BORANG PENGESAHAN STATUS TESIS

Transcription

1 UNIVERSITI TEKNOLOGI MALAYSIA PSZ 19:16 BORANG PENGESAHAN STATUS TESIS JUDUL : DUAL BED CATALYTIC REACTOR SYSTEM FOR DIRECT CONVERSION OF METHANE TO LIQUID HYDROCARBONS SESI PENGAJIAN : 2004/2005 Saya SRIRAJ A/L AMMASI (HURUF BESAR) mengaku membenarkan tesis (PSM/Sarjana/Doktor Falsafah)* disimpan di Perpustakaan Universiti Teknologi Malaysia dengan syarat-syarat kegunaan seperti berikut : 1. Tesis adalah hakmilik Universiti Teknologi Malaysia. 2. Perpustakaan Universiti Teknologi Malaysia dibenarkan membuat salinan untuk tujuan pengajian sahaja. 3. Perpustakaan dibenarkan membuat salinan tesis ini sebagai bahan pertukaran di antara institusi pengajian tinggi. 4. ** Sila tandakan ( / ) SULIT (Mengandungi maklumat yang berdarjah keselamatan atau kepentingan Malaysia seperti yang termaktub di dalam AKTA RAHSIA RASMI 1972) TERHAD (Mengandungi maklumat TERHAD yang telah ditentukan oleh organisasi/badan di mana penyelidikan dijalankan) / TIDAK TERHAD Disahkan oleh (TANDATANGAN PENULIS) Alamat Tetap : NO 32, TAMAN DESA MELINTANG BARU, HUTAN MELINTANG, PERAK (TANDATANGAN PENYELIA) PROF DR. NOR AISHAH SAIDINA AMIN Nama Penyelia Tarikh : 19 DECEMBER 2005 Tarikh : 20 DECEMBER 2005 CATATAN : * Potong yang tidak berkenaan ** Jika tesis ini SULIT atau TERHAD, sila lampirkan surat daripada pihak berkuasa/organisasi berkenaan dengan menyatakan sekali sebab dan tempoh tesis ini perlu dikelaskan sebagai SULIT atau TERHAD. Tesis dimaksudkan sebagai tesis bagi ljazah Doktor Falsafah dan Sarjana secara penyelidikan, atau disertai bagi pengajian secara kerja kursus dan penyelidikan atau Laporan Projek Sarjana Muda (PSM).

2 I hereby declare that I have read this thesis and in my opinion this thesis is sufficient in terms of scope and quality for the award of the degree of Master of Engineering (Chemical). Signature : ASSOC.PROF.DR.NOR AISHAH SAIDINA A Supervisor : Prof. Dr. Nor Aishah Saidina Amin NA NA Date : ASSOC.PROF.DR.NOR 20 December 2005 AISHAH SAIDINA A

3 DUAL BED CATALYTIC REACTOR SYSTEM FOR DIRECT CONVERSION OF METHANE TO LIQUID HYDROCARBONS SRIRAJ A/L AMMASI A thesis submitted in fulfilment of the requirements for the award of the degree of Master of Engineering (Chemical) Faculty of Chemical and Natural Resources Engineering Universiti Teknologi Malaysia DECEMBER 2005

4 ii I declare that this thesis entitled Dual Bed Catalytic Reactor System for Direct Conversion of Methane to Liquid Hydrocarbons is the result of my own research except as cited in the references. The thesis has not been accepted for any degree and is not concurrently submitted in candidature of any other degree. Signature : Name : SRIRAJ A/L AMMASI Date : 19 DECEMBER 2005

5 iii To my parents and all of my family for their love. All the teachers I have had and to whom I owe so much

6 iv ACKNOWLEDGEMENTS I owe this thesis to a number of people who has contributed directly and indirectly to the successful completion of this thesis. Firstly, I would like to thank my supervisor, Prof. Dr. Nor Aishah Saidina Amin for giving me the opportunity to work on this interesting project. Her many words of wisdom, her scientific insights and sharp criticism, her relentless scrutiny of experimental results and her high ethical standards are especially valuable for the implementation of this work. I am also grateful for her patience in correcting my papers/thesis and for keeping the pressure on me to finish the thesis within the time frame. Very special thanks are given to Mr. Didik for his helpful comments and advice on zeolite phsiochemical properties and measurement; Shamsina for skillful TPD-NH 3 and NA techniques; Kalai and KKJ for pointing out that you can always do more; Soon for his ready-for-action help with lots of experimental issues; Bidin and Lafti for kind assistance with the inevitable laboratory work, Ira and Putri for their administrative help, answering my questions from every day life and cheer-up smiles; Kusmiyati for useful discussions and information; Faridahamin for setting a nice office atmosphere; Tirena for her funny complaints, as well as nice conversations; Istadi for excellent computer techniques; Tung for a very interesting share of the same house; Harjit, Siti and Zam for their companionship; other members from Chemical Reaction Engineering Group (CREG), Ruzinah, Tutuk, Zaki and Dr Didi for all sorts of help inside the lab. Working in such a big group with each member specializing in one or two areas, allowed me to benefit so much from the group. I would like to express my deep sense of gratitude to my best friend, Yoges for her friendship and continuous support. I extend my special thanks to Chong for his kind help whenever I needed it. I also owe thanks to PTP-UTM and MOSTI for their financial support. Finally, my heartfelt thanks go to my parents, for always standing behind me and for their patience when I am trying to discuss something about my work (a subject which is completely unknown to them). Thanks to my family members for the support and especially to my sister for her long-distance chats.

7 v ABSTRACT The feasibility of upgrading natural gas that primarily consists of methane to valuable chemicals, especially liquid fuel has been investigated for years. However, the high cost and inefficient processes have hampered the widespread exploitation of natural gas. Accordingly, a system comprising of a dual-bed catalytic has been investigated in this study to overcome the limitations and permit the direct conversion of methane to liquid hydrocarbons. In this dual-bed system, methane is converted in the first stage to Oxidative Coupling of Methane(OCM) products over La/MgO and the second bed comprises of H-ZSM-5 that has been tested as an oligomerization function to convert the OCM products to liquid hydrocarbons. The influence of SiO 2 /Al 2 O 3 ratio of H-ZSM-5, temperature and CH 4 /O 2 ratio on the process has been studied. The results implied that the Bronsted acid sites of H-ZSM- 5 were the active centers responsible for the oligomerization of primary ethylene products. Oxygen was absolutely necessary for the formation of the methyl radicals from methane, but it should be provided at a controllable manner in order to avoid undesired oxidation. The partial destruction and dealumination of H-ZSM-5 at higher temperature had caused the deactivation of the H-ZSM-5 catalyst for the oligomerization reaction. Investigation on the catalytic activity of various metals loaded H-ZSM-5 showed that incorporating nickel into H-ZSM-5 significantly enhances the yield of liquid hydrocarbons. The central composite design (CCD) coupled with response surface methodology (RSM) was successfully applied to map the response and to obtain the optimal reaction design. The results indicated that the optimum C 5+ yield of 8.91% was attained at reaction temperature = 742 ºC, CH 4 /O 2 ratio = 9.7 and Ni loading = 0.67 wt%. This exploration suggests that the concept of this dual-bed catalytic system is an interesting candidate for application in methane utilization to produce liquid hydrocarbons.

8 vi ABSTRAK Keupayaan untuk mempertingkatkan gas asli yang mempunyai metana sebagai komponen utama kepada bahan kimia yang bernilai terutamanya cecair bahanapi telah diuji beberapa tahun yang lalu. Namun demikian, kos pengendalian yang tinggi dan ketidakcekapan proses telah menghadkan pengeksploitasian meluas gas asli. Sejajar dengan itu, sistem yang mengandungi dua lapisan mangkin telah dicadangkan dalam penyelidikan ini untuk mengatasi masalah tersebut dan membolehkan penukaran terus metana kepada cecair hidrokarbon. Dalam sistem dua lapisan mangkin ini, metana ditukarkan kepada produk pengoksidaan berpasangan metana (OCM) pada peringkat pertama dan seterusnya lapisan kedua yang mengandungi H-ZSM-5 telah diuji untuk berperanan dalam tindak balas pengoligomeran untuk menukarkan produk OCM kepada cecair hidrokarbon. Pengaruh nisbah SiO 2 /Al 2 O 3 pada H-ZSM-5, suhu dan nisbah CH 4 /O 2 ke atas proses telah dikaji. Keputusan eksperimen membuktikan bahawa asid Bronsted pada H- ZSM-5 bertindak sebagai pusat yang aktif yang berfungsi kepada tindak balas oligomerasi etena. Oksigen sangat diperlukan untuk pembentukan radikal metil daripada metana tetapi ia perlu dibekalkan pada tahap yang sesuai untuk mengelakkan pengoksidaan yang tidak diingini. Suhu tindak balas yang tinggi akan mengakibatkan kemusnahan separa dan dealuminasi HZSM-5 yang mengurangkan keaktifan mangkin untuk tidak balas oligomerasi. Kajian ke atas aktiviti mangkin H- ZSM-5 yang mengandungi logam sebagai fungsi tambahan kepada fungsi asid menunjukkan bahawa penambahan nikel pada H-ZSM-5 jelas meningkatkan kepemilihan cecair hidrokarbon. Rekabentuk Central composite design (CCD) bergandingan dengan response surface methodology (RSM) telah digunakan dengan jayanya untuk melakarkan respon dan untuk mendapatkan rekabentuk optimum tindak balas. Keputusan menunjukkan bahawa nilai optimum kepemilihan C 5+ ialah 8.91% yang dicapai pada suhu= C, nisbah CH 4 /O 2 =9.7 dan muatan nikel=0.67%berat. Kajian ini mencadangkan bahawa konsep yang mempunyai dua lapisan mangkin adalah satu rekabentuk yang menarik untuk digunakan dalam penukaran metana kepada cecair hidrokarbon.

9 vii TABLE OF CONTENTS CHAPTER TITLE PAGE TITLE DECLARATION DEDICATION ACKNOWLEDGEMENTS ABSTRACT ABSTRAK TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF SYMBOLS LIST OF ABBREVIATIONS LIST OF APPENDICES i ii iii iv v vi vii xiii xiv xvii xviii xx 1 INTRODUCTION General Introduction Natural Gas Natural Gas Conversion Technology Gas to Liquids Technology Oxidative Coupling of Methane Hydrogen From Methane Conversion

10 viii 1.4 Heterogeneous Catalysis and Catalyst Zeolite History and Zeolite 25 Development Structure of Zeolites Mechanism of the Shape 30 Selectivity ZSM-5 Zeolite Catalyst Magnesium Oxide-based Catalysts for OCM Reactions The Statement of Problems Hypothesis Research Objectives Scope and Structure of this Thesis 43 2 EXPERIMENTAL APPROACH Introduction Catalyst Preparation Zeolite Preparation of Metal/H-ZSM-5 based 49 Catalysts Preparation of La/MgO Catalyst Catalyst Characterization Temperature Programmed Desorption 51 Ammonia Zeolite Surface Area by Nitrogen Adsorption Reaction Unit and Product Analysis Experimental Rig Reactant Delivery System Reactor System 57

11 ix Product Separation System Analytical Procedure General Testing Procedure 59 3 ACIDITY VERSUS ACTIVITY IN DUAL BED CATALYTIC SYSTEM FOR DIRECT CONVERSION OF METHANE TO LIQUID HYDROCARBONS Introduction Dual Bed Catalytic System Acid Properties of Zeolite Bronsted Acid Sites Lewis Acid Sites Experimental Catalyst Preparation Catalyst Characterization Catalyst Activity Measurement Results and Discussion Catalysts Characterization SiO 2 /Al 2 O 3 Ratio Effect Thermal Treatment 74 Analysis of the H-ZSM-5 Samples Catalytic Performance Effect of Temperature Effect of Oxygen 80 Concentration Effect of Acid Site Concentration Summary 87

12 x 4 CATALYTIC DIFFERENT FUNCTION OF THE METALS ADDED TO H-ZSM-5 CATALYST IN DUAL BED CATALYTIC SYSTEM Introduction Catalysis of Metal Loaded H-ZSM Experimental Catalyst Preparation Catalytic Activity Measurement Results Methane Conversion Selectivity of Ethane and Ethylene Selectivity of C 3 and C 4 Products Ratio of CO 2 to CO Yield of C 5+ Products Discussion Summary PROCESS OPTIMIZATION STUDIES OF DUAL BED CATALYTIC REACTOR SYSTEM FOR CONVERSION OF METHANE TO LIQUID HYDROCARBONS Introduction Overview of the Nickel based 112 Catalysts Design of Experiment (DOE) Central Composite Design (CCD) Response Surface Methodology Analysis of the Experimental Data Experimental 125

13 xi Catalyst Preparation Catalyst Characterization Catalytic Activity Measurement Selection of Appropriate 126 Parameter Reaction Conditions Experimental Design and Statistical Analysis Results and Discussions Catalyst Characterization Textural Properties Acid Properties Preliminary Studies of the Reaction 133 Parameters Effect of Temperature Effect of CH 4 /O 2 Ratio Effect of Nickel Loading Response Surface Methodology and 145 Analysis of Variance Model Development Response Surface 151 Contours Verification of Model and Optimum Condition Summary GENERAL CONCLUSIONS AND RECOMMENDATIONS Conclusions Recommendations 161

14 xii REFERENCES 163 APPENDICES A-C 200

15 xiii LIST OF TABLES NO. TITLE PAGE The typical composition of natural gas before it is refined Characteristics of some typical zeolite porous material Zeolite HZSM-5 and their manufacture assay data Metal salts and their manufacture assay data Physical and chemical properties of the reactant gases Acidity of H-ZSM-5 catalysts with different SiO 2 /Al 2 O 3 73 ratios by TPD-NH NH 3 sorption capacity of the H-ZSM-5 samples treated at 76 various temperatures 5.1 The ANOVA table Textural properties of the nickel loaded materials as 130 measured by N 2 adsorption-desorption isotherms at 77K 5.3 Acidity of Ni-containing H-ZSM-5 catalysts by TPD-NH Independent variables and their coded levels for the 145 central composite design used in the present study 5.5 Central composite design and experimental results Analysis of variance (ANOVA) for model regression Verification of experimental and predicted values of C yield under the optimal conditions predicted by RSM 5.8 The composition of liquid hydrocarbons at optimum 156 conditions Al GC Analysis report of the gas standard 202

16 xiv LIST OF FIGURES NO. TITLE PAGE Geographical distribution of proven natural gas World primary energy demand Methane to liquid hydrocarbons conversion routes Typical stages in GTL technology Flow diagram of the GTL process Reaction scheme of the OCM over the basic oxides 16 catalyst Thermodynamics of some methane conversion reactions which produce hydrogen A catalyst provides an alternative route for the reaction with a lower activation energy 1.9 Schematic representation of building zeolites Schematic diagram of shape selectivity of zeolite Illustration of primary unit of TO 4 tetrahedral, secondary building units, the MFI framework and channel structure of ZSM-5 zeolite 1.12 Cubic (rocksalt) MgO crystal: different planes Model of oxygen atom flux in MgO-based catalysts Model of active center generation and the OCM reaction 38 model 2.1 Flow chart of overall experimental work Schematic diagram of ammonium ions replaced by hydrogen ions 48

17 xv 2.3 Schematic illustration of the procedures for preparing 49 metal containing HZSM Typical NH 3 -TPD profile A schematic diagram of laboratory reaction units set up for 55 catalyst 2.6 Schematic presentation of the dual-bed catalyst reactor 57 arrangement 3.1 Simplified reaction scheme for the dual bed catalytic 64 system over La/MgO and H-ZSM-5 catalysts 3.2 Neutral All-Silica Framework Formation of acid sites in zeolites Inter-conversion of Bronsted and Lewis acid sites Model of Brönsted acid sites in zeolites Temperature programmed desorption of ammonia from H- 71 ZSM-5 with different SiO 2 /Al 2 O 3 ratios 3.7 NH 3 -TPD profiles of H-ZSM-5 catalysts treated at 75 different temperatures 3.8 Influence of reaction parameters on the catalytic activity 81 and products distribution ( methane conversion, ethylene to ethane ratio, selectivity of C 3, selectivity of C 4, selectivity of C 5+ and CO to CO 2 ratio) 4.1 Methane conversion over metal loaded H-ZSM-5 catalysts Selectivity of ethane and ethylene over metal loaded H- 95 ZSM-5 catalysts( ethane and ethylene) 4.3 Selectivity of C 3 and C 4 products over metal loaded H- 97 ZSM-5 catalysts ( C 4 and C 3 ) 4.4 The ratio of CO 2 to CO over metal loaded H-ZSM-5 98 catalysts 4.5 Yield of C 5+ products over metal loaded H-ZSM-5 99 catalysts 4.6 Simplified reaction scheme for dual bed catalytic system Generation of a central composite design for two factors Central composite design for three factors 119

18 xvi 5.3 The flow chart of RSM General response surface types: A) Peak, B) Hillside, 122 C) Rising Ridge, and D) Saddle A Black Box process model schematic NH 3 -TPD profiles of nickel-modified and unmodified H ZSM-5 catalysts [(a)h-zsm-5, (b) 1%Ni/H-ZSM-5 and (c) 3% Ni/H-ZSM-5] 5.7 Influence of reaction parameters on the catalytic activity 136 and products distribution ( methane conversion, ethylene to ethane ratio, selectivity of C 3, selectivity of C 4, selectivity of C 5+ and CO to CO 2 ratio) 5.8 Pareto chart for the evaluation of parameter effects Parity plot of the predicted and observed C 5+ yield The response surface plot of C 5+ yield as a function of 152 reaction temperature (x 1 ) and CH 4 /O 2 (x 2 ) 5.11 The response surface plot of C 5+ yield as a function of 153 reaction temperature (x 1 ) and % Ni loading (x 3 ) 5.12 The response surface plot of C 5+ yield as a function of 154 reaction %Ni (x 3 ) and CH 4 /O 2 (x 2 ) A1 Gas Chromatogram of gas standard supplied by Scott 200 Specialty Gases A2 Gas Chromatogram of PIONA standard 201

19 xvii LIST OF SYMBOLS ai - The number of atoms of the carbon presents in each molecule of C the chemical species i Å - Angstrom 0 C - Degree Celsius F - The ratio of mean squares due to regression to mean squares due to residual Fi - Gas species i flowrate k - Number of independent variables K - Kelvin degree n o - The number of experiments repeated at the center point p - The number of terms in the fitted model R 2 - The determinant of coefficient w - The number of water molecules per unit cell x 1 - The coded forms of input variables for operating temperature x 2 - The coded forms of input variables for CH 4 /O 2 ratio x 3 - The coded forms of input variables for amount of nickel loading X i, X j - The independent variables Y - Experimental values of yield C 5+ y - Mol fraction - The predicted yield of C 5+ - Star point value 1, 2, 3 - The linear terms 11, 22, 33 - The squared terms 12, 13, 23 - The interaction terms o - The offset term

20 xviii LIST OF ABBREVIATIONS AGC-21 - Advanced Gas Conversion for the 21st Century ANOVA - Analysis of variance BET - Brunauer-Emmett-Teller's surface area model CCD - Central composite design CNG - Compressed natural gas DOE - Design of experiments ESR - Epsilon Serializability Resonance FCC - Fluid catalytic cracking FID - Flame Ionization Detector GC - Gas chromatography GTL - Gas to Liquids technologies HT - Hydrogen transfer I.D. - Internal diameter LNG - Liquefied natural gas M - Metal MFI MTG - - Mobil Five Topology Methanol to gasoline NA - Nitrogen adsorption NGV - Natural gas vehicles NMR - Nuclear Magnetic Resonance NNN - Nearest neighbour positions OCM ODH - - Oxidative Coupling of Methane Oxidative Dehydrogenation NODH PIONA - Non-Oxidative Dehydrogenation Paraffins, Iso-paraffins, Olefins, Naphtha, and Aromatics REM - Rare-earth-metal RON - Research Octane Number

21 xix RSM - Response surface methodology S.S - Sum of square SBU - Smallest number of TO 4 units SHOP - Shell higher olefin process SMDS - Shell Middle Distillate Synthesis SSE - Squares of the error SSR - Squares due to regression SST - Sum of squares Syngas - Synthesis gas TCD - Thermal conductivity detector TPD - Temperature programmed desorption TPR - Temperature programmed reduction XRD - X-ray diffraction ZSM-5 - Zeolite Socony Mobil Five

22 xx LIST OF APPENDICES APPENDIX TITLE PAGE A Gas Chromatogram of Standard and the 200 Calculation of Gas Sample B Calculation of Conversion, Yield and Selectivity 203 Using Internal Standard C List of Publications 205

23 CHAPTER 1 INTRODUCTION 1.1 General Introduction Natural gas may become more important both as an energy source and as the source of organic chemical feedstock in the 21 st century (Javanmardi et al., 2005). In 2001, natural gas accounted for a 24% share of the world s primary energy consumption (Aguiar et al., 2005). In addition, economically recoverable natural gas reserves have risen continuously over the last decades. However, the world s natural gas reserves are located in only a few countries. The geographical distribution of natural gas is given in Figure 1.1. The largest reserves are located in the countries of the former Soviet Union and the Middle East. Future demands will be met by volumes from reserves located at even greater distances, and the investment associated with gas procurement is therefore expected to rise. Furthermore, the decline in oil production, which is forecast to occur from onwards, will result in an increasing role for other fuels, most of which will be significantly more expensive (Ashman and Mullinger, 2005). Therefore, the evolution of the known natural gas reserves worldwide indicates a dramatic increase and this trend is expected to continue, which will, in addition to the price development with respect to the crude oil based upgrading-most likely generate a gradual shift towards the application of natural gas as a feedstock for the production of fuels and petrochemicals (Siddiq, 2002). This situation has forced an enhanced global interest in processes which can convert natural gas into liquids and higher added value products.

24 As the concerns about global warming have increased, there is more Figure 1.1 Geographical distribution of proven natural gas (The BP Statistical Review of World Energy 2004) On the other hand, natural gas, which is basically methane, is a colourless, odourless fossil fuel that is already widely used in homes and industry, and will be used to a greater extent in the future. It is fairly inexpensive, clean-burning, and abundant, and in recent years, it has made headlines as a solution to many environmental and economic concerns facing this country. Because it has received so much positive attention, people are now looking at ways to increase our ability to recover and utilize this resource. Although it does have some drawbacks compared to other fuels, mainly issues dealing with its volume, it promises to be an increasingly important energy source in years to come. Natural gas has shown to be a viable substitute for other fossil fuels, namely oil and coal, as concerns over the supply of these other nonrenewable fuels grow. Natural gas is also nonrenewable; however, we have large, untouched deposits in Malaysia that, if developed, will make the sale of natural gas easier in a recognized commodities market. Also, natural gas burns cleaner and produces less pollutant than other fossil fuels. It produces 45% less CO 2 than coal for a comparable amount of energy, and also emits considerably less nitrogen oxide and sulphuric acid

25 3 (Spivey, et al., 2005). This is a good thing, because carbon dioxide has been blamed to be the cause of the greenhouse effect. Figure 1.2 World Primary Energy Demand (Birol and Argiri, 1999) Projections of world energy demand by fuel are illustrated in Figure 1.2. Oil continues to dominate world energy consumption. However, recent studies, based on the current energy consumption growth, have shown that the world petroleum reserve will be depleted within the next century (Lunsford, 2000). Thus, for the survival of the world, research on alternative feedstock utilization has become one of the prime technologies of the twenty first century and natural gas seems very promising at this stage. It burns with less emissions than gasoline, and has a higher fuel efficiency (a higher percentage of the energy in the fuel can be used). However, one concern with using natural gas is the volume of the gas, since it is much greater than that of liquid fuels. Even using CNG, (compressed natural gas), the fuel tanks of NGV s (natural gas vehicles) must be quite large, and even then, fuel range is not as

26 4 much as if it had gasoline. On larger vehicles, such as buses, LNG (liquid natural gas) can be used. LNG is natural gas that is cooled to 166 degrees Farenheit, so that it condenses to a liquid (Kaneko et al., 2004). Thus, it is much more dense than CNG, although still not as energy dense as gasoline or diesel. However, in order to use LNG, the vehicle must be equipped with a large cryogenic cooling system to keep the natural gas as a liquid, so LNG is not really feasible for smaller vehicles like cars. Because the attention on natural gas as an alternative to oil or coal is relatively new, researchers are now working on ways to increase the production, processing, and distributing abilities of methane. As demand for natural gas rises in the future, probably sharply, a greater demand will be put on our natural gas infrastructure. When this occurs, a closer look will be directed to options that were previously thought unviable. Another area that is currently being developed in anticipation of the future demand is GTL (gas-to-liquid) technology. There are many potentially rich gas fields that are not drilled because of the need to lay expensive steel pipes to transport the gas. If the field is very far away from the storage reservoir, the cost of the pipeline could discourage any speculators on the field. Transporting by ship, train, or truck is much less costly, but methane s high volume as a gas prevents it from being transported in bulk. GTL is technology that allows the producer to convert the gas into a liquid fuel, allowing it to be transported much more compactly and hence more easily (Wilhelm et al.,2001). As mentioned before, liquid natural gas (LNG), is one way of accomplishing this. However, the need to constantly keep LNG at a very low temperature is a drawback, sometimes a serious one. This leads to the other area of GTL, chemically reacting methane with other compounds, to produce a stable liquid methane-based fuel. This new liquid form is ready to be used for transportation, as well as to directly supply power to machinery, in the same manner gasoline is used. Producing GTL fuel is very costly at this time; however, because a methane-based liquid fuel has a large amount of positive characteristics, it is being heavily researched, and in the future, as costs go down, we should see more of these liquid hydrocarbon fuels.

27 5 1.2 Natural Gas Natural gas is a combustible mixture of hydrocarbon gases. It is colorless, shapeless, and odorless in its pure form. As a safety measure, natural gas companies add an odorant, mercaptan, to the gas so that leaking gas can be detected. While natural gas is formed primarily of methane, it can also include ethane, propane, butane and pentane. The composition of natural gas can vary widely. Table 1.1 shows outlines the typical makeup of natural gas before it is refined. Natural gas is considered 'dry' when it is almost pure methane, having had most of the other commonly associated hydrocarbons removed. When other hydrocarbons are present, the natural gas is 'wet' (Thomas and Dawe, 2003). The primary component of natural gas is methanet (CH 4 ), the shortest and lightest hydrocarbon molecule. It may also contain heavier gaseous hydrocarbons such as ethane (C 2 H 6 ), propane (C 3 H 8 ) and butane (C H ), as well as other sulphur containing gases, in varying amounts, see also natural gas condensate. Organosulfur compounds and Hydrogen sulfide (H S see acid gas) are common contaminants, 2 which must be removed prior to most uses. Gas with a significant amount of sulfur impurities is termed "sour". Moreover, combustion of one cubic meter of commercial quality natural gas yields 38 MJ (10.6 kwh) (Yagi et al., 2005) Methane is also an extremely efficient greenhouse gas which may contribute to enhanced global warming when free in the atmosphere, and such free methane, would then be considered a pollutant rather than a useful energy resource. However, methane in the atmosphere reacts with ozone, producing carbon dioxide and water, so that the greenhouse effect of released methane is relatively short-lived. As a pollutant, significant biological sources of methane are termites, cattle (ruminants) and cultivation (estimated emissions are 15, 75 and 100 million tons per year respectively) (Aguiar et al., 2005). Natural gas is a fossil fuel like oil and coal, this means that it is, essentially, the remains of plants and animals and microorganisms that lived millions and millions of years ago. There are many different theories as to the origins of fossil fuels. The most widely accepted theory says that fossil fuels are formed when organic matter (such as the remains of a plant or animal) is compressed under the

28 6 earth, at very high pressure for a very long time. This is referred to as thermogenic methane (Lunsford, 2000). Similar to the formation of oil, thermogenic methane is formed from organic particles that are covered in mud and other sediment. Table 1.1 The typical composition of natural gas before it is refined (Kvenvolden, 1995) Typical Composition of Natural Gas Methane CH % Ethane C 2 H % Propane C 3 H % Butane C 4 H % Carbon Dioxide CO 2 0-8% Oxygen O % Nitrogen N 2 0-5% Hydrogen Sulphide H 2 S 0-5% Rare Gases A, He, Ne, Xe trace This sediment and debris puts a great deal of pressure on the organic matter, which compresses it. This compression, combined with high temperatures found deep underneath the earth, break down the carbon bonds in the organic matter. As one gets deeper and deeper under the earths crust, the temperature gets higher and higher (Lunsford, 2000). At low temperatures (shallower deposits), more oil is produced relative to natural gas. At higher temperatures, however, more natural gas is created, as opposed to oil. 1.3 Natural Gas Conversion Technologies Natural gas is found in considerable amounts in oil fields, often at remote locations, where the construction of pipelines to transport the gas is not economical. Under these circumstances, it is usually flared, vented, or pumped undergrounds

29 7 (Thomas and Dawe, 2003). Therefore, finding an efficient process for utilizing such stranded gas receives considerable attention. In this direction, the conversion of methane, the main constituent of natural gas, into more valuable hydrocarbons is a topic of great interest. Up to now, indirect processes, involving partial oxidation and Fischer-Tropsch synthesis, are being used for this purpose. Recently many attempts have been made to develop a direct process for converting methane into higher hydrocarbons and the effective chemical activation of the methane molecule remains the most challenging step in such a process. In this research, an attempt is made to initiate a route for direct liquid hydrocarbons (C 5+ ) production by catalytic reaction oxidation of methane with oxygen. Before that, various technology of methane conversion to higher hydrocarbon will be discussed further in this chapter to obtain a complete understanding about the processes involved. The abundance of natural gas (the main constituent of which is methane), together with steadily depleting oil reserves, highlights methane conversion to higher hydrocarbons as an extremely attractive goal. Activation of methane is very challenging due to its refractory nature and has made this one of the most extensively investigated processes in catalysis. The processes of conversion of natural gas into liquid higher hydrocarbons products can be divided into two types: direct conversion and indirect conversion as shown in Figure 1.3. The primary step of the indirect route involves the production of syn-gas (a mixture of CO and H 2 ) by steam reforming, partial oxidation, autothermal reforming; following this step the syn-gas is converted to hydrocarbons directly via Fischer Tropsch catalysis or indirectly via other intermediates such as methanol. The direct conversion processes utilize catalysts and specific synthesis routes to chemically transform the molecules of methane, the main component of natural gas, into more complex chained substances with heavier molecules. The desired products that can be obtained include the alcohols (mainly methanol), the olefins (ethylene and acetylene) and the aromatics (benzene, toluene and naphthalene). However, the high stability of the methane molecule creates a series of technical problems to make the chemical reactions involved viable. Thus, the research and development efforts of the direct conversion processes are focused on the

30 8 improvement of the catalysts in the elucidation of the mechanisms of reaction and the development of new equipment. Taylor et al. (1988) explained that methane, oxygen, and hydrogen chloride (mixed CuCI, KCI, LaCI 3 ) are reacted over an oxyhydrochlorination catalysts. In the first stage produce methyl chloride and water. In second stage, the methyl chloride is converted to higher hydrocarbons, namely paraffins, olefins, aromatics and cycloparaffins, over zeolite, such ZSM-5, as shown by the chemical equation in (Equation 1.1 and Equation 1.2 ). CuCl, KCl, LaCl 3 CH 4 + O 2 + HCI CH 3 CI + H 2 O (1.1) ZSM-5 CH 3 Cl Gasoline + HCI (1.2) They reported that the conversion of methlychloride to gasoline range hydrocarbon (31-42 %) accurse under conditions similar to those for methanol to gasoline conversion. The oxyhydrochlorination catalysts was prepared in a nonaqueous solvent by successive impregnation of metal chloride salts onto a silica support. Final weight composition was as 41.7 % CuCI, 37.5% SiO 2, 11.5 % KCI and 9.4 % LaCI 3. the ZSM-% catalysts was obtained from Mobil Oil Cooperation in the ammonium form with silica to alumina ratio of 70:1 (Tylor et al.,1991). The ammonium form was converted to the hydrogen form by calcaning in air at 538 o C for 16 hrs. Senkan (1987) has patent a process for oxidative pyrolysis of halogenated methanes in the gas phase and under non-flame conditions in the presence of oxygen which significantly reduces the formation of carbonaceous deposits, such as tars, solid carbon and soot, while maintaining high yields, in the order of 20 to 80 percent, of desired higher molecular weight hydrocarbon products. According to a preferred embodiment, the process of that invention may be a two step process wherein the first step involves the halogenation of methane using a halogen containing gas or oxyhalogenation using hydrogen halide in the presence of oxygen. Then in a second step, the halogenated methanes are oxidatively pyrolyzed under non-flame conditions

31 9 in the presence of an oxygen containing gas. The process of that invention may also be a single step process wherein the methane halogenation and the oxidative pyrolysis of halogenated methanes are accomplished in a single vessel. According to that invention, oxygen converts to carbon monoxide hydrocarbon compounds which would otherwise result in the formation of carbonaceous deposits during conversion of halogenated methanes. Oxidative pyrolysis of the halogenated methanes yields higher molecular weight hydrocarbons such as acetylene and ethylene. A process for oxidative pyrolysis of halogenated methanes in the presence of oxygen-containing gas under non-flame conditions is provided whereby the formation of solid carbonaceous materials is significantly reduced, while high yields of desired higher molecular weight hydrocarbons such as acetylene and ethylene are maintained. However, the oxyhydrochlorination process is not economical, because the chloride acid is very corrosive and the cost of special materials were replaced with carbon steel throughout (Cavani and Trifiro, 1997). The indirect route is technically easier than that of the direct conversion processes. The technologies of the processes involved have been better studied and several pilot and commercial plants are already in operation (Wilhelm, 2001). The processes of indirect conversion are characterized by a preliminary stage of transformation of natural gas into synthesis gas - syngas - (a mixture of carbon monoxide - CO and hydrogen - H 2 ). Synthesis gas production requires either steam (steam reforming) or oxygen (partial oxidation) as a co-reactant. In either case, generation of these reactants is extremely energy and capital intensive and, as a result, the major cost of converting natural gas to liquid fuels lies in the initial synthesis gas production step. After being produced, the syngas is converted into liquid hydrocarbons through the Fischer-Trospch Process (FT) (Gradassi and Green, 1995; and Vosloo, 2001). On the other hand, hydrogen is perceived as an ideal energy carrier for a clean and sustainable energy future in the world, and fuel cells will play a significant role in this hydrogen based energy system as flexible, clean, and efficient energy devices suitable for use in a wide range of transportation and stationary/portable power generation applications. Among many potential sources for hydrogen

32 10 METHANE INDIRECT ROUTE DIRECT ROUTE STEAM REFORMING DRY REFORMING TWO STEP SINGLE STEP PARTIAL OXIDATION ATR SYNGAS PARTIAL OXIDATION OXIDATIVE COUPLING METHANOL HALOGENATION ETHYLENE FISCHER- TROPCSH MTG PROCESS CLOROMETHANE LIQUID FUELS & CHEMICALS Figure 1.3 Methane to liquid hydrocarbons conversion routes

33 11 generation, natural gas, which is clean hydrocarbon fuel, abundant, and welldistributed, is considered to be one of the ideal fuels for hydrogen source for fuel-cell stationary applications. Many companies are developing natural-gas fuel-cell systems for residential/stationary applications and commercial/industrial markets. Reforming natural gas is a well-established technology in the petrochemical and fertilizer industries for the production of hydrogen for use in the manufacture of ammonia, methanol, and other chemical products (Lee et al., 2005) This technology has been reviewed in detail and includes three basic processes: steam reforming (SR), partial oxidation reforming (POR), and auto thermal reforming (ATR). In recent years, the catalytic ATR process has received much research and development attention as a viable process for hydrogen generation for fuel-cell systems because it offers advantages of simpler design (smaller and lighter unit), lower operating temperature (easier start-up and a wider choice of materials), and flexible load following characteristic as compared to the SR process. It also has higher energy efficiency than the POR process (Lee et al., 2005). The methane-steam reforming reactions are strongly endothermic and the thermodynamic equilibrium is favored by high temperature. The conventional steam reformers are mainly fixed bed catalytic reactors. The catalyst is normally loaded into a number of tubes placed in a furnace, which operates at high temperature varying from 850 to 1100 K. The reactors suffer from uniform distribution of heat fluxes and profound effect of high temperature on the tubes and catalyst life. Moreover, the diffusion limitations are severe with very low effectiveness factor in the order of ( ) (Abashar, 2004). In addition, the reactions produce significant quantities of carbon dioxide (important greenhouse gas), which is harmful to the environment. In recent years various experimental and theoretical attempts have been made to improve different aspects of processing methane to produce hydrogen. Considerable attention has been paid to the fluidized bed membrane reactors (FBMRs) as multi-functional reactors (Abashar, 2004). Their main advantages are: shift of the thermodynamic equilibrium, enhancement of methane conversion, simultaneous reaction and separation of hydrogen, elimination of diffusion limitations, good heat transfer capability and a more compact design (Laosiripojana

34 12 and Assabumrungrat, 2005). Also, the carbon dioxide reforming of methane, the socalled dry reforming of methane (CO 2 reforming) has gained increasing importance. The process of great significance for environmental protection, it enables the transformation of CO 2 into synthesis gas (syngas). Recently, several studies have shown that the performance of the fixed bed catalytic reactors has improved significantly by introduction of the dual-functionality via structured pattern (either mixed or layered configuration) of the catalyst bed. The purpose of these patterns is to couple the reactions in different ways to achieve certain degree of integration of heat, further shift of thermodynamic equilibrium and simultaneous production of more than one product (Abashar, 2004). The following sub-chapters will deal with the GTL - technology as well as the oxidative coupling of methane (OCM), where zeolites and related MgO based material respectively have been demonstrated to be superior catalysts Gas to Liquids (GTL) Technology GTL technology is based on the conversion of natural gas to liquid fuels in three stages: synthesis gas generation, production of heavy-chain hydrocarbons by Fischer-Tropsch synthesis and heavy fraction hydrocracking for production of useful compounds such as naphtha, diesel and lubricants (Figure 1.4) (Aguiar et al., 2005; and Yagi et al., 2005). Although the three stages of the GTL process by indirect means have been individually well studied and are available for commercialisation, an optimal combination that permits the reductions of costs of the commercial production plants still does not exist. Additionally, the synthesis gas generation stage corresponds to the greatest costs in investments and operation, being responsible for around 50 75% of the capital costs (Vosloo, 2001). Therefore, many large companies are seeking to optimize the synthesis gas generation stage, in order to make the GTL technology commercially viable (Wittcoff et al., 2004).

35 13 Figure 1.4 Typical stages in GTL technology (Yagi et al., 2005) On the other hand, fuels production directly from syngas (in former times obtained from coal) has been reported by Fischer and Tropsch in 1923 for the first time (Wilhelm et al., 2001), using an alkali-promoted iron catalyst. Fuels manufactured via the Fischer Tropsch route reveal an excellent quality since they consist mainly of linear paraffins and olefins and do not contain sulfur and aromatics. A Co-containing catalyst is applied for the production of heavy paraffins via the Fischer Tropsch route starting with natural gas, a technology developed by Shell and named the Shell Middle Distillate Synthesis (SMDS) route (Lunsford, 2000). In addition, diesel fuel (or gasoline) is produced by hydrocracking of the more or less sulfur and nitrogen-free wax obtained through the SMDS process using noble metal containing zeolites. The more restricted fuel specifications currently introduced in order to reduce the environmental impact of hazardous emissions represent a driving force with respect to an increased use of fuels prepared via the Fischer Tropsch route as a blending component of the gasoline and diesel pools in the future (Aguiar et al., 2005). Figure 1.5 is a block flow diagram of the GTL process.

36 14 Figure 1.5 Flow diagram of the GTL process (Aguiar et al., 2005) Besides the SMDS technology, an alternative has been presented by SASOL/Chevron termed as the Slurry-Phase-Distillate process, again based on the Fischer Tropsch route producing wax (using a Co-containing catalyst) followed by a hydrocracking step in order to get diesel or gasoline (Espinoza et al., 1999). The methanol to gasoline (MTG) plant in New Zealand has been combined with a methane steam reforming unit for production of synthesis gas and a methanol plant to produce gasoline from natural gas. The process economics can be improved considerably by a clever combination and close integration of the different steps. Finally, ExxonMobil has introduced the so-called Advanced Gas Conversion for the 21st Century (AGC-21) technology, again based on the Fischer Tropsch route (Ashley et al., 2003; and Wittcoff, et al., 2004). In any manner, the reduction of costs of any stage is still the principal objective of all companies developing GTL. The use of catalytic membranes and compact reformers represents considerable advances in the area. Another possibility is in the elimination of the isomerization unit, which would occur if the Fischer Tropsch catalysts were able to generate isomerized products. The addition of

37 15 Bronsted acid sites to the traditional cobalt catalysts appears to be an interesting research option that is still under-explored. This catalyst was tested and refined to improve performance; thus creating an ultra-stable catalyst with high activity and selectivity. Using an integrated simulation model for reforming and F-T synthesis, designers developed detailed mass balance flowsheets, engineering line diagrams, utility line diagrams, plant layouts, process and mechanical specifications of all equipment and instrumentation and comprehensive schedules of all bulk materials Oxidative Coupling of Methane Conversion of methane into liquid products is traditionally achieved via steam reforming to produce CO and H 2, followed by transformation of these diatomic molecules into oxygenates and higher hydrocarbons. Because the energy input required for steam reformation is exorbitant, alternative routes to methane conversion have been pursued. One approach which leads to direct formation of C 2 products is oxidative coupling of methane (OCM). This oxidative coupling of methane (OCM) is a promising route for the conversion of natural gas to ethylene that can be used for the production of petrochemicals or fuel. The reaction takes place in the presence of catalysts at temperatures from 650 to 850 o C (Zaman, 1997). In the oxidative coupling reaction, CH 4 and O 2 react over a catalyst at elevated temperatures to form C 2 H 6 as a primary product and C 2 H 4 as a secondary product. The reaction network is interesting from a fundamental perspective because it is an example of a heterogeneous homogeneous system. Oxides of alkali, alkaline earth and rare earth metals, singly or in combination, catalyze the reaction to the desired path. Maitra (1993) analyzed the performance of the different types of OCM catalysts and concluded that only the basic oxides constitute good catalysts. According to this mechanism, the methane activation proceeds via hydrogen abstraction on a basic O 2 site producing CH 3 which, in a secondary reaction with oxygen, produces CH 3, simultaneously

38 16 converting O 2 to O 2. Methyl radicals that are formed at the surface of the catalyst enter the gas phase where they couple to form ethane. Over basic oxides catalysts, it is indeed observed that secondary reactions of ethane are responsible for ethylene formation (Martin and Mirodatos, 1995). In addition to coupling, the gas phase radicals may enter into chain reactions that result in the formation of CO and subsequently CO 2. Isotopic labeling experiments have demonstrated that at small conversion levels, most of the CO 2 is derived from CH 4, but at commercially significant conversion levels, C 2 H 4 would be the dominant source of CO 2 (Martin and Mirodatos, 1995). Additional experiments have shown that this occurs mainly via a heterogeneous reaction. One of the challenges in catalyst development is to modify a material so that the secondary reaction of C 2 H 4 will be inhibited while the activation of CH 4 will still occur. There is no inherent reason that these two reactions should take place on the same types of sites. A complex model of the OCM reaction over a lanthanum basic oxides catalyst is summarized in Figure 1.6 (Mleczko et al., 1995). Figure 1.6 Reaction scheme of the OCM over the basic oxides catalyst (Mleczko et al., 1995) The scheme includes the nine heterogeneously catalyzed and one homogeneous non-catalytic (reaction 8) reactions:

39 17 (1) total oxidation of methane to carbon dioxide, (2) oxidative coupling of methane to ethane, (3) partial oxidation of methane to carbon monoxide, (4) oxidation of carbon monoxide to carbon dioxide, (5) oxidative dehydrogenation of ethane to ethylene, (6) oxidation of ethylene to carbon monoxide, (7) steam reforming of ethylene, (8) thermal dehydrogenation of ethane to ethylene, (9/10) conversion of CO to CO 2 and vice versa Hydrogen from Methane Conversion Reaction Hydrogen is used as both a chemical feed and a fuel, and is generally produced by steam reforming of hydrocarbons, especially natural gas, from water gas shift reactions, from catalytic reforming of petroleum stocks and by electrolysis. It has been reported that hydrogen can be formed by the decomposition of methane over iron catalysts. This process results in the formation of coke on the catalyst, which ultimately destroys the activity of the catalyst. It may be assumed that decomposition of methane produces hydrogen and carbon black by the following single reaction (Cho et al., 2004): CH (kj/mol) C + 2H 2 (1.3) According also to Cho et al. (2004), the total enthalpy of methane decomposition at C is kj/mol, and the energy related to carbon mass varies approximately between 4 and 7 kw h per kg of carbon produced or 1 and 1.9 kw h per normal cubic meter of hydrogen produced. In the case of OCM and oxidative dehydrogenation (ODH) it is very likely that a serious problem will be faced when trying to extract the intrinsic kinetic constants of elementary heterogeneous interactions from experimental data. The rates

40 18 of reactant conversion and product formation are a complex combination of terms attributed to numerous elementary reactions localized both on the catalyst surface and in the gas phase. However, such a conversion is highly endothermic; under the high temperature reaction conditions required therefore, control of the reaction to prevent the formation of unwanted by products is difficult. As Sinev et al.(2003) demonstrated earlier, free-radical gas-phase processes play a very important role in product formation during light alkane oxidation. The development of chain process in the gas phase leads to the formation of various radical and molecular species further transformations of which can produce hydrogen. First of all, alkyl radicals C n H 2n+1 formed as primary radical species during alkane activation in heterogeneous or homogeneous steps can undergo consecutive scission evolving H-atom which then reacts with any H-containing molecules RH (first of all, with initial alkane) with H 2 molecule formation: [O] S + C n H 2n+2 [OH] S + C n H 2n+1 (1.4) X + C n H 2n+2 XH + C n H 2n+1 (1.5) C n H 2n+1 H + C n H 2n (1.6) RH + H R + H 2 (1.7) where [O] S is the active surface site; X is any gas particle which has an affinity to hydrogen atom (e.g. O 2, OH, HO 2, H, alkyl, alkoxy, alkylperoxy species, etc.). As can be seen, the sequence of steps (1.4), (1.6) and (1.7) leading to the formation of at least one olefin and one H 2 molecules and to the development of homogeneous chain process starts from the oxidative abstraction of H-atom from alkane molecule by a surface active site. If a chain initiation takes place in the gas phase, the first step (3) (in this case X O 2 ) is also an oxidative H-atom abstraction. The same can be said about alkane molecule reaction (1.5) with any oxidative particles present in the gas phase during a developed chain reaction. In other words, this sequence of steps can be considered as an oxidative process leading to molecular hydrogen forming in parallel with olefin. In the work dedicated to the OCM process, Sinev (2003) explained the formation of high concentrations of molecular hydrogen

41 19 during methane oxidation at moderate temperatures (below 650 C) by the formation and rapid decomposition of formaldehyde. The overall reaction of oxidative coupling is: 4CH 4 + O 2 2 C 2 H 6 +2H 2 O (1.8) followed by dehydrogenation of ethane to ethylene: C 2 H 6 C 2 H 4 + H 2 (1.9) However the detailed reaction mechanism is very complex. The many possible side reactions produce other oxidation products such as CO and CO 2. Although hydrogen is also a product of the OCM reaction, it is normally converted back to other products via complex reaction mechanism. In the conditions of optimal production of olefins during oxidation of light alkanes over typical OCM catalysts molecular hydrogen forms as one of the major products. Variations of kinetic features depending on the initial alkane, catalyst and reaction conditions are indicative for the existence of multiple pathways of hydrogen formation. If almost total conversion of oxygen takes place in the front part of the reactor, non-oxidative dehydrogenation (NODH) process can proceed in the remaining part of catalyst bed. Over some catalysts hydrogen formation can also accompany the intense coking. In addition to non-oxidative routes, some oxidative processes, including those proceeding in the gas phase, can lead to the production of hydrogen which in this case is accompanied by both additional olefin formation and total oxidation. However, reactions followed by the hydrogen oxidation by oxygen and hydrogenation with primarily formed olefins decreased the yield of hydrogen in the OCM final reaction products. Further elucidation of the reaction pathways can help to improve the target olefin production in the OCM reaction. On the other hand, oxidative dehydrogenation (ODH) of light alkanes attracts a considerable attention as a potentially efficient way to produce olefins from relatively cheap and abundant feedstocks. The main usually discussed advantage of

42 20 ODH over NODH of the same alkanes is the absence of thermodynamic limitations for per-pass olefin production since in the case of NODH an upper limit of yield determined by equilibrium cannot be exceeded. The thermodynamics of CH 4 conversion is shown in Figure 1.7. Coke formation by reaction (1) in (Figure1.7) is most favorable, it can become one of the steps whereby supported noble metal catalysts can be inactivated. Figure 1.7 Thermodynamics of some methane conversion reactions which produce hydrogen (Gesser and Hunter, 1998) In principle, the removal of hydrogen from the reaction system by any means helps to eliminate the thermodynamic limitations for olefin formation. For instance, selective in situ burning of hydrogen formed over NODH catalyst enables to increase olefin yields. However, catalytic ODH is considered as the most promising way to avoid thermodynamic limitations and necessity of external heating during the

43 21 production of olefins from light alkanes. Indeed, if oxidant (e.g. molecular oxygen) is added to the feed gas the process becomes thermodynamically favorable at any level of alkane conversion and olefin yield. The only condition which has to be provided is water formation instead of H 2. In real practice, however, neither water, nor hydrogen are usually measured experimentally when the reaction mixture is analyzed by on-line gas chromatography (GC). An accurate analysis of water requires additional complications of sampling system. As to the GC analysis of molecular hydrogen, some problems are discussed below. As a result, the analysis of experimental sections of publications dealing with ODH of light alkanes shows that in most cases in these studies only carbon balance is a subject of concern and the fate of hydrogen abstracted from alkane molecule cannot be traced because neither H 2, nor water are analyzed. On the other hand, it is well known that substantial amount of hydrogen are formed during the oxidative coupling of methane (OCM) which is relative to ODH process (both reactions require elevated temperatures; they both proceed via the abstraction of H-atom from C H bond; OCM catalysts are very efficient in ODH). The detailed studies of hydrogen formation during oxidation of C 1 C 4 alkanes over a series of typical ODH catalysts undertaken in some OCM works indicates that the reaction pathways are much more complicated than it has been believed previously. However, the analysis of product yields shows that even if NODH is the only process leading to the formation of olefin, hydrogen production exceeds by far the value corresponding to the formation of olefin and its possible sequential overoxidation. This indicates that certain pathways of hydrogen formation become more important at increasing length of carbon chain in hydrocarbon molecules. The trends described above are suggestive for an existence of multiple pathways for hydrogen formation in the conditions of alkane oxidation in OCM reactions. These pathways should follow different kinetic regularities when reaction parameters, such as temperature, residence times and reactant concentrations, are varied. Moreover, since the highest concentrations of hydrogen in the reaction mixture are observed in the conditions close to optimal olefin formation, the elucidation of the process pathways is worth for further optimization of target product yield. In fact, one possibility is that a substantial fraction of olefin forms

44 22 simultaneously with H 2 via NODH over reduced catalyst at the expense of heat produced in total oxidation reaction (in which the most part of oxygen is consumed). If so, there is no practical reason to run the reaction in the oxidative mode and a more economical way to supply heat for NODH should be found. On the other hand, if hydrogen formation is a result of some by-processes, special efforts should be mounted to suppress them. In the latter case, monitoring of H 2 concentration in the reaction mixture is a good way to control the efficiency of these efforts. Under the usual reaction conditions for the OCM, H 2 is a major reaction product although early studies failed to observe or seemed to have overlooked this due to analytical difficulties, as mentioned in above. Therefore, not many studies seem to have reported the production of H 2 among the numerous studies on the OCM (Choudhary et al., 1997 and Lacombe et al.,1994). H 2 can be formed via a number of pathways, such as steam reforming reactions, the water-gas shift reaction, homogeneous gas-phase methane oxidation, and decomposition of ethyl radicals (Heracleous and Lemonidou, 2004). From a detailed study on the OCM over MgO, Hargreaves et al. (2002) have reported that the water-gas shift reaction is an important H 2 source and that partial oxidation of hydrocarbons becomes more significant at high flow rates and low O 2 conversions; also involved is thermal cracking of ethane to ethene. Hutchings et al. (2002) have proposed from a study on the OCM over MgO, La 2 O 3 and Sm 2 O 3 that gas-phase radical oxidation which is linked directly to the formation of carbon oxides is important but dehydrogenation of ethane is discounted as a major route due to the thermodynamic limitation. Possible reactions producing H 2 in the OCM process are listed as follows (Hong and Yoon, 2001): Steam reforming of hydrocarbons: CH 4 + H 2 O= CO + 3H 2 ; H 2 : CO = 3 (1.10) C 2 H 6 + 2H 2 O = 2CO + 5H 2 ; H 2 : CO = 2:5 (1.11) C 2 H 4 + 2H 2 O = 2CO + 4H 2 ; H 2 : CO = 2 (1.12) Partial oxidation of hydrocarbons:

45 23 CH 4 + 1/2O 2 = CO + 2H 2 ; H 2 : CO = 2 (1.13) C 2 H 6 + O 2 = 2CO + 3H 2 ; H 2 : CO = 1:5 (1.14) C 2 H 4 + O 2 = 2CO + 2H 2 ; H 2 : CO = 1 (1.15) CO 2 reforming of hydrocarbons: CH 4 + CO 2 = 2CO + 2H 2 ; H 2 : CO = 1 (1.16) C 2 H 6 + 2CO 2 = 4CO + 3H 2 ; H 2 : CO = 0:75 (1.17) C 2 H 4 + 2CO 2 = 4CO + 2H 2 ; H 2 : CO = 0:5 (1.18) Water-gas shift reaction: CO + H 2 O = CO 2 + H 2 (1.19) In summary, it is proposed that the major pathways for the production of H 2 over the OCM catalysts are the dehydrogenation of ethane, steam reforming and partial oxidation of ethane and ethene which accompany simultaneous production of CO. 1.4 Heterogeneous Catalysis and Catalyst A catalyst is a substance which controls the rate of reaction without itself undergoing a permanent chemical change and the process is called catalysis. The modern concept of catalysis was defined by Berzelius as follows: Catalysis is a process whereby a reaction occurs faster than the uncatalyzed reaction, the reaction being accelerated by the presence of a catalyst (Krische, 2005; Somorjai and McCrea, 2000; Haller, 2003; and Ponec, 1998). There are two types catalyst, which are positive catalyst and negative catalyst or inhibitor. A positive catalyst increases the rate of reaction by lowering the energy of activation. Thus, in the presence of a positive catalyst, the greater fraction of the total molecule will posses lower energy of activation and collided successfully in a

46 24 short period of time, there by increasing the rate of reaction. A positive catalyst functions by providing an alternate path to the reaction or by the formation of a transition (intermediate) compound having low energy of activation as shown in Figure 1.8. The activation energy of this path is lower. As a result the rate of reaction is increased. A negative catalyst retards the rate of reaction. A negative catalyst does not lower the energy of activation; rather it is combined with reactant molecule thus decreasing the number of colliding reactant molecules. This decreases the effective collisions, hence the rate of reaction (Fogler, 1999). Catalysis can be classified as homogeneous catalysis, in which only one phase is involved, and heterogeneous catalysis, in which the reaction occurs at or near an interface between phases (Bond, 2000; and Chen et al., 2005a). The choice for a certain catalyst for a catalytic conversion is not only determined by the reaction itself but also by the industrial process (Taguchi and Schuth, 2005). For example, heterogeneous catalysts are used for the production of bulk chemicals because the solid catalyst materials are unmixable with products. This facilitates the separation of products and catalyst material, especially if gaseous products are involved. Consequently, the reaction can be performed under continuous flow conditions which permits the scaling up of production processes to achieve high rates. This advantage of heterogeneous catalysis is of exceptional importance because it allows the production of gasoline, fuel oil and other bulk chemicals on a large scale which is essential to provide sufficient bulk chemicals to satisfy the vast demands of the world market (Holzwarth et al., 2001; Blaser, 2000; and Kerby et al., 2005). Because most of the work presented in this thesis focuses on zeolite material and MgO based OCM catalyst, a brief description of these catalysts is presented in the next sections.

47 25 Figure 1.8 A catalyst provides an alternative route for the reaction with a lower activation energy (Somorjai and McCrea, 2000) Zeolite History and Zeolite Development The word Zeolite was first used by a Swedish mineralogist Axel Cronstedt. In 1756, he discovered the very first zeolite mineral stilbite. This mineral visibly lost water when heated. Accordingly, he named this class of mineral zeolite from the classical Greek words zeo, meaning to boil, and lithos, meaning stone. Since then, about 40 different kinds of natural zeolites have been identified (Erdem et al., 2004). They vary in crystal structure and chemical composition, and some physical properties as well. Today, most natural zeolites have been successfully synthesized in laboratories or in commercial plants. Besides natural zeolites, man-made or synthetic zeolites have also added many new members to the zeolite family. Scientists have synthesized many zeolites that do not exist naturally or at least have not been discovered on the earth (Casci, 2005).

48 26 Zeolite is defined by Smith (1963) as an aluminosilicate with a framework structure enclosing cavities occupied by large ions and water molecules, both of which have considerable freedom of movement, permitting ion-exchange and reversible dehydration (Murakami et al., 1986). Other definitions can also be found. Jacobs (1977) gave the following definition: Zeolites as synthesized or formed in nature are crystalline, hydrated alumino-silicates of group I and II elements. Structurally, they comprise a framework based on an infinitely extending threedimensional network of SiO 4 and AlO 4 tetrahedra linked together through common oxygen atoms (Jeanette, 1980). Szostak (1989) stated that structurally, zeolite is a crystalline aluminosilicate with a framework based on an extensive threedimensional network of oxygen ions (Ribeiro et al., 1995). Recently, Corma (2003) defined zeolite as crystalline silicalites and aluminosilicates linked through oxygen atoms, producing a three-dimensional network containing channels and cavities of molecular dimensions. Although the four definitions are stated differently, the main characteristics of a zeolite are either clearly expressed or implied as a crystalline material of aluminosilicate featured by a three-dimensional microporous framework structure built of the primary SiO 4 and AlO 4 tetrahedra, and ion-exchange capability. Nowadays, zeolites are available on a large scale and in a variety of applications. The major use of zeolites is as ion exchangers in laundry detergents where they remove calcium and magnesium from water by exchanging it for sodium present in the zeolite. Furthermore, zeolites are applied as adsorbents in the purification of gas streams to remove water and volatile organic species, and in the separation of different isomers and gas-mixtures. Moreover they are applied in the clean up of radioactive wastes. However, in this thesis the focus will entirely be on the application of zeolites as catalysts for the conversion of hydrocarbons. Catalysis by zeolites with focus on hydrocarbon conversion and formation covers, nowadays, a broad range of processes related to the upgrading of crude oil and natural gas. This includes, among others, fluid catalytic cracking (FCC), hydrocracking, dewaxing, aliphate alkylation, isomerisation, oligomerisation, transformation of aromatics, transalkylation, hydrodecyclisation as well as the conversion of methanol to hydrocarbons (Stocker, 2005). All these conversions are catalyzed by zeolites or related microporous materials, based both on the acid properties and shape-selective behavior of this type of materials.

49 27 The ordered, crystalline structures imply that the micropores in zeolites have very well-defined dimensions and connectivities. In addition, the framework composition can be tuned to enhance chemical selectivity. These features combine to ensure that zeolites form a very important category of cheap, highly reproducible size and shape selective molecular sieves and adsorbents which find widespread use in industry (Weitkamp, 2000). Moreover, as extensively used porous materials, zeolites show marked advantages over other solid materials (Stocker, 2005; and Corma, 2003): Well defined structure which could be clearly related to the activity and selectivity; Well defined inner pores in which active species can be hosted; Adjustable frame-work composition and cations associated with different stability; hydrophilicity/hydrophobicity and acid-based properties A large amount of structures that can be chosen as shape-selective catalysts for different reactions Structure of Zeolites The elementary building units of zeolites are SiO 4 and AlO 4 tetrahedra. Adjacent tetrahedra is linked at their corners via a common oxygen atom, and this results in an inorganic macromolecule with a structurally distinct three-dimensional framework. It is evident from this building principle that the net formulae of the tetrahedra are SiO 2 and AlO - 2, i.e. one negative charge resides at each tetrahedron in the framework which has aluminum in its center (Ali et al., 2003). The framework of a zeolite contains channels, channel intersections and/or cages with dimensions from ca. 0.2 to 1 nm. So, aluminosilicates have negatively charged oxide frameworks (one charge per framework Al 3+ ) which require charge-balancing by cations. Typical cations include the alkaline (Li +, Na +, K + ) and the alkaline earth (Mg 2+, Ca 2+, Ba 2+ ) ions, quaternary ammonium ions, protons (in the acid form of zeolites), and the rare

50 28 earth and noble metal ions. The cations are quite mobile and may be exchanged by other cations, which determine the ion-exchange properties of the zeolites (Nowinska et al., 2003). Zeolites are crystalline networks of alumino-silicate, where the structural formula of zeolites is best expressed for the crystallographic unit cell as (Hashimoto, 2003): M x/n [ (AlO - 2 ) x (SiO 2 ) y ]. wh 2 O (1.20) Where M is the cation of valence n, the portion with [ ] represents the framework composition. The sum (x + y) is the total number of tetrahedral in the unit cell. The ratio x/y must be smaller than or equal to 1, since AlO - 2 tetrahedral can only join to SiO 2 tetrahedra according to the Lowenstein rule. w is the number of water molecules per unit cell. The water in many zeolites can be reversibly removed by calcinations, leaving an open host structure. Structurally, zeolites are built of primary and secondary building units. The primary unit is simply the SiO4 or AlO4 tetrahedron. Si or Al atom sits at the center of the tetrahedron with 4 oxygen atoms co-valently bonded to the centered Si or Al atom (so-called T-atom). From this primary unit, a number of secondary building units can be built by a linkage through the oxygen atom covalent bonding, which is called an oxygen bridge (Weitkamp, 2000).The secondary building units are featured by simple geometric shapes such as those shown in Figure 1.9. A zeolite structure is finally constructed from the secondary units. A pore channel system is formed during the systematic packing of the secondary units (Ribeiro et al., 1995). A schematic representation of the structuring process of a zeolite is shown in Figure 1.9. Furthermore, zeolites were traditionally used as highly selective adsorbents, ion exchangers and catalysts with high activity and selectivity in a wide range of reactions. The unique properties of zeolites arise from their uniformity in pore size and the chemical composition. The pore size of zeolites was determined by the number of T atoms in the ring openings, where T is Si 4+ or Al 3+. Zeolites with pores comprising of 8, 10, and 12 T atoms have historically been classified as small (8- member ring), medium (10-member ring), and large-pore zeolites (12-member ring)

51 29 respectively (Weitkamp, 2000).These pores may extend in one-dimensional tunnels, two-dimensional intersecting cages and, finally, three-dimensional connected channel system of nanometer scale networks. This microporous pore system renders zeolites high surface area and allows zeolites to recognize, discriminate, and organize molecules with precisions that can be less than 1 Å (Roldan et al., 2005). The molecular shape selectivity of zeolites results from the interaction of adsorbed molecules with zeolite channels or cages of atomic size. In addition, the characteristics of some typical zeolites are listed in Table 1.2. Figure 1.9 Schematic representation of building zeolites (Rabo, 1976)

52 30 Table 1.2: Characteristics of some typical zeolite porous material (Weitkamp, 1998) Zeolite Number of rings Pore size (A) Pore/Channel Structure 8-membered oxygen ring Erionite x 5.1 Intersecting 10-membered oxygen ring ZSM x 5.5 Intersecting 5.3 x 5.6 ZSM x 5.4 Intersecting Dual pore system Ferrierite 10,8 4.2 x 5.4 One dimensional 3.5 x : 8 intersecting Mordenite x 7.0 One dimensional x : 8 intersecting 12-membered oxygen ring ZSM x 5.9 One dimensional Faujasite Intersecting 7.4 x : 12 intersecting Mesoporous system VPI One dimensional MCM41-S One dimensional Mechanisms of the Shape Selectivity In particular, zeolites usually show specific selectivity in some reactions due to their microporous properties. Because the pores of the zeolites are similar to many organic molecules of practical interest, it became possible to design novel catalysts which control the progress of the reactants and products through them via selecting the molecules by their size and shape, so called shape selective catalysts (Ramaswamy, 2003). It is generally believed that the majority of the active sites is located in the pores of zeolite (Corma, 2003).

53 31 Typically, five fundamental steps are involved in an overall reaction using zeolite as a catalyst : (i) diffusion of molecules from gas phase into the pores of the zeolites; (ii) adsorption of molecules on active sites; (iii) conversion of reactants molecules into product molecules; (iv) desorption of product molecules from active sites and eventually; (v) diffusion of product molecules into gas phase from the pores of the zeolite. The shape selectivity of the zeolite for a specific reaction may be possibly involved in step (i), (iii) or / and (v) (Krishna and Baur, 2003). Based on different controlling steps, the shape selectivity of zeolites usually observed can be classified into three different types, including reactant shape selectivity, product shape selectivity and transition state selectivity, which are described in Figure Reactant shape selectivity Product shape selectivity Transition state selectivity Figure 1.10 Schematic diagram of shape selectivity of zeolite (Krishna and Baur, 2003). Reactant shape selectivity can be observed in the case of different participating reactant molecules with smaller and bulkier diameters compared to the pore entrance of zeolite. Reactants with smaller kinetic diameters can penetrate into

54 32 the interior pores of zeolite and access the active sites located in the pores of the zeolite. Reactants with larger diameter, however, cannot diffuse into the pores of zeolite and the reaction with bulky reactant molecules involved can be encumbered. Product shape selectivity occurs when some of the reaction products cannot escape from the zeolite due to their size or shape. Then these molecules can undergo a secondary reaction to form molecules that are able to leave the catalyst. Transition state selectivity is observed when both the reactant and product molecules fit well within the pores of the zeolite, while the reaction intermediates are larger than them and are spatially constrained either by their size or by orientation ZSM-5 Zeolite Catalyst The type of zeolite catalyst used in this project is called Mobil Synthetic Zeolite-5 (ZSM-5), which was developed by Argauer and Landolt in 1972 (Gayubo et al., 2003). Since its inception, ZSM-5 has been widely used for hydrocarbon inter conversion in the petroleum industry. ZSM-5 zeolites in protonic type (H-ZSM-5) have been extensively used in acid catalyzed reactions. H-ZSM-5 zeolites can be synthesized with a broad range of Si-Al ratio from 6 to infinity. In principle, the acid strength and acid types are the key properties of zeolites, which play a crucible role in the activity and selectivity of the zeolites. The range of zeolite-catalyzed reactions can be extended by the incorporation of metal atoms both inside and outside the zeolite framework (Goursot, et al. 2003). The topology of the zeolite framework is given by a unique three-letter code, which is not related to the composition of the material. Generally, based on the pore openings, zeolites are referred to as small (8-member ring), medium (10-member ring) and large (12- member ring) pore zeolites. Thus, ZSM-5 is a medium pore zeolite materials with Mobil Five (MFI) topology (Rabo, 1976). The MFI structure is built up by 5-1 secondary building units ( SBU; the smallest number of TO 4 units, where T is Si or Al, from which zeolite topology is built) which are linked together to form a chain (Figure 1.11) and the interconnection of these chains leads to the

55 33 formation of the channel system in the structure (Klinowski and Barrie, 1988). The MFI structure has a three-dimensional pore system consisting of sinusoidal 10-ring channels (5.1 x 5.5 Å) and intersecting straight 10-ring channels (5.3 x 5.6 Å) and these two different channels are perpendicular to each other and generate intersections with diameters of 8.9 Å (Hashimoto, 2003). For the Methanol to Gasoline (MTG) process, it is the pores created by these 10-oxygen rings, along with the zig-zag pores intersecting them that are essential to the formation of products that are desirable components of gasoline (Grobet, 1988). An 8-oxygen ring zeolite will not produce molecules with 6 or more carbons as molecules of this size will not fit into the small pores of these zeolites. The large pores of a 12-oxygen ring zeolite produce large amounts of C-11 and C-12 compounds, which are undesirable products for gasoline range hydrocarbons (Rabo, 1976). On the other hand, most of ZSM-5 catalysts are used in oil refining and gasconversion processes, such as the conversion and upgrading of the various fractions into transportation fuels, methanol-to-gasoline, conversion of syngas, light paraffins and olefins into gasoline and gasoil (Kerby et al., 2005). A huge interest in ZSM-5 is due to its unique properties that have provided them a vast niche in the industry : High thermal and hydrothermal stability in the industrial environment, especially highly siliceous zeolites High internal surface area Presence of ion exchange ability allows the formation of highly dispersed catalytically active sites, such as highly acidic sites Pores structure provides shape selectivity (which has been shown to limit chain length growth to gasoline range hydrocarbons in methane conversion) High acidity promotes the oligomerisation, isomerisation, cracking and aromatisation reactions The ZSM-5 type zeolite, which has all these properties, is the catalyst for this work due to its importance in the oligomerization processes.

56 34 Figure 1.11 Illustration of primary unit of TO 4 tetrahedral, 5-1 secondary building units, the MFI framework and channel structure of ZSM-5 zeolite (Klinowski and Barrie, 1988)

57 Magnesium Oxide-based Catalysts for OCM Reactions Metal oxides act as important catalysts in the reaction of oxidative coupling of methane (OCM). The catalytic reaction of OCM over metal oxide surfaces has been intensively investigated over the past 20 years (Lintuluoto and Nakamura, 2004). There are still ongoing studies to find good catalysts to achieve more active and especially higher selectivity toward C 2 hydrocarbon (especially ethylene). Among the solids used in OCM as the main components or supports of numerous active and selective oxide catalysts, magnesium oxide occupies a special position. Catalysts based on MgO promoted with alkali oxides, alkaline earth oxides, rare earth oxides, etc., have been frequently used in the research. It is widely accepted that the increase in surface basicity caused by a promoter enhances the activity and C 2 selectivity (the selectivity to C 2 H 4 and C 2 H 6 ) of MgO-based catalysts (Maitra, 1993). Only limited attention had been given to the activity of unpromoted MgO itself in the OCM process and to the role of MgO as a carrier in MgO-based catalysts. As noticed by some authors, the samples of MgO differing in origin and properties usually also varied in their catalytic performance. This phenomenon was usually related to the differences in morphology, the specific surface area, surface structure and properties, a type of precursor, the calcination time and temperature, etc. (Burrows et al., 1998; Lacombe et al., 1995; Kus et al., 2003; Kus et al., 2002b, Kus and Taniewski, 2002; and Maksimov et al., 1998). Burrows et al. (1998) have pursued the structure sensitivity aspect of OCM catalyst design for MgO and doped MgO catalysts for a number of years. A perfect structure of MgO as a cubic (rocksalt) crystal is illustrated in Figure However, under certain conditions or situations (such as addition of lithium), magnesium oxide has a highly defective surface structure showing steps, kinks, corners, and others which provide O 2- sites of low coordination. These low co-ordinated O 2- sites (O 2- on faces, O 2- on edges and O 2- on corners) are expected to be responsible for the strength of the basic sites (Pinarello, et al., 2001).

58 36 Figure 1.12 Cubic (rocksalt) MgO crystal: different planes (Hermann, 2002) By comparing the microstructure and performance of pure MgO catalysts prepared by a number of very different routes, Burrows et al. (1998) have demonstrated that the selectivity and specific catalytic activity are unaffected by the number of corner or edge sites that the MgO exposes. That suggested that vacancies on the planar {100}-type surfaces of MgO are likely to be active sites for methane activation. They have also shown that the addition of lithium carbonate to MgO results in subtle changes in the surface structure. The improved catalytic performance of Li doped MgO was associated with, amongst other effects, the presence of dislocations creating active sites on the MgO surface. The MgO properties/performance relationship still remains a controversial question. Although intensive investigations have been carried out, the nature of active sites formed in these oxides at high temperature are still a matter of debate. However, it is clear that the lattice defects of doped oxides play an important role in the reaction mechanism (Balint and Aika, 2001). This is not surprising since the lattice defects of a large number of oxide systems have been reported to play an important role in molecular activation at high temperatures. A part of the surface

59 37 oxygen becomes active at high temperatures, reacting with methane even on irreducible oxides such as MgO. The special oxygen structure such as O -, has been proposed as the active species. If O - is the active species, O - or the surface structure accepting O - must be generated only at high temperature on MgO catalytic systems. Thus, it can be seen that the OCM reaction occurs on MgO only at high temperatures, usually K. Accordingly, Karasuda and Aika (1997) concluded that the OCM reaction occurs through such an activation step as is shown in Figure 1.13 when applying MgO based catalyst. Figure 1.13 Model of oxygen atom flux in MgO-based catalysts where the fast exchange with an a-group O (O a ) in the MgO surface and the slow exchange with a b-group O (O b ) in MgO lattice ( Karasuda and Aika, 1997) Karasuda and Aika (1997) also focused their attention on MgO as a representative irreducible oxide for the OCM catalyst and have attempted to clarify the production mechanism of the active site. The authors have reported that the conductivity of MgO and Li/MgO is related to the defect structure produced by the decomposed impurity H 2 O dissolved in the MgO lattice. The electric conduction is considered due to proton jumping at low temperatures, while hole conductivity prevails at high temperatures, where H 2 O is dissolved in MgO (step 1 in Figure 1.14)

60 38 and the hole (O 2- + h + = O - ) is produced through H 2 evolution (step 2 in Figure 1.14). During the OCM reaction, O - (or hole + O 2- ) reacts with methane to produce methyl radical and OH - (step 3 in Figure 1.14). Two methyl radicals combine to form ethane (and then ethene). The produced OH - is removed as H 2 O leaving the defect (step 4 in Figure 1.14) and dioxygen reproduces O - again (step 5 in Figure 1.14). If methane is not present but dioxygen is at the OCM temperature (step 5 in Figure 1.14), then the reverse reaction (step 6 in Figure 1.14) can be observed through the isotopic exchange. Thus, the isotopic exchange between dioxygen and MgO provides significant information on the active center of the OCM reaction. Figure 1.14 Model of active center generation and the OCM reaction model (Karasuda and Aika, 1997) Currently, rare-earth-metal (REM) oxides, like La 2 O 3, are among the most active catalysts for OCM, and a number of experimental studies centered around the mechanistic features of their chemistry have been reported. In the same way, Choudhary et al. (2000b) and Traykova et al. (1998) suggested that La-promoted MgO catalysts showed high activity and selectivity in the OCM process at a very high space velocity and also high catalyst stability in the process. The promotion effect of lanthanum was attributed to a large increase in the strong basicity of the catalyst because of the doping of MgO by lanthanum. Lacombe et al., (1994)

61 39 suggested that surface intermediates such as methoxide ions M-O-CH 3 are formed on the lanthanum oxide catalysts in the course of the OCM reaction. The formation of these species would be related to the dehydroxylation of terminal La-OH groups leading to the formation of a complex site {O 2-, vacancy}; this site would allow the adsorption of a methyl radical colliding with the surface. Once stabilized, the La-O- CH 3 species would undergo a further oxidation. Thus, specific hydroxyl groups from the lanthana surface could act as methyl radical scavengers. Because of its high activity/selectivity and productivity (i.e. operation at high space velocity or low contact time), and also high thermal stability, the La-promoted MgO catalyst has a high potential for use in the OCM process. 1.4 The Statement of Problem Efficient utilization of methane, the primary component of natural gas, is becoming an immediate goal as we face issues of diminishing natural resources. Unfortunately, the remote locations of most natural gas reserves make transportation of this fuel economically unreasonable. One means to exploit the remote gas sources is to first convert methane to more useful higher carbon products that can easily be transported to the consuming facility. However, existing methane conversion methods have proven to be prohibitively expensive. Thus, developing a less expensive catalytic conversion method is of great interest and technologically significant. In particular, the direct catalytic conversion of methane into desired chemicals or liquid fuels is still a great challenge for the utilization of natural gas in the field of catalysis. The two approaches involved are direct conversion of methane with the assistance of oxidants and direct conversion of methane under nonoxidative conditions. For example, the heterogeneously catalyzed oxidative coupling of methane (OCM) seems to be one of the most promising routes to convert methane directly from the abundant natural gas into ethylene which is a feed-stock for chemical industry and/or by oligomerization of ethylene into transportation fuels. A wide variety of catalysts, mainly metal oxide based catalysts, have been studied and

62 40 reaction mechanisms have been proposed. Although significant progress has been achieved in the research and development of oxidative coupling of methane (OCM), it is still difficult to overcome the barriers of technical and economic problems. Two factors handicap commercialization of this process: (1) C 2+ selectivity and yield are too low and (2) technological novelty results in uncertainty for scale-up. Higher C 2+ selectivities and yields have to be achieved mainly by development of more selective catalysts, but optimization of reaction conditions, reactor design and operation also offers possibilities to improve catalytic performance. The technological uncertainty is due to the fact that all results available until now have been achieved in laboratory-scale units, mainly in microcatalytic fixed-bed and small-scale fluidized-bed reactors. Due to the complexity of the OCM reaction scale-up can be associated with a loss in catalytic performance. Moreover, the severe reaction conditions, i.e. high temperature and exothermicity require new reactor designs. Therefore, experience from other selective oxidation reactions can be used only to a limited extent. Reactor development includes not only the choice of the reactor type but also the selection of materials and of the method of temperature control. Against this background, reaction engineering aspects of the OCM reaction play an important and crucial part in the attempt to make this process commercially viable. In addition, one of the serious limitations of the OCM process is the low concentration of ethylene in the product stream. Separation of ethylene at a low concentration is not at all economical. One way to overcome this limitation of the OCM process is to convert the dilute ethylene present in the OCM product streams directly into much less volatile product streams, such as aromatics and/or gasoline range hydrocarbons, which can be easily separated. On the other hand, although OCM results in the production of hydrocarbons, this product is characterized by the lack of selectivity for a particular hydrocarbon type or carbon chain length. Hence, there is a great need for a highly active and selective ethylene transformation catalyst useful for converting ethylene at very low concentrations or partial pressures into liquid hydrocarbons. Consequently, a second catalyst becomes essential if selectivity for a particular product is required.

63 41 During the early days, H-ZSM-5 zeolite, which is one of the synthetic zeolite, was found to be a suitable catalyst for the conversion of methane to higher hydrocarbons. Later, ZSM-5 was modified to increase the conversion of methane and the selectivity of targeted hydrocarbons products. But, this catalytic process has not reached the commercializing stage and it needs development to obtain a more promising result. However, some earlier studies have demonstrated that acidic H-ZSM-5 zeolite catalyst has shown a reasonably good oligomerization performance for olefin products to higher hydrocarbons. Taking advantages of both the exclusive functionality of OCM catalyst and good oligomerization activity of H-ZSM-5 catalyst, we propose here the dual catalyst bed concept. In the present study, we have applied these catalysts to a serial dual-bed system as an alternative methane conversion route to liquid hydrocarbons in the presence of oxygen. The development of a combination of an OCM catalysts and H- ZSM-5 would overcome the inherent limitations and permit the direct conversion of methane to liquid hydrocarbons via a single reactor process. Therefore, a reactor system with such a function would be the focus of in this research. 1.5 Hypothesis Taking advantages of both the exclusive functionality of OCM catalyst and good oligomerization activity of H-ZSM-5 catalyst, we propose here the dual catalyst bed concept. In the present study, we have applied these catalysts to a serial dual-bed system as an alternative methane conversion route to liquid hydrocarbons in the presence of oxygen. The development of a combination of an OCM catalysts and H- ZSM-5 would overcome the inherent limitations and permit the direct conversion of methane to liquid hydrocarbons in order to solve the mystery of directly converting methane to liquid hydrocarbons via a single step reactor process. Therefore, a reactor system with such a function would be the focus of in this research. The first catalyst comprises rare earth and alkaline earth metal oxides that are good materials for producing good-performance OCM catalysts and methane

64 42 conversion. Rare earth oxides have been thought to be good catalysts for the OCM reaction, and alkaline earth oxides have been found to be good promoting materials for rare earth oxides. Therefore, the first catalyst is preferably lanthanum promoted MgO catalyst that is believed to give better C 2 selectivity. The second catalyst comprises an intermediate pore size zeolite, HZSM-5 which further provides shape selectivity (which has been shown to limit chain length growth to gasoline range hydrocarbons) and its high acidity is effective to promote the oligomerization, isomerization and aromatization reactions involved in the restructuring of the primary OCM products. Apart from that, recent research showed that that metal loaded HZSM-5 catalysts are suggested to be potential catalysts for direct conversion of methane to liquid hydrocarbons. Therefore, metal-containing HZSM-5 catalysts can produce C 5+ liquids from OCM intermediate products if dehydrogenation and oligomerization functions of the metal are in balance. Hence, performance of the HZSM-5 catalysts with combination metals of (Ni, Mo and Cu) is tested through this research. Furthermore, the process of this study is directed to converting high percentages of the methane present in mixed methane-oxygen containing feeds to OCM products at the first stage, while simultaneously converting mainly olefins to gasoline range liquid hydrocarbons at the second stage. In this way, the complexity of the separation processes needed in the ethylene recovery process also could be minimized. 1.6 Research Objectives The main objectives of this research are: i) To develop a lab scale dual-bed catalytic reactor system that is capable of catalyzing the production of liquid hydrocarbons from methane ii) To elucidate the effect of reaction parameter on the dual- bed system

65 43 iii) To modify H-ZSM-5 with different kinds of metal and investigate the performance of this modified H-ZSM-5 catalysts on the products distribution as a function of types of metal. iv) To find the optimum operating conditions required for maximum catalyst activity 1.7 Scope and Structure of this Thesis This thesis focuses on the experimental study of a dual bed catalytic system over La-MgO and H-ZSM-5 based catalysts in accordance with the purposes of the present invention as embodied and broadly described herein, to produce liquid hydrocarbons from the methane gases. Thus, in this work, the possibility of liquid hydrocarbon production by means of the dual-bed system was investigated, and the effects of some reaction conditions were also explored. The focus is also given on the modification of the second bed catalyst of H-ZSM-5 by introducing some metal via incipient impregnation method in order to improve the yield of targeted products. Furthermore, it also involved the characterization of some catalysts in order to determine catalyst behavior. In addition, we report for the sequential optimization strategy for liquid hydrocarbons production through statistically designed experiments as an effective tool for medium engineering. The thesis is divided into six chapters. Basically, an introduction with a short overview of the literature data on the particular topic is given in each chapter. In Chapter 1 the literature review which is related to the subject of this thesis has been reviewed briefly. The background regarding the properties of the zeolites and OCM catalysts are included in Chapter 1 as well as some theory on aspects of this research. The relevance of heterogeneous catalysis and the necessity to understand catalytic processes on microscopic scale are also accentuated. Furthermore, the fundamental concepts of catalytic conversion of methane is offered in this chapter. In addition, the applicability of our experimental approach to solve problems in applied catalysis is discussed. The objective of the research has also been elaborated in this section.

66 44 Chapter 2 outlines the experimental apparatus used and the experimental procedures. This chapter provides a brief description of the catalyst preparation procedures using conventional impregnation method. Moreover, the additional sections of this chapter focus on the general experimental procedures used throughout this work and the analytical techniques employed to interpret catalytic activity of all the catalysts. This chapter also describes the basic concept and the detailed procedures of the characterization method employed in this study. Chapter 3 deals with the structural characterization and study the influence of the acidic behavior of the H-ZSM-5 catalyst in explaining the catalytic reaction performance of the dual bed system. Moreover, the influence of temperature and the effect of oxygen concentration on the activity and products selectivity as well as the result of the use of different SiO 2 /Al 2 O 3 ratio zeolite matrix were studied. A systematic approach is undertaken to study the influence of SiO 2 /Al 2 O 3 ratio and thermal treatment on the acidity and its nature, and the implication thereof for the intermediate OCM products conversion. The catalysts are characterized by TPDammonia techniques and the observed differences are correlated with the structural properties of the respective materials and catalytic activity. Chapter 4 deals with the same system as Chapter 3 but focuses mainly on the role of metallic sites in order to obtain relations between the structural properties and the catalytic activity of the modified H-ZSM-5 catalysts in the conversion of methane through the dual bed strategy. The experimental phase was carried out in a dual-bed reactor in order to screen the catalysts and to identify the best one among the eleven metal-based H-ZSM-5 catalysts. The screening tests were conducted at fixed temperature of C and CH 4 /O 2 ratio of 10. In this chapter, some of the important matter are addressed in explaining the advanced performance of metal based H-ZSM-5 catalysts in this dual bed system, compared with unloaded H-ZSM-5 when applied as second bed catalysts, under fixed other reaction parameters. Moreover, the simplified reaction scheme to describe the reaction path in this dual bed system proposed by employing the recent products spectra in light of the previous literature is also reported.

67 45 Chapter 5 describes a study to develop an approach that would enable us to better understand the relationships between the variables (temperature, oxygen concentration and amount of Ni loading (wt%) and the response (yield of C 5+ hydrocarbons); and to obtain the optimum conditions for liquid hydrocarbons production using central composite design (CCD) and response surface methodology (RSM). The CCD has the advantage to predict responses based on a few sets of experimental data in which all factors are varied within a chosen range. The range of each variables was chosen based on the finding from one-factor at a time techniques. An empirical model including the effects of independent variables has also been developed through utilized software to represent the response surface. Also, TPD ammonia and NA characterization analysis were performed on the Ni loaded H- ZSM-5 catalysts to explicate the effect of modification of Ni on the catalytic reaction through the proposed system. Finally, the results and findings of this thesis are summarized in Chapter 6. The future work possibilities are also suggested in this chapter.

68 CHAPTER 2 EXPERIMENTAL APPROACH 2.1 Introduction In this chapter, the experimental techniques employed for the preparation, characterization and testing of the catalysts through this research are summarized and descriptions of rig setup are also presented. As the most important experimental systems used for the studies reported in Chapters 3-5 were the same, the experimental sections of the following chapters will therefore only contain information that is either specific for the experiments or deviate from the general procedure. The flow chart of the experimental work is shown in Figure 2.1. The first stage of this experimental work was preparation of catalysts. Two types of catalyst were prepared namely La-MgO and zeolite-based H-ZSM-5 catalyst. Afterwards, both prepared catalysts were put in series through dual bed catalysts strategy to evaluate the performance on this approach for direct conversion of methane to liquid hydrocarbons. Furthermore, the characterization of the H-ZSM-5 zeolite materials using a series of NA and TPD-NH 3 techniques was undertaken in order to provide some information about their composition, properties and their behavior on the activity and product selectivity in the dual bed system reaction. Then, the optimization of reaction parameters was carried out using response surface methodology (RSM) based on central composite design (CCD). Finally, conclusions were drawn based on the findings obtained from this study.

69 47 METHODOLOGY CATALYST PREPARATION La-MgO CATALYST H-ZSM-5 CATALYST DUAL BED CATALYTIC SYSTEM CATALYTIC EVOLUTION CATALYST CHARACTERIZATION OPTIMIZATION CONCLUSION Figure 2.1 Flow chart of overall experimental work

70 Catalysts Preparation Zeolite Figure 2.2 Schematic diagram of ammonium ions replaced by hydrogen ions (Amin and Anggoro, 2002) Zeolite H-ZSM-5 with different SiO 2 /Al 2 O 3 ratios was purchased from Zeolytst TM International Inc. It is in powdered form. Zeolite ZSM-5 samples were in ammonium form, so they were converted into H-form by calcining at C for 5 hours before any modification and catalytic evaluation were performed (Zhu et al., 2005). A schematic diagram of the phenomena is depicted in Figure 2.2. The properties of the zeolite samples are listed in Table 2.1. Table 2.1: Zeolite H-ZSM-5 and their manufacture assay data Zeolyst TM Product SiO 2 /Al 2 O 3 Mole ratio Nominal Cation Form Content of Na 2 O Surface Area (m 2 /g) (% wt) CBV 3020E 30 Ammonium CBV 5524G 50 Ammonium CBV Ammonium CBV Ammonium

71 Preparation of Metal/H-ZSM-5-based Catalysts Different metal/h-zsm-5 catalysts were prepared starting from decomposition of the ammonium form of ZSM-5 (Zeolyst, CBV 3020E) with a SiO 2 /Al 2 O 3 ratio of 30. These resulting materials were then further used for impregnation with an appropriate metal salts. The metal salts which used to modify the H-ZSM-5 in this study, are summarized in Table 2.2. H-ZSM-5(30) catalysts with the metal loading 1 wt% were prepared according to the incipient wetness method, followed by drying at C overnight and then was calcined at C for 4 hours (Yin et al., 2005). The freshly prepared metal loaded H-ZSM-5 samples were then crushed and sieved to 20/45 mesh sizes for catalytic evaluation. These H-ZSM-5 zeolites are expressed as M/H-ZSM-5 (M: metal cations). This procedure, summarized in Figure 2.3, was followed when preparing all metal ion modified H-ZSM-5 catalysts in this study. Metal precursor solution Drying (110 0 C, 12 hrs) Calcination (500 0 C, 4 hrs) H-ZSM-5 Incipient wetness impregnation M/H-ZSM-5 Figure 2.3 Schematic illustration of the procedure for preparing metal containing H-ZSM Preparation of La/MgO Catalyst Magnesium oxide (MgO, supplied by GCE with purity >98%) was used as the support of the La/MgO catalyst. The La promoted MgO catalyst was prepared by mixing powdered magnesium oxide with lanthanum (III) nitrate hexahydrate (La(NO 3 ) 3.6H 2 O, which were supplied by Merck with purity %) (La/MgO=

72 ratio) in deionized water just sufficient to form a thick paste and drying at C for 12 hours. The resulting dried mass was then calcined at C for 10 hours in static air. Table 2.2: Metal salts and their manufacture assay data Name of Salt Molecular Formula Molecular Weight (g/mol) Purity (%) Supplier Chromium (III) Nitrate Nanohydrate Cr(NO 3 ) 3.9H 2 O R&M Chemicals Gallium (III) Nitrate Hydrate Ga(NO 3 ) 3.xH 2 O ACROS Organics Cadmium Carbonate CdCO Fluka Chemika Ammonium(meta) Tungstate Hydrate (NH 4 ) 6 W 12 O 40.xH 2 O Fluka Chemika Nickel(II) Nitrate Hexahydrate Ni(NO 3 ) 2.6H 2 O Fluka Chemika Silver Nitrate AgNO R&M Chemicals Cupric Nitrate Cu(NO 3 ) 2.3H EMORY Reagents Cobalt (II) Nitrate Co(NO 3 ) 2.6H 2 O MERCK Hexahydrate Ammonium (NH 4 ) 6 Mo 7 O 24.4H 2 O MERCK Molybdate Iron(III) nitrate Fe(NO 3 ) MERCK Nonahydrate Mangan(II) Nitrate Mn(NO 3 ) MERCK

73 Catalysts Characterization The catalyst used in this research was characterized for texture and acidity. To elucidate the texture of a zeolite and its bound counterpart, BET surface area, micropore surface area, micropore volume and pore size distribution were analyzed using a BET machine. To express a solid acid in term of acid site density and acid strength, an ammonia Temperature Programmed Desorption (NH 3 -TPD) technique was used Temperature Programmed Desorption Ammonia Temperature-Programmed Desorption (TPD) is one of the most widely used and flexible techniques for characterization of site densities in solid acids due to the simplicity of the technique (Choi et al., 1996). Determining the quantity and strength of the acid sites on alumina, amorphous silica-alumina, and zeolites is crucial to understanding and predicting the performance of a catalyst. The catalytic activity depends on many factors, but the Brønsted-acid site density is usually one of the most crucial parameters (Borges et al., 2005). TPD of ammonia is frequently used for the characterization of zeolite acidity for the analysis to measure not only the number of acid sites but also its strength. The amount of acid sites can be determined from the amount of desorbed ammonia molecules adsorbed directly on the acid sites. The strength of the sites, on the other hand, can be measured based on the theoretical equation as the heat of ammonia adsorption. With these methods, therefore, the solid acidity is characterized precisely, and the obtained information is utilized for designing an active and selective zeolite catalyst. Its small molecular size allows ammonia to penetrate into all pores of the solid where larger molecules commonly found in cracking and hydrocracking reactions only have access to large micropores and mesopores. Also, ammonia is a very basic molecule which is capable of titrating weak acid sites which may not

74 52 contribute to the activity of catalysts. The strongly polar adsorbed ammonia is also capable of adsorbing additional ammonia from the gas phase (Pinto et al., 2005; and Lonyi and Valyon, 2001). Figure 2.4 Typical NH 3 -TPD profile Furthermore, Figure 2.4 shows a typical NH 3 -TPD spectrum. The weak acid site density and strong acid site density were determined from NH 3 -TPD profiles, for example, from the areas under weak acid peak and strong acid peak shown in Figure 2.4 (Landi et al., 2004). Total acid site density is then the summation of weak and strong acid site densities (Lonyi and Valyon, 2001). In this study, the measurements for this TPD- NH 3 analysis were carried out using a Micromeritics TPD/TPR 2900 Instrument. The catalyst utilized was 300 mg for each experiment. First, the sample was heated up to 773K under a flow of dry nitrogen and kept at that temperature for 1 hour in order to desorb all traces of desorbed water. The sample was then cooled down to ambient temperature. Next, the sample was saturated in ammonia gas flow for 30 minutes, and subsequently, loosely bound ammonia was removed by purging the catalyst with nitrogen stream at

75 53 a flow rate of 20 ml/min for 30 minutes, and finally cooled down to room temperature. The amount of desorbed ammonia was determined by a linear heating rate of approximately 15K/min from 393 K to 823 K while the evolved ammonia concentration was monitored using thermal conductivity and mass spectrometric detectors Zeolite Surface Area by Nitrogen Adsorption Gas adsorption is of major importance for the characterization of a wide range of porous materials. Of all the many gases and vapors which are readily available and could be used as adsorptives, nitrogen has remained universally preeminent (Stefaniak et al., 2003). With the aid of user-friendly commercial equipment and on-line data processing, it is now possible to use nitrogen adsorption at 77 K for both routine quality control and the investigation of new materials (Sing, 2001). In view of the importance of the technique, it is of interest to trace its historical development. Generally, nitrogen adsorption (NA) at boiling temperature (77 K) is a common characterization technique employed to examine the surface area, pore size distribution and/or shape. Adsorption is defined as the concentration of gas molecules near the surface of a solid material (Serrano et al., 2001). Basically, the adsorbed gas is denoted as adsorbate and the solid where adsorption takes place is called adsorbent. As solid is exposed to N 2, the Van der Waals force acts between the exposed surface of the solid and the gas molecules. This gas physisorption is a reversible phenomenon due to the weak bonds involved between gas molecules and the solid surface (Richardson et al., 1994). For the use of zeolite as molecular sieve material or catalyst, it is essential to characterize the pore structure system. The adsorption measurement can give direct information about the texture properties of catalyst including surface area, dimension of pore volume and the pore system by using different sizes of adsorbed molecules.

76 54 Molecules with kinetic diameters smaller than the pore opening can be adsorbed but larger molecules cannot be adsorbed (Stefaniak et al., 2003). During this gas physisorption, the gas molecules non-selectively fill up the solid surface layer by layer, comprising of external surface and internal surface. The formation of saturated monolayer on the solid surface enables the measurement of the entire surface area. The complete adsorption/desorption analysis is known as adsorption isotherm. Numerous kinds of calculation models have been derived to evaluate the adsorption isotherm obtained such as BET, BJH and Langmuir. In this study, the evaluations of the pore structure of the samples were measured by nitrogen adsorption-desorption isotherms using the conventional technique on a Micromeritics ASAP 2000 instrument at liquid nitrogen temperature (77K). Prior to the adsorption measurements, the samples (0.3 g) were outgassed under vacuum at 110 o C overnight in order to ensure a dry clean surface, free from any loosely bound adsorbed species. Surface area (S BET ) and average pore diameter (W A ) were calculated by means of the Brunauer-Emmet-Teller (BET) equation. The possible presence of microporosity in the samples was investigated by applying the the t-plot method, which provided the values of both micropore volume(v p ) and micropore surface (A m ). 2.4 Reaction Unit and Product Analysis Experimental Rig A schematic representation of the the experimental set-up apparatus is shown in Figure 2.5. This experimental set-up was design to investigate the performance of dual bed catalytic approach for conversion of methane liquid hydrocarbons in this study. The experimental rig consisted of three major parts: (1) reactant delivery system and gas flow system, (2) catalytic reactor system and (3) products separation and analysis system. A description and elaboration of each of these systems are given in the next section.

77 55 1. Methane Gas Tank 8. 3-way Mixing Valve 15. Condenser 2. Oxygen Gas Tank 9. Pressure Gauge 16. Gas Products 3. Nitrogen Gas Tank 10. Switching Valve 17. GC-TCD 4. Pressure Regulator 11. Safety Vent 18. Liquid Products 5. Needle Valve 12. Quartz Tube Reactor 19. GC-FID 6. Flow Controller 13. Catalyst 20. Computer 7. Check Valve 14. Furnace Figure 2.5 A schematic diagram of laboratory reaction units set up for catalyst testing

78 Reactant Delivery System The setup was composed of three gas stream lines (i) two reactants line (oxygen and methane) (ii) one blank line (nitrogen). The physical and chemical properties of the gases are shown in Table 2.3. The nitrogen gas in line 3 was used to activate the catalysts and to purge the system and also act as a carrier gas. A set of three pressure regulators, needle valves, check valves and one set of 3-way plug valve, switching valve and safety vent controlled the lines between the reactor and the gas tanks. All the tubing, tubing fittings, valves, and pressure gauges which were used to set up the rig were made from stainless steel SS 316 supplied by Kuala Lumpur Valve & Fitting Sdn. Bhd. The volumetric flow rate of each gas into the reactor was controlled using the individual Alicat volumetric flow controller with a flow rate range of ccm. These flow controllers were supplied by Front Instruments Sdn. Bhd. The total pressure of the reactant gases prior to the reactor was measured by a pressure gauge with a range of 0-4 atmg. Table 2.3: Physical and chemical properties of the reactant gases (Air Liquide Ltd., 2002) Gases Methane Nitrogen Oxygen Chemical Formula CH 4 N 2 O 2 Molecular weight Melting point (ºC) Boiling point (ºC) Relative density Solubility mg/l water Appearance/colour colourless colourless colourless Odour none none none Autoignition temperature (ºC) Flammability range (%vol in air) oxidizer

79 Reactor System The reactor system was entirely placed in a temperature-controlled oven so that the feed can be preheated and mixed well. The reaction was performed using a fixed-bed quartz reactor at atmospheric pressure. It was made of quartz tube with an inside diameter of 9 mm and a length of 350 mm. The temperatures profiles of catalyst bed were measured with a thermocouple. The thermocouple well was made of a stainless steel tube of 1/8 inch (3.2 mm) outside diameter. The dual bed catalysts reactor configurations used in this study are schematically shown in Figure 2.6. In this dual bed catalyst approach, the feed gas mixture first passed through the La-MgO catalyst bed and then to the H-ZSM-5 based catalyst bed. The catalyst inventory was about 1-3 grams, depending on the experimental requirement. The pelletized catalyst particles were placed in the reactor tube. The rest of the reactor space (the top and bottom sections) and between the catalyst were filled with a thin layer of quartz wool. A Carbolite tube furnace (Model: MTF 10/15/130) was used to heat up the reactor to the required operating temperature. The temperatures reported were those for the OCM bed; the zeolite upper bed (minimum) temperature was measured to be about 115 o C lower than the OCM bed. This difference of 115 o C remained constant over the most important range of OCM bed temperatures, from o C. Quartz Wool Quartz tube Furnace Reactant Gases Inlet FIRST CATALYST BED (La-MgO) SECOND CATALYST BED (HZSM-5) Final Products Figure 2.6 Schematic representation of the dual bed catalyst reactor arrangement

80 Product Separation System The product separation system consists of a U-shaped glass liquid collector tube and a water bath. The U-shaped glass liquid collector tube was fabricated at the Glass Laboratory, Faculty of Science, UTM. The water bath temperature was set to -10ºC and the liquid inside the bath was a mixture of 60 vol% tap water and 40 vol% inhibited ethylene glycol (acts as an antifreeze agent) Analytical Procedure The reaction products were separated into liquid and gas fraction through the condenser. The gas fraction was analyzed by an on-line Hewlett Packard Agilent 6890N GC system equipped with thermal conductivity detector (TCD) and 4 series column (UCW 982, DC 200, Porapak Q and Molecular Sieve 13A). The composition of the liquid fraction was determined with the same GC by manually injecting the liquid into the back inlet of the GC. The liquid was then separated by a capillary column and detected by a flame ionization detector (FID). The peaks detected by TCD were identified by matching their retention times with Scott gas standard (P/N: ). In the same way, the peaks detected by FID were identified by matching their retention times with Supelco PIONA (Paraffins, Iso-paraffins, Olefins, Naphtha, and Aromatics) standard (Cat No ). After the GC analysis was over, the retention times and peak areas of components in a sample were printed out for both TCD and FID. Based on the peak areas, the mixture composition can be calculated. Appendix A shows the chromatograms of both standards as well as the calculation method of the gas concentration using the GC. All CH 4 conversions, CO x and hydrocarbon yields and selectivities were calculated based on the carbon basis. The detailed calculation is shown in Appendix B. The Microsoft Excel programme is used to perform the calculation. Furthermore,

81 59 the conversions, selectivities and yield were then calculated using Equations 2.1, 2.2 and 2.3, respectively. Conversion (%) = Amount of Methane Reacted Amount of Methane Input X 100 (2.1) Selectivity (%) = a Carbon of desired product Carbon of methane reacted X 100 (2.2) Carbon of desired product Yield of desired product (%) = Carbon of Methane Input X 100 (2.3) General Testing Procedure The catalyst was first placed in the reactor and supported by quartz wools. Then the catalyst was preheated in situ at C in a flow of nitrogen for an hour to activate the catalyst and to remove the absorbed water. The feed gas mixture of required composition was fed into the reactor system at atmospheric pressure and the reactor was heated to the desired reaction temperature. The reactant streams were controlled by using the volumetric flow controllers. The product mixture during reaction was passed through the U-shaped glass liquid collector tube and a water bath to separate the gaseous and liquid products for analysis. The composition of the output gas stream was analyzed on-line by gas chromatography. The needle valve (refer to Figure 2.5) near the gas chromatography (GC) was adjusted carefully so that the outlet stream pressure does not exceed 0.2 atmg whereby it is still sufficient enough to flow into the GC. The compositions of the reaction products from the 40th minute to the 240th minute were determined every 40 minutes by GC. The detailed descriptions of the testing procedures for each reaction are reported in the

82 60 related chapters. Furthermore, the reaction conditions chosen are described in the experimental sections of corresponding chapters.

83 CHAPTER 3 ACIDITY VERSUS ACTIVITY IN THE DUAL BED CATALYTIC SYSTEM FOR DIRECT CONVERSION OF METHANE TO LIQUID HYDROCARBONS 3.1 Introduction The direct catalytic conversion of methane into desired chemicals or liquid fuels is still a great challenge for the utilization of natural gas in the field of catalysis due to the possibility of developing new processes of lower environment impacts and of lower cost. Many reviews and monographs have been published which analyze the fundamental aspects related to the oxidative activation and transformation of light alkanes over heterogenous catalysts. The general picture that can be drawn on the basis of the most important factors which have been examined in these reviews clearly shows that the problem of paraffin conversion and of selectivity to the desired products have to be solved within a complex framework of inter-related aspects. On the other hand, oxidative coupling of methane (OCM) is a futuristic process of great practical importance. However, one of the limitations of this process is a very low concentration of ethylene in the product stream making the ethylene separation uneconomical. Dilute ethylene streams are also obtained from several petroleum refining and hydrocarbon conversion processes. It is therefore of great practical interest to convert dilute ethylene without its separation from the products, which can be separated easily. This idea has been exploited through this study, which disclosed a two-step process for the conversion of methane or natural gas into

84 62 liquid range hydrocarbons (C 5+ ) by using a single reactor. In the first step, methane or natural gas is converted into ethylene and other lower olefins and in the second step the ethylene and other lower olefins in the product stream of the Step-I are directly converted into liquid hydrocarbons over a pentasil zeolite catalyst. The focus is given on this type of approach for overcoming the limitation of the OCM process through this study. Therefore, the main objective of this work was to investigate the performance of the dual-bed catalytic process for the production of liquid hydrocarbons from methane, which is commercially viable and energy saving. Attention was paid to study the role of the acidity of H-ZSM-5 in the dual bed catalytic system. The characterization of the H-ZSM-5 materials using a series of TPD-NH 3 techniques was undertaken in order to provide some information about composition, properties, and their behaviour of H-ZSM-5 catalysts on the reaction. The relation between the acidity and the activity of H-ZSM-5 was studied by means of TPD-NH 3 techniques. The influence of reaction temperature ( o C), CH 4 /O 2 ratio (4-10) and the SiO 2 /Al 2 O 3 ratio of HZSM-5 (30, 50, 80 and 280) on the performance of the dualbed catalytic system was also studied The Dual Bed Catalytic System The concept of the double-bed catalysts configuration is schematically shown in Figure 3.1. The system comprises of a dual- bed catalytic system for the production of liquid hydrocarbons from methane gas. In the first bed, methane is converted to C 2 (mostly ethylene and ethane) and other gaseous products by oxidative coupling of methane over La/MgO. Subsequently, the olefins, formed in the first bed, are converted to liquid hydrocarbons without having to separate them from the first bed. In literature, the investigation of this type of reactor concepts, which combine at least two catalyst or process functionalities has been the subject of considerable attention in the recent industrial and academic researches. The enhanced reaction

85 63 dual-bed reactor concept offers high potential for process simplification and energy conservation. For example, Pérez et al. (2000) proved that a dual-bed catalytic system, consisting of a deno x and den 2 O catalyst bed in series, showed to operate with high and stable activity for successive removal of NO and N 2 O. Furthermore, Lee et al. (2003) was investigated a practical uses of a Co-zeolite as a deno catalyst at a lower temperature in dual-bed system, consisting of deno and deno 2 catalyst beds in a series. In addition, syngas has been produced from the gasification of various biomasses such as jute stick, bagasse, rice straw, and saw dust of cedar wood using a dual-bed gasifier combined with a novel catalyst and this concept gave the high-technology biomass-upgrading systems to gaseous and liquid fuels such as gasification (Asadullah et al. 2003). While, Zhu et al. (2004) proposed a system with two catalyst beds instead of one single metal catalyst bed is proposed for catalytic partial oxidation of methane (CPOM ) to synthesis gas. The most important advantage of this approach is the prevention of any contact between oxygen and the metal catalyst, because O 2 is completely converted on the first bed catalyst. Therefore, metal loss via evaporation of oxides can be excluded. Thus, the previous research on the dual-bed catalytic system provides an incentive to develop a conceptually this type of catalytic system in the present work. On the other hand, a number of earlier studies demonstrated that La/MgO is a reasonably good catalyst for the OCM reaction (Traykova and Davidova, 1998; Choudhary et al., 1997; Lacombe et al., 1997; Kus et al., 2003; Choudhary et al., 2000b; Lacombe et al., 1995 and Kus et al., 2002a). Choudhary et al. (2000b) reported that La promoted MgO exhibited high activity and C 2 selectivity (with a methane conversion 27.6 %, C 2 selectivity 51.2%, 0.8 C 2 H 4 /C 2 H 6 ratio and C 2+ yield of about 1.29% at 700 o C) and also high catalyst stability in the OCM process. Their study also indicated that promotion effect of lanthanum was attributed to a large increase in the strong basicity of the catalyst because of the doping of MgO by lanthanum. Because of its high activity/selectivity and high thermal stability, the Lapromoted MgO catalyst has a high potential for use in the OCM process. Hence, La- MgO has been chosen as the potential candidate for producing a high concentration of olefins in the first bed for the dual -bed catalytic system in this study.

86 64 Methane Gases La-MgO 1P stp Catalyst Bed Intermediate OCM Products HZSM-5 2P ndp Catalyst Bed Liquid Hydrocarbons Figure 3.1 Simplified reaction scheme for the dual bed catalytic system over La/MgO and H-ZSM-5 catalysts Subsequently, the function of the second catalytic bed to oligomerize the olefins to higher hydrocarbon products is handed over to H-ZSM-5 catalyst. Some earlier studies have demonstrated that acidic H-ZSM-5 zeolite catalyst is a reasonably good oligomerization catalyst for converting C 2 products to higher hydrocarbons (Wang et al., 1990; Jeong et al., 1998; Yarlagadda et al., 1990; Baba and Abe, 2003; and Alkhawaldeh et al., 2003). Alkhawaldeh et al.(2003) reported that ethylene can be activated over HZSM-5 catalyst with a conversion of 99 % and sixty eight percent of the product stream is C 5 to C 8+. In another process developed by Babe and Abe, (2003) showed that the conversion of ethylene in the presence of methane over HZSM-5 catalyst was 93.9% with 91.3% selectivity for C 3+ hydrocarbons. In this work, La/MgO is placed in the first layer whilst H-ZSM-5 in the second so that the affiliation of basicity and acidity of the catalytic reaction could be explored. The two catalysts are separated by glass wool.

87 Acid Properties of Zeolite Among the most important properties of zeolites with respect to their use as catalysts are their surface acidity. Therefore, the "acidity" of a zeolite has been the subject of considerable research and conjecture. It is to be noted that product selectivity is characterized as being determined by pore geometry and size and its impact on diffusion limitations to selective internal sites rather than being determined or influenced by acidity. Figure 3.2 Neutral All-Silica Framework Zeolites and related microporous molecular sieves consist of a threedimensional network of metal-oxygen tetrahedral which provides the periodically sized microporous structure, in which the active sites are part of the structure. Acid sites result from the imbalance of the metal and the oxygen formal charge in the primary building unit. This can easily be recognized in the case of zeolites, which consist of a three-dimensional network of Si-O tetrahedral (Murakami et al., 1986). A lattice comprising of only Si-O tetrahedra is neutral (Figure 3.2 ) (the 4 + charge at the silicon is balanced by four oxygen atoms with each 2 - charge, however, belonging to two tetrahedra). Replacing one Si 4+ atoms by Al 3+ causes a formal charge on the tetrahedron of 1 - (Brunner et al., 1995). Positive charges are required to balance the negative framework charges to allow the zeolite to be a neutral material. Such positive charges can be metal ions or protons. Thus, this negative charge is then balanced by a proton or metal cation forming an acid site (Figure 3.3) (Weitkamp, 1998). Because of the presence of metal cations or protons in zeolite structures, zeolites have cation-exchange ability or acidic properties. The ion-exchange properties of zeolites play an important role

88 66 when zeolites are used as catalysts or adsorbents (Klinowski and Barrie, 1988). The acidic property of a zeolite takes the major role in zeolitic catalytic materials for acid-catalyzed reactions. HP +P +P Figure 3.3 Formation of acid sites in zeolites Futhermore, zeolites have Bronsted and Lewis acid sites. Bronsted acid can donate protons, while Lewis acid can accept a pair of electrons. Bronsted acid sites in zeolites can change into Lewis acid sites through dehydroxylation on heating. And the reverse reaction takes place when water is present in the zeolite, as shown in Figure 3.4 (Grobet, 1988). Figure 3.4 Inter-conversion of Bronsted and Lewis acid sites. As it is not presently possible to directly measure the concentration of hydroxyl groups, this measurement must be done indirectly. The analytical method of choice is to measure the acidity of the zeolite. This is believed to correlate directly to hydroxyl group concentration. Temperature-programmed desorption of ammonia (NH 3 -TPD) is widely used and is considered to be an effective procedure for

89 67 determining the acidity of zeolites. Ammonia is an excellent probe molecule for the TPD procedure. The molecule's small size allows exceptionally good pore penetration while its strong basicity enables ready stabilization on acid sites. The TPD procedure is preferred because it is relatively simple, consumes a minimum amount of time, and can be accurately reproduced Brønsted Acid Sites Zeolite structure containing only SiO 2 tetrahera would be electrically neutral and no acidity would be developed on its surface. Bronsted acid sites are developed when Si 4+ is isomorphically substituted by a trivalent metal cation, for instance Al 3+, and a negative charge is created in the lattice, which is compensated by a proton. The proton is attached to the oxygen atom connected to neighbouring silicon and aluminum atoms, resulting in the so-called bridged hydroxyl group which is the site responsible for the Bronsted acidity of zeolites (Feng and Chao, 1996). This means that the activity requested is based on the formation of Bronsted acid sites arising from the creation of bridging hydroxyl groups within the pore structure of the zeolites. These bridging hydroxyl groups are usually formed either by ammonium or polyvalent cation exchange followed by a calcination step. The bridging hydroxyl groups, which are protons associated with negatively charged framework oxygens linked into alumina tetrahedral, are the Bronsted acid sites, as demonstrated in Figure 3.5 (Woolery et al.,1997). Figure 3.5 Model of Brönsted acid sites in zeolites

90 68 However, the ability to dissociate these protons in the zeolite framework is influenced by various factors, which could be grouped into the following two main categories (Creyghton et al., 1996): a) geometric factors, such as : the length of the bonds between bridged atoms, the angle size of bonds from the Al-OH-Si group (e.g., it was established that the energy of deprotonation decreases the angle of T- O-T bonds increase), preferential localization of protons at particular oxygen atoms from the secondary building units. b) chemical factors:, such as: chemical composition, e.g. Si/Al ratio the nature of substitution element in the framework (Ga, B, etc.) The highly acidic sites, combined with the high selectivity arising from shape selectivity and large internal surface area make the zeolite an ideal industrial catalyst. In addition, the methanol to gasoline (MTG) process is highly dependent on the presence of the Brønsted proton. If the catalytic H-ZSM-5 is replaced by the purely siliceous analogue of ZSM-5, which has no Brønsted protons, the reaction does not take place at all (Jeanette, 1980). It is the initial steps of reactions such as this that need to be understood so that the zeolite structure and activity can be modified so as to increase the catalytic effects of the zeolites, which is a very economically important step for industry Lewis Acid Sites Lewis acid sites (electron pair acceptor sites) are related to the formation of positively charged oxide clusters or ions within the porous structures of the zeolites. These species are typically alumina or silica/alumina, formed by extraction of

91 69 aluminum from the lattice, or metal ions exchanged for the protons of acid sites. Depending on the nature of the metal cation, these may contain hydroxyl groups (by partial hydrolysis of water) (Brunner et al., 1995). Note, that these metal cations together with the adjacent framework oxygens will act as Lewis acid/base pair and may polarize bonds in reacting molecules. The former type of Lewis acidity, i.e. aluminum oxide clusters containing alumina in octahedral and tetrahedral coordination will usually be a stronger Lewis acid than exchangeable metal cations ( true Lewis acids) (Woolery et al., 1997). 3.2 Experimental Catalyst Preparation The La-MgO catalyst preparation procedures are illustrated in Chapter 2. Four H-ZSM-5 samples, with different SiO/Al 2 O 3 ratios, were used and designated as HZ-(30), HZ-(50), HZ-(80), and HZ-(280), where parenthesis indicates the SiO 2 /Al 2 O 3 molar ratio. Zeolytst TM International supplied all the ZSM-5 zeolite in the ammonium form. By calcinations at C for 5 hours, the ammonium form was converted to H-ZSM-5. In addition, H-ZSM-5 (SiO/Al 2 O 3 =30) were also pretreated at different thermal conditions (T=550, 600, 650, 700, 750 and C) in order to study the effect of temperature on the acidity of the H-ZSM-5. These zeolites were coded as HZT-(X) [HZT-(550), HZT-(600), HZT-(650), HZT-(700), HZT-(750) and HZT-(800)] where X refers to the temperature where thermal treatment were performed Catalyst Characterization All the catalysts in this chapter were characterized using TPD-NH 3. The technique for the catalyst characterization was described in Chapter 2.

92 Catalytic Activity Measurement Catalytic tests were carried out over La-MgO and H-ZSM-5 catalysts in the dual bed catalytic system under atmospheric pressure in a continuous flow fixed bed quartz reactor, which were charged with 1g of La-MgO and 2 g of H-ZSM-5 (20-45 meshes) catalysts. The catalytic testing were carried out to study the influence of reaction temperature ( C), CH 4 /O 2 ratio (4-10) and the SiO 2 /Al 2 O 3 ratio of H-ZSM-5 (30, 50, 80 and 280) on the performance of the dual bed catalytic system. The product gases leaving the reactor separated into liquid and gas fraction through water bath. The reactor effluent gases were analyzed on-line by means of a Hewlett Packard Agilent 6890N gas chromatograph system equipped with thermal conductivity detector and four series column (UCW 982, DC 200, Porapak Q and Molecular sieve 13A). The methane conversion and the product selectivity were calculated on a carbon number base using N 2 as inner reference (coke was not taken into account). Furthermore, the experimental rig setup and detailed procedures employed in the catalytic activity measurements have been described in Chapter Results and Discussion Catalysts Characterization SiO 2 /Al 2 O 3 Ratio Effect The NH 3 -TPD profiles of the H-ZSM-5 catalysts with different SiO 2 /Al 2 O 3 ratios are depicted in Figure 3.6. All samples exhibit two well resolved desorption peaks: the low desorption temperature peak ( ) and the high desorption temperature peak ( ) are in good agreement with other previous reports (Tynjala et al., 1998; and Guo et al., 2004). The peak is assigned to desorption of ammonia from silanol and other weak acidic centers; the peak is assigned to desorption of ammonia from strong acid sites (Wang et al., 1990; Lonyi and Valyon, 2001; Amin and Kusmiyati, 2004; and Schwarz et al., 1989). Based on the peak intensities, the relative areas of these two peaks ( and ) are proportional to the SiO 2 /Al 2 O 3 ratio. There is a gradual

93 71 HZ-(30) HZ-(50) HZ-(80) HZ-(280) T/K Figure 3.6 Temperature programmed desorption of ammonia from H-ZSM-5 with different SiO 2 /Al 2 O 3 ratios

94 72 decrease in the intensity of the peak with increasing SiO 2 /Al 2 O 3 ratio. A drastic reduction in the intensity of the signal is observed for the high silica sample HZ(280). Moreover, increasing SiO 2 /Al 2 O 3 ratio from 30 to 280 results in a nonzero background level at peak. The less significant intensities of the and peaks with increasing SiO 2 /Al 2 O 3 ratio support the attribution of the H-ZSM-5 SiO 2 /Al 2 O 3 ratio to the distribution and density of the acid sites. The acidity properties of the different SiO 2 /Al 2 O 3 ratio of the H-ZSM-5 catalysts are summarized in Table 3.1. The amount of desorbed ammonia can be considered as the amount of acid sites of the sample. The results in Table 3.1 show that the total number of acid sites (the strong and the weak acid sites) increases with aluminum content. A general decrease of the total amount of desorbed ammonia was found for the increasing SiO 2 /Al 2 O 3 ratio, indicating a global decrease in the concentration of the acid sites. The facts above also indicate that the total acidity and structural aluminum content of the zeolite are linearly related as it is expected that acidity decreases for samples with high SiO 2 /Al 2 O 3 ratios, which is confirmed by these results. The amount of ammonia (Table 3.1) also allows us to evaluate the concentration of accessible weak and strong acid sites. It is observed that the amount of ammonia desorbed at lower temperature is always higher than that desorbed at higher temperature. Moreover, the number of acid sites that corresponds to the peak of NH 3 -TPD corresponds to the framework Al atoms of fresh H-ZSM-5 zeolites. The strong acid sites increase with Al content. Theoretically, one proton should be introduced for each framework Al 3+, and therefore the larger the number of framework aluminum atoms, the higher the potential number of acid sites would be in zeolite (Corma, 1995; and Inanov et al., 1999). Strong acid sites (Bronsted) are generated by tetrahedrally coordinated Al atoms forming Al-O(H)-Si bridges. The increase in the strong acid sites with Al content implies that most of the Al is tetrahedrally coordinated in the framework. Therefore, it is clear that the total number of Bronsted acid sites present in a zeolite catalyst depends on the framework Si/Al ratio, or more generally speaking, on the ratio of framework M 4+ /M 3+ cations.

95 73 Table 3.1 Acidity of H-ZSM-5 catalysts with different SiO 2 /Al 2 O 3 ratios by TPD-NH 3 Sample Total acidity Weak Acidity Strong Acidity (mmol NH -1 3 gcat ) (mmol NH -1 T 3 gcat ) d (K) (mmol NH -1 3 gcat ) T d (K) HZ-(30) b HZ-(50) HZ-(80) HZ-(280) b The SiO 2 /Al 2 O 3 ratio of the materials are given in the parentheses Lobree et al. (1999) reported that the strong acid sites peak is assigned to NH 3 strongly adsorbed on Bronsted acid sites. Since it is agreed that the strong acid sites peak is due to NH 3 strongly adsorbed on Bronsted acid sites, this peak can be used to observe qualitative changes in the concentration of Bronsted acid sites by varying the SiO 2 /Al 2 O 3 ratio. Assuming that the strong acid sites are directly related to Bronsted acid sites in the H-ZSM-5 samples, the results suggested that the amount of Bronsted acid sites increases with decreasing SiO 2 /Al 2 O 3 ratio. The strength of the acid sites can be determined by the desorption temperature. In agreement with the literature, all the NH 3 -TPD profiles obtained for our catalysts exhibited two main peaks corresponding to two types of active sites with different strengths. Each peak is characterized by the T d value (Table 3.1). The temperature range for the maximum peak was set between 465 K and 500 K and above 640 K for the weak and strong acid sites respectively. As clarified earlier, the peak which represents the family of strongly acidic sites depends on the SiO 2 /Al 2 O 3 ratio. This peak slightly shifted toward the higher temperature range as the aluminum contents of H-ZSM-5 decrease. Among the zeolites, HZ(280) is found to have higher acidic strength; the order of the acidic strength order is listed as: HZ(280) >HZ80)> HZ (50) >HZ(30). Therefore, a comparison of the strength of the bands for samples with different SiO 2 /Al 2 O 3 ratio reveals a clear connection between the strength of these peaks and the SiO 2 /Al 2 O 3 ratio. Similar observations were reported by Forni et al. (1995) and Weitkamp (2000) where a high SiO 2 /Al 2 O 3 ratio led to an increase in acid sites strength. The

96 74 shift of the second peak as the aluminum content decreases indicates an increase in the acid strength of the sites and, according to Camiloti et al. (1999), can be attributed to a higher proton activity as a consequence of the lower negative charge density of the zeolite framework. On the other hand, Corma (1995) concluded that the acid strength of a given site increased when there is a decrease in the number of aluminum atoms in the nearest neighbouring positions (NNN) of the aluminum atom which supports the strongest type of framework Bronsted site. Therefore, he suggested that acid strength can be varied by changing the chemical composition of zeolite. Accordingly, it is clear that the chemical composition of SiO 2 /Al 2 O 3 of the H-ZSM-5 is a parameter that controlled the total number and strength of the acid sites in the zeolite Thermal Treatment Analysis of the H-ZSM-5 Samples The thermal analysis was carried out to determine the thermal stability of the optimum H-ZSM-5 (SiO 2 /Al 2 O 3 =30) zeolites and to measure the loss of acidity as a function of temperature. The samples investigated in this work were obtained by heating in air the parent H-ZSM-5 catalyst at different temperatures. They are referred to as HZ-550, HZ-600, HZ-650, HZ-700, HZ-750 and HZ-800, respectively. Figure 3.7 reveals the TPD of NH 3 thermograms for all the H-ZSM-5 zeolites treated at different temperatures. It can be observed that for most of the catalysts, two main bands occur in the NH 3 -TPD curve, with the one at ca. 465 K corresponding to a weak acid site, and the one at ca. 640 K to a strong acid site. From the intensities of the signals in Figure 3.7, we can deduce that the amount of available acids depends on the treatment temperature although the samples posses the same SiO 2 /Al 2 O 3 ratio. The density of the strong acid sites is significantly reduced, as indicated by the height (or the area under curve) and the temperature of the high temperature peaks, respectively, as shown in Figure 3.7. However, the density and strength of weak acids do not change as significantly as those of the strong acids. The TPD profile of HZ-550 resembles the HZ-600 and HZ-650 samples.

97 75 HZT-(550) HZT-(600) HZT-(650) HZT-(700) HZT-(750) HZT-(800) T/K Figure 3.7 NH 3 -TPD profiles of H-ZSM-5 catalysts treated at different temperatures

98 76 As stated earlier, the desorption temperature and the amount of ammonia desorbed are considered as indexes of acid strength and total number of acid sites, respectively. Therefore, the total desorbed ammonia or total area of the TPD peaks decreased monotonously with increased temperature of the thermal treatment, indicating a decrease in the total number of acid sites (Table 3.2). On the other hand, HZ-550 possesses the largest amount of available strong and weak acidity among the six samples. However, the number of both strong and weak acid sites decreases remarkably, when H-ZSM-5 undergoes heat treatment. Finally the TPD profile of H- ZSM-5 demonstrates that the heat treatment at above C is more likely to decrease the number of acid sites where only one peak at low temperature (~460 K) is observed, indicating a weak acidity. Table 3.2: NH 3 sorption capacity of the H-ZSM-5 samples treated at various temperatures Total acid Loss of strong acid sites after Sample Temperature ( o C) a Density (mmol/g) thermal treatment(%) b HZT HZT HZT HZT HZT HZT a temperature of thermal treatment analysis b loss of strong acid sites at each temperature calculated relative to amount of strong acid sites at HZ- 550 The trend of the acidity values measured by this technique agrees favourably with that obtained by other researchers. According to Xing et al. (2005), the increase in calcinations temperature would attribute to the removal of water molecules, resulting in the decrease of Bronsted acid sites with constant increase in Lewis acid sites. Some literature also showed (Schwarz, 1989; and Xing et al., 2005) that when H-ZSM-5 was thermally treated above 550 o C, more and more Bronsted-type acid sites were converted into Lewis acid sites with higher temperatures and prolonged calcinations time, which resulting in the desorption of water molecules. Therefore,

99 77 temperatures have an influence on the strength and distribution of acidic sites. With the increase of temperature, the number of Bronsted acids will decrease slowly, while the number of Lewis acids increases slightly. Additionally, the decrease of the higher temperature desorption peak of acid sites concentration is much more remarkable at the temperature above 700 o C as shown in Table 3.2. Campbell et al. (1996) pointed out that dealumination of zeolite lattice occurred at T c = 725 o C for H-ZSM-5 and there would be the breaking of the Si-H-Al bonds and the formation of extralatice aluminum species. At T c 750 o C, Tan et al. (2002) observed dealumination and suggested with the partial destruction of the H-ZSM-5 structure and the occupancy of the channels by extralattice aluminum, there would be a reduction mainly in the number of Bronsted acid sites. Similar results are also observed in our present study. Ultimately, these results suggest that the total amount of ammonia adsorbed continuously decreased with increasing treatment temperature due to the decrease in framework alumina of the treated ZSM-5 samples that is caused by increased dealumination with increase in temperature Catalytic Performance Effect of Temperature Figure 3.8(a) displays the variation in the rate of reaction of methane over the dual bed catalysts and the corresponding product distribution as a function of temperature. A ratio of CH 4 to O 2 about 10:1 (in volume) was selected, and the temperature was varied from 650 o C to 800 o C. Figure 3.8 (a) exhibits the slight increase in the conversion of methane obtained when the temperature increased from 650 o C to 800 o C, but only to a small extent. It also noted that by increasing the temperature, the ratio of ethylene/ethane ratio increased monotonously. The selectivity of C 5+ reached a maximum temperature of 700 o C and then decreased dramatically. At T=800 o C, H-ZSM-5 (SiO 2 /Al 2 O 3 =30) catalyst failed to function as a catalyst for oligomerization reaction. As the temperature increased, selectivity of

100 78 C 3 remained almost constant. The selectivity of C 4 was constant until 750 o C, and the selectivity leveled up significantly at 800 o C. The CO 2 /CO ratio was nearly constant at the temperature range of o C but increased remarkably to a stable value at higher temperatures. The influence of temperature on methane conversion is similar to most of the results investigated for OCM catalysts. With increasing temperature, methane conversion passes through a maximum or reaches a plateau. The temperature of maximum conversion of methane is specific for each catalyst and depends on the partial pressures of reactants as well as on mixing pattern, which are different in various reactor types. However, in this dual bed reactor, by increasing temperature numerous simultaneous reactions involved in the complex reaction pathway of methane and intermediate/final products transformation can also produce methane, decreasing the net conversion of methane (Erena et al., 2000; and Choudhary et al., 2000). In addition to catalytic reactions in dual bed, a large number of homogeneous free radical reactions (Samarth et al., 1994; Pareja et al., 1998; and Sinev, 1995) is also involved, depending upon the process conditions. Hence, the profile of methane conversion as a function of temperature in this study is slightly different compared with single bed La/MgO catalysts as reported previously (Traykova et al., 1998; Choudhary et al., 1997 and Choudhary et al., 2000). The increase in the ethylene/ethane ratio as a function of temperature indicates that the conversion of ethane to ethylene is favoured at higher temperatures. It is expected due to the thermal decomposition of ethyl radicals and thermal cracking of ethane at higher temperatures (Choudhary et al., 1997). On the other hand, increased selectivity of C 2 olefin formation with an increase in temperature can be explained on the basis that an increase in temperature increases the rate of dehydrogenation reactions. This results in an increase in the formation of C 2 olefins at the expense of C 2 paraffins (Sethuraman et al., 2001). Ethylene that is produced in the first stage may be reacting with other intermediate species, undergoing numerous secondary transformations in the second bed catalyst. Thus, the presence of strong acid sites in the second bed H-ZSM-5 which is efficiently introduced into the oligomerization, dehydrogenation,

101 79 cyclization, aklylation, isomerization and cracking reactions in complex mechanism produces C 5+ hydrocarbons. Nevertheless, deactivation of H-ZSM-5 catalysts at high temperature as proven in this study, affects the product distribution and properties in the catalytic reaction. Thus, a significant change in acidic property at higher temperature causes the H-ZSM-5 catalyst to be less effective for oligomerization reaction. This phenomenon exerts more negative effect in terms of ethylene oligomerization and it proved that the presence of Bronsted acid sites is essential for ethylene oligomerization. Hence, the ratio of ethylene to ethane increased markedly and selectivity of C 5+ decreased drastically within the range of reaction temperature above 700 o C. Based on these results, the effect of thermal deactivation of the H- ZSM-5 in the second bed catalyst demonstrated that the amount of ethylene formed in the first bed catalyst was higher than the amount of ethylene consumed at higher temperature in the second bed. In this dual bed process, the visible increase in selectivity of C 4 products over the investigated La-MgO and H-ZSM-5 samples is displayed only when the temperature is elevated to 800 o C. The enhancement in C 4 -hydrocarbon selectivity implies that reaction temperatures at above 750 o C exert an effect more negative on oligomerization reaction than on C 2 activation in the second bed, and the presence of Bronsted acid sites is essential for oligomerization. In addition, the marked increase of C 4 -hydrocarbon selectivity coupled with the decrease in the amount of acid sites demonstrates that C 4 -hydrocarbons is an initial product in C 2 -hydrocarbon activation through the second bed reaction. On the other hand, the rise of CO 2 /CO ratio in the whole investigated temperature range indicates the formation of CO 2 is more favoured at higher temperatures (Mallens et al., 1995). This implies that reaction steps that lead to the formation of CO 2 increased at a faster rate with temperatures than those that lead to the formation of hydrocarbons. Accordingly, this observation is consistent with results in the literature and implies a higher rate of carbon oxidation with an increase in temperature and the CO formed is converted to CO 2 at higher temperatures (Martin and Mirodatos, 1995 and Corma et al., 1997).

102 Effect of Oxygen Concentration In the second series of experiments, we studied the effect of varying the methane to oxygen (CH 4 /O 2 ) ratio at a reaction temperature of 700 o C on the combination of both La/MgO and H-ZSM-5 (SiO 2 /Al 2 O 3 =30) catalysts. As shown in Figure 3.8(b), with increasing CH 4 /O 2 ratio, the methane conversion and ethylene to ethane ratio decrease constantly. However, the selectivity of C 4 and the ratio of CO 2 /CO go through a maximum for a ratio of CH 4 /O 2 about 8 and then decrease dramatically. In contrast, the selectivity of C 5+ increases remarkably and the selectivity of C 3 remains almost constant with increasing CH 4 /O 2 ratio. Chaundary et al. (2000c), Trayhora et al. (1998) and Davydov et al. (1995) have reported that the promotion effect of the lanthanum on MgO is attributed to a large increase in the strong basicity of the catalyst. This property is responsible for the La-MgO catalyst to exhibit better catalytic performance in the OCM process, giving higher values of methane conversion and C 2 (C 2 H 4 + C 2 H 6 ) selectivity. However, the second acidic layer of the H-ZSM-5 catalyst which is tested as an oligomerization function catalyst was reported to have shown diminishing activities on the methane conversion compared to the OCM catalysts (Mleczko and Baerns, 1995 and Dehertog and Fromen, 1999). Based on this understanding, it can be concluded that in our dual-bed catalytic system, La/MgO catalyst is particularly responsible for the methane conversion to produce OCM products at the first stage. Therefore, discussions of the effect on the conversion of methane in dual bed catalytic system as well as by the CH 4 /O 2 ratio in the feed will correlate with OCM reactions over La-MgO catalyst. Anshits et al. (1998) and Takashi and Aika (1997) have demonstrated that the presence of gas phase oxygen is essential for favorable methane activity in order to achieve a high level methane conversion via OCM reaction. The equilibrium adsorption reaction between gas phase O 2 and the La-MgO catalyst result in a form of surface oxygen that is capable of abstracting a hydrogen atom from methane where methyl radicals formed on the surface coupled in the gas phase to produce ethane (Vedeneev et al., 1995; Mallens et al., 1995 and Martin and Mirodatos, 1995). Consequently, oxygen is absolutely required as a key in the formation of methyl

103 81 a) EFFECT OF TEMPERATURE b) EFFECT OF CHB 4B /OB 2B RATIOc) EFFECT OF SiOB 2B /AlB 2B OB 3 BRATIO Figure 3.8 Influence of reaction parameters on the catalytic activity and product distribution ( methane conversion, ethylene to ethane ratio, selectivity of C 3, selectivity of C 4, selectivity of C 5+ and CO to CO 2 ratio)

104 82 radicals. Therefore, the clear correlation between the methane conversion and oxygen concentration could be found in Figure 3.8, suggesting that the rate of methyl radical formation (rate determining step of this reaction) is exclusively related to the methane to oxygen ratio, in accordance with previous conclusions drawn from the study of the oxidative coupling over alkali earth metal oxides and rare earth metal oxides catalysts (Lacombe et al., 1994; Kus et al., 2003; and Borchert and Baerns, 1997). It is interesting to note that, at least for the selective route leading to C 2 products, methane must be activated into methyl radicals released into the gas phase. Ethylene as well as ethane are primary reaction products from methane but at higher oxygen concentration, ethylene was predominantly produced from ethane in a secondary reaction in accordance with previous studies (Mleczko and Baerns,1995 and Martin and Mirodatos, 1995). The increase in the ethylene/ethane with decreasing CH 4 /O 2 ratio which is also observed in the earlier studies (Choudhary et al., 1997; and Choudhary et al., 2000b), is most probably due to the availability of O 2 at higher concentration for the following gas phase reactions involved in the formation of ethyl radicals and ethylene from ethane. Therefore, the results in this study indicate that ethylene formation is not a result of dehydrogenation of ethane and suggests that the main production of secondary product ethylene is by oxydehydrogenation of the primary product, ethane (Traykova et al., 1998; and Choudhary and Mulla, 2001). This phenomenon explains why the profile of the ethylene to ethane ratio increased with the decrease of the CH 4 /O 2 ratio as illustrated in Figure 3.8. In the presence of oxygen, numerous studies reported the ease of gas phase oxidation of ethylene and this is not surprising owing to the fact that the double bond in ethylene is much more susceptible to be activated to undergo both catalytic and non-catalytic combustion reaction to form CO x products (Choudhary et al., 2000a; and Pak et al., 1998). In addition, high oxygen concentrations cause some of the intermediate and desired final hydrocarbon products to become more combustible and it is also attributed solely to the lost in selectivity of the products (Mallens et al., 1996). Therefore, with an increase in oxygen concentration, it is observed that C 4 and C 5+ selectivity decreases drastically, suggesting that the combustion reaction is

105 83 more favored than oligomerization reaction. Therefore, the concentration of oxygen should be kept at low levels to minimize the undesired oxidation of ethylene. The dependence of CO 2 /CO ratio on CH 4 /O 2 ratio clearly indicates consecutive oxidation of CO to CO 2. From the experimental results, and also based on the data published in the literature, it is possible to make some considerations that the availability of greater oxygen favours the conversion of CO to CO 2 and formation of more CO 2 as compared with that of CO in the process under the oxygenadvantaged conditions. Conversely, at the higher oxygen concentration (CH 4 /O 2 = 4 and 6), production of both CO and CO 2 involves a complex reaction system, with the reaction steps occurring in series and parallel. In such a situation, the response of CO and CO 2 products to change in CH 4 /O 2 ratio does not necessarily produce any definite trends. The methane activation reactions are generally considered to occur by initial homolytic cleavage of the methane molecule at the surface catalyst, followed by subsequent gas phase recombination of methyl radicals to form the desired C 2 products. However, in oxidation catalysis the possibility of total oxidation is always present, so the following unselective reactions are also likely to occur (Burrows et al., 1998): CH 4 + 3/2O 2 CO + 2H 2 O [3.1] CH 4 + 2O 2 CO 2 + 2H 2 O [3.2] In addition, the influence of CH 4 /O 2 ratio on the CO 2 /CO ratio is, however, found to vary from catalyst to catalyst. The CO 2 /CO ratio passed through a maximum with decreasing the CH 4 /O 2 ratio, which has also been observed earlier in the OCM over single bed La-promoted MgO catalyst by Choudhary et al. (1997). Thus, based on the findings, it is clear that a rise in oxygen concentration gives a positive effect in terms of methane conversion but is not beneficial for C 5+ selectivity, which is our main desired product in this study.

106 Effect of Acid Site Concentration The catalytic properties of H-ZSM-5 materials with different SiO 2 /Al 2 O 3 ratios were evaluated for the same reaction conditions at temperature C and CH 4 /O 2 ratio 10. This study was carried out in an attempt to test the hypothesis that the Bronsted acid sites of H-ZSM-5 participate actively in the oligomerization reaction step in the dual bed catalytic system. It can be seen that the methane conversion and selectivity of C 3 hydrocarbon remain almost constant for the whole SiO 2 /Al 2 O 3 range studied (Figure 3.8c). The ratio of ethylene to ethane and selectivity of C 4 increase when SiO 2 /Al 2 O 3 ratio is increased from 30 to 80. Nonetheless, these trends slightly decrease with SiO 2 /Al 2 O 3 ratio 280. Furthermore, increasing SiO 2 /Al 2 O 3 ratio of H-ZSM-5 results in a sharp decrease in selectivity of C 5+ and a gradual increase in the ratio of CO 2 to CO. As was presumed, the H-ZSM-5 catalysts in the second bed are much less active toward unreacted methane molecules from the first bed. Additionally, not much variation in the methane conversion was observed even by changing the acid properties of the second bed by varying the SiO 2 /Al 2 O 3 ratio of the H-ZSM-5 zeolite. The result seems to suggest that H-ZSM-5 used as second bed in this approach is not an active phase to alter the methane molecules and thus methane conversion is independent of the H-ZSM-5 catalyst. Based on the above consideration, it is evident from this study that acidic H-ZSM-5 catalyst applied as second bed is less effective to dissociate the C-H bond in methane and the C-H bond activation of methane mainly depends on the first bed of La-MgO catalyst. These results also help to conclude that the La-MgO catalyst is crucial to catalyze the methane reaction. On the other hand, the knowledge on the structure of active centre is of prime importance for elucidation of H-ZSM-5 catalytic activity in the dual bed catalytic system. As pointed out earlier by using TPD-ammonia characterization, the density of strong acid sites is dependent upon the value of the SiO 2 /Al 2 O 3 ratio, whereby lower SiO 2 /Al 2 O 3 gives higher amount of strong acid sites. In addition, the literature indicates that the chemical uniformity of the Bronsted -type acid sites (framework hydroxyl groups) in H-ZSM-5 contents has been proven in the oligomerization test

107 85 reaction, where the oligomerizing activity is linearly related to Bronsted acid sites content (Schwarz et al., 1989; Sano et al., 1999 and Kresnawahjuesa et al., 2002). In a previous work, Schwarz et al. (1989) showed that the structure of Bronsted acid centres is responsible for oligomerization reaction and there is a significant change in products distribution with an increase in SiO 2 /Al 2 O 3 ratio of H- ZSM-5. Furthermore, Vereshchagin et al. (1995) suggested that the reaction of ethylene oligomerization via H-ZSM-5 catalyst is a sequence of reactions that include a slow step of active surface species formation followed by fast process of chain growth. With respect to the studies (Wang et al., 1990; Yarlagadda et al., 1990; Schwarz et al., 1989; and Vereshchagin et al., 1995) the Bronsted acid sites of H- ZSM-5 in the second bed are most probably responsible for oligomerizing the initially formed ethylene by the OCM catalyst in the first bed to produce C 5+ hydrocarbon products in this catalytic system. Consequently, one can observe a monotonic increase in C 5+ selectivity with a decrease in the bulk SiO 2 /Al 2 O 3 ratio. This increase becomes more noticeable for the H-ZSM-5 with the highest aluminum content, namely H-ZSM-5 with SiO 2 /Al 2 O 3 of 30 and 50. This trend can be related to the high hydrogen transfer [HT] capability of low SiO 2 /Al 2 O 3 that results in the increase of the rate of oligomer formation as previously reported (Sano et al., 1999; and Vereshchagin et al., 1995). Therefore, it is demonstrated that C 5+ product selectivity in this study depends strongly on the distribution of Bronsted acid sites, which is related to the extent of hydrogen transfer reactions. The linear correlation between selectivity of C 5+ hydrocarbon products and concentration of Bronsted acidic site confirms the crucial role of acid sites in oligomerization. Another interesting observation that has been found in this study is the nearly opposite variation in selectivity of C 4 and C 5+ hydrocarbon as a function of SiO 2 /Al 2 O 3 ratio. The result suggests that the C 4 hydrocarbon is produced at the initial stage of the oligomerization reaction due to the coupling of ethyl radicals both on the catalyst surface and in the gas phase reaction (Anunziata et al., 2000; and Barbero et al., 2001). This C 4 species hydrocarbon which is more reactive than C 2

108 86 species will oligomerize further to produce C 5+ hydrocarbon products in the dualbed catalytic system. One can speculate as to why the ethylene to ethane ratio goes through a maximum. The most likely hypothesis is that the double bond in ethylene is much more susceptible to be activated than the C-H bonds in ethane. Therefore, H-ZSM-5 with SiO 2 /Al 2 O 3 30 which has high acid density sites will increase the rate of ethylene consumption which reflects the rate of oligomer formation (Schwarz et al., 1989; Kucherov et al., 1989; and Hu and Ruckenstein, 1996).This is a possible reason to low ethylene to ethane ratio at low SiO 2 /Al 2 O 3 ratio. Accordingly, an increase in ethylene to ethane ratio in the final products, coupled with the increase in the SiO 2 /Al 2 O 3 ratio of H-ZSM-5 is limited by the oligomerization. Changing SiO 2 /Al 2 O 3 ratios in the range of and other reaction parameters do not show a significant effect on selectivity of C 3. Therefore, it is not clear whether C 3 species hydrocarbon play any direct or indirect role in the catalytic oligomerization reaction in the second bed H-ZSM-5 catalyst. Likewise, H-ZSM-5 zeolite shows a great activity in the oxidation of light hydrocarbons leading to the formation of total oxidation products, mainly carbon oxide. Consequently, the ratio of CO 2 to CO increases as the SiO 2 /Al 2 O 3 ratio of H-ZSM-5 increase. This result indicates that H-ZSM-5 with relatively strong acidity lead to CO formation as observed previously (Lucas et al., 1998). This study clearly suggests that the activity of the oligomerization sites is predominantly affected by the SiO 2 /Al 2 O 3 ratio, possibly reflecting a high degree of Bronsted acid sites which is mainly responsible for the formation of larger hydrocarbons. Thus, H-ZSM-5 with low SiO 2 /Al 2 O 3 ratios in the second bed gives higher activity and selectivity to C 5+ liquid hydrocarbons, which is our desired product in this study.

109 Summary The results of this study demonstrated that the catalytic behavior of a dualbed system strongly depended on the reaction temperature and oxygen concentration. The catalytic activity and selectivity of C 5+ were directly proportional to the number of Bronsted acid sites. The results implied that the Bronsted acid sites of H-ZSM-5 were the active centers responsible for the oligomerization of primary ethylene products. Oxygen was absolutely necessary for the formation of the methyl radicals from methane, but it should be provided at a controllable manner in order to avoid undesired oxidation of intermediate OCM products. In addition, the partial destruction and dealumination of H-ZSM-5 at higher temperature had caused the deactivation of the second bed with the H-ZSM-5 catalyst for the oligomerization reaction. The activity of the oligomerization sites was unequivocally affected by the SiO 2 /Al 2 O 3 ratio.

110 CHAPTER 4 DIFFERENT FUNCTION OF METALS LOADED TO H-ZSM-5 CATALYST IN THE DUAL BED CATALYTIC SYSTEM 4.1 Introduction As pointed out by Choudhary et al. (2000b) the application of La-MgO catalyst, which was used as the first catalyst bed in this study, resulted in the high selectivity of ethane and ethylene products. Besides, the study in the previous chapter in this thesis clearly showed that the introduction of the second layer bed catalyst, which comprises H-ZSM-5 catalyst, could encourage supplementary reaction to form liquid products of C 5+ hydrocarbon. Furthermore, the H-ZSM-5 catalyst with high acid site concentration is more tolerant to promote the formation of C 5+ products. It is interesting to note that, in spite of its very strong acidity, some researchers have found a drastic increase in the catalytic activity of H-ZSM-5 zeolite which has metal functions in addition to their acid function. Thus, the present investigation was undertaken to develop the bifunctional metal loading H-ZSM-5 catalyst concept by modifying H-ZSM-5 surface while introducing some metal particles via impregnation method in order to maximize the yield of C 5+ products in this dual bed catalytic system. This study aims to evaluate the performance of metal loaded H-ZSM-5 as a potential catalyst for oligomerized the intermediate OCM product and compare it with unloaded metal zeolite. The variations of product selectivity in this dual bed system were also carefully examined in order to elucidate the possible reaction path in this system.

111 Catalysis of Metal Loaded H-ZSM-5 Catalysis by metal loaded H-ZSM-5 is extensively used. The presence of a metal on zeolite is one of research interest and such a conception provides a new means to tailor catalytic performance by altering the exposure of the metal sites and opens the way for creation of new types of catalysts. This class of catalysts and model surfaces can also provide a fundamental implication on the molecular-level mechanism and essential controlling factors for catalysis is to address important issues of activity and selectivity in the catalytic reaction, in particular the principles of tuning metal reactivity and activating reaction intermediate at zeolite surfaces. Therefore, the modification of the pore and acidity of ZSM-5, then, came to our attention. The methods for modifying zeolites are usually by impregnation and ionexchange. The impregnation method is widely applied to the preparation of various metal loaded zeolite catalysts, and the ion-exchange technique is used for the supports having ion-exchange ability. Many researchers studied the applicability of modified ZSM-5 zeolite for the conversion of lower hydrocarbons to gasoline and other liquid C + 5. Han et al.,(1994) demonstrated the production of higher hydrocarbons from methane oxidation using ZSM-5 containing catalyst metal oxides. Under their test conditions, Ni, Cu, Zn, Ga- ZSM-5 produced C 5+ hydrocarbons from a CH 4 /O 2 feed containing no additives. Cr- Mn-, Co-, Fe-ZSM-5 catalysts yielded exclusively CO x and C 2 -C 4 products. They showed that Pt- and Ag-ZSM-5 catalysts did not produce significant quantities of higher hydrocarbon from CH 4 and O 2 and explained that noble metal catalysts would be expected to have high oxidation activities. Finally they clarified that the direct partial oxidation of CH 4 with O 2 to higher hydrocarbons and in particular, C 5+ liquids, occurs over ZSM-5 catalysts containing various metal oxides and concluded that dehydrogenation and oxidation activities of metal oxides need to be balanced for C 5+ production to occur via the CH 4 oxidation/mtg reaction sequence over metalcontaining ZSM-5 catalysts. In Baba and Abe s (2003) work, the conversion of not only CH 4 but also 13 CH 4 was examined over various metal cations-loaded H-ZSM-5 (Ag, Mo, In, Fe, V, Ga, Pd, Pb, La, Zn and Cu) in the presence of ethylene to show the experimental

112 90 evidence that methane was activated on metal cation-acidic proton bifunctional catalyst. Their studies showed that the catalytic activity and the selectivity for hydrocarbons depended on the kind of metal cations loaded on H-ZSM-5. They concluded that the bifunctionality of acidic protons and metal cations is essential for the activation of methane and suggested their concept can be applied to activation of other alkanes. In their work, Alkhawaldeh et al., (2003) converted a mixture of methane and ethylene or acetylene was reacted over H-ZSM-5 and different metal-loaded H-ZSM-5 catalysts. They noticed that the reaction produced high molecular weight hydrocarbons with a carbon number of five and more. In another process developed by Timmons et al., (1992) reported the reaction of acetylene over nickel or cobaltcontaining zeolite catalyst in the presence of a hydrogen donor co-reactant to produce higher molecular weight hydrocarbons and the catalyst/reactant feed process described eliminates rapid catalyst deactivation. Yin et al. s (2005) study found that olefins in the cracked naphtha can be possibly transformed into higher hydrocarbon over Ni/H-ZSM-5 catalyst to convert the olefins as well as to improve the octane number. Hulea and Fajula (2004) had publicized that Ni-exchanged well-mesostructured silica-alumina catalysts revealed very interesting properties for ethylene oligomerization. They found that both nickel and acid sites are required for the activation of this reaction. Moreover, the acid density plays a significant role in determining the activity, stability and selectivity. They have shown in their work that it is possible to achieve a desired balance between acid and nickel ion sites so that high catalysts stabilities and high selectivities to suitable products could be achieved. Amin and Anggoro (2003) investigated the direct conversion of methane using a metal-loaded ZSM-5 zeolite prepared via acidic ion exchange. This was investigated to elucidate the roles of methane and acidity in the formation of liquid hydrocarbons. They illustrated that if the strength of the Bronsted acid sites is strong, the cracking of oligomers is easy. However, they indicated that new medium-strength Bronsted acid sites are generated when modified with metal and this corresponds to the highest percentage of C 5 -C 10. They also showed the RON of the liquid products

113 91 obtained over Ga-ZSM-5 is the highest due to its low paraffin and high aromatic content. Thus, they concluded that successful production of gasoline from methane depends on the amount of aluminum in the zeolite framework and the mediumstrength Bronsted acid sites. Weckhuysen et al. (1998) found that Mo-, Fe-, V-, W-, and Cr-loaded H- ZSM-5 zeolites were able to selectively convert methane to aromatics, and the activity decreased in order: Mo> W> Fe> V> Cr. They proved that the catalytic acivity is directly proportional to the number of Bronsted acid sites. These sites are responsible for the oligomerization and dehydrocyclization of the primary ethylene product. They concluded the catalytic performance is strongly dependent on the preparation method, the treatment condition, and the metal loading. They also proved that the activity of catalysts prepared by impregnation is always higher than that of solid state ion-exchanged materials. Recently, it has been shown that Mo loaded medium pore zeolites are active and selective for conversion of methane to arenes (Sarioglan et al., 2004; and Lacheen and Iglesia, 2005). Wang et al., (1996) studied the conversion of methane to arenes over Mo/H-ZSM-5 based catalyst under non-oxidative condition and found that Mo species is transformed to Mo carbide species which is an active phase for methane transformation to higher hydrocarbons. It is generally reported that the activation of methane occurs on molybdenum active species forming ethylene as primary product (Chen et al., 1995; Solymosi et al., 1997; and Xu et al., 2003). In Ha et al. s (2002) work, it was shown by contrast that acetylene was formed as a primary product along with possibly ethylene in the reaction of methane over Mo/H- ZSM-5. The reactions of ethane and propane have also been studied extensively using the monometallic ZSM-5. For example, ZSM-5 zeolites loaded with Ga or Zn (Ga- ZSM-5 or Zn-ZSM-5) are known to catalyze the conversion of lower alkanes such as ethane and propane into aromatics (Heemsoth et al., 2001), but they show practically no catalytic activity for methane conversion. Chang et al., (1995) reported that the activity of catalysts in facilitating the reaction of ethane with oxygen decreases in the order: Ru III -Na-ZSM-5 > Cu II -Na-ZSM-5 ~ Cu I -H-ZSM-5 > Co II -Na-ZSM-5> V V -

114 92 Na-ZSM-5 ~ H-ZSM-5 ~ Co II -H-ZSM-5 > Cr III -H-ZSM-5 ~ Fe III -H-ZSM-5 > Na- ZSM-5. They concluded that the performance of the catalysts is governed by the nature of the metal cations and the acidity of the parent zeolite precursor. Both Cu I - H-ZSM-5 and Cu II -Na-ZSM-5 catalysts are very active for the reaction of ethane with oxygen, but their selectivities to hydrocarbons are very low and they are excellent oxidation catalysts but not for the oxydehydrogenation of ethane. However, Co-H-ZSM-5 gives the best performance of the catalyst tested for the oxydehydrogenation of ethane. They concluded that moderate acidity of ZSM-5 is needed for high selectivity to hydrocarbons. In recent times, Choudhary et al. (2002) reported the Ga-oxide containing H-ZSM-5 catalyst showed much better performance than the parent zeolite for the conversion of dilute ethylene into liquefied higher hydrocarbons. Based on the work cited above, the bifunctional metal loaded H-ZSM-5 system has been promising for the catalytic transformation of light hydrocarbon. However, the problem is that the yield of higher hydrocarbons is still rather low at present; more catalysts that are effective are needed. Therefore, further investigations on the dual bed catalytic system to achieving higher C 5+ yield, namely the additional promoter action due to the presence of metal cation introduced to H- ZSM-5 are studied. To the best of our knowledge, this present account investigating behaviours of such catalytic systems is the first time which involves La/MgO and monometallic H-ZSM-5. The present work is concerned with rationalizing the activities of Mo, Ga, Cr, Co, Ni, Fe, Mn, Cd, Cu, W, and Ag compared to unloaded H-ZSM-5 zeolite in the dual bed catalytic system. 4.2 Experimental Catalyst Preparation The La-MgO catalyst was prepared by the method of Choudhary et al. (2000). A series of H-ZSM-5 (SiO 2 /Al 2 O 3 =30) zeolites loaded with metal cations (Ga, Mo, Cr, Co, Fe, Mn, Cd, Cu, W, Ag and Ni) were prepared by an incipient

115 93 wetness method. Each H-ZSM-5 support was loaded with 1 wt% metal and the metal salts used to prepare these catalysts were listed in Table 2.2. Finally, all the prepared catalysts were pressed, crushed and sieved to 20/45 mesh sizes for evaluating the performance. Furthermore, the details of catalyst preparation procedures were illustrated in Chapter Catalytic Activity Measurement The catalytic testing was carried out to study the influence of metal loaded H- ZSM-5 on the performance of the dual bed catalytic system. The experimental set up was presented earlier in Chapter 2. Only the main features are mentioned here. The first bed catalyst was operated with 1 g La-MgO, and the second bed contained 2 g metal loaded H-ZSM-5 catalyst. The reactor was operated at atmospheric pressure. N 2 was used as an internal standard. Before starting the catalytic test, the catalyst was preheated with nitrogen stream at a temperature of 550 o C for an hour. Testing of all the catalyst samples for catalytic activity was conducted at identical experimental conditions in a fixed bed reactor at a reaction temperature 700 o C, and CH 4 /O 2 ratio of 10. The dual-bed catalytic approach was performed by using the appropriate catalyst combination of La-MgO with different types of metal loaded H- ZSM-5. An analysis of the effluent product from the reactor was performed gas chromatographically using GC-TCD and GC-FID. The detailed procedures employed in the catalytic activity measurements were described in Chapter 2.

116 Results Methane Conversion Figure 4.1 Methane conversion over metal loaded H-ZSM-5 catalysts Figure 4.1 depicts conversion of methane profiles as a function of the type of metal loaded on the H-ZSM-5 based catalysts. As can be seen in Figure 4.1, methane is considerably a stable molecule and difficult to be activated. Therefore, the variation is not much in terms of methane conversion. The addition of nickel to the H-ZSM-5 catalyst results in a significant increase of the conversion of methane than other tested catalysts when applied as second bed catalyst in this catalytic system. Mo/H-ZSM-5 and Co/H-ZSM-5 also possess much better methane conversion when acting as a second layer catalyst compared to unloaded H-ZSM-5. Conversely, it can

117 95 be observed that, Ag/H-ZSM-5, Fe/H-ZSM-5 and Cr/H-ZSM-5 catalysts show slightly lower methane conversion compared to unloaded H-ZSM-5 catalyst. On comparing the results presented in Figure 4.1, one can easily see that the profiles of methane conversion of Ga/H-ZSM-5, Mn/H-ZSM-5, Cd/HZSM-5, Cu/H-ZSM-5 and W/H-ZSM-5 are very similar with unloaded H-ZSM-5 which is nearly 22 % Selectivity of Ethane and Ethylene Figure 4.2 Selectivity of ethane and ethylene over metal loaded H-ZSM-5 catalysts( ethane and ethylene) The present study focuses on the impact of type of metal species on the H-ZSM-5 catalyst on the selectivites of ethane and ethylene, which are mainly

118 96 primary products from the first bed La-MgO catalyst. The selectivity of ethane and ethylene is presented in Figure 4.2. Based on Figure 4.2, one can observe that the positive effect of adding metal is the largest in the case of Ni/H-ZSM-5. For the activity of ethylene conversion in second bed catalyst, Ni species plays very important roles compared to other metal cations which are tested in this study. For the ethylene selectivity, it is almost 4.1%, whereas this number climbed to 13.8% for Mo-loaded H-ZSM-5, 12.7% for Ga-loaded H-ZSM-5, 17.0% for Cr-loaded H-ZSM- 5, 14.9% for Co-loaded H-ZSM-5, 16.9% for Fe-loaded H-ZSM % for Mnloaded H-ZSM-5, 19.0% for Cd-loaded H-ZSM-5, 18.1% for Cu-loaded H-ZSM-5, 12.2% for W-loaded H-ZSM-5 and 14.5% for Ag-loaded H-ZSM-5, 14.9% for Coloaded H-ZSM-5. Therefore, we need to acknowledge the ability of Ni/H-ZSM-5 to catalyze the OCM intermediate in this system by itself. In addition, the Cd/H-ZSM-5 catalyst is obviously less active than other metal loaded zeolite catalyst because showed less activity toward C 2 products. These results reveal that under the same operating condition, metal loaded H-ZSM-5 partly alters the selectivity of C 2 products Selectivity of C 3 and C 4 Products The selectivity of C 4 and C 3 products as a function of the type of metal loaded H-ZSM-5 catalysts compared to unloaded H-ZSM-5 is shown in Figure 4.3. As in the case of bifunctional H-ZSM-5, the C 3 and C 4 products distribution were also affected by the nature of metal used to modify H-ZSM-5. These results indicate a similar trend where selectivity of C 4 is higher than C 3 for all tested metal loaded H- ZSM-5 catalysts except Ag/H-ZSM-5, which showed a reverse trend. Surprisingly, Cr/H-ZSM-5 gave the highest selectivity of C 4 products followed by Fe/H-ZSM-5, Mn/H-ZSM-5 and Cd/H-ZSM-5. On the other hand, Ag/H-ZSM-5 gave the lowest selectivity toward C 4, which is 2.5%. Moreover, the distribution of C 3 product of reaction was not significantly different and has a minor effect on the change in the type of the metal in structure of zeolite as demonstrated in Figure 4.3. However it showed a slight effect when applied Ni loaded H-ZSM-5.

119 97 Figure 4.3 Selectivity of C 3 and C 4 products over metal loaded H-ZSM-5 catalysts ( C 4 and C 3 ) Ratio of CO 2 to CO The results of Figure 4.4 clearly show that the ratio of CO 2 to CO is strongly dependent on the identity of the metal ion. In this study, CO and CO 2 are undesired products and these two products are formed by combustion and oxidation reactions of hydrocarbon products. The bifunctional material in Figure 4.4 shows a significantly different oxidation behavior during the similar test in the dual bed catalytic system. The ratio CO 2 /CO decreases in the sequence: Cd/H-ZSM-5 > Mn/H-ZSM-5 > Cr/H-ZSM-5 > Cu/H-ZSM-5 > Ga/H-ZSM-5 > Mo/H-ZSM-5 > Fe/H-ZSM-5 > Co/H-ZSM-5 > W/H-ZSM-5 > Ag/H-ZSM-5 > H-ZSM-5 > Ni/H- ZSM-5. Another unique property observed in the use of Ni/HZSM-5 gave the lowest

120 98 CO 2 /CO ratio about This result indicates the considerable high formation of CO when Ni is applied as a modifier metal for H-ZSM-5 catalyst in the second layer catalyst. Figure 4.4 The ratio of CO 2 to CO over metal loaded H-ZSM-5 catalysts Yield of C 5+ Products Figure 4.5 shows the comparison of the yield of C 5 or higher hydrocarbons products over metal loaded H-ZSM-5 samples. From the results given in Figure 4.5, it is clear that the yield of C 5+ products is affected by the nature of the metal type

121 99 over HZSM-5. The yield of C 5+ product can be arranged in the order: Ni/H-ZSM-5 > Ga/H-ZSM-5 > H-ZSM-5 > Mo/H-ZSM-5 > W/H-ZSM-5 > Co/H-ZSM-5 > Ag/H-ZSM-5 > Cu/H-ZSM-5 > Cd/H-ZSM-5 > Cr/H-ZSM-5 > Mn/H-ZSM-5 > Fe/H-ZSM-5. As can be seen from Figure 4.5, introducing nickel into the H-ZSM-5 has significantly increased the yield C 5+ products. The yield of C 5+ products is about 6.2% when H-ZSM-5 catalyst is used, but once Ni is introduced this number jumps to 7.5%. Gallium loaded zeolite also showed a better performance than that of the non-loaded H-ZSM-5. Since the activity of H-ZSM-5 alone was low, Ni and Ga loaded H-ZSM-5 could be an important key component to increase yield of C 5+ in this dual-bed system. Figure 4.5 Yield of C 5+ products over metal loaded H-ZSM-5 catalysts

122 Discussion The catalytic activity of metal supported on zeolite catalysts depends on several factors. The type of metal, the degree of the metal dispersion, method of catalysts preparation and the acidity strength of the zeolite support has all been recognized as very important factors in catalysis (Okamoto, 1997; Ohman et al., 2002; and Akhmedov et al., 2003). Therefore, the present discussion is focused on these factors by reviewing some previous characterization studies and catalytic behaviour in order to elucidate reaction properties in the dual catalytic system when various types of metal loaded H-ZSM-5 are applied as second bed catalyst. A few papers of more general character are reviewed to begin with, before the main chapters dealing with the more specific items related to metal loaded zeolites and their trend on the catalytic reaction in this dual-bed system will be addressed. In general, in real catalytic systems, there are several states where the size of the metal species and the oxidation state of the ions differ. Therefore, adsorption studies as well as interaction energies and charge transfer studies are of particular interest in understanding the mechanisms of hydrocarbon transformation by metal loaded zeolite. A fairly large number of publications on this topic (Chatterjee and Vetrivel, 1996; Slinkin and Kucherov, 1997; Limtrakul et al., 2001; Viswanadham et al., 2004; and Marquis et al., 2005) demonstrates that considerable effort has been devoted to understand the hydrocarbon interactions with the different metal loaded zeolite systems and the elementary steps of the reaction mechanisms. Therefore, to explain the reaction or in other words to correlate the activity order in this dual bed system, it is necessary to consider the electronic factors affecting the activity of different metal loaded zeolites. Likewise, pioneering studies on the zeolites by XPS performed by Shpiro et al. (1994) and Barr (1990) concluded that electron transfer occurs between the metallic clusters and the zeolitic support, resulting in strong electron donor properties of the support and this responds to structural changes in the surface region of the zeolite samples which depend on the type of metal incorporated. Furthermore, several metal loaded H-ZSM-5 (Cu 2+, Mn 2+, Cr 3+, Fe 3+, V 4+, Cr 5+ and Mo 5+ ) were investigated by Slinkin and Kucherov (1997. Catalytic oxidation of alkanes traces

123 101 with oxygen is enhanced using cation-containing H-ZSM-5 zeolites instead of H- ZSM-5 and the increased activity was due to an electronic metal-support effect. The authors found that the system permits to compare intrinsic activities of different ions or of the same ions in different local environments and concluded that the electronic structure of zeolites depended strongly on both the composition and the cation involved. In some cases, these electronic properties are found to alter various aspects of the adsorptive and catalytic properties of the metal loaded H-ZSM-5 catalysts. Stocker (1996) had shown that the characteristics of zeolites and related materials in relation to the surface properties and their redox behavior are primarily depend on the nature and electronic state of metals and finally concluded the heteroatoms help in fine-tuning the strength of the acid sites and in introducing bifunctional features to zeolite catalysts. On the other hand, Viswanadham et al. (2004) observed that acidity as well as the different electronic environments present in metal loaded in zeolite plays the key role in the transformation of organic molecule. Goursot et al. (2003) explained that the change in the type of the metal support interface was considered to be responsible for the observed changes in the catalytic and chemisorption properties of the metal and also demonstrate very well the role of electronic structure calculation in explaining the catalytic reaction. Therefore, it is likely that the electronic configuration of the metallic ion change the property of the zeolite surface. It is now to explain the reactivity order, we felt from some earlier finding that influence of metal electronic interaction of between framework and organic molecule may play key role in the behaviour of the corresponding metal loaded zeolite to promote the types of reaction mechanisms. These differences we believed mainly caused the diverse distribution of selectivity of the products when tested with different metal modifier in this dual bed catalytic system. Apparently, the location, e.g. inside or outside the zeolite, of finely dispersed metallic species are of significant importance to catalyst performances for catalytic reactions and could be another reason for variation in products in this dual bed system when H-ZSM-5 is modified with different metals. According to literature, the metal phase being introduced by incipient impregnation is found to be restricted on the outer surface of zeolite crystals and only small amounts of metal are found

124 102 inside the zeolite cages. The metal cations are too large to gain access to the exchangeable protons of Bronsted sites inside the zeolite pores during impregnation step and the population of metal particles within the channels in the impregnated sample is lower than in the ion exchanged one (Arcoya et al., 2003). On the other hand, Kucherov et al., (1989) observed that the impregnated zeolites had lower specific surface area than the zeolites prior to impregnation. This confirmed the expectation that the impregnated metal presumably blocked some of the pores of the zeolite network. Hu and Ruckenstein (1999) concluded that the concentrations of the metals in the surface layer and the bulk zeolite material are significantly different, with higher transition metal concentrations in the surface. Conversely, Slinkin and Kucherov (1997) found that metal particles could be able to penetrate into the internal void space of the zeolite and the preferred locations lay in straight channels in the vicinity of the intersection with zigzag channel. Additionally, the bigger metal particles may block the accessibility to the Bronsted acid sites. This appears to be an important mode of deactivation, reducing the available metal surface area, both by virtue of the lower surface/volume ratio of larger particles and also by filling the zeolite channels and blocking the orifices and cavities access. Moreover, Xiao et al. (1998) had demonstrated that the salts dispersed inside of zeolite pores and they could design the pore size of zeolites to various degrees by dispersing various loadings of inorganic salts into zeolites, which are very important for the application on catalysis by zeolites because the reactants required suitable pores of zeolites. On the other hand, the acidity of these zeolites modified with metal should change accordingly (Qian and Yan, 2001). Furthermore, by the combination of MAS NMR, ESR and TPD experiments, Zhou et al. (2001) have proved that molybdenum do imagrate into the internal channels of H-ZSM-5 during the impregnation and calcinations processed, leading to a decrease of the amount of Bronsted acid sites and the appearance of super-hyperfine structure in the Mo 5+ signal in ESR measurements. There is clear evidence that during the high temperature pretreatment, small amounts of molybdenum initially present on the external surface of the zeolite will migrate into the zeolite pores, and react with Bronsted acid sites to form bonded Mo species. The mobile Mo species must be small enough to penetrate the zeolite pores, thus, it

125 103 should be in a somewhat lower molecular weight form than the heptamolybdate ion. This is also consistent with the conclusion reported previously that Bronsted sites serve as a tray for the migration of metal ions into the zeolitic channels (Kucherov et al., 1997) which, in the end, can be taken as the general feature of zeolite-supported metal bifunctional catalysts. In addition, modification with some metal could generate new acid sites over zeolite matrix and this effect seems to have a great influence on the total acid site density and catalytic performance over the metal loaded zeolite catalyst. However, this acid characteristic of the metal loaded zeolite system is affected by the nature of the metal components. Therefore, in this study, different types of metals used to modify the H-ZSM-5 catalyst would generate different acid environments, which introduce into different types of reaction mechanisms. With this idea in mind, thus, the results in this dual bed system did not always point in the same direction, but depends on the type of metals applied for modifying the H-ZSM-5 zeolite support. Based on the earlier discussions from the previous studies, some important suppositions can be drawn. The variation of the products in the dual bed system was influenced by electronic properties, locations, different acid environments and surface properties of the metal used to modify the second bed H-ZSM-5 catalyst. In addition, the information obtained from the activity and selectivity of products distribution in the dual bed catalytic system by using metal loaded H-ZSM-5 also gives valuable information on the functionalities of the acid protons and metal species. Based on the results, one can indicate that the metal cations in zeolite topology play a significant role on both the rate of intermediate OCM products depletion and higher hydrocarbon products formation. This suggests that the metal species and channel structure participate concurrently on the reactivity for transformation of intermediate OCM products in the presence of oxygen into C 5+ range liquid hydrocarbon products. It appears that the selectivity of products is different between H-ZSM-5 and metal loaded H-ZSM-5, depending on the nature of the metal used during the reaction. In an effort to understand the nature of metal sites operating during the transformation OCM intermediate products in this dual bed catalytic system, more elaboration and discussion on the basis of some other previous findings and the results of the present catalytic tests will be discussed further.

126 104 It was established in this study that activity and selectivity of the H-ZSM-5 containing monometallic catalytsts undergo significant changes depending on the nature of the metal phase on the catalyst surface. In view of the knowledge that unpromoted H-ZSM-5 is not the best dehydrogenation catalyst, addition of promoters is required for such application (Creyghton and Downing, 1998). The combination of H-ZSM-5 with some metal species have high activity for the dehydrogenation of paraffins but are also an active hydrogenolysis catalyst, while some are assumed to have moderate or low activity in hydrogenolysis (Rice and Keptner, 2004; Araujo and Schmal, 2002; Bond and Hooper, 2000; Blowers and Masel, 1998; and Jentys et al., 1997). Defined by Bond et al. (1996) as a structure sensitive reaction, the specific rates of alkane hydrogenolysis reactions highly dependent on type of metal and the metal particle size of the active phase. Jackson et al. (1998) also claimed that the metals W, Fe and Co all have a high tendency for total degradation to methane. This depth of hydrogenolysis, however, has no relation to the selectivity of the reaction, that is, the tendency for particular bonds to be broken (Jackson et al., 1998). It was concluded that under hydrogenolysis conditions, other reactions such as isomerisation, oxidation etc. of alkanes can also take place. Therefore, the type of metal function of zeolite supported catalysts has a large influence on the hydrogenolysis characteristics. However, in this dual bed catalyst, the hydrogenolysis has to be minimised to achive a better selectivity of production of desired C 5+ hydrocarbon products. Considering the above, we believe that the small amounts of methane arose from hydrogenolysis, which took place on the metal sites caused in drop off net conversion of methane. Therefore, the observed small increase and decrease of methane conversion in metal exposure of the H-ZSM-5 catalysts (Figure 4.1), respectively, as compared with the unloaded H-ZSM-5 is consistent with the order of impregnation of the metal phase in each catalyst. It is therefore proposed that the change in hydrogenolysis activity is related to the change in the structure of the metal-zeolite support interface and the reactive catalyst surface. The hydrogenolysis reaction is therefore generated in situ and contain a carbonaceous component (Solymosi and Szoke, 1998 and; Bond and Slaa, 1996). Results obtained at a lower

127 105 conversion level for some type of metals confirmed the possibility of hydrogenolysis inhibition by means of metallic dispersion increment, promoted by metal addition to H-ZSM-5. The presence of metal sites on the exterior surface of the zeolite also changes in the reaction mechanism, going from an acidic to a bifunctional mechanism (Iglesia et al., 1997). These changes are the result of the dehydrogenating activity. The literature has reported that the effect of a metal in the metal/h-zsm-5 system was beneficial for ethane conversion or the C 5+ range hydrocarbons formation. In the case of ethane, however, reaction rates remain low over undoped H-ZSM-5 (Cheung and Gates, 1997). Here, doping the zeolite to catalyze the initial dehydrogenation to ethylene is absolutely mandatory. Furthermore, the dopant generally promotes direct dehydrogenation of hydrocarbons in all stages of the oligomerization-aromatization reaction with accompanying formation of molecular hydrogen. To support this idea, we cite results from a study of the reaction of ethane catalyzed by molybdenum compounds deposited on ZSM-5 which has been investigated at K (Solymosi and Szoke, 1998). That investigation emphasized that the role of the Mo compounds is mainly the activation of ethane and the aromatization-oligomerization of ethylene proceed on the acidic sites on ZSM-5. In accord once with the data, they found that the dehydrogenation of ethane is the main reaction on the metal surface, C 2 H 6 C 2 H 4 + H 2 (4.1) and also the side reaction at a higher temperature is the formation of CH 4, C 2 H 6 CH 4 + H 2 + C (4.2) C 2 H 6 + H 2 2CH 4 (4.3) that come into prominence. In addition to reaction 4.2, the decomposition of ethylene C 2 H 4 CH 4 + C (4.4)

128 106 also contributes to the deposition of surface carbon. In the light of their results, they concluded that deposition of MoO 3 on ZSM-5 improved the catalytic performance of ZSM-5 and carbon formed in the reaction of ethane poisons the acidic sites of ZSM- 5 and prevents the oligomerization and aromatization of ethylene. This point will be discussed further later. As mentioned above, it is important to emphasize that the role of the metal compounds is mainly the activation of ethane and the promotion of the formation of ethylene. Further reactions, namely the oligomerization and the aromatization of ethylene proceed on the acidic sites on ZSM-5. Choudhary et al. (2001) described the role of bifunctionality of the monometallic deposited H-ZSM-5 catalysts, where the hydrogenation/dehydrogenation activities are provided by the metallic sites, whereas the oligomerization, isomerization and aromatization activity is provided by the acidic sites of the catalysts. Based on this idea, it is believed that a certain amount of ethane, which was produced in the first bed, will convert to ethylene when metal cations were introduced on the zeolite surface. However, according to the quoted authors, the rate of formation of this reaction was dependent on the nature of metallic sites. Hence, the changes in selectivity towards ethane (Figure 4.2) in the presence of some metal cation obtained from this study can also be indirectly related to this case. In the light of these results, it is interesting that a relatively large amount of ethylene is released from H-ZSM-5 based catalysts or, in other words, a large fraction of ethylene produced escape oligomerization-aromatization (Figure 4.2). This suggests that the carbon formed in the reaction poisons the acidic sites of H-ZSM-5 and prevents the oligomerization-aromatization of ethylene. It is a common problem with zeolite catalysts used in hydrocarbon conversion is deactivation by coke deposits which are formed from undesirable side reactions. The chemical nature of the carbon deposits has been concluded to depend very much on how they are formed, the conditions of temperature and pressure, the age of the catalyst and the chemical nature of the feed and products formed (Praserthdam and Ayutthaya, 2004; Guo et al., 1998; and Pradhan et al., 1997). In the same way, Sahoo et al. (2001) and Pradhan et al. (1997) concluded that the rate of coke formation is also strongly dependent on the zeolite pore structure and sensitivity of

129 107 metal modifier and very closely related to the acidity of the catalyst. The trend of carbon formation can be explained merely based on total acidity of the metal loaded H-ZSM-5 samples in this study, but further, more thorough studies on this problem are beyond the scope and limits of the present work. However, based on Nagamori and Kawase s (1998) investigation, it is well recognized that deactivation of zeolite catalyst for transformation of lower hydrocarbons into gasoline and aromatics is attributed to poisoning of the external surface by coke deposition, accompanied by the restricted access of reactants to inner acidic sites by pore blocking. On the other hand, the selectivity of the C 4 products distribution also changed (Figure 4.3) when tested with series of bifunctional catalyst in which metals are deposited on the H-ZSM-5. Although present data do not allow us to clarify the reasons, some hypotheses may be advanced. Bokhoven et al. (2004) and Boudart, (1997) claimed that these differences are certainly related to differences in the catalytic behaviour of metallic sites, in the significance in turnover frequencies of C 2 coupling reaction. This latter effect, already shown by various authors (Ivanov et al., 1999, Dehertog and Fromen, 1999; Nascimento et al., 1999; and Li and Armor, 1998), could be related to the ability and capacity of hydrocarbon mainly ethane and ethylene to adsorb on the metal surface. The participation in C 2 coupling reaction is to reactive some metals located near the acid sites, followed by blockading in zeolite pores attributed to increase in C 4 selectivity in the final products stream. Whereas the much higher C 4 formation activity of Cr/H-ZSM-5 is clearly related to its lower oligomerizing activity. It is more difficult to explain the origin of the differences between C 4 formation/ oligomerization rate reaction found with Cr/H-ZSM-5. However, complementary experiments are needed to specify the origin of these differences. Moreover, for the catalysts being considered in this fragment, the product spectra clearly showed low production of C 3 products (Figure 4.3). The intricate interplay between metal surface and gas-phase reactions is also demonstrated for the case of OCM final products mainly ethylene and ethane oxidation. Qian and Yan (2001) and Li (2003), claimed that the incorporation of metal led to the formation of some metal oxides clusters and probably outside the zeolite surface. Panov et al. (1998) illustrated that the presence of metal oxides on zeolite surface increases the amount of oxygen on zeolite surface. It is attributed to

130 108 the high reactivity of the oxygenated radicals on the metal surface. Some researchers reported that the presence of transition metal oxide on the surface increased the concentration of surface oxygen and hence the oxidative reaction is facilitated (selectivity of CO x is increased)(neyestanaki et al., 1995; Panov et al., 1998; Yoo 1998; and Stockenhuber et al., 2001). This explained that insertion of the external metal surface by impregnation on the H-ZSM-5 surface could increase in amount of oxygen on zeolite surface and contribute in oxidation of intermediate products to CO x to some extent. However, the oxygen absorbed on the surface was varied depending on the nature of the metals. We can therefore conclude that the surface of some metal compounds are capable of changing the rate of oxidation of the intermediate hydrocarbons products (Martin and Mirodatos, 1995). They can either accelerate the gas-phase oxidation or inhibit it, depending on the individuality of the metal. This background varied the amount of carbon oxides (CO and CO 2 ) formation in this dual-bed system where it is dependent on the kind of metal modifier which used to construct the second bed H- ZSM-5 catalyst (Figure 4.4). Therefore, the properties of the metal loaded H-ZSM-5 surface seem to be an important factor in determining the catalytic activities of these zeolites in the oxidation of hydrocarbon in this dual bed system. However, it is emphasized again that the existence of the metal oxides also helps the oxidation of light hydrocarbon to radical species and triggers other reactions such as oligomerization and dehydrogenation (Baerns and Buyevskaya, 1998; Banares,1999; Zhang et al., 2002; Hollens et al., 2003; and Heracleous and Lemonidou, 2004). When the metal is introduced on the surface of H-ZSM-5, a significant amount of C 5+ products was produced. Amin and Anggoro (2003) reported that the optimal oligomers yield resulted from a good balance between the dehydrogenation function of the metal and the acid function of H-ZSM-5. Although the yield of C 5+ in the products stream was lowered when metal cations were introduced to zeolite, a significant increase in the production of C 5+ has been used achieved when Ni and Ga loaded H-ZSM-5 was applied. Furthermore, the results of Figure 4.5 clearly show that the catalytic activity to form C 5+ products strongly depend on the identity of the metal which used to modify the second bed of H-ZSM-5 catalyst.

131 109 Based on the above results, combined with the OCM process, an overall simplified mechanistic scheme for the main reactions is depicted in Figure 4.6 to better rationalize the reaction pathways of intermediate OCM products involved in this dual bed catalytic system when tested with different metal loaded H-ZSM-5 as a second bed catalyst. With such a reaction network, one can expect that the complex reactions take place in this dual-bed catalytic system. 4.5 Summary Based on this observation, it is reasonable to conclude that the differences in the structure, electronic properties, dispersion, chemisorption and the activity for oligomerization and hydrogenolysis described above can partially explain the change in the trend of selectivity of products of the metal-zeolite supported catalyst because the catalysts sequence for these characterization parameters may differs depending on the kind of metal species. These observation and the fact that catalyst acidity and activity in this dual bed catalyst are not only determined by the type of the metal dispersion nor the acidity, but by a combination of all of these factors. Therefore, by incorporating proper additives metal, the concentration and intensity of the surface acid sites could be refined and controlled in order to maximize the formation of the C 5+ products in the dual bed catalytic system. Furthermore, nickel was selected as a suitable metal for modifying the H-ZSM-5 catalyst for application in the dual bed system as the second bed catalyst because it exposes a special catalytic characteristic compared to other metals tested through this investigation. This is consistent with some earlier reports that Ni/H- ZSM-5 is an effective catalyst for oligomerizing ethylene and other light alkanes. Therefore, Ni/H-ZSM-5 samples are selected to do further investigation in Chapter 5 to understand the role of metal ions and their relative contribution to the various reaction steps involved in the conversion of methane in the dual bed catalytic system.

132 110 METHANE OXYGEN FIRST BED CATALYST La-MgO First Stage Reactions SECOND BED CATALYST Metal Loaded HZSM-5 Dehydrogenation/ Hydrogenation Oligomerization Oxidation Hydrogenolysis/ Cracking Second Stage Reactions Isomerization Aromatization FINAL PRODUCTS Figure 4.6 Simplified reaction scheme for the dual bed catalytic system