Catalyst Design Strategies for Chemicals and

Fuels

Mark A. Barteau

Director, University of Michigan Energy Institute

DTE Energy Professor of Advanced Energy Research

Department of Chemical Engineering

Department of Chemistry

barteau@umich.edu

energy.umich.edu

What We Do, Where We Focus

• Carbon-free Energy Sources

• Energy Storage and Utilization

• Transportation Systems and Fuels

• Policy, Economics and Societal Impact

We build on the University of Michigan’s strong energy research heritage at the heart of the nation’s automotive and manufacturing industries to develop and integrate science, technology and policy solutions to the world’s pressing energy challenges.

More than 140 faculty affiliates across the University in the sciences, engineering, policy, economics and social research

Energy Institute Reports

pdfs can be downloaded at energy.umich.edu

Where are the greatest needs and opportunities related toCatalysis?

1. Fuels and chemicals- “Similar” feedstocks, processes and product transmission

- Fuels production >> Chemicals production, but are there lessons from the latter? (Especially for new resources such as biomass and shale gas)

2. Carbon-free hydrogen

3. Cross-over of other renewables from “stationary” to “transport”, i.e., from power to fuels/chemicals

Oil

Gas

Coal

Biomass

Hydrocarbon fuels

Organic Chemicals

CO2

electrons photons

H2 CO

H2O

• New catalytic upgrading and conversion processes• Co-processing of carbon-based resources – The “Omnivorous Refinery”• Integration of carbon-free resources

Understanding mechanisms and dynamics of catalyzed transformationsConnect catalytic and photocatalytic reaction rates and selectivities to the kinetics, energetics and dynamics of individual elementary steps and relate these to the structure and dynamics of the catalytic sites involved.

Design and controlled synthesis of catalytic structuresUse theory as a predictive design tool, develop systematic approaches to construct and to characterize at the atomic level the molecules and materials designed, and develop the necessary understanding to control or direct chemical reactions in complex media

“Basic Research Needs: Catalysis for Energy”

Office of Basic Energy Sciences, US Department of Energy (2008)

Grand Challenges:

• Model

• Make

• Maintain

The 3Ms - with apologies to

Case Studies1.Ketene synthesis catalyst (from surface science experiments)2.Ethylene epoxidation catalyst (from computational predictions)3.“Flying carpet catalysts” (by assembling known catalytic functions)

Representative Metal Oxide Surfaces

ZnO

MgO(100)

Rutile TiO2(110)

RCH2COOH

RCHO

Reduced TiO2 Surfaces

ZnO(0001)

CO2 + hydrocarbons

R

HC=C=O

TiO2(001)–{011} facets; MgO(100)

TiO2(001)–{114} facets

RCH2CCH2R||

O

Summary of surface science results for reactions of carboxylic acids on metal oxide single crystals

RCH2COOH

RCHO

Reduced Surfaces(TiOx)

Reducible Surfaces(ZnO)

CO2 + hydrocarbons

R

HC=C=O

Surface cations with single coordination vacancy(TiO2(001)–{011} facets; MgO(100))

Surface cations with two coordination vacancies(TiO2(001)–{114} facets)

RCH2CCH2R||

O

Critical properties of surface sites on oxides•Coordinative unsaturation of surface cations•Oxidation state, extent of surface reduction•Redox properties

Design Principles for Metal Oxide Catalyst for Ketene Synthesis:

1. Surface cations must be coordinatively unsaturated in order to dissociate the acid.

2. Unimolecular decomposition of adsorbed carboxylates to form ketenes occurs at cations with a single coordination vacancy.

3. Ketene production is favored on non-reducible oxides.

4. The catalyst must be insoluble in carboxylic acids.

Ketene Production with silica catalysts

• Ketenes can be produced catalytically by dehydration of carboxylic acids on silica

T = 700KConversion = 40 to 85% (increases with

carbon number)Selectivity = 35 to 50% (linear acids)

80% (isobutyric acid)

• Activity tracks population of isolated hydroxyl groups on the surface (from IR and gravimetry)

Transmission Infrared SpectroscopyHydroxyl Peak Area (arbitrary units)

Hydroxyl Population (OH/nm2)from Thermogravimetry

Acetate P

opulation (acetate/nm2

)from

Therm

ogravimetry

Transm

ission Infrared Spectroscopy

Acetate P

eak Height (arbitrary units)

First Generation Catalyst:

High surface area silica with controlled hydroxyl

populations (US 5,475,144)

Proposed Mechanism:

CH3COOH + OH (ad) CH3COO (ad) + H2O

CH3COO (ad) H2C=C=O + OH (ad)

Surface hydroxyls serve as active sites by providing “place holders” for coordination of individual carboxylates

Series and Parallel Reactions in Ketene Synthesis

CH4 + CO2

CH3COOH H2C=C=O CH3COOH

½ CH4 + ½ C + CO

½ C2H4 + CO

(CH3)2C=O + CO2 + H2O

+ CH3 COOH

- H2O + H2O

Short Contact Time Reactions over Monoliths

• Negligible pressure drop• Nearly isothermal (400–600°C)• Short contact times (20 milliseconds and up)• Derivatization with high surface area SiO2 can be done

in liquid or slurry processes

0

0.2

0.4

0.6

0.8

1

Acetic Acid Isobutyric Acid

Silica Powder

MonolithCatalyst

Se

lect

ivity

to K

ete

ne

an

d D

MK

Performance of Functionalized Silica Monoliths for Ketene

Synthesis

Surface Science Experiments

Reaction Studies on High Surface Area Catalyst

Spectroscopic Studies of Active Sites

Reactor/Catalyst Configuration Optimization

Prototype Process Schemes

Mechanism and Site Requirementsof Carboxylic Acid Conversion to Ketenes

Materials Selection

Prototype Catalyst

Catalyst Optimization, Structure-Function Studies

First Generation Catalyst US 5,475,144

Functionalized Monolith Catalyst US 6,232,504

Direct Utilization of Ketenes for Product Synthesis

Olefin Epoxidation with Silver Catalysts

12

CH2 CH2

O

Ethylene

C2H4 + O2

Ethylene oxide (EO)

Catalyst: Silver (large particles – up to 1 )

Supported on: -Al2O3 (low surface area < 1m2/g)

Promoted with: Alkalis (K, Cs, Rb)

Halides (ppm of C2H4Cl2 added to feed)

Reactor: 8000 packed tubes, 1˝ ID 30´ long

Hypothesis

Catalytic epoxidation of olefins on silver involves surface oxametallacycles

C

C

O

Et (ads)+ O (ads)

TS1

EO

TS2

17 (measured)16 (calc.)

32.3a

14.9

5

Oxametallacycle

24

Reaction Coordinate

Energy

O2

17.3a

2 O (ads)

Reaction Coordinate, Kinetics and Selectivity

Is this triangular network necessary or correct?

Ethylene Oxide

Acetaldehyde

k1

k2

CO2

ΔG‡1H – ΔG‡

2H = 0.3 kcal/mol (500 K)

ΔG‡2D – ΔG‡

2H = 0.8 kcal/mol (420 K)

Reactant

C2H4

C2H4 C2D4

Predicted Measured

Kinetic Isotope Effect and Mechanism

0.42 0.30-0.50

up by 73% up by 74%

Selectivity to EO:

S. Linic and M. A. Barteau, JACS,125 (2003) 4034.

-1

-0.5

0

0.5

1

1.5

2

Cu/Ag Pd/Ag Au/Ag

Material

TS_AA

Rational Design

EO

AA

TS_EO

CO2Oxametallacycle

{Ea(AA)-Ea(EO)}(alloy) - {Ea(AA)-Ea(EO)}(Ag)

(kcal/mol)

Pure Ag

Photograph of -Al2O3 foam monolith catalyst

Catalyst Preparation

10-14% Ag on -Al2O3

Surface area = 0.2 m2/g Void Fraction of 0.8 Negligible Pressure Drop

Catalysts are comprised of an α-Al2O3 support in monolith form

Catalysts are prepared by wet impregnation a soluble Ag salt

Catalysts are activated via thermal reduction or chemical reduction of the Ag salt

Cu-Ag alloys are prepared by wet impregnation of Cu salt to activated Ag monolith, followed by chemical reduction

0.2% Cu/Ag; 19%Cu surface0.1% Cu/Ag; 9%Cu surfaceAg

0.13

0.27

0.40

0.54

35

40

45

50

55

60

0.13

0.27

0.40

0.54

Feed: 8/1/1 N2/O2/C2H4

1.3 atm, 1.8% conversion

Selectivity on Cu-Ag is up to 1.5 times greater than on pure Ag.

Selectivity advantage of Cu-Ag maintained when promoted with Cs and/or Cl

Selectivity Profiles of Ag and Cu-Ag Catalysts

Bulk XPS

Catalyst Ag AgCu AgPd AgPt AgCd AgAu AgRh AgIr AgZn AgNi AgOs

E1 13.54 16.56 11.28 10.70 23.79 11.37 18.22 19.11 25.47 22.73a 25.83a

ΔH1 7.73 11.97 5.99 4.03 17.47 4.33 15.31 13.78 20.06 17.87 21.30b

E2 15.91 20.53 16.17 14.42 18.80 13.13 19.47 18.59 22.44 23.33 26.59

ΔH2 -22.85 -18.55 -24.82 -27.44 -12.81 -26.36 -16.74 -17.78 -10.24 -12.85 -9.43

E 2.38 3.97 4.88 3.72 -4.99 1.76 1.25 -0.52 -3.03 0.60 0.75

E 0.00 1.60 2.50 1.34 -7.37 -0.62 -1.12 -2.90 -5.41 -1.77 -1.62

Oxametallacycle reaction energetics on Ag14M1 clusters

Full Microkinetic Model vs. Reactor Results

28

Nominal Promoter

concentration (ppm)

Ethylene Conversion

(%)

EO Selectivity

(%)Ag 0 6.4 31.5

Cu-Ag 300 12.2 51.7Pd-Ag 500 14.7 41.8Pt-Ag 100 1.0 0*Cd-Ag 250 31.4 40.0Au-Ag 100 1.4 22.4

*EO below detection limit

All reaction data at 267oC Calcination at 400oC for 12 hoursReduction at 300oC for 12 hours

Ag Cu-Ag Pd-Ag Pt-Ag Au-Ag Cd-Ag0

10

20

30

40

50

60

70

80 Experiment Model

EO

Se

lect

ivity

(%

)

Catalyst Type

100ppm Au-Ag SEM Images

Whenever we observe discrete particles of second metal decorating silver, EDAX signal for second metal, catalyst selectivity is poor:

AuPtCu (high loadings)

Model predictions may fail when the working state of the catalyst is too dissimilar

Polyoxometalate–functionalized, Dispersible

Graphene Catalysts

+ = ?

H3PMo12O40

STM image of H3PMo12O40 array self-assembled on graphite (5.07nm x 5.07nm)

Polyoxometalate-functionalized Graphene Catalysts

• Concept:– POMs are acid catalysts and can also catalyze both low temperature

(liquid phase) and high temperature (gas phase) oxidations– POMs can be dispersed as monolayers on graphite– POM redox properties are tunable

• Objective:– Create POM-functionalized graphene, dispersible in organic solvents,

with controlled and characterizable loading, structure and reactivity

• Benefit:May provide an advantageous approach to dispersing nanoparticle

catalysts compared to other methods (dendrimers, surfactants)

(Nano)catalysts as fuel additivesFuel (Gasoline, Diesel, Jet-A) = hydrocarbons + additives

Additives:

Gas: antiknock agents, antioxidants, detergents, dyes, etc.Jet-A: antioxidant, corrosion inhibitor, antistatic agents, detergents (icing inhibitor), biocides

(Nano)catalysts as additive?

Typically organometallics or metal nanoparticles.Expensive, efficiency is not obvious.

Preliminary results:

Aksay group (Princeton): 175% increase in nitromethane burning rate with 0.075 wt.% functionalized graphene (Vor-X, Vorbeck Materials).

But:

Heptane/air Heptane/air + FGS

1. Defects are necessary - to prevent sheets from re-stacking - improve the POM dispersion - (maybe) to initiate the combustion reaction

2. Graft a fuel-soluble group - commercial additives are known

(often polymers) - instinctively: the groups need to be

chemically similar to the fuel - rationalized by Hansen’s solubility theory

3. Choice: C10 and C20 aliphatic chains - simple, easy characterization

- similar chain length than the fuel fractions of gasoline, Diesel and Jet-A

Making the FGS dispersible in non-polar solvents (fuels)

Grafting of C10 and C20 on FGS

1H NMR

Similar spectra for C10-FGS and C20-FGS.Confirms the grafting of C10 and C20.The integration does not lead to the expected number of each H. Possible grafting of C4.

ATR-FTIR

C-H stretch2954-2852

CH2 bend 1457-1375

Dispersion in water-toluene

Starting FGS are hydrophilic. Grafting C10 chains makes them hydrophobic. The chemical anchoring of the decyl groups was confirmed by NMR, ATR-FTIR and TG-MS.

(a) FGS and (b) C10-FGS dispersions in water-toluene 50:50.(c) C10-FGS dispersion in toluene without sonication.

Toluene

J.-P. Tessonnier and M. A. Barteau, Langmuir 2012, 28, 6691

in DMSO, DMF, CHCl3, ethanol, methanol, water, toluene, heptane and decane (from left to right). The FGS and C10-FGS concentration is 1 mg/ml. The pictures show the redispersion of both materials using thermally driven motion, by simply placing each vial on at hot plate at 70 °C for 5 min.

Dispersibility of FGS (top) and C10-FGS (bottom)

SEM – C10- and C20-FGS

1 m

BET surface area: 415 m /g2

POM deposition

H3PMo12O40

Drying

100 °C POM/Cxx-FGSCxx-FGS

Incipient wetness impregnation

VPOM solution = VFGS pores

Solvent: H2O-EtOH

pHPOM solution < pHIEP

STEM-HAADF of 20 wt.% PMo/C10-FGSUDel058

≈ 1 nm

Combustion TestsFGS

H3PMo12O40/C10-FGS

FGS aggregates form incandescent particles in the flame.

Grafting of alkyl chains significantly improves the dispersion in the fuel. Only 1 incandescent particle observed throughout the tests.

Dispersion in toluene and methylcyclohexane (MCH).

H3PMo12O40/C10-FGS in MCH

Spray Flame Combustion Results

0

5

10

15

20

Fla

me

Sta

nd

off

Hei

gh

t, m

m

45 ml/min fuel25.0 kW flame

Baseline MCH flame (no catalyst)

50 ppmw (C10 or C20 alkyl-grafted FGS) in MCH

50 ppmw (H3PMo12O40/Cxx-FGS) in MCH

50 ppmw (H3PMo11VO40/C20-FGS) in MCH

Potential applications:

• Bi-phasic/phase-transfer catalysis

• Homogeneous catalysis by dispersed nano-particles

• Electrode modification

• Energy storage: battery/supercapacitor

“Flying Carpet Catalysts”

• Each involved new combinations of components• The guiding principles behind these came from

diverse sources: Chemical Reaction Engineering Computational Chemistry Solubility Theory

Case Studies1.Ketene synthesis catalyst (from surface science experiments)2.Ethylene epoxidation catalyst (from computational predictions)3.“Flying carpet catalysts” (by assembling known catalytic functions)