Catalytic Reactors 9783110332964

Table of contents :
Cover
Half Title
Also of Interest
Catalytic Reactors
Copyright
List of contributing authors
About the editor
Preface
Contents
1. Catalysis in Multifunctional Reactors
1.1 Introduction
1.2 Reactive Distillation (RD)
1.2.1 Homogeneous catalysis
1.2.2 Heterogeneous catalysis
1.2.3 Catalysts used in reactive distillation
1.3 Reactive Stripping
1.3.1 Esterification
1.3.2 Aqueous phase reforming (APR) of sorbitol
1.3.3 Dehydration of xylose to furfural
1.3.4 Catalytic exchange of hydrogen isotopes
1.4 Catalytic membrane reactors
1.4.1 Biodiesel production
1.4.2 Dehydrogenation
1.4.3 Oxidative coupling of methane (OCM)
1.4.4 Partial oxidation of methane to synthesis gas
1.5 Chromatographic Reactor
1.5.1 Concept of a Chromatographic Reactor
1.5.2 Types of Chromatographic Reactor
1.5.3 Applications of Liquid Chromatographic Reactor
1.6 Summary
Acknowledgment
References
2. Biocatalytic membrane reactors (BMR)
Nomenclature
2.1 Introduction
2.2 Role of membrane in biocatalytic membrane reactors (BMRs)
2.3 Membrane separation reactors (MSRs)
2.3.1 Concept
2.3.2 Application
2.4 Membrane aeration bioreactors (MABR)
2.5 Extractive membrane bioreactors (EMBR)
2.5.1 Concept
2.5.2 Application
2.6 Enzyme immobilization techniques in membrane reactor systems
2.6.1 Physical adsorption
2.6.2 Entrapment
2.6.3 Cross-linking
2.6.4 Encapsulation
2.6.5 Segregation by membranes
2.6.6 Covalent binding
2.7 Laminated (multilayer) enzyme membrane reactors
2.7.1 Concept
2.7.2 Application
2.8 Biphasic (multiphase) membrane bioreactors
2.8.1 Concept
2.8.2 Application
2.9 Phase transfer catalysis in multiphase membrane reactors
2.9.1 Concept
2.9.2 Application
2.10 Conclusions
References
3. Metallic nanoparticles made in flow and their catalytic applications in micro-flow reactors for organic synthesis
3.1 Introduction
3.2 Metal nanoparticles in a microfluidic reactor
3.2.1 Gold
3.2.2 Silver
3.2.3 Palladium
3.2.4 Platinum
3.2.5 Copper
3.3 Metal nanoparticles in a millifluidic reactor
3.4 Outlook – metal nanoparticles generated in flow and used in situ
3.5 Conclusions
Acknowledgment
References
4. Application of multi-objective optimization in the design and operation of industrial catalytic reactors and processes
4.1 Introduction
4.2 Multi-objective optimization
4.2.1 Concept of multi-objective optimization
4.2.2 MOO methods
4.3 No-preference methods
4.3.1 Neutral compromised solution
4.4 A priori methods
4.4.1 Method of Weighted global criterion
4.4.2 Lexicographic method
4.4.3 Goal Programming (GP)
4.5 A posteriori methods
4.5.1 ε-Constraint Method
4.6 Interactive methods
4.7 Genetic algorithms
4.7.1 About binary-coded variables
4.7.2 Simple Genetic Algorithm (SGA)
4.7.3 Use of GA in MOO
4.7.4 Constraint handling in GA
4.8 Simulated annealing
4.9 MOO problems in chemical engineering
4.9.1 Petroleum Processing Engineering
4.9.2 Steam Reforming
4.9.3 Polymer industry
4.10 Conclusions
References
5. Design of catalytic micro trickle bed reactors
Nomenclature
5.1 Introduction
5.2 Hydrodynamics
5.2.1 Flow regimes
5.2.2 Pressure drop
5.2.3 Liquid holdup
5.2.4 Flow maldistribution and start-up effects
5.2.5 Axial dispersion
5.3 Mass and heat transfer in micro trickle bed reactors
5.3.1 Mass transfer
5.3.2 Heat transfer
5.3.3 Scale up
5.4 Periodic operation
5.5 Applications
5.5.1 Micro trickle bed reactors
5.5.2 Semi-structured and structured trickle bed reactors
5.6 Outlook
References
6. Three-phase catalytic reactors for hydrogenation and oxidation reactions
Nomenclature
6.1 Introduction
6.2 Slurry Reactors
6.2.1 Theory: Determination of Controlling Resistance
6.2.2 Mixing and mass transfer in the stirred tank slurry reactor
6.2.3 Hydrogenation reactions in the stirred tank slurry reactor: kinetics and effect of operating variables
6.2.4 Catalysts for hydrogenation reactions: overview and novel biomass supported metal catalysts
6.2.5 Oxidation reactions in the slurry reactor
6.2.6 Bubble column reactors
6.3 Trickle bed reactors
6.3.1 Theory and flow regimes
6.3.2 Overall rate model
6.3.3 Imaging of gas-liquid flows
6.3.4 Scale up and modeling
6.3.5 Enantioselective Hydrogenation reactions
6.3.6 Industrial applications in heavy oil upgrading
6.4 Structured monolith reactors
6.4.1 Flow patterns in the single capillary
6.4.2 Applications in hydrogenation reactions
6.5 Conclusions
References
7. Design and modeling of laboratory scale three-phase fixed bed reactors
Nomenclature
7.1 Background
7.2 Reactor set- up
7.3 Physical and chemical phenomena in fixed bed reactors
7.3.1 Overview
7.4 Research targets and topics
7.4.1 Catalyst selection
7.4.2 Reaction kinetics
7.4.3 Mass transfer effects
7.4.4 Heat effects
7.4.5 Physical properties of gases and liquids
7.4.6 Reactor design and operation policy
7.4.7 Modeling options
7.5 Experimental design
7.5.1 Targeted products
7.5.2 Catalyst screening
7.5.3 Experimental productivity and selectivity optimization
7.5.4 Particle geometry
7.5.5 Feed distribution and flow regimes
7.5.6 Bed dilution
7.5.7 Residence time distribution
7.6 Chemical kinetics
7.6.1 Topics of the kinetic studies
7.6.2 Reaction scheme simplifications
7.6.3 Rate expressions
7.6.4 Qualitative reaction rate comparison: fixed bed against batch reactors
7.6.5 Catalyst deactivation
7.6.6 Truly intrinsic kinetics
7.7 Mass and heat transfer effects
7.7.1 Mass transfer resistances
7.7.2 Gas-liquid mass transfer
7.7.3 Liquid-solid mass and heat transfer
7.7.4 Internal diffusion – pore diffusion
7.7.5 Effectiveness factors for particles
7.8 Physical properties of gas mixtures and solutions
7.8.1 Density and viscosity
7.8.2 Diffusivity
7.8.3 Gas solubility
7.8.4 Thermal conductivity
7.8.5 Reaction enthalpy
7.9 Liquid flow effects
7.9.1 Qualitative flow arrangement comparison
7.9.2 External wetting of the catalyst
7.9.3 Radial flow
7.9.4 Pressure drop
7.9.5 Liquid saturation (hold-up)
7.10 Reactor modeling steps
7.10.1 Overview
7.10.2 Selected modeling policy
7.10.3 Studies of simplified reaction systems
7.10.4 Ways how to rule out phenomena by experimental design
7.10.5 Sensitivity studies
7.11 Balances for the generic three-phase fixed bed model
7.11.1 Mass balances for gas, liquid and solid phases
7.11.2 Energy balances for gas-, liquid- and solid phases
7.11.3 Boundary conditions
7.11.4 Sub-model examples
7.12 Axial dispersion modeling and experiments
7.12.1 Classical axial dispersion model
7.12.2 Alternative modeling approaches for back-mixing
7.13.1 Model classification
7.13.2 Benefits of dynamic models
7.13.3 Solvers and solution algorithms
7.13.4 Numerical method of lines
7.13.5 Hoyos method for particles
7.13.6 Parameter optimization methods
7.13.7 Parameter number reduction
7.14 Scale-up issues of fixed beds
7.14.1 Overview
7.14.2 Large scale operation
7.14.3 Gradients in scale up
7.14.4 Flow regime in scale-up
7.14.5 Back mixing in scale-up
7.15 Examples
7.15.1 Citral hydrogenation
7.15.2 Direct synthesis of hydrogen peroxide
7.16 Conclusions
Acknowledgment
References
Index

Citation preview

Basudeb Saha (Ed.) Catalytic Reactors De Gruyter Graduate

Also of Interest

Engineering Catalysis Murzin, 2013 ISBN 978-3-11-028336-5, e-ISBN 978-3-11-028337-2

Flow Chemistry: Vol. 1 Fundamentals Darvas, Hessel, Dorman, 2014 ISBN 978-3-11-028915-2, e-ISBN 978-3-11-028916-9

Flow Chemistry: Vol. 2 Applications Darvas, Hessel, Dorman, 2014 ISBN 978-3-11-036707-2, e-ISBN 978-3-11-036750-8

Process Technology: An Introduction De Haan, 2015 ISBN 978-3-11-033671-9, e-ISBN 978-3-11-033672-6

Enantioselectice Catalysis: From Principles to Industrial Production Goldfuß, 2016 ISBN 978-3-11-034040-2, e-ISBN 978-3-11-038959-3

Reviews in Chemical Engineering Dan Luss, Neima Brauner (Editors-in-Chief) ISSN 0167-8299, e-ISSN 2191-0235

Catalytic Reactors

| Edited by Basudeb Saha

Editor Professor Basudeb Saha School of Engineering London South Bank University 103 Borough Road London, SE1 0AA UK

ISBN 978-3-11-033296-4 e-ISBN (PDF) 978-3-11-033298-8 e-ISBN (EPUB) 978-3-11-039012-4 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2016 Walter de Gruyter GmbH, Berlin/Boston Cover image: Eric Middelkoop Typesetting: PTP-Berlin, Protago-TEX-Production GmbH, Berlin Printing and binding: CPI books GmbH, Leck ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

List of contributing authors Volkan Degirmenci School of Engineering University of Warwick Coventry, CV4 7AL, UK [email protected] Chapter 5 Kari Eränen Department of Chemical Engineering PCC Åbo Akademi FI-20500 Turku/Åbo, Finland [email protected] Chapter 7 Volker Hessel Laboratory of Chemical Reactor Engineering/ Micro Flow Chemistry and Process Technology Department of Chemical Engineering and Chemistry Eindhoven University of Technology P.O. Box 513, 5600 MB Eindhoven The Netherlands [email protected] Chapter 3 Stanislav Y. Ivanov Department of Chemical and Biochemical Engineering University of Western Ontario London, ON N6A5B9, Canada [email protected] Chapter 4 Teuvo Kilpiö Department of Chemical Engineering PCC Åbo Akademi FI-20500 Turku/Åbo, Finland [email protected] Chapter 7

Sanjay M. Mahajani Department of Chemical Engineering Indian Institute of Technology Bombay, Powai Mumbai 400 076, India [email protected] Chapter 1 Timothy Noël Laboratory of Chemical Reactor Engineering/ Micro Flow Chemistry and Process Technology Department of Chemical Engineering and Chemistry Eindhoven University of Technology P.O. Box 513, 5600 MB Eindhoven The Netherlands [email protected] Chapter 3 Ajay K. Ray Department of Chemical and Biochemical Engineering University of Western Ontario London, ON N6A5B9, Canada [email protected] Chapter 4 Evgeny V. Rebrov School of Engineering University of Warwick Coventry, CV4 7AL, UK and Department of Biotechnology and Chemistry Tver State Technical University A. Nikitina str., 22 Tver, 170026, Russia [email protected] Chapter 5 Vincenzo Russo Chemical Sciences Department University of Naples ‘Federico II’ IT-80126 Naples, Italy [email protected] Chapter 7

VI | List of contributing authors

Basudeb Saha School of Engineering London South Bank University 103 Borough Road London, SE1 0AA, UK [email protected] Chapter 1

Goran T. Vladisavljević Chemical Engineering Department Loughborough University Loughborough Leicestershire, LE11 3TU, UK [email protected] Chapter 2

Tapio Salmi Department of Chemical Engineering PCC Åbo Akademi FI-20500 Turku/Åbo, Finland [email protected] Chapter 7

Qi Wang Laboratory of Chemical Reactor Engineering/ Micro Flow Chemistry and Process Technology Department of Chemical Engineering and Chemistry Eindhoven University of Technology P.O. Box 513, 5600 MB Eindhoven The Netherlands [email protected] Chapter 3

Elnaz Shahbazali Laboratory of Chemical Reactor Engineering/ Micro Flow Chemistry and Process Technology Department of Chemical Engineering and Chemistry Eindhoven University of Technology P.O. Box 513, 5600 MB Eindhoven The Netherlands [email protected] Chapter 3

Joseph Wood School of Chemical Engineering University of Birmingham Edgbaston, Birmingham, B15 2TT, UK [email protected] Chapter 6

About the editor Basudeb Saha is Professor of Chemical and Process Engineering and Head of Energy and Environment Research Group at London South Bank University (LSBU), UK. He is founding Director of the Centre for Green Process Engineering at LSBU. Previously (1997–2010), he worked at Loughborough University, UK as a Research Associate, Lecturer, Senior Lecturer and a Reader in Chemical Engineering. He has a proven track record of delivery on research contracts in terms of the scientific and training aspects, has established collaborative research programmes with several leading international research groups, and his services are in demand for lecture, peer review, examination, and teaching activities, both in the UK and overseas. He is recognised as a leading expert in his field and has been appointed as a Visiting Professor in the University of Barcelona, Spain; Saga University, Japan and University of Burgos, Spain. He has published over 150 research papers (including heavily cited research articles), book chapters and several patents. He has received many invitations to visit laboratories overseas and to give invited/keynote lectures at national and international conferences. He is a winner of The Royal Society Brian Mercer Award. He was an Associate Editor of The Canadian Journal of Chemical Engineering (2009–2012) and is currently an International Advisory Board Member of The Canadian Journal of Chemical Engineering (2013–present). He is an Editor of Green Processing and Synthesis Journal (2012–present) and an Editorial Board Member of Process Safety and Environmental Protection Journal (2014–present). He is a Fellow (FIChemE) of The Institution of Chemical Engineers and a Senior Fellow (SFHEA) of The Higher Education Academy.

Preface The only true wisdom is in knowing you know nothing. Socrates (470–399 B.C.)

Catalytic reactors have numerous applications in the production of petrochemicals, bulk chemicals, pharmaceuticals and speciality chemicals. They are at the heart of producing almost all industrial chemicals. The use of catalytic reactor technology is essential for the economic viability of the chemical manufacturing industry. Appropriate selection of a reactive system that operates in the most efficient and benign manner can be crucial to the economic success of a chemical plant. This book aims to collate into a comprehensive and well-informed work of leading researchers and focuses on the state-of-the-art applications of novel catalytic reactors. The book contains seven comprehensive chapters encompassing a wide gamut of topics – they truly reflect the diversity in the field of novel catalytic reactors and processes. This graduate level text book is intended for students, scientists and practising engineers. The level of application will depend on the choice of chapters and the materials to be covered. A multifunctional reactor is broadly defined as a multifaceted reactor system that combines a conventional reactor with any physical process to enhance the overall performance of the process to bring cost-effectiveness and/or compactness to a chemical plant. This multi-functionality can exist either on micro (catalyst) level or on macro (reactor) level. There is substantial information available on several ways to achieve this task. Combining reaction with separation is one such popular approach. Chapter 1 reviews the recent literature on catalysts and their modified forms used in multifunctional reactors that combine reaction and separation. This chapter is mainly focused on four of the most studied multifunctional reactors: reactive distillation, reactive stripping, membrane reactor and chromatographic reactor. Chapter 2 describes basic concepts, operation principles, and applications of biocatalytic membrane reactors (BMRs) in versatile fields ranging from treatment of wastewaters and waste vapour streams to resolution of amino acids and enzymatic hydrolysis of triglycerides and biomacromolecules. BMRs are multifunctional devices in which biochemical reactions catalysed by enzymes or cells are integrated with mass transfer through a membrane. The most common techniques of enzyme immobilisation on/in polymeric and inorganic membranes have been reviewed including physical adsorption, entrapment within membrane pores, segregation, crosslinking and covalent binding. Novel microfluidic routes for preparation of enzyme-loaded microcapsules and microgels have also been discussed. Chapter 3 presents recent developments on the synthesis of noble metal nanoparticles in micro and millifluidic devices and their catalytic applications in organic flow synthesis. A variety of synthesis methods using microfluidics is presented for gold,

Preface

| IX

silver, palladium, platinum, and copper nanoparticles, including the formation in single-phase and multiphase flows. In the field of organic chemistry, metal nanoparticles can be used as catalysts. This can lead to remarkably improved reaction performance in terms of minimizing the reaction time and higher yields. In this context, various applications of those metal nanoparticles as catalysts in microfluidic devices are highlighted at selected examples. To underpin the capabilities of microfluidics and as a new operational window, nanocatalysts may be even synthesized in situ in flow (especially when difficult to make conventionally) and directly utilized in an organic synthesis – a kind of flashed nanocatalyst organic synthesis. Industrial chemical reactors are very intricate and expensive units. It is of high importance to know the best operation and/or design parameters in order to provide economically efficient manufacturing of final product. Moreover, it is also important to account for safety and environmental aspects in such complex systems. Application of single objective optimization allows one to formulate the only objective function which relates overall process efficiency with its multiple parameters. However, in reality complex chemical processes require addressing multiple objectives that are time as well as site specific. Methods of multi-objective optimization offer a more rigorous approach to find optimum solutions of real-world industrial optimization problems that are more meaningful thereby overcoming the drawbacks of single-objective optimization. Chapter 4 is organized to cover the concept of multiobjective optimization (Pareto optimality) followed by mathematical definition and brief explanation of existing multi-objective methods especially emphasizing Genetic Algorithms (GAs) due to its robustness. Lastly, a number of recent applications of multi-objective optimization to different chemical reactors published in recent years are discussed. This chapter will be helpful for readers to understand the importance of application of multi-objective optimization and stimulates an interest in application to different chemical engineering problems. Small-scale and distributed chemical manufacturing systems offer significant advantages such as compactness of equipment, lower operating costs and reduced byproduct formation. Process intensification via development of compact and highly efficient reactors with channel diameters in the micrometre and millimetre range becomes a regular tool in multi-phase catalytic reaction engineering. These reactors have attracted interest for their ability to decrease physical limitations for heat and mass transfer. Chapter 5 illustrates the design of three phase reactors with characteristic channel dimensions in micro and millimetre range. Micro trickle bed reactors are examined in terms of their internal structure such as randomly packed microparticle beds; semi-structured reactors filled with foams and meshes and structured reactors made of periodic arrays of etched pillars. Engineering correlations for pressure drop, liquid hold-up, axial dispersion and heat and mass transfer are reviewed. Cyclic operation of micro trickle bed reactors and their applications in catalytic reactions are described.

X | Preface

Chapter 6 introduces three-phase reactors, which are widely used for carrying out a range of industrial catalytic reactions from hydrogenation to oxidation. Firstly, slurry reactors are discussed, encompassing stirred vessels and bubble columns. Mass transfer correlations for gassed vessels containing catalyst particles are presented and experimental measurements of mixing patterns and velocity maps in stirred slurry reactors by non-invasive imaging techniques are discussed. A detailed review of catalysts, effect of operating variables and mechanisms of hydrogenation reactions are considered, before describing research efforts to develop bio-metal nanoparticulate catalysts, which have been applied in reactions ranging from hydrogenation of alkynes to the Heck reaction. Oxidation reactions in the slurry reactor are next discussed, focussing in particular upon oxidation of secondary alcohols. The slurry bubble column reactor is reviewed as a case study, covering the study of hydrodynamics by non-invasive imaging and Computational Fluid Dynamics. Particular attention is paid to the Cocurrent Downflow Contactor Reactor, which has been utilised as a photocatalytic reactor for the oxidation of water pollutants. Trickle bed reactors are next described including a review of theory and flow regimes, followed by presentation of the overall rate model. A review of imaging of trickle beds is presented then scale up of trickle beds is discussed. Finally structured catalysts are discussed, focussing in particular upon monoliths. Studies of flow regimes in a single capillary and recirculation patterns in liquid slugs are presented and discussed before considering the range of potential applications of monoliths in three phase reactions such as the hydrogenation of alkynes. Chapter 7 deals with the design and modeling of laboratory scale three-phase fixed bed reactors. Continuous trickle bed reactors (TBRs) are practical for catalyst screening and especially well-suited for studying catalyst long-term activity. Catalyst development is conducted by screening potential candidates for catalyst/support and then by revealing the operating conditions that maximize the productivity and the selectivity. Catalyst selection, revelation of reaction kinetics as well as mass and heat transfer effects are discussed. Ways as how to minimize the effects of undesirable heat and mass transfer and flow related phenomena by proper selections of reactor design and operation conditions are also addressed. The mathematical models have their origins in mass and energy balances which are given in generic form. Momentum balances are often replaced by dispersion models and semi-empirical expressions for liquid hold-up and pressure drop. Sensitivity studies are conducted in order to test the reliability of the model and to get a feeling of how the system behaves. The solution of balance equations is based on numerical mathematics and fine-tuning is performed by parameter estimation. Qualitative reaction rate comparison (fixed bed against batch reactors), is presented because truly intrinsic kinetics is best revealed in batch reactors using highly intensive mixing and very small catalyst particles. Two modelling examples, citral hydrogenation and direct synthesis of hydrogen peroxide studies are discussed. The sensitivity of the product concentration on the changes of various key parameters is illustrated. Hydrogen peroxide production with the conven-

Preface

| XI

tional production technology is relatively expensive. Direct synthesis from hydrogen and oxygen using methanol as solvent and a selective catalyst is one of the attractive alternatives leading to a much simpler process. Parameter estimation results for direct synthesis and decomposition reactions are provided. The book has been written in a simple manner and the materials presented in this book should be readily understandable by graduate and postgraduate students. In closing, I would like to thank all of the individuals named as authors in this book for their valuable contributions. I would like to acknowledge the work of many postgraduate and post-doctoral researchers who have contributed to studies in my research group over a period of many years. I hope that you, the readers, enjoy tucking into this feast of world class catalytic reactors research and find answers to your questions. Finally, I would like to dedicate this book to my loving parents, and to my students whose enthusiasm and research accomplishments have been a continuing inspiration. Basudeb Saha

Contents List of contributing authors | V About the editor | VII Preface | VIII Sanjay M. Mahajani and Basudeb Saha 1 Catalysis in Multifunctional Reactors | 1 1.1 Introduction | 1 1.2 Reactive Distillation (RD) | 1 1.2.1 Homogeneous catalysis | 2 1.2.2 Heterogeneous catalysis | 3 1.2.3 Catalysts used in reactive distillation | 8 1.3 Reactive Stripping | 13 1.3.1 Esterification | 14 1.3.2 Aqueous phase reforming (APR) of sorbitol | 17 1.3.3 Dehydration of xylose to furfural | 17 1.3.4 Catalytic exchange of hydrogen isotopes | 18 1.4 Catalytic membrane reactors | 19 1.4.1 Biodiesel production | 20 1.4.2 Dehydrogenation | 24 1.4.3 Oxidative coupling of methane (OCM) | 25 1.4.4 Partial oxidation of methane to synthesis gas | 26 1.5 Chromatographic Reactor | 26 1.5.1 Concept of a Chromatographic Reactor | 27 1.5.2 Types of Chromatographic Reactor | 28 1.5.3 Applications of Liquid Chromatographic Reactor | 34 1.6 Summary | 41 Goran T. Vladisavljević 2 Biocatalytic membrane reactors (BMR) | 51 2.1 Introduction | 52 2.2 Role of membrane in biocatalytic membrane reactors (BMRs) | 52 2.3 Membrane separation reactors (MSRs) | 57 2.3.1 Concept | 57 2.3.2 Application | 60 2.4 Membrane aeration bioreactors (MABR) | 64 2.5 Extractive membrane bioreactors (EMBR) | 65 2.5.1 Concept | 65 2.5.2 Application | 67

XIV | Contents

2.6 2.6.1 2.6.2 2.6.3 2.6.4 2.6.5 2.6.6 2.7 2.7.1 2.7.2 2.8 2.8.1 2.8.2 2.9 2.9.1 2.9.2 2.10

Enzyme immobilization techniques in membrane reactor systems | 70 Physical adsorption | 72 Entrapment | 72 Cross-linking | 73 Encapsulation | 73 Segregation by membranes | 77 Covalent binding | 77 Laminated (multilayer) enzyme membrane reactors | 81 Concept | 81 Application | 83 Biphasic (multiphase) membrane bioreactors | 86 Concept | 86 Application | 88 Phase transfer catalysis in multiphase membrane reactors | 92 Concept | 92 Application | 93 Conclusions | 94

Elnaz Shahbazali, Volker Hessel, Timothy Noël, and Qi Wang 3 Metallic nanoparticles made in flow and their catalytic applications in micro-flow reactors for organic synthesis | 103 3.1 Introduction | 103 3.2 Metal nanoparticles in a microfluidic reactor | 105 3.2.1 Gold | 105 3.2.2 Silver | 114 3.2.3 Palladium | 117 3.2.4 Platinum | 120 3.2.5 Copper | 122 3.3 Metal nanoparticles in a millifluidic reactor | 123 3.4 Outlook – metal nanoparticles generated in flow and used in situ | 127 3.5 Conclusions | 128 Stanislav Y. Ivanov and Ajay K. Ray 4 Application of multi-objective optimization in the design and operation of industrial catalytic reactors and processes | 134 4.1 Introduction | 134 4.2 Multi-objective optimization | 136 4.2.1 Concept of multi-objective optimization | 136 4.2.2 MOO methods | 137 4.3 No-preference methods | 138 4.3.1 Neutral compromised solution | 138

Contents |

4.4 4.4.1 4.4.2 4.4.3 4.5 4.5.1 4.6 4.7 4.7.1 4.7.2 4.7.3 4.7.4 4.8 4.9 4.9.1 4.9.2 4.9.3 4.10

A priori methods | 139 Method of Weighted global criterion | 139 Lexicographic method | 140 Goal Programming (GP) | 141 A posteriori methods | 141 ε-Constraint Method | 141 Interactive methods | 142 Genetic algorithms | 142 About binary-coded variables | 143 Simple Genetic Algorithm (SGA) | 144 Use of GA in MOO | 145 Constraint handling in GA | 146 Simulated annealing | 148 MOO problems in chemical engineering | 148 Petroleum Processing Engineering | 149 Steam Reforming | 150 Polymer industry | 152 Conclusions | 168

Volkan Degirmenci and Evgeny V. Rebrov 5 Design of catalytic micro trickle bed reactors | 174 5.1 Introduction | 175 5.2 Hydrodynamics | 179 5.2.1 Flow regimes | 179 5.2.2 Pressure drop | 187 5.2.3 Liquid holdup | 191 5.2.4 Flow maldistribution and start-up effects | 193 5.2.5 Axial dispersion | 194 5.3 Mass and heat transfer in micro trickle bed reactors | 197 5.3.1 Mass transfer | 197 5.3.2 Heat transfer | 201 5.3.3 Scale up | 204 5.4 Periodic operation | 206 5.5 Applications | 207 5.5.1 Micro trickle bed reactors | 208 5.5.2 Semi-structured and structured trickle bed reactors | 212 5.6 Outlook | 212 Joseph Wood 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions | 220 6.1 Introduction | 222

XV

XVI | Contents

6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.2.5 6.2.6 6.3 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5 6.3.6 6.4 6.4.1 6.4.2 6.5

Slurry Reactors | 223 Theory: Determination of Controlling Resistance | 223 Mixing and mass transfer in the stirred tank slurry reactor | 226 Hydrogenation reactions in the stirred tank slurry reactor: kinetics and effect of operating variables | 235 Catalysts for hydrogenation reactions: overview and novel biomass supported metal catalysts | 238 Oxidation reactions in the slurry reactor | 245 Bubble column reactors | 248 Trickle bed reactors | 253 Theory and flow regimes | 253 Overall rate model | 255 Imaging of gas-liquid flows | 256 Scale up and modeling | 258 Enantioselective Hydrogenation reactions | 261 Industrial applications in heavy oil upgrading | 264 Structured monolith reactors | 267 Flow patterns in the single capillary | 268 Applications in hydrogenation reactions | 271 Conclusions | 274

Teuvo Kilpiö, Vincenzo Russo, Kari Eränen, and Tapio Salmi 7 Design and modeling of laboratory scale three-phase fixed bed reactors | 283 7.1 Background | 285 7.2 Reactor set- up | 287 7.3 Physical and chemical phenomena in fixed bed reactors | 288 7.3.1 Overview | 288 7.4 Research targets and topics | 289 7.4.1 Catalyst selection | 289 7.4.2 Reaction kinetics | 290 7.4.3 Mass transfer effects | 290 7.4.4 Heat effects | 290 7.4.5 Physical properties of gases and liquids | 290 7.4.6 Reactor design and operation policy | 290 7.4.7 Modeling options | 291 7.5 Experimental design | 291 7.5.1 Targeted products | 292 7.5.2 Catalyst screening | 292 7.5.3 Experimental productivity and selectivity optimization | 292 7.5.4 Particle geometry | 293 7.5.5 Feed distribution and flow regimes | 293

Contents |

7.5.6 7.5.7 7.6 7.6.1 7.6.2 7.6.3 7.6.4 7.6.5 7.6.6 7.7 7.7.1 7.7.2 7.7.3 7.7.4 7.7.5 7.8 7.8.1 7.8.2 7.8.3 7.8.4 7.8.5 7.9 7.9.1 7.9.2 7.9.3 7.9.4 7.9.5 7.10 7.10.1 7.10.2 7.10.3 7.10.4 7.10.5 7.11 7.11.1 7.11.2 7.11.3 7.11.4 7.12 7.12.1 7.12.2

Bed dilution | 294 Residence time distribution | 294 Chemical kinetics | 295 Topics of the kinetic studies | 295 Reaction scheme simplifications | 295 Rate expressions | 296 Qualitative reaction rate comparison: fixed bed against batch reactors | 297 Catalyst deactivation | 298 Truly intrinsic kinetics | 299 Mass and heat transfer effects | 299 Mass transfer resistances | 299 Gas-liquid mass transfer | 300 Liquid-solid mass and heat transfer | 301 Internal diffusion – pore diffusion | 302 Effectiveness factors for particles | 303 Physical properties of gas mixtures and solutions | 303 Density and viscosity | 303 Diffusivity | 304 Gas solubility | 305 Thermal conductivity | 305 Reaction enthalpy | 306 Liquid flow effects | 306 Qualitative flow arrangement comparison | 306 External wetting of the catalyst | 307 Radial flow | 307 Pressure drop | 308 Liquid saturation (hold-up) | 309 Reactor modeling steps | 311 Overview | 311 Selected modeling policy | 311 Studies of simplified reaction systems | 311 Ways how to rule out phenomena by experimental design | 312 Sensitivity studies | 312 Balances for the generic three-phase fixed bed model | 312 Mass balances for gas, liquid and solid phases | 312 Energy balances for gas-, liquid- and solid phases | 313 Boundary conditions | 314 Sub-model examples | 315 Axial dispersion modeling and experiments | 315 Classical axial dispersion model | 315 Alternative modeling approaches for back-mixing | 316

XVII

XVIII | Contents

7.13 7.13.1 7.13.2 7.13.3 7.13.4 7.13.5 7.13.6 7.13.7 7.14 7.14.1 7.14.2 7.14.3 7.14.4 7.14.5 7.15 7.15.1 7.15.2 7.16

Numerical strategies | 316 Model classification | 316 Benefits of dynamic models | 317 Solvers and solution algorithms | 317 Numerical method of lines | 318 Hoyos method for particles | 318 Parameter optimization methods | 319 Parameter number reduction | 319 Scale-up issues of fixed beds | 320 Overview | 320 Large scale operation | 320 Gradients in scale up | 321 Flow regime in scale-up | 321 Back mixing in scale-up | 321 Examples | 322 Citral hydrogenation | 322 Direct synthesis of hydrogen peroxide | 323 Conclusions | 327

Index | 333

Sanjay M. Mahajani and Basudeb Saha

1 Catalysis in Multifunctional Reactors 1.1 Introduction A multifunctional reactor is broadly defined as a multifaceted reactor system that combines a conventional reactor with any physical process to enhance the overall performance of the process to bring cost-effectiveness and/or compactness to a chemical plant. This multi-functionality can exist either on micro (catalyst) level or on macro (reactor) level [1]. There is substantial information available on several ways to achieve this task. Combining reaction with separation is one such popular approach. Here, when separation is performed in situ, several benefits like an increase in per-pass conversion and/or selectivity, energy integration, longer catalyst life, etc. are attained. When a separation process – e.g. distillation, adsorption, etc. – is to be performed simultaneously with a reaction, it imposes more restrictions on the reactor design so as to meet possible conflicting requirements that result from the reaction and separation. The existence of multiple phases as well as problems associated with heat and momentum transfer, mixing issues, etc. make the process complex, thereby attracting the attention of experts in reaction engineering, catalysis, modeling and simulation, and process design. Since catalysts are an integral part of a reactor system, many efforts have been made to manipulate its design to meet the above-mentioned challenges. A few examples are inserting special catalyst-filled envelopes into a distillation column to reduce pressure drop, manipulating the hydrophobicity of ion exchange resin in reactive chromatography for selective separation, grafting the catalyst in membrane material, etc. In this chapter, we review the recent literature on catalysts and their modified forms used in multifunctional reactors that combine reaction and separation. We restrict ourselves to the four most studied multifunctional reactors: reactive distillation, reactive stripping, membrane reactors and chromatographic reactors.

1.2 Reactive Distillation (RD) Reactive distillation is a multifunctional reactor in which chemical reaction and a fractional distillation can be performed simultaneously. It is generally applied to a reversible reaction in which at least one of the products has a different volatility with respect to the other compounds. The most general configuration of a reactive distillation unit consists of: (i) a rectification section in the upper zone; (ii) a reactive section in the middle; and (iii) a stripping section in the lower zone. Due to the simultaneous operation of reaction and separation, this process offers smaller plant size, lower operating costs, higher yields and energy savings. The basic requirement for the success

2 | 1 Catalysis in Multifunctional Reactors

of reactive distillation is a reasonable reaction rate in the temperature and pressure ranges that are governed by the vapor-liquid equilibrium. It is particularly advantageous for equilibrium limited reactions in which the separation of at least one of the products as it is formed can drive the reaction to near completion. Reactive distillation allows the reaction to be carried out much closer to the stoichiometric ratio of the feed flows. Further, it is useful in the case of reactions in which a high concentration of the product or one of the reactants can cause undesired side reactions. Literature is replete with the information on various aspects of reactive distillation. Sharma and Mahajani [2] have reviewed various applications of reactive distillation. The important applications include etherification, dimerization, oligomerization, condensation, esterification, trans-esterification, hydrolysis of esters, hydration, hydro-desulfurization, alkylation, acetalization, ketalization, etc. Successful commercialization of RD technology requires special attention to the hardware design. This means that standard designs used for conventional distillation may not work in the case of RD. The column should provide favorable conditions for both reaction and distillation. The catalyst used in RD columns (RDCs) can be either homogeneous or heterogeneous. The homogeneous catalysts generally offer a high activity; however, a separation of the catalyst from the product mixture incurs an additional cost. However, heterogeneous catalysts such as anion and cation exchangers, zeolites, etc. are preferred over their homogeneous counterparts. These catalysts offer various advantages that include the elimination of separation and recycling of catalysts, elimination of acid disposal problems, exactly defined position of the height of the reaction zone in the column, less corrosion problems, lower investment costs and relatively easier operation. The challenge in heterogeneously catalyzed reactive distillation is the decision of how, where and which type of catalyst should be placed in the reactor to achieve the desired performance. The location of the reaction zone inside the column depends on the type of reaction and the relative volatilities of the components. The column internals for reactive distillation should be designed in such a way that there is an efficient contact between solid catalysts and the liquid phase; one achieves efficient separation by distillation with a high capacity and low pressure drop. A liquid hold-up higher than what is necessary for normal distillation columns is required if the reaction is slow [3, 4]. The optimal solution must be a compromise between these requirements. The mechanical arrangement of the catalyst inside the column and its shape are of primary importance in achieving an optimal performance for both the reaction and distillation. In the following section, RD column hardware for both homogeneous reactions and heterogeneous reactions are explained.

1.2.1 Homogeneous catalysis RD columns, in which a reaction takes place in the liquid phase, are operated countercurrently; a sufficient degree of staging can be achieved in a multi-tray column or in

1.2 Reactive Distillation (RD) | 3

a column with random or structured packing. The packing in this case is inert and serves only to provide even liquid distribution in the column and to suppress liquid phase back-mixing. To increase the productivity of an RD column, it is important to maximize the liquid hold-up in the column as the Hatta number is usually less than unity in most cases [5]. Packed columns usually have much lower hold-up than tray columns, so for a homogeneous RD, tray columns are preferred. The tray column can be operated in the spray, mixed froth or bubbly flow regimes. As higher liquid hold-up and higher residence time are desired, the preferred regime of operation is the bubbly flow regime; it can be achieved by operating the column at lower superficial vapor velocities. The higher weir height ensures higher liquid hold-up on the tray. The bubble cap trays provide higher liquid hold-up, and reverse flow trays with additional sumps can be used to increase the liquid residence time. Eastman Kodak uses a specially designed tray for the manufacture of methyl acetate [6]. Computational fluid dynamics (CFD) can provide better insight into the flow pattern and hence the column performance based on liquid hold-up, pressure drop, residence time distribution and mass transfer aspects. It can predict whether there is an internal circulation within the liquid flowing on the tray [7].

1.2.2 Heterogeneous catalysis For heterogeneously catalyzed processes, hardware design poses a considerable challenge. The catalyst size, hold-up in the column, low pressure drop, good vapor-liquid and liquid-solid contact, and mass transfer are the basic criteria for design. The catalyst particle sizes used in such operations are usually in the range of 1–3 mm. Larger particles are not preferred because they lead to intra-particle diffusion limitations, whereas dumped packing of these small particles can create flooding problems. To overcome these limitations, the catalyst particles are generally enveloped within a wire gauze structure. A variety of envelope geometries have been developed for this purpose. An excellent review of various structures that are used as column packings is given by Taylor and Krishna [8]. Some of the important packing geometries are as follows: (i) porous spheres filled with catalyst inside them; (ii) cylindrical shaped envelopes with catalyst inside them; (iii) wire gauze envelopes with various shapes: spheres, tablets, doughnuts, etc. (Fig. 1.1 (a)) [9]; (iv) horizontally disposed wire-mesh “gutters” filled with catalyst; (v) horizontally disposed wire-mesh tubes containing catalyst; (vi) catalyst particles enclosed in cloth wrapped in the form of bales; (vii) catalyst particles sandwiched between corrugated sheets of wire gauze [10]; (viii) catalyst coated on suitable supports; (ix) internally finned monoliths (Fig. 1.1 (b)) [11]; (x) micro engineered catalysts (Fig. 1.2) [12]. It is worth mentioning that there have been efforts by researchers from ABB Lummus to improve the integration of catalyst and reactor engineering in case of reactive distillation [12]. The novel catalyst system, termed micro-engineered catalyst (MEC),

4 | 1 Catalysis in Multifunctional Reactors Reflux condensor

Non-reactive packing enriching section

Distillate

F1 F2 F3

Catalyst section

F4

(Or)

F5

Non-reactive packing stripping section

F6 F7 Reboiler Bottoms (a) Wired mesh tea bags

(b) Internally finned monoliths

Fig. 1.1: Schematic representation of a reactive distillation column.

allows one to maximize internal and external mass transfer while reducing the hydrodynamic limitations at the same time. MEC systems consist of a web of very thin microfibers that enclose small (< 50 μ) catalyst particles (See Fig. 1.2). It can be either in the form of microencapsulation or thin coating. The metal sheet used as a support here has a very high porosity (~ 90 %) and it can be made in the form of structure that is suitable for the required application, i.e. column packing, in such a way that external mass transfer is also minimized. MEC internal made from highly porous, metal fiber sheets

Encapsulated micro-catalyst particles

Coated micro-catalyst Fig. 1.2: Micro-engineered catalyst concept [12].

1.2 Reactive Distillation (RD) | 5

The important advantage of the structured catalyst sandwich structures is that the liquid follows a criss-crossing flow path that improves radial dispersion. In addition to this, the frequent criss-crossing leads to significant improvement in mass transfer operations within the sandwich structures. The direct coating of normal distillation packings – as Raschig rings or Berl saddles – with a catalytic layer can also be used, but such a solution is probably feasible only for fast reactions because of the limited catalytic surface area available for the reaction [8]. The important catalysts and structures are elaborated in the subsequent sections.

1.2.2.1 Alternate reaction-distillation zones The column is provided with alternate spaces that offer different structures for reaction and distillation. In a reactive distillation column, it is important for the liquid phase reaction that the liquid comes in maximum possible contact with the solid catalyst. The main hurdle for this is the presence of upward moving vapor flow that causes disturbance. In order to overcome this problem, the catalyst may only be placed in the down-comer of a conventional tray distillation column. However, this option is not attractive; it causes a tremendous drop in pressure. An alternate arrangement is shown in Fig. 1.3. In the reaction zone, a separate channel is provided for the vapor flow, while the catalyst is fully surrounded by liquid. The distillation zone is typically comprised of conventional non-reactive trays. A number of such separate reaction and distillation zones may be provided as required.

1.2.2.2 Emulsion or block polymerization In this category, the polymeric catalysts are either prepared by conventional emulsion polymerization or molded with a thermoplastic-like polyethylene, or it is prepared by block polymerization in molds. These catalytic packings may be given a desired shape and size. Moreover, the gas-liquid mass transfer area can be considerably improved. The widespread commercialization of these types of catalysts, especially the block polymers, has been probably impeded by the mechanical stresses developed due to osmotic swelling. The advantage of these packings is that the catalytic activity can still be maintained closer to the original catalytic material [14].

1.2.2.3 Tray columns The catalyst particles can be used in the suspended (fluidized) form over a distillation tray, or catalyst envelopes can be placed on a tray or in the down comers in the RD column along the direction of the liquid flow path across a tray. These envelopes are almost completely immersed in the froth on the tray, ensuring good contact between the liquid and catalyst. The structure of such reactive distillation column is shown in Fig. 1.4. Typically two zones exist: viz. a liquid-solid catalytic reaction zone

Reaction zone

Distillation zone

Reaction zone

6 | 1 Catalysis in Multifunctional Reactors

Fig. 1.3: Alternate regions for reaction and distillation in a column [13].

and vapor-liquid separation zone. Such an arrangement circumvents the need to specially manufacture catalytic packings and obviates the problems associated with a shut-down of the operation to replace the spent catalyst. It may offer better mass and heat transfer characteristics due to the presence of fine catalyst particles. However, the process may suffer from high-pressure drop and operational problems like blockage of screen support by catalytic particles. Syntheses of cumene [15, 16], linear alkylbenzene [17, 18] and methyl acetate hydrolysis [19] have been successfully tested in such a suspended catalytic distillation column. The arrangements that can be used to place a catalyst in the tray towers are as follows [8]:

1.2 Reactive Distillation (RD) | 7

1.

Place catalyst envelopes along the liquid flow path. Weirs provide a liquid level to cover the containers. In this way, only the liquid phase makes full contact with the catalyst while moving across the tray, and the gas flows axially through the tray holes. 2. Place catalyst envelopes within the down comers. The primary drawback with this method is the limited volume available for catalyst inventory. 3. Place the catalyst envelopes near the down-comer. In this case, the catalyst inventory is also limited. 4. Alternate the arrangement of mass transfer trays and packed catalyst sections.

Vapor–liquid separation zone

Downcomer

Solid–liquid reaction zone

Bubbles

Catalyst

Vapor tube Fig. 1.4: Structure of suspended catalytic distillation column [19].

8 | 1 Catalysis in Multifunctional Reactors

1.2.3 Catalysts used in reactive distillation The main challenges for R&D in heterogeneous catalysis in a reactive distillation are to develop catalysts that are active under the operating conditions of distillation. Furthermore, these catalysts should have a relatively long life, so as to avoid excessive shutdowns, catalyst replacements and start-ups that are more complex than in conventional operation.

1.2.3.1 Ion exchange resins Among the various heterogeneous catalysts, ion exchange resin (IER) catalysts are the ones that are used the most. This is probably because a majority of liquid phase reversible reactions, benefited by RD, are acid catalyzed; they occur at relatively low temperatures. IERs are temperature sensitive and hence cannot be used in the case of reactions involving less volatile components. The ion exchange resins such as Amberlyst-15, Amberlite IR-120, Dowex 50W, Indion 130, etc. have been used as catalysts in several applications of reactive distillation [2, 20]. Almost all the configurations discussed above have been investigated with IER as catalysts. Tab. 1.1 summarizes the important applications that are based on ion exchange resins as catalysts. The applications include dehydration, etherification, esterification, trans-esterification, hydrolysis, acetalization, aldol condensation, etc. The recent literature on the catalysts other than ion exchangers is also reviewed. Tab. 1.2 lists these catalysts and their corresponding reactions. A few examples are elaborated below.

1.2.3.2 Zeolites Zeolites, because of their molecule-sieving properties, offer additional opportunities to achieve higher selectivity. Moreover, they withstand a relatively high temperature. A recently studied example is the conversion of phenyl-ethanol to styrene (Eq. (1.1)); here, reactive distillation with a highly active catalyst helps surpass equilibrium and achieve higher conversion of phenyl-ethanol. However, the selectivity towards styrene is the main concern. Medium-pore zeolite H-ZSM-11 showed higher styrene yields than large pore zeolites H-Beta [53]. This is because dehydration is favored in micro-pores that are large enough to accommodate reactants but small enough to inhibit side reactions. It is concluded that the zeolites with a pore size of around 5.5 Å can accommodate mono aromatics and reject large molecules such as polyaromatics. Hence, these catalysts in RD offer not only a higher conversion but also a higher yield of styrene. Other parameters of importance are the Si/Al ratio and crystal size. Interestingly, crystal size increases with an increase in the Si/Al ratio, because Al-rich gels are known to form smaller crystals than Al-lean gels. Styrene yield decreases with an increase in crystal size and the Si/Al ratio. However, the performance of large-pore zeolites is

1.2 Reactive Distillation (RD) | 9

Tab. 1.1: List of various reactive distillation applications based on ion exchange resins. Application

Catalyst

Reference [21] [22] [23] [24, 25] [9, 26]

2-ethylhexyl acetate 2-methyl propyl acetate n-hexyl acetate n-butyl acrylate cyclohexene esterification methyl dodecanoate

Dowex 50 W X-8 Amberlyst-15 Dowex 50 W Amberlyst-15 cationic exchange resins (Indion 130, Amberlyst 15) Amberlyst-15W Amberlyst-15 Amberlyst 46 Amberlyst-15 cation exchange resins (Indion 130, Amberlyst 15) Amberlyst-15 ion exchange resin Amberlyst CSP2 ion exchange resin ion exchange resins Amberlyst-15

[31] [32] [33] [34] [35] [36]

Etherification methyl tert butyl ether (MTBE) tert amyl methyl ether (TAME) ethyl tert butyl ether (ETBE) di isopropyl ether (DIPE) 2-methoxy-2,4,4-trimethyl pentane tert-amyl ethyl ether methylal ethylal 3-methoxy-3-methylpentane isobutyl tert-butyl ether

Amberlyst-15 Amberlite XAD Amberlyst-15 Amberlyst-36 Amberlyst-35 Amberlyst-15 anion exchange resin Indion 130 cation exchange resins ion exchange resins

[37] [38] [39] [40] [41] [42] [43] [44] [45] [46]

Hydrolysis methyl acetate methyl formate

ion exchange resin ion exchange resin

[47] [48]

Hydration/Dehydration ethylene oxide isobutene

cation/anion exchange resin Amberlyst-15

[49] [50]

Acetalization ethylene glycol propylene glycol

Amberite IR-120 Amberite IR-120

[51] [52]

Trans-esterification butyl acetate from methyl acetate

Amberlyst-15

[47]

Esters Synthesis methyl acetate methyl lactate ethyl acetate n-butyl acetate i-propyl acetate n-propyl acetate n-propyl propionate n-amyl acetate i-amyl acetate

[27] [25] [28] [29] [9, 25, 30]

Application

dehydration of phenyl-ethanol to produce styrene

trans esterification of ethyl butyrate with n-butanol

trans esterification of sec-butyl acetate (SBAC) with methanol

trans esterification of propylene carbonate with methanol

conversion of bio oils to biodiesel

fatty acids esterification

synthesis of biodiesel from waste cooking oil

amination of benzene to aniline

bio oil up gradation

Catalyst

zeolites

lipase CALB enzyme

1-(3-sulfopropyl)-3methylimidazolium hydrogen sulfate ionic liquid

sodium methoxide

carbon nano tubes with cobalt and iron metals

sulfated zirconia

12-tungstophosphoric acid

2.5 wt % Cu on h-TS-1 zeolite

p-toluene sulfonic acid supported on activated carbon

353 K, abs 20 kPa

343 K

1 atm

343–483 K and 1 atm

388–493 K

333–433 K

338–373 K, 6 bar

333 K, abs 15 kPa

443 K

Temperature, Pressure

Tab. 1.2: Catalysts used in the various applications of reactive distillation.

pH value increased to 6 after up gradation

12.4 % yield with 84.7 % selectivity towards aniline is obtained

FAME yield up to 93.98 % is achieved

near complete conversion was achieved

efficiency is increased and stability of biodiesel is enhanced

sodium methoxide homogeneous catalyst is only found to active enough to do this reaction in RD

conversion up to 97.72 % is achieved in batch distillation

conversion beyond equilibrium up to 98 %

higher yield of styrene is obtained by medium-pore zeolite with smaller crystal size particles; large pore zeolites show similar activity but lesser selectivity

Performance

acetic acid converted to ethyl ester

these are very active heterogeneous catalysts

highly active and thermally stable catalyst

glycerol, which is a side product, is also simultaneously converted into a valuable chemical

well-known acidic/basic heterogeneous catalysts are not active enough for the reaction

easier catalyst separation makes this process environmentally friendly

enzymes are sensitive to higher temperature

formation of diphenylethyl ethers reduces the selectivity with large pore size zeolites

Remarks

[87]

[86]

[85]

[84]

[83]

[82]

[55]

[54]

[53]

Reference

10 | 1 Catalysis in Multifunctional Reactors

1.2 Reactive Distillation (RD) | 11

independent of crystal size. Hence, catalyst pore size and crystal size are the important parameters that influence the overall performance of this process. OH CH3

(1.1)

CH2 –H2O

1-Phenyl ethanol

Stryrene

1.2.3.3 Enzyme catalysts Another interesting case is the use of bio-catalysts (e.g. enzymes) in RD. An enzyme can be used in an RD column in two ways; lipase CALB immobilized on a silica-gel matrix and in the form of a granule. Trans-esterification of ethyl butyrate with n-butanol (Eq. (1.2)) in a reactive distillation mode has been studied in the presence of these catalysts [54]. Simultaneous removal of ethanol helped in achieving higher conversions (up to 98 %) beyond equilibrium. It has also been proven that the leaching of a catalyst is minimal. Thermal stability of bio-catalyst is the main factor that would determine their potential for commercialization. O H3C

O O

C2H5

n–C4H9 + n–C4H9

Ethyl butyrate

H3C

OH

n–Butanol

+

O C2H5

Butyl butyrate

(1.2) OH

Ethanol

1.2.3.4 Ionic liquids Ionic liquids have higher thermal stability and they stand temperatures as high as 593 K. Because of their low volatility, separation is easier. These catalysts, in spite of being in liquid form, can be conveniently used and would prove to be costeffective. In reactive distillation, the activity of different acidic imidazolium (ionic liquids) catalysts has been tested for trans-esterification of 2-butyl acetate (SBAC) with methanol (Eq. (1.3)) [55]. It has been found that 1-(3-sulfopropyl)-3-methylimidazolium hydrogen sulfate [HSO3 -PMIM]HSO4 has much higher activity than other catalysts such as [BMIM]HSO4 , [HSO3 -BMIM]HSO4 , [HSO3 -PMIM]p-TSA and [HSO3 PMIM]CH3 SO3 . This is mainly because the acidity value is higher than in all the other catalysts. The performance of the catalyst has been tested in an RD column and a conversion of up to 97 % is realized. O H3C

CH3 O 2–Butyl acetate

O

HO

CH3 + CH3OH Methanol

H3C

CH3 2–Butanol

CH3 + H3C

O

Methyl acetate

(1.3)

12 | 1 Catalysis in Multifunctional Reactors

1.2.3.5 Desulfurization of gasoline The desulfurization process of gasoline mainly consists of acid catalyzed alkylation of thiophenic sulfur followed by a product separator. Desulfurization happens via an olefinic reaction to form alkylated thiophenes (Eq. (1.4a)), which have higher boiling point than sulfur components. Hence, the separation of sulfur compounds becomes easier using distillation. The reaction is also accompanied by undesired oligomerization as a side reaction (Eq. (1.4b)). By adjusting operating conditions in an RD column, one can reduce the olefin oligomerization; this helps increase the selectivity for the alkylated product [56]. The catalysts used in this reactor include: zeolites (Naβ [57], HY [58–60], HMCM-22 [59, 61], MCM-41 [62, 63], HBEA [59]), sulfonic resins (NKC-9 [56, 64, 65], Amberlyst-35 [65–67]), solid phosphoric acid [65], silica-supported H3 PW12 O40 6 H2 O (HPW) [68], HPW on γ-Fe2 O3 [69], HPW supported on MgF2 [70], etc. Interestingly, reactive distillation can replace the two-step process with a single unit that combines reaction and distillative separation [56]. RD experiments were mostly performed using ion exchange resins such as NKC-9 [56, 64] and Amberlyst-35 [66, 67]. Among all the operating and design parameters, reactive stage location is the sensitive parameter that decides the performance of desulfurization in a RD column [66]. CH3

CH3

CH3 +

H3C

2 H2C

S

CH3

3-Methylthiophene Isobutylene

CH3

H3C

CH3

H3C

CH3

CH3 2 H2C CH3

(1.4a)

CH3

S

H 3C H3C

CH3

(1.4b)

CH3

Isobutylene

In addition to acid catalyzed reactions, there is a large range of other reactions such as hydrogenation, chlorinations, etc. that can benefit from an RD configuration [2].

1.2.3.6 Epoxidation of alkenes/terpenes Epoxides are valuable building blocks for organic synthesis, particularly for the production of commercially important products for pharmaceuticals, plastics, fragrances, food additives, paints and adhesives [71, 72]. The conventional methods for the industrial production of epoxides employ either stoichiometric peracids or chlorohydrin as an oxygen source [73, 74]. However, both methods have a serious environmental impact. The former produces an equivalent amount of acid waste, whilst the latter yields chlorinated by-products and calcium chloride waste. In addition, there are safety issues associated with the handling and storage of peracids [75].

1.3 Reactive Stripping

| 13

Hence, there is a strong need for cleaner catalytic epoxidation methods that use safer oxidants and produce little waste. The new route includes developing a greener epoxidation process by utilizing a heterogeneous catalyst and a benign oxidant, such as tert-butyl hydroperoxide (TBHP), as it is environmentally benign, safer to handle and possesses good solubility in polar solvents [73, 76]. A number of authors reported a novel and greener solvent-free process for alkene epoxidation using environmentally benign tert-butyl hydroperoxide (TBHP) as an oxidant [74–78]. In this process, polybenzimidazole supported molybdenum complex (PBI.Mo) and a polystyrene 2-(aminomethyl) pyridine-supported molybdenum complex (Ps.AMP.Mo) were used as catalysts for the epoxidation of alkenes/terpenes. During the epoxidation reaction, tert-butanol is also formed as a co-product; this is thus termed an atom-efficient process. Furthermore, tert-butanol can be efficiently recycled through hydrogenolysis and oxidation [79]. Recently, continuous epoxidation of alkenes and terpenes (e.g. cyclohexene, 1-octene, limonene, α-pinene, etc.) with TBHP using PBI.Mo and Ps.AMP.Mo catalysts have been successfully conducted in an RDC [80, 81]. The remarkable catalytic performance of this catalyst has been confirmed in continuous epoxidations of cyclohexene, limonene and α-pinene employing TBHP as the oxidant [80, 81]. A very high conversion of TBHP to cyclohexene oxide (> 98 %) and 100 % selectivity towards cyclohexene epoxide was achieved during epoxidation of cyclohexene in the RDC. This study confirmed that the energy efficiency of this process is about 4–6 fold of conventional processes.

1.3 Reactive Stripping Reactor stripping (RS) is a multifunctional reactor in which reaction and separation are carried out in a single piece of equipment just like RD. However, it offers a greater freedom in the choice of temperature and pressure conditions than RD [88]. RS is particularly advantageous for applications when the reactants and products are temperature-sensitive and the common window for reaction and separation using RD does not exist because some of the reactants are destroyed or degraded in side reactions if heated up to boiling temperatures. For example, condensation of phenol and acetone to produce bisphenol-A; here the boiling points of a majority of the components, except acetone, are relatively high [89]. RS can be operated in a concurrent or counter current or mixed mode type of operation depending on the application of interest. One of the products is stripped off simultaneously from the reacting liquid phase, either using the reactant gas stream itself or by means of an inert gas stream to overcome the reaction equilibrium limitations. There are various applications reported in literature for which RS is more efficient than its more popular counterpart, i.e. RD. A few important applications are Knoevenagel condensation of aldehydes or ketones, esterification of fatty acids, etc. [90]. Some of the reported applications of reactive stripping and the catalyst used are

14 | 1 Catalysis in Multifunctional Reactors

summarized in Tab. 1.3. The catalyst required for given applications plays an important role in the design and performance of the RS process. In general, all the possible catalyst bed configurations that can be used for a packed bed-type reactive distillation system can also be applied to reactive stripping with greatly improved hydrodynamic parameters such as low column hold-up and pressure drop, higher catalytic activity and long catalyst life. Recently, monolithic catalytic packing has also been widely investigated for RS. Conventionally, monolith catalysts have been developed for gas-solid systems and used extensively for exhaust gas cleaning in automobiles and the removal of VOCs and NOx, considering their large geometrical surface area, relatively low cost, low pressure drop and efficient catalyst utilization. For application in reactive stripping, these monolithic catalyst supports are being modified to accommodate two-phase flow by introducing larger internal channels with or without fines to provide an improved gasliquid interfacial area as well as a less irrigated bed pressure drop. These catalysts have been studied extensively for esterification of carboxylic acids with alcohols. In industry, these reactions are catalyzed by using a solid-acid catalyst such as zeolites and ion exchange resins in slurry form [96]. The monolithic support coated with a solid-acid catalyst or ion exchange resin catalyst provides an attractive process alternative as it does not involve the separation of the catalyst after the completion of the reaction. Furthermore, the packed column reactor configuration also enables the introduction of inert gas flow for stripping water from the reaction zone to shift the equilibrium of the reaction to achieve a better yield and selectivity [97]. Flow patterns in a reactor and also the gas-liquid mass transfer are very important factors. A few important applications of reactive stripping are explained in the following sections.

1.3.1 Esterification Esterification, like in RD, has also been investigated in the RS mode. RS is applicable especially when either alcohol or the acid is less volatile and conducting the reaction in RD mode is not advisable. Beers et al. [97] found that the esterification rate of 1-octanol with hexanoic acid (Eq. (1.5a)) increased by 15–20 times in the presence of a catalyst such as zeolite BEA (Si/AL = 37.5) or a Nafion/silica composite with 13 % (SAC13). Complete conversion was achieved with both of the catalysts, but the selectivity towards ester is highest with the SAC13 catalyst. It may be due to the formation of ethers in the side reaction (Eq. (1.5b)) over the strongly acidic catalytic sites. This confirms the selection of the catalyst and operating conditions are very important to achieve the conversion and also selectivity of the required product.

Application

aqueous phase reforming of sorbitol

aqueous phase reforming of sorbitol

catalytic exchange of hydrogen isotopes

dehydration of pentoses and hexoses to furanic aldehydes

dehydration of xylose to furfural

Catalyst

Pt on AlOOH

Pt-Ru on AlOOH

platinum–carbon–PTFE (0.001 : 0.009 : 0.99 w/w)

hydroxylated MgF2

Amberlyst 70

448 K, 8 bar

453 K, 10 bar

333 K, 1.47 bar

51 K, 35 bar

493 K, 35 bar

Temperature, Pressure

Tab. 1.3: Catalysts used in various applications of reactive stripping.

high furfural yield

higher conversion and selectivity compared to other catalysts Amberlyst70 and Nb2 O5

enrichment of deuterium is achieved

higher production rate than monometallic catalyst

reaction is kinetically controlled

Performance

Brønsted sites favor higher yield of furfural

control over acidic and basic sites ration yield optimum conversion and selectivity

customized wet proof catalyst prevents activity and design of catalyst favors gas liquid mass transfer

synergy of catalyst design and reactor design implies maximum production rate

better mass transfer characteristics in micro reactor imply reaction to be kinetically controlled

Remarks

[95]

[94]

[93]

[92]

[91]

Reference

1.3 Reactive Stripping | 15

Application

dehydration of xylose to furfural

esterification of 1-octanol with hexanoic acid

esterification of 1-octanol with hexanoic acid

decarbonylation of aldehydes

synthesis of dimethyl carbonate (DMC) from urea and methanol

Catalyst

Nb2 O5

zeolite BEA (Si/AL = 37.5)

SAC13: Nafion/silica composite with 13 %

Rhodium precursors

Samarium nitrate solution

Tab. 1.3 (continued)

improved in conversion and selectivity due to in situ stripping of CO higher DMC selectivity (75 %) and methyl carbanate (MC) conversion (34 %)

453 K, 20–24 bar

complete conversion and selectivity is around 97 % towards ester

complete conversion and selectivity around 94 % towards ester; main side reaction being etherification of alcohol

high xylose conversion

Performance

453–473 K, 6.8–12.5 bar

428 K, 10 bar

428 K, 10 bar

448 K, 8 bar

Temperature, Pressure

[95] [96–98]

[97]

[99]

[100]

counter current stripping helps in simultaneous removal of water formation of ethers over the strongly acidic catalytic sites selectivity is improved compared to zeolite catalysts many side reactions form carbonyl complexes are reduced because of less concentration of CO efficient removal of ammonia and DMC using superheated methanol increases selectivity to DMC and avoid dimethoxyethane (DME) formation

Reference

Lewis sites favor higher conversion of xylose

Remarks

16 | 1 Catalysis in Multifunctional Reactors

1.3 Reactive Stripping

O C8H17

OH

+ C5H11

O C5H11 O

OH 1–Octanol

Hexanoic acid 2 C8H17

OH

| 17

+ C8H17

Octyl hexanoate C8H17

O

H 2O

(1.5a) Water

C8H17 + H2O

(1.5b) 1–Octanol

Dioctyl ether

1.3.2 Aqueous phase reforming (APR) of sorbitol Hydrogen may be produced by the aqueous phase reforming of sorbitol in the presence of Pt catalysts (Eq. (1.6)). Higher hydrogen concentration in the reactor favors the reverse reaction as well as the side reactions that consume hydrogen. An in-situ stripping of hydrogen thus enhances conversion and also selectivity. An excellent mass transfer attribute offered by micro reactors implies that the overall reaction be kinetically controlled [91, 92]. This demands a highly active catalyst to enhance the production rate. A bimetallic catalyst (Pt-Ru) offers higher conversion but reduces the selectivity for hydrogen, ultimately implying a neutral effect in terms of production. However, continuous hydrogen stripping increases conversion and also gives nearly the same hydrogen selectivity thereby resulting in a higher hydrogen production rate. This is a good example of synergism between the catalyst design and reactor configuration leading to an enhanced performance (0.2 to 6.6 moles of H2 /moles of metal/min). Cn H2m On + n H2 O = n CO2 + (m + n) H2 (1.6)

1.3.3 Dehydration of xylose to furfural The stripping of water during the course of the reaction for the dehydration of xylose (Eq. (1.7)) produces a higher yield of furfural. The activity of catalysts such as Nb2 O5 and Amberlyst-70 for the formation of furfural in the RS mode has been studied in the past [94, 95]. It was concluded that Lewis acidic sites help in converting xylose and strong Brønsted sites are responsible for higher furfural yield. To maximize the conversion and yield, a hydroxylated MgF2 catalyst is used with which one can tune the Lewis/Brønsted acid sites ratio. This helps in achieving a higher conversion and higher yield of furfural.

18 | 1 Catalysis in Multifunctional Reactors OH HO

O

O

CHO

–3 H2O

HO

(1.7)

OH Xylose

Furfural

1.3.4 Catalytic exchange of hydrogen isotopes An interesting case of a process intensification involving an equilibrium reaction and separation is the catalytic hydrogen-water isotope exchange process used for the separation of hydrogen isotopes. In this process, the catalytic isotope exchange reaction takes place in the vapor phase over a wet-proofed solid catalyst between the strippedoff component and gas. The isotope exchange reaction is carried out in a trickle bed reactor (TBR) filled with a wet-proofed catalyst in which hydrogen and liquid water are contacted in countercurrent mode over the catalyst bed to transfer the heavier isotope of hydrogen, i.e. deuterium from liquid water to hydrogen through an exchange reaction that takes place in the gas phase [99]. The overall process may be written as: HDO(L) + H2 (g) = HD(g) + H2 O(L) .

(1.8)

Conventionally, a noble metal catalyst such as platinum supported on a high surface area support such as alumina, silica or activated carbon are found to be the most effective for the exchange reaction. However, these catalysts were found to be catalytically active only in the gas phase as they lose their catalytic activity, and when they come in contact with liquid water due to wetting of the catalyst surface, this prevents reactants from reaching the active sites. Hence, a Pt catalyst on carbon is blended with PTFE (platinum–carbon–PTFE (0.001 : 0.009 : 0.99 w/w) to render hydrophobicity, thereby providing for the direct exposure of catalytic sites to the reacting gas phase [93]. On the other hand, because of poor wetting, the gas-liquid mass transfer necessary to strip HDO off the liquid phase is adversely affected. To circumvent this problem, a mixture of a wet-proof catalyst and ceramic Raschig rings was used in a trickle bed reactor as shown in Fig. 1.5. An enrichment from 30 ppm deuterium to 200 ppm deuterium was achieved in a single pass. The interplay between reaction rate and mass transfer rate leads to an optimum proportion of catalytic and non-catalytic packings in the column. It is also reported that the catalyst used in this study imposes severe pore diffusion resistances, thereby implying an internal effectiveness factor to be not more than 0.2; this provides for a further scope for improvement in catalyst design.

1.4 Catalytic membrane reactors |

H2(g)+HD(g)

19

HDO(L)

HDO(L) H2(g) H2O(g)+HD(g)

H2(g) H2(g)+HDO(g) HDO(L) Wettable inert packing Wet proofed catalyst packing H2(g)

HDO(L)+H2O(L)

Fig. 1.5: Schematic of hydrogen isotopic exchange process [93].

1.4 Catalytic membrane reactors Catalytic membrane reactors (CMRs) combine reaction with membrane separation. In a catalytic membrane reactor, the selective removal of a reaction product through the membrane or a controlled addition of reactant through the membrane helps in shifting the equilibrium. In addition to this, these reactors are also useful in enhancing the selectivity of catalytic reactions [102]. Based on their material of construction, the membranes used for CMR applications can be classified into two categories: organic and inorganic. Organic membranes have good separation properties as compared to inorganic membranes. However, the drawback of organic membranes is that they generally do not withstand temperatures above 473 K, which limits their use in catalytic processes. Inorganic membranes are preferred for severe process conditions and high temperature applications. Various arrangements have been proposed to combine a catalyst and a membrane in a catalytic membrane reactor (CMR). They are mainly classified into three categories as shown in Fig. 1.6. In the first case, the membrane is permeable only to products while retaining reactants and the catalyst (Fig. 1.6 (a)). Another type of application is one in which the membrane acts as a contactor between two reactants (Fig. 1.6 (b)). In the third category, one of the products or reactants is permeable through the membrane (Fig. 1.6 (c)). In the latter case, this property of the membrane can be used to meet several objectives, i.e. the conversion or selectivity enhancement through prod-

20 | 1 Catalysis in Multifunctional Reactors

Homogeneous catalyst

Reactants

JA

A A+B

Membrane Products

(a)

Catalyst

B

C

(b) Catalyst A B+C

A

C

JB Sweep

A

A+B

D

D+B

U

JD B

Sweep (d)

(c)

B

C JB

Jj (i ≠ (Catalyst)

A

C

A+B

D

D+B

U

D

JB

D

(e) Fig. 1.6: Catalytic membrane reactors [103].

uct withdrawal (Fig. 1.6 (d)) and selectivity enhancement through an optimal reactant dosage (Fig. 1.6 (e)), etc. A few important applications of membrane reactors and the associated details are given in Tab. 1.4. In the following section, we review some selected applications in which the catalyst in the membrane reactors plays an important role.

1.4.1 Biodiesel production CMR offers unique way to remove the large molecules of oil that cannot pass through the membrane pores. On the other hand, the produced biodiesel, which is essentially a mixture of fatty acid alkyl esters with relatively small molecular sizes, is able to pass through the membrane along with alcohol, glycerol and the catalyst. The equilibrium of the trans-esterification reaction (Eq. (1.9)) is thus shifted to the product side. Several catalysts have been used for the production of biodiesel in a membrane reactor: KOH

U

D

hollow fiber membrane reactor

Pt (1 wt %)/SBA-15

dehydrogenation of propane

823 K

773 K and 0.1 MPa

packed-bed catalytic membrane reactor

Cr2 O3 /Al2 O3

dehydrogenation of propane

catalyst embedded membrane

Mg-Al hydrotalcites

353 K, 4.15 ml/min

333 K, batch reactor

biodiesel synthesis

micro porous ceramic

p-TSA/MCM-41

343 K, 0.21 cm/s

Temperature, circulation velocity

biodiesel synthesis

biodiesel synthesis

micro porous TiO2 /Al2 O3

KOH supported on activated carbon

Application

Type of membrane

Catalyst

Tab. 1.4: Catalysts used in the membrane reactor for different applications.

propane conversion 75.3 %

~ 36 % conversion of propane at 773 K

supported HT are twenty times more active than unsupported HT

highest biodiesel yield of 84.1 % was obtained at optimum reaction condition

94 % conversion of oil at optimum reaction condition

Performance

[105]

[106]

membrane pore size does not have significant effect on biodiesel yield increase of membrane hydrophobicity seems to increase catalytic activity but decrease FAME equilibrium yield

HFMR-II can be applied to other catalytic reactions with less coking problems, such as the water gas shift reaction and steam reforming, etc.

[108]

[107]

[104]

catalyst was reusable up to three times but its activity reduces from the fresh catalyst

separation of H2 from membrane double the conversion from its equilibrium value

Reference

Remarks

1.4 Catalytic membrane reactors |

21

Type of membrane

ionic-electronic conducting membrane

porous γ-alumina membrane

dual-phase composite membrane reactor

ceramic membrane reactor

Ru deposited ceramic membrane reactor

Catalyst

Na-W-Mn/SiO2

Mn-W-Na/SiO2

Perovskite SrTiO3

Ni/Al2 O3

Ru deposited on membrane

Tab. 1.4 (continued)

at higher pressures, a decrease in the methane conversion takes place, due to the unfavorable equilibrium shift

96.4 % methane conversion

59 % methane conversion

1073 K, atmospheric pressure

773 K, 0.12 MPa

partial oxidation of methane to synthesis gas

partial oxidation of methane to synthesis gas

methane conversion is higher than equilibrium conversion (54 %)

catalytic performance strongly depends on both reaction temperature and CH4 feed rate

74 % CO and 75 % H2 selectivity at 17 % CH4 conversion

1223 K and CH4 feed rate of 20 ml/min

partial oxidation of methane to synthesis gas

[113]

[112]

[111]

[110]

C+2 yields 27.5 %

[109]

Reference

yield of C+2 as 34.7 %, methane conversion of 51.6 %, C+2 selectivity of 67.4 %

Remarks

1078 K, He flow 212 ml/min

1173 K, sweep gas flow rate 100 ml/min

oxidative coupling of methane

Performance

oxidative coupling of methane

Temperature, circulation velocity

Application

22 | 1 Catalysis in Multifunctional Reactors

553–603 K, 22 ml/min

methanol steam reforming

Al2 O3 supported Pd based membrane

Pd-Ag membrane

nickel alumina membrane

Pd membrane

CuO/ZnO/ Al2 O3 catalyst

Pt/20 % CeZrO2 / Al2 O3 catalyst

nickel membrane

1 % Pd/CuOZnO catalyst

steam reforming of methanol

combined steam and CO2 reforming of methane

CO2 reforming of methane

623–923 K, 20 ml/min

ethanol steam reforming

Pd-Ru membrane

0.3 % Pt-Ru on detonation nanodiamonds

reaction rate is 50– 100 % faster than for reaction without Pdmembrane

conversions up to 96 % was achieved

923–1023 K

583 K

conversion is 60 % higher than in fixed bed with 75 % hydrogen recovery

highly pure hydrogen is produced with 85 % methanol conversion

Pt-Ru catalysts perform better than Pt-Ni catalysts

Ru/ La2 O3 (50 wt %)– SiO2 , exhibited the highest turnover frequency, having 38 % Ru dispersion

Performance

823 K, 10–90 ml/min

823 K, 10–70 ml/min

dry reforming of methane

Pd-Ag membrane

Temperature, circulation velocity

0.6 % Ru on La2 O3 and SiO2

Application

Type of membrane

Catalyst

Tab. 1.4 (continued)

[114]

La2 O3 and SiO2 supports imply stable and active support

[118]

[119] hydrogen spill over from the catalysts to membrane favors catalytic activity

[117]

catalyst with 20 % CeZrO2 on alumina is a stable formulation carbon deposition was not observed

[116]

Pd based membranes has long life time

[115]

Reference

Remarks

1.4 Catalytic membrane reactors |

23

24 | 1 Catalysis in Multifunctional Reactors

supported on activated carbon [104]; p-TSA supported on MCM-41 [105]; Mg-Al hydrotalcites [106]; etc. Biodiesel has also been produced using poly vinyl alcohol (PVA) membranes with solid base catalysts for the trans-esterification of soybean oil with methanol. The membranes were prepared by dispersing hydrotalcite in the polymer solution. The hydrophobicity of membrane may be increased by adding succinic anhydride in the PVA. It is observed that by increasing the hydrophobicity of the PVA membranes, its catalytic activity increases at the cost of FAME yield. OCH3 R1 O

R1 OH

O O R2

O R3

O

O

Triglyceride

+ OH

+ 3 HC 3 Methanol

O OCH3

+ R2

OH

O

HO Glycerol

+

(1.9)

OCH3

R3 O Esters

1.4.2 Dehydrogenation The increasing demand for propene and its derivatives requires the further development of available technologies that offer a higher efficiency and improved propene selectivity. It can be produced by dehydrogenation of alkane (Eq. (1.10)) using a catalytic membrane reactor. A conventional way of dehydrogenating propane is to use a tubular flow reactor combined with a gas-separation membrane [107]. The reaction may be performed at high temperatures (773–873 K) using a Cr2 O3 /Al2 O3 catalyst and gas-separation membrane (silica/alumina). Hydrogen formed during the reaction can be separated through the membrane to overcome the equilibrium limitation. A maximum of 36 % conversion of propane at 773 K, which is double the equilibrium value, has been reported. In another study, a hollow fiber membrane reactor (HFMR-I) using Pt (0.5 wt %)/γ-Al2 O3 catalyst was developed for the dehydrogenation of propane to propene [120]. It is further improved to a highly compact multifunctional Pd/alumina hollow fiber membrane reactor (HFMR-II) that consists of a thin and defect-free Pd membrane coated directly onto the outer surface of an alumina hollow fiber substrate. It has a unique asymmetric pore structure, i.e. a sponge-like outer layer and a fingerlike inner layer where Pt (1 wt %)/SBA-15 catalyst is deposited. SBA-15 offers high catalyst loading and a high surface area for Pt as compared to γ-alumina. With a hollow fiber membrane reactor (HFMR), propene selectivity and space-time yield is 10 times higher than that obtained in a conventional fixed bed reactor.

1.4 Catalytic membrane reactors |

C3 H8 󳨀󳨀→ C3 H6 + H2

25

(1.10)

In another example, styrene is produced by catalytic dehydrogenation of ethylbenzene. This reaction is also accompanied by side reactions that produce benzene and toluene (Eq. (1.11)). The membrane reactor helps increase the conversion of ethylbenzene and improve selectivity. There are various membranes that were successfully applied for this reaction, for example, ceramic membranes [121], zeolite membranes [122], palladium membranes [123, 124] etc. In all the studies, a significant increase in conversion and selectivity were observed by using a membrane reactor when compared to a fixed bed reactor. For example, She et al. [123] found an increase in conversion up to 10 % and an increase in selectivity up to 15 % using palladium supported on porous stainless steel tubes. −H2

+H2

ethylbenzene ←󳨀→ Styrene 󳨀󳨀󳨀→ Benzene + C2 H4 +H2

(or) Toluene + CH4

(1.11)

1.4.3 Oxidative coupling of methane (OCM) A catalytic membrane reactor is a promising way of coupling methane to C2+ or forming ethane + ethylene from methane (Eq. (1.12)). CMR has major advantage in terms of the selectivity of a C2+ product as compared to co-feed packed bed reactor. The experimental results clearly demonstrated that it was beneficial to distribute the feed of oxygen along the reactor length for methane oxidative coupling reactions [110]. Out of many inorganic membranes, the ionic-electronic conducting membrane [109], porous γ-alumina membrane [61] and catalytic perovskite hollow fiber membrane [125] are the few membranes that showed higher activity for the OCM reaction. +O2

CH4 󳨀󳨀󳨀→ C2 H2 , C2 H4 , C2 H6 + H2 O

(1.12)

The outer surface of CMR consists of a mixed ionic-electronic conducting membrane (MIECM) Ba0.5 Ce0.4 Gd0.1 Co0.8 Fe0.2 O3-δ (BCGCF) coated using the sol-gel method; the inner surface of the membrane tube was coated with an Na-W-Mn/SiO2 catalyst [109]. This CMR performed best in terms of a C2+ yield of 34.2 % with methane conversion of 51.6 % and C2+ selectivity of 67.4 %, which is higher than a conventional reactor. OCM experiments were also conducted in a porous γ-alumina membrane reactor using an Mn-W-Na/SiO2 catalyst [110]. A maximum of 27.5 % C2+ yield was obtained at optimal reaction conditions. It was observed that a higher helium flow rate produced a higher C2 selectivity and yield, whereas changing the methane flow rate did not significantly affect the reactor performance. A dense La0.6 Sr0.4 Co0.2 Fe0.8 O3 (LSCF) hollow fiber membrane was used for the OCM reaction, a maximum 21 % yield of C2+ was obtained using SrTi0.9 Li0.1 O3 catalyst [125], although it is still lower than the 30 % threshold for commercial feasibility.

26 | 1 Catalysis in Multifunctional Reactors

1.4.4 Partial oxidation of methane to synthesis gas The partial oxidation of methane (POM) (Eq. (1.13)) has attracted growing attention for the production of synthesis gas from methane with a ratio of H2 /CO equal to two, which is a suitable feedstock for the synthesis of hydrocarbons or methanol and the subsequent production of liquid fuels. A catalytic membrane reactor allows us to separate the oxygen and catalytic POM process in a single step. There are many types of catalytic membrane reactors used for PMO process, few important CMRs are dualphase composite membrane reactor [111], ceramic membrane reactor [112] and Ru deposited ceramic membrane reactor [113]. The catalytic partial oxidation of methane to syngas over perovskite SrTiO3 was studied with a Ce0.8 Sm0.2 O2-δ -La0.8 Sr0.2 CrO3-δ dual-phase composite membrane reactor [111]. Approximately 74 % CO and 75 % H2 selectivity at 17 % CH4 conversion can be achieved under the optimum operating conditions at 950 °C and CH4 feed rate of 20 ml min−1 . The catalytic POM for the synthesis of syngas was also performed using a two-zone fixed bed of Ni/Al2 O3 catalyst inside a modified ceramic membrane [112]. A maximum 96.4 % conversion of methane (more than equilibrium conversion, 94 %) was achieved at 1073 K and at atmospheric pressure and a CH4 , O2 , and N2 feed ratio of 2/1/1. A catalytic ruthenium membrane reactor (CRMR) has also been investigated [113]. Two layers of Ru nanoparticles were coated over the ceramic reactor giving maximum methanol conversion over 59 % with H2 selectivity over 71 % at 773 K and feed ratio of CH4 /O2 /N2 = 2/1/14 (total feed rate = 381.5 ml/min). CH4 + 21 O2 󳨀󳨀→ CO + 2 H2 (1.13)

1.5 Chromatographic Reactor A chromatographic reactor is one of the multifunctional reactors in which reaction and separation can be carried out in a single vessel or equipment. The concept of chromatographic reactor lies in the adsorptivity of the different species involved rather than differences in their volatility as in the case of reactive distillation (RD). A chromatographic reactor is particularly advantageous to use, as an alternative to RD, when the difference in the volatility of the species are small or the components are nonvolatile or sensitive to temperature. Ideally, a chromatographic reactor employs a mixture of a solid catalyst and an adsorbent. Nowadays, it is common to use a solid catalyst that also acts as an adsorbent. This section will explain the concept, types and applications of liquid chromatographic reactors.

1.5 Chromatographic Reactor

|

27

1.5.1 Concept of a Chromatographic Reactor The basic concept of a chromatographic reactor can be easily understood by studying a single chromatographic column, operated in the conventional batch mode. Consider an equilibrium-limited reaction in a single chromatographic column given by reaction (1.14): A ⇔ C+D, (1.14) where D is the most strongly adsorbed on the catalyst while A and C are less adsorbed on the catalyst. The reaction (1.14) takes place under diluted conditions in an inert solvent. Consider a pulse of A being injected into a fixed bed comprised of an adsorbent of high affinity towards D, and a lower one towards A and C. As the reaction takes place, both product species migrate through the reactor at different velocities, with D being retained more strongly than A and C, and thus staying behind the reactive front. This continuous separation of the two products leads to a suppression of the backward reaction, thus driving the conversion towards completion and enabling the withdrawal of two high-purity product fractions at the column outlet. The following things are worth mentioning: (i) in a chromatographic reactor, a separation of the two products can only be achieved if complete conversion has been achieved; (ii) it is not necessary to have a reactant of intermediate affinity towards the stationary phase as long as C and D can be separated. The extension of the above concept to bimolecular reactions is straightforward. However, it must be noted that in a bimolecular reaction, one has to avoid the separation of the reactants by choosing a suitable stationary phase and solvent. This can easily be achieved in applications by using one of the reactants as solvent in order to ensure its availability at the reaction locus [126, 127]. There are several classes of reactions to which reactive chromatography can be applied. Probably, the widest one is given by esterification reactions, catalyzed, for example, by acidic ion exchange resins [126, 128] as here the obvious polarity difference between the two products (i.e. ester and water) make their separation easy on different adsorbents, for example ion exchange resins which act both as a catalyst and an adsorbent. In general, the packing of a chromatographic reactor column is carried out with two different solid particles; one is the catalyst and the other is the adsorbent [129]. Amberlyst-15, a cation exchange resin catalyst which acts both as the catalyst for the reaction and adsorbent for the chromatographic reactor was used to carry out the reactions in chromatographic reactors [126–128, 130–138]. Similarly, Amberlyst-31 [127], DOWEX 50W-X8 [139, 140] and TS-1 [141] have been used in the chromatographic reactor experiments for their dual characteristics of both catalyst and adsorbent. The industrial application of chromatographic separation is in the field of chiral and bioseparations [142]. Chromatographic reactors can be operated in both batch and continuous modes.

28 | 1 Catalysis in Multifunctional Reactors

1.5.2 Types of Chromatographic Reactor A chromatographic reactor mainly consists of a stationary phase and a mobile phase. It can be operated in either batch (discontinuous) mode or continuous mode, i.e. the reactant or the reactant mixture are fed continuously or discontinuously to the chromatographic reactor. The stationary phase acts both as an adsorbent and catalyst. As a result, the reaction and the separation take place simultaneously inside the chromatographic reactor. The common stationary phases are solids typically in the form of porous media with large specific surface areas. The solid phases can be individual adsorbent for self-catalytic or homogeneous catalyzed reactions, adsorbent activated by metal ions or functional groups, or a mixture of catalyst and adsorbent. The stationary phase may be a liquid coated on a solid support or a liquid retained by centrifugal force [143]. Based on the employed phases and the working principles, a chromatographic reactor can be classified as explained in the following section.

1.5.2.1 Batch Chromatographic Reactor (BCR) Batch chromatographic reactors (BCR) consist of a fixed volume packed with catalyst and/or adsorbent. For most of BCR experiments, the catalyst also acts as an adsorbent. In BCR, the adsorption is followed by the catalytic reaction after which the unreacted feed along with the products are eluted from the chromatographic reactor one after the other depending on the affinity towards the adsorbent [139]. Fig. 1.7 shows the schematic of a batch chromatographic reactor (BCR) experimental set-up. Consider the reactions (1.15) and (1.16): A ⇔ C+D,

(1.15)

A+B ⇔ C+D.

(1.16)

Two kinds of experiments are carried out in a batch chromatographic reactor for each of the above two reactions mentioned. One is the reaction step and other is the regeneration step. During the reaction step, a known amount of reactant mixture (A or A + B, depending on the reaction (1.15) or (1.16)) is continuously fed to the column which is initially saturated with a desorbent (B, in case of the reaction (1.16) and inert solvent in case of the reaction (1.15)). As soon as the reactants enter the column, A or A + B reacts in the presence of the catalyst to produce C and D. D, which has a higher affinity towards resin, is readily adsorbed on the resin while C, with lesser affinity, soon gets desorbed. As a result, C gets separated with D, moves towards the exit of the column along with A and B. Since C is separated from the reaction, the reaction proceeds until consumption of the limiting reactant (A). With any specific location in the column, this process continues until the resin catalyst becomes saturated with A, C and D. When the resin catalyst becomes saturated, chemical equilibrium is achieved and the composition at every point in the column remains the same. At this point, the reaction step is stopped and the regeneration step

1.5 Chromatographic Reactor

|

29

Desorbent [Solvent, B (reactant or inert solvent) Samples for analysis GC/GC–MS Reactants feed (A or A+B) Chromatographic reactor

HPLC pump Catalyst/Adsorbent

Oven

Fraction collector

Fig. 1.7: Schematic of a batch chromatographic reactor (BCR) experimental set-up.

is started. Pure B or inert solvent (depending on the reaction (1.15) or (1.16) is continuously fed to the column for the regeneration by removing A, C and D which are present in the column. When the composition of the exit of the column becomes the same as the composition of the inlet B, the regeneration step is considered to be complete. In recent works [144–147], acetic acid (A) is used as a reactant, n-hexanol (B) is used as a solvent (desorbent), n-hexyl acetate (C) and water (D) are the products. For bimolecular reactions, separation of the reactants should be avoided by choosing a suitable stationary phase and solvent, as well as proper operating conditions [128]. This can easily be achieved by using one of the reactants as the solvent in order to ensure its availability at the reaction locus [127]. The major disadvantages of batch chromatographic reactors are low yields due to poor utilization of the stationary phase, high dilution of products due to high desorbent required and discontinuous operation. A comparison of batch chromatographic reactor and plug flow reactor (PFR) was reported by Falk and Seidal-Morgenstern [139]. The comparison of these reactors showed that the conversion of both reactors seems to be similar at the same dilution ratio, but the fixed-bed chromatographic reactor (FBCR) is more attractive for products separation without assistance of a downstream separator.

1.5.2.2 True Counter-Current Chromatographic Reactor (TCR) Fig. 1.8 shows the schematic of a true counter-current chromatographic reactor (TCR). Consider the reaction (1.17): A+B ⇔ C+D, (1.17) where component D is strongly adsorbed onto the adsorbent/catalyst followed by components A and B, while C is the least adsorbed component. In TCR, the direc-

30 | 1 Catalysis in Multifunctional Reactors

Fluid flow Raffinate (C,B)

Feed (A,B) 3

4 C

Extract (D,B) 2

D C

Eluent (B)

1 D C

D

Solid flow Fig. 1.8: Concept of a true counter-current chromatographic reactor (TCR).

tion of the fluid flow is opposite to the direction of the true solid flow. The TCR is divided into four different sections. Each section performs a specific function so that complete conversion and separation could be achieved. Section I is located between the desorbent (eluent) and the extract port. The flow rate is higher in this section compared to other sections. The higher flow rate is necessary to remove strongly adsorbed component (D) from the adsorbent/catalyst so that regeneration of the chromatographic reactor column is carried out. In section II, between the extract and the feed port, components C and D are formed. For a complete conversion of A and separation of products (i.e. C and D), the flow rate of this section should be adjusted such that C gets desorbed from the solid phase before reaching the extract port while D remains adsorbed in the solid phase which leads to a higher purity of D at the extract port. Section III is located between the feed and raffinate ports. The reaction takes place in this section and hence this section is also referred to as the reactive section. Like section II, flow rate in this section is adjusted so that reaction time should be sufficient and the component D could get adsorbed on the solid phase. Thus the fluid collected at the raffinate port has a very high concentration of the least adsorbed component (C). Section IV is placed between the raffinate and the desorbed ports. Before the fluid is recycled, it is cleaned in section IV. The component (C) is adsorbed from the fluid phase and transported back to section III along with the solid phase so that desorbent (B) is purified before being recycled to section I. True countercurrent chromatographic reactor (TCR) suffers from practical problems with handling the solid phase as well as its requirement to achieve a large rate of solid movement. Also, the abrasion between the solid particles and back mixing are difficult to avoid. Hence, a simplified version of TCR, i.e. continuous chromatographic reactor (CCR) also known as simulated moving bed reactor (SMBR), is more popular and considered to be a better option; it is explained in the following section.

1.5.2.3 Continuous Counter-Current Chromatographic Reactor (CCR) A continuous counter-current chromatographic reactor (CCR) is also referred to as a simulated moving bed reactor (SMBR). The CCR was first developed by UOP (Univer-

1.5 Chromatographic Reactor

|

31

sal Oil Products) in 1960 for the non-reactive separation of mixed xylenes [148]. This process was licensed as the SORBEX process. Fig. 1.9 shows the schematic of a continuous counter-current chromatographic reactor (CCR) experimental set-up. It consists of several batch chromatographic reactor (BCR) columns packed with adsorbents/ catalysts connected in series. The inlet and outlet ports of the reactants and products are switched at regular intervals. The time period between the successive switches of ports is called switching time (t s ). Consider the reaction (1.18): A+B ⇔ C+D,

(1.18)

where component D is strongly adsorbed onto the adsorbent/catalyst followed by the components A and B, while C is the least adsorbed component. There are two inlet fluid streams (feed (A, B) and eluent (B)) and two outlet fluid streams (raffinate (C, B) and extract (D, B)) as shown in Fig. 1.9. These two inlet and outlet fluid ports divide the unit into four sections (i.e. I, II, III and IV). Each section of CCR is replaced with several subsections packed with catalyst/adsorbent and each section performs a specific job so that complete conversion and separation could be achieved. Component B acts both as feed reactants along with A as a desorbent. For complete separation of products, component C should be least adsorbed and component D should be strongly adsorbed. The outlet ports are switched in the direction of fluid flow by one bed volume length. The working principle of CCR remains the same as TCR except that there is no movement of the solid phase in the CCR. Depending on the reactive system, different CCR configuration set-ups can be found in the literature. If the least adsorbed product (component C) does not get adsorbed, the recycling of the pure desorbent is not possible; section IV can be Raffinate (C,B)

Eluent (B)

Section 4 3

4

2

5 Direction of fluid flow and port switching

Section 3 1

6

8 Feed (A,B)

Section 1

7 Section 2

Extract (D,B)

Fig. 1.9: Schematic of a continuous counter-current chromatographic reactor (CCR).

32 | 1 Catalysis in Multifunctional Reactors

eliminated [127]. Also, if the regeneration of the adsorbent requires a change in the operating conditions, i.e. temperature, pressure or change in desorbent, then it would be more convenient to remove section I and perform the regeneration of the solid phase separately [149]. It may be noted that the adsorbent and catalyst materials used for CCR may be the same, different or a mixture of the two. A comparison of CCR and TCR processes that have been discussed in detail by Lode et al. [150] can be summarized as follows: (i) TCR does not really apply to CCR units with finite number of columns per section, i.e. CCR tends to behave like a TCR only for infinite columns in each section of infinite lengths; (ii) the two reactors (TCR and CCR) exhibit different residence time distributions and lead to different degrees of conversion; and (iii) TCR reaches true steady state while the CCR only reaches cyclic steady state.

1.5.2.4 Centrifugal Partition Chromatographic Reactor (CPCR) A centrifugal partition chromatographic reactor (CPCR) is an integration of reaction and centrifugal partition chromatographic separation where counter-current distribution of species takes place in the absence of an adsorbent or catalyst. In this chromatographic reactor, two immiscible liquid phases with different densities are separated. The stationary phase with higher density is retained in the column by a combination of centrifugal force and a geometric channel, while the mobile phase passes through the column as micro-droplets. A schematic of a centrifugal partition chromatographic reactor (CPCR) column is shown in Fig. 1.10. A mixture of components A and B is separated due to their different affinities towards the two-liquid phases. The advantage of this chromatographic reactor is the large capacity of the stationary phase compared to the conventional technique with liquid on solid support (i.e. liquid-liquid extraction). CPCR has been successfully applied for enzymatic reactions [151, 152].

Time

A

B

Time

Fig. 1.10: Schematic of a centrifugal partition chromatographic reactor (CPCR).

1.5 Chromatographic Reactor |

33

1.5.2.5 Continuous Rotation Annular Chromatographic Reactor (CRACR) In a continuous rotation annular chromatographic reactor (CRACR), the stationary phase is packed into the annulus of two concentric cylinders, rotating continuously about the common axis. The mobile phase (eluent) is fed uniformly over the whole cross-section at the top of the annular space while the reactant is fed to a fixed feed inlet port. The reacting species are conveyed along the longitudinal axis of the annular space due to mobile phase flow, whereas they have a circumferential displacement by adsorption, desorption and rotational movement of the annular. Hence, the components are separated and eluted from the reactor in different angles, compared to the fixed feed port. A schematic of a continuous rotation annular chromatographic reactor (CRACR) [153] is shown in Fig. 1.11. The performance of a continuous rotation annular chromatographic reactor (CRACR) is similar to that of a batch chromatographic reactor (BCR) system, except that the operation is carried out in a continuous mode in the CRACR [154]. The continuous operation of CRACR results in an inefficient utilization of the solid phase and high desorbent consumption because complete separation is achieved without recycling mixed fractions. Fixed feed inlet

Eluent

Fig. 1.11: Schematic of a continuous rotation annular chromatographic reactor (CRACR).

1.5.2.6 Reversed Flow Chromatographic Reactor (RFCR) The concept of a reversed flow chromatographic reactor is similar to that of an adsorptive reactor that is used for heat accumulation for regeneration, but it is related to mass accumulation [143]. The reverse-flow chromatographic reactor (RFCR) is a fixedbed reactor packed with suitable adsorbent/catalysts. One of the reactants is fed at

34 | 1 Catalysis in Multifunctional Reactors

the middle of the reactor and the flow direction of the carrier is periodically switched. A schematic of a reversed flow chromatographic reactor (RFCR) is shown in Fig. 1.12. If the reactants are strongly adsorbed onto the adsorbent/catalyst whilst products are less adsorbed onto the adsorbent/catalyst, the periodic switching of the carrier could lead to trapping the strongly adsorbed reactant within the reactor. Two three way valves are controlled to keep the concentration profile of the reactant propagated in both the directions, but not out of the column. It was first applied by Agar and Ruppel [155] for the reduction of NOx with NH3 , where only NH3 is adsorbed and the reactant is fed at the middle of the reactor. Recently, RFCR has been reviewed and it was shown that it can significantly improve conversion and yield for equilibrium or selectivity limited reactions [156–158]. Very little or no experimental information is available for this case [143]. A

B

C

Fig. 1.12: Schematic of a reversed flow chromatographic reactor (RFCR).

1.5.3 Applications of Liquid Chromatographic Reactor There are several applications of liquid chromatographic reactors that are explained in detail in the following sections:

1.5.3.1 Esterification Reaction 1.5.3.1.1 Triacetine Synthesis (Esterification of acetic acid with glycerol) Triacetine (glycerol triacetate) is used as a plasticiser including filters in cigarettes; hence it is required to be of food grade quality. Gelosa et al. [134] studied the synthesis of triacetine; it is a series of three steps esterification of glycerol with acetic acid in the presence of an acidic polymeric resin (Amberlyst-15) in a chromatographic reactor. Water molecules formed in each step of esterification along with monoacetine

1.5 Chromatographic Reactor |

35

in the first step, diacetine in the second step and finally triacetine in the third step of esterification. The kinetics of esterification was studied in a batch reactor with and without Amberlyst-15. Gelosa et al. [134] reported that the conversion of a non-catalyzed esterification reaction was lower than 10 % in one hour as compared to catalyzed esterification, which had already attained equilibrium within that time. It was further reported that the effect of interphase mass transport resistances on the kinetics of triacetine synthesis was negligible by carrying out experiments using different stirrer speeds that gave similar results. However, it was also reported that intraphase mass-transport resistances may affect the kinetics of the process by conducting a non-reactive experiment where a small amount of water was added to the batch reactor containing only resin and glycerol at 333 K. They also studied multi-component adsorption equilibria for three binary mixtures (i.e., water-acetic acid, water-glycerol and acetic acid-triacetine) and reported that water has a higher affinity for the resin due to the strong polarity that was made inside the resin by the sulfonic acid groups, followed by glycerol, acetic acid, monoacetine, diacetine and then triacetine. The effect of various parameters such as the reaction temperature, feed molar ratio of the reactants (acetic acid to glycerol) and catalyst to reactant ratio (resin to glycerol) was studied for the modeling of the chromatographic reactor. They concluded that there was a good agreement between the experimental findings and predicted values. One important concern was addressed by Gelosa et al. [134] regarding the regeneration of the resin. Since acetic acid was used as a desorbent to remove a high degree of water from the column in the regeneration step, it had to be separated before recycling it back into the column; this is an expensive operation. Otherwise, the water left in the resin after the regeneration step (largely because of recycling dilute acetic acid) may react in the breakthrough experiments with the purified triacetine to produce diacetine, which could affect process efficiency. This will affect the performance of the chromatographic reactor, where a good separation between triacetine and diacetine is achieved when a well-regenerated column was used. Therefore, the usage of either a dry acetic acid or desulfonated resins (lesser affinity towards water) was proposed by Gelosa et al. [134]. Gelosa et al. [134] carried out esterification of glycerol with acetic acid in the presence of Amberlyst-15 in a chromatographic reactor to address the concern of the column regeneration mentioned above with a possible solution. For these experiments, a reactive adsorbent (i.e. a mixture of acetic acid and acetic anhydride) were used as desorbents instead of pure acetic acid. Acetic anhydride reacts with water to produce acetic acid (the desorbent itself) in the regeneration step; this resulted in the enhancement of process efficiency and a reduction in the desorbent requirements.

36 | 1 Catalysis in Multifunctional Reactors

1.5.3.1.2 Methyl Acetate Synthesis (Esterification of acetic acid with methanol) Yu et al. [133] carried out the synthesis of methyl acetate in a chromatographic reactor in the presence of Amberlyst-15. Both reactive and non-reactive experiments were performed in a chromatographic reactor. Adsorption parameters were calculated from non-reactive experiments, while kinetic parameters were obtained from reactive experiments. Methanol was used as a carrier solvent. A mixture of methyl acetate and water dissolved in methanol were used as a feed for the non-reactive breakthrough experiments, while a binary mixture of acetic acid and water were used as a feed for reactive breakthrough experiments. The experiments were carried out at different temperatures, feed concentrations and flow rates. The samples were taken from the column outlet at regular intervals and were then analyzed. Methanol was used for the regeneration of the resin. A quasi-homogeneous (QH) kinetic model was developed that assumed the reaction in the polymer phase to be homogeneous because of the presence of a large volume of methanol in the reaction mixture. It was observed that methyl acetate has less affinity towards the resin than water, and that the calculated adsorption constants of water and methyl acetate decreased with an increase in the temperature (since adsorption is an exothermic process). The model predicted the experimentally measured breakthrough curves very well. However, for non-reactive breakthrough curves, the model was able to predict the experimental results for methyl acetate very well – but not for water. The reason reported was that the tailing effect may be responsible for this; the use of non-linear adsorption isotherm was purposed. Yu et al. [133] carried out synthesis of methyl acetate in the presence of Amberlyst-15 in a simulated moving-bed reactor (SMBR). Four jacketed steel columns in series were used and each column was connected to four rotary valves actuated by the control system. Methanol was used as a mobile phase. The effect of different switching times, feed and desorbent flow rates were studied in detail; observations in respect to various parameters were reported.

1.5.3.1.3 n-Hexyl Acetate Synthesis (Esterification of acetic acid with n-hexanol) Synthesis of n-hexyl acetate by esterification of acetic acid with n-hexanol using a gelular ion exchange resin catalyst (Purolite® CT-124) was studied in batch and continuous chromatographic reactors [144–147]. Batch chromatographic reactor column (BCRC) experiments were carried out using different parameters such as feed flow rate, feed mole ratio of n-hexanol to acetic acid, desorbent (n-hexanol) flow rate and reaction step time to maximize the formation of n-hexyl acetate as well as to achieve complete conversion of acetic acid. It was found that an increase in reaction step time increases unreacted acetic acid, whereas an increase in feed molar ratio of n-hexanol to acetic acid decreases unreacted acetic acid at the BCRC outlet. It was also observed that an increase in temperature increases the desorption rate of the product (n-hexyl acetate) so that it reaches the BCRC outlet quicker. Also, an increase in feed flow rate was found to

1.5 Chromatographic Reactor

|

37

decrease the residence time of acetic acid to convert to n-hexyl acetate. Complete conversion (100 %) of acetic acid was achieved in the BCRC at a reaction temperature of 353 K, feed molar ratio (n-hexanol:acetic acid) of 3 : 1, feed flow rate of 0.2 ml/min, and reaction and regeneration step time of 75 min each, respectively. The outlet fluid from the BCRC can be collected as products from both the reaction and the regeneration step experiments for the above-mentioned optimized parameters. The variable flow rate of n-hexanol in the regeneration step experiment was found to be the best way to minimize the solvent used in the BCRC. A continuous chromatographic reactor column (CCRC) was designed and constructed on the basis of BCRC experimental results. Since the CRC used for BCRC was fully packed with ion exchange resin, the bed volume is actually the volume of the CRC used, and hence, the bed volume was used as a scale up for the design of CCRC. An experiment with the optimized parameters with three CRCs for the reaction step experiment and one CRC for the regeneration step experiment was used for the synthesis of n-hexyl acetate in CCRC. The experiments carried out in a CCRC correlate very well with the results obtained from the optimized reaction condition in a BCRC for maximum formation of n-hexyl acetate and complete conversion of acetic acid.

1.5.3.1.4 β-Phenethyl Acetate Synthesis (Esterification of acetic acid with β-Phenethyl alcohol) Synthesis of β-phenethyl acetate by esterification of acetic acid and β-phenethyl alcohol was investigated by Kawase et al. [128] in a simulated moving-bed reactor (SMBR). Amberlyst-15 catalyst was packed in eight stainless steel columns (in series) for this purpose. Each column had five solenoid valves to which desorbent, feed, extract, raffinate and effluent lines were connected. 1,4-Dioxane was used as a desorbent. It was reported that overall conversion in the range of 100 % was achieved by the application of simulated moving-bed reactor if the following criteria was satisfied – the products should be separated out chromatographically and the reaction rate should be fast enough so that the reactant does not elute from the column outlet.

1.5.3.1.5 Methyl Acrylate Synthesis (Esterification of acrylic acid with methanol) Acrylic esters are monomers that are widely used for the production of coatings, adhesives, plastics, etc. A conventional acrylic ester production involves several distillation columns (a reactor column, a water removal column, an azeotropic column and a column to separate the desired product from the undesired or by-products). In addition, an inhibitor has to be used in the whole process to minimize polymerization of acrylic acid acrylic ester and to avoid its local depletion. Moreover, reduced pressures are employed in the distillation columns to reduce the boiling temperatures. Besides polymerization and fouling, one of the major problems of this process is the thermodynamic limitation due to the reaction equilibrium. Due to the drawbacks of

38 | 1 Catalysis in Multifunctional Reactors

the current production process, an increasing interest can be seen to develop alternative production technologies. The option to employ a heterogeneous catalyst, e.g. an ion exchange resin, makes the use of integrated reactor-separator processes like reactive distillation or reactive chromatography feasible. Since the former process would have similar drawbacks as the conventional one, i.e., fouling and polymerization due to elevated temperatures, the latter could be a viable option since the separation is accomplished by selective adsorption in the liquid phase. For the above reason, esterification of acrylic acid with methanol for the production of methyl acrylate in the presence of Amberlyst-15 in a chromatographic reactor was carried out by Strohlein et al. [136]. A jacketed batch column was filled with Amberlyst-15 in the hydrogen form immersed in methanol. The composition of the column outlet was analyzed by gas chromatography. Methanol was used for the regeneration of the column. The batch column property was characterized by determining the total column porosity using tracer experiments. The effect of various parameters such as feed compositions, flow direction and flow rates have been studied for modeling the chromatographic reactor. A heterogeneous kinetic model, lumped kinetics and a linear driving force transport model have been developed. A more dispersed breakthrough and a sharp desorption profile was seen for the top-down flow and viceversa for the bottom-up flow as observed by Strohlein et al. [136]. Also, it was found that water adsorbs more strongly than the other components.

1.5.3.2 Hydrolysis Reaction Hydrolysis of four esters (i.e. methyl formate, methyl acetate, ethyl formate and ethyl acetate) was carried out in a chromatographic reactor with a Dowex 50W-X8 catalyst [159]. An HPLC column was used for carrying out the experiments. Water was used as the mobile phase. Various parameters such as temperature, flow rate, feed concentration and injection volume were varied. Following observations were reported by Mai et al. [159]: (i) the hydrolysis reaction of methyl formate and ethyl formate was faster when compared to methyl acetate and ethyl acetate; (ii) even at a low flow rate, methyl acetate and ethyl acetate eluted from the column again, thus confirming that these were the slowest reactions compared to methyl formate and ethyl formate; (iii) the peaks width was reduced considerably as the temperature was increased and in contrast, there was no major effect of temperature on the retention times in the studied range, i.e. 298–328 K, and (iv) the heterogeneous rate constant decreases in the order of methyl formate, ethyl formate, methyl acetate and ethyl acetate. Similarly, heterogeneously catalyzed hydrolysis reaction of methyl formate and methyl acetate in the presence of Dowex 50W-X8 was studied by Vu et al. [160]. It was reported that for the reaction 2A ⇔ B + C, complete conversion and separation were only possible if reactant A has an intermediate adsorptivity. On the other hand, for the reactions A ⇔ B + C, complete conversion and separation were possible for any order of adsorptivities.

1.5 Chromatographic Reactor

|

39

Falk and Seidel-Morgenstern [161] also carried out hydrolysis of methyl formate in the presence of Dowex 50W-X8 in a chromatographic reactor. The effect of temperature, residence time, feed, concentration and cycle time on the performance of the reactor was evaluated. The pseudo-homogeneous (PH) model was able to predict the experimental data over a wide range of parameters. Yu et al. [133] carried out hydrolysis of methyl acetate in a chromatographic reactor in the presence of Amberlyst-15. Both reactive and non-reactive experiments were performed in a chromatographic reactor. Adsorption parameters were calculated from non-reactive experiments while kinetic parameters were obtained from reactive experiments. Water was used as a carrier solvent. A mixture of methanol (or acetic acid) dissolved in water was used as a feed for the non-reactive breakthrough experiments, while a mixture of acetic acid and methanol dissolved in water or a binary mixture of methyl acetate and water was used as a feed for reactive breakthrough experiments. Experiments were carried out at different temperatures, feed concentrations and flow rates. The samples were taken from the column outlet at regular intervals and were analyzed. Water was used for the regeneration of the resin. The quasi-homogeneous (QH) kinetic model was developed assuming the reaction in the polymer phase is homogeneous because of the presence of a large volume of water in the reaction mixture. The model predicted the experimentally measured breakthrough curves very well. It was reported that the reaction equilibrium constant of the hydrolysis of methyl acetate increased with an increase in the temperature, since the backward reaction is an endothermic reaction.

1.5.3.3 Hydroxylation Hydroquinone (HQ) and catechol (CT) are used in photographic processing and polymerization inhibitors. Both HQ and CT can be produced by hydroxylation of phenol along with benzoquinone (BQ). Rangsunvigit and Kulrathipanja [141] studied phenol hydroxylation in the presence of TS-1 catalyst (structure similar to silicalite) in a chromatographic reactor for the production of HQ and CT. Stainless steel columns were used and the samples from the column outlet were collected in a fraction collector, which was analyzed by an HPLC. Unreacted H2 O2 was analyzed by a H2 O2 kit. Aqueous mixture of H2 O2 or water was used as a desorbent. It was reported that TS-1 could be easily regenerated by using pure water. Rangsunvigit and Kulrathipanja [141] also reported that the separation of each product can be achieved with water as a desorbent, depending upon the amount of phenol in the feed and that the selectivity of CT on TS-1 was found to be concentration-dependent. Tab. 1.5 shows the summary of some of the applications of chromatographic reactors.

Catalyst/Adsorbent Dextransucrose Amberlyst-15 Amberlyst-15 Finex KEF76 Amberlyst-15 Amberlyst-15 Purolite® CT-124 Amberlyst-15 Proton type ion exchange resins Amberlyst-15 Amberlyst-15 Amberlyst-15 Platinum supported on alumina Dowex 50W-X8 Dowex 50W-X8 Invertase β-fructofuranosidase Metal catalyst Oxide type of catalyst Immobilized isomerise ZSM-4 Amberlyst-15 Amberlyst-31 Amberlyst-15 Amberlyst-15 Amberlyst-15

Systems

Biosynthesis of dextran from sucrose Esterification of acetic acid and butyl cellosolve Esterification of acetic acid and ethanol (ethyl acetate synthesis) Esterification of acetic acid with ethanol/methanol (ethyl acetate/methyl acetate synthesis) Esterification of acetic acid and 2-ethylhexanol (2-ethylhexyl acetate synthesis) Esterification of acetic acid and glycerol (glycerine acetate synthesis) Esterification of acetic acid and n-hexanol (n-hexyl acetate synthesis) Esterification of acetic acid and methanol (methyl acetate synthesis) Esterification of acetic acid and β-phenethyl alcohol (β-phenethyl acetate synthesis) Esterification of acetic acid and n-propanol (n-propyl acetate synthesis) Esterification of acrylic acid and methanol (methyl acrylate synthesis) Esterification of lactic acid and ethanol (ethyl lactate synthesis) Hydrogenation of 1,3.5-trimethyl benzene Hydrolysis of methyl formate Hydrolysis of methyl acetate / methyl formate Inversion of sucrose Lactosucrose synthesis Methanol synthesis from syngas Oxidative coupling of methane Isomerization of glucose Isomerization of p-xylene Synthesis of Acetals from acetaldehyde and ethanol/butanol Synthesis of bisphenol-A from acetone and excess phenol Synthesis of diethylacetal from acetaldehyde and ethanol Synthesis of diethylacetal from acetaldehyde and methanol Synthesis of MTBE

Tab. 1.5: Some of the applications of chromatographic reactors.

[162] [163] [126] [164] [137] [135] [144–147] [150, 154] [128] [138] [136] [165] [166, 167] [161] [160] [168] [169] [130] [170] [171] [172] [173] [127] [174, 175] [176] [31]

References

40 | 1 Catalysis in Multifunctional Reactors

1.6 Summary |

41

1.6 Summary There has been considerable progress in the development of catalysts or catalytic systems used in multifunctional reactors. Special consideration needs to be given to their design to meet requirements from the point of view of both catalysis and reactor engineering. The associated effects on the reactor level such as heat/mass transfer effects, pressure drop and mixing become crucial in many cases. The literature on catalysis in reactive distillation is dominated by ion exchange resin catalysts. However, new catalysts such as ionic liquid, biocatalyst, zeolites micro-engineered systems have also been tested successfully in laboratory and/or on pilot scales. Computational tools such as CFD are being used effectively to meet the above-mentioned objectives. Wetting characteristics may be modified conveniently to meet the separation requirement as explained in the case of reactive stripping for an isotope exchange. Membrane reactors offer unique opportunities from the point of view of clever catalyst design, and its integration into the membrane structure thus brings compactness to the unit. Chromatographic reactors are particularly advantageous to use as an alternative to RD when the difference in the volatility of the species are small or the components are non-volatile or sensitive to temperature. The catalyst/adsorbent used for the chromatographic reactor may be the same or a mixture of solid catalyst and adsorbent. Also, the catalyst itself may act both as a catalyst for the reaction mixture and an adsorbent for the separation. In spite of the considerable work in the last two decades, there are enough challenges posed by the reaction of specific complexities and hence the field is still wide open for further research and development.

Acknowledgment We wish to acknowledge the work of many postgraduate and post-doctoral researchers who have contributed to studies in our research groups over a period of many years. In particular, we would like to mention the important contributions by Dr. Bhoja Reddy (IIT Bombay), Dr. Misbahu Ladan Mohammed (LSBU), Dr. Dipesh Patel (LSBU), Dr. Rene Mbeleck (LSBU), Dr. Krzysztof Ambroziak (Loughborough University), Dr. Praveen Ghodke (IIT Bombay) and Victor Nnamdi Onyenkeadi (LSBU).

References [1] [2]

Agar DW. Multifunctional reactors?: Old preconceptions and new dimensions. Chem Eng Sci 1999,54:1299–1305. Sharma MM, Mahajani SM. Industrial applications of reactive distillation in reactive distillation: Status and future directions. Weinheim, Germany, Wiley VCH, 2003.

42 | 1 Catalysis in Multifunctional Reactors

[3] [4] [5] [6] [7] [8] [9] [10] [11] [12]

[13] [14]

[15] [16] [17] [18] [19] [20]

[21] [22]

[23] [24]

Ertl G, Knozinger H, Weitkamp J. Handbook of heterogeneous catalysis volume 3. Weinheim, Germany, Wiley VCH, 1997. Noeres C, Kenig EY, Gorak A. Modelling of reactive separation processes: Reactive absorption and reactive distillation. Chem Eng Process 2003,42:157–178. Sundmacher K, Rihko LK, Hoffmann U. Classification of reactive distillation processes by dimensionless numbers. Chem Eng Commun 1994,127:151–167. Agreda VH, Partin LR. Reactive distillation process for the production of methyl acetate, US4,435,595 (Patent), 1984. Krishna R, Baten JMV, Ellenberger J, Higler AP, Taylor R. CFD simulations of sieve tray hydrodynamics. Chem Eng Res Des 1999,77:639–644. Taylor R, Krishna R. Modelling reactive distillation. Chem Eng Sci 2000,55:5183–5229. Saha B, Chopade SP, Mahajani SM. Recovery of dilute acetic acid through esterification in a reactive distillation column. Catal Today 2000,60:147–157. Moritz P, Hasse H. Fluid dynamics in reactive distillation packing katapak-s. Chem Eng Sci 1999,54:1367–1374. Lebens PJM, Kapteijn F, Sie ST, Moulijn JA. Potentials of internally finned monoliths as a packing for multifunctional reactors. Chem Eng Sci 1999,54:1359–1365. Trubac RE, Dautzenberg FM, Griffin TA, Paikert B, Schmidt VR, Overbeek RA. Micro-engineered catalyst systems: ABB’s advancement in structured catalytic packings. Catal Today 2001, 69:17–24. Quang DV, Amigues P, Gaillard J, Leonard J, Nocca J. Process for manufacturing a tertiary alkyl ether by reactive distillation, US4847430 (Patent), 1989. Sundmacher K, Hoffmann U. Multicomponent mass and energy transport on different length scales in a packed reactive distillation column for heterogeneously catalysed fuel ether production. Chem Eng Sci 1994,49:4443–4464. Wang EQ, Li CY, Wen LY, Du ZX, Zhang YQ. Simulation of cumene synthesis by suspension catalytic distillation. Adv Mater Res 2012,557-559,2243–2248. Lei Z, Li C, Li J, Chen B. Suspension catalytic distillation of simultaneous alkylation and transalkylation for producing cumene. Sep Purif Technol 2004,34:265–271. Wang E, Li C, Wen L, Min E. Study on suspension catalytic distillation for synthesis of linear alkylbenzene. AIChE J 2005,51:845–853. Erqiang W, Chengyue L. Simulation of suspension catalytic distillation for synthesis of linear alkylbenzene. Chinese J Chem Eng 2003,11:520–525. Han SJ, Jin Y, Yu ZQ. Application of a fluidized reaction-distillation column for hydrolysis of methyl acetate. Chem Eng J 1997,66:227–230. Hiwale RS, Bhate N V, Mahajan YS, Mahajani SM. Industrial applications of reactive distillation: Recent trends industrial applications of reactive distillation: Recent trends. Int J Chem Kinet 2004,2:1–52. Nuemann R, Sasson Y. Recovery of dilute acetic acid by esterification in a packed chemorectification column. Ind Eng Chem Process Des Dev 1984,23:654–659. Pöpken T, Steinigeweg S, Gmehling J. Synthesis and hydrolysis of methyl acetate by reactive distillation using structured catalytic packings: experiments and simulation. Ind Eng Chem Res 2001,40:1566–1574. Choi JI, Hong WH. Recovery of lactic acid by batch distillation with chemical reactions using ion exchange resin. J Chem Eng Japan 1999,32:184–189. Calvar N, González B, Dominguez A. Esterification of acetic acid with ethanol: Reaction kinetics and operation in a packed bed reactive distillation column. Chem Eng Process 2007,46:1317–1323.

References |

[25] [26]

[27] [28] [29] [30] [31] [32] [33] [34] [35]

[36] [37] [38] [39]

[40] [41]

[42] [43]

[44]

[45]

43

Patidar P, Mahajani SM. Esterification of fusel oil using reactive distillation – part I: reaction kinetics. Chem Eng J 2012,207-208,377–387. Gangadwala J, Mankar S, Mahajani S, Kienle A, Stein E. Esterification of acetic acid with butanol in the presence of ion-exchange resins as catalysts. Ind Eng Chem Res 2003,42: 2146–2155. Gadewar SB, Malone MF, Doherty MF. Feasible region for a countercurrent cascade of vaporliquid CSTRs. AIChE J 2002,48:800–814. Buchaly C, Kreis P, Andrzej G. n-Propyl propionate synthesis via catalytic distillation – experimental investigation in pilot-scale. Ind Eng Chem Res 2012,51:891–899. Chiang S, Kuo C, Yu C, Wong DSH. Design alternatives for the amyl acetate process?: Coupled reactor/column and reactive distillation. Ind Eng Chem Res 2002,41:3233–3246. Saha B, Teo HTR, Alqahtani, A. iso-Amyl acetate synthesis by catalytic distillation, International Journal of Chemical Reactor Engineering 2005, 3 (A 11), 1-14. Patidar P, Mahajani S. Entrainer-based reactive distillation for the synthesis of 2-ethylhexyl acetate. Ind Eng Chem Res 2012,51:8748–8759. Smejkal Q, Hanika J, Kolena J. 2-Methylpropylacetate synthesis in a system of equilibrium reactor and reactive distillation column. Chem Eng Sci 2001,56:365–370. Schmitt M, Hasse H, Althaus K, Schoenmakers H, Götze L, Moritz P. Synthesis of n-hexyl acetate by reactive distillation. Chem Eng Process 2004,43:397–409. Zeng KL, Kuo CL, Chien IL. Design and control of butyl acrylate reactive distillation column system. Chem Eng Sci 2006,61:4417–4431. Saha B, Sharma MM. Esterification of formic acid, acrylic acid and methacrylic acid with cyclohexene in batch and distillation column reactors?: Ion-exchange resins as catalysts. React Funct Polym 1996,28:263–278. Steinigeweg S, Gmehling J. Esterification of a fatty acid by reactive distillation. Ind Eng Chem Res 2003,42:3612–3619. Jacobs R, Krishna R. Multiple solutions in reactive distillation for methyl tert-butyl ether synthesis. Ind Eng Chem Process Des Dev 1993,32:1706–1709. Bravo JL, Pyhalahti A. Investigations in a catalytic distillation pilot plant: Vapor/liquid equilibrium, kinetics, and mass-transfer issues. Ind Eng Chem Res 1993,32:2220–2225. Oudshoom OL, Janissen M, Kooten WEJ Van, Jansen J, Bekkuin H Van, Bleek CM Van Den, et al. A novel structured catalyst packing for catalytic distillation of ETBE. Chem Eng Sci 1999, 54:1413–1418. Marker TL, Funk GA, Barker PT, Hammershaimb HU. Two-stage process for producing diisopropyl ether using catalytic distillation, US5,744,645 (Patent), 1998. Rihko-Struckmann LK, Karinen RS, Jakobsson K, Aittamaa JR. Process configurations for the production of the 2-methoxy-2, 4, 4- trimethylpentane - a novel gasoline oxygenate. Chem Eng Process 2004,43:57–65. Aiouache F, Goto S. Reactive distillation–pervaporation hybrid column for tert-amyl alcohol etherification with ethanol. Chem Eng Sci 2003,58:2465–2477. Hagen GP, Spangler MJ. Preparation of polyoxymethylene dialkane ethers, by catalytic conversion of formaldehyde formed by dehydrogenation of methanol or dimethyl ether. US6,350,919 (Patent), 2002. Chopade SI, Sharma MM. Reaction of ethanol and formaldehyde: Use of versatile cationexctiange resins as catalyst in batch reactors and reactive distillation columns. React Funct Polym 1997,32:53–64. Hendriksen DE, Mcglamery GG, Keenan, Michael J, Pete DD. Process for preparing 3-methyl2-pentene. US5731486 (Patent), 1998.

44 | 1 Catalysis in Multifunctional Reactors

[46] [47]

[48] [49] [50] [51]

[52] [53] [54]

[55]

[56]

[57]

[58] [59]

[60]

[61]

[62] [63] [64]

Bessling B, Knab W, Brox W, Lohe B. Method and device for obtaining isobutenes from conjugated hydrocarbons. US6,362,386 (Patent), 2002. Jimenez L, Costa-Lopez J. The production of butyl acetate and methanol via reactive and extractive distillation. II Process modeling, dynamic simulation and control strategy. Ind Eng Chem Res 2002,41:6735–6744. Saari K, Tirronen E, Vuori A, Lahtinen M. Method for preparing formic acid. US6,429,333 (Patent), 2002. Ciric AR, Miao P. Steady state multiplicities in an ethylene glycol reactive distillation column. Ind Eng Chem Res 1994,33:2738–2748. Zhang CM, Adesina AA, Wainwright MS. Isobutene hydration over Amberlyst-15 in a slurry reactor. Chem Eng Process 2003,42:985–991. Chopade SP, Sharma MM. Acetalization of ethylene glycol with formaldehyde using cationexchange resins as catalysts?: Batch versus reactive distillation. React Funct Polym 1997,34:37–45. Broekhuis RR, Lynn S, King CJ. Recovery of propylene glycol from dilute aqueous solutions via reversible reaction with aldehydes. Ind Eng Chem Res 1994,33:3230–3237. Lange J, Otten V. Dehydration of phenyl-ethanol to styrene?: Zeolite catalysis under reactive distillation 2006,238:6–12. Heils R, Sont A, Bubenheim P, Liese A, Smirnova I. Integration of enzymatic catalysts in a reactive distillation column with structured packings. Ind Eng Chem Res 2012,51:11482– 11489. Wang H, Wu C, Bu X, Tang W, Li L, Qiu T. A benign preparation of sec-butanol via transesterification from sec-butyl acetate using the acidic Imidazolium ionic liquids as catalysts. Chem Eng J 2014,246:366–372. Liu Y, Yang B, Li S. Process simulation based on experimental investigations for 3methylthiophene alkylation with isobutylene in a reactive distillation column. Ind Eng Chem Res 2012,51:9803–9811. Zekai Z, Hui J, Shenglin L, Qingxia W, Longya X. Alkylation performance of thiophene and its derivatives during olefinic alkylation of thiophenic sulfur in gasoline. Chinese J Catal 2006, 27:309–313. Bellie V, Geantet C, Vrinat M. Alkylation of 3-methylthiophene with 2-methyl-2-butene over a zeolitic Catalyst. Energy & Fuels 2004,18:1806–1813. Dupuy B, Laforge S, Morais C, Bachmann C, Magnoux P, Richard F. Alkylation of 3methylthiophene by 2-methyl-1-pentene over HY, HBEA and HMCM-22 acidic zeolites. Appl Catal A Gen 2012,413–414,192–204. Richard F, Boita T, Moreau M, Bachmann C, Pérot G. Transformation of thiophenic compounds over HY zeolite: Study of the acid-catalyzed isomerization and disproportionation mechanisms by quantum chemical calculations. J Mol Catal A Chem 2007,273:48–54. Dupuy B, Laforge S, Bachmann C, Magnoux P, Richard F. Desulfurization of model FCC feedstocks by alkylation: Transformation of thiophenic compounds in presence of 2-methyl1-pentene over acidic zeolites. J Mol Catal A Chem 2012,363-364,273–82. Wang R, Li Y. Preparation of MCM-41 supported phosphoric acid catalyst for thiophenic compounds alkylation in FCC gasoline. Catal Commun 2010,11:705–709. Shi R, Li Y, Wang R, Guo B. Alkylation of thiophenic compounds with olefins and its kinetics over MCM-41 supported phosphoric acid in FCC gasoline. Catal Letters 2010,139:114–122. Liu Y, Yi CH, Hu JY, Yang BL. Gasoline desulfurization with two catalytic distillation columns. Adv Mater Res 2012,550-553,550–3.

References |

[65]

[66]

[67] [68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78]

[79] [80] [81]

45

Guo B, Wang R, Li Y. The performance of solid phosphoric acid catalysts and macroporous sulfonic resins on gasoline alkylation desulfurization. Fuel Process Technol 2010,91:1731– 1735. Ha Y, Guo B, Li Y. Sensitivity and economic analysis of a catalytic distillation process for alkylation desulfurization of fluid catalytic cracking (FCC) gasoline. J Chem Technol Biotechnol 2014, accepted. Guo B, Li Y. Analysis and simulation of reactive distillation for gasoline alkylation desulfurization. Chem Eng Sci 2012,72:115–125. Arias M, Laurenti D, Bellière V, Geantet C, Vrinat M, Yoshimura Y. Preparation of supported H3PW12O40•6H2O for thiophenic compounds alkylation in FCC gasoline. Appl Catal A Gen 2008,348:142–147. Xu C, Yang K, Liu Z, Qin Z, He W, Dai Q, et al. Superparamagnetic supported catalyst H3PW12O40/γ-Fe2O3 for alkylation of thiophene with olefine. Chinese J Chem Eng 2014, 22:305–311. Richard F, Célérier S, Vilette M, Comparot JD, Montouillout V. Alkylation of thiophenic compounds over heteropoly acid H3PW12O40 supported on MgF2. Appl Catal B Environ 2014, 152–153:241–249. Mbeleck R, Ambroziak K, Saha B, Sherrington DC. Stability and recycling of a polymersupported Mo(VI) alkene epoxidation catalyst. Reactive and Functional Polymers 2007,67: 1448–1457. Ambroziak K, Mbeleck R, Saha B, Sherrington DC. Epoxidation of limonene by tert-butyl hydroperoxide catalyzed by polybenzimidazole-supported Mo(VI) complex. Journal of Ion Exchange 2007,18(4):452–457. Ambroziak K, Mbeleck R, Saha B, Sherrington D C.Greener and sustainable method for alkene epoxidations by polymer-supported Mo(VI) catalysts. International Journal of Chemical Reactor Engineering 2010,8(A125):1–13. Mohammed M L, Patel D, Mbeleck R., Niyogi D, Sherrington DC, Saha B. Optimisation of alkene epoxidation catalysed by polymer supported Mo(VI) complexes and application of artificial neural network for the prediction of catalytic performances. Applied Catalysis A 2013,466:142–152. Mohammed ML, Mbeleck R, Patel D, Sherrington DC, Saha B. Greener route to 4-vinyl-1-cyclohexane 1,2-epoxide synthesis using batch and continuous reactors. Green Processing and Synthesis 2014,3:411–418. Ambroziak K, Mbeleck R, He Y, Saha B, Sherrington DC. Investigations of batch alkenes epoxidation catalayzed by novel polymer-supported Mo(VI) complexes. Industrial and Engineering Chemistry Research 2009,48:3293–3302. Mohammed ML, Patel D, Mbeleck R, Niyogi D, Sherrington DC, Saha B, Greener and efficient epoxidation of 4-vinyl-1-cyclohexene with polystyrene 2-(aminomethyl)pyridine supported Mo(VI) catalyst in batch and continuous reactors. Chemical Engineering Research and Design 2015,94:194–203. Mbeleck R, Mohammed ML, Ambroziak K, Sherrington DC, Saha B. Efficient epoxidation of cyclododecene and dodecene catalysed by polybezimidazole-supported Mo(VI) complex. Catalysis Today 2015,256:287–293. Oku N, Seo T. Process for producing propylene oxide. US6646138 (Patent), 2003. Saha B, Ambroziak K, Sherrington DC, Mbeleck R. Liquid phase epoxidation process. US8,759,552 B2 (Patent), 2014. Saha B, Ambroziak K, Sherrington DC, Mbeleck R. A continuous process for the liquid phase epoxidation of an olefinic compound with an oxidant. ZL201080044175.2 (Patent), 2015.

46 | 1 Catalysis in Multifunctional Reactors

[82]

Holtbruegge J, Leimbrink M, Lutze P, Górak A. Synthesis of dimethyl carbonate and propylene glycol by transesterification of propylene carbonate with methanol. Catalyst screening, chemical equilibrium and reaction kinetics. Chem Eng Sci 2013,104:347–360. [83] Zhang D, Wei D, Ding W, Zhang X. Carbon-based nanostructured catalyst for biodiesel production by catalytic distillation. Catal Commun 2014,43:121–125. [84] Kiss AA, Rothenberg G. Sustainable biodiesel production by catalytic reactive distillation in catalysis of organic reactions. Taylor & Francis, USA, 2009. [85] Noshadi I, Amin NAS, Parnas RS. Continuous production of biodiesel from waste cooking oil in a reactive distillation column catalyzed by solid heteropolyacid: Optimization using response surface methodology (RSM). Fuel 2012,94:156–164. [86] Yu T, Zhang Q, Xia S, Li G, Hu C. Direct amination of benzene to aniline by reactive distillation method over copper doped hierarchical TS-1 catalyst. Catal Sci Technol 2014,4:639–647. [87] Wang C, Hu Y, Chen Q, Lv C, Jia S. Bio-oil upgrading by reactive distillation using p-toluene sulfonic acid catalyst loaded on biomass activated carbon. Biomass and Bioenergy 2013,56: 405–411. [88] Mueller I, Schildhauer TJ, Madrane A, Kapteijn F, Moulijn JA, Kenig EY. Experimental and theoretical study of reactive stripping in monolith reactors. Ind Eng Chem Res 2007,46:4149– 4157. [89] Yu S, Zhou A, Tan Q. Simulation of multistage catalytic stripping with a nonequilibrium stage model. Comput Chem Eng 1997,21:409–415. [90] Sundmacher K, Kienle A, Seidel-Morgenstern A. Integrated chemical processes. Weinheim, Germany, Wiley-VCH, 2005. [91] D’Angelo MFN, Ordomsky V, Van der Schaaf J, Schouten JC, Nijhuis TA. Aqueous phase reforming in a microchannel reactor: The effect of mass transfer on hydrogen selectivity. Catal Sci Technol 2013,3:2834–2842. [92] Neira D’Angelo MF, Ordomsky V, Van der Schaaf J, Schouten JC, Nijhuis TA. Continuous hydrogen stripping during aqueous phase reforming of sorbitol in a washcoated microchannel reactor with a Pt–Ru bimetallic catalyst. Int J Hydrogen Energy 2014,39:18069–18076. [93] Kumar R, Mohan S, Mahajani SM. Reactive stripping for the catalytic exchange of hydrogen isotopes. Ind Eng Chem Res 2013,52:10935–10950. [94] Agirrezabal-Telleria I, Guo Y, Hemmann F, Arias PL, Kemnitz E. Dehydration of xylose and glucose to furan derivatives using bifunctional partially hydroxylated MgF2 catalysts and N2 -stripping. Catal Sci Technol 2014,4:1357–1368. [95] Agirrezabal-Telleria I, García-Sancho C, Maireles-Torres P, Arias PL. Dehydration of xylose to furfural using a Lewis or Brönsted acid catalyst and N2 stripping. Chinese J Catal 2013, 34:1402–1406. [96] Nijhuis TA, Beers AEW, Kapteijn F, Moulijn JA. Water removal by reactive stripping for a solidacid catalyzed esteriÿcation in a monolithic reactor. Chem Eng Sci 2002,57:1627–1632. [97] Beers AEW, Spruijt RA, Nijhuis TA, Kapteijn F, Moulijn JA. Esterification in a structured catalytic reactor with counter-current water removal. Catal Today 2001,66:175–181. [98] Schildhauer TJ, Hoek I, Kapteijn F, Moulijn JA. Zeolite BEA catalysed esterification of hexanoic acid with 1-octanol: Kinetics, side reactions and the role of water. Appl Catal A Gen 2009,358: 141–145. [99] Gutmann B, Elsner P, Glasnov T, Roberge DM, Kappe CO. Shifting chemical equilibria in flowefficient decarbonylation driven by annular flow regimes. Angew Chem Int Ed Engl 2014,126: 11741–11745. [100] Ranade VV, Kelkar AA, Pane VH, Kinage AK, Shingote SK, Roy LS. Synthesis of methyl carbamate and dimethyl carbonate (DMC) in presence of stripping with inert gas or superheated vapours and reactor for the same. WO2014072803A1 (Patent), 2014.

References |

47

[101] Butler JP. Hydrogen isotope separation by catalyzed exchange between hydrogen and liquid water. Sep Sci Technol 1980,15:371–396. [102] Armor JN. Overcoming equilibrium limitations in chemical processes. Appl Catal A Gen 2001, 222:91–99. [103] Seidel-Morgenstern A. Membrane reactors: Distributing reactants to improve selectivity and yield. Weinheim, Germany, Wiley VCH, 2010. [104] Baroutian S, Aroua MK, Raman AA A, Sulaiman NMN. A packed bed membrane reactor for production of biodiesel using activated carbon supported catalyst. Bioresour Technol 2011, 102:1095–1102. [105] Xu W, Gao L, Wang S, Xiao G. Biodiesel production in a membrane reactor using MCM-41 supported solid acid catalyst. Bioresour Technol 2014,159:286–291. [106] Guerreiro L, Pereira PM, Fonseca IM, Martin-Aranda RM, Ramos AM, Dias JML, et al. PVA embedded hydrotalcite membranes as basic catalysts for biodiesel synthesis by soybean oil methanolysis. Catal Today 2010,156:191–197. [107] Weyten H, Keizer K, Kinoo A, Luyten J, Leysen R. Dehydrogenation of propane using a packedbed catalytic membrane reactor. AIChE J 1997,43:1819–1827. [108] Gbenedio E, Wu Z, Hatim I, Kingsbury BFK, Li K. A multifunctional Pd/alumina hollow fibre membrane reactor for propane dehydrogenation. Catal Today 2010,156:93–99. [109] Bhatia S, Thien CY, Mohamed AR. Oxidative coupling of methane (OCM) in a catalytic membrane reactor and comparison of its performance with other catalytic reactors. Chem Eng J 2009,148:525–532. [110] Lu Y, Dixon AG, Moser WR, Ma YH. Oxidative coupling of methane in a modified γ-alumina membrane reactor. Chem Eng Sci 2000,55:4901–4912. [111] Tian T, Wang W, Zhan M, Chen C. Catalytic partial oxidation of methane over SrTiO3 with oxygen-permeable membrane reactor. Catal Commun 2010;11:624–628. [112] Santos A, Coronas J, Santamaria J. Catalytic partial oxidation of methane to synthesis gas in a ceramic membrane reactor. Catal Letters 1995,30:189–199. [113] Paturzo L, Gallucci F, Basile A, Pertici P, Scalera N, Vitulli G. Partial oxidation of methane in a catalytic Ruthenium membrane reactor. Ind Eng Chem Res 2003,42:2968–2974. [114] Faroldi BM, Lombardo E A, Cornaglia LM. Ru/La2O3–SiO2 catalysts for hydrogen production in membrane reactors. Catal Today 2011,172:209–217. [115] Mironova EY, Ermilova MM, Orekhova NV, Muraviev DN, Yaroslavtsev AB. Production of high purity hydrogen by ethanol steam reforming in membrane reactor. Catal Today 2014,236: 64–69. [116] Liguori S, Iulianelli A, Dalena F, Piemonte V, Huang Y, Basile A. Methanol steam reforming in an Al2O3 supported thin Pd-layer membrane reactor over Cu/ZnO/Al2O3 catalyst. Int J Hydrogen Energy 2014,39:18702–18710. [117] Silva F A, Hori CE, da Silva AM, Mattos LV, Múnera J, Cornaglia L, et al. Hydrogen production through CO2 reforming of CH4 over Pt/CeZrO2/Al2O3 catalysts using a Pd–Ag membrane reactor. Catal Today 2012,193:64–73. [118] Ryi SK, Lee SW, Park JW, Oh DK, Park JS, Kim SS. Combined steam and CO2 reforming of methane using catalytic nickel membrane for gas to liquid (GTL) process. Catal Today 2014, 236:49–56. [119] Rei BMH, Yeh GT, Pan CW. Catalysis-spillover-membrane-2 [1]. The rate enhancement of methanol steam reforming reaction in a membrane catalytic reactor. Catal Today 2004,97: 167–172. [120] Wu Z, Hatim IMD, Kingsbury BFK, Gbenedio E, Li K. A novel inorganic hollow fiber membrane reactor for catalytic dehydrogenation of propane. AIChE J 2009,55:2389–2398.

48 | 1 Catalysis in Multifunctional Reactors

[121] Wu JCS, Gerdes TE, Pszczolkowski JL, Bhave RR, Liu PKT, Martin ES. Dehydrogenation of ethylbenzene to styrene using commercial ceramic membranes as reactors. Sep Sci Technol 1990, 25:1489–1510. [122] Kong C, Lu J, Yang J, Wang J. Catalytic dehydrogenation of ethylbenzene to styrene in a zeolite silicalite-1 membrane reactor. J Memb Sci 2007,306:29–35. [123] She Y, Han J, Ma YH. Palladium membrane reactor for the dehydrogenation of ethylbenzene to styrene. Catal Today 2001,67:43–53. [124] Yu C, Xu H. An efficient palladium membrane reactor to increase the yield of styrene in ethylbenzene dehydrogenation. Sep Purif Technol 2011,78:249–252. [125] Tan X, Pang Z, Gu Z, Liu S. Catalytic perovskite hollow fibre membrane reactors for methane oxidative coupling. J Memb Sci 2007,302:109–114. [126] Mazzotti M, Neri B, Gelosa D, Morbidelli M. Dynamics of a Chromatographic Reactor: Esterification Catalyzed by Acidic Resins. Industrial and Engineering Chemistry Research 1997,36(8): 3163–3172. [127] Kawase M, Yasunobu I, Araki T, Hashimoto K. The Simulated Moving-Bed Reactor for production of bisphenol A. Catalysis Today 1999,48(1–4);199–209. [128] Kawase M, Suzuki TB, Inoue K, Yoshimoto K, Hashimoto K. Increased esterification conversion by application of the simulated moving-bed reactor. Chemical Engineering Science 1996, 51(11):2971–2976. [129] Ray AK, Carr RW, Aris R. The simulated counter-current moving-bed chromatographic reactor: A novel reactor-separator. Chemical Engineering Science 1994,49(4):469–480. [130] Mazzzoti M, Kruglov A, Neri B, Gelosa D, Morbidelli M. A continuous chromatographic reactor: SMBR. Chemical Engineering Science 1996,51(10):1827–1836. [131] Zhang Z, Hidajat K, Ray AK. Application of simulated counter-current moving-bed chromatographic reactor for MTBE synthesis. Industrial and Engineering Chemistry Research 2001, 40(23):5305–5316. [132] Yu W, Hidajat K, Ray AK. Modelling, simulation, and experimental study of a simulated moving-bed reactor for the synthesis of methyl acetate ester. Industrial and Engineering Chemistry Research 2003,42(96):674 –6754. [133] Yu W, Hidajat K, Ray AK. Determination of adsorption and kinetic parameters for methyl acetate esterification and hydrolysis reaction catalyzed by amberlyst-15. Applied Catalysis A General, 2004,260(2):191–205. [134] Gelosa D, Ramaioli M, Valente G. Chromatographic reactors: Esterification of glycerol with acetic acid using acidic polymeric resins. Industrial and Engineering Chemistry Research 2003,42(25):6536–6544. [135] Gelosa D, Sliepcevich A., Morbidelli M. Chromatographic reactors with reactive desorbents. Industrial and Engineering Chemistry Research 2006,45(11):3922–3925. [136] Strohlein G, Assuncao Y, Dube N, Bardow A, Mazzotti M, Morbidelli M. Esterification of acrylic acid with methanol by reactive chromatography: Experiments and simulations. Chemical Engineering Science 2006,61(16):5296–5306. [137] Gyani V C, Mahajani SM. Reactive chromatography for the synthesis of 2-ethylhexyl acetate. Separation Science and Technology 2008,43(9-10):2245–2268. [138] Reddy B, Mahajani S. Feasibility of reactive chromatography for the synthesis of n-propyl acetate. Industrial and Engineering Chemistry Research 2014,53(4):1395–1403. [139] Falk T, Seidal-Morgenstern A. Comparison between a fixed-bed reactor and a chromatographic reactor. Chemical Engineering Science 1999,54(10):1479–1485. [140] Vu TD, Seidal-Morgenstern A, Gruner S, Kienle A. Analysis of ester hydrolysis reactions in a chromatographic reactor using equilibrium theory and a rate model. Industrial and Engineering Chemistry Research 2005,44(25):9565–9574.

References |

49

[141] Rangsunvigit P, Kulprathipanja S. Phenol hydroxylation with TS-1 in a chromatographic reactor. Separation Science and Technology 2004,39(14):3301–3315. [142] Juza M, Mazzotti M, Morbidelli M. Simulated moving-bed chromatography and its application to chirotechnology. Trends in Biotechnology 2000,18(3):108–118. [143] Gyani VC (Supervisor – Mahajani, SM). Reactive chromatography. PhD Thesis, Indian Institute of Technology, Mumbai, India, 2010. [144] Patel D, Saha B, Wakeman R. Optimisation of n-hexyl acetate synthesis in a chromatographic reactor. Journal of Ion Exchange 2010,21(4):388–391. [145] Patel D (Supervisor – Saha B). Synthesis of n-hexyl acetate in batch and chromatographic reactors. PhD Thesis, Loughborough University, UK, 2011. [146] Patel D, Saha B. Esterification of acetic acid with n-hexanol in batch and continuous chromatographic reactors using a gelular ion exchange resin as a catalyst. Industrial and Engineering Chemistry Research 2012a,51(37):11965–11974. [147] Patel D, Saha B. Synthesis of n-hexyl acetate in batch and continuous chromatographic reactors using a dual property based gelular ion exchange resin catalyst. IEX 2012 – Developments and Applications of Ion Exchange, Cambridge, UK, 2012b,73-74 [CD-ROM]. [148] Broughton DB, Gerhold CG. Continuous sorption process employing fixed-bed of sorbent and moving inlets and outlets. US2985589 (Patent), 1961. [149] Shieh MT, Barker PE. Saccharification of modified starch to maltose in a semi-continuous counter-current chromatographic reactor-separator (SCCR-S). Journal of Chemical Technology and Biotechnology 1995,63:125–134. [150] Lode F, Francesconi G, Mazzoti M, Morbidelli M. Synthesis of methyl acetate in simulated moving-bed reactor: Experiment and modelling. AIChE Journal 2003,49(6):1516–1524. [151] Den Hollander J L, Stribos BI, Van Buel MJ, Luyben K, Ch AM., Van Der Wielen LAM. Centrifugal partition chromatographic reaction for the production of chiral amino acids. Journal of Chromatography B 1998,711(1–2):223–235. [152] Den Hollander JL, Wong YW, Luyben K, Ch AM, Van Der Wielen LAM. Non-separating effects in a centrifugal partition chromatographic reactor for the enzymatic production of l-amino acids. Chemical Engineering Science 1999,54(15–16):3207–3215. [153] Howard AJ, Carta G, Byers C H. Separation of sugars by continuous annular chromatography. Industrial and Engineering Chemistry Research 1988,27:1873–1882. [154] Lode F, Houmard C, Migliorini C, Mazzoti M, Morbidelli M. Continuous reactive chromatography. Chemical Engineering Science 2001,56(2):269–291. [155] Agar DW, Ruppel W. Extended reactor concept for dynamic denox design. Chemical Engineering Science 1988,43(8):2073–2078. [156] Jeong YO, Luss D. Pollutant destruction in a reverse flow chromatographic reactor. Chemical Engineering Science 2003,58(7):1095–1102. [157] Viecco GA, Caram HS. Analysis of the reverse flow chromatographic reactor. AIChE Journal 2004,50(9):2266–2275. [158] Viecco GA, Caram HS. Use of a reverse flow chromatographic reactor to enhance productivity in consecutive reaction systems. Industrial and Engineering Chemistry Research 2005,44:3396–3401. [159] Mai PT, Vu TD, Mai KX, Seidal-Morgenstern A. Analysis of heterogeneously catalyzed ester hydrolysis performed in a chromatographic reactor and in a reaction calorimeter. Industrial and Engineering Chemistry Research 2004,43(16):4691–4702. [160] Vu TD, Seidal-Morgenstern A. Quantifying temperature and flow rate effects on the performance of a fixed-bed chromatographic reactor. Journal of Chromatography A 2011,1218: 8097–8109.

50 | 1 Catalysis in Multifunctional Reactors

[161] Falk T, Seidal-Morgenstern A. Analysis of a discontinuously operated chromatographic reactor. Chemical Engineering Science 2002,57(9):1599–1606. [162] Zafer I, Barker PE. An experimental and computational study of a biochemical polymerisation reaction in a chromatographic reactor separator. Chemical Engineering Science 1988,43(9): 2369–2375. [163] Deshmukh KS, Gyani VC, Mahajani SM. Esterification of butyl cellosolve with acetic acid using ion exchange resin in fixed bed chromatographic reactors. International Journal of Chemical Reactor Engineering 2009, 7, Article A2. [164] Sainio T, Zhang L, Seidel-Morgenstern A. Adiabatic operation of chromatographic fixed-bed reactors. Chemical Engineering Journal 2011,168:861–871. [165] Pereira CSM, Zabka M., Silva VMTM, Rodrigues AE. A novel process for the ethyl lactate synthesis in a simulated moving bed reactor (SMBR). Chemical Engineering Science 2009,64: 3301–3310. [166] Fish BB, Carr RW. An experimental study of the counter-current moving-bed chromatographic reactor. Chemical Engineering Science 1989,44(9):1773–1783. [167] Ray AK, Carr RW, Aris R. The simulated counter-current moving-bed chromatographic reactor: A novel reactor-separator. Chemical Engineering Science 1994,49(4):469–480. [168] Meurer M, Altenhoner U, Strube J, Unitied A, Schmidt-Traub H. Dynamic simulation of a simulated moving-bed chromatographic reactor for the inversion of sucrose. Starch–Starke 1996, 48(11–12):452–457. [169] Kawase M, Pilgrim A, Araki T, Hashimoto K. Lactosucrose production using simulated movingbed reactor. Chemical Engineering Science 2001,56(2):453–458. [170] Tonkovich ALY, Carr RW. A simulated counter-current moving-bed chromatographic reactor for the oxidative coupling of methane – experimental results. Chemical Engineering Science 1994,49(24A):4647–4656. [171] Hoshimoto K, Adachi S, Nougima H, Ueda Y. A new process combining adsorption and enzyme reaction for producing higher-fructose syrup. Biotechnology and Bioengineering 1983, 25(10):2371–2393. [172] Minceva M, Gomes PS, Meshko V, Rodrigues A E. Simulated moving-bed reactor for isomerisation and separation of p-xylene. Chemical Engineering Journal 2008,140(1–3):305–323. [173] Pereira CSM, Rodriques AE. Process intensification: New technologies (SMBR and PermSMBR) for the synthesis of acetals. Catalysis Today 2013,218-219:148–152. [174] Silva VMTM, Rodrigues AE. Dynamics of a fixed bed adsorptive reactor for synthesis of diethylacetal. AIChE Journal 2002,48(3):625–634. [175] Silva VMTM, Rodrigues AE. Novel process for diethylacetal synthesis. AIChE Journal 2005, 51(10):2752–2768. [176] Pereira CSM, Sa Gomes P, Gandi GK, Silva VMTM, Rodrigues AE. Multi-functional reactor for the synthesis of dimethylacetal. Industrial and Engineering Chemistry Research 2008,47(10): 3515–3524.

Goran T. Vladisavljević

2 Biocatalytic membrane reactors (BMR) Nomenclature Abbreviations of Chemicals AAm acrylamide 6-APA 6-aminopenicillanic acid ATEE N-acetyl tyrosine ethyl ester bPG benzylpenicillin CDI carbonyl diimidazole cis-DE cis-1,2-dimethylcyclohex-4-ene-1,2-dicarboxylate DCP 2,4-dichlorophenol DD dodecane HT hydrotalcite 1-PEA 1-phenylethyl acetate PAA poly(acrylic acid) PhAA phenylacetic acid PAN polyacrylonitrile PBA poly(normal-butyl acrylate) PEI polyethyleneimine PES polyethersulfone PET poly(ethylene terephthalate) PFMD perfluoromethyldecalin PDMS polydimethylsiloxane POPC 1-palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine POPS 1-palmitoyl-2-Oleoyl-sn-Glycero-3-Phospho-L-Serine PS polysulfone PP polypropylene PTFE polytetrafluoroethylene SO silicone oil T toluene TG triglycerides THF tetrahydrofuran TPP tripolyphosphate Other abbreviations EMBR extractive membrane bioreactor BEMBR biphasic extractive membrane bioreactor FFMD flow focusing microfluidic device IEM ion exchange membrane ILM immobilised liquid membrane MF microfiltration MSR membrane separation reactor UF ultrafiltration PTC phase-transfer catalysis or catalyst SC supercritical VOC volatile organic campound

52 | 2 Biocatalytic membrane reactors (BMR)

2.1 Introduction The term membrane reactor first began to appear in the chemical engineering literature around 1980 [1]. Although there is no commonly accepted definition of a membrane reactor, the term usually refers to membrane devices whose function is to perform net chemical conversion under conditions in which the unique contacting and separation features of membranes and membrane devices are exploited. In particular, the term membrane reactor is reserved for those processes wherein the membrane functions as more than simply a reactive membrane, i.e. a membrane matrix used for catalyst immobilization [1]. The biocatalytic membrane reactor (BMR) is a device in which biochemical transformations catalyzed by enzymes or cells are combined with permeation or mass transfer through the membrane. Since the catalytic efficiency of biochemical systems in vivo can often be attributed to the action of membrane-bound enzymes, it is not surprising that membrane-aided biocatalysis using synthetic membranes is an increasingly popular topic of research in engineering circles. As noticed by Matson and Quinn [1], any aspects of the structure and function of biological membranes that have been elucidated over the past several decades make it clear that nature has designed a highly integrated chemical plant that engineers might profitably mimic.

2.2 Role of membrane in biocatalytic membrane reactors (BMRs) The most common types of BMRs are summarized in Fig. 2.1. The role of synthetic membranes in BMRs is to serve: (a) solely as a separation barrier to achieve separation of the product from a biocatalyst; (b) both as a separation barrier and a medium for biocatalyst immobilization; (c) as a medium for creating and maintaining high interfacial area per unit volume to conduct non-dispersive gas/liquid and liquid/liquid mass-transfer processes; and (d) as a medium (carrier) for enzyme or cell encapsulation. If the membrane serves only as a semipermeable barrier, there are two modes of operations with regard to the location of the biocatalyst in the membrane system: (i) The biocatalyst is suspended in a stirred tank reactor and the reaction mixture is continuously brought into contact with the membrane for product removal. In this mode of operation, the stirred tank functions as a stirred tank reactor (STR); (ii) the biocatalyst is suspended only within the membrane module, where a biochemical conversion takes place. In the latter case the membrane acts as a medium for biocatalyst segregation; the membrane module functions as a reactor [2]. If the membrane is a medium for encapsulation, enzymes or whole cells are entrapped within semipermeable micro/nano-particles or vesicles such as liposomes [3], polymersomes [4], colloidosomes [5], microgels [6] and polymeric microspheres [7]. Biocatalyst can also be immobilized in or onto the microporous membrane by physical entrapment [8], gela-

2.2 Role of membrane in biocatalytic membrane reactors (BMRs) | 53

Biocatalyst segregated

Biocatalyst flushed

Biocatalyst entrapped within pores

Membrane serves only for separation

Biocatalyst gelled on membrane

Membrane serves for catalyst immobilisation

BMR Membrane promotes nondispersive interfacial contact

Extractive reactor

Waste water

Membrane

Aeration reactor

Biomass

Bio-film

O2

Biphasic reactor

Waste water

Bio-film

Organic phase

Aqueous phase

Fig. 2.1: Classification of biocatalytic membrane reactors.

tion [9], physical adsorption [10], ionic binding [11], covalent binding [12] or copolymerization [13]. In addition to the type of immobilization, BMRs can be classified according to their operational mode into ultrafiltration (UF) or microfiltration (MF) membrane reactors, biphasic (organic and aqueous) membrane reactors, membrane aeration reactors and extractive membrane reactors [14]. UF membrane reactors are used when the substrate has a significantly higher molecular weight compared to that of the product, and both substrate and product are soluble in the same solvents. In that case, the substrate

54 | 2 Biocatalytic membrane reactors (BMR)

molecules are transferred to the enzyme immobilized in or onto the membrane, but they cannot pass through the membrane, whilst the product can freely pass through and be recovered from the permeate side of the membrane. If the substrate and product have a similar molecular size, they can both pass through the membrane. In that case, the membrane type and operating conditions should be chosen in such a way that the permeation rate matches the reaction rate to ensure that the substrate will be fully converted into the product as it passes through the membrane. If the substrate has a different solubility to the product (e.g. an oil-soluble ester and its water-soluble hydrolysis products), a biphasic membrane reactor might be a good choice [15]. In this type of reactor system, the enzyme-loaded membrane is located between two immiscible liquid phases, an organic and an aqueous phase. The organic phase, which contains the substrate, circulates on one side of the membrane; the substrate is diffused to the hydrophilic membrane wetted by the aqueous phase, where the reaction takes place in an aqueous environment, and the product is extracted by the aqueous phase circulated on the other side of the membrane. If the biocatalyst is selective for only one of the two enantiomers present in a racemic mixture, a biphasic system is particularly useful for production of pure enantiomers [16]. Pure isomers are required in biomedical applications, where they have to be administered to humans and animals. The membrane in BMRs may also serve as a medium for a bubbleless transfer of gas (oxygen or air) into a bioreactor to augment microbial degradation [17] or as a medium for a non-dispersive extraction of organic pollutants from wastewater [18]. Another classification of membrane reactors is based on the relative position of the two most important elements of the reactor: the membrane and the catalyst. The three main configurations are shown in Fig. 2.2: (a) the catalyst is suspended in a reaction mixture separated from the membrane; (b) the catalyst is suspended in a liquid core which is surrounded by the membrane (a core-shell microcapsule); and (c) the catalyst is incorporated within the membrane, i.e. the membrane is inherently catalytic. The membrane in core-shell microcapsules can be composed of lipid bilayers, a hydrogel, synthetic polymer, fused particles, etc. Hybrid systems are also possible, for example the matrix-type porous microcapsule shown in Fig. 2.2 (d) combines the structures shown in Fig. 2.2 (b) and (c). Although membranes are more expansive as enzyme supports than porous beads shown in Fig. 2.2 (b) and (d), their advantages may in many applications overcome this limitation (Tab. 2.1). Membranes permit a much higher order and generally smaller scale of organization than do particles [1]. The fluids on either side of a membrane can be separated by appropriate membrane module design; this then provides the engineer with an additional degree of freedom in reactor design. While there is only a single interface between a porous catalyst particle and the solution surrounding it (Fig. 2.2 (d)), two membrane/solution interfaces and two liquid compartments exist that can be exploited in a number of ways. For example, a membrane reactor might be used to catalyze the reaction between two reactants present in separate streams that cannot be mixed for some reason, perhaps to

2.2 Role of membrane in biocatalytic membrane reactors (BMRs) | 55

avoid a subsequent separation problem or a non-catalytic side reaction, or reactants or products are soluble in different solvents (organic and aqueous). When porous beads are used, separation of the products from the reactants is impossible because there is only one liquid compartment. For the same reason, the product concentration in the porous beads reactor is limited by the stoichiometry of the reaction, conversion and reactant concentrations. When a catalyst is encapsulated within a porous membrane, the product concentration can be controlled by the flow rate ratio of the product and feed stream, which makes it possible to achieve high product concentrations irrespective of the reactant concentration. In addition, the rate of mass transfer within a porous membrane can be enhanced by creating a pressure difference across the membrane, which will lead to convective transport of species across the membrane. On the other hand, within a porous bead, mass transfer can only occur by molecular diffusion. Tab. 2.1: Comparison of porous membranes and porous particles as enzyme supports. Function

Type of catalyst support

Immobilization Mass transfer by diffusion Mass transfer by convection Liquid/liquid separation Enzyme/product separation Product concentration

Membrane

Membrane

Porous particles or capsules

Yes Yes Yes Yes Yes Yes

Yes Yes No No No No

Membrane

Membrane

Interface 1 (a)

(b)

(c)

Single interface

Matrix

Interface 2 (d)

Fig. 2.2: Typical configurations of membrane biocatalytic reactors according to the relative position of the membrane and the catalyst: (a) the catalyst is physically separated from the membrane; (b) the catalyst is encapsulated in a core-shell microcapsule; (c) the catalyst is incorporated within the membrane wall; (d) the catalyst is encapsulated in a matrix-type microcapsule.

56 | 2 Biocatalytic membrane reactors (BMR)

Substrate

Recycling enzyme

Enzymes Substrate Membrane

Cofactors

Product

Cofactors

Product Membrane Enzyme

(a)

(b) Substrate Cofactors

Membrane Product (c) Fig. 2.3: Typical biocatalytic membrane reactor systems based on configurations shown in Fig. 2.2: (a) Stirred tank reactor combined with membrane module for enzyme recycle and product withdrawal; (b) Stirred tank with enzymes entrapped within core-shell microcapsules; (c) Stirred tank with enzymes immobilized within the membrane matrix.

Fig. 2.3 shows typical BMRs based on the configurations presented in Fig. 2.2. The BMR shown in Fig. 2.3 (a) consists of a continuous stirred tank reactor combined with an ultrafiltration (UF) or microfiltration (MF) membrane module for the product removal. The reactor presented in Fig. 2.2 (b) is a conventional stirred tank reactor in which biocatalysts are encapsulated within microcapsules with semi-permeable walls. In a membrane immobilization approach shown in Fig. 2.3 (c), enzymes are bound to the membrane by physical adsorption, gel formation, electrostatic deposition or covalent attachment. Substrates are transferred across the membrane wall to the catalyst and the products diffuse from the reaction site to the other side of the membrane; they are recovered here as a permeate. In general, it is the mass-transport resistance that primarily influences the performance of these reaction systems. In order for a reactor to function at its optimal performance, it should work in a reaction-limited regime rather than a diffusion-limited regime. The parameter that can give a measure of this

2.3 Membrane separation reactors (MSRs)

| 57

condition is the Thiele modulus, which is given by: ϕ = L(

rmax 1/2 , ) Deff K m

(2.1)

where L is the length, r max is the maximum reaction rate, K m is the Michaelis-Menten constant and Deff is the effective diffusivity. This has the physical meaning of a ratio between the reaction rate and the diffusion rate. For ϕ ≤ 1, the system is essentially controlled by chemical kinetics and the mass-transfer limitation is negligible [19].

2.3 Membrane separation reactors (MSRs) 2.3.1 Concept A membrane separation reactor (MSR) can be considered as a combination of a reaction vessel, either a continuous stirred tank reactor (CSTR) or a batch stirred tank reactor, and a membrane module. The purpose of the membrane is to contain a dissolved or dispersed biocatalyst (enzyme molecules or cells) in the reaction vessel, while products and unreacted permeable reactants are allowed to leave the reactor. In a continuous flow process, the solution is supplied continuously to the vessel, while the product, together with excess solvent, is withdrawn in a feed-and-bleed operation [14, 20, 21]. The process results in the product removal without a loss of the biocatalyst and high-molecular weight substrate. The biocatalytic reaction usually occurs in an aqueous phase, although organic solvents or supercritical CO2 (SC CO2 ) could be more appropriate medium than water for conversion of hydrophobic substrates [22, 23]. The advantages of using SC CO2 over organic solvents include environmental, health and safety, and process benefits. The use of SC CO2 can allow higher mass transfer rates because of the lower viscosity of SC CO2 compared to that of organic solvents and accompanying high molecular diffusivities. On the other hand, the problem when using biocatalysts in SC CO2 is that high pressure, law water activity, increased temperature, and pressurization/depressurization steps may adversely affect the catalyst stability and activity. The main advantages of MSRs over conventional biocatalytic reactors are prolonged biocatalyst activity, reduction in costs and energy due to biocatalyst recycling, greater yields and selectivities due to continual removal of products, higher capacity at the same reactor volume due to high biocatalyst loadings, and simple scale-up to large systems due to modular design of membrane units. In addition, the product stream is typically free from all suspended matter including bacteria and viruses [24]. Along with the complete removal of bacteria and viruses, MSRs are compact devices with a smaller footprint than conventional biocatalytic systems, since they can achieve simultaneous bioconversion, separation and concentration in a single unit.

58 | 2 Biocatalytic membrane reactors (BMR)

(a)

(b) ct

du

Pro

e rat bst Su

(c)

Ox yge n

(d)

Fig. 2.4: Different modes of operation of membrane separation reactors: (a) External membrane loop (‘sidestream’); (b) Internal membrane (‘submerged’); (c) segregation in a single-set hollow fiber module; (d) segregation in a two-set hollow fiber module.

The separation of low-molecular-weight products from biocatalysts can be achieved using either an external loop with a separate membrane module (Fig. 2.4 (a)) or a membrane submerged inside the reaction vessel (Fig. 2.4 (b)) [25, 26]. In the set-up illustrated in Fig. 2.4 (a), the content of the vessel is circulated continuously through the external membrane module where an enzyme-free permeate is removed from the reaction mixture. The submerged system differs in that there is no recirculation loop and the separation occurs within the bioreactor. In either case, surface shear and/or backflushing is applied to control cake formation and fouling. In the submerged system, the membrane is mounted in the stirred tank; stirring or bubbling is usually applied to produce the surface shear and the permeate is removed by suction with a vacuum pump or by a compressed gas. The membranes used in the submerged systems are typically hollow fibers or flat sheets, aligned vertically or horizontally. When the surface shear is provided by bubbling, gas is introduced below the membrane assembly and is distributed to optimize the air scouring action across the vertical membrane surface. The two methods based on internal or external membranes in conjunction with the reaction vessel both have their advantages and drawbacks. The external loop membrane can easily be changed if severe clogging occurs. Furthermore, in the external membrane loop it is easier to increase the filtration area if necessary. On the other hand, when operating with high biocatalyst densities using an external loop,

2.3 Membrane separation reactors (MSRs)

| 59

there is a risk of the occurrence of oxygen limitation. One way to reduce this risk is to increase the rate of circulation through the membrane module, but high shear stress can damage shear-sensitive biocatalysts. One option to prevent damages to shear-sensitive cells and enzymes is to segregate a biocatalyst within a hollow fiber membrane module. In the system shown in Fig. 2.4 (c), the feed stream is introduced through the lumen of the fibers and biocatalysts are localized in the extra-capillary space, where the reaction proceeds in a low shear-stress environment. Substrate molecules diffuse through the walls of the fibers from inside to outside and the products diffuse in the opposite direction. A transverse flow hollow fiber module containing two independent sets of hollow fibers placed within the same frame element (Fig. 2.4 (d)) is a potential alternative to the conventional parallel flow module, suitable to meet very high oxygen demands in the reactor. In this case, substrate is fed and products are removed via the lumen of one set of hollow fibers and oxygen is supplied through the lumen of the second set of the fibers, while the biocatalyst is dispersed in the extra-capillary space. An additional advantage of this design is that the membrane area can easily be increased by adding more frame elements, similar to a plate-and-frame-filter press [27]. The main benefits of using MSRs are schematically shown in Fig. 2.5 [28]. (a) Increase of conversion in reversible reactions: the major role of the membrane here is to selectively remove a product of a reversible reaction from the reaction mixture (product C in Fig. 2.5 (a), thereby shifting the equilibrium towards higher product yields (i.e. higher conversions of A). The reactions of this type include dehydrogenation and esterification reactions. (b) Enhancement of selectivity through controlled product removal: the role of the membrane in Fig. 2.5 (b) is to suppress the further decomposition of a desirable intermediate product (C) by removing this product through the membrane. (c) Enhancement of selectivity through controlled addition of reactants: in this approach, the concentration of a reactant B in the reaction mixture (A + B) is kept low by a controlled supply of B through the membrane, so that the rate of formation of a by-product D is kept low (Fig. 2.5 (c)). Controlled addition can also be useful to prevent catalyst deactivation and to avoid a dangerous increase of temperature in exothermic reactions. Partial oxidation of hydrocarbons is the most relevant application of this approach where a controlled addition of the oxidant through the membrane results in a better yield of the intermediate oxidation products. A

B+C

A+B

C (a)

C+B

D

B

A+B C A + nB D

C (b)

(c)

Fig. 2.5: Major benefits of using membrane reactors: (a) improved conversion; (b), (c) increased selectivity.

60 | 2 Biocatalytic membrane reactors (BMR)

2.3.2 Application A number of nutritionally important amino acid have been synthesized or optically resolved on an industrial scale using enzymatic reactions in membrane separation reactors [1]. Wichmann et al. [29] have conducted multienzyme (multistep) bioconversions with simultaneous cofactor regeneration in a continuously operated MSR shown in Fig. 2.6. For the continuous operation of cofactor dependent systems, it is critically important to retain and regenerate the cofactor in the reaction mixture. The retention of native cofactors such as NAD/NADH by a membrane would require a reverse osmosis membrane that has a very low permeability to water as compared to an UF membrane. Moreover, it would be difficult, if not impossible, to obtain the product without leakage of the cofactor through the membrane, due to the small difference in size between the solvent and the cofactor molecules. This problem was solved by enlarging the cofactor molecule to a size similar to that of the enzyme. It was achieved by covalent attachment of the cofactor (NAD+ /NADH) to poly(ethylene glycol) (PEG), a linear polymer with only two terminal reactive groups on the polymer chain, so steric hindrance was low [29]. The PEGylated cofactor was used to catalyze the reductive amination of α-ketoisocaproate by L-leucine dehydrogenase (enzyme E2 in Fig. 2.6) to L-leucine. Formate dehydrogenase (enzyme E1 in Fig. 2.6) was used for regeneration of NAD+ from NADH. The use of charged UF membranes is an alternative method for the retention of the native coenzyme NADH in a membrane reactor [30]. ( ) NAD+

( ) HCOO–

( ) L-Amino acid + H2O

E1 CO2 ( )

(

E2

)

NADH ( )

(

)

+ NH4 α-Keto acid ( )

Fig. 2.6: Membrane separation reactor for continuous enzymatic synthesis of L-amino acid from α-keto acid with coenzyme (NADH/NAD+ ) regeneration. Adopted from [29].

In addition to the synthesis reactions, membrane reactors are extensively used in the enzymatic hydrolysis of natural macromolecules (starch, cellulose, proteins, pectin, chitosan, etc.), oligo- and di-saccharides, and bioflavonoids, especially for biomedical and pharmaceutical applications (Tab. 2.2) and in the agro-food sector (Tab. 2.3).

2.3 Membrane separation reactors (MSRs)

| 61

Tab. 2.2: Examples of medical and pharmaceutical applications of membrane reactors (CO = cut-off). Enzyme or cells

Membrane

Reaction

Purpose

Reference

Arginase, asparaginase

Hemophan (cellulose derivative)

Hydrolysis of arginine and asparagines

Cancer therapy

Shettigar [31]

Heparinase, tripsin, pronase

Polyester membrane

Hydrolysis of blood toxins, such as heparin

Removal of blood toxins in hemodialasys

Ameer et al. [32]

Hepatocytes

Hollow fiber or flat plate

Liver-specific functions

Extracorporeal bioartificial liver devices

Allen et al. [33] 2001

Pancreatic endocrine cells

Nucleopore, Insulin secretion pore size = 0.1 μm

Bioartificial pancreas

Ohgawara et al. [34]

Bacterial protease

PES, CO = 3 kDa

Hydrolysis of whey proteins

Production of low allergenicity hydrolysates for enteral nutrition

Guadix et al. [35]

Amidase from Microbacterium imperiale

Fluoropolymer, CO = 20 kDa

Deamidation of nicotinamide to nicotinic acid

Production of nicotinic acid (Bcomplex vitamin)

Cantarella et al. [36]

Glucose-fructose oxidoreductase from Zymomonas Mobilis

Milipore, CO = 10 kDa

oxidation of glucose

Production of lactobionic acid for medical use

Satory et al. [37]

More recently, membrane reactors have been used successfully for the treatment of wastewaters. The constantly increasing degree of industrialization and urbanization, rising standards of living, increasing population growth and agricultural activities are strongly impacting on the use of available water sources and on the quality of water that is found therein. This exhaustive use of limited resources and energy by modern society implies a need for changes in present and future urban water and wastewater treatment systems [51]. Membrane reactors enable degradation but also the recuperation of valuable components from effluent streams (Tab. 2.4). They are also useful for reusing contaminated process water [60, 61].

62 | 2 Biocatalytic membrane reactors (BMR)

Tab. 2.3: Examples of agro-food applications of membrane reactors (CO = cut-off). Enzyme

Membrane

Reaction

Purpose

Reference

Trypsin from bovine pancreas

Cellulose acetate, CO = 3 kDa

Hydrolysis of caseinomacropeptide

Production of peptides for functional food

Prata-Vidal et al. [38]

Trypsin from bovine pancreas

Cerasep® nanofiltration

Hydrolysis of caseinomacropeptide

Production of peptides for functional food

Martin-Orue et al. [39]

Enzymes from Trichoderma reesei fungi

Carbosep® M5, CO = 10 kDa

Hydrolysis of olive mill solid residue

Ethanol production

Mameri et al. [40]

Glucoamylase from Aspergillus niger

Amicon Diaflo® H1510-43, CO = 10 kDa

Hydrolysis of cassava flour starch

Production of pure dextrose, high fructose syrups, etc.

López-Ulibarri and Hall [41]

Glucoamylase

regenerated cellulose, CO = 10 kDa

Hydrolysis of corn starch

Production of glucose or maltose syrups

Singh and Cheryan [42]

β-Galactosidase from Kluyveromyces fragilis

Cuprophan® and PS hollow fiber, CO = 5 kDa

Hydrolysis of lactose

Milk or whey delactosization

Jurado et al. [43]

naringinase from Aspergillus niger

Romicon HF 1.1-43-PM10

Hydrolysis of naringin

Debittering of grapefruit juice

Gray and Olson [44]

polygalacturonase from Aspergillus niger

regenerated cellulose, CO = 30 kDa

Hydrolysis of pectin

Clarification of fruit juice

Bélafi-Bakó et al. [45]

polygalacturonase from Aspergillus niger

polysulfone, CO = 10 and 50 kDa

Hydrolysis of pectin

Clarification of wine and fruit juice

RodriguezNogales et al. [46]

Invertase immobilised on polystyrene beads

regenerated cellulose, CO = 100 kDa or 5 μm

Hydrolysis of sucrose

Production of high-fructose syrup

Tomotani and Vitolo [47]

xylose reductase from Candida tenuis

Nitto NTR 7430, CO = 1 kDa

Reduction of xylose to xylitol

Production of xylitol, a natural food sweetener

Nidetzky et al. [48]

chitosanases and chitinase from Bacillus cereus

Amicon Diaflo® H1510-43, CO = 10 kDa

Hydrolysis of chitosan

Production of chitooligosaccharides

Kuo et al. [49]

Pronase from Streptomyces griseus

Amicon Diaflo® H1510, CO = 10 kDa

Hydrolysis of soy protein isolate

Production of soy protein hydrolysates for functional food

Deeslie and Cheryan [50]

2.3 Membrane separation reactors (MSRs)

| 63

Tab. 2.4: Examples of applications of membrane reactors to the wastewater treatment. Enzyme

Source

Reactor

Membrane

Application

Reference

Immobilized enzyme membrane reactor Crude enzyme extract

Pseudomonas sp.

Batch UF cell

Flat polyacrylonitrile

Phenolic effluent

Bohdziewicz [8], Bodzek [52]

Polyphenol oxidase (EC 1.14.18.1)

Agaricus bisporum

Hollow fiber membrane reactor

Polysulphone

Coal-gas conversion plant effluents (phenols)

Edwards et al. [53]

Laccasse (EC 1.10.3.2)

Pyricularia oryzae

Spiral-wound module

Polyethersulphone

Synthetic industrial wastewater (phenols)

Lante et al. [54]

Laccasse

Trametes versicolor

Frame plate reactor module

Modified polyvinylidene difluoride (PVDF)

Herbicide N󸀠 ,N󸀠 (dimethyl)N-(2-hydroxyphenyl) urea (2-HF)

Jolivalt et al. [55]

Extractive membrane bioreactor Glucose oxidase

Aspergilus Niger

Hollow fiber

Polyethersulphone

Synthetic industrial wastewater (glucose)

Kojima et al. [56]

Glycerol dehydrogenase

Enterobacter aerogenes

Stirred tank reactor and hollow fiber

Polyethersulphone and hydrophobic membrane

Ethanol oxidation

Liese et al. [57]

Membrane separation reactor Soybean peroxidase

Ground soybean seed-hulls

Stirred tank reactor

—

Synthetic industrial wastewater (phenols)

Flock et al. [58]

Pectolyase

Aspergilus japonicus

Stirred tank reactor

Polyethersulphone

Depolymerization of polygalacturonic acid

Gallifuoco et al. [59]

Manganese peroxidase

Bjerkandera sp.

Stirred tank reactor

Polyethersulphone

Dye decolorization

López et al. [20]

64 | 2 Biocatalytic membrane reactors (BMR)

2.4 Membrane aeration bioreactors (MABR) A basic requirement for the aerobic degradation of organic material is oxygen, which is required to support the life and growth of the microbes performing the degradation. It is therefore imperative that these systems receive sufficient oxygen as, without it, a rapid deterioration in the quality of effluent will occur [62]. The rate of oxygen mass transfer can be greatly improved by using high purity oxygen rather than atmospheric air (approximately by a factor of 5), but these oxygenation devices require a great deal of power in order to efficiently mix the gas into the solution and obtain a high oxygen transfer efficiency; they thus cannot be used in conjunction with biofilm processes in which the reactor should remain static [20]. The membrane aeration bioreactor (MABR) concept was developed in response to the need for increased oxygen mass transfer into wastewaters in cases where the oxygen requirements for degradation of the pollutant were too high for conventional aeration processes. It is also applicable when the bubbling of air can result in either stripping of volatile organic compounds (VOCs) [63], foaming [17] of industrial wastewaters or the damage to shear-sensitive cell cultures [64]. The membrane itself can play a dual role in the reactor, namely as a means for supplying oxygen and a substrate for biofilm formation. Fig. 2.7 is a schematic diagram outlining the principle of the MABR. A membrane (dense gas permeable, microporous or composite, i.e. microporous coated with a nonporous gas permeable) is used to transfer oxygen to the bacteria present in the bioreactor without forming bubbles and with 100 % oxygen transfer efficiency. The membrane also acts as a support for biofilm growth. Wastewater flows over the outer surface of the biofilm and counter-diffusion of oxygen and pollutant occurs, as shown in Fig. 2.7 (a). Oxygen transferred through the membrane is utilized in the degradation of pollutants in the biofilm. MABRs have been use to treat a wide variety of wastewater types and have been shown to be particularly effective in treating high oxygen demanding wastewaters [65], biodegradation of VOCs [66, 67], combined nitrification, denitrification and/or organic carbon degradation in a single biofilm [68, 69]. When a MABR system is used for the treatment of waste vapor streams, volatile organic compounds (VOCs) and oxygen are transferred concurrently through the membrane to the biofilm and a liquid phase flowing on the opposite side of the membrane is a source of mineral salts for the culture growth. The most commonly used membranes for MABRs are silicon rubber tubing [70] and microporous polypropylene hollow fiber membranes [67]. For the biotreatment of vapor streams containing aromatic contaminants, the nonporous silicone membrane system delivered superior performance over the microporous membrane system in terms of surface-area-based removal rates, longterm operational stability and maintenance [67].

2.5 Extractive membrane bioreactors (EMBR) | 65

Gas-filled pore

Gas phase

Gas-filled pore

CO₂

Wastewater

O₂

Organics

Waste vapour stream

Membrane

CO₂ O₂ VOCs

Mineral salts medium

Membrane

Biofilm

Biofilm (b)

(a)

Fig. 2.7: Schematic of the MABR system utilizing a hydrophobic microporous membrane: (a) biodegradation of organics from a wastewater stream; (b) removal of VOCs from a waste vapor stream.

2.5 Extractive membrane bioreactors (EMBR) 2.5.1 Concept The principle of an EMBR is illustrated in Fig. 2.8. An aqueous feed stream (usually wastewater) containing the organic compound(s) to be degraded is passed over one surface of a nonporous semipermeable membrane, while a biocatalyst (usually microbial culture) is maintained in an oxygenated aqueous biomedium at the other surface. The pH and ionic strength of the wastewater have no influence on the makeup of the biomedium as the membrane is impermeable to any inorganic or charged species in the wastewater. Thus the biomedium composition can be controlled and optimized independently of the composition of the wastewater to provide optimal growth conditions for the microbial culture in spite of the biologically hostile makeup of the wastewater [71]. The driving force for the mass transfer across the membrane is the Aqueous stream containing inorganic and organic species

Biofilm

Suspended biomass

Organics

Exiting stream containing only inorganic species

Membrane

Fig. 2.8: Schematic of the EMBR process. Organic pollutants selectively diffuse through the nonporous membrane material into the biomedium phase where they are biodegraded in the biofilm.

66 | 2 Biocatalytic membrane reactors (BMR)

concentration difference between the organic pollutant in the wastewater and in the biomedium, which is maintained by biological degradation. The EMBR systems are usually composed of silicone rubber (polydimethylsiloxane) tubing or plates. The pilot unit, schematically represented in Fig. 2.9, consists of two membrane modules, each with eight coils of silicone tubing linked in series and submerged in a reactor fitted with a mechanical stirrer. Wastewater was fed through the membrane tubes. Air was bubbled through the reactor to maintain the dissolved oxygen concentration at the optimum level. Low pressure steam was fed directly into the reactor tank to maintain temperature at around 30 °C. pH was controlled by the automatic dosing of either sulphuric acid or sodium hydroxide solution. pH control

Acid or base dosing Wastewater tank

Outlet wastewater Biomedium overflow

Nutrient tank

Bioreactor

Membrane tube Steam Air

Fig. 2.9: Schematic of the pilot-scale EMBR unit [18].

Due to bacterial attachment, a membrane-attached biofilm forms on the shell side of the membrane. A thin biofilm (200–400 μm) is advantageous because it limits the air stripping of volatile organic compounds [72]. However, an increase in biofilm thickness results in a decrease in organic pollutant(s) flux across the membrane. The EMBR shares this problem with membrane aeration reactors, where membraneattached biofilms also create mass transfer limitations [73, 74]. It was confirmed by mathematical modeling of biofilm growth in an EMBR that due to the limited solubility of oxygen in water, oxygen could only penetrate thin biofilms [75]. Therefore, the active layer is very close to the biofilm/biomedium interface. Consequently, an inverse relationship between the organic flux and biofilm thickness was observed. Several modifications have been attempted to suppress the metabolism of the biofilmforming bacteria, such as addition of sodium chloride [76], increasing the shear stress on the membrane surface [77], using a biphasic system [78] and the addition of nitrate as an electron acceptor instead of oxygen [79]. Nitrate is highly soluble in water and therefore, the nitrate concentration in the biomedium can be sufficiently high to fully

2.5 Extractive membrane bioreactors (EMBR) | 67

penetrate the biofilm, ensuring that nitrate is present at the biofilm/membrane interface. Theoretically, the substrate could then be biodegraded at the biofilm/membrane interface, alleviating the decrease in pollutant flux with biofilm growth. Another approach is based on the modification of the interfacial properties that regulate microbial attachment to the membrane through addition of surfactants [80].

2.5.2 Application EMBR was used for degradation of organic pollutants and solvents from synthetic and industrial wastewaters. More than 99 % of 1,2-dichloroethane (1,2-DCA) was removed from a synthetic wastewater containing 1600 mg l−1 of 1,2-DCA [63]. 2,4-dichlorophenoxyacetic acid, a component of commercial herbicides, has also been successfully degraded at laboratory scale using a suitably acclimated microbial culture [81]. In addition, toluene and dichloromethane, the two most commonly used solvents in the chemical and pharmaceutical industries, were being degraded by a commensal microbial culture [82]. When EMBR was used with industrial wastewaters, the results were also encouraging. Wastewater from a 3-chloronitrobenzene-manufacture plant was successfully remediated with the removal efficiencies of greater than 99 % at residence times of ~ 30 min [83]. Brookes and Livingston [84, 85] operated a laboratory-scale reactor continuously for 5 months and consistently removed aniline, 4-chloroaniline, 2,3-dichloroaniline and 3,4-dichloroaniline at efficiencies greater than 99 %. Conventional direct biological treatment of such effluents cannot be implemented without pretreatment or dilution because of the hostile inorganic composition of the wastewater. Other examples of successful application include the removal and degradation of compounds such as monochlorobenzene, where 98 to 99 % of the pollutant was destroyed at a flow rate of 50 l/h [71], 1,2-dichloroethane (94.5 % removal) with negligible air stripping [63] and various other chemical and pharmaceutical waste streams with similar results. Particularly, the EMBR system is suitable for the treatment of wastewaters in which a recalcitrant hydrophobic toxic compound is mixed with high concentrations of an easily biodegradable hydrophilic compound. Biodegradation of such wastewaters is difficult in conventional bioreactors, since the microbial cultures tend to grow on the easily biodegradable compound in preference to the toxic compounds. An example of this type of wastewater is the industrial wastewater produced from a hydrogenation process that contains 2 g l−1 of toluidines such as 3-chloro-4-methylaniline and methylammines (which are hydrophobic and not easily biodegradable) and up to 2–3 vol % of methanol (which is hydrophilic and easily biodegradable) [18]. The EMBR technology is suitable for treating this kind of waste, since the hydrophobic target compounds can rapidly cross the membrane and be degraded in the bioreactor, whilst the hydrophilic methanol is retained in the EMBR effluent and goes on to further treatment.

68 | 2 Biocatalytic membrane reactors (BMR)

A number of variations of the basic EMBR configuration have been attempted. Splendiani et al. [78] have developed the biphasic extractive membrane bioreactor (BEMBR) in which biofilm accumulation was controlled by preventing direct contact between microorganisms and the membrane. In BEMBR systems, the two main constituents of the process, membrane and bacteria, are kept separated and interact via a suitable recirculating solvent, as shown in Fig. 2.10. The BEMBR system consists of a biphasic bioreactor and a membrane module with a silicone rubber tube fitted coaxially in a cylindrical shell. The nonporous membrane selectively extracts the organic pollutants from the waste streams. On the shell side, the pollutants are desorbed into the organic solvent, which is re-circulated between the membrane module and the bioreactor. In the bioreactor, the pollutant-rich solvent is mixed with the aqueous biomedium, and the pollutants are transferred from the solvent to the aqueous biomedium where they are metabolized by suspended bacteria. The lean solvent is continuously separated from the aqueous biomedium and recycled back into the membrane shell. The organic solvent must not be toxic for bacteria or inhibit the pollutant metabolization; it must not be utilized by bacteria as a source of carbon and energy – it must exhibit low microbial adhesivity, low membrane swelling and emulsionforming tendency. According to Splendiani et al. [78], the solvent that best satisfies these requirements is perfluoromethyldecalin (PFMD). They tested the BEMBR system with PFMD as a solvent for degradation of monochlorobenzene using a synthetic wastewater. The system was operated over a one-month period with negligible biofilm accumulation. Throughout the period of operation, the overall mass transfer coefficient remained stable at approximately the value obtained for mass transfer through the bare membrane.

Rich solvent Wastewater

O₂

Membrane module Wastewater reservoir Membrane tube

Lean solvent

Aqueous overflow Mineral medium Solvent Biphasic reactor

Fig. 2.10: Schematic of the biphasic EMBR process [78].

2.5 Extractive membrane bioreactors (EMBR) | 69

Liu et al. [86, 87] coupled EMBR with the conventional solvent extraction and stripping to degrade chlorophenolic compounds to levels lower than 100 mg l−1 . The process was operated in four units shown in Fig. 2.11: (1) Extraction unit. The acidic organic pollutant in the acidic wastewater (stream a, containing 1000 mg l−1 2,4-dichlorophenol (DCP), and 5 % NaCl at pH < 1) is extracted by a water insoluble organic solvent. The raffinate is discharged after filtering through a hydrophilic microfiltration membrane that minimizes loss of organics in droplet form (stream b, containing < 100 mg l−1 DCP and 5 % NaCl at pH < 1). (2) Stripping unit. The extract is pumped from the extraction unit to the stripping unit where organic pollutant is transferred into an alkaline aqueous phase and its concentration in the raffinate out of the stripping unit (stream c) is much higher than that in the treated water (stream b). (3) Membrane separation of O/W emulsion and biomass. Organic droplets in the raffinate from the stripping unit (stream c) are filtered out by the hydrophilic ceramic membrane, thus the organic pollutant is fed into the bioreactor but not the solvent droplets. The outlet stream of the bioreactor is filtered from the reverse direction in a second membrane for the purpose of removing the reaction products while retaining the biocatalysts. These units are periodically switched in function. After the switch the filtration of biomass separation serves to backflush the membrane previously used for filtering the O/W emulsion and vice versa. In this way, membrane fouling is reduced. (4) Bioreactor. The organic pollutant is fed to the bioreactor as the carbon and energy source for the bacteria and may be supplied to be present at optimum concentrations and suitable pH and temperature. When the pollutants are chlorinated hydrocarbons, hydrochloric acid is released as the biodegradation by-product. The hydroxide ion present in the stripping solution is consumed to neutralize the hydrochloric acid. Wastewater (stream a)

Extraction (pH 0. The denominator in each term plays the role of a scaling factor for minimizing the distance between upper I i,up and lower I i,lower values for

4.4 A priori methods | 139

each objective function. Also, a method of global criterion is one of such methods but will be considered in section on a priori methods below with some remarks.

4.4 A priori methods A priori methods require the DM to state his preference in a MOO problem. This has to be done prior to determining the Pareto set. One can specify the priority of objectives (or aims) to be achieved. Since a DM is a person familiar with a particular problem, sometimes it becomes possible to single out more important objectives or put them in an order of preference.

4.4.1 Method of Weighted global criterion This method with some variations is the most popular technique for a MOO. The idea is to transform objective functions into a single one, thereby scalarizing the search space. In the most general form, this method can be written as n

minimize ∑ F(I j (x), w j ) . x∈S

(4.3)

j=1

A scalarized function represents the sum of composite functions of objective I i (x) and weighting factor w i . The latter itself is a measure of the DM’s preferences for a particular objective. Usually weighting factors are assigned in such a way that ∑nj=1 w j = 1 and w j > 0. In a simplest form, the expression (4.3) can be written as a weighted exponential sum [4]: n

minimize ∑ w j [I j (x)]p , x∈S

I j (x) > 0 ,

j=1

(4.4)

n

minimize ∑ [w j I j (x)]p , x∈S

I j (x) > 0 .

j=1

Note that in the case of Eq. (4.3) with p = 1 (because of its simplicity), it is called the method of a weighted sum and widely used in applied chemical engineering problems [6]. Some other modifications are required for the idea of a utopia point, I utopia (x); the imaginary point in a search space where all the objectives reach a minimum value simultaneously. The aim is to minimize the weighted distance between the objectives and that point [7]. Different metrics can be used as a distance measure. Often this

140 | 4 Multi-objective optimization in chemical engineering

group of techniques is called weighted metrics [3]. Here we provide some of them: 1/p

n

utopia

minimize [ ∑ w j (I j (x) − I j x∈S

,

j=1 n

minimize [ ∑ x∈S

(x))p ]

(4.5)

1/p p w j (I j (x)

−

utopia Ij (x))p ]

.

j=1

Note here that instead of a utopia point, the DM can determine a set of objectives that one desires to reach. This makes sense from a practical point of view, or when the real utopia point is unknown. Remarks: A group of weighted global criterion methods is also a popular a posteriori technique. By varying the weights, it is possible to obtain a Pareto set instead of a single point. These methods always converge to a Pareto optimal solution, but an entire Pareto set can not be found if the problem is not convex [3]. If one assigns all weights, w i , equal to 1, the approach can be classified as no-preference; but the drawback remains the same. In addition, the magnitudes of objective values should be commensurable with each other to avoid overemphasis of one over the other. Hence, normalization is required for applying this technique [3].

4.4.2 Lexicographic method Lexicographic methods require the DM to sequentially organize objectives from 1 to N in terms of preferences [8]. The following problem has to be solved [4]: minimize I i (x) , x∈S

(4.6)

subject to: I k (x) < I k (x∗k )

k = 1, 2, . . . , i − 1; i = 1, 2, . . . , n ,

where k is the function order in a preference list, I k (x∗k ) the constraint’s limit received at kth step. The first objective in the list should be minimized with the original constraints. If the DM obtains a single solution, one can accept it as an optimum. If not, the new constraint I k (x∗k ) has to be accepted to keep the kth objective’s optimal values. The procedure continues with the next objective function (e.g. second function in a list, third function in a list, etc.), until the optimum is reached. In reality, it is often difficult for a DM to distinctly organize objectives in an order of importance on account of the complexity of a MOO problem. Another drawback with this technique is that a unique solution is often found before the best optimal solution is reached. It means that some of the objectives are not taken into consideration at all [3].

4.5 A posteriori methods | 141

4.4.3 Goal Programming (GP) This method was developed by Charnes and Cooper [9]. The DM defines a set of goals G that should be achieved for each objective I i (x). Even if all these goals are unattainable simultaneously, it is still desired to reach them “as close as possible”. It is proposed to minimize the distance between vectors I(x) and G. Such a weighted GP problem formulation is written as: n

minimize ∑ w i δ i , x∈S

δ i = I i (x) − g i ,

(4.7)

i=1

i = 1, 2, . . . , n ,

where w i is a weighting factor for objective I, is a deviation of objective I i (x) from goal g i . The formulation of this goal programming problem doesn’t necessarily require the solution to be Pareto optimal. The solution obtained can be referred to as: (a) efficient; (b) inefficient; or (c) an unbounded solution. An efficient solution belongs to a Pareto front while an inefficient solution can be improved for two or more objectives simultaneously. The latter case is a solution located too far from a Pareto front [10]. Setting goals is a clear approach for a DM (unlike, for example, the use of a utopia point in the global criterion method). However, the further procedure for an optimum search is not necessarily easy, e.g. weights assignment can be more difficult. Some GP methods are combined with a lexicographic method, where deviations are structured in preference order and then minimized. The DM has to be aware of all the drawbacks of GP methods and choose the proper technique for finding an optimal solution.

4.5 A posteriori methods In contrast to other methods discussed so far, an a posteriori method generates a Pareto set first, when the DM is given the opportunity to choose acceptable ones. It is reasonable if the DM is unsure about his/her preferences, or the problem definition is vague about the relative importance of objectives.

4.5.1 ε-Constraint Method The ε-constraint method is a non-scalarizing approach. The original idea was reported by Yacov Haimes [11]. The more comprehensive explanation is provided by Chankong and Haimes [12]. It is proposed to solve the following n-objective problem (Eq. (4.1)) to define a Pareto set: minimize I i (x) , x∈S

(4.8)

142 | 4 Multi-objective optimization in chemical engineering

subject to: I m (x) ≤ ε m ,

i = {1, 2, . . . , n \ m ≠ i} ,

g k (x) ≤ 0,

i = 1, 2, . . . , K ,

h j (x) = 0,

j = 1, 2, . . . , J ,

where ε m are user defined constraints. Note that any of the objective functions can be chosen for minimization. By varying ε m , the Pareto set can be reached. It is reported by authors that the current method can deal with non-convex problems. However, drawbacks still exist. The choice of ε m is not as easy for DM; the technique also significantly increases computation time if the total number of equations (objectives and constraints) is relatively high.

4.6 Interactive methods As it follows from the name, interactive methods require some sort of interaction between the DM and the MOO algorithm. Initially, no a priori information is required, and the DM specifies some objective-related preference information during a search process. Solutions in interactive methods move iteratively, providing the DM with some new solution(s) and allowing the re-specification of his/her preferences, if needed. The interactive methods outcome is one or more Pareto optimal solutions, but not the entire Pareto set. Generally, many other variations exist, which are a kind of extension of classical methods described here with the way how DM should interact with an algorithm. There is a variety of such methods and we will not discuss it here providing only references on some original sources and reviews: – interactive Surrogate Worth Trade-off (ISWT) [12]; – reference point methods [13]; – non-differentiable Interactive Multi-objective Bundle-based Optimization System (NIMBUS) [14]; – step method (STEM) [15]. For an overview of interactive methods, we refer the reader to outstanding reviews by Miettenen [16] and Branke et al. [3].

4.7 Genetic algorithms Genetic algorithms (GAs) are currently one of the most developing groups of methods in MOO. They are “based on the mechanics of natural selection and natural genetics” [17]. Here, we would like to emphasise the power of GAs and discuss them in more details. However, genetic algorithms belong to a posteriori methods; we discuss GAs in

4.7 Genetic algorithms |

143

an individual sub-chapter on account of their fundamental difference to the methods discussed above. The original idea was proposed by Holland [18] as an adaptation concept. Thereafter, Goldberg evolved this theory and formulated general regulations of GAs [17]. GAs have been developed intensively in recent years, but the main principles remain the same. As indicated by Goldberg, main distinctions from classical methods are: – GAs work with a number of points (population) instead of a single one; – GAs treat objective functions directly; there is no need for derivatives, utility functions, or any other auxiliary knowledge; – GAs operators are probabilistic in nature in contrast to deterministic ones used in all classical methods. GAs are notable for their robustness. It is a superior search procedure in many aspects. Unlike many derivative-based methods that can be trapped around local optima, GAs are a global optimum search procedure. They can also treat discontinuous or discrete functions. they overcome issues with the convexity of a Pareto set as well as deal with multi-modal objective functions [19].

4.7.1 About binary-coded variables Preceding the explanation of GAs’ working principles, one has to know about binarycoded variables. The most common representation of a variable utilized by GAs is a binary string. That variable is simply a certain length sequence of ones and zeros (e.g. 1001). If a user deals with continuous variable (e.g. length, product yield, time, etc.), it is required to discretize the variable. The procedure is quite simple. For example, the decision variable x ∈ [Xmin , Xmax ] has to be mapped into a binary string. The user decides to use 4 bits for each variable, in other words, the length of binary string is set to 4 digits. Thereby, we have 24 = 16 possible combinations of strings. Lower and upper bounds are assigned with the values Xmin → 0000 and Xmax → 1111. All other values are mapped in between these two values (Fig. 4.4).

π Xmin

Xmax

0000 – 0001 – 0010 – ........................................... –1111 24 = 16 Fig. 4.4: Mapping of real value into binary variable.

144 | 4 Multi-objective optimization in chemical engineering

The precision of discretization of variables is directly dependent on the string length; the longer the length, the more binary variables can be mapped between the lower and upper limit. The precision π may be calculated as [17]: π=

xmax − xmin . 2lstr − 1

(4.9)

4.7.2 Simple Genetic Algorithm (SGA) For a better understanding the GAs’ principle, let us consider a simple genetic algorithm (SGA) first. The main components of a SGA include: (a) reproduction; (b) crossover; and (c) mutation of genetic operators. At the beginning, the initial population is generated randomly. The population is a set of individuals; each of which represents a single decision variable (or a vector). The reproduction operator is applied to the population to create a “mating pool”. Individuals with a higher objective function value have a higher chance of being copied into a matting pool. Classical and simple way to perform a reproduction operator is a roulette wheel [17]. Once the mating pool is formed, crossover and mutation operations are executed. In a single point crossover, two individuals (called parent chromosomes) are chosen randomly to exchange “information” with each other. They swap binary sequences after the arbitrary position p (which is randomly selected) and then generate “daughter chromosomes” (Fig. 4.5). Crossover point 1111

1100

0000

0011

Parents

Daughters

Fig. 4.5: Representation of single point crossover between two binary strings.

Mutation is also aimed at altering the daughter chromosomes’ binaries but in a different manner. Like mutation in nature, it occurs with a very small probability. Mathematically, it alters one cell in a sequence each time from 0 to 1 or vice versa. It is absolutely necessary to keep diversity in the population [20]. For example, let’s assume a case where all individuals in a population have 0 at kth position, under these conditions the crossover operator cannot create 1 at this point. Mutation allows one to overcome this issue. The best n daughter individuals are taken to form a new mating pool where crossover and mutation are carried out again. This procedure repeats until the termination criterion is satisfied. Below we provide a generalized scheme of SGA (Fig. 4.6).

4.7 Genetic algorithms |

Random generation Population of N individuals 0101 0010 1110 0001 0111

Selection of the “best” individuals

No

Crossover with probability Pcros 10 10

10 00

00 00

00 10

Termination?

145

Mutation with probability Pmut 1 010

0 010

Evaluation of objectives and selection

Yes Pareto optimal front Fig. 4.6: Simple genetic algorithm.

4.7.3 Use of GA in MOO If one has a SOO problem, it is easy to choose the best solutions from the population by comparing the single objective values of individuals. When one deals with multiple objectives, it is not clear how to compare them. To deal with this, Goldberg introduces the concept of non-dominated vectors [17]. Vector a is said to be less than vector b if and only if these two conditions are satisfied simultaneously: – all components of a are less or equal to corresponding components of b; – at least one component of a is strictly less than corresponding element of b; or, in other words (for a minimization MOO problem), a dominates b. If for the vector a there is no such vector c that dominates it, vector a is called non-dominated. From this point of view, a Pareto set is a non-dominated set. Recent GAs’ modifications are more complex than a SGA. There is diversity of different algorithms presented in the open literature: Vector Evaluated Genetic Algorithm (VEGA) [21], Multi-objective GAs (MOGA) [22], Strength Pareto Evolutionary Algorithm (SPEA) [23], Niched-Pareto GAs [24], Predator-Prey Evolution Strategy [25], Rudolph’s Elitist Evolutionary Algorithm [26], NSGA-II [27], Differential Evolution (DE) [28] based methods and many others. We will not discuss most of them here, but refer readers to original sources. We would like to emphasize one of the state-of-the-art algorithms: the non-dominated sorting genetic algorithm II (NSGA-II). The reader can note that this algorithm was used in a majority of MOO problems solved in the literature (Tab. 4.1). After development by Deb [27], it has been widely propagated in optimization problems for chemical engineering as well as for many other fields. NSGA-II is notable for its characteristics, especially its ability to find diverse solutions close to a real Pareto set and

146 | 4 Multi-objective optimization in chemical engineering

the speed of convergence [27]. Here are the elements that contribute to its high performance. – This algorithm uses the concept of elitism. After mating pool formation, N parents and N daughters’ chromosomes are united into a single group of 2N. Selection is carried out over this pool and not only from the original mating pool. If parents are better than their daughters, it allows them to not be excluded them from population, but carry on in the next generation. This allows diversity. – The Non-dominated Sorting Approach is used as a selection procedure. It divides an entire population into groups of non-dominated individuals (non-dominated fronts). Any solution in front 1 is superior to any solution in front 2, and so on. – To maintain the diversity of the population, authors introduced crowding distance. If some region in an objective domain is too populated with individuals, it is reasonable to exclude some of them from the population. The crowding distance of point i represents an average side length of n-dimensional cuboid in objective space, drawn out around point i where two neighboring points are taken as vertices. The higher the crowding distance, the less crowded a region. Points from the same front but with less values than this parameter have less chance to carry on into the next generation. A step-by-step guide to execute NSGA-II, a performance of the algorithm in test problems or other characteristics can be found elsewhere [19, 27].

4.7.4 Constraint handling in GA There are different techniques aimed at constraint handling in GAs. Constraints impose extra conditions on a MOO problem, thereby limiting the search space. Based on this, solutions are divided into feasible and infeasible regions. An infeasible solution cannot be neglected in GAs in order to maintain diversity. Even if a particular solution violates constraints, it should have a chance to remain in the population in order to have a chance to move to a feasible region [19]. To do this, many techniques evaluate the extent of violation from a feasible region. Two noteworthy techniques are discussed below, which have been utilized more frequently while solving applied MOO problems in chemical engineering.

Penalty function approach The penalty function approach modifies the original objective functions by adding a constraint violation to them as follows [20]: minimize P(x) = I i (x) + Ω(R, g(x), h(x)) , x∈S

(4.10)

4.7 Genetic algorithms |

147

where I i (x) is the original objective function, Ω is a penalty term, R is a penalty parameter. The penalty term represents the sum of constraint violations v i (x) from a feasible region: Ω = R ∑ v i (x) . Constraint violations v i (x) could be defined as:

or

{|g k (x)|, v i (x) = { 0, { v i (x) = |h i (x)|2 .

if g k (x),

(4.11)

otherwise, (4.12)

The penalty parameter R is used to have values I i (x) and Ω of a similar magnitude. Hence, if a particular solution overruns a feasible region, the value of the penalty function P(x) increases even if the value of the original objective function I i (x) is small. The solution becomes inferior and has a higher chance of being excluded from the population. One of the main drawbacks of this method is that the penalty function distorts the Pareto front of the original function which cause difficulties finding a true Pareto set.

Constrained tournament method The constrained tournament method is a methods developed for use with GAs only. The approach can treat constraints directly instead of using any objective function transformation. It modifies the tournament selection of individuals for the formation of a mating pool. Now solutions are checked for constraint violation in addition to dominance. Between two infeasible solutions, the one chosen is the one with less constraint violations. When two individuals are picked for a tournament selection, the following “constraint-domination” rules have to be kept: – the feasible individual is always superior to the infeasible individual; – between two infeasible individuals the one with smaller constraint violation should be given priority; and – if both individuals are feasible, the regular (non-constraint) approach should be applied. The generic “constraint-domination” principle can be used with any GAs and doesn’t require extra computational time [19].

148 | 4 Multi-objective optimization in chemical engineering

4.8 Simulated annealing Simulated annealing (SA) is another stochastic-based method of search and, like GAs, belongs to a posteriori methods. The procedure mimics the behavior of molten metals cooling. At high temperatures, metals behave like a liquid where atoms are in chaotic motion. When the cooling is started, atoms lose mobility and begin to form crystalline lattice of solid metal. The rate of cooling strongly affects the structure of crystal, the slower the rate, the more uniformed the structure. Uniformed mono-crystalline structure is more stable (i.e. has minimum energy). SA for optimization was considered in [29]. They applied principles of statistical mechanics of systems in thermal equilibrium to solve the optimization problem. The main principle is based on the Boltzmann probability distribution function. At a given temperature T, the probability of the system to have energy E1 is proportional to 1 exp ( −E kT ), where k is the Boltzmann constant. In this context, probability for a system to move from state 1 to state 2 is given as: −(E2 − E1 ) state 1 = exp ( ). state 2 kT

(4.13)

Hence, if E2 is lower than E1 , then the system definitely turns to state 2. At the same time, if E2 − E1 > 0 a finite probability for transition from 1 to 2 still exists. The higher temperatures T correspond to higher probabilities of state 2 to exist. For energies in Boltzmann distribution equations, the reader has to consider objective values. SA in the simplest form can be described in the following way: the algorithm starts with an initial point x0 (usually random). The random point x1 is generated in the neighborhood of x0 and the objective values are compared at these points. If a new point improves our objectives, it is accepted instead of x0 . If not, the point x1 is ac2 −E 1 ) cepted with the probability exp ( −(EkT ). During the search, the temperature T is slowly decreased (“cooling”) which reduces the probability of a new point with worse objective being accepted. The search continues until some termination criteria are reached, for example, it can be an error between points in subsequent iteration or minimal temperature. One run of SA yields one Pareto optimal solution. Thus multiple simulations are required to obtain a Pareto set. The same principle with some modifications can be applied for MOO problems [30–33]. Algorithms could differ in probability functions or stopping criteria, or they have some operators for a better Pareto distribution. The current technique is less popular than GAs but still has a significant interest in modern MOO applications.

4.9 MOO problems in chemical engineering The popularity of GA methods experienced significant growth since the end of the 1990s when they began to be implemented. The majority of research in chemical en-

4.9 MOO problems in chemical engineering |

149

gineering optimization used GAs as a main technique in search of Pareto optimal solutions, however some other techniques were also used. Some comprehensive reviews of early applications are provided at Bhaskar et al. and Nandasana et al. [6, 34]. Here we will focus on a review of MOO problems and their solutions made in a recent decade for chemical reactors and process engineering. Some more widely presented MOO problems in open literature will be described below. Other MOO problems with objectives, methods and general remarks will be summarized in an auxiliary table at the end of chapter. We recommend readers to refer to original sources if interested in order to have a detailed description of a particular problem. Here we just summarize the main applied issues and concepts of MOO in chemical engineering and of their solutions. It can be seen that even though reactors have different designs and arrangements, the processes have different foundations – continuous or batch type, homogeneous or heterogeneous, in gas or liquid phases, etc. Despite this fact, there are some general concepts that are used for MOO. For example, the most desired intention is to increase the production of a main product (e.g. in terms of yield or selectivity) and minimize side product formation. This usually affects some quality parameters of the product, which becomes a conflicting objective. Additionally, it could be related to a change in heat duties of heat exchangers, the fuel rate into furnaces or other similar parameters. Different scenarios and conflicting objectives could make the formulation very complex. Researchers/engineers are free to choose which objectives have higher importance and have to be given more consideration from practical point of view. In some way, proper MOO formulation itself is an “art” and can play a key role in finding meaningful and appropriate solutions.

4.9.1 Petroleum Processing Engineering Noticeable contribution to GA and its application to MOO in petroleum processing was made by Kasat et al. [35]. They introduced a genetic operator called a Jumping Gene (JG). A JG (or transposon) mimics the real nature phenomena discovered in 1987 by McClintock. The main point of the discovery was that transposon is a DNA sequence which can randomly migrate among chromosomes and replace existing sequences. One of the transposon roles is providing for diversity in genotype. The jumping gene was introduced as a binary sequence that can replace a part of the original individual. First, the chromosome is checked for carrying JG out with some probability PJG . If the condition is satisfied, two positions of binaries, p and q, are randomly chosen in the current chromosome with a total length lstr (p < q ≤ lstr ). The random binary string of length (q − p − 1) is generated and inserted between p and q. Another alternative for JG is to inverse binary sequences between chosen locations. It is reported that these two modifications have the same performance. Authors suggested to implement a JG operator after mutation and combine this operator with NSGA-II (NSGA-II-JG). Using

150 | 4 Multi-objective optimization in chemical engineering

benchmark problems, they demonstrated that the proper choice of PJG (≈ 0.5 or more) provides a faster convergence to a Pareto front and better distribution of the population along it. The JG concept was developed in some later works [36–38] where new modifications with improved characteristics were introduced. We will not describe all of them in detail; we refer readers to the original articles. In their work, NSGA-II-JG was applied for multi-objective optimization of an industrial fluidized catalytic cracking unit (FCCU). The FCC is very relevant for the petroleum processing industry since it is the main process for gasoline production. Industrial FCCUs consist of a reactor-riser and catalyst regenerator. Authors used a five-lump kinetic scheme with two steady state models of these units, previously developed and verified by Arbel et al. [39] and Krishna and Parkin [40]. The two objectives were to maximize the yield of gasoline from the FCCU and minimize coke content on the catalyst. The decision variables used were feed temperature and the catalyst flow rate into the reactor, as well as air temperature and flow rate into the regenerator with lower and upper bounds based on process technology. Again, the problem was solved with both NSGA-II and NSGA-II-JG. The obtained results were compared with their previous work [41], where optimization was carried out with original NSGA-II. The generated Pareto fronts ware similar but with a wider distribution of solutions for NSGA-II-JG. Authors emphasized computational efficiency and the speed of convergence and proposed methods for MOO problems in chemical engineering. Some other petroleum processing MOO are presented in the open literature. Various researchers investigate MOO problems for different types of naphtha catalytic reformers, such as conventional catalytic naphtha reactor (CR) or the thermally coupled fluidized bed naphtha reactor (TCFBNR) [42–45]. Besides the designation for feed conversion into products, the naphtha reforming process could be aimed at a refinery’s hydrogen supply. Because of this, objectives can vary from one reformer to another, depending on their roles in particular productions. In the majority of research, it was proposed to maximize the production of aromatic compounds and hydrogen while other objectives differed. However, only Weifeng et al. [42] treated objectives directly with a Neighbourhood and Achieved Genetic Algorithm (NAGA) to generate an entire Pareto set. Other researchers used summation method to form the SO function, and solve it with methods of differential evolution. All of them could provide improved objectives and propose a better operation conditions for naphtha reformers than current ones.

4.9.2 Steam Reforming The first multi-objective optimization of a side-fired steam reformer was performed by Rajesh et al. [46]. They combined the kinetic model of main reactions, a heat transfer model through a furnace tube wall and the diffusion model in a catalyst pellet. The complex model was utilized for optimization. Authors assumed that the rate of hy-

4.9 MOO problems in chemical engineering |

151

drogen production was kept at a required level. The main operational costs of steam reforming are: (a) methane feed; (b) furnace fuel; and (c) steam. The first objective used was the minimization of the methane feed rate. The second objective used was the maximization of CO in the reformer outflow. The reason for this was that the higher the CO % in outflow, the more heat can be generated at the shift converter and, consequently, more steam can be produced in heat exchangers at the exit of the unit. Decision variables used were the temperature and pressure of the feed flow and its rate, steam/methane ratio (S/C), recycled hydrogen/methane ratio (H/C) and temperature of the furnace gas. Additionally, the process was constrained by a maximum possible furnace wall temperature. Thus, they came up with two objective problem formulations subjected to lower and upper boundaries for decision variables based on process technology and one constraint. The objectives and constraints were treated in the form of a penalty function. A Pareto set was obtained. It was noticed that most of the decision variables didn’t differ significantly for an entire Pareto, but that the S/C ratio makes a significant contribution to the objectives value and for the Pareto distribution. They also studied the effect of catalyst deactivation on the change in optimal parameters. This change wasn’t important due to thermodynamically controlled reactions. Generally it was shown in the work how to apply MOO with GAs to optimize the steam reformer. More precise problem formulations (e.g. constraints, process parameters limits, etc.) for a particular steam reformer can bring different results. The work of Nandasana et al. [47] extended the optimization of a steam reformer dynamic regime. The existing model was modified as a non-steady state to study the effect of disturbances on the reformer. The objectives of MOO were to minimize the reduction of loss of the total (a) hydrogen and (b) steam production caused by a sudden change in some process parameters. Two disturbances were independently introduced to the system: a step decrease of methane feed; and a drop in feed temperature. Authors reported the high computational intensity of MOO problems. They could carry out 9 and 18 generations for the problem respectively. In more recent research, Ebrahimi et al. [48] performed MOO of a steam reforming arrangement for the synthesis gas production (mixture CO and H2 ). They modeled two combinations of top-fired methane steam and auto-thermal reformers, parallel and in series. They formulated objectives similar to Rajesh et al. [46]: (a) maximize production of syngas; (b) minimize furnace fuel consumption; and (c) minimize CO2 releases. Like in previous research, the main constraint for the steam reforming operation was the maximum tube wall temperature. Obtained Pareto sets showed that a parallel arrangement is superior for higher syngas production while the configuration in series allows for a decrease of fuel consumption and CO2 release.

152 | 4 Multi-objective optimization in chemical engineering

4.9.3 Polymer industry MOO in polymer manufacturing has been an intensive research field in so far as such processes with multiple objectives result in more meaningful solutions. Many works had been aimed at the optimization of polymerization reactors’ and processes’ performances. One of the first MOO problems was solved for the Nylon 6 reactor by Mitra et al. [49]. They utilized a kinetics scheme combined with a batch reactor model. Objectives to minimize were (a) reaction time and (b) undesired product concentration. Constraints implemented were desired monomer conversion and the degree of polymerization. For a solution of a current MOO problem, they used NSGAs, and constraints were handled by penalty functions. Authors varied different GA parameters (e.g. number of generations, crossover probability) to show the stability of obtained Pareto sets because no significant changes were observed. Earlier authors tried to carry out SOO for the same system with Pontryagin’s principle; this failed due to some numerical complexity. It was emphasized that NSGA allowed for the overcoming of previous problems and generation of a reasonable set of optimal solutions. An interesting discovery was found in Bhaskar et al. [50]. The authors carried out MOO for a polyethylene terephthalate wiped-film reactor. They chose to minimize the (a) acid and (b) vinyl end group concentration in the polymer product for a better polymer quality. Optimization with NSGA showed that the problem had a unique solution instead of a Pareto set. To confirm the results, they carried out the SOO problem for each objective independently; it resulted in the same solution. Later, in another work by Bhaskar et al. [51], it was pointed out that the unique solution was dependent on the seed random generator (a number used in computer code to execute randomization). By varying this number, they always obtained different single optimal points. Also they showed that NSGAs didn’t obtain an optimal point if more than one decision variable was used. The conclusion was made that NSGAs failed to converge to global optimal solutions and some other search technique is required. Current issues were resolved in Babu et al. [52], in which the authors used a multi-objective differential evolution (MODE) for the optimization of the same system. Different MOO cases were considered and MODE converged to a Pareto front in each of them. Many other similar works for optimization of industrial continuous or batch polymerization processes are made. In general, it can be noted that the main objectives in MOO problems could be: – maximization of monomer conversion; – minimization of the concentration of side products or some functional groups; and – maintainence of quality-related parameters on a desired level (e.g. molecular weight, degree of polymerization).

4.9 MOO problems in chemical engineering |

153

Some MOO problems include a design stage and significant improvement in the reactor’s performance is reported. For example, in the work by Agrawal et al. [53] the operation of another type of polyethylene reactor, tubular, was optimized. Two objectives were to maximize monomer conversion and minimize the concentration of side products. The reactor and jacket diameter, and length of reactor zones were included as decision variables. So they carried out optimization of both design-stage and operation-stage optimization. They reported improved results in objectives when compared to the case when design variables are not included into the MOO problem [54]. For MOO problems in polymerization reactors and processes, researchers mentioned significant computational issues such as: to obtain global optimal solutions, large computation time is required. The mathematical models are relatively complex for such processes, because they include mass, heat and momentum balance equations that could include comprehensive equations involving partial derivatives or other intensive mathematical variables involving highly non-linear equations. To carry out MOO using GAs, it is required to perform a number of simulation runs to evaluate objectives for one population, while a MOO search requires a number of generations to obtain convergence to a global optimal Pareto front. All together, it increases computational time up to some hours or even days and hence necessitates the use of supercomputers. Besides the polymerization processes, there are works made in monomer production. Most commonly used objectives used in these MOO problems are monomer’s yield and selectivity. Among the works, there is a group of researchers who optimized styrene production. They carried out various MOOs for different reactor types using different algorithms. Firstly, Yee et al. [55] performed two-objective optimization for the operation of adiabatic and steam-injected reactors. The same work was done by Li et al. [56], but including reactor design parameters into decision variables. Both used NSGAs with a penalty function approach and obtained smooth Pareto fronts. Babu et al. [57] performed a MOO for an adiabatic styrene reactor with the same problem formulation but used a multi-objective differential evolution. They reported a better Pareto front. However, it can be noted that MODE hadn’t affected the Pareto front significantly, but there are still improvements in objective values for some MOO cases. Tarafder et al. [58] compared performance of three types, single-, double-bed and steam-injected reactors, with NSGA-II in a three-objective problem formulation (maximization of yield and selectivity of styrene plus minimization of heat-exchanger duty). They showed better objective values for the double-bed reactor. It can provide better productivity for styrene with a higher selectivity at the same time.

FCC reactor-regenerator

Petroleum processing

Process/unit

The same two objectives plus – minimize air feed rate to regenerator

– maximize gasoline yield – minimize % CO in flue gas constrained by coke content on catalyst

– maximize yield of gasoline – minimize coke percentage on catalyst

The same two objectives plus – minimize air feed rate to regenerator

– maximize gasoline yield – minimize air feed rate to regenerator constrained by CO % in flue gas

– maximize gasoline yield – minimize CO % in flue gas constrained by coke content on catalyst

Objectives/constraints

MOSA

NSGA-II-JG

NSGA-II

Optimization method

Tab. 4.1: Application of GAs in chemical reactors and processes engineering.

[35]

[59]

Pareto set is comparable with ones obtained with NSGA-II.

[41]

Reference

Obtained Pareto set with better distribution and faster convergence than at Kasat et al. [41].

The satisfying optimal solution can be chosen from the obtained Pater set by DM.

Remarks and comments

154 | 4 Multi-objective optimization in chemical engineering

[45] Both arrangements perform similarly but SMS has some design advantages and proposed as better one.

Maximize: – hydrogen flow rate – aromatics flow rate Two reactor arrangements in series – SMS and SMM – are investigated.

Maximize yield of: – ethylene – propylene

Spherical (S) and Tubular Membrane Naphtha (M) Reforming Reactor

Naphtha Pyrolysis

Optimal solutions obtained. MOPDECES performs slightly better.

[44]

Single optimal solution obtained which allowed improving reactors performance.

Maximize: – hydrogen production – aromatics production – nitrobenzene conversion – aniline flow rate

MOPDE-CES, NSGA-II

[43]

Reactor performance compared to conventional naphtha reformer.

Objective sum method, solved by differential evolution

Maximize: – hydrogen production – aromatics production and selectivity – aniline flow rate

Naphtha Catalytic Reformer with Thermally Coupled Fluidized Bed Heat Exchanger

[60]

[42]

Pareto set obtained which is superior to current unit operation performance.

NAGA

– maximize light aromatics yield – Minimize heavy aromatics yield

Naphtha Catalytic Reforming Reactor

Reference

Remarks and comments

Optimization method

Objectives/constraints

Process/unit

Tab. 4.1 (continued)

4.9 MOO problems in chemical engineering |

155

Paraffin dehydrogenation reactor of LAB plant

3 MOO cases:

HVGO Hydrocracker

For process product (olefins) maximize: – production rate – selectivity

constrained by inlet temperature at hydrocracker and outlet temperature at beds, feed conversion

– maximize the sum of all desired products – maximize the sum of heavy desired products

2 MOO cases:

constrained by inlet temperature at hydrocracker and outlet temperature at beds, liquid velocity rate, feed conversion

– minimize light products flow rate – maximize heavy products flow rate

– maximize diesel flow rate – minimize hydrogen flow rate

– maximize kerosene flow rate – minimize hydrogen flow rate

Objectives/constraints

Process/unit

Tab. 4.1 (continued)

NSGA-II with crowding tournament selection operator

GA with artificial neural network model

Real-coded NSGA-II with simulated binary crossover

Optimization method

Dynamic optimization was carried out. Shift of Pareto from is shown due to catalyst deactivation.

Shown possibility to improve reactors performance up to 16 %.

Pareto set obtained for all cases. Wide range of equally optimal solution are presented for DM.

Remarks and comments

[63]

[62]

[61]

Reference

156 | 4 Multi-objective optimization in chemical engineering

Objectives/constraints

Industrial Steam Reformer

The satisfying optimal solution can be chosen from obtained Pater set by DM.

The satisfying optimal solution can be chosen from obtained Pater set by DM.

Comparable Pereto set to the one obtained using NSGA-II. Authors used different operating conditions and transition between them. Final problem is formulated in form of singleobjective function.

NSGA-II

MOSA-JG, MOSA-aJD GA

For a step disturbances of (a) methane feed (b) temperature minimize deviation from steady-state values for – hydrogen production – steam production

Same as Nandasana et al. [47] but only for case (a) step decrease in feed

– maximize methane conversion – Maintain desired ratios for H2 /CO2 and H2 /CO MOO problem solved for dynamic model.

Remarks and comments

NSGA with penalty function approach

Optimization method

For a required hydrogen rate production – minimize methane feed – maximize CO at reactor’s outflow constrained by maximum tube wall temperature

Steam reforming processes

Process/unit

Tab. 4.1 (continued)

[65]

[64]

[47]

[46]

Reference

4.9 MOO problems in chemical engineering |

157

Nylon 6 semibatch reactor

Polymers synthesis Superior approach comparing to previous attempt to carry out MOO.

NSGA-II-aJG has a better distribution of individuals in Pareto front.

NSGA with penalty function approach

NSGA-II-aJG, MOSA-aJG

For a required monomer conversion minimize: – dimensionless reaction time – dimensionless side product concentration constrained by required values for average polymer length

Cases 1 and 2: same as Mitra et al. [49] but different decision variables Case 2: – Same as case 2 – maximize monomer conversion

Parallel configuration is better for syngas. Arrangement in series is superior for lower fuel consumptions and CO2 release.

NSGA-II

For two arrangements of reactors – in parallel and in series: – maximize production of syngas – minimize furnace fuel consumption

Methane and autothermal steam reformers

Pareto set is obtained. Among it, authors chose one point with H2 /CO ratio = 1 as an optimal operating point.

NSGA-II

– maximize methane conversion – maximize CO selectivity – minimize CO2 feed rate

Autothermal reformer

Remarks and comments

Optimization method

Objectives/constraints

Process/unit

Tab. 4.1 (continued)

[67]

[49]

[48]

[66]

Reference

158 | 4 Multi-objective optimization in chemical engineering

Decision support system narrowed the Pareto set.

Diploid GA followed by decision support system

– maximize styrene conversion minimize deviations from desired values for: – Polymer average molecular weight – Number of particles per liter

The satisfying optimal solution can be chosen from obtained Pater set by DM.

Authors’ version of MOGA (includes real-coded variables, elitism, niche count) with fuzzy penalty function approach

Styrene emulsion homopolymerization

[69]

Pareto set obtained in contrast to Bhaskar et al. [50].

MODE with penalty function approach

Same as Bhaskar et al. [50]

– maximize styrene conversion – minimize remaining initiator concentration in final product

Fails to converge the optimum solution for multiple decision variables.

NSGA with penalty function approach

Same as Bhaskar et al. [50] plus additional constraint for di-ethylene glycol group concentration

Isothermal polystyrene reactor

[52]

Single optimal solution.

NSGA with penalty function approach

Minimize: – acid – vinyl groups in the product constrained by desired degree of polymerization

Poly-ethylene wiped-Film reactor

[70, 71]

[51]

[50]

[68]

The satisfying optimal solution can be chosen from obtained Pater set by DM.

NSGA with penalty function approach

– maximize monomer conversion – minimize length of film reactor constrained by the end value of polymer molecular weight

Sheet-molding for poly(methyl methacrylate)

Reference

Remarks and comments

Optimization method

Objectives/constraints

Process/unit

Tab. 4.1 (continued)

4.9 MOO problems in chemical engineering |

159

Epoxy polymerization

Process/unit

Tab. 4.1 (continued) Remarks and comments Optimization for design and control is carried out.

Pareto set obtained for each case and the satisfying optimal solution can be chosen from obtained Pater set by DM. It is found that for molecular weight vs. polydispersity index, set is non-convex. Binary and real coded NSGA-II performs similarly.

Optimization method Mixed-integer dynamic optimization, ε-constraint approach

NSGA-II with crowding tournament selection operator

NSGA-II with crowding tournament selection operator

Objectives/constraints

– minimize operating cost of reactor – minimize integral square difference of average molecular weight from its desired value

For polymer product: – maximize molecular weight – minimize reaction time Constrained by minimum desired molecular weight and maximum desired polydispersity index

Case 1: – maximize polymer molecular weight – minimize polymer polydispersity index Case 2: – maximize concentration of species with glycidyl ether groups at both ends – minimize polymer chain propagation Case 3: – Same as case 2 – + minimize total addition of NaOH

[74]

[73]

[72]

Reference

160 | 4 Multi-objective optimization in chemical engineering

Styrene and acrylonitrile copolymerization in semi-batch reactor

Process/unit

Tab. 4.1 (continued)

[76]

[77]

Pareto set obtained in both cases. Process control policies are defined.

Dynamic optimization is carried out.

NSGA-II, Real-coded NSGA-II

NSGA-II with crowding tournament selection operator

Differential evolution

3 MOO problems for following objectives: – maximize polymer’s average molecular weight – minimize polydispersity index – minimize reaction time

Case 1: – maximize monomer conversion – minimize polydispersity index of final product Case 2: – Same as case 1 – minimize presence of unreacted monomer at reactor

Minimize deviations from desired: – Copolymer molecular weight – Copolymer composition

[78]

[75]

Real-coded NSGA-II

Reference

Case 1: – maximize concentration of particular species – minimize polymer chain propagation – minimize reaction time Case 2: – minimize total addition of NaOH – + last 2 objectives from case 1

Remarks and comments

Optimization method

Objectives/constraints

4.9 MOO problems in chemical engineering |

161

A single solution was chosen from obtained Pareto set with decision support system.

Evolutionary algorithm followed by multi-attributive utility theory

– maximize monomer conversion – deviation from desired glass temperature profile

Styrene and butyl acrylate emulsion copolymerization reactor

[81]

[80]

[79]

Improved performance of method comparing to NSGA-II.

NSGA-II combined neural network

– maximize monomer conversion – minimize the difference between real and desired molecular weight of polymer

[53]

Improved reactor performance comparing to operation MOO only. Constraint-dominance approach is better than penalty function.

NSGA-II and JD adaptations with: – Penalty function approach – Constraint dominance approach

Two MOO problems: Same as Agrawal et al. [53, 54] Same as Agrawal et al. [54] +1 objective to minimize compressor operating cost * Design parameters as reactor length and diameter were included as decision variables.

Polysiloxane synthesis

[54]

All algorithms provide similar, but NSGA-II converges faster.

NSGA and JD adaptations, NSGA-II and JD adaptations with penalty function approach

For poly-ethylene: – maximize monomer conversion – minimize side products constrained by desired range for product molecular weight and maximum process temperature

Poly-ethylene tubular reactor

Reference

Remarks and comments

Optimization method

Objectives/constraints

Process/unit

Tab. 4.1 (continued)

162 | 4 Multi-objective optimization in chemical engineering

Objectives/constraints

Adiabatic Styrene Reactor

Adiabatic and Steam-Injected Styrene Reactors

MODE provides improved Pareto set comparing to the one from Yee et al. [55]. Better objective values as well as less computational time for Tabu Search.

MODE with penalty function approach Tabu Search, GA

2 MOO problems: – maximize styrene production – minimize undesired products for new and deactivated catalyst

The satisfying optimal solution can be chosen from obtained Pater set by DM.

NSGA with penalty function approach

For styrene maximize: – productivity – selectivity constrained by steam feed rate and inlet streams temperature plus exit pressure for steam-injected reactor * Design parameters as reactor length and diameter were included as decision variables.

Same as Yee et al. [55]

The satisfying optimal solution can be chosen from obtained Pater set by DM.

Remarks and comments

NSGA with penalty function approach

Optimization method

4 MOO problems: for styrene maximize either two of three objectives or all of them: – productivity – selectivity – yield constrained by steam feed rate and inlet streams temperature

Monomers for polymer industry

Process/unit

Tab. 4.1 (continued)

[82]

[57]

[56]

[55]

Reference

4.9 MOO problems in chemical engineering |

163

Double-bed reactor has higher productivity.

NSGA-II with crowding tournament selection operator

MODE with penalty function approach

For styrene: – maximize productivity – maximize selectivity – heat duty of heat exchanger constrained by inlet streams temperatures, reactor pressure * Design parameters such as reactor length and diameter were included as decision variables.

Same as at Yee et al. [55]

Single-bed, steaminjected and double-bed styrene reactors

Adiabatic and steam-injected styrene reactor

Obtained Pareto set has better objective values than the one with NSGA.

Feed temperature and reactor length are mostly affect Pareto optimal solutions.

NSGA-II with crowding tournament selection operator

For ethylene maximize: – flow rate – conversion – selectivity constrained by reactor pressure and temperature

Proposed algorithm compared with other well-known MOEA and showed better performance.

Author’s Hybrid-MODE

Same as Yee et al. [55]

Remarks and comments

Optimization method

Objectives/constraints

Ethylene Reactor

Process/unit

Tab. 4.1 (continued)

[85]

[58]

[84]

[83]

Reference

164 | 4 Multi-objective optimization in chemical engineering

NSGA with penalty function approach NSGA with penalty function approach NSGA with penalty function approach

Same as at Rajesh et al. [86] plus minimize heat duty of reformer tubes

Same as at Oh et al. [87]

Same as at Rajesh et al. [86]

Purified terephtalic acid oxidation

4 MOO cases: – minimize concentration of intermediate product in outflow – maximize feed rate to reactor with different number of decision variables

NAGA

NSGA with penalty function approach

Optimization method

Maximize production of: – hydrogen – steam constrained by maximum tube wall temperature, H2 O/H2 ratio and some other limitations for equipment operating condition

Objectives/constraints

Other processes an reactors

Hydrogen plant with absorber and methanator instead of PSA unit

Hydrogen plant (natural gas operating)

Hydrogen production

Process/unit

Tab. 4.1 (continued)

Pareto sets obtained for each case. The more decision variables are taken into account the better objectives are reached.

The satisfying optimal solution can be chosen from obtained Pater set by DM.

Pareto set is affected by origin of feed (comparing to Oh et al. [87]).

More practical information about Pareto front for three-objective optimization problem.

The satisfying optimal solution can be chosen from obtained Pater set by DM.

Remarks and comments

[90]

[89]

[88]

[87]

[86]

Reference

4.9 MOO problems in chemical engineering |

165

Oxidative coupling of methane in simulated moving bed reactor

Membrane hydrogen synthesis reactor

NSGA-II-JG

Same as at Kundu et al. [93]

MOO problem is similar to [93], but reactor configuration is different.

Pareto-optimal sets are provided.

NSGA-II-JG

– maximize methane conversion – maximize selectivity for ethane and ethylene operation and design MOO are carried out.

[94]

[93]

[92] Pareto set for hydrogen reactor is very clear and readable while the one for methanol has scatter data.

NSGA-II

– maximize desired product rate – minimize feed rate – minimize exergy loss in reactor

Membrane methanol synthesis reactor

[37]

Guided NSGA-II needs proper choice of genetic parameters but provides faster convergence to Pareto.

NSGA-II-aJG, Guided NSGA-II-aJG with penalty function approach

For 2 different reactor arrangements: – maximize product yield – minimize catalyst mass

Phthalic anhydride catalytic reactor

[91]

Empirical process model was utilized. Better objective values are reported comparing with previous SOO.

Real-coded NSGA

– maximize methane conversion – maximize total selectivity for CO production – keep H2 /CO molar ratio around required value constrained by O2 /CH4 molar ratio, gas stream velocity

Syngas production using CO2 reforming and natural gas (methane) partial oxidation

Reference

Remarks and comments

Optimization method

Objectives/constraints

Process/unit

Tab. 4.1 (continued)

166 | 4 Multi-objective optimization in chemical engineering

[96, 97]

[98]

Pareto set obtained for each MOO case. Three-objective problems provide better range of solutions. Design optimization provides better objective values.

Both methods provided the same Pareto sets.

NSGA-II-aJG

Normalized normal constraint and normal boundary intersection methods

Operation and design MOO problems using 2 or 3 objectives from list are solved: maximize: – ethylene production – propylene production – ethylene selectivity – furnace run length minimize: – severity – heat duty

– maximize styrene yield – maximize nitrobenzene conversion Optimization problem included design variables

Thermal cracker for LPG

Autothermal membrane reactor for simultaneous dehydrogenation of ethylbenzene to styrene with the hydrogenation of nitrobenzene to aniline

[95]

Reactor length and diameter were included into MOO on design stage. MOO of design stage showed significant improvement in objectives.

NSGA-II-aJG

3 objectives are: – maximize methane conversion – maximize selectivity for ethane and ethylene – maximize yield for ethane and ethylene and 2 MOO cases for operation stage and 2 MOO cases for design case are solved for two out of three objectives

Porous ceramic membrane reactor for oxidative coupling of methane

Reference

Remarks and comments

Optimization method

Objectives/constraints

Process/unit

Tab. 4.1 (continued)

4.9 MOO problems in chemical engineering |

167

168 | 4 Multi-objective optimization in chemical engineering

4.10 Conclusions The multi-objective optimization approach is superior to classical single-objective optimizations. It can take into account more than one objective; this is very important when objectives conflict with one another. Reactors and processes systems in chemical engineering include many parameters (qualitative or quantitative parameters that characterize the process performance), which cannot be improved without any detriment to others. So the application of MOO can play a vital role in making process operation improvements. In summary, the general ideas of MOO techniques as well as the advantages in applying the concept of MOO in the design and operation of chemical reactors and processes engineering are reported. – MOO is based on the concept of Pareto optimality. In contrast with SOO, no single optimal solution but a set of optimal solution results are more meaningful. This set is called a Pareto set. None of the solutions in a Pareto set is better than any others in the set. – Various MOO methods are discussed that helps in search of solutions in the form of a Pareto set. There is a variety of these methods available. Each has its own advantages and disadvantages. Researchers are free to choose any of them depending on the particular problem he/she is trying to solve. – In the field of chemical reactors and processes engineering, a group of stochastic optimization methods, called genetic algorithms, showed robustness in finding Pareto-optimal solutions. – Genetic algorithms are not based on a deterministic mechanism of search, and require no extra a priori knowledge (like weighting information of preference order) about MOO of conflicting objectives. Also, GAs work with a population of solutions simultaneously, not a single one; hence they search a global space for optimum solutions. – Many of the reported work carried out by researchers shows the significance of MOO in chemical engineering. It can be done at the operation stage level, because many industrial reactors operate in non-optimal regimes and there’s still room for improvements. It’s also useful for the design stage, which can significantly improve a reactor/system performance when designing reactors. – Among GAs, there are some more advanced algorithms that are able to converge to a Pareto front in less computation time while providing better distribution of solutions. Researchers should take it into account when applying them to MOO problem. – A majority of Pareto fronts in chemical engineering are convex in nature; however, it’s not an absolute rule. – Many problems considered only 2, or at most 3, objectives. Researchers are trying to pick out more important ones based on their knowledge about particular systems. Also, it is more difficult to visualize and analyze results if more than 3 objective functions are chosen.

References |

169

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

[14] [15] [16] [17] [18] [19] [20] [21] [22] [23]

Edgar TF, Himmelblau DM, Lasdon LS. Optimization of chemical processes. McGraw Hill, 2001. Miettinen K. Nonlinear Multiobjective Optimization. Kluwer Academic Publishing, 1999. Branke J, Deb K, Miettinen K, Slowinski R. Multiobjective Optimization. Interactive and Evolutionary Approaches. Springer, 2008. Marler R, Arora J. Survey of multi-objective optimization methods for engineering. Struct Multidiscip Optim 2004;26:369–395. Wierzbicki A, Makowski M, Wessels J. Model-based decision support methodology with environmental applications. Dordrecht; Boston: Kluwer Academic Publishers, 2000. Bhaskar V, Gupta SK, Ray AK. Applications of multiobjective optimization in chemical engineering. Reviews in Chemical Engineering 2000;16:1–54. Yu PL, Leitmann G. Compromise Solutions, Domination Structures, and Salukvadze’s Solution. J Optimiz Theory Appl 1974;13:362–378. Fishburn PC. Lexicographic Orders, Utilities and Decision Rules: A Survey. Management Science 1974;20:1442–1471. Charnes A, Cooper WW. Management models and industrial applications of linear programming. New York: Wiley, 1961. Tamiz M, Jones D, Romero C. Goal programming for decision making: An overview of the current state-of-the-art. Eur J Oper Res 1998;111:569–581. Haimes Y. On a bicriterion formulation of the problems of integrated system identification and system optimization. IEEE transactions on systems, man, and cybernetics 1971:296–297. Vira Chankong, Haimes YY. Multiobjective decision making :theory and methodology. New York: North Holland, 1983. Wierzbicki A. The use of reference objectives in multiobjective optimization. In: Fandel G, Gal T, eds. Multiple Criteria Decision Making Theory and Application. Berlin: Springer Berlin Heidelberg, 1980:468–486. Miettinen K, Mäkelä MM. Interactive bundle-based method for nondifferentiable multiobjective optimization: nimbus §. Optimization 1995;34:231–246. Benayoun R, de Montgolfier J, Tergny J, Laritchev O. Linear programming with multiple objective functions: Step method (stem). Math Program 1971;1:366–375. Miettinen K. Nonlinear Multiobjective Optimization. Dordrecht: Kluwer Academic Publishers, 1999. Goldberg DE. Genetic algorithms in search, optimization, and machine learning. Reading, Mass; Don Mills, Ont.: Addison-Wesley Pub. Co., 1989. Holland JH. Adaptation in natural and artificial systems :an introductory analysis with applications to biology, control, and artificial intelligence. Cambridge, Mass.: MIT Press, 1992. Deb K. Multi-objective optimization using evolutionary algorithms. Chichester, England; New York: John Wiley & Sons, 2001. Deb K. Optimization for engineering design: Algorithms and Examples. New Delhi: PHI Learning Private Limited, 1995. Schaffer JD. Some Experiments in Machine Learning Using Vector Evaluated Genetic Algorithms (Artificial Intelligence, Optimization, Adaptation, Pattern Recognition). 1984. Fonseca CM, Fleming PJ. Genetic Algorithms for Multiobjective Optimization: Formulation, Discussion and Generalization. 1993. Zitzler E, Thiele L. An evolutionary Algorithm for Multiobjective Optimization: The Strength Pareto Approach. 43. Zurich: Swiss Federal Institute of Technology, 1998.

170 | 4 Multi-objective optimization in chemical engineering

[24] Horn J, Nafpliotis N, Goldberg DE. A niched Pareto genetic algorithm for multiobjective optimization. Evolutionary Computation, 1994 IEEE World Congress on Computational Intelligence, Proceedings of the First IEEE Conference on 1994;1:82–87. [25] Laumanns M, Rudolph G, Schwefel H. A spatial predator-prey approach to multi-objective optimization: A preliminary study. 1998:241–249. [26] Rudolph G. Evolutionary Search under Partially Ordered Fitness Sets. 2001:818–822. [27] Deb K, Pratap A, Agarwal S, Meyarivan T. A fast and elitist multiobjective genetic algorithm: NSGA-II. Ieee Transactions on Evolutionary Computation 2002;6:182–197. [28] Storn R, Price K. Differential evolution - A simple and efficient heuristic for global optimization over continuous spaces. J Global Optimiz 1997;11:341–359. [29] Kirkpatrick S, Gelatt CD, Vecchi MP. Optimization by Simulated Annealing. Science 1983;220: 671–680. [30] Suppapitnarm A, Seffen KA, Parks GT, Clarkson PJ. A simulated annealing algorithm for multiobjective optimization. Engineering Optimization 2000;33:59–85. [31] Czyzżak P, Jaszkiewicz A. Pareto simulated annealing—a metaheuristic technique for multipleobjective combinatorial optimization. J Multi-Crit Decis Anal 1998;7:34–47. [32] Suman B. Study of simulated annealing based algorithms for multiobjective optimization of a constrained problem. Comput Chem Eng 2004;28:1849–1871. [33] Suman B. Study of self-stopping PDMOSA and performance measure in multiobjective optimization. Comput Chem Eng 2005;29:1131–1147. [34] Nandasana A, Ray A, Gupta S. Applications of the Non-Dominated Sorting Genetic Algorithm (NSGA) in Chemical Reaction Engineering. International Journal of Chemical Reactor Engineering 2003;1:Review R2. [35] Kasat R, Gupta S. Multi-objective optimization of an industrial fluidized-bed catalytic cracking unit (FCCU) using genetic algorithm (GA) with the jumping genes operator. Comput Chem Eng 2003;27:1785–1800. [36] Guria C, Bhattacharya PK, Gupta SK. Multi-objective optimization of reverse osmosis desalination units using different adaptations of the non-dominated sorting genetic algorithm (NSGA). Comput Chem Eng 2005;29:1977–1995. [37] Bhat GR, Gupta SK. MO optimization of phthalic anhydride industrial catalytic reactors using guided GA with the adapted jumping gene operator. Chemical Engineering Research & Design 2008;86:959–976. [38] Agarwal A, Gupta SK. Jumping gene adaptations of NSGA-II and their use in the multi-objective optimal design of shell and tube heat exchangers. Chemical Engineering Research & Design 2008;86:123–139. [39] Arbel A, Huang Z, Rinard IH, Shinnar R, Sapre AV. Dynamic and Control of Fluidized Catalytic Crackers. 1. Modeling of the Current Generation of FCC’s. Ind Eng Chem Res 1995;34:1228– 1243. [40] Krishna A, Parkin E. Modeling the Regenerator in Commercial Fluid Catalytic Cracking Units. Chem Eng Prog 1985;81:57–62. [41] Kasat RB, Kunzru D, Saraf DN, Gupta SK. Multiobjective optimization of industrial FCC units using elitist nondominated sorting genetic algorithm. Ind Eng Chem Res 2002;41:4765–4776. [42] Hou Weifeng, Su Hongye, Mu Shengjing, Chu Jian. Multiobjective optimization of the industrial naphtha catalytic reforming process. Chin J Chem Eng 2007;15:75–80. [43] Rahimpour MR, Iranshahi D, Pourazadi E, Bahmanpour AM. Boosting the gasoline octane number in thermally coupled naphtha reforming heat exchanger reactor using de optimization technique. Fuel 2012;97:109–18.

References |

171

[44] Pourazadi E, Vakili R, Iranshahi D, Jahanmiri A, Rahimpour MR. Optimal design of a thermally coupled fluidised bed heat exchanger reactor for hydrogen production and octane improvement in the catalytic naphtha reformers. Can J Chem Eng 2013;91:54–65. [45] Iranshahi D, Rahimpour MR, Paymooni K, Pourazadi E. Utilizing DE optimization approach to boost hydrogen and octane number, through a combination of radial-flow spherical and tubular membrane reactors in catalytic naphtha reformers. Fuel 2013;111:1–11. [46] Rajesh J, Gupta S, Rangaiah G, Ray A. Multiobjective optimization of steam reformer performance using genetic algorithm. Ind Eng Chem Res 2000;39:706–717. [47] Nandasana A, Ray A, Gupta S. Dynamic model of an industrial steam reformer and its use for multiobjective optimization. Ind Eng Chem Res 2003;42:4028–4042. [48] Ebrahimi H, Behroozsarand A, Zamaniyan A. Arrangement of primary and secondary reformers for synthesis gas production. Chemical Engineering Research & Design 2010;88:1342–1350. [49] Mitra K, Deb K, Gupta SK. Multiobjective dynamic optimization of an industrial nylon 6 semibatch reactor using genetic algorithm. J Appl Polym Sci 1998;69:69–87. [50] Bhaskar V, Gupta SK, Ray AK. Multiobjective optimization of an industrial wiped-film pet reactor. AICHE J 2000;46:1046–58. [51] Bhaskar V, Gupta S, Ray A. Multiobjective optimization of an industrial wiped film poly(ethylene terephthalate) reactor: some further insights. Comput Chem Eng 2001;25:391– 407. [52] Babu BV, Mubeen JHS, Chakole PG. Simulation and optimization of wiped-film poly-ethylene terephthalate (PET) reactor using multiobjective differential evolution (MODE). Mater Manuf Process 2007;22:541–552. [53] Agrawal N, Rangaiah GP, Ray AK, Gupta SK. Design stage optimization of an industrial lowdensity polyethylene tubular reactor for multiple objectives using NSGA-II and its jumping gene adaptations. Chemical Engineering Science 2007;62:2346–2365. [54] Agrawal N, Rangaiah G, Ray A, Gupta S. Multi-objective optimization of the operation of an industrial low-density polyethylene tubular reactor using genetic algorithm and its jumping gene adaptations. Ind Eng Chem Res 2006;45:3182–3199. [55] Yee A, Ray A, Rangaiah G. Multiobjective optimization of an industrial styrene reactor. Comput Chem Eng 2003;27:111–130. [56] Li Y, Rangaiah GP, Ray AK. Optimization of Styrene Reactor Design for Two Objectives using a Genetic Algorithm. Int J Chem React Eng 2003;1:A13. [57] Babu B, Chakole P, Mubeen J. Multiobjective differential evolution (MODE) for optimization of adiabatic styrene reactor. Chem Eng Sci 2005;60:4822–4837. [58] Tarafder A, Rangaiah G, Ray A. Multiobjective optimization of an industrial styrene monomer manufacturing process. Chem Eng Sci 2005;60:347–363. [59] Sankararao B, Gupta SK. Multi-objective optimization of an industrial fluidized-bed catalytic cracking unit (FCCU) using two jumping gene adaptations of simulated annealing. Comput Chem Eng 2007;31:1496–1515. [60] Wang X, Tang L. Multiobjective Operation Optimization of Naphtha Pyrolysis Process Using Parallel Differential Evolution. Ind Eng Chem Res 2013;52:14415–14428. [61] Bhutani N, Ray AK, Rangaiah GP. Modeling, simulation, and multi-objective optimization of an industrial hydrocracking unit. Ind Eng Chem Res 2006;45:1354–1372. [62] Bhutani N, Rangaiah GP, Ray AK. First-principles, data-based, and hybrid modeling and optimization of an industrial hydrocracking unit. Ind Eng Chem Res 2006;45:7807–7816. [63] Bayat M, Dehghani Z, Rahimpour MR. Dynamic multi-objective optimization of industrial radial-flow fixed-bed reactor of heavy paraffin dehydrogenation in LAB plant using NSGA-II method. Journal of the Taiwan Institute of Chemical Engineers 2013.

172 | 4 Multi-objective optimization in chemical engineering

[64] Sankararao B, Gupta SK. Multiobjective optimization of the dynamic operation of an industrial steam reformer using the jumping gene adaptations of simulated annealing. Asia-Pacific Journal of Chemical Engineering 2006;1:21–31. [65] Alizadeh A, Mostoufi N, Jalali-Farahani F. Multiobjective dynamic optimization of an industrial steam reformer with genetic algorithms. Int J Chem React Eng 2007;5:A19. [66] Behroozsarand A, Ebrahimi H, Zamaniyan A. Multiobjective Optimization of Industrial Autothermal Reformer for Syngas Production Using Nonsorting Genetic Algorithm II. Ind Eng Chem Res 2009;48:7529–7539. [67] Ramteke M, Gupta SK. Multiobjective Optimization of an Industrial Nylon-6 Semi Batch Reactor Using the a-Jumping Gene Adaptations of Genetic Algorithm and Simulated Annealing. Polym Eng Sci 2008;48:2198–2215. [68] Zhou FB, Gupta SK, Ray AK. Multiobjective optimization of the continuous casting process for poly (methyl methacrylate) using adapted genetic algorithm. J Appl Polym Sci 2000;78:1439– 1458. [69] Silva C, Biscaia E. Genetic algorithm development for multi-objective optimization of batch free-radical polymerization reactors. Comput Chem Eng 2003;27:1329–1344. [70] Massebeuf S, Fonteix C, Hoppe S, Pla F. Development of new concepts for the control of polymerization processes: Multiobjective optimization and decision engineering. I. Application to emulsion homopolymerization of styrene. J Appl Polym Sci 2003;87:2383–2396. [71] Fonteix C, Massebeuf S, Pla F, Kiss L. Multicriteria optimization of an emulsion polymerization process. Eur J Oper Res 2004;153:350–359. [72] Asteasuain M, Bandoni A, Sarmoria C, Brandolin A. Simultaneous process and control system design for grade transition in styrene polymerization. Chem Eng Sci 2006;61:3362–3378. [73] Mitra K, Majumdar S, Raha S. Multiobjective dynamic optimization of a semi-batch epoxy polymerization process. Comput Chem Eng 2004;28:2583–2594. [74] Mitra K, Majumdar S, Raha S. Multiobjective Optimization of a Semibatch Epoxy Polymerization Process Using the Elitist Genetic Algorithm. Ind Eng Chem Res 2004;43:6055–6063. [75] Majumdar S, Mitra K, Raha S. Optimized species growth in epoxy polymerization with realcoded NSGA-II. Polymer 2005;46:11858–11869. [76] Deb K, Mitra K, Dewri R, Majumdar S. Towards a better understanding of the epoxypolymerization process using multi-objective evolutionary computation. Chemical Engineering Science 2004;59:4261–4277. [77] Nayak A, Gupta SK. Multi-objective optimization of semi-batch copolymerization reactors using adaptations of genetic algorithm. Macromolecular Theory and Simulations 2004;13:73–85. [78] Anand P, Venkateswarlu C, Rao MB. Multistage dynamic optimization of a copolymerization reactor using differential evolution. Asia-Pacific Journal of Chemical Engineering 2013;8:687– 698. [79] Furtuna R, Curteanu S, Leon F. An elitist non-dominated sorting genetic algorithm enhanced with a neural network applied to the multi-objective optimization of a polysiloxane synthesis process. Eng Appl Artif Intell 2011;24:772–785. [80] Benyahia B, Latifi MA, Fonteix C, Pla F. Multicriteria dynamic optimization of an emulsion copolymerization reactor. Comput Chem Eng 2011;35:2886–2895. [81] Benyahia B, Latifi MA, Fonteix C, Pla F. Modeling and Multiobjective Optimization of A FedBatch Emulsion Copolymerization Process to Control the Resulting Particles Core-Shell Morphology. Macromolecular Symposia 2011;302:142–150. [82] Shahhosseini S, Vakili S. Optimization of Styrene Reactor Using Tabu Search and Genetic Algorithm Methods. Int J Chem React Eng 2011;9:A64. [83] Gujarathi AM, Babu BV. Optimization of Adiabatic Styrene Reactor: A Hybrid Multiobjective Differential Evolution (H-MODE) Approach. Ind Eng Chem Res 2009;48:11115–11132.

References |

173

[84] Tarafder A, Lee B, Ray A, Rangaiah G. Multiobjective optimization of an industrial ethylene reactor using a nondominated sorting genetic algorithm. Ind Eng Chem Res 2005;44:124–141. [85] Gujarathi AM, Babu BV. Multi-objective optimization of industrial styrene reactor: Adiabatic and pseudo-isothermal operation. Chemical Engineering Science 2010;65:2009–2026. [86] Rajesh J, Gupta S, Rangaiah G, Ray A. Multi-objective optimization of industrial hydrogen plants. Chem Eng Sci 2001;56:999–1010. [87] Oh P, Ray A, Rangaiah G. Triple-objective optimization of an industrial hydrogen plant. J Chem Eng Japan 2001;34:1341–1355. [88] Oh P, Rangaiah G, Ray A. Simulation and multiobjective optimization of an industrial hydrogen plant based on refinery off-gas. Ind Eng Chem Res 2002;41:2248–2261. [89] Montazer-Rahmati MM, Binaee R. Multi-objective optimization of an industrial hydrogen plant consisting of a CO2 absorber using DGA and a methanator. Comput Chem Eng 2010;34:1813– 1821. [90] Mu SJ, Su HY, Jia T, Gu Y, Chu J. Scalable multi-objective optimization of industrial purified terephthalic acid (PTA) oxidation process. Comput Chem Eng 2004;28:2219–2231. [91] Mohanty S. Multiobjective optimization of synthesis gas production using non-dominated sorting genetic algorithm. Comput Chem Eng 2006;30:1019–1025. [92] Cheng S, Chen H, Chang H, Chang C, Chen Y. Multi-objective optimization for two catalytic membrane reactors - Methanol synthesis and hydrogen production. Chem Eng Sci 2008;63:1428–1437. [93] Kundu PK, Zhang Y, Ray AK. Multi-objective optimization of simulated countercurrent moving bed chromatographic reactor for oxidative coupling of methane. Chem Eng Sci 2009;64:4137– 4149. [94] Kundu PK, Ray AK, Elkamel A. Numerical simulation and optimisation of unconventional threesection simulated countercurrent moving bed chromatographic reactor for oxidative coupling of methane reaction. Can J Chem Eng 2012;90:1502–1513. [95] Quddus MR, Zhang Y, Ray AK. Multiobjective Optimization of a Porous Ceramic Membrane Reactor for Oxidative Coupling of Methane. Ind Eng Chem Res 2010;49:6469–6481. [96] Nabavi SR, Rangaiah GP, Niaei A, Salari D. Multiobjective Optimization of an Industrial LPG Thermal Cracker using a First Principles Model. Ind Eng Chem Res 2009;48:9523–9533. [97] Nabavi R, Rangaiah GP, Niaei A, Salari D. Design Optimization of an LPG Thermal Cracker for Multiple Objectives. Int J Chem React Eng 2011;9:A80. [98] Abo-Ghander NS, Logist F, Grace JR, Van Impe JFM, Elnashaie SSEH, Lim CJ. Optimal design of an autothermal membrane reactor coupling the dehydrogenation of ethylbenzene to styrene with the hydrogenation of nitrobenzene to aniline. Chem Eng Sci 2010;65:3113–3127.

Volkan Degirmenci and Evgeny V. Rebrov

5 Design of catalytic micro trickle bed reactors Nomenclature CP CFD dP dR DL DM h g MTBR k kL L n ΔP q RF T∞ T TBR u x

Heat capacity Computational fluid dynamics Particle diameter Reactor diameter Axial dispersion coefficient Molecular diffusion coefficient Heat transfer coefficient Gravity Micro trickle bed reactor Rate constant Mass transfer coefficient Reactor length Reaction order Pressure drop Volumetric heat generation rate Radio frequency Ambient temperature Temperature Trickle bed reactor Superficial velocity Fraction

Greek Letters εB Bed porosity εL Liquid holdup ε Ld Dynamic liquid holdup ε Ls Static liquid holdup ε L,t Total liquid holdup μ Micron η Effectiveness factor ρ Density Ω Energy dissipation factor λeff Effective conductivity of the reactor bed βL Liquid saturation σ Surface tension Subscripts L G p

Liquid Gas Particle

5.1 Introduction

|

175

Dimensionless numbers Bo Bodenstein number (uL/D L ) Eö Eötvös number ((ρ − ρ f )L2 /σ) Fr Froude number (u2 /gL) Mo Morton number (gμ4 Δρ/ρ2 σ 3 ) Nu Nusselt number (hL/λ) Pe Peclet number (uL/D L ) Pr Prandtl number (ν/α) Re Reynolds number (uLρ/μ) Sc Schimdt number (ν/D M ) We Weber number (ρu2 L/σ)

5.1 Introduction Gas-liquid-solid reactions can be performed in various types of multiphase reactors such as stirred slurry reactors, bubble columns, ejector loop reactors, fluidised bed reactors or trickle bed reactors. Slurry reactors could be considered as an alternative to fixed bed reactors for highly exothermic reactions, because the high heat capacity and possibly high thermal conductivity of liquids makes it easier to achieve uniform temperatures. Besides, the small size of the catalyst particles often makes diffusional effects negligible. However, slurry reactors have various disadvantages. Retention and separation of catalyst particles in the reactor vessel poses great operational challenges such as clogging in filters. Furthermore, solid loadings are limited in stirred slurry and ejector loop reactors; therefore, their application is limited to fast reactions. Gas and liquid reactants can be fed to a fixed bed reactor packed with solid catalyst particles either concurrently or counter currently. In concurrent feeding, the flow may be either downward or upward. Packed bed reactors with a concurrent gas-liquid downward flow are called trickle bed reactors. Counter current operation is sometimes preferable for selective removal of by-products, such as hydrogen sulphide removal in desulfurization processes. Upflow concurrent packed bed reactors are called packed bubble column reactors. Bubble columns may handle solid loadings up to 25–30 wt % and they offer excellent mass and heat transfer rates. However, there is a significant back mixing in these reactors, which results in lower selectivity in consecutive reactions. The advantage of down flow trickle bed reactors with respect to an up flow bubble column is the fact that there is no limitation on the flow rates imposed by flooding limits in TBRs. The flow rates are only limited by the available pressure head at the inlet. Moreover, the liquid is much more evenly and thinly distributed as compared to bubble flow reactors. Furthermore, for slow reactions that require high catalyst loadings, it is important to use trickle bed reactors. These reactors can also be employed where direct contact between gas and a catalyst may benefit the overall performance. Another important advantage of trickle bed reactors is their simplicity in operation at elevated temperatures and/or pressures. Overall energy consumption is often lower as

176 | 5 Design of catalytic micro trickle bed reactors

compared with other reactor types, as the catalyst is not suspended in fluid as opposed to slurry, bubble column or stirred tank reactors. Low axial dispersion can be achieved in trickle bed reactors as compared to other three-phase flow reactors, therefore higher conversions and selectivities can be reached. Trickle bed reactors are used in many chemical processes due to their advantages over slurry reactors such as easy handling of catalysts and operation at elevated pressures. Traditionally, petrochemical industries use TBRs for hydrocracking and hydrotreating; the fine chemical industries use them for hydrogenation, alkylation and oxidation [1, 2]. In a typical TBR, gas and liquid flow concurrently downward over a packed solid bed of catalyst. In some applications, such as desulfurization, a counter current operation is preferred where gas flows upward and liquid downwards. In both operation modes, the reaction takes place between the dissolved gas and the reactants in the liquid phase on the catalyst surface. Therefore, the mass and heat transfer limitations dramatically affect the reactor’s performance. Several extensive reviews [1] and books on trickle bed reactors have been published in the recent years [3]. The performance of trickle bed reactors depends on hydrodynamics, fluid phase mixing, interphase and intraparticle heat and mass transfer rates, and reaction kinetics. In turn, the heat and mass transfer rates depend on particle shape and particle size distribution, wetting of catalyst particles and the mode of operation. Conventional methods of design and optimization of trickle bed reactors often rely on empirical methods. The conventional design methods usually provide only global descriptions. Averaged properties and empirically evaluated model parameters are used in such models. These correlations may not be valid for larger scale reactors because hydrodynamics is different from that in laboratory reactors. These methods have limited applicability; engineering correlations often cannot be extended to different operational conditions. The uncertainties of these models are related to lumped descriptions of processes taking place at different spatial and temporal scales (from a molecular scale to macro scale). Therefore, CFD methods have often been employed in recent studies for proper reactor design. Transient reactor operation requires detailed knowledge of micro scale processes. Micro scale processes are associated with a single particle and its surroundings. Key processes that occur at this scale are energy and mass exchange between the phases. Analysis at this scale is performed using simulations in porous media. The micro scale modelling could provide significant insight into reactor dynamics and may enhance the applicability of macro-scale models. Trickle bed reactor configurations can be classified into four types: 1. Conventional trickle bed reactors: these are comprised of randomly packed beds of catalyst particles; 2. Micro trickle bed reactors comprising packed beds of microparticles: reactor functionality is similar to conventional trickle bed reactors; however, the particle size is several orders of magnitude smaller when compared to conventional reactors.

5.1 Introduction

|

177

These reactors have high surface-to-volume ratio; therefore, they provide substantial enhancement in mass and heat transfer as compared to conventional reactors. They are useful in controlling temperature in fast exothermic reactions, especially when the products are sensitive to high temperatures [4]. The hydrodynamics in MTBR often make the scale-up or scale-down rather straightforward; 3. Semi-structured micro trickle bed reactors: these are comprised of non-randomly packed particles or catalysts coated on structured packing or metal foams. This structured packing requires a lower pressure drop to operate. However, liquid flow distribution at the inlet is very crucial; 4. Structured micro trickle bed reactors: in these reactors, the micro channels are patterned with arrays of columns where the catalyst is coated. This uniform arrangement of pillars in the reactor channel mimics the packed bed with the additional benefit of control in uniformly distributed packing. Current and future emphasis on green routes for production of intermediates for fine chemicals and the pharmaceutical industry require an in-depth understanding of the underling physical phenomena to improve the performance of novel concepts of trickle bed reactors, including micro trickle bed reactors (MTBR) and mesh reactors, as those have many features characteristic to their size. It is therefore essential to clearly understand the fundamentals of contacting fluid phases to realize the advantages of these types of reactors for existing and emerging applications. Flow regimes in TBR can vary from a bubble flow, pulsing flow, trickling and spray flow regime. In the bubble flow regime, gas is dispersed in a continuous liquid. In a pulse flow regime, alternating gas-rich and liquid-rich segments are prevalent in the reactor. In the trickle flow regime, a continuous gas phase and discontinuous liquid film exists. In a spray flow regime, liquid drops are dispersed in a continuous gas flow. In a conventional TBR liquid, holdup is a major design parameter that defines the residence time of the reactant in the liquid phase. The flow regime and the relative gas and liquid flow rates determine the liquid holdup. We studied the hydrodynamics and the heat transfer of a MTBR [5]. The MTBR has a diameter of 26 mm with a length of 86 mm packed with 110 micron catalyst particles. The liquid holdup was in the range of 0.88–0.90. The residence time distribution showed that the gas flow rate does not affect the liquid holdup, dispersion and mean residence time. This is a unique feature of micro trickle bed reactors. Designing a reactor requires the precise knowledge of various parameters such as the reaction rate and enthalpy, hydrodynamics and heat transfer properties of the reactor. The presence of hysteresis in pressure drop, liquid holdup and wetting efficiency poses a challenge for the design of TBRs. Hysteresis is the difference in pressure drop, liquid holdup and wetting efficiency between increasing and decreasing modes of operation. Maiti et al. [6–8] reviewed the parameters controlling the hysteresis in TBRs which are start-up procedures, cycling of flow, particle size, particle properties, flow ranges, column size and inlet liquid distribution. The unique advantage of MTBRs is

178 | 5 Design of catalytic micro trickle bed reactors

that the liquid holdup does not depend on gas flow rate and a zero dynamic hold-up is observed. Therefore the mode of operation does not affect the hydrodynamics of the MTBR. In other words, the hysteresis is not observed in MTBRs. High liquid holdup ensures the uniform wetting of the catalyst; this simplifies the prediction of the mass transfer resistances to the catalyst particle. For the gaseous reactants, the transport limitations occur due to the necessity that they must dissolve in a liquid phase and reach the catalyst surface. The mass transfer limitations for the gaseous reactants affect the reactor performance. In order to decrease this mass transport limitation, cyclic operation of a TBR could be applied by alternating between different liquids at the reactor inlet. This will eventually induce a change in the wetting properties of the catalyst and provide an ease of mass transfer for the gas phase reactants. Atta et al. [9] recently reviewed the cyclic operation of trickle bed reactors. In gas limited reactions, partial wetting of the catalyst is desired; this could be achieved by alternating the liquid phase without sacrificing the reactor performance due to liquid maldistribution. In liquid limited reactions, the maximum mass transfer performance can be achieved in complete catalyst wetting. The periodic operation provides a pulse flow regime at lower flow rates, decreasing the cost of reactor operations. Another advantage of the periodic operation is the possibility to supress the side reactions by limiting the contact time of reactants/products. In MTBRs, the periodic operation could be useful – especially to prevent hot spot formation, specifically for highly exothermic reactions. The periodic variation of the catalyst wetting properties minimizes the risk of hot spot formation. Besides, the relatively slow mode of operation provides the necessary time for heat removal through conduction and convection. The MTBR designed by Chatterjee et al. [5] uses an external magnetic field and magnetic particles as the heating medium; this provides almost instantaneous heating of the catalyst. Such a reactor design could be utilized to induce periodic operations of heating and cooling cycles of the catalyst itself by controlling the external magnetic field rather than the liquid flow of species at controlled periods. This will enable the decoupling of the hydrodynamics from the adsorption/desorption kinetics of the reactants and products. The prospects of this type of reactor design are discussed in detail in Section 5.4. In this chapter, the hydrodynamics, mass and heat transfer, and periodic operation of micro trickle bed reactors have been discussed. Several examples of emerging applications of multiphase micro trickle bed reactors have been highlighted. It is hoped that this chapter will provide insight in advancing our understanding of micro trickle bed reactors.

5.2 Hydrodynamics

| 179

5.2 Hydrodynamics Different flow regimes may exist with different contacting and mixing characteristics. Each flow regime corresponds to a specific gas-liquid interaction, thus having a great influence on parameters such as liquid holdup, pressure drop and mass and heat transfer rates. The volumetric gas and liquid flow rates as well as catalyst wettability determine the boundaries of flow regimes. Therefore hydrodynamics need to be carefully addressed for proper reactor design.

5.2.1 Flow regimes Downward flow packed bed reactors, namely trickle bed reactors (TBR), allow for a variety of flow regimes. In general, four distinct flow regimes exist in TBRs: 1. Trickle flow regime: this regime exists when both gas and liquid flow rates are relatively low. In this regime, gas phase is continuous, and the liquid phase is dispersed. It is also known as a low interaction regime, since the flow in one phase does not significantly affect the flow in the other phase; 2. Pulsed flow: a higher gas flow rate results in pulsed flow, where the interaction between the phases is also higher; 3. Spray flow: for a given liquid flow rate, if the gas flow rate is relatively increased too much, spray flow will be obtained; 4. Bubble flow: in contrast to the spray flow, if the liquid throughput is relatively high in comparison to the gas flow, the liquid phase is continuous and the gas phase is dispersed; this is the bubble flow. Schematic representation of bubble flow, pulsing flow, trickling regime and spray flow are given in Fig. 5.1 [6]. The knowledge of transition between flow regimes is critical to design TBRs. A flow map for trickle bed reactors for foaming and non-foaming liquids was developed by Charpentier and Favier [10] and reported in various text books [11, 12]. Various hydrodynamic correlations are summarized in extensive reviews of twophase flow systems in packed beds [1, 13], and the role of hydrodynamics on conventional trickle bed reactors was covered in extensive reviews [14–18]. The transition from trickle to pulse flow is generally characterized by a sharp increase in the root mean square pressure fluctuation for a small increase in gas or liquid flow rate. At low gas velocities (Re G < 400), the flow regimes change consecutively from trickle flow to pulse flow and from pulse flow to bubble flow by increasing the liquid flow rate. For high gas flow rates, the transition is in the order of wavy, spray, pulse and bubble flow by increasing the liquid flow rate. Many correlations and models have been proposed in recent years to predict the regime boundaries. However, none of them has been entirely successful. Larachi et al. [19] systematically analyzed the prediction

180 | 5 Design of catalytic micro trickle bed reactors

Gas Column wall

Liquid film

Liquid slug

Catalyst particle (a) Trickling

Gas pulse

(b) Pulsing in a small column

Gas Liquid slug Trickling flow

Column wall

(c) Pulsing in a large column

Catalyst particle

Liquid droplet

Bubble

(d) Spray

(e) Bubble flow

Fig. 5.1: Schematic representations of flow regimes in trickle-bed reactors. Reprinted with permission from “Ind Eng Chem Res 2006;45(15):5185–5198” [6]. © 2006, American Chemical Society.

performances of all the transition models and correlations against all the transition data published in the literature since 1964. It was seen that the use of available phenomenological and semi-theoretical models for predicting flow transition leads to unacceptable errors. Based on all this information, Larachi et al. [19] recommended the use of a neural network correlation that turned out to be the most general correlation for predicting the trickle-to-pulse flow transition.

5.2.1.1 Micro trickle bed reactors The trickle flow regime prevails at relatively low gas and liquid flow rates. The liquid flows as a laminar film and/or in rivulets over the catalyst particles, while gas passes through the remaining void space. Trickle flow regime exists at low liquid and moderate gas flow rates where liquid flows as film over catalyst particles. In this regime, inertial forces are weaker compared to the interfacial forces, and liquid hold-up is controlled by capillary pressure. The trickle flow regime region widens with a decrease in liquid viscosity and/or surface tension. Heat and mass transfer rates are slower when compared to other possible flow regimes. Transition from a trickle to pulse flow regime occurs with an increase in either the gas or liquid flow rate. In the pulse flow regime, local flow paths for gas are blocked by liquid pockets, which results in the formation of alternate gas and liquid-enriched zones. A pulse flow regime has advantages in terms of the utilization of a catalyst bed and higher heat and mass transfer rates. However, the operating window of this

5.2 Hydrodynamics

| 181

regime is relatively small and decreases as the particle diameter increases. The concept of flow passage blockage at large liquid holdups is reported by several authors [20, 21]. This blockage generates disturbances that propagate and grow with the length of the reactor. In other studies, the appearance of a pulse flow regime is thought to be related to the instability that occurs in the liquid film due to the shear exerted by the gas phase [22, 23]. Several experimental methods were used to detect the transition line from a trickle to pulse flow regime. A change in slope of measured pressure drop versus gas or liquid flow rate indicates the transition to the pulse flow regime [24]. Another group of methods is based on the application of different on-line transducers and imaging techniques. Gas and liquid properties have significant effects on the transition boundary. Cyclohexane has a three times lower surface tension than water. As a result, a transition to a pulse flow regime occurs at lower liquid flow rates (Fig. 5.2) [25]. The influence of gas viscosity and density on the transition boundary is relatively small when compared to that of liquid. Higher liquid viscosity leads to an increase in liquid holdup and therefore earlier transition to the pulse flow regime. Faridkhou et al. [26] employed digital image analysis to characterize the two flow regimes, hysteresis, and transition thereof. The effect of gas density on the onset of flow regime transition was studied by comparing the curves for air-water and argon-

Gas mass surperficial velocity GG (kg/m2s)

1

Données de Tosun (1984) Glass beads (dp=1,9 mm and ɛ=0,344)

0,8

28% Glycerine-Air Grosser et al. (1988) Houb et al. (1993) Present model Ng (1986)

0,6

0,4

Pulse flow Trickle flow

0,2

0 0

2 4 6 8 10 12 Liquid mass surperficial velocity GL (kg/m2s)

14

Fig. 5.2: Effect of gas and liquid throughputs on trickle to pulse flow regime transition boundary. Reprinted from “Chem. Eng. Sci., volume 55, Attou A, Ferschneider G., a two-fluid hydrodynamic model for the transition between trickle and pulse flow in a cocurrent gas-liquid packed-bed reactor, 491–511” [25]. © 2000, with permission from Elsevier.

182 | 5 Design of catalytic micro trickle bed reactors

water systems. In MTBRs, the low-to-high interaction flow regime transition at a given liquid superficial velocity shifts toward higher gas superficial velocities the higher the pressures (or the gas densities) are. This makes the trickle flow operating region wider at elevated pressures or with gases with higher molecular weights. As stated by Attou et al. [25], an increase in gas density leads to a decrease in inertial forces and therefore transition occurs at higher gas velocities. Faridkhou et al. [27] developed a method to study RTD in a MTBR. They used pellets of two different sizes: 53–63 μm and 106–125 μm and co-current operation. Due to relatively high pressure gradients of 500–3000 kPa/m, they carefully densified the bed prior to the experiments. The gas and liquid phases were brought into contact by a T-junction located upstream of the packed bed consisting of a coaxial arrangement. The liquid flows through the inner tube while the gas flows within the annular area between liquid line and the outer tube. To avoid flow instabilities within the bed, it is of importance that the contact between the two phases takes place at the start of the packed bed. They used two point electrodes inserted into the reactor wall via two holes. The authors experimentally confirmed theoretically predicted maximum liquid velocity in the high porosity zone close to the wall. The solid holdup was 0.592 and 0.576 for the 53–63 μm and 106–125 μm pellet beds, respectively. The liquid holdup was in the range 0.22–0.25 for the 53–63 μm pellet bed and this value was by 10–15 % higher than that for the 106–125 μm pellet bed at the same gas and liquid velocity.

5.2.1.2 Semi-structured micro trickle bed reactors Since more than a decade, various micro and milli reactor concepts have been developed for gas-liquid-solid reactions. Many of these reactors are becoming a key technology for the industry to face market pressures and match increased environmental considerations. Due to their high mass and heat transfer performances and the intrinsically safe design inherent to microstructures, they allow isothermal operation under otherwise explosive conditions. Furthermore, the small material inventory reduces the costs optimization of catalyst and/or operating conditions that often decrease time to the market. These reactors have been applied in kinetic studies of highly exothermic fast reactions, catalyst screening and process optimization. Structured packings have the advantage that they are made to fit the dimensions of the reactor in which they are placed and thus avoid the flow maldistribution due to a channelling or bypassing of the solids. Solid foam packings represent a generation of materials combining relatively high specific surface area with low pressure drop per unit height. This is largely due to the open-celled structure with very high voidages (up to 97 vol %). The geometric surface areas of the solid foam packings increase as the voidage decreases (solids holdup increasing) because the struts making up the unit cells increase in diameter. The ppi number (pore per linear inch) of the solid foam packings is an independent parameter to describe the average cell size.

5.2 Hydrodynamics

(a) Monolith

(b) Internally finned monolith

(c) Mellapak (Sulzer)

| 183

(d) Katapak-S (Sulzer)

Fig. 5.3: Schematic representation of commonly available structured reactor packings. (a) Monolith, (b) internally finned monolith, (c) Mellapak (Sulzer), (d) Katapak-S (Sulzer).

These packing materials commonly have a two-dimensional structure that redirects the liquid and gas flow in planar directions. Several examples of structured packing materials are shown in Fig. 5.3. In monoliths (including internally finned monoliths), the liquid and gas flows in separated channels, while in Mellapak, the fluids flow down corrugated sheets of gauze stacked to form open channels between these sheets, and in Katapak some of the open channels are filled with spherical particles. In the development of these packings, the aim was to increase their relatively low surface area, while maintaining a low pressure drop (high voidage) and adequate contact between the flowing phases. Solid foams have been produced in different materials (metal, ceramics, carbon, SiC, polymers, etc.). Banhart et al. [28] outlined the methods and procedures for producing these and many other solid foams. In the particular field of chemical engineering, ceramic foams found application as heat exchangers, solar receivers, gas filters, packing columns or catalyst support. A review of this field is proposed by Twigg et al. [29]. The surface area of random or structured packings can be enlarged by depositing catalytic coatings. However, due to most of the area being internal, it is not hydrodynamically accessible, and diffusion limitations within the pores may still affect the transfer of components to the active catalyst. This mass transfer of components to the catalyst located on the solid support is essential for the operation of multiphase catalytic reactions. A clear understanding of the corresponding mass transfer resistances is vital.

5.2.1.3 Structured micro trickle bed reactors A novel MTBR was developed where arrays of posts were built in each channel imitating the packing in conventional reactors [30] (Fig. 5.4). These posts in the reactor channels provide a bed porosity of around 0.4 and enhance the overall heat and mass transfer analogous to packing in conventional reactors. In a follow up study, Wada et al. reported a multichannel microreactor in which the channels comprise uniformly distributed pillars [31]. The multi-channel microreactor was built as an integrated system with a pressure drop section at the inlet

184 | 5 Design of catalytic micro trickle bed reactors

(a)

50 μm (b)

25 μm

Fig. 5.4: Micro-structured catalyst supports prior to porous silicon formation. (a) Channel view. (b) Micron scale striations produced by the etch process. © 2002, IEEE. Reprinted with permission from “J. Microelectromechanical Syst. 2002;11(6): 709–17” [30].

(Fig. 5.5). Flow regimes were studied by a reaction between oxygen and ethyl acetate (Fig. 5.6). At low liquid to gas flow rates and all gas flow rates, an annular flow was observed. At high liquid to gas flow ratios, slug flow existed. At high gas and liquid flow rates, churn flow was observed with rapidly undulating gas-liquid interface shapes. In a follow up study, an integrated MTBR with cooling was developed [32]. In order to achieve an overall device capacity of 1 ml min−1 , the reactor consisted of 32 parallel channels with a width of 650 μm, a depth of 300 μm that were packed with 5000 posts with a diameter of 70 μm. The 8 : 1 width-to-packing diameter ratio provided uniform flow distribution. The authors observed significant differences in transition lines between different flow regimes as compared with the flow regime maps for open channels, such as the Baker coordinate map [33], Charpentier-Favier coordinate map [10] and Talmor coordinate map [34]. They also compared the data obtained from similar studies by Losey et al. [30] and Wada et al. [31]. In all cases, substantial differences in the position of regime boundaries were observed as compared to those observed in conventional TBRs. In the case of the Baker coordinate map, the annular regime was observed in microchannel systems at conditions corresponding to a dispersed regime in conventional systems. This is the direct result of substantial differences in the relative strength of capillary and gravity forces which dictate the flow regime [35]. More accurate hydrodynamic models are required for predicting flow regimes in a structured MTBR. In consecutive hydrogenation reactions, the order of selectivity to the intermediate product follows the pattern of an increased mass transfer rate [36]. Therefore, a MTBR packed with small catalyst particles could over perform monolith and other structured reactors. Krishnamurthy et al. [37] investigated a two-phase flow across a bank of staggered circular micropillars, 100 μm long with a diameter of 100 μm and a pitch-todiameter ratio of 1.5. The device was sealed from the top with a Pyrex cover which al-

5.2 Hydrodynamics

Reaction channel

| 185

Pressure drop channel

(a)

0 1.0 0.8 0.6 0.4 0.2 0

100

Pressure

200 300 6.00 4.00 2.00 Distance (mm)

Liquid Depth (μm)

Pressure (psig)

Gas

0 (b)

16 mm

40 mm

(c) Fig. 5.5: Multichannel microreactor and design of the pressure drop zone. (a) Relationship between structure and normalized pressure drop. The depth was measured from the surface of the Si wafer. The depths of manifold and reaction channels were 300 μm. The pressure drop across the shallow channels (25 μm deep) dominates the total pressure drop. (b) Fabricated structure. (c) Picture of microfabricated multichannel microreactor. Reprinted with permission from “Ind. Eng. Chem. Res. 2006;45(24):8036–42” [31]. © 2006, American Chemical Society.

lows flow visualization. Pressure measurements were performed with pressure transducers placed at the inlet, exit, and in three locations along the channel length. A two phase gas-liquid flow was obtained by passing the two phases through a micromixer, which was located upstream of the main pillar array and was fabricated as a part of the device. The mixer has two inlets, one for the water and one for the nitrogen, and a series of closely spaced 50 μm diameter circular pillars with a pitch-to-diameter ratio of 1.3 (Fig. 5.7). Four different flow patterns were observed, namely bubbly slug, gas-slug, bridge and annular flows. In a follow up study, the authors concluded that no significant deviations were observed between water and ethanol with respect to flow patterns [38]. However, the reduction of surface tension affected the flow pattern transition lines. The authors modified constant B in Eq. (5.7) to account for the effect of surface tension on pressure drop. The resulting correlation was able to predict the combined experimental data for ethanol to within ± 10 %.

186 | 5 Design of catalytic micro trickle bed reactors

100.00

(a) Churn

Slug

JL (mm/s)

10.00 (b) 1.00

0.10 (c) Annular 0.01 1

10

100

1000

JG (mm/s) Fig. 5.6: Gas-liquid flow regimes observed in the multichannel microreactor with posts: (a) slug flow (gas as dark area); (b) churn flow (the interface fluctuates rapidly and liquid periodically spans the entire channel); (c) annular flow (gas flows at the center of reactor). The conditions used in the oxidation experiments fall within the rectangle in dashed lines. Reprinted with permission from “Ind. Eng. Chem. Res. 2006;45(24):8036–42” [31]. © 2006, American Chemical Society.

750 μm Water in

2–phase flow out

6900 μm

N2 gas in

5000 μm

5000 μm Pressure ports

S1=150 μm

ST=150 μm D=100 μm Fig. 5.7: Schematic view of pillared microreactor. Reprinted with permission from “Phys. Fluids. 2007;19(4):043302” [37]. © 2007, AIP Publishing LLC.

5.2 Hydrodynamics

| 187

5.2.2 Pressure drop The calculation of pressure drop in trickle bed reactors is an important design parameter. It depends on the particle packing characteristics and the fluid properties. Many earlier correlations are based on the data obtained with non-spherical particles such as Raschig rings and cylinders. These correlations under-predict the pressure drop for spherical particles. Therefore, it is important to use a proper correlation based on flow regime, particle size and operating pressure and temperature.

5.2.2.1 Micro trickle bed reactors Most correlations for pressure drop in MTBRs are expressed in terms of Re numbers of gas and liquid phases. Kan et al. [39] proposed a correlation which is valid for trickle and pulse flow regimes at standard temperature and pressure. ( ΔP L )LG ( ΔP L )G

= 0.024d P μ L (

−1/3 3 Re We εB G G . ) ( ) 1 − εB Re L

(5.1)

ε B is the bed porosity, Re is the Reynolds number based on particle diameter, We is the Weber number. The subscripts L and G denote liquid and gas phases, respectively. Several correlations are available for a wider range of temperatures and pressures. Another class of correlations uses modified Lockhart-Martinelli number (Eq. (5.2)) instead of a single phase pressure drop. XL =

1 UL ρL = (√ ) . XG UG ρG

(5.2)

Ellman et al. [40] proposed a correlation for a wider range of operating pressures of 0.1–10 MPa for low interaction regime: ΔP 2ρ G U G2 [200(X G ξ1 )−1.2 + 85(X G ξ1 )−0.5 ] , = L dP

(5.3)

where ξ1 is defined as follows; ξ1 =

Re 2L (0.001 + Re 1.5 L )

.

(5.4)

The pressure drop along the reactor was not influenced by the gas flow rate [40]. The steady state residence time and liquid hold up were not changed whether starting from a fully liquid filled or fully gas filled reactor. This shows an important characteristic of MTBRs: the hysteresis is not applicable.

5.2.2.2 Semi-structured micro trickle bed reactors Mohammed et al. [41] studied pressure drop using three different liquid distributors, namely a single point distributor, spray nozzle and multipoint distributor in a cylindri-

188 | 5 Design of catalytic micro trickle bed reactors

cal column (polyvinyl chloride) with an internal diameter of 100 mm and a height of 163 cm. The highest pressure drop was observed for the single point liquid distributor due to a high degree of liquid maldistribution. They observed an accumulation of the liquid phase in the center region in the upper part of the packing. A more uniform initial distribution of the liquid phase was observed with the multipoint and spray nozzle distributors. The authors observed higher static liquid holdups at higher foam pore density. Moreover, especially for solid foams with high pore density (e.g. 25 and 30 ppi), the static liquid holdup increased with increasing superficial flooding velocity. The authors observed different values of the static holdup at high superficial flooding velocity in experiments with foams of high pore density. They concluded that the rapid filling of the small pores may cause forming of immobile gas pockets that cannot be replaced by the liquid phase. The number of gas pockets and their location, in turn, is of a random nature and possibly affected by local geometry-flow interactions. They also used two different pre-wetting modes, which are defined as follows: 1. “Kan-liquid” mode: adjusting the flow rates to the desired gas and liquid superficial velocity after operating the packed bed for 10 min in the high-interaction regime. This mode represents the upper hydrodynamic boundary [42]; 2. “Levec” mode: adjusting the flow rates to the desired gas and liquid superficial velocity after initial liquid flooding of the packed bed followed by draining the bed under gravity for 15 min according to Levec et al. [43] until only the residual liquid holdup remained. This mode represents the lower hydrodynamic boundary. A distinctive pressure drop multiplicity was found for the single point distributor. A higher pressure drop branch was found for the “Kan-liquid” pre-wetting mode, which can be described as rapid flooding. This mode of operation leads to an increased static liquid holdup especially in the wall region in the upper part of the column. They concluded that uniform initial liquid distribution could dampen the occurrence of hydrodynamic multiplicity. The authors developed correlations to predict the pressure drop and liquid holdup model in tubular reactors with solid foam packings. To account for the influence of different velocities and physical properties of the fluids, the liquid Reynolds number, the liquid Galileo number and gas Weber number were applied. Furthermore, the correlations include geometric properties of the foam, such as window diameter, specific surface area and porosity. dp dL

ρL g

b

c

= a1 Re L 1 Re L1 (a s d w

1 − ε d1 ) ε dp

ε L = a2 (

(5.5) c2

Re G We L b2 1 − ε d2 ) ) Ga L ( dL ) ( Re L ρL g ε

(5.6)

5.2 Hydrodynamics

| 189

Tab. 5.1: Coefficients and exponents for pressure drop correlation (Eq. (5.5)). Pre-wetting

Pore density, ppi

Flow regime

a1

Levec Levec Kan-liquid Kan-liquid

10, 20 25 10, 20 25

Trickle Pulse Trickle Pulse

21.56 0.21 0.03 0.06

b1 0.15 −0.15 0.20 0.04

c1

d1

3.45 0.53 3.02 0.46

9.80 0.37 6.86 0.26

Tab. 5.2: Coefficients and exponents for total liquid holdup correlation (Eq. (5.6)). Pre-wetting

Pore density, ppi

Flow regime

a2

b2

Levec Kan-liquid

10, 20 10, 20

Trickle Trickle

2426 170.3

0.78 0.60

c2 −0.18 −0.16

d2 0.82 0.35

The window diameter (d w ) was used as a characteristic linear dimension of the foams for the dimensionless numbers. The coefficients and exponents of the correlations are listed in Tables 5.1 and 5.2. In the co-current upflow configuration, Stemmet et al. [44–46] observed bubble and pulsing regimes in a 1 cm width reactor filled with 10 ppi solid foam packing with a solid holdup of 7 %. The liquid holdup increases with increasing liquid velocity and decreasing gas velocity, up to the maximum voidage of the solid foam packing. (Fig. 5.8 (a)). The liquid holdup increases as the ppi number of the solid foam increases due to higher capillary pressures and higher static liquid holdup. Similar behavior was observed in the co-current downflow configuration; however, the absolute value of the liquid holdup at low gas velocities was much lower and it did not exceed 0.4 (Fig. 5.8 (b)). As the liquid viscosity increases from 0.8 to 2.0 mPa/s, the liquid holdup slightly increases by 10–15 %. This increase in liquid holdup is more pronounced in the 10 ppi solid foam packing due to the higher wetting of the packing material, resulting in more effective drainage of the solid foam packing.

5.2.2.3 Structured micro trickle bed reactors Krishnamurthy et al. [37] studied the pressure drop in for a two-phase flow across a bank of staggered circular micropillars. An improved pressure drop model has been developed based on both the relevant scaling effects observed in their study and adoption from previous studies: ΔpLG B ⋅ Re d 1 = 02L = 1 + + 2 , Δp L XW XW

(5.7)

190 | 5 Design of catalytic micro trickle bed reactors

Liquid holdup, ɛL [m3L mp–3]

1 Bubble

0.8

Liquid velocity uL = 0.02 ms–1 uL = 0.04 ms–1 uL = 0.10 ms–1

Pulse

0.6

0.4

0.2

0 0

0.2

0.4

0.6

0.8

1

Gas velocity, uG [m s–1]

(a)

Liquid holdup, ɛL [m3L mp–3]

1 Bubble

0.8

Pulse

10 ppi, uL = 0.02 ms–1 10 ppi, uL = 0.04 ms–1 40 ppi, uL = 0.02 ms–1 40 ppi, uL = 0.04 ms–1

0.6

0.4

0.2

0 0 (b)

0.2

0.4

0.6

0.8

1

Gas velocity, uG [m s–1]

Fig. 5.8: (a) Liquid holdup in a 10 ppi solid foam packing in the co-current upflow configuration; (b) Liquid holdup in 10 and 40 ppi solid foam packings in the co-current downflow configuration. Reprinted from “Chem. Eng. Sci., volume 62, Stemmet CP, Meeuwse M, van der Schaaf J, Kuster BFM, Schouten JC., gas–liquid mass transfer and axial dispersion in solid foam packings, 5444– 5450” [46]. © 2007, with permission from Elsevier.

where X W is the Martinelli parameter XW =

( ΔP ΔL )L ( ΔP ΔL )G

,

(5.8)

and ( ΔP ΔL )L is the frictional pressure gradient across the channel that would result if the liquid flowed through the channel alone at a mass flow rate equal to: G ⋅ (1 − x) ⋅ A, and ( ΔP ΔL )G is the frictional pressure gradient across the channel that would result if

5.2 Hydrodynamics

| 191

the gas flowed through the channel alone at a mass flow rate of G ⋅ x ⋅ A. Here, x is the quality of the two-phase flow and A is the minimum cross-sectional flow area, which for the staggered pillar device with a transverse pitch, S T , pillar diameter, D, width, w, and height, H, is given by ST − D w⋅H. (5.9) A= ST Constant B is an empirically defined constant (Tab. 5.3). 90 % of the data fall within ± 20 % of the model when using a flow-pattern-dependent model, while 90 % of the data fall within ± 25 % of the predicted values while using all of the data. Thus the model is less sensitive to flow pattern and more sensitive to the liquid Reynolds number, as opposed to some conventional scale studies. Tab. 5.3: The values of the constant B for different flow patterns and their respective accuracy. Flow patterns

B

Mean average error, %

Bubbly/gas slug flow Bridge flow Annular flow All the flow patterns

0.0152 0.0256 0.0860 0.0358

12 14 1.7 17.8

The multiphase flow of gas and liquid through packing materials can occur in three configurations: co-current upflow, downflow and counter-current flow. A plug flow regime can be realized in the counter-current flow configuration allowing for high conversion and selectivity. However, flooding may limit the reactor productivity at high gas and liquid flow rates. The co-current upflow configuration demonstrates a large pressure drop as compared to other configurations. This may cause a large concentration gradient of gas phase reactants over the length of the reactor. The co-current downflow configuration could result in catalyst densification if rather soft catalyst pellets are used.

5.2.3 Liquid holdup The liquid volume fraction in trickle bed reactors is characterized by dynamic liquid holdup and static liquid holdup. The latter is proportional to the number of stagnant liquid pockets. The difference between these two parameters often determines the effective liquid residence time distribution in trickle bed reactors. For kinetically controlled reactions, the reaction rate are directly proportional to the extent of internal wetting of particles.

192 | 5 Design of catalytic micro trickle bed reactors

Liquid holdup can be expressed in two ways: 1. total liquid holdup (ε L ) defined as volume of liquid per reactor volume; 2. liquid saturation (β L ) defined as a volume of liquid per void volume. The liquid hold-up consists of two parts: dynamic liquid holdup (ε Ld ) and static liquid holdup (ε Ls ). The latter is the volume of liquid which remains in the bed after draining the bed. The liquid hold-up controls the liquid residence time and therefore reactant conversion. Marquez et al. [47] studied the dispersion and holdup in MTBRs. They used two reactors with lengths of 7 and 97 cm and with a diameter of 2 mm packed with particles of 106–125 μm. They observed a rather high liquid holdup between 0.65 and 0.85. Under these conditions, the particles were fully wetted, which is a beneficial effect for the catalytic applications. Besides, the gas flow rate affected neither the holdup nor the dispersion. No liquid flow was observed from the bed when both gas and liquid flow was stopped simultaneously, suggesting zero dynamic holdup. Thus the flow in MTBR is comparable with the packed beds and liquid flow only, and the operation mode does not affect the hydrodynamics in MTBRs. High liquid holdup values translate into good wetting characteristics of catalysts and are highly desired for catalytic reactor applications. These values are considerably higher as compared to those in industrial TBR [48]. A similar effect was observed by Yang et al. [49]. The use of diluent particles with a diameter of 300 μm increased liquid holdup from 0.05 to 0.20. In beds diluted with small particles, the liquid is more easily retained between the particles due to the increased liquid-solid contact and high capillary forces that lead to higher total liquid hold up. A similar observation was reported by Kulkarni et al. [50]: the dynamic hold up almost doubled and the wetting efficiency reached 90 %. Bej et al. [51] studied the effect of the diluting the catalyst bed with non-porous inert particles on the performance of MTBR. By testing various sizes of non-porous silicon carbide particles, it was determined that the 160–190 μm diameter particles packed in a relatively small MTBR with a volume of 5 ml considerably increased the liquid hold up and provided a comparable reactor performance with a larger (100 ml) TBR. MTBRs are employed in fine chemicals and pharmaceutical synthesis. Al-Herz et al. [52] studied the hydrogenation of 1-heptyne in a trickle bed reactor over a 1 wt % Pd/Al2 O3 catalyst. Various solvents were screened and 95 % selectivity was achieved in isopropanol. They observed that the liquid flow rate significantly affects the reactor performance. The gas flow rate was 10 ml min−1 and the liquid flow rate was increased gradually from 5 ml min−1 to 20 ml min−1 . The liquid hold up is very low and thus a partial wetting of the catalyst is expected. The wetting efficiencies were calculated by the correlation of Lebigue et al. [53] which is given below: f = 1 − (−1.986Fr 0.139 Mo 0.0195 ε−1.55 ), L L B

(5.10)

5.2 Hydrodynamics

| 193

where Fr L (liquid Froude number), Mo L (liquid Morton number) and ε B (bed porosity) are representative of the flow behaviour at the pellet scale, the physical properties of the fluid and the bed topology. In all liquid flow rates the catalyst wetting efficiency was above 91 % and increasing up to 97 % with the highest liquid flow rate. The total liquid holdup was calculated by correlation from Lange et al. [54]; ε L,t = 0.16 (

d R 0.33 0.14 Re L , ) dP

(5.11)

where d P and d R are the particle diameter and reactor diameter respectively. Total liquid holdup changes from 0.28 to 0.34 with the increasing liquid flow rate. Partial catalyst wetting is expected in such low liquid holdups. Higher liquid flow rate increased the hydrogenation rate and it was leveled off after 15 ml min−1 . A similar trend was observed for the 1-heptene selectivity. Improper catalyst wetting could be the culprit for lower observed reaction rates at lower flow rates, and the reaction may fall within the kinetic regime above the liquid flow rate of 15 ml min−1 .

5.2.4 Flow maldistribution and start-up effects Liquid phase maldistribution is an important factor in the design, scale-up and operation of trickle-bed reactors [55]. Large catalyst particles, uneven catalyst loading and a non-uniform liquid inlet distribution enhance channeling. For a longer reactor, a small variation in vertical orientation during installation of reactor could also lead to liquid maldistribution. Liquid maldistribution is often related to non-complete wetting of the reactor zone near the inlet. However, no quantitative information is available in the literature. For smaller particles, larger surface to volume ratio leads to better liquid spreading [56]. Therefore inert fine particle are used to improve flow distribution. Wu et al. [57] suggested the use of an inert fine particle along with catalyst particles to improve wetting in the reactor bed. Use of spherical particles with a uniform particle size can provide better control on local packing characteristics and therefore on local mixing and transport rates. Van Herk et al. [58] performed a flow image analysis; it revealed that segregation of inert and catalyst pellets leads to preferential pathways in the bed, where the zones with the smallest particles were filled with stagnant liquid that was refreshed much less often. The initial conversion level and the deactivation rate were only reproducible when the segregation was prevented by matching free-fall velocities of particles during the filling procedure of the reactor. Fishwick et al. [59] compared several different reactor types in selective hydrogenation of 2-butyne-1,4-diol into the corresponding alkene over a 0.5 wt % Pd/Al2 O3 catalyst: a stirred tank reactor; wall coated monoliths with a diameter of 5 and 10 cm; a wall coated capillary microreactor with a 2 mm diameter and a 50 mm trickle bed reac-

194 | 5 Design of catalytic micro trickle bed reactors

tor packed with 6 mm pellets diluted with 200 μm silicon carbide. The highest initial reaction rate was observed in the TBR. However, the selectivity to alkene measured at a 90 % conversion was 100 % in the monolith, capillary and stirred tank reactors; however, it was only 93 % in the TBR. The poor selectivity was due to the presence of stagnant zones in TBR broadening the residence time distribution. All reactors operated closed to a plug flow behavior under the used flow conditions [36]. Marquez et al. [60] studied different start up procedures and step changes in flow in MTBRs. As the Bond number was in the order of 10−2 to 10−3 , the up flow and down flow patterns were identical and hysteresis was not observed. However, a little hysteresis was observed in the pressure drop during step changes in liquid (and gas) flow rates. The characteristic time to reach a new steady state after start up or after a step change in gas or liquid flow rates was the same. This time depends on the pressure drop and can exceed the liquid residence time by a factor of 3.2. The volume of the gas feed section, which is the volume of the tubing and the connections after the mass flow controller to the top of the bed, should be minimized to reduce the characteristic time to reach the new steady state after a step change.

5.2.5 Axial dispersion Measurement of residence time distribution is the main method to determine axial dispersion. The axial dispersion model is used to describe all deviations from an ideal plug flow mode via a single parameter called dispersion coefficient (D L ), which is often expressed in terms of the liquid phase Peclet number (Pe = uL/D L ). The main factor contributing to the deviation from plug flow behavior is non-uniform porosity distribution which could lead to channeling and short-circuiting. Non-uniform bed porosity can be induced by a wide particle size distribution or bed assembly methods resulting in non-uniform bed densification. Mears [61] proposed a criterion to estimate the minimum reactor length required to avoid dispersion effects: L Cin 20n > ln ( ), dP Pe L Cout

(5.12)

where n is the reaction order, and C is the concentration. It can be seen that dispersion effects can be more pronounces at higher conversions or higher reaction orders, other parameters being the same. At low liquid flow rates (Re L < 4), Fu et al. [62] showed that the dispersion depends on particle diameter Bo = 0.00014(d h )−0.75 (ε)−1 ,

(5.13)

where Bo is the Bodenstein number (Bo = uL/D L ). MTBRs usually have short lengths that require accounting for axial dispersion. In conventional TBRs, the particle size has the same order of magnitude with capillary

5.2 Hydrodynamics

| 195

length. Thus, the gravity is an important factor; it should be taken into consideration during the reactor design. However, particle size in MTBRs is much smaller than the capillary length that makes the effect of gravity negligible. The analysis of the multiphase flow becomes the analysis of the perturbation from the single-phase flow. The single-phase dispersion in catalytic reactors is well understood; the reactor models are developed based on the dispersion coefficient in which the Peclet number appears in the dimensionless axial dispersion reactor model equation. The Peclet number has two limits: at very low liquid velocities, the dispersion is dominated by molecular diffusion where the Peclet number becomes uL/D M , where D M is the diffusion coefficient. The other limit is the fully developed turbulent flow, where the particle Peclet number is around unity and independent of diffusion. Marquez et al. [47, 60] studied the intermediate regime in a 2 mm inner diameter, 97 cm long MTBR packed with 106–125 μm particles. The liquid was fed into the packed bed with a needle and the dispersion due to the packing was analyzed by subtracting the spread observed in the empty reactor (Fig. 5.9). Van Herk et al. [63] studied the scaling down of TBR for hydrodesulphurisation reactions. Their reactor setup involved six parallel MTBRs with a diameter of 2 mm packed with particles of 100 μm. They studied the residence time distributions with a pulse of 10 μl of colored dye. The inlet and outlet effects were eliminated by performing A 10μL

Liquid feed

Gas feed

B C ΔP

Gas outlet P=Patm D

E

Liquid outlet

Fig. 5.9: Micropacked bed for stability and transient times studies. (A) 10 μl tracer injection loop. (B) Zoom showing the way that gas and liquid are introduced in the bed. (C) Differential pressure transmitter. (D) Gas-liquid separator. (E) Refractive-index cell. Reprinted with permission from “Ind. Eng. Chem. Res. 2010;49(3):1033–1040” [60]. © 2010, American Chemical Society.

196 | 5 Design of catalytic micro trickle bed reactors

the RTD experiments with and without the reactor. In the single–phase experiments, they found an accurate agreement with the expected residence time within 5 % error. They claimed that the origin of the error was due to small bubbles present in the space between the catalyst particles. The liquid residence time was independent of the gas and the liquid flow rates. Thus it could be concluded that gravity plays little role in the hydrodynamics of MTBRs. In the MTBR at the Reynolds numbers around 1, the flow patterns were either bubble flow or segregated flow. In the latter case, the gas flow could bypass the liquid resulting in low conversions. The Peclet numbers were above 100 [63]. The high value of particle Peclet number implies the validity of a plug flow estimation with short reactor lengths and an order of magnitudes of 10–20 particle diameters. The dispersion in a MTBR and in a liquid-only flow fixed bed reactor turned out to be similar. Maiti et al. [8] reported the effect of porosity on the hysteresis observed in TBRs. They found out that the hysteric behavior of porous particles is much different from that of nonporous particles. While a hysteric behavior can expected with porous microparticles, however the effect is too small to be observed for microparticles packed in MTBRs. Kulkarni et al. found out that the use of 6 mm porous particles rather than the nonporous glass beads have an impact on the residence time distribution in a laboratoryscale TBR with diameter of a 5 cm and 35 cm length [50]. Higher dispersion was observed with the porous particles due to the diffusion of the tracer into the catalyst pores. Moreover, the addition of 200 μm inert fine particles increased the Peclet number. The Peclet number increased as the superficial liquid velocity increased. Bodenstein numbers were reported in the range of 10–100 in a semistructured MTBR, which indicates the importance of dispersion effects [32]. The Peclet number was studied in a MTBR with a diameter of 10 mm and a 5 cm length packed with 150– 250 μm catalyst particles at temperatures of 25–65 °C and 6.1 bar [64]. The experiments were performed in step-response measurements with an inert tracer, which showed that the axial dispersion should be taken into account. Several authors reported that effect of gas flow rate on dispersion is almost negligible [50, 65, 66]. However supercritical fluids behave like a gas rather than a fluid. Jin et al. [67] studied the hydrodynamics in a MTBR with a diameter of 12.5 mm, a length of 110 cm under supercritical conditions. The reactor was filled with particles in the range of 0.67–1.32 mm in diameter. N-hexane was utilized as the supercritical fluid at pressures of 35–45 bar and temperatures between 140 and 280 °C. Nitrogen was used as the gas. It was observed that the residence time distribution changes drastically with the gas flow rate. The mean residence time and the dispersion decreased with the increasing gas flow rate under supercritical conditions.

5.3 Mass and heat transfer in micro trickle bed reactors |

197

5.3 Mass and heat transfer in micro trickle bed reactors 5.3.1 Mass transfer Micro trickle bed reactors operate with a relatively small energy input as compared to other reactors as there is no rigorous mixing mechanism present. Due to complete wetting, two types of mass transfer rate are relevant for MTBR, i.e gas-liquid and liquidsolid.

5.3.1.1 Gas-liquid mass transfer In a gas-liquid mass transfer process, the liquid side mass transfer rate is often a ratelimiting step. The gas-liquid mass transfer rate depends on particle diameter, flow rates, fluid properties and flow regime. Reactor geometry has virtually no effect on mass transfer rates. The gas-liquid mass transfer rate increases with decreases in particle size. Several correlations have been reported in literature. Some of them include pressure drop [68] kGL aGL = 0.0036 (

U L ΔP 0.35 . ) ε L,d L

(5.14)

In this correlation, pressure drop should be known or estimated using other correlations. In another approach, [69] expressed the gas-liquid mass transfer rate in terms of dimensionless numbers. They proposed an empirical correlation for low interaction regime: 1/4 kGL aGL d2k 1/4 1/5 1/5 1/2 a V d k = 2 × 10−4 (X G Re L We L Sc L ( ) ) DAL 1−ε

3.4

.

(5.15)

The overall mass transfer coefficient for gases with limited solubility is typically controlled by liquid-side film resistance. Therefore, the overall mass transfer coefficient value is close to that for the liquid-side mass transfer coefficient. The gas-liquid mass transfer coefficient (k L aGL ) depends on the interfacial area (aGL ), which for structured packings is a function of the specific geometric surface area (a S ) and the liquid holdup. The physical properties of the two phases, the lyophobicity of the solid and the solid holdup influence the value of k L aGL . The gas-liquid surface area for gauze-type packings is correlated by Rocha et al. [70] as a function of the Froude number: aGL = 1 − 1.203Fr 0.111 , (5.16) L aS where a S is the specific surface area of the packing, and Fr L is the liquid Froude number, ε2 ν2L . (5.17) Fr L = dp g Usually specific geometric surface area of the packing is in the order of 103 to 104 m2 m−3 .

198 | 5 Design of catalytic micro trickle bed reactors

In the co-current downflow configuration, the gas-liquid mass transfer coefficient increases as the liquid velocity increases. The mass transfer is constant in the trickle flow regime with increasing gas velocity, while it decreases with increasing ppi number. This can be explained by the increase in the strut size of the solid foam packing decreasing ppi number. The larger the obstruction, the higher the degree of turbulence and hence an increase in the refreshment of the liquid at the gas-liquid interface and enhanced gas-liquid mass transfer. For multiphase systems, it is difficult to define the value of characteristic length as boundary layer analogies with heat transfer are not applicable. Stemmet et al. [45] proposed the following correlation for a 10 ppi solid foam packing k L aGL ε L u L ρ L 1.16 0.5 = 3.68 − ( Sc L . ) DL μL

(5.18)

The results for the gas-liquid mass transfer coefficient in the co-current upflow configuration were correlated with a similar equation, where the influence of the gas velocity is included, similar to the correlations for packed beds of spherical particles proposed in Fukushima and Kusaka [71]: k L aGL ε L u L ρ L 0.92 0.5 = 311 ⋅ u0.44 −( Sc L . ) G DL μL

(5.19)

Jensen et al. [72] studied the hydrogenation of cyclohexene as a model reaction with well-known kinetics in a MTBR that contained 10 parallel channels with a width of 625 μm filled with catalyst pellets of 50–75 μm. Every channel was equipped with individual gas and liquid inlets and filters at the outlet to prevent the catalyst from leaking out. Splitting the inlet flow into multiple channels maintained the high surface-tovolume ratio and allowed operation at a low pressure drop at 1.7 bar and at a liquid flow rate of 0.1 ml min−1 . A conventional mass transfer analysis approach was applied including the resistances of gas absorption into the liquid and diffusion of gas from the liquid to the catalyst and gas diffusion inside the catalyst pellet. The overall mass transfer coefficient was in the range of 5–15 s−1 which is more than two orders of magnitude higher than the trickle bed reactor systems packed with particles of 4–9 mm. In the presence of hydrogen excess at a pressure of 6 bar, the reaction rate was limited by pore diffusion rather than by gas-liquid mass transfer [73]. In this example, better mass transfer rate comes with the expense of the relatively high pressure drop per unit volume of the reactor which was characterized by the energy dissipation factor (Ω = ΔP ⋅ u L /L) which is the power input per unit volume of the reactor. The MTBR requires 2–5 kW m−3 , which is an order of magnitude higher as compared to laboratory scale TBR (0.01–0.2 kW m−3 ). The external mass transfer resistance was studied in hydrogenation of o-nitroanisole to o-anisidine in a MFBR over a Pd/zeolite catalyst [74]. Different bed lengths were used to change the residence time. The authors concluded that mass transfer is much faster than intrinsic kinetics. However, in case of a fast reaction, mass transfer

5.3 Mass and heat transfer in micro trickle bed reactors |

199

limitations could have a considerable impact on reactor performance. Metaxas et al. [75] employed a laboratory TBR with a diameter of 25 mm and a length of 475 mm for catalytic hydrogenation of benzene over a Ni/Al2 O3 catalyst particles with a size of 350 μm. Mass transfer rates for hydrogen and benzene were observed on the gas and liquid sides. Gas-side mass transfer limitations were higher in the case of a dilution of catalyst with fine particles of 250 μm [76].

5.3.1.2 Liquid-solid mass transfer It is difficult to measure the mass transfer resistance (k L and a) in conventional TBR therefore mass transfer correlations for the low interaction regime involve the wetting efficiency term as a correction factor [77]. Van Houwelingen et al. [78] reported a method for the determination of the wetting efficiency and liquid-solid mass transfer resistance. The approach involves the use of hydrogenation of linear-octenes and iso-octenes, which are first order in terms of olefin concentration, i.e. liquid-limited reactions. The deviations from first-order behavior were interpreted as a combined effect of resistance to mass transfer and incomplete wetting. Authors derived an equation for the catalyst activity to fit the conversion data with the assumption that liquid-solid mass transfer and the reaction rate are linearly dependent on wetting efficiency. Losey et al. [72] proposed a mass transfer film model to compare the performance of microreactors and conventional multiphase reactors in hydrogenation reactions (Eq. (5.20)): C H2 R= 1 , (5.20) 1 1 k L a i + k c a s + ηk where, R is the overall reaction rate, CH2 is the saturated solubility of hydrogen in the liquid, k L a i is the external mass transfer resistance for transport of gas due to absorption to liquid, k c a s is the external mass transfer resistance for transport due to the diffusion of dissolved gas that forms bulk liquid on the surface of the catalyst, and ηk is the internal mass transfer resistance for diffusion of the reactant inside of the porous catalyst, which is given by the effectiveness factor η. They obtained the value of overall mass transfer coefficient k L a = 5–15 s−1 for multiphase microreactors which are much larger than laboratory batch reactors for multiphase systems where k L a = 0.01–0.08 s−1 . Durante et al. [79] employed a MTBR for sugar hydrogenation, namely L-arabinose to arabitol. The gas-liquid and liquid-solid mass transfer resistances were evaluated. The numerical analysis showed a good agreement with the experimental values and theoretical predictions. It turned out that the gas-liquid mass transfer resistance is more significant than the liquid-solid mass transfer resistance in their MTBR. In the semi-structured trickle bed reactor comprised of metal foam loadings, the liquid side mass transfer coefficient (k L ) is mainly determined by the slip velocity between the gas and liquid phases in the co-current upflow configuration. The value

200 | 5 Design of catalytic micro trickle bed reactors

of k L remains rather constant over the entire range of gas velocities for most of the solid foam packing. As the ppi number of the solid foam increases, the value of k L decreases due to an increased number of small restrictions to the flow that decrease the local velocity of the liquid. This reveals a lower turbulence and hence a lower k L value. It may also be argued that the lower liquid side mass transfer coefficient at high ppi numbers is a result of thinner walls, incapable of inducing large eddies in the liquid phases, leading to a reduced level of turbulence in the flowing liquid. Characteristic values for k L were reported to be in the range between 450 and 500 s−1 for a 10 ppi solid foam and between 50 and 150 s−1 for a 40 ppi solid foam for liquid flow rate between 0.02 and 0.04 m s−1 [44]. Mohammed et al. [80] proposed a correlation that considers the effect of the gas and liquid superficial velocities, the physical properties of the fluids and characteristic geometric features of the foam material on the liquid-slid mass transfer coefficient in co-current flow configuration: 0Sh 1 − ε d3 b3 c3 = a Re Re d ) . (a 3 s w L G ε Sc 0.33

(5.21)

The values of the parameters are given in Tab. 5.4. Tab. 5.4: Coefficients and exponents for the liquid–solid mass transfer correlation (Eq. (5.21)). Pre-wetting

Pore density, ppi

Flow regime

a3

b3

Levec Levec

10, 20 25

Trickle Pulse

0.13 1.58

0.805 0.54

c3

d3

−0.89 −0.09

−1.34 0.06

Eq. (5.21) shows that higher liquid superficial velocity favors the liquid–solid mass transfer, while increasing foam pore density lowers the transfer rate. In the co-current down flow configuration, the liquid-side mass transfer coefficient depends on both the liquid velocity and the ppi number of the solid foam packings, but it does not depend on the gas velocity [45]. The value of k L is four times higher for the 10 ppi solid foam packing than for 40 ppi. This higher rate of mass transfer is a result of a higher local mixing within the film flowing over the 10 ppi solid foam packing due to the larger strut thickness. Characteristic values for k L were reported to be in the range between 10 and 30 s−1 for a 10 ppi solid foam. The value of k L for the co-current downflow configuration is an order of magnitude lower than that for the co-current upflow configuration. Volumetric mass transfer coefficients, k L aGL , are not a function of the ppi number of the solid foam packing, but increase with increasing gas and liquid velocities. Experimentally found values of the volumetric mass transfer coefficients were in the range of 0.1 to 1.3 s−1 [44]. These results indicated a high potential for application of structured packings in MTBR as the obtained values were one order of magnitude higher than in conventional packed bed reactors.

5.3 Mass and heat transfer in micro trickle bed reactors |

201

Tab. 5.5: Performances of various microreactors in gas-liquid-solid reactions.

aGL (m2 m−3 liquid ) m−3 liquid )

aLS

(m2

kL

(s−1 at 298 K)

ΔP (kPa m−1 ) L εL

(m−3 m−3 reactor )

Mesh reactor [82]

Micro trickle bed reactor [27, 72]

Monolith reactor [83, 84]

2500

—

—

6500

1.5–3 × 104

1500–2500

1–2

2–6

0.03–0.12

—

500–3000

< 10

< 0.02

0.15–0.27

0.25–0.75

Tourvielille et al. [81] compared the performance of different gas-liquid-solid microreactors (Tab. 5.5). Wall coated monolith present lowest mass transfer coefficient and low pressure drop. Packed microreactor provides good mass transfer rates; however, it is accompanied with high pressure drop as a trade-off.

5.3.2 Heat transfer Many reactions performed in MTBRs are often highly exothermic, and heat released in these reactions is transported by conduction and convection. This may lead to nonuniform temperature distribution in the reactor bed. Local hot spots may form in the micro packed bed due to poor fluid distribution. It could also deactivate the catalyst and reduce conversion and selectivity [85]. Besides, solid-liquid heat transfer limitations affect the kinetics of reactions. Heat management inside the catalyst bed without causing catalyst deactivation and/or by-product formation is a critical task in the design of a MTBR. A near isothermal temperature distribution can be achieved by structuring catalyst and inert material zones, and removing heat via cooling. In some cases, the gas or liquid flows can be recycled to reduce conversion per pass and control temperature. Correct calculation of heat transfer rates is crucial for design and scale-up of MTBR. The particle size in MTBR is usually very small, therefore no intraparticle temperature gradient inside the catalyst particle is observed. Few studies in literature deal with particle-liquid heat transfer rates in trickle-bed reactors. The main reason is probably the difficulty to find accurate experimental methods. Heat transfer rates increase with both increasing gas and liquid flow rate. In the trickle flow regime, the effect of the gas flow rate is rather weak, while the transition to pulsing flow results in an increase in the local average heat transfer rates [86]. Boelhouwer et al. [87] proposed a correlation for the Nu number Nu = 0.4 + 0.1Re 0.8 Pr 0.33 ,

(5.22)

202 | 5 Design of catalytic micro trickle bed reactors

where the Re number is based on the linear liquid velocity and the particle diameter. Ruether et al. [88] presented a correlation for particle-liquid mass transfer in which a power of 0.77 for the Re number was found. Heidari et al. [89] have developed CFD models to describe micro and meso scale heat transfer in TBRs at the gas-liquid interphase. Their micro scale model was a modified version of a double-slit [90] model implemented to determine the interfacial heat transfer. The meso-scale model was solved numerically to determine the effect of particle shape on an interfacial heat transfer. They calculated the interfacial Nusselt number as a function of pressure drop, liquid holdup and wetting efficiency. The Nusselt number increased with an increase in gas phase Reynolds and Prandtl numbers and decreased as the liquid phase Prandtl, Reynolds and Eötvös numbers increased. Fedorov et al. [91] investigated heat transfer in a heat sink with rectangular micro channels by developing a numerical model for fluid flow and the heat conduction in a silicon substrate. They observed rather complex heat flux patterns because of a strong coupling between convection in the fluid and conduction in the substrate. They concluded that axial heat conduction in micro packed beds should be taken into account. Lee et al. [92] experimentally studied the heat transfer in rectangular micro channels with a width from 194 to 534 μm and the depth of five times the width in each case. In case of a 1D heat transfer model with constant temperature or constant heat flux boundary conditions, the deviations from the 3D full conjugate analysis were 12.4 % and 7.1 %, respectively, demonstrating the importance of the use of numerical simulations, instead of 1D correlation. Norton et al. [93] reported that between 60 to 80 % of the heat generated in microstructured reactors is lost to the surroundings regardless of the insulation. These losses could only be avoided by application of special insulation techniques, such as vacuum insulation. Therefore microreactors, even with thick insulating packaging, cannot be operated adiabatically; heat losses to the environment should be taken into account. A pseudo-homogeneous two parameter model with an effective thermal conductivity (λeff ) and heat transfer coefficient at the wall (h w ) could satisfactory describe heat transfer in a MTBR [94]. Dudas et al. [95] studied a highly exothermic reaction of hydrogenation of thymol on a Ni/Al2 O3 catalyst (1.2 mm) in a 40 mm internal diameter MTBR. The temperature profile was controlled with a pre-heater. A hot zone was observed in the reactor. The position of the hot zone moved along the bed during start up due to different catalyst wetting at low liquid flow rates. However, at a high liquid flow rate of 2.85 kg h−1 , a considerable temperature gradient was observed along the reactor lengths. Due to the hot zone in the beginning of the reactor bed, thermal decomposition of reaction products occurred with formation of several by-products. To reduce temperature gradients, we proposed to use radio frequency (RF) heating instead of a preheater [5]. The reactor configuration was composed of alternating catalyst and heating zones. The heat was generated inside the reactor by magnetic particles placed in an external RF field. The heating zones contained nickel ferrite

5.3 Mass and heat transfer in micro trickle bed reactors

|

203

microparticles with a diameter of 110 μm. The catalytic zones were packed with catalyst particles of the same size. The temperature profile was studied along the reactor length after fast heating of a part containing magnetic particles (Fig. 5.10). The effective thermal conductivity of the bed (λeff ) and convective heat transfer coefficient for the heat losses to the environment (hext ) were calculated from a 1D transient heat transfer model (Eq. (5.23)). ∂T ∂T ∂ , (λeff ) + q + hext a(T∞ − T) = ρC P ∂z ∂z ∂t

(5.23)

where a is the external wall area per unit of the reactor volume, C p is the heat capacity and q is the volumetric heat production rate. The conduction and convective losses contributed 30 and 70 % of the total energy losses, respectively. The number and relative positions of heating zones within the reactor were optimized to obtain a near-isothermal temperature profile (Fig. 5.11). The position of heating zones depends on the heat losses to the environment and heat generation rate. Three heating zones, positioned at specific locations inside the bed, provided a temperature nonuniformity of 2 °C over a length of 50 mm.

Temperature (°C)

50

Temperature at interface (°C)

Time, s 0 10 30 60

60

54.0 53.6 53.2 52.8 52.4 52.0 10

12 14 Time (s)

40

30

20 30

35

40

50 45 Axial distance (mm)

55

60

65

Fig. 5.10: Temperature of reactor bed along the axial direction as a function of time (symbols). Lines are the guide for an eye. The interface between the heated and non-heated zones is located at x = 40 mm. Reprinted from “Chem. Eng. J., volume 243, Chatterjee S, Degirmenci V, Aiouache F, Rebrov EV. Design of a radio frequency heated isothermal micro-trickle bed reactor, 225–33” [5]. © 2014, with permission from Elsevier.

204 | 5 Design of catalytic micro trickle bed reactors

80 ±2°C c

Temperature (°C)

60 b a 40

20

0 0

15

30 45 Axial distance (mm)

60

75

Fig. 5.11: Steady state fluid temperature along the axial direction of the micro trickle bed reactor for (a) single zone, (b) two zone, and (c) three zone configuration. Reprinted from “Chem. Eng. J., volume 243, Chatterjee S, Degirmenci V, Aiouache F, Rebrov E V. Design of a radio frequency heated isothermal micro-trickle bed reactor, 225–233” [5]. © 2014, with permission from Elsevier.

5.3.3 Scale up For large-scale processing, multiple MTBR may be connected in series or in parallel. For a number of fine chemical syntheses that require large liquid residence time, TBR are operated in a semi-batch mode when the liquid flow has complete recycle with a recycle to feed ratio of above 50. Hickman et al. [96] scaled up a 14 inch TBR packed with 200–400 μm particles from the laboratory scale. They observed a substantial loss in the efficiency due to inefficient catalyst wetting. By adjusting the superficial liquid velocities for efficient catalyst wetting and after a thorough investigation on the catalyst deactivation, the reactor was scaled up by a factor of 3 × 106 . The hydrodynamics of the TBRs are very sensitive to the scale of the reactor, which brings about the challenge of transposing laboratory data into a commercial-sized reactor. The scale up methods relying on 1D models are not reliable because they use pressure drop and liquid holdup; this changes within the reactor in a complex fashion with the change in a reactor’s scale. Large bed diameters lead to liquid maldistribution. Ranade et al. [97] proposed a general scale up approach (Fig. 5.12) based on CFD models that account for the porosity and thus consider liquid maldistribution and local hot spot formation in hydrodesulfurization and hydrodearomatization reactions.

5.3 Mass and heat transfer in micro trickle bed reactors

|

205

Scale-up and scale-down issues in TBR Evaluation of kinetic parameters in absence of hydrodynamic effects Step–I Process development on laboratory scale (Process optimization)

Mathematical modeling and pilot plan experiments

Model accounts for reactor hardware + hydrodynamic parameters

Mathematical modeling and pilot plan experiments

Scale down

Step–II

Scale up

Laboratory scale reactors • Low D/dp • High porosity variation • Partial wetting • Wall flow

Both are equally important and complicated

Commercial scale reactors • Large diameters • Liquid maldistribution • Poor heat transfer and local hotspots

Step–III

Step–III CFD plays a vital role in Step-II

Fig. 5.12: Schematic of Steps in Scale-up and Scale-down. Reprinted from “Chem. Eng. Sci., volume 62, Gunjal PR, Ranade V V. Modeling of laboratory and commercial scale hydro-processing reactors using CFD, 5512–5526” [97]. © 2007, with permission from Elsevier.

Simulations were then carried out at temperatures between 320 and 380 °C, pressures of 200–800 bar, and initial H2 S concentration between 1.0 and 2.5 vol %. In MTBRs, superficial gas-liquid velocities were much lower; this reduced the overall performance and mass transfer rates due to increased dispersion. The predicted results were compared with the experimental data. While the results were not conclusive, it was shown that slight alteration in hydrodynamic parameters led to very different predictions in the reactor performance. Simulation showed that the sensitivity of conversion at high temperature and pressure is rather low; however, it becomes very sensitive to any alteration in hydrodynamic parameters at low temperatures and high liquid flow rates. Therefore, it is crucial to develop correlations under realistic operating conditions and reactor sizes that adequately represent hydrodynamic and multi-scale transport processes in order to achieve predictive models for scale up.

206 | 5 Design of catalytic micro trickle bed reactors

5.4 Periodic operation Trickle bed reactors are widely used in industrial applications [98]. A great economic potential still exists for improving their performance either by process intensification in MTBRs or by novel operational methods [99]. The latter includes, among others, the cyclic operation of trickle bed reactors [100]. This approach has been presented in several reviews [9]. In the cyclic operation, the inlet flow is forced to contact with another liquid stream periodically switched between a low level (base) and high level flow rate. In the extreme case, the minimum flow rate could be set at zero; this is known as on-off cycling. As a result, a variation of catalyst wetting is achieved and the gas transport to the catalyst surface is improved due to partial wetting of catalyst particles. This results in higher reaction rates in gas phase limited reactions. For the liquid phase limited reactions, the pulse flow results in higher reactor performance due to a higher catalyst wetting at the maximum flow rate. Despite these advantages, periodic operation is still not a viable alternative to conventional steady-state TBR operation in industrial applications due to a considerably higher investment and operational costs. Predicting the performance of a periodically operated TBR by theoretical models remains a challenging task. Lange et al. [101] observed higher time-average conversions as compared to steady-state operation in catalytic hydrogenation of alphamethylstyrene. They developed a reactor model for the unsteady state operation of TBRs. In this model, they used conventional correlations for liquid holdup, wetting efficiency and axial dispersion to determine mass transfer coefficient. However, their model failed to describe all of the main transient behavior observed. Improved correlations for the liquid hold up and mass transfer rate, determined under actual conditions, are required to improve the model. The complexity of modeling of TBR hydrodynamics arouses from the presence of multiple hydrodynamic states showing different pressure gradients and liquid hold ups for identical gas and liquid flow rates. This poses a particular challenge for a successful modeling of reactor performance. Borren et al. [102] studied the influence of periodic operation on TBR hydrodynamics and the gas-liquid mass transfer. They used a laboratory scale TBR with a diameter of 40 mm packed with a γ-Al2 O3 catalyst with pellets of 2.57 or 3.14 mm. Saturated aqueous solutions were used as the fluid medium and volumetric gas-liquid transfer coefficient (k L a) was calculated from the oxygen mass balance in the liquid phase. They found out that the operating mode influenced the two-phase pressure drop and gas-liquid mass transfer coefficient, but it has little effect on the dynamic liquid hold up. Simulations also showed that a significant increase in conversion could be achieved by the increase of gas-liquid mass transfer resistance for the gas-limited reactions. However, fluid modulation alters the liquid-solid mass transfer too, which should be carefully weighed for the overall performance of a TBR. An advantage of periodic operation of TBRs is fully utilized when the reaction is either highly exothermic or endothermic. The fast removal of products with an in-

5.5 Applications | 207

ert stream of fluid facilitates the heat removal and prevents the further reactions of the products. Liu et al. [103] studied the hydrogenation of 2-ethylanthraquinoes on a Pd/Al2 O3 catalyst in a periodically operated TBR where the liquid feed was controlled in an on-off mode. The performance of the reactor was analyzed with an operating period of 40–480 s at 300 kPa. The conversion increased by 8 and 4 % at 333 and 343 K respectively as compared to steady-state operation. In a follow up study [104], the effect of the cycle period, pressure, temperature and time-average flow rate on the performance was investigated and compared with the steady-state operation. Their results showed that under optimal operating conditions, the conversion and the selectivity were improved by 21 and 12 % respectively. Periodic temperature oscillations could result in desired surface coverages, at least for a part of the cycle and the performance could be improved in terms of selectivity and catalyst stability. Habtu et al. [105] investigated the effect of constant and modulated feed temperature on the MTBR performance in phenol oxidation over an activated carbon with a particle size of 500 μm. The study was conducted at Re L in the range from 0.15 to 1.0 and Re G from 0.35 to 4.5 in a 25 cm long reactor with a diameter of 9.3 mm. An unsteady state pseudo-homogeneous heat transfer model was used in the absence of chemical reaction. The overall heat transfer coefficient increased from 1.25 to 2.5 W m−2 K−1 with an increase in pressure and/or liquid flow rate. The heat transfer is particularly sensitive to the value of the dynamic liquid hold up. The gas flow rate only marginally affected the heat transfer. Other recent parametric studies for the understanding of temperature on the hydrodynamics during periodic operation by flow modulation were performed by Aydin et al. [106, 107]. Under MTBR operating conditions, at high liquid hold up and low Reynolds numbers, the periodic modulation of inlet liquid flow rates becomes negligible in terms of hydrodynamics. Massa et al. [108] studied phenol oxidation over a CuO/Al2 O3 catalyst in a periodically operated TBR in an on-off liquid flow mode. During the dry cycle, external mass transport coefficients were increased, resulting in higher oxygen concentrations at the catalyst surface which in turn favors total oxidation. While the phenol conversion remained virtually the same under cyclic operation, but the product distribution changed towards total oxidation products.

5.5 Applications Hotz et al. [109] developed a disk-shaped packed bed microreactor containing a Rh/ceria/zirconia catalyst for butane-to-syngas processing at a temperature of 550 °C (Fig. 5.13). This microreactor was a part a butane fuel processor that can be integrated into a micro solid oxide fuel cell (SOFC) system. A small reactor volume, a highly compact design and a low pressure drop are crucial requirements for the integration of a fuel processor into an entire micro SOFC system. The disk-shaped packed bed

208 | 5 Design of catalytic micro trickle bed reactors

Syngas C4H10

Quartz glass wool Quartz glass tubes

Air

Syngas Packed bed

Fig. 5.13: Schematic of the disk-shaped packed bed microreactor. Reprinted from “Chem. Eng. Sci., volume 63, Hotz N, Osterwalder N, Stark WJ, Bieri NR, Poulikakos D. Disk-shaped packed bed micro-reactor for butane-to-syngas processing, 5193–5201” [109]. © 2008, with permission from Elsevier.

reactor demonstrated a 6.5 times lower pressure drop compared to an equivalent tubular packed bed reactor at similar operating conditions.

5.5.1 Micro trickle bed reactors Conversion of biomass to fuels and chemicals has attracted great attention as one of the future technologies for a sustainable the low carbon society [110]. Hydrogenation of biomass-derived molecules into sugar alcohols is an important step for the production of value added chemicals and intermediates for the chemical industry [111]. In the future, trickle bed reactors might be a choice for various reactions. Jeon et al. [112] described the hydrogenation of C9 -aldehyde into C9 -alcohol over a supported Ni-MgO catalyst. Their reactor has a diameter of 0.5 inch and a length of 80 cm. A selectivity of 90 % to C9 -alcohol was observed at full conversion at 130 °C and 400 psi. Kilpio et al. [113] investigated the hydrogenation of glucose into sorbitol in a MTBR (Scheme 5.1). OH

OH

OH

O H2

HO

CH3 OH

OH

Ru/C, 90–130 °C

OH

HO

CH3 OH

OH

Scheme 5.1: Hydrogenation of glucose to sorbitol.

Sorbitol is an alternative sweetener and a platform chemical for a wide variety of compounds. Their reactor has a diameter of 10 mm and a length of 70 mm packed with a commercial Ru/C catalyst pellets of 100–250 μm or 330–500 μm in diameter. 90 % selectivity to sorbitol was obtained in the 90–130 °C range. The catalyst deactivation was clearly pronounced, especially at the entrance region where the highest reactant concentration was observed. They combined reaction kinetics and deactivation kinetics in a single model that agreed well with the observed reaction behavior. The model is based on an axial dispersion model using temperature-dependent kinetics and deac-

5.5 Applications |

209

tivation. Their simulations predicted no temperature gradient rise inside the particles. The effectiveness factor was around 70 % due to internal diffusion limitations. Xi et al. [114] used MTBRs with a diameter of 13 mm a length of 45 cm for a ruthenium-catalyzed hydrogenolysis of lactic acid (2-hydroxypropanoic acid) to propylene glycol (1,2-propanediol), which is an important reaction step for the biomass utilization (Scheme 5.2). Lactic acid can be produced from renewable sources such as carbohydrates derived from agricultural crops and waste biomass. Lactic acid is very reactive due to its hydroxyl and carboxyl functional groups and it can be converted into various chemicals through polymerization, esterification, dehydration and oxidation [115]. O H3C

OH OH

Lactic acid

+2H2 –H2O

H3C

OH OH

Propylene glycol

Scheme 5.2: Hydrogenolysis of lactic acid into propylene glycol.

This reaction is usually performed in a batch reactor, because mass transfer limitations could be eliminated via vigorous agitation. The MTBR was packed with catalyst particles and inert glass beads of the size 200 μm and operated at a liquid flow rate of 100 ml h−1 . The translation of a batch protocol to a MTBR demonstrated similar values for the reaction rate of 4.9 × 10−4 kmol m−3 cat s−1 when the both reactors operated in the intrinsic kinetic regime. The MTBR approaches fully wetted behavior and essentially acts as a differential reactor. A detailed model was developed for MTBR that accounts for interphase mass transfer, temperature gradients and partial wetting of catalyst particles. However, the model predictions must be handled cautiously, since predicting fractional wetting and mass transport coefficients did not always match with the experimental observations. Son et al. [116] employed a MTBR with a diameter of 16 mm and 200 mm length for the transesterification of sunflower oil with methanol to produce biodiesel over a CaO catalyst with the size of 1–2 mm. The reactor allowed for a separation of the products and reactant (methanol) when it was operated in a semi batch mode where the condensed products were collected in a reservoir attached to the bottom of the reactor and the methanol was collected in a condenser at the exit. A 98 % yield of fatty acid methyl esters was achieved at 373 K with oil and methanol flow rates of 3.8 and 4.1 ml h−1 . Glycerol is the side product of biodiesel production. The growing production of biodiesel as a renewable source-based fuel leads to an increased amount of glycerol. The utilization of glycerol strongly affects the process economy. Brander et al. [117]

210 | 5 Design of catalytic micro trickle bed reactors

employed a commercial MTBR (Vinci Technologies) to investigate various possible glycerol reaction pathways such as oxidation over a Pt-Bi catalyst, aqueous phase reforming and hydrogenolysis over Ru/C and Ru/TiO2 catalysts. Glycerol is a highly functionalized molecule and thus a large number of value added chemicals can be obtained from glycerol by a variety of chemical reactions (Scheme 5.3). Xi et al. [118] reported the hydrogenolysis of glycerol where a MTBR was used with a diameter of 12.5 mm and 61 cm length in a similar study. A one dimensional, non-isothermal reactor model was developed based on the kinetic model, intra particle mass transport; liquid-solid and gas-liquid mass transfer coefficients were calculated using correlations for conventional reactors. They applied the reactor model to a large set of steady state trickle bed reactor experiments at a variety of operating conditions to predict glycerol conversion. The three kinetic constants were adjusted and all other constants and coef?cients governing hydrodynamics and mass/heat transfer in the model are taken straight from literature or textbook correlations. As a result, good agreement was observed between the predicted and experimental results. O HO

OH

Oxidation HO

HO

OH

Hydrogenolysis HO OH

CH3

Amination

Glycerol

H2N HO Etherification OH HO

O OH

CH3 O CH3 H3C

Scheme 5.3: Transformations of glycerol into high value chemicals.

The conventional batch synthesis of 6-hydroxybuspirone, an important pharmaceutical product, suffers from low process safety, difficulties in upscaling of cryogenic equipment and long reaction times. Therefore a continuous process is highly desired. LaPorte et al. [119] successfully scaled up the process of formation of 6-hydroxybuspirone from buspirone, which includes enolization, oxidation and quench steps in a TBR (Scheme 5.4). At first, the authors used a two stage CPC CYTOS microreactor that provided a 85–92 % conversion with a residence time of 5 min corresponding to a 300 g of product per day. Then a pilot-plant TBR with four columns was constructed. A yield of 83 % (41.4 kg) was obtained in steady state operation for 72 h. This time was limited by the amount of the available feed solution and potentially can be extended.

5.5 Applications | 211

1. 3.5 eq. (EtO)3P THF, –10 to –30 °C 2. 1.0 eq. NaHMDS 3. Oxygen 4. 3 eq. 2.5 N HCl, pH=2.0

O N N

N

O N N

N N

5. Heat at 60 °C (24-48 h) 6. Neutralize with 3.5 N NaOH/14 wt % brine (pH = 6.8) 7. 25 % brine wash 8. Solvent swap to IPA and crystallization

O Buspirone

N

N

N O

HO

6-hydroxybuspirone

Scheme 5.4: Preparation of 6-hydroxybuspirone. Adapted from “Org. Process. Res. Dev. 2008;12(5): 956–966” [119]. © 2008, American Chemical Society.

Adipic acid is a commercially important intermediate for the production of nylon-6,6. The commercial processes include a two-step oxidation of cyclohexane that results in large amounts of N2 O, a major air pollutant and greenhouse gas. The synthesis based on oxidation with hydrogen peroxide is considered a greener alternative. An adipic acid yield of 50 % in the direct oxidation of cyclohexene by hydrogen peroxide was reported by Shang et al. [120] at 100 °C and a residence time of 20 min in a MTBR with a diameter of 10 mm and a length of 20 cm packed with 212–300 μm inert glass beads. OH Na2WO4

+ 4 H2O2

O

O [CH3(C8H17)3N]HSO4 Adipic acid

OH

Scheme 5.5: One-step oxidative cleavage of cyclohexene to synthesize adipic acid.

Selective hydrogenation of α,β-unsaturated aldehydes, such as citral, leads to a wide range of fine chemical intermediates. A simplified reaction scheme of successive citral hydrogenation is shown in Scheme 5.6. Wörz et al. [121] replaced a batch operation into a continuous process. They employed 1 wt % Pd/SiO2 and 5 wt % Pd/Al2 O3 catalysts covered with ionic liquid layer ([BMIM][N(CN)2 ]), 1-butyl-3-methylimidazolium dicyanamide in a MTBR with a diameter of 16 mm packed with catalyst pellets of 300– CH3

CH3 O

CH3

CH3

O H2

H2

H2 OH

H3C

CH3

Citral

H3C

CH3

Citronellal

H 3C

CH3

Citronellol

Scheme 5.6: Consecutive reactions in citral hydrogenation.

OH H3C

CH3

3,7-dimethyloctanol

212 | 5 Design of catalytic micro trickle bed reactors

350 μm [121]. The use of ionic liquids enhanced the reaction selectivity towards citronellal to 100 % at 70 °C at 52 % conversion with a liquid flow rate of 1 ml min−1 . Maki-Arvela et al. [73] employed MTBRs for the parallel screening of catalysts. They used six parallel reactors, each of which had the internal diameter of 10 mm packed with two different catalyst particle sizes of 63–90 μm and 150–200 μm. The model reaction used to observe the performance of the reactors was citral hydrogenation. They obtained excellent reproducibility.

5.5.2 Semi-structured and structured trickle bed reactors A semi-structured MTBR based on monolithic structures [122, 123], filled with catalyst particles of 0.8 mm in diameter, was used in hydrogenation of alpha-methylstyrene over a Pd catalyst [124]. The monolith segments had a square cross section of 10 × 10 mm2 , a length of 5.0 cm and contained 64 parallel flow channels with a hydraulic diameter of 1 mm. The authors compared the semi-structured MTBR, a wall coated monolith and a larger scale TBR. The monolithic reactor and semi-structured MTBR demonstrated reaction rates that were at least two times higher than those achieved in the conventional TBR. Slug flow was the dominant flow regime observed at the liquid and gas superficial velocities in the range of 0.02 to 0.2 m s−1 and 0.04 to 0.2 m s−1 respectively. Van Herk et al. [58] studied the hydrogenation of biphenyl over a Pt-Pd/Al2 O3 catalyst in a MTBR with a diameter of 2.2 mm and a length of 475 mm. They have used microfabricated beds where pillars in 100–150 μm diameter were uniformly distributed. Phosphite, amine, and olefin oxidation with ozone were studied in a structured MTB and up to 100 % increase in selectivity was observed. [30, 31] A structured MTBR was used for the production of singlet oxygen from chlorine gas and hydrogen peroxide at different pressures, flow rates and chlorine mole fraction [32]. A singlet-delta oxygen yield of 78 % was obtained at the optimal reactor performance, occurring at a gas flow rate of 75 ml min−1 and an outlet pressure of 0.13 bar. A complete bottom-to-top modeling, design, construction and flow testing of a fully MTBR system were developed.

5.6 Outlook Two phase micro and meso scale packed bed reactors provide significant advantages of performing industrial relevant reactions [125] due to high catalyst wetting, absence of hysteresis effects and avoiding hot spot formation by the fast removal of heat. High liquid hold ups are observed and gas flow rates hardly effect either the dispersion or the mean flow rate in MTBRs. This feature simplifies the reactor models as the hydrodynamics approach the plug flow behavior of single phase packed bed reactors. The

References |

213

catalyst particle sizes are typically below 250 μm and the reactor diameters are below 30 mm. In recent years, much attention has been placed on the use of MTBRs as platforms for testing the commercial catalysts in the laboratory scale and developing methods for obtaining reliable data for conventional size industrial reactors especially in the field of hydrodesulphurization. MTBRs possess great potential in the pharmaceutical and fine chemicals industry as they are offering greener processes. In future sustainable chemical industries, TBRs will play an increasingly important role for the conversion of renewable feedstocks. The literature is limited, and more understanding is necessary for scale up and process intensification of MTBRs combined with a holistic approach to the process industry [126]. Reliable correlations for the parameter estimation are essential to develop predictive models for scale up. The bottleneck for commercial application is the energy requirement per reactor volume for high gasliquid superficial velocities to keep high throughputs. Scaling up strategies need to be developed either in axial dimensions or by multiplying the reactors in parallel; this could provide high throughputs required by industrial-scale production.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]

[13]

Duduković MP, Larachi F, Mills PL. Multiphase catalytic reactors: a perspective on current knowledge and future trends. Catal Rev 2002;44(1):123–246. Mills PL, Chaudhari R V. Multiphase catalytic reactor engineering and design for pharmaceuticals and fine chemicals. Catal Today 1997;37(4):367–404. Ranade V, Chaudhari R, Gunjal PR. Trickle Bed Reactors. Amsterdam, The Netherlands: Elsevier; 2011. Hessel V, Angeli P, Gavriilidis A, Löwe H. Gas–Liquid and Gas–Liquid–Solid Microstructured Reactors: Contacting Principles and Applications. Ind Eng Chem Res 2005;44(25):9750–9769. Chatterjee S, Degirmenci V, Aiouache F, Rebrov E V. Design of a radio frequency heated isothermal micro-trickle bed reactor. Chem Eng J 2014;243:225–233. Maiti R, Khanna R, Nigam KDP. Hysteresis in Trickle-Bed Reactors: A Review. Ind Eng Chem Res 2006;45(15):5185–5198. Maiti R, Atta A, Nigam KDP. Effect of particle porosity on hysteresis in trickle-bed reactors. Ind Eng Chem Res 2008;47(21):8126–8135. Maiti R, Khanna R, Nigam KDP. Trickle-bed reactors: Porosity-induced hysteresis. Ind Eng Chem Res 2005;44(16):6406–6413. Atta A, Roy S, Larachi F, Nigam KDP. Cyclic operation of trickle bed reactors: A review. Chem Eng Sci 2014;115:205–214. Charpentier JC, Favier M. Some liquid holdup experimental data in trickle-bed reactors for foaming and nonfoaming hydrocarbons. AIChE J 1975;21:1213–1218. Froment GF, Bischoff KB. Chemical reactor analysis and design. 2nd ed. New York, NY, USA: John Wiley & Sons Inc; 1979. Ramachandran PA, Chaudhari R V. Topics in chemical engineering volume 2: three-phase catalytic reactors. 1st ed. New York, NY, USA: Gordon and Breach, Science Publishers Inc.; 1983. Wang Y, Chen J, Larachi F. Modelling and simulation of trickle-bed reactors using computational fluid dynamics: A state-of-the-art review. Can J Chem Eng 2013;91(1):136–180.

214 | 5 Design of catalytic micro trickle bed reactors

[14] [15]

[16] [17] [18] [19] [20] [21] [22] [23] [24] [25]

[26] [27] [28] [29] [30]

[31] [32]

[33] [34] [35]

Sundaresan S. Role of hydrodynamics on chemical reactor performance. Curr Opin Chem Eng 2013;2(3):325–330. Hamidipour M, Chen J, Larachi F. CFD study and experimental validation of trickle bed hydrodynamics under gas, liquid and gas/liquid alternating cyclic operations. Chem Eng Sci 2013;89:158–170. Schwidder S, Schnitzlein K. A new model for the design and analysis of trickle bed reactors. Chem Eng J 2012;207-208:758–765. Gunjal PR, Kashid MN, Ranade V V, Chaudhari R V. Hydrodynamics of trickle-bed reactors: Experiments and CFD modeling. Ind Eng Chem Res 2005;44(16):6278–6294. Attou A, Boyer C, Ferschneider G. Modelling of the hydrodynamics of the cocurrent gas-liquid trickle flow through a trickle-bed reactor. Chem Eng Sci 1999;54(6):785–802. Larachi F, Iliuta I, Chen M, Grandjean BPA. Onset of pulsing in trickle beds: Evaluation of current tools and state-of-the-art correlation. Can J Chem Eng 1999;77(4):751–758. Ng KM. A model for flow regime transitions in cocurrent down-flow trickle-bed reactors. AIChE J 1986;32(1):115–122. Sicardi S, Gerhard H, Hoffmann H. Flow regime transition in trickle-bed reactors. Chem Eng J 1979;18(3):173–182. Grosser K, Carbonell RG, Sundaresan S. Onset of pulsing in two-phase cocurrent downflow through a packed bed. AIChE J 1988;34(11):1850–1860. Holub RA, Dudukovic MP, Ramachandran PA. Pressure drop, liquid holdup, and flow regime transition in trickle flow. AIChE J 1993;39(2):302–321. Bansal A, Wanchoo RK, Sharma SK. Flow regime transition in a trickle bed reactor. Chem Eng Commun 2005;192(7-9):1046–1066. Attou A, Ferschneider G. A two-fluid hydrodynamic model for the transition between trickle and pulse flow in a cocurrent gas-liquid packed-bed reactor. Chem Eng Sci 2000;55(3):491– 511. Faridkhou A, Larachi F. Hydrodynamics of Gas–Liquid Cocurrent Flows in Micropacked Beds – Wall Visualization Study. Ind Eng Chem Res 2012;51(50):16495–16504. Faridkhou A, Hamidipour M, Larachi F. Hydrodynamics of gas–liquid micro-fixed beds – Measurement approaches and technical challenges. Chem Eng J 2013;223:425–435. Banhart J. Manufacture, characterisation and application of cellular metals and metal foams. Prog Mater Sci 2001;46(6):559–632. Twigg MV, Richardson JT. Fundamentals and Applications of Structured Ceramic Foam Catalysts. Ind Eng Chem Res 2007;46(12):4166–4177. Losey MW, Jackman RJ, Firebaugh SL, Schmidt MA, Jensen KF. Design and fabrication of microfluidic devices for multiphase mixing and reaction. J Microelectromechanical Syst 2002; 11(6):709–717. Wada Y, Schmidt MA, Jensen KF. Flow Distribution and Ozonolysis in Gas−Liquid Multichannel Microreactors. Ind Eng Chem Res 2006;45(24):8036–8042. Wilhite BA, Jensen KF, Hill TF, Velásquez-García LF, Epstein AH, Livermore C. Design of a silicon-based microscale trickle-bed system for singlet-oxygen production. AIChE J 2008; 54(9):2441–2455. Baker D. Simultaneous flow of oil and gas. Oil Gas J 1954;53:183–195. Talmor E. Two-phase downflow through catalyst beds: part 1. flow maps. AIChE J 1977;23(6): 868–874. Günther A, Jensen KF. Multiphase microfluidics: from flow characteristics to chemical and materials synthesis. Lab Chip 2006;6(12):1487–1503.

References |

[36]

[37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53]

[54] [55] [56] [57]

215

Kulkarni R, Natividad R, Wood J, Stitt EH, Winterbottom JM. A comparative study of residence time distribution and selectivity in a monolith CDC reactor and a trickle bed reactor. Catal Today 2005;105(3-4):455–463. Krishnamurthy S, Peles Y. Gas-liquid two-phase flow across a bank of micropillars. Phys Fluids 2007;19(4):043302. Krishnamurthy S, Peles Y. Surface tension effects on adiabatic gas–liquid flow across micro pillars. Int J Multiph Flow 2009;35(1):55–65. Kan KM, Greenfield PF. Pressure drop and holdup in two-phase cocurrent trickle flows through beds of small packings. Ind Eng Chem Process Des Dev 1979;18(4):740–745. Ellman MJ, Midoux N, Laurent A, Charpentier JC. A new, improved pressure drop correlation for trickle-bed reactors. Chem Eng Sci 1988;43(8):2201–2206. Mohammed I, Bauer T, Schubert M, Lange R. Hydrodynamic multiplicity in a tubular reactor with solid foam packings. Chem Eng J 2013;231:334–344. Kan KM, Greenfield PF. Multiple hydrodynamic states in cocurrent two-phase downflow through packed beds. Ind Eng Chem Process Des Dev 1978;17(4):482–485. Levec J, Sáez AE, Carbonell RG. The hydrodynamics of trickling flow in packed beds. Part II: Experimental observations. AIChE J 1986;32(3):369–380. Stemmet CP, Van Der Schaaf J, Kuster BFM, Schouten JC. Solid Foam Packings for Multiphase Reactors. Chem Eng Res Des 2006;84(12):1134–1141. Stemmet CP (Promotor – Schouten JC). Gas-liquid solid foam reactors: hydrodynamics and mass transfer. PhD Thesis, Technische Universiteit Eindhoven, The Netherlands, 2008. Stemmet CP, Meeuwse M, van der Schaaf J, Kuster BFM, Schouten JC. Gas-liquid mass transfer and axial dispersion in solid foam packings. Chem Eng Sci 2007;62(18-20):5444–5450. Márquez N, Castaño P, Makkee M, Moulijn JA, Kreutzer MT. Dispersion and Holdup in Multiphase Packed Bed Microreactors. Chem Eng Technol 2008;31(8):1130–1139. Colombo AJ, Baldi G, Sicardi S. Solid-liquid contacting effectiveness in trickle bed reactors. Chem Eng Sci 1976;31(12):1101–1108. Yang X-L, Euzen J-P, Wild G. Residence time distribution of the liquid in gas–liquid cocurrent upflow fixed-bed reactors with porous particles. Chem Eng Sci 1990;45(11):3311–3317. Kulkarni RR, Wood J, Winterbottom JM, Stitt EH. Effect of fines and porous catalyst on hydrodynamics of trickle bed reactors. Ind Eng Chem Res 2005;44(25):9497–9501. Bej SK, Dabral RP, Gupta PC, et al. Studies on the Performance of a Microscale Trickle Bed Reactor Using Different Sizes of Diluent. Energy & Fuels 2000;14(3):701–705. Al-Herz M, Simmons MJH, Wood J. Selective hydrogenation of 1-heptyne in a mini trickle bed reactor. Ind Eng Chem Res 2012;51(26):8815–8825. Julcour-Lebigue C, Augier F, Maffre H, Wilhelm A-M, Delmas H. Measurements and Modeling of Wetting Efficiency in Trickle-Bed Reactors: Liquid Viscosity and Bed Packing Effects. Ind Eng Chem Res 2009;48(14):6811–6819. Lange R, Schubert M, Bauer T. Liquid Holdup in Trickle-Bed Reactors at Very Low Liquid Reynolds Numbers. Ind Eng Chem Res 2005;44(16):6504–6508. McManus RL, Funk GA, Harold MP, Ng KM. Experimental study of reaction in trickle-bed reactors with liquid maldistribution. Ind Eng Chem Res 1993;32(3):570–574. Baussaron L, Julcour-Lebigue C, Wilhelm AM, Delmas H, Boyer C. Wetting topology in trickle bed reactors. AIChE J 2007;53(7):1850–1860. Wu Y, Khadilkar MR, Al-Dahhan MH, Duduković MP. Comparison of upflow and downflow twophase flow packed-bed reactors with and without fines: Experimental observations. Ind Eng Chem Res 1996;35(2):397–405.

216 | 5 Design of catalytic micro trickle bed reactors

[58]

[59]

[60] [61] [62] [63] [64]

[65] [66] [67] [68]

[69]

[70]

[71] [72] [73] [74] [75] [76] [77] [78]

Van Herk D, Castaño P, Makkee M, Moulijn JA, Kreutzer MT. Catalyst testing in a multipleparallel, gas–liquid, powder-packed bed microreactor. Appl Catal A Gen 2009;365(2):199– 206. Fishwick RP, Natividad R, Kulkarni R, et al. Selective hydrogenation reactions: A comparative study of monolith CDC, stirred tank and trickle bed reactors. Catal Today 2007;128(1-2 SPEC. ISS.):108–114. Márquez N, Castaño P, Moulijn JA, Makkee M, Kreutzer MT. Transient Behavior and Stability in Miniaturized Multiphase Packed Bed Reactors. Ind Eng Chem Res 2010;49(3):1033–1040. Mears DE. The role of axial dispersion in trickle-flow laboratory reactors. Chem Eng Sci 1971;26(9):1361–1366. Fu MS, Tan CS. Liquid holdup and axial dispersion in trickle-bed reactors. Chem Eng Sci 1996;51(24):5357–5361. Van Herk D, Kreutzer MT, Makkee M, Moulijn JA. Scaling down trickle bed reactors. Catal Today 2005;106(1-4):227–232. Kilpiö T, Mäki-Arvela P, Rönnholm M, Sifontes V, Wärnå J, Salmi T. Modeling of a three-phase continuously operating isothermal packed-bed reactor: Kinetics, mass-transfer, and dispersion effects in the hydrogenation of citral. Ind Eng Chem Res 2012;51(26):8858–8866. Pant HJ, Saroha AK, Nigam KDP. Measurement of liquid holdup and axial dispersion in trickle bed reactors using radiotracer technique. Nukleonika 2000;45(4):235–241. Saroha AK, Nigam KDP, Saxena AK, Kapoor VK. Liquid distribution in trickle-bed reactors. AIChE J 1998;44(9):2044–2052. Jin J, Cheng ZM, Li JG, Wu SC. Hydrodynamics of multiphase flow in a trickle-bed filled with small particles under the supercritical condition. Chem Eng Sci 2013;100:69–73. Iliuta I, Thyrion FC. Flow regimes, liquid holdups and two-phase pressure drop for two-phase cocurrent downflow and upflow through packed beds: air/Newtonian and non-Newtonian liquid systems. Chem Eng Sci 1997;52(21-22):4045–4053. Wild G, Larachi F, Charpentier JC. Heat and mass transfer in gas-liquid-solid fixed bed reactors. In: Quintard M, Todorovic M, editors. Heat and Mass Transfer in Porous Media. Amsterdam: Elsevier; 1992. Rocha JA, Bravo JL, Fair JR. Distillation columns containing structured packings: A comprehensive model for their performance. 2. Mass-transfer model. Ind Eng Chem Res 1996;35(5):1660–1667. Fukushima S, Kusaka K. Gas-liquid mass transfer and hydrodynamic flow region in packed columns with cocurrent upward flow. J. Chem. Eng. Japan. 1979;12(4):296–301. Losey MW, Schmidt MA, Jensen KF. Microfabricated Multiphase Packed-Bed Reactors: Characterization of Mass Transfer and Reactions. Ind Eng Chem Res 2001;40(12):2555–2562. Mäki-Arvela P, Eränen K, Alho K, Salmi T, Murzin DY. Multitubular reactor design as an advanced screening tool for three-phase catalytic reactions. Top Catal 2007;45(1-4):223–227. Tadepalli S, Halder R, Lawal A. Catalytic hydrogenation of o-nitroanisole in a microreactor: Reactor performance and kinetic studies. Chem Eng Sci 2007;62(10):2663–2678. Metaxas KC, Papayannakos NG. Kinetics and mass transfer of benzene hydrogenation in a trickle-bed reactor. Ind Eng Chem Res 2006;45(21):7110–7119. Metaxas K, Papayannakos N. Gas-liquid mass transfer in a bench-scale trickle bed reactor used for benzene hydrogenation. Chem Eng Technol 2008;31(10):1410–1417. Saroha AK. Solid-liquid mass transfer studies in trickle bed reactors. Chem Eng Res Des 2010;88(5-6):744–747. Van Houwelingen AJ, Nicol W. Parallel hydrogenation for the quantification of wetting efficiency and liquid-solid mass transfer in a trickle-bed reactor. AIChE J 2011;57(5):1310–1319.

References |

[79]

217

Durante D, Kilpiö T, Suominen P, et al. Modeling and simulation of a small-scale trickle bed reactor for sugar hydrogenation. Comput Chem Eng 2014;66:22–35. [80] Mohammed I, Bauer T, Schubert M, Lange R. Liquid–solid mass transfer in a tubular reactor with solid foam packings. Chem Eng Sci 2014;108:223–232. [81] Tourvieille JN. Innovating microstructured gas-liquid-solid reactors?: a contribution to the understanding of hydro dynamics and mass transfers. 2014;171. [82] Abdallah R, Magnico P, Fumey B, De Bellefon C. CFD and kinetic methods for mass transfer determination in a mesh microreactor. AIChE J 2006;52(6):2230–2237. [83] Kreutzer MT, Du P, Heiszwolf JJ, Kapteijn F, Moulijn JA. Mass transfer characteristics of threephase monolith reactors. Chem Eng Sci 2001;56(21-22):6015–6023. [84] Cybulski A, Moulijn JA. Structured Catalysts and Reactors. 2nd ed. CRC Press; 2005. [85] Watson PC, Harold MP. Dynamic effects of vaporization with exothermic reaction in a porous catalytic pellet. AIChE J 1993;39(6):989–1006. [86] Marcandelli C, Wild G, Lamine AS, Bernard JR. Measurement of local particle-fluid heat transfer coefficient in trickle-bed reactors. Chem Eng Sci 1999;54(21):4997–5002. [87] Boelhouwer J, Piepers H, Drinkenburg AA. Particle–liquid heat transfer in trickle-bed reactors. Chem Eng Sci 2001;56(3):1181–1187. [88] Ruether JA, Yang CS, Hayduk W. Particle mass transfer during cocurrent downward gas-liquid flow in packed beds. Ind Eng Chem Process Des Dev 1980;19(1):103–107. [89] Heidari A, Hashemabadi SH. Numerical evaluation of the gas-liquid interfacial heat transfer in the trickle flow regime of packed beds at the micro and meso-scale. Chem Eng Sci 2013; 104:674–689. [90] Iliuta I, Larachi F, Al-Dahhan MH. Double-slit model for partially wetted trickle flow hydrodynamics. AIChE J 2000;46(3):597–609. [91] Fedorov AG, Viskanta R. Three-dimensional conjugate heat transfer in the microchannel heat sink for electronic packaging. Int J Heat Mass Transf 2000;43(3):399–415. [92] Lee P-S, Garimella SV, Liu D. Investigation of heat transfer in rectangular microchannels. Int J Heat Mass Transf 2005;48(9):1688–1704. [93] Norton DG, Wetzel ED, Vlachos DG. Thermal Management in Catalytic Microreactors. Ind Eng Chem Res 2006;45(1):76–84. [94] Mariani NJ, Bressa SP, Mazza GD, Martínez OM, Barreto GF. A bem formulation for evaluating conductive heat transfer rates in particulate systems. Numer Heat Transf Part B Fundam 1997;31(4):459–476. [95] Dudas J, Hanika J, Lepuru J, Barkhuysen M. Thymol hydrogenation in bench scale trickle bed reactor. Chem Biochem Eng Q 2005;19(3):255–262+i. [96] Hickman DA, Holbrook MT, Mistretta S, Rozeveld SJ. Successful scale-up of an industrial trickle bed hydrogenation using laboratory reactor data. Ind Eng Chem Res 2013;52(44):15287–15292. [97] Gunjal PR, Ranade VV. Modeling of laboratory and commercial scale hydro-processing reactors using CFD. Chem Eng Sci 2007;62(18-20):5512–5526. [98] Charpentier J-C. In the frame of globalization and sustainability, process intensification, a path to the future of chemical and process engineering (molecules into money). Chem Eng J 2007;134(1-3):84–92. [99] Nigam KDP, Larachi F. Process intensification in trickle-bed reactors. Chem Eng Sci 2005; 60(22):5880–5594. [100] Wang Y, Chen J, Munteanu M, Larachi F. CFD modeling and simulation of forced periodic operation of trickle bed reactors. Salt Lake City, UT: 2010. [101] Lange R, Schubert M, Dietrich W, Grünewald M. Unsteady-state operation of trickle-bed reactors. Chem Eng Sci 2004;59(22-23):5355–5361.

218 | 5 Design of catalytic micro trickle bed reactors

[102] Borren B, Zeitler O, Schmaus C, Agar DW. Gas-liquid mass transfer in periodically operated trickle-bed reactors: Influence of the liquid-phase physical properties and flow modulation. Chem Eng Process Process Intensif 2010;49(9):923–929. [103] Liu G, Mi Z. Hydrogenation of 2-ethylanthraquinones in a periodically operated trickle-bed reactor. Chem Eng Technol 2005;28(8):857–862. [104] Liu G, Zhang X, Wang L, Zhang S, Mi Z. Unsteady-state operation of trickle-bed reactor for dicyclopentadiene hydrogenation. Chem Eng Sci 2008;63(20):4991–5002. [105] Habtu N, Font J, Fortuny A, et al. Heat transfer in trickle bed column with constant and modulated feed temperature: Experiments and modeling. Chem Eng Sci 2011;66(14):3358–3368. [106] Aydin B, Larachi F. Structure of trickle-to-pulse flow regime transition and pulse dynamics at elevated temperature in slow-mode cyclic operation. Chem Eng Sci 2008;63(6):1510–1522. [107] Aydin B, Cassanello MC, Larachi F. Influence of temperature on fast-mode cyclic operation hydrodynamics in trickle-bed reactors. Chem Eng Sci 2008;63(1):141–152. [108] Massa P, Ayude MA, Ivorra F, Fenoglio R, Haure P. Phenol oxidation in a periodically operated trickle bed reactor. Catal Today 2005;107-108:630–636. [109] Hotz N, Osterwalder N, Stark WJ, Bieri NR, Poulikakos D. Disk-shaped packed bed microreactor for butane-to-syngas processing. Chem Eng Sci 2008;63(21):5193–5201. [110] Alonso DM, Bond JQ, Dumesic JA. Catalytic conversion of biomass to biofuels. Green Chem 2010;12(9):1493–1513. [111] Rinaldi R, editor. Catalytic Hydrogenation for Biomass Valorization. Cambridge: Royal Society of Chemistry; 2014. [112] Jeon JK, Yim JH, Park YK. C9-aldehyde hydrogenation over nickel/kieselguhr catalysts in trickle bed reactor. Chem Eng J 2008;140(1-3):555–561. [113] Kilpiö T, Aho A, Murzin D, Salmi T. Experimental and modeling study of catalytic hydrogenation of glucose to sorbitol in a continuously operating packed-bed reactor. Ind Eng Chem Res 2013;52(23):7690–7703. [114] Xi Y, Jackson JE, Miller DJ. Characterizing lactic acid hydrogenolysis rates in laboratory trickle bed reactors. Ind Eng Chem Res 2011;50(9):5440–5447. [115] Jang H, Kim S-H, Lee D, et al. Hydrogenation of lactic acid to propylene glycol over a carbon-supported ruthenium catalyst. J Mol Catal A Chem [Internet] 2013 [cited 2015 Jan 29];380:57–60. Available from: http://www.scopus.com/inward/record.url?eid=2-s2.084885337491&partnerID=tZOtx3y1 [116] Son SM, Kusakabe K. Transesterification of sunflower oil in a countercurrent trickle-bed reactor packed with a CaO catalyst. Chem Eng Process Process Intensif 2011;50(7):650–654. [117] Brandner A, Lehnert K, Bienholz A, Lucas M, Claus P. Production of biomass-derived chemicals and energy: Chemocatalytic conversions of glycerol. Top Catal 2009;52(3):278–287. [118] Xi Y, Holladay JE, Frye JG, Oberg AA, Jackson JE, Miller DJ. A kinetic and mass transfer model for glycerol hydrogenolysis in a trickle-bed reactor. Org Process Res Dev 2010;14(6):1304– 1312. [119] LaPorte TL, Hamedi M, DePue JS, Shen L, Watson D, Hsieh D. Development and scale-up of three consecutive continuous reactions for production of 6-hydroxybuspirone. Org Process Res Dev 2008;12(5):956–966. [120] Shang M, Noël T, Wang Q, Hessel V. Packed-Bed Microreactor for Continuous-Flow Adipic Acid Synthesis from Cyclohexene and Hydrogen Peroxide. Chem Eng Technol 2013;36(6):1001– 1009. [121] Wörz N, Arras J, Claus P. Continuous selective hydrogenation of citral in a trickle-bed reactor using ionic liquid modified catalysts. Appl Catal A Gen 2011;391(1-2):319–324. [122] Kapteijn F, Nijhuis T, Heiszwolf J, Moulijn J. New non-traditional multiphase catalytic reactors based on monolithic structures. Catal Today 2001;66(2-4):133–144.

References | 219

[123] Gascon J, van Ommen JR, Moulijn JA, Kapteijn F. Structuring catalyst and reactor – an inviting avenue to process intensification. Catal Sci Technol 2015;5:807–817. [124] Bauer T, Haase S. Comparison of structured trickle-bed and monolithic reactors in Pd-catalyzed hydrogenation of alpha-methylstyrene. Chem Eng J 2011;169(1-3):263–269. [125] Hessel V, Hardt S, Löwe H. Chemical Micro Process Engineering: Fundamentals, Modelling and Reactions. 1st ed. Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. KGaA; 2006. [126] Ehrfeld W, Hessel V, Lowe H. Microreactors: new technology for modern chemistry. 1st ed. Weinheim: Wiley-VCH Verlag GmbH; 2000.

Joseph Wood

6 Three-phase catalytic reactors for hydrogenation and oxidation reactions Nomenclature a a ab A B b C1 C󸀠 C ∗A Ca C Ab C Bb C Ag C AS CH2 Ci Cj d dp d32 D D D AB FA FB H H k ki k1 k cA kg k 󸀠gA k lA K K1 , K2 , K4 KA KB KH Ki

Exponent; Eq. (6.12) Area; Eq. (6.36) Gas-liquid interfacial area Concentration of reactant A; Eqs. (6.18)–(6.23) Concentration of reactant B; Eqs. (6.18)–(6.23) Exponent; Eq. (6.12) Constant; Eq. (6.12) Constant; Eq. (6.13) Equilibrium concentration of dissolved gas Capillary number Concentration of reactant A in the bulk Concentration of reactant B in the bulk Concentration of reactant A in the gas phase Concentration of reactant A at the catalyst surface Concentration of hydrogen (dissolved) Concentration of component i; Eqs. (6.22), (6.23) Concentration of component j; Eqs. (6.22), (6.23) Impeller diameter Particle diameter Mean bubble size Reactor diameter or diameter of capillary Diffusion coefficient; Eqs. (6.13), (6.38) Diffusivity of A through B Molar flowrate of A Molar flowrate of B Henry’s law constant Liquid height in the reactor; Eq. (6.11) Reaction rate constant; Eqs. (6.18)–(6.23) Rate constant reaction i; Eq. (6.23) Pseudo first order rate constant Liquid-solid mass transfer coefficient Gas-liquid mass transfer coefficient, gas side Overall coefficient in trickle bed design; Eq. (6.26) Gas-liquid mass transfer coefficient, liquid side Consistency index; Eq. (6.13) Adsorption coefficients of alcohol,oxygen and product; Eq. (6.16) Adsorption coefficient of A; Eqs. (6.18)–(6.23) Adsorption coefficient of B; Eqs. (6.18)–(6.23) Adsorption coefficient of hydrogen; Eqs. (6.22), (6.23) Adsorption coefficient component i; Eqs. (6.22), (6.23)

Nomenclature | 221

L m n n cr P P PH q rb rc rr −r AW I −r AW R R󸀠A Re Sc Sh T U Ug Ut v sg V VL W XA

Length of liquid slug; Eq. (6.37) Mass loading of catalyst in the reactor Exponent; Eq. (6.13) Critical impeller speed Power input per unit volume Pressure; Eqs. (6.18)–(6.23) Pressure of hydrogen; Eq. (6.22) Volumetric gas flowrate External resistance at the gas bubble External resistance at the catalyst particle Internal resistance at the catalyst particle Actual observed rate per unit mass of catalyst Intrinsic rate of reaction per unit mass of catalyst Ideal gas constant; Eq. (6.22) Rate of reaction per unit bed volume Reynolds number Schmidt number Sherwood number Temperature Fluid velocity; Eq. (6.16) Superficial gas velocity; Eq. (6.13) Terminal velocity of a bubble in free rise; Eq. (6.13) Superficial gas velocity Velocity Liquid volume Mass of catalyst Conversion of A

Greek Letters α Exponent δ Film thickness ε Voidage εb Bed density εT Mean specific energy dissipation rate η Catalyst effectiveness factor μ Liquid viscosity μa Apparent viscosity of slurry in stirred tank μw Viscosity of slurry in water ρ Liquid density ρl Apparent density of slurry; Eq. (6.13) ρp Particle density σ Liquid surface tension ϕ Thiele modulus

222 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

Additional Subscripts C Capillary G Gas L Liquid OV Overall S Solid slug Slug TP Two phase

6.1 Introduction The term “three phase catalytic reactors” generally refers to gas-liquid-solid reaction systems, in which the solid is typically a heterogeneous catalyst [1]. A fundamental division of three-phase reactors may be made by whether the solid phase is suspended in the liquid as in a slurry reactor, present as a fixed bed typical of a trickle bed reactor, or attached to some kind of structured support, for example a monolith reactor. Three-phase reactors have found extensive application in industry, for example in the production of bulk and fine chemicals, production of petroleum product and biochemical reactions. Winterbottom and King [2] have listed some typical applications of three-phase reactors as: triglyceride hydrogenation, hydrodesulphurization, hydrocracking, methanol synthesis and fine chemical synthesis. Examples of the latter include geraniol hydrogenation for the production of fragrances, conversion of nitriles to primary amines (for use in the production of pharmaceuticals), manufacture of dyestuffs, specialty chemical production, such as the alkylation of phenol, reaction of glucose to produce sorbitol. A number of textbooks have been devoted to the subject of multiphase catalytic reactors [3, 4]. However, the subject remains an active field of research for a number of reasons. These include the development of new catalysts with enhanced selectivity and reaction rate, the emergence of a range of non-invasive imaging techniques that lead to new insights into reactor behavior and the new generation of reactor modeling capabilities such as computational fluid dynamics. Industrial drivers towards the study of three-phase reactors include a requirement to increase productivity and profit margins, maximize catalyst lifetime, as well as environmental considerations such as the reduction of waste and efficient use of raw materials. Emerging industries such as the conversion of biofeedstocks and utilization of heavy fossil fuels from the bottom of the oil barrel have also driven the need for new applications of three-phase reactors. The aim of this chapter is to provide a brief review of the theory of three-phase reactors, although more detailed texts are referred to for a thorough treatment, before reviewing a range of relevant research in the field of three-phase reactors. The chapter is divided according to slurry, trickle bed and structured catalysts.

6.2 Slurry Reactors |

223

6.2 Slurry Reactors The slurry reactor may be defined as a three phase catalytic reactor in which the catalyst is in the form of small particles, which can be porous or non-porous, with a size typically in the range 50–200 μm or even less, where the fluid motion suspends the particles in the liquid [1]. Such reactors can be further sub-divided into stirred tank reactors, where fluid motion is induced by an impeller, or bubble column reactors, where the fluid motion occurs as a result of vigorous bubbling of gas through the liquid phase. Slurry type reactors are often operated as semi-batch or batch reactors, which offer good flexibility for processes such as fine chemicals and pharmaceuticals, where product recipes may regularly change and batch traceability is required. The process engineering must be considered when selecting a suitable type of reactor, for example the slurry reactor requires the catalyst to be separated after reaction, representing an additional cost of filtration. Slurry reactors are attractive for highly exothermic reactions because of their high liquid holdup/catalyst mass. Bubbling slurry reactors are useful for processes such as fermentation, which are catalyzed by biocatalysts.

6.2.1 Theory: Determination of Controlling Resistance Although slurry reactors have been an industry workhorse for many years, their design is complicated due to the interplay between physical processes such as mixing, diffusion and mass transfer with chemical reaction processes. Therefore, gas-liquid and liquid-solid mass transfer, pore diffusion and suitable reaction kinetic models must be taken into consideration. The analysis of such reactors typically follows the reaction of a gas (A) with liquid reactant (B) to produce a liquid phase product (P). A number of steps occur in the course of this process, which include: 1. Absorption of A from the gas phase in to the liquid; 2. Diffusion of A across a boundary layer at the gas-liquid interface; 3. Diffusion of A through the bulk liquid; 4. Diffusion of A across a boundary at the external surface of the catalyst particle; 5. Internal diffusion of A and B through the catalyst pores; 6. Reaction of A and B at the active sites of the porous catalyst; 7. Diffusion of products out of the catalyst pellet. Fig. 6.1 depicts the various mass transfer and reaction steps that must take place in order for the reactants to reach the catalyst surface and react, together with the relevant interface concentrations of A. The equations associated with each step are outlined in Tab. 6.1. The reader is referred to Winterbottom and King [2] or Fogler [5] for a full derivation of the relevant equations. These simple models are useful in understanding the parameters that control the three-phase reactor and hence how the design may be optimized to achieve the maximum reaction rate, as briefly discussed below.

224 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

1

2

3

4

C Ag C Ag = Bulk gas phase concentration of A

CA* C Aig

C Ab

C Aig = Concentration of A at gas-side of liquid interface

S1

C A* = Concentration of A at liquid-side of liquid interface C Ab = Bulk liquid concentraion of A C AS = Concentration of A at catalyst surface

S2

C AS Gas Film

Molecular and Knudsen Diffusion from site S1 to site S2

Liquid Film

Bubble Interface

Fig. 6.1: Resistances at the gas-liquid and liquid-catalyst interfaces in a three-phase reactor system.

Gas absorption It is observed from Eq. (6.1) of Tab. 6.1 that the concentration of dissolved gas at the interface is dependent upon Henry’s law, where Henry’s law constant (H) is a measure of the solubility of gas in the particular solvent or reactant used and varies as a function of reactor operating temperature. Here it is assumed that concentration gradients in the gas phase are negligible, if the gas bubbles contain only small amounts of components other than A, or if A is only sparingly soluble in the liquid phase. Diffusion processes Eqs. (6.2) and (6.3) represent the respective gas-liquid and liquid-solid mass transfer steps, encompassing the mass transfer coefficients for the gas-liquid (k lA ) and liquidsolid (k cA ), the gas-liquid interfacial area (a b ), catalyst specific surface area (a c ) and mass loading of catalyst in the reactor (m). Reaction in the catalyst particle Eqs. (6.4)–(6.8) represent the diffusion/reaction processes, the actual observed rate (−r AW ) expressed in Eq. (6.4) with the intrinsic rate that would occur over a finely crushed catalyst without diffusion limitations (−r IAW ) and the catalyst effectiveness factor (η) being expressed in Eq. (6.5). In Eq. (6.8), it is assumed that the liquid phase reactant is present at a sufficiently high concentration that the kinetics become first order in terms of hydrogen concentration at the catalyst surface (C AS ) with pseudo first order rate constant.

6.2 Slurry Reactors |

225

Overall slurry reactor model Eq. (6.9) represents a combination and rearrangement of Eqs. (6.1)–(6.5) and (6.8) to provide an overall slurry reactor model, where r b is a mass transfer resistance at the gas bubble, r c is a mass transfer resistance at the external surface of the catalyst particle and r r represents internal diffusion/reaction resistance inside the catalyst particle. The slurry reactor ‘model’ can be expressed as a graphical plot of C∗A /R A against reciprocal catalyst loading (1/m), as shown in Fig. 6.2. Thus the slope of the graph represents the combined external diffusion resistance r c and internal diffusion/reaction resistance r r at the catalyst particle, whilst the intercept of the graph represents the gas absorption resistance at the gas-liquid interface, r b . The intercept in Fig. 6.2 is determined only by the gas absorption resistance, which could be altered by the use of an improved sparger design or higher gas pressure. The slope of the graph, representing the resistances at the catalyst particle is influenced by the catalyst particle size, and therefore internal and external diffusion resistances can be decreased by using a more finely crushed catalyst particle. Tab. 6.1: Summary of expressions for mass transfer, diffusion and reaction parameters in the threephase catalytic reactor, following analysis of Winterbottom and King [2]. Process

Expression

Dissolution of gas at the gas-liquid interface, according to Henry’s law

C ∗A =

Rate of mass transfer from gas-liquid interface to bulk liquid

R A = k lA a b (C ∗A − C Ab )

(6.2)

Rate of mass transfer across boundary layer at external surface of catalyst pellets

R A = k cA a c m(C Ab − C AS )

(6.3)

Rate of diffusion and reaction in the catalyst pores

R A = m(−r AW )

(6.4)

Relation of actual reaction rate to intrinsic rate Expression for catalyst effectiveness factor for first order reaction in a spherical catalyst pellet

RA =

Eq.

C Aig H

(6.1)

I mη(−r AW )

(6.5)

3 1 1 η= [ − ] ϕ tanh ϕ ϕ

(6.6)

d p ρ P k1 √ 2 Deff

Thiele modulus

ϕ=

Pseudo first order reaction rate in hydrogen

I (−r AW ) = k 1 C AS

(6.8)

Overall slurry reactor model

C ∗A 1 1 1 1 = + + ( ) RA k lA a b m k cA a c ηk 1

(6.9)

(6.7)

or C ∗A RA

= rb +

1 (r c + r r ) m

(6.10)

226 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

rcr = rc + rr C*A RA

Diffusion/reaction resistance

rb =

1 kb ab = gas absorption resistance 1/m

Fig. 6.2: Plot of equilibrium dissolved gas concentration/reaction rate versus reciprocal catalyst loading, showing magnitude of resistances at the gasliquid interface and catalyst particle.

6.2.2 Mixing and mass transfer in the stirred tank slurry reactor The introductory sections of this chapter showed that the slurry reactor performance may be influenced by both reaction kinetics and various mass transfer resistances, associated with gas absorption and diffusion at the catalyst particle. In order to design or predict the performance of the stirred tank reactor, the gas-liquid and liquid-solid mass transfer coefficients must be calculated. Markopoulos et al. [6] has reviewed mass transfer coefficients in mechanically agitated gas-liquid contactors. There are many correlations reported for the prediction of gas-liquid (liquid-side) mass transfer coefficients in stirred vessels, such that no single correlation is suitable for all situations. Rather the type of fluid (Newtonian or non-Newtonian), geometry of the system and type of impeller must be considered in selecting a suitable correlation. For lowviscous and coalescing Newtonian fluids a number of correlations have been proposed in the literature, which are dependent upon material properties, geometry of the system and operating conditions, of the form: k l a = f(P/V L , v sg , H/D, d/D, q/V L , ρ, μ, σ)d ,

(6.11)

where k l a is the volume based liquid side mass transfer coefficient, P/V L is the power input, P is per unit liquid volume, V L , v sg is the superficial gas velocity, H is the liquid height in the reactor, d/D is the ratio of the impeller diameter to the reactor diameter, q/V L is the volumetric gas flow rate, q, per unit liquid volume, ρ is the liquid density, μ is the liquid viscosity and σ is the liquid surface tension. For a given geometry, a critical impeller speed exists, n cr . For impeller speed n < n cr , the interfacial area and volume-related mass transfer coefficient depends on the gas load and not on the speed of the agitator. For n > n cr , the volume-related mass transfer coefficient increases linearly with impeller speed. In this region, there is no effect of the gas flow rate on a and k l a. Fig. 6.3 shows a cut-away diagram of the stirred tank reactor. Various types of impellers are available, including pitched blade turbines and Rushton turbines, as shown in Fig. 6.4. For given geometry and fluid properties, the mass transfer coefficient

227

6.2 Slurry Reactors |

is often expressed as: b

k l a = C1 (P/V L )a1 v sg1 .

(6.12)

The exponents, a and b vary for different systems, depending upon the stirrer and gas distributor designs, whilst the constant C is influenced by the liquid type and properties. Three phase slurry reactors contain a dispersion of solid catalyst particles, whilst many gas-liquid mass transfer correlations do not allow for the presence of the particles upon the mass transfer process. Kawase et al. [7] found that volumetric mass transfer coefficients in water decreased due to the presence of solid particles at constant impeller speed and superficial gas velocity, and proposed an extended mass transfer correlation for gas-liquid-solid three-phase systems: k l a = C󸀠 √D

−3/5 (9+4n)/10(1+n) ε

ρl

(K/ρ l )1/2(1+n) σ3/5

(

U g 1/2 μ a −0.25 ) ( ) Ut μw

(6.13)

Where C󸀠 is a constant, D is diffusivity, ρ l is apparent density of slurry, K is a consistency index of slurry, σ is surface tension, U g is superficial gas velocity, U t is terminal velocity of a bubble in free rise, μ a is apparent viscosity of slurry in stirred tank and μ w is viscosity of slurry in water. Hydrogenation reactions are an important class of industrial reaction, which may be strongly influenced by gas-liquid mass transfer considerations. They are often carried out in a solvent in order to dissolve the reactant, to absorb heat of reaction and to solubilize coke precursors from the catalyst surface. In recent years, there has been

Fig. 6.3: Schematic of a stirred tank reactor with Rushton turbine and gas sparger.

228 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

Sensitive to sparger design Form large cavities Down-Pumping pitched blade turbine

Suffer power instanbilities Rushton turbine

Up-Pumping pitched blade turbine

‘Flat’ power characteristics

Fig. 6.4: Pitched blade and Rushton turbines and their characteristics.

a motivation to use ‘greener’ solvents rather than volatile organic carbons (VOCs) for environmental considerations, for example water or supercritical carbon dioxide. It has also been shown that the type of solvent used can influence the rate and selectivity of the reaction. A number of reasons have been suggested to explain the solvent effects observed [8]: 1. The solubility of hydrogen in the reaction media; 2. Competitive adsorption of solvent at active catalyst sites; 3. Agglomeration of catalyst in some solvents; 4. Intermolecular interaction between the reactant and solvent molecules. The reaction rate in a particular system may have a complex dependence upon the system properties, including the selection of solvent [9]. In addition to the solvent effects listed above, under typical industrial conditions the reaction rate may be mass transfer controlled and thus dependent on the gas-liquid interfacial area. Hu et al. [9] studied the effect of solvent composition, including water, 2-propanol and mixtures of the two upon three-phase hydrogenation of 2-butyne-1,4-diol on Pd/Al2 O3 catalysts in a stirred reactor with mean specific energy dissipation rates, ε T , up to 50 W kg−1 . The investigations were carried out using a 3 l reactor operated at 1 barg pressure, 35 °C and with a 12 % v/v headspace prior to sparging. Pd/Al2 O3 particles and alumina particles were mixed to give an overall metal content of 1 wt % Pd whilst reducing the number of dark Pd particles to improve the observation of bubbles. For the first time, bubble sizes were measured in situ with reaction rates using a video-microscope-computer system [10] during reaction via a viewing window protruding very slightly in to the vessel, at a constant flow rate of 0.25 l/min hydrogen at 1 barg. Typical bubble images are shown in Fig. 6.5. The stirred tank was agitated using an up-pumping pitch blade

6.2 Slurry Reactors |

400 μm

229

400 μm

(a)

(b)

Fig. 6.5: Bubble images at ε T = 3.3 W/kg during hydrogenation of 2-butene-1,4-diol in water: (a) 20 min after beginning the reaction; (b) post-reaction. Reprinted from Hu et al. © [9], with permission from Elsevier.

turbine for the two-step hydrogenations, whilst the second step of 2-butene-1,4-diol to the alkane used a Rushton turbine. The hydrogenation of alkynes is often studied as a model for the investigation of three-phase reactions. The reaction scheme is shown in Fig. 6.6. The desirable product in alkyne hydrogenations is often the cis-alkene, where 2-butene-1,4-diol is used in the manufacture of products such as vitamins and insecticide, the over hydrogenation to the alkane or formation of side products being undesirable. Hu et al. [11] also reported bubble size measurements for air-water systems; they indicated that in the single component solvents, irregular, relatively large bubbles with a wide size distribution were observed. In the mixed aqueous/organic solvents, and especially at lower concentrations of IPA (1 %, 5 %, 10 % v/v), the bubbles were spherical, much smaller and with HOH2C –C=C– CH2OH 2-butyne-1,4-diol 1 1% Pd/AI2O3 +H2 +½H2 HOH2C –CH=CH–CH2OH cis and/or trans-2-butene-1,4-diol 2 +H2 HOH2C –CH2–CH2–CH2OH 1,4-butanediol

CH2–CH=CH–CH2OH 2-buten-1-ol –H2O +H2 CH3–CH2–CH2–CH2OH 1-butanol +2H2 –H2O HOH2C–CH2–CH2–CHO 4-hydroxybutyraldehyde +½H2 –H2O CH3–CH2–CH2–CHO butanal

Fig. 6.6: Reaction scheme for the hydrogenation of 2-butyne-1,4-diol. Reprinted from Hu et al. © [9], with permission from Elsevier.

230 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

a narrow size distribution. This was explained as being due to interaction between the organic and aqueous phase leading to lower incidence of coalescence in mixed solvents. A similar effect was observed during and after the reaction in the hydrogenation of 2-butyne-1,4-diol, as shown by the Sauter mean bubble size, d32 , versus solvent concentration in Fig. 6.7 [9]. The bubble size is observed to pass through a minimum at a concentration of 5 % v/v 2-propanol in water. Fig. 6.8 shows the average uptake of hydrogen for the two-step reaction as a function of solvent concentration. For both steps of the reaction, there is a peak at around 5 % v/v 2-propanol, followed by a fall to 30 % v/v before rising steadily to pure 2-propanol, where a rate approximately double that observed in water is achieved, although the solubility of hydrogen in 2-propanol is around six times higher than in water. The first step of the reaction was found to be under kinetic control, but the second step was determined to be under gas-liquid mass transfer control. The peaks in reaction rate observed in Fig. 6.8 could be explained in terms of the bubble size and solubility of hydrogen. At a solvent concentration of 5 % v/v the smallest bubble size led to the highest interfacial area available to mass transfer of gas to liquid, and thus giving a peak in reaction rate at the corresponding concentration. As the amount of 2-propanol in the solvent mixture increased, the reaction rate initially decreased due to a larger bubble size and lower interfacial area. However, at the highest concentrations of 2-propanol it increased further due to the higher solubility of hydrogen in the organic solvent compared with water. This trend for reaction rate as a function of solvent composition was predicted by the parameters bubble size and hydrogen solubility. When the reaction is under gas-liquid mass transfer control, the overall rate can be expressed by the proportional relationship: rαCH2 /d32 .

(6.14)

Liquid-solid mass transfer resistance at the outside surface of the catalyst particle may be a significant consideration for large particles or slowly agitated systems. If shear stress occurs between the particle and fluid, an appropriate form of the Frössling equation may be used to calculate the liquid-solid mass transfer coefficient: Sh = 2.0 + 0.6Re 1/2 Sc 1/3 ,

(6.15)

k cA d p D AB . At sufficiently high shear stresses such

where Sh is the Sherwood number, Sh = that the Reynolds number has the dominant effect upon the Sherwood number, it is found that the mass transfer coefficient is mainly influenced by fluid velocity and particle diameter, according to the relationship k cA α

U 1/2 1/2

.

(6.16)

dp

Thus increasing the fluid velocity, U, or using a smaller catalyst particle diameter, d p , can be used to increase the liquid-solid mass transfer coefficient. Fishwick et al. [12] asserted that the design of stirred vessels has been more of an art than science, relying on “black box” type approaches. Poorly understood mixing

6.2 Slurry Reactors |

231

600

d32(μm)

500

400

300

200 0

20 40 60 80 100 Concentration of 2-propanol in water (vol%)

Fig. 6.7: Bubble diameter, d32 , versus concentration of 2-propanol in water during (∘) and post reaction (∙). Reprinted from Hu et al. © [9], with permission from Elsevier.

Hydrogen uptake (dm3min–1)

0.8

0.6

0.4

0.2

0.0 0

20 40 60 80 100 Concentration of 2-propanol in water (vol%)

Fig. 6.8: Reaction rate versus concentration of 2-propanol in water for hydrogenation of 0.2 M 2-butyne-1,4-diol: ∙ first step; ∘ second step. Reprinted from Hu et al. © [9], with permission from Elsevier.

issues may lead to the loss of several billion dollars per year to the related process industries. However, this approach is starting to change owing to various non-invasive imaging techniques such as Electrical Capacitance Tomography (ECT), Positron Emission Particle Tracking (PEPT) and Particle Image Velocimetry (PIV). Such techniques can be used to obtain information such as fluid trajectories, velocity distributions, velocity gradients and shear rates. The design of stirred reactors presents a challenge in order to achieve bulk fluid motion to guarantee good circulation in the mixing vessel, whilst also achieving suspension of the catalyst particles and dissipation of the gas

232 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

bubbles throughout the fluid, requiring adequate local velocities to be achieved in the parts of the vessel remote from the impeller. The majority of earlier two-phase studies investigated the velocity fields occurring with Rushton turbines, using Laser Doppler Velocimetry (LDV) [13] or PIV [14]. However, such optical techniques have mainly been carried out at low gassing rates and without particles, since once the fluid in the vessel becomes opaque due to gas and particle dispersion, such optical techniques cannot be used. Radioactive particle tracking techniques developed at Washington University (Computer Automated Radioactive Particle Tracking CARPT [15]) and at the University of Birmingham (Positron Emission Particle Tracking, PEPT [16]) provided the advantage of being able to study fluid trajectories in opaque systems. Also, because the tracked particle is carried with the flow field, Lagrangian data are collected. In the study of Fishwick et al. [17] the liquid phase mixing in gassed vessels agitated by upand downward pumping 30° and 45° pitched-blade turbines was investigated using PEPT. The PEPT technique uses a radioisotope that decays by β + decay, which involves the emission of a positron. The emitted positron is annihilated by an electron, leading to the release of energy in the form of two back-to-back gamma rays, as shown in Fig. 6.9. The γ-rays are detected with two γ-ray cameras positioned either side of the vessel and their paths are reconstructed. From several such events the location of the tracer can be determined. A computer algorithm removes erroneous detections, allowing the tracer to be tracked within the vessel ~ 250 times per second within to 0.5 mm at the time of the experiment in 2007. A neutrally buoyant tracer was used based on an anion exchange resin labeled with 18F. Tracer particle

Emitted positron

Positron annihilated by electron

Energy released as two back-to-back 511 keV photons

Fig. 6.9: Formation of back-to-back γ-rays from an emitted positron. From several such events, the tracer can be located by triangulation. Reprinted from Fishwick et al. © [12], with permission from American Chemical Society.

The effect of gassing on mean liquid velocity fields is shown in Fig. 6.10, for four impellers. In Fig. 6.10 (a) i the flow pattern is typical for down-pumping pitched blade turbines, with the discharge at ~ 45° to the vertical. A primary circulation loop is formed in the lower part of the vessel, with a weak secondary circulation loop in the upper part, circulating clockwise. Below the impeller the liquid circulates upward into the

6.2 Slurry Reactors |

233

impeller to form another loop. Fig. 6.10 (a) (ii) shows the effect of gassing on the system, whereby the injection of gas at a specific flowrate of 1.5 vvm was studied. After the introduction of gas the flow field changed considerably, and the impeller discharge became essentially radial, suggesting that the effect of gassing had a more dominant effect than the impeller. In Fig. 6.10 (b), the effect of gas introduction had a weaker effect upon the flow pattern for the 30° PBTD. Fig. 6.10 (c) (i) presents the flow patterns for the 45° PBTU, showing a strong circulation loop in the flower part of the vessel near the impeller discharge. In the upper part of the vessel, a weaker secondary circulation loop was induced. The addition of gas was found to destabilize this loop and the upward flow of gas at the wall forced the eye of the secondary loop slightly inward towards the center of the vessel (Fig. 6.10 (c) (ii)). Similar results were obtained with the 30° PBTU (Fig. 6.10 (d) (i) and (ii)). In order to further interpret the data, a fluid circulation index was defined as the ratio of the average particle circulation velocity to the impeller tip speed: vm Ic = . (6.17) vtip

(a)(i)

(a)(ii)

(c)(i)

(c)(ii)

0.4 m.s-1

(b)(i)

(b)(ii)

(d)(i)

(d)(ii)

Fig. 6.10: Mean radial-axial velocity vector plots for (a) a 45° PBTD, (b) a 30° PBTD, (c) a 45° PBTU, and (d) a 30° PBTU ((i) without gassing and (ii) with a specific gas flow rate of 1.5 vvm) as measured using PEPT. Reprinted from Fishwick et al. © [12], with permission from American Chemical Society.

234 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions 0.18 0.17

Circulation index [–]

0.15

0.18

0.14 0.16

0.13 0.12

0.14 0.11 0.10

Mean tracer speed [m.s–1]

0.20

0.16

0.12

0.09 0.10

0.08 0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Gas flow rate [dm3.min–1] Fig. 6.11: Circulation index (mean tracer speed/impeller tip speed) at different volumetric gas flow rates at N = 483 min−1 for (◼) 45° PBTD, (󳵳) 30° PBTD, (◻) 45° PBTU, and (△) 30° PBTU. Reprinted from Fishwick et al. © [12], with permission from American Chemical Society.

Fig. 6.11 shows the circulation index for the four impellers as a function of the gas flow rate. The effect of gas flow rate has only a negligible effect upon the circulation index for the 30° PBTU, whilst the 30° PBTD showed only a slight decrease of circulation index with increased gassing. The 45° PBTU showed an initial decrease in circulation index with initial gassing, but this leveled out at gas flows above 0.4 dm3 min−1 . The most dramatic effect was observed upon the 45° PBTU, which showed a significant decrease of ~ 38 % with the injection of gas over the range of gas flow rates studied, indicating decreased momentum transfer capacity, being the poorest of the impellers studied in the gassed system. Electrical Resistance Tomography presents an alternative technique for the study of gassing of mixing vessels, which can be applied to plant scale systems [18]. The goal of electrical resistance tomography is to obtain the resistance distribution in the domain of interest. It can be obtained in a particular cross section by injecting currents on the domain and measuring voltages on it via a number of spaced electrodes, which are mounted non-invasively on its boundary. UMIST developed the necessary sensor, data acquisition and reconstruction systems, incorporating and 8-plane 16-electrode ring sensor which were applied to measurements on a 1.5 m-diameter pilot plant stirred tank. The system was used to study the impact of stirring on gas-liquid mixing with a Rushton turbine. The impeller created a strong radial outflow, giving a lower gas hold-up in the impeller plane. The gas hold-up immediately below the impeller was higher and there was also a higher gas hold-up in the liquid being drawn back towards the agitator in the upper part of the vessel. Some asymmetry in the gas hold-up was observed, resulting from the baffles of the tank

6.2 Slurry Reactors |

235

disrupting the tangential flow. The measurements were used to validate networks of zone models of stirred tank mixing.

6.2.3 Hydrogenation reactions in the stirred tank slurry reactor: kinetics and effect of operating variables The slurry reactor is often used for carrying out hydrogenation reactions, particularly for the case of fine chemicals and pharmaceuticals, where changes to product requirements and batch traceability are required. Mills and Chaudhari [19] reviewed the use of multiphase reactors in the pharmaceuticals and fine chemical production, wherein for gas-liquid-solid systems the reaction essentially occurs at the catalyst surface. These reaction mechanisms are often described using classical Langmuir-Hinshelwood reaction models. Such models consider the various reversible adsorption, formation of intermediates, reaction and desorption steps that could occur in the course of the reaction, together with the surface fractional coverage of the catalyst surface with each component. The rate-determining step is used in order to derive the overall reaction rate, as outlined by Froment [20]. Mills and Chaudhari [19] summarized some examples of kinetic rate equation models for gas-liquid-solid catalytic reactions that are relevant to fine chemicals and pharmaceuticals, a selection of which are given in Tab. 6.2. Tab. 6.2: Kinetic expressions for some typical hydrogenation reactions. Adapted from Mills et al. [21], with permission of Elsevier. Reaction System

Catalyst

Kinetic Model

Eq.

Hydrogenation of glucose [22]

Raney-Ni

Hydrogenation of o-nitroanisole [23].

Pd/C

Hydrogenation of 2-ethylhexenal [24].

Pd-SiO2 monolithic

(1 + √ K A A + K B B + K P P)

Hydrogenation of m-nitrochlorobenzene [25].

Pt-S/C

WkAB (1 + K B B)

Hydrogenation of 2,4dinitrotoluene [152].

Pd/Al2 O3

ri =

Hydrogenation of 1,5,9cyclododecatriene [153].

Pd/Al2 O3

ri =

Wk11 AB (1 + K A A) WkAB (1 + K A A)(1 + K B B)

(6.18) (6.19)

WkAB 3

kKH K i (PH /RT)0.5 C i [1 + KH (P/RT)0.5 ][1 + K i C i + ∑ji K j C j ] Wk i K j A α C j (1 + ∑3j=1 K j C j )

(6.20)

(6.21)

(6.22)

(6.23)

236 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

The expressions listed in Tab. 6.2 represent intrinsic kinetics; any associated mass transfer resistances also need to be evaluated in order to determine the actual rate of reactions. For reactions involving dissolved gas and liquid phase reactants and products, the complications of competitive adsorption, solvent effects and solubility of the reacting gas in the solvent should also be considered. In applications where more than one elementary step can be rate limiting, the equations in Tab. 6.2 are not valid and the development of a suitable kinetic model becomes a challenge. The early understanding of the mechanisms and kinetics of hydrogenation of alkynes and alkenes was largely contributed by research in the 1960s by Bond and Wells [26, 27] and Feidlin and Kaup [28], particularly in the case of ethyne and ethene. The hydrogenation of ethyne on Pd/Al2 O3 catalysts is a well-studied gas-solid reaction of industrial importance, as it is used for the removal of ethyne impurities from ethane prior to polymerization. The understanding of selectivity issues in ethyne hydrogenation requires an understanding of the adsorption models of ethyne and ethane upon the catalyst. Arafa and Webb [29] suggested that three types of active sites exist on the catalyst surface, including Type I sites which are active for ethyne hydrogenation to ethane; Type II sites which are active for direct hydrogenation of ethyne to ethane and Type III sites that are active for hydrogenation of ethane to ethane. A number of studies reported mechanisms of the reaction, incorporating different hydrogenation and oligomerization routes [30, 31]. Borodzinski and Cybulski [32] reported a model for ethyne hydrogenation that allowed for the reaction over a surface on which some carbonaceous deposits already exist. They assumed that there are three types of active sites, created on the palladium surface by carbonaceous deposits. Their model was novel in the consideration of the role in the kinetic analysis of (1) carbonaceous deposits creating a heterogeneous surface and (2) the transfer of hydrogen atoms from the carbon to hydrogenate ethyne. During ethyne hydrogenation a number of surface polymers may be formed [33], as well as butane and benzene [34, 35]. A known side product of ethyne hydrogenation is green oil, which is comprised of alkanes and alkenes of high molecular weight. The green oil can poison or block access to the ethyne hydrogenation sites of the catalyst. In more recent times, researchers have sought to understand whether the same effects occur in the hydrogenation of higher hydrocarbons [36]. Palladium is a highly selective catalyst for the hydrogenation of unsaturated hydrocarbons, offering high selectivity in alkyne over alkene hydrogenation and preferential cis/trans ratio of alkene products [37]. The hydrogenation of 1-pentyne and 2-pentyne over three different types of palladium catalysts was studied by Jackson et al. [36]. It was found that the internal triple bond hydrogenated faster than the terminal bond over a Pd/C catalyst. A support particle size effect was suggested. Boitiaux et al. [38] studied the effect of metal particle size and dispersion, and found that increasing metal dispersion above 20 % led to a reduction in the activity of but-1-yne hydrogenation. This resulted from a stronger complexation of alkyne to the metal over smaller particles, which tend to be electron deficient. Boitiaux et al. [39] reported

6.2 Slurry Reactors |

237

increased activity of supported palladium catalysts upon the addition of an electrondonating additive, such as piperidine in the hydrogenation of 1-butyne. The additive was proposed to modify the complexation strength of the alkyne with the metal, similar to the effect of a ligand in homogeneous catalysis. Increased alkene selectivity was observed as the rate of alkyne hydrogenation accelerated, but alkene hydrogenation decelerated. Competitive effects may occur in hydrogenation of multiple alkyne substrates, as investigated by Hamilton et al. [40] over a Pd/C catalyst. The reaction of pairs of 1-pentyne, phenyl acetylene, 2-pentyne and 1-phenyl-1-propene were studied and it was revealed that a lower hydrogenation rate typically occurred for both alkynes compared to either single reactant. However, a rate enhancement was observed for the 1-pentyne/2-pentyne system, which was through to be a result of enhanced hydrogen transfer on the surface. The hydrogenation of 1-pentyne over a range of palladium catalysts was studied by Teschner et al. [41], applying techniques such as catalytic pulse experiments, tapered element oscillating microbalance and catalytic pulse experiments. A carbonaceous Pd-C surface phase was found to build up in the early stages of the reaction, upon which the reaction subsequently occurred. Similar observations were made in the hydrogenation of C2 alkynes, where the reaction is well known to occur upon a layer of surface deposited carbon [26, 27]. Competitive effects were also studied in the hydrogenation of 1-pentene, cis-2pentene and trans-2-pentene over a 1 % Pd/alumina catalyst by Canning et al. [42]. Cis-2-pentene hydrogenated faster than 1-pentene, which in turn was faster than trans2-pentene. The rate of 1-pentene hydrogenation was greatly accelerated by 20-fold in a competitive system between alkene and alkyne, even after complete consumption of alkyne. Enhanced hydrogen transfer was believed to occur through a modified layer of hydrocarbon deposits upon the catalyst surface. Bennett et al. [43] studied the hydrogenation of 2-pentyne over Pd/Al2 O3 catalysts, with a view to understanding scale up and effect of operating variables in the stirred tank reactor. The research was also intended to contribute towards gaps in knowledge of the hydrogenation mechanisms and reactivity for > C4 alkynes in the liquid phase. The reactions were carried out in a 2.65 l baffled stirred vessel, using 1 wt % Pd/Al2 O3 catalyst. The effects upon rate and selectivity of stirring speed, catalyst loading, substrate concentration, hydrogen pressure, solvent type and catalyst preparation method were investigated. Increasing the stirring speed from 445 rpm to 1100 rpm was found to lead to a substantial increase in reaction rate from 2.49 × 106 mol dm−3 s−1 to 4.38 mol dm−3 s−1 . The effect of stirring speed upon gas-liquid mass transfer was investigated by calculating the value of k l a from the correlation of Kawase et al. [7] (Eq. (6.13)) and liquid-solid mass transfer coefficient was calculated from the Frössling equation. It was found that the gas-liquid mass transfer coefficient k l a increased from a value of 7.93 × 10−4 to 4.63 × 10−3 s−1 , by a factor of 5.8 upon increasing the stirring speed, whereas the solid-liquid mass transfer coefficient, k c , increased from

238 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions 4.89 × 10−4 to 9.04 × 10−4 m s−1 , by a factor of 1.84. Calculation of the rates of hydrogen mass transfer indicated that the rate based upon liquid-solid mass transfer was close (within 12 %) to the observed reaction rate; therefore, it was assumed that the reaction was under external mass transfer control at the catalyst particle. Owing to the likeliness of external diffusion control, Bennett et al. [43] found that using a reduced size of the catalyst support particle gave improved reaction rate. Particles with diameter in the range 45–75 μm rather than ~ 100 μm led to a substantial increase in reaction rate at a stirring speed of 1100 rpm, by a factor of ~ 10.3. The increase of reaction rate was greater than would be expected from the increase in liquid-solid mass transfer alone, although it should be noted that, with a given amount of catalyst, using particles of smaller size also gives a large surface area and thus the value of the specific surface area of the catalyst, a c , also increases. The combined effect of liquid-solid mass transfer coefficient and increased surface area of particles could explain the observed increase in reaction rate. Pre-treatment of a catalyst by reduction in hydrogen is commonly used to activate the catalyst by converting the catalytic metals from the oxide to metallic form. However, Bennett et al. [43] studied the effect of pre-treatment with hydrogen upon the product profiles during the hydrogenation reaction of 2-pentyne. It was found that when the catalyst is not pre-reduced prior to the addition of the reagents, the selectivity of the reaction is altered significantly. The rates of alkyne consumption and pentane formation are similar to those found over the pre-reduced catalyst, but the trans-alkene is formed to a much greater extent when the catalyst is not pre-reduced. This effect was explained, since when the catalyst is pre-reduced under hydrogen, the plentiful availability of hydrogen chemisorbed on the catalyst surface leads to the main reaction being hydrogenation to form cis-2-pentene. When the catalyst is not prereduced and has a lower coverage of hydrogen, isomerization reactions are favored over hydrogenations, and the observed production of trans-2-pentene increases. On the pre-reduced catalyst, the reaction is thought to involve the formation of a carbonaceous overlayer; but when the catalyst is not pre-reduced, the surface consists of Pd metal or oxide.

6.2.4 Catalysts for hydrogenation reactions: overview and novel biomass supported metal catalysts A variety of catalyst preparation techniques are available, depending upon the required properties of the finished catalyst. The properties of the heterogeneous catalyst are influenced strongly by each step of the preparation method and the quality of the raw materials. Perego and Villa [44] have summarized the unit operations involved in catalyst preparation, as shown in Tab. 6.3.

6.2 Slurry Reactors |

239

Tab. 6.3: Typical steps involved in catalyst preparation. Reprinted from Perego and Villa © [44], with permission of Elsevier. 1. 2. 3. 4. 5. 6.

Precipitation Gelation Hydrothermal transformation Decantation, filtration, centrifugation Washing Drying

7. 8. 9. 10. 11. 12.

Calcination Forming operation Impregnation Crushing and grinding Mixing Activation

Most catalyst formulations comprise a combination of some or all of the above operations. Whilst methods of preparation vary widely, three broad classifications of catalyst can be made according to their preparation procedure, namely: (1) Bulk catalysts and supports; (2) Impregnated catalyst; and (3) Mixed-agglomerated catalysts. Bulk catalysts are mainly manufactured from active substances, for example the use of silica-alumina for catalytic cracking reactions. Impregnated catalysts are obtained by impregnating a pre-formed support with the active phase, with the method being commonly used in the preparation of hydrotreatment catalysts. Mixed-agglomerated catalysts can be manufactured by mixing a powdered support with the active substance and combining them by agglomeration, as is quite often performed in the preparation of hydrotreatment catalysts. Bennett et al. [43] studied the effect of catalyst preparation technique upon the product profiles and rates of 2-pentyne hydrogenation. They used a 1 wt % Pd/Al2 O3 catalyst supplied by Johnson Matthey. Catalyst Type A consisted of θ alumina particles that were ground to a powder before impregnation with palladium, whilst catalyst Type B was comprised of the same support, but ground after impregnation. It was found that catalyst Type B led to a much higher rate of reaction owing to a higher metal dispersion of palladium. Chemisorption measurements revealed that Pd on catalyst Type A had an active particle diameter of 8.1 nm and dispersion of 13.8 % whilst on catalyst Type B the active particle diameter was 1.6 nm and metal dispersion was 68.4 %. Solvent selection was shown to strongly influence the reaction rate and selectivity in heptane, isopropanol and 50/50 mixtures of these solvents. The reaction in heptane was much faster than in isopropanol, which was through to be due to enhanced substrate solubility. A mixed solvent consisting of 50/50 heptane/isopropanol was found to lead to intermediate rates to those of the pure solvents, but selectivity was superior, with cis/trans and alkene/alkane selectivities much higher than the pure solvents. The effect was further investigated by carrying out adsorption experiments of the reagents in the different solvents. These indicated that the 50/50 heptane/ isopropanol gave a higher degree of adsorption of alkyne on the catalyst than in either of the pure solvents. This can help to explain the high alkene/alkane selectivity, since a high coverage of the surface with alkyne is less likely to lead to re-adsorption of free alkene and further hydrogenation. Also, a strong alkyne adsorption may lead to lower

240 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

concentration of adsorbed hydrogen on the surface, and may slow the hydrogenation rate whilst improving the selectivity. The goal of increasing reaction rate, selectivity and catalyst lifetime drives the requirement for further research into new catalytic materials. Almost monodisperse metal nanoparticles of less than 100 nm size have considerable potential as novel and highly active and selective catalysts [45]. Since these particles have a much higher surface-to-volume ratio than their bulk counterparts, they have a larger fraction of catalytically active atoms on their surface. These atoms are not ordered in the same manner as those in the bulk metal and, since the electrons in nanoparticles are confined in small spaces of a few atoms width, they can give rise to quantum size effects. The methods of synthesizing metal nanoparticles also provide greater control over the size and surface composition of the particles than bulk scale catalysts. Zahmakiran and Ozkar [45] noted that there are two general approaches to the preparation of metal nanoparticles, amongst numerous detailed procedures, these being “topdown” and “bottom-up” approaches. Large particles of bulk metal are disintegrated into nanoparticles using thermal, chemical or mechanical methods in the “top-down” approach. The generation of metal atoms from a metal salt or precursor is followed by agglomeration into nanoparticles in the more common and practical “bottom up” approach. A challenging issue in the synthesis of metal nanoparticles is achieving well-defined shape and close control of particle size, since these parameters control the surface structure, electronic and oxidation states of the particle and thus strongly impact upon the catalytic rate and selectivity. Classical methods of preparing supported metal nanoparticles include ion-exchange followed by chemical reduction or thermal reduction, co-precipitation, deposition-precipitation, impregnation, electrochemical and photochemical methods. Chemical vapor deposition and atomic layer deposition are gas phase techniques used to encapsulate metal NPs within various types of porous materials. Catalytic nanoparticles find applications in aqueous, organic, biphasic and ionic liquid media. The use of palladium nanoparticles in hydrogenation and other organic syntheses has been summarized in a number of review articles [46, 47]. Creamer et al. [48] demonstrated the use of bacterial cells as a catalyst support using various bacterial strains, including Desulfovibrio Desulfuricans. The ability of some bacteria to reduce metallic ions has been applied to the recovery of previous metals (Pd(II), Pt(IV), Au(III)) from pure solutions and reprocessing wastes. During bioreduction of metal ions from solution, a layer of metallic nanoparticles becomes deposited upon the bacterial cell [49]. For palladium, a high catalytic activity is displayed by biogenic Pd(0) nanoparticles; ‘bioPd’, as it has become known, is effective in a range of reactions including the hydrogenation of itaconic acid, 2-butyne-1,4-diol, 2-pentyne, dehalogenation of flame retardant materials and Heck coupling reactions [48–53]. The catalyst preparation involves growing the bacterial culture, harvesting by centrifugation and washing the live cells. They are then treated with an acidified solution of sodium tetrachloropalladate (pH 2–3) prepared to the desired Pd loading.

6.2 Slurry Reactors |

(a)

241

(b)

Fig. 6.12: STEM images of D. desulfuricans bacteria (a) before and (b) after impregnation with palladium metal. Reprinted from Bennett et al. © [54], with permission from Elsevier.

The cells and solution are stirred under nitrogen followed by hydrogen flows, the latter acting as a reducing environment. The resulting material is recovered washed and dried to produce a powdered catalyst, supported upon dead bacterial biomass. A TEM micrograph showing the bacterial cell with deposits of Pd in the cell membrane is shown in Fig. 6.12. Studies indicated that hydrogenase enzymes near the surface of the bacterial cell are associated with the formation of nanoparticles. They are formed within the peptidoglycan matrix (which contains a sugar-amino acid polymer), located between the two layers of the cell membrane. Particle size may be regulated in the preparation by the ratio of mass of cells to palladium salt. The formed nanoparticles are immobilized and protected from dissolving by becoming entangled within the peptidoglycan matrix. Biomass has a number of attractive features for use as a catalyst support. The precious metals could be obtained from waste sources, including spent car catalytic converters or electronic scrap, since an acidic leachate is required in the preparation process. After the useful lifetime of the catalyst, the metal can be easily recycled by microwaving, sonication or incineration of the biomass. Bennett et al. [54] sought to demonstrate the use of bioPd catalyst in a multi-product reaction, the hydrogenation of 2-pentyne, and to study reaction rate and selectivity obtained using this type of catalyst in the stirred tank reactor. The catalyst used was 5 wt % Pd loaded upon Desulfovibrio Desulfuricans. The metal particles were found to be evenly dispersed in the cell wall with a fairly even particle size of ~ 1.7 nm. Reaction rates and selectivity were also measured over a conventional Pd/Al2 O3 catalyst for the purposes of comparison. Fig. 6.13 shows the reactant and product profiles over the 5 wt % bioPd catalyst. The 5 wt % bioPd catalyst was found to give a slower initial rate of reaction in isopropanol solvent, with only 30 % of the rate observed over a 5 wt % Pd/Al2 O3 catalyst. However, a high selectivity was displayed by the biomass-supported catalyst, with pentene/pentane ratios in the range of 8–14 and cis-trans product ratios in the

242 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

100 Trans-2-pentene Pentene 1-Pentene Cis-2-pentene 2-pentyne

% Component

80 60 40 20 0 0

1

2 3 Time (hours)

4

5

Fig. 6.13: Concentration profile for 2-pentyne hydrogenation in isopropanol using 5 wt % bio-Pd as a catalyst and a stirring speed of 445 rpm. The bacterial strain used in the catalyst preparation was D. desulfuricans. Reprinted from Bennett et al. © [54], with permission from Elsevier.

range of 6–14. Fig. 6.14 shows the pentene/pentane and cis/trans selectivity ratios over 5 wt % bioPd, 1 wt % and 5 wt % Pd/Al2 O3 in isopropanol solvent. The conventional 5 % Pd/Al2 O3 catalyst gave very high pentene/pentane and cis/trans ratios at low conversions of 2-pentyne, but these fell sharply as the reaction proceeded and became slightly lower than those observed with the bioPd at alkyne conversions greater than 70 %. At high alkyne conversion of 92 %, the 5 wt % Pd/Al2 O3 gave a cis/trans ratio of 2.0 and pentene/pentane selectivity of 2.0, whereas the 5 wt % bioPd gave respective selectivity values of 2.5 and 3.3. Although the reaction rate could not match the Pd/Al2 O3 catalyst, it had been demonstrated that bioPd could be applied with favorable selectivity in a multi-product hydrogenation reaction. It was also demonstrated that the bioPd could be separated from the product mixture and recycled for use in a subsequent hydrogenation, in which it remained active. The concept of rational design of catalyst surfaces may be extended to cover engineering of active site distribution to try to achieve the desired product selectivity in a particular reaction. Recent work showed that selectivity control can be achieved for heterogeneous catalysts by control of metal particle size and topography. Particular crystal facets upon a metal surface may influence the reaction behavior; for example, in partial hydrogenation of alkynes, it is known that alkyne hydrogenation is favored on terraces, while alkene hydrogenation or isomerization is favored on surface defect sites, such as corners or edges, as illustrated in Fig. 6.15 [55]. Schmidt et al. [56] reported differences in the catalytic activity of equally sized nanoparticles of platinum with different shapes, namely cubic, cubooctahedral and octahedral. In the absence of a chiral modifier, each of the three types of NPs gave similar activity and selectivity in the racemic hydrogenation of ethyle pyruvate. However the addition of a chiral modifier such as cinchonidine or quinine were able to give different effects depend-

6.2 Slurry Reactors |

243

30 5% bio-Pd in IPA 1% Pd/AI2O3 in IPA 5% Pd/AI2O3 in IPA

Pentene/pentane

25 20 15 10 5 0 0

20

(a)

40 60 Pentyne conversion (%)

80

100

12 5% bio-Pd in IPA 1% Pd/AI2O3 in IPA 5% Pd/AI2O3 in IPA

10

Cis/trans

8 6 4 2 0 0 (b)

20

40 60 Pentyne conversion (%)

80

100

Fig. 6.14: (a) Pentene/pentane versus 2-pentyne conversion for hydrogenations catalyzed by 5 wt % bio-Pd and 1 % and 5 % Pd/Al2 O3 in isopropanol. (b) Cis/trans-2-pentene ratio versus 2-pentyne conversion for hydrogenations catalysed by 5 wt % bio-Pd and 1 % and 5 % Pd/Al2 O3 in isopropanol. Reprinted from Bennett et al. © [54], with permission from Elsevier.

ing upon the NP shape, with the reaction being observed to be faster with increasing ratio of Pt{111]/Pt{100}. The cinchonidine modifier was found to adsorb more strongly to certain types of active site (Pt{100}) than other sites (Pt{111} face), leading to hydrogenation and degradation of the modifier. Therefore an ideal catalyst for activated ketone hydrogenation should contain mainly Pt{111} terraces, since this type of crystallographic face gives rise to higher enantioselectivity with higher stability of the modifier. Bennett et al. [50] carried out a study of selectivity control of Pt catalysts in 2-butyne-1,4-diol hydrogenation using two routes: (1) In conventional Pt/C catalysts bismuth was used as a site blocking agent, which was able to selectively cover step sites. (2) In bioPt catalysts prepared using Eschericihia coli, residual molecular frag-

244 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions Pd(100)

Low coordination (edge) sites Pd(111) Pd(111)

Pd(100) Fig. 6.15: Illustration of the crystallographic facets of palladium, showing edge and terrace sites (Pd(100) and Pd(111)). Adapted from Crespo-Quesada et al. © [55], with permission of American Chemical Society.

Butenediol selectivity

1.0 0.8 0.6 0.4 Pt/C unsintered bioPt chemically processed bioPt Pt/C unsintered + 0.5 monolayers Bi

0.2 0 0

20

40 60 Alkyne conversion (%)

80

100

Fig. 6.16: Butenediol selectivity versus alkyne conversion using four different platinum catalysts with increasing occupation/blockage of defect sites. Reprinted from Attard et al. © [52], with permission from American Chemical Society.

ments, left over after chemical cleaning and subsequent separation from the bacterial support were used as a site blocker, being found to accumulate preferentially at crystal defect sites. In the second case, the bioPt is prepared by depositing the Pt on the bacterial cell, followed by the usual drying and grinding, then chemical processing, to remove most of the biomass, particularly from terrace crystal sites. Selective poisoning of a catalyst achieved using residual biomass, rather than irorganic counterpart chemical modifiers such as lead or bismuth, could be an advantage because of reduced toxicity. Fig. 6.16 illustrates the butenediol selectivity as a function of alkene conversion for both the Pt/C system with and without Bi poison and the BioPt catalyst before and after chemical cleaning. Bi-poisoned 5 % Pt on graphite gave butenediol

6.2 Slurry Reactors |

245

selectivity greater than 90 % at low alkyne conversion below 20 %, but only a low rate of reaction (20 % of the alkyne converted after 2 hours). The butenediol selectivity with chemically processed bioPt was of the order of 70 % but with a faster rate (45 % of the alkyne converted after 2 hours). The butene diol selectivity of the Pd/C treated with Bi (~ 90 %) was substantially higher than Pd/C without modifier (~ 55 %), as was the selectivity over the chemically processed bioPt (~ 70 %) compared with the unmodified bioPt (~ 50 %). It was suggested that biogenic catalysts may function similarly to Lindlar catalysts, which are often used in the selective hydrogenation of alkynes, where a poison of the metal sites is used, but the biogenic catalyst avoid the potential risks of using toxic heavy metals such as lead. Related work by Attard et al. [52] showed that the use of defect site blocking modifiers such as bismuth or polyvinylpyrrolidone (PVP) could be used to effect large increases in selectivity to the semi-hydrogenation product during the hydrogenation of alkynes. This was again a result of selective site blocking, in particular covering defect sites to leave behind mainly Pt{111} terrace sites. It was found that alkynes strongly associate with defect sites to produce a durable surface complex that allows for over-hydrogenation of the intermediate to produce an alkane. Covering the defect sites using bismuth or PVP was found to eliminate the defect sites associated with over-hydrogenation and thus a substantial increase in selectivity towards the alkene product was observed. The results were corroborated by advanced spectroscopic data obtained using shell-isolated Raman spectroscopy (SHINERS) and electrochemical data obtained by cyclic voltammetry.

6.2.5 Oxidation reactions in the slurry reactor Over the last twenty years, there has been increasing interest in the development of catalytic oxidation processes for the production of bulk and fine chemicals. Catalytic oxidation covers a vast range of applications, with some of these being reviewed by Mills and Chaudhari [19]. Some examples of liquid phase oxidation in industry include oxidation of p-xylene to terephthalic acid, cyclohexane to adipic acid, n-butane to acetic acid, epoxidation of propylene to propylene oxide and hydroxylation of phenol to hydroquinone and catechol, as well as a range of catalytic oxidation processes for the treatment of organic and inorganic wastes. The traditional approach in oxidation reactions was to use stoichiometric oxidizing agents such as permanganate, dichromate, manganese dioxide and nitric acid, which are toxic and create large volumes of waste material. Therefore, catalytic oxidation has provided an attractive alternative to these processes, driving the development of new catalysts capable of operating under milder conditions with a minimal amount of undesirable by-products being formed. The multiphase character of working with liquid phase reactants, products and air or oxygen poses similar challenges to the design of hydrogenation reactors, in terms of mixing, mass transfer and kinetic processes. Safety must be paid particular attention due to the constraint in limits of operation with regard to oxygen and organic

246 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

concentrations employed. Oxidation processes are distinguished by a number of features, which include the use of critical oxygen concentration, below the flammability limit of the oxygen-organic mixture; this may be a limiting factor in terms of the reaction rate achieved. However, the use of molecular oxygen via air or enriched air is relatively cheap and much less toxic than strong stoichiometric oxidizing agents. Heterogeneous catalysts can efficiently use molecular oxygen to produce only water as a side-product. This proves to be cost efficient because supported catalysts can be recovered and recycled, or can be used in continuous flow reactors [57]. Control of reaction temperature is another significant consideration, since many oxidation reactions are strongly exothermic. Since the mid-1990s, much attention has been focused on the use of supported gold catalysts for oxidation reactions [58]. The breakthrough work of Haruta et al. [59] showed that nanocrystals of Au supported on oxides with a very effective catalyst for CO oxidation at low temperature, whilst Hutchings and Grady [60] predicted – and later the group of Hutchings verified [61] – that cationic gold would be an effective catalyst for the hydrochlorination of acetylene. Since then, research on gold catalysis has been greatly increased, and gold catalysts developed encompass reactions such as the oxidation of glycerol, cyclohexane and alcohols [58]. Heterogeneous catalysts for oxidation reactions are attracting the interest of industry and academics; one particular class of such reactions being the oxidation of alcohols to carbonyl compounds [62–64]. The acknowledged reaction mechanism for platinum group metals is the loss of two hydrogen atoms from alcohol, forming the carbonyl compound via a dehydrogenation pathway before the adsorbed hydrogen reacts with molecular oxygen to form water [62, 65]. A crucial issue concerns the best catalyst selection and design, and how to ensure the catalyst has a long lifetime without significant deactivation problems. PtBi/Carbon is also a promising catalyst for alcohol oxidations, since bismuth acts as a promoter of selectivity for the alcohol oxidation [66, 67]. However, the catalyst can be deactivated through various processes such as over-oxidation, leaching or agglomeration of metal and by-product poisoning [63, 65]. The strong adsorption of an excess of oxygen upon the metal surface occurs in overoxidation, preventing the alcohol from adsorbing on the active sites. Leaching of the active metal into solution may occur, depending on the substrate used, whilst sintering leads to the migration of small particles to form large agglomerates with decreased surface area and activity [66, 68]. Component inhibition may occur when the reactant, product or by-products adsorb on the metal surface leading to blockage of the active sites [69, 70]. This can occur is the carbonyl product strongly adsorbs on the metal. The oxidation of alcohols is an important class of oxidation reactions, although different alcohols such as vinyl, allylic, aromatic and aliphatic display differing levels of reactivity over noble metal catalysts [64]. Aliphatic alcohols are known to be particularly resistant to oxidation [71, 72]. Mounzer et al. [73] investigated the heterogeneous oxidation of 2-octanol on 5 wt % Pt-1 wt % Bi/Carbon catalyst as a case study. Reactions were carried out in a baffled stirred autoclave of 500 ml volume, whilst studies of ke-

6.2 Slurry Reactors |

247

2-octanone adsorption coeffecient (1/(mol.L))

0.8

0.08 K4 adsorption coefficient Reaction rate

0.7

0.07 0.06

0.6

0.05 0.04

0.5

0.03 0.4

0.02

0.3 0

20

40 60 % v/v Dioxane

80

0.01 100

Reaction rate after 15 minutes (M/hr)

tone adsorption upon the catalyst were carried out in a laboratory scale glass flask fitted with a condenser. Oxidation of 2-octanol was carried out at atmospheric pressure with air as an oxidant, temperature at 343 K and using solvents heptane, p-xylene and dioxane and mixtures of heptane-dioxane. Using pure solvents such as heptane and p-xylene, the reaction rates were shown to start quite fast initially, but rapidly decrease shortly after the start of the reaction due to poisoning by product adsorption. It was shown that pre-treatment of the catalyst surface with ketone also led to a deactivation of the catalyst, demonstrating that this effect could also occur during the reaction due to product formation. To decrease the amount of ketone adsorption, a solvent was sought with greater solubility of 2-octanone to enhance ketone desorption and leave more active sites available for reaction. Various solvents including DMSO, p-xylene, dioxane and dioxane/heptane mixtures were investigated for this purpose and adsorption studies carried out to determine the equilibrium uptake of 2-octanone in each solvent or mixture. Fig. 6.17 displays the effects of heptane-dioxane solvent mixture composition on the reaction rate and ketone adsorption coefficient. The lowest amount of ketone adsorption occurred at a 16–18 % v/v mixture of dioxane in heptane, and corresponded to the maximum observed rate of oxidation. The effect of using mixed solvents could influence oxygen solubility in the liquid phase, although the effect of pressure in the range 1–4 bar upon the reaction showed a weak effect of oxygen solubility upon the reaction rate. Therefore, it was concluded that the maximum reaction rate occurred because the ketone was removed from the catalyst surface by the corresponding particular solvent composition. The use of mixed solvents may be used to tune the product adsorption properties, with 2-octanone having a long carbon chain resulting in a lipophilic character but also the carbonyl bond exhibiting polarity that is responsible for bonding with the metal. The experimental rates were fitted by

Fig. 6.17: Effects of heptane–dioxane solvent composition upon reaction rate and ketone adsorption coefficient. Reaction conditions: T = 343.15 K; PAir = 1 bar; C i = 0.248 M; stirring speed = 22.5 s−1 . Reprinted from Mounzer et al. © [73], with permission from Elsevier.

248 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

a Langmuir-Hinshelwood expression, which include a ketone adsorption term in the denominator: k3 K1 [A](K2 [O2 ])0.5 R= , (6.24) 2 (1 + K1 [A] + (K2 [O2 ])0.5 + [P] ) K4 where k3 , K1 , K2 , and K4 are the reaction rate, alcohol adsorption, oxygen adsorption and product adsorption coefficients respectively. [A] is the alcohol concentration and [P] is the product concentration. Although particular to a case study of 2-octanol oxidation, the results illustrate the complex interplay between kinetics, product adsorption and catalyst deactivation that may occur in selective oxidations.

6.2.6 Bubble column reactors Bubble column reactors are a class of slurry reactor in which a gas is bubbled through a column of liquid where it reacts via a chemical or biochemical reaction, typically in the presence of a suspended catalyst. Normally no agitator is used, which offers the advantage of no moving parts, together with good heat and mass transfer, ease of operation and low operating and maintenance costs. However, backmixing can occur, with the expense of a reduction in product conversion. A high superficial gas velocity (up to 50 cm/s) is used to impart momentum transfer to the slower moving liquid (up to 2 cm/s) to promote mixing and circulation patterns [74]. Bubble columns find application in a range of processes for hydrogenation and oxidation, and are also used for hydroformylation, chlorination, cell growth and bioremediation. Industrial applications include partial oxidation of ethylene to acetaldehyde, liquid phase methanol synthesis, wet air oxidation, Fischer-Tropsch synthesis, maleic acid hydrogenation and a range of biochemical processes including cultivation of bacteria and various cell cultures, also the treatment of sewage. However, bubble columns can be difficult to scale up, since the extrapolation of laboratory scale data to pilot and industrial scales requires similarity criteria that would achieve similar mixing, hydrodynamic performance and thus match conversion and selectivity. Shaik and Al-Dahhan [74] have provided a recent detailed review of the state-of-the-art in scale up methods for bubble column reactors, based on extensive experimental and computation studies. Advances in the understanding of fluid dynamics in the bubble column have resulted from novel measurements and computation modeling efforts. A number of techniques including hot-wire anemometry, PIV and LDA have been applied to the measurement of velocity and turbulent stresses in the bubble column [75]. However, optical techniques can only be used for low solids loading and gas fraction; therefore, non-invasive imaging techniques for opaque vessels are required. Devanathan et al. [15] and Dudukovic et al. [75] introduced radioactive particle tracking studies to investigate liquid motion in bubble columns. The technique produces Lagrangian data for the whole column, similar to the PEPT technique reviewed in Section 6.2.2. The instantaneous velocities, time averaged flow patterns, turbulent

6.2 Slurry Reactors |

249

stresses and turbulent kinetic energy can be derived from the tracer particle trajectories. CARPT is used to measure the position of a single radioactive particle using a series of scintillation detectors. In order to measure the motion of catalyst particles in slurries or fluidized beds, a tracer particle of equal density to the particles in the system is used, or if measuring the fluid velocity, a neutrally buoyant particle is used. The system is able to measure motion up to frequencies of 20–30 Hz. Computer tomography (CT) was also used to obtain time averages gas holdup profiles in the column cross sections over a range of elevations. The combination of CARPT-CT provided the time averaged flow field and gas holdup distribution in bubble columns. Using the technique, the radial gas holdup profile was observed to be almost flat, with slightly more gas near the center at low gas superficial velocities, whilst the gas holdup profile became almost parabolic in churn turbulent flow. Velocity maps recorded at U g = 2.4 cm/s, in a 14 cm diameter column [76] showed a single-cell flow circulation pattern, as had also been earlier observed by Devanathan et al. [15]. The flow pattern was observed to be quite symmetric about the column axis. Visual observation of the flow in the column indicated that gas bubbles tend to form “swarms” that follow a spiraling motion, rocking from one side of the wall to the other and becoming gradually more evenly distributed at higher levels in the column. Computational fluid dynamics (CFD) has also been used for the modeling of bubble columns. A review of modeling of bubble column reactors is provided by Jakobsen et al. [77]. Two approaches are commonly used: the Euler-Euler formulation and the Lagrange-Euler approach. The Eulerian description of the fluid flow is based on the concept of pseudo-continuum and thus defines a point volume fraction for each of the phases that is representative of the probability of that phase to be present at that particular point in multiple realizations of flow. Each of the phases shares the same pressure field. The force interactions between the phases are allowed by using various effective “volumetric” force functions including drag force, lift and added mass force [78]. Such a model was formulated and solved by Gupta and Roy [78] and validated using data from radioactive particle tracking. Very good agreement was observed for velocity profiles at a high superficial gas velocity of 2.38 mm/s with the selection of suitable drag (Schiller Nauman) and turbulence (RNG k-ε) model. The Lagrange-Euler method involves the solution of the Navier-Stokes equation for the continuous phase and solution for the motion of each of the bubbles by the application of Newton’s second law, with the various forces upon the bubble being calculated using the local velocity patterns of the continuous phase. The density and viscosity of the continuous phase are often modified to allow for the low volume fraction of the dispersed phase (bubbles). The two approaches have various advantages and disadvantages, for example the Lagrange-Euler model requires some additional tuning parameters, such as effective diffusivity for the dispersed phase and effective viscosity for the continuous phase [79]. A remaining challenge in the modeling of bubble column reactors concerns the treatment of bubble coalescence and breakage processes, although population balance modeling is starting to be used to represent these phenomena [77].

250 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

Many industrial bubble column reactors work in upflow mode, with gas and sometimes liquid (cocurrent operation) being introduced via a sparger at the base of the reactor. In countercurrent operation, the liquid would be introduced at the top of the reactor and flow towards the bottom. The Cocurrent Downflow Contactor Reactor (CDCR) is an exception to the aforementioned operation, in which both the gas and liquid stream are introduced concurrently via an orifice and entry zone at the top of a fully flooded column. The device was developed at the University of Birmingham by Boyes et al. [80]. The CDCR was found to give high gas-liquid mass transfer coefficients (k la ) via high interfacial area (1000–1500 m2 /m3 ) and gas hold up (0.5–0.6), thus making it suitable for application in catalytic hydrogenation reactions, where is it possible to overcome the hydrogenation mass transfer resistance and make the most effective use of the catalyst [81]. In the upper section of the column, a gas-liquid dispersion is formed by a high velocity liquid jet inlet stream, which prevents a gas pocket from forming. The dispersion is comprised of almost uniformly sized bubbles depending on the liquid and gas used (e.g. H2 /H2 O 3–4 mm). Further advantages of the CDCR include: low power consumption, small containment volume, easy scale up, 100 % gas utilization with operation close to equilibrium, inherent reliability due to no moving parts and tolerance to particles, therefore suitable for use as a slurry reactor [80]. Increasingly stringent regulations regarding wastewater treatment led to the need to develop innovative and more efficient wastewater treatment technologies, required to decompose toxic industrial effluents and to improve discharge water quality. The CDCR has been demonstrated to be an effective technology for wastewater degradation. The use of Advanced Oxidation Processes (AOPs) includes photocatalysis and oxidation to treat organic pollutants. Key to the effectiveness of AOPs is the formation of hydroxyl radicals (HO · ), which are a powerful oxidizing species for organic compounds in the aqueous phase, with stronger oxidising power than that achievable using single oxidizing chemicals such as chlorine, hydrogen peroxide and ozone [82]. A number of different techniques can be used to generate hydroxyl radicals, with each method involving the use of an oxidant (H2 O2 , O3 , O2 ), together with an activating system such as a catalyst, UV light, alkali or another oxidizing species. In photocatalysis, a photocatalytic semiconductor, such as TiO2 , is activated by radiation of suitable energy (hv) and wavelength to promote the excitation of electrons, which upon acceptance by oxygen on the semiconductor surface yield unstable superoxide radical ions (O2· – ) and further yielding hydroxyl radicals HO · . Winterbottom et al. [82] gave further details of the photocatalytic mechanism. Photocatalysis was investigated for the destruction of phenol using the CDCR by Winterbottom et al. [82]. Allowing for all the possible reactions, the general reaction schemes for phenol destruction by photocatalysis could be represented as: C6 H5 OH + · OH 󳨀󳨀→ intermediates 󳨀󳨀→ CO2 + H2 O; and C6 H5 OH + HO2· 󳨀󳨀→ intermediates 󳨀󳨀→ CO2 + H2 O.

6.2 Slurry Reactors |

251

In order to use the CDCR as a photocatalytic reactor, UV lamps were fitted into the base section, with powers of either 30 W or 1.0 kW. The reaction of phenol was carried out at temperatures of 40–50 °C and 202.6 kPa, with operation of the reactor under a closed loop recycle mode with recirculation of the suspended TiO2 catalyst. Complete degradation of phenol was achieved in as little as 30 minutes reaction time using O2 /UV/TiO2 with a pH of 7.0 or below using the higher powered lamp from solutions containing 100 mg/dm3 phenol. An important class of chemicals of environmental concern include nitrogencontaining compounds which are resistant to biological treatment and thus persist after passing through conventional sewage treatment works. These compounds are often encountered in pharmaceutical effluents, generally at microconcentrations. European Community regulations have stipulated discharge limits for a number of such pollutants [83]. 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) is a recalcitrant compound present in waste water produced from pharmaceutical processes, consisting of a nitrogen-containing tertiary amine that was tested for degradation in a CDCR operating as photocatalytic reactor. The effects of photocatalyst loading, initial reactant concentration, temperature, pH and various combinations of AOPs including UV, O2 , H2 O2 , and TiO2 upon photocatalytic degradation of DBU were investigated. A slurried granular catalyst (Degussa VP Aeroperl P25/20) was selected for ease of filtration from the treated water. The reactions were carried out at 40–60 °C and 1 barg using a 1.0 kW UV lamp mounted in the base of the reactor, and operation of the reactor in closed loop recirculation mode. It was found that a combination of TiO2 , UV radiation, and O2 gave the most rapid degradation and mineralization of DBU in comparison with other combinations of AOPs. Optimization of the reaction conditions led to 100 % degradation of DBU in 45 min, using a 1 kW lamp, 0.5 g/dm3 TiO2 , 100 mg/dm3 DBU, 1 barg, 50 °C, and pH of 3.17. Fig. 6.18 (a) displays the reactor set up, in which a slurry catalyst may be optionally used in place of the reticulated foam monolith, and Fig. 6.18 (b) shows the degradation profiles of DBU with the slurry catalyst. The formation of intermediates in the degradation process was considered, with use of GC-MS to detect the various compounds occurring during oxidation. It should be noted that some of the intermediates formed during photocatalytic degradation of wastes could be more hazardous than the starting material, so it is important to measure the destruction of total organic carbon (TOC), rather than just the starting compound. A first order reaction model was able to describe the degradation process. Ochuma et al. [84] also carried out a similar optimization study for the destruction of 2,4,6-trichlorophenol as an alternative model pollutant in industrial wastewater. A potential drawback of using slurry catalysts is the necessity for installation of a filtration system to separate the catalyst from treated wastewater. Ochuma et al. [85] demonstrated the application of a reticulated foam monolith catalyst, comprised of TiO2 -coated alumina, installed in the annular space between the centrally installed UV lamp and internal wall of the reactor (Fig. 6.18 (a)). The oxidation of the model pharmaceutical waste DBU was again tested for evaluation of the reactor performance. The results indicated that a

252 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions Liquid input High velocity gas-liquid jet

O2 input Turbulent gas-liquid mixing zone

Reservoir with cooling coil UV lamp and pyrex sheath

Pump (a)

Treated effluent

Photocatalytic reaction zone Reticulated foam monolith Lamp power source

Conversion [%]

100.00 80.00 60.00 40.00 20.00

DBU conversion TOC conversion

0.00 0 (b)

50 100 Reactor operating time [min]

150

Fig. 6.18: (a) Experimental set-up of the photocatalytic CDCR showing optional reticulated foam monolith installed around the UV lamp. (b) Degradation profile of DBU in the photocatalytic slurry bubble column reactor (pH = 3.17, T = 50 °C, flow rate = 0.1 dm3 /min, catalyst loading = 0.5 g/dm3 ). Reprinted from Ochuma et al. © [83, 85], with permission from Elsevier.

12 wt % TiO2 coated reticulated foam catalytic reactor could achieve a TOC conversion of approximately 23 % within 60 minutes reaction time. Depending on the basis for comparison, as to whether equivalent mass concentration in slurry form as the monolith TiO2 loading or optimal slurry concentration for the quickest degradation is used, the former comparison indicates that the reticulated foam is more effective than the slurry reactor, whilst the latter indicates that the slurry reactor is preferable. The use of a reticulated foam monolith has the advantage of avoiding catalyst filtration and separation problems from the treated water.

6.3 Trickle bed reactors

|

253

6.3 Trickle bed reactors Trickle bed reactors are packed bed catalytic reactors operating in concurrent downflow, which have attracted a large amount of industrial interest owing to their importance in the chemical, petroleum and petrochemical industries. As such, trickle bed reactors have been the subject of many reviews [79, 86]. Trickle beds are typically used in the hydrotreatment of fuels, including hydrodesulphurization, hydrodemetallation, hydrodenitrification [87], a range of selective hydrogenation of substrates including as selected examples; C4 olefins [88], maleic anhydride [89], α-methyl styrene [90] and biological processes such as wet air oxidation of waste water and model pollutant effluents [91].

6.3.1 Theory and flow regimes Trickle bed reactors may be operated under a range of flow regimes that may be summarized as: – Spray flow: where liquid exists as droplets and gas flows continuously; – Trickle flow: liquid rivulets and films flow over catalyst particles with gas as continuous phase; – Pulse flow: consisting of intermittent flows of gas and liquid rich zones through the reactor; – Bubble flow: where the liquid flow is continuous with dispersed bubbles. A detailed review of liquid distribution and flow texture in the trickle bed has been provided by Maiti et al. [92]. Fig. 6.19 shows a cut-away diagram of the trickle bed reactor. The flow regime in the trickle bed reactor is typically determined using a flow map, as shown in Fig. 6.20 (adapted from Baker [93]), where the axes are presented in terms of the liquid and gas mass velocities. Laboratory scale reactors typically operate within the lower range of the trickle flow, whilst many industrial processes such as hydrotreatment operate in pulsed flow due to the higher energetic interaction between the phases [94]. The pulsed regime is attractive for industrial operation due to the intensified processes for heat and mass transfer, where the periodic passage of liquid and gas waves promotes uniform liquid distribution through the packing [95]. This is important to achieve a high wetting of the exterior surface of the catalyst particle, since incomplete wetting effects can lead to a lower reaction rate than expected [86]. Due to the complex flow regimes that may occur, the design of trickle bed reactors is complex, but the key parameters that need to be considered for design include pressure drop, liquid holdup and wetting efficiency. A detailed review of trickle bed hydrodynamics is beyond the scope of this chapter, and therefore this review considers some of the important factors that may need to be considered in the scale up and design of trickle bed reactors from lab to pilot and industrial scale.

254 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

Liquid distributor

Catalyst packing

Catalyst support

Fig. 6.19: Schematic diagram of a trickle bed reactor.

1000

Dispersed bubble flow C

100 L'/G'

Trickle flow A

B

Pulsed (& foaming) flow

10 Trickle flow A = Pilot reactors B = Industrial reactors C = Flow transition Low → High interaction

1 0.01

0.1

Spray flow 1

10

100

G'/e (kg/sqm/s=m2) Fig. 6.20: Flow map for a trickle bed reactor, after Baker [93]. Notation: L󸀠 , liquid mass velocity; G󸀠 , gas mass velocity; e, voidage.

6.3 Trickle bed reactors |

255

6.3.2 Overall rate model The basic transport and reaction steps for trickle bed reactors are identical to those already presented for slurry reactors, although the hydrodynamics of contacting the various reacting phases is different. The main differences in the overall design equations result from: (i) the different correlations required to calculate the mass transfer coefficients; (ii) the greater likelihood that gas absorption resistance is significant, since dilute gases are sometimes used to prevent runaway of exothermic reactions; and (iii) the need to consider effects of diffusion resistance of the liquid component B as well as the dissolved gas phase component A (Winterbottom and King [2]). The derivation steps of an overall reaction rate model for the trickle bed have been presented by Winterbottom and King [2] and Fogler [5]. It should be noted that the reaction rates for the trickle bed are often presented per unit mass of catalyst, whilst for the slurry reactor the rates are per unit volume of slurry. This results in the introduction of the term 1/((1 − ε b )ρ P ) in the below equations, where ε b is the bed voidage and ρ p is the density of the catalyst pellets. The additional step that must be considered for the trickle bed compared to the slurry reactor is the transport from the bulk gas phase to the gas-liquid interface. The rate of transport per unit mass of catalyst is: R󸀠A = k g a i

1 [C Ag − C Ai ] , (1 − ε b )ρ p

(6.25)

where k g is the gas phase mass transfer coefficient, C Ag is the bulk gas phase concentration of A and C Ai is the concentration of A at the interface. Similar equations may be written for the other mass transport and reaction steps, as for the slurry reactor. Reaction is assumed to be first-order in dissolved gas A and in liquid phase reactant B. Combining the various mass transfer and reaction equations leads to the overall rate equation for gaseous reactant: R󸀠A =

1/H (1−ε b )ρ c Hk gA a b

+

(1−ε b )ρ c k l Aa b

+

1 k cA a c

+

1 ηkC Bs

C Ag = k󸀠gA C Ag ,

(6.26)

the symbols are as defined for the slurry reactor in Section 6.2.1, with k being a second order rate constant and C Bs the concentration of liquid phase reactant at the exterior surface of the catalyst particle. k󸀠gA is an overall transfer coefficient for the gas A into the pellet. In terms of design, the catalyst weight must be calculated for a particular flowrate and conversion. Considering the trickle bed as a heterogeneous plug flow reactor, the differential material balance then gives, for the change dF A in the molar flow rate of A on passing through the catalyst mass dW: − dF A (6.27) = R󸀠A = k󸀠gA C Ag . dW In the case of reactant A being mass transfer limiting the catalyst mass may be found by integration of the above equation to give: −v g 1 (6.28) W = 󸀠 ln [ ], 1 − XA k gA

256 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

where v g is the volumetric flow of the gas stream, and X A is the conversion of reactant A. A similar analysis for the rate of diffusion and reaction of B inside the catalyst pellet leads to the expression: R󸀠B =

1 k cB a c

1 +

1 ηkC As

C Bb = k󸀠lB C Bb ,

(6.29)

where k cB is the external mass transfer coefficient from the liquid to solid for reactant B, and C Bb is the bulk concentration of reactant B in the liquid phase. k󸀠lB is the overall transfer coefficient of reactant B. A differential balance for B leads to: − dF B = R󸀠B = k󸀠lB C Bb . dW

(6.30)

In case of B becoming mass transfer limiting, the catalyst mass required may be calculated by integration of the above equation to give: W=

−v l 1 ln [ ], 󸀠 1 − XB k lB

(6.31)

where v l is the liquid volumetric flowrate, and X B is the conversion of liquid phase reactant B. Summaries of some of the many correlations for the mass transfer coefficients for trickle beds have been presented by Winterbottom and King [2] and Fogler [5]. The above equations are shown here to demonstrate that the conversion in the trickle bed reactor is a complex interplay between various mass transfer steps and kinetics. Gas-liquid, liquid-solid or internal diffusion in the catalyst pellets may influence the reaction rate, in addition to the intrinsic reaction kinetics, and therefore scale up of the reactor requires a detailed understanding of the particular regime of operation.

6.3.3 Imaging of gas-liquid flows As outlined in Section 6.3.1, understanding the flow regime of operation and transition between trickle and pulsing flow is an important consideration for the operation of industrial reactors, which benefit from enhanced heat and mass transfer effects of operation in a high interaction regime, such as pulsed flow. The Gladden group has developed Magnetic Resonance Imaging (MRI) methods to probe hydrodynamic features of trickle bed reactors such as the onset of pulsing and changing liquid holdup during periodic operation, and mapping of chemical conversion in gas-liquid-solid reactors. To study the onset of pulsing in the trickle bed, rapid 3D MRI was applied to investigate the spatial distribution of liquid in the bed as a function of time, which is able to clearly determine the regions of the bed that are in trickle flow, with constant gas-liquid distribution, and the regions that have moved into unstable flow, with changing gas-liquid distribution characteristic of pulsing [96]. The work has been reviewed by Gladden

6.3 Trickle bed reactors |

257

et al., including details of the development of the MRI techniques used [97, 98]. The implemented trickle bed had a column diameter of 45 mm and catalyst support pellets of 1–3 mm dimension. The formation and development of these pulses with increasing liquid velocity was detected by MRI. The technique was also able to detect fluctuations in the liquid films flowing over the catalyst pellets, which may lead to the formation of local scale pulsing in the reactor. The basis of the technique briefly involved acquiring successive 3D images of the bed, showing the liquid distribution at that instant. A series of 3D images of a 60 mm3 bed volume with resolution 3.75 mm (x) × 3.75 mm (y) × 1.87 mm (z) were acquired at a rate of 3.6 frames per second, with six series of eight consecutive images being acquired for each flowrate investigated. The standard deviation of the signal intensity associated with each voxel of the image through the series of acquired images was calculated in order to quantify the stability of the liquid distribution in the bed, thus building up a 3D map of standard deviation values. The trickle flow regime is represented by a standard deviation value of ~ 0, whilst unsteady gasliquid distribution associated with the onset of pulsing is represented by a standard deviation of ≥ 1. A typical standard deviation map is shown in Fig. 6.21, from which the spatial extent of local pulsing regions can be identified as the bed moves from the trickle to pulsing regime. The number of voxels associated with unsteady state liquid content was found to increase with increasing liquid velocity. Film instabilities in localized areas of the bed were found to propagate with increasing liquid velocities, such that local regions of instability merged together at the transition from a trickle to pulse flow regime. Related studies considered how the MRI results relate to pressure drop and conductance measurements of the onset of pulsing that would be performed for larger, pilot and industrial scale reactor beds [99]. It was found that the pressure drop and conductance measurements only respond after the bed has moved into the fully

(a)

(b)

(c)

(d)

Fig. 6.21: Identification of the location and size of local pulses within the trickle bed determined by MRI. A high spatial resolution image (in-plane spatial resolution 175 μm × 175 μm; slice thickness 1 mm) was overlayed with a standard deviation map calculated from images acquired at a spatial resolution of in-plane spatial resolution 1.4 mm × 2.8 mm and slice thickness 2 mm. The standard deviation maps were linearly interpolated to the same in-plane spatial resolution as that of the high-resolution data. Images are shown for a constant gas velocity of 112 mm s−1 , and the data were recorded as a function of decreasing liquid velocity. The liquid velocities are (a) 2.8, (b) 3.7, (c) 6.1, and (d) 7.6 mm s−1 . Reprinted from Gladden et al. © [98], with permission from American Chemical Society.

258 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

pulsing region, whereas the MRI technique is able to detect the onset of pulsing. Similar techniques were also applied to the study of periodic operation of the trickle bed reactor. Since reactions in the trickle bed may be controlled by mass transfer processes outlined in Section 6.3.2, the operation in the periodic model offers the possibility of enhancing the reaction performance by switching on and off the gas or liquid flow, so as to periodically reduce the mass transfer resistances. MRI imaging of the bed was used to calculate the drainage profiles of the bed as a function of time, together with drainage profiles of local channels within the bed. The overall bed showed a gradual and smooth drainage profile after the liquid flow was turned off; however, while some local liquid-catalyst drainage patterns followed the overall pattern for the bed, other voids in the bed showed marked variation from the overall drainage profile, illustrating that local heterogeneities of behavior can occur. The development of magnetic resonance was also applied to chemical mapping of conversion and selectivity within the trickle bed reactor [100], which offered the potential to understand spatial variations in activity and selectivity within the bed for the first time. Such heterogeneities could result from spatial variations in hydrodynamics and mass transfer, as mentioned above. Rather than observing the 1H nucleus, 13C observation was used due to easier spectral assignment, with less overlapping of the peaks compared with 1H. 13C Distortion Enhancement Polarization Transfer (DEPT) MRI was thus applied to the study of 1-octene isomerization and hydrogenation in the trickle bed reactor over a 1 wt % Pd/Al2 O3 catalyst. This particular reaction was chosen for study since it occurs rapidly under conditions of low temperature, so that a substantial amount of isomerization and hydrogenation occurred during a single pass over ~ 3 cm bed length during the experiment. Upon the liquid reaching the catalyst bed, the spectra displayed olefinic and aliphatic peaks, which were quantified to determine the components present within the corresponding region of the bed. The profiles of 1-octene, n-octane, 2-octene and 3,4-octene as a function of position in the trickle bed were calculated, illustrating the conversion levels at different penetration distances into the catalyst bed. Such non-invasive measurements are likely to be highly useful in verifying models of the trickle bed reactor.

6.3.4 Scale up and modeling The issue of scale up is important for the successful design of commercial reactors based on laboratory data, whilst ‘scale down’ refers to testing of a new catalyst or feedstock in a laboratory reactor for potential use in an existing industrial scale reactor. It is desirable to match the space velocities of large and small reactors in order to relate the reactant conversion in laboratory scale reactors to those occurring in industrial reactors [101]. Dilution of the catalyst bed with fines is a widely used method to try to achieve this aim, whereby voids between the catalyst pellets in the laboratory 1 scale reactor are filled with small, inert, nonporous particles of approximately 10 of

6.3 Trickle bed reactors |

259

the catalyst pellet diameter. Kulkarni et al. [101] carried out a study of residence time distribution in trickle beds with porous and non-porous pellets, with and without fines dilution. The dispersion coefficient in the bed was found to decrease by up to 50 % for the bed with fines, compared with an undiluted bed. This suggests that the flow tends closer towards plug flow upon dilution of the catalyst bed with fines, compared with the undiluted bed. The RTD for porous particles showed a longer tail than for nonporous particles because of the holdup of liquid in the catalyst pores. Iliuta et al. [102] developed an axial dispersion-exchange model (ADEM) allowing for transient diffusion within the pores of catalysts. In the trickle bed reactor, films of liquids may be flowing over the outside surfaces of catalyst pellets, whilst liquid bridges may also become trapped in confined voids where particles come into contact with each other. Iliuta et al. [102] denoted these regions of liquid as respectively dynamic and static zones. Mass transfer was assumed to occur from the dynamic liquid to static liquid and in turn to liquid inside the catalyst pores. Kulkarni et al. [101] found that the model could represent a good fit of RTD data in beds of porous particles diluted with fines. Iliuta et al. [103] subsequently formulated a 1D two-fluid hydrodynamic model to approximate two-phase flow in the trickle bed, allowing for the effects of partially wetted catalyst. Significant complications in design of trickle bed reactors can occur due to a change in phase of the reactants and/or products. A design of the catalytic reactor based on the assumption of liquid phase wetting of the catalyst particles may lead to erroneous results if some of the reacting components are volatile and prone to evaporation [104]. In the event of an exothermic reaction, evaporation of the reactants may occur, leaving previously wetted catalyst particles with dry zones and depletion of the liquid phase reactant. On the other hand, when operating with vapors at high pressure within the confines of a porous catalyst, capillary condensation may lead to the formation of liquid pockets below the normal dew point of the bulk vapor. A review of this phenomenon and its effect upon catalyst particles has been provided by Ostrovskii and Wood [105]. Wood and co-workers utilized pore network modeling in combination with the Kelvin equation to predict the distribution of catalyst pores filled with liquid by capillary condensation in vapor phase reactors [106, 107]. They solved diffusionreaction network equations and determined – using percolation theory – the fraction of pores accessible to the vapor phase within the catalyst pellet. Assuming that the condensed liquid blocked regions of the catalyst where vapor phase reactions would otherwise occur, a reduction in the catalyst effectiveness factor was observed at low values of the Thiele modulus. A model proposed by Khadilkar et al. [104] allowed for wet and dry zones of the catalyst surface together with phase changes occurring due to evaporation and condensation effects, such as may be encountered in the hydrogenation of cyclohexane. The model incorporated pellet and reactor scale submodels. Flow, reaction and transport phenomena were accounted for using multicomponent diffusion theory based on the Stefan-Maxwell formulations. It was shown that multiple steady states may occur, depending on the operating conditions and history of

260 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

the reactor operation, which were confirmed by comparison with experimental data. For example, at low hydrogen to cyclohexene feed ratios, a lower conversion and reaction rate occurred (Fig. 6.22), showing that the catalyst remains in an internally fully wetted conditions throughout the reactor. Low rates are observed for the mostly fully wetted catalyst pellets. However, the upper branch in Fig. 6.22 illustrates that at high hydrogen to cyclohexane ratios, the catalyst in the entire reactor dries out. This leads to much higher rates and a greater temperature increase. The wet and dry branches are also influenced by phase holdup and velocities. DB - Dry branch WB - Wet branch 1.2 DB-II 1 DB-II Conversion

0.8 0.6 WB-II 0.4 0.2

WB-III

Wet branch (Expt) Dry branch (Expt)

0 0

5 10 Hydrogen feed ratio, N

15

Fig. 6.22: Multiplicity behavior of trickle bed reactors with a volatile liquid phase: conversion dependence on hydrogen to cyclohexene ratio. Reprinted from Khadilkar et al. © [104], with permission from Elsevier.

The application of CFD modeling to trickle bed reactors is a complex subject, owing to the discontinuous distribution of the gas and liquid, non-ideal flow behaviors including flow maldistribution, channeling and partial catalyst wetting. A very detailed review has been provided by Wang et al. [108]. Similar to the bubble column reactor, the typical modeling approaches may include Euler-Lagrange and Euler-Euler treatments, whilst the volume of fluid (VOF) approach is a surface tracking technique applied to a fixed Eulerian mesh when the locus of the interface between two or more immiscible fluids is of interest, such as gas-liquid interfaces encountered in the trickle bed reactor. Moving grid or fixed grids can be employed to simulate the interface. The “effective porous medium” approach is a method of treating the trickle bed as an effective porous medium, with lumped parameters for dispersion and heat transfer, in which case modified Ergun equation or momentum balance is used to obtain the

6.3 Trickle bed reactors

|

261

velocity field. In the “discrete particle approach”, interstitial flow is modeled which accounts for the geometric complexity of the packing structure. For the trickle bed, this approach has mainly been confined to small sections or periodic regions of the bed. Various constitutive models are required to represent the momentum exchange between the phases, such as the relative permeability, slit or fundamental force balance models. Capillary pressure effect also needs to be allowed for in modeling the trickle bed reactor, whilst closure relationships are required to represent macroscopic turbulence. CFD models have been used with some success to predict liquid phase holdup, pressure drop and gas-liquid flow maldistribution in trickle beds, although prediction of flow regime transitions is very computationally intensive. Full models incorporating mass transfer and chemical reactions effect into CFD simulations has presently achieved limited success because of the very intensive computational effort that would arise from coupling the necessary conservation of species and energy equations with the CFD model. Development of improved turbulence models, efficient parallel computational algorithms to improve solution time, and validation of the CFD models using data from imaging and non-invasive monitoring of multiphase flows are some areas identified for future consideration and development. Reports of actual industrial experience in scale up tend to be relatively scarce because of commercial intellectual property considerations, but Hickman et al. [109] recently reported the scale up of an industrial hydrogenation process for a proprietary hydrogenation reaction over a palladium catalyst. A laboratory scale reactor was used to screen the catalyst formulation and to obtain kinetic data for scale up. The final reactor design was scaled up from the laboratory by a factor of ~ 3 × 106 . The main problems encountered included incomplete wetting of the laboratory catalyst bed, catalyst productivity being lower than the value required for attractive economic production and difficulties associated with catalyst deactivation. The aspect ratio of the catalyst bed was increased in order to increase the liquid superficial velocity, while smaller catalyst pellets helped to avoid the problem of incomplete wetting. The issues of catalyst deactivation were studied by a range of characterization techniques upon spent catalyst samples and iron poisoning of Pd was found to be a problem leading to catalyst deactivation. Avoidance of iron contaminated feed was found to lead to a less severe deactivation rate, with sintering, fouling and leaching of Pd occurring at a slow enough rate for an economically attractive catalyst lifetime to be achieved.

6.3.5 Enantioselective Hydrogenation reactions Although heterogeneous catalysts provide many advantages such as ease of separation and ability to be operated in flow reactors, homogeneous catalysts may sometimes offer enhanced activity and selectivity, particularly for synthesis of fine chemicals and pharmaceuticals, which may contain stereo, chemo and enantioselective products. Stereoisomers have the same molecular formula and sequence of bonded atoms, but

262 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

differ only in the three-dimensional orientations of their atoms in space. Enantiomers consist of two stereoisomers that are mirror images of each other; they are not superimposable, similar to one’s left and right hands being the same except for their opposite orientation. The enantioselectivity of supported metal catalysts may be controlled by the addition of a very small quantity of an adsorbed chiral modifier, which is deposited on the heterogeneous catalyst surface in a special pre-treatment step before reaction. The modifier is adsorbed on the active metal surface and controls the enantioselectivity by interacting with the reactants. Homogeneous catalysts often consist of a metal center with different ligands attached to it, which may be comprised of ions or molecules. The ligands may be designed to control the reactivity of the central atom. Owing to the potential advantages of heterogeneous and homogeneous catalysts, a relatively new field of research is the preparation and application of heterogenized metal complexes, which combine the ease of separation and recycling of heterogeneous catalysts with the high activity and selectivity of heterogeneous catalysts. Several approaches for the preparation of immobilized complexes are available, including covalent binding, adsorption, ion pair formation and entrapment or the ‘ship in a bottle’ approach [110]. Despite this range of immobilization procedures, it is still challenging to synthesize efficient immobilized catalysts, since the immobilized complex must maintain its activity and selectivity, be easily recovered and not leach under the reaction conditions. Augustine et al. [111] developed a new technique of immobilization that involves attaching catalytically active complexes to solid supports via heteropolyacids (HPA). The metal atom of the complex is attached to the support via the HPA, and thus avoids the use of the ligand for attachment, which may be prone to leaching. The enantioselective hydrogenation of β-ketoesters to chiral β-hydroxyesters and the hydrogenation of α-ketoesters or acids to chiral α-hydroxyesters are two commonly studied enantioselective reactions [112]. In terms of homogeneous catalyst design, various chiral phosphine ligands have been developed following Wilkinson’s discovery of [RhCl(PPh3 )3 ] as an homogeneous hydrogenation catalyst for unhindered alkenes [113]. Several authors have then shown that the enantioselective asymmetric hydrogenation of prochiral C−C double bonds can be carried out using optically active phosphines as ligands in complexes of rhodium [114]. The production of enantiopure products is of paramount importance for applications such as pharmaceuticals. Baiker [115] asserted that research in enantioselective catalysis has been driven by the well-known fact that the incorrect enantiomer of a chiral product may have negative side effects that outweigh the beneficial value of the correct enantiomer, the birth defects caused by the drug thalidomide being an example. Currently the production of enantiopure compounds is carried out at the industrial scale using batch processing. However, the replacement of batch processes with continuous would be natural step forward as the demand for increased production volumes of fine chemicals is expected to increase [116]. Some examples of large scale production of fine chemicals

6.3 Trickle bed reactors |

263

include the painkiller Ibuprofen (10 000 tonnes/annum), sweetener S,S-aspartame (14 000 t/a) and herbicide (R)-mecropol (14 000 t/a). Only a few examples of the application of continuous processes for enantioselective catalysis are reported in the literature. Kunzle et al. [117] reported the continuous enantioselective hydrogenations of some substrates including ethyl pyruvate, ketopantolactone and 1-phenyl-1,2-propanedione over cinchona modified Pt/alumina catalysts. However, a drawback of their approach was that a continuous small top-up feed of cinchona modifier was required to be added to the inlet stream to maintain enantioselectivity. The catalyst used was in the form of very small particles, which could lead to a higher pressure drop in the trickle bed; a higher cinchona alkaloid/substrate ratio was required compared with batch processes [118]. In order to solve the problem of high pressure drop across the catalyst bed, Toukoniitty et al. [116] used a knitted silica fiber support material impregnated with platinum for the enantioselective hydrogenation of 1-phenyl-1,2-propanediol in a fixed bed reactor. Although the initial enantioselectivity of 23 % was very low, it increased to a steady state value of 57 %, with continuous feed of modifier to the reactor being necessary. Al Herz et al. [119] studied the catalytic hydrogenation of dimethyl itaconate in a laboratory scale trickle bed reactor with a liquid feed recycle. The enantioselectivity or enantiomeric excess (ee) of the reaction was determined as (S − R)/(S + R), where S and R refer to the product enantiomers. A catalyst complex [Rh((R,R)-Me-DuPhos)(COD)]BF4 was used in an immobilized form by attachment to alumina support powder or pellets using phostungstic acid (PTA) as an anchoring agent. Preliminary experiments were carried out in a shake flask to determine the kinetics: substrate to catalyst molar ratio 60, atmospheric pressure, room temperature (293.15 K), H2 flow rate of 100 ml min−1 , agitation speed of 200 rpm, whereupon a turnover frequency of 50 h−1 was obtained with the powdered alumina support and enantioselectivity 96 %. The reaction rate data were fitted using Osborne-Wilkinson kinetics [120]. The rate determining step can be one of two possible paths: (1) attack of the uncomplexed olefin on the dihydrocomplex at the vacant site giving a transition state in which both hydrogen and olefin are bound to the metal; and (2) attack of molecular hydrogen on the olefin complex leading to the same transition state. The rate may be expressed as: R=−

d[S] (k󸀠 K1 + k󸀠󸀠 K2 )p[S][A] = , dt 1 + K1 p + K2 [S]

(6.32)

where k󸀠 and k󸀠󸀠 are rate constants, K1 and K2 are equilibrium constants [S] and [A] are concentrations of substrate and catalyst respectively, p is the concentration of hydrogen in the solution. The constants were fitted using experimental data, where it was shown the value of K2 is much greater than K1 , indicating that the formation of the olefin complex is favored. The reaction in the fixed bed was carried out in the trickle flow regime using a laboratory scale set up with a bed of 6 mm diameter and a total height of 100 mm. Gas and liquid flowrates were found to have a noticeable effect upon the initial reaction rate and enantioselectivity, and optimization of flowrates

264 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

was studied. The liquid flowrate was found to have the stronger effect, being directly proportional to the wetting efficiency of the catalyst bed and rate of mass transfer of hydrogen from the bulk liquid to the catalyst surface. Optimal conditions were found to be: substrate/catalyst molar ratio of 223, atmospheric pressure, room temperature (293.15 K), 100 ml min−1 gas flowrate, 20 ml min−1 liquid flowrate, at which a conversion of DMI of 99 % and enantioselectivity of 99.9 % were achieved. The results showed some promising behavior for possible scale up of enantioselective reactions for application in the trickle bed; however, for truly continuous operation, conversion per pass of reactants through the bed would need to be increased instead of the batch recycle mode.

6.3.6 Industrial applications in heavy oil upgrading Trickle bed reactors have traditionally been used in petroleum refining processes such as hydrocracking, hydrodesulphurization and hydrodearomatization, for example. However, with the decline in the reserves of light oils, attention is switching to alternative energy sources such as heavy crude oil and bitumen, for short to medium-term energy needs. Heavy oils are asphaltic and dense; viscous oils have an API gravity between 10 and 20 °API. The American Petroleum Institute gravity or API gravity is a measure of how heavy or light a petroleum liquid is compared to water, with higher values representing lighter oils. Bitumens or oil sands have similar attributes to heavy oil but are even more viscous and dense, with viscosities usually greater than 10 000 cP and API gravity less than 10 °API. Such oils require more energy intensive operations for their production, upgrading and transportation. Due to the potentially higher cost of such extraction methods, a range of Enhanced Oil Recovery (EOR) techniques have been proposed and developed to try to extract as much of the oil in place as possible. These include miscible displacement with gases such as CO2 , flooding with chemicals, surfactants or polymers, microbial enhanced recovery methods and thermal methods [121]. The thermal methods encompass technologies requiring steam injection, such as steam flooding, cyclic steam stimulation and steam assisted gravity drainage (SAGD). Alternatively, in situ combustion can be used, which works by burning a small fraction of the oil by injecting air into the reservoir in order to enable flow of the remaining oil and enhance the oil production rate. This is normally achieved by the injection of an oxidizing gas such as air. ‘Toe-to-heel’ Air Injection (THAI) combines ISC with a horizontal production well, as shown in Fig. 6.23. The well is initially steamed to raise its temperature; thereafter, air is injected to ignite the in situ combustion process. The flame front moves gradually along the horizontal well from the ‘toe’ position to the ‘heel’. The process offers good scope for control of the combustion since a coke layer forms as the combustion front passes, preventing the bypass of gas over the heavy oil layer. Up to 80–85 % recovery of the oil in place may be possible, which is significantly higher than alternative techniques. A catalytic add-on

6.3 Trickle bed reactors |

265

to the process was also developed, termed CAPRI, which involves the packing of an annular layer of refinery hydrotreatment catalyst surrounding the horizontal well: in effect, the well becomes the reactor with in situ upgrading taking place. The upgrading reactions occurring in the THAI process are thought to occur in the Mobile Oil Zone (MOZ). The reactions taking place include (i) carbon rejection reactions as a result of thermal cracking, and (ii) hydrogen addition when the already lighter, cracked components contact the hydrotreating catalyst in the annular layer (CAPRI) surrounding the producer well. Field trials of THAI have been carried out by Petrobank at their White Sands pilot project in 2006, at Christina Lake, Alberta, Canada. Maximum gross oil production per well pair was reported as 2000 barrels/day with a bitumen cut of around 55 %. The produced oil was found to be partially upgraded in the range of 10.6 to 16.1 °API. The THAI-CAPRI upgrading mechanism can simply be represented by the following equations: Thermal cracking (Pyrolysis): Heavy residue

󳨀󳨀→ Light oil + coke

Oxidation of coke (high temperature oxidation): Coke + O2

󳨀󳨀→ CO + CO2 + H2 O

Oxidation of heavy residue: Heavy residue + O2 󳨀󳨀→ CO + CO2 + H2 O Carbon rejection: CHx

󳨀󳨀→ CHx1 + C

(x1 > x)

󳨀󳨀→ CHx1

(x1 > x)

Hydrogen addition: CHx + H2

Upgrading of heavy oil/natural bitumen is a very complex process, so the hydrogen responsible for the upgrading of the THAI-CAPRI produced oil is thought to be formed due to the water gas shift reaction, and is represented with the following simple equations: Gasification of hydrocarbon: CHx

󳨀󳨀→ C + 2x H2

C + H2 O (steam) 󳨀󳨀→ CO + H2 C + CO2

󳨀󳨀→ CO

CO + H2 O

󳨀󳨀→ CO2 + H2

Water-gas shift:

266 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions Injection well

Production well

Combustion front Mobile Horizontal well oil zone ‘Toe’

Cold, heavy oil ‘Heel’

Fig. 6.23: Schematic diagram showing heavy oil recovery using Toe-to-Heel Air Injection method (THAI). Reprinted from Shah et al. © [121], with permission from Royal Society of Chemistry.

It can be observed that coke is produced by the reactions, which may lead to possible deactivation of the catalyst. Therefore Shah et al. [122] made a laboratory-based study of the optimization of process conditions for in situ heavy oil and bitumen recovery using the THAI-CAPRI process. A rig was developed consisting of two microreactors for the purpose of simulating in situ oil upgrading conditions in well. Each reactor contains a fixed bed of catalyst to represent a cylindrical core of 10.2 mm diameter taken in a radial direction through the annular layer of catalyst packed around the producer well. Associated equipment includes a gas supply system, trace heating, furnace, temperature control system and gas-liquid separator. Experiments were carried out on a feed oil supplied from the THAI trials at the White Sands field by Petrobank. The microreactors were packed with 10 g of catalyst samples such as CoMo, NiMo and ZnO/CuO. An oil flow of 1 ml/min and gas flow of 0.5 l/min were used under different temperatures, pressures and gas environments. Selected experiments were carried out using a gas mixture representative of the combustion gases from the THAI process, namely 80 % N2 , 13 % CO2 , 3 % CO and 4 % CH4 . Other experiments were carried out under a flow of nitrogen. It was found that temperature has a strong effect upon the upgrading occurring and catalyst lifetime. Operation at 500 °C and 20 bar led to an average of 6.1 °API upgrading of THAI feed oil to 18.9 °API, however catalyst lifetime was limited to just 1.5 hours. Reduction of the temperature to 420 °C was found to lead to a lower extent of upgrading of an average of 1.6 °API, and sometimes up to 3 °API, but catalyst lifetime was considerably extended up to 77.5 hours. The process was less sensitive to changes in gas-flow rate and pressure. Substantial catalyst deactivation occurred in the first few hours of the reaction, such that the pore space became blocked. In order to try to mitigate the effects of coke deposition, [123] studied the effect of guard bed and hydrogen upon the upgrading process. A guard bed involves the packing of inert porous particles above the catalyst in order to adsorb or filter coke precursors from the feed before they reach the catalyst. Hart et al. [123] used a layer of activated carbon particles for this purpose, over the CoMo catalyst. Furthermore, the addition of hydrogen was studied to augment catalytic hydroconversion and hydrocracking reactions in order to achieve a higher level of upgrading than would be achieved with inert gases. It was found that depending on the upgrading temperature, the viscosity of the produced oil reduced significantly by 42–82 % and API gravity in-

6.4 Structured monolith reactors

|

267

creased by ~ 2 to 7 °API relative to the feedstock of 0.49 Pa s and 13 °API, respectively. Conversely, the use of hydrogen further increased the API gravity by 2 °API and the viscosity by 5.3 %. Notably, the coke content of the catalyst reduced from 57.3 wt % in nitrogen to 34.8 wt % in hydrogen atmosphere. The use of a guard bed increased the API gravity of the produced oil by a further 2° and reduced the viscosity by an average of 8.5 % further than achieved with the active HDS catalyst CoMo/alumina. Catalytic hydrocracking is thought to involve the breaking of larger molecules to give fragmented free radicals, which then react with hydrogen radicals in order to stabilize the hydrocarbon chains and terminate the reaction. In the absence of hydrogen, the active chains keep reacting with each other, resulting in the formation of higher molecular weight compounds by polymerization, increased coke formation and adverse impact of viscosity and API upgrading of the produced oil.

6.4 Structured monolith reactors Monolithic catalysts were initially developed for use in automotive exhaust emission control, but during the past 10–15 years increasing interest has occurred in their application to a more diverse range of areas in the chemical and allied industries [57]. Structured honeycomb monolith catalyst supports are usually comprised of parallel, usually straight capillary channels upon which the catalyst is coated within a wash coat layer of porous material. In the field of energy, monolith reactors have been applied for hydrogen production from steam reforming [124], water gas shift reaction [125], methanation of carbon dioxide to produce methanol [126] and manufacture of synthetic fuels using Fischer-Tropsch synthesis [127]. The purification of air or water by monolithic catalysts has been studied [128, 129] as well as the catalytic reduction of pollutants such as NOx from stationary sources [130, 131]. Although originally applied to gas phase systems, new applications of monoliths in three-phase gas-liquid-solid applications have been considered [132, 133]. Reaction systems of interest include hydrogenations, for example of 2-butyne-1,4-diol [134], nitrobenzoic acid [135], α-methyl styrene [136], benzaldehyde [137], pyrolysis gasoline [138] and oxidations such as of glycerol [139], glucose [140] and ethanol [141]. However, one of the few known scaled up industrial applications is in the production of hydrogen peroxide [142]. The monolith reactor has the advantage that intensified performance can be achieved with much lower power inputs than some alternative process intensification devices. Key to this enhancement of performance is the length scales over which reaction and mixing occur in the monolith reactor. Although a range of flow patterns may occur within the channel, one of the most desirable regimes for enhanced mass transfer is ‘Taylor flow’ or bubble train flow, in which liquid slugs separated by gas bubbles flow along each channel. This creates a thin film of liquid at the sides of the gas bubbles in contact with the catalyst coating on the channel wall, the thickness of which may range from 5–50 μm for fluids of low viscosity [143]. Typically, the

268 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

channel side length in monoliths is of the order of 1–2 mm and optimum washcoat thickness is of the order 70 μm, for example in the case of Syngas production [144]. These scales mean that the momentum balance and bulk mass transfer occur over the characteristic length of the channel diameter of the order of millimeters, whilst diffusion across the liquid film and into the washcoat occurs over distances of the order of tens of microns [145]. Reaction occurs at catalytic sites in pores of the order of tens of nanometers, within the washcoat. These small length scales can be contrasted with the much larger sizes in traditional reactors such as fixed and trickle beds, which utilize extrudate pellets of several millimeters length within reactors on the scale of several meters. By minimizing the transport and diffusion distances, utilizing a thin catalyst washcoat, enhancing mixing within the channels and bringing the reactants together in intimate contact, monolith reactors represent an advantageous technology for process intensification. They can lead to improved reactant conversion, selectivity towards desirable products and improved process safety [132].

6.4.1 Flow patterns in the single capillary 6.4.1.1 Flow regimes The performance of the monolith reactor as a high-intensity mass transfer device is very dependent upon the flow pattern of gas and liquid within the monolith channels. Therefore, it is important to understand the conditions under which certain flow patterns occur and to be able to predict mass transfer parameters for these flow regimes. In some earlier studies, Satterfield and Ozel [146] showed that the flow regimes are strongly influenced by the wettability between the liquid and the capillary wall as well as the surface tension between the liquid and gas phases. A wealth of literature exists for gas-liquid two-phase flows in pipes and channels, for example Taitel et al. [147] presented a comprehensive model for flow pattern transitions in large diameter tubes for vertical upflow. However, such studies were not representative of the correct length scales of monolith channels with diameters of 1–2 mm. Thulasidas et al. [143] carried out some of the first detailed hydrodynamic studies at scales relevant to monolith channels operating in upflow mode and determined the main mass transfer parameters, such as bubble size, shape, velocity and volume fraction of gas inside capillaries of circular or square cross section on the basis of the superficial flow rates of gas and liquid in the feed. However, monolith reactors such as the CDCR operate in downflow mode and later studies of downflow in monolith channels were carried out by Tsoligkas et al. [148] for 1.5–2 mm square glass capillaries, which represent a single channel in a monolith reactor. Using a high-speed camera, five flow patterns and three transitional flow patterns were observed as shown in Fig. 6.24. The regimes most worthy of note are (a) annular Flow, (c) Taylor flow, also known as regular slug or bubble-train flow, (e) bubbly flow (h) churn flow, with the others representing transitional regimes. Flow regime maps were determined for air-water and

6.4 Structured monolith reactors

|

269

air-water/isopropanol mixtures, which may be used to guide the selection of gas and liquid velocities in order to achieve the desired flow regime within the monolith. Taylor flow is a desirable regime in which to operate in order to maximize the transfer of a gas such as hydrogen to the catalyst surface within the three-phase reactor, owing to the thin liquid films formed at the sides of the gas bubble.

(a)

(b)

(c)

(d1)

(d2)

(e)

(f)

(g)

(h)

Fig. 6.24: Flow patterns in a square capillary tube for vertical down-flow. (a) annular flow; (b) slug– annular flow; (c) Taylor (regular slug) flow; (d) slug–bubbly flow; (e) bubbly flow; (f) irregular slug flow; (g) slug–churn flow; and (h) churn flow. Reprinted from Tsoligkas et al. © [148], with permission of Elsevier.

6.4.1.2 Mixing and mass transfer In order to design and optimize the monolith reactor in three-phase operation, it is necessary to understand the different modes of mass transfer which may be (1) gasliquid-solid (GS) via the liquid film, (2) gas-liquid (GL) via the gas bubble caps and (3) liquid-solid (LS) within the liquid slugs. In this section, flow studies of the liquid slugs are reviewed, followed by the mass transfer characteristics of the three-phase monolith reactor.

6.4.1.3 Recirculation in liquid slugs Within the liquid slugs that occur in Taylor flow, mass transfer processes (2) GL and (3) LS as mentioned above occur and therefore it is important to understand the mechanisms of mixing which assist this mass transfer. Thulasidas et al. [143] used Particle Image Velocimetry (PIV) to study the velocity distributions in liquid slugs in Taylor flow operated in ‘upflow’ mode, and showed that recirculating patterns occur with a high degree of mixing. The detailed behavior is a function of the capillary number,

270 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

where

μV . (6.33) σ Depending on the capillary number of the flow, counter rotating vortices or a complete bypass flow inside the liquid slug were observed. Such recirculation patterns can play an important role in transporting dissolved gas from the bubble cap to the catalyst coated upon the channel walls. Tsoligkas et al. [148] also used PIV to study flow in liquid slugs, but in downflow mode. They found that short slugs (slug length less than the tube hydraulic diameter) led to a relatively flat velocity profile, where the axial velocity is only a function of the position in the tube cross-section, as shown in Fig. 6.25. By contrast, in long slugs the axial velocity component depends upon both the axial position in the tube and the tube cross-section. Parabolic velocity profiles are approximated for Vmax /V b ≈ 1.1–1.7. Significant differences from upflow operation were observed with upflow recirculation times being three times faster than downflow; this has implications for the models used to predict mass transfer and residence time distribution. Ca =

3 2

–z

z (mm)

1

z (mm)

0.5 –z 0

–x

0

Vel mag 0.507949 –x

0 +x

0.304769

–1

0 +z

+x –2

+z

0.20318 0.10159

–0.5 –3 –0.5 (a)

0.406359

0 x (mm)

0.5 (b)

0 –0.6 0 0.6 x (mm)

Fig. 6.25: Flow fields within the liquid slugs determined by PIV in the W = 1.5 mm tube for (a) L S < D c , 30 % v/v isopropanol/water–air mixture, Ca = 0.008, VGS = 0.384 m s−1 , VLS = 0.111 m s−1 , V B = 0.576 m s−1 ; (b) L S > D c , water–air mixture, Ca = 0.0287, VGS = 0.209 m s−1 , VLS = 0.106 m s−1 , V B = 0.343 m s−1 . Units of velocity are m s−1 . Reprinted from Tsoligkas et al. © [148], with permission of Elsevier.

6.4 Structured monolith reactors

|

271

6.4.1.4 Mass transfer processes A detailed analysis of the mass transfer steps involved in Taylor flow within the monolith reactor has been given by Kreutzer et al. [133], and is summarized in this section. The three mass transfer modes of steps (1)–(3) from above may be combined, with the (2) GL and (3) LS steps considered as resistances in series; they are in parallel with respect to (1) GS mass transfer. For the overall mass transfer, the following expression can be used: −1 1 1 kOV a = kGS aGS + ( + (6.34) ) . kGL aGL kLS aLS The individual mass transfer coefficients and areas are determined as: kGS = aGS kGL aGL

D

, δ 4(1 − ε L ) = , Dc 1.2 0.133UTP = , L0.5 slug

kLS aLS =

D 4ε L

δ Dc

.

(6.35) (6.36) (6.37) (6.38)

In Eqs. (6.35) and (6.38), the film thickness δ is an important parameter, which strongly influences the rate of mass transfer. The film thickness is influenced by the Capillary number and thus from Eq. (6.33) it can be observed that the unit cell velocity, as well as the viscosity and surface tension of the fluid have a strong effect upon the mass transfer rates in the reactor.

6.4.2 Applications in hydrogenation reactions Monolith reactors have been studied for an increasing range of applications in threephase reactions, and are particularly effective for hydrogenation, dehydrogenation and oxidation reactions [137, 149, 150]. However, rational evaluation of the performance of such reactors must be made on the basis of a comparison with other traditional reactor designs. Fishwick et al. [17] made a comparison of selective hydrogenations in the monolith CDC, stirred tank and trickle bed reactors with configurations shown in Fig. 6.26. The selective hydrogenation of 2-butyne-1,4-diol (B3D) was used as the test reaction, which has the reaction scheme shown in Fig. 6.6. This is a consecutive reaction, in which the alkene intermediate, 2-butene-1,4-diol (B2D) is an important chemical in the manufacture of vitamins, pharmaceuticals and insecticides. The fully hydrogenated product 2-butane-1,4-diol (B1D) is a raw material used in the polymer industry and manufacture of tetrahydrofuran. Five reactors were compared, which included a stirred tank, trickle bed reactor, single capillary reactor, 5 cm diameter monolith CDCR and 10 cm diameter monolith CDCR. All of the reactors were operated

272 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

at a pressure of 200 kPa, except the single capillary which was operated at 100 kPa and a temperature of 55 °C was used in each case. Industrial Pd/Al2 O3 catalysts, supplied by Johnson Matthey, were tested, which included a powdered 1 % γ-Pd/Al2 O3 (stirred tank), 0.5 % Pd/Al2 O3 pellet (trickle bed) and a washcoated Pd/α-Al2 O3 monolith. Each catalyst was typical of the industrial catalyst used in that type of reactor, and in order to facilitate a comparison of the different reactors, rates were reported as normalized per gram of palladium. Solvents used in the reaction included water, 2-propanol and a 30 % v/v mixture of 2-propanol in water. Tab. 6.4: Initial reaction rates and selectivity towards cis-2-butene-1,4-diol for CDC monolith, stirred tank (STR) and trickle bed reactors (TBR). Reprinted from Fishwick et al. © [17], with permission of Elsevier. Solvent

Reactor

Initial Rate [mol s−1 g−1 ] Pd

5 cm monolith 10 cm monolith STR TBR

0.186 × 10−3

30 % v/v 2-Propanol in water (M) 2-Propanol

Water

(90 % conversion)

(100 % conversion)

0.186 × 10−3 0.264 × 10−3 0.033 × 10−3

1.00 1.00 0.99 0.87

1.00 0.96 0.93 0.86

5 cm monolith 10 cm monolith STR TBR

0.206 × 10−3 0.206 × 10−3 0.319 × 10−3 0.330 × 10−3

1.00 1.00 1.00 0.93

1.00 1.00 0.95 0.90

5 cm monolith STR

0.268 × 10−3 0.522 × 10−3

0.99 0.97

0.90 0.87

Gas in

Liquid in

Gas in

Liquid in

Gas in

Bubble dispersion

Liquid in

Selectivity S

Gas Liquid

Liquid out Slurry column

Liquid out Packed bed

Liquid out Monolith

Taylor flow

Fig. 6.26: Schematic diagram illustrating the CSTR, packed bed, single capillary, 5 cm diameter monolith and 10 cm diameter monolith operation in the CDCR.

6.4 Structured monolith reactors |

273

Tab. 6.4 displays the initial rate and selectivity towards B2D at B3D conversions of 90 % and 100 % for a range of different solvent compositions and reactor types. Selectivity, S, is defined at the mole fraction of the intermediate (B2D) in the total number of moles of product. From Tab. 6.4 it can be observed that the B2D selectivity is generally very high, but that the structured monoliths lead to higher selectivities than both the stirred tank and trickle beds, regardless of which solvent is used. High selectivity towards B2D was thought to be associated with effective mass transfer of hydrogen to the catalyst surface. This ensures that the active sites are sufficiently populated with hydrogen so that hydrogenation occurs in preference to the formation of side products by reactions that do not consume hydrogen, for example 4-hydroxybutanal, 2-buten-1-ol or the derivatives of these products shown in Fig. 6.6. However, if the concentration of hydrogen at the surface is too high, it could lead to over-hydrogenation of the alkene to produce the alkane and thus decrease the selectivity. Also a close approximation to plug flow, as expected in the monoliths, could reduce the amount of time that the intermediate alkene spends in contact with the catalyst and thus decrease the extent of over-hydrogenation to form the alkane. In the monoliths, the mass transfer coefficients of hydrogen at both the gas-liquid and liquid-solid interfaces were substantially higher than in the stirred tank and trickle bed, which would be expected to lead to better availability of hydrogen at the catalyst surface and thus increase selectivity towards the hydrogenation products, rather than promoting side reactions. The performance of the monolith was also found to be dependent upon the hydrodynamic mode of operation, that is whether the gas bubble dispersion is allowed to penetrate in to the monolith channels, such that Taylor flow occurs in the channels, or liquid is pre-saturated with hydrogen by controlling the bubble dispersion to remain in the upper part of the column. Higher mass transfer rates were observed when the monolith is operated in two-phase flow. Referring again to Tab. 6.4, for the use of the mixed water/2-propanol solvent, complete selectivity towards the alkene is achieved without the requirement of using dopants or poisoning the catalyst with addition of bases. The effect of a solvent type upon the reaction behavior could be partly explained in terms of higher hydrogen solubility in 2-propanol compared with water, leading to higher rate of hydrogenation. Also, in mixtures of 2-propanol, the bubble size is influenced by the solvent composition, which in turn affects the gas-liquid interfacial area. Of the three solvents studied, the smallest bubbles observed by a video-microscope-computer system occur for a mixture of 2-propanol in water mixture; therefore, a higher gas-liquid interfacial area is observed [9]. These studies show that suitable selection of reactor type, operating conditions and solvent can lead to improved selectivity towards the desirable product, therefore decreasing the production of undesirable waste. In some cases, structured reactors open up new reaction routes or make reactions feasible that would otherwise have low rates and selectivities in the autoclave. The oxidation of glycerol, a by-product of the biodiesel industry, can be used to produce a range of useful products. Glycerol is available from sustainable sources, but

274 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

traditional oxidation processes used in industry sometimes involve environmentally unfriendly reagents such as dichromate, permanganate, manganese and peroxides [151]. Hutchings et al. [58] has shown that gold catalysts are active for selective oxidation reactions, using molecular oxygen as the oxidizing agent and is very selective for the oxidation of glycerol to glycerate. The addition of base, such as sodium hydroxide, is required to accelerate the rate-limiting deprotonation step. Pollington et al. [139] studied the selective oxidation of glycerol in several multiphase reactors. The flow reactors consisted of a laboratory scale loop system operated in batch-recycle mode. The catalyst or support used included an Au/C catalyzed monolith or a cordierite monolith without metal coating, in which case the Au/C catalyst powder was placed in the feed tank with the liquid such that slurry flowed through the uncoated monolith during operation, denoted as a meso-structured slurry bubble column (MSSBC). Comparisons were made with studies of the same reaction in an autoclave. It was observed that the highest rate occurred in the MSSCB (109–201 mmol glycerol s−1 m−3 bar−1 ), followed by the catalyst coated monolith (36–65 mmol glycerol s−1 m−3 bar−1 ), and lastly the stirred reactor (4 mmol glycerol s−1 m−3 bar−1 ). Selectivity towards glyceric acid was 100 % for the monolith and autoclave, but in the MSSBC the production of dihydroxyacetone and glyceric acid was approximately equal, thus selectivity towards glyceric acid was in the range 47–54 %. The lower selectivity in the MSSBC was attributed to better oxygen availability than the monolith and autoclave reactors. The study illustrated the benefits of structuring catalysts, where the monolith gave rise to enhanced reaction rates while maintaining high selectivity, which arises from enhanced mass transfer via thin liquid films when the gas-liquid mixture flows through the channels of the monolith catalyst.

6.5 Conclusions This chapter has reviewed research covering a range of three-phase reactors including stirred tanks, bubble columns, trickle beds and monolith reactors. Case studies have been described to illustrate the application of these reactors in a range of hydrogenation and oxidation reactions from the laboratory to industrial scale operations. The interplay between hydrodynamics, mass transfer, mixing and reaction kinetics has been highlighted in the discussion. The selection of a suitable catalyst must also be accompanied by analysis of transport processes in the reactor to determine the limiting step, and issues concerning catalyst deactivation must also be understood. The development of non-invasive imaging techniques such as PEPT, CARPT, MRI and PIV has led to improved understanding of the mixing processes occurring in a range of three-phase reactors. Results from such experimental studies have been used to validate CFD models of three-phase flows of increasing complexity. The combination of these tools has led to increased confidence for scale up of processes for industrial application.

References |

275

References [1] [2] [3] [4]

[5] [6] [7] [8] [9]

[10] [11]

[12]

[13] [14] [15] [16]

[17] [18] [19] [20] [21]

Biardi G, Baldi G. Three-phase catalytic reactors. Catal Today 1999, 52, 223–234. Winterbottom JM, King MB. Reactor design for chemical engineers. Cheltenham, UK, Stanley Thornes, 1999. Ramachandran PA, Chaudhari RV. Three-phase catalytic reactors. New York, Gordon and Breach Science Publishers, 1983. Gianetto A, Silveston PL, Baldi G. Multiphase chemical reactors: theory, design, scale up. Washington; Berlin; New York, Hemisphere Pub. Corp. ; Distribution outside North America, Springer-Verlag, 1986. Fogler HS. Elements of chemical reaction engineering. Upper Saddle River, N.J Prentice Hall PTR, 1999. Markopoulos J, Christofi C, Katsinaris I. Mass transfer coefficients in mechanically agitated gas-liquid contactors. Chem Eng Technol 2007, 30, 829–834. Kawase Y, Araki T, Shimizu K, Miura H. Gas-liquid mass transfer in three-phase stirred tank reactors: Newtonian and non-Newtonian fluids. Can J Chem Eng 1997, 75, 1159–1164. Rajadhyaksha RA, Karwa SL. Solvent effects in catalytic-hydrogenation. Chem Eng Sci 1986, 41, 1765–1770. Hu B, Fishwick RP, Pacek AW, et al. Simultaneous measurement of in situ bubble size and reaction rates with a heterogeneous catalytic hydrogenation reaction. Chem Eng Sci 2007, 62, 5392–5396. Hu B, Pacek AW, Stitt EH, Nienow AW. Bubble sizes in agitated air-alcohol systems with and without particles: Turbulent and transitional flow. Chem Eng Sci 2005, 60, 6371–6377. Hu BJ, Nienow AW, Stitt EH, Pacek AW. Bubble sizes in agitated solvent/reactant mixtures used in heterogeneous catalytic hydrogenation of 2-butyne-1,4-diol. Chem Eng Sci 2006, 61, 6765–6774. Fishwick RP, Winterbottom JM, Parker DJ, Fan XF, Stitt EH. Hydrodynamic measurements of up- and down-pumping pitched-blade turbines in gassed, agitated vessels, using positron emission particle tracking. Ind Eng Chem Res 2005, 44, 6371–6380. Mahouast M, Cognet G, David R. 2-Component LDV measurements in a stirred tank. Aiche Journal 1989, 35, 1770–1778. Sheng J, Meng H, Fox RO. Validation of CFD simulations of a stirred tank using particle image velocimetry data. Can J Chem Eng 1998, 76, 611–625. Devanathan N, Moslemian D, Dudukovic MP. Flow mapping in bubble-columns using CARPT. Chem Eng Sci 1990, 45, 2285–2291. Leadbeater TW, Parker DJ. A modular positron camera for the study of industrial processes. Nuclear Instruments & Methods in Physics Research Section a-Accelerators Spectrometers Detectors and Associated Equipment 2011, 652, 646–649. Fishwick RP, Natividad R, Kulkarni R, et al. Selective hydrogenation reactions: A comparative study of monolith CDC, stirred tank and trickle bed reactors. Catal Today 2007, 128, 108–114. Wang M, Dorward A, Vlaev D, Mann R. Measurements of gas-liquid mixing in a stirred vessel using electrical resistance tomography (ERT). Chemical Engineering Journal 2000, 77, 93–98. Mills PL, Chaudhari RV. Reaction engineering of emerging oxidation processes. Catal Today 1999, 48, 17–29. Froment GF, Bischoff KB. Chemical reactor analysis and design. New York, Wiley, 1990. Mills PL, Chaudhari RV. Multiphase catalytic reactor engineering and design for pharmaceuticals and fine chemicals. Catal Today 1997, 37, 367–404.

276 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

[22]

[23] [24] [25] [26] [27] [28] [29] [30] [31]

[32]

[33]

[34]

[35] [36] [37]

[38]

[39] [40] [41]

Brahme PH, Doraiswamy LK. Modeling of a slurry reaction - hydrogenation of glucose on raney-nickel. Industrial & Engineering Chemistry Process Design and Development 1976, 15, 130–137. Chaudhari RV, Mills PL. Multiphase catalysis and reaction engineering for emerging pharmaceutical processes. Chem Eng Sci 2004, 59, 5337–5344. Irandoust S, Andersson B. Mass-transfer and liquid-phase reactions in a segmented 2-phase flow monolithic catalyst reactor. Chem Eng Sci 1988, 43, 1983–1988. Gupte SP, Chaudhari RV. Carbonylation of 2,4-dinitrotoluene using homogeneous Pd and Rh complex catalysts. Journal of Molecular Catalysis 1984, 24, 197–210. Bond GC, Wells PB. Hydrogenation of acetylene .2. Reaction of acetylene with hydrogen catalyzed by alumina-supported palladium. J Catal 1966, 5, 65–73. Bond GC, Wells PB. Hydrogenation of acetylene .3. Reaction of acetylene with hydrogen catalyzed by alumina-supported rhodium and iridium. J Catal 1966, 5, 419–427. Freidlin LK, Kaup, II. Selectivity and stereospecificity in hydrogenation of acetylene hydrocarbons on metal catalysts. Doklady Akademii Nauk Sssr 1963, 152, 1383 ff. Arafa EA, Webb G. A C-14 radiotracer study of the hydrogenation of ethyne over EUROPT-1. Catal Today 1993, 17, 411–418. Margitfalvi J, Guczi L, Weiss AH. Reaction routes for hydrogenation of acetylene-ethylene mixtures using a double labeling method. React Kinet Catal Lett 1980, 15, 475–479. Guczi L, Tetenyi P. Some new aspects of mechanism of hydrogen exchange and hydrogenation reaction over metal catalyst .2. Hydrogenation of ethylene by tritiated hydrogen on ni catalyst. Zeitschrift Fur Physikalische Chemie-Leipzig 1968, 237, 356 ff. Borodzinski A, Cybulski A. The kinetic model of hydrogenation of acetylene-ethylene mixtures over palladium surface covered by carbonaceous deposits. Appl Catal A-Gen 2000, 198, 51– 66. Duca D, Arena F, Parmaliana A, Deganello G. Hydrogenation of acetylene in ethylene rich feedstocks: Comparison between palladium catalysts supported on pumice and alumina. Appl Catal A-Gen 1998, 172, 207–216. Kaltchev M, Stacchiola D, Molero H, Wu G, Blumenfeld A, Tysoe WT. On the reaction pathway for the formation of benzene from acetylene catalyzed by palladium. Catalysis Letters 1999, 60, 11–14. Stacchiola D, Calaza F, Zheng T, Tysoe WI. Hydrocarbon conversion on palladium catalysts. Journal of Molecular Catalysis a-Chemical 2005, 228, 35–45. Jackson SD, Hamilton CA, Kelly GJ, de Bruin D. The hydrogenation of C-5 alkynes over palladium catalysts. React Kinet Catal Lett 2001, 73, 77–82. Boitiaux JP, Cosyns J, Robert E. Hydrogenation of unsaturated-hydrocarbons in liquid-phase on palladium, platinum and rhodium catalysts .3. Quantitative selectivity ranking of platinum, palladium and rhodium in the hydrogenation of 1-butene, 1,3-butadiene and 1-butyne using a single reaction scheme. Applied Catalysis 1987, 35, 193–209. Boitiaux JP, Cosyns J, Vasudevan S. Hydrogenation of highly unsaturated-hydrocarbons over highly dispersed palladium catalyst .1. Behavior of small metal particles. Applied Catalysis 1983, 6, 41–51. Boitiaux JP, Cosyns J, Derrien M, Leger G. Newest hydrogenation catalysts. Hydrocarbon Processing 1985, 64, 51–59. Hamilton CA, Jackson SD, Kelly GJ, Spence R, de Bruin D. Competitive reactions in alkyne hydrogenation. Appl Catal A-Gen 2002, 237, 201–209. Teschner D, Vass E, Havecker M, et al. Alkyne hydrogenation over Pd catalysts: A new paradigm. J Catal 2006, 242, 26–37.

References |

[42]

[43] [44] [45] [46] [47]

[48] [49]

[50] [51]

[52]

[53] [54]

[55]

[56] [57]

[58] [59] [60] [61]

277

Canning AS, Jackson SD, Mitchell S. Identification, by selective poisoning, of active sites on Ni/Al2O3 for hydrogenation and isomerisation of cis-2-pentenenitrile. Catal Today 2006, 114, 372–376. Bennett JA, Fishwick RP, Spence R, et al. Hydrogenation of 2-pentyne over Pd/Al2O3 catalysts: Effect of operating variables and solvent selection. Appl Catal A-Gen 2009, 364, 57–64. Perego C, Villa P. Catalyst preparation methods. Catal Today 1997, 34, 281–305. Zahmakiran M, Ozkar S. Metal nanoparticles in liquid phase catalysis; from recent advances to future goals. Nanoscale 2011, 3, 3462–3481. Tsuji Y, Fujihara T. Homogeneous nanosize palladium catalysts. Inorganic Chemistry 2007, 46, 1895–1902. Astruc D. Palladium nanoparticles as efficient green homogeneous and heterogeneous carbon-carbon coupling precatalysts: A unifying view. Inorganic Chemistry 2007, 46, 1884– 1894. Creamer NJ, Mikheenko IP, Yong P, et al. Novel supported Pd hydrogenation bionanocatalyst for hybrid homogeneous/heterogeneous catalysis. Catal Today 2007, 128, 80–87. Deplanche K, Snape TJ, Hazrati S, Harrad S, Macaskie LE. Versatility of a new bioinorganic catalyst: Palladized cells of Desulfovibrio desulfuricans and application to dehalogenation of flame retardant materials. Environmental Technology 2009, 30, 681–692. Bennett JA, Attard GA, Deplanche K, et al. Improving Selectivity in 2-Butyne-1,4-diol Hydrogenation using Biogenic Pt Catalysts. Acs Catalysis 2012, 2, 504–511. Depanche K, Bennett JA, Mikheenko IP, et al. Catalytic activity of biomass-supported Pd nanoparticles: Influence of the biological component in catalytic efficacy and potential application in ’green’ synthesis of fine chemicals and pharmaceuticals. Applied Catalysis B-Environmental 2014, 147, 651–665. Attard GA, Bennett JA, Mikheenko I, et al. Semi-hydrogenation of alkynes at single crystal, nanoparticle and biogenic nanoparticle surfaces: the role of defects in Lindlar-type catalysts and the origin of their selectivity. Faraday Discussions 2013, 162, 57–75. Wood J, Bodenes L, Bennett J, Deplanche K, Macaskie LE. Hydrogenation of 2-Butyne-1,4-diol Using Novel Bio-Palladium Catalysts. Ind Eng Chem Res 2010, 49, 980–988. Bennett JA, Creamer NJ, Deplanche K, Macaskie LE, Shannon IJ, Wood J. Palladium supported on bacterial biomass as a novel heterogeneous catalyst: A comparison of Pd/Al2O3 and bioPd in the hydrogenation of 2-pentyne. Chem Eng Sci 2010, 65, 282–290. Crespo-Quesada M, Yarulin A, Jin MS, Xia YN, Kiwi-Minsker L. Structure Sensitivity of Alkynol Hydrogenation on Shape- and Size-Controlled Palladium Nanocrystals: Which Sites Are Most Active and Selective? Journal of the American Chemical Society 2011, 133, 12787–12794. Schmidt E, Vargas A, Mallat T, Baiker A. Shape-Selective Enantioselective Hydrogenation on Pt Nanoparticles. Journal of the American Chemical Society 2009, 131, 12358–12367. Kreutzer MT, Kapteijn F, Moulijn JA, Heiszwolf JJ. Multiphase monolith reactors: Chemical reaction engineering of segmented flow in microchannels. Chem Eng Sci 2005, 60, 5895– 5916. Hutchings GJ, Carrettin S, Landon P, et al. New approaches to designing selective oxidation catalysts: Au/C a versatile catalyst. Topics in Catalysis 2006, 38, 223–230. Haruta M, Kobayashi T, Sano H, Yamada N. Novel gold catalysts for the oxidation of carbonmonoxide at a temperature far below 0-degrees-C. Chemistry Letters 1987, 405–408. Hutchings GJ, Grady DT. Hydrochlorination of acetylene – the effect of mercuric-chloride concentration on catalyst life. Applied Catalysis 1985, 17, 155–160. Nkosi B, Coville NJ, Hutchings GJ, Adams MD, Friedl J, Wagner FE. Hydrochlorination of acetylene using gold catalysts - a study of catalyst deactivation. J Catal 1991, 128, 366–377.

278 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

[62] [63] [64]

[65] [66] [67] [68]

[69] [70] [71]

[72]

[73] [74] [75] [76] [77] [78]

[79] [80]

[81]

[82]

Mallat T, Baiker A. Oxidation of alcohols with molecular-oxygen on platinum metal-catalysts in aqueous-solutions. Catal Today 1994, 19, 247–283. Kluytmans JHJ, Markusse AP, Kuster BFM, Marin GB, Schouten JC. Engineering aspects of the aqueous noble metal catalysed alcohol oxidation. Catal Today 2000, 57, 143–155. Keresszegi C, Grunwaldt JD, Mallat T, Baiker A. In situ EXAFS study on the oxidation state of Pd/Al2O3 and Bi-Pd/Al2O3 during the liquid-phase oxidation of 1-phenylethanol. J Catal 2004, 222, 268–280. Gallezot P. Selective oxidation with air on metal catalysts. Catal Today 1997, 37, 405–418. Venema FR, Peters JA, Vanbekkum H. Platinum-catalyzed oxidation of aldopentoses to aldaric acids. Journal of Molecular Catalysis 1992, 77, 75–85. Pinxt H, Kuster BFM, Marin GB. Promoter effects in the Pt-catalysed oxidation of propylene glycol. Appl Catal A-Gen 2000, 191, 45–54. Schuurman Y, Kuster BFM, Vanderwiele K, Marin GB. Selective oxidation of methyl alpha-dglucoside on carbon supported platinum .3. Catalyst deactivation. Appl Catal A-Gen 1992, 89, 47–68. Mallat T, Bodnar Z, Baiker A. Promotion and deactivation of platinum catalysts in liquidphase oxidation of secondary alcohols. Acs Symposium Series 1993, 523, 308–317. Mallat T, Bodnar Z, Hug P, Baiker A. Selective oxidation of cinnamyl alcohol to cinnamaldehyde with air over Bi-Pt/alumina catalysts. J Catal 1995, 153, 131–143. Mori K, Yamaguchi K, Hara T, Mizugaki T, Ebitani K, Kaneda K. Controlled synthesis of hydroxyapatite-supported palladium complexes as highly efficient heterogeneous catalysts. Journal of the American Chemical Society 2002, 124, 11572–11573. Anderson R, Griffin K, Johnston P, Alsters PL. Selective oxidation of alcohols to carbonyl compounds and carboxylic acids with platinum group metal catalysts. Advanced Synthesis & Catalysis 2003, 345, 517–523. Mounzer HN, Wood J, Stitt EH. Heterogeneous oxidation of 2-octanol on 5 wt % Pt1 wt % Bi/Carbon catalyst. Chem Eng Sci 2010, 65, 179–185. Shaikh A, Al-Dahhan M. Scale-up of Bubble Column Reactors: A Review of Current State-ofthe-Art. Ind Eng Chem Res 2013, 52, 8091–8108. Dudukovic MP, Devanathan N, Holub R. Multiphase reactors – models and experimentalverification. Revue De L Institut Francais Du Petrole 1991, 46, 439 ff. Degaleesan S, Dudukovic M, Pan Y. Experimental study of gas-induced liquid-flow structures in bubble columns. Aiche Journal 2001, 47, 1913–1931. Jakobsen HA, Lindborg H, Dorao CA. Modeling of bubble column reactors: Progress and limitations. Ind Eng Chem Res 2005, 44, 5107–5151. Gupta A, Roy S. Euler-Euler simulation of bubbly flow in a rectangular bubble column: Experimental validation with Radioactive Particle Tracking. Chemical Engineering Journal 2013, 225, 818–836. Dudukovic MP, Larachi F, Mills PL. Multiphase reactors – revisited. Chem Eng Sci 1999, 54, 1975–1995. Boyes AP, Chugtai A, Khan Z, Raymahasay S, Sulidis AT, Winterbottom JM. The cocurrent downflow contactor (CDC) as a fixed-bed and slurry reactor for catalytic-hydrogenation. Journal of Chemical Technology and Biotechnology 1995, 64, 55–65. Boyes AP, Lu XX, Raymahasay S, Saremento S, Tilston MW, Winterbottom JM. The cocurrent downflow contactor (CDC) - mass-transfer and reaction characteristics in unpacked and packed-bed operation. Chemical Engineering Research & Design 1991, 69, 200–202. Winterbottom JM, Khan Z, Boyes AP, Raymahasay S. Photocatalyzed oxidation of phenol in water using a cocurrent downflow contactor reactor (CDCR). Environmental Progress 1997, 16, 125–131.

References |

[83]

279

Ochuma IJ, Fishwick RP, Wood J, Winterbottom JM. Optimisation of degradation conditions of 1,8-diazabicyclo 5.4.0 undec-7-ene in water and reaction kinetics analysis using a cocurrent downflow contactor photocatalytic reactor. Applied Catalysis B-Environmental 2007, 73, 259– 268. [84] Ochuma IJ, Fishwick RP, Wood J, Winterbottoom JM. Photocatalytic oxidation of 2,4,6-trichlorophenol in water using a cocurrent downflow contactor reactor (CDCR). Journal of Hazardous Materials 2007, 144, 627–633. [85] Ochuma IJ, Osibo OO, Fishwick RP, et al. Three-phase photocatalysis using suspended titania and titania supported on a reticulated foam monolith for water purification. Catal Today 2007, 128, 100–107. [86] Al-Dahhan MH, Khadilkar MR, Wu Y, Dudukovic MP. Prediction of pressure drop and liquid holdup in high-pressure trickle-bed reactors. Ind Eng Chem Res 1998, 37, 793–798. [87] Handbook of petroleum refining processes. McGraw-Hill, 2004. at http://site.ebrary.com/id/ 10180024.) [88] Vergel C, Euzen JP, Trambouze P, Wauquier JP. 2-phase flow catalytic reactor, influence of hydrodynamics on selectivity. Chem Eng Sci 1995, 50, 3303–3312. [89] Herrmann U, Emig G. Liquid phase hydrogenation of maleic anhydride and intermediates on copper-based and noble metal catalysts. Ind Eng Chem Res 1997, 36, 2885–2896. [90] McManus RL, Funk GA, Harold MP, Ng KM. Experimental-study of reaction in trickle-bed reactors with liquid maldistribution. Ind Eng Chem Res 1993, 32, 570–574. [91] Fortuny A, Font J, Fabregat A. Wet air oxidation of phenol using active carbon as catalyst. Applied Catalysis B-Environmental 1998, 19, 165–173. [92] Maiti RN, Sen PK, Nigam KDP. Trickle-bed reactors: Liquid distribution and flow texture. Reviews in Chemical Engineering 2004, 20, 57–109. [93] Baker D. Simultaneous Flow of Oil and Gas. Oil and Gas J 1954, 53, 183–195. [94] Al-Naimi SA, Al-Sudani FTJ, Halabia EK. Hydrodynamics and flow regime transition study of trickle bed reactor at elevated temperature and pressure. Chemical Engineering Research & Design 2011, 89, 930–939. [95] Burghardt A, Bartelmus G, Janecki D, Szlemp A. Hydrodynamics of a three-phase fixed-bed reactor operating in the pulsing flow regime at an elevated pressure. Chem Eng Sci 2002, 57, 4855–4863. [96] Anadon LD, Sederman AJ, Gladden LF. Mechanism of the trickle-to-pulse flow transition in fixed-bed reactors. Aiche Journal 2006, 52, 1522–1532. [97] Gladden LF, Anadon LD, Dunckley CP, Mantle MD, Sederman AJ. Insights into gas-liquid-solid reactors obtained by magnetic resonance imaging. Chem Eng Sci 2007, 62, 6969–6977. [98] Gladden LF, Anadon LD, Lim MHM, Sederman AJ, Stitt EH. Insights into the mechanism of the trickle-to-pulse transition in trickle-bed reactors. Ind Eng Chem Res 2005, 44, 6320–6331. [99] Anadon LD, Sederman AJ, Gladden LF. Rationalising MRI, conductance and pressure drop measurements of the trickle-to-pulse transition in trickle beds. Chem Eng Sci 2008, 63, 4640–4648. [100] Sederman AJ, Mantle MD, Dunckley CP, Huang ZY, Gladden LF. In situ MRI study of 1-octene isomerisation and hydrogenation within a trickle-bed reactor. Catalysis Letters 2005, 103, 1–8. [101] Kulkarni RR, Wood J, Winterbottom JM, Stitt EH. Effect of fines and porous catalyst on hydrodynamics of trickle bed reactors. Ind Eng Chem Res 2005, 44, 9497–9501. [102] Iliuta I, Thyrion FC, Muntean O. Residence time distribution of the liquid in two-phase cocurrent downflow in packed beds: Air/Newtonian and non-Newtonian liquid systems. Can J Chem Eng 1996, 74, 783–796.

280 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

[103] Iliuta I, Larachi F, Al-Dahhan MH. Multiple-zone model for partially wetted trickle flow hydrodynamics. Chemical Engineering Research & Design 2000, 78, 982–990. [104] Khadilkar MR, Mills PL, Dudukovic MP. Trickle-bed reactor models for systems with a volatile liquid phase. Chem Eng Sci 1999, 54, 2421–2431. [105] Ostrovskii NM, J. W. Reaction and Capillary Condensation in Dispersed Porous Particles. Surfactant science series 2006, 130, 601–640. [106] Wood J, Gladden LF. Modelling diffusion and reaction accompanied by capillary condensation using three-dimensional pore networks. Part 1. Fickian diffusion and pseudo-first-order reaction kinetics. Chem Eng Sci 2002, 57, 3033–3045. [107] Wood J, Gladden LF, Keil FJ. Modelling diffusion and reaction accompanied by capillary condensation using three-dimensional pore networks. Part 2. Dusty gas model and general reaction kinetics. Chem Eng Sci 2002, 57, 3047–3059. [108] Wang YN, Chen JW, Larachi F. Modelling and simulation of trickle-bed reactors using computational fluid dynamics: A state-of-the-art review. Can J Chem Eng 2013, 91, 136–180. [109] Hickman DA, Holbrook MT, Mistretta S, Rozeveld SJ. Successful Scale-up of an Industrial Trickle Bed Hydrogenation Using Laboratory Reactor Data. Ind Eng Chem Res 2013, 52, 15287–15292. [110] De Vos DE, Jacobs PA. Heterogenization of Mn and Fe complex oxidation catalysts. Catal Today 2000, 57, 105–114. [111] Augustine RL, Tanielyan SK, Mahata N, Gao Y, Zsigmond A, Yang H. Anchored homogeneous catalysts: the role of the heteropoly acid anchoring agent. Appl Catal A-Gen 2003, 256, 69– 76. [112] Augustine RL, Goel P, Mahata N, Reyes C, Tanielyan SK. Anchored homogeneous catalysts: high turnover number applications. Journal of Molecular Catalysis a-Chemical 2004, 216, 189–197. [113] Young JF, Osborn JA, Jardine FH, Wilkinso.G. Hydride intermediates in homogeneous hydrogenation reactions of olefins and acetylenes using rhodium catalysts. Chemical Communications 1965, 131 ff. [114] Vries JGdE, Cornelis J. The handbook of homogeneous hydrogenation. Weinheim; [Great Britain], Wiley-VCH, 2007. [115] Baiker A. Progress in asymmetric heterogeneous catalysis: Design of novel chirally modified platinum metal catalysts. Journal of Molecular Catalysis a-Chemical 1997, 115, 473–493. [116] Toukoniitty E, Maki-Arvela P, Neyestanaki AK, Salmi T, Murzin DY. Continuous hydrogenation of 1-phenyl-1,2-propanedione under transient and steady-state conditions: regioselectivity, enantio selectivity and catalyst deactivation. Appl Catal A-Gen 2002, 235, 125–138. [117] Kunzle N, Soler JW, Baiker A. Continuous enantioselective hydrogenation in fixed-bed reactor: towards process intensification. Catal Today 2003, 79, 503–509. [118] Kunzle N, Soler JW, Mallat T, Baiker A. Enantioselective hydrogenation on palladium – Limitations of continuous fixed-bed reactor operation. J Catal 2002, 210, 466–470. [119] Al Herz MA, Tsoligkas AN, Simmons MJH, Wood J. Enantioselective hydrogenation of dimethyl itaconate with immobilised rhodium-duphos complex in a recirculating fixed-bed reactor. Appl Catal A-Gen 2011, 396, 148–158. [120] Osborn JA, Jardine FH, Young JF, Wilkinson G. The preparation and properties of tris(triphenylphosphine)halogenorhodium(I) and some reactions thereof including catalytic homogeneous hydrogenation of olefins and acetylenes and their derivatives. Journal of the Chemical Society A: Inorganic, Physical, Theoretical 1966, 1711–1732. [121] Shah A, Fishwick R, Wood J, Leeke G, Rigby S, Greaves M. A review of novel techniques for heavy oil and bitumen extraction and upgrading. Energy & Environmental Science 2010, 3, 700–714.

References |

281

[122] Shah A, Fishwick RP, Leeke GA, Wood J, Rigby SP, Greaves M. Experimental Optimization of Catalytic Process In Situ for Heavy-Oil and Bitumen Upgrading. Journal of Canadian Petroleum Technology 2011, 50, 33–47. [123] Hart A, Shah A, Leeke G, Greaves M, Wood J. Optimization of the CAPRI Process for Heavy Oil Upgrading: Effect of Hydrogen and Guard Bed. Ind Eng Chem Res 2013, 52, 15394–15406. [124] Christian, Mitchell M, Kenis PJA. Ceramic microreactors for on-site hydrogen production from high temperature steam reforming of propane. Lab on a Chip 2006, 6, 1328–1337. [125] Ruettinger W, Ilinich O, Farrauto RJ. A new generation of water gas shift catalysts for fuel cell applications. Journal of Power Sources 2003, 118, 61–65. [126] Brooks KP, Hu J, Zhu H, Kee RJ. Methanation of carbon dioxide by hydrogen reduction using the Sabatier process in microchannel reactors. Chem Eng Sci 2007, 62, 1161–1170. [127] Guettel R, Kunz U, Turek T. Reactors for Fischer-Tropsch synthesis. Chem Eng Technol 2008, 31, 746–754. [128] Sung M, Kato S, Kawanami F, Sudo M. Evaluation of an air-cleaning unit having photocatalytic sheets to remove acetaldehyde from indoor air. Building and Environment 2010, 45, 2002– 2007. [129] Van Gerven T, Mul G, Moulijn J, Stankiewicz A. A review of intensification of photocatalytic processes. Chemical Engineering and Processing 2007, 46, 781–789. [130] Lisi L, Pirone R, Russo G, Stanzione V. Cu-ZSM5 based monolith reactors for NO decomposition. Chemical Engineering Journal 2009, 154, 341–347. [131] Dhanushkodi SR, Mahinpey N, Wilson M. Kinetic and 2D reactor modeling for simulation of the catalytic reduction of NOx in the monolith honeycomb reactor. Process Safety and Environmental Protection 2008, 86, 303–309. [132] Cybulski A, Stankiewicz A, Albers RKE, Moulijn JA. Monolithic reactors for fine chemicals industries: A comparative analysis of a monolithic reactor and a mechanically agitated slurry reactor. Chem Eng Sci 1999, 54, 2351–2358. [133] Kreutzer MT, Du P, Heiszwolf JJ, Kapteijn F, Moulijn JA. Mass transfer characteristics of threephase monolith reactors. Chem Eng Sci 2001, 56, 6015–6023. [134] Tsoligkas AN, Simmons MJH, Wood J, Frost CG. Kinetic and selectivity studies of gas-liquid reaction under Taylor flow in a circular capillary. Catal Today 2007, 128, 36–46. [135] Tsoligkas AN, Simmons MJH, Wood J. Two phase gas-liquid reaction studies in a circular capillary. Chem Eng Sci 2007, 62, 5397–5401. [136] Mogalicherla AK, Kunzru D. Performance of monolithic reactors in film flow. Chemical Engineering Research & Design 2010, 88, 1057–1066. [137] Nijhuis TA, Kreutzer MT, Romijn ACJ, Kapteijn F, Moulijn JA. Monolithic catalysts as more efficient three-phase reactors. Catal Today 2001, 66, 157–165. [138] Enache DI, Landon P, Lok CM, Pollington SD, Stitt EH. Direct comparison of a trickle bed and a monolith for hydrogenation of pyrolysis gasoline. Ind Eng Chem Res 2005, 44, 9431–949. [139] Pollington SD, Enache DI, Landon P, et al. Enhanced selective glycerol oxidation in multiphase structured reactors. Catal Today 2009, 145, 169–175. [140] Patrick TA, Abraham MA. Evaluation of a monolith-supported Pt/Al2O3 catalyst for wet oxidation of carbohydrate-containing waste streams. Environmental Science & Technology 2000, 34, 3480–3488. [141] Rodrigues CP, da Silva VT, Schmal M. Partial oxidation of ethanol over cobalt oxide based cordierite monolith catalyst. Applied Catalysis B-Environmental 2010, 96, 1–9. [142] Albers RE, Nystrom M, Siverstrom M, et al. Development of a monolith-based process for H2O2-production: from idea to large-scale implementation. Catal Today 2001, 69, 247–252. [143] Thulasidas TC, Cerro RL, Abraham MA. The monolith froth reactor - residence time modeling and analysis. Chemical Engineering Research & Design 1995, 73, 314–319.

282 | 6 Three-phase catalytic reactors for hydrogenation and oxidation reactions

[144] Stutz MJ, Poulikakos D. Optimum washcoat thickness of a monolith reactor for syngas production by partial oxidation of methane. Chem Eng Sci 2008, 63, 1761–1770. [145] Kreutzer MT, Kapteijn F, Moulijn JA. Shouldn’t catalysts shape up? Structured reactors in general and gas-liquid monolith reactors in particular. Catal Today 2006, 111, 111–118. [146] Satterfield CN, Ozel F. Some characteristics of 2-phase flow in monolithic catalyst structures. Industrial & Engineering Chemistry Fundamentals 1977, 16, 61–67. [147] Taitel Y, Bornea D, Dukler AE. Modeling flow pattern transitions for steady upward gas-liquid flow in vertical tubes. Aiche Journal 1980, 26, 345–354. [148] Tsoligkas A, Simmons MJH, Wood J. Influence of orientation upon the hydrodynamics of gasliquid flow for square channels in monolith supports. Chem Eng Sci 2007, 62, 4365–4378. [149] Stitt EH, Jackson SD, Shipley DG, King F. Modelling propane dehydrogenation in a rotating monolith reactor. Catal Today 2001, 69, 217–226. [150] Subramanian R, Panuccio GJ, Krummenacher JJ, Lee IC, Schmidt LD. Catalytic partial oxidation of higher hydrocarbons: reactivities and selectivities of mixtures. Chem Eng Sci 2004, 59, 5501–5507. [151] Hoelderich WF. Environmentally benign manufacturing of fine and intermediate chemicals. Catal Today 2000, 62, 115–130. [152] Molga EJ, Westerterp KR. Kinetics of the hydrogenation of 2,4-dinitrotoluene over a palladium on alumina catalyst. Chem Eng Sci 1992, 47, 1733–1749. [153] Benaissa M, LeRoux GC, Joulia X, Chaudhari RV, Delmas H. Kinetic modeling of the hydrogenation of 1,5,9-cyclododecatriene on Pd/Al2O3 catalyst including isomerization. Ind Eng Chem Res 1996, 35, 2091–2095.

Teuvo Kilpiö, Vincenzo Russo, Kari Eränen, and Tapio Salmi

7 Design and modeling of laboratory scale three-phase fixed bed reactors Nomenclature a, a0 , a f a A Ap av B Ci C i,G C i,L C i,F IN C i,F C i,S IN C i,S c P,F c P,L c P,S D A,B D0A,B , D0B,A D0A,i D A,M D a,L D A,L D e,i D i,L D r,F D z,F dK

Activity, initial, final Area-to-volume ratio Parameter External surface area of particle Specific surface area of particles Parameter Local concentration of component i in liquid, i = A, B, C, D, H Concentration of component i in gas Concentration of component i in liquid, i = A, B, C Concentration of component i in the fluid phase Initial concentration of component i in the fluid phase Concentration of component i in the solid Initial concentration of component i in the solid Specific heat of the fluid, liquid Specific heat of the solid Diffusivity of A in solvent B Infinite dilution diffusivity of A in B and B in A Infinite dilution diffusivity of A in i Diffusivity in mixture Axial dispersion coefficient in liquid Diffusivity of component A in liquid Effective diffusivity of the compound i, i = A, B, C Diffusivity of component i in liquid Radial dispersion coefficient of the fluid Axial dispersion coefficient of the fluid Krischer-Kast hydraulic diameter dK = dp √ 3

dp E1 , E2 f g Ga G , Ga L Geo H GL hw i J GL k, k j kd

[] [m2 /m3 ] [] [m2 ] [m2 /m3 ] [] [mol/l] [mol/l] [mol/l] [mol/m3 ] [mol/m3 ] [mol/m3 ] [mol/m3 ] [J/(kg K)] [J/(kg K)] [m2 /s] [m2 /s] [m2 /s] [m2 /s] [m2 /s] [m2 /s] [m2 /s] [m2 /s] [m2 /s] [m2 /s] [m]

16ε3b 9π(1 − ε b )2

Particle diameter Ergun equation constants Friction factor Gravity constant Galileo number for gas and liquid Dimensionless geometry dependent variable Gas-liquid heat transfer rate Wall heat transfer coefficient Reaction rate order Gas-liquid mass transfer rate Reaction rate constant, units to express the rate, j for reaction j Deactivation rate constant units to express the activation change

[m] [] [] [m2 /s] [] [] [J/(m3 s)] [W/(m2 K)] [] [mol/(m3 s)] [mol/(g s)] [1/s]

284 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

kl kl a kr Kj Keq L M MB mL n ni Nu m p p1 , p4 , p5 Pe r ri r i,j r i,S r i,x rp R RP Rg Re L Re G s Sc L Sh T TF T FIN TS T SIN Tw t uF vA VA , VB VL , VG Vp , Vx wF wG wL We L xA , xB xi

Liquid-solid mass transfer coefficient Combined gas-liquid mass transfer coefficient Radial thermal conductivity Adsorption parameters for component j = A, B, C, H, K, L Equilibrium constant Reactor length Total number of volume elements Molar mass of solvent Mass flux of liquid Exponent Adsorption exponent for component i Nusselt number Exponent Pressure Reaction rate constant, deactivation rate constant and final activity in Figs. 7.13–7.17 Peclet number Radial location, in particle or catalyst bed Reaction rate for component i ; i = A, B, C Reaction rate of component i in reaction j Reaction rate at the surface of the particle, A, B, C Reaction rate at location x Distance from the particle centre Reactor radius Particle radius Ideal gas constant Reynolds number of liquid Reynolds number of gas Surface shape factor Schmidt number of liquid Sherwood number Temperature Temperature of the fluid Initial temperature of the fluid Temperature of the solid Initial temperature of the solid Temperature of the wall Time Fluid velocity Stoichiometric coefficient of component A Molar volume of component A and B Volumetric flowrate of liquid and gas Volume of the particle and volume element x Superficial fluid velocity Superficial gas velocity Superficial liquid velocity Weber number for liquid Mole fractions of component A and solvent Volume fraction

[1/s] [1/s] [W/(m K)] [l/mol] [] [m] [] [kg/kmol] [kg/m2 s] [] [] [] [] [Pa]

[] [m] [mol/(g s)] [mol/(g s)] [mol/(g s)] [mol/(g s)] [] [m] [m] [J/(mol K)] [] [] [] [] [] [K] [K] [K] [K] [K] [K] [s] [m/s] [] [cm3 /mol] [ml/min] [m3 ] [m/s] [m/s] [m/s] [] [] []

7.1 Background

XG z βnc ΔHads,i −ΔH r dp/dz, Δp/Δz αi α α α β βi γI δi εL εF εB εP η e,i λS , λB , λp , λF , λL λ z,F λ r,F μL , μG , μM , μi μB ν ρL , ρG , ρB , ρF ρcat σ υ i,j ϕL ϕ ψL , ψG

Lockhardt-Martinelli ratio Axial location Non capillary liquid hold-up Adsorption energies for component i, i = A, B, C Reaction enthalpy Pressure gradient Exponent for component i Parameter Maximum coverage Thermodynamic correction factor for diffusivity in liquid Parameter Exponent for component i Exponent for component k Exponent for component l Liquid hold-up Fluid void fraction Porosity of the catalyst bed Particle porosity Effectiveness factor of component i Conductivity of solid, bed, particles, fluid, and liquid Axial heat conductivity Radial heat conductivity Liquid, gas and mixture and component i viscosities Viscosity of solvent Kinematic viscosity Density of liquid, gas, particles and fluid Catalyst mass concentration Surface tension Stoichiometric coefficient for component i in phase j Dynamic liquid fraction Dissociation factor Ergun equation left hand sides for gas and liquid phases

| 285

[] [m] [] [J/mol] [J/mol] [Pa/m] [] [] [] [] [] [] [] [] [] [] [] [] [] [W/(m K)] [W/(m K)] [W/(m K)] [kg/(m s)] [cP] [m2 /s] [kg/m3 ] [kgcat /m3 ] [N/m] [] [] []

7.1 Background Heterogeneous catalysis is one of the key technologies of our modern, highly industrialized society. Solid heterogeneous catalysts enhance the rates of chemical reactions without being consumed in them. This very fundamental definition, which dates back to the famous Swedish chemist Jöns Jacob Berzelius (1835), is still valid. The development of solid catalysts, such as iron-based catalysts for ammonia synthesis, nickel and noble metal catalysts for hydrogenation, cobalt-molybdenium catalysts for hydrodesulphurization and hydrodeoxygenation and zeolite catalysts for hydrocarbon transformation, has guaranteed us a continuous and increasing delivery of fertilizers and fuel components, it has enabled a high standard of living for millions of people and it has also enabled the enormous growth of the human population, with all its advan-

286 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

tages and disadvantages. Besides the production of bulk components, heterogeneous catalysis plays an enormous role in the preparation of fine and specialty chemicals, as well as pharmaceuticals and alimentary products. Drugs, perfumes, sweeteners, herbicides and pesticides are produced with the aid of heterogeneous catalysts. The chemical applications are wide, but from the viewpoint of chemistry, physics and chemical engineering, all the man-made catalytic processes have similar features. It is usually claimed that more than 90 % of the industrially applied chemical processes are based on the use of catalysts, but 100 % of them take place in chemical reactors, i.e. in equipment particularly designed to carry out the catalytic transformation in a best possible way, so that a high productivity and product selectivity is achieved. Good productivity combined with high selectivity is the primary goal for any catalyst development. These two targets are hard to achieve simultaneously for most applications and, therefore, an enormous amount of experimental effort is demanded worldwide, in academia and in industry. As a reward of the collective effort, catalysts have managed to revolutionize individual chemical production processes. High selectivity often implies a simpler process concept and, consequently, lower investment and operating costs. In most cases, however, chemical industry has to cope with multi-component feedstock and, in addition, by-product formation is often inevitable. Therefore, complicated separation steps are often required for final recovery of a desired main product and valuable by-products. Many of the chemical applications of heterogeneous catalysis comprise threephase systems, in which a gas and a liquid phase interact with a solid catalyst. Typical examples are catalytic hydrogenation, hydrodesulphurization and hydrodeoxygenation processes, in which gaseous hydrogen react with an organic phase. A multitude of reactor constructions are available for continuous, heterogeneously catalyzed gasliquid reactors. All these reactors have to satisfy the following two demands. First, the gas has to be brought into a very good contact with the liquid to minimize the gasliquid mass transfer resistance. Secondly, liquid has to be brought into a good contact with the catalyst, since the reaction takes place only when the dissolved reactants in the liquid bulk phase reach the active sites on the catalyst surface. The workhorse for catalytic three-phase processes is a fixed bed reactor, in which the solid catalyst exists either in form of particles or fixed structures, such as structured elements or monoliths. Fixed bed reactors try to meet the respective demands by having the liquid flow on particle surfaces, thus providing a large surface area for mass transfer to occur, or by using porous catalysts, the pores of which are filled with the liquid by capillary forces. Continuous fixed bed reactors are practical for long duration catalyst screening because they do not demand frequent filling and emptying as batch and semi-batch reactors do. Fixed beds are especially well suited for studying how well the solid catalyst keeps its long-term activity over the time. If a parallel set of multiple continuous reactors is used, operating in such a way that each reactor is either under different process conditions (pressure, temperature, liquid and gas flows) or loaded with different

7.2 Reactor set- up

|

287

catalysts, a large amount of experimental data can be generated simultaneously, although the sampling has yet to be taken sequentially. Fixed bed reactors are also used in order to eliminate the possible losses of catalyst.

7.2 Reactor set- up A fixed bed reactor system designed by us for catalyst and flow screening as well as kinetic studies is displayed in Fig. 7.1. Several tubular reactors are placed in parallel in an oven to guarantee a homogeneous background temperature and to keep the fluidity of the liquid-phase components high enough for feeding. Each reactor is heated separately by a heating element, which implies that several reaction temperatures can be screened simultaneously. The gas and liquid feeds are controlled by mass flow controllers and HPLC pumps, respectively. In this way, the gas and liquid feeds can be altered independently. Catalyst particles or structured elements are placed in the reactor tubes. Typically, the catalyst is diluted with inert material to keep the individual reactor as isothermal as possible. The temperature in the bed is measured by thermocouples. Pressure and temperature data are stored on a computer. Gas and liquid-phase samples are analyzed on-line or off-line, depending on the chemical case. The equipment is particularly suitable for the investigation of catalytic two and threephase systems, but it can be used for non-catalytic applications, too. The solid catalyst Liquid feed

Heater

Inert material for liquid distribution

Gas feed

I.D. = 10 mm

Catalyst mixed with inert material Stainless steel net

Thermocouple Fig. 7.1: Parallel reactor set-up for reasonable throughput screening in laboratory scale [46]. Reprinted with permission from {Ind. Eng. Chem. Res., 2012, 51 (26), pp 8858–8866}. Copyright {2012} © American Chemical Society.

288 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

can be efficiently pretreated before its use; for instance, a hydrogenation catalyst can be reduced without any contamination at elevated temperatures under hydrogen flow before switching the reactant flow on. By varying the flow rates, the effect of the liquid and gas residence times is screened and the time-on-stream behaviors of the outlet concentrations reveal the catalyst stability. The role of internal diffusion in the catalyst particles is investigated by placing particles of different sizes in the parallel reactor tubes. Because the chemical analysis is the bottle-neck in the parallel screening of catalysts, it does not make sense to couple too many reactors in the system; in any case, off-line analysis becomes the limiting step in the research project. Therefore, instead of high throughput screening, we call the concept reasonable throughput screening.

7.3 Physical and chemical phenomena in fixed bed reactors 7.3.1 Overview The major issues concerned in fixed bed studies are summarized in Fig. 7.2. Fixed bed reactors operating under the trickling flow regime are among the most commonly selected continuous reactor alternatives and are, therefore, the focal point of interest in both experimental and modeling studies. With the experience presented in open literature [1–8], the modeling of three-phase fixed bed reactors still remains challenging for every new reaction system, because the extent and the importance of each individual effect is very case-specific and strongly dependent on actual operating conditions (temperature, pressure, liquid and gas flow rates and feed concentrations), physical properties of gases and liquids in question and the properties of the catalyst (activity, particle size and shape, porosity, pore structure, metal loading, metal dispersion, oxidation state). Fig. 7.2 summarizes the key issues to be faced and the means to cope with these challenges in the experimental and modeling studies of continuous fixed bed reactors operating at the trickling flow regime. Just a glance at this overview of topics gives an immediate reminder that these kinds of reactors are surprisingly complicated systems, both to study experimentally and especially to model mathematically, although they initially might look like simple ones (just a pipe filled with particles and flows). The complicated nature arises from the facts that three phases are always involved, that multiple reactions proceed simultaneously and that the particles of various sizes typically have irregular shapes in a laboratory scale. Although three-phase systems are so commonly used in the chemical industry, they are among the most challenging ones to be modeled. Complicated geometry is also well known to add further challenges especially for detailed description of fluid dynamics. This implies that several simplifications are required to progress with the modeling effort of a continuous fixed bed reactor operating in the trickling flow regime.

7.4 Research targets and topics |

A. 1. 2. 3. 4. 5.

Reaction kinetics Mechanism, orders Rate constants Adsorption T dependency Deactivation

B. Mass transfer 1. Gas-liquid, liquid-solid 2. Internal

G. Modelling steps 1. Verification with simpler cases 2. Modelling of individual phenomenon 3. Modelling of a simplified reaction system 4. Sensitivity studies for each variable prior to estimation 5. Narrowing of search ranges 6. Minimisation of number of parameters 7. Improving identification 8. Parameter estimation

E. Experimental design 1. Desirable product(s) 2. Catalyst, support, active metal 3. Concentration range 4. Operation conditions: T,p, flows 5. Reactor geometry: d, h 6. Particle geometry: dp, shape, size distribution 7. Liguid feed distribution 8. Start-up procedure. Pulsed flow 9. Catalyst dilution 10. Residence time of liquid

High productivity & selectivity

C. Heat transfer 1. Reactor, reactor section 2. Within particles D. Physical Properties 1. Densities, viscosities, solubilities 2. Conductivities, diffusivities 3. Properties of particles and beds

F. Liquid Flow 1. Direction (up/down) 2. Chaneling & stagnant zones 3. External wetting, wall flow 4. Pressure drop 5. Liquid hold-up

289

H. 1. 2. 3.

Numeric methods Method of lines Hoyos step-by-step Solvers: ODESSA, Euler

I. 1. 2. 3. 2.

Optimisation LevMarq, Simplex Target function Least squares Weighing, data filtration

J. Balances 1. Mass, heat & momentum 2. Balance areas: whole reactor, reactor section, particle 3. Separate or simultaneous solution 4. Simplifying assumptions Semiemphirical for holdup and Δp mass balance based reactor and particle model simultaneous heat & mass balances: for reactor section and for particle

Fig. 7.2: Major issues of continuous fixed bed reactor studies [9].

7.4 Research targets and topics 7.4.1 Catalyst selection High productivity and selectivity are the obvious goals for reaction engineering and reactor design. The intrinsic kinetics are strongly dependent on the catalyst used. Thus, the catalyst selection and preparation both play a vital role in the success of all the conducted studies. The catalysts used in laboratory-scale reactors are either tailormade or commercial ones. The key issues in the operation and modeling of fixed beds for three-phase processes are summarized below.

290 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

7.4.2 Reaction kinetics Temperature and concentration-dependent multiple reaction kinetics determine largely both the productivity and selectivity (if the system is not under mass transfer control). Therefore, modeling of the reaction kinetics is an essential part of all our studies. Each reactant must reach and adsorb the active sites of the catalyst to react there. This may limit the reaction rates, especially, when dealing with concentrated solutions. Therefore, also the adsorption kinetics are to be considered in the modeling study. Deactivation is a phenomenon to be expected in long-term continuous operation sooner or later.

7.4.3 Mass transfer effects The required steps of mass transfer (from gas to liquid phase, from liquid to solid surface and finally internal diffusion) can limit both productivity and selectivity, the more so, the faster the intrinsic reaction rates are.

7.4.4 Heat effects Whenever exothermic reactions produce too much heat to be effectively transferred away, temperature gradients emerge. The most severe ones typically are radial temperature gradients inside the reactor and within individual particles (in normal cases where the axial temperature is tried to be kept constant by means of a cooling jacket).

7.4.5 Physical properties of gases and liquids Physical properties of the solutions determine not only how well the fluids flow and how much liquid is retained in the fixed bed, but also how well mass and heat are transferred. By a proper selection of the operation conditions and catalyst/support properties, the physical properties can be adjusted to be more favorable; in that way, the reactor performance can be enhanced.

7.4.6 Reactor design and operation policy The details of the reactor design (particle sizes and shapes, feed distribution, start-up procedure, catalyst loading, bed dilution) all play an essential role in determining how ideal flow fields (without channeling and wall flow) are generated and how effectively the reactor is used. One can suppress and minimize the effects of undesirable heat and mass transfer as well as flow related phenomena by proper selections of reactor de-

7.5 Experimental design

| 291

sign and operation conditions. Mathematical modeling in many cases works hand in hand with experimental design. On the one hand, it sets demands for the experimental work and, on the other hand, it tries to extract a maximum amount of know-how from experiments conducted to figure out and explain what really took place in the reactor. A very essential part of the experimental design is to generate ideas on how the interrelated phenomena could be separated from each other and how an extensive modeling task can be divided into multiple smaller sub-tasks. Gravity driven down-flow of a liquid was used in all our cases. The formation of ideal flow fields of liquid was, on the one hand, aimed at by using a fine sand bed feed distributor; on the other hand, by diluting the catalyst with small inert particles.

7.4.7 Modeling options The solutions of the modeling tasks are strongly based on the application of numerical mathematical methods and in most cases demand case-specific parameter optimizations. Many of the models for individual phenomena are to be fine-tuned by parameter adjustment. Efficient algorithms for the minimization of the least square differences between observations and predictions are widely used. Data filtration is used whenever experimental points are clearly out of the pattern. Weighing is in some cases applied to improve the model accuracy in the vicinity of the most interesting results (results providing the highest productivities and selectivities). The models have their origins in the conservation laws of nature: mass, energy and momentum balances. Momentum balances are often systematically replaced by semi-empirical expressions, and the liquid flow is described by a simple interpretation of results by using an axial dispersion model for backmixing and semi-empirical equations for liquid hold-up and pressure drop. In many isothermal cases, mass balances and semi-empirical expressions together with kinetic expressions (including catalyst deactivation) can be used to describe the reaction system in a satisfactory manner. A simultaneous solution of dynamic and steady state mass and heat transfer models is demanded when heat generation via reactions is considerable.

7.5 Experimental design Experimental design is a crucially important issue for the success of a research project devoted to catalytic three-phase systems. The main features of experimental design are summarized in Fig. 7.3.

292 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

Experimental design A. Desirable targets a) Directly value adding materials. b) Platform chemicals. c) Simplified production processes.

F. Particle geometry a) The smaller, the less serious T and C gradients. b) Smaller particles evens flow. c) higer Δp. c) Large surface shape factor desirable.

B. Catalyst selection a) Based on activity and selectivity. b) Candidates mentioned in literature. c) Expertise in our laboratory.

G. Feed distribution a) To establish fine flow patterns. b) Fine sand bed an example.

C. Concentration range a) Dilute: Eliminates undesired T and C gradients. Well suited for catalyst screening studies. b) Concentrated: Favored by industry. Exception: Recovery from waste streams.

H. Start-up procedure a) Flow patterns remain to stay. b) Pulsing feed flow leads to higher productivity. c) Initial flooding recommended.

D. Operation conditions p,T a) Temperature influences often selectivity. Max. temperature giving good selectivity. b) Pressure improves gas solubility.

I. Catalyst dilution a) Less severe T and C gradients. b) May improve flow patterns. c) Danger: Catalyst by-passing.

E. Diameter, height a) (Particle diameter)/(Reactor diameter): to eliminate chanelling and wall flow. b) Height to offer reasonable conversion.

J. Residence time a) Long enough for reactions to occur. b) Distribution unavoidable.

Fig. 7.3: Key issues in experimental design [9].

7.5.1 Targeted products Desirable targets are typically value-added product compounds (and/or) product compounds that lead the way to simplified production processes.

7.5.2 Catalyst screening The goal is first to screen best candidates for catalyst/support and then to reveal how the productivity and selectivity can be further enhanced by optimizing the operating conditions. In catalyst screening, experience on systems of a similar nature helps considerably by limiting the number of potential candidates.

7.5.3 Experimental productivity and selectivity optimization When making kinetic experiments in continuous fixed bed reactors, complete conversion is to be avoided in order to reveal the reaction kinetics. The operation conditions, T, p, u L , and u G can be selected according to the following reasoning. Typically, the

7.5 Experimental design |

293

temperature rise accelerates reaction rates exponentially. When reaching an acceptable productivity, the selectivity becomes the major issue according to which the operation temperature is to be optimally selected. The reason why the selectivity of a desirable product compound is often temperature dependent is that activation energies for various reactions are different. The generation of undesirable side products is to be avoided, because they may demand additional separation steps in down-stream processing. In order to have a high gas solubility, high partial pressure of the gaseous component is required. In shallow laboratory reactors, the liquid flow is adjusted to low in order to provide reactions enough residence time to progress in detectable extents. Gas is often fed in excess. Diluted streams are often used to reduce the undesirable mass and heat transfer effects.

7.5.4 Particle geometry In our catalyst screening tests, typically, the reactor diameter was around 1 cm or less and the reactor length was case dependent. Small catalyst particles were used to generate better flow fields and to provide a higher specific surface area without generating an unacceptably high drop in pressure within the short laboratory reactors. In our studies, the catalyst particles were sieve fractions of crushed particles. The second largest dimension was their characteristic mean size. The external surface area of a particle is well known to be proportional to the 2nd and the volume to the 3rd power of characteristic size, thus making the smallest and the largest particles of a sieve fraction very different. Mass transfer limitations normally become the most severe within the largest particles. Krisher-Kast hydraulic diameter (d K ) is frequently used as the characteristic particle size for packed beds, because it accounts for the bed porosity effect [10]. Mean sizes were calculated straight from the largest and the smallest sizes, when the particle size distribution was not available. If the particle size distribution were known, it could have been included in the reactor model [11]. The geometry of randomly packed trickle bed reactors is very complicated, especially when the catalyst bed consists of particles having a distribution of sizes and even more so when particle shapes are irregular. In CFD studies, the detailed fixed bed geometry can be taken into account only for more regular particle shapes.

7.5.5 Feed distribution and flow regimes The liquid feed is targeted to be well distributed. A simple yet effective feed distributor for small-scale operation is a fine sand bed. Complete external wetting of all catalyst particles is easier to get when liquid is initially well distributed. Fixed bed reactors are commonly operated under the trickling flow regime. Under these conditions, the way the operation is started determines to a large extent how

294 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

fine liquid flow fields are generated. Initial flooding of the bed or a short visit to the pulse flow regime a priori helps to reduce channeling and can do it, surprisingly, permanently [12]. The pulse flow regime can be a more effective operation mode than the trickling flow regime, since pulses continually even the mal-distribution of the flow. However, only higher superficial velocities of liquid and gas can generate pulsed flow. This shortens residence times, if a longer reactor is not used. The artificial generation of pulsed flow at lower flow rates has been studied simply by making feed flow variations [13]. There, high and low liquid holdup periods were forced to take their turns repeatedly with adjustable period lengths. However, the fluctuations in the liquid flow did not progress far inside the reactor. In pilot scale styrene hydrogenation, liquid feed fluctuation has been reported to give a higher yield.

7.5.6 Bed dilution Temperature gradients can be lowered by diluting the catalyst bed with highly conductive inert particles. The generated heat per unit volume of the bed depends directly on the bed dilution. The more the catalyst bed is diluted, the more difficult it becomes to avoid catalyst bypassing. Small particles are especially well suited for dilution because they improve catalyst wetting [10].

7.5.7 Residence time distribution Typically, any back-mixing decreases the productivity. Therefore, it is essential to include it in the reactor models. If dispersion takes place and is not included, this leads to an underestimation of the productivity (and erroneous values of kinetic parameters). The single parameter axial dispersion model is the standard way of taking into account the back-mixing effect and the consequent broadening of the liquid residence time distribution [3]. The axial dispersion model adds a second order spatial derivative term in the liquid mass balances. It works punctually in the two extreme cases (for the most productive mode of operation (plug flow), and the least productive mode (continuous ideal mixing of liquid)). The axial dispersion model is unable to predict a sometimes visible tail in the distribution caused by the presence of stagnant zones. For such cases, the piston flow exchange model is a more precise modeling option. It has two adjustable parameters. The liquid flow in this model is thought to consist of two streams (one flowing rapidly, the other flowing slowly) and mass exchange is modeled to take place in between these streams.

7.6 Chemical kinetics |

295

7.6 Chemical kinetics 7.6.1 Topics of the kinetic studies The intrinsic rates of heterogeneously catalyzed reactions are described with chemical kinetics; this explains how the reaction rates depend on concentrations, temperature and the nature of catalyst. Chemical kinetics together with other means of studying catalytic reactions, such as spectroscopy of catalysts, molecular modeling and calculation of the thermodynamics of reactants, intermediates and products, form the basis for understanding chemical processes. Experimental kinetic investigations are the basis of revealing reaction mechanisms. The following problems can be solved by using a kinetic model: – choosing the catalyst and comparing the selectivity and activity of various catalysts and their performance under optimum conditions for each catalyst; – the determination of the main and by-products formed during the process; – the determination of the optimum sizes and structures of catalyst grains and the necessary amount of the catalyst to achieve the specified values of the selectivity of the product and conversion of the starting material; – the determination of the short and long time stability of the catalyst; – the determination of the stability of steady states and parametric sensitivity; that is, the influence of deviations of all parameters on the steady state regime and the behavior of the reactor under unsteady state conditions; – the study of the dynamics of the process and decision if the process should be carried out under unsteady state conditions; – the study of the influence of mass and heat transfer processes on the chemical reaction rates and product selectivities as well as the determination of the kinetic region of the process; – selection of the type of a reactor and structure of the contact unit that provide the best approach to the optimum conditions. The method which simplifies the kinetic treatment and reduces the number of parameters in the mathematical model is the quasi-steady state approximation (QSSA) originally developed by Bodenstein in 1913.

7.6.2 Reaction scheme simplifications In general, a reaction is considered to be at steady state, if the concentrations of all species in each element of the reaction space (volume for a homogeneous reaction or surface for a heterogeneous reaction) do not change in time. In general, such conditions are fulfilled in open systems, such as continuous tank and tubular reactors and in flow circulation reactors. Grouping the reaction species in two categories – reaction

296 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

intermediates in one, and reactants (substrates) and products in another one – it can be stated, for reaction participants at a steady state, that if a species enters an element of the reaction space, its concentration does not change with time, while for intermediates, the production rate is equal to the consumption rate. Bodenstein proposed that there are intermediates in chemical reactions that are present in “inferior” amounts, e.g. in much lower concentrations than the major species in the mechanism. If this condition is met, the rate of change of the concentration of the intermediate can be considered negligible. The application of the steady state approximation to heterogeneous catalysis does not require that the surface concentrations are low, but implies that they do not essentially change with time. Another frequently used approximation in catalytic kinetics is the quasi-equilibrium approximation, which implies that one step in a multistep catalytic reaction sequence is much slower than the other ones. In such a case, the other steps can be considered as being close to the equilibrium, as their forward and backward rates will be close to each other. The quasi-equilibrium approach, utilized very often, limits the description of catalytic kinetics as it discards transient processes. It is, however, frequently used to obtain simplified and useful expressions for reaction rates. Typically the surface reaction rate is assumed to be rate-limiting, whereas the adsorption and desorption steps are presumed to be rapid. The approach should be supported by experimental evidence obtained by measurements of the reaction and adsorption rates. In the case of complex reactions, several steps are possible and a rigorous treatment is required comprising a multistep rate control, as the quasi-equilibrium approach cannot be applied.

7.6.3 Rate expressions The rate equations depend heavily on the particular chemical case of interest. Typically, the process involves a reaction between a large organic molecule and a much smaller reagent molecule (e.g. oxygen or hydrogen). In this case, the overall rate is often controlled by the surface reaction step, whereas the adsorption and desorption steps are assumed to be rapid. Further assumptions are needed concerning the details of the reactant adsorption. Both competitive and non-competitive adsorption models have been proposed and used to describe the kinetics. A non-competitive adsorption model can be justified by the size difference of the reacting molecules. A general overview of kinetic models is provided by [4] and [14], where also a semi-competitive adsorption model is discussed. It has been successfully applied to sugar hydrogenation [15]. Some generalization of classical kinetic models is possible. For a bimolecular and reversible reaction A + B = C + D, the rate takes the form of Eq. (7.1). R=

kinetic factor ⋅ driving force , (adsorption term)n

(7.1)

7.6 Chemical kinetics

|

297

where the adsorption term includes K A C A , etc. for molecular adsorption or (K A C A )1/2 for dissociative adsorption, and the power in denominator corresponds to the number of species in the rate determining steps, while the driving force is (1 − C C C D )/ (Keq C A C B ). The adsorption of molecules can be of a competitive nature, but in cases when the molecules represent very different sizes, such as catalytic hydrogenation of organic components, non-competitive adsorption is assumed. A rather general rate equation comprising the rate limitation by the surface reaction step as well as both competitive and non-competitive adsorption can be written as Eq. (7.2): βi

rj =

k j (∏ C αi i − ∏ C i /K j ) 𝛾k

m

(∑ K k C k + 1) (∑ K l C δl l + 1)

n

.

(7.2)

For competitive adsorption, either m = 0 or n = 0. Of course, rate equations of this kind should be derived case by case, starting from a plausible mechanism. Models with semi-competitive adsorption have been applied in the hydrogenation of aromatics and sugar molecules [16]. The principle of semi-competitive adsorption is that the large organic molecules always leave some empty space, i.e. vacant interstitial sites, where the smaller molecule (typically hydrogen and oxygen) can adsorb. Compared to the classical model, the semi-competitive adsorption model includes one additional parameter, namely the maximum coverage of the organic molecule [15]. The concept of semi-competitive adsorption of a large molecule and a smaller one implies that the maximum coverage of the large molecule is less than one, and thus some interstitial sites always remain accessible for the smaller molecule (such as hydrogen). For example, for irreversible catalytic hydrogenation of a large organic molecule (A) with molecular hydrogen (H), the following rate equation can be derived for the rate control of the surface reaction, Eq. (7.3): r=

kK A K H C A C H α((1 − α)K A C A + 1) ((1 − α)K A K H C A C H + K A C A + K H C H + 1)

2

,

(7.3)

where α is the maximum coverage of A (< 1). For the limit case α = 1 and the standard Langmuir-Hinshelwood model is obtained from Eq. (7.3). Generalizations of this concept are presented in literature [15, 16]. For highly non-ideal systems, activities should be used in the rate equations instead of concentrations.

7.6.4 Qualitative reaction rate comparison: fixed bed against batch reactors Theoretically, the modeling of an ideal plug flow reactor and an ideally mixed batch reactor is almost identical. Similar mass balance equations can be used. The reaction time in the batch reactor corresponds to the space/time in the plug flow reactor. According to this analogy, the space/time represents a measure of how far the observation point has moved following the flow. The situation demands adaptations when

298 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

focusing on multiphase flow in trickle beds. Liquid flow occupies only a part of the pipe volume and this is typically accounted for by introducing the liquid hold-up in the mass balances. The analogy still holds well if gas is abundantly available to saturate the liquid. In theory, the intrinsic kinetics on the active metal sites of the catalyst is exactly the same for both reactors, when exactly the same catalyst is used. The questions are whether all the reactants can reach the active sites and how active the sites are. Uneven liquid flow (resulting from too low liquid velocity, too large particles or complex detailed geometry) is the key reason behind the deviations between the reaction rates of trickle bed and batch reactors. Low flow velocities may allow mass transfer resistances to arise and can also cause more severe local deactivation. In a fixed bed reactor, deactivation is location-dependent and often it is at its strongest close to the feed inlets where the concentration is on top. The catalyst deactivation in a batch reactor is much more uniform since all fluid elements face similar conditions. The main reasons discussed for deviations are summarized in Fig. 7.4.

1. Hold-up local 2. Wetting incomplete 3. Residence time uneven 4. Axial dispersion present 5. Stagnant zones prevail 6. Particle size distribution 7. Particle contact places 8. Pore diffusion 9. Gas-Liquid mass transfer 10. Deactivation local

Fig. 7.4: Reasons for kinetic deviations between a fixed bed reactor and a slurry reactor [9].

7.6.5 Catalyst deactivation Sooner or later, the deactivation of a heterogeneous catalyst takes place. Generally, there are multiple reasons for it, but in our cases, the most probable reason was coking of organic reactant, which is well known to take place at elevated temperatures. Coke simply blocks pores and reactants cannot reach the active sites. Typically, the concentration curves show that after being initially fast, later on, the deactivation slows down and often finally becomes hardly detectable. Then it can be modeled using the concept of final activity. An analytical expression, Eq. (7.4) is valid for first order power law kinetics. For cases following the Langmuir-Hinshelwood-Hougen-Watson kinetics, a numerical solution is the only alternative, since coking can then only be expressed with ordinary differential equations [9]. The equations are solved together

7.7 Mass and heat transfer effects |

299

with the dynamic mass balances. A couple of equations, Eqs. (7.4) and (7.5), for the catalyst activity (a), are listed below. Analytical expression: a = a f + (a0 − a f ) exp(−k d t) Reactant concentration dependent deactivation: (a − a f )k d C A,L da − = dt (1 + K A C A,L + K B C B,L )

(7.4)

(7.5)

In practice, the reaction rate equations are corrected for the catalyst deactivation by multiplying them with the activity factor (a). The deactivation rate parameter (k d ) is temperature-dependent.

7.6.6 Truly intrinsic kinetics Highly intensive mixing and very small catalyst particles are to be used to reveal truly intrinsic kinetics. An isothermal semi-batch slurry reactor, equipped with an adjustable stirrer speed, is an exceptionally well-suited device for that purpose. From these results, the truly intrinsic activation energies can also be obtained. Otherwise, various kinds of mass transfer resistances (gas-liquid, liquid-solid or internal) may be rate-limiting. Intrinsic kinetic rate constants depend typically far stronger on temperature than mass transfer coefficients. Therefore, in cases where the reported activation energies are low, it is possible that some of the mass transfer steps may be rate-limiting [4]. Since in many published reaction rate studies it is not absolutely clear, whether or not the kinetics were truly intrinsic, the safe option is to take only the forms of the kinetic expressions from literature.

7.7 Mass and heat transfer effects 7.7.1 Mass transfer resistances In Fig. 7.5, the mass transfer effects present in a three-phase system are qualitatively illustrated. The figure shows how the reactant concentration decreases due to mass transfer resistances along the way from bulk gas down to the active sites within a porous catalyst particle. In order to react, a gas-phase compound has to go all the way from the gas to liquid, from liquid to solid and finally penetrate inside the porous solid until it reaches the active site. Gas-liquid and liquid-solid mass transfer is typically expressed by power laws of dimensionless numbers. The steady state internal diffusion is described by an ODE.

300 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

Gas

G-L mass transfer resistance L-S mass transfer resistance Liquid

Drop due to solubility

Solid Internal diffusion

Fig. 7.5: Mass transfer steps in a three-phase fixed bed reactor [9].

7.7.2 Gas-liquid mass transfer For rapid reactions, gas-liquid mass transfer may become rate-limiting if mixing is not intensive. This can happen especially when the partial pressure of the limiting gas reactant is low. Then an increase in operating pressure may help, because it directly increases the solubility and thus the mass flux. Another way to increase the partial pressure of the rate-limiting gas reactant is to increase its mass fraction in the gas feed. This cannot always be done freely, since the explosive range may set limits to the maximum gas contents. Direct synthesis of hydrogen peroxide serves as an example of this kind of behavior. Gas-liquid mass transfer restrictions can be overcome by providing a large interfacial mass transfer area and by using high superficial velocities to enhance the mass transfer. The liquid flows as a film on the surfaces of the particles and, therefore, the mass transfer area depends strongly on particle sizes, shapes and roughness of the external surfaces. For providing a reasonable residence time, the higher the superficial liquid velocity is, the longer the required reactor should be. Another way to increase liquid flow is to use liquid circulation. It suits best for a low yield system, the productivity of which is less sensitive to back-mixing. When using gas-liquid mass transfer expressions from literature [10], in the range of very low superficial velocities, the equations give very different values [19, 20]. Most of the given mass transfer correlations are power laws of dimensionless numbers. Each dimensionless number represents either a geometric or a kind of force ratio. General agreement concerning which dimensionless numbers to include is still lacking. Differences can also clearly be seen in the values of gas and liquid flow exponents. If studies cover only a narrow temperature range, the physical properties can be treated as constants. Then superficial gas and liquid velocities become the only variables on which the mass transfer coefficients depend. Gas-liquid mass transfer does not easily become a limiting factor for dilute systems when the intrinsic reaction rates are low. Mass transfer coefficient correlations are given in a review [10]. One typical gasliquid mass transfer correlation especially for the low interaction regime is given in Eq. (7.6). The united gas-liquid mass transfer coefficient is expressed as a function of

7.7 Mass and heat transfer effects |

301

various dimensionless numbers, characteristic particle dimension and molecular diffusivity. The example equation is given in the following form: k l ad2K 1/4 1/5 1/5 1/2 = 2.8 ⋅ 10−4 (X G Re L We L Sc L Geo1/4 )3.4 D A,L w G √ρ G w L √ρ L ρL wL dp Re L = μL XG =

(7.6) (7.7) (7.8)

ρ L w2L d p σ μL Sc L = ρ L D A,L av dk Geo = 1 − εB

We L =

(7.9) (7.10) (7.11)

Lockhardt Martinelli ratio, Reynolds, Weber, and Schmidt numbers of liquid and finally a geometric ratio (representing the catalyst properties) are used to calculate the united mass transfer coefficient (k l a), Eqs. (7.7)–(7.11). The gas-liquid mass transfer coefficient depends on the diffusivity of the compound in question. Gas-liquid mass transfer rates for a reactive and a non-reactive system differ from each other [21]. For a reactive system, rates are higher than for a corresponding absorption system. The reason behind the difference is probably the presence of stagnant liquid zones. In the case of absorption, these zones become saturated with gas. The saturation does not take place in a reactive system so easily because the reaction consumes the dissolved gas reactant.

7.7.3 Liquid-solid mass and heat transfer Liquid-solid mass transfer depends on local liquid velocities and, therefore, in a trickle bed, it may become restricting especially in the places where liquid flows slowly. Correlations for liquid-solid mass transfer coefficients are provided in literature. The general principle for the correlations currently available is that the liquid-solid mass transfer coefficient for the liquid film surrounding the solid catalyst particles depends on the Reynolds (Re) and Schmidt (Sc) numbers [4], Eqs. (7.12)– (7.15). From these dimensionless numbers, the Sherwood (Sh) number is calculated and the liquid-solid mass transfer coefficient (k) is obtained: β

Sh = A + BRe αL Sc L , Sh =

kL dp , D i,L

(7.12) (7.13)

302 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors ρL wL dp , μL νL . Sc = D i,L

Re L =

(7.14) (7.15)

In an analogous manner, the heat transfer coefficient (h) for the liquid film can be calculated; the Nusselt number (Nu) is obtained from the Reynolds (Re) number and the Prandtl number (Pr), Eqs. (7.16)–(7.19): Nu = A + BRe α Pr β , hL dp , λL wL dp wL dp ρL = , Re = νL μL c p,L μ L . Pr = λL

Nu =

(7.16) (7.17) (7.18) (7.19)

Coefficient A corresponds to the situation of stagnant flow and has a value around 2. The other parameters are of empirical character, typically α = 0.6–0.7; β = 13 . Coefficient B is close to 1. Liquid-solid mass and heat transfer can depend a lot also on the pre-wetting procedure. This supports the finding that, if by initial wetting, the extent of channeling can be decreased; the generated flow fields will also remain to stay.

7.7.4 Internal diffusion – pore diffusion The reaction rates are at their highest close to the feed entrance when the liquid feed stream is saturated with the gas reactants. Especially there, the reactants may not be able to diffuse as fast as the reactions could consume them and internal diffusion may become rate-limiting. Small particles can be used in laboratory-scale trickle beds, because there they do not generate a high pressure drop. In large-scale operation, the pressure drop is an operating cost to be minimized. The use of small particles cannot be afforded then, and under these circumstances the particle shape becomes the primary factor with which to increase the surface-to-volume ratio [5]. By having particles with a high surface area-to-volume ratio, the importance of internal mass transfer resistance can be reduced. According to open literature [1], the most common regular shapes are spheres, cubes, hollow cubes, cylinders, hollow cylinders, four holes cylinders, single rings, cross webs, grooved cylinders, Pall rings, Intalox saddles and Berl saddles. One surface-to-volume ratio comparison study is reported for particles originating from a cylindrical structure. Discs, cylindrical extrudates, quadrulopes, rings, hollow extrudates, wagon wheels and miniliths had the

7.8 Physical properties of gas mixtures and solutions

| 303

surface area-to-volume ratio in increasing order. The ratio for the miniliths went up to 20. Various catalyst suppliers have also special geometries that, compared to cylinders, provide a manifold external surface-to-volume ratio. The length of the diffusion path is another factor that directly determines the severity of internal mass transfer resistance. Short diffusion paths are highly desirable. They can be obtained in two ways: by well-selected particle geometry or by locating the active sites close to the external surface of the particles. In order to determine the severity of the internal diffusion, the effective diffusivity needs to be known. The most common way to calculate its value is to start from molecular diffusivities in liquid and make a simple porosity and tortuosity correction [4, 7].

7.7.5 Effectiveness factors for particles The aim of an internal diffusion and heat conduction study is to determine the effectiveness factor for each component. It is a direct measure that tells how well the active sites of a catalyst particle are in use [4, 6]. If the diffusion is rapid and provides reactants to active sites so that the surface concentrations prevail everywhere, effectiveness factor is 1. Effectiveness factors are calculated because they are directly comparable for different particle geometries while internal concentration profiles are not. The effectiveness factor is defined as: V

η e,i =

p ∫0 r i,x dV

r i,S V p

.

(7.20)

Eq. (7.20) tells that the effectiveness factor is obtained by integrating the rate obtained from the concentration profiles in the porous catalyst particle.

7.8 Physical properties of gas mixtures and solutions The operating conditions (T, p), and the composition determine the physical properties of gas and liquid mixtures. When studying new reaction systems (multi-component mixtures at specified process conditions), the odds are frequently in favor of not finding exact values directly from open literature.

7.8.1 Density and viscosity The proper approaches are either to use extrapolation or to take values from a resembling mixture. For extrapolation and interpolation of highly nonlinear liquid viscosities, one can use the functions from Properties of Gases and Liquids [22]. For estimates of viscosity, at least at two temperatures are required. The viscosities of a resembling

304 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

mixture can also be taken as a starting point of extrapolation. Resembling organic compounds are the ones that have nearly the same chain length and contain similar functional groups. Linear extrapolation and interpolation work well for liquid densities. The most reliable way is experimental determination of unknown densities and viscosities.

7.8.2 Diffusivity In the review of [23], several correlation equations were proposed and compared for liquid-phase diffusivity. Most of the correlations are based on the original Stokes and Einstein equation, where for practical purposes the radius of the solute molecule can be replaced by the molar volume of the solute. The Wilke-Chang equation has been favorably compared to other expressions; it has become popular [24]. Diffusivity of a component is mainly dependent on the molecule size, temperature and solvent viscosity. Some common correlations are listed below (A = solute, B = solvent). Wilke-Chang Eq. (7.21): 1.2 ⋅ 10−16 ⋅ (ϕM B / (kg/kmol))0.5 (T / K) D0A,B / (m2/s) = . (μ B / Pa s)(V A / (m3/kmol))0.6

(7.21)

Scheibel equation for organic solvents Eq. (7.22): 2/3

D0A,B / (m2/s) =

8.2 ⋅ 10−16 (T / K) 3(V B / (m3/kmol)) + (1 ( ) (μ B / Pa s)(V A / (m3/kmol))1/3 (V A / (m3/kmol))

). (7.22)

Reddy and Doraiswamy equation for water as the solute in various solvents Eq. (7.23): (T / K)(M B / (kg/kmol))0.5 , (7.23) D0A,B / (m2/s) = β (μ B / Pa s)((V A / (m3/kmol)(V B / (m3/kmol))1/3 β = 10 ⋅ 10−17 , V B /V A < 1.5,

β = 8.5 ⋅ 10−17 , V B /V A ≥ 1.5.

Fedors equation for viscous liquids, Eq. (7.24): (T / K) D0A,B / (m2/s) = 4.5 ⋅ 10−15 (μ B / Pa s) ⋅

(V Bc / (m3/kmol) − V B / (m3/kmol))3/2 3 (V Bc / (m /kmol)4/3 (V Ac / (m3/kmol) − V A / (m3/kmol))1/2

. (7.24)

Many of the equations demand an estimate of molar volume of the solute and the solvent at the normal boiling point, which is not always easy to get, but it can be estimated from atomic increments. For polar components, the value of association factor is uncertain in the Wilke-Chang equation. Diffusivity in a highly viscous solution becomes low. Higher temperatures improve the diffusivities for two reasons: through the

7.8 Physical properties of gas mixtures and solutions

| 305

direct effect of temperature; and through the effect of temperature on viscosity. A critical comparison of 13 different correlations for liquid viscosities is provided by [23]. Infinite dilution diffusivities can be used for the cases in which the solutions are dilute. For high-yield systems, the diffusivity in an infinite dilution cannot directly be used, because the diffusivities in liquid-phase are basically concentration-dependent. The concentration dependency of the liquid diffusivity can be estimated from the Eq. (7.25) [25]: D A,B = (D0A,B )x B (D0B,A )x A α . (7.25) The thermodynamic activity correction (α) is not demanded if the product and reactant have almost similar chain lengths and infinite dilution diffusivities are not far away from each other. For solvent mixtures (m), the individual diffusivities of the solute can be estimated from the Eq. (7.26) [26]: 0 0.8 D Am μ0.8 . m = ∑ x i D Ai μ i

(7.26)

7.8.3 Gas solubility Gas solubilities are required for the evaluation of gas-liquid mass transfer fluxes. Although solubilities can be estimated from thermodynamic theories, a direct experimental approach is preferred. Henry’s law is the standard way of describing the solubilities of sparingly soluble gases in liquids. Temperature-dependent empirical correlations for solubilities of many gases are available [27]. In the presence of electrolytes, the salting-out effect should be included (preferably according to the theory in [28]).

7.8.4 Thermal conductivity Whenever an excessive amount of heat is generated and the reaction rates are temperature-dependent, a simultaneous solution of heat and mass balances is required. Conduction, convection and dispersion are the ways in which heat transfers. Since the conductivities of each phase may differ greatly from each other, the detailed heat conduction geometry becomes very complex and it is not included in the modeling. Instead, an effective conductivity for particles or the whole bed is used to simplify the conduction modeling. The effective conductivity within particles has often been expressed to be a function of the particle porosity and the conductivities of solid and liquid [29]. Effective conductivity of the whole bed must include the effect of gases as well. Provided that the conductivity of the catalyst is close to the conductivity of the liquid, an average thermal conductivity works well in the calculation of heat transfer within particles. For regularly shaped particles, it allows the use of simple geometry, although the particles are porous and pore structures may be very complex. The assumption is that the pores are completely filled with liquid. Generally, within particles,

306 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

most of the heat is conducted through the better conducting material. When metalbased supports are used, the conductivity of the support material may be superior to that of liquid. In some cases, however, the support is of a low-conductivity material (e.g. active carbon) and the conduction in the liquid dominates. The bed conductivity is a function of the fluid conductivity, the solid conductivity and the bed void fraction as shown below (Eq. (7.27)): λ B = λ F (λ S /λ F )(1−ε B ) .

(7.27)

The effective conductivity is between dry and wet bed conductivities. Superficial velocities of liquid and gas have been observed to have an impact on the effective thermal conductivity. However, in lab-scale packed bed reactors under the low flow interaction regime, practically no promotion of effective conductivity by fluid flows is to be expected.

7.8.5 Reaction enthalpy Heats of formation of individual components can be used as a basis when evaluating the reaction enthalpies. Another option is to use group contribution methods. Again, one potential alternative is to use heats of formations of resembling compounds.

7.9 Liquid flow effects The phenomena related to liquid flow are summarized in Fig. 7.6.

7.9.1 Qualitative flow arrangement comparison The co-current downward flow arrangement is the most typical selection for trickle bed reactors. Liquid flows down the external surfaces of the particles selecting freely flow paths of least resistance. Stagnant zones in between the particles may be generated. Compared to upward liquid flow arrangement, the advantages of a downward liquid flow are a lower extent of back-mixing and a lower pressure drop, whereas the disadvantage is a lower liquid hold-up (shorter mean residence time). In the downward operating mode, liquid takes multiple paths of the lowest resistance, thus generating a residence time distribution. The stagnant liquid zones may produce a tail in this distribution. In a shallow reactor equipped with a fine feed distributor, the distribution becomes narrow. Channeling and catalyst by-passing can be avoided in lab-scale operation. In a lab-scale reactor, complete wetting can be approached by the combined effect of the fine sand bed feed distributor and the initial flooding procedure.

7.9 Liquid flow effects |

307

Liquid flow 1. Flow direction (up/down) Upward: a) +Higher liquid hold-up. b) –More extensive backmixing. Downward: Gravity driven. 2. Chanelling & stagnant zones a) Both present when downward flow. b) Liquids selects the way of least resistance. c) Can be reduced by having small particles. 3. External wetting & wall flow a) Active sites reachable only if externally wetted. b) Wall flow and chanelling main reasons for partial external wetting. 4. Pressure drop a) Small particles, large velocities and high viscosity increse the Δp. 5. Liquid hold-up a) Total, static (not moving), dynamic (flows), external = static + dynamic (=non–capillary). b) Small particles, rough surfaces, high liquid flowrate increase hold-up. c) Where does the reactions take place? In which part of liquid? To be specified.

Fig. 7.6: Flow-related phenomena in fixed beds [9].

7.9.2 External wetting of the catalyst Correlations are available to predict the external wetting efficiency [30]. The correlations are of a semi-empirical nature and they include parameters, the values of which have been optimized using original data. The correlations show correct trends and tendencies, but should be regarded as only indicative when starting to model a new system. The wetting efficiency is strongly liquid flow dependent. When parts of the external particle surface remain dry, it means that all the active sites in the pores of the dry section become unreachable to liquid reactants. The external wetting efficiency thus directly influences the catalyst effectiveness factor. In literature, incomplete wetting is discussed for most typical regular geometries, spherical and cylindrical shapes [31, 32], respectively.

7.9.3 Radial flow Laboratory-scale trickle bed reactors operate quite close to the plug flow conditions. This means that the major concentration gradients are observed in the axial direction. Therefore, in mass balance based models the radial direction is frequently excluded. An experimental study of radial mass distribution in a larger scale reactor has been

308 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

reported. There, the reactor bottom was divided into multiple annuli-shaped sections, and the liquid flow from each section was measured. This way the wall flow and its’ fraction of the total volume flow could be revealed. Wall flow effect is especially strong when the particle-to-reactor diameter ratio is not small enough. Particle shapes and physical properties of the fluids are among the other factors determining the extent of wall flow. The wall flow effect is reduced when the liquid surface tension is low and when the liquid density is high [12].

7.9.4 Pressure drop The two-phase pressure drop can be calculated with the general pressure drop equation for a single gas-phase flow by using a two-phase flow friction factor, Eq. (7.28). A collection of pressure drop correlations is given in a trickle bed review article [10]. As a selected example, Larachi’s equation for the friction factor, Eq. (7.28), is given below: 1 17.3 f = (7.28) (31.3 + ), (X G (Re L We L )1/4 )3/2 √ X G (Re L We L )1/4 w G √ρ G , w L √ρ L ρL wL dp , Re L = μL XG =

We L =

ρ L w2L d p , σ

(7.29) (7.30) (7.31)

The friction factor depends on the Lochardt-Martinelli ratio as well as the Reynolds and Weber numbers of the liquid, Eqs. (7.28)–(7.31). The use of this equation is straightforward. The real challenge is how to estimate all the physical properties under true process conditions. Especially, the surface tensions are often missing for multicomponent systems. The equation above is tailor-made for the low interaction regime. If superficial flow velocities are low and the reactor short, the pressure drop becomes only a small fraction of the operating pressure. The pressure drop can also result in a decrease in productivity by indirectly lowering the partial pressure of the gas reactant. More precise pressure drop estimates can possibly be obtained for a regular geometry by solving the mass and momentum balances simultaneously using CFD. For various flow regimes, a collection of semi-empirical pressure drop correlations is given in [10]. In one study [33], an extensive two-phase pressure drop database was used for teaching a neural network and in that way to obtain a semi-empirical model for pressure drops. The pressure drop then became a function of both gas and liquid Reynolds numbers, the Galileo number of liquid, densities, superficial velocities and mean residence times in both phases, porosity of the bed, liquid hold-up and length of the reactor, Eqs. (7.32)–(7.37) and the single phase pressure drop of liquid, which can be

7.9 Liquid flow effects |

309

evaluated with Ergun equation Eq. (7.38). All these variables are practically attainable. The equations are listed below: ψL =

E2 Re 2L Δp/Δz εB 3 E1 Re L + 1 = ( ) (( )+( )) , ρL g εL (1 − ε B )Ga L (1 − ε B )2 Ga L

3 E2 Re 2G εB E1 Re G Δp/Δz +1=( ) (( )+( )) , ρG g εB − εL (1 − ε B )Ga G (1 − ε B )2 Ga G ρG (ψ G − 1) ; ψL = 1 + ρL ρL wL dp , Re L = μL

ψG =

Ga L = Ga G =

ρ L gd3p ε3B μ2L (1 − ε B ) ρ G gd3p ε3B μ2G (1 − ε B )

(7.32) (7.33) (7.34) (7.35)

,

(7.36)

,

(7.37)

E1 , E2 = constants of the Ergun equation. The Ergun equation (single phase pressure drop): Δp μ L ṁ L (1 − ε B ) ṁ L (1 − ε B ) + E2 . = E1 3 Δz gρ L d p gρ εB ε3B L dp

(7.38)

7.9.5 Liquid saturation (hold-up) Perhaps the most natural way to specify the volumetric reaction rates is to express them on total external liquid hold-up basis. The capillary-liquid does not flow and is therefore not included. The liquid hold-up depends on the superficial liquid and gas velocities, the physical properties of the fluids and the properties of the bed and particles. For a new application, new parameter optimization is recommended. The most popular liquid hold-up expressions are power laws of dimensionless numbers. Each dimensionless number is either an aspect ratio or ratio of characteristic forces faced by the liquid volume element. These ratios determine how much liquid is held in a fixed bed. Power laws are popular because they are practical, mathematically well-behaving functions and they are also scientifically acceptable. After all, it is true that forces generate the flows. The liquid hold-up correlations given here are taken from an extensive review article [10]. The liquid saturation (volume fraction of noncapillary liquid) equations are Eqs. (7.39)–(7.42): Kohler-Richard equation Eq. (7.39) [34]: β nc

a v d p 0.65 ρ2L gd3p = 0.71 ( ) ( ) εB μ2L

−0.42

(

ρ L w L d p 0.53 ρ G w G d p −0.31 . (7.39) ) ( ) μL μG

310 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

Ellman equation, Eq. (7.40) [35]: log (β nc ) = −0.42 X 0.24 Re 0.14 ( L G

a v d K −0.163 . ) 1 − εB

(7.40)

Wammes equations, Eqs. (7.41) and (7.42) [36], valid for low interaction regime, Re L < 11, and no gas flow: −0.39

β nc

ρ L w L d p 0.36 ρ2L gd3p = 16.3 ( ) ( ) μL μ2L

,

(7.41) −0.42

β nc = 3.8 (

ρ L w L d p 0.55 ρ2L gd3p Δp )) ) ( (1 + 2 μL ρ L gΔz μL

(

a v d p 0.65 . ) εB

(7.42)

Physical properties of fluids are more strongly temperature and less strongly pressure dependent (gas density being the exception). The liquid hold-up becomes simply a function of superficial velocities in nearly isothermally operating short reactors with a negligible pressure drop. The most strongly temperature dependent physical property is the liquid viscosity and change in it is the main reason why the liquid saturation depends on temperature. High hold-ups (with same bed volume fraction) are obtained with viscous liquids using small particles having rough surfaces. In the first liquid saturation expression, Eq. (7.39), the terms on the right hand side are the catalyst property term, one kind of Galileo number and Reynolds numbers of liquid and gas. This equation has five parameters. At a constant temperature and pressure, this equation becomes a function of superficial velocities of liquid and gas only and then the total number of parameters is reduced to three. Since new reaction systems and operation conditions differ from the original ones, the original values are only indicative. The liquid hold-up has been studied at very low liquid and gas velocities [37]. For their system, the liquid saturation became a function of mainly the Reynolds number of liquid and the particle-to-reactor diameter ratio. The total hold-up turned out to be almost constant, 0.4, and the static hold-up 0.1. With CFD, there are two ways for modeling the liquid holdup. The first one is to model it with the porous media concept (without giving the detailed geometry at all), and the other is a free surface model based on giving the detailed geometry. The porous media concept has been reported to be successfully utilized for evaluating the total liquid hold-up and the pressure drop in the bed [38]. CFD with a simplified geometry as an input and a free surface model is the most sophisticated and heaviest computational way to treat flow phenomena and could in principle be used for calculating the liquid hold-up (Lopes et al. 2007) for a bed packed with equal-sized spheres. This kind of modeling suits best for cases where particles have regular shape and flow is laminar. When catalyst particles are sieve fractions of crushed particles, no simple geometry exists. CFD, while being highly successfully applied in single-phase applications, is still struggling in multi-phase applications. The tolerances easily obtainable for a single-phase flow are hard to reach in multi-phase applications.

7.10 Reactor modeling steps

| 311

7.10 Reactor modeling steps 7.10.1 Overview An overview of the model development stage is given in Fig. 7.7. The model development progresses gradually from simpler cases to more advanced ones. The verification task aims at checking that the individual parts of the program code work reliably.

Modelling steps 1. Verification with simpler cases (Plug-flow unit, ideally stirred tank, only single reaction, few reactions, only mass balances, only heat balances, only heat conduction, particles of ideal shape, etc.). 2. Modelling of individual phenomenon separately a) Each phenomenon = research field of its own. b) Everything can not be included in full detail. c) One has to make decisions (what to include, in what degree of details). 3. Modelling of a simplified reaction system Often demands experimental data from simpler systems where some of the reactions are supressed. 4. Sensitivity studies prior to estimation (single variable) 5. Minimisation of number of parameters Narrowing of search ranges, starting with a model having fewer parameters and gradually adding more. 6. Improving identification (Modest identification study) 7. Parameter estimation Fig. 7.7: Modeling steps of laboratory-scale fixed bed reactors [9].

7.10.2 Selected modeling policy When developing three-phase fixed bed reactor models, lots of phenomena are simultaneously present, and one has to make simplifications and decisions on how detailed one should be. In our modeling, decisions were made not to deeply progress into details of particle technology and fluid dynamics.

7.10.3 Studies of simplified reaction systems Our modeling study of the direct synthesis of hydrogen peroxide shows how sometimes it is possible to study the intrinsic kinetics of a simplified reaction system separately. Direct decomposition and total decomposition reactions could be studied separately from other reactions. This allowed one to determine the hydrogen peroxide

312 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

decomposition parameters independently and to divide the extensive parameter estimation to smaller sub-tasks.

7.10.4 Ways how to rule out phenomena by experimental design By proper selections of process conditions and reactor set-up, most heat and mass transfer related phenomena can be ruled out in preliminary studies. For example, by deciding to work with dilute concentrations and a short reactor, the effects of adsorption terms in the reaction equation can be suppressed. As a bonus, the effects of heat and mass transfer become suppressed as well. The effect of internal diffusion can be suppressed by using small catalyst particles. The intrinsic kinetics are revealed by studying them in a slurry reactor equipped with intensive mixing. The effects of impurities are avoided by working with pure solutions. Criteria of how to avoid the generation of serious temperature gradients have been presented in literature [8]. The severity of temperature effects directly depends on catalyst loading. The temperature gradients can be reduced by diluting the catalyst bed with inert particles. Internal heat transfer effects can be reduced in trickle bed operation by selecting to work with smaller particles, although at the expense of an increased pressure drop. By having support materials of high thermal conductivity, the bed works more isothermally. A common practice, especially in preliminary laboratory tests and catalyst screening studies, is to work with dilute streams – also to suppress temperature gradients.

7.10.5 Sensitivity studies Sensitivity studies are performed for two reasons: first, in order to test and make sure that the program can operate reliably in broad ranges of variables; and, second, to get a feeling of how the system behaves and reacts to individual parameter changes.

7.11 Balances for the generic three-phase fixed bed model 7.11.1 Mass balances for gas, liquid and solid phases The general model equations presented here are suitable for calculations of the timedependent concentrations of all compounds and temperature at each location inside a trickle bed reactor. In particular, three different locations are taken into account: axial and radial position inside the reactor and position within each particle in every location of the packing. The compressed mass balances of the components in gas and liquid, here presented for fluid, are given in Eq. (7.43). This mass balance states that the accumulation is a net effect of gas-liquid mass transfer, convection, axial and ra-

7.11 Balances for the generic three-phase fixed bed model | 313

dial dispersion and finally the mass flux to/from the particles (the last term is usually present only for liquid balances). This last term is necessary because the reactions are assumed to take place here within the solid catalyst particles. εF

∂C i,F ∂2 C i,F ∂(w F C i,F ) = ± JGL − + ε F D z,F ∂t ∂z ∂z2 󵄨 ∂2 C i,F 1 ∂C i,F D e,i s ∂C i,S 󵄨󵄨󵄨 󵄨󵄨 + ε F D r,F ( + − ) r ∂r R p ∂r p 󵄨󵄨󵄨r ∂r2

(7.43) p =R p

The mass balance for the solid phase, Eq. (7.44) states that the accumulation of each component depends on the internal diffusion and reaction terms. The equations demand the effective diffusivities and rate equations to be known for each component. The adjustable surface shape factor, together with well-defined characteristic particle dimension, are required for taking the particle geometry into account. The shape factor, s, for an arbitrary geometry can be estimated using Eq. (7.45). ∂C i,s D e,i ∂2 C i,s εF s ∂C i,s + = ( ) + ρcat ⋅ ∑ (ν i,j r i,j ) ∂t εp r p ∂r p εp ∂r2p Ap s + 1 = (R p ⋅ ) Vp

(7.44) (7.45)

7.11.2 Energy balances for gas-, liquid- and solid phases The energy balances for fluids (gas and liquid), Eq. (7.46), state that the heat accumulation is the consequence of convective heat flux, heat conduction to axial and radial directions, heat fluxes caused by axial and radial dispersion, and the heat flux from the particles (last term for liquid only). Among the key parameters for a correct description of temperature fields inside the reactor are the effective conductivities to the radial and axial direction of the reactor tube. Since three phases are involved in heat conduction, it is always a challenge to find well representative values for trickle beds. λ z,F ∂T F ∂(w F T F ) ±HGL ∂2 T F − + D z,F ) ⋅ = +( ∂t ε F ρ F c p,F ∂z ε F ρ F c p,F ∂z2 +(

󵄨 λp s λ r,F ∂2 T F 1 ∂T F ∂T S 󵄨󵄨󵄨 󵄨󵄨 + D r,F ) ( 2 + . )− ε F ρ F c p,F r ∂r R p ε F ρ F c p,F ∂r p 󵄨󵄨󵄨r =R ∂r p p

(7.46)

The energy balance of the solid phase, Eq. (7.47), states that heat accumulation is the net result of heat conduction to characteristic direction and heat generation by reactions. Here the challenge is to estimate the effective particle conductivity, to select representative values for characteristic particle radius and surface shape factor and to have an accurate description of the reaction kinetics. λp ∂T s ∂2 T s s ∂T s 1 εF ⋅ = ( 2 + )+ (∑ r j (−ΔH r )) . ∂t ρcat c p,s ∂r p r p ∂r p c p,s ε p

(7.47)

314 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

7.11.3 Boundary conditions Since the system consisted of a collection of PDEs, it became crucially important to define proper initial and boundary conditions. Boundary conditions are listed in full details in Eqs. (7.48)–(7.67). 󵄨 C i,F 󵄨󵄨󵄨z=0 = C IN i,F ,

(7.48)

󵄨 T F 󵄨󵄨󵄨z=0 = T FIN ,

(7.49)

∂C i,F 󵄨󵄨󵄨 󵄨󵄨 ∂z 󵄨󵄨z=L ∂T F 󵄨󵄨󵄨 󵄨󵄨 ∂z 󵄨󵄨z=L ∂C i,F 󵄨󵄨󵄨 󵄨󵄨 ∂r 󵄨󵄨r=0 ∂T F 󵄨󵄨󵄨 󵄨󵄨 ∂r 󵄨󵄨r=0 ∂C i,F 󵄨󵄨󵄨 󵄨󵄨 ∂r 󵄨󵄨r=R ∂T F 󵄨󵄨󵄨 󵄨󵄨 −k r ⋅ ∂r 󵄨󵄨r=R 󵄨 C i,s 󵄨󵄨󵄨z=0

= 0,

(7.50)

= 0,

(7.51)

= 0,

(7.52)

= 0,

(7.53)

= 0,

(7.54)

󵄨 = h w ⋅ (T F 󵄨󵄨󵄨r=R , −T w )

(7.55)

= C IN i,s ,

(7.56)

󵄨 T s 󵄨󵄨󵄨z=0 = T sIN ,

(7.57)

∂C i,s 󵄨󵄨󵄨 󵄨󵄨 ∂z 󵄨󵄨z=L ∂T s 󵄨󵄨󵄨 󵄨󵄨 ∂z 󵄨󵄨z=L ∂C i,s 󵄨󵄨󵄨 󵄨󵄨 ∂r 󵄨󵄨r=0 ∂T s 󵄨󵄨󵄨 󵄨󵄨 ∂r 󵄨󵄨r=0 ∂C i,s 󵄨󵄨󵄨 󵄨󵄨 ∂r 󵄨󵄨r=R ∂T s 󵄨󵄨󵄨 󵄨󵄨 −k r ⋅ ∂r 󵄨󵄨r=R ∂C i,s 󵄨󵄨󵄨 󵄨󵄨 ∂r p 󵄨󵄨r p =0 ∂T s 󵄨󵄨󵄨 󵄨󵄨 ∂r p 󵄨󵄨r p =0 󵄨 C i,s 󵄨󵄨󵄨r p =R p 󵄨 T s 󵄨󵄨󵄨r p =R p

= 0,

(7.58)

= 0,

(7.59)

= 0,

(7.60)

= 0,

(7.61)

= 0,

(7.62)

󵄨 = h w ⋅ (T s 󵄨󵄨󵄨r=R − T w ) ,

(7.63)

= 0,

(7.64)

= 0,

(7.65)

= C i,F ,

(7.66)

= TF .

(7.67)

7.12 Axial dispersion modeling and experiments | 315

Boundary conditions are the feed concentrations and temperature at the reactor inlet, axial derivatives of concentrations and temperature at zero at the reactor outlet, radial derivatives of concentration and temperature as zero at reactor center due to symmetry and at the reactor wall concentration derivatives as zero and only heat flux to jacket. Concerning the particles, the boundary conditions are liquid concentrations and liquid temperature at the surface, zero derivative of both concentration and temperature in the center of the particles due to symmetry. At the reactor outlet, they are called Danckwerts closed boundary conditions. Reactor inlet boundary conditions are plug flow conditions due to downward gravity driven feed flows.

7.11.4 Sub-model examples The model equations can be considered to be general for trickle bed reactors, because by switching on/off several parameters or terms, it is possible to describe a wide range of sub-models. Here some cases are depicted. 1. Reduction of the number of phases: by simply switching off the mass transfer terms and properly adjusting hold-up, it is possible to reduce the number of the phases to produce two-phase models in gas-liquid or liquid-solid systems. 2. Isothermal/adiabatic system: the wall heat transfer parameter can be adjusted in order to regulate the right heat transfer of the fixed bed. Of course, the two extremes can be either an isothermal or an adiabatic reactor, simply by choosing an infinite or zero value for this parameter and having a negligible heat generation rate. 3. Axial dispersion model: the radial dispersion can be switched off by neglecting the relative terms. 4. Laminar flow model: by adjusting the velocity field and switching of the dispersion terms, a laminar flow can be easily described. It is clear, that the total number of sub-models is large because a big amount of terms is considered in the mass and heat balances, so the model equations can be considered general and a starting point for the description of experiments performed in fixed bed.

7.12 Axial dispersion modeling and experiments 7.12.1 Classical axial dispersion model The axial dispersion coefficient for the fixed bed reactor can be obtained from experiments carried out with inert tracers, which implies that the diffusion term originating from reaction rates in Eq. (7.68) becomes zero. Furthermore, if radial dispersion is negligible and the tracer remains in liquid phase, the equation for a liquid-phase tracer is

316 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors ∂C i,L ∂2 C i,L w L ∂C i,L − = D a,L . ∂t ε L ∂z ∂z2

(7.68)

The step or impulse changes for tracer are used in all the axial dispersion experiments. The most natural selection for the tracer is to use a component that is originally present in the feed or product stream and which do not significantly change the physical properties of the solution. The dimensionless Peclet number is obtained from the axial dispersion coefficient, Eq. (7.69). Pe = w L L/D a,L . (7.69)

7.12.2 Alternative modeling approaches for back-mixing Some alternative modeling approaches have been proposed, for example the piston exchange model and modeling with neural networks. The piston exchange model is based on the solution of twice the mass balances, one for the dynamic part of liquid and the other for the static part of liquid [39]. The two parameters that the piston exchange model has are dynamic liquid fraction and united mass exchange coefficient between the dynamic and static part of the liquid. Combined dimensional analysis and neural network modeling was used for an extensive amount of data from the open literature for evaluating the Pe number with a semi-emphirical expression [40]. This model was claimed to be capable of predicting the axial dispersion coefficient better than any other tested model. The neural network model used the Reynolds number of liquid, the Eötvos number of liquid, the Galileo numbers of liquid and gas, wall factor and mixed Reynolds number as inputs. A collection of the equations from open literature for evaluating the axial dispersion coefficient for specified systems under specified operation conditions was also presented. The two key variables that were present in all these models were the superficial velocity of liquid and the diameter of particles.

7.13 Numerical strategies 7.13.1 Model classification Most of the models consist of parabolic partial differential equations (PDEs) that are solved numerically. Some of the models are ordinary differential equations (ODEs), either initial value problems (IVP) or boundary value problems (BVP). The mathematical structures of the models are summarized in Tab. 7.1. For each kind of model, an appropriate numerical solution strategy should be selected. Some common and robust numerical methods are reviewed in this section.

7.13 Numerical strategies

| 317

Tab. 7.1: Models and their mathematical structures. Fixed bed model

Mathematical structure

Concurrent, pseudo-homogeneous, steady state/dynamic Countercurrent, pseudo-homogeneous, steady state/dynamic Concurrent, heterogeneous, steady state/dynamic Countercurrent, heterogeneous, steady state/dynamic

ODE (IVP or BVP) / PDE (IVP) ODE (BVP) / PDE (IVP) ODE (IVP + BVP or BVP) / PDE (IVP) ODE (BVP) / PDE (IVP)

Tab. 7.1 was prepared for cases in which the model only takes one spatial co-ordinate into account. As the table reveals, most of the mathematical structures are partial differential equations, because the concentrations and temperatures are time and spacedependent. Ordinary differential equations are obtained for steady state models. They are boundary value problems in most cases because of axial and/or radial dispersion effects. The models of catalyst particles are boundary value problems if steady state conditions prevail; but under transient conditions, the catalyst particle models are initial value problems.

7.13.2 Benefits of dynamic models Dynamic models are preferable for several reasons: dynamic, time-dependent models can describe the reactor dynamics and catalyst activity changes, and they can be solved in a more robust manner than steady state models. The boundary value problems appearing in the steady state models are often tricky to solve, because some kind of iterative approach is needed (shooting methods), or approximation functions must be used such as orthogonal polynomials, which leads to a set of non-linear equations.

7.13.3 Solvers and solution algorithms The numerical mathematics in the solution of ordinary differential equations (ODEIVP) are advanced, and very robust and efficient numerical algorithms, such as backward difference methods and semi-implicit Runge-Kutta methods, have been developed for stiff ODEs, which appear in chemical systems [4]. The stiffness of the ODE system is the key issue having two principal origins: the rates of the chemical reactions can vary very much because a system can comprise both very rapid and very slow reactions, and the stiffness can increase due to the discretization of the spatial coordinates in the porous pellet or in the reactor system. Because many efficient algorithms have been developed for stiff ODEs, it is very attractive to transform the PDEs appearing in three-phase reactor models to ODEs by applying a discretization procedure on the spatial coordinate. In this way, a system of parabolic PDEs is transformed to a large set of ODEs (IVP), which can be solved with

318 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

stiff ODE algorithms. The discretization can be done in several ways, the most common ones are discretization with finite differences (method of lines) and discretization with approximation functions, such as orthogonal collocation. Orthogonal collocation or finite element methods are in principle more accurate than finite differences, but the implementation of finite differences is easier. This implies that the method of lines with finite differences is very often used even in spite of its disadvantages.

7.13.4 Numerical method of lines Both the dynamic mass and energy balances can be solved by applying the numerical method of lines [41]. In the numerical method of lines, the spatial derivatives (i.e. longitudinal and radial concentration and temperature gradients) are replaced by their finite difference approximations. The simplest way is to use two-point approximations for the first derivatives and three-point difference for the second derivatives. By doing so, each partial differential equation is reduced into a set of ordinary differential equations. The problem can thereafter be expressed as a set of ordinary differential equations for which reliable solvers are readily available. It is possible to choose different discretization methods, such as the central finite difference method, the backward finite difference method, the forward finite difference method, the orthogonal collocation or finite elements method. The choice of the right method is not trivial and sometimes can even lead to physically unrealistic solutions. For example, backward differences should be used for the plug flow terms and central differences for the diffusion and dispersion terms to obtain a correct numerical solution. The number of discretization steps for each direction defines the grid where the calculations are performed. A large number of elements lead to a more accurate solution, consequently increasing the execution time of the code, a fact that is extremely important in parameter estimations problems in which the reactor model is invoked even thousands of times. A compromise between execution time and accuracy must be chosen by defining the right number of discretization elements. Approximation order and polynomial degree parameters have a great influence on the accuracy of the solution overall in coarse grids in which only a small number of elements are used. Stiff ODE solvers [42], always adjust the duration of each time step freely; when it faces numerical challenges, it simply starts to use smaller and smaller time steps. Then a computation, which initially starts well, has a tendency to slow down.

7.13.5 Hoyos method for particles A numerical calculation method for evaluating the steady state diffusion and heat conduction within a cylindrical or a hollow cylindrical particle to both axial and radial directions was given originally in [43]. This Hoyos method can be extended to become

7.13 Numerical strategies

| 319

valid for other geometries by adjusting the surface shape factor and diffusion path length properly. The method can be extended to cover particles of different shapes and used even for a rough estimate evaluation of a much more complicated geometry by simply using the adjustable surface shape factor and diffusion path length to compensate for the geometry. From the viewpoint of numerical efficiency, orthogonal collocation on finite elements is a very superior method in obtaining concentration and temperature profiles in porous catalyst particles and layers.

7.13.6 Parameter optimization methods Estimation of kinetic and transfer parameters from experimental data implies always a use of an optimization algorithm, because the optimal values of the parameters are searched. Mostly the Levenberg-Marquardt method [44] is frequently used in parameter estimation tasks. Always, prior to the estimation tasks, multiple simulation studies over the range of interest for each parameter should be conducted. This is done to reveal the concentration responses and their sensitivity a priori and also in order to make sure that the possible convergence troubles can be faced and solved in the simulation stage to avoid surprises from arising in the more time-consuming parameter estimation tasks. Convergence problems can be surmounted by finding the combinations of variables in the simulation stage that caused the trouble and reasons for that. With this kind of strategy, the estimations can be carried out without significant disturbances using still rather broad ranges of parameter values, selected based on experience gained from the multiple simulation stage. The target function for the optimization is usually the sum of the least square deviation between the experimental and calculated concentrations. Separate weighting factors can be given to each and every observation point. Modest software was used in the examples (Citral hydrogenation, direct synthesis of hydrogen peroxide) presented in this review [45]. It has built-in ordinary differential equation solvers, simulation, sensitivity analysis and optimization routines.

7.13.7 Parameter number reduction The number of parameters in each optimization task should be minimized without decreasing the accuracy. For the phenomena that can be studied separately, this should be done. Tracer tests for revealing the magnitude of axial dispersion coefficient serve as an excellent example. In many cases, a single phenomenon can become different when it is studied alone. Then it is not absolutely correct first to study the effect separately and then to use the obtained parameters in context that is different. An often-discussed example of this is the issue whether or not the mass transfer should be studied alone in a fixed bed operating at trickling flow regime [21]. Extrapolation

320 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

of flow-related phenomena of a fixed bed in the trickling flow regime to different gas and liquid flow ranges is risky [47].

7.14 Scale-up issues of fixed beds 7.14.1 Overview Fig. 7.8 summarizes the major issues that appear in the scale-up of continuous fixed bed reactors [2–7]. Scale-up

Lab scale

Industrial scale

1. Reactor

Small diameter, short

Large diameter, long

2. Catalyst particles

Small, irregular shapes, size distribution

Large, regular tailor made shapes, single size

3. Catalyst dilution

Provides residence time, Higher loading preferred Mild conditions

4. Concentration range

Narrow, dilute

Broad, concentrated

5. Pressure drop

Negligible

Operation cost, aimed low

6. Feed distribution

Fine

Challenging

7. Chanelling and stagnant zones Minor challenge

Hard to avoid

8. Radial T and C gradients

Mild

More severe

9. Internal diffusion

Controlled

More severe

10. Internal heat conduction

Minor challenge

May become a problem, hot spots

11. Liquid velocities

Low

Higher

12. Operation conditions

Global

Local

13. Axial dispersion

Present

Different

Fig. 7.8: Scale-up issues [9].

7.14.2 Large scale operation In large-scale operation, it is not as easy to avoid the generation of temperature and concentration gradients. One of the main reasons for this is that chemical industry prefers to work with concentrated streams to intensify the process. High pressure drop is not acceptable because it directly affects the operating costs. Therefore, large catalyst particles have to be used. Industry also pushes towards more compact reactor sizes. Then catalyst dilution is not desirable. Due to the large reactor diameter, even distribution of liquid feed (and also redistribution) becomes more challenging. Superficial velocities of gas and liquid are typically set at higher levels in industrial units. As a consequence of all these factors, industrial units typically work in many ways under

7.14 Scale-up issues of fixed beds | 321

more severe conditions, which makes the undesirable flow, heat and mass transfer phenomena to emerge.

7.14.3 Gradients in scale up When temperature gradients arise in a trickle bed reactor, they normally appear first in a radial direction. The heat transfer area per unit volume becomes smaller, the larger the diameter of the reactor is required. Then one of the remaining options to decrease severe temperature gradients and to improve the radial heat recovery is to select a multi-tubular reactor. Axial temperature gradients become more severe if the radial heat cannot be effectively recovered. If the liquid flow is channeled and stagnant places do exist, uniform temperature is hard to get and hot spots can emerge.

7.14.4 Flow regime in scale-up In the reactor scale-up, it is extremely important that also in the bigger scale, the reactor operates at the same flow regime. Otherwise, unexpected behavior may arise, since many of the flow-related phenomena are strongly flow-regime dependent. A lot of studies have been conducted to evaluate the borderlines between flow regimes. The type of flow regime depends on the magnitudes and the ratio of superficial gas and liquid velocities, the physical properties of the fluids involved and the properties of the catalyst bed. Flow maps do exist, but they are case specific and, therefore, mostly indicative. Both experimental correlations and semi-empirical expressions to figure out the flow regime can be found.

7.14.5 Back mixing in scale-up The extent of axial dispersion in large-scale operations is hard to predict based on small-scale experiments alone. The real challenge in the scale-up using an axial dispersion model is faced when trying to figure out how to get an estimate of the axial dispersion coefficient for a large unit without large-scale experiments. The axial dispersion coefficient can hardly be kept constant in scale-up, since this single parameter takes into account all kinds of flow non-idealities, and the extent of these are strongly scale dependent.

322 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

7.15 Examples 7.15.1 Citral hydrogenation The use of renewable raw materials instead of petroleum-based ones is one of the main research fields of green chemical processes for the future and is, therefore, studied intensively worldwide. An example from this field is the production of citronellal from citral, a renewable compound from essential oils. Citronellal is a key compound that can serve as a starting material for synthesis of a wide spectrum of organic compounds, the most famous of which is perhaps menthol. An extensive review of the organic compounds that can be synthesized from citronellal does exist [48]. The reaction system is a complex one, but under the experimental conditions used by us, the system could be described with two reactions only: Citral + H2 󳨀󳨀→ Citronellal; Citronellal + H2 󳨀󳨀→ Citronellol. This implies that the system is a consecutive reaction, in which the first hydrogenation dominates; citronellal was the main product, while citronellol was a side-product. The catalyst was supported by nickel. The reactor flowsheet is displayed in Fig. 7.9. Figs. 7.10 and 7.11 present a set of results from the conducted sensitivity study. The sensitivity of the citronellal concentration on the changes of various key parameters is illustrated. The selected key parameters were: the reaction rate constant (in the absence and the presence of the deactivation), the gas-liquid mass transfer coefficient,

Filters Liq. 1 Liq. 2 Liq. 3 Liq. 4

Pump

Catalyst screening reactor system, ÅA PIC

Mixer/evaporator

Liquid distribution Gas distribution

FIC

Gas 1 FIC

Liquid separators

Capillaries

Gas 2 FIC Gas 3

TIC Gas stream selection value

TIC

(All reaction separately)

Gas sampling Filters

TIC

TIC

Fig. 7.9: Experimental flowsheet for citral hydrogenation [49].

Heated oven

GC/MS

PIC

7.15 Examples | 323

Concentration [mol/l]

0.02 5.0 4.5 4.0 3.50

0.015 3.0

0.01

2.5 2.0 1.5

0.005

0

1.0

p1/(reaction rate constant) 0

1000

2000

3000 4000 Time [s]

5000

6000

Fig. 7.10: Example results from the sensitivity study. The effect of the reaction rate constant in the presence of catalyst deactivation [46]. Reprinted with permission from { Ind. Eng. Chem. Res., 2012, 51 (26), pp. 8858–8866}. Copyright {2012} © American Chemical Society.

the deactivation rate constant and the Peclet number. All the details of this study can be found from the corresponding article [46]. The final results of the parameter estimation study are given in Fig. 7.12, which shows that the model very adequately describes the time-on-stream behavior of the citral hydrogenation system. The dominating effects for citral hydrogenation were the reaction kinetics along with the raw material dependent deactivation. The reaction system was very dilute, thus the mass transfer effects were suppressed.

7.15.2 Direct synthesis of hydrogen peroxide Hydrogen peroxide is a chemical component used in many industrial applications. It is used as a strong oxidant by the chemical industry and as a bleaching agent by the pulp and paper industry. With the conventional production technology, the anthraquinone process, hydrogen peroxide becomes relatively expensive. If hydrogen peroxide could be produced with a simpler process, it could be available with a more competitive price and could then be used more extensively in current and totally new applications. Reviews that highlight the scientific community’s pursuits to produce hydrogen peroxide in new ways exist [50, 51]. Direct synthesis from hydrogen and oxygen using a proper solvent and well-selected catalyst is one of the attractive alternatives. Direct synthesis in the presence of a Pd catalyst also served as our modeled example system. If it could be managed to be done safely with a catalyst exhibiting both high productivity and selectivity, the production process of this environmentally friendly component could be essentially simplified. Due to the explosive nature of the H2 and O2 gas mixture, the direct synthesis experiments are usually carried out using a sol-

324 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors 0.02 Concentration [mol/l]

10 1

0.015

Pe

0.01

0.005

0

0

1000

2000

3000 4000 Time [s]

5000

6000

5000

6000

0.02 Concentration [mol/l]

kgla/[1/s] 0.10

0.015 0.02

0.01

0.03

0.01

0.005 0.001

0

0

1000

2000

3000 4000 Time [s]

Fig. 7.11: Example results from the sensitivity study. Effects of Peclet number and the gas-liquid mass-transfer coefficient on product concentration [46]. Reprinted with permission from { Ind. Eng. Chem. Res., 2012, 51 (26), pp. 8858–8866}. Copyright {2012} © American Chemical Society.

vent. Methanol was used in our experiments. It was originally selected since, in the comparison of five potential solvents [51–53], methanol is mentioned as giving the highest rate of reaction. Our direct synthesis experiments were carried out outside the explosive regime. This implied that hydrogen content was low and, consequently, the gas-liquid mass transfer became also one issue of importance to be checked. The reaction system was a demanding one, since together with the desirable main reaction, also oxidation, direct decomposition and hydrogenation reactions took place, all leading to water formation. The system was experimentally studied under various superficial gas and liquid velocities and optimum velocity combinations had been observed both for the productivity and the selectivity viewpoints. In the experimental study, a very short reactor was used and the productivity in all the experiments was consequently low. Direct decomposition and total decomposition reactions were stud-

7.15 Examples | 325

0.016 0.014

Concentration [mol/l]

0.012 0.01 0.008 0.006 0.004

40°C 45°C 50°C 55°C 60°C

0.002 0

0

2000

4000

6000

Time [s] Fig. 7.12: Example results from parameter estimation: citronellal concentration as a function of time-on-stream [46]. Reprinted with permission from { Ind. Eng. Chem. Res., 2012, 51 (26), pp. 8858–8866}. Copyright {2012} © American Chemical Society.

ied separately and modeled individually to improve the reliability of the model. The parallel-consecutive reaction system is displayed below in Fig. 7.13. The liquid hold-up parameters were optimized together with the kinetic parameters of the direct decomposition reaction by using direct decomposition data. Gasliquid mass transfer parameters were estimated together with H2 O2 hydrogenation parameters using total decomposition data. Finally, synthesis and oxidation reaction parameters were estimated from the direct synthesis experiments when all the four reactions were taking place simultaneously. The modeling results of direct decomposition and total decomposition studies after the parameter estimation were in line with the experimental observations. The accuracy of the modeling results for the whole reaction system was improved by introducing an adaptive element to the model, namely a gas and liquid flow rate-dependent correction to all reactions. The flow range in the H2 +O2

H2O2

+0.5O2

H2 O

H2

2H2O H2O+0.5O2 Fig. 7.13: Reaction in direct synthesis of hydrogen peroxide.

326 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors 5

1

18

19 20

16

7 6

F

2

11 13

7 F

9

6 3

12

9

14 F

15

7 7 F

6 4

9

17

10 Liquid sample

Vent gas

7 Liquid exhaust

Fig. 7.14: Fixed bed reactor set-up for direct synthesis of hydrogen peroxide [47]. Reprinted with permission from { Ind. Eng. Chem. Res., 2012, 51 (41), pp. 13366–13378}. Copyright {2012} © American Chemical Society.

experimental studies of the H2 O2 decomposition was narrower than in the full reaction set studies. This might be one reason behind the observed deviations. The reactor set-up is shown in Fig. 7.14. The reactor was equipped with large inlet and outlet connections consisting of the pipelines and fine sand bed sections before and after the reaction section. Therefore, the whole reactor set-up approached plug flow behavior, Fig. 7.15, although inside the reaction section, axial dispersion effects were clearly present. Details of the experimental axial dispersion studies and their numerical modeling can be found in article [47]. The direct decomposition reaction was studied separately as a function of liquid and gas velocities. The parameter estimation results, Fig. 7.16, show both the experimental data (points) and model predictions for various combinations of gas and liquid flow. The parameter estimation for the case when all the four reactions were progressing was finally made in Fig. 7.17. The results illustrate how closely the final model could follow the experimentally observed behavior.

7.16 Conclusions | 327

Dimensionless tracer concentration

1 0.9 0.8 0.7 0.6

Pe = 46 VI = 1.0 ml/min VG = 2.7 ml/min

0.5 0.4 0.3 0.2 0.1 0

0

1500 1000 Time [s]

500

2000

2500

Fig. 7.15: Tracer experiments and modeling: Pe number evaluation for the whole reactor [47]. Reprinted with permission from { Ind. Eng. Chem. Res., 2012, 51 (41), pp. 13366–13378}. Copyright {2012} © American Chemical Society.

0.126 H2O2 concentration [mol/l]

0.124 0.122 0.120

VI=1 l/min

VI=2 l/min

VI=0.5 l/min

0.118 0.116 Vg=0.5,1,2.9, 4, and 6 ml/min

0.114 0.112 0.110 0.108

kdir= 0.0074 aepsI= 0.25 bepsI= –0.469

Vg=0.5, 1, 2.9, 4 and 6 ml/min 0

1

2 3 4 Cumulative time [s]

5

6 x 104

Fig. 7.16: Parameter estimation results for the hydrogen peroxide decomposition study [47]. Reprinted with permission from { Ind. Eng. Chem. Res., 2012, 51 (41), pp. 13366–13378}. Copyright {2012} © American Chemical Society.

7.16 Conclusions Design and modeling of laboratory-scale three-phase fixed bed reactors is a demanding task, comprising all the essential elements of chemical reaction engineering: thermodynamics, kinetics, fluid and catalyst properties, gas solubilities, mass and heat transfer effects and multiphase flow patterns. Recent advances in reactor technology, particularly the development of small and very precise experimental devices, with

328 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors 0.008

Exp. VI=0.5 ml/min Exp. VI=0.75 ml/min Exp. VI=1 ml/min Model VI=0.5 ml/min Model VI=0.75 ml/min Model VI=1 ml/min

0.007

Concentration [mol/l]

0.006 0.005 0.004 0.003 0.002 0.001 0

5

10 Gas flow rate [ml/min]

15

Fig. 7.17: Parameter estimation results for the whole fixed bed reactor [47]. Reprinted with permission from { Ind. Eng. Chem. Res., 2012, 51 (41), pp. 13366–13378}. Copyright {2012} © American Chemical Society.

which several catalysts can be evaluated under varying conditions, implies a breakthrough in the efficiency of research efforts. Experimental data can be produced faster than ever. Experimental design becomes more important in order to keep the amount of experimental data within reasonable limits and to extract the best possible information. In order to progress in a rational way, these data should be systematically analyzed by mathematical modeling. In our opinion, the experimental data should be modeled based on physico-chemical theories, starting from the active site of the solid catalyst and ending up at the description of the flow behavior of the fixed bed reactor.

Acknowledgment The work is a part of the activities of Åbo Akademi Process Chemistry Centre (PCC), a center of excellence for scientific research. Financial support from the Academy of Finland is gratefully acknowledged.

References |

329

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

[12] [13]

[14] [15] [16]

[17] [18] [19] [20] [21] [22] [23] [24] [25]

Krishna R, Sie S. Strategies for multiphase reactor selection. Chem Eng Sci, 1994, 49, 4029– 4065. Froment G, Bishoff K. Chemical reactor analysis and design. New York, USA, John Wiley & Sons, 1979. Levenspiel O. Chemical reaction engineering. New York, USA, John Wiley & Sons, 1999. Salmi T, Mikkola JP, Wärnå J. Chemical reaction engineering and reactor technology. New York, USA, CRC Press, 2010. Hagen J. Industrial catalysis, A practical approach, 2nd Edition, Weinheim, Germany, Wiley-VCH Verlag GmpH, 1999. Belfiore L. Transport phenomena for chemical reactor design: New Jersey, USA, John Wiley& Sons, 2003. Butt J. Reaction kinetics and reactor design: New York, USA, Marcel Dekker Inc., 2000. Mederos F, Ancheyta J, Chen J. Review on criteria to ensure ideal behaviours in trickle-bed reactors. Applied catalysis A: General, 2009, 355, 1–19. Kilpiö T. Mathematic modeling of laboratory scale three-phase fixed bed reactors. Thesis, Åbo, Finland, Åbo Akademi University, Department of Chemical Engineering, 2013. Al-Dahhan M, Larachi F, Dudukovic M. Laurent A. High-pressure trickle bed reactors: A review. Ind Eng Chem Res, 1997, 36, 3292. Leveneur S, Wärnå J, Salmi J, Murzin YM, Estel L. Interaction of intrinsic kinetics and internal mass transfer in porous ion-exchange catalysts: Green synthesis of percarboxylic acids. Chem Eng Sci, 2009, 64, 4101–4114. Møller L, Halken C, Hansen J, Bartholdy J. Liquid and gas distribution in trickle-bed reactors. Ind Eng Chem Res, 1996, 35, 926–930. Tukac V, Simickova M. Chyba V, Lederer J, Kolena J, Hanika J, Jiricny V, Stanek V, Stavarek P. The behaviour of pilot trickle-bed reactor under periodic operation. Chem Eng Sci, 2007, 62, 4891–4895. Murzin D, Salmi T. Catalytic kinetics, ISBN: 0-444-51605-0, Amsterdam, Netherlands , Elsevier, 2005. Mikkola JP, Salmi, T, Sjöholm, R. Modelling of kinetics and mass transfer in the hydrogenation of xylose over Raney nickel catalyst. J Chem Technol Biotechnol, 1999, 74, 655–662. Salmi T, Murzin DY, Mikkola JP, Wärnå J, Mäki-Arvela P, Toukoniitty E, Toppinen S. Advanced kinetic concepts and experimental methods for catalytic three-phase processes. Ind Eng Chem Res 2004, 43, 4540–4550. Cabrera MI, Grau, RJ. J Mol Catal A Chem, 287, 2008, 24. Cabrera MI, Grau RJ. Trends in Chemical Engineering, 39–53, 2008. Metaxas K, Papayannakos N. Gas-liquid mass transfer in a bench-scale trickle bed reactor used for benzene hydrogenation. Chemical Engineering & Technology, 2008, 31, 410–417. Metaxas K, Papayannakos N. Kinetics and mass transfer of benzene hydrogenation in a tricklebed reactor. Ind Eng Chem Res, 2006, 45, 7110–7111. Koning B. Heat and mass transport in tubular fixed bed reactors at reacting and non-reacting conditions. Thesis, Twente, Holland, University of Twente, 2002. Poling, B, Prausnitz, J, O’Connell J. Properties of gases and liquids. 5:th Ed. McGraw-Hill 2001. Wild G, Charpentier, JC. Diffusivité des gaz dans les liquids, Paris, France, Techniques d’ingenieur, 1987. Fei W, Bart HJ. Prediction of diffusivities in liquids. Chem Eng Technol, 1998, 21, 659–665. Vignes A. Diffusion in binary solutions. Ind Eng Fundam, 1966, 5, 189–196.

330 | 7 Design and modeling of laboratory scale three-phase fixed bed reactors

[26] Perkins LR, Geankoplis, CJ. Molecular diffusion in a ternary liquid with the diffusing component dilute. Chem Eng Sci, 1969, 24, 1035–1042. [27] Fogg P, Gerrard W. Solubility of gases in liquids, Chichester, England, Wiley, 1992. [28] Estimation of gas solubilities in salt solutions at temperatures from 273 to 363, AIChE J, 1996, 42, 298–300. [29] Woodside W, Messmer J. Thermal conductivity of porous media. I. unconsolidated sands. J Appl Phys 1961, 32, 1688. [30] Lappalainen K, Alopeus V, Manninen M, Aittamaa J. Improved hydrodynamic model for wetting efficiency, pressure drop and liquid holdup in trickle bed reactors. Ind Eng Chem Res, 2008, 47, 8436–8444. [31] Sakornwimon W, Sylvester, N. Effectiveness factors for partially wetted catalysts in trickle-bed reactors. Ind Eng Chem Process Des Dev, 1982, 21, 16–25. [32] Tsamatsoulis D, Papayannakos N. Partial wetting of cylindrical catalytic carriers in trickle-bed reactors. AIChE J, 1996, 42, 1853–1863. [33] Holub A. Thesis: Hydrodynamics of trickle bed reactors. Thesis, Washington, USA, Washington University, 1990. [34] Kohler M, Richarz W. Investigation of liquid hold-up in trickle bed reactors. Ger Chem Eng, 1985, 8, 295–300. [35] Ellman M, Midoux N, Wild G, Laurent A, Charpentier J. A new, improved liquid hold-up correlation for trickle-bed reactors. Chem Eng Sci, 45, 1677. [36] Wammes W, Mechielsen S, Westerterp K. The influence of pressure on the liquid hold-up in a cocurrent gas-liquid trickle-bed reactor operating at low gas velocities. Chem Eng Sci, 1991, 46, 409. [37] Lange, R, Schubert, M, Bauer T. Liquid hold-up in trickle bed reactors at very low liquid Reynolds numbers. Ind Eng Chem Res, 2005, 44, 6504–6508. [38] Atta A, Roy S, Nigam K. CFD of multiphase flow in high pressure trickle bed reactors: porous media concept, proceedings of European Congress of Chemical Engineering (ECCE-6). Copenhagen, 2007. [39] Stegeman D, van Rooijen F, Kamperman A, Weijer S, Westerterp, K. Residence time distribution in the liquid phase in a cocurrent gas-liquid trickle bed reactor. Ind Eng Chem Res, 1996, 35, 378–385. [40] Piche S, Larachi F, Iliuta I, Grandjean B. Improving the prediction of liquid back-mixing in trickle-bed reactors using a neural network approach. Journal of Chemical Technology and Biotechnology, 2002, 77, 989–998. [41] Schiesser W. The numerical method of lines: integration of partial differential equations. San Diego, California, USA, Academic Press, 1991. [42] Hindmarsh AC, ODEPACK-A Systematized Collection of ODE-Solvers. In: Stepleman R. Scientific Computing, Holland, IMACS/North Holland Publishing Company, 1983, 55–64. [43] Hoyos B, Cadavid J, Rangel H. Formulation and numeric calculation of non-Isothermal effectiveness factor for finite cylindrical catalysts with bi-dimensional diffusion. Latin American Applied Research, 2004, 34, 17. [44] Marquardt D. An algorithm for the least-squares estimation of nonlinear Parameters. SIAM Journal of Applied Mathematics, 1963, 11, 431–441. [45] Haario H. Modest user’s guide. Profmath Oy, Helsinki, 2007. [46] Kilpiö T, Mäki-Arvela P, Rönnholm M, Sifontes V, Wärnå J, Salmi T. Modelling of a three-phase continuously operating isothermal packed-bed reactor: kinetics, mass-transfer, and dispersion effects in the hydrogenation of citral. Ind Eng Chem Res, 2012, 51, 8858–8866. [47] Kilpiö T, Biasi P, Bittante A, Salmi T, Wärnå J. Modelling of direct synthesis of hydrogen peroxide in a fixed bed reactor. Ind Eng Chem Res, 2012, 51, 13366–13378.

References |

331

[48] Leonardão E, Botteselle G, Azumbuja F, Perin G, Jacob R. Citronellal as key compound in organic synthesis. Tetrahedron 2007, 60, 6671–6712 [49] Mäki-Arvela P, Eränen K, Alho K, Salmi T, Murzin D. Multi-tubular reactor design as an advanced screening tool for three-phase catalytic reactions. Topics in Catalysis, 2007, 45, 223. [50] Campos-Martin J, Blanco-Brieva G, Fierro J. Hydrogen peroxide synthesis: An outlook beyond the anthraquinone process. Angew Chem Int Ed, 2006, 45, 6962–6984. [51] Samanta C. Direct synthesis of hydrogen peroxide from hydrogen and oxygen: an overview of recent developments in the process. Applied Catalysis A: General 2008, 350, 133–149. [52] Biasi P, Menegazzo F, Pinna F, Eränen K, Canu P, Salmi T. Hydrogen peroxide direct synthesis: Selectivity enhancement in a trickle bed reactor. Ind Eng Chem Res, 2010, 49, 10627–10632. [53] Biasi P. Combination of catalyst development and chemical reaction engineering: a key aspect to improve the hydrogen peroxide direct synthesis. Doctoral Thesis, Laboratory of Industrial Chemistry and Reaction Engineering, Åbo Akademi University, 2013.

Index 1,2-dichloroethane 67 2-butyne-1,4-diol 271

bubble size 230, 268, 273 bubbling 58, 64

a posteriori method 141 a priori method 139 A3 -reaction mechanism 111 active site 236, 298 activity 295 advanced oxidation process 250 alkene/terpene 13 amination 60 amino acid 60, 83–85, 90 amino terminal group 78, 79 anion exchange resin 9 annular flow 185 aqueous phase reforming 17 – of sorbitol 15 asymmetric vesicle 73, 75 atomization 77 axial dispersion 108, 194, 259, 294, 315, 321, 326

capillary condensation 259 capillary number 269 carbonaceous deposit 236 carboxylic acid terminal group 78, 79, 91 catalyst – deactivation 59, 266, 298 – envelope 6–8 – immobilization 52 – lifetime 240 – particle 54, 82, 293 – preparation technique 238 – screening 292 – selection 292 – washcoat 268 – /support 292 catalytic – distillation column 7 – exchange 18 – of hydrogen isotopes 15 – membrane reactor 19 – oxidation 245 – packing 5, 6, 18 – perovskite hollow fiber membrane 25 cation exchange resin 9, 27 centrifugal partition chromatographic reactor 32 ceramic membrane 25, 69, 71, 78, 80 CFD 310 chemical kinetics 295 chemical reaction 285 chromatographic reactor 26–28, 30, 32, 35, 36, 38–41 cis-trans product ratio 241 citral hydrogenation 211, 322 citronellal 322 co-current 306 – downflow contactor reactor 250 cofactor regeneration 60 coil-to-helix transition 76 colloidosome 52 computational fluid dynamics 249 concentrated stream 320

β-phenethyl acetate 37 backmixing 300 backflushing 58, 88 bacterial cell 240 batch chromatographic reactor 28, 29, 31, 33 bed porosity 193 binary metal nanoparticle 116 binary-coded variable 143 biocatalyst immobilization 52 biocatalytic membrane reactor 52, 53, 56 biochemical conversion 52, 94 biodiesel 10, 20, 21, 24 biofilm 64–68 biological membrane 52, 72 bioPd 240 biosynthesis of nanoparticles 116 biphasic extractive membrane bioreactor 68 biphasic membrane reactor 54 block polymerization 5 boundary condition 314 bubble column 175 – reactor 248 bubble flow 179

334 | Index

concentration gradient 224 concurrent 175 constrained tournament 147 constraint 146 continuous – chromatographic reactor 30 – counter-current chromatographic reactor 30, 31 – flow reactor 246 – process 263 – rotation annular chromatographic reactor 33 – stirred tank reactor 56, 57 controlling resistance 223 convective transport 55, 86 coordinate map 184 copolymerization 53 copper – nanocatalyst 123 – nanocluster 126 – nanoparticle 122 core-shell microcapsule 54–56 core/shell 116 correlation 300, 308 counter current 175 covalent attachement 56, 60, 80, 88, 91 covalent binding 53, 71, 77, 78 cross coupling reaction 118 cross-linking 72, 73, 78 crossover 144 crowding distance 146 crystallinity 103 cyclic operation 178 cyclic voltammetry 245 cyclopropanation reaction 113 deactivation rate constant 323 dehydration 17 – of pentoses and hexoses to furanic aldehydes 15 dehydrogenation 24, 59 dendrimer encapsulated Au 112 density 304 desulphurization of gasoline 12 diblock copolymer 74, 75, 95 diffusion path 303 diffusion-limited regime 56 diffusivity 304 direct decomposition 326

direct synthesis 323 – of hydrogen peroxide 323 discrete slug 108 disk-shaped packed bed microreactor 208 dispersion 195 downflow 191, 270, 306 droplet-based flow 109 droplet-based microreactor 118 dry reforming of methane 23 dynamic model 317 ε-constraint method 141 effective conductivity 305 effective diffusivity 57 effectiveness factor 199, 303 electrical resistance tomography 234 electrostatic deposition 56 elitism 146 emulsion 5 enantiomer 54, 83, 85, 89 enantioselective hydrogenation 262 encapsulation 52, 73, 74 energy balance 291 – for fluids 313 – of the solid phase 313 enhanced oil recovery 264 entrapment 52, 71, 72, 91 enzyme 52, 54–63, 70–81, 83–91, 95 – catalyst 11 epoxidation 13 epoxy terminal group 78, 79 equilibrium uptake 247 esterification 14, 34–37, 59, 80, 91 ethanol steam reforming 23 ethyne hydrogenation 236 Euler-Lagrange 260 experimental design 291, 292, 312, 328 experimental flowsheet 322 experimental productivity and selectivity optimization 292 extent of axial dispersion 321 external diffusion control 238 external wetting efficiency 307 extractive membrane reactor 53 FAME 10, 21, 24 feed distribution 293 film resistance 197 fine chemical 261

Index |

fixed-bed – reactor 287, 288 – chromatographic reactor 29 flash catalyst 127 flash chemistry 127 flow focusing microfluidic device 73, 75 flow maldistribution 193 flow regime 293, 294, 321 – map 184, 268 fluidized catalytic cracking 150 foam packing 188 formate dehydrogenase 60 four reactions 325 furfural 17 gas solubility 305 gas-liquid – flow regime 186 – mass transfer 197, 227, 300 gas-liquid-solid 175 gelation 53, 73, 76, 77 general model equation 312 general rate equation 297 genetic algorithm 142 geometry 307 glucose oxidase 77 goal programming 141 gold – catalyst 246 – nanoparticle 105 good productivity 286 ‘greener’ solvent 228 growth 107 growth-to-seeds solution 109 heat effect 290 heat transfer 197, 201, 299 Henry’s law 224, 305 heterogeneous catalyst 2, 3, 8, 10, 285 hexyl acetate 36 hierarchical catalyst 120 high selectivity 286 hollow fiber membrane reactor 24 hollow fiber module 58, 59 homogeneous catalyst 3, 262 hot spot 178 hydrodesulphurization 195, 253 hydrodynamics 176, 178, 179 hydrogen bonding 76

335

hydrogen isotope 18 hydrogenation 208, 222, 322 – of alkynes 229 – of nitrobenzene 121 hydrogenolysis 209 hydrolysis 38 hydroxylation 39 hysteresis 177, 181 imidazol terminal group 78, 79 immobilization technique 121 impeller 226 in situ 127 – combustion 264 – upgrading 265 incomplete wetting 261 industrial experience 261 industrial reaction 227 industrial unit 320 initial condition 314 inorganic nanoparticle 104 interactive method 142 interfacial effect 116 interior wall of the capillary 115 internal diffusion 288, 302 internal mixing 108 internally finned monolith 4 intrinsic 295 – kinetics 236, 299 ion exchange resin 8, 9, 12, 14, 27, 36–38, 40, 41 ionic binding 53, 70 ionic liquid 11 ionic-electronic conducting membrane 25 ionophore Au nanoparticle 113 ionotropic gelation 76 irreversible gelation 77 isomer 54, 84, 85, 88–91 isothermal 201 isothermal/adiabatic 315 jumping gene 149 kinetic model 296 kinetics 290 L-leucine 60 laminated (multilayer) enzyme membrane reactor 81

336 | Index

Langmuir-Hinshelwood model 235, 248, 297 large-scale operation 320 Levenberg-Marquardt method 319 lexicographic method 140 lipid bilayer 54, 95 liposome 52, 74 liquid – chromatographic reactor 34 – distributor 187 – flow 306, 307 – holdup 188, 191, 309 – slug 269 liquid-solid – heat transfer 301 – mass transfer 199, 230, 301 liver esterase 71, 78, 79, 90, 91 long time stability 295 magnetic resonance imaging 256 maldistribution 193 mapping 143 mass balance 291 – for the solid phase 313 – of the components 312 mass transfer 197, 223, 290, 299, 300 – area 300 – coefficient 197, 224 – limitation 293 – rates for a reactive and a non-reactive system 301 – resistance 299 mating pool 144 membrane – aeration reactor 53, 66 – module 52, 54, 56–59, 66, 68, 93 – reactor 1, 20, 26, 41, 52–54, 56, 59–63, 70, 71, 81–92, 94 metal 104 – nanoparticle 105, 128, 240 methanol 324 – steam reforming 23 method of weighted global criterion 139 methyl acetate 36 methyl acrylate 37 metrics 139 Michaelis-Menten constant 57 micro trickle bed reactor 174, 176–178, 180, 187, 208 micro-engineered catalyst 4, 5

microfiltration membrane reactor 53 microfluidic 103, 128 microgel 52, 73, 76, 77 microparticle 73 microreactor 103, 266 millifluidic – chip 124 – device 123 – mixer 125 miniaturization 128 mixed solvent 239 model development 311 modeled based on physico-chemical theories 328 momentum balance 291, 308 monolith 183 monolithic catalyst 267 morphology 103 multi-lamination micromixer 106 multi-objective optimization 134, 136 multichannel microreactor 185, 186 multifunctional reactor 1, 13, 26, 41 multiphase 175 – flow 191, 195 multiple reactor 286, 288 multistep rate control 296 mutation 144 nanobar 118 nanocube 118 nanoparticle 71, 103, 104 nanorod 109 neutral compromised solution 138 no-preference method 138 non-dispersive gas/liquid mass-transfer process 52 non-dispersive liquid/liquid mass-transfer process 52 non-dominated sorting 145 non-dominated vector 145 non-invasive imaging technique 248 NSGA-II 145 nucleation 105, 107 numerical method 291 – of lines 318 numerical strategy 316 operation condition 291 optical transducer 113

Index |

optimized 325 organic reaction 105 orthogonal collocation 318 Osborne-Wilkinson kinetics 263 overall design equation 255 overall rate 230 oxidation of alcohols 246 oxidation of glycerol 274 oxidative coupling of methane 22, 25 oxygen transfer efficiency 64 packed-bed microreactor 121 palladium – catalyst 237, 323 – membrane 25 – nanoparticle 117 parabolic velocity profile 270 parameter estimation 323 parameter optimization method 319 Pareto optimality 137 partial differential equation 316 partial oxidation of methane 22, 26 particle – geometry 292, 293 – growth 105 – size 201 penalty function 146 percolation theory 259 perfluoromethyldecalin 68 periodic modulation 207 periodic operation 206 permeate 54, 56, 58 peroxidase 63, 77 petroleum processing 149 pH 105 photocatalysis 250 photoreduction 117 physical adsorption 53, 56, 70–72 physical entrapment 52, 71, 91 physical properties of mixtures 303 pinching 118 piston exchange model 316 piston flow exchange model 294 plasmonic absorption 116 plasmonic hollow gold 111 platinum nanoparticle 120 plug flow 273, 307 polyacrylamide microgel 73 poly(ethylene glycol) 60

337

polymer industry 151 polymeric microsphere 52 polymersome 52, 75 polyol process 120 polypropylene hollow fiber 64, 91 population 144 pore diffusion 302 porous media concept 310 positron emission particle tracking 231 power law 309 pressure drop 177, 187, 202 process intensification 18, 206, 213, 267 productivity 289 pulsed flow 179 quasi-homogeneous kinetic model 36 radial flow 307 radio frequency 202 rate expression 296 rate-limiting 299 rational design of catalyst surfaces 242 reaction – enthalpy 306 – kinetics 290 – -limited regime 56 reactive distillation 1, 2, 5, 6, 8, 11, 12, 14, 26, 38, 41 reactive stripping 13 reactor – design 290 – design and operation policy 290 – modeling 311 – set-up 312, 326 – stripping 13 recent advances 327 recirculation 269 – loop 58 reducing agent 105 reductive precipitation 118 reproduction 144 residence time distribution 191, 194, 196, 294 reversed flow chromatographic reactor 33, 34 scale down 177, 205, 258 scale up 111, 204, 205, 258 – issues of fixed beds 320 seedless approach 109 segmented flow 108, 115

338 | Index

segregation 52, 58, 77 selective hydrogenation 193, 211 selective poisoning 244 selectivity 239, 273, 289, 293 – control 242 self-assembly 73 semi-hydrogenation 245 semi-structured 212 – micro trickle bed reactor 177, 182, 187 – trickle bed reactor 212 sensitivity study 312 silicon rubber 64 silver – nanoparticle 114 – nanowire 125 – pentafluoropropionate 114 simple genetic algorithm 144 simplified geometry 310 simplified process 292 simplified reaction system 311 simulated annealing 147 single capillary reactor 271 single-objective optimization 134 single-phase flow 195 site blocker 244 slip velocity 108 slug 185 slurry reactor 223, 225 solution algorithm 317 sorbitol 17 spiral microreactor 115 spray flow 179 start-up 177 static micromixer 106 steady state 296 – diffusion and heat conduction 318 steam reforming 150 step or impulse change 316 stereocomplexation 76 stereoisomer 262 stirring speed 237 structured 212 – catalyst 222 – micro trickle bed reactor 177, 183 – trickle bed reactor 212 sub-model 315 supercritical CO2 57, 80 surface area-to-volume ratio 302

Suzuki-Miyaura coupling 127 synthesis gas 26 Taylor flow 267 temperature gradient 321 temperature oscillation 207 thermal – conductivity 305 – decomposition 118 – reduction 114 Thiele modulus 57 three-phase – fixed bed model 312 – fixed bed reactor 283, 327 – reactor 222 – system 286 ‘toe-to-heel’ air injection 264 transform PDE to ODE 317 tray column 6 tray tower 7 triacetine 34 trickle bed 175 – reactor 176, 253 trickle flow 179, 253 true counter-current chromatographic reactor 29, 30 turbulence 200 two-phase pressure drop 308 Ullmann-type C–O coupling 123 uniform size distribution 103 unstable catalytic intermediate 127 upflow 191 value-added product 292 van der Waals force 72 velocity map 249 viscosity 304 volatile organic compound 64, 66 wall flow 308 wastewater 54, 61, 63–69 wetting efficiency 177 W/O/W emulsion 75 xylose 17 zeolite 8 – membrane 25