Aspects on reaction intensification

Åbo Akademi, Faculty of TechnologyDepartment of Chemical Engineering

Process Chemistry Centre,Laboratory of Industrial Chemistry

and Reaction EngineeringFI-20500 Turku / Åbo Finland

Tapio Salmi

Process intensification

Structures and methods which lead to a considerable compression of the equipment size, and to a more efficient, selective and clean production

Raw material ProductsProcess

Process intensification

Intensification methods

Ultrasonic techniquesMicrowave techniquesIntegrated reaction and separationReactions under extreme conditionsUnsteady state operationEnhancement of mass and heat transfer...

Process intensification

Intensification equipment

Structured reactors (monoliths, foams, fibre structures, columns)Reactors with internal heat exchangersMicroreactorsEquipment for reactive separation Spinning disk reactors...

Reaction intensification

Intensification of the chemical reactor performance is a vital part of chemical reaction

engineering and process intensification

Reactant molecules Product moleculesChemical reactor

Batch and continuous reactors

Three-phase reactors in laboratory scale

New catalyst materials have emerged

monoliths fibres/cloths foams

Benefit: low pressure drop, suppressed diffusion resistance inside the catalyst particle

Challenge: activity, selectivity, metal particle size, chemical state

Monolith catalysts

Continuous reactor –KATAPACK column

Fibre catalystsFibre catalysts

4 nm

TEM image of the 5 wt.% Pt/Al2O3 (Strem Chemicals) catalyst

SEM image of the 5 wt.% Pt/SiO2 fiber catalyst

1 mm

125-90 m particles

D = 27% dPt= 4 nm

D = 40% dPt= 2.5 nm

Catalysts in micro and nanoscale

5 m

TEM imageSEM image

4 nm5 m

• 5wt.% Pt/SF (Silica fibre) 5wt.% Pt/Al2O3 (Strem)

4 nm

Reaction and mass transfer in three-phase reactors - bottlenecks

Bulk Film Bulk Film Solid

Reaction and diffusion

Even though the governing phenomena of coupled reaction and mass transfer in porous media are principally known since the days of Thiele and Frank-Kamenetskii

Reaction and diffusion

They are still not frequently used in the modeling of complex organic systems, involving sequences of parallel and consecutive reactions.

Evaluation of Thiele modulus and Biot number for first-order reactions are not sufficient for such a network comprising slow and rapid steps with non-linear reaction kinetics.

ei

GiM D

RkBi

Biot

eD

kR

'

Porous particle

dr

rNdrr

dt

dc SiS

pipi 1

j

jjiji aRr

Reaction, diffusion and catalyst deactivation in porous particles

Particle model

Rates

Separable and reversible deactivation kinetics

tkjjjj

jeaaaa *0

*

1,1 1

1'1*

0*

ntnkaaaa n

j

n

jjj

Fresh catalyst sample Recycled catalyst sample

Raney-NickelCatalyst

Porous particle model

dr

dcDN i

eii

dr

dc

r

s

dr

cdDaR

dt

dc iieijjijpp

i2

21

, where Dei=(p/p)Dmi

0dr

dci 0r

Rcckdt

dcD iiLi

Rr

iei

Rr

Boundary conditions

Batch reactor

GLGLipii aNaN

dt

dc

RcckN iiLii

GiLi

i

bLii

bGi

GLi

kk

KcKc

N1

Liquid phase mass balance

Liquid-solid flux

Gas-liquid flux

Example systems

Hydrogenation of

Citral

LactoseO O

OH

OH

OH

OHHO

OHOH

OHOH

Lactose Lactitol

O O OHOH

OH

OHHO

OHOH

OHOH

LactulitolLactulose

OOH

OH

HOOH

OH

Galactose

+Glucose

OHOH

OH

HOOH

OH

Galactitol

OH

OH

OH

OHOH

HO

Sorbitol

H2

H2

H2

H2

H2

IsomerizationOHˉ

Hydrolysis

O

O

O

OHHO

OH

OHHO

HO

HO OH

O

O

O

OH

OH

OH

OHHO

OHOH

OH

Lactobionic acid(Na-salt)

DehydrogenationHydrogenation

O

OH

OHHO

OH

OH

O

OH

O OHOH

OH

OHHO

OHOH

OH

O

O O

OH

OH

OH

OHHO

OHOH

OHOH

O O

OH

OH

OH

OHHO

OHOH

OHOH

Lactose Lactitol

O O OHOH

OH

OHHO

OHOH

OHOH

O O OHOH

OH

OHHO

OHOH

OHOH

LactulitolLactulose

OOH

OH

HOOH

OH

Galactose

+Glucose

OHOH

OH

HOOH

OHOH

OH

OH

HOOH

OH

Galactitol

OH

OH

OH

OHOH

HOOH

OH

OH

OHOH

HO

Sorbitol

H2

H2

H2

H2

H2

IsomerizationOHˉ

Hydrolysis

O

O

O

OHHO

OH

OHHO

HO

HO OH

O

O

O

OHHO

OH

OHHO

HO

HO OH

O

O

O

OH

OH

OH

OHHO

OHOH

OH

O

O

O

OH

OH

OH

OHHO

OHOH

OH

Lactobionic acid(Na-salt)

DehydrogenationHydrogenation

O

OH

OHHO

OH

OHO

OH

OHHO

OH

OH

O

OH

O OHOH

OH

OHHO

OHOH

OH

O

Modelling resultsCitral hydrogenation

0 50 100 150 200 250 300 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

time (min)

Re

lativ

e co

nce

ntr

atio

n

citronellal

citral

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0

0.05

0.1

0.15

0.2

0.25

Hydrogen

Citral

Citronellal

c m

ol/l

x

trilobic catalyst particle

0 50 100 150 0

0.2

0.4

0.6

0.8

1

x

Intrinsic kinetics,thin catalyst layers

concentration profiles inside a trilobic catalyst particle

Lactose hydrogenation to lactitolIntrinsic kinetics

0 50 100 150 200 250 3000

5

10

15

20

25

30

35

40

45

time (min)

w-%

20 bar

20 bar

70 bar

70 bar

T=110C, P=70 bar

Lactose

Lactitol

Lactose hydrogenation - diffusion

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.2

0.4

0.6

0.8

1

1.2

1.4

x

c (

mo

l/l)

0.03 mm

0.3 mm

0.3 mm

0.03 mm1.0 mm

1.0 mm

3.0 mm

3.0 mm

Concentration profilesinside catalyst particlewith different particle sizes

SurfaceCenter

LactoseLactitol

0200

400600

0

0.5

1

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0200

400600

0.5

1

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Concentration profiles of lactose inside a catalyst particle

No deactivation Deactivation

surface

surface

center center

timetimex

x

4FeSO4 + O2 + 2HSO-4 + 2H3O- ®

2Fe2(SO4)3 + 4H2O

Proceeds as a gas-liquid reaction but catalyst helpsCatalyst: Active carbon

Oxidation of ferrous sulphate to ferric sulphate

Simultaneous catalytic and non-catalytic reaction

FeSO4 oxidation in Katapack

Non-catalytic process can be enhanced by addition of a heterogeneous catalyst

Diffusion resistance in catalyst particle Gas-liquid equilibria

Kinetic model

c ck 1 + c K

)K

c c cc c( )c

c(k = r FeO

OO

C

3/2wFe

+w¼OFe

+w

¼O

II

III

II

2/12/1

Non-catalytic reaction rate

Catalytic reaction rate

) f (1

c a

a + 1

c c a = r

Fe1

2

O2Fe2

Catalytic and noncatalytic oxidation

0 50 100 150 200 250 3000

1

2

3

4

5

6

7

8

9

10

time/min

Fe(

II)/

wei

ght-

%

Experiments at 4 bar : from top 60oC, 80oC (o), 100oC (x) and 10 bar 100oC (+). Lines by simulation.

Catalyst particles

0 0.2 0.4 0.6 0.8 10.035

0.04

0.045

0.05

0.055

0.06

0.065

0.07

2

510

=

Simulated concentration profiles of oxygen inside catalyst particle

0 0.2 0.4 0.6 0.8 10

0.2

0.4

0.6

0.8

1

1.2

1.4

Catalyst particle coordinate

2

510

J =

Simulated concentration profiles of Fe2+ inside catalyst particle

r + aN aN +

dl

cdD +

dldc

w1

= dt

dcnoncatiLvLiv

bLi2

Li2

LLLi

LL

Li,

aN

dl

cdD+

dt

dcc

dldw+ w

dldc1

= dt

dcv

bLi2

Gi2

GGG

Gi2GiG

1GGi

G

Gi

r + aN aN

dl

cdD +

dldc

w1

= dt

dcinoncat,vLippL2

Li2

LLLi

LL

Li ''''

''

'

'

Dynamic column model

Liquid phase

Gas phase

Liquid inside the net

Gas-liquidmasstransfer

MasstransferThrough thenet

Reaction in particle

Reaction inBulk liquid

Model simulation and verification

0

5

10

15

20

05

1015

2025

0.60.8

11.21.41.61.8

22.22.4

mol/l

timelength

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250

time (min)

120 C 4.7 bar kla=0.116 1/s

Independent model verification in a pilot reactor

Concentration of ferrous sulphate

Catalyst reactivation and ultrasound

Deactivation is serious problem in heterogeneous catalysis

- help is needed

The aim of the study:

a) to determine the effect of ultrasound on deactivation kinetics

b) to model quantitatively the effect of ultrasound on kinetics

0 2 4 6

Time

CA

1st

3rd

2nd

4th Batch

Catalyst performance in batch reactor

• The catalyst activity declines at each batch even at low temperatures

• The reason is very often the fouling by organic compounds

• More catalyst is added -> mass transfer limitations enter

• Can ultrasound prolong the catalyst lifetime?

A + H2 -> P

In situ ultrasound equipment

Electricconnection

HornM odifiedRushton

turbine

Pressurizedautoclave

High-pressure autoclave with in situ ultrasonic irradiation system (in-house design) :

Power input 0-100 W

Operating frequency 20kHz

Slurry reactor

Multi-transducer set-up

Generator (0-600W)

20 kHz

Reactor pot inserted

6 transducers

A time-variable

power input

Case studies

Case Catalysts

D-fructose hydrogenation Raney-Ni, Cu/SiO2,

Cu/ZnO/Al2O3

1-phenyl-1,2-propanedione hydrogenation

Pt/SF, Pt/Al2O3 , Pt/ SiO2 ,

Pt/C

D-fructose hydrogenation

CH2OH

C

CH

HC

O

HO

OH

HC OH

CH2OH

H2

CH2OH

CH

CH

HC

HC

CH2OH

HO

HO

OH

OH

CH2OH

HC

CH

HC

HC

CH2OH

HO

OH

OH

OHH2

D-sorbitol

D-mannitol is a low caloric sweetener widely used in pharmaceutical and alimentary industry.

D-mannitol

D-fructose

Reaction conditions

• Pressurised batch autoclave (VL=250ml)

• Stirring rate 1800 rpm

• pH2 = 30 bar

• T=110C

• Nominal ultrasound intensity 130 Wcm-2

• Solvent: deionised water

• Catalyst: Raney-Ni

Catalyst characterisation

The spent Raney-Ni catalyst treated in the absence of ultrasound.

The ultrasonic treated spent Raney-Ni catalyst.

0

20

40

60

80

100

0 20 40 60 80 100 120

time (min)

Con

vers

ion

(%)

3rd reuse

3rd reuse Sono

fresh Sonofresh Silent

Deactivation of re-used catalyst

)( 21 rrdt

dCB

F

F

FC

CX0

1 , tckk

F

F HBeC

C 221 )(

0

tktckkX HB''

21 2)()1ln(

Kinetic modelling

Simple kinetic model

Test of pseudo-first order kinetics. T = 110°C, pH2 = 30 bar. X = conversion of D-fructose.

0

0.5

1

1.5

2

2.5

3

3.5

0 20 40 60 80

reaction time / [min]

-ln

(1-X

)

silent, fresh

silent, 3rd reuse

sono, fresh

sono, 3rd reuse

BHCkk

2'''

k’silent k’sono

fresh 2.27 2.19

3rd re-use 1.06 1.61

min)('catg

mlk

Ultrasound effect and model simulation

Example fit using the graphically obtained rate constants, T = 110°C, pH2 = 30 bar. (Fructose: ○ = silent, ● = ultrasound;

Mannitol: □ = silent, ■ = ultrasound; Sorbitol: ◊ = silent, ♦ = ultrasound).

0 20 40 60 80 100 1200

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

reaction time / [min]

conc

entr

atio

n /

[mol

/l]

3rd reuse

Mannitol & Sorbitol

Fructose

1-Phenyl-1,2-propanedione hydrogenation

O

O

O

OH

O

OH

O

HO

O

HO

OH

HO

(B)

(C)

(D)(E)

OH

HO

HO

OH

(I) (F)

(G)(H)

(A)

OH

HO

Used for the synthesis of several pharmaceuticals e.g.

ephedrine...

Reaction conditions

Pressurised batch autoclave (VL=200ml)

• Stirring rate 2000 rpm

• pH2 = 10 bar

• T=15C

• Nominal ultrasound intensity 78 Wcm-2

• Catalyst modifier (M): cinchonidine cM = 0.1mg/ml

• Solvents: toluene, methyl acetate, mesitylene

N

N

OHH

Catalyst modifier cinchonidine

The spent Pt/SF catalyst treated in the absence of ultrasound (SEM image).

The ultrasonic treated spent Pt/SF catalyst (SEM image).

Catalyst characterization

Catalyst deactivation – experiments in a fixed bed

0

10

20

30

40

0 20 40 60 80

Time-on-stream (min)

c (m

mo

l l-1

)

A

R

S

Solvent and ultrasound effect on conversion on Pt/Silica fibre catalyst

0

20

40

60

80

100

0 20 40 60 80 100 120

time (min)

Con

vers

ion

(%)

US effect

US effect

US effect

Methyl acetate

Toluene

Me

Mesitylene

Hydrogenation kinetics of 1‑phenyl‑1,2‑propanedione

0.000

0.005

0.010

0.015

0.020

0.025

0 20 40 60 80 100 120

time (min)

c (m

ol /

dm

3 )

0.000

0.005

0.010

0.015

0.020

0.025

0 20 40 60 80 100 120

time (min)

c (m

ol /

dm

3 )

(R )- enantiomer

(S )- enantiomer

Hydrogenation kinetics of 1‑phenyl‑1,2‑propanedione (A) at 15°C and 6.5 bar hydrogen

AB (R)

C (S)

H2

H2

B + C

B

C

AB

C

2

' HAjj cckr

))1(exp((

1

1'0

' tkdkk jj

dUS kk

A + * A* A’* P + * kd US

Modelling of ultrasound and deactivation

Parameter values

Solvent kd

(100*min-1)

kd,US

(100*min-1)

kd-kd,US

(100*min-1)

Toluene 1.74 1.22 0.52

Mesitylene 14.1 13.0 1.1

Methyl acetate

1.45 1.07 0.39

Data fitting and model simulation

Batch reactor

0 20 40 60 80 100 1200

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

US

No US

time (min)

c (m

ol/l)

A

B

C

DE A

B

C

DE

A: 1-Phenyl-1,2-propanedione, B: (R)-1-hydroxy-1-phenylpropanone, C: (S)-1-hydroxy-1-phenylpropanone, D+E: (R)+(S)-2-hydroxy-1-phenylpropanone

0 20 40 60 80 100 1200

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

c (m

ol/l)

time (min)

A

BC

D+E

A, no US

A, US

Continuous, dynamic fixed bed with axial dispersion

Liquid phase

Gas phase

The effectiveness factor ( ) is obtained from the pellet model ( ).

Special cases: : plug flow reactor all flows zero : batch reactor

GVGiGi

GGGi

GGGGi N

z

c

z

cPe

t

c /)()( 12

21

BieiLVLiLi

LLLi

LLLLi rN

z

c

z

cPe

t

c

/)()( 12

21

eiiN

GL PePe ,),( GL

C

B

Dynamic modelling of continuous bed

Non-steady state kinetics in continuous fixed bed

c (m

ol/d

m-3)

Experiences from ultrasound

Acoustic irradiation

• Can sometimes improve reaction rate and selectivity

• Prevents catalyst deactivation by surface cleaning and smoothening

• Effects are solvent dependent

• Catalyst and reaction specific effects are visible

• The applications are not limited to catalysis –

the approach works for liquid-solid reactions

Conclusions and future aspects

Reaction intensification is a part of process intensification – keep the entire process in mind

Fundamental understanding on kinetics, thermodynamics, flow structures should be the basis of reaction intensification

A lot of reactor and catalyst structures are available – they shoud be evaluated critically

Intensification methods are very promising, but scale-up is a challenge Implementation of catalysts is an intensification approach as such Active search for new application areas is needed More imagination is needed

A structured reactor

Design by Victor Vasarely (1908-1997 )A famous Hungarian-French painter

”Colours are the vitamines of our life”

Art shows the way

Thank you for your kind attention

Laboratory of Industrial Chemistry and Reaction Engineering