Synthesis of alternative fuels and chemicals from fossil and renewable feedstocks

Andrei Y. Khodakov

UCCS, National Centre for Scientific Research (CNRS),

Université Lille -ENSCL-EC Lille, 59655 Villeneuve d’Ascq, France

University of Lille, ENSCL and EcoleCentrale de Lille

Campus UCCS Pilot Hall

Lille, Grande Place

RealCat HTE centre

Lille

3

SYNGASimportant intermediate can be produced from

all fossil and renewable feedstocks

CO2

CH4

H2

H2SNH3

Major components Minor components Impurities

HCN HCl

Tar

Which feedstocks to produce syngas?

4

5

Energy and Chemical Feedstocks

1G, 2G & 3G Biomasse

Organic and chemical waste

Natural gas

Coal

Oil

Biogas

Energy transition

Unconventional gas

Shale gas and shale oil are hydrocarbons in argillaceous sedimentary rocks, located between 1 and 3 km from depth, which are both compact and waterproof

“New” raw materials: shale gas (USA)

20-30% of gas production in 2015 in the USA compared to only 1% in 2000.

“New” raw materials: coalbed gas: relation between coal and gas

7

Gas extraction by pumping

Gas Concentrations

“New” energy feedstocks: Biogas

8

Constituants % volume

CH4 55-75

CO2 25-45

H2S 0-1.5

NH3 0,05

Biogas composition

Fire-wood

50% of wood production is intended for energy, 50% industrial use

Residue or agricultural waste

Dedicated crops

Crops with short rotation (annual species crops) Perennial undergrowth off wood (miscanthus...)

Waste, the wheat, corn stalks, wood, fibrous biomass (e.g. miscanthus) or macro-algae cultures.

“New” raw materials: 2G biomass, ligno -celullose

3rd generation biofuels differ from the 2nd generation by the type of biomass used. •Microalgae and macroalgae in autotrophic condition (capacity to synthesize organic matters from mineral matter). •In addition to microalgae, 3rd generation include all biofuels which are produced using biomass from water resources

“New” raw materials: 3 GbiomassPhotobioreactors

Open ponds or closed-loop systems

• 17th century first experiments with syngas Thomas Shirley, Dean Clayton

• 1840 First commercially used gasifier in France

• 1850 Streets of London lighted with syngas

• 1878 Gasifiers were successfully used with engines for power generation

• 1901 Passenger vehicle with syngas

• 1940-1945 1 000 000 gasifiers, several 105

cars and tractors

11

History of syngas

Renault AFVH with a Gasifier

Imbert [1941-45]. This tractor

was started with gasoline.

London gas lighting

12

Syngas from conventional and nonconventional gas

Steam reformingendothermicCH4+H2O=CO+3H2

Partial oxidationexothermic 2CH4+O2=2CO+4H2

Autothermic reforming

Dry reformingendothermicCH4+CO2=2CO+2H2

Water gas shift reactionCO+H2O=CO2+H2

Tubular reactor700-1100°C, 3-25 barsNi, Rh, Pt, Pd, Ir, Ru

Non-catalytic PO

Methane,Lighter alkanes

Hydrogen,carbon monoxide,carbon dioxide and impurities

Wabash River Clean Coal Power Plant in USA

Gussing (Austria) biomass gasification plant

CxHyOz=CHs+CO/CO2+CkHlOm

C+H2O=CO+H2

C+CO2=2CO

CO+1/2O2=CO2

H2+1/O2=H2O

CnHm+xO2=nCO2+m/2 H2O

Syngas from biomass and coal gasification

14

Other technologies of syngas production

•Biomass catalytic partial oxidation

•Solar upload (thermochemical cycle)

•Photochemistry

• Combustion, Combined Heat & Power (Gas turbine, electricity generation)

Better than coal!Environmental BenefitsExtremely low SO x, NOx and particulate emissions from burning coal-derived gases.

Carbon dioxide in concentrated gas stream can be captured and sequestered more easily and at lower costs.

Efficiency Benefits35% typical plant efficiency of conventional subcritical pulverized coal (PC) power plant

> 50% fuel efficiency of coal gasification plants

15

Use of Syngas as Fuel

Heavy pollution because of

traditional coal combustion

in China

Puertollano integrated

gasification combined cycle

(IGCC) Plant

Fischer-Tropsch (FT) synthesis is the key step in transformation of syngas into othervaluable products.

16

Biomass

Coal

Natural GasShale GasCoal-bed gas

Syngas

Syngas is an important intermediate for fuels and chemicals

Rh

Cu Fe

Co RuMo

Exothermic reactions

Gas-To-Liquids, Coal-To-Liquids, Biomass –To Liquids : GTL industrial reality

Syngas

H2 + CO

Natural gas (CH4)

Fischer-Tropsch process

Syngas conversion into

hydrocarbons

(LT-FT)

Natural gas Syngas

Production

- Steam reforming

- Partial oxidation

- Autothermal reforming

Final products

-Hydrocracking

-Hydroisomerisation

Pearl (Qatar) - Shell

Oryx (Qatar) – Sasol/QP

140 000 b/j

30 000 b/j

≈ 1100 billon barrels eq.

crude oil

AN INDUSTRIALLY FEASIBLE SUPPORTED COBALT SLURRY PHASE FISCHER-TROPSCH SYNTHESIS (FTS) CATALYST WAS DEVELOPED AND COMMERCIALIZED AT THE ORYX GTL PLANT

S.34

Two slurry phase FTS reactors(Height = 60m; Diameter = 10m)

• Plant was inaugurated on6 June 2006

Copyright reserved 2008, Sasol Technology R&D

Methanol demand Capacity 100 million tons per year

Methanol Synthesis from Syngas

CO+2H2=CH3OH exothermic

CO2+3H2= CH3OH + H2O exothermic

CO+H2O=CO+H2 exothermic

Catalyst: Cu/Zn/Al2O3

T=240-270°C (thermodynamic limitations)P=50-100 bars

3000-5000 t/jICI, Lurgi, Topsoe

Catalysts for syngas conversion: metals, sulfides and carbides

21

Catalyst Products

Fe/FexCy gasoline, olefins, oxygenates

Co diesel and waxes

Ru too expensive and volatile

Ni methanation

Rh ethanol, C2+ oxygenates

Pd methanol

Cu methanol

MoS2/Mo2C alcohols, olefins

Selectivity ?

Selectivity challenge in syngas conversion

From J W (Hans) Niemantsverdriet

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+other oxygenates

Hydrocarbons or oxygenates?

tp

p

vvv+=α

α is the probability of chain growth,

vp and v t propagation and termination rates

ααα

2

)1( −=

n

n

nS Mass fraction of Cn hydrocarbon

Model of linear polymerization (Schulz, Flory 1935-1936) adapted by Anderson

24

The mechanism of polymerizationAnderson-Schulz-Flory(ASF)

Chain growth factor α

H2 + CO → CnH2n+2 (Paraffins)

H2 + CO → CnH2n (Olefins)

H2 + CO → CnH2n+1OH (Oxygenates)

Main reactions:

Video from ICMS, Eindhoven University of TechnologyChallenge: increasing the C2-4

=, C5-11 and C12-C20 selectivities

Broad Anderson -Schulz-Flory (ASF) distribution

Production of water and carbon dioxide in syngas conversion

25

Water gas shift reaction

CO+H2O↔ CO2+H2

CO2 is thermodynamically favored at FT reaction temperatures (200-300°C)

• CO+2H2=-CH2-+ H2O

• 2CO+H2=-CH2-+CO2

Cobalt, rutheniumcatalysts

Iron, molybdenumsulphide catalysts….

Deactivation as a major challenge

� Higher operational cost� Lower catalyst productivity

� Several possible mechanisms

CHALLENGE: DEACTIVATION OF

CATALYSTS FOR SYNGAS CONVERSION

Two-step catalysts deactivation process in demonstration plants for both cobalt and iron based catalysts (van Berge and Everson, 1997)

Catalyst and reactor effects

27

Deactivation mechanisms

(1) Poisoning

Al2O3

Co 0

S

(5) Carbon deposition

Co 0

C

(3) Mixed compounds

CoAl2O4CoOCo0

Al2O3

oxidation diffusion

(7) Catalyst attrition (catalyst loss in slurry

reactor)

Before After

R. Zhao et al. Applied Catalysis A: General 189 (1999) 99–116

(6) Surface reconstruction

Before

After

J. Wilson, C. de Groot, J. Phys. Chem. 99 (1995) 7860–7866

(4) Sintering

Co 0

Co 0

Co 0

S. Soled, et al, Proc. 11th Int. Symp. 21st

North American Meeting 2009

(2) Metal re-oxidation

Al2O3

Co0 CoO

H2O H2

Selectivity and catalyst stability

28

Which are the important parameters which influence the selectivity and stability?

Examples

Crucial challenges in syngas conversion

29

Tools: Laboratory Reactors for Syngas Conversion

Milli-fixed bed

ID = 0.14 cmL = 32 cm

mcat = 0.5 gNo SiC dilution

Double-shell heatexchanger

Mechanically stirred (100 mL)mcat = 5 g

Catalyst suspended in wax

Slurry stirred tank reactor

High Throughput Flowrence Unitfrom Avantium

16 parallel reactorsID = 1.3 cm

L = 5 cmmcat = 1 g

SiC/Cat = 5:1Electricalheating

Centimetric fixedbed

A High Throughput Technologies REALCAT Platform: Catalyst Synthesis, Characterization and Test

31

●Bifunctional catalysts

Well-dispersed metal particles + Mordenite, Beta and ZSM-5 zeolite

●Using the two stage process =Two catalysts and two

reactors

Fischer-Tropsch synthesis+Hydrocracking

Industry (Oryx, Pearl Qatar)

●Nanoreactors

Metal particle encapsulation, steric effects on chain growth

●Promotion (Iron catalysts)

Selectivity to olefins and alcohols to olefins, alkanes and alcohols

32

Two stage process =Two catalysts and two reactors

Sasol-Qatar Petroleum Oryx PlantQatar

0 10 20 30 40Carbon number

10

8

6

4

2

0

Pro

duct

com

positio

n,

wt

%

after medium-severity hydrocracking

Fischer-Tropschproduct

after high-severity hydrocracking

Hydrocarbon distribution of the Fischer-Tropschproducts over cobalt-based catalysts and by additional hydrocracking (adapted from Sie and Senden)

SyngasFischer-

Tropsch

Hydro-carbons

Hydro-

processing

Refined fuels

-diesel

-Jet fuel

-kerosene

33

Bifunctional catalystsCo/Zeolite

Two reactions ( Fischer-Tropsch and cracking/isomerisation ) on the same catalyst

BetaMordeniteZSM-5

Pore diameter = 5.5 Å Pore diameter = 6.5 ÅPore diameter = 7.6x6.4 Å

Pore diameter

“Small sizes of the pore is a handicap for the ZSM-5 zeolite…”

C. Marcilly, Zeolites in the Petroleum Chemistry

34

Bifunctional catalysts: H ydrocarbon distribution in heavier products

Catalysts prepared by mechanical mixing and by impregnation

● Zeolite porous structure

●The most significant decrease in the fraction of long chain hydrocarbons on the BEA based catalysts, which have the most open pore system .

BEA>MOR>ZSM-5

Wax over Co/BEA is liquid in comparison withsolid wax formed over Co/ZSM-5 and Co/MOR!

35

Nanoreactors: Shape Selectivity Concept

The shape selectivity concept suggests that “the tra nsformation of reactants into products depends on how the processed molecule s fit the active site of the catalyst” [B. Smit, T. L. M. Maesen, Nature 200 8].

Reactant Transition state Products

Shape selectivity concept for Fischer-Tropsch synth esisDeveloped with Dr. V. Ordomsky (UCCS)

commonly used for zeolites

36

Nanoreactors: Synthesis

.

●Surfactant (CTAB) dispersed in hexanol● addition of Co (NO 3)2.6H2O ● Clear emulsion was mixed with NaBH4.● After stirring the appropriate amount (to obtain 90 wt.% of SiO 2) of TEOS was added and allowed to hydrolyze during 5 h for Co@SiO 2-5 or 48 h for Co@SiO 2-48. ● The metal nanoparticles coated with silica were washed thoroughly using ethanol and water, dried and calcined.● Co/SiO 2 prepared using nitrate impregnationCo@SiO2 prepared via impregnation with cobalt nanoparticles

Impregnated samples

Nanoreactors

37

Nanoreactors: Catalytic performance in Fischer-Tropsch synthesis

Slices redrawn from the volume reconstructed from the electron tomography analysis under STEM-HAADF(S. Moldovan and O. Ersen).

240°C, 20 bar, H2/CO= 2, GHSV of 67 L/gCo·hHigh activity per cobalt and better stability

GHSV takes into account different cobalt loading

nanoreactors

Impregnatedcatalysts

38

Nanoreactors: Hydrocarbon selectivity

ASF distribution for both the Co/SiO 2 and Co@/SiO 2 catalysts prepared by impregnation. Significant deviations from the ASF distribution fo r Co@SiO 2-5. Lower selectivities to heavier hydrocarbons than cou ld be expected from the ASF model. No increase in methane selectivity (S CH4=2-4%)Growth of hydrocarbon chain restricted by the nanor eactor volume (C 30H62, d=3.5 nm).

39

Results based on Fe/SiO2 (Commericial silica)

0 10 20 30 40 50 600

20

40

60

80

100

C

O c

onve

rtio

n (%

)

Time on stream (h)

Activated with CO

Activated with H2

CO CO2

H2 H2OFe3O4FeOFe

Selectivity control on iron catalysts : Active phase in iron carbide catalysts

Activation conditions : H2 or CO flow at 623 K for 10 h.Reaction conditions : catalyst, 1.0 g; H2/CO = 2.1; GHSV, 3.6 NL g-1h-1; temperature, 573 K; pressure, 20 bar; time on stream, 60 h.

Supposed to be active phase for FTS.

40

General trends in CO hydrogenation on iron catalysts – HTE Results UCCS(series of mildly promoted Fe/SiO2, Fe/P=100:2, H2/CO=1, p=10 bar)

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70 80 90

CO

2 s

ele

ctiv

ity,

%

Conversion, %

0

10

20

30

40

50

60

70

0 20 40 60 80 100

C2-C

4o

lefi

n s

ele

ctiv

ity,

%

Conversion, %

0

10

20

30

40

50

0 20 40 60 80 100

CH

4 s

ele

ctiv

ity,

%

Conversion, %

0

5

10

15

20

25

30

35

40

45

50

0 20 40 60 80 100

C5

+se

lect

ivit

y, %

Conversion, %

Almost Stoichiometric

WGS: CO+H2O=CO2+H2

0

5

10

15

20

25

30

35

40

45

50

0 20 40 60 80 100

C2-C

4p

ara

ffin

se

lect

ivit

y, %

Conversion, %

41

General trends in CO)Trends in CO hydrogenation on Fe/SiO 2

catalysts(promoted Fe/SiO2, H2/CO=1, p=10 bar)

Low conversion High conversion

Very high olefin selectivity Low olefin selectivity

High methane selectivity Moderate methane

selectivity

Low C5+ selectivity High C5+ selectivity

Little trend in light paraffin selectivity

H2/COC2-C4

Cn-Cm

C2-C4 olefins C2-C4 paraffins

oligomerizationC1

Conversion

CH4

Simplified schemaCn-Cm hydrocarbons

hydrogenation

How to improve selectivity to light olefins?

• Slow down surface polymerization

• Decrease the catalyst hydrogenation activity

42

Different effect of Na on the catalytic performance depending on the support

Na addition causes activity decreasing for Fe/CNT, Fe/SiO2, and Fe/CMK-3

Increase in the activity of Fe/Al2O3

Increase in ole/paraffin ratio after promotion with Na

Increase in chain growth probability

CO conversion and ole/para ratio in C2-C4 products depending on Na/Fe ratio

(P = 2 MPa, H2/CO=2.1, GHSV=16 L h-1 g-1, T = 573 K)

Fe/Al2O3

Fe/CNT

Na promotion: Decrease in hydrogenating activity, enhancement of chain growth

Effect of the support

Another way to FT control selectivity: iron -zeolite bifunctional catalysts

43Chain growth factor α

55%

ASF Hydrocarbons

Syngas

Iron carbide Zeolite

FT synthesisCracking/isomerisation

Non ASF Hydrocarbons

•Suitable to reduce the hydrocarbon chain•Not suitable for olefin synthesis because of strong zeolite hydrogenation activity

44

Challenges in synthesis of fuels and chemicals: Catalyst deactivation

Identification of initial deactivation mechanisms using operando techniques

45

Reactor and setup for in-situ synchrotron based

XRD/XAS/Raman measurements

BM01B (SNBL) beamline (ESRF, Grenoble,France)

� Real plug flow reactor (20 bar, 200-300 ° C) with comparable GHSV withlab-scale reactor

� Possibility to follow each step (from the reduction to the re action)

� XRD and EXAFS/XANES operando measurements

25%Co0.1%Pt/Al 2O3 catalyst: Catalytic performance under realistic conditions

46

P=20 bar, T=220°C, GHSV= 25000 Nml.h-1/gcat

XCO = 19% and SCH4 = 4.5%Steady state

0

10

20

30

40

50

60

70

80

100 200 300 400 500

Reaction time (min)

Con

vers

ion

& s

elec

tivity

(%

)Conversion, %

CH4 selectivity, %

Considerable � in the activity during first 4 hours of reaction

XRD synchrotron

operando study

46

Evolution of cobalt phases under standard reaction

conditions

47

13.0 13.5 14.0 14.5 15.0

Inte

nsité

2 Theta

t=478min

t=0 min

Al2O3

Phase HCP du Co0

Phase CFC du Co0

Narrowing XRD patterns

Changes of XRD patterns

Sintering of cobalt particles, traces of

cobalt carbide 47

48

Co FCC

40

45

50

55

60

65

70

75

80

0 100 200 300 400

Time, min

Cry

stal

lite

size

, Å

Evolution of cobalt particle size during the reaction under standard conditions

Modeling operando XRD data

using « Whole Pattern

Matching »

TEM data with spent CoPt/Al2O3

catalyst from slurry pilot plant

(Saib, Catal Today 2011)

0

2

4

6

8

10

12

14

0 20 40 60 80 100 120 140 160 180 200

Days on stream

Po

pu

lati

on

-wei

gh

ted

av

era

ge c

ryst

all

ite

dia

mete

r (n

m)

Karaca, Sadeqzadeh, 2011

48

Few hours of the reaction in fixed bed, longer sintering time in

slurry reactor?

0.0

0.8

1.6

2.4

3.2

4.0

4.8

1 2 3 4 5 6 7 8

ϒ(J

/m2)

dCo (nm)

Surface energy vs. particle size

CoO

Co

Co

Co

Co

CoCo Co

CoCo

Co

Co

Co

Co

CoCoCoCo

Co

Co

CoCo

Co

Co

CoCo Co

Co

Co

Co

CoCo

Co

Co

Co Co

Co

Co

Co

CoCo

Co

Co

CoCo Co

Co

Co

Co

CoCo

Co

Co

Co Co

Co

Co

Co

DIFFUSION

COLLISION/

AGGLOMERATION

CoCo

Co

Co

CoCo Co

Co

Co

Co

CoCo

Co

Co

Co Co

Co

Co

Co

dj

H2

H2OCo

Co

Co

Co

CoCo Co

Co

Co

Co

CoCo

Co

Co

Co Co

Co

Co

Co

OXIDE

FORMATION

CoCo

Co

Co

CoCo Co

Co

Co

Co

CoCo

Co

Co

Co Co

Co

Co

Co

H2O

H2

di

H2O

H2

OXIDE REMOVAL

(di3+dj

3)1/3

Oxidation on particle surface, bulkremains metallic

Sintering mechanism: our concept

*Inspired from (J. Sehested, J. Catal. 2003) for nickel catalysts

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

0 100 200 300 400

Re

lati

ve

CO

co

nv

ers

ion

time (min)

4

5

6

7

8

9

0 100 200 300 400

Me

an

Cry

sta

llit

e S

ize

(n

m)

time (min)

02

46

8

0

10

20

30

40

50

60

70

0

60

120

180

240

300360

CO

Co

nve

rsio

n

tim

e on s

trea

m (m

in)

Reactor length (mm)

Effect of sintering on the catalytic activity (gPROMS model)

Typical Simulation Results

Fixed Bed Reactor Modeling

51

Deactivation of Copper-ZSM -5 catalysts for direct DME synthesis from syngas

30%

40%

50%

60%

70%

80%

90%

0 10 20 30 40 50 60 70 80 90 100

CO

co

nv

ers

ion

Time on stream / h

CZA-ZSM-5@95 CZA-ZSM-5@110 CZA-ZSM-5@80

CZA-ZSM-5@65 CZA-ZSM-5@500 CZA-ZSM-5@800

0%

10%

20%

30%

40%

50%

60%

70%

0 10 20 30 40 50Reaction time /h

CO

Con

vers

ion

water injection for 24 h0.2ml/min without CO

Syngas

(CO+H2)

Hybrid bifunctional catalysts showed noticeable deactivation

Water injection leads to a major drop in the activity

T=260 °C, P=20 bar, GHSV=3600 cm3/gcat h

52

Copper sintering under the reaction conditions: in-situ XRD

X-ray patterns recorded between five XANES scans with position sensitive detector

Reduction in H2 at 290 C, DME Reaction (H2=4 ml/min,CO=2ml/min), P=20 bar, T=260 °C

copper particles size calculated using Scherrer equation at 2θ=43.0

Cu sintering

13 14 15 16 17

13 h

11 h

12 h

9 h

8 h

5 h

3 h

2 h

Inte

nsity

, a.u

.

2 Theta

Reduced

Deactivation mechanism of CZA-HZSM-5

53

Time, h

Cu-Zn-Al-HZSM-5

70

60

50

CO

co

nve

rsio

n, %

2CO + 4H2 CH3OCH3 + H2O

DME

synthesis

H+ H+

H+

H+

H+

H+Zn-O-Al

Cu H+

H+Zn-O-Al

CuCu2+

Cu2+

Cu

H+H+

H+HZSM-5

The reason of deactivation over the hybrid catalyst CZA-HZSM5 is mainly Cu sintering and ion exchange of Cu2+ with the protons of Zeolite HZSM-5

How to solve the problem of Cu sintering and ion exchange ?

Cu

54

TEOSH-ZSM5

SiO2

SiO2

H+ H+

H+

H-ZSM5

H+

H+ H+Liquid phase deposition

Si Si

H-ZSM5

O O O OOSi Si SiSi

OOOOO

Calcination

� Silylated HZSM-5 by tetraethoxysilane (TEOS) modification

� Silylated HZSM-5 by Hexadecyltrimethoxysilane (HTDMS) without calcination

HZSM5

hexadecyl

Hydrophobic HZSM5

HZSM5

HTDMS

Liquid phase deposition

Modified HZSM5 by silane reagents

SiO2/HZSM5

Most important parameters affecting the selectivity and selectivity have been identified

55

Highly selective catalysts

Very stable catalysts

Acknowledgements

56

Thank you for your attention!

UCCS CatEn group

Dr. V. OrdomskyDr. M. VirginieDr. A. GribovalDr. Y. LuoDr. M. Sadeqzadeh (PhD student)Dr. C. Liu (PhD student)Dr K. Cheng (PhD student)Dr. G. de Souza (PhD student)Ms. M. Cai(PhD student)Mr. A. Carvalho (PhD student)

Dr. V. Valtchev, LCS, CaenDr. C. Pham-Huu, ICPEES, StrasbourgProf. O. Ersen, IPCMS, StrasbourgProf. Wang Ye, XiamenProf. P.A. Chernavskii, MoscowDr. O. Safonova, ESRFDr. W. van Beek, ESRFProf. N. R. Marcillio, UFRGSProf. M. Oberson de Souza, UFRGS

Deactivation of CZA-X hybrid catalysts

57

Conditions: Cu-based catalyst / solid acid catalyst(weight ratio) =5/3;

T= 260 C, P=20 bar, H2 /CO=2, SV=3600 ml/(h·gcat）