Heterogeneous CatalysisOpportunities and challenges

• Challenges–Societal needs–Developing the basic understanding

• Opportunities–Designing at the nano-scale

J. K. Nørskov Center for Atomic-scale Materials Physics

Technical University of Denmarknorskov@fysik.dtu.dk

Challenges I

Jens Rostrup-Nielsen: XVII Sympósio Iberoamericano de Catálisis, July 16-21, 2000

Dream reactions waiting for a catalyst:

Dreaming on ….

• Heterogeneous catalysts for assymmetric synthesis• Photolytic water splitting (hydrogen economy)• Biomimetics, synthetic enzymes• Non-thermal processes in general

(e.g. electro- and photocatalysis)• …

See: E. Derouane, CATTECH 5, 226 (2001)

Challenges II

Challenges III

The science of heterogeneous catalysis:

• A comprehensive scientific basis– Much has been done

– Much more is needed (oxides, size effects, photocatalysis, electrocatalysis, relation to homogeneous and enzyme catalysis …)

• Making the insight useful!– The ultimate test

Opportunities- design at the nano-scale

• Rational catalyst design- Discovery on the basis of insight

• Data-driven methods- Accelerated discovery by access to

large amounts of data

• Bio-inspired catalysis

Rational catalyst design

1. What determines the catalytic activity/selectivity/lifetime ?

2. How can we affect it?

- We have tremendous new possibilities

Ammonia synthesisN2+3H2 2NH3

Ozaki and Aika, Catalysis 1 (Anderson and Boudart, Ed.)

Ammonia synthesis over Ru

Ru(0001)

step

Logadottir, Nørskov

Steps do everything

Dahl, Logadottir, Egeberg, Larsen, Chorkendorff, Törnqvist, Nørskov, Phys.Rev.Lett. 83, 1814 (1999)

Au decorates steps:Hwang, Schroder, Gunther, Behm, Phys. Rev. Lett. 67, 3279 (1991) 

Logatottir, Rod, Nørskov, Hammer, Dahl, Jacobsen, J. Catal. 197, 229 (2001)

The Brønsted-Evans-Polanyi relation

-0.8 -0.4 0.0 0.4 0.8[E-E(Ru)](eV/N2)

10-5

10-4

10-3

10-2

10-1

100

101

TO

F(s

-1)

Fe

Mo

Ru

Co

Ni

Os

Calculated ammonia synthesis rates400 C, 50 bar, H2:N2=3:1, 5% NH3

Logatottir, Rod, Nørskov, Hammer, Dahl, Jacobsen, J. Catal. 197, 229 (2001)

Interpolation in the periodic table

Jacobsen, Dahl, Clausen, Bahn, Logadottir, Nørskov, JACS 123 (2001) 8404.

Jacobsen, Dahl, Clausen, Bahn, Logadottir, Nørskov, JACS 123 (2001) 8404.

Interpolation in the periodic table

Jacobsen, Dahl, Clausen, Bahn, Logadottir, Nørskov, JACS 123 (2001) 8404.

Measured ammonia synthesis rates 400 C, 50 bar, H2:N2=3:1

Data driven methods

• High throughput screening– Direct testing of many catalysts, fast,

efficiently

• Data mining – Correlating catalytic activity/selectivity/

durability to descriptors that can be tabulated

The object of the game…

• Find sets of descriptors {Dik} of solid materials Mi , and a mathematical model F such that Aij being the Turn Over Frequency of Mi as catalyst for the reaction j at operationg conditions Cj one has:

• Identify ranges of Dik that maximize F• Screen Databases of Materials Properties before

screening real materials• Better if one descriptor is sufficient, but do not take it

for granted• Much better if F has a sound physical basis• Adsorbate/substrate bond strengths should provide

good descriptors according to the Sabatier principle

jikjiij CDFCMA ,,

Using DFT calculations

in the search of prospective catalysts

H. Toulhoat and P. Raybaud

Workshop Catalysis from First Principles Vienna 02/02

Periodic Trends for E MC in Fm-3m (NaCl) carbides

ScCr

VMn

CoNi

Y

Zr

TcRu

Rh

Pd

Ta

ReOs

Ir

Ti

Fe

Cu

NbMo

La

Hf

W

Pt

AuAg

0

20

40

60

80

100

120

140

160

E M

C Y

Y/P

AW

/GG

A/S

P (

kJ

/mo

l)

Using DFT calculations

in the search of prospective catalysts

H. Toulhoat and P. Raybaud

Workshop Catalysis from First Principles Vienna 02/02

Adsorption of C2H4 100K

: di-s bound : p bound

: No ads.

Sc Ti V Cr Mn Fe Co Ni Cu

Y Zr Nb Mo Tc Ru Rh Pd Ag

La Hf Ta W Re Os Ir Pt Au

Ru

Rh

Fe

CrW

Ni

Ta

Pt (Diss.)

Ag

AuCu Pt(molec)

Pd

y = 6,267x - 287,99

R2 = 0,894

y = 1,9625x - 29,758

R2 = 0,791

0

100

200

300

400

500

600

0 20 40 60 80 100 120 140

E MC YY (@NaCl/PAW/GGA/SP) (kJ/mol)

Qa

ds

ex

p. C

2H

4/M

(k

J/m

ol)

• E MC @ Fm-3m carbides is rather consistent with simple chemisorption models

• Onset of dissociative chemisorption as MC bond strength increases

Using DFT calculations

in the search of prospective catalysts

H. Toulhoat and P. Raybaud

Workshop Catalysis from First Principles Vienna 02/02

Re3Ir

Ir3Re

Ir3Cu

Cu3Ir

Cu

Pd

Ru

IrPt

Os

Re

Co

Ni

1,E+00

1,E+01

1,E+02

1,E+03

1,E+04

1,E+05

10 12 14 16 18 20 22 24 26 28 30E MC YY (kCal.mol-1)

Ra

te i

n H

yd

rog

en

ati

on

of

C6

H6

(s

-1) TOF@30°C (M/Al2O3

Brunelle et al., 1977)

V théor.

• The experimental Alloying effects is correctly predicted

Using DFT calculations

in the search of prospective catalysts

H. Toulhoat and P. Raybaud

Workshop Catalysis from First Principles Vienna 02/02

Getting data/descriptors

• Structure (in situ)

• Spectroscopy (in situ)

• Surface thermochemistry

• Calculations

• …

There is a large need for systematic data

- and for good descriptors

Structure-activity CorrelationHydrodesulfurization of thiophene

1.5

10

0.5

00 1 2 3Number of Co edge atoms

(x1020/g catalyst)

HD

S a

cti

vit

y

(x10

2/m

ol/g

/h)

Topsøe, Clausen, MassothHydrotreating Catalysis, Science and Technology(Anderson and Boudart (Eds.), Springer (1996).

Descriptors from spectroscopyCO TPD shift Core level shift

Goodman and Rodriguez, Science 279 (1992) 897

Single crystal microcalorimerty

Cu/MgO

Ag/MgO

Pb/MgO

Larsen, Starr, Campbell, Chem.Thermodyn. 33, 333 (2001)Brown, Kose, King, Chem. Rev. 98, 797 (1998).

Descriptors from

DFT

Correlation between adsorptionenergies and activation barriersand the d-band center

Mavrikakis , Hammer, NørskovPhys. Rev. Lett. 81, 2819 (1998)

CO tolerance of Pt alloy anodes for PEM fuel cells

Pt

M

-0,5

0,0

0,5

1,0

1,5

2,0

Au Ir Ag Pd Rh Ru Cu Ni Co Fe Pt

EC

O ,

eV

- d,

eV

Substrate M

-0,2

0,0

0,2

0,4

0,6

0,8

1,0

0,0

0,2

0,4

0,6

0,8

1,0

Measured overages of CO on the alloy electrodes with 100 ppm CO/H2

M. Watanabe et al., Phys. Chem. Chem. Phys. 3 (2001) 306

1- co

Calculated changes in CO adsorption energy

S. Gottesfeld et al., J. Electrochem. Soc. 148 (2001) A11.

   Christoffersen, Liu, Ruban, Skriver, Nørskov, J.Catal. 199, 123 (2001)

How can the d-band center be changed?Calculated d band shifts:

Ruban, Hammer, Stoltze, Skriver, Nørskov, J.Mol.Catal. A 115, 421 (1997)

Overlayer

Host

Methane activation

Bengaard, Rostrup-Nielsen, Nørskov

bTransition state for CH4

dissociation on Ni(211)

Methane activation on Ni/Ru

Ni Coverage [ML]

0 1 2

Initi

alst

icki

ngpr

obab

ility

0

1e-7

2e-7

3e-7

4e-7

5e-7

Thermal dissociation of CH4 at T = 530 K

Egeberg, Chorkendorff, Catal. Lett. 77, 207 (2001)

Lessons from biology

• Catalysis at ambient temperature and pressure

• Extreme selectivity

• Direct coupling of energy into the important reaction coordinate (non-thermal catalysis)

Nitrogenase 8e8HN2 23 H2NH

nitrogenase

ATP

FeP +2(MgATP) MoFeP 1k

1-kFeP 2(MgATP) MoFeP

4k 2knucleotidereplacement

ATP cleavageelectron transfer

oxFeP 2(MgADP) + MoFeP 3k

3-k oxFeP 2i )P(MgADP, MoFeP

reduction

complex formation

complex dissociation

-4AlFADP

Fe protein

Fe protein

MoFe protein 4Fe-4S cluster

P-cluster

FeMo cofactor

Burgess, Lowe, Chem. Rev. 96, 2983 (1996)Schindelin, Kisker, Schlessman, Howard, Rees, Nature 387, 370 (1997)

N2 hydrogenation on FeMoco

Rod, Nørskov JACS 122, 12751 (2000)

The Fe Protein cycle

1)

2)

3)

4)

E

E

E

MoFe protein

FeMoco P-cluster

Fe protein

4Fe-4S cluster

ATP

ADP 24HPO

See also: Spee, Arendsen, Wassnik, Marrit, Hagen, Haaker, FEBS Lett. 432, 55 (1998)

Comparing the FeMoco and Ru(0001)

Rod, Logadottir, Nørskov J.Chem.Phys. 112, 5343 (2000)

Status

• Well developed basic understanding – theory-experiment

• Beginning to be able to use it directly in catalyst design

• Some activity-descriptor correlations

• Host of new in situ methods for catalyst characterization

• New very powerful screening methods

• We have a starting point which is radically different from the situation 5 or 10 years ago!

Moving forward

• More basic understanding –theory-experiment

• Integration of the conceptual framework for heterogeneous, homogeneous and enzyme catalysis

• More systematic data (descriptors)

• Better synthesis methods

• Better coupling of catalyst design and process engineering

• INTEGRATION

Promoting development

Synthesistestingcharacterization

Experiments, models

Theory

An integrated approach: