Joris W. Thybaut - UGent Advisory... · Joris W. Thybaut 1 Methusalemadvisory board meeting, Ghent, Belgium, 19 June 2012 Single-Event MicroKinetics(SEMK) in complex reaction mixtures,

Laboratory for Chemical Technology, Ghent University

http://www.lct.UGent.be

Joris W. Thybaut

1

Methusalem advisory board meeting, Ghent, Belgium, 19 June 2012

Single-Event MicroKinetics (SEMK) in complex

reaction mixtures, catalyst design based on catalyst

descriptors & Adsorption by nanoporous materials

(P1 - P3)

rational catalyst design


catalyst library activity library

optimizeddescriptors

newconcept

industrialapplication

performance testing

design

synthesis

kinetic and catalystdescriptors

modelling

1

2

3

4

H

H

H

H

H

H

H

H

H

H

H

H

H

H

H

H

H

H

H

H

HH

H H

H

H

H

H

**

*

*

**

*

* *

*

*

*

H

H

H

H

H

H

H

H

H

H

H

*

1

3 6

2 4

5

7

8

9

10

11

12 13k(2,2) k(2,2)

k(2,2)k(2,2)

k(2,2) k(2,2)k(1,2)

k(1,2)

k(1,2)

k(1,2)

k(1,2)

k(1,2)

k(1,2)

k(1,2)

k(0,2) k(0,2)

k(0,2)k(0,2)

k(0,2) k(0,2)

2

Single-Event MicroKinetics (SEMK)


+ +

reactant 0,# 0,#global b

#

global

σ k T S Hk exp exp

σ h R RT

∆ ∆= −

%

3

rational ZSM-22 design


related patents:- US20100181229- US20110042267publications:- Hayasaka et al. Chem. Eur. J. (2007)- Choudhury et al. J. Catal. (2012)

0

20

40

60

80

100

423 473 523 573

Iso

mer

Yie

ld /

mo

l%

Temperature / K 4

outline

• introduction• µkinetic engine• catalyst design

– methanol-to-olefins– xylene isomerization/ethylbenzene

dealkylation

• adsorption characterization– hydrocracking– transesterification

• conclusions5


µkinetic engine: background

• generic methodology for kinetic model construction

• no user intervention is needed in programming

• reaction network is automatically converted into rate equations

6


high-throughput experimentation

microkinetic modeling(includes all elementary steps)

1. optimize catalyst properties and kinetic parameters

2. predict behavior for reactions / compounds of the same family

µkinetic engine: workflow

7


Operating conditions

Experimental data

No. experimentsNo. input variables

No. responsesReactor typeInitial values

DatafileGeneration

Reaction Network

Standalone software tool

Input data

Kinetic parametersStatistical analysis of results

Generate plotsModel predictions

detailed flow scheme

8


• writing section• simulation results• statistical interpretation

• overall information• experimental• regression

• perform simulationfor all experiments

• compare calculatedand experimentaloutlet flow rates

• adjust parameter values

fit ok?yes no

• start main program

• reading section• reaction network

• thermochemistry

• regression section• Rosenbrock• Levenberg-

Marquardt

• thermochemicalcalculations

• parameter constraints

• solution set of equations:

• initialization• integration

calculated versus experimental values

9


WSSQ Weight to the reponse

Experimental flowrates

Error

Parameters to be

estimated

Flowratescalculated using

model

� ��

��

��

��

��

solution set of equations: reactor model

10


• Plug Flow reactor (differential equations)

• Continuous Stirred Tank Reactor (algebraic equations)

solution set of equations: kinetic model

11


• net production rate of component i

• reaction rate (A* + B � D*, A*: intermediate, B: response)

• rate coefficient

,=∑αi i j jj

R r

, 1...

( ), 1...

n ms response tot elem steps

j paramete

s

s rs

r p C s nk

j nk f

θβ

= =

= =intermediate

where, αi,j is stoichiometric coefficient of component i in reaction j

1 1exp

ave

as T

ave

Ek k

R T T

= − −

β

solution set of equations

12


• Mass balance for the catalyst’s active sites

• analytical solution only possible for simple reaction mechanisms

• steady state approximation:• intermediates are considered as highly reactive, i.e.,

net rate of formation equals zero:

• DDASPK2.0 used as solver for the set of equations

*totC C C= +∑ intermediate

0R =intermediate

graphical user interface

13


µkinetic engine: summary

• performs microkinetic modeling adopting complex networks in heterogeneous catalysis

• no programming required by the end user

• incorporates differential and algebraic solvers + deterministic & stochastic optimization routines

• able to provide information about quasi-equillibruim steps.

14


µkinetic engine: summary

• reaction orders can also be estimated

• no rate determining step. • able to plot the agreement

between model and experimental data points with residual error and save them as images

• Provide statistical analysis of results:• 95% confidence interval,• t-value, • F value • etc.

15


outline



dealkylation


• conclusions16


methanol to olefinsMTO process provides an alternative route to produce olefins/gasoline.

Methanol/Oxygenates

synthesis

Synthesis Gas Production

Coal

Natural gas

Biomass

MTO/MTH

and higher olefins


17

(*) Formation and consumption of hydrocarbon pool species not considered.

DME Formation

Primary Hydrocarbons Formation

Higher Olefins Formation

CH3OH

DMO+

CH3OH2+

CH4 + HCHO + H+

H+ + DME

CH3+ HYDROCARBON

POOL*

C2=

C3=

C4 Olefinsisomerization




β-s

ciss

ion

Me

thyl

atio

n &

Alk

yla

tion

reaction scheme hydrocarbons over ZSM-5

18


Ethene/Propeneand aromatics

Alkanes

Propene andhigher alkenes

Cyclization and hydride transfers

Toluene

Trimethylbenzene

Exocyclicmethylation

cycle

Alkenehomologation

cycle

Propene

Higher alkenes

MeOH

overall summary of the reaction network

Bjorgen et al., J. Catal. 249 (2007) 19519


Reaction network on H-ZSM-5 catalyst

20


Surface methoxy and DME formation:2 quasi-equilibrium reactions (protonation of MeOH and DME)2 reversible reactions6 adjustable parametersMethane formation:Irreversible reaction1 adjustable parametersPrimary olefins formation:8 reversible reactions10 adjustable parameters1 total concentration of hydrocarbon poolHigher olefins formation:1 Irreversible reaction (methylation)1 reversible reaction (alkylation)12 adjustable parameters- 6 activation energies based on stability

of carbenium ions-6 olefin protonation enthalpies depending onnumber of carbon atoms from O2 to O7

Reaction network in terms of elementary steps for ZSM-5 catalyst:

Number of Species DME

formation

Primary olefins

formation

Higher olefins

formation

(Cyclic) Olefins/DME/H2O 3 4 50

Carbenium ions 3 5 41

Aromatics 1

Total 6 10 91

Number of elementary steps

Protonation 3 3 71

Deprotonation 3 3 71

Hydride shift 40

Methyl shift 15

PCP branching 54

Methylation 3 22

Demethylation 3

Alkylation 2 15

Dealkylation 2

β-scission 6

hydration 1

dehydration 1

Total 8 16 294

Simulation results

21


Parity diagram of experimental vs calculated outlet MTO products flow rates at T = 360 - 480 0C, Pt = 1.04 bar and W/FMeOH = 0.5 - 6.5 kgcat.s mol-1. line: experimental; ♦: obtained from model regression

model parameter values

22


Activation energy

(kinetic descriptors)

Forward

(kJ/mol)

Reverse

(kJ/mol)

Protonation enthalpy

(catalyst descriptors)

(kJ/mol)

Surface methoxy, and DME formation

Dehydration 222.7±8.1 170.4±15.6 Methanol -61.8±3.8

Protonation with MeOH 138.6±7.5 163.7±14.8 DME -42.8±8.1

Methane formation

Methane formation 121.1±18.4

Hydrocarbon pool species formation

Methylation of p-xylene 101.5±1.1 123.2±17.1

Deprotonation of TMeB+ and DMeEtB+ 156.7+9.6 123.1±16.8

Methylation of DMeMCHDE and DMeEtCHDE 75.5±16.9 151.8±17.4

Dealkylation of DMeEtB+ and PDMeB+ 84.2±8.2 14.1±2.4

Deprotonation of PX+ 143.4±20.8 123.9±16.8

model parameter values

23


Activation energy

(kinetic descriptors)

(kJ/mol) Protonation enthalpy/ CHP

(catalyst descriptors)

(kJ/mol)/(mol/ kgcat)

Methylation (p-p) 131.9±28.2 Ethene -11.1±0.16

Methylation (p-s) 92.8±10.9 Propene -42.5±7.6

Methylation (p-t) 54.9±9.2 Butene -53.9±9.0

Alkylation (s-s) 138.0±9.4 Pentene -61.6±15.1

Alkylation (s-t) 119.7±17.6 Hexene -67.7±2.1

Alkylation (t-s) 167.6±28.6 Heptene -70.3±12.4

CHP(Total concentration of hydrocarbon pool species)

3.47×10-2

Performance curves

24


Experimental and model calculated yield of MTOproducts at T = 400 0C and Pt = 1.04 bar. Symbols:experimentally observed values, ethene (�), propene(▲), butene (■); lines: calculated

Experimental and model calculated yield of MTOproducts at T = 400 0C and Pt = 1.04 bar.Symbols: experimentally observed values,pentene (�), hexene (▲), heptene (■), methane(�); lines: calculated

Experimental and model calculated yield of primaryolefins at methanol conversion = ~65% and Pt = 1.04bar. Symbols: experimentally observed values, ethene(�), propene (▲); lines: calculated

Experimental and model calculated yield of olefinicproducts at T=360 0C, Pt = 1.04 bar and W/FMeOH = 4.28 kgcat.s mol-1.

Contribution analysis

25


Contribution analysis for MTH on H-ZSM-5 catalystat space time of 2.21 kgcat.s/mol and at 400 0C, conversion=66.9%

contribution analysis: temperature effect

26


Contribution analysis MTH on H-ZSM-5 catalystat space time of 1.40 kgcat.s/mol and at 480 0C,conversion=73.6%

Contribution analysis for MTH on H-ZSM-5 catalystat space time of 2.21 kgcat.s/mol and at 400 0C,conversion=66.9%

MTO framework structure effects

ZSM-5 (MFI) ZSM-23 (MTT)

27


interconnected straight and sinusoidal channels

monodimensional, straight channels

contribution analysis: framework effect

28


Contribution analysis for MTO on ZSM-23 catalyst at T= 400 0C and Weff/FMeOH = 24.8 kgcat.smol-1

conversion=50.2%

Contribution analysis for MTH on H-ZSM-5 catalystat space time of 2.21 kgcat.s/mol and at 400 0C,conversion=66.9%

Xylene isomerization on a bifunctional Pt/H-ZSM-5 catalyst

Reaction network consists out of:

• acid catalyzed reactions:

(de-)protonation,

alkyl shift (MS),

dealkylation, (DA)

transalkylation (TA)

• metal catalyzed reactions:

Hydrogenation (HYD)

• physisorption

xylene isomerization: SEMK model

29


CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

CH3

+

CH3

CH3

+

CH3

CH3

+

CH3

CH3

CH3CH3

CH3

C2H6

CH3+

CH3

CH3 CH3

+

CH3

CH3

CH3

Physisorption

Physisorption

(de-)Protonation

Physisorption

Physisorption

(de-)ProtonationMethylshift

Transalkylation

Dealkylation

(de-

)Hyd

rog

enat

ion

Metal sites Acid sites

Zeolite

CH3

CH3CH3

(de-)Protonation

Chemisorption

Chemisorption

CH3

K. Toch et al. Appl. Catal. A-Gen 425 (2012) 130-144

xylene isomerization: results

30


Successful estimation of the kinetic parameters

• small confidence interval

• physically meaningful

importance of the reactions

HYD << TA << DA ~ MS

order of activation energies:

Ea,DA >> Ea,MS ~ Ea,TA >> Ea,HYD

Successful description of the responses

∆S A

DA > 0 105 Aref Monomolecular

MS 0 Aref Monomolecular

TA < 0 10-3 Aref Bimolecular

HYD - 102 Aref Bimolecular, numberof active sites


xylene isomerization: catalyst optimization

31


acid strength modifications: ∆Hpr is varied from -60 … -110 kJ mol-1

What is of industrial relevance?maximization of PX content minimization of xylene losses maximization of benzene yield

→ definition of profit function Ψ:

→ ∆Hpr(opt) ≈ ∆Hpr(est)

(catalyst used = industrial catalyst)

max →=Ψ ∆ prH

XYL

BPX

X

SATE

0

1000

2000

3000

4000

5000

6000

7000

-110-100-90-80-70-60

Ψ

ΔHpr (kJ mol-1)

633K, 1 MPa

653K, 1 MPa

673K, 1 MPa

0

10

20

30

40

50

60

70

80

90

100

-110-100-90-80-70-60

YB

(%)

ΔHpr (kJ mol-1)

633K, 1MPa

653K, 1MPa

673K, 1MPa

0

20

40

60

80

100

120

-110-100-90-80-70-60

AT

EP

X(%

)

ΔHpr (kJ mol-1)

633K, 1MPa

653K, 1MPa

673K, 1MPa

0

2

4

6

8

10

12

14

16

-110-100-90-80-70-60

XX

YL

(%)

ΔHpr (kJ mol-1)

633K, 1MPa

653K, 1MPa

673K, 1MPa


outline



dealkylation


• conclusions32


33

long alkane hydrocracking over BETA


USY n-octane BETA n-hexadecane

B. Vandegehuchte et al. Appl. Catal. A-Gen (accepted)

shape selectivity

+

+

1,2 alkyl shift

rAS = kAS CR+

kAS = k0 exp-(Ea;AS + ∆Ea)

R TAS

+

+

Ene

rgy

Reaction coordinate

∆Ea

Ea;AS79.8 kJ mol-1

21.9 (± 1.0) kJ mol-1

Transition state shape selectivity duringEthyl branch formation

0.0E+00

5.0E-08

1.0E-07

1.5E-07

2.0E-07

2.5E-07

0.0E+00 5.0E-08 1.0E-07 1.5E-07 2.0E-07 2.5E-07

Fm

od

(mo

l s-1

)

Fexp (mol s-1)

b

34



0

10

20

30

40

50

60

70

80

4 6 8 10 12 14

↗

CN ↗

θθθθ ↗

θθθθ ↘

adsorption saturation: size entropy

Krishna et al. Chem. Eng. J., 2002

Entropic effects favor the component with the lowest carbon number atsaturation loadings as smaller molecules fit more easily in the gapswithin the zeolite matrix.

Linear approximation of -∆Ssiz0

= 0

35



transesterification

• ethylacetate + methanol over LewatitK1221 (sulphonic acid ion exchange resin)

36


+ +

E. Van de Steene et al. J. Mol. Catal. A-Chem. 359 (2012) 57-68

model discrimination

37


Simulated values (lines) exhibit a good agreement with the experimental results

(temperature effect and initial reactant molar ratio effect)

1

1

SR MeOH MeOH EtOAc MeOAc EtOHeq

MeOH MeOH EtOH EtOH

k K a a a aK

ra K a K

−

=+ +

Kinetic model: ER_MeOH_SR: Eley-Rideal mechanism with the reaction of EtOAc from

the bulk with adsorbed MeOH on the catalyst surface as rate-determining step.


model assessment: reaction mechanism

38



outline



dealkylation


• conclusions39


conclusions

• versatility of the SEMK methodology has been demonstrated

• µKE: generic platform has been createdrequiring a minimum user intervention

• zeolite framework and acid strengtheffects have been quantitatively assessed

• peculiar adsorption effects may govern the overall conversion

40


perspectives

• further integration of tools in the µKE– reaction network generation– thermochemical calculations– benchmarking against the ‘competition’

• ‘Pleiade’ of future applications– methane aromatization– ethylene oligomerization– ethanol to hydrocarbons– aldol condensations– hydroisomerisation including diffusion– … 41


acknowledgements

• Methusalem program• EC FP7• FWO• IAP• SRF• Shell, BP• …

42


questions?

43


Documents

Joris W. Thybaut - UGent Advisory... · Joris W. Thybaut 1 Methusalemadvisory board meeting, Ghent, Belgium, 19 June 2012 Single-Event MicroKinetics(SEMK) in complex reaction mixtures,