Near-Surface Alloys forImproved Catalysis

Manos Mavrikakis

Department of Chemical and Biological EngineeringUniversity of Wisconsin – Madison

Madison, Wisconsin 53706

Acknowledgements

Ratko Adzic (BNL)

J. Chen (U of Delaware)

Jeff Greeley, Ye Xu, Anand Nilekar

Jens Nørskov and colleagues @ CAMP – Denmark

DoE (BES-Chemical Sciences)

NSF - CTS

NPACI, DOE-NERSC, PNNL (supercomputing resources)

Outline

• H2 catalytic chemistry:– Identifying Promising Catalysts from 1st Principles:

H2 and H on Bimetallic Near Surface Alloys (NSAs)

• Oxygen Reduction Reaction (ORR)– Improved catalysts with ML structures– Further improvements with mixed-metal ML structures

1. J. Kitchen, N. Khan, M. Barteau, J. Chen, B.

Yashhinskiy, and T. Madey, Surf. Sci. 544 (2003) 295

1Ni/Pt(111)

H2/Rh(111)

2. R. Schennach, G. Krenn, B. Klötzer, K. Rendulic, Surf. Sci. 540 (2003) 237

Hydrogen on NSA’s

2V/Rh(111 )

MethodsDensity Functional Theory – DACAPO total energy code 1,2

Periodic self-consistent PW91-GGA 3

Ultra-soft Vanderbilt pseudo-potentials 4

Plane wave basis sets with 25-Ry kinetic energy cut-off

Spin polarization as needed

Four-metal-layer slabs; (2x2) unit cell; top two layers relaxed

First Brillouin zone sampled at 18 k-points

Nudged Elastic Band method for reaction paths 5

1. B. Hammer, L. B. Hansen, J. K. Nørskov, Phys. Rev. B 59, 1999, 7413.2. J. Greeley, J. K. Nørskov, M. Mavrikakis, Annu. Rev. Phys. Chem. 53, 2002, 319.3. J. P. Perdew et al., Phys. Rev. B 46, 1992, 6671.4. D. H. Vanderbilt, Phys. Rev. B 41, 1990, 7892.5. G. Henkelman, H. Jónsson, J. Chem. Phys. 113, 2000, 9978.

Ideal Bimetallic Near Surface Alloys • Segregation properties of two metals are

critical• Consider two special classes:

– Overlayers– Subsurface Alloys

Overlayers* Subsurface Alloys

Eseg

-1.25 -1.00 -0.75 -0.50 -0.25 0.00 0.25 0.50 0.75 1.00 1.25

|B.E

. Hso

l | - |B

.E. H

host|

-1.25

-1.00

-0.75

-0.50

-0.25

0.00

0.25

0.50

0.75

1.00

1.25

Au*/Pt

Cu/Pt

Ir/Pt

Ni/Pt

Ru/Pt

Co/Pt

Rh/Pt

Mo/PtFe/Pt

Ta/PtV/Pt

W/Pt

Re/Pt

Mo/Pd

Fe/Pd

Ir/Pd

Ru/Pd

V/PdTa/Pd

Mo/Ru

W/Ru Ru/Rh

Ir/Rh

Mo/Rh Re/Rh

W/RhTa/Rh

Pt*/Ru

Pd*/Ru

Cu*/Ru

Ni*/RuCo*/Ru

Fe*/Ru

Cu*/Re

Mo*/Re

Rh*/ReNi*/Re

Pd*/Re Pt*/Re

Pt subsurface alloysPd subsurface alloysIr subsurface alloysRu subsurface alloysRh subsurface alloysRu-based overlayersRe-based overlayers

H-induced segregation

Stability of NSA’s with Respect to Hydrogen-induced Segregation

B.E.H (eV)

-3.3 -3.2 -3.1 -3.0 -2.9 -2.8 -2.7 -2.6 -2.5 -2.4 -2.3 -2.2 -2.1

AuIr CuNiCoMo FeWTaV Pt

Cu/

Pt

Ir/P

tR

h/P

tN

i/Pt

Ru/

Pt

Re/

Pt

Co/

Pt

Fe/P

t

W/P

t

Mo/

Pt

V/P

tTa

/Pt

Pd

Ir/P

d

Ru/

Pd

Re/

Pd

Fe/P

d

Mo/

Pd W

/Pd

V/P

d

Ta/P

d

Ru/

Rh

Rh

Ir/R

h

Mo/

Rh

Re/

Rh

Re

Ni*/

Re

Rh*

/Re

Cu*

/Re

Pt*

/Re

Pd*

/Re

Ni*/

Ru

Pd*

/Ru

Cu*

/Ru

Pt*

/Ru

Ru

B.E.H on Close-Packed Metal SurfacesOverlayers* Subsurface Alloys

Correlation of B.E.H with Clean Surface Properties

d ε (eV)

-3.4 -3.2 -3.0 -2.8 -2.6 -2.4 -2.2 -2.0

B.E

. H(e

V)

-3.2

-3.0

-2.8

-2.6

-2.4

-2.2

-2.0

-1.8

Pt subsurface alloysPd subsurface alloys

Rh subsurface alloys

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

-6.0 -5.0 -4.0 -3.0 -2.0 -1.0 0.0

EbFS (eV, PW91)

Eb

TS (e

V, P

W91

)

O2 dissociation: Does EbTS follow Eb

FS?

Pt

Cu

Ir

Ag

Au +10%

Pt3Co

Pt3Co skin

Pt -2%

Pt3Fe

Pt3Fe skin

Y. Xu, A. V. Ruban, M. Mavrikakis, JACS 126, 4717 (2004).

Hydrogen B.E. (eV)

-0.3 -0.2 -0.1 0.0 0.1 0.2

ETS

(eV

)

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Au

Cu

Noble metals

V/Pd

Pd*/Re

Re/Pd

Ta/PdPd-terminated alloys

Ta/Pt

Mo/Pt

Re/Pt

Ru/PtPt-terminated alloys

BEP Plot for H2 Dissociation onNSA’sJ. Greeley, M. Mavrikakis, Nature Materials 3, 810 (2004)

Pt2+

Pd Pd

CuUPD/Pd + Pt 2+ Pt/Pd + Cu2+

Cu2+

Cu Pt

Pt

Ru

Pt

Ru

Metal monolayer deposition by galvanic displacement of a less noble

metal monolayer deposited at underpotentials

Electroless (spontaneous) deposition of one metal on

another metal

Brankovic, S. R.; Wang, J. X.; Adzic, R. R. Surf. Sci. 2001, 474, L173

Brankovic, S. R.; McBreen, J.; Adzic, R. R. J. Electroanal. Chem. 2001, 503, 99

Sasaki, K.; Wang, J. X.; Balasubramanian, M.; McBreen, J.; Uribe, F.; Adzic, R. R. Electrochim. Acta 2004, 49, 3873

Zhang, J.; Vukmirovic, M.; Xu, Y.; Mavrikakis, M.; Adzic, R. R. Angew. Chem. Int. Ed. 2005, 44, 2132

A new kind of Minority/Defect site for TM Catalysis

Isolated metal hetero-atoms near metal surfaces can be viewed as generating an alternative type of “defect/minority” surface sites, associated with a different kind of near-surface “impurity”.

These sites could be more poison-resistant and possess better catalytic kinetics than the rest of the surface sites.

Pt

V

J. Greeley, M. Mavrikakis, Catalysis Today 111, 52 (2006)

NSA’s - Summary• First-Principles Methods can help with identifying promising

bimetallic NSAs with interesting catalytic properties

• Example: 1. H and H2 on NSA’s: Fine-tuning BEH is possible2. Some NSA’s: Activate H2 easily AND bind atomic H weakly useful for highly selective low T H-transfer reactions

• Developing Catalyst Preparation Techniques with Layer-by-Layer control of metal deposition (ALD-like) is critical for making the desired NSAs

• New type of “NSA” - defect site

Low Temperature Fuel cells Proton Exchange Membrane

(PEM) fuel cellCathodeAnode

Representative Catalysis

Anode:

2H2 ↔ 4 H+ + 4 e-

Cathode:

O2 ↔ 2 O-

2O- +2H+ +2e-↔ 2OH-

2OH- + 2H+ ↔ 2H2O

Oxygen Reduction Reaction (ORR)- Very slow kinetics

Pt monolayers on transition metals

• Pt monolayers on

– Ru(0001), Ir(111), Rh(111), Au(111) and Pd(111)

– Ru(0001), Ir(111) and Rh(111) lead to compression of Pt overlayer

– Au(111) leads to expansion of Pt overlayer

– Pd(111) has almost same lattice constant as Pt(111)

Pt monolayer

Substrate(Ru, Ir, Rh, Pd, Au)

O2 Dissociation Barrier

0.30

0.50

0.70

0.90

1.10

1.30

-4.5 -4.0 -3.5 -3.0

E bO/eV

E bTS

/eV

PtML/Ir(111)

PtML/Au(111)

Pt(111) PtML/Pd(111)/

TSbEeV

/ObE eV

Compressive

strain

Expansive

strain

J Zhang, MB Vukmirovic, Y Xu, M Mavrikakis, RR Adzic, Angew. Chemie, Int. Ed., 44, 2132 (2005)

OH Formation Barrier

0.30

0.50

0.70

0.90

1.10

1.30

-4.5 -4.0 -3.5 -3.0

E bO/eV

E bTS

/eV

PtML/Ir(111)

PtML/Au(111)

Pt(111)

PtML/Pd(111)

/

TSbEeV

/ObE eV

Expansive

strain

Compressive

strain

J Zhang, MB Vukmirovic, Y Xu, M Mavrikakis, RR Adzic, Angew. Chemie, Int. Ed., 44, 2132 (2005)

The best catalyst performs a Balancing Act

0-4.5 -4.0 -3.5 -3.0

Eb /eV

0.3

0.5

0.7

0.9

1.1

1.3

Ea

/eVPd

Au Pt Ir- 2- jK

/mA cm

3

6

9

12

15

18 Pd

Ru

Rh

Ea for O2dissociation (filled red circles)

Ea for OH formation (open circles)

Kinetic currents (jK; blue squares)jk value for Pt(111) (black square) obtained from Markovic et.al., J. Electroanal. Chem. 1999, 467, 157

1--PtML/Ru(0001)

2--PtML/Ir(111)

3--PtML/Rh(111)

4--PtML/Au(111)

5--Pt(111)

6--PtML/Pd(111)

Pt(111)

PtML/Pd(111)

O

J Zhang, MB Vukmirovic, Y Xu, M Mavrikakis, RR Adzic, Angew. Chemie, Int. Ed., 44, 2132 (2005)

Further Decrease of Pt -loading:The role of “OH poisoning” in ORR

• Pt binds OH strongly OH blocks active sites on Pt

• Replace part of the PtML on Pd(111) with another transition metal M

HypothesisM will attract initial OH and induce repulsion on neighbouring Pt-OH decrease OH binding and OH coverage on Pt

OHPt

Pd

OM

Pd

ORR Experiments on (MxPt1-x)ML/Pd(111): Current .vs. Surface Composition

0 20 40 60 80 100

2

4

6

8

10(IrxPt1-x)ML / Pd (111)

(RuxPt1-x)ML / Pd (111)

Pt molar percentage %020406080100

-j k/ m

Acm

-2 a

t 0.8

5V R

HE

M (Ru or Ir) molar percentage %

• Effect of addition of different amount of M: Ir & Ru in PtML/Pd(111)

– Around 20% M gives highest ORR activity

J Zhang, MB Vukmirovic, K. Sasaki, A. U. Nilekar, M Mavrikakis, RR Adzic, JACS, 127, 12480 (2005)

Mixed metal Pt monolayer• Modeling these systems with

DFT:– (Pt3M)ML/Pd(111) to get

θM = 0.25 ML

• Calculate: OH+OH or O+OHrepulsion with M being: Pt monolayer

Pd Substrate

Other metal (M)

Au Pd Pt RhRu

Os Re

OH+OH or O+OH optimal structures

M with much higher reactivity than Pt in PtML/Pd(111): Os, Re (break the OH bond to form H2O and O)

O on M-top and 1st OH on Pt-Pt-bridge site

M with reactivity higher than Pt in PtML/Pd(111): Rh, Ru, Ir

1st OH on M-top and 2nd OH on Pt-top

M with reactivity less or equal compared to Pt in PtML/Pd(111): Au, Pd, Pt

1st OH on Pt-top and 2nd OH on Pt-Pt-bridge site

R2 = 0.99

12

17

22

27

32

37

42

47

52

-0.2 0 0.2 0.4 0.6 0.8 1

OH-OH or OH-O repulsion relative to OH-OH repulsion on PtML/Pd(111) /eV

-j k a

t 0.8

V / m

A.c

m-2

Au

Ru

Ir Os

Re

Pd

Pt

Rh

Comparison of theory with experiment

(Pt3M)ML/Pd(111)

400% increase in ORR activity for (Pt3Re)ML/Pd(111) compared to Pt(111)

M

J Zhang, MB Vukmirovic, K. Sasaki, A. U. Nilekar, M Mavrikakis, RR Adzic,

JACS, 127, 12480 (2005)

ORR - Summary:• A combination of first-principles studies with experiments can identify key

ORR reactivity descriptors:

• Kinetics of O-O Bond-breaking

• Kinetics of O-H Bond-making

• OH-poisoning

• Use less Pt: PtML/Pd(111) 30% more current [compared to Pt(111)]

• Use even less Pt: Pt3M/Pd(111) up to 400% increase in current [compared to Pt(111)]

• First-principles can help with further improvements of ORR catalysts, by identifying materials which are:

– Cheaper (low Pt loading)

– More active

– Increased stability of Pt against oxidation

BEO

Anand Nilekar

University of Wisconsin-Madison

Department of Chemical and Biological Engineering

Dr. Ye XuDr. Jeff Greeley

Acknowledgements

Ratko Adzic (BNL)

J. Chen (U of Delaware)

Jeff Greeley, Ye Xu, Anand Nilekar

Jens Nørskov and colleagues @ CAMP – Denmark

DoE (BES-Chemical Sciences)

NSF - CTS

NPACI, DOE-NERSC, PNNL (supercomputing resources)