3 +CH Fundamentals of CH Heterogeneous Catalysis · Heterogeneous Catalysis a molecular view of catalysis by metals Hans Niemantsverdriet Schuit Institute of Catalysis . Contents

1

CH

2+C

H3

*Sasol Technology R&D

Fundamentals of Heterogeneous Catalysisa molecular view of catalysis by metals

Hans NiemantsverdrietSchuit Institute of Catalysis

Contents • Introduction

• Catalysis in terms of kinetics; coverage is the key

• Bonding on surfaces: Molecular Orbital View

• Catalysts break bonds!….How? Dissociative Adsorption

• potential energy description

• molecular orbital picture

• thermodynamics

• Elementary reaction steps and kinetics

• Adsorbate interactions and the effect on kinetics

• Concluding Remarks

What is Catalysis?

bonding

reaction

separation

AB

catalyst

catalyst

catalyst

P

P

A B

• Catalysis is a cycle of elementary steps (at least three) • Catalytic sites are regenerated

What is a catalyst?

Catalysts• increase the rate of a reaction• without being consumed in the process

offer alternative, energetically favorable pathways for reactions

enable reactions to occur under industrially achievable conditions

allow selective production routes without or with less undesirable byproducts

are the work horses of the chemical industry

are the key enablers for sustainable (green)production

supported catalyst

catalystpellets and extrudates

CourtesyHaldor Topsoe

catalyticsurface

catalytically active particles on a support

shaped catalyst particles

catalyst bed in a reactor

1 nm

10 mm

1 µm

1 m

microscopic mesoscopic macroscopic

What is an ‘energetically favourable’ reaction path?How to understand a catalytic reaction in terms of potential energies?

non-catalytic energy barrier

energy barrier catalytic route much lower!

The Sabatier Effect

metal - adsorbate bond strength

cata

lytic

act

ivity optimum interaction

catalyst - adsorbate:• not too strong• not too weak

optimum coverageat the rightsurface

Example: CO Oxidation

+

adsorption reaction desorption

CO

O2

CO2

catalyst

What is the most essential thing that the catalyst does ?

a catalyst breaks bonds……...

…...and lets other bonds form







• thermodynamics




Langmuir - Hinshelwood Kinetics

Irving Langmuir1881 - 1957

Nobel Prize 1932

Cyril NormanHinshelwood

1897 - 1967Nobel Prize 1956

1915 Langmuir: Adsorption Isotherm

1927 Hinshelwood:Kinetics of Catalytic Reactions

• Consistent with Sabatier’s Principle

• Coverage dependence: Volcano plot

• Temperature dependence: Volcano plot

Irving Langmuir(1881 - 1957)

• worked at General Electrics• oxygen adsorption on tungsten

filaments of light bulbs• 1932: Nobel Prize in Chemistry• Langmuir Adsorption Isotherm:

0A = KA pA

1 + KA pA

Pressure (bar)0 1 2 3 4 5 6 7 8 9 10

θ co

vera

ge

0.0

0.2

0.4

0.6

0.8

1.0

K=0.1 bar-1

K=1 bar-1

K=10 bar-1

The Langmuir Adsorption Isotherm

0A = KA pA

1 + KA pA

θA = KA pAθ*

Reaction Mechanism:

A + * ⇔ Aads equilibrium; KA

B + * ⇔ Bads equilibrium; KB

Aads + Bads → ABads + * r.d.s; k

ABads → AB + * fast

Coverages:

θA = KA pAθ*

θB = KB pB θ*

θ*

= 1

1+KApA+KBpB

Reaction rate: r=N* k KAKB pApB

(1 + KApA + KBpB)2

θA

θB

rate

θ, norm

aliz

ed

rate

pA/pB

Eads(A) = Eads(B) = 125 kJ/mol

s(B) = s(A) ; Eact = 50 kJ/mol

T = 600 K; pB is fixed

θ*

Rate of a Catalytic Reaction:Pressure Dependence

reaction orderpositive in pAnegative in pB

1.0

0.8

0.6

0.4

0.2

0.0

θAθB

rate

reaction ordernegative in pApositive in pB

0.1 1.0 10

0,0

0,2

0,4

0,6

0,8

1,0

100 300 500 700 900

θ, norm

aliz

ed

rate

T (K)

Eads (A) = 135 kJ/molEads (B) = 125 kJ/mols(B) pB = 10 s(A) pAEact = 50 kJ/mol

θA

θB

θ*rate

Rate of a Catalytic Reaction:Temperature Dependence

reaction ordernegative in pApositive in pB

reaction orderpositive in pA and pB

The Sabatier Effect

metal - adsorbate bond strength

cata

lytic

act

ivity optimum interaction

catalyst - adsorbate:• not too strong• not too weak

optimum coverageat rightsurface

r=N

* k KAKB pApB

(1 + KApA + KBpB)2

Rate of reaction, Activation energy, Order of reaction:

AA

AA

A prp

prn θ21....ln

lnln

−==∂∂

=∂∂

=

( )BBAA

rdsa

appa

HHET

rRTTrRE

∆−+∆−+=

==∂

∂=

∂∂

−=

)21()21(

......ln/1

ln 2

θθ

Kinetics of catalytic reactions:

Sabatier’s Principle: Volcano Plot Methanation

Metal oxide formationper oxygen atom

-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

TOF(

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)

Sabatier’s Principle: Volcano Plot Ammonia Synthesis

Courtesy Jens Nørskov

Catalysis by Metals: Trends in Reactivity

stable againstoxide, carbide, nitride formation

stable oxides, carbides, nitridesstrong, dissociative adsorption

Weak, molecular adsorption

Cr Mn Fe Co Ni Cu

Mo Tc Ru Rh Pd Ag

W Re Os Ir Pt Au

Tren

ds

in c

hem

isorp

tion







• thermodynamics




atom molecule atom

atomic molecular atomicorbital orbitals orbital

antibonding

bonding

much / little overlap

strong weak no bond

The minimum you need to know about . . . . . .

Molecular OrbitalsOverlap:

Filling:

Presenter

Presentation Notes

Figure 6.8 Summary of molecular orbital theory for homonuclear molecules. Note how the stability of a chemical bond depends both on the interaction strength and the filling of the orbitals.

Formation of an electron band by addition of atoms and their orbital.

Note that the splitting between the bonding and anti bonding level increases by increasing the overlap.

Eventually when a high number of orbitals are added a continuum band is formed.

…...

and about . . . . .

Bonding in Metals

the minimum you need to know about . . . . .

Bonding in Metals

sp - band

d-banden

ergy

density of statesatom metal

4p

4s

3dFermi level

Presenter

Presentation Notes

Figure 6.10 Schematic representation of the energy levels of a typical 3d- transition metal. The extended s and p orbitals forms the broad sp-band shown in the panel to the right. The more localized d-orbital lead to a narrow d band.

d-metal adsorbed freeatom atom

Evac

EF

a) b)

d-metal adsorbed freeatom atom

antibonding

bonding

antibonding

bonding

d-band

s-band

Atom on d-metal:Evac

EF

Presenter

Presentation Notes

Figure 6.24 Interaction between an atomic adsorbate with one valence level and a transition metal, which possesses a broad sp-band and a narrow d-bandlocated at the Fermi level. The strong interaction with the d-band causes splitting of the adsorbate level into a bonding and an anti-bonding level. The part of the adsorbate levels below the Fermi level are occupied by electron density.

Cr Mn Fe Co Ni Cu

Mo Tc Ru Rh Pd Ag

W Re Os Ir Pt Au

Strong atomic adsorption

Weaker adsorption

Tren

ds

in c

hem

isorp

tion

d-band < half filledstrong bond

d-band > half filledweaker bond

585 564

543

531 531 C/Metal, eV

σ*

σ

1s 1s

Evac

EF

d-metal free molecule

antibonding

bonding

antibonding

bonding

σ-orbitals σ*-orbitals

Molecular Adsorption on a d-metal

adsorbed molecule

“relieved repulsion”favors on-top adsorptionoften called “donation”

“back donation”binds molecule to surfaceweakens internal CO bond!favors multiple coordination

This

pic

ture

is t

he k

ey t

o un

ders

tand

ing

cata

lysi

s in

ter

ms

of o

rbit

al t

heor

y

2300 2200 2100 2000 1900wave number (cm-1)

abso

rban

ce

CO gas

2143 cm-1

CO/ Ir/SiO2

SiO2

FTIR of CO 1) The infrared spectrum of gas phase CO shows rotational fine structure, which isabsent in the spectrum of CO adsorbed on an Ir/SiO2catalyst

2) The IR absorption frequency of adsorbed CO is lowered, mainly due to electron back donation into its 2π* orbital

IR of Gas Phase and Adsorbed CO

IR spectra by Leo van Gruijthuijsen, TU Eindhoven

a catalyst breaks bonds……...

…...and lets other bonds form

dissociativechemisorption physisorption

freemolecule

E=0

distance from the surface

energ

y

Heat ofphysisorption

Heat of chemisorption2 H atoms

σ*

σ

1s 1s

Evac

EF

d-metal free molecule

antibonding

bonding

antibonding

bonding

σ-orbitals σ*-orbitals

Molecular orbital picture of dissociation:

adsorbed molecule

J.W. Niemantsverdriet, Spectroscopy in Catalysis, Wiley-VCH, Weinheim, 1993 & 2000

causes H-H bond to break

Potential energy and orbitals of CO dissociation

∆Hads (AB)–150 kJ/mol Eact

75 kJ/mol

∆Hads (A+B)–600 kJ/mol

Energetics of Dissociationon a transition metal such as Fe, Ru

DrivingForce

Cr Mn Fe Co Ni Cu

Mo Tc Ru Rh Pd Ag

W Re Os Ir Pt Au

easy dissociation

no dissociation

Tren

ds

in c

hem

isorp

tion

∆Hads (AB) δEact

δ ∆Hads (A+B)

Dissociation on Different Metalse.g. Rh and Fe

δEact ≈ ½δ ∆Hads (A+B)Bronstedt-Polanyi Relation

Rh

Fe

Dissociation of CO on Fe(100)

adsorbed CO

C

Eact = 1.14 eVExp: 110 kJ/mol


transition state

∆H = - 0.34 eV


dissociated CO

∆H = - 0.34 –0.82 eV


dissociated COrepulsion relieved


adsorbed CO


-0.34 eV

1.14 eV(exp 110 kJ/mol)

-1.16 eV

2.30 eV

0.82 eV


dissociated CO

transition state

adsorbed CO

T.C. Bromfield, D. Curulla Ferre, J.W. Niemantsverdriet, ChemPhysChem, 6 (2005) 254

Compare CO dissociation on different metals

Energy Scaling Relations

Freek Scheijen, Dani Curulla, Hans Niemantsverdriet, J. Phys. Chem. C 113 (2009) 11041

Catalysis by Metals: Trends in Reactivity

Cr Mn Fe Co Ni Cu

Mo Tc Ru Rh Pd Ag

W Re Os Ir Pt Au

stable againstoxide, carbide, nitride formation

stable oxides, carbides, nitrides

strong, dissociative adsorption

Weak, molecular adsorption







• thermodynamics




Catalytic Reaction:

a cycle of elementary reaction steps

CO + NO Reaction Mechanism

CO + * → COads

NO + * → NOads

NOads + * → Nads + Oads

CO + Oads → CO2 + 2 *

2 Nads → N2 + 2 *

Can we determine the kinetics of each step?

Catalytic Reaction:

a cycle of elementary reaction steps

CO + NO Reaction on Rh(100) & (111)

only 30 % CO → CO2

CO2 formation slowN2 formation fast

200 300 400 500 600 700 800 900

0.15 ML 13

CO + 0.20 ML NO on Rh(111)

N2O

N2

13CO

NO

13CO2

Des

orp

tion r

ate

(a.u

.)

Temperature (K)

80 % CO → CO2

CO2 formation fastN2 formation slow

200 300 400 500 600 700 800 900

0.20 ML CO + 0.26 ML NO/Rh(100)

Temperature (K)

N2

CO

NO

CO2

Des

orp

tion r

ate

(a.u

.)

200 400 600 800

Temperature (K)

1000

Nads

NOads

TPD

SIMS

TPD+SIMS: NO Dissociation on Rh(100)

Rh(100)


37 ± 3 kJ/mol

N2,gas

NOgas

Nads + Nads → N2,gas + 2*

215 ± 10 kJ/mol

CO + NO Reaction Mechanism

CO + * → COads

NO + * → NOads


CO + Oads → CO2 + 2 *

2 Nads → N2 + 2 *

Can we determine the kinetics of each step?

2.1 2.4 2.7 3.0-6

-4

-2

0

2

1000 / T (K-1)

300 400 500 600

CO 2

form

atio

n ra

te (a

.u.)

Temperature (K)

COads+ Oads

Rh (100)θo = 0.16 MLθco= 0.07 ML

Rate equation:

r = k θO θCO = ν θOθCO e-Eact /RT

Plot:

ln (r/θO θCO) vs 1/T

Eact = 103 ± 5 kJ/mol

ν = 1012.7±0.7 s-1

M.J.P. Hopstaken, W.E. van Gennipand J.W. Niemantsverdriet,

Surface Sci., 433-435 (1999) 69

Kinetics of COads + Oads = CO2ln

r/θ

o θ

co

(s

-1)

dissociation of NOads

desorption of NOads

desorption of COads

reaction COads+Oads = CO2

reaction Nads + Nads = N2

37± 3

106± 10*

139± 3

103± 5

215± 10

1011± 1

1013.5± 1

1014± 0.3

1012.7± 0.7

1015.1± 0.5

65± 6

113± 10*

155± 5

67± 3

118± 10

1011± 1

1013.5± 1

1015± 1

107.3± .2

1010± 1

Eact ν Eact νkJ/mol s-1 kJ/mol s-1

Rh(100) Rh(111)

Kinetic ParametersCO + NO

H.J. Borg, J. Reijerse, R.A. van Santen, and J.W. Niemantsverdriet, J. Chem. Phys. 101 (1994) 10052

M.J.P. Hopstaken and J.W. Niemantsverdriet, J.Phys.Chem. B104 (2000) 3058 & J.Chem.Phys. 113 (2000) 5457







• thermodynamics




Adsorbate –adsorbate interactions

Li Be B C N O F-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.5In

tera

ctio

n en

ergy

C

O –

X (e

V)

CO - Xattraction

CO - Xrepulsion

COXX

XX

COXX

XX

COXX

XX

CO

XX

X X

CO

XX

X X

CO

XX

X X

0.50 ML X0.25 ML CO

0.25 ML X0.25 ML CO

Li Be B C N O F-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

CO - Xattraction

CO - Xrepulsion

Li

B

C

N

O

F

LiBe B C

N

OF

Be

Atom–CO interactions on Rh(100):>0.5 ML

Nearest neighbour interactions >> next nearest neighbour interactionsD.L.S. Nieskens, D. Curulla, J.W. Niemantsverdriet, Chem Phys Chem 7 (2006) 1075

-0.34 eV


-1.16 eV

2.30 eV

0.82 eV

T.C. Bromfield, D. Curulla Ferre, J.W. Niemantsverdriet, ChemPhysChem, 6 (2005) 254


dissociated CO

transition state

adsorbed CO

“Ensemble Requirement” CO Dissociation

DFT CalculationCO on Fe(100)

Ensemble neededfor dissociation

“Ensemble Size” determined by repulsion between the dissociated atoms


0.82 eV

Dani Curulla, Ashriti Governder, Tracy Bromfield, Hans Niemantsverdriet, J. Phys. Chem. B 110 (2006) 13897


dissociated CO

adsorbed CO

transition state

1.29 eV

• sulfur retards CO dissociation• mainly by S-C and S-O repulsion• effect on CO is 0.26 eV• S blocks sites for C and O atoms

Sulfur Poisoning of CO Dissociation DFT CalculationFe(100) – (2x2)S







• thermodynamics




AcknowledgementsThe Fischer-Tropsch Synthesistechnology – mechanisms – catalysts

Hans Niemantsverdriet

Schuit Institute of Catalysis

Eindhoven University of Technology

Crude Oil

Gas

Coal

Biomass

Sun; H2O

Catalysis for Energy

Fuelsgasoline kerosine

diesel

CH3OHdimethyl ether

H2NH3

Energy Sources Catalytic Processes Energy Carriers

The Fischer-Tropsch Synthesis

• Fischer-Tropsch reactions & technology

• Mechanisms: iron and iron carbides

• GTL: cobalt catalysts and their stability

• Nano particle model systems

• Conclusions and outlook

FTSGas

Coal

Biomass

CxHyFuels

gasoline kerosine

Diesel

OygenatesOlefins

CO +H2

synthesisgas

Energy Source Catalytic Conversion Processes Energy Carriers

gasification


GTL = Gas to Liquids (Sasol, Shell)CTL = Coal to Liquids (Sasol)

BTL = Biomass to Liquids

Photos: courtesy of Haldor Topsoe A/S

steam reforming of natural gas

CH4 + H2O = CO + 3H2 Ni on Al2O3 or MgAl2O4

How is synthesis gas produced?

water gas shift reaction

CO + H2O = CO2 + H2 Fe3O4 or Cu/ZnO on Al2O3

Alternatively: coal, oil fractions or biomass

more H2:

Cobalt catalystCO + 2H2 → - CH2- + H2O ΔH = -165 kJ/mol

H2/CO ratio at least 2 (CnH2n+2)

Iron catalystWater-Gas Shift (WGS) reaction:

CO + H2O → CO2 + H2 ΔH = -42 kJ/mol

FTS Catalysts with WGS activity:2CO + H2 → - CH2- + CO2 ΔH = -204 kJ/mol

H2/CO ratio > 0.5

Operating conditionsTemperatures: 200-350°C Pressures: 15-40 bar

Fischer-Tropsch Synthesis:SynFuels from Natural Gas, Coal or Biomass via Syn Gas

Reaction Mechanism of FTS: PolymerizationInitiation and building blocks

COads + * → Cads + Oads

H2 + 2 * → 2 Hads

Oads + 2 Hads → H2O + 3 *

Cads + 2 Hads → CH2ads + 2 *

CH2,ads + Hads → CH3,ads + *

Chain growth and terminationa simple mechanism

CH3,ads + CH2,ads → CH3CH2,ads + *

CH3,ads + Hads → CH4 + 2 *

CH3CH2,ads + CH2,ads → CH3CH2CH2,ads + *

CH3CH2,ads ± Hads → C2H5 ± 1 + 2 *

etc.

1- α

α

α

1- α

α probability of growth1- α termination to

paraffin or olefin

CH3

CH2 CH2

Chaingrowth byinsertion of CH2

into metal-C bond

leads to Flory-Schulzpolymerization kinetics

important implications for selectivity!

Selectivity of FTS Chain growth and termination

CH3,ads + CH2,ads → CH3CH2,ads + *

CH3,ads + Hads → CH4 + 2 *

CH3CH2,ads + CH2,ads → CH3CH2CH2,ads + *

CH3CH2,ads ± Hads → C2H5 ± 1 + 2 *

etc.

1- α

α

α

1- α

CH3

CH2 CH2


into metal-C bond

Anderson-Schulz-Flory Distribution

wn = n (1-α)2 αn-1

C7-C11

≥ C20

≤ C2

C12-C19

C3-C4

Fischer-Tropsch: Polymerization to –(CH2)n–

Presenter

Presentation Notes

New CTL Plants: LTFT at 230 C, have higher chain growth (even than Co) LTFT = Thys Botha Sasolburg 2500 bbl per day HTFT = Tracy Bromfield Fe-Co vergelijking: zie Peter van Berge (low conversion Fe is better, at high conversion Co is better)

Cr Mn Fe Co Ni Cu

Mo Tc Ru Rh Pd Ag

W Re Os Ir Pt Au

CO dissociation = hydrocarbons

no CO dissociation = methanol

Metal Catalysts in CO Hydrogenation

Fe gasoline range, oxygenatesCo diesel and waxes

Ru too expensive and difficult too handleNi methanation

Rh ethanol, C2 oxygenatesPd methanolCu methanol

Franz Fischer; 1918

Courtesy of Prof. Calvin H Bartholomew, Brigham Young University, Utah, USA

Catalysts:

Fischer-Tropsch Synthesis

Coal based

Presenter

Presentation Notes

Workforce numbers (2004)

Since oil crises of the 80s:

Natural Gas to Liquidscobalt catalysts

40%

Cobalt catalystCO + 2H2 → - CH2- + H2O ΔH = -165 kJ/mol

H2/CO ratio at least 2 (CnH2n+2)

Iron catalystWater-Gas Shift (WGS) reaction:

CO + H2O → CO2 + H2 ΔH = -42 kJ/mol

FTS Catalysts with WGS activity:2CO + H2 → - CH2- + CO2 ΔH = -204 kJ/mol

H2/CO ratio > 0.5

Operating conditionsTemperatures: 200-350°C Pressures: 15-40 bar

Fischer-Tropsch Synthesis:SynFuels from Natural Gas, Coal or Biomass via Syn Gas

Removal of heatis a key issue in

FTS process design

lurry Bubble ColumnProf. Krishna - UvA

Fischer-Tropsch Reactor TechnologyFixed Bed versus Slurry Bubble Column

Fixed bed SlurryTemperature/partial pressure gradients

Isothermal, gradient less, well-mixed

High pressure drop Low pressure drop <2 bar

Extended shut down for catalyst removal

On-line catalyst removal/addition

Limited for scaling up Significant scope for scaling up (20 000 bbl/day)

Shell Sasol

Gas-To-Liquids process (GTL)

80% Sasol SPD™ DieselMost important productHigh performance fuel

Low emissionsEnvironmentally friendly

20% Sasol SPD™ NaphthaLow octane number

Mostly alkanesExcellent feedstock for

chemicals

Copyright reserved 2007, Sasol Technology R&D

oxygen

natural gassteam

Synthesis gas

Sasol slurry phase distillate process™

hydrocarbons

Sasol-Qatar Petroleum Oryx GTL plant


O2

Gasheater

ATR

FTS

Sasol-Qatar Petroleum Oryx GTL plant


34000 bbl/day







www.fischer-tropsch.org

Bulletin 580Physical Chemistry of the Fischer-Tropsch Synthesis

R.B. AndersonJ.F. SchultzL.J.E. HoferH.H. Storch

1959

www.fischer-tropsch.org

Mechanism based on oxygenates

etc.

Mechanistic developments FTS

FTS via dissociated CO1980s: Ponec, Biloen-Sachtler, Rabo,

>2005: H-assisted CO dissociation (King, sev

Several detailed mechanisms proposals involving Biloen-Sachtler, Gaube, Schulz, Maitliss, Brad

Most popular mechanism: but also o intermediates considered

(Pichler-Sc

CH3

CH2 CH2


into metal-C bond

CH3

CH2 CO

Chaingrowth byinsertion of CO

into metal-C bond

MossbauerSpectra:

mostlyFe5C2

Iron Converts into Carbides during FTS !Is FTS activity related to carbide?

G.B. Raupp and W.N. Delgass

J.B. Butt and L.H. Schwartz

P. Bussiere, G. leCaer, et al.

J.W. Niemantsverdriet,A.M. van der Kraan

W.L. van DijkH.S. van der Baan

J.Phys.Chem. 84 (1980) 3363

Behavior iron in FTS:FT

S A

ctiv

ity

time on stream

fast consumption of C from dissociated CO

by iron interior, (slows down rapidly)

slow build up ofsurface carbon

causes deacivation

J.W. Niemantsverdriet & A.M. van der Kraan, J. Catal. 72 (1981) 385

Questions: Are carbides essential for FTS on Fe catalysts?What is the FTS mechanism on iron carbide?

Aim: FTS Mechanism on Iron Carbide Surface

Questions:

How does CO adsorb and dissociate on a carbon containing surface?

How do hydrocarbons form on iron carbide surfaces?

Method: Computational Chemistry, DFT and UBI-QEPwith Dani Curulla and Jose Gracia

DFT = Density Functional Theory

UBI-QEP = unity bond index – quadratic exponential potentialE. Shustorovich, H. Sellers, Surf. Sci. Rep. 31 (1998) 1

DFT CO Dissociation on Fe(100)

-0.34 eV


-1.16 eV

2.30 eV

0.82 eV

Tracy Bromfield, Dani Curulla, Hans Niemantsverdriet, ChemPhysChem, 6 (2005) 254


dissociated CO

transition state

adsorbed CO

size of the catalytic ensemble

for dissociationdetermined by

the atoms!!

Ener

gy

(eV)

-8,28 eV

-6,82 eVtransition state

-5,34 eVtransition state-6,54 eV

-6,02 eV

-5,22 eV

-2,59 eV -2,55 eV-2,03 eV

transition stateH

O

C

C, O, H - Atoms on Fe(100)

Atoms prefer high coordination; Mobility: H > O >> CNote that diffusion into the lattice has been excluded

activation energydiffusion ≈ 1.5 eV

Water formation on Fe(100)(with reference to water in the gas phase and atomic oxygen adsorbed on the slab)

Two reaction pathways:

OH + OH more likely than OH + HAshriti Govender, Dani Curulla, Hans Niemantsverdriet, 2007

-1.0

0.0

1.0

2.0

CCH

CH2

CH3 CH4

Ener

gy

(eV)

Top

Top

Top

Top

Bridge BridgeBridge

Bridge

HollowHollow

Hollow

Non specific

H addition: stabilizes top/bridge, destabilizes hollowmobility increases

CH3 is a highly mobile species

CHx intermediates on Fe(100)

Presenter

Presentation Notes

Top becomes more stabilised by addition of hydrogen. However, only the CH2 at the top site is a local minima. Bridge is largely unaffected – CH2 and CH3 are local minima. Hollow becomes less stable as C is hydrogenated. C, CH,CH2 are most stable adsorption sites with all real frequencies – more stable than CH4 in the gas phase and adsorbed on the surface. CH3 does not exist in the hollow site. C and CH are predominant. Only 3 are more stable than CH4 adsorbed on the surface.

Surface Reaction C + H = CH

Reaction Coordinate

Ener

gy

(eV)

1) Endothermic reaction

2) Activation barrier corresponds

roughly to a reaction temperature

of ~300 K0.74 eV

0.44 eV

fill emptysite with H

0.42 eV

or exothermic reaction!

Presenter

Presentation Notes

Include activation energy of trans step. But TS is second order saddle point. So consider diffusion of H to cis hollow – Eact very small. Get a true transition state from cis config. So that’s more reactive site. True for all of these hydrogenations. Most stable site was not most reactive.

-1.5

-1.0

-0.5

0.0

0.5

1.0

C+4H

(C+H)+3H

Transition state

CH+3H

(CH+H)+2H

Transition state

CH2+2H

(CH2+H)+H

Transition state

CH3+H

(CH3+H) Transition state

CH4(a) CH4(g)

Ener

gy (e

V)

∆Er = +0.31 eV

0.71 eV

∆Er = +0.50 eV

0.81 eV

∆Er = +0.19 eV

0.77 eV

∆Er = -0.23 eV

1.5

(C+H) +3H

(CH+H) +2H

(CH2+H) +H(CH3+H)

0.75 eV

0.44 eV

0.21 eV

0.62 eV1.00 eV

Transition state I

Transition state III

Transition state II

CH4(a)

Transition state IV

Methane formation on Fe(100) from C + 4 H(with reference to methane in the gas phase and the clean slab)

In agreement with DC Sorescu, Phys Rev B 73 (2006)155420

-2.0

-1.5

-1.0

-0.5

0.0

0.5

C+4H

(C+H)+3H

CH+4H

(CH+H)+3H

CH2+4H

(CH2+H)+3H

CH3+4H

(CH3+H)+3H

CH4(a)+4H

CH4(g)+4H

Ener

gy (e

V)

∆H = -0.11 eV

0.71 eV

∆H = +0.08 eV

0.81 eV

∆H = -0.23 eV

0.77 eV

∆H = -0.65 eV

(C+H) +3H (CH+H) +2H (CH2+H) +H (CH3+H)

0.75 eV0.86 eV 0.63 eV 1.04 eV

1.42 eV

Transition state I Transition state IIITransition state II Transition state IV

Methane formation on Fe(100) from C + H(with reference to methane in the gas phase and the clean slab)

Ashriti Govender, Dani Curulla, Hans Niemantsverdriet, 2008

Example of a C-C coupling reaction stepCH3+CH2 CH3CH2

CH2+CH3

TS[CH2+CH3]

CH2CH3

CH2CH3+H

TS[CH2CH3+H]

CH3CH3

TS[CH2CH2+H]

CH2CH2

-1.00

-0.50

0.00

0.50

1.00

1.50

Ene

rgy

(eV

)

0.72 0.75

0.67

0.87 1.17

0.59

0.10

… several other mechanisms with similar barriersare conceivable

Ethylene formation pathways on Fe(100)

CHCHCCH2

CHCH+H

[CHCH+H]

CHCH2CHCH2+H

[CHCH2+H]

CHCH3

CCH2+H

[CCH2+H]

CCH3 CCH3+H

[CCH3+H][CH+CH3]

[H+CCH2]CH2+CH2

[CH2+CH2]

CH2CH2

CH+CH3CH+CH2

[CH+CH2][H+CHCH2]

C+CH3

[C+CH3]

-2.00

-1.50

-1.00

-0.50

0.00

0.50

1.00

1.50

Ener

gy (e

V)

CCH2

CHCH

CHCH+H

CH+CH2

CCH2+H

C+CH3

TS[CHCH+H]

TS[CH+CH2]

TS[CCH2+H]

TS[H+CCH2]

CCH3

CHCH2

CH2CH2(g)

CHCH2+H

CH+CH3

CCH3+H

TS[H+CHCH2]

TS[CHCH2+H]

TS[CCH3+H]

TS[CH+CH3]

CH2CH2

CHCH3

0.76

0.69

0.43

1.27

0.48

1.55

0.57

0.55

0.72

0.78

0.10

-0.11

-0.08

Transition states

Transition states

TS[C+CH3]

CH2+CH2

TS[CH2+CH2]

DFT activation energies (eV) on Fe(100)

• FTS on Fe(100): several pathways possible• reactions involving CH3 are almost always favored

• but…. all surface reactions to hydrocarbons are endothermic!• and… iron is a carbide in FTS ………

C CH CH2 CH3 H

C 1.97 1.37 0.68 0.48 0.75

CH 1.37 1.14 1.27 0.57 0.71

CH2 0.68 1.27 1.55 0.72 0.81

CH3 0.48 0.57 0.72 high 0.77

in good agreement with DC Sorescu, Phys Rev B 73 (2006)155420JMH Lo, T Ziegler, J Phys. Chem. C 111 (2007) 13149

J Cheng, P Hu, P Ellis, S French, G Kelly, CM Lok, J Phys Chem C 112 (2008) 6082

FTS mechanism on Fe5C2(100)Jose Gracia, Frans Prinsloo, Hans Niemantsverdriet, TU/e + Sasol, 2009

P. J. Steynberg, J. A. van den Berg, W. J. van Rensburg, J. Phys.: Condens. Matter. 20 (2008) 064238.

top view side view‘4-fold hollow sites’ for Cc-atoms

carbide lattice

‘cartoon’:

CO adsorbs on Fe5C2(100)…Ccarbide =C=O Ccarbide + C=O (TOP)

a) C carbide =C + O (BRIDGE) b) C carbide =C + O (BRIDGE)

Eads = -0.51 eV Eads = -1.29 eV

or

but dissociation is not feasible

ΔH = +2.4 eV endothermic

in agreement with D.-B. Cao, F.-Q. Zhang, Y.-W. Li, J. Wang, H. Jiao, J. Phys. Chem. B 109 (2005) 10922

Hydrogenation Ccarb - exchange with CO

However, • formation of CH2 is endothermic (+0.33 eV)• reaction on to CH3 is favorable (-0.53 eV)

• hence: CH2 coverage low, coupling unlikely

C carbide H2 (4 - fold) + CO (top) CO (4 - fold) + C carbide H2 (top)

Hydrogenated Ccarb exchanges with CO

• formation CH3 is exothermic (-0.53 eV):→ CH3 abundantly present on the surface

• exchange CH3 - CO also exothermic (-0.63 eV)with 0.2 eV activation energy only

• places CO in a 4-fold site, where it may dissociate

C carbide H3 (bridge- ) + CcarbCO CO (4 - fold) + C carbide H3 (bridge)

Catalytic Cycle Methanation on Carbide

Similar to ‘Mars - van Krevelen’ mechanism

+CH4

in a simplified cartoon

H-assisted CO dissociation



+CH4





+CH4





+CH4





+CH4



Jose Gracia, Frans Prinsloo, Hans Niemantsverdriet, 2009

DFT – UBI-QEP Methanation on Fe5C2

ENERGY PROFILE ON FE5C2 (100)– 0.05 FOR THE REACTION:

CO + 3H2 CH4 + H2O

-3.5

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

CC-H

CC-H3

H2O*

1H2

CC-H3 + CO(Fe-4- fold)

H2 O formation

C-OH bondactivation

eV

CO + 3 H2 CH4 + H2O

direct C-O bond activation

CH4

CO4-fold

COH

Cc +OH

½H2

½H2

½H2

½H2

H2O

all barriers below 0.7 eV


0.5 eV

0.65 eV

DFT – UBI-QEP: H-assisted CO dissociation

Direct CO dissociation: Eact = 1.4 eV

Towards a FTS mechanism on iron carbides Fe5C2(100)

Conclusion

Feasible and plausible mechanism for methane formation on iron carbide

• Direct dissociation of CO on carbide difficult

• Ccarb from hollow site becomes CH3

• Empty hollow site available for CO

• H-assisted dissociation of CO = Ccarb + OH

Mars –van Krevelen likereaction cycle

Energy Profile Methanation Reaction on Group VIII and IB metals

G Jones, T Bligaard, F Abild-Pedersen, JK Norskov, J.Phys: Condens. Matter 20 (2008) 064239

Ag

Au

Cu

Pd, Pt

RuFe

W

Prediction:Ni, Rh, Co

can domethanation

as clean metal

Ni, RhCo

Ni, RhCo

Energy Profile Methanation Reaction on Group VIII and IB metals

G Jones, T Bligaard, F Abild-Pedersen, JK Norskov, J.Phys: Condens. Matter 20 (2008) 064239

Ag

Au

Cu

Pd, Pt

RuFe

W

Prediction:Ni, Rh, Co

can domethanation

as clean metal

but Iron worksonly as carbide

FTS Mechanism over Iron Carbide: longer hydrocarbons








Co/Al2O3 catalyst

20wt.% Co;

0.04 Pt

prepared by

slurry impregnation,

calcined at 250 oC

reduced at 425 oC


0

0.05

0.1

0.15

0.2

0.25

0.3

0 5 10 15 20Co size (nm)

Nor

mal

ized

pop

ulat

ion

TEM size distribution

TEM, XRD, Hydrogen Chemisorption: Co ∼ 6nm

G.L. Bezemer, J.H. Bitter, H.P.C.E. Kuipers, H. Oosterbeek, J.E. Holewijn,X. Xu, F. Kapteijn, A.J. van Dillen and K.P. de Jong,

J. Am. Chem. Soc. 128 (2006) 3956

Particle Size Dependence FTS Cobalt on Carbon Nano Fibres - 35 bar, 210 °C

C5+ selectivityTOF

Optimum: Co particles of 6-8 nm

Dissociation easier on steps

K. Honkala, J. Norskov et al. Science 307, 555 - 558 (2005)

Diameter approximately 6 nm:Smallest size that supports steps

needed to dissociate CO

Long term catalyst performance testingunder realistic Fischer-Tropsch synthesis

100 bbl/day slurry bubble column reactor, 230 °C, 20 bar, (H2+CO) conversion: 50-70 %,

feed gas: 50 vol. % H2, 25 vol. % CO, PH2O = 4-6 bar)

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60Time on line (days)

RIAF

rela

tive

activ

ity

0 10 20 30 40 50 60 Time on line (days)

Cobalt is expensive; need to maximize catalyst life

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60Time on line (days)

RIA

F

Co/Al2O3

Deactivation mechanisms:

(postulated)

• Oxidation

• Poisoning (S, HCN,NH3)

• Sintering

• Carbon deposition

Long term catalyst performance testing

Time on stream (days)

Nor

mal

ized

act

ivity

What causes the deactivation of Co FTS catalysts?

100 bbl/day slurry bubble column reactor, 230°C, 20 bar, H2+CO conversion: 50-70 %, feed gas: 50 vol. % H2, 25 vol. % CO, PH2O = 4-6 bar

free atom

atoms in a lattice

Ek = hν - Eb

hν

hν

abso

rption a

bso

rption

hν

hν

Eb

Eb

EXAFS

XANES

edgepreedge

EXAFS and XANES

/Al2O3

53% Co0

80% Co0

85%

88%

89%

XANES of Cobalt Phases

LURE, ORSAY LURE, ORSAY

XANES:

• phase identification

• oxidation state

• in situ measurement

• at synchrotron

• quantitation

straightforward

/Al2O3

53% Co0

80% Co0

85%

88%

89%

XANES of wax coated/protected catalystsfrom FT demonstration reactor

A.M. Saib, A. Borgna, J. van de Loosdrecht, P.J. van Berge, J.W. Niemantsverdriet Appl. Catal. A: General 312 (2006) 12

No oxidation of Co> 6 nm, instead reduction of unreduced cobalt.

LURE, ORSAY LURE, ORSAY

AFM Co/SiO2/Si(100) Model Catalysts

Abdool Saib, Armando Borgna, Jan van de Loosdrecht, Peter van Berge, Hans Niemantsverdriet,

J. Phys. Chem. B 110 (2006) 8657

XANES of Co/SiO2/Si(100) Oxidation

Question:Can cobalt FTS catalysts

oxidize under FTS?

Conclusion:Co/SiO2

highly resistant to oxidation by water

Abdool Saib, Armando Borgna, Jan van de Loosdrecht, Peter van Berge, Hans Niemantsverdriet,

J. Phys. Chem. B 110 (2006) 8657

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60Time on line (days)

RIA

F

Co/Al2O3

Deactivation mechanisms:

(postulated)

• Oxidation

• Poisoning (S, HCN,NH3)

• Sintering




Nor

mal

ized

act

ivity



Deactivation mechanisms: poisoning

Sulphur: irreversible effect

0 10 20 30

Time (days)

Act

ivity

(a.u

.)

Sulphur in syngas

J. van de Loosdrecht, M.M. Hauman, D.J. Moodley, S.D. Nthute, A.M. Saib, B.H. Sigwebela, Sasol Technology (Pty) Ltd, South Africa, presented at ACS Philadelphia, 2008.

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60Time on line (days)

RIA

F

Co/Al2O3Deactivation mechanisms:

(postulated)

• Oxidation

• Poisoning (S, HCN,NH3) V• Sintering




Nor

mal

ized

act

ivity



fresh catalyst after 3 days after 14 days

HAADF-TEM of CoPt/Al2O3

Sintering Cobalt Catalyst in FTSM.J. Overett, B. Breedt, E. du Plessis, W. Erasmus, J. van de Loosdrecht,

Prepr. Pap.-Am. Chem. Soc., Div. Petr. Chem. 2008, 53(2), 126

Sintering Cobalt Catalyst in FTSM.J. Overett, B. Breedt, E. du Plessis, W. Erasmus, J. van de Loosdrecht,

Prepr. Pap.-Am. Chem. Soc., Div. Petr. Chem. 2008, 53(2), 126

35% surface area loss due to sinteringin the first 20 days

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60Time on line (days)

RIA

F


(postulated)

• Oxidation

• Poisoning (S, HCN,NH3) V

• Sintering V• Carbon deposition



Nor

mal

ized

act

ivity



CATALYST COVERED IN WAX…

CATALYST POWDER

Denzil Moodley

1 2 3

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60Time on line (days)

RIA

F


(postulated)

• Oxidation

• Poisoning (S, HCN,NH3) V

• Sintering V• Carbon deposition V



Nor

mal

ized

act

ivity



Co/Al2O3 FTS catalyst

• Cobalt metal is active phase

• Contains reduction promotor (Pt)

• Particles in the 6-10 nm size range

• Catalyst deactivates due to a number of factors

• Sintering

• Poisons

• Inactive carbon formation (at a very slow rate)

• Oxidation is not a factor in deactivation







Synthesis of FeO (wustite) particles

0

10

20

30

40

50

60

31 32 34 36 39

Diameter (nm)

Part

icle

s (

%)

Synthesis from thermal decomposition of iron carboxylate salts (22nm)

0

5

10

15

20

25

30

18 20 21 22 23 24 26-29

Diameter (nm)

Par

ticle

s (%

)

Synthesis of 12nm FeOx - T Hyeon

05

1015202530354045

13-14 15-16 17-18 19-20Diameter (nm)

% P

artic

les

Synthesis of 5nm particles -T.Hyeon

0

510

15

20

2530

35

40

6 7 8 9 10 11Diameter (nm)

% P

artic

les

Seed mediated growth after 110 min. reaction time

0

10

20

30

40

50

60

6 7 8

Diameter (nm)

Par

ticle

s (%

)

Particle growth after 30 minutes

0

10

20

30

40

50

2 3 4 5 6 7

Diameter (nm)

Part

icle

s (

%)

Particle growth after 10 minutes

0

10

20

30

40

50

60

2 3 4 5

Diameter (nm)

Par

ticle

s(%

)

4 nm

± 15%

5 nm

± 17%

7 nm

± 8%

9 nm

± 11%

iron oxide nano particlesseveral diameters - with narrow size distribution

Prabashini Moodley, Freek Scheijen, Peter Thüne

16 nm± 9%

22 nm± 7%

35 nm± 6%

Variety of preparations

iron oxide; cobalt oxide

silicon wafer

SiO2 / SiNx / SiO2membrane15 nm thick

Electronbeam-

Silicon wafer

9 TEM grids

TEM Grid: Support for Model Catalyst

Presenter

Presentation Notes

Side view of a silica TEM support showing the transmission window. This window is a 15 nm thick silicon nitride membrane that has been calcined in air to produce a surface oxide layer.

Particles deposited by spin coating

Spin Coater

TEMgrid

Procedure:1. Add 600 µl iron particle suspension 2. Rotate at 2800 RPM for ca. 5 sec3. Remove disk; place in oven 450°C

4. Surfactant burns off5. Iron oxide binds to silica surface

Thermal stability iron oxide particles (oxidation 500˚C)

A

B C

E

DA

BC

E

D

9.5 nm particles

Inter-particle distance (spincoated)A1-A2 = 39.2nmB1-B2 = 14.4nmC1-C2 = 12.7nmE1-E2 = 12.8nmD1-D2 = 13.0nm

Inter particle distance (calcined)A1-A2 = 39.2nmB1-B2 = 12.3nmC1-C2 = 12.3nmE1-E2 = 11.7nmD1-D2 = 11.1nm

spin coated calcined

Super-imposed: spin coated and calcined particles

spin coated calcined

Part. SC (nm) Calc (nm)A1 10.3 10.5A2 11.6 12.5B1 10.0 10.3B2 13.2 14.2C1 9.0 9.8C2 10.3 11.2D1 10.6 11.4D2 9.0 10.3E1 10.6 11.4E2 10.0 10.6

A1 A2D1

D2

B1

B2

C1C2

E1 E2spin coatedcalcined

Super-imposed: spin coated and calcined particles

Diameter increases by about 10% upon calcination

Particle pair after spin coatingCenter 1 - Center 2 = 13nm

Particle pair A after calcinationCenter 1 – Center 2 = 11nm

• particles make contact after calcination• no major rearrangements during calcination

• no sintering

Thermal Stability of Iron Oxide Particles

1

2

1

2

During heating to 500 ˚Cthe particles flatten out

(diameter increases by ~10%)

silica support silica support

If the particles are close enough they will make contact

Tentative explanation how particles come together

(a) (b)20 nm

Reduced in H2 at 700 ºC for 45 min

Rearrangements upon reduction Calcined at 500 ºC

9.5 nm iron oxide particles core shell morphologycore sizes from 4 - 20 nm

Prabashini Moodley, TU/e

Presenter

Presentation Notes

TEM images (using the silica TEM grid) of (a) calcined iron oxide particles; (b) reduced particles with the inset showing more clearly the core shell structure.

Reduction in H2 – passivation in air320 °C 500 °C 700 °C 800 °C

No reduction Appearance of hollow particles

Some particles unaffected

Partial reduction

Complete reduction

Sintering

Appearance of small particles

Appearance of core shell particles

Reduction incomplete

reduction temperature

Freek Scheijen, TU/e

TEM on the same particles after treatments

28 nm iron oxide particles on SiO2/Si(100)impregnated calcined & reduced after syngas 270ºC

20 nm

20 nm10 nm

Rearrangements: Tentative Explanation

key ingredient: iron diffusion during oxidation

fresh calcined reduced and reoxidized

core and shell hollow donut like

silicon

SiO2

We acknowledge valuable discussions with Prof Abhaya Datye on particle rearrangements







Crude Oil

Gas

Coal

Biomass

Sun; H2O

Catalysis for Energy

Fuelsgasoline kerosine

diesel

CH3OHdimethyl ether

H2NH3

Energy Sources Catalytic Processes Energy Carriers


• old technology – many new opportunities (GTL, CTL, BTL, SNG)

• mechanistically: metals are too reactive

• many endothermic surface reactions

• iron carbide: energetically favorable MvK mechanism

• Fe FTS catalyst: self assembling system

• other metals…?

• cobalt catalyst:

• stability is the major challenge

• sintering, poisoning, carbon deposition (?) contribute to deactivation

• oxidation is not a factor in deactivation

• nano particle models reveal massive rearrangement in FTS

ammonia

synthesis

Langmuir

HinshelwoodIR

Surface

ScienceComputational

chemistryBerzelius

Equilibrium

Thermodynamics

1800 1900 2000

TST

Know

ledge

Fundamental & Applied CatalysisHow much do we know?

prac

tical cat

alys

is

unde

rsta

nding

Heterogeneous Catalysis:

• having the right species

• with the right coverages

• at the right temperature

• on the right surface

Acknowledgements

Cobalt• Abdool Saib (TU/e - Sasol)

• Denzil Moodley (TU/e – Sasol)• Kees-Jan Weststrate (TU/e – Sasol))• Armando Borgna (ICES – Singapore)

• Jan van de Loosdrecht (Sasol)• Peter van Berge (Sasol)

• Abhaya Datye (Univ New Mexico)• Tiny Verhoeven (TU/e)

Iron and Iron CarbidesAshriti Govender (TU/e – Sasol)

Prabashini Moodley (TU/e – Sasol)Dani Curulla (TU/e)

Tracy Bromfield (Sasol)Freek Scheijen (TU/e)

Peter Thüne (TU/e)Frans Prinsloo (Sasol)

Jose Gracia (TU/e – Sasol)

National Computer FacilityThe Netherlands

Documents

3 +CH Fundamentals of CH Heterogeneous Catalysis · Heterogeneous Catalysis a molecular view of catalysis by metals Hans Niemantsverdriet Schuit Institute of Catalysis . Contents