Idea Transcript
FCTO Monthly Webinar Series: Past, Present and Future Challenges in Electrocatalysis for Fuel Cells
Presenter: Voya (Vojislav) Stamenkovic Argonne National Laboratory 1 | Fuel Cell Technologies Office
DOE Host: Dimitrios Papageorgopoulos Fuel Cell Technologies Office
U.S. Department of Energy Fuel Cell Technologies Office October 13th, 2016 eere.energy.gov
Question and Answer • Please type your questions into the question box
2 2 | Fuel Cell Technologies Office
eere.energy.gov
Past, Present and Future Challenges in Electrocatalysis for Fuel Cells Materials Science Division A r g o n n e N a t i o n a l Laboratory Vojislav Stamenkovic
DOE EERE Fuel Cell Technology Office Webinar Series October 13, 2016
https://www.toyota.com/mirai/fcv
2016 Toyota Mirai $499/month 36-month lease $57,500 MSRP
HYDROGEN SAFETY Science behind the safety Safety was our primary objective when we engineered the Mirai. Our proprietary safety system is based on four fundamental principles.
The first production fuel cell vehicle The Mirai is at the forefront of a new age of hydrogen fuel cell cars that allows you to enjoy long distance zero-emissions driving. As well as only producing water from its tailpipe – which means no impact on our planet when you're driving – Mirai brings the unique Toyota Hybrid driving experience to a new level.
’60s PEM FC unsupported Pt catalyst loading: 28 mgPt/cm2 membrane: Nafion T ~ 21oC
’60-70s Alkaline FC Ni based catalyst T ~ 250oC Operating time: 400-690h
’80-2011 Alkaline FC Pt and PtAu catalysts loading: 10 and 20 mgPt/cm2 T ~ 93oC Operating time: > 16 days
Particle Size Effect 10 % Pt/Vulcan 3 nm
30 nm (Pt/C)
cubo-octahedral
particles
5
ORR: cathode limitations 2020 DOE Technical Targets • Mass activity @0.9V: 0.44 A/mgPt
losses ~ 30-40% (!!!)
Cathode kinetics (?!)
• Electrochemical area loss: < 40% • PGM Total content: 0.125 g/kW • PGM Total loading: 0.125 mg/cm2electrode • Durability w/cycling (80oC): 5000 hrs
Main limitations in PEM fuel cell technology: 1) Activity: Pt/C = Pt-poly/10 2) Durability: (Pt catalyst dissolves) 3) Pt loading: Cost Issues 7
Activity | Durability | Cost Specific/Mass Activity
Electrochemical Stability
Loading
1o Structure-Function 2o Composition-Function 3o Surface Modifications 4o Tailored Electrolytes
STM: Pt Single Crystals As prepared
Pt(111)
(b)
(c)
Current Density (a. u.)
(a)
Pt(100) 12.0
Pt(110)
Kinetic Current Density [mA/cm^2]
As prepared
10.0
8.0
Pt(110) Pt(111) Pt(100) Series4 Pt-poly average
6.0
4.0
2.0
0.0
0.925 V
0.900 V
0.875 V
0.850 V
9
In-Situ EC-ICP-MS
Pt(hkl)-Surfaces vs. Pt/C Quadrupole mass filter
Electrochemical Cell
Horizontal torch
Total Pt loss over one potential cycle up to 1.05 V for distinct Pt surface morphologies, indicating the stability trend follows the coordination number of the surface sites
Pt Surface
Dissolved Pt per cycle [µML]
Pt(111)
2
Pt(100)
7
Pt(110)
83
Pt-poly
36
Pt/C
| 103*
P. P. Lopes, D. Strmcnik, J. Connell, V. R. Stamenkovic and N.M. Markovic
ACS Catalysis, 6 (4), 2536-2544, 2016
10
Surface Structure + Composition: Pt3Ni[hkl] Surfaces LEED
STM
(d)
Pt3Ni(111)
LEIS
p(1x1)
(e)
[01
-1 ]
[0 11 ]
Sputtered
Pt3Ni(100)
c(5x1)
(f)
Pt3Ni(110)
Annealed
[1-10]
[001]
(110)-(1x1)
11
Activity: ORR Platinum Alloy Surfaces Pt(hkl)
Pt3Ni(hkl) Pt3Ni(hkl)
0.1 M HClO4
0.1 M HClO4 0.95 V vs. RHE 20 °C
Pt3Ni(hkl) Pt (111) (100) (110) (111)
Science, 315 (2007) 493
Pt3Ni(110)
Pt3Ni(100)
Pt3Ni(111)
Pt3Ni(111)/Pt-Skin Surface is the most active catalyst for the ORR 12
Subsurface Composition + Surface Structure: Pt3Ni(111) 0.6 0.8 1.0
E [V] vs. RHE 0.2
0.4
(a)
7.0
6.0
(a’)
80
20
60
40
40
60
20
80
1
60
SXS
100
0 2
3
4
5
y Atomic layer
6
(b)
7
Pt3Ni(111)
2
i [µA/cm ]
0
Ni [at. %]
6.5
%]
100
Pt [at. Pt %] [at.
Intensity [arb. units]
0.0
30 0
Pt(111)
-30
CV
2
i [mA/cm ]
H2O2 [%]
60
(c)
60
30
30
0
0
50 0
ΘOxd [%]
ΘHupd [%]
-60
(d)
-2
∆E~100 mV
-4
ORR
Pt poly
-6
II
I 0.0
0.2
0.4
III 0.6
E [V] vs. RHE
0.8
1.0
Science, 315(2007)493
Pt3Ni(111)/Pt-Skin Surface is the most active catalyst for the ORR (100-fold enhancement) 13
Subsurface Composition + Surface Structure: Pt3Ni(111) 0.6 0.8 1.0
E [V] vs. RHE 0.2
0.4
(a)
7.0
6.0
(a’)
80
20
60
40
40
60
20
80
1
60
SXS
100
0 2
3
4
5
y Atomic layer
6
(b)
7
Pt3Ni(111)
2
i [µA/cm ]
0
Ni [at. %]
6.5
%]
100
Pt [at. Pt %] [at.
Intensity [arb. units]
0.0
30 0
Pt(111)
-30
CV
2
i [mA/cm ]
H2O2 [%]
60
(c)
60
30
30
0
0
50 0
ΘOxd [%]
ΘHupd [%]
-60
(d)
-2
∆E~100 mV
-4
ORR
Pt poly
-6
II
I 0.0
0.2
0.4
III 0.6
E [V] vs. RHE
0.8
1.0
Science, 315(2007)493
Pt3Ni(111)/Pt-Skin Surface is the most active catalyst for the ORR (100-fold enhancement) 13
Controlled Synthesis: Multimetallic Nanocatalysts Colloidal solvo - thermal approach has been developed for monodispersed PtM NPs with controlled size and composition
Efficient surfactant removal method does not change the catalyst properties
PtMControlled Alloy NPsSynthesis: | Distribution: Elements and Particle Size Multimetallic Nanocatalysts HAADF - STEMhas solvo - thermal approach
Colloidal been developed for monodispersed PtM NPs with controlled size and composition
1o Particle size effect applies to Pt-bimetallic NPs Specific Activity increases with particle size: 3 < 4.5 < 6 < 9nm
Mass Activity decreases with particle size
Particle size distribution
Optimal size particle size ~5nm J. Phys. Chem. C., 113(2009)19365 100
davg = 5.9segregation nm 2o Temperature induced in Pt-bimetallic NPs Agglomeration prevented 80
Counts
(b)
60
Optimized annealing temperature 400-500oC Phys.Chem.Chem.Phys., 12(2010)6933 40
3o Composition effect in Pt-bimetallic NPs Pt3M
20
PtM2
PtM
PtM3
0.22 nm
1 nm
0
1 nm
1 nm
1 2 of 3 Pt-bimetallic 4 5 6 NPs7 is PtM 8 9 Optimal composition
10
Adv. Funct. Mat., 21(2011)147 Size (nm)
4o Surface chemistry of homogeneous Pt-bimetallic NPs PtxM(1-x) NPs
Efficient surfactant removal method does not change the catalyst properties
ACS Catalysis, 1 (2011) 1355
Dissolution of non Pt atoms forms Pt-skeleton surface
14
Controlled Synthesis: Multimetallic Nanocatalysts Colloidal solvo - thermal approach has been developed for monodispersed PtM NPs with controlled size and composition
1o Particle size effect applies to Pt-bimetallic NPs Specific Activity increases with particle size: 3 < 4.5 < 6 < 9nm
Mass Activity decreases with particle size Optimal size particle size ~5nm J. Phys. Chem. C., 113(2009)19365
2o Temperature induced segregation in Pt-bimetallic NPs Agglomeration prevented (b)
Optimized annealing temperature 400-500oC Phys.Chem.Chem.Phys., 12(2010)6933
3o Composition effect in Pt-bimetallic NPs Pt3M
PtM2
PtM
PtM3
0.22 nm
1 nm
1 nm
1 nm
Optimal composition of Pt-bimetallic NPs is PtM Adv. Funct. Mat., 21(2011)147
4o Surface chemistry of homogeneous Pt-bimetallic NPs PtxM(1-x) NPs
Efficient surfactant removal method does not change the catalyst properties
Dissolution of non Pt atoms forms Pt-skeleton surface
PtxNi1-x: Surface Chemistry and Composition Effect
Adv. Funct. Mat., 21 (2011) 14715
15
PtxNi1-x: Surface Chemistry and Composition Effect
Adv. Funct. Mat., 21 (2011) 14715
15
PtNi with multilayered Pt-Skin Surfaces : Tailoring Nanoscale Surfaces Temperature annealing protocol used to transform PtNi1-x skeletons to multilayered PtNi/Pt NPs with 2-3 atomic layers thick Pt-Skin PtNi/Pt core/shell As-prepared PtNi PtNi1-y Skeleton
400oC
HRTEM
Intensity (a. u.)
Line profile
Intensity (a. u.)
Pt Ni
0
1
2
3
Position (nm)
4
5
0
1
2
3
Position (nm)
4
5
16
NPs with multilayered Pt-Skin Surface: PtNi/C Leached
As Synthesized
Annealed
Multilayered Pt-skin NP
∆
TEM
Catalysts with multilayered Pt-skin surfaces exhibit substantially lower coverage by Hupd vs. Pt/C (up to 40% lower Hupd region is obtained on Pt-Skin catalyst)
Pt-NPs
Surface coverage of adsorbed CO is not affected on Pt-skin surfaces
PtNi-NPs
davg = 5 nm
Ratio between QCO/QHupd>1 is indication of Pt-skin formation Catalyst Pt/C PtNi/C PtNi-skin/C
QH (µC) 279 292 210
ECSAH (cm2) 1.47 1.54 1.10
QCO (µC) 545 615 595
ECSACO (cm2) 1.41 1.60 1.54
QCO/2QH 0.98 1.05 1.42
Electrooxidation of adsorbed CO (CO stripping) has to be performed for Pt-alloy catalysts in order to avoid underestimation electrochemically active surface area and overestimation of specific and mass activities
17
PtNi Catalyst: RDE Studies of Multilayered Pt-Skin Surfaces
0.0 8340
8360
E (eV)
8380
0.5
acid treated acid treated/annealed Pt
0.0 11540
11560
11580
E (eV)
11600
11620
1.0
-0.1 1.00
d 2 ) i (mA/cmgeo
1.0
0.8
0.0
8400
Pt L3 at 1.0 V
0.6
0
-2
e
0.95
0.90
0.1
1 2
jk (mA/cm )
-4
1.0
before 0.8
after
6
0.6
8
6
4
0.4
4
2
0.2 0.0
10
Improvement Factor (vs. Pt/C)
acid treated acid treated/annealed Ni NiO
0.4
6 nm Pt/C acid treated PtNi/C acid treated/annealed PtNi/C
E (V vs. RHE)
0.5
1.5
Normalized Absorption µ(E)
0.2
0.1
1.0
8320
b
0.0
Specific Activity (mA/cm2)
Ni K at 1.0 V
1.5
E (V vs. RHE)
c
i (mA/cm2Pt)
Normalized Absorption µ(E)
a
Pt/C
PtNi/C Skeleton
before
PtNi/C Skin
0
2
0
after
-6 0.6
0.7
0.8
0.9
1.0
E (V vs. RHE)
TEM/XRD: Content of Ni is maximized and allows formation of the multilayered Pt-skin by leaching/annealing RDE: PtNi-Skin catalyst exhibits superior catalytic performance for the ORR and is highly durable system In-Situ XANES: Subsurface Ni is well protected by less oxophilic multilayered Pt-skin during potential cycling Durability: Surface area loss about 10%, SA 8 fold increase and MA 10 fold increase over Pt/C after 20K cycles
J. Amer. Chem. Soc. 133 (2011) 14396
18
Mass Activity Enhancement by 3D Surfaces: Multimetallic Nanoframes In collaboration with Peidong Yang, UC Berkeley
- H2PtCl6 and Ni(NO3)2 react in oleylamine at 270oC for 3 min forming solid PtNi3 polyhedral NPs - Reacting solution is exposed to O2 that induces spontaneous corrosion of Ni - Ni rich NPs are converted into Pt3Ni nanoframes with Pt-skeleton type of surfaces - Controlled annealing induces Pt-Skin formation on nanoframe surfaces
Science , 343 (2014) 1339
20
Compositional Profile: PtNi Nanoframes with Pt-skin Surfaces
- Narrow particle size distribution - Hollow interior - Formation of Pt-skin with the thickness of 2ML - Surfaces with 3D accessibility for reactants - Segregated compositional profile with overall Pt3Ni composition Science , 343 (2014) 1339
21
Multimetallic Nanoframes with 3D Electrocatalytic Surfaces
Science , 343 (2014) 1339
22
Improving the ORR Rate by Protic Ionic Liquids [MTBD][beti]
Chemically Tailored Interface High O2 solubility along with hydrophobicity yield improved ORR kinetics
CO2 ,[MTBD][beti ] = 2.40 ± 0.013 CO2 ,HClO4
0.95 V vs. RHE
Pt/C
Erlebacher and Snyder, Advanced Functional Materials, 23(2013)5494
np-NiPt/C
np-NiPt/C+IL
23
Tailoring Activity: PtNi Nanoframes as the ORR Electrocatalyst
RDE @ 0.95V vs. RHE 0.1M HClO4 1600 rpm
- No change in activity after 10K cycles 0.6 – 1.0 V - Specific activity increase over 20-fold vs. Pt/C
C
RDE @ 0.90V vs. RHE 0.1M HClO4 1600 rpm
- Mass activity increase over 35-fold vs. Pt/C - Increase in mass activity over 15-fold vs. DOE target
Science , 343 (2014) 1339
24
DURABILITY: Pt/C
Initial morphology
After 60,000 cycles Potential Range: 0.6-1.0V
25
Commercial Pt/C Catalysts
c
d
•
Commercial catalysts are usually made by impregnation methods.
•
Poor control – Broad size distribution – Different, undefined morphologies, but everyone calls them “cubo-octahedral”, which is, in fact, not correct!
e
26
DURABILITY: Pt/C
Initial morphology
Potential Range: 0.6-1.0V
After 60,000 cycles
27
DURABILITY: Pt/C
Initial morphology
Potential Range: 0.6-1.0V
After 60,000 cycles
27
SIZE EFFECT(s)? Pt/C
28
SHAPE EFFECT(s)? Pt/C
20 nm
cubo-octahedron
truncated cube
cube
• Uniform morphology • Size control octahedron (?) cubo-octahedral Cube 29
HR-TEM: Characterization of Nanoscale Pt/C Catalyst 1) Shape: cubooctahedron 2) Size distribution: 2-15 nm 3) Composition: Pt, C
x 15
4) Side Orientation: [111], [100]
x5
3 nm
30
APPROACH: Well-Defined Systems (ext & nano) (111) (100) [001]
Kinetic Current Density [mA/cm^2]
12.0
0.08 0.06
8.0
6.0
4.0
2.0
2.0x10-4
0.0
0.04
0.925 V
0.900 V
0.875 V
0.850 V
0.02 0.00
I/A
i [mA/cm2]
10.0
Pt(110) Pt(111) Pt(100) Series4 Pt-poly average
0.0
-0.02 -0.04 -0.06
-2.0x10-4
-0.08 0.0
0.2
0.4
0.6
E [V vs RHE]
0.8
1.0
0.0
0.2
0.4
0.6
E/V vs. RHE
31
ICP-MS: Principle of Operation Quadrupole mass filter
Horizontal torch
Quadrupole
Quadrupole allows single m/z pass at any given time Time frame for the range of m/z = 1 – 240 is 0.1sec
Detector
32
RDE/ICP-MS: Pt(hkl)
Rotating Disk Electrode 333
In-Situ RDE / ICP-MS: Standard Deviation
Pt(hkl)-Surfaces vs. Pt/C Total Pt loss over one potential cycle up to 1.05 V for distinct Pt surface morphologies, indicating the stability trend follows the coordination number of the surface sites
σ=(Σ(µ-xi)2/n-1)1/2 DL=3σbckg(λ)
Pt Surface
Background
λ (amu)
Dissolved Pt per cycle [µML]
Pt(111)
2
Pt(100)
7
Pt(110)
83
Pt-poly
36
Pt/C
| 103*
Method detection limit (DL) is applied to blank sample prepared with all analytical steps related to given methodology, since each step is potential error source Limit of quantification (LOQ) is 3 times DL
34
In-Situ RDE / ICP-MS:
Pt/C, Pt and Pt/Au Well-Defined Surfaces
GC-Au-Pt(4ML)
GC-Pt(4ML)
Pt/C
| 103*
35
DURABLE NPs: Core-Interlayer-Shell Particles 0.2 A
0.4
0.6
I
II
E (V vs RHE) 0.8
1.0
1.2
1.4
III
1.6
1.8
IV
40 x3
I (µA cm-2)
0 -40
I would love double this area to blank out H Au-OH Pt-OH layer I would love this area to blank out 3 -80 love this area to blank out Au(111) 2I would Au(111)-Pt 1I would loveNiPt this area to blank out 8 ORR @ 0.9V Au(111)-FePt
i (mA cm )
-2
-2 -3 -4 -5 -6 0.2
0.4
0.6
0.8
1.0
E (V vs RHE)
C 3
6 2 4 2 0
Pt -p Au oly (1 11 Fe )-Pt P Au 3t -p o (1 11 ly )-F eP t 3
33
B 0 -1
0.0
ad
ad
1
Improvement Factor vs Pt Poly
upd
Specific Activity (ikin / mA cm-2)
4
I want 1.4 this blanked 1.2 1.6 out! 1.8
Nano Letters, 11 (2011) 919
36
DURABLE NPs: Core-Interlayer-Shell Particles 0.2 A
0.4
0.6
I
II
E (V vs RHE) 0.8
1.0
1.2
1.4
III
1.6
1.8
IV
40 x3
I (µA cm-2)
0 -40
I would love double this area to blank out H Au-OH Pt-OH layer I would love this area to blank out 3 -80 love this area to blank out Au(111) 2I would Au(111)-Pt 1I would loveNiPt this area to blank out 8 ORR @ 0.9V Au(111)-FePt
i (mA cm )
-2
-2 -3 -4 -5 -6 0.2
0.4
0.6
0.8
1.0
E (V vs RHE)
C 3
6 2 4 2 0
Pt -p Au oly (1 11 Fe )-Pt P Au 3t -p o (1 11 ly )-F eP t 3
33
B 0 -1
0.0
ad
ad
1
Improvement Factor vs Pt Poly
upd
Specific Activity (ikin / mA cm-2)
4
I want 1.4 this blanked 1.2 1.6 out! 1.8
Nano Letters, 11 (2011) 919
36
Compositional Profile: Core/Interlayer/Shell Electrocatalysts Au interlayer Ni core
2 nm
2 nm
4nm
PtNi shell
~ 5-6nm
Monodisperse , Core/Interlayer/Shell NPs: Ni core / Au interlayer / PtNi shell
37
Durable ORR Catalysts: Core/Shell NPs with Au Interlayer Stabilization mechanism
- Pt
- Au
- Ni - Fe
-O
segregation trend of Pt into the bulk segregation trend of Au onto surface driving force that diffuses Pt into the bulk driving force induced by strong Pt - OHad interaction
Nano Letters, 14 (2014) 6361
38
Durable Systems: PtNi Nanoframes in MEA in collaboration with Debbie Myers, ANL - CSE
Nanoframes in 5cm2 MEA ANL and ORNL
Cathode Loading: 0.035 mg-Pt/cm2, I/C = 0.8 H2/O2, 80°C, 150 kPa(abs), 100%RH ORR Activity @ 0.9V: Mass Activity x3.5 Specific Activity x6.5 TKK 20 wt%Pt/C: 0.22 A/mg-Pt 0.39 mA/cm2-Pt PtNi Nanoframes: 0.76 A/mg-Pt 2.60 mA/cm2-Pt
39
Durable Systems: PtNi Nanoframes (before and after) Nanoframes in 5cm2 MEA ANL and ORNL BF-STEM
catalyst layer
HAADF-STEM
0.1 µm
membrane
0.1 µm
BF-STEM
5 nm
membrane Pt Ni
5 nm
40
Durable Systems: PtNi with Multilayered Pt-Skin Surfaces 1.00
IR corrected Cell Voltage (V)
0.95
Commercial TKK Pt/C PtNi/C
0.90 0.85
H2/O2 Performance
0.80 0.75 0.70 0.65 0.60 0.55 0.50
0
500
1000
1500
2000
2500
3000
Cell Voltage without IR compensation (V)
in collaboration with Debbie Myers, ANL - CSE 1.0 Commercial TKK Pt/C PtNi/C
0.8
H2/Air Performance 0.6
0.4
0.2
0
TKK 20 wt%Pt/C PtNi 16.7 wt%Pt/C
600
900
1200
PtNi
TKK Pt
Pt loading
mgPGM/cm²geo
0.045
0.045
Mass Activity (H2-O2)
A/mgPGM @ 0.9 ViR-free
0.60
0.27
Specific Activity (H2-O2)
mA/cm2PGM @ 0.9 ViR-free
1.85
0.39
mA/cm2 @ 0.8 V
101
47
m2/gPGM
35.10
52.5
ECSA
1800
Current Density (mA/cm )
Units
MEA performance (H2-Air)
1500 2
Current Density (mA/cm2)
Cathode Loading: 0.046 mg-Pt/cm2 I/C = 1, H2/O2 (or Air), 80°C, 150 kPa(abs), 100%RH
300
41
TM doped Pt3Ni Octahedra
SA and MA over 70-fold vs. Pt/C
Huang et al. Science , 348 (2015) 1230
Hierachical PtCo Nanowires
SA and MA over 30-fold vs. Pt/C
Bu et al. Nat. Commun. , (2016) DOI 10.1038.natcomms11850
Dealloying of PtNi octahedrons
SA and MA over 10-fold vs. Pt/C
Cui et al. Nature Mat., 12 (2013) 765
Alloying Pt with Rare Earth Elements
SA over 6-fold vs. Pt/C Greeley et al. Nature Chem., 1 (2009) 552
Escudero-Escribano et al. Science, 352 (2013) 73
Pt/CuNW and PtNT
Pt/CuNW
PtNT
SA and MA over 3-fold vs. Pt/C
Alia et al. ACS Catalysis., 3 (2013) 358
BNL: Electrocatalyst Development
Remaining Challenges and Barriers
1) Durability of fuel cell stack (