gas-phase microreactor technology some experiences
TRANSCRIPT
Gas-phase microreactor
technology – some experiences
Tapio Salmi, Dmitry Murzin, Kari Eränen
José R. Hernández Carucci, Sabrina Schmidt, Ville
Halonen Åbo Akademi, Teknisk kemi och reaktionsteknik
Microreactors
• Microstructured reactor: – At least one inner dimension
in the micrometre range
• Benefits of microreactors: High heat transfer rates
Short diffusion distances
Small inner volume: Safety
Efficient kinetic investigation and catalyst screening
2
Microreactor - advantages
• Faster transfer of research results into on-site
production
• High safety – small amounts of components
• Easy number-up to production capacity
• Smaller plants for production at distributed sites
• Smaller consuption of chemicals
• High surface-to-volume ratio
• Narrow residence time distribution (RTD)
Microchannels Coatings Micropores Clusters Reactor plates
10-3 10-4 10-5 10-6 10-7 10-8 10-9 m
Microreaction technology: total reaction control at
all length scales
Level 1 Level 2 Level 3
Chemical engineering Organic
chemistry
Catalysis
Mechanical engineering
Materials science
Microreactors
– During the past 25
years processes
have been
developed for the
fabrication of
three-dimensional
microdevices from
a wide variety of
materials based on
electronic
technologies
Microstructures = 100 – 500 mm
1 mm = 10-6 m
Objectives
• To design and build the equipment
• To develop a catalyst (preparation, screening,
characterization)
• To study the SCR of NOx and preparation of
chemicals in microreactors
• To develop kinetic and reactor models
Level 3. Catalyst deposition (impregnation)
Post-treatments
Level 1: Micro-channel formation (EDM)
Level 2: Washcoating deposition (Anodic oxidation)
Stainless Steel plate
SS plate with alumina layer
Stacking & Bonding
Aluminum plate with coated micro-channels
‘Monolith’ micro-reactor
Aluminum plate with micro-channels
Plates manufacture
Plates impregnation
Washcoating
Al2O3
Ag
impregnation
Ionic liquids Ag
Ammonia-based Au
Characterization
SEM pictures of the plates SEM pictures of raw platelet – 50X
SEM pictures of raw platelet –
200X
460 mm
70 mm
Modelling of microchannels
• Modelling of microreactors still inmature.
• Nevertheless, reliable results due to
dimensions
• Laminar flow (almost no turbulent flow
observed)
• Big differences in Lab-on-a-chip and
micro total analysis (mTAS) with not-so-
micro channels
Geometry for microchannel
analysis
SEM picture of a microplate - 50X
Geometry for the analysis
H=460 mm
L=12500 mm
Lappeenranta, 17.03.08
Microreaction engineering
Differences between a macro- and
microflow
• Flow in microchannels is usually laminar but
turbulent in macrochannels
• Diffusion paths in microchannels for heat and
mass transfer are short
• High surface-to-volume ratio
• Solid wall material are important. Surface heat
transfer effects
Dimensions and fluid properties
Parameters Value
Length 125000 mm
Width 460 mm
Height 75 mm
Pressure 101 kPa
Temperature 373 K
Temperature at the wall 373 K
Viscosity 2e-5 Pa*s
Molecular mass 28 kg/kmol
Kn all the time in non-slip domain,
usual continuum description and all
the components of the velocity
are zero close to the walls
Flow in microchannels
• What are the
boundaries?
• What is
microscale?
• Are the classical
equations valid?
Gas-phase microreactor system with spare parts
What is Knudsen number?
• Continuum flow with no-
slip boundary conditions
(Kn < 10-2, 10 -3)
• Continuum flow with slip
boundary conditions (10-2
< Kn < 10-1)
• Transition flow (10-
1<Kn<10)
• Free molecular flow
(Kn>10)
300 400 500 600 700 800 9002 10
4
3 104
4 104
5 104
Kn vs. Temperature
Temperature (K)
Knudse
n n
um
ber
22 pd
TkB
LKn
Macroflows
relationships are valid!
j
j
ik
i
k
k
i
k
i
ij
ij
i
x
u
x
u
x
u
dxg
dx
p
dx
uu
dt
um
3
2
Convection
Pressure
difference
Gravity Viscous dissipation
b= 230µm a= 70-75 µm
The total area
of the channels
was
approximated
to that of an
ellipse
Parameters:
Navier-Stokes (2D)
Boundary settings:
2 & 19 No slip
Inflow Velocity
According to
figure
Flow properties:
N2 ρ = f (T, p) ~ 1.251 Kg/m^3
h = f (T) ~ 2e-5 Pa.s
Mesh
Transient
Analysis Type
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
20 40 60 80 100
Volumetric Flow (mL/min)
Inle
t V
elo
city
(m
/s)
Inflow
Velocity
2
baπA
CFD
Reactor modelling
Models
1. Axial dispersion
2. Laminar flow and radial diffusion
• Nevertheless of the thickness of the catalytic
layer and depending on the reaction conditions,
diffusion limitation inside the microchannels
might play a role in the system
• Mass transfer limitation from the bulk phase to
the surface of the coating could appear, mainly
via molecular diffusion
Diffusion effects in modelling
Catalyst layer and microchannels
dx
dc
x
s
dx
cdDr
dt
dc iiepi
ip
'
2
'2
2
'
2,0s
1,0x
dl
wcd
dr
dc
rdr
cdD
dt
dc iiii
i
12
2
0
2
1 wR
rw
0,0 rdr
dci
Rr
ii
x
ie
dr
dcD
dx
dcD
0
'
Boundary conditions:
Catalyst layer
Mass balance in the microchannels:
dci'
dx
x1
0
Layer thickness
dl
wcd
dr
dc
rdr
cdD
dt
dc iiii
i
12
2
0
2
1 wR
rw
De
dc i'
dx
x 0
A distribution function w() as a function of the thickness :
max
min 0
'
)(
w
ddx
dcD
x
ie=
Equation to be solved:
Preparation of ethylene oxide
in a microreactor
Possibility for on-site production
Ethylene oxide
• Important organic intermediate
– global production ca. 19 million metric tons
– used in e.g. cellulose derivates
• Produced in gas-phase by partial oxidation
of ethylene over silver.
– exothermic reaction and toxic compenents
– safety issues
Reaction scheme
• Two parallel competing oxidation reactions
– partial and complete oxidation
• Total oxidation of ethylene oxide considered to
be negligible
Experimental work
• Utilizing microreactor technology in partial
oxidation of ethylene.
• Preparation of silver catalysts for
microreactor
• Obtaining kinetic data and determining a
kinetic model for partial oxidation of
ethylene
Catalysts
• Silver catalysts
• Preparation
– incipient wetness
– washcoating
• Characterization
– BET and SEM-EDS
Catalysts - summary
• Four catalysts
– Three silver-alumina catalysts
• Silver content from 14 wt-% to 18 wt-%
– One commercial pure silver catalyst
Experiments
• A series of experiments to determine the effect of e.g. reaction temperature
• Range of reaction conditions
– pressure 1 atm
– temperature range 220 - 300°C
– total flow range 6 - 10.5 ml/min
– ethylene and oxygen concentration varied from 5 vol-% to 25 vol-%
The effect of pretreatment
0,00
0,06
0,12
0,18
0,24
0,30
0 2 4 6 8 10 12 14 16 18 20
Time [min]
Co
nv
ers
ion
[%
]
0,0
10,0
20,0
30,0
40,0
50,0
Sele
ctiv
ity
[%
]
Conversion of ethylene (ethylene) Conversion of ethylene (oxygen)
Selectivity to ethylene oxide (ethylene) Selectivity to ethylene oxide (oxygen)
The effect of temperature
0,00
0,06
0,12
0,18
0,24
0,30
200 220 240 260 280 300 320
Temperature [Celsius]
Co
nv
ers
ion
, Y
ield
[%
]
0,0
12,0
24,0
36,0
48,0
60,0
Sele
ctiv
ity
[%
]
Conversion of ethylene Yield of ethylene oxide Selectivity to ethylene oxide
The effect of total flow
0,00
0,04
0,08
0,12
0,16
0,20
4,5 6 7,5 9 10,5 12
Total flow [ml/min]
Co
nv
ers
ion
, Y
ield
[%
]
0,0
10,0
20,0
30,0
40,0
50,0
Sele
ctiv
ity
[%
]
Conversion of ethylene Yield of ethylene oxide Selectivity to ethylene oxide
The effect of oxygen
concentration
0,0E+00
2,0E-08
4,0E-08
6,0E-08
8,0E-08
1,0E-07
1,2E-07
0 5 10 15 20 25 30
Concentration of oxygen in feed [%]
Fo
rma
tio
n o
f et
hy
len
e o
xid
e [m
ol/
s]
Formation of ethylene oxide (ethylene 5% in feed)
Formation of ethylene oxide (ethylene 10% in feed)
Formation of ethylene oxide (ethylene 15% in feed)
Formation of ethylene oxide (ethylene 20% in feed)
Formation of ethylene oxide (ethylene 25% in feed)
The effect of ethylene
concentration
0,0E+00
2,0E-08
4,0E-08
6,0E-08
8,0E-08
1,0E-07
1,2E-07
0 5 10 15 20 25 30
Concentration of ethylene in feed [%]
Fo
rma
tio
n o
f eth
yle
ne o
xid
e [
mo
l/s]
Formation of ethylene oxide (oxygen 5% in feed)
Formation of ethylene oxide (oxygen 10% in feed)
Formation of ethylene oxide (oxygen 15% in feed)
Formation of ethylene oxide (oxygen 20% in feed)
Formation of ethylene oxide (oxygen 25% in feed)
Kinetic model
• Two models for the formation of the
ethylene oxide were developed
– Langmuir-Hinshelwood mechanism
– rate-limiting step: surface reaction
– competitive adsorption of ethylene and
oxygen on the silver surface
– the effects of products and internal diffusion
assumed to be negligible
Kinetic model
• Model 1 assumed a competitive
adsorption of ethylene and dissociatively
adsorped oxygen on the surface
• Model 2 assumed a competitive
adsorption of ethylene and molecular
oxygen on the surface
Kinetic model
• Best fit to the experimental data was
achieved with model 2 and following rate
equation was proposed for the formation
of ethylene oxide:
21)1( OOEE
OEOE
cKcK
ccKkKr
Kinetic model – parity plot
0,0E+00
2,0E-08
4,0E-08
6,0E-08
8,0E-08
1,0E-07
1,2E-07
0,0E+00 2,0E-08 4,0E-08 6,0E-08 8,0E-08 1,0E-07 1,2E-07
Experimental formation of ethylene oxide [mol/s]
Est
ima
ted
fo
rma
tio
n o
f et
hy
len
e o
xid
e [m
ol/
s]
Kinetic model – fit to
experimental data
0,0E+00
2,0E-08
4,0E-08
6,0E-08
8,0E-08
1,0E-07
1,2E-07
0 5 10 15 20 25 30
Concentration of oxygen in feed [%]
Fo
rma
tio
n o
f eth
yle
ne o
xid
e [
mo
l/s]
Formation of ethylene oxide (ethylene 5% in feed)
Formation of ethylene oxide (ethylene 10% in feed)
Formation of ethylene oxide (ethylene 15% in feed)
Formation of ethylene oxide (ethylene 25% in feed)
Ethylene oxide - conclusions
• Ethylene oxide was succesfully
synthesized in a microreactor over silver-
alumina and silver catalysts.
• The microreactor technology was suitable
for studying fast gas-phase heterogeneous
catalytic reaction and many advantages
could be employed e.g. inherent safety.
Microreactors in selective
catalytic reduction of NOx from
biofuels
Catalyst development and
modelling of kinetics
and flow pattern
Set-up
Agilent 3000
Micro GC
F
F
F
F
F
TIC
TIC
6
1 2 3 4 5
6
6
7
8
9
6
11
10
12 13
14 15
16
1 Helium
2 NO/He Gas Mixture
3 Oxygen
4 Calibration Gas (GC)
5 Calibration Gas (NOx Analyzer)
6 Mass Flow Controller (Gas)
7 Mass Flow Controller (Water)
8 Evaporator
9 Microdialysis Pump (Water, Hexadecane)
10 Pre-Heater
11 Microreactor
12 Condenser (-5°C)
13 Condenser (-20°C)
14 Gas Chromatograph
15 NOx Analyzer
16 Computer
Gas-phase microreactor
Inlet I Inlet II
Outlet
Heating/C
ooling
Lower
part
Heating Cartridges
Production of bio-diesel
Renewable source
Natural fats & oils
Biodiesel
Bio
Catalyst
HC-SCR of NOx
C2H4 C8H18
*Hernández Carucci et al. Catalysis Today, Vol. 133-135, (2008) 448
Biofuels-assisted SCR of NOx in
microchannels
First generation biodiesel-FAME
Second generation biodiesel-
Decarboxylation of vegetable oils,
e.g., (NExBTL)
*Hernández Carucci et al. European congress on catalysis VIII, Turku, Finland (2007)
Experimental conditions
• Total flow rate 50 ml/min
• 150 < T < 550 C
• Carrier gas: He
• 500-2000 ppm NO
• 187-1500 ppm Hexadecane
• 3-15% O2
• 12% H2O hexadecane, C16H34
Model paraffin diesel compound (cetane number 100)
Variation of NO concentration
NO to N2 conversion (%) C16H34 to CO2 + CO
conversion (%)
SCR of NOx with hexadecane, O2=6%, 187.5 ppm C16H34, H2O=12%
*Hernández Carucci et al. XVIII International Conference on Chemical Reactors, Malta (2008)
*Hernández Carucci et al. submitted to JCE (2008)
Hexadecane concentration
NO to N2 conversion (%) C16H34 to CO2 + CO
conversion (%)
SCR of NOX with hexadecane, O2=6%, NO=500ppm, H2O=12%
*Hernández Carucci et al. XVIII International Conference on Chemical Reactors, Malta (2008)
*Hernández Carucci et al. submitted to JCE (2008)
Oxygen concentration
NO to N2 conversion (%) C16H34 to CO2 + CO
conversion (%)
SCR of NOX with hexadecane, NO=500ppm, C1/NO=6 H2O=12%
*Hernández Carucci et al. XVIII International Conference on Chemical Reactors, Malta (2008)
*Hernández Carucci et al. submitted to JCE (2008)
Proposed mechanism
C16H34
O2
NO
Ag/Al2O3 surface
Adsorption:
+O* + *
C16H33* + OH*
+ 2*
2O*
+ *
NO*
C16H33* + NO* → C16H33NO* + * → ... →
N2 + CO2 + H2O
C16H33* + O* → ... →
CO2 + H2O
Desorption:
NO concentration
NO3*
NO3*
NO3*
NO3*
NO3*
NO3*
Kinetic analysis
• Adsorption, surface reaction and
desorption steps, were included in the
model.
• Plug flow model used for the microchannel
• Mass balances for every
reactant/intermediate/product
• Based on the data of the components in
the gas phase
HC-SCR Model
JQRP
P: # of basic routes (9)
Q: # balance equations (1)
J: # intermediates (13)
R: # of steps (21)
OHCOOHC 2223416 3432492
OHCOOHC 223416 3432332
22223416 5.0171624 NOHCOONOHC
2223416 5.0171616 NOHCOONOHC
222223416 171623 NOHCOONONOHC
22223416 171615 NOHCOONONOHC
225.0 COOCO
225.0 NOONO
22 222 CONCONO
1.-
2.-
3.-
4.-
5.-
6.-
7.-
8.-
9.-
*Hernández Carucci et al. XVIII International Conference on Chemical Reactors, Malta (2008)
*Hernández Carucci et al. submitted to JCE (2008)
Proposed
model
Step Reaction Rate Equation
1 ** NONO NOvNO krckr 1111
2 **** ONNO vNOkr 22
3 *2*22 OO 2
33
2
33 2 OvO krckr
4 *2*2 2 NN 2
44 Nkr
5 *2** 2 NOONO ONOkr 55
6 ** COCO COvCO krckr 6666
7 *2** 2 COOCO OCOkr 77
8 **** 33163416 OHHCOHC vOhexckr 88
9 **** 33163316 NOHCNOHC NOhexkr 99
10 *2*2 331623316 OHCNNOHC 2
1010 NOhexkr
12 **** 33163316 OHCOHC Ohexkr 1212
13 **2 2 OOHOH 2
1313 OHkr
14 **** 331623316 OOHCNNONOHC NONOhexkr 1414
15 ** 22 NONO 22 15151515 NOvNO krckr
16 **** 331623316 ONOHCNOHC 21616 NOhexkr
17 *** NCOCON CONkr 1717
20 **** 2 ONNON NONkr 2020
21 *** 22 ONON vONkr 22121
Model equations
vNONO pK 1¨ vOO pK 23
vCOCO pK 6 vNONO pK 22 15
22
2
151631219
38
NOONO
vOhex
hexpkkpKkpKk
pKpk
v
NOONO
NONO
OhexNONO
NOhexk
pkkpKkpKk
pkkpkkpkpkkpKkpKk
10
151631219
151619
3810
22
1
2
14114
4
8
22
2
2
v
Ohex
OHk
pkpk
13
38
2
2
v
NONOCONOCO
Nk
pkkkpkkpkkpkkpkk
4
124
2
120617120617
4
8
NOHNOhexhexNOCOONO
v
2
1
1
Adsorption steps Coverage of hexadecane: Coverage of OH:
Coverage of hex-NO:
Coverage of N: Coverage of vacant sites:
*Hernández Carucci et al. XVIII International Conference on Chemical Reactors, Malta (2008)
*Hernández Carucci et al. submitted to JCE (2008)
Mathematical tools (MODEST)
• ODE-solver suitable for
stiff differential
equations. Temperature
dependence expressed
as:
• Simplex-Levenberg-
Marquardt algorithm
minimized the residual
sum of squares.
2
ii yy
ave
aveTTR
Eakk
11exp
aveRT
Ea
ave Aek
Parameter estimation
A total of
30
parameters
estimated!
Pre-exponential
factor/
Equilibrium Constant
Estimated values Activation Energy
[J/mol]
Estimated values
A1 [m3/mol] 2.98 ± 0.04
A2 [mol/m3s] 5.13 ± 0.08 Ea2 (1.02 ± 0.052)E+05
A3 [m3/mol] 13.0 ± 0.17
A4 [mol/m3s] 0.527 ± 0.004 Ea4 (0.785 ± 0.493)E+05
A5 [mol/m3s] 0.0225 ± 0.0143 Ea5 (1.299 ± 0.129)E+05
A6 [m3/mol] 0.962 ± 2.260
A7 [mol/m3s] 2.12 ± 8.73 Ea7 (0.252 ± 0.384)E+05
A8 [mol/m3s] 0.519 ± 0.070 Ea8 (0.386 ± 0.134)E+05
A9 [mol/m3s] 9.74 ± 9.94 Ea9 (0.581 ± 0.404)E+05
A10 [mol/m3s] 25.5 ± 42.5 Ea10 (0.613 ± 0.677)E+05
A12 [mol/m3s] 3.35 ± 15.9 Ea12 (0.291 ± 0.919)E+05
A13 [mol/m3s] 5.14 ± 4.81 Ea13 (0.498 ± 0.304)E+05
A14 [mol/m3s] 11.4 ± 22.9 Ea14 (1.241 ± 0.381)E+05
A15 [m3/mol] 25.4 ± 13.8
A16 [mol/m3s] 3.60 ± 3.94 Ea16 (0.692 ± 0.439)E+05
A17 [mol/m3s] 14.2 ± 44.8 Ea17 (0.887 ± 0.270)E+05
A20 [mol/m3s] 14.3 ± 1.51 Ea20 (0.699 ± 0.250)E+05
Laminar flow with radial diffusion Mass balance in the catalytic layer:
dx
dc
x
s
dx
cdDr
dt
dc iie
pi
i
p
'
2
'2
2
'
, 2,0s and 1,0x
Mass balance in the microchannel:
dl
rwlrcd
dr
lrdc
rdr
lrcdD
dt
lrdc iii
iM
i )(),(),(1),(),(2
2
,
with the axial flow velocity as: 0
2
21)( wR
rrw
Nomenclature:
ci = concentration of species in the bulk phase
c’i = concentration of species in the catalyst layer
De
Effective diffusion coefficient
DM,I Molecular diffusion coefficient
L
= axial coordinate
R =
radial coordinate
ri
= reaction rates
R
= reactor radius
s
= form factor
w(r)
= flow velocity in the l direction
x
=
radial coordinate along the washcoat dep th
thickness of the catalytic layer
p
= porosity of the coating
p
= density of the catalyst particle
Boundary conditions:
Symmetry: 0,0 rdr
dci for all l
No gradients when l=1: 0
1
'
x
i
dx
dc
Diffusion: Rr
iiM
x
ie
dr
dcD
dx
dcD
,
0
'
*Hernández Carucci et al., International Symposium on Chemical Reaction Engineering, Japan (2008)
Profiles in microchannel
*Hernández Carucci et al., International Symposium on Chemical Reaction Engineering, Japan (2008)
Flow patterns inside the microchannels have been
modelled. At higher flows the pressure drop augments.
A model for the RTD has been developed.
Kinetic phenomena have been discovered
Laminar flow with radial diffusion is a good model for
gas-phase microreactors
Summary
Synthesis of chemical
intermediates in microreactors
Sabrina A. Schmidt, Tapio Salmi, Dmitry Murzin,
José Hernández Carucci, Narendra Kumar, Kari Eränen
Teknisk kemi & reaktionsteknik
Åbo Akademi
Methyl and ethyl chloride
ICIS Chemical Business Americas; 3/19/2007, Vol. 271 Issue 11, p50-50, 1p;
Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, New York, 2004.
Pictures: Wikipedia
69
~106 tons/year to important
everyday products
~100.000 tons/year
direct use and ethyl
cellulose production
Production
• Hydrochlorination of ethanol and methanol
• R-OH + HCl → R-Cl + H2O
– Ether as side-product formed
– In case of ethanol also ethylene,
acetaldehyde
– T ~ 300 °C, catalyst: Alumina, Zinc chloride /
Alumina
– Very rapid gas phase reactions
70
Why microreactor: safety
71
• Highly flammable and toxic gases
• Transportation and storage = / a risk and a cost
• Failure (e.g. runaway) of a big unit is dangerous
• Idea: produce alkyl chloride on-site in a microreactor
in the amounts needed
• ”Keep the tiger in the cage!”
Why microreactor: diffusion
• Efficiency: EtCl / MeCl formation are very fast!
– Low diffusion distances
– Increased catalyst and space efficiency
– Ideal tool for kinetic studies
72
Research strategy
• Catalyst studies
• Catalyst coating technique for
microchannels
• Kinetic and thermodynamic investigations
• Mathematical modelling
• Product separation
The microreactor
• IMM GPMR-mix : Gas phase microreactor with mixing and catalyst
zone
• Material: stainless steel
74
Catalysts • Neat Alumina
• Active sites: Lewis acid sites (LAS, e.g. Al3+ centres)
• Alternatively: ZnCl2/Alumina – Introduction of zinc based LAS
• ZnCl2/ Zeolites
– Tunable acidity
77
Effects of zinc support
• Support: 5 wt% Zn on alumina and
zeolites
Zn/H-Beta 25
Zn/H-Beta 150
Zn/H-Beta 300
Zn/alumina
Zn/H-ZSM5
80
Zn/H-ZSM5 280
Zn/H-ZSM5
23
0
10
20
30
40
50
60
70
1 3 5 7 9
Co
nve
rsio
n /
%
LAS / μmol [in reactor]
S.A. Schmidt, N. Kumar, A., Shchukarev, K. Eränen, J.-P. Mikkola, D. Murzin, T. Salmi Preparation and characterization
of neat and ZnCl2 modified zeolites and alumina for methyl chloride synthesis Appl. Catal. A. 2013, 468, 120-134. 78
Effects of zinc support
• Support: 5 wt% Zn on alumina and
zeolites
Zn/H-Beta 25
Zn/H-Beta 150
Zn/H-Beta 300
Zn/alumina
Zn/H-ZSM5
80
Zn/H-ZSM5 280
Zn/H-ZSM5
23
0
10
20
30
40
50
60
70
1 3 5 7 9
Co
nve
rsio
n /
%
LAS / μmol
Zn/H-Beta 25
Zn/H-Beta 150
Zn/H-Beta 300
Zn/alumina
Zn/ H-ZSM5 80 Zn/H-
ZSM5 280
Zn/ H-ZSM5 23
60
65
70
75
80
85
90
95
100
0.3 0.5 0.7 0.9
Se
lec
tivit
y
weak LAS /total LAS
S.A. Schmidt, N. Kumar, A., Shchukarev, K. Eränen, J.-P. Mikkola, D. Murzin, T. Salmi Preparation and characterization
of neat and ZnCl2 modified zeolites and alumina for methyl chloride synthesis Appl. Catal. A. 2013, 468, 120-134. 79
Effects of zinc loading
• Loading: 0-21 wt% Zn on alumina
0
5
10
15
20
25
30
35
40
45
80 100 120 140 160
Meth
an
ol
co
nvers
ion
/ %
LAS / μmol g-1
20.4 wt%
21.1 wt%
19.5 wt%
14.9 wt%
9.4 wt%
2.4 wt%
5 wt%
neat Alumina
S. A. Schmidt, M. Peurla, N. Kumar, K. Eränen, D. Yu. Murzin, T. Salmi, Preparation of selective ZnCl2/alumina
catalysts for methyl chloride synthesis: Influence of pH, precursor and zinc loading (submitted). 80
Effects of zinc loading • Loading: 0-21 wt% Zn on alumina
0
5
10
15
20
25
30
35
40
45
80 100 120 140 160
Me
tha
no
l c
on
ve
rsio
n /
%
LAS / μmol g-1
20.4 wt%
21.1 wt%
19.5 wt%
14.9 wt%
9.4 wt%
2.4 wt%
5 wt%
neat Alumina
S. A. Schmidt, M. Peurla, N. Kumar, K. Eränen, D. Yu. Murzin, T. Salmi, Preparation of selective ZnCl2/alumina
catalysts for methyl chloride synthesis: Influence of pH, precursor and zinc loading (submitted). 81
Catalyst of choice • Activity and selectivity can be improved by addition of zinc chloride
• Zeolites are the most active but least stable and selective catalysts
• Zinc modified alumina is stable in the tubular reactor but selectivity
decreases in the microreactor
• Neat alumina is least active but selective, stable and inexpensive
catalyst
• Catalyst of choice
82
Catalyst coating
• Binder-free slurry coating method
• Adhesion through:
– Thermal surface treatment
– Ball milled catalyst (<32 µm)
• Amount of catalyst in one microreactor:
3.4 mg
S. A. Schmidt, N. Kumar, B. Zhang, K. Eränen, D. Yu. Murzin, and T. Salmi, Preparation and Characterization of Alumina-Based
Microreactors for Application in Methyl Chloride Synthesis, Ind. Eng. Chem. Res. 2012, 51, 4545
83
84
C h a r a c t e r i s a t i o n o f c o a t i n g
S. A. Schmidt, N. Kumar, B. Zhang, K. Eränen, D. Yu. Murzin, and T. Salmi, Preparation and Characterization of Alumina-Based
Microreactors for Application in Methyl Chloride Synthesis, Ind. Eng. Chem. Res. 2012, 51, 4545
• Confocal microscopy: Morphology, thickness and surface roughness
• Coating thickness : 15 μm, channel depth: 90 μm
Methanol hydrochlorination
• Hydrochlorination of methanol at 300 °C
• Lightly exothermic, main reaction: -30 kJ/mol
• The reactions are not completely irreversible!
86
ClCHOHCHHClOCHCH
OHOCHCHOHCH
OHClCHHClOHCH
3333
2333
233
2
(I)
(II)
(III)
Keq
398
12
36
S. A. Schmidt, N. Kumar, A. Reinsdorf, K. Eränen, J. Wärnå, D., Murzin, T., Salmi, Methyl chloride synthesis on Al2O3 in a
microstructured reactor – thermodynamics, kinetics and mass transfer, Chem. Eng. Sci 2013, 95, 232-245.
Performance of microreactors
87
70
75
80
85
90
95
100
280 300 320 340S
ele
cti
vit
y t
ow
ard
s M
eC
l / %
Temperature / °C
one reactor
two reactors
• A very high conversion and selectivity can be reached
with two microreactors
Methanol conversion Selectivity towards methyl chloride
S. A. Schmidt, Z. Vajglova, K. Eränen, D. Murzin, T. Salmi, Microreactor technology for on-site production of methyl chloride. Green
Process. Synth. 2014, Advance online publication, DOI: 10.1515/gps-2014-0039
Reaction modeling
-catalyst layer
88
• Kinetic model: Langmuir-Hinshelwood
• Plug flow model for the reactor
2
1
233
11
)(
D
K
cccc
kr
OHClCHHClOHCH
2
2
22
22
)(
D
K
ccc
kr
OHDME
MeOH
2
333
)(
D
K
cccc
kr
MeClMeOHHClDME
1 HClHClcKD
S. A. Schmidt, N. Kumar, A. Reinsdorf, K. Eränen, J. Wärnå, D., Murzin, T., Salmi, Methyl chloride synthesis on Al2O3 in a
microstructured reactor – thermodynamics, kinetics and mass transfer, Chem. Eng. Sci 2013, 95, 232-245.
0
0.4
0.8
1.2
0 0.02 0.04 0.06 0.08 0.1
c(M
eC
l) [m
ol/
m3
]
residence time [s]
280 °C
300 °C
320 °C
340 °C
0
0.02
0.04
0.06
0.08
0.1
0 0.05 0.1c
(DM
E)
[m
ol/
m3
] residence time [s]
DME MeCl
• Detailed description of MeCl formation
• DME formation shows deviation
Significantly lower concentration, rough description of reaction 3
Kinetic model
89 S. A. Schmidt, N. Kumar, A. Reinsdorf, K. Eränen, J. Wärnå, D., Murzin, T., Salmi, Methyl chloride synthesis on Al2O3 in a
microstructured reactor – thermodynamics, kinetics and mass transfer, Chem. Eng. Sci 2013, 95, 232-245.
Reaction modeling
- catalyst layer
90
• Obtained activation energy for MeCl formation is double of previously
published
• Suggests internal diffusion limitations
• Diffusion modelling in the catalyst layer
• Mean transport pore model, Catalyst shape: slab
i
p
p
ei DD )(
S. A. Schmidt, N. Kumar, A. Reinsdorf, K. Eränen, J. Wärnå, D., Murzin, T., Salmi, Methyl chloride synthesis on Al2O3 in a
microstructured reactor – thermodynamics, kinetics and mass transfer, Chem. Eng. Sci 2013, 95, 232-245.
)(1
2
2
pii
ei
p
i rdr
cdD
dt
dc
Reactant concentration profiles
inside the catalyst layer
91
Conventional
fixed bed
Effectiveness factor:
0.1-0.06 (low)
Microreactor
Effectiveness factor:
0.93 (high)
• Wrong activation energies reported in literature
Ethanol hydrochlorination
EtOH (+HCl)
EtCl
EtOEt
C2H4
CH3CHO + H2
+ HCl
-H2O
- H2O - HCl
(At high T)
-EtOH
+ HCl
- H2O
- H2O
Reversible reactions
92 S. A. Schmidt, Q. Balme, N. Gemo, N. Kumar, K. Eränen, D. Yu. Murzin, T. Salmi, Kinetics of ethanol
hydrochlorination over γ-Al2O3 in a microstructured reactor (Chemical Engineering Science).
EtOH conversion
0
10
20
30
40
50
60
70
80
90
100
0.00 0.05 0.10 0.15 0.20
Co
nve
rsio
n [
%]
Residence time [s]
T=
240
260
280
300
°C
93 S. A. Schmidt, Q. Balme, N. Gemo, N. Kumar, K. Eränen, D. Yu. Murzin, T. Salmi, Kinetics of ethanol
hydrochlorination over γ-Al2O3 in a microstructured reactor (submitted).
0
5
10
15
20
0 20 40 60 80
C2
H4
O e
lect
ivit
y
EtOH Conversion
Selectivity to side products
• DEE: decrease with time and increases
with temperature
• C2H4: increase with temperature and
time ( at T≥300°C )
DEE + HCl → EtCl+ EtOH
DEE → C2H4 + EtOH ( at high T)
EtCl→ C2H4 + HCl ( at high T)
• Acetaldehyde: very low
0
5
10
15
20
0 50 100
DEE
se
lect
ivit
y
EtOH Conversion
T=
240
260
280
300
°C
DEE C2H4
C2H4O
94 S. A. Schmidt, Q. Balme, N. Gemo, N. Kumar, K. Eränen, D. Yu. Murzin, T. Salmi, Kinetics of ethanol
hydrochlorination over γ-Al2O3 in a microstructured reactor (Chemical Engineering Science).
Kinetic model
• Reactions in kinetic model :
1) EtOH + HCl ↔ EtCl + H2O
2) 2 EtOH ↔ DEE + H2O
3) EtOH ↔ C2H4 + H2O
4) EtOH ↔ CH3CHO + H2
5) DEE + HCl ↔ EtCl+ EtOH
2
2
11D
K
cccc
kr I
OHEtClHClEtOH
)1( HClHClEtOHEtOH cKcKD
95
• Langmuir-Hinshelwood :
S. A. Schmidt, Q. Balme, N. Gemo, N. Kumar, K. Eränen, D. Yu. Murzin, T. Salmi, Kinetics of ethanol
hydrochlorination over γ-Al2O3 in a microstructured reactor (submitted).
Model fit
• Variation of concentration profiles with
temperature was well described
• Variation of ethanol/HCl concentration:
– Standard Langmuir-Hinshelwood model does
not describe well the product distribution
96 S. A. Schmidt, Q. Balme, N. Gemo, N. Kumar, K. Eränen, D. Yu. Murzin, T. Salmi, Kinetics of ethanol
hydrochlorination over γ-Al2O3 in a microstructured reactor (submitted).
HCl & ether and ethene formation
• Solution: C2H4 and DEE formation are catalyzed by HCl
on the alumina surface due to increased acidity
• Expression:
2
22
22D
K
ccc
kr II
OHDEEEtOH
**
*
** )1(* EtOHHCl
HCl
cc
c
HClc
α: Parameter : impact of HCl catalysis
For cHCl or α → 0 the term approaches 1
97
Improved kinetic model
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 0.05 0.1 0.15 0.2
EtC
l co
nce
ntr
atio
n [
mo
l/m
3]
Residence time [s]
a)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0 0.05 0.1 0.15 0.2
DEE
co
nce
ntr
atio
n [
mo
l/m
3]
Residence time [s]
b)
0
0.01
0.02
0.03
0.04
0.05
0.06
0 0.05 0.1 0.15 0.2
Eth
yle
ne
co
nce
ntr
aito
n [
mo
l/m
3]
Residence time [s]
c)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0 0.05 0.1 0.15 0.2
Eth
yle
ne
co
nce
ntr
aito
n [
mo
l/m
3]
Residence time [s]
d)
Improved /precise description
Dependence of product
distribution on reactant
concentration is improved but
not precise
Exact dependence of kinetics
on catalyst surface is complex
EtCl DEE
Ethylene Ethylene
98
Product separation
• Aim: At the outlet of the reactor: only
traces of MeOH, HCl and DME due to
maximum conversion
• Methanol and water separation by
condensation
• Glass made condenser, coolant:
glycerin/water -10 °C
• Ongoing work with metal condenser Cooling surface: 210 cm2
99 S. A. Schmidt, Z. Vajglova, K. Eränen, D. Murzin, T. Salmi, Microreactor technology for on-site production of
methyl chloride. Green Process. Synth. 2014, Advance online publication, DOI: 10.1515/gps-2014-0039.
• Composition of the gas phase at maximum conversion (97.6%)
• MeCl and DME are efficiently separated from the liquid
Efficiency: gas phase
/wt%
2.5 1.6
95.9
Prior to condenser
MeOH
DME
MeCl
/wt%
0.12 0.74
99.14
After condenser
100 S. A. Schmidt, Z. Vajglova, K. Eränen, D. Murzin, T. Salmi, Microreactor technology for on-site production of
methyl chloride. Green Process. Synth. 2014, Advance online publication, DOI: 10.1515/gps-2014-0039.
Efficiency: liquid phase
101
Composition of the liquid phase phase (at 83.3 % conversion)
48
18.2
33.8
total condensation (calc.)
58
13
29
0.04
Dimroth condenser
H2O
MeOH
HCl
MeCl+DME
• Methanol cannot be completely separated
• Largest part of HCl is contained in the condensate
• Water is well separated
• Only trace amounts of MeCl and DME in the condensate
S. A. Schmidt, Z. Vajglova, K. Eränen, D. Murzin, T. Salmi, Microreactor technology for on-site production of
methyl chloride. Green Process. Synth. 2014, DOI: 10.1515/gps-2014-0039.
Conclusions
• Neat alumina is the most stable catalyst
• Binder free slurry coating method for stable and
uniform catalyst coating
• Microreactor suppresses severe diffusion
limitations in methanol and ethanol
hydrochlorination
• Detailed kinetic models were developed for
methanol and ethanol hydrochlorination
• Separation of MeCl and DME from water,
methanol and HCl is efficient at high conversion
(97.6 % conversion; > 99 wt% MeCl)
102
Microprocess technology is a great challenge
Microprocess technology is strongly multidisciplinary:
manufacturing, characterization of materials, screening,
kinetics, mass and heat transfer, flow measurement,
modelling
Take the challenge, search for new applications !
Summary
Thank you!
Internship students: Arne Reinsdorf, Zuzana Vajglova and Quentin Balme
Financial support: