R2 = 0.9883 y = 0.8084x + 0.8138 R2 = 0.9964
Normalized loading (measured, Fe XRF, or S2 XRF)
5
6
7
1
1.2
7
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
Nanotechnology: Catalyst Loading Effect
0.04
Loading
0.02
F C9890 F C9974 F C9975
0.004
0.003
0
Z” (Ohms)
I (Amps/cm2)
50cm2 FC; 75ºC cell; 0/0 psig; 500 sccm H2/500 sccm dilute H2 in N2; 150%/150%RH
0.002
Z''
0.02
50cm2 FC; 75ºC cell; 0/0 psig; 500 sccm H2/500 sccm N2; 150%/150%RH
0 0.2 0 .4 0.6 0.8 1 .0
0.001
Rs
0
Rp Loading
0.04
E (Volts)
0.001
0
0.001
0.002
0.003
0.004
0.005
The effect of catalyst loading can be seen both in the increase in the relative magnitude of the CVs (taken under nitrogen) and in the change in impedance (under hydrogen). With more catalyst, the ohmic resistance, Rs, increases, while the polarization resistance, Rp, decreases.
Z’ (Ohms)
Z'
Rp Rs Normalized Loading (ohm*cm 2) (ohm*cm 2) 1 0.095 0.27 2.67 0.115 0.09 5.25 0.170 0.06
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
8
Vacuum Processes: The CNxFe Space
Multiple series of CNxFe compositions were produced (Dalhousie University). Stability is tested in liquid acid and activity in 64-channel fuel cell (for acid soak protocol, see J. Electrochem. Soc., 152 (1) A61-A72 (2005)).
Current Density (mA/cm2) Current Density (mA/cm )
2
Dodelet’s FeN2/C and FeN4/C sites
0
C
1 0.8
At
nF
0.2
om ic F
0.4
0.6
Co-C-N: Highly active Co-C-N: Highly active && stable inalkaline, stable in alkaline, dissolved in acid dissolved in acid (Dignam et al.)
At
0.6
0.4
0.8
0.2 0
0 -5 -10 0 -5 -10 0 -5 -10 0 -5 -10 0 -5 -10 0 -5 -10 0 -5 -10 0 -5 -10 0 0.3 0.6 0 0.3 0.6 0 0.3 0.6 0 0.3 0.6 0 0.3 0.6 0 0.3 0.6 0 0.3 0.6 0 0.3 0.6 Cell Potential (V)
e
F ic om
rac tio
tio rac
nC
Cell potential (V)
Fe
0
0.2 0.4 0.6 0.8 Atomic Fraction N
1
N
FeN system as a baseline was completed first (manuscript submitted)
Current Density (mA/cm2)
1
Fe
10 0 -5 -10
0 0.3 0.6 0
Volts
9
3
Novel Approach to Non-Precious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
Vacuum Processes: Effect of Iron on Nitrogen in CNxFe
Addition of iron to nitrogenatedcarbon alters the chemical environment of the nitrogen atoms.
• NPM catalyst literature - ESCA spectra of N1s core at 398.5 eV labeled as “pyridinic” • However, no direct evidence of surface terminated nitrogen that is sp2-bonded to carbon has been provided. • Alternatively, nitrogen substituted into an sp3bonded carbon environment can also give a 398.5 eV N1s binding energy component. • ESCA of N1s does not give unique material fingerprint for ORR catalytic tendency. • In spite of a strong 398.5 eV component, no substantial activity in our material was found.
07 405.5
C-Nx (~15%N)
Min: 0
C-Nx:Fe(7%)
Max: 3109
N(E)
404
402.5
401
399.5
398
396.5
395
393.5
392
410
408.2
406.4
404.6
402.8 399.2 401 Binding Energy (eV)
397.4
395.6
393.8
392
N1s Component Ratio
6
Ratio (Peak 1/Peak 2)
5 4 3 2 1 0 0 2 4 6 8
ore M
i rid y “p
N c” ni
Fe content (at %)
10
3
Novel Approach to Non-Precious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
Vacuum Processes: Fuel Cell Characterization
Sample made via vacuum synthesis showed an ORR catalytic activity an order of magnitude greater than the previous best result.
Catalyst on whisker substrate
0.0002
High Frequency AC Impedance
I (Amps/ cm2)
0
N2 baselines
|Z|
2.0 1.5 1.0 0.5 0 103
2004
FC9845 FC9807
0.0002
O2 response 2004
50cm2 FC; 75ºC; 0/0 psig; 500 sccm H2/500 sccm N2 or O2; 150%/150%RH; CVs at 5 mV/s
0.6 0.8 1.0 1.2
Today
Frequency (Hz)
104
0.0004
0.0006
0.0008
FC9845 N2 O2 response FC8945 O2 FC9807 N2 FC9807 O2 Today
Impedance improved by a factor of 2.5; however, remains high.
0
0 .2
0.4
E (Volts)
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
11
Vacuum Processes: Catalyst Impedance
Posttreatment reduces impedance (interim milestone #1) and improves oxygen response.
101
Oxygen Reduction
0
F C9523 F C9833 F C9843
AC Impedance
|Z|
100
Whiskers Carbon substrate Posttreated carbon substrate
I (Amps/ cm2)
FC9523 FC9833 FC9843
0.0004
0.0008
101 10 10
2
0.0012
0.0016
3
50cm2 FC; 75ºC; 0/0 psig; 500 sccm H2/500 sccm O2; 150%/150%RH; CVs at 5 mV/s
0 0 .2 0.4 0.6 0.8 1.0 1.2
103
Frequency (Hz)
104
0.0020
• Impedance values of catalyst depend on the type of substrate. • Certain postprocess treatments of the coated carbon substrate decrease the impedance even more.
E (Volts)
Oxygen response increases as the impedance decreases. However, the decrease in impedance cannot fully account for the increase in activity.
12
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
UPS & EXAFS: PostTreatment Effects
Synchrotron UPS and EXAFS reveal electronic and structural changes from post treatment of the catalyst.
• Valence estructure significantly modified by posttreatment. • Fe dstate intensity at Fermi level is reduced.
As received Post Treated
(Photon energy is 55 eV )
4000
• Fe nearest neighbor distance decreases ~0.15 Å • Increased atomic order
Heat treatment effect
pd states
Photoelectron Intensity
3500
Transform Magnitude Transform magnitude
3000
B
2500
CNxFe B251C9 B251C9F Posttreated
Fe d states
2000
1500
Post treated As received
2 4
C
1000
A
500
0
20
15
Binding Energy (eV)
10
5
0
R (Angstroms) R (Angstroms)
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
13
SIMS Characterization and Modeling
SIMS Characterization
1000000 100000 10000 Counts 1000 100 10 1 0 30 Time (seconds) 60
The Models
C C
Si substrate
Fe(CN)2
N
H C O CN Si S Cl CNO C2N3 FeO FeOH C4N3 Fe(CN)2
N Fe Fe C N
N
C
Model catalyst (Dodelet)
90
Lower in energy by 0.99 eV (ab initio)
• SIMS characterization showed the coating is uniform throughout the sample. Several samples gave reproducible results. • Fragments indicating CN2Fe structures have not been detected.
• Modeling work indicates that literature proposed CNFeNC catalyst structure is not the lowest in energy. The structure depicted on the right, where Fe is connected to one nitrogen and one carbon, CNFeCN, is lower.
14
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
Modeling: Iron – Nitrogen Substitution
d c b a
v a c u u e m
a b e c d N C
Fe
N N
Relaxation does not make a big difference for relative stability of only N substituted on the edge of graphene sheet. For Fe substituted sheets, it not only changes the magnitude of relative energies, but also the order.
Fe
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
15
Modeling: Fe Coordination and Effect of Disorder
FeN2C4 with 3 NH3 ligands
NFeN
CFeN
N
Relative Energy 1.72 eV
N
N
Fe
Fe
Considering disordered systems Completing the Fe coordination generated using ab initio molecular does not change the relative dynamics. energetics. (manuscript submitted)
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
16
Summary of Accomplishments
Catalytic activity has progressed during this reporting period and now can be compared with data reported elsewhere.
1.0 0.8 0.6 0.4 0.2 0 1 05
50 cm2
FC9414 FC; 80ºC; 30/50 psig; 180 sccm FC9644 H2/335 sccm O2; CDP80ºC/80ºC; FC9700 FC9644 CVs at 5 mV/s FC9807 FC9890 FC9890D FC9895 FC9895D FC9975 QD FC9994 QD
1.13 cm2; 80ºC; 30/60 psig; 180 sccm H2/335 sccm O2; CDP105ºC/105ºC
E (Volts)
Interim Milestone #2
Best Vacuum
104 103
I (Amps/cm2)
102
101
100
This Project: Selected from over 120 50cm2
FCs.
Dodelet, J. Electrochem. Soc., 151 (2004) A1507 3M present status
17
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
Summary (continued)
• The tremendous activity during the past year has produced an extremely large number of catalyst samples synthesized, screened, characterized, and fully tested. This resulted in improved catalyst performance and is reflected in the quality and quantity of materials characterized by stateoftheart techniques. • The Project has been greatly aided and expanded by the introduction of nanoparticlebased catalysts. • The modeling and characterization work has produced some unexpected results that could provide a lead to the synthetic effort, common to both the vacuum and the nanoparticlebased processes. • The origin of the high coating impedance has been well understood and the catalysts’ stability region has been mapped. • Task oriented, interactive collaboration with Universities and National Labs has been established. • Two manuscripts and one patent application have been submitted.
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
18
Response to Reviewer’s Comments
1. Technology Transfer/Collaboration • Collaboration: Interactive and fruitful collaborations have been established with universities and national labs. • Transfer: Approach emphasizes ease of technology transfer to product commercialization.
Processes used for catalyst synthesis are amenable to scaleup. Produced catalyst in quantities sufficient for 2kW stacks as proved by making multiple MEA’s for 50cm2 fuel cell testing.
2. Planning/Milestones • Besides the decision for scaleup, interim milestones have been established. 3. Initial Performance/Activity Needs Improvement • Performance improved orders of magnitude. Root causes of high impedance understood and ways for overcoming it are being implemented.
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
19
Future Work
Remainder of fiscal 2005
• Direct the synthetic effort towards achieving Milestone #2 performance. • Implement the most recent findings by combining the two synthetic processes to eliminate the impedance issue and achieve a synergetic effect.
2006
• In the nanotechnology area, combine the most promising nanoparticle precursors, appropriately pretreated substrates, application procedures, and thermal treatment processes for best synergetic effects. • In the vacuum processes, utilize the advanced instrumental techniques and the modeling effort for process parameter changes and the nanotechnology accomplishments for the best synergetic effects between the two synthesis paths. • On the fundamental level, point out with more certainty the possible active sites via modeling. Continue and expand the effort to experimentally confirm/identify the nature of the new ORR catalysts. • For best performing catalysts, test stability, peroxide (RRDE, fluoride), etc. • Downselect the catalyst for scaleup and 1 to 2kW stack testing by 3/06.
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
20
Publications and Presentations
Publications: 1. E.B. Easton, T. Buhrmester, J.R. Dahn: “Preparation and Characterization of Sputtered Fe(1 – x) N(x) Films”, submitted to Thin Solid Films 2. M. Jain, S.H. Chou, A. Siedle:
“Structure of FeN C 2 4
Moiety from Quantum Mechanic Study”,
submitted to J. Electrochem. Soc. Presentations: 1. R. Atanasoski: “Recent Advances in the 3M MEA Technology for PEMFC: The Catalysts”, Departmental Seminar, Chemical Engineering Department, Univ. of South Carolina, Columbia, SC, 21 April 2005
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
21
Hydrogen Safety
The most significant hydrogen hazard associated with this project is: • Accidental H2 release in cylinder closet leading to ignition from:
H2 line or manifold breach Accident during replacement of cylinders
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
22
Hydrogen Safety
Our approach to deal with this hazard is: • Design
Hydrogen cylinder closet and gas distribution system adhere to codes. Reduction in number of cylinders in the closet 2step regulators (less susceptible to failure and designed to fail closed) H2 sensors in all labs and cylinder closet, alarm system Automatic shutoff of H2 gas supply if sensors detect H2 release
• Procedures
SOP’s for cylinder changing, alarm responses, test station operation Cylinder changing restricted to highly trained personnel Regular maintenance checks – sensors, leak check of valves, etc.
• Installing H2 Generator (in noninhabited mechanical room) to significantly reduce total volume of H2 in facility
3
Novel Approach to NonPrecious Metal Catalysts – 2005 DOE Hydrogen Program Review, May 23 – 26
23