Yu Zhengnan 201406 MSC PDF

Ethane Dehydrogenation Using a Catalytic Membrane Reactor
by
Zhengnan Yu
A thesis submitted in partial fulfillment of the requirements for the degree of
Master of Science
in
Chemical Engineering
Department of Chemical and Materials Engineering

University of Alberta
© Zhengnan Yu, 2014

Abstract
The steam cracking of hydrocarbons is the dominant technology for ethylene
production and is a highly energy intensive process. The increasing demand for
ethylene has stimulated substantial research into the development of new process
routes to reduce energy consumption and other costs. Catalytic dehydrogenation of
ethane using a membrane reactor is an attractive solution because the cracking
equilibrium can be shifted in favor of ethylene by selectively removing hydrogen.
Lower temperatures can thus be used to generate comparable ethylene yields. Using
natural mordenite as membrane and Pt/Al2O3 as catalyst, a reactor with membrane
area to reactor volume ratio of 0.16 m-1 improved ethylene yield by 15.6%
comparing to the conventional packed-bed reactor at 500°C. A novel Pt-Zn/ETS-2
catalyst for ethane dehydrogenation was also developed. With this catalyst, unlike
the Pt/Al2O3, side reactions could be completely suppressed and the ethylene
selectivity could reach 100% while conversion was at equilibrium.
i
Acknowledgements
I would like to thank my supervisor Professor Steven M. Kuznicki for his
directions, guidance, and support. Without his help and supports, this thesis would
not have been possible.
I appreciate all of the help I have received from Dr. Kuznicki’s group.
Specifically, I wish to thank Dr. James Sawada for his guidance and support of my
research as well as his help in the revision of my thesis. I also wish to thank Dr.
Adolfo Avila for valuable discussions, assistance with data analysis, and help in
conducting the membrane reactor testing. I want to extend my thanks Weizhu An
for her insights, suggestions, and varieties of help in the lab. I also wish to thank
Lan Wu and Tong Qiu for their support of my research efforts in the lab. I would
like to extend my appreciation to Albana Zeko and Dr. Tetyana Kuznicki for their
efforts to improve my technical writing skills.
Jing Shen, PhD candidate in Dr. Natalia Semagina’s group, conducted the
TPR tests for me and helped guide the XPS data analysis. Dr. Lianhui Ding
graciously offered to help revise my thesis and provide valuable suggestions about
both catalyst characterization and technical writing. I appreciate their valuable
assistance.
I also thank Lily Laser, Graduate Administrative Assistant, for her kind
assistance through the final stages of my thesis.
ii
Table of Contents
Table of Contents
List of Tables ....................................................................................................... vii
List of Figures ..................................................................................................... viii
Chapter 1. Introduction........................................................................................ 1
1.1 Overview of Global Ethylene Production ......................................................... 1
1.2 Feedstock of Ethylene Production .................................................................... 6
1.3 Technologies for Ethylene Production.............................................................. 9
1.4 Challenges for Tube Furnace Pyrolysis Process ............................................. 12
1.4.1 Coking ................................................................................................... 12
1.4.2 Energy Consumption of Pyrolysis Furnace .......................................... 14
1.5 Catalytic Dehydrogenation of Ethane to Ethylene by Membrane Reactor ..... 15
1.5.1 Gas Separation Mechanisms of Porous Inorganic Membranes ............ 15
1.5.2 Zeolite Membrane Reactor Configurations .......................................... 19
1.6 Research Objectives ........................................................................................ 23
Chapter 2. Characterizations and Product Analysis ....................................... 24
2.1 Characterization methods................................................................................ 24
2.1.1 X-ray diffraction (XRD) ....................................................................... 24
2.1.2 Pore Size Distribution, Specific Surface Area, and Pt Particle Dispersion
.............................................................................................................. 24
2.1.3 X-ray photoelectron spectroscopy (XPS) ............................................. 25
2.1.4 TPR ....................................................................................................... 26
iii
Table of Contents
2.1.5 SEM and EDX ...................................................................................... 26
2.1.6 TEM ...................................................................................................... 26
2.1.7 TGA and TG-MS .................................................................................. 27
2.1.8 Dilatometry test..................................................................................... 28
2.2 Gas Product Analysis ...................................................................................... 28
Part I: The Use of Natural Zeolite Mordenite in a Membrane Reactor for
Ethane Dehydrogenation ........................................................................ 29
Chapter 3. Characterization of Natural Mordenite Zeolite ............................ 30
3.1 Introduction ..................................................................................................... 30
3.2 Channel Sizes of Zeolite Mordenite................................................................ 31
3.3 XRD ................................................................................................................ 34
3.4 Scanning Electron Microscopy (SEM) ........................................................... 35
3.5 Thermal Analysis ............................................................................................ 36
3.5.1 Thermogravimetric Analysis (TGA)..................................................... 36
3.5.2 Dilatometry Test ................................................................................... 38
3.6 BET of Mordenite ........................................................................................... 41
3.7 Chemical compositions of natural mordenite zeolite ...................................... 46
Chapter 4. Packed Bed Membrane Reactor for Ethane Dehydrogenation ... 49
4.1 Introduction ..................................................................................................... 49
4.2 Experimental ................................................................................................... 53
4.2.1 The Catalyst .......................................................................................... 53
4.2.2 Membrane Preparation .......................................................................... 53
iv
Table of Contents
4.2.3 Membrane Reactor Configuration ........................................................ 54
4.2.4 Experiment Conditions ......................................................................... 55
4.2.5 Switching Membrane Reactor Mode to Conventional Reactor Mode .. 55
4.2.6 Mass Balance of Species at a Reactor and Membrane Reactor Mode .. 56
4.3 Results and Discussions .................................................................................. 60
4.3.1 Reversibility in the Reactor-Membrane Reactor Mode ........................ 60
4.3.2 The Conventional Reactor and Membrane Reactor Modes: Dynamics of
Molar Fractions of Reaction Species at the Reactor Outlet ................. 61
4.3.3 Increasing the Ratio of Permeation Area/Reactor Volume (𝑨/𝑽𝑹) of the
Membrane Reactor ............................................................................... 64
4.3.4 Effect of Reaction Temperatures .......................................................... 67
4.3.5 Membrane Reactor Effectiveness ......................................................... 69
4.4 Conclusions ..................................................................................................... 70
Part II: Ethane Dehydrogenation Catalyst Development ............................... 71
Chapter 5. Ethane Dehydrogenation Catalyst Development .......................... 72
5.1 Introduction ..................................................................................................... 72
5.2 Experimental ................................................................................................... 79
5.2.1 Catalyst Preparation .............................................................................. 79
5.2.2 Catalysts Testing in Ethane Dehydrogenation Reaction ....................... 80
5.3 Results and Discussion ................................................................................... 81
5.3.1 Catalyst Characterization ...................................................................... 81
v
Table of Contents
5.3.2 Effect of Main Metals on the Activity and Selectivity of Ethane
Dehydrogenation .................................................................................. 95
5.3.3 Effect of the Supports on the Activity and Selectivity of Ethane
5.3.4 Effect of Promoters on the Activity, Selectivity and Stability of Ethane
5.4 Conclusions ................................................................................................... 112
Chapter 6. Conclusions and Suggestions for Future Work .......................... 113
6.1 General conclusions ...................................................................................... 113
6.2 Suggestions for future work .......................................................................... 114
References .......................................................................................................... 117
vi
Table of Contents
List of Tables
Table 1-1 Main technology licensors for ethylene production ............................. 12
Table 1-2 Transportation regimes in porous membrane ....................................... 16
Table 3-1 Chemical composition of natural zeolite mordenite ............................. 47
Table 4-1 Dimensionless numbers defined for the comparison between reactor and
membrane reactor mode ....................................................................... 59
Table 4-2 Experimental values of the dimensionless numbers defined in Table 4-1
for R and MR modes at 500 °C and 550 °C ......................................... 64
Table 4-3 Ethylene yield enhancements achieved with the membrane reactor at two
different reaction temperatures and two 𝐴/𝑉𝑅 ratios. The corresponding
Damköhler-Peclet numbers were calculated with reaction conditions.
WHSV = 0.4 h-1 .................................................................................... 70
Table 5-1 A product yield comparison using different feedstocks [13] .................. 73
Table 5-2 BET Surface Area and Pt Dispersion of Catalysts ............................... 82
vii
Table of Contents
List of Figures
Figure 1-1 The Main Applications of Ethylene ...................................................... 2
Figure 1-2 Global Ethylene Demand by End Use .................................................. 3
Figure 1-3 Regional Ethylene Capacity Growth ..................................................... 5
Figure 1-4 Global Ethylene Production by Feedstock [6] ........................................ 6

[11]
Figure 1-5 U.S. Ethylene Feedstock Sources (Percent of Ethylene Production)
.............................................................................................................. 7
Figure 1-6 Global Ethylene Cash Cost by Region .................................................. 8
Figure 1-7 Global Ethylene Cash Cost Index: 2010 [12] .......................................... 9
Figure 1-8 Steam Cracking Process ...................................................................... 10
Figure 1-9 Optimal conditions of ethylene production ......................................... 11
Figure 1-10 Types of Zeolite Catalytic Membrane Reactors................................ 22
Figure 3-1 Localities of natural zeolite mordenite ................................................ 31
Figure 3-2 Framework and ring sizes or mordenite .............................................. 33
Figure 3-3 One-dimensionally accessible channel system in mordenite .............. 33
Figure 3-4 XRD patterns of natural zeolite mordenite before calcination............ 34
Figure 3-5 XRD patterns of natural zeolite mordenite treated at different
temperatures for 4 hours .................................................................... 35
Figure 3-6 SEM images of cross-section of natural zeolite mordenite ................. 36
Figure 3-7 TGA profile of natural mordenite ....................................................... 38
Figure 3-8 Dilatometry test of natural zeolites mordenite .................................... 39
viii
Table of Contents
Figure 3-9 Refined natural mordenite structure viewed along [001] at different
temperatures: (a) 25 °C, (b) 200 °C, (c) 450 °C, (d) 830 °C .............. 41
Figure 3-10 BET surface area of natural mordenite treated at different temperatures
............................................................................................................ 42
Figure 3-11 Pore size distribution of natural mordenite calculated by HK method
............................................................................................................ 44
Figure 3-12 Pore size distribution of natural mordenite calculated by BJH method
............................................................................................................ 46
Figure 4-1 Illustration of membrane reactor concept for ethane dehydrogenation50
Figure 4-2 Steps of membrane disk preparation ................................................... 53
Figure 4-3 Packed-bed membrane reactor configuration ...................................... 54
Figure 4-4 Schematic demonstration of switching between membrane reactor and
conventional reactor mode ................................................................. 56
Figure 4-5 Dynamics of molar fractions of H2, C2H4, and C2H6 in the reactor outlet
as the system switched reversibly between a membrane reactor and a
reactor operating mode at 500 °C, WHSV = 0.4 h-1 .......................... 61
Figure 4-6 Dynamics of outlet molar fractions of C2H6, H2, and C2H4 between
membrane reactor and reactor mode at 500 °C, WHSV = 0.4 h-1...... 62
Figure 4-7 Dynamics of outlet molar fractions of H2 and C2H4 between membrane
reactor and reactor mode at 500 °C. , WHSV = 0.4 h-1 ..................... 63
ix
Table of Contents
reactor and membrane reactor mode at 500 °C, WHSV = 0.4 h-1 with
bigger membrane area (𝐴/𝑉𝑅 ≅ 0.16) ............................................... 66
reactor and membrane reactor mode at 550 °C, WHSV = 0.4 h-1...... 68
Figure 5-1 The scheme of packed-bed reactor for ethane dehydrogenation ......... 81
Figure 5-2 Pt mean particle size and dispersion change with Zn content ............. 83
Figure 5-3 Pt particle size change with dispersion on catalysts............................ 84
Figure 5-4 XRD patterns of ETS-2 dried at 150 °C and calcined at 600 °C for four
hours ................................................................................................... 85
Figure 5-5 XRD patterns of: (a) ETS-2 dried at 150 °C , (b) ETS-2 treated at 600 °C,
(c) 5%Zn, 1%Pt/ETS-2 (500 °C calcined), (d) 5%Zn, 1%Pt/ETS-2
(500 °C tested), (e) 5%Zn, 1%Pt/ETS-2 (600 °C calcined), (f) 5%Zn,
1%Pt/ETS-2 (600 °C tested) .............................................................. 86
Figure 5-6 Industrial ETS-2 (a) dried at 150°for 4 hours; (b) calcined at 600 °C for
4 hours ................................................................................................ 87
Figure 5-7 Catalyst 5%Zn, 1%Pt/ETS-2 (a) calcined at 500 °C for 4 hours; (b) tested
at 500 °C............................................................................................. 87
Figure 5-8 Catalyst 5%Zn, 1%Pt/ETS-2 (a) calcined at 600 °C for 4 hours; (b) tested
at 650 °C............................................................................................. 88
Figure 5-9 Catalyst 1%Pt/ETS-2 calcined at 500 °C for 4 hours ......................... 88
x
Table of Contents
Figure 5-10 TEM images of the catalyst 1%Pt, 5%Zn/ETS-2: (a) calcined at 500 °C,
(b) spent (run at 500 °C), (c) calcined at 600 °C, (d) spent (run at 600 °C)
............................................................................................................ 89
Figure 5-11 XPS of the Zn 2p3/2 of catalyst 5%Zn, 1%Pt/ETS-2 calcined at 600 °C
............................................................................................................ 91
Figure 5-12 XPS of the Zn 2p3/2 of catalyst 5%Zn, 1%Pt/ETS-2 calcined at 500 °C
............................................................................................................ 92
Figure 5-13 H2-TPR profiles of ETS-2 supported catalysts ................................. 93
Figure 5-14 TG-MS analysis of catalyst 5%Zn, 1%Pt/ETS-2 tested at 650 °C ... 94
Figure 5-15 TG-MS analysis of catalyst 1%Zn, 1%Pt/Al2O3 tested at 500 °C .... 95
Figure 5-16 Product spectrum obtained with the 1%Pd/ETS-2 and 1%Pt/ETS-2
catalysts at 509 °C, and ethane WHSV = 0.28h-1 .............................. 96
Figure 5-17 Product yield change with space velocity at 509 °C over catalysts
1%Pd/ETS-2 and 1%Pt/ETS-2........................................................... 97
Figure 5-18 Product spectrum obtained with the 1%Pt/ETS-2, 1%Pt/Al2O3, and
1%Pt/clinoptilolite catalysts at 509 °C, and ethane WHSV = 0.28 h-1
............................................................................................................ 98
Figure 5-19 Product concentration change with WHSV at 509 °C over catalysts
1%Pt/Al2O3 and 1%Pt, 1%Zn/Al2O3 ................................................. 99
Figure 5-20 Product concentration changes with time over 1%Pt/ETS-2 at 509 °C,
WHSV = 0.80 h-1 ............................................................................. 100
xi
Table of Contents
Figure 5-21 Product concentration changes with time over 0.5%Zn, 1%Pt/ETS-2 at
509 °C, WHSV = 1.13 h-1 ................................................................ 101
Figure 5-22 Product concentration changes with time over 1%Zn, 1%Pt/ETS-2 at
509 °C, WHSV = 0.88 h-1 ................................................................ 102
Figure 5-23 Product concentration changes with time over 1%Zn, 1%Pt/ETS-2 at
509 °C, WHSV = 1.14 h-1 ................................................................ 102
Figure 5-24 Product yield changes with time over 2%Zn, 1%Pt/ETS-2 at 509 °C,
WHSV = 0.87 h-1 ............................................................................. 103
WHSV = 1.13 h-1 ............................................................................. 104
WHSV = 0.89 h-1 ............................................................................. 105
Figure 5-27 Product yield changes with time over 7.5%Zn, 1%Pt/ETS-2 at 509 °C,
WHSV = 0.86 h-1 ............................................................................. 105
Figure 5-28 Product concentration change with Zn content on catalyst 1%Pt/ETS-
2 at 509 °C, WHSV = 0.87h-1 .......................................................... 106
Figure 5-29 C2H4, H2, and CH4 yield change with space velocity on different
content of Zn modified catalysts at 509 °C ...................................... 107
Figure 5-30 Effects of Zn content of catalysts 1%Pt/ETS-2 on selectivity of
ethylene at WHSV=0.87 h-1 and 1.13h-1, 509 °C ............................. 108
Figure 5-31 C2H4, H2, and CH4 yields change with WHSV at 509 °C over
1%Pt/ETS-2 modified with different promoters .............................. 109
xii
Table of Contents
Figure 5-32 C2H4, H2, and CH4 yields change with space velocities at 509 °C over
1%Pt/Al2O3 catalyst modified with Zn and Ce promoters ............. 110
Figure 5-33 Product spectrum obtained with the 1%Pd, 1%Sn/Al2O3 and 1%Pd,
1%Ce, 1%Zn, 1%Sn/Al2O3 catalysts at 509 °C, and ethane WHSV =
0.28 h-1 ............................................................................................. 111
xiii
Table of Contents
Abbreviations
Symbol Description
AGO Atmospheric gas oil
BET Brunauer-Emmett-Teller
BJH Barret-Joyner-Halenda
EB Ethylbenzene
EDC Ethylene dichloride
EDX Energy dispersive x-ray
EO Ethylene oxide
ETS-2 Engelhard titanosilicate-2
GC Gas chromatography
HDPE High density polyethylene
HK Horvath-Kawazoe
ID Inner diameter, mm
LDPE Low density polyethylene
LLDPE Linear low density polyethylene
LPG Liquefied petroleum gas
OD Outer diameter, mm
PBMR Packed-bed membrane reactor
PE Polyethylene
PVC Polyvinyl chloride
SEM Scanning electron microscopy
xiv
Table of Contents
TCD Thermal conductivity detector
TEM Transmission electron microscopy
TGA Thermogravimetric analysis
TG-MS Thermogravimetric analysis-mass spectrometry
TPR Temperature-programmed reduction
WHSV Weight hourly space velocity, h-1
XPS X-ray photoelectron spectroscopy
XRD X-ray diffraction
Symbols
Symbol Description Unit

𝑐 Speed of light 𝑚/𝑠
𝐶 BET constant
𝑑 Pore diameter 𝑚
𝑑𝐴 Diameter of adsorbate molecule 𝑚
𝑑𝑝 Particle diameter 𝑚
𝑑𝑠 Diameter of adsorbent molecule 𝑚
𝐸1 Heat of adsorption for the 1st layer 𝐽
𝐸𝐿 Heat of adsorption for the 2nd and higher layers, 𝐽
heat of liquefaction
𝐹𝐴 Permeation flux of gas species A 𝑚𝑜𝑙⁄(𝑚2 ∙ 𝑠)
𝐹𝐾 Permeance of membrane 𝑚𝑜𝑙⁄(𝑚2 ∙ 𝑠 ∙ 𝑃𝑎)
𝑘 Boltzmann constant, 𝐽/𝐾
xv
Table of Contents
𝐾 Avogadro’s number, 𝑚𝑜𝑙−1

𝑙 Distance between two layers of adsorbent 𝑚
𝐿 Thickness of porous medium 𝑚
𝑚 Mass of an electron 𝑘𝑔
𝑀𝐴 Molecular weight of gas species A 𝑔/𝑚𝑜𝑙
𝑁𝐴 Number of molecules per unit area of adsorbate
𝑁𝑠 Number of atoms per unit area of adsorbent
𝑃 Pressure, or equilibrium pressure 𝑃𝑎
𝑃0 Saturation pressure of adsorbates at the 𝑃𝑎
temperature of adsorption
𝑟 Radius of pore 𝑚
𝑅 Gas law constant 𝐽⁄(𝑚𝑜𝑙 ∙ 𝐾)
𝑇 Absolute temperature 𝐾
𝑊 Weight of gas adsorbed at a relative pressure 𝑔
𝑊𝑚 Weight of adsorbate constituting a monolayer of 𝑔
surface coverage
𝛼𝐴 Polarizability of adsorbate
𝛼𝑠 Polarizability of adsorbent
𝜀 Porosity
𝜆 Mean free path 𝑚
𝜏 Tortuosity
𝜒𝐴 Magnetic susceptibility of adsorbate
𝜒𝑠 Magnetic susceptibility of adsorbent
xvi
Chapter 1. Introduction
1.1 Overview of Global Ethylene Production
Ethylene is a versatile petrochemical intermediate, and it plays a vital role to
modern society. As Figure 1-1 shows, a wide series of daily items such as plastic
bags, milk and food cartons, washing-up liquids, paints, antifreeze, toys, pipes,
window frames, car components, and many other industrial materials are produced
from ethylene. Based on Chemical Market Associates Inc. (CMAI) data (Figure
1-2), the majority of global ethylene is for the production of polyethylene (PE),
which is primarily used in a wide variety of non-durable goods applications, such
as packaging materials. Ethylene oxide (EO), ethylene dichloride (EDC), and
ethylbenzene (EB)/styrene are also important ethylene derivatives. The figure also
shows a continuous increase of ethylene demand in the near future as PE feedstock.
According to Oil & Gas Journal’s survey, global ethylene capacity has been
increasing since 1995 or earlier [1], as shown in Figure 1-3. Asia-Pacific and Middle
East were the two fastest increasing regions. The ethylene capacity in North
America and West Europe has been almost unchanged in recent years. Survey data
[2]
revealed that as of January 1, 2013 the global ethylene production was more than
143 million tons versus 141 million tons in 2012. Formosa Petrochemical Corp. had
the largest ethylene complex in Mailiao, Taiwan with the capacity of 2.94 million
tons. Nova Chemicals Corp. had the second largest complex in Joffre, Alberta with
1
the capacity of 2.81 million tons. Complex capacity varied yearly, but the rank did
not change since 2010 [1-3].
packaging film,
LDPE,
toys, diapers,
LLDPE
housewears
polymerization
bottles, food
HDPE containers, drums,
crates, housewears
ethylene-
co-polymerization propylene rubbers, paint
rubber
oligomerization 1-butene monomer
surfactant
ethylene ethylene
oxide glycol
oxidation
ethanol-
Natural gas amines
Ethylene
/Crude oil
acetic fibers,
acet-
polyster resin,
aldehyde acid misc.
pipes,
ethylene vinyl
chlorination PVC window
dichloride chloride
frames
polystyrene, styrene
ethyl-
alkylation styrene acrylonitrile, styrene
benzene butadiene rubber
hydration ethanol
linear
detergents
alcohols
adhesives, coatings,
vinyl acetate textile finishing,
flooring
Figure 1-1 The Main Applications of Ethylene
Saudi Basic Industries Corp. (SABIC) replaced Dow Chemical Co. in 2010 to
become the largest ethylene producer with the total capacity of 23.67 million tons
2
per year. Dow Chemical Co. became the second largest one worldwide with the
total capacity of 23.57 million tons per year. The capacity rank of the two
companies did not change since 2010. As of January 2013, the United States was
the largest producer, with the capacity of 28.12 million tons. China and Saudi
Arabia were the second and third ones, with the capacity of 13.78 and 13.16 million
tons respectively. Canada was the seventh with the capacity of 5.53 million tons.
PE EO EDC/VCM/PVC Styrene/EB Other
160
Ethylene demand, million tons
140
120 14.3 14.3
15.1 15.1
13.6 21.1 22.6
100 13.6 13.6 21.1
14.3 11.3 18.1 19.6
80 14.3 13.6 16.6 17.4
14.3 14.3 14.3 16.6
60
86.8 92.1 95.8
40 68.7 74.0 75.5 79.2 82.3
64.2 64.2 67.2
20
0
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
Source: CMAI, 2012 World Light Olefins Analysis

Figure 1-2 Global Ethylene Demand by End Use
As shown in Figure 1-3, global ethylene capacity reached 147.7 million tons
in 2011, and was predicted at over 170 million tons in 2016 at 3.4% annual growth
[4]
. Relying on its huge market advantage and rapid growth of demand, Asia-Pacific
becomes a region attractive to investors for ethylene production. Even though the
newly added production capacity was small, the advantage of low production costs
3
could make, Middle East the world’s most investment-intensive region in the
ethylene industry, and the main ethylene exporting area [5].
4
North America West Europe Middle East Asia-Pacific Rest of World

180
160
21.2 23.5
19.4
Production capacity, million tons
140 18.8 18.8

18.2
17.1 16.5
120 16.5
16.5 52.4 53.5
15.3 46.5 50.0
14.7 15.3 43.5 45.3
100 14.7
14.7 36.5 37.6
14.1 32.4 35.3
12.4 12.9 30.6
80 12.4 27.1 28.2 28.8
11.8 10.6 24.7
22.4 22.9 26.5 29.4 28.2 30.6 30.6
21.8 12.9 15.3 17.6 21.2 25.9
60 19.4 19.4 10.6 10.6 10.0 12.4
17.1 7.1 9.4
5.9 5.9
4.7 4.1 6.5
22.4 22.9 23.5 24.7 23.5 24.7 23.5 24.7 24.1 24.1 24.7 22.9 24.1 23.5 24.1
40 20.6 21.8 21.8
19.4 20.0 19.4
20 34.1 33.5 34.1 34.1 34.7 34.1 33.5

27.1 27.6 28.2 29.4 30.6 31.8 32.9 32.9 32.9 31.8 32.4 32.9 32.9 32.9
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Year
Figure 1-3 Regional Ethylene Capacity Growth
5
1.2 Feedstock of Ethylene Production
Natural gas (ethane) and petroleum distillates (naphtha) are the main
feedstocks for ethylene production. The trend of their ratios over the years are
illustrated in Figure 1-4. The figure indicates naphtha is the major feedstock for
ethylene production globally. However, ethane’s use as feedstock has gradually
increased. The cost of feedstock accounts for 60% - 80% of ethylene production
costs. The main factors driving feedstock changes for petrochemical plants are the
supply and the price of oil and natural gas.
Ethane Propane Butanes Naphtha Gas Oil Others
200
180
Feedstock, million tons
160
140
79.0
120
67.7 69.7
100 62.1 64.1 65.1
61.0 61.5 60.5
80 60.5 55.4 55.9 13.8
60 11.3 11.3 11.8
10.8 10.8 11.3 10.8
40 8.7 9.7
8.2 8.2 71.3
45.1 47.2 49.7 52.8 55.4 57.9
20 33.8 33.3 36.4 40.0 44.1
0
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2022
Year
Figure 1-4 Global Ethylene Production by Feedstock [6]
In natural gas-rich regions, like North America, ethane was the main feedstock
for ethylene production in recent years [4, 7, 8]. As showed in Figure 1-5, ethane was
the main feedstock for ethylene production in the United States from 2005 to 2011.
6
Regional cash cost and global cash cost of ethylene production are illustrated in
Figure 1-6 and Figure 1-7 respectively. As revealed from the two figures, ethane as
ethylene feedstock was more profitable than LPG and naphtha. The higher margins
of ethane-to-ethylene have resulted in significant feedstock shifts in the U.S. over

[9, 10]
the past few years. Numerous chemical producers are expanding ethylene
capacity to take advantage of the low-cost and increased availability of ethane from
natural gas. A vast number of chemicals and products are made from ethylene. It is
the most produced organic chemical worldwide. In 2012, global ethylene sales
reached about 150 billion US dollars [7].
Ethane Propane Butanes Naphtha & C5+
100
90 16.1 17.0 14.7
28.7 28.0 24.1
32.6 3.7 3.2 3.2
80
Ethylene Feedstock, wt%
4.4 18.3 17.0 17.4

70 4.6 4.1
3.9
60 18.3
18.3 18.1
17.4
50
40
30 61.9 62.8 64.7
48.4 49.8 53.2
46.1
20
10
0
2005 2006 2007 2008 2009 2010 2011
Year
Figure 1-5 U.S. Ethylene Feedstock Sources (Percent of Ethylene Production)

[11]
7
Ethane is the most competitive feedstock for ethylene production due to its
cost and the highest ethylene yield among all other feedstocks. Generally, ethylene
yields decrease with the increase in the density of feedstocks. Heavier feedstocks
produce more low-value by-products. To produce the same amount of ethylene, the
use of AGO feed was about three times that of ethane. Furthermore, the utilities
consumption is higher for heavy feeds.
1200
1006
951 967
1000
883
799 824 830
Cash Cost, USD/ton
800
600
400 318
200 103
0
MDE NAM NEA NAM MDE MDE NEA SEA WEP
Ethane Ethane LPG Naphtha LPG Naphtha Naphtha Naphtha Naphtha
Source: IHS as of July 2012.

MDE: Middle East, NAM: North America, NEA: Northeast Asia, SEA: Southeast
Asia, WEP: Western Europe.
Figure 1-6 Global Ethylene Cash Cost by Region
8
350
313 319
303 307
300 277 285
Index, Canada Cost = 100
250 232
200
150
100
100
50 28
0
MDE Canada NAM MDE MDE WEP NEA NAM SEA
Ethane Ethane Ethane Naphtha LPG Naphtha Naphtha Naphtha Naphtha
MDE: Middle East, NAM: North America, NEA: Northeast Asia, SEA: Southeast
Asia, WEP: Western Europe.
Figure 1-7 Global Ethylene Cash Cost Index: 2010 [12]
1.3 Technologies for Ethylene Production
Driven by the pursuit of lower investment cost the tube furnace pyrolysis
technology has continually improved over the course of several decades. Currently,
tube furnace steam cracking is the dominant technology in ethylene production.
About 99% of global ethylene production uses tube furnace pyrolysis method. A
simplified process flow diagram is shown in Figure 1-8.
9
Convection
Feedstock Zone
BFW
Steam
Superheated
Steam
~850°C Steam
Drum
< 0.1s
Radiation
Zone
350~600°C
To Quench
Steam Cracker Transfer-Line
Exchanger
Figure 1-8 Steam Cracking Process
First, preheated hydrocarbon feedstock (500 to 680 °C) is mixed up with
dilution steam in the convection zone and then quickly discharged to the radiation
zone (750 to 875 °C) where the feed is cracked to produce ethylene, other small
olefins, di-olefins, and by-products. The method is also known as steam cracking
process. The residence time is 0.1 – 0.5 seconds. Due to the high reactivity of the
products, the high temperature effluent has to be quenched within 0.02 – 0.1
seconds in the transfer-line exchanger to avoid further unwanted side reactions. The
cooled products are separated by distillation towers.
Using ethane as feedstock, typical hydrocarbon conversion levels reach 70%
with olefin yields of around 50% [13]. Single-pass conversion and yields are lower
in naphtha crackers.
10
Industrial ethylene production requires dehydrogenation and/or cracking
reactions of hydrocarbon molecules at elevated temperatures because the cracking
reactions are highly endothermic, and also equilibrium limited. Unfortunately,
ethylene and feed hydrocarbons are very reactive at high temperature, which means
they can easily convert to undesirable by-products. Both dehydrogenation and
cracking are expansion reactions, which according to the Le Chatelier’s principle
means that lower product partial pressure is favored for the higher conversion. In
the steam cracking process, steam is used as a diluent to lower reaction partial
pressure. Lower hydrogen partial pressure also reduces unwanted hydrogenation
reactions, which increases yields of desired products—olefins. Steam can also
suppress coking in radiant coils. The design constraints for increasing ethylene
yields are shown in Figure 1-9.
Figure 1-9 Optimal conditions of ethylene production
As shown in Figure 1-9, parameters that influence the performance and
product distribution are residence time, the partial pressures of the feedstock and
steam, and the process temperature (profile) in the reactor.
The main six technology licensors for ethylene production and their main
technology characteristics are summarized in Table 1-1. All technology providers
11
have their own proprietary equipment, which is designed to maximize olefin
(ethylene) yields with different feedstock by varying operation conditions.
Table 1-1 Main technology licensors for ethylene production

Companies ABB Stone & Technip Kellogg Mitsubishi Selas-Linde
Lummus Webster KTI Brown & Petrochemical
Root Company
Pyrolysis SRT-I, II, USC-U, W, SMK, USRT, M-TCF PyroCrack,
furnace III, IV, V, M USC M- SCORE LSCC
VI, VII coil, GK6,
USC U-
coil
Quench boiler Quicker USX, M- U coil/ Kellogg M-TLX Linde quench
Quencher SLE SLE casing exchanger
TLE TLX
Furnace outlet 830~840 820~860 Diesel: 871~927 820~850
temperature, °C 800
Naphtha:
840
Residence time, 0.2~0.6 0.2~0.3 0.3 0.03~0.1 0.3~0.6
s
Thermal SRT-II: 86~91 90 93~94 89~91
efficiency, % 87-91
SRT-III:
92-93
Feedstock Ethane to Ethane to Ethane to Heavy Ethane to Ethane to
diesel diesel diesel diesel diesel light diesel
1.4 Challenges for Tube Furnace Pyrolysis Process
1.4.1 Coking
Coking is one of the challenges to radiant coils and quench exchangers. There
[14]
are two types of coking, tube metal catalyzed coking and pyrolysis coking .
Carburized metals are problematic as they can boost the catalyzed coking process.
Catalytically formed coke is about 80% of the total coke when using gas feedstocks,
like ethane, propane, and butane. For liquid feedstock, catalytically formed coke is
about 30% to 40% of the total coke. Pyrolysis coking is the lesser coke product and
12
is caused by dehydrogenation of small unsaturated hydrocarbons like acetylene, and
polycyclic aromatics by condensation, polymerization, and dehydrogenation
reactions.
The thickness of coke is a function of tube metal temperature and conversion.
Coke layer could reach 10 mm or thicker depending on the type of feedstock and
operation severity. Coking in furnace tubes can reduce product yields, decrease
production capacity, increase energy consumption, shorten tube service life, and
increase maintenance and operation costs.
A variety of anti-coking technologies have been developed to reduce coking.
The goal is to engineer a metal surface by applying a layer or multiple layers of
coating. The coating materials could be silicon-based, like GE Betz’s Py-CoatTM,
Chevron Phillips’ CCA 500, or phosphorus-based, like Nalco’s COKE-LESS®.
These coating agents can be applied on-line to make a stable layer coated on the
inner surface of the tubes in the furnace. The coated layer can reduce coking
effectively, protect the tube metal base from oxidation, and improve tube life and
extend the de-coking cycle. Westaim’s CoatAlloyTM and Nova Chemicals
/Kubota’s ANK 400 coating agents can be applied on the inner surfaces of new
furnace tubes and transfer line exchangers. Coking inhibitors can also reduce
coking. The agents are the chemicals containing S, P, Sn, Sb, and Cr among others.
Typical compounds are DMDS, triphenylphosphine, triphenyl phosphorus oxide,
and chromium phosphate. Both coating agents and inhibitors are effective in
reducing coking and extending operation cycles. However, these technologies only
mitigate coking issues, but cannot eliminate the coking problems.
13
1.4.2 Energy Consumption of Pyrolysis Furnace
Steam cracking for light olefins such as ethylene and propylene production is
the most energy consuming process in the chemical industry. During ethylene
production, the cracking furnace is a major energy consumer. Its energy
consumption is about 50-60% of the total energy of an ethylene complex. A naphtha
steam cracker consumes about 65% of the total process energy and approximately
[15]
75% of the total energy loss . It is a challenge to increase ethylene production
while saving energy simultaneously. Thermodynamically, higher temperatures
enhance endothermic reaction to reach higher equilibrium conversion. Kinetically,
higher temperatures also accelerate the relative rate of the primary reaction
(producing target product) to secondary reaction (producing by-products). However,
higher temperatures also cause deep cracking of hydrocarbons and accelerate
coking because these secondary reactions are more thermodynamic favorable. This
results in ethylene yield reduction and other problems. A solution is to shorten
residence time. In order to maintain conversion in shorter residence time, higher
tube and furnace temperatures are required, which increases the energy
consumption of the cracking furnace.
In summary, ethane is the most economically competitive feedstock for
ethylene production. However, ethane dehydrogenation is an endothermic and
equilibrium limited process so that high reaction temperature is required for
industrial production operations. Although hydrocarbon steam cracking is the main
industrial process for ethylene manufacture, it has its disadvantages. First, steam
cracking is a highly energy intensive process. When using ethane as feedstock, the
14
[13]
total energy consumption is 16 GJ/ton versus 23 GJ/ton when using naphtha .
Second, fired furnace produces CO2 and NOx emissions. approximately 1 – 1.6 tons
of CO2 are produced per ton of ethylene. In addition, coking and high temperature
create maintenance challenges for furnace tubes.
High temperature operation also produces low value by-products. According
to the mechanism of ethane dehydrogenation, if one of the products hydrogen can
be selectively removed from the reaction system, it can break the thermodynamic
barrier and will increase the conversion even at lower temperature and achieve more
economical production of ethylene. An effective catalyst can also help the
dehydrogenation reaction reach equilibrium quickly at low temperature.
1.5 Catalytic Dehydrogenation of Ethane to Ethylene by
Membrane Reactor
1.5.1 Gas Separation Mechanisms of Porous Inorganic Membranes
Pore size is fundamental to the separation properties of porous membranes.
According to the IUPAC classification of pore diameter [16], there are three general
types of porous membranes: macroporous membranes with average pore diameter
larger than 50 nm, mesoporous membranes with average pore diameter in the range
between 2 nm and 50 nm, and microporous membranes with average pore diameter
less than 2 nm. The micropores can be further classified into subgroups by the
diameter: supermicropores (1.4 to 3.2 nm), micropores (0.5 to 1.4 nm), and
ultramicropores (<0.5 nm) [17].
15
Pore size and distribution, openness (volume fraction) and the nature of the
pores to a great extent decide a porous inorganic membrane’s permeability and
permselectivity. Higher porosity membranes have a more open porous structure,
thus generally leading to higher permeation rates for the same pore size. The nature
of the pores determines the mechanisms of gas stream transport through a porous
membrane, and consequently the membrane’s separation properties. Generally,
inorganic membranes do not exhibit a high permeability and a high permselectivity
simultaneously.
In gas separation, there are six possible transportation mechanisms [18-20]. An
overview of the transportation mechanisms is shown in Table 1-2.
From the table, it can be seen that viscous flow, molecular diffusion, and
Knudsen diffusion occur in macro- and meso-pores, but viscous flow and molecular
diffusion do not separate species. However they can affect the total flow resistance
of the membrane system. Knudsen diffusion, surface diffusion, multilayer diffusion,
capillary condensation, and molecular sieving are responsible for gas separations.
Table 1-2 Transportation regimes in porous membrane

Transport type Pore diameter Selectivity
Viscous flow 𝑑 > 20 𝑛𝑚 -
Molecular diffusion 𝑑 > 10 𝑛𝑚 -
Knudsen diffusion 2 < 𝑑 < 100 𝑛𝑚 1/√𝑀
Surface diffusion +
Capillary condensation/
++
multilayer diffusion
Molecular sieving
𝑑 < 1.5 𝑛𝑚 ++
(Size exclusion)
16
The separation mechanism of Knudsen diffusion is based on the collisions
between molecules and the pore surface. When the mean free path length of the
molecules is larger than the characteristic pore diameter, the permeating species
collide with the pore surfaces more frequently than with other molecules. As shown
in equation (1-1), smaller molecules have longer mean free path length because the
diameter is shorter. Mean free path length is proportional to temperature and
inversely proportional to pressure.
𝑘𝑇
𝜆= 2𝑃 (1-1)
√2𝜋𝑑𝑝
As shown in equation (1-2), the separation factor of Knudsen diffusion
depends on the square root of the molecular weight ratio of the gases separated,
where F and M represent the permeation fluxes and molecular weights of the gas
species A and B.
𝐹𝐴 𝑀𝐵
𝛼∗ = =√ (1-2)
𝐹𝐵 𝑀𝐴
2𝜆
𝐾𝑛 = (1-3)
𝑑𝑝
Knudsen number 𝐾𝑛 in equation (1-3) generally defines the three regimes of
gas diffusion (also refer to Table 1-2): (1) for viscous flow, 𝐾𝑛 ≪ 1 ; (2) for
17
Knudsen diffusion, 𝐾𝑛 ≫ 1; and (3) for transition flow, 0.01 ≤ 𝐾𝑛 ≤ 10. Viscous
and transition flow are non-selective. Generally speaking, when is ten times
[18]
larger than 𝑑𝑝 , Knudsen diffusion becomes dominant . In the Knudsen flow
regime, the mean free path of gases is comparable to the pore diameter.
As equation (1-2) implies, the separation is only effective for those gases with
large difference in molecular weights. Since the selectivity between molecules in
this regime is usually low, Knudsen flow is not effective in gas separation. As
[21]
equation (1-4) shows , Knudsen permeance is not a function of pressure. It is
proportional to 1⁄√𝑇.
2𝜀𝛾 8
𝐹𝐾 = √ (1-4)
3𝜏𝐿 𝜋𝑅𝑇𝑀
Surface diffusion, multilayer diffusion and capillary condensation are more
effective separation mechanisms than Knudsen diffusion. Surface diffusion occurs
when gas mixtures interact with the inner wall materials. Chemical properties of
the membrane surface play an important role in gas surface diffusion.
In porous membranes meso-pores and micro-pores contribute to selectivity.
In mesoporous membranes, the dominant transport mechanism relies on the pore
size and the permeating molecules which limits the selectivity of the membrane.
The dominant transport mechanism in microporous membranes is molecular
sieving, an activated transport mechanism, which is capable of permselectivities
much greater than Knudsen.
18
Three main parameters are used to evaluate the performances of a gas
separation membrane system: (1) the permselectivity or selectivity toward the gases
to be separated. This property affects the recovery of the valuable gas in the feed
and, for the most part, directs the process economics. (2) the permeate flux or
permeability which is related to productivity and determines the membrane area
required. (3) the membrane stability or service life which has a strong impact on
the replacement and maintenance costs of the system.
1.5.2 Zeolite Membrane Reactor Configurations
By the definition of IUPAC [22], a membrane reactor combines two functional
parts in one unit: a chemical reaction and a membrane-based separation. The
membrane used in the reactor can be polymeric or inorganic, homogeneous or
heterogeneous, symmetric or asymmetric in structure. A zeolite membrane reactor
can be configured in one or more modules in these categories.
For applications of membrane-based reactive separation at high temperature,
zeolite based membranes overcome the limitations associated with organic and
metal membranes. Zeolites have a highly uniform pore structure, good thermal
stability, and strong mechanical properties. Zeolitic materials have been used for
inorganic membranes since 1930s. Zeolites that have been extensively studied for
membrane applications include zeolite X, Y, ZSM-5, Silicalite-1, Na A, and
sodalite [23]. However, gas separation and reactive separation applications of zeolite
membranes depend strongly on the selectivity, permeability and stability
characteristics of these membranes. In addition, the development of a high-
performance and energy-efficient membrane-based process will depend on the

19
availability of highly-selective and robust zeolite membranes, fabricated with
simple and cost-effective protocols [24].
Unlike traditional separation processes governed by thermodynamic
equilibrium relationships, membrane separation is based primarily upon the relative
rates of mass transfer in membranes. For gas separation, the transport occurs by
adsorption-diffusion mechanism. Membrane selectivity is based upon the relative
permeation rates of the components through the membrane. Gas diffusion obeys
Fick’s Law. The driving force of gas diffusion in a porous membrane is the
concentration gradient through the membrane.
The catalytic function can be introduced in one of two ways: inert zeolite
catalytic membrane reactor and zeolite catalytic membrane reactor. In an inert
catalytic membrane, the zeolite membrane is not catalytically active and does not
participate in the reaction, and simply acts as a selective separation unit for the
desired products. A separate packed bed of catalyst particles promotes the chemical
reaction on the feed side and the membrane selectively removes one of the products
of reaction. For a catalytic zeolite membrane reactor, the membrane is catalytically
functionalized with a precious metal, and the membrane effectively performs the
catalytic reaction and separation simultaneously because the catalysis takes place
on the surface of the membrane.
Based on the configurations of the zeolite membranes and the catalysts, there
[25]
are three types of zeolite catalytic membrane reactor modules , as shown in
Figure 1-10. An extractor-type zeolite catalytic membrane reactor increases the
conversion by selectively removing a product or an intermediate from the reaction
20
system whose conversion in a traditional reactor is limited by the thermodynamic
equilibrium (cases (a) and (b) in Figure 1-10). In the examples shown in case (a),
the conversion to olefins from alkanes or other hydrocarbons by dehydrogenation
was enhanced by the removal of product hydrogen from the feed side. This reactor
is ideal for the ethane dehydrogenation to ethylene [26].
As shown in case (c) in Figure 1-10, the reactant distributor type membrane
reactor works like an extractor type membrane reactor inversely. A typical example
of this reactor’s application is partial oxidation reaction of butane to maleic
anhydride [27]. By controlling lower partial pressure of one reactant, it is expected
to improve the selectivity of the target product. Oxidative dehydrogenation of lower
alkanes like ethane [28-33] and propane [34, 35] to make olefins were carried out in this
type of reactors by conducting oxygen or oxygen anions trough the membrane to
the reaction system. There are many benefits to using a distributor membrane for
oxidation reactions, such as achieving higher selectivity, more even reaction heat
distribution, and working beyond flammability regions. The membrane in the
contactor type of reactor shown in (d) is catalytically functionalized. It functions as
catalyst and separator at the same time. In this configuration the reactants are
located at different sides of the membrane and must diffuse through the zeolite layer
to react.
21
Figure 1-10 Types of Zeolite Catalytic Membrane Reactors
22
1.6 Research Objectives
The objective of this work was to demonstrate the concept of using natural
zeolites in a membrane reactor to increase the conversion of ethane and yield of
ethylene at temperatures below conventional thermal cracking conditions. This
work also sought to demonstrate the benefits of using ETS-2 as a catalyst substrate
for ethane dehydrogenation.
23
Chapter 2. Characterizations and Product Analysis
2.1 Characterization methods
2.1.1 X-ray diffraction (XRD)
The phase identification and structural analysis of zeolites and catalyst
materials were characterized by X-ray diffraction (XRD) using Rigaku Gigerflex
2173 diffractometer with Co tube x-ray source (Co 𝐾𝛼 , 𝜆 = 1.79021Å), and a
vertical goniometer with a graphite monochromator as 𝐾𝛽 filter.
Testing conditions were: 38 kV, 38 mA, scan speed 2 degree/min, step width
0.02 degree, and scan range 5 to 90 degree. XRD samples data were compared to
the database from the International Center for Diffraction Data-Powder Diffraction
File (ICDD-PDF) to identify crystalline phases present in the samples.
For natural zeolites, both powder and thin surface samples were tested. For
catalysts and supports, all samples were fine powder. All XRD data were collected
at room temperature.
2.1.2 Pore Size Distribution, Specific Surface Area, and Pt Particle
Dispersion
The BJH (Barret-Joyner-Halenda) and Horvath-Kawazoe (HK) pore size
distribution (PSD) of the natural zeolite modernite were calculated from the N2
adsorption isotherm at 77K using a Quantachrome Instruments AUTOSORB-1.

24
The specific surface area of samples was calculated using the Brunauer-Emmett-
Teller (BET) method. For the natural zeolite mordenite, the sample dried at 120 °C
was degassed at 150 °C for two hours. Other samples were degased at 200 °C for
two hours before adsorption.
Hydrogen chemisorption measurements for the Pt catalysts were performed
on a Micrometrics ASAP 2020C instrument. Samples of ~ 0.5 g were first reduced
in hydrogen at 400 °C for 30 minutes, cooled down under vacuum to 35 °C, and
then evacuated for an additional 10 minutes at this temperature before the hydrogen
isotherm was collected at the same temperature. A total of 14 points were collected
for the hydrogen isotherm between 0.25 and 0.60 kPa. Dispersions were estimated
assuming the adsorption of one hydrogen atom per surface Pt atom.
2.1.3 X-ray photoelectron spectroscopy (XPS)
The XPS measurements were carried out on a Kratos Axis 165 X-ray
photoelectron spectrometer with a Mono Al Kα source (hν = 1486.6 eV) (operated
at 15 mA and 14 kV). The survey spectra were collected with analyzer pass energy
of 160 eV and a step of 0.4 eV; the high resolution spectra were scanned with a pass
energy of 20 eV and a step of 0.1 eV. During the acquisition of a spectrum, charge
neutralization was applied to compensate the insulating problem of the sample. XPS
signals were fitted by mixed Lorentzian–Gaussian curves using Casa XPS software.
All binding energies are reported after calibration for C1s peak to match 284.8 eV.
25
2.1.4 TPR
Hydrogen temperature-programmed reduction (H2-TPR) was carried out on
an AutoChem 2950 HP instrument (Micromeritics, USA) equipped with thermal
conductivity detector (TCD) under atmospheric pressure. About 0.5 grams of the
catalyst calcined in air, at 200 °C, for one hour was used for the analysis. The
catalysts were reduced in a flow of 10% H2/Ar (50 ml/min) from room temperature
to 500 °C with a heating rate of 5 °C/min.
2.1.5 SEM and EDX
The SEM used was a Hitachi S-2700 Scanning Electron Microscope (SEM)
equipped with a PGT (Princeton Gamma-Tech) IMIX digital imaging system and
a PGT PRISM IG (Intrinsic Germanium) detector for Energy Dispersive X-Ray
Analysis (EDX). Backscattered Electron Detector - GW Electronics System 47 four
quadrant solid state Backscattered Electron Detector. Samples for EDX were coated
with carbon.
The high-resolution scanning electron micrographs were obtained using a
JEOL 6301F field emission scanning electron microscope. Samples for high-
resolution images were coated with gold.
2.1.6 TEM
The morphology of Pt particles on catalysts were characterized by
transmission electron microscopy (TEM) on a JEOL 2100 transmission electron
26
microscope operating at 200 kV. Fine powder samples were loaded on copper mesh
and coated with carbon.
2.1.7 TGA and TG-MS
The thermogravimetric analysis of natural zeolites was conducted on TA
Instruments TGA Q500. N2 was used as the inert sweeping gas at a flow rate of 40
ml/min while between 15 to 25 mg of fine powder sample was used for analysis.
Heating rate was 10 °C/min from room temperature to 1000 °C.
The analysis of coke deposits on spent catalysts was conducted using TG-MS.
The TG-MS plots were collected using a TA Instruments Q500 TGA coupled to a
Pfeiffer Omnistar QMA 200 residual gas analyzer. TG scans were run using a
balance purge rate of 5 ml/min Ar and a sample (air) purge rate of 180 ml/min.
Samples were loaded on platinum pans and heated at a rate of 10 °C/min from
ambient to 600 °C.
The mass spectrometer was configured with a stainless steel capillary heated
to 200 °C and connected to the TGA by means of a stainless steel adapter and a ¼”
branch T-fitting. The capillary was positioned at the center of the tee close to the
furnace exhaust to sample the evolved gases. The exhaust end of the T-fitting was
connected to a 6 m length of ¼” OD tubing that prevented back-diffusion of
atmospheric components and led to a fume hood exhaust. Experiments were run in
MID mode with fragments at m/z = 18 and 44 tracked over time. These fragments,
associated with water and carbon dioxide respectively, were found to be the most
diagnostic signals associated with the combustion of the coke. To start data
27
collection, the two instruments were manually synchronized. An uncertainty of a
few tenths of a degree for the starting temperature can be expected between the TG
and MS plots.
2.1.8 Dilatometry test
The dilatometric behavior of natural zeolite mordenite was carried out on a
L75 PT Vertical Dialtometer (Linseis Thermal Analysis). About 0.3 g of each
powder sample was pressed into cylindrical rods using a special die and an isostatic
press (Top Industrie) under a pressure of 300 MPa. The rods were about 10 mm in
length and 5 mm in diameter. The pressed rods were heated from room temperature
to 1000 °C under air with a rate of 10 °C/min. After maintaining a temperature of
1000 °C for one hour, the sample was cooled down at a rate of 10 °C/min. The
thermodilatometric data were processed under the form of curves, which report the
percent variation of cylinder length (ΔL/L0, %) as a function of temperature.
2.2 Gas Product Analysis
An on-line GC (Bruker 450-GC) with automatic sampling valves was used to
analyze gas products. The packed column was Hysep D (80-100 mesh) stainless
steel ( 1/8”, 3 meters), equipped with a thermal conductivity detector (TCD).
Argon (99.99%, Praxair) was used as GC carrier gas.
28
Part I: The Use of Natural Zeolite Mordenite in a
Membrane Reactor for Ethane Dehydrogenation

Chapter 3. Characterization of Natural Mordenite Zeolite
3.1 Introduction
Zeolites are hydrated crystalline alumina-silicates that contain alkali and
alkaline earth metals. The name zeolite was created by the Swedish mineralogist
[36]
Axel Fredrick Cronstedt in 1756 and it means “boiling stone” in Greek . More
than 40 naturally occurring zeolites have been identified in the past two centuries.
Chabazite, clinoptilolite, erionite, ferrierite, mordenite, and phillipsite are the six
[37]
most industrially important natural zeolites . Mordenite was first discovered in
[38]
1864 in Morden, King’s County, Nova Scotia, Canada by How and, as per
tradition, was named after the locality. Mordenite mineral deposits occur in more
than 348 locations in 39 countries including Antarctic [39], as shown in Figure 3-1.
New Zealand is one of the countries with the most abundant mordenite
deposits. About 250,000 years ago, massive volcanic ash formed thick sediment
beds in lakes. Hot water transformed those clay beds into soft rocks with ordered
internal structures. Over time the lakes drained due to severe block faulting and the
deeply buried structured clay rocks became zeolites. Mordenite and clinoptilolite
are the predominant natural zeolites in New Zealand. Due to their diagenesis, they
are highly matured and mechanically strong and cannot be broken down into
constituent clays.
All samples of natural zeolite mordenite used in the thesis work are from New
Zealand.
30
Mordenite was chosen as potential membrane reactor material because it is
physically and chemically stable at the required working temperatures and it has
suitable pore structures to achieve good selectivity for hydrogen removal. In
addition, it is very economical due to its abundance in nature. Its pore structure,
unique to naturally occurring zeolites, is only large enough to admit molecules
and smaller making it ideal for hydrogen separation from larger hydrocarbons.
Figure 3-1 Localities of natural zeolite mordenite
3.2 Channel Sizes of Zeolite Mordenite
Zeolite mordenite has a big elliptical 12-ring channel and an 8-ring (2.6 ×
5.7Å) channel along [001] with two interconnected 8-ring channels (3.4 × 4.8Å)
along [010]. Figure 3-2 shows an ideal framework of mordenite. The 8-ring holes
of successive sheets do not align to make channels along [010]. Since the 8-ring
31
channel along [001] is smaller than most gases’ kinetic diameter [40] (He: 2.6 Å, H2:
2.8 Å, CO2: 3.3 Å, O2: 3.4Å, N2: 3.6Å), the 8-ring opening acts as a connecting
window. Practically, mordenite is one-dimensionally permeable along [001]. A
more visual demonstration is shown in Figure 3-3 [41]. The figure clearly shows that
the 8-ring channels are highly compressed and not accessible to most molecules or
ions, and there is no direct connections between 12-ring channels. Although the
framework structure appears like it would admit larger molecules, the position and
types of the cations inside the framework prevent molecules larger than about 3 Å
from entering the structure.
Framework viewed along [001]
12-ring viewed along [001]
32
8-ring viewed along [001] between 12-ring channels
8-ring viewed along [010]
Figure 3-2 Framework and ring sizes or mordenite
Figure 3-3 One-dimensionally accessible channel system in mordenite
33
3.3 XRD
Both the powder and the thin membrane surface samples of natural zeolite
mordenite were treated at different temperatures.
Figure 3-4 shows that the sample of natural zeolite mordenite contained a
quartz impurity. From Figure 3-5, it can be seen that the natural mordenite samples
were thermally stable at elevated temperatures. The XRD patterns show that the
crystalline framework remained unchanged up to 750 °C.
Figure 3-4 XRD patterns of natural zeolite mordenite before calcination
34
Figure 3-5 XRD patterns of natural zeolite mordenite treated at different
temperatures for 4 hours
3.4 Scanning Electron Microscopy (SEM)
In Figure 3-6, cross-section images clearly show that there were plenty of
needle-like crystals in natural mordenite samples. Well-matured natural zeolite
mordenite has a needle-like crystal structure while the layers stacked on each other
indicate the presence of multiple crystals. This multiple-layer structure reveals the
growing process of naturally occurring zeolite mordenite.
35
Figure 3-6 SEM images of cross-section of natural zeolite mordenite
3.5 Thermal Analysis
3.5.1 Thermogravimetric Analysis (TGA)
Figure 3-7 shows the TGA curve of the natural zeolite mordenite. The total
weight loss was about 12.1 wt%. The weight losses at ≤ 150 °C, 150-200 °C, 200-
300 °C, 300-400 °C, 400-670 °C were 6.2 wt% (about 51% of the total weight loss),
1.1 wt% (9% of the total weight loss), 1.3 wt% (10.7% of the total weight loss), 1.1
wt% (9% of the total weight loss), and 2.2 wt% (18.2% of the total weight loss)
respectively. After 660 °C, there was nearly no weight loss (~ 0.2 wt%).
36
The weight loss at < 200 °C is caused by dehydration of the zeolite. For the
zeolites with high kinetic diameter, like mordenite, the main dehydration
temperature is lower than 200 °C. The so-called zeolitic water is free water
absorbed in capillary channels and external surface as well. The water dipole is
attracted to the oxygen electron density in the SiO4 and AlO4 tetrahedra. Free water
in large channels is easy to remove because it lacks strong associations, and so the
dehydration temperature is in the low range (~ 100 °C). Water associated with the
cations and bound to them through ion-dipole interactions is more difficult to
remove and is desorbed at higher temperatures (100-200 °C). Once all of the water
associated with the cations is desorbed the remaining weight loss is due to
“chemical” and “structural” water. Hydroxyl groups present in defects and at the
terminus of SiO4 and AlO4 tetradedra can be removed at sufficiently high
temperatures by condensing those hydroxyl groups to water molecules. The total
amount of water content and dehydration profile of natural mordenite were in good
agreement with literature reports [42]. The loss of water can cause zeolite membrane
to shrink. This will be illustrated in the dilatometric analysis.
37
100
98
96
Weight (%)
94
92
90
88
86
0 100 200 300 400 500 600 700 800 900 1000
Temperature (C)
Figure 3-7 TGA profile of natural mordenite
3.5.2 Dilatometry Test
In order to understand the thermal mechanical behavior of natural zeolite
mordenite, a dilatometry test was conducted.
38
20
-20
∆L/L0, ×10-3
-40
-60
-80
-100
-120
0 100 200 300 400 500 600 700 800 900 1000
Temperature (°C)
Figure 3-8 Dilatometry test of natural zeolites mordenite
Dilatometric curves of sintering for mordenite between room temperature and
1000 °C under atmospheric conditions is shown in Figure 3-8. In the region between
room temperature and around 800 °C, the negative slope curve shows shrinkage of
sintering behavior of the sample. The TGA curve in Figure 3-7 shows that from
room temperature up to 700 °C the mordenite sample was continuously losing water,
and the weight loss at 700 °C was about 98% of total weight loss. From the XRD
profile in Figure 3-5, it can be seen the mordenite retained its crystal frameworks
without obvious structural collapse and phase transformation. So it appears that the
continuous dehydration with increased temperature was the main reason for the
shrinkage. In the region between 800 °C and 1000 °C, shrinkage increased
39
[43]
significantly and thermal collapse of the structure occurred. Cruciani reported
mordenite exhibited good thermal stability above 800 °C.
For the negative thermal expansion behaviors, Sleight proposed four possible
[44]
mechanisms : (a) decrease of M-O bond length leading to a net cell volume
contraction; (b) anisotropic thermal behavior of M-O bonds; (c) extrinsic effect of
interstitial cations relocating as temperature changes; (d) transverse thermal
vibrations of oxygen in M-O-M linkages.
For zeolites, the M-O bonds in the frameworks are hinged by two-coordinated
oxygen and are relatively rigid. Their thermal motions have limited freedom in
order to maintain structural framework. Hammonds et al developed a rigid unit
mode (RUM) model to interpret and quantify the flexibility possessed by zeolites
[45, 46]
, but it was a debated issue [43].
Martucci et al [47] carried out an in situ study of dehydration process of natural
mordenite by X-ray synchrotron powder diffraction from 25 to 830 °C. Structure
refinements indicated that the slight cell-volume contraction (~1.9 %) is related to
the release of water molecules from the channels. The cell parameters b and c
decreased regularly as the temperature rose. The removal of water molecules also
relocated the initial Ca sites into many positions bonded to the framework oxygen.
The increased interaction with the framework oxygen of Ca sites was intimately
related to the distortion of the 12-ring which is in turn related to the lengthening of
the a cell parameter. Figure 3-9 demonstrates the initial cations relocating and water
molecules removing out the 12-ring channel.
40
Figure 3-9 Refined natural mordenite structure viewed along [001] at different
temperatures: (a) 25 °C, (b) 200 °C, (c) 450 °C, (d) 830 °C
3.6 BET of Mordenite
Specific surface area calculated with the Brunauer-Emmett-Teller (BET)
method. The BET equation is shown in Equation (3-1).
1 1 𝐶−1 𝑃
𝑃 = + ( ) (3-1)
𝑊( 0 −1) 𝑊𝑚 𝐶 𝑊𝑚 𝐶 𝑃0
𝑃
41
𝐸1 −𝐸𝐿
𝐶 = 𝑒𝑥𝑝 ( ) (3-2)
𝑅𝑇
Figure 3-10 shows the BET surface area changes of the mordenite treated at
different temperatures. In general, the measured surface area was similar to

[48]
literature reports . From the figure, it can be seen that the sample calcined at
200 °C for four hours has the largest surface area. As water was removed out from
[47]
the inner channels, the surface area increased. Martucci and coworkers found
that at 200 °C the diameter of the 8-ring along [001] was the largest, decreasing
with further temperature raise. A big ring aperture exposes more internal surfaces
to N2 and may be the cause of surface area increase.
40
35
30
Surface Are, m2/g
25
20
15
10
5
0
0 200 400 600 800 1000
Temperature, °C
Figure 3-10 BET surface area of natural mordenite treated at different
temperatures
42
Pore size distribution determined with Horvath-Kawazoe (HK) method [49] is
shown in Figure 3-11. Many pore size distribution methods are derived from the
Kelvin equation that describes the phenomena of capillary condensation. The HK
method was derived independently from the Kelvin equation, which more
accurately describes micropore (< 20 Å) size distribution. The HK equation is
shown in Equation (3-3).
𝑃 𝑁𝑠 𝐴𝑠 +𝑁𝐴 𝐴𝐴 𝜎4 𝜎 10 𝜎4 𝜎 10
𝑅𝑇 𝑙𝑛 ( ) = 𝐾 [ 𝑑 3
− 𝑑 9
− 𝑑 3
+ 𝑑 9
] (3-3)
𝑃0 𝜎 4 (𝑙−𝑑) 3(𝑙− ) 9(𝑙− ) 3( ) 9( )
2 2 2 2
6𝑚𝑐 2 𝛼𝑠 𝛼𝐴
𝐴𝑠 = 𝛼𝑠 𝛼𝐴
(3-4)
( + )
𝜒𝑠 𝜒𝐴
3𝑚𝑐 2 𝛼𝐴 𝜒𝐴
𝐴𝐴 = (3-5)
2
0.858𝑑
𝜎= (3-6)
2
𝑑 = 𝑑𝑠 + 𝑑𝐴 (3-7)
43
4 6 8 10 12 14 16 18 20
1.6E-3
1.1E-3 120°C
5.4E-4
0.0
5.1E-2
200°C
3.4E-2
1.7E-2
Dv(w) (cc/Å/g)
0.0
1.8E-3 400°C
1.2E-3
6.0E-4
0.0
6.0E-4 800°C
4.0E-4
2.0E-4
0.0
12.3 1000°C
8.2
4.1
0.0
4 6 8 10 12 14 16 18 20
Pore Width (Å)
Figure 3-11 Pore size distribution of natural mordenite calculated by HK
method
Figure 3-11 clearly shows the micropore size distribution of the natural
mordenite treated at different temperatures. As treating temperature increased, the
mean pore size increased, and the size distribution became wider. Both pore size
and distribution were well predicted by the removal of water adsorbed in channels
and redistribution of initial cations through water removal, which made N2 more
accessible to the micropores. At 800 °C, a higher number of large pores with a wide
range of sizes are present. These big pores may be caused by the collapse of some
of the micro frameworks. When temperature increased to 1000 °C, micropore size
44
distribution narrowed down to two main sizes: and . This phenomenon
may further confirm the collapse of micro structures.
The Barrett-Joyner-Halenda (BJH) method provides quantitative specific
surface area and pore size distribution information primarily for mesoporous and
macroporous materials but also some semi-quantitative information on
microporous materials. The equations for pore size distribution calculations are
shown in equations (3-8) and (3-9).
2
𝑟𝑝𝑛
𝑉𝑝𝑛 = ( 𝛥𝑡 ) (𝛥𝑉𝑛 − 𝛥𝑡𝑛 ∑𝑛−1
𝑗=1 𝐴𝑐𝑗 ) (3-8)
𝑟𝐾𝑛 + 𝑛
2
𝑟̅ ̅̅̅−𝑡
𝑟𝑝 𝑟̅
𝑐 = ̅̅̅𝑐 = ̅̅̅
(3-9)
𝑟𝑝 𝑟𝑝
Figure 3-12 shows the mesopore size (> 20 Å) distribution of mordenite
treated at different temperatures. The figure clearly indicates that low temperature
treated samples lack mesopores. However, when treated at 800 °C and 1000 °C, 20
Å and 35 Å mesopores appeared respectively. The mesopores present at higher
temperatures provide further evidence of the structure collapse.
45
0 10 20 30 40 50
2E-2
120°C
1E-2
8E-3
4E-3
0
5.4E-2 200°C
3.6E-2
1.8E-2
0.0
Dv(d) (cc/Å/g)
400°C
5.7E-3
3.8E-3
1.9E-3
0.0
1.2E-3 800°C
8.2E-4
4.1E-4
0.0
5.4E-4 1000°C
3.6E-4
1.8E-4
0.0
0 10 20 30 40 50
Pore Diameter (Å)
Figure 3-12 Pore size distribution of natural mordenite calculated by BJH
method
3.7 Chemical compositions of natural mordenite zeolite
The chemical composition of natural zeolite mordenite from New Zealand is

Si
summarized in Table 3-1. Data show that Si/Al molar ratio is 5.10 (Al+Fe 3+ = 4.36),
46
[50]
falling in the typical range of Si/Al of mordenite . The sample had low
concentration of alkali metals like sodium and potassium, but very high calcium
concentration, which was double that of the typical samples [51].
Table 3-1 Chemical composition of natural zeolite mordenite

Compound wt %
LOI* 15.36
Na2O 1.60
MgO 0.57
Al2O3 9.38
SiO2 56.42
P2O5 0.12
SO3 0.080
K2O 0.70
CaO 12.24
TiO2 0.27
Cr2O3 0.0035
MnO 0.072
Fe2O3 2.49
NiO 0.0034
CuO 0.0038
ZnO 0.0078
Rb2O 0.0015
SrO 0.040
Y2O3 0.0052
ZrO2 0.011
Total 99.38
* LOI = loss on ignition at 950 °C for 1 hour
47
Natural zeolite mordenite is classified in the high thermal stability category

[52]
. XRD and dilatometry results show that it is stable up to 800 °C. Pore size
distribution measurements show that mordenite has uniform pore size. All these
features make the mordenite a strong candidate as a high temperature membrane
material for membrane reactors.
48
Chapter 4. Packed Bed Membrane Reactor for Ethane Dehydrogenation
Chapter 4. Packed Bed Membrane Reactor for Ethane
Dehydrogenation
4.1 Introduction
Ethylene, the simplest olefin, is a fundamental intermediate in petrochemical
industry for the productions of plastics, solvents, cosmetics, pneumatics, paints,
packaging and more. The global ethylene capacity reached 150 million tons in 2012,
[4]
and is predicted to have a growth rate of 3.4% per year . The global sales of
ethylene reach $148 billion annually [8].
Currently, tube furnace steam cracking process is the dominant technology for
ethylene production, which accounts for about 99% of global ethylene production
process. Among the hydrocarbon feedstock for ethylene production, ethane has the
lowest cost and the highest ethylene yield. Heavier feedstock produces more low
value by-products. Steam cracking process for ethylene production occurs at high
temperature because hydrocarbon dehydrogenation is an endothermic reaction
(Equation (4-1).

Published: A. M. Avila, Z. Yu, S. Fazli, J. A. Sawada, and S. M. Kuznicki: "Hydrogen-
Selective Natural Mordenite in a Membrane Reactor for Ethane Dehydrogenation", Microporous
and Mesoporous Materials, 2014, 190 (0), 301-308
49
(4-1)
The ethane diluted with steam flows through high temperature tubes at high
velocity to achieve a short residence time (less than 1 second). Typical
decomposition reaction temperatures are between 800 °C to 850 °C. To maintain
the high reaction temperature, furnace skin temperature is usually between 1100 °C
to 1200 °C. Such high temperature operation causes serious side reactions in
addition to being energy intensive.
Figure 4-1 Illustration of membrane reactor concept for ethane
dehydrogenation
As illustrated in Figure 4-1, selective removal of the product hydrogen from
the reaction system enables shifting the ethane dehydrogenation reaction to
production of ethylene. The dehydrogenation catalysts help the reaction reach
equilibrium quickly and suppress side reactions at relatively low temperature.
50
A membrane reactor that selectively removes hydrogen can shift the
conversion of ethane dehydrogenation to values above the thermodynamic
equilibrium. High conversion levels at lower temperature result in increased energy
efficiency of the system [53-56].
Ethane dehydrogenation has been investigated with various catalytic

[57]
membrane reactors. Champagnie and coworkers studied ethane
dehydrogenation in a tubular alumina ceramic membrane reactor. The ceramic
membrane was impregnated with 5wt% Pt as a cracking catalyst. The feedstock
ethane was diluted with argon and hydrogen. The driving force for hydrogen
separation was Knudsen diffusion. Gobina and coworkers [58-61] studied the packed-
bed membrane reactor. The membrane was a thin layer of Pd-Ag alloy supported
on porous Vycor glass tube. Ethane was diluted with nitrogen. Szegner and
coworkers [62] studied ethane dehydrogenation in a packed-bed membrane reactor.
The catalyst was Pt-Sn/Al2O3 pellets. Membrane used in the reactor was inert and
only functioned as a separator. Feed gas was diluted with hydrogen and argon. The
results showed that the conversion of ethane dehydrogenation was higher than that
of equilibrium. Ethane dehydrogenation is a volume expansion reaction, so diluting
the reaction system with inert gas could also shift the conversion above the
equilibrium.
Among the membrane reactors for ethane dehydrogenation, the extractor-type
membrane reactors is best suited, because the conversion is enhanced by selectively
removing hydrogen from the reaction system shifting the reaction to the favorable
direction. Hydrogen removal by the selective membrane is based on the fact that
51
hydrogen permeation is faster than that of hydrocarbons. However, at high
temperatures the perm-selectivity of hydrogen is just slightly higher than that in the
Knudsen flow, because the Knudsen separation factor is proportional to the square
root of the ratio of molecular weights. In order to increase conversion further at
these temperatures a higher selective membrane is required.
The unique internal channel system and pore sizes of some zeolitic
membranes allow them higher selectivity to hydrogen permeation, making them
good candidates as high temperature membrane reactors. Natural zeolite mordenite
has uniform a one-dimensional channel system which has pores that can be small
enough to admit only hydrogen leading to a higher selectivity toward hydrogen.

[37]
The studies on gas separation proved its perm-selectivity to hydrogen in the
separation of hydrogen and ethane mixture [63].
The main objective of this work is to investigate ethane dehydrogenation
reaction in a packed-bed membrane reactor (PBMR) that uses natural zeolite
mordenite as membrane to extract hydrogen from the reaction system. The Pt-
Sn/Al2O3 pellets were used as the catalyst. The PBMR operated in both the
membrane reactor and a conventional packed-bed reactor mode by dramatically
switching the sweeping gas on and off to evaluate the relative effect of the
membrane on ethane conversion.
52
4.2 Experimental
4.2.1 The Catalyst
The catalyst used for the PBMR was 1wt%Pt-0.3wt%Sn/Al2O3 (Alfa Aesar
product, USA, reduced). The spherical alumina pellet size ranged from 1-2 mm.
The catalyst was dried at 110 °C for four hours before use.
4.2.2 Membrane Preparation
Mordenite membrane disks were prepared from the raw natural monolithic
mordenite fragments from Paradise Quarry Limited, New Zealand. The rock pieces
were shaped to 5 × 5 × 5 𝑐𝑚 cubes using a Rock Rascal Model TM Combination
Trim Saw/Grinder equipped with a MK-303 diamond lapidary saw blade. The
cubes were sliced into ∅19 × 1.5 𝑚𝑚 disks by water jet cutting or by hand. The
disks were polished with a rotating diamond polishing lapidary disk (180 mesh,
Fac-Ette Manufacturing Inc.), washed in ultrasonic bath for 30 minutes, and then
dried at 110 °C for four hours. Finished membrane disks were activated at 750 °C
for four hours before use. Figure 4-2 show the steps of membrane disks preparation
from the natural mordenite zeolite.
Figure 4-2 Steps of membrane disk preparation
53
4.2.3 Membrane Reactor Configuration
The packed-bed membrane reactor configuration is shown in Figure 4-3. The
outer tube was made of quartz (ID 22 mm) and used as the feed side of the
membrane reactor. The permeate side consisted of two concentric ceramic tubes.
The sweep gas flowed through the inner tube and the permeate flux flowed out
through the outer tube (OD 12.6 mm). The catalyst pellets were packed in the feed
side in contact with the membrane and held by the quartz wool. The membrane was
sealed on the end of the ceramic outer tube with high temperature ceramic adhesive
(REFRACTOBOND ALP, Accumet Materials Co., USA) by the curing procedures
described in the next section. All tube connections were sealed with rubber O-rings
and Swagelok fittings. Outlet flows of both feed side and permeate side were
measured with a bubble meter.
Figure 4-3 Packed-bed membrane reactor configuration
54
4.2.4 Experiment Conditions
Ethane dehydrogenation reactions were carried out at 500 °C and 550 °C. The
amount of catalyst loaded was 2.2 g at both temperatures. An on-line GC (Bruker
450-GC) with automatic sampling valves and TCD detector was used to analyze
both the permeate and the feed side streams. Inlet flow rates (feed and sweep gas)
were controlled by mass flow controllers (Sierra Instruments Inc.). All gas flow
rates (both inlets and outlets) were measured under lab conditions (21 °C,
atmosphere pressure). Inlet flow rates were controlled by mass flow meters. Outlet
flow rates were measured with bubble meters. All gases used were from Praxair
Canada.
Ethane (99.5%) flow rate was 12.0 ml/min. Argon (99.99%) was used as the
sweep gas and its flow rate was 54.0 ml/min. A higher sweep gas flow rate did not
improve the hydrogen removal from the feed side. Feed side pressure was
maintained at 2.7 kPa(g) with a back pressure regulator valve while the permeate
side pressure was atmospheric. Argon was not detected in the retentate stream when
pure helium was used as a feed stream in the membrane reactor at the same
operating conditions. Feed dilution effects were negligible in the membrane reactor
experiments.
4.2.5 Switching Membrane Reactor Mode to Conventional Reactor Mode
To quantify the enhancement of ethane conversion and evaluate the
performances of the membrane reactor, the ethane dehydrogenation reaction was
switched from packed-bed membrane reactor mode to conventional packed-bed
55
reactor mode by closing off the inlet and outlet of the sweep gas stream. A
schematic demonstration is shown in Figure 4-4. When both inlet and outlet of the
permeate side were shut off, the concentration of permeated species eventually
equalized in both sides. The membrane reactor set-up essentially performed like a
conventional packed-bed reactor. The performances of membrane reactor mode and
conventional reactor mode for ethane dehydrogenation can be fairly compared to
each other.
Figure 4-4 Schematic demonstration of switching between membrane reactor
and conventional reactor mode
4.2.6 Mass Balance of Species at a Reactor and Membrane Reactor Mode
The reaction rate enhancement achieved in the membrane reactor mode could
be quantified considering the mass balance of species in the reaction zone. The
integral form of the continuity equation for species 𝑖 in steady-state condition is
expressed as:
56
∫𝐴 𝑁̅𝑖 ∙ 𝑛̅ 𝑑𝐴 = 𝑊 ∫𝑉 𝑟∗𝑖 𝑑𝑉 (4-2)

𝑅 𝑅
where the term on the left-hand side represents the net molar flow rate of
species i through the boundaries of the reaction zone, 𝑟𝑖∗ is the local reaction rate in
𝑚𝑜𝑙/(𝑐𝑚3 ∙ 𝑔𝑐𝑎𝑡 ∙ 𝑠) and is the catalyst weight in the reactor. Integrated both
sides, the equation (4-2) became:
𝐹𝑖,𝑜𝑢𝑡 − 𝐹𝑖,𝑖𝑛 = 𝑊〈𝑟∗𝑖 〉𝑉𝑅 (4-3)
where 𝐹𝑖 is the molar flow rate of each reacting species and 〈𝑟𝑖∗ 〉𝑉𝑅 is the
volume-integrated reaction rate in 𝑚𝑜𝑙/(𝑔𝑐𝑎𝑡 ∙ 𝑠).
The integral mass balance was applied for each product species of ethane
dehydrogenation (H2 and C2H4) in both the conventional reactor and the membrane
reactor mode. The combination of each resulting expression gives:
〈𝑟∗𝑖 〉𝑉𝑅,𝑀𝑅 [𝑄𝑜𝑢𝑡 𝑥𝑖,𝑜𝑢𝑡 ]𝑀𝑅 [𝑄𝑝 𝑥𝑖,𝑝 ]

= + (4-4)
〈𝑟∗𝑖 〉𝑉𝑅,𝑅 [𝑄𝑜𝑢𝑡 𝑥𝑖,𝑜𝑢𝑡 ]𝑅 [𝑄𝑜𝑢𝑡 𝑥𝑖,𝑜𝑢𝑡 ]𝑅
Equation (4-4) can also be expressed in terms of dimensionless numbers
𝜏𝑖,𝑀𝑅/𝑅 = 𝛾𝑖,𝑀𝑅/𝑅 + 𝛽𝑖,𝑃/𝑅 (4-5)
57
Where 𝜏𝑖,𝑀𝑅/𝑅 is the reaction rate ratio of product species between the
membrane reactor (MR) and the conventional reactor (R) mode, 𝛾𝑖,𝑀𝑅/𝑅 is the
outlet flow rate ratio for product species between the MR and the R mode, and
𝛽𝑖,𝑃/𝑅 is the ratio between product permeation flow rate over its formation rate in
the R mode. The dimensionless number 𝜏𝑖,𝑀𝑅/𝑅 represents the reaction rate
enhancement achieved by the system as it switches from R to MR mode. It can be
estimated in terms of the flow rate leaving the reaction zone, 𝑄𝑜𝑢𝑡 ; the outlet
permeate flow rate, 𝑄𝑝 and the outlet compositions from the reaction and permeate
zones, 𝑥𝑖,𝑜𝑢𝑡 and 𝑥𝑖,𝑝 .
All the dimensionless numbers defined for the analysis of the experimental
data are summarized in Table 4-1. They are also related to the Damköhler and
membrane Peclet numbers [64] estimated under reaction conditions. The Damköhler
number relates the rate of reaction to the rate of reactant feed. The membrane Peclet
number is defined in terms of the reactant feed rate and permeation rate of each
product.
58
Table 4-1 Dimensionless numbers defined for the comparison between reactor
and membrane reactor mode
Dimensionless number Symbol Definition
〈𝑟𝑖∗ 〉𝑉𝑅,𝑀𝑅
Reaction rate ratio 𝜏𝑖,𝑀𝑅/𝑅 𝜏𝑖,𝑀𝑅/𝑅 =
〈𝑟𝑖∗ 〉𝑉𝑅,𝑅
[𝑄𝑜𝑢𝑡 𝑥𝑖,𝑜𝑢𝑡 ]𝑀𝑅

Outlet flow rate ratio 𝛾𝑖,𝑀𝑅/𝑅 𝛾𝑖,𝑀𝑅/𝑅 =
[𝑄𝑜𝑢𝑡 𝑥𝑖,𝑜𝑢𝑡 ]𝑅
Permeation/reaction rate 𝛽𝑖,𝑃/𝑅 𝛽𝑖,𝑃/𝑅 =

[𝑄𝑝 𝑥𝑖,𝑝 ]
=
1
ratio [𝑄𝑜𝑢𝑡 𝑥𝑖,𝑜𝑢𝑡 ]𝑅 𝐷𝑎𝑅 𝑃𝑒
𝑥H2,𝑜𝑢𝑡
H2/ethane outlet molar ratio 𝜑H2⁄C2 𝜑 H2 ⁄C 2 =
𝑥C2,𝑜𝑢𝑡
〈𝑟𝑖∗ 〉𝑅,𝑅 𝑊 1 𝐹𝑖,𝑝

𝐷𝑎𝑅 = ; =
𝐹𝑖𝑛 𝑃𝑒 𝐹𝑖𝑛
The reaction enhancement can also be evaluated using ethane conversion,
ethylene yield and ethylene selectivity. Ethane conversion was calculated in terms
of the reaction products. For the reactor mode the ethane conversion is expressed
as
𝑥𝐶2 ,𝑜𝑢𝑡
𝛼𝐶2 = (1 − ) × 100% (4-6)
𝑥𝐶2 ,𝑜𝑢𝑡 +𝑥𝐶= ,𝑜𝑢𝑡 +0.5𝑥𝐶𝐻4 ,𝑜𝑢𝑡
2
where C2 and C2= represent the ethane and ethylene species respectively. The
ethylene yield was expressed as:
59
𝑥𝐶= ,𝑜𝑢𝑡
2
𝑌𝐶2= = ( ) × 100% (4-7)
𝑥𝐶2 ,𝑜𝑢𝑡 +𝑥𝐶= ,𝑜𝑢𝑡 +0.5𝑥𝐶𝐻4 ,𝑜𝑢𝑡
2
Ethylene selectivity can be expressed as:
𝑌𝐶= 𝑥𝐶= ,𝑜𝑢𝑡

2 2
𝑆𝐶2= = =( ) × 100% (4-8)
𝛼𝐶2 𝑥𝐶= ,𝑜𝑢𝑡 +0.5𝑥𝐶𝐻4 ,𝑜𝑢𝑡
2
4.3 Results and Discussions
4.3.1 Reversibility in the Reactor-Membrane Reactor Mode
The system functioned as a membrane reactor (MR) by introducing the sweep
gas. When a steady state was reached, it was switched to the reactor mode (R) by
shutting off the sweeping gas. Experimental results are shown in Figure 4-5. When
the sweep gas was shut off, hydrogen concentration on the feed side increased, and
ethylene concentration decreased accordingly. When the system was switched back
to MR mode by turning on the sweeping gas, the concentration profiles gradually
resumed back to the levels at the beginning of the MR steady state. The
experimental data demonstrates the reversibility between the conventional reactor
and membrane reactor modes.
60
10.0
Membrane reactor Reactor Membrane reactor
9.0
C2H6 x 1/10
Outlet molar fraction (%) 8.0
7.0
6.0
H2
5.0
4.0
C2H4
3.0
2.0
1.0
0.0
0 20 40 60 80 100 120 140 160
Time Lapse ∆t (min)
Figure 4-5 Dynamics of molar fractions of H2, C2H4, and C2H6 in the reactor outlet as the
system switched reversibly between a membrane reactor and a reactor operating mode at 500
°C, WHSV = 0.4 h-1
4.3.2 The Conventional Reactor and Membrane Reactor Modes: Dynamics
of Molar Fractions of Reaction Species at the Reactor Outlet
Figure 4-6 shows the steady state concentrations of ethane, hydrogen, and
ethylene for both the membrane reactor (MR) and conventional reactor mode (R).
When switched from membrane reactor mode to conventional reactor mode,
hydrogen concentration increased, ethylene concentration decreased
simultaneously, and hydrogen concentration became higher than that of ethylene.
Figure 4-7 reveals the concentration changes of hydrogen and ethylene. Hydrogen
molar fraction increased from 4.5% to 4.8% when the system was switched from
membrane reactor mode to conventional reactor mode. Ethylene molar fraction was
61
higher in the membrane reactor mode than in conventional reactor mode. The
measured data were: [𝑄𝑜𝑢𝑡 ]𝑀𝑅 = 12.6 𝑚𝑙/𝑚𝑖𝑛 , [𝑄𝑜𝑢𝑡 ]𝑅 = 12.9 𝑚𝑙/𝑚𝑖𝑛 ,
permeate rate: 55, permeate compositions: 𝑥H2,𝑜𝑢𝑡 = 0.19 𝑣% , 𝑥C2=,𝑜𝑢𝑡 =
0.025 𝑣%, 𝑥C2 ,𝑜𝑢𝑡 = 0.69 𝑣%.
9.0 Equilibrium
C2H6 x 1/10
8.0
Membrane reactor Reactor
Outlet molar fraction (%)
7.0
6.0
H2
5.0
4.0 Equilibrium
C2H4
3.0
2.0
1.0
0.0
0 20 40 60 80 100 120 140
Figure 4-6 Dynamics of outlet molar fractions of C2H6, H2, and C2H4 between membrane
reactor and reactor mode at 500 °C, WHSV = 0.4 h-1
62
6.0
Membrane reactor Reactor
5.5
Outlet molar fraction (%) 5.0

C2H4 H2
4.5
Equilibrium
4.0
3.5
3.0
0 20 40 60 80 100 120
Figure 4-7 Dynamics of outlet molar fractions of H2 and C2H4 between membrane reactor
and reactor mode at 500 °C. , WHSV = 0.4 h-1
As shown in Table 4-2, the reaction rate ratio, 𝜏C=2 ,𝑀𝑅/𝑅 , calculated with
Equation (4-5) in terms of ethylene was 1.05, which indicates that the ethylene
production rate in the MR mode was higher than that in the R mode. The ethylene
yields (𝑌C=2 ) calculated with Equation (4-7) in the R and MR mode were 4.6% and
4.9% respectively. The results indicate that an ethylene yield enhancement of 6.5%
was reached in the MR mode. Ethane conversion increased from 4.9% in the R
mode to 5.2% in the MR mode at 500 °C.
The conversion improvement can also be analyzed in terms of the hydrogen
permeation rate in relation to its generation rate. The ratio of permeation/reaction
rate, 𝛽H2 ,𝑝/𝑅 , was 16%, which meant that about 16% of the hydrogen generated in
the reaction was extracted through the membrane. The value 𝛽H2 ,𝑝/𝑅 =
63
16% corresponds to a Damköhler-Peclet number of 𝐷𝑎𝑅 𝑃𝑒 = 6.1 This number is

[65]
bigger than the optimal value of membrane reactor effectiveness (𝐷𝑎𝑅 𝑃𝑒~1) .
The number approaching one means all the hydrogen produced in the
membrane reactor is extracted through the membrane.
Table 4-2 Experimental values of the dimensionless numbers defined in Table 4-1 for R
and MR modes at 500 °C and 550 °C
Mode R MR R MR R MR
T, ℃ 500 500 500 500 550 550
𝐴/𝑉𝑅 , 𝑚−1 0.04 0.04 0.16 0.16 0.04 0.04
𝛽H2 ,𝑝/𝑅 , % - 16.7 - 36.6 - 24.2
𝜑H2⁄C2 , % 5.3 5.1 5.8 4.6 11.2 9.9
𝛽C2=,𝑝/𝑅 , % - 2.4 - 9.4 - 8.7
𝜏C=2 ,𝑀𝑅/𝑅 - 1.05 - 1.15 - 1.09
𝛼C2 , % 4.9 5.2 4.8 5.5 9.7 10.5
𝑌C=2 , % 4.6 4.9 4.5 5.2 9.0 9.8
𝑆𝐶2= , % 94.0 94.1 92.4 94.8 92.2 93.7
𝑆CH4 , % 6.0 5.9 7.6 5.2 7.8 6.3
4.3.3 Increasing the Ratio of Permeation Area/Reactor Volume (𝑨/𝑽𝑹 ) of the
Membrane Reactor
Increasing the effective permeation area of the membrane in a membrane
reactor can extract more hydrogen from the reaction system while the hydrogen
generation rate in the conventional reactor mode remains the same under the same
64
reaction temperature. Therefore, the ratio of permeation rate/reaction rate, 𝛽H2 ,𝑝/𝑅 ,
increased. The measured data is: [𝑄𝑜𝑢𝑡 ]𝑀𝑅 = 12.1 𝑚𝑙/𝑚𝑖𝑛, [𝑄𝑜𝑢𝑡 ]𝑅 = 12.5 𝑚𝑙/
𝑚𝑖𝑛, permeate rate: 55.8, permeation compositions: 𝑥H2 ,𝑜𝑢𝑡 = 0.43 𝑣%, 𝑥C2=,𝑜𝑢𝑡 =
0.09 𝑣%, and 𝑥C2 ,𝑜𝑢𝑡 = 1.33 𝑣%.
Figure 4-8 shows the feed and product concentration changes in the two
operation modes of ethane dehydrogenation. The membrane reactor had a larger
membrane area. With the large-area membrane, the hydrogen molar fraction
dropped considerably when the system was switched from conventional packed-
bed reactor mode to membrane reactor mode. The ethylene molar fraction values
increased beyond the thermodynamic equilibrium.
65
10.0
Reactor Membrane reactor
9.0 Equilibrium
C2H6 x 1/10
8.0
Outlet molar fraction (%)
7.0
6.0
H2
5.0
4.0 Equilibrium
C2H4
3.0
2.0
1.0
0.0
0 10 20 30 40 50 60 70 80
reactor and membrane reactor mode at 500 °C, WHSV = 0.4 h-1 with bigger
membrane area (𝑨/𝑽𝑹 ≅ 𝟎. 𝟏𝟔)
For the larger ratio of effective membrane area to reactor volume, 𝐴/𝑉𝑅 =
0.16 , the dimensionless number is 𝛽H2 ,𝑝/𝑅 = 36% , while for the smaller one
(𝐴/𝑉𝑅 = 0.04), the dimensionless number is 𝛽H2 ,𝑝/𝑅 = 16%, as shown in Table
4-2. Similarly, the H2/C2H6 molar ratio at the reactor outlet 𝜑H2⁄C2 decreased 80%
of its original value when the system was switched from the conventional packed-
bed reactor mode to the membrane reactor mode, while the 𝜑H2⁄C2 value only
decreased 3.8% (from 5.3% in the conventional reactor mode to 5.1% in the
membrane reactor mode) for the lower ratio of 𝐴/𝑉𝑅 = 0.04. Larger permeation
area favored more hydrogen pass through the membrane, and thus more hydrogen
66
was extracted from the reaction zone. As a result, the reaction rate, 〈𝑟𝑖∗ 〉𝑉𝑅,𝑀𝑅 , in the
membrane reactor mode increased and the ethane conversion, 𝛼C2 , improved in
comparison to the membrane reactor with a smaller permeation area. The
corresponding value for the dimensionless number is 𝜏C=2 ,𝑀𝑅/𝑅 = 1.15, indicating a
15% increase in the reaction rate for a membrane reactor mode. Ethylene yield
improved about 17%, from 4.4% in the conventional reactor mode to 5.2% in the
membrane reactor mode.
4.3.4 Effect of Reaction Temperatures
Figure 4-9 shows the molar fractions of the products of ethane
dehydrogenation in the conventional reactor mode and the membrane reactor mode
at 550 °C. The ethylene selectivity at 550 °C decreased compared to that at 500 °C
(see Table 4-2) because more side reactions occurred at the higher temperature. The
reaction rate ratio, 𝜏C=2 ,𝑀𝑅/𝑅 = 1.09, was larger than the corresponding value at
500 °C. The ethane conversion at 550 °C increased from 9.7% in reactor mode to
10.5% in the membrane reactor mode. Other measured data were: [𝑄𝑜𝑢𝑡 ]𝑅 =
12.4 𝑚𝑙/𝑚𝑖𝑛 , [𝑄𝑜𝑢𝑡 ]𝑀𝑅 = 12.2 𝑚𝑙/𝑚𝑖𝑛 , Permeate rate: 54.2, permeate
composition [%v/v]: 𝑥H2 ,𝑜𝑢𝑡 = 0.50 𝑣% , 𝑥C=2 ,𝑜𝑢𝑡 = 0.16 𝑣% , and 𝑥C2 ,𝑜𝑢𝑡 =
0.60 𝑣%.
67
11.0 9.0
Outlet molar fractions of H2 and C2H4 (%)

10.0 8.0
Outlet molar fraction of C2H6 (%)

C2H6 x 1/10
9.0 H2 7.0
8.0 6.0
C2H4
7.0 5.0
Reactor Membrane reactor
6.0 4.0
5.0 3.0
0 10 20 30 40 50 60 70 80
reactor and membrane reactor mode at 550 °C, WHSV = 0.4 h-1
Higher reaction temperature led to higher product concentraion, and thus the
driving force for product permeation increased. Hydrogen and ethylene permeate
fluxes at 550 °C increased comparing to experimental data at 500 °C. The
dimensionless numbers 𝛽H2,𝑝/𝑅 = 24.2% , and 𝛽C2=,𝑝/𝑅 = 8.7% at 550 °C were
larger than the corresponding values at 500 °C with the same 𝐴/𝑉𝑅 ratio (see Table
4-2). At 500 °C, 𝜑H2⁄C2 ratio dropped about 3.8% of the original value when the
system was switched from the conventional ractor mode to the membrane reactor
mode, while it dropped about 11.6% of the original value at 550 °C. Product
extraction effectiveness and product concentration of the membrane reactor system
increased at the higher temperature. Ethylene yield increased by 6.5% of the
original value by switching from reactor mode to membrane reactor mode at 500 °C,
68
while the yield increased 8.9% of the original value at 550 °C. The ethylene
selectivity slightly increased in the membrane reactor mode than in the
conventional reactor mode at 550 °C.
4.3.5 Membrane Reactor Effectiveness
As shown in Table 4-3, when the temperature increased from 500 °C to 550
°C, the Damköhler number (𝐷𝑎𝑅 ) doubled as the reaction rate increased twice with
the temperature. The increase of 𝐷𝑎𝑅 is also reflected in the conversion values risen
from ~5% at 500 °C to ~10% at 550 °C, as shown in Table 4-2. However, the larger
formation rate of hydrogen was compensated with a higher hydrogen removal rate.
The hydrogen permeation rate was three times larger at 550 °C than at 500 °C. The
increase of hydrogen permeation is also reflected in the 1⁄𝑃𝑒 values shown in
Table 4-3. Consequently, the Damköhler-Peclet number (𝐷𝑎𝑅 𝑃𝑒) decreased from
6.1 to 4.2. Thus the effectiveness of the membrane reactor increased at higher
temperature. The ethylene yield enhancement at 550 °C was 9.4%, which was
higher than that at 500 °C. When the membrane area-to-volume ratio 𝐴/𝑉𝑅
increased from 0.04 to 0.16, the permeation/reaction rate ratio 𝛽H2 ,𝑝/𝑅 increased up
to 36% at 500 °C. This is equivalent to the 𝐷𝑎𝑅 𝑃𝑒 number decreasing to 2.7 . The
membrane reactor can work more effectively as 𝐷𝑎𝑅 𝑃𝑒 number approaches 1. The
ethylene yield enhanced about 17% at 500 °C in the membrane reactor mode.
Higher hydrogen removal rate verse its formation rate resulted in a larger ethylene
yield enhancement. The membrane reactor mode with the smallest 𝐷𝑎𝑅 𝑃𝑒 number
led to the largest ethylene yield improvement.
69
Table 4-3 Ethylene yield enhancements achieved with the membrane reactor at two different
reaction temperatures and two 𝑨⁄𝑽𝑹 ratios. The corresponding Damköhler-Peclet numbers
were calculated with reaction conditions. WHSV = 0.4 h-1
Temperature, °C 500 550 500
𝐴/𝑉𝑅 , 𝑚−1 0.04 0.04 0.16
1
,× 102 0.8 2.4 2.0
𝑃𝑒
𝐷𝑎𝑅 ,× 102 4.9 10.0 5.4
𝐷𝑎𝑅 𝑃𝑒 6.1 4.2 2.7
𝑌C=2 ,𝑒𝑛ℎ𝑎𝑛𝑐𝑒𝑚𝑒𝑛𝑡 , % 6.1 9.4 16.5
4.4 Conclusions
A membrane reactor with a natural mordenite membrane disk coupled with a
Pt/Al2O3 catalyst packed bed was able to selectively extract hydrogen and to shift
the equilibrium of ethane dehydrogenation reaction at 500-550 °C. By increasing
the area/volume ratio of the membrane reactor module, a 15.6% ethylene yield
enhancement was achieved in comparison to that of the conventional packed bed
reactor at 500 °C. The effectiveness of the membrane reactor improved since more
hydrogen than the amount formed was removed from the system. The membrane
reactor showed a higher ethylene yield enhancement at 550 °C than at 500 °C, due
to the increase in hydrogen permeation rate. Lower Damköhler-Peclet number
favored the higher ethylene yield improvement. The larger membrane disk at
500 °C provided the largest enhancement in ethylene yield.
70
Part II: Ethane Dehydrogenation Catalyst Development
Chapter 5. Ethane Dehydrogenation Catalyst Development
Chapter 5. Ethane Dehydrogenation Catalyst
Development
5.1 Introduction
Ethylene is the most important organic chemical for the petrochemical
industry. It is mainly manufactured by steam cracking of hydrocarbon feedstocks,
ranging from natural gas to heavy gas oils. The steam cracking technology is a
[15] [66]
highly energy intensive process with low selectivity (~75% per pass) .
Ethylene production from ethane (main component in natural gas or shell gas) is
the most economical and efficient process due to the abundant reserves and easy
exploitation of natural gas, and the high selectivity as shown in Table 5-1 . However,
compared to other feedsotcks, ethane cracking occurs at a much higher reaction
temperature. For example, a typical furnace outlet temperature for cracking ethane
is approximately 800 °C, while the temperature for cracking naphtha or gas oils is
about 675-700 °C.
In order to lower the energy consumption and increase the conversion of
ethane dehydrogenation at lower temperatures, the catalytic membrane reactor was
developed as a potential technology for commercial ethylene manufacturing. The
use of the membrane reactor can improve conversion by removing hydrogen from
the reaction system, and thus shifting the equilibrium to the direction favoring the
production of ethylene.
72
Table 5-1 A product yield comparison using different feedstocks [13]

Ethane Propane Butane Naphtha AGO
P/E, kg/kg 65 90 0.35 0.55 0.54
Steam dilution, kg/kg 0.3 0.3 0.486 0.45 0.8
Residence time, s 0.3451 0.3337 0.344 0.3572 0.32
Yield, wt%
H2 4.05 1.55 1.09 0.92 0.71
CH4 3.52 23.27 20.29 14.91 10.58
C2H2 0.47 0.51 0.42 0.47 0.38
C2H4 52.31 37.51 35.81 28.45 25.93
C2H6 35.03 2.8 4.16 4.23 2.82
C3H4 0.02 0.57 0.93 0.76 0.63
C3H6 1.13 14.82 17.24 15.64 14.07
C3H8 0.12 9.96 0.35 0.51 0.37
C4H4 0.05 0.08 0.1 0.13 0.12
C4H6 1.8 2.9 4.08 4.79 5.73
C4H8 0.19 1 2.83 4.52 3.61
C4H10 0.21 0.04 6.07 0.44 0.05
Benzene 0.47 2.12 2.69 7.12 5.44
Toluene 0.07 0.4 0.77 3.1 3.49
Xylenes 0 0.05 0.11 1.16 0.92
Ethylbenzene 0 0.01 0.01 0.62 0.37
Styrene 0.02 0.2 0.24 1.14 1.18
Pyrolysis gasoline 0.32 1.26 1.81 7.96 7.28
Pyrolysis fuel oil 0.16 0.89 0.94 3.09 16.3
E+P+B olefin 53.63 53.33 55.88 48.61 43.61
73
The ethane dehydrogenation reaction is more endothermic than that of heavier
alkanes like butanes, ethylbenzene because methyl C-H bonds are stronger than
methylene C-H bonds. It needs much higher reaction temperature to achieve the
practical conversions. The use of catalyst can lower activation energy because the
catalysts provide an alternative route to steam cracking by decreasing the activation
energy of C-H bonds. Moreover, C-H bonds in ethane (423 kJ/mol) are much
stronger than C-C bond (347~356 kJ/mol). C-H bonds in ethane are among the
[67]
strongest single bonds in nature . In order to achieve a high yield of ethylene
production, C-H bonds need to be selectively activated while keeping the C-C bond
stable. In this way, the side reactions such as coking and methane production are
suppressed and high selectivity can be achieved.
Catalytic dehydrogenation of ethane and other light alkanes at elevated
temperatures traces back to early 1900s when transition metals such as Co, Ni, Cu,
[68]
Fe, Ag, Pt, and Pd were used as catalysts. Frey and Huppke carried out the
equilibrium dehydrogenation of ethane, propane, and butane from 300 °C to 500 °C
on chromic oxide gel catalyst in 1933, and found the catalysts lost activity at higher
[69]
temperature. Fabian and Robertson examined ethane dehydrogenation in a
vacuum on an incandescent Pt filament at temperatures between 1300 K to 1800 K.
A mass spectrometry was used to detect free radicals and atoms evaporating from
the filament. At very low pressure, the dehydrogenation reaction was a first order
reaction. Pt filament lost reactivity quickly and residual oil vapor was observed.
They found water and heavy water (D2O) vapor could reactivate the poisoned
filament, but oxygen and hydrogen could not.
74
Instead of unsupported catalysts, Pt was ussed on a variety of supports as

[70]
ethane dehydrogenation catalyst. Cusumano et al. revealed that Pt was more
highly dispersed on alumina than on silica-alumina. The higher reactivity of
catalyst for cyclohexane dehydrogenation was attributed to the greater
[71-73] [70]
dispersion of Pt on alumina support. , , , ,
[74]
and were studied as the supports for the dehydrogenation catalysts.
[75]
, carbon [76], hydrotalcite ( ) [77-81], monoliths [82, 83], [84]
, and
[85]
carbon nanotubes were also exploited as the supports for the dehydrogenation
catalysts.
Since 1960s the bimetallic platinum catalyst systems were developed to
improve the performances of reforming catalysts. A second and/or more metal
components were used as promoters. For ethane direct catalytic dehydrogenation,
Sn is a widely used promoter for the supported Pt catalyst. Bell’s group [80, 81, 86, 87]
developed catalysts for ethane, propane, and butane
dehydrogenation. The addition of a second or third component improved ethylene
selectivity significantly up to 100%. The bimetallic system also used in

[88] [77]
ethane oxidative dehydrogenation reactions and propane dehydrogenation .
Besides Sn, Fe [89], Rh [90], and Au [91] were also investigated as a second component
of Pt dehydrogenation catalysts.
Mechanisms of ethane dehydrogenation on Pt and supported Pt catalysts have
been extensively investigated. Loaiza et al. [92] proposed that ethane was adsorbed
dissociatively into ethyl, which dehydrogenated into ethylene ( )
followed by a transformation into ethylidyne ( ) and bond scission.
75
[66]
Vincent et al. proposed a mechanism of generating methane by-product and
[93]
coke deposit. Adlhart and Uggerud used a Fourier transform ion cyclotron
resonance mass spectrometer (FTICR) for investigated the dehydrogenation
reaction of methane, ethane, and propane on gaseous platinum cluster cations
( ). Ethane dehydrogenation occurred through
[94]
elimination mechanism. Ju et al. proposed a and
bond activation model on a di-platinum cluster. They found the
dehydrogenation pathway ( ) was favorable in thermodynamics, and
the deethylenation pathway ( ) was favorable in kinetics. Both
and cleavages were thermodynamically favorable, but scission was
kinetically preferred and scission was kinetically hindered. They also
concluded that the two bonds from the two carbon atoms synchronously broke
on cluster and then ethylene kinetically released. General elemental reaction
steps proposed by Loaiza, Vincent, Adlhart and Uggerud, and Ju et al. are consistent
with each other.
Yagasaki and Masel [95] found that coke and methane were mainly formed on
the surface by the decomposition of adsorbed ethylene ( )
at elevated temperatures. The key to supress coke deposits and methane by-product
was to prevent the re-adsorption of ethylene onto the Pt surface [80]. Rodriguez and
Kuhn [96] investigated chemical and electronic properties of Pt bimetallic surfaces,
and concluded that the formation of Pt-Zn bimetallic bond on surfaces
caused the density of Pt 5d electrons to shift to 6s and 6p. Zn has the similar
electronic perturbations as early transition metals (Group 3, 4 metals) to Pt, but
76
larger than late transition metals (Group 8-11 metals, by IUPAC definition). The
electronic perturbations caused by Zn on Pt weakened the strength of the
bonding interactions, which reduce the carbon monoxide chemisorption ability of
Pt.
Some general reaction schemes of ethane dehydrogenation on Pt are
illustrated in Reactions (5-1) to (5-12). Ethane adsorbed on Pt surface and
dissociated hydrogen atom and ethyl group (Reactions (5-1), (5-2), and (5-4)).
Hydrogen atom reacted with absorbed ethane to produce hydrogen gas (Reaction
(5-3)).
(5-1)
(5-2)
(5-3)
Adsorbed ethyl group further dissociated to form ethylene (Reaction
(5-5), then converted to ethylene (Reaction (5-6), followed by a kinetic
desorption step (Reaction (5-7)).
77
(5-4)
(5-5)
(5-6)
(5-7)
The ethylene transformed to ethylidyne (Reaction (5-8) on ) and
followed a scission to form carbon deposit and methane.
(5-8)
(5-9)
When ethylene re-adsorbed on (Reaction (5-9)), it transformed
into a strained group (Reaction (5-10), and followed a disproportionation
to yield methane and carbon deposit [95] (Reaction (5-11)).
(5-10)
(5-11)
(5-12)
78
Heterogeneous catalysts are a complex system. Not only the supported active
metal components but also the support itself plays a key role in determining the
catalyst performances. During the reactions, both the active components and the
support interact with reactants. Catalyst performances can be optimized by
enhancing the synergetic effects and adjusting the interactions among active
components and the supports [97, 98]. This work reports the development and a study
of a novel catalyst with high reactivity and selectivity for ethane dehydrogenation:
Zn promoted Pt supported on ETS-2.
5.2 Experimental
5.2.1 Catalyst Preparation
A series of Pt-Zn/ETS-2 and Pd-Zn/ETS-2 catalysts were prepared by the
incipient wetness impregnation method. ETS-2 is a sodium titanate prepared by
alkaline digestion of nano-particles of titania. The caustic digestion converts the
surface titania atoms to charged titanate species whose anionic charge is offset by
sodium or hydrogen ions. A commercial sample marked “filter cake” was used to
prepare the catalysts. The filter cake represents the material directly discharged
from the reactor, filtered, and washed without any other pre-treatment steps. The
filter cake was calcined at 600 °C for four hours and crushed into 30 to 50 mesh
pellets as catalyst support. The ETS-2 support pellets were modified with ZnO and
CeO2 by incipient wetness impregnation method.
Metal precursors were prepared from aqueous solutions: 0.1M H2PtCl6
(Hydrogen hexachloroplatinate (IV) hydrate, Pt 40%, Acros Organics), 0.2 M

79
Zn(NO3)2 (Zinc nitrate hexahydrate crystalline, Fisher Scientific Canada), 0.1M
(NH4)2Ce(NO3)6 (Ammonium cerium(IV) nitrate, 98.5%, Acros Organics), 0.1M
PdCl2 (Palladium (II) chloride, Pd 59.5%, Alpha Aesar, USA) (in 28wt%
ammonium hydroxide solution, Fisher Scientific Canada).
All ETS-2 supports were calcined at 600 °C for four hours before
impregnating metal components. Support ETS-2 was modified by Zn with different
contents and then impregnated with 1 wt% of Pt. All catalysts were calcined at
500 °C for four hours except for the specified temperatures (600 and 650 °C). In
the same procedure, for the other metal catalysts, the ETS-2 support was
impregnated with the solutions mentioned above respectively, dried at room
temperature overnight (> 8 hours), and then calcined at 500 °C for four hours. All
catalysts mentioned in this work were prepared by the same procedure.
5.2.2 Catalysts Testing in Ethane Dehydrogenation Reaction
To evaluate the catalyst performance in ethane dehydrogenation reaction, a
reactor as shown in Figure 5-1 was set up. The reactor was a quartz tube (ID 22
mm). The catalyst bed was supported by quartz wool on both sides. Both reduced
and non-reduced catalysts were dried at 110 °C for four hours before use. The
catalysts were reduced in situ with argon diluted hydrogen (40 v% H2). The
reduction conditions were: diluted hydrogen flow rate 40 ml/min; temperature
range from room temperature to 300 °C with the heating rate 2 °C/min; at 300 °C
for one hour.
80
The amount of the catalyst loaded on each run was 3.0 g. An on-line GC
(Bruker 450-GC) was used. The ethane feed (99.5%, Praxair) flow rate was
controlled by a mass flow meter (Sierra Instruments Inc.). The outlet flow rate was
measured by bubble meter under lab conditions.
The selectivity of ethylene was calculated by Equation (5-13).
[𝐶2 𝐻4 ]
𝑆𝐶2 𝐻4 = [𝐶 × 100% (5-13)
2 𝐻4 ]+0.5[𝐶𝐻4 ]
Figure 5-1 The scheme of packed-bed reactor for ethane dehydrogenation
5.3 Results and Discussion
5.3.1 Catalyst Characterization
5.3.1.1 BET and Pt Dispersion
Table 5-2 summarizes the BET specific surface areas of ETS-2, Al2O3
supported Pt catalysts, and Pt dispersion on the catalysts. From the table, it can be
seen that the commercial ETS-2 filter cake dried at 150 °C has a high surface area.
81
When the sample was calcined at 600 °C for four hours, the surface area dropped
significantly (about 36%). The huge difference of surface areas implies
microstructure changes at elevated temperature.
When 1 wt% of Pt was introduced in the ETS-2 support, the specific surface
area of 1%Zn/ETS-2 catalyst slightly increased. As the Zn content added into the
support increased, the specific surface areas of the catalysts gradually decreased,
and leveled off after the Zn content reached 5 wt%. The surface area decrease could
be caused by the micropores plugged by ZnO. Solid state reaction occurred at
higher temperature, as shown by the XRD results in Figure 5-4.
Table 5-2 BET Surface Area and Pt Dispersion of Catalysts

Catalyst (support) Specific Pt surface Pt mean
surface area, dispersion, crystal size,
(m2/g) % nm
ETS-2 (150 °C dried) 189.7 - -
ETS-2 (600 °C calcined) 68.5 - -
1%Pt/ETS-2 83.3 2.02 56.0
0.5%Zn, 1%Pt/ETS-2 50.0 8.21 13.8
1%Zn, 1%Pt/ETS-2 51.4 10.75 10.5
2%Zn, 1%Pt/ETS-2 42.4 3.54 32.0
5%Zn, 1%Pt/ETS-2 28.4 5.72 19.8
7.5%Zn, 1%Pt/ETS-2 28.4 3.42 33.0
5%Zn, 1%Pt/ETS-2 (500 °C, spent) 28.7 3.41 33.2
5%Zn, 1%Pt/ETS-2 (600 °C 24.3 0.97 116.4
calcined)
5%Zn, 1%Pt/ETS-2 (650 °C, spent) 22.4 0.05 2464.2
0.5%Pt/Al2O3 (standard) 21.72 4.40
1%Zn, 1%Pt/Al2O3 41.14 2.75
82
The relationship between Zn content and Pt surface dispersion and particle
sizes is illustrated in Figure 5-2. From the figure, it can be seen that the Pt dispersion
on the catalysts dramatically increased with Zn content, and reached the maximum
at 1wt% of Zn. Further increase in the Zn content led to rapid reduction in Pt
dispersion. When Zn content reached 5 wt% any additional Zn could not cause any
change in the dispersion, but the dispersion of Pt on the catalyst was still higher
than that on the catalyst without Zn. The Pt crystal size change followed the
opposite trend. Pt particle size decreased exponentially with the Pt dispersion, as
shown in Figure 5-3. The introduction of Zn changed the ETS-2 surface properties,
and therefore enhanced the Pt dispersion on the surface and Pt particle size
reduction. The higher dispersion of Pt on the surface, the more active sites they can
provide, and a higher activity of the catalyst can be achieved.
60 12
Pt mean particle size
Pt Dispersion on Catalysts, %
50 10
Pt dispersion
Pt Mean Particle Size, nm
40 8
30 6
20 4
10 2
0 0
0 1 2 3 4 5 6 7 8
Zn Content in the Catalysts 1%Pt/ETS-2, wt%
Figure 5-2 Pt mean particle size and dispersion change with Zn content
83
60
50
Pt Particle Size, nm
40
30
20
10
0
0 2 4 6 8 10 12
Pt Dispersion, %
Figure 5-3 Pt particle size change with dispersion on catalysts
For alumina supported Pt catalysts, adding Zn enhanced the Pt dispersion
significantly. The dispersion results of Pt on ETS-2 and alumina indicate that Zn
could improve Pt dispersing on supports. When the catalysts with 5wt% of Zn
calcined and/or tested at 600 °C or higher the Pt dispersion decreased significantly,
but the specific surface area dropped slightly.
5.3.1.2 XRD
Figure 5-4 shows the XRD patterns of ETS-2 dried at 150 °C and calcined at
600 °C for four hours. For the dried sample, the TiO2 was in the form of anatase.
After calcined at 600 °C for four hours, some anatase phase was converted into
rutile phase. Figure 5-5 indicates the XRD patterns of the catalyst 5%Zn,
1%Pt/ETS-2 treated at different temperatures. From the figure, it can be seen that
new characteristic peaks are present in all catalyst’s patterns, which imply loaded
84
metal components reacted with the support at elevated temperatures. It is unclear
whether those reactions will affect catalyst performance until further experiments
are conducted.
● ▲: Rutile
●: Anatase
●
Intensity (Counts)
ETS-2, 600°C
▲ ▲
● ●● ●
▲
▲▲ ●●
●
●
ETS-2, 150°C
●● ● ●●
0 10 20 30 40 50 60 70 80 90
2-Theta (degree)
Figure 5-4 XRD patterns of ETS-2 dried at 150 °C and calcined at 600 °C for
four hours
85
(f)
(e)
Intensity (Counts)
(d)
(c)
(b)
(a)
0 10 20 30 40 50 60 70 80 90
2-Theta (degree)
Figure 5-5 XRD patterns of: (a) ETS-2 dried at 150 °C , (b) ETS-2 treated at
600 °C, (c) 5%Zn, 1%Pt/ETS-2 (500 °C calcined), (d) 5%Zn, 1%Pt/ETS-2
(500 °C tested), (e) 5%Zn, 1%Pt/ETS-2 (600 °C calcined), (f) 5%Zn,
1%Pt/ETS-2 (600 °C tested)
5.3.1.3 Microstructure and Morphology of the Catalysts
Figure 5-6 shows the SEM images of ETS-2 before calcination (a) and after
calcination at 600 °C (b). From the two images, it can be seen that the texture of
ETS-2 with and without calcination is made of 1-2 µm of irregular fine particles.
Low temperature dried samples are present as more flaky grains. After calcination
at the elevated temperature the particle of the sample became granular. The change
of morphology may count for the change of the specific surface area.
Figure 5-7 shows the images of the catalyst 5%Zn, 1%Pt/ETS-2 (a) calcined
and (b) tested at 500 °C. There are no visible aggregated Pt particles observed in
86
either image. Figure 5-8 shows the images of the catalyst with the same
compositions but treated at 600 °C. After testing, aggregated Pt particles were
present, which was confirmed by the EDX analysis.
Figure 5-6 Industrial ETS-2 (a) dried at 150°for 4 hours; (b) calcined at 600 °C
for 4 hours
Figure 5-7 Catalyst 5%Zn, 1%Pt/ETS-2 (a) calcined at 500 °C for 4 hours; (b)
tested at 500 °C
87
Figure 5-8 Catalyst 5%Zn, 1%Pt/ETS-2 (a) calcined at 600 °C for 4 hours; (b)
tested at 650 °C
Figure 5-9 shows the image of catalyst 1%Pt/ETS-2 calcined at 500 °C. In the
figure, aggregated Pt particles can be identified. Aggregation of active metal
component will cause activity loss and changes in both stability and selectivity.
Figure 5-9 Catalyst 1%Pt/ETS-2 calcined at 500 °C for 4 hours

88
Figure 5-10 shows TEM images of catalyst 1%Pt, 5%Zn/ETS-2 treated at
different conditions. From Figure 5-10 (a), it can be seen that Pt particles have
aggregated on the catalyst surface. After the catalyst was evaluated for ethane
dehydrogenation at 500 °C for four hours, Pt particles grew larger (b). When the
catalyst calcined at 600 °C, the aggregated Pt particle size were larger than at 500 °C.
Figure 5-10 TEM images of the catalyst 1%Pt, 5%Zn/ETS-2: (a) calcined at
500 °C, (b) spent (run at 500 °C), (c) calcined at 600 °C, (d) spent (run at 600 °C)
5.3.1.4 XPS
The catalyst 5%Zn, 1%Pt/ETS-2 was calcined at 500 °C and 600 °C for four
hours respectively, followed by hydrogenation reduction under the conditions: 40
v% H2, 40ml/min, and 300 °C. Figure 5-11 shows XPS spectrum of the catalyst
89
calcined at 600 °C before reduction. The deconvolutions of the spectrum for core
level Zn 2p3/2 revealed that the binding energies of Zn 2p3/2 could be deconvoluted
to 1022.5 eV and 1021.1 eV. The Zn precursor was Zn(NO3)2. When calcined at
elevated temperature, it was decomposed to ZnO. The binding energy 1022.5 eV is
associated to Zn species in ZnO, which showed good agreement with literature data
[99, 100]
. The binding energy 1021.1 eV corresponds to metallic Zn [99-101] generated
from the reduced ZnO. Large negative shift of the binding energy implied the
formation of Pt-Zn alloy [96]. By integrating the two peak areas, it can be calculated
that 81% of ZnO was reduced to metallic Zn.
XPS spectrum of the catalyst calcined at 500 °C before reduction was different
from that of the catalyst treated at 600 °C, as shown in Figure 5-12. Only one single
peak at 1021.3 eV was observed, which was ascribed to metallic Zn. Based on the
negative binding energy shift, it was predicted that probably Pt-Zn alloy was
formed.
Thermodynamic calculations indicate that cannot be reduced by
hydrogen due to the positive free Gibbs energy under the present conditions.
However, the free Gibbs energy becomes negative if hydrogen atoms take part in
the reduction. The reduction of can be explained as follows: Pt cations were
reduced first by hydrogen during the reduction process, and then, the reduced
metallic Pt acted as the active sites to dissociate hydrogen. Afterwards, hydrogen
atoms spilled over from Pt [102-106] to the adjacent zinc cations and reduced them to
their metallic state. Furthermore, when Pt and Zn formed an alloy, the enthalpy of
alloy formation was negative. The reduction of zinc cation was thermodynamically
90
favorable. The similar binding energy of metallic Zn and Zn2+ made it difficult to
assess the zinc state by XPS. The Zn LMM Auger transition analysis is more
distinguishable [100].
The differences of reducibility of ZnO in the catalysts treated under different
temperatures indicate that ZnO reacted with support at higher temperature (600 °C)
and formed a new species phase, which is harder to be reduced.

Intensity, Counts
1025 1024 1023 1022 1021 1020 1019 1018

Binding Energy, eV
Figure 5-11 XPS of the Zn 2p3/2 of catalyst 5%Zn, 1%Pt/ETS-2 calcined at
600 °C
91
Intensity, counts
1025 1024 1023 1022 1021 1020 1019 1018 1017

Binding Energy, eV
Figure 5-12 XPS of the Zn 2p3/2 of catalyst 5%Zn, 1%Pt/ETS-2 calcined at
500 °C
5.3.1.5 TPR
Figure 5-13 illustrates the H2-TPR profiles of the three catalysts: (1) no Zn
modified, only Pt loaded 1%Pt/ETS-2; (2) no Pt, only Zn modified 5%Zn/ETS-2;
and (3) Zn modified and Pt loaded 1%Pt, 5%Zn/ETS-2.
92
1%Pt/ETS-2 1%Pt,5%Zn/ETS-2 5%Zn/ETS-2
H2 Consumption, a.u.
0 100 200 300 400 500

Temperature, °C
Figure 5-13 H2-TPR profiles of ETS-2 supported catalysts
TPR curves reflected the difference in the reduction ability of the catalysts
modified with different metals. The Pt metal on the catalyst 1%Pt/ETS-2 was much
easier to be reduced. The Pt started to be reduced at room temperature. The
hydrogen uptake gradually decreased. Starting from 130 °C for 5%Zn/ETS-2, the
Zn was not reduced until 330 °C. A gradual increased curve after 330 °C revealed
that increasing amounts of Zn were reduced at relatively high temperatures. For the
Pt catalyst with Zn modification, 1%Pt, 5%Zn/ETS-2, the reduction of Zn changed
their reduction behaviors. Two reduction peaks (H2 consumption) appeared at low
temperatures (110 °C and 160 °C), and one peak at 360 °C.
93
5.3.1.6 TG-MS Analysis of Coke on Spent Catalysts
The TG-MS analysis of the spent 5%Zn, 1%Pt/ETS-2 catalyst is shown in
Figure 5-14. In the temperature range from room temperature to 600 °C, the total
weight loss was only 1.4 wt%. The MS analysis proved that all the weight loss was
associated with the water that was due to the desorption of the physically adsorbed
water and –OH decomposition. No carbon deposited on 5%Zn, 1%Pt/ETS-2 after
reactions was identified. However, a carbon dioxide emission peak around 400 °C
in the TG-MS analysis of the spent alumina supported catalyst was identified, as
shown in Figure 5-15. Carbon dioxide emission indicated there was coke deposit
on the spent catalyst. The amount of water loss of the alumina supported catalyst
was also higher than that of the ETS-2 supported catalyst.
H₂O CO₂ wt%

2.0E-13 100.0%
1.5E-13
99.5%
1.0E-13
Weight Loss
Ion Current
5.0E-14 99.0%
0.0E+00
98.5%
-5.0E-14
-1.0E-13 98.0%
0 100 200 300 400 500 600
Temperature, °C
Figure 5-14 TG-MS analysis of catalyst 5%Zn, 1%Pt/ETS-2 tested at 650 °C
94
H₂O CO₂ wt%
3.0E-13 100.0%
2.0E-13 98.0%
96.0%
1.0E-13
Weight Loss
Ion Current
94.0%
0.0E+00
92.0%
-1.0E-13
90.0%
-2.0E-13 88.0%
-3.0E-13 86.0%
0 100 200 300 400 500 600
Temperature, °C
Figure 5-15 TG-MS analysis of catalyst 1%Zn, 1%Pt/Al2O3 tested at 500 °C
5.3.2 Effect of Main Metals on the Activity and Selectivity of Ethane
Dehydrogenation
The noble metals, Pt and Pd, are the main catalytic components in ethane
dehydrogenation reactions. To study the Pt and Pd effect on catalyst performances,
the same amount of platinum and palladium were loaded on ETS-2 support
respectively as the catalysts for ethane dehydrogenation. Figure 5-16 shows the
ethane dehydrogenation product distributions on the two catalysts. From the figure,
it can be seen that the Pt/ETS-2 catalyst exhibited a higher ethylene and lower
hydrogen and methane yields at steady state. About 0.5 wt% of by-product methane
95
was produced on both catalysts. Pt was a better active component for ethylene
production compared to Pd.
CH₄ C₂H₄ H₂
7.0
6.0
5.0
Yields, %
4.0
3.0
2.0
1.0
0.0
1%Pd/ETS-2 1%Pt/ETS-2
Catalyats
Figure 5-16 Product spectrum obtained with the 1%Pd/ETS-2 and 1%Pt/ETS-
2 catalysts at 509 °C, and ethane WHSV = 0.28h-1
Figure 5-17 shows the yield changes using 1%Pt/ETS-2 as the catalyst at
different space velocities. The product distribution of 1%Pd/ETS-2 at WHSV=0.27
h-1 is also shown in the figure. The product yields on 1%Pt/ETS-2 decreased with
the space velocity increasing. Even at higher space velocity (WHSV=0.80 h-1), the
ethylene yield of 1%Pt/ETS-2 was still higher than that of 1%Pd/ETS-2 at the lower
space velocity (WHSV=0.27 h-1). The result further proved that the Pt catalyst had
higher activity and ethylene selectivity than that of the Pd catalyst.
96
5
1% Pt/ETS-2
4
Yields, %
3
H2
2 C2H4
1 CH4
0
6
5
1% Pd/ETS-2
Yields, %
4
3
2
1
0
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
-1
WHSV, h
Figure 5-17 Product yield change with space velocity at 509 °C over catalysts
1%Pd/ETS-2 and 1%Pt/ETS-2
5.3.3 Effect of the Supports on the Activity and Selectivity of Ethane
Dehydrogenation
To study the effect of the supports on the reaction performances of the ethane
dehydrogenation catalysts, three catalysts were prepared with three different
supports, ETS-2, Al2O3, and natural zeolite clinoptilolite (impregnated with 1 wt%
Pt). The product distributions of the three catalysts in ethane dehydrogenation
reaction under the same conditions are illustrated in Figure 5-18. Catalyst
1%Pt/ETS-2 gave the lowest yield of by-product methane, but the highest yield of
ethylene. Catalyst 1%Pt/Al2O3 generated the highest methane. Catalyst
1%Pt/clinoptilolite had the highest hydrogen yield and the lowest ethylene yield.
97
Obviously, ETS-2 was the best support used on the catalyst for ethane
dehydrogenation.
CH₄ C₂H₄ H₂
8.0
7.0
6.0
Yields, %
5.0
4.0
3.0
2.0
1.0
0.0
1%Pt/ETS-2 1%Pt/Al₂O₃ 1%Pt/clinoptilolite
Catalysts
Figure 5-18 Product spectrum obtained with the 1%Pt/ETS-2, 1%Pt/Al2O3,
and 1%Pt/clinoptilolite catalysts at 509 °C, and ethane WHSV = 0.28 h-1
5.3.4 Effect of Promoters on the Activity, Selectivity and Stability of Ethane
Dehydrogenation
5.3.4.1 Zn as the promoter
The 1%Pt/Al2O3 was modified with 1 wt% of Zn, and a comparison is
illustrated in Figure 5-19. It is noted that Zn improved the ethylene concentration,
and simultaneously suppressed the production of methane generated from the side
reactions. As space velocity increased, the methane yield decreased significantly.
Hydrogen yield decreased slightly, while ethylene yield increased slightly. The
98
results in Figure 5-19 suggest Zn is an effective promoter for ethane
dehydrogenation reaction.
7
H2
C2H4
Yields, %
6
CH4
5
4
1% Pt/Al2O3
7
6
5
Yields, %
4
3
1% Pt, 1% Zn/Al2O3
2
1
0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50
WHSV, h-1
Figure 5-19 Product concentration change with WHSV at 509 °C over catalysts
1%Pt/Al2O3 and 1%Pt, 1%Zn/Al2O3
The product yield changes with time on 1%Pt/ETS-2 and 0.5%Zn, 1%Pt/ETS-
2 are shown in Figure 5-20 and Figure 5-21 respectively. The conversion on
1%Pt/ETS-2 catalyst without Zn was lower than the equilibrium value. The side
reaction occurred (producing CH4), and activity was not stable.
99
H₂ CH₄ C₂H₄ Equilibrium@509°C
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
20 30 40 50 60 70 80 90 100
Time on Stream, min
Figure 5-20 Product concentration changes with time over 1%Pt/ETS-2 at
509 °C, WHSV = 0.80 h-1
The addition of 0.5wt% Zn into 1%Pt/ETS-2 did not promote the conversion
since the ethylene yield was still below equilibrium, and side reactions were still
present (CH4 still presenting) even at relatively high feed flow rate. However, the
Zn modification improved the catalyst stability.
Increasing the Zn content to 1wt%, the activity increased and stability was
improved. The conversion reached equilibrium, as shown in Figure 5-22. Since
ethane dehydrogenation is an endothermic reaction, the temperature measured in
the middle of the catalyst bed was lower than that near the reactor wall because of
the temperature gradient. So, the overall product yields were between the
equilibrium and the highest point near the wall. However, side reactions (CH4 by-
product) still occurred at relatively lower space velocity.
100
H₂ CH₄ C₂H₄ Equilibrium@509°C
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
75 85 95 105
Time on Stream, min
Figure 5-21 Product concentration changes with time over 0.5%Zn,
1%Pt/ETS-2 at 509 °C, WHSV = 1.13 h-1
When the WHSV increased from 0.88h-1 to 1.14 h-1, the by-product methane
disappeared, as shown in Figure 5-23. The ethylene yield still stayed around
equilibrium. Meanwhile, hydrogen yield was gradually reduced with reaction time
and close to ethylene yield.
101
H₂ CH₄ C₂H₄
Equilibrium@509°C Equilibrium@514°C
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
10 20 30 40 50
Time on Stream, min
Figure 5-22 Product concentration changes with time over 1%Zn, 1%Pt/ETS-
2 at 509 °C, WHSV = 0.88 h-1
H₂ CH₄ C₂H₄
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
50 60 70 80 90
Time on Stream, min
Figure 5-23 Product concentration changes with time over 1%Zn, 1%Pt/ETS-
2 at 509 °C, WHSV = 1.14 h-1

102
As shown in Figure 5-24and Figure 5-25, when the Zn content increased to 2
wt%, the hydrogen and ethylene yields were similar at low feed flow rate, even
though side reactions still existed. However, when feed flow rate increased,
methane disappeared, and the ethylene and hydrogen yield became identical.
H₂ CH₄ C₂H₄
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
15 25 35 45 55 65
Time on Stream, min
Figure 5-24 Product yield changes with time over 2%Zn, 1%Pt/ETS-2 at
509 °C, WHSV = 0.87 h-1
103
H₂ CH₄ C₂H₄
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
70 80 90 100
Time on Stream, min
509 °C, WHSV = 1.13 h-1
Further increase of Zn content to 5wt%, suppressed side reactions even at low
space velocity (0.89 h-1). The conversion reached equilibrium state, as shown in
Figure 5-26.
As shown in Figure 5-27, when Zn content increased to 7.5 wt%, the side
reactions disappeared at low space velocity, the product yields were maintained
around the equilibrium, and the activity increased further. The hydrogen yield was
slightly higher than that of the ethylene. Zn modification enhanced the catalyst
stability.
104
H₂ CH₄ C₂H₄
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
15 25 35 45 55
Time on Stream, min
509 °C, WHSV = 0.89 h-1
H₂ CH₄ C₂H₄
6.0
5.0
4.0
Yields, %
3.0
2.0
1.0
0.0
10 20 30 40
Time on Stream, min
Figure 5-27 Product yield changes with time over 7.5%Zn, 1%Pt/ETS-2 at
509 °C, WHSV = 0.86 h-1

105
The production yield changes with the Zn content were summarized in Figure
5-28. It can be seen that, to reach the reaction equilibrium, more than 1 wt% of Zn
needed to be added. Higher Zn content (5 wt% or higher) helped to suppress the
side reactions even at lower space velocity (0.87 h-1).
H₂ CH₄ C₂H₄
6.0
5.0
Yields, %
4.0
3.0
2.0
1.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Zn, wt%
Figure 5-28 Product concentration change with Zn content on catalyst
1%Pt/ETS-2 at 509 °C, WHSV = 0.87h-1
For Zn modified 1%Pt/ETS-2 catalysts, the yield changes of the main product
ethylene, hydrogen, and by-product methane with different WHSV are shown in
Figure 5-29. As expected, the product yields decreased with WHSV for all catalysts.
However, the extent was quite different. From Figure 5-29, it can be seen that the
106
ethylene yields for all Zn modified catalysts were higher than that for the catalyst
without Zn modification (1%Pt/ETS-2) in the whole range of space velocities
studied. Among the Zn modified catalysts, 1%Pt/ETS-2 with 1 wt%, 2 wt%, and 5
wt% of Zn had similar high ethylene yields at all space velocities, but 1 wt% and 2
wt% of Zn modified catalysts had relatively higher methane yield at low space
velocity. The catalyst 1%Pt, 5%Zn/ETS-2 had relatively stable ethylene yields at
low space velocities, and its methane yields were suppressed to zero at all space
velocities. The catalyst 1%Pt, 7.5%Zn/ETS-2 had the similar properties as 1%Pt,
5%Zn/ETS-2. Hydrogen yield had a similar trend to that of ethylene.
6 6
5 5
4 4
C2H4, %
H2, %
3 3
2 2
0.5 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.4 0.6 0.8 1.0 1.2 1.4 1.6
-1
0.4 WHSV, h
0.3
1%Pt/ETS-2
CH4, %
0.2 1%Pt, 0.5%Zn/ETS-2

1%Pt, 1%Zn/ETS-2
0.1
1%Pt, 2%Zn/ETS-2
0.0 1%Pt, 5%Zn/ETS-2
1%Pt, 7.5%Zn/ETS-2
0.4 0.6 0.8 1.0 1.2 1.4 1.6
WHSV, h-1
Figure 5-29 C2H4, H2, and CH4 yield change with space velocity on different
content of Zn modified catalysts at 509 °C
107
100
99
Selectivity, %
98
97
WHSV=0.87h-1
WHSV=1.13h-1
96
0 1 2 3 4 5 6 7 8
Zn, wt%
Figure 5-30 Effects of Zn content of catalysts 1%Pt/ETS-2 on selectivity of
ethylene at WHSV=0.87 h-1 and 1.13h-1, 509 °C
Figure 5-30 shows the effects of Zn content on the selectivity of ethylene at
the space velocities 0.87 and 1.13 h-1. From the figure, it can be seen that selectivity
of ethylene increased with Zn content at both space velocities. Ethylene selectivity
was able to reach 100% even at lower Zn content for high space velocity (1.13 h-1).
5.3.4.2 Ce or/and Zn as Promoters
To study the promotion effect with the other metals, 1%Pt/ETS-2 was
modified with Zn and Ce. The ethylene, hydrogen and methane yields at various
108
WHSVs are shown in Figure 5-31. From the figure, it can be seen that Zn enhanced
ethylene yield, while Ce it. However, Ce could not eliminate side reactions even at
high space velocity. The ethylene yield on Zn modified Catalyst (1%Pt,
1%Zn/ETS-2) was higher than that on only Zn and Ce modified catalysts (1%Pt,
1%Zn, 2.5%Ce/ETS-2). Zn and Zn-Ce modified catalysts could suppress the side
reactions and minimize by-product methane at higher space velocities. Zn could
also stabilize ethylene yields at higher space velocities. Hydrogen yield change was
similar to that of ethylene.
6 6
5 5
C2H4, %
4 4
H2, %
3 3
2 2
1 1
0.6 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
0.5 WHSV, h-1

0.4
CH4, %
0.3 1%Pt/ETS-2
0.2
1%Pt, 2.5%Ce/ETS-2
1%Pt, 1%Zn/ETS-2
0.1
1%Pt, 1%Zn, 2.5%Ce/ETS-2
0.0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

-1
WHSV, h
Figure 5-31 C2H4, H2, and CH4 yields change with WHSV at 509 °C over
1%Pt/ETS-2 modified with different promoters
109
To investigate the Ce modification effect for the support Al2O3, the catalysts
modified with 1%Zn and 1%Zn+2.5%Ce were prepared. The ethylene, hydrogen,
and methane yields at the various WHSVs are shown in Figure 5-32. Ethylene yield
increased with space velocity for both catalysts, but both hydrogen and by-product
methane yields decreased with space velocity. For the catalyst 1%Pt, 1%Zn/Al2O3,
Ce inhibited the catalyst activities and thus decreased the ethylene yield. However,
Ce modification did not help suppress by-product methane production for the
alumina supported Pt catalysts.
5.2 7.2
5.1 7.0
5.0 6.8
4.9 6.6
C2H4, %
H2, %
4.8 6.4
4.7
6.2
4.6
6.0
4.5
5.8
4.4
4.0 1.0 1.5 2.0 2.5 3.0 1.0 1.5 2.0 2.5 3.0
-1
3.5 WHSV, h
3.0
CH4, %
2.5
2.0 1% Pt, 1% Zn/Al2O3

1.5 1% Pt, 1% Zn, 2.5% Ce/Al2O3
1.0
0.5
1.0 1.5 2.0 2.5 3.0
-1
WHSV, h
Figure 5-32 C2H4, H2, and CH4 yields change with space velocities at 509 °C
over 1%Pt/Al2O3 catalyst modified with Zn and Ce promoters
110
5.3.4.3 Ce, Zn, and Sn Promoters
Ethane dehydrogenation testing results of the Zn and Ce modified Pd-
Sn/Al2O3 catalyst are shown in Figure 5-33. From Figure 5-33, it can be seen that
Zn and Ce enhanced the activity of Pd-Sn/Al2O3 catalyst. However, the side
reaction increased as well. For 1%Pt/ETS-2 catalyst, Zn and Ce had similar
promoting effects (Figure 5-31). However, for 1%Pt/Al2O3 catalyst, Zn acted as a
promoter, while Ce acted as an inhibitor for catalyst activities (Figure 5-32).
CH₄ C₂H₄ H₂
7.0
6.0
5.0
Yield, %
4.0
3.0
2.0
1.0
0.0
1% Pd, 1% Sn/Al₂O₃ 1% Pd, 1% Sn, 1%Ce, 1% Zn/Al₂O₃
Catalysts
Figure 5-33 Product spectrum obtained with the 1%Pd, 1%Sn/Al2O3 and
1%Pd, 1%Ce, 1%Zn, 1%Sn/Al2O3 catalysts at 509 °C, and ethane WHSV =
0.28 h-1
111
5.4 Conclusions
The Zn modification helped the Pt disperse on ETS-2 and Al2O3 supports, and
therefore resulted in a significant improvement in ethane dehydrogenation
performances. Pt was aggregated at high temperature. ZnO could react with the
support to form new species. In the Pt catalysts, ZnO can be easily reduced by
hydrogen. The new phase formation and strong interaction with the support of ZnO
could lead to the decrease the reductibility of ZnO.
For the ethane dehydrogenation reaction, Pt showed a better performance than
Pd; ETS-2 support was superior in reaction specificity to Al2O3 and clinoptilolite.
The addition of Zn in the Pt/ETS-2 catalyst improved the ethylene selectivity
and stability significantly. The side reactions like coking and methane formation
were effectively suppressed at a wider range of space velocities. With the Zn
content increase, the activity of the catalysts increased, and the methane production
was greatly suppressed. Pt-Zn alloy was formed and might act as the active species
for ethane dehydrogenation that helped to reduce side reactions and shorten the time
to reach the equilibrium. Residence time was a main factor that affected selectivity
and activity.
The addition of Ce had detrimental effects on the Pt catalysts supported on
both Al2O3 and ETS-2 supports.
112
Chapter 6. Conclusions and Suggestions for Future Work
6.1 General conclusions
Natural zeolite mordenite is abundant in nature. Monolithic mordenite rock
fragments procured from New Zealand have good machinability and thermal
stability. The crystal structure is stable up to 800 °C. The zeolite mordenite has
uniform pore size distribution and has internal channels which can selectively
separate hydrogen. All these features make the mordenite a good candidate as a
high temperature membrane material for membrane reactors.
A membrane reactor with a natural mordenite membrane disk coupled with a
Pt/Al2O3 catalyst packed bed was able to selectively extract hydrogen and shifted
equilibrium of ethane dehydrogenation reaction at 500-550 °C. By increasing the
area/volume ratio of the membrane reactor module from 0.04 m-1 to 0.16 m-1, the
ethylene yield increased from 6.5% to 15.6%. The membrane reactor showed a
higher ethylene yield enhancement at 550 °C than at 500 °C, due to the increase in
hydrogen permeation rate. Ethylene yield increased from 6.5% at 500 °C to 8.9%
at 550 °C compared to the conventional packed bed reactor with the area/volume
ratio 0.04 m-1. Lower Damköhler-Peclet number favored the higher ethylene yield
improvement.
For ethane dehydrogenation reaction, Pt showed a better performance than Pd;
ETS-2 support was superior to Al2O3 and clinoptilolite.
113
The Zn modification promoted the Pt dispersion on ETS-2 and Al2O3 supports,
and therefore significantly improved ethane dehydrogenation performance. The Pt-
Zn alloy phase generated on catalyst surface could be active species for ethane
dehydrogenation that helped to reduce side reactions. The addition of Zn in the
Pt/ETS-2 catalyst improved the ethylene selectivity and stability significantly. The
side reactions like coking and methane formation were effectively suppressed. As
the Zn content increased, the activity of the catalysts increased as well, while the
methane production was greatly suppressed. Residence time was one of the main
factors that affected selectivity and activity. Pt particle aggregation at high
temperature caused catalyst deactivation.
The addition of Ce had detrimental effects on the Pt catalysts supported on
both Al2O3 and ETS-2 supports.
6.2 Suggestions for future work
Selectivity and permeance are key factors for membrane performance and
effectiveness. Knudsen flow in membrane has limited selectivity since the
selectivity depends on √𝑀𝐵 ⁄𝑀𝐴 , because the square root of the molecular weight
ratio of the two gas molecules is a small number. However, Knudsen flow occurs
in pores 2 < 𝑑 < 100 𝑛𝑚 , which is in meso-pores (2 < 𝑑 < 50 𝑛𝑚) and small
macro-pores range ( 𝑑 > 50 𝑛𝑚 ). Knudsen flow significantly contributes to
permeance. Permeance is a crucial parameter of process economics. Knudsen
permeance is not a function of pressure. It is proportional to √1⁄𝑇. Viscous flow
also occurs in the range (𝑑 > 20 𝑛𝑚) that Knudsen flow does, but it is not selective.
114
High temperature membranes require a material with good thermal stability,
physically and chemically stable at high working temperatures. The membrane
material’s coefficient of thermal expansion (CTE) should be close to the sealing
parts, so a tight sealing can be maintained.
Therefore, a good high temperature membrane should have selectivity better
than Knudsen diffusion, high permeance, and good thermal stability at working
temperatures.
Pt-Zn/ETS-2 catalyst has the ability to suppress side reactions including
coking and methane formation. However, its activity could be improved (higher
space velocity). Compared to alumina support, its specific surface area is very low,
which affects the dispersion of active metal species like Pt, reduces the active sites
on the surface, and thus limits the activity. Therefore, further modification of the
catalysts is a topic that should be pursued in the future.
ETS-2 is not stable at high temperature. It can easily change phase above
600 °C, and it is also reactive with supported metal components at high temperature.
These will cause the catalyst to lose activity and will reduce the run cycles. This is
another area of potential future improvement.
Active metal components aggregating at high temperature is a major issue for
catalyst deactivation. Aggregation of Pt particles still occurred on ETS-2 support
even with Zn modification. Future research is needed in order to eliminate or
mitigate the problem.
A more systematic and thorough investigation and characterization of the
catalyst and the support, such as optimal metal compositions and chemistry of
115
catalyst preparation, will help understand reaction mechanisms and control catalyst
quality.
116
References
[1] W. R. True: "Global Ethylene Production Continues Advance in 2009", Oil
& Gas Journal, 2010, 108 (27), 34-38.
[2] W. R. True: "Global Ethylene Capacity Poised for Major Expansion", Oil &
Gas Journal, 2013, 111 (7), 90-95.
[3] W. R. True: "Global Ethylene Capacity Continues Advance in 2011", Oil &
Gas Journal, 2012, 110 (7), 78-84.
[4] L. Bewley: "Ihs Wpc 2012: Shale Reshapes Ethylene Markets", Chemical
Week, 2012, 174 (10), 15.
[5] L. Bewley and R. Westervelt: "Mideast Producers Reap Benefits of Low-Cost
Production. (Cover Story)", Chemical Week, 2011, 173 (29), 32-33.
[6] C. Carr: "Global Ethylene Outlook: One Product....... Many Strategies", 2013,
http://ihsglobalevents.com/wpc2013/files/2013/03/Carr_20131.pdf.
[7] S. Jenkins: "Shale Gas Ushers in Ethylene Feed Shifts", Chemical
Engineering, 2012, October, 17-19.
[8] K. Bullis: "Shale Gas Will Fuel a U.S. Manufacturing Boom", MIT
Technology Review, 2013, 116 (2), 73-74.
[9] J. Chang: "Nova Explores Ethylene Expansions", ICIS Chemical Business,
2013, 283 (12), 12.
[10] The Dow Chemical Company: "Dow to Build New Ethylene Production Plant
at Dow Texas Operations", 2012.
117
[11] Nexant: "North American Shale Gas: Opportunity or Threat for Global
Ethylene Producers?", 2012,
http://www.chemsystems.com/reports/search/docs/prospectus/STMC12_Sha
le%20Gas_pros.pdf.
[12] K. Holmquist: "Global Ethylene Surplus to Last through 2011", Oil & Gas
Journal, 2010, 108 (27), 46-52.
[13] H. Zimmermann and R. Walzl: 'Ethylene', in "Ullmann's Encyclopedia of
Industrial Chemistry", 6th edn, 465-529; 2009, Wiley-VCH Verlag GmbH &
Co. KGaA.
[14] J. Towfighi, M. Sadrameli, and A. Niaei: "Coke Formation Mechanisms and
Coke Inhibiting Methods in Pyrolysis Furnaces", Journal of Chemical
Engineering of Japan, 2002, 35 (10), 923-937.
[15] T. Ren, M. Patel, and K. Blok: "Olefins from Conventional and Heavy
Feedstocks: Energy Use in Steam Cracking and Alternative Processes",
Energy, 2006, 31 (4), 425-451.
[16] K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J.
Rouquerol, and T. Siemieniewska: "Reporting Physisorption Data for
Gas/Solid Systems with Special Reference to the Determination of Surface
Area and Porosity", Pure and Applied Chemistry, 1985, 57 (4), 603-619.
[17] M. Huttepain and A. Oberlin: "Microtexture of Nongraphitizing Carbons and
Tem Studies of Some Activated Samples", Carbon, 1990, 28 (1), 103-111.
118
[18] H. P. Hsieh: 'Chapter 4 Physical, Chemical and Surface Properties of
Inorganic Membranes', in "Membrane Science and Technology Series, 3:
Inorganic Membranes for Separation and Reaction", 93-147; 1996, Elsevier.
[19] A. J. Burggraaf: 'Chapter 9 Transport and Separation Properties of
Membranes with Gases and Vapours', in "Membrane Science and
Technology Series, 4: Fundamentals of Inorganic Membrane Science and
Technology", (eds. A. J. Burggraaf and L. Cot), 331-433; 1996, Elsevier.
[20] Y. Cen and R. N. Lichtenthaler: 'Chapter 3 Vapor Permeation', in "Membrane
Science and Technology Series, 2: Membrane Separations Technology--
Principles and Applications", (eds. D. N. Richard and S. A. Stern), 85-112;
1995, Elsevier.
[21] B. Fayyaz-Najafi: "Nanoporous Silicon Carbide Membranes: Preparation and
Reactive Applications", Ph.D. thesis, University of Southern California, Ann
Arbor, 2005.
[22] W. J. Koros, Y. H. Ma, and T. Shimidzu: "Terminology for Membranes and
Membrane Processes (Iupac Recommendation 1996)", Journal of Membrane
Science, 1996, 120 (2), 149-159.
[23] M. Daramola, E. Aransiola, and T. Ojumu: "Potential Applications of Zeolite
Membranes in Reaction Coupling Separation Processes", Materials, 2012, 5
(11), 2101-2136.
[24] S. Tennison: "Current Hurdles in the Commercial Development of Inorganic
Membrane Reactors", Membrane Technology, 2000, 2000 (128), 4-9.
119
[25] C. Téllez and M. Menéndez: '8 Zeolite Membrane Reactors', in "Membranes
for Membrane Reactors: Preparation, Optimization and Selection", (eds. A.
Basile and F. Gallucci), 243-273; 2011, John Wiley & Sons, Ltd.
[26] M. E. E. Abashar and A. A. Al-Rabiah: "Production of Ethylene and
Cyclohexane in a Catalytic Membrane Reactor", Chemical Engineering and
Processing, 2005, 44 (11), 1188-1196.
[27] J.-A. Dalmon, A. Cruz-López, D. Farrusseng, N. Guilhaume, E. Iojoiu, J.-C.
Jalibert, S. Miachon, C. Mirodatos, A. Pantazidis, M. Rebeilleau-
Dassonneville, Y. Schuurman, and A. C. van Veen: "Oxidation in Catalytic
Membrane Reactors", Applied Catalysis A: General, 2007, 325 (2), 198-204.
[28] L. Chalakov, L. K. Rihko-Struckmann, B. Munder, and K. Sundmacher:
"Oxidative Dehydrogenation of Ethane in an Electrochemical Packed-Bed
Membrane Reactor: Model and Experimental Validation", Chemical
Engineering Journal, 2009, 145 (3), 385-392.
[29] M. P. Lobera, M. Balaguer, J. Garcia-Fayos, and J. M. Serra: "Rare Earth-
Doped Ceria Catalysts for Odhe Reaction in a Catalytic Modified Miec
Membrane Reactor", Chemcatchem, 2012, 4 (12), 2102-2111.
[30] M. L. Rodriguez, D. E. Ardissone, E. Heracleous, A. A. Lemonidou, E. López,
M. N. Pedernera, and D. O. Borio: "Oxidative Dehydrogenation of Ethane to
Ethylene in a Membrane Reactor: A Theoretical Study", Catalysis Today,
2010, 157 (1–4), 303-309.
[31] H. Q. Jiang, Z. W. Cao, S. Schirrmeister, T. Schiestel, and J. Caro: "A
Coupling Strategy to Produce Hydrogen and Ethylene in a Membrane
120
Reactor", Angewandte Chemie-International Edition, 2010, 49 (33), 5656-
5660.
[32] H. H. Wang, C. Tablet, T. Schiestel, and J. Caro: "Hollow Fiber Membrane
Reactors for the Oxidative Activation of Ethane", Catalysis Today, 2006, 118
(1-2), 98-103.
[33] D. Ahchieva, M. Peglow, S. Heinrich, L. Mörl, T. Wolff, and F. Klose:
"Oxidative Dehydrogenation of Ethane in a Fluidized Bed Membrane
Reactor", Applied Catalysis A: General, 2005, 296 (2), 176-185.
[34] S. Crapanzano, I. V. Babich, and L. Lefferts: "Selection of Mixed Conducting
Oxides for Oxidative Dehydrogenation of Propane with Pulse Experiments",
Applied Catalysis a-General, 2011, 391 (1-2), 70-77.
[35] O. Czuprat, J. Caro, V. A. Kondratenko, and E. V. Kondratenko:
"Dehydrogenation of Propane with Selective Hydrogen Combustion: A
Mechanistic Study by Transient Analysis of Products", Catalysis
Communications, 2010, 11 (15), 1211-1214.
[36] D. S. Coombs, A. Alberti, T. Armbruster, G. Artioli, C. Colella, E. Galli, J.
D. Grice, F. Liebau, J. A. Mandarino, H. Minato, E. H. Nickel, E. Passaglia,
D. R. Peacor, S. Quartieri, R. Rinaldi, M. Ross, R. E. Sheppard, E. Tillmanns,
and G. Vezzalini: "Recommended Nomenclature for Zeolite Minerals: Report
of the Subcommittee on Zeolites of the International Mineralogical
Association, Commission on New Minerals and Mineral Names",
Mineralogical Magazine, 1998, 62 (4), 533-571.
121
[37] M. W. Ackley, S. U. Rege, and H. Saxena: "Application of Natural Zeolites
in the Purification and Separation of Gases", Microporous and Mesoporous
Materials, 2003, 61 (1–3), 25-42.
[38] D. C. L. How: "Xi.-on Mordenite, a New Mineral from the Trap of Nova
Scotia", Journal of the Chemical Society, 1864, 17 (0), 100-104.
[39] "Mordenite", http://www.mindat.org/min-2779.html.
[40] S. A. Hoseinzadeh Hejazi: "Characterization of Natural Zeolite Membranes
for Molecular Hydrogen/Carbon Dioxide Separations by Single Gas
Permeation", M.S. thesis, University of Alberta (Canada), Ann Arbor, 2011.
[41] P. Simoncic and T. Armbruster: "Peculiarity and Defect Structure of the
Natural and Synthetic Zeolite Mordenite: A Single-Crystal X-Ray Study",
American Mineralogist, 2004, 89 (2-3), 421-431.
[42] M. Földvári: "Handbook of Thermogravimetric System of Minerals and Its
Use in Geological Practice", 101-102; 2011, BUDAPEST, Geological
Institute of Hungary (=Magyar Állami Földtani Intézet).
[43] G. Cruciani: "Zeolites Upon Heating: Factors Governing Their Thermal
Stability and Structural Changes", Journal of Physics and Chemistry of Solids,
2006, 67 (9–10), 1973-1994.
[44] A. W. Sleight: "Negative Thermal Expansion Materials", Current Opinion in
Solid State and Materials Science, 1998, 3 (2), 128-131.
[45] K. D. Hammonds, V. Heine, and M. T. Dove: "Rigid-Unit Modes and the
Quantitative Determination of the Flexibility Possessed by Zeolite
122
Frameworks", The Journal of Physical Chemistry B, 1998, 102 (10), 1759-
1767.
[46] V. Heine, P. R. L. Welche, and M. T. Dove: "Geometrical Origin and Theory
of Negative Thermal Expansion in Framework Structures", Journal of the
American Ceramic Society, 1999, 82 (7), 1793-1802.
[47] A. MARTUCCI, M. SACERDOTI, G. CRUCIANI, and C. DALCONI: "In
Situ Time Resolved Synchrotron Powder Diffraction Study of Mordenite",
European Journal of Mineralogy, 2003, 15 (3), 485-493.
[48] E. Chmielewská: 'Chapter 18 an Update of Zeolitic and Other Traditional
Adsorption and Ion Exchange Materials in Water Cleanup Processes', in
"Handbook of Natural Zeolites", (eds. V. J. Inglezakis and A. A. Zorpas),
436-452; 2012, Bentham eBooks, Bentham Science Publishers.
[49] G. Horvath and K. Kawazoe: "Method for the Calculation of Effective Pore
Size Distribution in Molecular Sieve Carbon", Journal of Chemical
Engineering of Japan, 1983, 16 (6), 470-475.
[50] R. L. Hay and R. A. Sheppard: "Occurrence of Zeolites in Sedimentary Rocks:
An Overview", Reviews in Mineralogy and Geochemistry, 2001, 45 (1), 217-
234.
[51] M. Shekarchi, B. Ahmadi, and M. Najimi: 'Chapter 27 Use of Natural Zeolite
as Pozzolanic Material in Cement and Concrete Composites', in "Handbook
of Natural Zeolites", (eds. V. J. Inglezakis and A. A. Zorpas), 665-694; 2012,
Bentham eBooks, Bentham Science Publishers.
123
[52] D. L. Bish and J. W. Carey: "Thermal Behavior of Natural Zeolites", Reviews
in Mineralogy and Geochemistry, 2001, 45 (1), 403-452.
[53] F. Gallucci, E. Fernandez, P. Corengia, and M. van Sint Annaland: "Recent
Advances on Membranes and Membrane Reactors for Hydrogen Production",
Chemical Engineering Science, 2013, 92 (0), 40-66.
[54] J. Gascon, F. Kapteijn, B. Zornoza, V. Sebastián, C. Casado, and J. Coronas:
"Practical Approach to Zeolitic Membranes and Coatings: State of the Art,
Opportunities, Barriers, and Future Perspectives", Chemistry of Materials,
2012, 24 (15), 2829-2844.
[55] M. O. Daramola, E. F. Aransiola, and T. V. Ojumu: "Potential Applications
of Zeolite Membranes in Reaction Coupling Separation Processes", Materials
(1996-1944), 2012, 5 (11), 2101-2136.
[56] S. M. Hashim, A. R. Mohamed, and S. Bhatia: "Catalytic Inorganic
Membrane Reactors: Present Research and Future Prospects", Reviews in
Chemical Engineering, 2011, 27 (3/4), 157-178.
[57] A. M. Champagnie, T. T. Tsotsis, R. G. Minet, and A. I. Webster: "A High
Temperature Catalytic Membrane Reactor for Ethane Dehydrogenation",
Chemical Engineering Science, 1990, 45 (8), 2423-2429.
[58] E. Gobina and R. Hughes: "Ethane Dehydrogenation Using a High-
Temperature Catalytic Membrane Reactor", Journal of Membrane Science,
1994, 90 (1–2), 11-19.
124
[59] E. Gobina, K. Hou, and R. Hughes: "Ethane Dehydrogenation in a Catalytic
Membrane Reactor Coupled with a Reactive Sweep Gas", Chemical
Engineering Science, 1995, 50 (14), 2311-2319.
[60] E. Gobina, K. Hou, and R. Hughes: "Equilibrium-Shift in Alkane
Dehydrogenation Using a High-Temperature Catalytic Membrane Reactor",
Catalysis Today, 1995, 25 (3-4), 365-370.
[61] E. Gobina, K. H. Hou, and R. Hughes: "A Comparative Evaluation of High-
Temperature Membrane Systems for Catalytic Processing", Chemical
Engineering Communications, 1998, 166, 157-181.
[62] J. Szegner, K. L. Yeung, and A. Varma: "Effect of Catalyst Distribution in a
Membrane Reactor: Experiments and Model", Aiche Journal, 1997, 43 (8),
2059-2072.
[63] W. An, P. Swenson, L. Wu, T. Waller, A. Ku, and S. M. Kuznicki: "Selective
Separation of Hydrogen from C1/C2 Hydrocarbons and Co2 through Dense
Natural Zeolite Membranes", Journal of Membrane Science, 2011, 369 (1–2),
414-419.
[64] S. T. Oyama and H. Lim: "An Operability Level Coefficient (Olc) as a Useful
Tool for Correlating the Performance of Membrane Reactors", Chemical
Engineering Journal, 2009, 151 (1–3), 351-358.
[65] S. Battersby, P. W. Teixeira, J. Beltramini, M. C. Duke, V. Rudolph, and J.
C. Diniz da Costa: "An Analysis of the Peclet and Damkohler Numbers for
Dehydrogenation Reactions Using Molecular Sieve Silica (Mss) Membrane
Reactors", Catalysis Today, 2006, 116 (1), 12-17.
125
[66] R. S. Vincent, R. P. Lindstedt, N. A. Malik, I. A. B. Reid, and B. E. Messenger:
"The Chemistry of Ethane Dehydrogenation over a Supported Platinum
Catalyst", Journal of Catalysis, 2008, 260 (1), 37-64.
[67] U. Fekl and K. I. Goldberg: 'Homogeneous Hydrocarbon C-H Bond
Activation and Functionalization with Platinum', in "Advances in Inorganic
Chemistry", 259-320; 2003, Academic Press.
[68] F. E. Frey and W. F. Huppke: "Equilibrium Dehydrogenation of Ethane,
Propane, and the Butanes", Industrial and Engineering Chemistry, 1933, 25,
54-59.
[69] D. J. Fabian and A. J. B. Robertson: "The Dehydrogenation of Ethane at Low
Pressure on Incandescent Platinum", Proceedings of the Royal Society of
London Series a-Mathematical and Physical Sciences, 1956, 237 (1208), 1-
16.
[70] J. A. Cusumano, G. W. Dembinski, and J. H. Sinfelt: "Chemisorption and
Catalytic Properties of Supported Platinum", Journal of Catalysis, 1966, 5 (3),
471-475.
[71] B. S. Gudkov, L. Guczi, and P. Tétényi: "Kinetics and Mechanism of Ethane
Hydrogenolysis on Silica-Supported Platinum and Platinum-Iron Catalysts",
Journal of Catalysis, 1982, 74 (2), 207-215.
[72] G. C. Bond, F. Garin, and G. Maire: "Characterization of the Standard
Platinum/Silica Catalyst Europt-1: Vi. Catalytic Properties", Applied
Catalysis, 1988, 41 (0), 313-335.
126
[73] C. De la Cruz and N. Sheppard: "Adsorption and Dehydrogenation of Ethane
on a Silica-Supported Platinum Catalyst Monitored in the Temperature Range
90-294 K by Ftir Spectroscopy", Physical Chemistry Chemical Physics, 1999,
1 (2), 329-332.
[74] H. E. Swift, F. E. Lutinski, and H. H. Tobin: "Catalytic Aluminas: Iii. Physical
and Catalytic Properties of Nickel and Platinum Impregnated Eta and Gamma
Alumina", Journal of Catalysis, 1966, 5 (2), 285-292.
[75] P. Mériaudeau, O. H. Ellestad, M. Dufaux, and C. Naccache: "Metal-Support
Interaction. Catalytic Properties of Ti02-Supported Platinum, Iridium, and
Rhodium", Journal of Catalysis, 1982, 75 (2), 243-250.
[76] R. Szymanski and H. Charcosset: "Preparation, Characterization and
Catalytic Properties of Carbon-Supported Platinum-Zirconium Alloy
Catalysts", Journal of Molecular Catalysis, 1984, 25 (1–3), 337-345.
[77] J. Salmones, J.-A. Wang, J. A. Galicia, and G. Aguilar-Rios: "H2 Reduction
Behaviors and Catalytic Performance of Bimetallic Tin-Modified Platinum
Catalysts for Propane Dehydrogenation", Journal of Molecular Catalysis A:
Chemical, 2002, 184 (1–2), 203-213.
[78] A. Virnovskaia, S. Jørgensen, J. Hafizovic, Ø. Prytz, E. Kleimenov, M.
Hävecker, H. Bluhm, A. Knop-Gericke, R. Schlögl, and U. Olsbye: "In Situ
Xps Investigation of Pt(Sn)/Mg(Al)O Catalysts During Ethane
Dehydrogenation Experiments", Surface Science, 2007, 601 (1), 30-43.
127
[79] A. Virnovskaia, E. Rytter, and U. Olsbye: "Kinetic and Isotopic Study of
Ethane Dehydrogenation over a Semicommercial Pt,Sn/Mg(Al)O Catalyst",
Industrial & Engineering Chemistry Research, 2008, 47 (19), 7167-7177.
[80] V. Galvita, G. Siddiqi, P. Sun, and A. T. Bell: "Ethane Dehydrogenation on
Pt/Mg(Al)O and Ptsn/Mg(Al)O Catalysts", Journal of Catalysis, 2010, 271
(2), 209-219.
[81] P. Sun, G. Siddiqi, W. C. Vining, M. Chi, and A. T. Bell: "Novel
Pt/Mg(in)(Al)O Catalysts for Ethane and Propane Dehydrogenation", Journal
of Catalysis, 2011, 282 (1), 165-174.
[82] S. F. Hakonsen, B. Silberova, and A. Holmen: "Oxidative Dehydrogenation
of Ethane on Pt-Sn Impregnated Monoliths", Topics in Catalysis, 2007, 45
(1-4), 61-67.
[83] S. F. Håkonsen, J. C. Walmsley, and A. Holmen: "Ethene Production by
Oxidative Dehydrogenation of Ethane at Short Contact Times over Pt-Sn
Coated Monoliths", Applied Catalysis A: General, 2010, 378 (1), 1-10.
[84] R. Gudgila and C. A. Leclerc: "Support Effects on the Oxidative
Dehydrogenation of Ethane to Ethylene over Platinum Catalysts", Industrial
& Engineering Chemistry Research, 2011, 50 (14), 8438-8443.
[85] L. M. Ombaka, P. Ndungu, and V. O. Nyamori: "Usage of Carbon Nanotubes
as Platinum and Nickel Catalyst Support in Dehydrogenation Reactions",
Catalysis Today, 2013, 217 (0), 65-75.
[86] A. T. Bell: "Thermal Dehydrogenation of Ethane on Pt and Pt-Sn Catalysts:
Effects of Particle Size and Sn/Pt Ratio on Catalyst Activity, Selectivity, and
128
Stability", Abstracts of Papers of the American Chemical Society, 2012, 244,
1.
[87] J. Wu, Z. Peng, P. Sun, and A. T. Bell: "n-Butane Dehydrogenation over
Pt/Mg(in)(Al)O", Applied Catalysis A: General, 2014, 470 (0), 208-214.
[88] E. A. de Graaf, G. Rothenberg, P. J. Kooyman, A. Andreini, and A. Bliek:
"Pt0.02sn0.003mg0.06 on Gamma-Alumina: A Stable Catalyst for Oxidative
Dehydrogenation of Ethane", Applied Catalysis a-General, 2005, 278 (2),
187-194.
[89] J. Jia, J. Shen, L. Lin, Z. Xu, T. Zhang, and D. Liang: "A Study on Reduction
Behaviors of the Supported Platinum–Iron Catalysts", Journal of Molecular
Catalysis A: Chemical, 1999, 138 (2–3), 177-184.
[90] R. Lødeng, O. A. Lindvag, S. Kvisle, H. Reier-Nielsen, and A. Holmen:
'Oxidative Dehydrogenation of Ethane over Pt and Pt/Rh Gauze Catalysts at
Very Short Contact Times', in "Studies in Surface Science and Catalysis", (eds.
D. S. F. F. A. V. A. Parmaliana and A. F), 641-646; 1998, Elsevier.
[91] P. Biloen, F. M. Dautzenberg, and W. M. H. Sachtler: "Catalytic
Dehydrogenation of Propane to Propene over Platinum and Platinum-Gold
Alloys", Journal of Catalysis, 1977, 50 (1), 77-86.
[92] A. Loaiza, M. Xu, and F. Zaera: "On the Mechanism of the H–D
Exchangereaction in Ethane over Platinum Catalysts", Journal of Catalysis,
1996, 159 (1), 127-139.
[93] C. Adlhart and E. Uggerud: "Mechanisms for the Dehydrogenation of
Alkanes on Platinum: Insights Gained from the Reactivity of Gaseous Cluster
129
Cations, Pt-N(+) N=1-21", Chemistry-a European Journal, 2007, 13 (24),
6883-6890.
[94] T. Y. Ju, H. Q. Yang, F. M. Li, X. Y. Li, and C. W. Hu: "Reaction Mechanism
on the Activation of Ethane C-H and C-C Bonds by a Diplatinum Cluster",
Theoretical Chemistry Accounts, 2013, 132 (9), 14.
[95] E. Yagasaki and R. I. Masel: "Ethylene Adsorption and Decomposition on
(1×1) Pt(110): Chemistry of the Coke Formation Site on Platinum?", Surface
Science, 1989, 222 (2–3), 430-450.
[96] J. A. Rodriguez and M. Kuhn: "Chemical and Electronic Properties of Pt in
Bimetallic Surfaces: Photoemission and Co‐Chemisorption Studies for
Zn/Pt(111)", The Journal of Chemical Physics, 1995, 102 (10), 4279-4289.
[97] S. J. Tauster, S. C. Fung, R. T. K. Baker, and J. A. Horsley: "Strong
Interactions in Supported-Metal Catalysts", Science, 1981, 211 (4487), 1121-
1125.
[98] R. M. Mironenko, O. B. Belskaya, V. P. Talsi, T. I. Gulyaeva, M. O. Kazakov,
A. I. Nizovskii, A. V. Kalinkin, V. I. Bukhtiyarov, A. V. Lavrenov, and V. A.
Likholobov: "Effect of Γ-Al2o3 Hydrothermal Treatment on the Formation
and Properties of Platinum Sites in Pt/Γ-Al2o3 Catalysts", Applied Catalysis
A: General, 2014, 469 (0), 472-482.
[99] S. W. Gaarenstroom and N. Winograd: "Initial and Final State Effects in the
Esca Spectra of Cadmium and Silver Oxides", The Journal of Chemical
Physics, 1977, 67 (8), 3500-3506.
130
[100] J. Silvestre-Albero, A. Sepúlveda-Escribano, F. Rodrı́guez-Reinoso, and J. A.
Anderson: "Influence of Zn on the Characteristics and Catalytic Behavior of
Tio2-Supported Pt Catalysts", Journal of Catalysis, 2004, 223 (1), 179-190.
[101] L. Ley, S. P. Kowalczyk, F. R. McFeely, R. A. Pollak, and D. A. Shirley: "X-
Ray Photoemission from Zinc: Evidence for Extra-Atomic Relaxation Via
Semilocalized Excitons", Physical Review B, 1973, 8 (6), 2392-2402.
[102] R. Prins: "Hydrogen Spillover. Facts and Fiction", Chemical Reviews, 2012,
112 (5), 2714-2738.
[103] V. V. Rozanov and O. V. Krylov: "Hydrogen Spillover in Heterogeneous
Catalysis", Russian Chemical Reviews, 1997, 66 (2), 107-119.
[104] U. Roland, R. Salzer, T. Braunschweig, F. Roessner, and H. Winkler:
"Investigations on Hydrogen Spillover. Part 1.-Electrical Conductivity
Studies on Titanium Dioxide", Journal of the Chemical Society, Faraday
Transactions, 1995, 91 (7), 1091-1095.
[105] W. C. Conner and J. L. Falconer: "Spillover in Heterogeneous Catalysis",
Chemical Reviews, 1995, 95 (3), 759-788.
[106] P. A. Sermon and G. C. Bond: "Hydrogen Spillover", Catalysis Reviews:
Science and Engineering, 1974, 8 (1), 211-239.
131

Yu Zhengnan 201406 MSC PDF

Uploaded by

Copyright:

Available Formats

Yu Zhengnan 201406 MSC PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Yu Zhengnan 201406 MSC PDF

Uploaded by

Copyright:

Available Formats

Ethane Dehydrogenation Using a Catalytic Membrane Reactor

A thesis submitted in partial fulfillment of the requirements for the degree of

Department of Chemical and Materials Engineering

© Zhengnan Yu, 2014

The steam cracking of hydrocarbons is the dominant technology for ethylene

routes to reduce energy consumption and other costs. Catalytic dehydrogenation of

ethane using a membrane reactor is an attractive solution because the cracking

equilibrium can be shifted in favor of ethylene by selectively removing hydrogen.

natural mordenite as membrane and Pt/Al2O3 as catalyst, a reactor with membrane

comparing to the conventional packed-bed reactor at 500°C. A novel Pt-Zn/ETS-2

selectivity could reach 100% while conversion was at equilibrium.

I would like to thank my supervisor Professor Steven M. Kuznicki for his

not have been possible.

conducting the membrane reactor testing. I want to extend my thanks Weizhu An

efforts to improve my technical writing skills.

both catalyst characterization and technical writing. I appreciate their valuable

assistance through the final stages of my thesis.

List of Tables ....................................................................................................... vii

List of Figures ..................................................................................................... viii

1.1 Overview of Global Ethylene Production ......................................................... 1

1.2 Feedstock of Ethylene Production .................................................................... 6

1.3 Technologies for Ethylene Production.............................................................. 9

1.4 Challenges for Tube Furnace Pyrolysis Process ............................................. 12

1.4.1 Coking ................................................................................................... 12

1.4.2 Energy Consumption of Pyrolysis Furnace .......................................... 14

1.5 Catalytic Dehydrogenation of Ethane to Ethylene by Membrane Reactor ..... 15

1.5.1 Gas Separation Mechanisms of Porous Inorganic Membranes ............ 15

1.5.2 Zeolite Membrane Reactor Configurations .......................................... 19

1.6 Research Objectives ........................................................................................ 23

Chapter 2. Characterizations and Product Analysis ....................................... 24

2.1 Characterization methods................................................................................ 24

2.1.1 X-ray diffraction (XRD) ....................................................................... 24

2.1.3 X-ray photoelectron spectroscopy (XPS) ............................................. 25

2.1.4 TPR ....................................................................................................... 26

2.1.5 SEM and EDX ...................................................................................... 26

2.1.6 TEM ...................................................................................................... 26

2.1.7 TGA and TG-MS .................................................................................. 27

2.1.8 Dilatometry test..................................................................................... 28

2.2 Gas Product Analysis ...................................................................................... 28

Part I: The Use of Natural Zeolite Mordenite in a Membrane Reactor for

Ethane Dehydrogenation ........................................................................ 29

Chapter 3. Characterization of Natural Mordenite Zeolite ............................ 30

3.1 Introduction ..................................................................................................... 30

3.2 Channel Sizes of Zeolite Mordenite................................................................ 31

3.3 XRD ................................................................................................................ 34

3.4 Scanning Electron Microscopy (SEM) ........................................................... 35

3.5 Thermal Analysis ............................................................................................ 36

3.5.1 Thermogravimetric Analysis (TGA)..................................................... 36

3.5.2 Dilatometry Test ................................................................................... 38

3.6 BET of Mordenite ........................................................................................... 41

3.7 Chemical compositions of natural mordenite zeolite ...................................... 46

Chapter 4. Packed Bed Membrane Reactor for Ethane Dehydrogenation ... 49

4.1 Introduction ..................................................................................................... 49

4.2 Experimental ................................................................................................... 53

4.2.1 The Catalyst .......................................................................................... 53

4.2.2 Membrane Preparation .......................................................................... 53

4.2.3 Membrane Reactor Configuration ........................................................ 54

4.2.4 Experiment Conditions ......................................................................... 55

4.2.5 Switching Membrane Reactor Mode to Conventional Reactor Mode .. 55